v4.19.13 snapshot.
diff --git a/drivers/spi/Kconfig b/drivers/spi/Kconfig
new file mode 100644
index 0000000..671d078
--- /dev/null
+++ b/drivers/spi/Kconfig
@@ -0,0 +1,820 @@
+#
+# SPI driver configuration
+#
+menuconfig SPI
+	bool "SPI support"
+	depends on HAS_IOMEM
+	help
+	  The "Serial Peripheral Interface" is a low level synchronous
+	  protocol.  Chips that support SPI can have data transfer rates
+	  up to several tens of Mbit/sec.  Chips are addressed with a
+	  controller and a chipselect.  Most SPI slaves don't support
+	  dynamic device discovery; some are even write-only or read-only.
+
+	  SPI is widely used by microcontrollers to talk with sensors,
+	  eeprom and flash memory, codecs and various other controller
+	  chips, analog to digital (and d-to-a) converters, and more.
+	  MMC and SD cards can be accessed using SPI protocol; and for
+	  DataFlash cards used in MMC sockets, SPI must always be used.
+
+	  SPI is one of a family of similar protocols using a four wire
+	  interface (select, clock, data in, data out) including Microwire
+	  (half duplex), SSP, SSI, and PSP.  This driver framework should
+	  work with most such devices and controllers.
+
+if SPI
+
+config SPI_DEBUG
+	bool "Debug support for SPI drivers"
+	depends on DEBUG_KERNEL
+	help
+	  Say "yes" to enable debug messaging (like dev_dbg and pr_debug),
+	  sysfs, and debugfs support in SPI controller and protocol drivers.
+
+#
+# MASTER side ... talking to discrete SPI slave chips including microcontrollers
+#
+
+config SPI_MASTER
+#	bool "SPI Master Support"
+	bool
+	default SPI
+	help
+	  If your system has an master-capable SPI controller (which
+	  provides the clock and chipselect), you can enable that
+	  controller and the protocol drivers for the SPI slave chips
+	  that are connected.
+
+if SPI_MASTER
+
+config SPI_MEM
+	bool "SPI memory extension"
+	help
+	  Enable this option if you want to enable the SPI memory extension.
+	  This extension is meant to simplify interaction with SPI memories
+	  by providing a high-level interface to send memory-like commands.
+
+comment "SPI Master Controller Drivers"
+
+config SPI_ALTERA
+	tristate "Altera SPI Controller"
+	help
+	  This is the driver for the Altera SPI Controller.
+
+config SPI_ATH79
+	tristate "Atheros AR71XX/AR724X/AR913X SPI controller driver"
+	depends on ATH79 && GPIOLIB
+	select SPI_BITBANG
+	help
+	  This enables support for the SPI controller present on the
+	  Atheros AR71XX/AR724X/AR913X SoCs.
+
+config SPI_ARMADA_3700
+	tristate "Marvell Armada 3700 SPI Controller"
+	depends on (ARCH_MVEBU && OF) || COMPILE_TEST
+	help
+	  This enables support for the SPI controller present on the
+	  Marvell Armada 3700 SoCs.
+
+config SPI_ATMEL
+	tristate "Atmel SPI Controller"
+	depends on ARCH_AT91 || COMPILE_TEST
+	help
+	  This selects a driver for the Atmel SPI Controller, present on
+	  many AT91 ARM chips.
+
+config SPI_AU1550
+	tristate "Au1550/Au1200/Au1300 SPI Controller"
+	depends on MIPS_ALCHEMY
+	select SPI_BITBANG
+	help
+	  If you say yes to this option, support will be included for the
+	  PSC SPI controller found on Au1550, Au1200 and Au1300 series.
+
+config SPI_AXI_SPI_ENGINE
+	tristate "Analog Devices AXI SPI Engine controller"
+	depends on HAS_IOMEM
+	help
+	  This enables support for the Analog Devices AXI SPI Engine SPI controller.
+	  It is part of the SPI Engine framework that is used in some Analog Devices
+	  reference designs for FPGAs.
+
+config SPI_BCM2835
+	tristate "BCM2835 SPI controller"
+	depends on GPIOLIB
+	depends on ARCH_BCM2835 || COMPILE_TEST
+	help
+	  This selects a driver for the Broadcom BCM2835 SPI master.
+
+	  The BCM2835 contains two types of SPI master controller; the
+	  "universal SPI master", and the regular SPI controller. This driver
+	  is for the regular SPI controller. Slave mode operation is not also
+	  not supported.
+
+config SPI_BCM2835AUX
+	tristate "BCM2835 SPI auxiliary controller"
+	depends on (ARCH_BCM2835 && GPIOLIB) || COMPILE_TEST
+	help
+	  This selects a driver for the Broadcom BCM2835 SPI aux master.
+
+	  The BCM2835 contains two types of SPI master controller; the
+	  "universal SPI master", and the regular SPI controller.
+	  This driver is for the universal/auxiliary SPI controller.
+
+config SPI_BCM63XX
+	tristate "Broadcom BCM63xx SPI controller"
+	depends on BCM63XX || COMPILE_TEST
+	help
+          Enable support for the SPI controller on the Broadcom BCM63xx SoCs.
+
+config SPI_BCM63XX_HSSPI
+	tristate "Broadcom BCM63XX HS SPI controller driver"
+	depends on BCM63XX || COMPILE_TEST
+	help
+	  This enables support for the High Speed SPI controller present on
+	  newer Broadcom BCM63XX SoCs.
+
+config SPI_BCM_QSPI
+	tristate "Broadcom BSPI and MSPI controller support"
+	depends on ARCH_BRCMSTB || ARCH_BCM || ARCH_BCM_IPROC || \
+			BMIPS_GENERIC || COMPILE_TEST
+	default ARCH_BCM_IPROC
+	help
+	  Enables support for the Broadcom SPI flash and MSPI controller.
+	  Select this option for any one of BRCMSTB, iProc NSP and NS2 SoCs
+	  based platforms. This driver works for both SPI master for spi-nor
+	  flash device as well as MSPI device.
+
+config SPI_BITBANG
+	tristate "Utilities for Bitbanging SPI masters"
+	help
+	  With a few GPIO pins, your system can bitbang the SPI protocol.
+	  Select this to get SPI support through I/O pins (GPIO, parallel
+	  port, etc).  Or, some systems' SPI master controller drivers use
+	  this code to manage the per-word or per-transfer accesses to the
+	  hardware shift registers.
+
+	  This is library code, and is automatically selected by drivers that
+	  need it.  You only need to select this explicitly to support driver
+	  modules that aren't part of this kernel tree.
+
+config SPI_BUTTERFLY
+	tristate "Parallel port adapter for AVR Butterfly (DEVELOPMENT)"
+	depends on PARPORT
+	select SPI_BITBANG
+	help
+	  This uses a custom parallel port cable to connect to an AVR
+	  Butterfly <http://www.atmel.com/products/avr/butterfly>, an
+	  inexpensive battery powered microcontroller evaluation board.
+	  This same cable can be used to flash new firmware.
+
+config SPI_CADENCE
+	tristate "Cadence SPI controller"
+	help
+	  This selects the Cadence SPI controller master driver
+	  used by Xilinx Zynq and ZynqMP.
+
+config SPI_CLPS711X
+	tristate "CLPS711X host SPI controller"
+	depends on ARCH_CLPS711X || COMPILE_TEST
+	help
+	  This enables dedicated general purpose SPI/Microwire1-compatible
+	  master mode interface (SSI1) for CLPS711X-based CPUs.
+
+config SPI_COLDFIRE_QSPI
+	tristate "Freescale Coldfire QSPI controller"
+	depends on (M520x || M523x || M5249 || M525x || M527x || M528x || M532x)
+	help
+	  This enables support for the Coldfire QSPI controller in master
+	  mode.
+
+config SPI_DAVINCI
+	tristate "Texas Instruments DaVinci/DA8x/OMAP-L/AM1x SoC SPI controller"
+	depends on ARCH_DAVINCI || ARCH_KEYSTONE
+	select SPI_BITBANG
+	help
+	  SPI master controller for DaVinci/DA8x/OMAP-L/AM1x SPI modules.
+
+config SPI_DESIGNWARE
+	tristate "DesignWare SPI controller core support"
+	help
+	  general driver for SPI controller core from DesignWare
+
+config SPI_DW_PCI
+	tristate "PCI interface driver for DW SPI core"
+	depends on SPI_DESIGNWARE && PCI
+
+config SPI_DW_MID_DMA
+	bool "DMA support for DW SPI controller on Intel MID platform"
+	depends on SPI_DW_PCI && DW_DMAC_PCI
+
+config SPI_DW_MMIO
+	tristate "Memory-mapped io interface driver for DW SPI core"
+	depends on SPI_DESIGNWARE
+
+config SPI_DLN2
+       tristate "Diolan DLN-2 USB SPI adapter"
+       depends on MFD_DLN2
+       help
+         If you say yes to this option, support will be included for Diolan
+         DLN2, a USB to SPI interface.
+
+         This driver can also be built as a module.  If so, the module
+         will be called spi-dln2.
+
+config SPI_EFM32
+	tristate "EFM32 SPI controller"
+	depends on OF && ARM && (ARCH_EFM32 || COMPILE_TEST)
+	select SPI_BITBANG
+	help
+	  Driver for the spi controller found on Energy Micro's EFM32 SoCs.
+
+config SPI_EP93XX
+	tristate "Cirrus Logic EP93xx SPI controller"
+	depends on ARCH_EP93XX || COMPILE_TEST
+	help
+	  This enables using the Cirrus EP93xx SPI controller in master
+	  mode.
+
+config SPI_FALCON
+	bool "Falcon SPI controller support"
+	depends on SOC_FALCON
+	help
+	  The external bus unit (EBU) found on the FALC-ON SoC has SPI
+	  emulation that is designed for serial flash access. This driver
+	  has only been tested with m25p80 type chips. The hardware has no
+	  support for other types of SPI peripherals.
+
+config SPI_FSL_LPSPI
+	tristate "Freescale i.MX LPSPI controller"
+	depends on ARCH_MXC || COMPILE_TEST
+	help
+	  This enables Freescale i.MX LPSPI controllers in master mode.
+
+config SPI_GPIO
+	tristate "GPIO-based bitbanging SPI Master"
+	depends on GPIOLIB || COMPILE_TEST
+	select SPI_BITBANG
+	help
+	  This simple GPIO bitbanging SPI master uses the arch-neutral GPIO
+	  interface to manage MOSI, MISO, SCK, and chipselect signals.  SPI
+	  slaves connected to a bus using this driver are configured as usual,
+	  except that the spi_board_info.controller_data holds the GPIO number
+	  for the chipselect used by this controller driver.
+
+	  Note that this driver often won't achieve even 1 Mbit/sec speeds,
+	  making it unusually slow for SPI.  If your platform can inline
+	  GPIO operations, you should be able to leverage that for better
+	  speed with a custom version of this driver; see the source code.
+
+config SPI_IMG_SPFI
+	tristate "IMG SPFI controller"
+	depends on MIPS || COMPILE_TEST
+	help
+	  This enables support for the SPFI master controller found on
+	  IMG SoCs.
+
+config SPI_IMX
+	tristate "Freescale i.MX SPI controllers"
+	depends on ARCH_MXC || COMPILE_TEST
+	select SPI_BITBANG
+	help
+	  This enables using the Freescale i.MX SPI controllers in master
+	  mode.
+
+config SPI_JCORE
+	tristate "J-Core SPI Master"
+	depends on OF && (SUPERH || COMPILE_TEST)
+	help
+	  This enables support for the SPI master controller in the J-Core
+	  synthesizable, open source SoC.
+
+config SPI_LM70_LLP
+	tristate "Parallel port adapter for LM70 eval board (DEVELOPMENT)"
+	depends on PARPORT
+	select SPI_BITBANG
+	help
+	  This driver supports the NS LM70 LLP Evaluation Board,
+	  which interfaces to an LM70 temperature sensor using
+	  a parallel port.
+
+config SPI_LP8841_RTC
+	tristate "ICP DAS LP-8841 SPI Controller for RTC"
+	depends on MACH_PXA27X_DT || COMPILE_TEST
+	help
+	  This driver provides an SPI master device to drive Maxim
+	  DS-1302 real time clock.
+
+	  Say N here unless you plan to run the kernel on an ICP DAS
+	  LP-8x4x industrial computer.
+
+config SPI_MPC52xx
+	tristate "Freescale MPC52xx SPI (non-PSC) controller support"
+	depends on PPC_MPC52xx
+	help
+	  This drivers supports the MPC52xx SPI controller in master SPI
+	  mode.
+
+config SPI_MPC52xx_PSC
+	tristate "Freescale MPC52xx PSC SPI controller"
+	depends on PPC_MPC52xx
+	help
+	  This enables using the Freescale MPC52xx Programmable Serial
+	  Controller in master SPI mode.
+
+config SPI_MPC512x_PSC
+	tristate "Freescale MPC512x PSC SPI controller"
+	depends on PPC_MPC512x
+	help
+	  This enables using the Freescale MPC5121 Programmable Serial
+	  Controller in SPI master mode.
+
+config SPI_FSL_LIB
+	tristate
+	depends on OF
+
+config SPI_FSL_CPM
+	tristate
+	depends on FSL_SOC
+
+config SPI_FSL_SPI
+	tristate "Freescale SPI controller and Aeroflex Gaisler GRLIB SPI controller"
+	depends on OF
+	select SPI_FSL_LIB
+	select SPI_FSL_CPM if FSL_SOC
+	help
+	  This enables using the Freescale SPI controllers in master mode.
+	  MPC83xx platform uses the controller in cpu mode or CPM/QE mode.
+	  MPC8569 uses the controller in QE mode, MPC8610 in cpu mode.
+	  This also enables using the Aeroflex Gaisler GRLIB SPI controller in
+	  master mode.
+
+config SPI_FSL_DSPI
+	tristate "Freescale DSPI controller"
+	select REGMAP_MMIO
+	depends on SOC_VF610 || SOC_LS1021A || ARCH_LAYERSCAPE || M5441x || COMPILE_TEST
+	help
+	  This enables support for the Freescale DSPI controller in master
+	  mode. VF610 platform uses the controller.
+
+config SPI_FSL_ESPI
+	tristate "Freescale eSPI controller"
+	depends on FSL_SOC
+	help
+	  This enables using the Freescale eSPI controllers in master mode.
+	  From MPC8536, 85xx platform uses the controller, and all P10xx,
+	  P20xx, P30xx,P40xx, P50xx uses this controller.
+
+config SPI_MESON_SPICC
+	tristate "Amlogic Meson SPICC controller"
+	depends on ARCH_MESON || COMPILE_TEST
+	help
+	  This enables master mode support for the SPICC (SPI communication
+	  controller) available in Amlogic Meson SoCs.
+
+config SPI_MESON_SPIFC
+	tristate "Amlogic Meson SPIFC controller"
+	depends on ARCH_MESON || COMPILE_TEST
+	select REGMAP_MMIO
+	help
+	  This enables master mode support for the SPIFC (SPI flash
+	  controller) available in Amlogic Meson SoCs.
+
+config SPI_MT65XX
+	tristate "MediaTek SPI controller"
+	depends on ARCH_MEDIATEK || COMPILE_TEST
+	help
+	  This selects the MediaTek(R) SPI bus driver.
+	  If you want to use MediaTek(R) SPI interface,
+	  say Y or M here.If you are not sure, say N.
+	  SPI drivers for Mediatek MT65XX and MT81XX series ARM SoCs.
+
+config SPI_NUC900
+	tristate "Nuvoton NUC900 series SPI"
+	depends on ARCH_W90X900
+	select SPI_BITBANG
+	help
+	  SPI driver for Nuvoton NUC900 series ARM SoCs
+
+config SPI_LANTIQ_SSC
+	tristate "Lantiq SSC SPI controller"
+	depends on LANTIQ || COMPILE_TEST
+	help
+	  This driver supports the Lantiq SSC SPI controller in master
+	  mode. This controller is found on Intel (former Lantiq) SoCs like
+	  the Danube, Falcon, xRX200, xRX300.
+
+config SPI_OC_TINY
+	tristate "OpenCores tiny SPI"
+	depends on GPIOLIB || COMPILE_TEST
+	select SPI_BITBANG
+	help
+	  This is the driver for OpenCores tiny SPI master controller.
+
+config SPI_OCTEON
+	tristate "Cavium OCTEON SPI controller"
+	depends on CAVIUM_OCTEON_SOC
+	help
+	  SPI host driver for the hardware found on some Cavium OCTEON
+	  SOCs.
+
+config SPI_OMAP_UWIRE
+	tristate "OMAP1 MicroWire"
+	depends on ARCH_OMAP1
+	select SPI_BITBANG
+	help
+	  This hooks up to the MicroWire controller on OMAP1 chips.
+
+config SPI_OMAP24XX
+	tristate "McSPI driver for OMAP"
+	depends on ARCH_OMAP2PLUS || COMPILE_TEST
+	select SG_SPLIT
+	help
+	  SPI master controller for OMAP24XX and later Multichannel SPI
+	  (McSPI) modules.
+
+config SPI_TI_QSPI
+	tristate "DRA7xxx QSPI controller support"
+	depends on ARCH_OMAP2PLUS || COMPILE_TEST
+	help
+	  QSPI master controller for DRA7xxx used for flash devices.
+	  This device supports single, dual and quad read support, while
+	  it only supports single write mode.
+
+config SPI_OMAP_100K
+	tristate "OMAP SPI 100K"
+	depends on ARCH_OMAP850 || ARCH_OMAP730 || COMPILE_TEST
+	help
+	  OMAP SPI 100K master controller for omap7xx boards.
+
+config SPI_ORION
+	tristate "Orion SPI master"
+	depends on PLAT_ORION || ARCH_MVEBU || COMPILE_TEST
+	help
+	  This enables using the SPI master controller on the Orion
+	  and MVEBU chips.
+
+config SPI_PIC32
+	tristate "Microchip PIC32 series SPI"
+	depends on MACH_PIC32 || COMPILE_TEST
+	help
+	  SPI driver for Microchip PIC32 SPI master controller.
+
+config SPI_PIC32_SQI
+	tristate "Microchip PIC32 Quad SPI driver"
+	depends on MACH_PIC32 || COMPILE_TEST
+	help
+	  SPI driver for PIC32 Quad SPI controller.
+
+config SPI_PL022
+	tristate "ARM AMBA PL022 SSP controller"
+	depends on ARM_AMBA
+	default y if MACH_U300
+	default y if ARCH_REALVIEW
+	default y if INTEGRATOR_IMPD1
+	default y if ARCH_VERSATILE
+	help
+	  This selects the ARM(R) AMBA(R) PrimeCell PL022 SSP
+	  controller. If you have an embedded system with an AMBA(R)
+	  bus and a PL022 controller, say Y or M here.
+
+config SPI_PPC4xx
+	tristate "PPC4xx SPI Controller"
+	depends on PPC32 && 4xx
+	select SPI_BITBANG
+	help
+	  This selects a driver for the PPC4xx SPI Controller.
+
+config SPI_PXA2XX
+	tristate "PXA2xx SSP SPI master"
+	depends on (ARCH_PXA || ARCH_MMP || PCI || ACPI)
+	select PXA_SSP if ARCH_PXA || ARCH_MMP
+	help
+	  This enables using a PXA2xx or Sodaville SSP port as a SPI master
+	  controller. The driver can be configured to use any SSP port and
+	  additional documentation can be found a Documentation/spi/pxa2xx.
+
+config SPI_PXA2XX_PCI
+	def_tristate SPI_PXA2XX && PCI && COMMON_CLK
+
+config SPI_ROCKCHIP
+	tristate "Rockchip SPI controller driver"
+	help
+	  This selects a driver for Rockchip SPI controller.
+
+	  If you say yes to this option, support will be included for
+	  RK3066, RK3188 and RK3288 families of SPI controller.
+	  Rockchip SPI controller support DMA transport and PIO mode.
+	  The main usecase of this controller is to use spi flash as boot
+	  device.
+
+config SPI_RB4XX
+	tristate "Mikrotik RB4XX SPI master"
+	depends on SPI_MASTER && ATH79
+	help
+	  SPI controller driver for the Mikrotik RB4xx series boards.
+
+config SPI_RSPI
+	tristate "Renesas RSPI/QSPI controller"
+	depends on SUPERH || ARCH_RENESAS || COMPILE_TEST
+	help
+	  SPI driver for Renesas RSPI and QSPI blocks.
+
+config SPI_QUP
+	tristate "Qualcomm SPI controller with QUP interface"
+	depends on ARCH_QCOM || (ARM && COMPILE_TEST)
+	help
+	  Qualcomm Universal Peripheral (QUP) core is an AHB slave that
+	  provides a common data path (an output FIFO and an input FIFO)
+	  for serial peripheral interface (SPI) mini-core. SPI in master
+	  mode supports up to 50MHz, up to four chip selects, programmable
+	  data path from 4 bits to 32 bits and numerous protocol variants.
+
+	  This driver can also be built as a module.  If so, the module
+	  will be called spi_qup.
+
+config SPI_S3C24XX
+	tristate "Samsung S3C24XX series SPI"
+	depends on ARCH_S3C24XX
+	select SPI_BITBANG
+	help
+	  SPI driver for Samsung S3C24XX series ARM SoCs
+
+config SPI_S3C24XX_FIQ
+	bool "S3C24XX driver with FIQ pseudo-DMA"
+	depends on SPI_S3C24XX
+	select FIQ
+	help
+	  Enable FIQ support for the S3C24XX SPI driver to provide pseudo
+	  DMA by using the fast-interrupt request framework, This allows
+	  the driver to get DMA-like performance when there are either
+	  no free DMA channels, or when doing transfers that required both
+	  TX and RX data paths.
+
+config SPI_S3C64XX
+	tristate "Samsung S3C64XX series type SPI"
+	depends on (PLAT_SAMSUNG || ARCH_EXYNOS || COMPILE_TEST)
+	help
+	  SPI driver for Samsung S3C64XX and newer SoCs.
+
+config SPI_SC18IS602
+	tristate "NXP SC18IS602/602B/603 I2C to SPI bridge"
+	depends on I2C
+	help
+	  SPI driver for NXP SC18IS602/602B/603 I2C to SPI bridge.
+
+config SPI_SH_MSIOF
+	tristate "SuperH MSIOF SPI controller"
+	depends on HAVE_CLK
+	depends on ARCH_SHMOBILE || ARCH_RENESAS || COMPILE_TEST
+	help
+	  SPI driver for SuperH and SH Mobile MSIOF blocks.
+
+config SPI_SH
+	tristate "SuperH SPI controller"
+	depends on SUPERH || COMPILE_TEST
+	help
+	  SPI driver for SuperH SPI blocks.
+
+config SPI_SH_SCI
+	tristate "SuperH SCI SPI controller"
+	depends on SUPERH
+	select SPI_BITBANG
+	help
+	  SPI driver for SuperH SCI blocks.
+
+config SPI_SH_HSPI
+	tristate "SuperH HSPI controller"
+	depends on ARCH_RENESAS || COMPILE_TEST
+	help
+	  SPI driver for SuperH HSPI blocks.
+
+config SPI_SIRF
+	tristate "CSR SiRFprimaII SPI controller"
+	depends on SIRF_DMA
+	select SPI_BITBANG
+	help
+	  SPI driver for CSR SiRFprimaII SoCs
+
+config SPI_SPRD_ADI
+	tristate "Spreadtrum ADI controller"
+	depends on ARCH_SPRD || COMPILE_TEST
+	depends on HWSPINLOCK || (COMPILE_TEST && !HWSPINLOCK)
+	help
+	  ADI driver based on SPI for Spreadtrum SoCs.
+
+config SPI_STM32
+	tristate "STMicroelectronics STM32 SPI controller"
+	depends on ARCH_STM32 || COMPILE_TEST
+	help
+	  SPI driver for STMicroelectonics STM32 SoCs.
+
+	  STM32 SPI controller supports DMA and PIO modes. When DMA
+	  is not available, the driver automatically falls back to
+	  PIO mode.
+
+config SPI_ST_SSC4
+	tristate "STMicroelectronics SPI SSC-based driver"
+	depends on ARCH_STI || COMPILE_TEST
+	help
+	  STMicroelectronics SoCs support for SPI. If you say yes to
+	  this option, support will be included for the SSC driven SPI.
+
+config SPI_SUN4I
+	tristate "Allwinner A10 SoCs SPI controller"
+	depends on ARCH_SUNXI || COMPILE_TEST
+	help
+	  SPI driver for Allwinner sun4i, sun5i and sun7i SoCs
+
+config SPI_SUN6I
+	tristate "Allwinner A31 SPI controller"
+	depends on ARCH_SUNXI || COMPILE_TEST
+	depends on RESET_CONTROLLER
+	help
+	  This enables using the SPI controller on the Allwinner A31 SoCs.
+
+config SPI_MXS
+	tristate "Freescale MXS SPI controller"
+	depends on ARCH_MXS
+	select STMP_DEVICE
+	help
+	  SPI driver for Freescale MXS devices.
+
+config SPI_TEGRA114
+	tristate "NVIDIA Tegra114 SPI Controller"
+	depends on (ARCH_TEGRA && TEGRA20_APB_DMA) || COMPILE_TEST
+	depends on RESET_CONTROLLER
+	help
+	  SPI driver for NVIDIA Tegra114 SPI Controller interface. This controller
+	  is different than the older SoCs SPI controller and also register interface
+	  get changed with this controller.
+
+config SPI_TEGRA20_SFLASH
+	tristate "Nvidia Tegra20 Serial flash Controller"
+	depends on ARCH_TEGRA || COMPILE_TEST
+	depends on RESET_CONTROLLER
+	help
+	  SPI driver for Nvidia Tegra20 Serial flash Controller interface.
+	  The main usecase of this controller is to use spi flash as boot
+	  device.
+
+config SPI_TEGRA20_SLINK
+	tristate "Nvidia Tegra20/Tegra30 SLINK Controller"
+	depends on (ARCH_TEGRA && TEGRA20_APB_DMA) || COMPILE_TEST
+	depends on RESET_CONTROLLER
+	help
+	  SPI driver for Nvidia Tegra20/Tegra30 SLINK Controller interface.
+
+config SPI_THUNDERX
+	tristate "Cavium ThunderX SPI controller"
+	depends on PCI && 64BIT && (ARM64 || COMPILE_TEST)
+	help
+	  SPI host driver for the hardware found on Cavium ThunderX
+	  SOCs.
+
+config SPI_TOPCLIFF_PCH
+	tristate "Intel EG20T PCH/LAPIS Semicon IOH(ML7213/ML7223/ML7831) SPI"
+	depends on PCI && (X86_32 || MIPS || COMPILE_TEST)
+	help
+	  SPI driver for the Topcliff PCH (Platform Controller Hub) SPI bus
+	  used in some x86 embedded processors.
+
+	  This driver also supports the ML7213/ML7223/ML7831, a companion chip
+	  for the Atom E6xx series and compatible with the Intel EG20T PCH.
+
+config SPI_TXX9
+	tristate "Toshiba TXx9 SPI controller"
+	depends on GPIOLIB && (CPU_TX49XX || COMPILE_TEST)
+	help
+	  SPI driver for Toshiba TXx9 MIPS SoCs
+
+config SPI_UNIPHIER
+	tristate "Socionext UniPhier SPI Controller"
+	depends on (ARCH_UNIPHIER || COMPILE_TEST) && OF
+	help
+	  This enables a driver for the Socionext UniPhier SoC SCSSI SPI controller.
+
+	  UniPhier SoCs have SCSSI and MCSSI SPI controllers.
+	  Every UniPhier SoC has SCSSI which supports single channel.
+	  Older UniPhier Pro4/Pro5 also has MCSSI which support multiple channels.
+	  This driver supports SCSSI only.
+
+	  If your SoC supports SCSSI, say Y here.
+
+config SPI_XCOMM
+	tristate "Analog Devices AD-FMCOMMS1-EBZ SPI-I2C-bridge driver"
+	depends on I2C
+	help
+	  Support for the SPI-I2C bridge found on the Analog Devices
+	  AD-FMCOMMS1-EBZ board.
+
+config SPI_XILINX
+	tristate "Xilinx SPI controller common module"
+	depends on HAS_IOMEM
+	select SPI_BITBANG
+	help
+	  This exposes the SPI controller IP from the Xilinx EDK.
+
+	  See the "OPB Serial Peripheral Interface (SPI) (v1.00e)"
+	  Product Specification document (DS464) for hardware details.
+
+	  Or for the DS570, see "XPS Serial Peripheral Interface (SPI) (v2.00b)"
+
+config SPI_XLP
+	tristate "Netlogic XLP SPI controller driver"
+	depends on CPU_XLP || ARCH_THUNDER2 || COMPILE_TEST
+	help
+	  Enable support for the SPI controller on the Netlogic XLP SoCs.
+	  Currently supported XLP variants are XLP8XX, XLP3XX, XLP2XX, XLP9XX
+	  and XLP5XX.
+
+	  If you have a Netlogic XLP platform say Y here.
+	  If unsure, say N.
+
+config SPI_XTENSA_XTFPGA
+	tristate "Xtensa SPI controller for xtfpga"
+	depends on (XTENSA && XTENSA_PLATFORM_XTFPGA) || COMPILE_TEST
+	select SPI_BITBANG
+	help
+	  SPI driver for xtfpga SPI master controller.
+
+	  This simple SPI master controller is built into xtfpga bitstreams
+	  and is used to control daughterboard audio codec. It always transfers
+	  16 bit words in SPI mode 0, automatically asserting CS on transfer
+	  start and deasserting on end.
+
+config SPI_ZYNQMP_GQSPI
+	tristate "Xilinx ZynqMP GQSPI controller"
+	depends on SPI_MASTER && HAS_DMA
+	help
+	  Enables Xilinx GQSPI controller driver for Zynq UltraScale+ MPSoC.
+
+#
+# Add new SPI master controllers in alphabetical order above this line
+#
+
+#
+# There are lots of SPI device types, with sensors and memory
+# being probably the most widely used ones.
+#
+comment "SPI Protocol Masters"
+
+config SPI_SPIDEV
+	tristate "User mode SPI device driver support"
+	help
+	  This supports user mode SPI protocol drivers.
+
+	  Note that this application programming interface is EXPERIMENTAL
+	  and hence SUBJECT TO CHANGE WITHOUT NOTICE while it stabilizes.
+
+config SPI_LOOPBACK_TEST
+	tristate "spi loopback test framework support"
+	depends on m
+	help
+	  This enables the SPI loopback testing framework driver
+
+	  primarily used for development of spi_master drivers
+	  and to detect regressions
+
+config SPI_TLE62X0
+	tristate "Infineon TLE62X0 (for power switching)"
+	depends on SYSFS
+	help
+	  SPI driver for Infineon TLE62X0 series line driver chips,
+	  such as the TLE6220, TLE6230 and TLE6240.  This provides a
+	  sysfs interface, with each line presented as a kind of GPIO
+	  exposing both switch control and diagnostic feedback.
+
+#
+# Add new SPI protocol masters in alphabetical order above this line
+#
+
+endif # SPI_MASTER
+
+#
+# SLAVE side ... listening to other SPI masters
+#
+
+config SPI_SLAVE
+	bool "SPI slave protocol handlers"
+	help
+	  If your system has a slave-capable SPI controller, you can enable
+	  slave protocol handlers.
+
+if SPI_SLAVE
+
+config SPI_SLAVE_TIME
+	tristate "SPI slave handler reporting boot up time"
+	help
+	  SPI slave handler responding with the time of reception of the last
+	  SPI message.
+
+config SPI_SLAVE_SYSTEM_CONTROL
+	tristate "SPI slave handler controlling system state"
+	help
+	  SPI slave handler to allow remote control of system reboot, power
+	  off, halt, and suspend.
+
+endif # SPI_SLAVE
+
+endif # SPI
diff --git a/drivers/spi/Makefile b/drivers/spi/Makefile
new file mode 100644
index 0000000..a90d559
--- /dev/null
+++ b/drivers/spi/Makefile
@@ -0,0 +1,113 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# Makefile for kernel SPI drivers.
+#
+
+ccflags-$(CONFIG_SPI_DEBUG) := -DDEBUG
+
+# small core, mostly translating board-specific
+# config declarations into driver model code
+obj-$(CONFIG_SPI_MASTER)		+= spi.o
+obj-$(CONFIG_SPI_MEM)			+= spi-mem.o
+obj-$(CONFIG_SPI_SPIDEV)		+= spidev.o
+obj-$(CONFIG_SPI_LOOPBACK_TEST)		+= spi-loopback-test.o
+
+# SPI master controller drivers (bus)
+obj-$(CONFIG_SPI_ALTERA)		+= spi-altera.o
+obj-$(CONFIG_SPI_ARMADA_3700)		+= spi-armada-3700.o
+obj-$(CONFIG_SPI_ATMEL)			+= spi-atmel.o
+obj-$(CONFIG_SPI_ATH79)			+= spi-ath79.o
+obj-$(CONFIG_SPI_AU1550)		+= spi-au1550.o
+obj-$(CONFIG_SPI_AXI_SPI_ENGINE)	+= spi-axi-spi-engine.o
+obj-$(CONFIG_SPI_BCM2835)		+= spi-bcm2835.o
+obj-$(CONFIG_SPI_BCM2835AUX)		+= spi-bcm2835aux.o
+obj-$(CONFIG_SPI_BCM63XX)		+= spi-bcm63xx.o
+obj-$(CONFIG_SPI_BCM63XX_HSSPI)		+= spi-bcm63xx-hsspi.o
+obj-$(CONFIG_SPI_BCM_QSPI)		+= spi-iproc-qspi.o spi-brcmstb-qspi.o spi-bcm-qspi.o
+obj-$(CONFIG_SPI_BITBANG)		+= spi-bitbang.o
+obj-$(CONFIG_SPI_BUTTERFLY)		+= spi-butterfly.o
+obj-$(CONFIG_SPI_CADENCE)		+= spi-cadence.o
+obj-$(CONFIG_SPI_CLPS711X)		+= spi-clps711x.o
+obj-$(CONFIG_SPI_COLDFIRE_QSPI)		+= spi-coldfire-qspi.o
+obj-$(CONFIG_SPI_DAVINCI)		+= spi-davinci.o
+obj-$(CONFIG_SPI_DLN2)			+= spi-dln2.o
+obj-$(CONFIG_SPI_DESIGNWARE)		+= spi-dw.o
+obj-$(CONFIG_SPI_DW_MMIO)		+= spi-dw-mmio.o
+obj-$(CONFIG_SPI_DW_PCI)		+= spi-dw-midpci.o
+spi-dw-midpci-objs			:= spi-dw-pci.o spi-dw-mid.o
+obj-$(CONFIG_SPI_EFM32)			+= spi-efm32.o
+obj-$(CONFIG_SPI_EP93XX)		+= spi-ep93xx.o
+obj-$(CONFIG_SPI_FALCON)		+= spi-falcon.o
+obj-$(CONFIG_SPI_FSL_CPM)		+= spi-fsl-cpm.o
+obj-$(CONFIG_SPI_FSL_DSPI)		+= spi-fsl-dspi.o
+obj-$(CONFIG_SPI_FSL_LIB)		+= spi-fsl-lib.o
+obj-$(CONFIG_SPI_FSL_ESPI)		+= spi-fsl-espi.o
+obj-$(CONFIG_SPI_FSL_LPSPI)		+= spi-fsl-lpspi.o
+obj-$(CONFIG_SPI_FSL_SPI)		+= spi-fsl-spi.o
+obj-$(CONFIG_SPI_GPIO)			+= spi-gpio.o
+obj-$(CONFIG_SPI_IMG_SPFI)		+= spi-img-spfi.o
+obj-$(CONFIG_SPI_IMX)			+= spi-imx.o
+obj-$(CONFIG_SPI_LANTIQ_SSC)		+= spi-lantiq-ssc.o
+obj-$(CONFIG_SPI_JCORE)			+= spi-jcore.o
+obj-$(CONFIG_SPI_LM70_LLP)		+= spi-lm70llp.o
+obj-$(CONFIG_SPI_LP8841_RTC)		+= spi-lp8841-rtc.o
+obj-$(CONFIG_SPI_MESON_SPICC)		+= spi-meson-spicc.o
+obj-$(CONFIG_SPI_MESON_SPIFC)		+= spi-meson-spifc.o
+obj-$(CONFIG_SPI_MPC512x_PSC)		+= spi-mpc512x-psc.o
+obj-$(CONFIG_SPI_MPC52xx_PSC)		+= spi-mpc52xx-psc.o
+obj-$(CONFIG_SPI_MPC52xx)		+= spi-mpc52xx.o
+obj-$(CONFIG_SPI_MT65XX)                += spi-mt65xx.o
+obj-$(CONFIG_SPI_MXS)			+= spi-mxs.o
+obj-$(CONFIG_SPI_NUC900)		+= spi-nuc900.o
+obj-$(CONFIG_SPI_OC_TINY)		+= spi-oc-tiny.o
+spi-octeon-objs				:= spi-cavium.o spi-cavium-octeon.o
+obj-$(CONFIG_SPI_OCTEON)		+= spi-octeon.o
+obj-$(CONFIG_SPI_OMAP_UWIRE)		+= spi-omap-uwire.o
+obj-$(CONFIG_SPI_OMAP_100K)		+= spi-omap-100k.o
+obj-$(CONFIG_SPI_OMAP24XX)		+= spi-omap2-mcspi.o
+obj-$(CONFIG_SPI_TI_QSPI)		+= spi-ti-qspi.o
+obj-$(CONFIG_SPI_ORION)			+= spi-orion.o
+obj-$(CONFIG_SPI_PIC32)			+= spi-pic32.o
+obj-$(CONFIG_SPI_PIC32_SQI)		+= spi-pic32-sqi.o
+obj-$(CONFIG_SPI_PL022)			+= spi-pl022.o
+obj-$(CONFIG_SPI_PPC4xx)		+= spi-ppc4xx.o
+spi-pxa2xx-platform-objs		:= spi-pxa2xx.o spi-pxa2xx-dma.o
+obj-$(CONFIG_SPI_PXA2XX)		+= spi-pxa2xx-platform.o
+obj-$(CONFIG_SPI_PXA2XX_PCI)		+= spi-pxa2xx-pci.o
+obj-$(CONFIG_SPI_QUP)			+= spi-qup.o
+obj-$(CONFIG_SPI_ROCKCHIP)		+= spi-rockchip.o
+obj-$(CONFIG_SPI_RB4XX)			+= spi-rb4xx.o
+obj-$(CONFIG_SPI_RSPI)			+= spi-rspi.o
+obj-$(CONFIG_SPI_S3C24XX)		+= spi-s3c24xx-hw.o
+spi-s3c24xx-hw-y			:= spi-s3c24xx.o
+spi-s3c24xx-hw-$(CONFIG_SPI_S3C24XX_FIQ) += spi-s3c24xx-fiq.o
+obj-$(CONFIG_SPI_S3C64XX)		+= spi-s3c64xx.o
+obj-$(CONFIG_SPI_SC18IS602)		+= spi-sc18is602.o
+obj-$(CONFIG_SPI_SH)			+= spi-sh.o
+obj-$(CONFIG_SPI_SH_HSPI)		+= spi-sh-hspi.o
+obj-$(CONFIG_SPI_SH_MSIOF)		+= spi-sh-msiof.o
+obj-$(CONFIG_SPI_SH_SCI)		+= spi-sh-sci.o
+obj-$(CONFIG_SPI_SIRF)		+= spi-sirf.o
+obj-$(CONFIG_SPI_SPRD_ADI)		+= spi-sprd-adi.o
+obj-$(CONFIG_SPI_STM32) 		+= spi-stm32.o
+obj-$(CONFIG_SPI_ST_SSC4)		+= spi-st-ssc4.o
+obj-$(CONFIG_SPI_SUN4I)			+= spi-sun4i.o
+obj-$(CONFIG_SPI_SUN6I)			+= spi-sun6i.o
+obj-$(CONFIG_SPI_TEGRA114)		+= spi-tegra114.o
+obj-$(CONFIG_SPI_TEGRA20_SFLASH)	+= spi-tegra20-sflash.o
+obj-$(CONFIG_SPI_TEGRA20_SLINK)		+= spi-tegra20-slink.o
+obj-$(CONFIG_SPI_TLE62X0)		+= spi-tle62x0.o
+spi-thunderx-objs			:= spi-cavium.o spi-cavium-thunderx.o
+obj-$(CONFIG_SPI_THUNDERX)		+= spi-thunderx.o
+obj-$(CONFIG_SPI_TOPCLIFF_PCH)		+= spi-topcliff-pch.o
+obj-$(CONFIG_SPI_TXX9)			+= spi-txx9.o
+obj-$(CONFIG_SPI_UNIPHIER)		+= spi-uniphier.o
+obj-$(CONFIG_SPI_XCOMM)		+= spi-xcomm.o
+obj-$(CONFIG_SPI_XILINX)		+= spi-xilinx.o
+obj-$(CONFIG_SPI_XLP)			+= spi-xlp.o
+obj-$(CONFIG_SPI_XTENSA_XTFPGA)		+= spi-xtensa-xtfpga.o
+obj-$(CONFIG_SPI_ZYNQMP_GQSPI)		+= spi-zynqmp-gqspi.o
+
+# SPI slave protocol handlers
+obj-$(CONFIG_SPI_SLAVE_TIME)		+= spi-slave-time.o
+obj-$(CONFIG_SPI_SLAVE_SYSTEM_CONTROL)	+= spi-slave-system-control.o
diff --git a/drivers/spi/internals.h b/drivers/spi/internals.h
new file mode 100644
index 0000000..4a28a83
--- /dev/null
+++ b/drivers/spi/internals.h
@@ -0,0 +1,43 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+/*
+ * Copyright (C) 2018 Exceet Electronics GmbH
+ * Copyright (C) 2018 Bootlin
+ *
+ * Author: Boris Brezillon <boris.brezillon@bootlin.com>
+ *
+ * Helpers needed by the spi or spi-mem logic. Should not be used outside of
+ * spi-mem.c and spi.c.
+ */
+
+#ifndef __LINUX_SPI_INTERNALS_H
+#define __LINUX_SPI_INTERNALS_H
+
+#include <linux/device.h>
+#include <linux/dma-direction.h>
+#include <linux/scatterlist.h>
+#include <linux/spi/spi.h>
+
+void spi_flush_queue(struct spi_controller *ctrl);
+
+#ifdef CONFIG_HAS_DMA
+int spi_map_buf(struct spi_controller *ctlr, struct device *dev,
+		struct sg_table *sgt, void *buf, size_t len,
+		enum dma_data_direction dir);
+void spi_unmap_buf(struct spi_controller *ctlr, struct device *dev,
+		   struct sg_table *sgt, enum dma_data_direction dir);
+#else /* !CONFIG_HAS_DMA */
+static inline int spi_map_buf(struct spi_controller *ctlr, struct device *dev,
+			      struct sg_table *sgt, void *buf, size_t len,
+			      enum dma_data_direction dir)
+{
+	return -EINVAL;
+}
+
+static inline void spi_unmap_buf(struct spi_controller *ctlr,
+				 struct device *dev, struct sg_table *sgt,
+				 enum dma_data_direction dir)
+{
+}
+#endif /* CONFIG_HAS_DMA */
+
+#endif /* __LINUX_SPI_INTERNALS_H */
diff --git a/drivers/spi/spi-altera.c b/drivers/spi/spi-altera.c
new file mode 100644
index 0000000..a5adf0d
--- /dev/null
+++ b/drivers/spi/spi-altera.c
@@ -0,0 +1,249 @@
+/*
+ * Altera SPI driver
+ *
+ * Copyright (C) 2008 Thomas Chou <thomas@wytron.com.tw>
+ *
+ * Based on spi_s3c24xx.c, which is:
+ * Copyright (c) 2006 Ben Dooks
+ * Copyright (c) 2006 Simtec Electronics
+ *	Ben Dooks <ben@simtec.co.uk>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/interrupt.h>
+#include <linux/errno.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/io.h>
+#include <linux/of.h>
+
+#define DRV_NAME "spi_altera"
+
+#define ALTERA_SPI_RXDATA	0
+#define ALTERA_SPI_TXDATA	4
+#define ALTERA_SPI_STATUS	8
+#define ALTERA_SPI_CONTROL	12
+#define ALTERA_SPI_SLAVE_SEL	20
+
+#define ALTERA_SPI_STATUS_ROE_MSK	0x8
+#define ALTERA_SPI_STATUS_TOE_MSK	0x10
+#define ALTERA_SPI_STATUS_TMT_MSK	0x20
+#define ALTERA_SPI_STATUS_TRDY_MSK	0x40
+#define ALTERA_SPI_STATUS_RRDY_MSK	0x80
+#define ALTERA_SPI_STATUS_E_MSK		0x100
+
+#define ALTERA_SPI_CONTROL_IROE_MSK	0x8
+#define ALTERA_SPI_CONTROL_ITOE_MSK	0x10
+#define ALTERA_SPI_CONTROL_ITRDY_MSK	0x40
+#define ALTERA_SPI_CONTROL_IRRDY_MSK	0x80
+#define ALTERA_SPI_CONTROL_IE_MSK	0x100
+#define ALTERA_SPI_CONTROL_SSO_MSK	0x400
+
+struct altera_spi {
+	void __iomem *base;
+	int irq;
+	int len;
+	int count;
+	int bytes_per_word;
+	unsigned long imr;
+
+	/* data buffers */
+	const unsigned char *tx;
+	unsigned char *rx;
+};
+
+static inline struct altera_spi *altera_spi_to_hw(struct spi_device *sdev)
+{
+	return spi_master_get_devdata(sdev->master);
+}
+
+static void altera_spi_set_cs(struct spi_device *spi, bool is_high)
+{
+	struct altera_spi *hw = altera_spi_to_hw(spi);
+
+	if (is_high) {
+		hw->imr &= ~ALTERA_SPI_CONTROL_SSO_MSK;
+		writel(hw->imr, hw->base + ALTERA_SPI_CONTROL);
+		writel(0, hw->base + ALTERA_SPI_SLAVE_SEL);
+	} else {
+		writel(BIT(spi->chip_select), hw->base + ALTERA_SPI_SLAVE_SEL);
+		hw->imr |= ALTERA_SPI_CONTROL_SSO_MSK;
+		writel(hw->imr, hw->base + ALTERA_SPI_CONTROL);
+	}
+}
+
+static void altera_spi_tx_word(struct altera_spi *hw)
+{
+	unsigned int txd = 0;
+
+	if (hw->tx) {
+		switch (hw->bytes_per_word) {
+		case 1:
+			txd = hw->tx[hw->count];
+			break;
+		case 2:
+			txd = (hw->tx[hw->count * 2]
+				| (hw->tx[hw->count * 2 + 1] << 8));
+			break;
+		}
+	}
+
+	writel(txd, hw->base + ALTERA_SPI_TXDATA);
+}
+
+static void altera_spi_rx_word(struct altera_spi *hw)
+{
+	unsigned int rxd;
+
+	rxd = readl(hw->base + ALTERA_SPI_RXDATA);
+	if (hw->rx) {
+		switch (hw->bytes_per_word) {
+		case 1:
+			hw->rx[hw->count] = rxd;
+			break;
+		case 2:
+			hw->rx[hw->count * 2] = rxd;
+			hw->rx[hw->count * 2 + 1] = rxd >> 8;
+			break;
+		}
+	}
+
+	hw->count++;
+}
+
+static int altera_spi_txrx(struct spi_master *master,
+	struct spi_device *spi, struct spi_transfer *t)
+{
+	struct altera_spi *hw = spi_master_get_devdata(master);
+
+	hw->tx = t->tx_buf;
+	hw->rx = t->rx_buf;
+	hw->count = 0;
+	hw->bytes_per_word = DIV_ROUND_UP(t->bits_per_word, 8);
+	hw->len = t->len / hw->bytes_per_word;
+
+	if (hw->irq >= 0) {
+		/* enable receive interrupt */
+		hw->imr |= ALTERA_SPI_CONTROL_IRRDY_MSK;
+		writel(hw->imr, hw->base + ALTERA_SPI_CONTROL);
+
+		/* send the first byte */
+		altera_spi_tx_word(hw);
+	} else {
+		while (hw->count < hw->len) {
+			altera_spi_tx_word(hw);
+
+			while (!(readl(hw->base + ALTERA_SPI_STATUS) &
+				 ALTERA_SPI_STATUS_RRDY_MSK))
+				cpu_relax();
+
+			altera_spi_rx_word(hw);
+		}
+		spi_finalize_current_transfer(master);
+	}
+
+	return t->len;
+}
+
+static irqreturn_t altera_spi_irq(int irq, void *dev)
+{
+	struct spi_master *master = dev;
+	struct altera_spi *hw = spi_master_get_devdata(master);
+
+	altera_spi_rx_word(hw);
+
+	if (hw->count < hw->len) {
+		altera_spi_tx_word(hw);
+	} else {
+		/* disable receive interrupt */
+		hw->imr &= ~ALTERA_SPI_CONTROL_IRRDY_MSK;
+		writel(hw->imr, hw->base + ALTERA_SPI_CONTROL);
+
+		spi_finalize_current_transfer(master);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int altera_spi_probe(struct platform_device *pdev)
+{
+	struct altera_spi *hw;
+	struct spi_master *master;
+	struct resource *res;
+	int err = -ENODEV;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct altera_spi));
+	if (!master)
+		return err;
+
+	/* setup the master state. */
+	master->bus_num = pdev->id;
+	master->num_chipselect = 16;
+	master->mode_bits = SPI_CS_HIGH;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(1, 16);
+	master->dev.of_node = pdev->dev.of_node;
+	master->transfer_one = altera_spi_txrx;
+	master->set_cs = altera_spi_set_cs;
+
+	hw = spi_master_get_devdata(master);
+
+	/* find and map our resources */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	hw->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(hw->base)) {
+		err = PTR_ERR(hw->base);
+		goto exit;
+	}
+	/* program defaults into the registers */
+	hw->imr = 0;		/* disable spi interrupts */
+	writel(hw->imr, hw->base + ALTERA_SPI_CONTROL);
+	writel(0, hw->base + ALTERA_SPI_STATUS);	/* clear status reg */
+	if (readl(hw->base + ALTERA_SPI_STATUS) & ALTERA_SPI_STATUS_RRDY_MSK)
+		readl(hw->base + ALTERA_SPI_RXDATA);	/* flush rxdata */
+	/* irq is optional */
+	hw->irq = platform_get_irq(pdev, 0);
+	if (hw->irq >= 0) {
+		err = devm_request_irq(&pdev->dev, hw->irq, altera_spi_irq, 0,
+				       pdev->name, master);
+		if (err)
+			goto exit;
+	}
+
+	err = devm_spi_register_master(&pdev->dev, master);
+	if (err)
+		goto exit;
+	dev_info(&pdev->dev, "base %p, irq %d\n", hw->base, hw->irq);
+
+	return 0;
+exit:
+	spi_master_put(master);
+	return err;
+}
+
+#ifdef CONFIG_OF
+static const struct of_device_id altera_spi_match[] = {
+	{ .compatible = "ALTR,spi-1.0", },
+	{ .compatible = "altr,spi-1.0", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, altera_spi_match);
+#endif /* CONFIG_OF */
+
+static struct platform_driver altera_spi_driver = {
+	.probe = altera_spi_probe,
+	.driver = {
+		.name = DRV_NAME,
+		.pm = NULL,
+		.of_match_table = of_match_ptr(altera_spi_match),
+	},
+};
+module_platform_driver(altera_spi_driver);
+
+MODULE_DESCRIPTION("Altera SPI driver");
+MODULE_AUTHOR("Thomas Chou <thomas@wytron.com.tw>");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:" DRV_NAME);
diff --git a/drivers/spi/spi-armada-3700.c b/drivers/spi/spi-armada-3700.c
new file mode 100644
index 0000000..7dcb14d
--- /dev/null
+++ b/drivers/spi/spi-armada-3700.c
@@ -0,0 +1,946 @@
+/*
+ * Marvell Armada-3700 SPI controller driver
+ *
+ * Copyright (C) 2016 Marvell Ltd.
+ *
+ * Author: Wilson Ding <dingwei@marvell.com>
+ * Author: Romain Perier <romain.perier@free-electrons.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_irq.h>
+#include <linux/of_device.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/spi/spi.h>
+
+#define DRIVER_NAME			"armada_3700_spi"
+
+#define A3700_SPI_MAX_SPEED_HZ		100000000
+#define A3700_SPI_MAX_PRESCALE		30
+#define A3700_SPI_TIMEOUT		10
+
+/* SPI Register Offest */
+#define A3700_SPI_IF_CTRL_REG		0x00
+#define A3700_SPI_IF_CFG_REG		0x04
+#define A3700_SPI_DATA_OUT_REG		0x08
+#define A3700_SPI_DATA_IN_REG		0x0C
+#define A3700_SPI_IF_INST_REG		0x10
+#define A3700_SPI_IF_ADDR_REG		0x14
+#define A3700_SPI_IF_RMODE_REG		0x18
+#define A3700_SPI_IF_HDR_CNT_REG	0x1C
+#define A3700_SPI_IF_DIN_CNT_REG	0x20
+#define A3700_SPI_IF_TIME_REG		0x24
+#define A3700_SPI_INT_STAT_REG		0x28
+#define A3700_SPI_INT_MASK_REG		0x2C
+
+/* A3700_SPI_IF_CTRL_REG */
+#define A3700_SPI_EN			BIT(16)
+#define A3700_SPI_ADDR_NOT_CONFIG	BIT(12)
+#define A3700_SPI_WFIFO_OVERFLOW	BIT(11)
+#define A3700_SPI_WFIFO_UNDERFLOW	BIT(10)
+#define A3700_SPI_RFIFO_OVERFLOW	BIT(9)
+#define A3700_SPI_RFIFO_UNDERFLOW	BIT(8)
+#define A3700_SPI_WFIFO_FULL		BIT(7)
+#define A3700_SPI_WFIFO_EMPTY		BIT(6)
+#define A3700_SPI_RFIFO_FULL		BIT(5)
+#define A3700_SPI_RFIFO_EMPTY		BIT(4)
+#define A3700_SPI_WFIFO_RDY		BIT(3)
+#define A3700_SPI_RFIFO_RDY		BIT(2)
+#define A3700_SPI_XFER_RDY		BIT(1)
+#define A3700_SPI_XFER_DONE		BIT(0)
+
+/* A3700_SPI_IF_CFG_REG */
+#define A3700_SPI_WFIFO_THRS		BIT(28)
+#define A3700_SPI_RFIFO_THRS		BIT(24)
+#define A3700_SPI_AUTO_CS		BIT(20)
+#define A3700_SPI_DMA_RD_EN		BIT(18)
+#define A3700_SPI_FIFO_MODE		BIT(17)
+#define A3700_SPI_SRST			BIT(16)
+#define A3700_SPI_XFER_START		BIT(15)
+#define A3700_SPI_XFER_STOP		BIT(14)
+#define A3700_SPI_INST_PIN		BIT(13)
+#define A3700_SPI_ADDR_PIN		BIT(12)
+#define A3700_SPI_DATA_PIN1		BIT(11)
+#define A3700_SPI_DATA_PIN0		BIT(10)
+#define A3700_SPI_FIFO_FLUSH		BIT(9)
+#define A3700_SPI_RW_EN			BIT(8)
+#define A3700_SPI_CLK_POL		BIT(7)
+#define A3700_SPI_CLK_PHA		BIT(6)
+#define A3700_SPI_BYTE_LEN		BIT(5)
+#define A3700_SPI_CLK_PRESCALE		BIT(0)
+#define A3700_SPI_CLK_PRESCALE_MASK	(0x1f)
+#define A3700_SPI_CLK_EVEN_OFFS		(0x10)
+
+#define A3700_SPI_WFIFO_THRS_BIT	28
+#define A3700_SPI_RFIFO_THRS_BIT	24
+#define A3700_SPI_FIFO_THRS_MASK	0x7
+
+#define A3700_SPI_DATA_PIN_MASK		0x3
+
+/* A3700_SPI_IF_HDR_CNT_REG */
+#define A3700_SPI_DUMMY_CNT_BIT		12
+#define A3700_SPI_DUMMY_CNT_MASK	0x7
+#define A3700_SPI_RMODE_CNT_BIT		8
+#define A3700_SPI_RMODE_CNT_MASK	0x3
+#define A3700_SPI_ADDR_CNT_BIT		4
+#define A3700_SPI_ADDR_CNT_MASK		0x7
+#define A3700_SPI_INSTR_CNT_BIT		0
+#define A3700_SPI_INSTR_CNT_MASK	0x3
+
+/* A3700_SPI_IF_TIME_REG */
+#define A3700_SPI_CLK_CAPT_EDGE		BIT(7)
+
+struct a3700_spi {
+	struct spi_master *master;
+	void __iomem *base;
+	struct clk *clk;
+	unsigned int irq;
+	unsigned int flags;
+	bool xmit_data;
+	const u8 *tx_buf;
+	u8 *rx_buf;
+	size_t buf_len;
+	u8 byte_len;
+	u32 wait_mask;
+	struct completion done;
+};
+
+static u32 spireg_read(struct a3700_spi *a3700_spi, u32 offset)
+{
+	return readl(a3700_spi->base + offset);
+}
+
+static void spireg_write(struct a3700_spi *a3700_spi, u32 offset, u32 data)
+{
+	writel(data, a3700_spi->base + offset);
+}
+
+static void a3700_spi_auto_cs_unset(struct a3700_spi *a3700_spi)
+{
+	u32 val;
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+	val &= ~A3700_SPI_AUTO_CS;
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+}
+
+static void a3700_spi_activate_cs(struct a3700_spi *a3700_spi, unsigned int cs)
+{
+	u32 val;
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CTRL_REG);
+	val |= (A3700_SPI_EN << cs);
+	spireg_write(a3700_spi, A3700_SPI_IF_CTRL_REG, val);
+}
+
+static void a3700_spi_deactivate_cs(struct a3700_spi *a3700_spi,
+				    unsigned int cs)
+{
+	u32 val;
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CTRL_REG);
+	val &= ~(A3700_SPI_EN << cs);
+	spireg_write(a3700_spi, A3700_SPI_IF_CTRL_REG, val);
+}
+
+static int a3700_spi_pin_mode_set(struct a3700_spi *a3700_spi,
+				  unsigned int pin_mode, bool receiving)
+{
+	u32 val;
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+	val &= ~(A3700_SPI_INST_PIN | A3700_SPI_ADDR_PIN);
+	val &= ~(A3700_SPI_DATA_PIN0 | A3700_SPI_DATA_PIN1);
+
+	switch (pin_mode) {
+	case SPI_NBITS_SINGLE:
+		break;
+	case SPI_NBITS_DUAL:
+		val |= A3700_SPI_DATA_PIN0;
+		break;
+	case SPI_NBITS_QUAD:
+		val |= A3700_SPI_DATA_PIN1;
+		/* RX during address reception uses 4-pin */
+		if (receiving)
+			val |= A3700_SPI_ADDR_PIN;
+		break;
+	default:
+		dev_err(&a3700_spi->master->dev, "wrong pin mode %u", pin_mode);
+		return -EINVAL;
+	}
+
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+
+	return 0;
+}
+
+static void a3700_spi_fifo_mode_set(struct a3700_spi *a3700_spi, bool enable)
+{
+	u32 val;
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+	if (enable)
+		val |= A3700_SPI_FIFO_MODE;
+	else
+		val &= ~A3700_SPI_FIFO_MODE;
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+}
+
+static void a3700_spi_mode_set(struct a3700_spi *a3700_spi,
+			       unsigned int mode_bits)
+{
+	u32 val;
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+
+	if (mode_bits & SPI_CPOL)
+		val |= A3700_SPI_CLK_POL;
+	else
+		val &= ~A3700_SPI_CLK_POL;
+
+	if (mode_bits & SPI_CPHA)
+		val |= A3700_SPI_CLK_PHA;
+	else
+		val &= ~A3700_SPI_CLK_PHA;
+
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+}
+
+static void a3700_spi_clock_set(struct a3700_spi *a3700_spi,
+				unsigned int speed_hz)
+{
+	u32 val;
+	u32 prescale;
+
+	prescale = DIV_ROUND_UP(clk_get_rate(a3700_spi->clk), speed_hz);
+
+	/* For prescaler values over 15, we can only set it by steps of 2.
+	 * Starting from A3700_SPI_CLK_EVEN_OFFS, we set values from 0 up to
+	 * 30. We only use this range from 16 to 30.
+	 */
+	if (prescale > 15)
+		prescale = A3700_SPI_CLK_EVEN_OFFS + DIV_ROUND_UP(prescale, 2);
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+	val = val & ~A3700_SPI_CLK_PRESCALE_MASK;
+
+	val = val | (prescale & A3700_SPI_CLK_PRESCALE_MASK);
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+
+	if (prescale <= 2) {
+		val = spireg_read(a3700_spi, A3700_SPI_IF_TIME_REG);
+		val |= A3700_SPI_CLK_CAPT_EDGE;
+		spireg_write(a3700_spi, A3700_SPI_IF_TIME_REG, val);
+	}
+}
+
+static void a3700_spi_bytelen_set(struct a3700_spi *a3700_spi, unsigned int len)
+{
+	u32 val;
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+	if (len == 4)
+		val |= A3700_SPI_BYTE_LEN;
+	else
+		val &= ~A3700_SPI_BYTE_LEN;
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+
+	a3700_spi->byte_len = len;
+}
+
+static int a3700_spi_fifo_flush(struct a3700_spi *a3700_spi)
+{
+	int timeout = A3700_SPI_TIMEOUT;
+	u32 val;
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+	val |= A3700_SPI_FIFO_FLUSH;
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+
+	while (--timeout) {
+		val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+		if (!(val & A3700_SPI_FIFO_FLUSH))
+			return 0;
+		udelay(1);
+	}
+
+	return -ETIMEDOUT;
+}
+
+static int a3700_spi_init(struct a3700_spi *a3700_spi)
+{
+	struct spi_master *master = a3700_spi->master;
+	u32 val;
+	int i, ret = 0;
+
+	/* Reset SPI unit */
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+	val |= A3700_SPI_SRST;
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+
+	udelay(A3700_SPI_TIMEOUT);
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+	val &= ~A3700_SPI_SRST;
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+
+	/* Disable AUTO_CS and deactivate all chip-selects */
+	a3700_spi_auto_cs_unset(a3700_spi);
+	for (i = 0; i < master->num_chipselect; i++)
+		a3700_spi_deactivate_cs(a3700_spi, i);
+
+	/* Enable FIFO mode */
+	a3700_spi_fifo_mode_set(a3700_spi, true);
+
+	/* Set SPI mode */
+	a3700_spi_mode_set(a3700_spi, master->mode_bits);
+
+	/* Reset counters */
+	spireg_write(a3700_spi, A3700_SPI_IF_HDR_CNT_REG, 0);
+	spireg_write(a3700_spi, A3700_SPI_IF_DIN_CNT_REG, 0);
+
+	/* Mask the interrupts and clear cause bits */
+	spireg_write(a3700_spi, A3700_SPI_INT_MASK_REG, 0);
+	spireg_write(a3700_spi, A3700_SPI_INT_STAT_REG, ~0U);
+
+	return ret;
+}
+
+static irqreturn_t a3700_spi_interrupt(int irq, void *dev_id)
+{
+	struct spi_master *master = dev_id;
+	struct a3700_spi *a3700_spi;
+	u32 cause;
+
+	a3700_spi = spi_master_get_devdata(master);
+
+	/* Get interrupt causes */
+	cause = spireg_read(a3700_spi, A3700_SPI_INT_STAT_REG);
+
+	if (!cause || !(a3700_spi->wait_mask & cause))
+		return IRQ_NONE;
+
+	/* mask and acknowledge the SPI interrupts */
+	spireg_write(a3700_spi, A3700_SPI_INT_MASK_REG, 0);
+	spireg_write(a3700_spi, A3700_SPI_INT_STAT_REG, cause);
+
+	/* Wake up the transfer */
+	complete(&a3700_spi->done);
+
+	return IRQ_HANDLED;
+}
+
+static bool a3700_spi_wait_completion(struct spi_device *spi)
+{
+	struct a3700_spi *a3700_spi;
+	unsigned int timeout;
+	unsigned int ctrl_reg;
+	unsigned long timeout_jiffies;
+
+	a3700_spi = spi_master_get_devdata(spi->master);
+
+	/* SPI interrupt is edge-triggered, which means an interrupt will
+	 * be generated only when detecting a specific status bit changed
+	 * from '0' to '1'. So when we start waiting for a interrupt, we
+	 * need to check status bit in control reg first, if it is already 1,
+	 * then we do not need to wait for interrupt
+	 */
+	ctrl_reg = spireg_read(a3700_spi, A3700_SPI_IF_CTRL_REG);
+	if (a3700_spi->wait_mask & ctrl_reg)
+		return true;
+
+	reinit_completion(&a3700_spi->done);
+
+	spireg_write(a3700_spi, A3700_SPI_INT_MASK_REG,
+		     a3700_spi->wait_mask);
+
+	timeout_jiffies = msecs_to_jiffies(A3700_SPI_TIMEOUT);
+	timeout = wait_for_completion_timeout(&a3700_spi->done,
+					      timeout_jiffies);
+
+	a3700_spi->wait_mask = 0;
+
+	if (timeout)
+		return true;
+
+	/* there might be the case that right after we checked the
+	 * status bits in this routine and before start to wait for
+	 * interrupt by wait_for_completion_timeout, the interrupt
+	 * happens, to avoid missing it we need to double check
+	 * status bits in control reg, if it is already 1, then
+	 * consider that we have the interrupt successfully and
+	 * return true.
+	 */
+	ctrl_reg = spireg_read(a3700_spi, A3700_SPI_IF_CTRL_REG);
+	if (a3700_spi->wait_mask & ctrl_reg)
+		return true;
+
+	spireg_write(a3700_spi, A3700_SPI_INT_MASK_REG, 0);
+
+	/* Timeout was reached */
+	return false;
+}
+
+static bool a3700_spi_transfer_wait(struct spi_device *spi,
+				    unsigned int bit_mask)
+{
+	struct a3700_spi *a3700_spi;
+
+	a3700_spi = spi_master_get_devdata(spi->master);
+	a3700_spi->wait_mask = bit_mask;
+
+	return a3700_spi_wait_completion(spi);
+}
+
+static void a3700_spi_fifo_thres_set(struct a3700_spi *a3700_spi,
+				     unsigned int bytes)
+{
+	u32 val;
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+	val &= ~(A3700_SPI_FIFO_THRS_MASK << A3700_SPI_RFIFO_THRS_BIT);
+	val |= (bytes - 1) << A3700_SPI_RFIFO_THRS_BIT;
+	val &= ~(A3700_SPI_FIFO_THRS_MASK << A3700_SPI_WFIFO_THRS_BIT);
+	val |= (7 - bytes) << A3700_SPI_WFIFO_THRS_BIT;
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+}
+
+static void a3700_spi_transfer_setup(struct spi_device *spi,
+				     struct spi_transfer *xfer)
+{
+	struct a3700_spi *a3700_spi;
+
+	a3700_spi = spi_master_get_devdata(spi->master);
+
+	a3700_spi_clock_set(a3700_spi, xfer->speed_hz);
+
+	/* Use 4 bytes long transfers. Each transfer method has its way to deal
+	 * with the remaining bytes for non 4-bytes aligned transfers.
+	 */
+	a3700_spi_bytelen_set(a3700_spi, 4);
+
+	/* Initialize the working buffers */
+	a3700_spi->tx_buf  = xfer->tx_buf;
+	a3700_spi->rx_buf  = xfer->rx_buf;
+	a3700_spi->buf_len = xfer->len;
+}
+
+static void a3700_spi_set_cs(struct spi_device *spi, bool enable)
+{
+	struct a3700_spi *a3700_spi = spi_master_get_devdata(spi->master);
+
+	if (!enable)
+		a3700_spi_activate_cs(a3700_spi, spi->chip_select);
+	else
+		a3700_spi_deactivate_cs(a3700_spi, spi->chip_select);
+}
+
+static void a3700_spi_header_set(struct a3700_spi *a3700_spi)
+{
+	unsigned int addr_cnt;
+	u32 val = 0;
+
+	/* Clear the header registers */
+	spireg_write(a3700_spi, A3700_SPI_IF_INST_REG, 0);
+	spireg_write(a3700_spi, A3700_SPI_IF_ADDR_REG, 0);
+	spireg_write(a3700_spi, A3700_SPI_IF_RMODE_REG, 0);
+	spireg_write(a3700_spi, A3700_SPI_IF_HDR_CNT_REG, 0);
+
+	/* Set header counters */
+	if (a3700_spi->tx_buf) {
+		/*
+		 * when tx data is not 4 bytes aligned, there will be unexpected
+		 * bytes out of SPI output register, since it always shifts out
+		 * as whole 4 bytes. This might cause incorrect transaction with
+		 * some devices. To avoid that, use SPI header count feature to
+		 * transfer up to 3 bytes of data first, and then make the rest
+		 * of data 4-byte aligned.
+		 */
+		addr_cnt = a3700_spi->buf_len % 4;
+		if (addr_cnt) {
+			val = (addr_cnt & A3700_SPI_ADDR_CNT_MASK)
+				<< A3700_SPI_ADDR_CNT_BIT;
+			spireg_write(a3700_spi, A3700_SPI_IF_HDR_CNT_REG, val);
+
+			/* Update the buffer length to be transferred */
+			a3700_spi->buf_len -= addr_cnt;
+
+			/* transfer 1~3 bytes through address count */
+			val = 0;
+			while (addr_cnt--) {
+				val = (val << 8) | a3700_spi->tx_buf[0];
+				a3700_spi->tx_buf++;
+			}
+			spireg_write(a3700_spi, A3700_SPI_IF_ADDR_REG, val);
+		}
+	}
+}
+
+static int a3700_is_wfifo_full(struct a3700_spi *a3700_spi)
+{
+	u32 val;
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CTRL_REG);
+	return (val & A3700_SPI_WFIFO_FULL);
+}
+
+static int a3700_spi_fifo_write(struct a3700_spi *a3700_spi)
+{
+	u32 val;
+
+	while (!a3700_is_wfifo_full(a3700_spi) && a3700_spi->buf_len) {
+		val = *(u32 *)a3700_spi->tx_buf;
+		spireg_write(a3700_spi, A3700_SPI_DATA_OUT_REG, val);
+		a3700_spi->buf_len -= 4;
+		a3700_spi->tx_buf += 4;
+	}
+
+	return 0;
+}
+
+static int a3700_is_rfifo_empty(struct a3700_spi *a3700_spi)
+{
+	u32 val = spireg_read(a3700_spi, A3700_SPI_IF_CTRL_REG);
+
+	return (val & A3700_SPI_RFIFO_EMPTY);
+}
+
+static int a3700_spi_fifo_read(struct a3700_spi *a3700_spi)
+{
+	u32 val;
+
+	while (!a3700_is_rfifo_empty(a3700_spi) && a3700_spi->buf_len) {
+		val = spireg_read(a3700_spi, A3700_SPI_DATA_IN_REG);
+		if (a3700_spi->buf_len >= 4) {
+
+			memcpy(a3700_spi->rx_buf, &val, 4);
+
+			a3700_spi->buf_len -= 4;
+			a3700_spi->rx_buf += 4;
+		} else {
+			/*
+			 * When remain bytes is not larger than 4, we should
+			 * avoid memory overwriting and just write the left rx
+			 * buffer bytes.
+			 */
+			while (a3700_spi->buf_len) {
+				*a3700_spi->rx_buf = val & 0xff;
+				val >>= 8;
+
+				a3700_spi->buf_len--;
+				a3700_spi->rx_buf++;
+			}
+		}
+	}
+
+	return 0;
+}
+
+static void a3700_spi_transfer_abort_fifo(struct a3700_spi *a3700_spi)
+{
+	int timeout = A3700_SPI_TIMEOUT;
+	u32 val;
+
+	val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+	val |= A3700_SPI_XFER_STOP;
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+
+	while (--timeout) {
+		val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+		if (!(val & A3700_SPI_XFER_START))
+			break;
+		udelay(1);
+	}
+
+	a3700_spi_fifo_flush(a3700_spi);
+
+	val &= ~A3700_SPI_XFER_STOP;
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+}
+
+static int a3700_spi_prepare_message(struct spi_master *master,
+				     struct spi_message *message)
+{
+	struct a3700_spi *a3700_spi = spi_master_get_devdata(master);
+	struct spi_device *spi = message->spi;
+	int ret;
+
+	ret = clk_enable(a3700_spi->clk);
+	if (ret) {
+		dev_err(&spi->dev, "failed to enable clk with error %d\n", ret);
+		return ret;
+	}
+
+	/* Flush the FIFOs */
+	ret = a3700_spi_fifo_flush(a3700_spi);
+	if (ret)
+		return ret;
+
+	a3700_spi_mode_set(a3700_spi, spi->mode);
+
+	return 0;
+}
+
+static int a3700_spi_transfer_one_fifo(struct spi_master *master,
+				  struct spi_device *spi,
+				  struct spi_transfer *xfer)
+{
+	struct a3700_spi *a3700_spi = spi_master_get_devdata(master);
+	int ret = 0, timeout = A3700_SPI_TIMEOUT;
+	unsigned int nbits = 0, byte_len;
+	u32 val;
+
+	/* Make sure we use FIFO mode */
+	a3700_spi_fifo_mode_set(a3700_spi, true);
+
+	/* Configure FIFO thresholds */
+	byte_len = xfer->bits_per_word >> 3;
+	a3700_spi_fifo_thres_set(a3700_spi, byte_len);
+
+	if (xfer->tx_buf)
+		nbits = xfer->tx_nbits;
+	else if (xfer->rx_buf)
+		nbits = xfer->rx_nbits;
+
+	a3700_spi_pin_mode_set(a3700_spi, nbits, xfer->rx_buf ? true : false);
+
+	/* Flush the FIFOs */
+	a3700_spi_fifo_flush(a3700_spi);
+
+	/* Transfer first bytes of data when buffer is not 4-byte aligned */
+	a3700_spi_header_set(a3700_spi);
+
+	if (xfer->rx_buf) {
+		/* Clear WFIFO, since it's last 2 bytes are shifted out during
+		 * a read operation
+		 */
+		spireg_write(a3700_spi, A3700_SPI_DATA_OUT_REG, 0);
+
+		/* Set read data length */
+		spireg_write(a3700_spi, A3700_SPI_IF_DIN_CNT_REG,
+			     a3700_spi->buf_len);
+		/* Start READ transfer */
+		val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+		val &= ~A3700_SPI_RW_EN;
+		val |= A3700_SPI_XFER_START;
+		spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+	} else if (xfer->tx_buf) {
+		/* Start Write transfer */
+		val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+		val |= (A3700_SPI_XFER_START | A3700_SPI_RW_EN);
+		spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+
+		/*
+		 * If there are data to be written to the SPI device, xmit_data
+		 * flag is set true; otherwise the instruction in SPI_INSTR does
+		 * not require data to be written to the SPI device, then
+		 * xmit_data flag is set false.
+		 */
+		a3700_spi->xmit_data = (a3700_spi->buf_len != 0);
+	}
+
+	while (a3700_spi->buf_len) {
+		if (a3700_spi->tx_buf) {
+			/* Wait wfifo ready */
+			if (!a3700_spi_transfer_wait(spi,
+						     A3700_SPI_WFIFO_RDY)) {
+				dev_err(&spi->dev,
+					"wait wfifo ready timed out\n");
+				ret = -ETIMEDOUT;
+				goto error;
+			}
+			/* Fill up the wfifo */
+			ret = a3700_spi_fifo_write(a3700_spi);
+			if (ret)
+				goto error;
+		} else if (a3700_spi->rx_buf) {
+			/* Wait rfifo ready */
+			if (!a3700_spi_transfer_wait(spi,
+						     A3700_SPI_RFIFO_RDY)) {
+				dev_err(&spi->dev,
+					"wait rfifo ready timed out\n");
+				ret = -ETIMEDOUT;
+				goto error;
+			}
+			/* Drain out the rfifo */
+			ret = a3700_spi_fifo_read(a3700_spi);
+			if (ret)
+				goto error;
+		}
+	}
+
+	/*
+	 * Stop a write transfer in fifo mode:
+	 *	- wait all the bytes in wfifo to be shifted out
+	 *	 - set XFER_STOP bit
+	 *	- wait XFER_START bit clear
+	 *	- clear XFER_STOP bit
+	 * Stop a read transfer in fifo mode:
+	 *	- the hardware is to reset the XFER_START bit
+	 *	   after the number of bytes indicated in DIN_CNT
+	 *	   register
+	 *	- just wait XFER_START bit clear
+	 */
+	if (a3700_spi->tx_buf) {
+		if (a3700_spi->xmit_data) {
+			/*
+			 * If there are data written to the SPI device, wait
+			 * until SPI_WFIFO_EMPTY is 1 to wait for all data to
+			 * transfer out of write FIFO.
+			 */
+			if (!a3700_spi_transfer_wait(spi,
+						     A3700_SPI_WFIFO_EMPTY)) {
+				dev_err(&spi->dev, "wait wfifo empty timed out\n");
+				return -ETIMEDOUT;
+			}
+		}
+
+		if (!a3700_spi_transfer_wait(spi, A3700_SPI_XFER_RDY)) {
+			dev_err(&spi->dev, "wait xfer ready timed out\n");
+			return -ETIMEDOUT;
+		}
+
+		val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+		val |= A3700_SPI_XFER_STOP;
+		spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+	}
+
+	while (--timeout) {
+		val = spireg_read(a3700_spi, A3700_SPI_IF_CFG_REG);
+		if (!(val & A3700_SPI_XFER_START))
+			break;
+		udelay(1);
+	}
+
+	if (timeout == 0) {
+		dev_err(&spi->dev, "wait transfer start clear timed out\n");
+		ret = -ETIMEDOUT;
+		goto error;
+	}
+
+	val &= ~A3700_SPI_XFER_STOP;
+	spireg_write(a3700_spi, A3700_SPI_IF_CFG_REG, val);
+	goto out;
+
+error:
+	a3700_spi_transfer_abort_fifo(a3700_spi);
+out:
+	spi_finalize_current_transfer(master);
+
+	return ret;
+}
+
+static int a3700_spi_transfer_one_full_duplex(struct spi_master *master,
+				  struct spi_device *spi,
+				  struct spi_transfer *xfer)
+{
+	struct a3700_spi *a3700_spi = spi_master_get_devdata(master);
+	u32 val;
+
+	/* Disable FIFO mode */
+	a3700_spi_fifo_mode_set(a3700_spi, false);
+
+	while (a3700_spi->buf_len) {
+
+		/* When we have less than 4 bytes to transfer, switch to 1 byte
+		 * mode. This is reset after each transfer
+		 */
+		if (a3700_spi->buf_len < 4)
+			a3700_spi_bytelen_set(a3700_spi, 1);
+
+		if (a3700_spi->byte_len == 1)
+			val = *a3700_spi->tx_buf;
+		else
+			val = *(u32 *)a3700_spi->tx_buf;
+
+		spireg_write(a3700_spi, A3700_SPI_DATA_OUT_REG, val);
+
+		/* Wait for all the data to be shifted in / out */
+		while (!(spireg_read(a3700_spi, A3700_SPI_IF_CTRL_REG) &
+				A3700_SPI_XFER_DONE))
+			cpu_relax();
+
+		val = spireg_read(a3700_spi, A3700_SPI_DATA_IN_REG);
+
+		memcpy(a3700_spi->rx_buf, &val, a3700_spi->byte_len);
+
+		a3700_spi->buf_len -= a3700_spi->byte_len;
+		a3700_spi->tx_buf += a3700_spi->byte_len;
+		a3700_spi->rx_buf += a3700_spi->byte_len;
+
+	}
+
+	spi_finalize_current_transfer(master);
+
+	return 0;
+}
+
+static int a3700_spi_transfer_one(struct spi_master *master,
+				  struct spi_device *spi,
+				  struct spi_transfer *xfer)
+{
+	a3700_spi_transfer_setup(spi, xfer);
+
+	if (xfer->tx_buf && xfer->rx_buf)
+		return a3700_spi_transfer_one_full_duplex(master, spi, xfer);
+
+	return a3700_spi_transfer_one_fifo(master, spi, xfer);
+}
+
+static int a3700_spi_unprepare_message(struct spi_master *master,
+				       struct spi_message *message)
+{
+	struct a3700_spi *a3700_spi = spi_master_get_devdata(master);
+
+	clk_disable(a3700_spi->clk);
+
+	return 0;
+}
+
+static const struct of_device_id a3700_spi_dt_ids[] = {
+	{ .compatible = "marvell,armada-3700-spi", .data = NULL },
+	{},
+};
+
+MODULE_DEVICE_TABLE(of, a3700_spi_dt_ids);
+
+static int a3700_spi_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct device_node *of_node = dev->of_node;
+	struct resource *res;
+	struct spi_master *master;
+	struct a3700_spi *spi;
+	u32 num_cs = 0;
+	int irq, ret = 0;
+
+	master = spi_alloc_master(dev, sizeof(*spi));
+	if (!master) {
+		dev_err(dev, "master allocation failed\n");
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	if (of_property_read_u32(of_node, "num-cs", &num_cs)) {
+		dev_err(dev, "could not find num-cs\n");
+		ret = -ENXIO;
+		goto error;
+	}
+
+	master->bus_num = pdev->id;
+	master->dev.of_node = of_node;
+	master->mode_bits = SPI_MODE_3;
+	master->num_chipselect = num_cs;
+	master->bits_per_word_mask = SPI_BPW_MASK(8) | SPI_BPW_MASK(32);
+	master->prepare_message =  a3700_spi_prepare_message;
+	master->transfer_one = a3700_spi_transfer_one;
+	master->unprepare_message = a3700_spi_unprepare_message;
+	master->set_cs = a3700_spi_set_cs;
+	master->mode_bits |= (SPI_RX_DUAL | SPI_TX_DUAL |
+			      SPI_RX_QUAD | SPI_TX_QUAD);
+
+	platform_set_drvdata(pdev, master);
+
+	spi = spi_master_get_devdata(master);
+	memset(spi, 0, sizeof(struct a3700_spi));
+
+	spi->master = master;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	spi->base = devm_ioremap_resource(dev, res);
+	if (IS_ERR(spi->base)) {
+		ret = PTR_ERR(spi->base);
+		goto error;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(dev, "could not get irq: %d\n", irq);
+		ret = -ENXIO;
+		goto error;
+	}
+	spi->irq = irq;
+
+	init_completion(&spi->done);
+
+	spi->clk = devm_clk_get(dev, NULL);
+	if (IS_ERR(spi->clk)) {
+		dev_err(dev, "could not find clk: %ld\n", PTR_ERR(spi->clk));
+		goto error;
+	}
+
+	ret = clk_prepare(spi->clk);
+	if (ret) {
+		dev_err(dev, "could not prepare clk: %d\n", ret);
+		goto error;
+	}
+
+	master->max_speed_hz = min_t(unsigned long, A3700_SPI_MAX_SPEED_HZ,
+					clk_get_rate(spi->clk));
+	master->min_speed_hz = DIV_ROUND_UP(clk_get_rate(spi->clk),
+						A3700_SPI_MAX_PRESCALE);
+
+	ret = a3700_spi_init(spi);
+	if (ret)
+		goto error_clk;
+
+	ret = devm_request_irq(dev, spi->irq, a3700_spi_interrupt, 0,
+			       dev_name(dev), master);
+	if (ret) {
+		dev_err(dev, "could not request IRQ: %d\n", ret);
+		goto error_clk;
+	}
+
+	ret = devm_spi_register_master(dev, master);
+	if (ret) {
+		dev_err(dev, "Failed to register master\n");
+		goto error_clk;
+	}
+
+	return 0;
+
+error_clk:
+	clk_disable_unprepare(spi->clk);
+error:
+	spi_master_put(master);
+out:
+	return ret;
+}
+
+static int a3700_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct a3700_spi *spi = spi_master_get_devdata(master);
+
+	clk_unprepare(spi->clk);
+
+	return 0;
+}
+
+static struct platform_driver a3700_spi_driver = {
+	.driver = {
+		.name	= DRIVER_NAME,
+		.of_match_table = of_match_ptr(a3700_spi_dt_ids),
+	},
+	.probe		= a3700_spi_probe,
+	.remove		= a3700_spi_remove,
+};
+
+module_platform_driver(a3700_spi_driver);
+
+MODULE_DESCRIPTION("Armada-3700 SPI driver");
+MODULE_AUTHOR("Wilson Ding <dingwei@marvell.com>");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:" DRIVER_NAME);
diff --git a/drivers/spi/spi-ath79.c b/drivers/spi/spi-ath79.c
new file mode 100644
index 0000000..3f6b657
--- /dev/null
+++ b/drivers/spi/spi-ath79.c
@@ -0,0 +1,323 @@
+/*
+ * SPI controller driver for the Atheros AR71XX/AR724X/AR913X SoCs
+ *
+ * Copyright (C) 2009-2011 Gabor Juhos <juhosg@openwrt.org>
+ *
+ * This driver has been based on the spi-gpio.c:
+ *	Copyright (C) 2006,2008 David Brownell
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/delay.h>
+#include <linux/spinlock.h>
+#include <linux/platform_device.h>
+#include <linux/io.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/bitops.h>
+#include <linux/gpio.h>
+#include <linux/clk.h>
+#include <linux/err.h>
+
+#include <asm/mach-ath79/ar71xx_regs.h>
+#include <asm/mach-ath79/ath79_spi_platform.h>
+
+#define DRV_NAME	"ath79-spi"
+
+#define ATH79_SPI_RRW_DELAY_FACTOR	12000
+#define MHZ				(1000 * 1000)
+
+struct ath79_spi {
+	struct spi_bitbang	bitbang;
+	u32			ioc_base;
+	u32			reg_ctrl;
+	void __iomem		*base;
+	struct clk		*clk;
+	unsigned int		rrw_delay;
+};
+
+static inline u32 ath79_spi_rr(struct ath79_spi *sp, unsigned int reg)
+{
+	return ioread32(sp->base + reg);
+}
+
+static inline void ath79_spi_wr(struct ath79_spi *sp, unsigned int reg, u32 val)
+{
+	iowrite32(val, sp->base + reg);
+}
+
+static inline struct ath79_spi *ath79_spidev_to_sp(struct spi_device *spi)
+{
+	return spi_master_get_devdata(spi->master);
+}
+
+static inline void ath79_spi_delay(struct ath79_spi *sp, unsigned int nsecs)
+{
+	if (nsecs > sp->rrw_delay)
+		ndelay(nsecs - sp->rrw_delay);
+}
+
+static void ath79_spi_chipselect(struct spi_device *spi, int is_active)
+{
+	struct ath79_spi *sp = ath79_spidev_to_sp(spi);
+	int cs_high = (spi->mode & SPI_CS_HIGH) ? is_active : !is_active;
+
+	if (is_active) {
+		/* set initial clock polarity */
+		if (spi->mode & SPI_CPOL)
+			sp->ioc_base |= AR71XX_SPI_IOC_CLK;
+		else
+			sp->ioc_base &= ~AR71XX_SPI_IOC_CLK;
+
+		ath79_spi_wr(sp, AR71XX_SPI_REG_IOC, sp->ioc_base);
+	}
+
+	if (gpio_is_valid(spi->cs_gpio)) {
+		/* SPI is normally active-low */
+		gpio_set_value_cansleep(spi->cs_gpio, cs_high);
+	} else {
+		u32 cs_bit = AR71XX_SPI_IOC_CS(spi->chip_select);
+
+		if (cs_high)
+			sp->ioc_base |= cs_bit;
+		else
+			sp->ioc_base &= ~cs_bit;
+
+		ath79_spi_wr(sp, AR71XX_SPI_REG_IOC, sp->ioc_base);
+	}
+
+}
+
+static void ath79_spi_enable(struct ath79_spi *sp)
+{
+	/* enable GPIO mode */
+	ath79_spi_wr(sp, AR71XX_SPI_REG_FS, AR71XX_SPI_FS_GPIO);
+
+	/* save CTRL register */
+	sp->reg_ctrl = ath79_spi_rr(sp, AR71XX_SPI_REG_CTRL);
+	sp->ioc_base = ath79_spi_rr(sp, AR71XX_SPI_REG_IOC);
+
+	/* TODO: setup speed? */
+	ath79_spi_wr(sp, AR71XX_SPI_REG_CTRL, 0x43);
+}
+
+static void ath79_spi_disable(struct ath79_spi *sp)
+{
+	/* restore CTRL register */
+	ath79_spi_wr(sp, AR71XX_SPI_REG_CTRL, sp->reg_ctrl);
+	/* disable GPIO mode */
+	ath79_spi_wr(sp, AR71XX_SPI_REG_FS, 0);
+}
+
+static int ath79_spi_setup_cs(struct spi_device *spi)
+{
+	struct ath79_spi *sp = ath79_spidev_to_sp(spi);
+	int status;
+
+	status = 0;
+	if (gpio_is_valid(spi->cs_gpio)) {
+		unsigned long flags;
+
+		flags = GPIOF_DIR_OUT;
+		if (spi->mode & SPI_CS_HIGH)
+			flags |= GPIOF_INIT_LOW;
+		else
+			flags |= GPIOF_INIT_HIGH;
+
+		status = gpio_request_one(spi->cs_gpio, flags,
+					  dev_name(&spi->dev));
+	} else {
+		u32 cs_bit = AR71XX_SPI_IOC_CS(spi->chip_select);
+
+		if (spi->mode & SPI_CS_HIGH)
+			sp->ioc_base &= ~cs_bit;
+		else
+			sp->ioc_base |= cs_bit;
+
+		ath79_spi_wr(sp, AR71XX_SPI_REG_IOC, sp->ioc_base);
+	}
+
+	return status;
+}
+
+static void ath79_spi_cleanup_cs(struct spi_device *spi)
+{
+	if (gpio_is_valid(spi->cs_gpio))
+		gpio_free(spi->cs_gpio);
+}
+
+static int ath79_spi_setup(struct spi_device *spi)
+{
+	int status = 0;
+
+	if (!spi->controller_state) {
+		status = ath79_spi_setup_cs(spi);
+		if (status)
+			return status;
+	}
+
+	status = spi_bitbang_setup(spi);
+	if (status && !spi->controller_state)
+		ath79_spi_cleanup_cs(spi);
+
+	return status;
+}
+
+static void ath79_spi_cleanup(struct spi_device *spi)
+{
+	ath79_spi_cleanup_cs(spi);
+	spi_bitbang_cleanup(spi);
+}
+
+static u32 ath79_spi_txrx_mode0(struct spi_device *spi, unsigned int nsecs,
+			       u32 word, u8 bits, unsigned flags)
+{
+	struct ath79_spi *sp = ath79_spidev_to_sp(spi);
+	u32 ioc = sp->ioc_base;
+
+	/* clock starts at inactive polarity */
+	for (word <<= (32 - bits); likely(bits); bits--) {
+		u32 out;
+
+		if (word & (1 << 31))
+			out = ioc | AR71XX_SPI_IOC_DO;
+		else
+			out = ioc & ~AR71XX_SPI_IOC_DO;
+
+		/* setup MSB (to slave) on trailing edge */
+		ath79_spi_wr(sp, AR71XX_SPI_REG_IOC, out);
+		ath79_spi_delay(sp, nsecs);
+		ath79_spi_wr(sp, AR71XX_SPI_REG_IOC, out | AR71XX_SPI_IOC_CLK);
+		ath79_spi_delay(sp, nsecs);
+		if (bits == 1)
+			ath79_spi_wr(sp, AR71XX_SPI_REG_IOC, out);
+
+		word <<= 1;
+	}
+
+	return ath79_spi_rr(sp, AR71XX_SPI_REG_RDS);
+}
+
+static int ath79_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct ath79_spi *sp;
+	struct ath79_spi_platform_data *pdata;
+	struct resource	*r;
+	unsigned long rate;
+	int ret;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*sp));
+	if (master == NULL) {
+		dev_err(&pdev->dev, "failed to allocate spi master\n");
+		return -ENOMEM;
+	}
+
+	sp = spi_master_get_devdata(master);
+	master->dev.of_node = pdev->dev.of_node;
+	platform_set_drvdata(pdev, sp);
+
+	pdata = dev_get_platdata(&pdev->dev);
+
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(1, 32);
+	master->setup = ath79_spi_setup;
+	master->cleanup = ath79_spi_cleanup;
+	if (pdata) {
+		master->bus_num = pdata->bus_num;
+		master->num_chipselect = pdata->num_chipselect;
+	}
+
+	sp->bitbang.master = master;
+	sp->bitbang.chipselect = ath79_spi_chipselect;
+	sp->bitbang.txrx_word[SPI_MODE_0] = ath79_spi_txrx_mode0;
+	sp->bitbang.setup_transfer = spi_bitbang_setup_transfer;
+	sp->bitbang.flags = SPI_CS_HIGH;
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	sp->base = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(sp->base)) {
+		ret = PTR_ERR(sp->base);
+		goto err_put_master;
+	}
+
+	sp->clk = devm_clk_get(&pdev->dev, "ahb");
+	if (IS_ERR(sp->clk)) {
+		ret = PTR_ERR(sp->clk);
+		goto err_put_master;
+	}
+
+	ret = clk_prepare_enable(sp->clk);
+	if (ret)
+		goto err_put_master;
+
+	rate = DIV_ROUND_UP(clk_get_rate(sp->clk), MHZ);
+	if (!rate) {
+		ret = -EINVAL;
+		goto err_clk_disable;
+	}
+
+	sp->rrw_delay = ATH79_SPI_RRW_DELAY_FACTOR / rate;
+	dev_dbg(&pdev->dev, "register read/write delay is %u nsecs\n",
+		sp->rrw_delay);
+
+	ath79_spi_enable(sp);
+	ret = spi_bitbang_start(&sp->bitbang);
+	if (ret)
+		goto err_disable;
+
+	return 0;
+
+err_disable:
+	ath79_spi_disable(sp);
+err_clk_disable:
+	clk_disable_unprepare(sp->clk);
+err_put_master:
+	spi_master_put(sp->bitbang.master);
+
+	return ret;
+}
+
+static int ath79_spi_remove(struct platform_device *pdev)
+{
+	struct ath79_spi *sp = platform_get_drvdata(pdev);
+
+	spi_bitbang_stop(&sp->bitbang);
+	ath79_spi_disable(sp);
+	clk_disable_unprepare(sp->clk);
+	spi_master_put(sp->bitbang.master);
+
+	return 0;
+}
+
+static void ath79_spi_shutdown(struct platform_device *pdev)
+{
+	ath79_spi_remove(pdev);
+}
+
+static const struct of_device_id ath79_spi_of_match[] = {
+	{ .compatible = "qca,ar7100-spi", },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, ath79_spi_of_match);
+
+static struct platform_driver ath79_spi_driver = {
+	.probe		= ath79_spi_probe,
+	.remove		= ath79_spi_remove,
+	.shutdown	= ath79_spi_shutdown,
+	.driver		= {
+		.name	= DRV_NAME,
+		.of_match_table = ath79_spi_of_match,
+	},
+};
+module_platform_driver(ath79_spi_driver);
+
+MODULE_DESCRIPTION("SPI controller driver for Atheros AR71XX/AR724X/AR913X");
+MODULE_AUTHOR("Gabor Juhos <juhosg@openwrt.org>");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:" DRV_NAME);
diff --git a/drivers/spi/spi-atmel.c b/drivers/spi/spi-atmel.c
new file mode 100644
index 0000000..3f890d1
--- /dev/null
+++ b/drivers/spi/spi-atmel.c
@@ -0,0 +1,1843 @@
+/*
+ * Driver for Atmel AT32 and AT91 SPI Controllers
+ *
+ * Copyright (C) 2006 Atmel Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/kernel.h>
+#include <linux/clk.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/spi/spi.h>
+#include <linux/slab.h>
+#include <linux/platform_data/dma-atmel.h>
+#include <linux/of.h>
+
+#include <linux/io.h>
+#include <linux/gpio.h>
+#include <linux/of_gpio.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/pm_runtime.h>
+
+/* SPI register offsets */
+#define SPI_CR					0x0000
+#define SPI_MR					0x0004
+#define SPI_RDR					0x0008
+#define SPI_TDR					0x000c
+#define SPI_SR					0x0010
+#define SPI_IER					0x0014
+#define SPI_IDR					0x0018
+#define SPI_IMR					0x001c
+#define SPI_CSR0				0x0030
+#define SPI_CSR1				0x0034
+#define SPI_CSR2				0x0038
+#define SPI_CSR3				0x003c
+#define SPI_FMR					0x0040
+#define SPI_FLR					0x0044
+#define SPI_VERSION				0x00fc
+#define SPI_RPR					0x0100
+#define SPI_RCR					0x0104
+#define SPI_TPR					0x0108
+#define SPI_TCR					0x010c
+#define SPI_RNPR				0x0110
+#define SPI_RNCR				0x0114
+#define SPI_TNPR				0x0118
+#define SPI_TNCR				0x011c
+#define SPI_PTCR				0x0120
+#define SPI_PTSR				0x0124
+
+/* Bitfields in CR */
+#define SPI_SPIEN_OFFSET			0
+#define SPI_SPIEN_SIZE				1
+#define SPI_SPIDIS_OFFSET			1
+#define SPI_SPIDIS_SIZE				1
+#define SPI_SWRST_OFFSET			7
+#define SPI_SWRST_SIZE				1
+#define SPI_LASTXFER_OFFSET			24
+#define SPI_LASTXFER_SIZE			1
+#define SPI_TXFCLR_OFFSET			16
+#define SPI_TXFCLR_SIZE				1
+#define SPI_RXFCLR_OFFSET			17
+#define SPI_RXFCLR_SIZE				1
+#define SPI_FIFOEN_OFFSET			30
+#define SPI_FIFOEN_SIZE				1
+#define SPI_FIFODIS_OFFSET			31
+#define SPI_FIFODIS_SIZE			1
+
+/* Bitfields in MR */
+#define SPI_MSTR_OFFSET				0
+#define SPI_MSTR_SIZE				1
+#define SPI_PS_OFFSET				1
+#define SPI_PS_SIZE				1
+#define SPI_PCSDEC_OFFSET			2
+#define SPI_PCSDEC_SIZE				1
+#define SPI_FDIV_OFFSET				3
+#define SPI_FDIV_SIZE				1
+#define SPI_MODFDIS_OFFSET			4
+#define SPI_MODFDIS_SIZE			1
+#define SPI_WDRBT_OFFSET			5
+#define SPI_WDRBT_SIZE				1
+#define SPI_LLB_OFFSET				7
+#define SPI_LLB_SIZE				1
+#define SPI_PCS_OFFSET				16
+#define SPI_PCS_SIZE				4
+#define SPI_DLYBCS_OFFSET			24
+#define SPI_DLYBCS_SIZE				8
+
+/* Bitfields in RDR */
+#define SPI_RD_OFFSET				0
+#define SPI_RD_SIZE				16
+
+/* Bitfields in TDR */
+#define SPI_TD_OFFSET				0
+#define SPI_TD_SIZE				16
+
+/* Bitfields in SR */
+#define SPI_RDRF_OFFSET				0
+#define SPI_RDRF_SIZE				1
+#define SPI_TDRE_OFFSET				1
+#define SPI_TDRE_SIZE				1
+#define SPI_MODF_OFFSET				2
+#define SPI_MODF_SIZE				1
+#define SPI_OVRES_OFFSET			3
+#define SPI_OVRES_SIZE				1
+#define SPI_ENDRX_OFFSET			4
+#define SPI_ENDRX_SIZE				1
+#define SPI_ENDTX_OFFSET			5
+#define SPI_ENDTX_SIZE				1
+#define SPI_RXBUFF_OFFSET			6
+#define SPI_RXBUFF_SIZE				1
+#define SPI_TXBUFE_OFFSET			7
+#define SPI_TXBUFE_SIZE				1
+#define SPI_NSSR_OFFSET				8
+#define SPI_NSSR_SIZE				1
+#define SPI_TXEMPTY_OFFSET			9
+#define SPI_TXEMPTY_SIZE			1
+#define SPI_SPIENS_OFFSET			16
+#define SPI_SPIENS_SIZE				1
+#define SPI_TXFEF_OFFSET			24
+#define SPI_TXFEF_SIZE				1
+#define SPI_TXFFF_OFFSET			25
+#define SPI_TXFFF_SIZE				1
+#define SPI_TXFTHF_OFFSET			26
+#define SPI_TXFTHF_SIZE				1
+#define SPI_RXFEF_OFFSET			27
+#define SPI_RXFEF_SIZE				1
+#define SPI_RXFFF_OFFSET			28
+#define SPI_RXFFF_SIZE				1
+#define SPI_RXFTHF_OFFSET			29
+#define SPI_RXFTHF_SIZE				1
+#define SPI_TXFPTEF_OFFSET			30
+#define SPI_TXFPTEF_SIZE			1
+#define SPI_RXFPTEF_OFFSET			31
+#define SPI_RXFPTEF_SIZE			1
+
+/* Bitfields in CSR0 */
+#define SPI_CPOL_OFFSET				0
+#define SPI_CPOL_SIZE				1
+#define SPI_NCPHA_OFFSET			1
+#define SPI_NCPHA_SIZE				1
+#define SPI_CSAAT_OFFSET			3
+#define SPI_CSAAT_SIZE				1
+#define SPI_BITS_OFFSET				4
+#define SPI_BITS_SIZE				4
+#define SPI_SCBR_OFFSET				8
+#define SPI_SCBR_SIZE				8
+#define SPI_DLYBS_OFFSET			16
+#define SPI_DLYBS_SIZE				8
+#define SPI_DLYBCT_OFFSET			24
+#define SPI_DLYBCT_SIZE				8
+
+/* Bitfields in RCR */
+#define SPI_RXCTR_OFFSET			0
+#define SPI_RXCTR_SIZE				16
+
+/* Bitfields in TCR */
+#define SPI_TXCTR_OFFSET			0
+#define SPI_TXCTR_SIZE				16
+
+/* Bitfields in RNCR */
+#define SPI_RXNCR_OFFSET			0
+#define SPI_RXNCR_SIZE				16
+
+/* Bitfields in TNCR */
+#define SPI_TXNCR_OFFSET			0
+#define SPI_TXNCR_SIZE				16
+
+/* Bitfields in PTCR */
+#define SPI_RXTEN_OFFSET			0
+#define SPI_RXTEN_SIZE				1
+#define SPI_RXTDIS_OFFSET			1
+#define SPI_RXTDIS_SIZE				1
+#define SPI_TXTEN_OFFSET			8
+#define SPI_TXTEN_SIZE				1
+#define SPI_TXTDIS_OFFSET			9
+#define SPI_TXTDIS_SIZE				1
+
+/* Bitfields in FMR */
+#define SPI_TXRDYM_OFFSET			0
+#define SPI_TXRDYM_SIZE				2
+#define SPI_RXRDYM_OFFSET			4
+#define SPI_RXRDYM_SIZE				2
+#define SPI_TXFTHRES_OFFSET			16
+#define SPI_TXFTHRES_SIZE			6
+#define SPI_RXFTHRES_OFFSET			24
+#define SPI_RXFTHRES_SIZE			6
+
+/* Bitfields in FLR */
+#define SPI_TXFL_OFFSET				0
+#define SPI_TXFL_SIZE				6
+#define SPI_RXFL_OFFSET				16
+#define SPI_RXFL_SIZE				6
+
+/* Constants for BITS */
+#define SPI_BITS_8_BPT				0
+#define SPI_BITS_9_BPT				1
+#define SPI_BITS_10_BPT				2
+#define SPI_BITS_11_BPT				3
+#define SPI_BITS_12_BPT				4
+#define SPI_BITS_13_BPT				5
+#define SPI_BITS_14_BPT				6
+#define SPI_BITS_15_BPT				7
+#define SPI_BITS_16_BPT				8
+#define SPI_ONE_DATA				0
+#define SPI_TWO_DATA				1
+#define SPI_FOUR_DATA				2
+
+/* Bit manipulation macros */
+#define SPI_BIT(name) \
+	(1 << SPI_##name##_OFFSET)
+#define SPI_BF(name, value) \
+	(((value) & ((1 << SPI_##name##_SIZE) - 1)) << SPI_##name##_OFFSET)
+#define SPI_BFEXT(name, value) \
+	(((value) >> SPI_##name##_OFFSET) & ((1 << SPI_##name##_SIZE) - 1))
+#define SPI_BFINS(name, value, old) \
+	(((old) & ~(((1 << SPI_##name##_SIZE) - 1) << SPI_##name##_OFFSET)) \
+	  | SPI_BF(name, value))
+
+/* Register access macros */
+#ifdef CONFIG_AVR32
+#define spi_readl(port, reg) \
+	__raw_readl((port)->regs + SPI_##reg)
+#define spi_writel(port, reg, value) \
+	__raw_writel((value), (port)->regs + SPI_##reg)
+
+#define spi_readw(port, reg) \
+	__raw_readw((port)->regs + SPI_##reg)
+#define spi_writew(port, reg, value) \
+	__raw_writew((value), (port)->regs + SPI_##reg)
+
+#define spi_readb(port, reg) \
+	__raw_readb((port)->regs + SPI_##reg)
+#define spi_writeb(port, reg, value) \
+	__raw_writeb((value), (port)->regs + SPI_##reg)
+#else
+#define spi_readl(port, reg) \
+	readl_relaxed((port)->regs + SPI_##reg)
+#define spi_writel(port, reg, value) \
+	writel_relaxed((value), (port)->regs + SPI_##reg)
+
+#define spi_readw(port, reg) \
+	readw_relaxed((port)->regs + SPI_##reg)
+#define spi_writew(port, reg, value) \
+	writew_relaxed((value), (port)->regs + SPI_##reg)
+
+#define spi_readb(port, reg) \
+	readb_relaxed((port)->regs + SPI_##reg)
+#define spi_writeb(port, reg, value) \
+	writeb_relaxed((value), (port)->regs + SPI_##reg)
+#endif
+/* use PIO for small transfers, avoiding DMA setup/teardown overhead and
+ * cache operations; better heuristics consider wordsize and bitrate.
+ */
+#define DMA_MIN_BYTES	16
+
+#define SPI_DMA_TIMEOUT		(msecs_to_jiffies(1000))
+
+#define AUTOSUSPEND_TIMEOUT	2000
+
+struct atmel_spi_caps {
+	bool	is_spi2;
+	bool	has_wdrbt;
+	bool	has_dma_support;
+	bool	has_pdc_support;
+};
+
+/*
+ * The core SPI transfer engine just talks to a register bank to set up
+ * DMA transfers; transfer queue progress is driven by IRQs.  The clock
+ * framework provides the base clock, subdivided for each spi_device.
+ */
+struct atmel_spi {
+	spinlock_t		lock;
+	unsigned long		flags;
+
+	phys_addr_t		phybase;
+	void __iomem		*regs;
+	int			irq;
+	struct clk		*clk;
+	struct platform_device	*pdev;
+	unsigned long		spi_clk;
+
+	struct spi_transfer	*current_transfer;
+	int			current_remaining_bytes;
+	int			done_status;
+	dma_addr_t		dma_addr_rx_bbuf;
+	dma_addr_t		dma_addr_tx_bbuf;
+	void			*addr_rx_bbuf;
+	void			*addr_tx_bbuf;
+
+	struct completion	xfer_completion;
+
+	struct atmel_spi_caps	caps;
+
+	bool			use_dma;
+	bool			use_pdc;
+	bool			use_cs_gpios;
+
+	bool			keep_cs;
+	bool			cs_active;
+
+	u32			fifo_size;
+};
+
+/* Controller-specific per-slave state */
+struct atmel_spi_device {
+	unsigned int		npcs_pin;
+	u32			csr;
+};
+
+#define SPI_MAX_DMA_XFER	65535 /* true for both PDC and DMA */
+#define INVALID_DMA_ADDRESS	0xffffffff
+
+/*
+ * Version 2 of the SPI controller has
+ *  - CR.LASTXFER
+ *  - SPI_MR.DIV32 may become FDIV or must-be-zero (here: always zero)
+ *  - SPI_SR.TXEMPTY, SPI_SR.NSSR (and corresponding irqs)
+ *  - SPI_CSRx.CSAAT
+ *  - SPI_CSRx.SBCR allows faster clocking
+ */
+static bool atmel_spi_is_v2(struct atmel_spi *as)
+{
+	return as->caps.is_spi2;
+}
+
+/*
+ * Earlier SPI controllers (e.g. on at91rm9200) have a design bug whereby
+ * they assume that spi slave device state will not change on deselect, so
+ * that automagic deselection is OK.  ("NPCSx rises if no data is to be
+ * transmitted")  Not so!  Workaround uses nCSx pins as GPIOs; or newer
+ * controllers have CSAAT and friends.
+ *
+ * Since the CSAAT functionality is a bit weird on newer controllers as
+ * well, we use GPIO to control nCSx pins on all controllers, updating
+ * MR.PCS to avoid confusing the controller.  Using GPIOs also lets us
+ * support active-high chipselects despite the controller's belief that
+ * only active-low devices/systems exists.
+ *
+ * However, at91rm9200 has a second erratum whereby nCS0 doesn't work
+ * right when driven with GPIO.  ("Mode Fault does not allow more than one
+ * Master on Chip Select 0.")  No workaround exists for that ... so for
+ * nCS0 on that chip, we (a) don't use the GPIO, (b) can't support CS_HIGH,
+ * and (c) will trigger that first erratum in some cases.
+ */
+
+static void cs_activate(struct atmel_spi *as, struct spi_device *spi)
+{
+	struct atmel_spi_device *asd = spi->controller_state;
+	unsigned active = spi->mode & SPI_CS_HIGH;
+	u32 mr;
+
+	if (atmel_spi_is_v2(as)) {
+		spi_writel(as, CSR0 + 4 * spi->chip_select, asd->csr);
+		/* For the low SPI version, there is a issue that PDC transfer
+		 * on CS1,2,3 needs SPI_CSR0.BITS config as SPI_CSR1,2,3.BITS
+		 */
+		spi_writel(as, CSR0, asd->csr);
+		if (as->caps.has_wdrbt) {
+			spi_writel(as, MR,
+					SPI_BF(PCS, ~(0x01 << spi->chip_select))
+					| SPI_BIT(WDRBT)
+					| SPI_BIT(MODFDIS)
+					| SPI_BIT(MSTR));
+		} else {
+			spi_writel(as, MR,
+					SPI_BF(PCS, ~(0x01 << spi->chip_select))
+					| SPI_BIT(MODFDIS)
+					| SPI_BIT(MSTR));
+		}
+
+		mr = spi_readl(as, MR);
+		if (as->use_cs_gpios)
+			gpio_set_value(asd->npcs_pin, active);
+	} else {
+		u32 cpol = (spi->mode & SPI_CPOL) ? SPI_BIT(CPOL) : 0;
+		int i;
+		u32 csr;
+
+		/* Make sure clock polarity is correct */
+		for (i = 0; i < spi->master->num_chipselect; i++) {
+			csr = spi_readl(as, CSR0 + 4 * i);
+			if ((csr ^ cpol) & SPI_BIT(CPOL))
+				spi_writel(as, CSR0 + 4 * i,
+						csr ^ SPI_BIT(CPOL));
+		}
+
+		mr = spi_readl(as, MR);
+		mr = SPI_BFINS(PCS, ~(1 << spi->chip_select), mr);
+		if (as->use_cs_gpios && spi->chip_select != 0)
+			gpio_set_value(asd->npcs_pin, active);
+		spi_writel(as, MR, mr);
+	}
+
+	dev_dbg(&spi->dev, "activate %u%s, mr %08x\n",
+			asd->npcs_pin, active ? " (high)" : "",
+			mr);
+}
+
+static void cs_deactivate(struct atmel_spi *as, struct spi_device *spi)
+{
+	struct atmel_spi_device *asd = spi->controller_state;
+	unsigned active = spi->mode & SPI_CS_HIGH;
+	u32 mr;
+
+	/* only deactivate *this* device; sometimes transfers to
+	 * another device may be active when this routine is called.
+	 */
+	mr = spi_readl(as, MR);
+	if (~SPI_BFEXT(PCS, mr) & (1 << spi->chip_select)) {
+		mr = SPI_BFINS(PCS, 0xf, mr);
+		spi_writel(as, MR, mr);
+	}
+
+	dev_dbg(&spi->dev, "DEactivate %u%s, mr %08x\n",
+			asd->npcs_pin, active ? " (low)" : "",
+			mr);
+
+	if (!as->use_cs_gpios)
+		spi_writel(as, CR, SPI_BIT(LASTXFER));
+	else if (atmel_spi_is_v2(as) || spi->chip_select != 0)
+		gpio_set_value(asd->npcs_pin, !active);
+}
+
+static void atmel_spi_lock(struct atmel_spi *as) __acquires(&as->lock)
+{
+	spin_lock_irqsave(&as->lock, as->flags);
+}
+
+static void atmel_spi_unlock(struct atmel_spi *as) __releases(&as->lock)
+{
+	spin_unlock_irqrestore(&as->lock, as->flags);
+}
+
+static inline bool atmel_spi_is_vmalloc_xfer(struct spi_transfer *xfer)
+{
+	return is_vmalloc_addr(xfer->tx_buf) || is_vmalloc_addr(xfer->rx_buf);
+}
+
+static inline bool atmel_spi_use_dma(struct atmel_spi *as,
+				struct spi_transfer *xfer)
+{
+	return as->use_dma && xfer->len >= DMA_MIN_BYTES;
+}
+
+static bool atmel_spi_can_dma(struct spi_master *master,
+			      struct spi_device *spi,
+			      struct spi_transfer *xfer)
+{
+	struct atmel_spi *as = spi_master_get_devdata(master);
+
+	if (IS_ENABLED(CONFIG_SOC_SAM_V4_V5))
+		return atmel_spi_use_dma(as, xfer) &&
+			!atmel_spi_is_vmalloc_xfer(xfer);
+	else
+		return atmel_spi_use_dma(as, xfer);
+
+}
+
+static int atmel_spi_dma_slave_config(struct atmel_spi *as,
+				struct dma_slave_config *slave_config,
+				u8 bits_per_word)
+{
+	struct spi_master *master = platform_get_drvdata(as->pdev);
+	int err = 0;
+
+	if (bits_per_word > 8) {
+		slave_config->dst_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		slave_config->src_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+	} else {
+		slave_config->dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		slave_config->src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	}
+
+	slave_config->dst_addr = (dma_addr_t)as->phybase + SPI_TDR;
+	slave_config->src_addr = (dma_addr_t)as->phybase + SPI_RDR;
+	slave_config->src_maxburst = 1;
+	slave_config->dst_maxburst = 1;
+	slave_config->device_fc = false;
+
+	/*
+	 * This driver uses fixed peripheral select mode (PS bit set to '0' in
+	 * the Mode Register).
+	 * So according to the datasheet, when FIFOs are available (and
+	 * enabled), the Transmit FIFO operates in Multiple Data Mode.
+	 * In this mode, up to 2 data, not 4, can be written into the Transmit
+	 * Data Register in a single access.
+	 * However, the first data has to be written into the lowest 16 bits and
+	 * the second data into the highest 16 bits of the Transmit
+	 * Data Register. For 8bit data (the most frequent case), it would
+	 * require to rework tx_buf so each data would actualy fit 16 bits.
+	 * So we'd rather write only one data at the time. Hence the transmit
+	 * path works the same whether FIFOs are available (and enabled) or not.
+	 */
+	slave_config->direction = DMA_MEM_TO_DEV;
+	if (dmaengine_slave_config(master->dma_tx, slave_config)) {
+		dev_err(&as->pdev->dev,
+			"failed to configure tx dma channel\n");
+		err = -EINVAL;
+	}
+
+	/*
+	 * This driver configures the spi controller for master mode (MSTR bit
+	 * set to '1' in the Mode Register).
+	 * So according to the datasheet, when FIFOs are available (and
+	 * enabled), the Receive FIFO operates in Single Data Mode.
+	 * So the receive path works the same whether FIFOs are available (and
+	 * enabled) or not.
+	 */
+	slave_config->direction = DMA_DEV_TO_MEM;
+	if (dmaengine_slave_config(master->dma_rx, slave_config)) {
+		dev_err(&as->pdev->dev,
+			"failed to configure rx dma channel\n");
+		err = -EINVAL;
+	}
+
+	return err;
+}
+
+static int atmel_spi_configure_dma(struct spi_master *master,
+				   struct atmel_spi *as)
+{
+	struct dma_slave_config	slave_config;
+	struct device *dev = &as->pdev->dev;
+	int err;
+
+	dma_cap_mask_t mask;
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	master->dma_tx = dma_request_slave_channel_reason(dev, "tx");
+	if (IS_ERR(master->dma_tx)) {
+		err = PTR_ERR(master->dma_tx);
+		if (err == -EPROBE_DEFER) {
+			dev_warn(dev, "no DMA channel available at the moment\n");
+			goto error_clear;
+		}
+		dev_err(dev,
+			"DMA TX channel not available, SPI unable to use DMA\n");
+		err = -EBUSY;
+		goto error_clear;
+	}
+
+	/*
+	 * No reason to check EPROBE_DEFER here since we have already requested
+	 * tx channel. If it fails here, it's for another reason.
+	 */
+	master->dma_rx = dma_request_slave_channel(dev, "rx");
+
+	if (!master->dma_rx) {
+		dev_err(dev,
+			"DMA RX channel not available, SPI unable to use DMA\n");
+		err = -EBUSY;
+		goto error;
+	}
+
+	err = atmel_spi_dma_slave_config(as, &slave_config, 8);
+	if (err)
+		goto error;
+
+	dev_info(&as->pdev->dev,
+			"Using %s (tx) and %s (rx) for DMA transfers\n",
+			dma_chan_name(master->dma_tx),
+			dma_chan_name(master->dma_rx));
+
+	return 0;
+error:
+	if (master->dma_rx)
+		dma_release_channel(master->dma_rx);
+	if (!IS_ERR(master->dma_tx))
+		dma_release_channel(master->dma_tx);
+error_clear:
+	master->dma_tx = master->dma_rx = NULL;
+	return err;
+}
+
+static void atmel_spi_stop_dma(struct spi_master *master)
+{
+	if (master->dma_rx)
+		dmaengine_terminate_all(master->dma_rx);
+	if (master->dma_tx)
+		dmaengine_terminate_all(master->dma_tx);
+}
+
+static void atmel_spi_release_dma(struct spi_master *master)
+{
+	if (master->dma_rx) {
+		dma_release_channel(master->dma_rx);
+		master->dma_rx = NULL;
+	}
+	if (master->dma_tx) {
+		dma_release_channel(master->dma_tx);
+		master->dma_tx = NULL;
+	}
+}
+
+/* This function is called by the DMA driver from tasklet context */
+static void dma_callback(void *data)
+{
+	struct spi_master	*master = data;
+	struct atmel_spi	*as = spi_master_get_devdata(master);
+
+	if (is_vmalloc_addr(as->current_transfer->rx_buf) &&
+	    IS_ENABLED(CONFIG_SOC_SAM_V4_V5)) {
+		memcpy(as->current_transfer->rx_buf, as->addr_rx_bbuf,
+		       as->current_transfer->len);
+	}
+	complete(&as->xfer_completion);
+}
+
+/*
+ * Next transfer using PIO without FIFO.
+ */
+static void atmel_spi_next_xfer_single(struct spi_master *master,
+				       struct spi_transfer *xfer)
+{
+	struct atmel_spi	*as = spi_master_get_devdata(master);
+	unsigned long xfer_pos = xfer->len - as->current_remaining_bytes;
+
+	dev_vdbg(master->dev.parent, "atmel_spi_next_xfer_pio\n");
+
+	/* Make sure data is not remaining in RDR */
+	spi_readl(as, RDR);
+	while (spi_readl(as, SR) & SPI_BIT(RDRF)) {
+		spi_readl(as, RDR);
+		cpu_relax();
+	}
+
+	if (xfer->bits_per_word > 8)
+		spi_writel(as, TDR, *(u16 *)(xfer->tx_buf + xfer_pos));
+	else
+		spi_writel(as, TDR, *(u8 *)(xfer->tx_buf + xfer_pos));
+
+	dev_dbg(master->dev.parent,
+		"  start pio xfer %p: len %u tx %p rx %p bitpw %d\n",
+		xfer, xfer->len, xfer->tx_buf, xfer->rx_buf,
+		xfer->bits_per_word);
+
+	/* Enable relevant interrupts */
+	spi_writel(as, IER, SPI_BIT(RDRF) | SPI_BIT(OVRES));
+}
+
+/*
+ * Next transfer using PIO with FIFO.
+ */
+static void atmel_spi_next_xfer_fifo(struct spi_master *master,
+				     struct spi_transfer *xfer)
+{
+	struct atmel_spi *as = spi_master_get_devdata(master);
+	u32 current_remaining_data, num_data;
+	u32 offset = xfer->len - as->current_remaining_bytes;
+	const u16 *words = (const u16 *)((u8 *)xfer->tx_buf + offset);
+	const u8  *bytes = (const u8  *)((u8 *)xfer->tx_buf + offset);
+	u16 td0, td1;
+	u32 fifomr;
+
+	dev_vdbg(master->dev.parent, "atmel_spi_next_xfer_fifo\n");
+
+	/* Compute the number of data to transfer in the current iteration */
+	current_remaining_data = ((xfer->bits_per_word > 8) ?
+				  ((u32)as->current_remaining_bytes >> 1) :
+				  (u32)as->current_remaining_bytes);
+	num_data = min(current_remaining_data, as->fifo_size);
+
+	/* Flush RX and TX FIFOs */
+	spi_writel(as, CR, SPI_BIT(RXFCLR) | SPI_BIT(TXFCLR));
+	while (spi_readl(as, FLR))
+		cpu_relax();
+
+	/* Set RX FIFO Threshold to the number of data to transfer */
+	fifomr = spi_readl(as, FMR);
+	spi_writel(as, FMR, SPI_BFINS(RXFTHRES, num_data, fifomr));
+
+	/* Clear FIFO flags in the Status Register, especially RXFTHF */
+	(void)spi_readl(as, SR);
+
+	/* Fill TX FIFO */
+	while (num_data >= 2) {
+		if (xfer->bits_per_word > 8) {
+			td0 = *words++;
+			td1 = *words++;
+		} else {
+			td0 = *bytes++;
+			td1 = *bytes++;
+		}
+
+		spi_writel(as, TDR, (td1 << 16) | td0);
+		num_data -= 2;
+	}
+
+	if (num_data) {
+		if (xfer->bits_per_word > 8)
+			td0 = *words++;
+		else
+			td0 = *bytes++;
+
+		spi_writew(as, TDR, td0);
+		num_data--;
+	}
+
+	dev_dbg(master->dev.parent,
+		"  start fifo xfer %p: len %u tx %p rx %p bitpw %d\n",
+		xfer, xfer->len, xfer->tx_buf, xfer->rx_buf,
+		xfer->bits_per_word);
+
+	/*
+	 * Enable RX FIFO Threshold Flag interrupt to be notified about
+	 * transfer completion.
+	 */
+	spi_writel(as, IER, SPI_BIT(RXFTHF) | SPI_BIT(OVRES));
+}
+
+/*
+ * Next transfer using PIO.
+ */
+static void atmel_spi_next_xfer_pio(struct spi_master *master,
+				    struct spi_transfer *xfer)
+{
+	struct atmel_spi *as = spi_master_get_devdata(master);
+
+	if (as->fifo_size)
+		atmel_spi_next_xfer_fifo(master, xfer);
+	else
+		atmel_spi_next_xfer_single(master, xfer);
+}
+
+/*
+ * Submit next transfer for DMA.
+ */
+static int atmel_spi_next_xfer_dma_submit(struct spi_master *master,
+				struct spi_transfer *xfer,
+				u32 *plen)
+{
+	struct atmel_spi	*as = spi_master_get_devdata(master);
+	struct dma_chan		*rxchan = master->dma_rx;
+	struct dma_chan		*txchan = master->dma_tx;
+	struct dma_async_tx_descriptor *rxdesc;
+	struct dma_async_tx_descriptor *txdesc;
+	struct dma_slave_config	slave_config;
+	dma_cookie_t		cookie;
+
+	dev_vdbg(master->dev.parent, "atmel_spi_next_xfer_dma_submit\n");
+
+	/* Check that the channels are available */
+	if (!rxchan || !txchan)
+		return -ENODEV;
+
+	/* release lock for DMA operations */
+	atmel_spi_unlock(as);
+
+	*plen = xfer->len;
+
+	if (atmel_spi_dma_slave_config(as, &slave_config,
+				       xfer->bits_per_word))
+		goto err_exit;
+
+	/* Send both scatterlists */
+	if (atmel_spi_is_vmalloc_xfer(xfer) &&
+	    IS_ENABLED(CONFIG_SOC_SAM_V4_V5)) {
+		rxdesc = dmaengine_prep_slave_single(rxchan,
+						     as->dma_addr_rx_bbuf,
+						     xfer->len,
+						     DMA_DEV_TO_MEM,
+						     DMA_PREP_INTERRUPT |
+						     DMA_CTRL_ACK);
+	} else {
+		rxdesc = dmaengine_prep_slave_sg(rxchan,
+						 xfer->rx_sg.sgl,
+						 xfer->rx_sg.nents,
+						 DMA_DEV_TO_MEM,
+						 DMA_PREP_INTERRUPT |
+						 DMA_CTRL_ACK);
+	}
+	if (!rxdesc)
+		goto err_dma;
+
+	if (atmel_spi_is_vmalloc_xfer(xfer) &&
+	    IS_ENABLED(CONFIG_SOC_SAM_V4_V5)) {
+		memcpy(as->addr_tx_bbuf, xfer->tx_buf, xfer->len);
+		txdesc = dmaengine_prep_slave_single(txchan,
+						     as->dma_addr_tx_bbuf,
+						     xfer->len, DMA_MEM_TO_DEV,
+						     DMA_PREP_INTERRUPT |
+						     DMA_CTRL_ACK);
+	} else {
+		txdesc = dmaengine_prep_slave_sg(txchan,
+						 xfer->tx_sg.sgl,
+						 xfer->tx_sg.nents,
+						 DMA_MEM_TO_DEV,
+						 DMA_PREP_INTERRUPT |
+						 DMA_CTRL_ACK);
+	}
+	if (!txdesc)
+		goto err_dma;
+
+	dev_dbg(master->dev.parent,
+		"  start dma xfer %p: len %u tx %p/%08llx rx %p/%08llx\n",
+		xfer, xfer->len, xfer->tx_buf, (unsigned long long)xfer->tx_dma,
+		xfer->rx_buf, (unsigned long long)xfer->rx_dma);
+
+	/* Enable relevant interrupts */
+	spi_writel(as, IER, SPI_BIT(OVRES));
+
+	/* Put the callback on the RX transfer only, that should finish last */
+	rxdesc->callback = dma_callback;
+	rxdesc->callback_param = master;
+
+	/* Submit and fire RX and TX with TX last so we're ready to read! */
+	cookie = rxdesc->tx_submit(rxdesc);
+	if (dma_submit_error(cookie))
+		goto err_dma;
+	cookie = txdesc->tx_submit(txdesc);
+	if (dma_submit_error(cookie))
+		goto err_dma;
+	rxchan->device->device_issue_pending(rxchan);
+	txchan->device->device_issue_pending(txchan);
+
+	/* take back lock */
+	atmel_spi_lock(as);
+	return 0;
+
+err_dma:
+	spi_writel(as, IDR, SPI_BIT(OVRES));
+	atmel_spi_stop_dma(master);
+err_exit:
+	atmel_spi_lock(as);
+	return -ENOMEM;
+}
+
+static void atmel_spi_next_xfer_data(struct spi_master *master,
+				struct spi_transfer *xfer,
+				dma_addr_t *tx_dma,
+				dma_addr_t *rx_dma,
+				u32 *plen)
+{
+	*rx_dma = xfer->rx_dma + xfer->len - *plen;
+	*tx_dma = xfer->tx_dma + xfer->len - *plen;
+	if (*plen > master->max_dma_len)
+		*plen = master->max_dma_len;
+}
+
+static int atmel_spi_set_xfer_speed(struct atmel_spi *as,
+				    struct spi_device *spi,
+				    struct spi_transfer *xfer)
+{
+	u32			scbr, csr;
+	unsigned long		bus_hz;
+
+	/* v1 chips start out at half the peripheral bus speed. */
+	bus_hz = as->spi_clk;
+	if (!atmel_spi_is_v2(as))
+		bus_hz /= 2;
+
+	/*
+	 * Calculate the lowest divider that satisfies the
+	 * constraint, assuming div32/fdiv/mbz == 0.
+	 */
+	scbr = DIV_ROUND_UP(bus_hz, xfer->speed_hz);
+
+	/*
+	 * If the resulting divider doesn't fit into the
+	 * register bitfield, we can't satisfy the constraint.
+	 */
+	if (scbr >= (1 << SPI_SCBR_SIZE)) {
+		dev_err(&spi->dev,
+			"setup: %d Hz too slow, scbr %u; min %ld Hz\n",
+			xfer->speed_hz, scbr, bus_hz/255);
+		return -EINVAL;
+	}
+	if (scbr == 0) {
+		dev_err(&spi->dev,
+			"setup: %d Hz too high, scbr %u; max %ld Hz\n",
+			xfer->speed_hz, scbr, bus_hz);
+		return -EINVAL;
+	}
+	csr = spi_readl(as, CSR0 + 4 * spi->chip_select);
+	csr = SPI_BFINS(SCBR, scbr, csr);
+	spi_writel(as, CSR0 + 4 * spi->chip_select, csr);
+
+	return 0;
+}
+
+/*
+ * Submit next transfer for PDC.
+ * lock is held, spi irq is blocked
+ */
+static void atmel_spi_pdc_next_xfer(struct spi_master *master,
+					struct spi_message *msg,
+					struct spi_transfer *xfer)
+{
+	struct atmel_spi	*as = spi_master_get_devdata(master);
+	u32			len;
+	dma_addr_t		tx_dma, rx_dma;
+
+	spi_writel(as, PTCR, SPI_BIT(RXTDIS) | SPI_BIT(TXTDIS));
+
+	len = as->current_remaining_bytes;
+	atmel_spi_next_xfer_data(master, xfer, &tx_dma, &rx_dma, &len);
+	as->current_remaining_bytes -= len;
+
+	spi_writel(as, RPR, rx_dma);
+	spi_writel(as, TPR, tx_dma);
+
+	if (msg->spi->bits_per_word > 8)
+		len >>= 1;
+	spi_writel(as, RCR, len);
+	spi_writel(as, TCR, len);
+
+	dev_dbg(&msg->spi->dev,
+		"  start xfer %p: len %u tx %p/%08llx rx %p/%08llx\n",
+		xfer, xfer->len, xfer->tx_buf,
+		(unsigned long long)xfer->tx_dma, xfer->rx_buf,
+		(unsigned long long)xfer->rx_dma);
+
+	if (as->current_remaining_bytes) {
+		len = as->current_remaining_bytes;
+		atmel_spi_next_xfer_data(master, xfer, &tx_dma, &rx_dma, &len);
+		as->current_remaining_bytes -= len;
+
+		spi_writel(as, RNPR, rx_dma);
+		spi_writel(as, TNPR, tx_dma);
+
+		if (msg->spi->bits_per_word > 8)
+			len >>= 1;
+		spi_writel(as, RNCR, len);
+		spi_writel(as, TNCR, len);
+
+		dev_dbg(&msg->spi->dev,
+			"  next xfer %p: len %u tx %p/%08llx rx %p/%08llx\n",
+			xfer, xfer->len, xfer->tx_buf,
+			(unsigned long long)xfer->tx_dma, xfer->rx_buf,
+			(unsigned long long)xfer->rx_dma);
+	}
+
+	/* REVISIT: We're waiting for RXBUFF before we start the next
+	 * transfer because we need to handle some difficult timing
+	 * issues otherwise. If we wait for TXBUFE in one transfer and
+	 * then starts waiting for RXBUFF in the next, it's difficult
+	 * to tell the difference between the RXBUFF interrupt we're
+	 * actually waiting for and the RXBUFF interrupt of the
+	 * previous transfer.
+	 *
+	 * It should be doable, though. Just not now...
+	 */
+	spi_writel(as, IER, SPI_BIT(RXBUFF) | SPI_BIT(OVRES));
+	spi_writel(as, PTCR, SPI_BIT(TXTEN) | SPI_BIT(RXTEN));
+}
+
+/*
+ * For DMA, tx_buf/tx_dma have the same relationship as rx_buf/rx_dma:
+ *  - The buffer is either valid for CPU access, else NULL
+ *  - If the buffer is valid, so is its DMA address
+ *
+ * This driver manages the dma address unless message->is_dma_mapped.
+ */
+static int
+atmel_spi_dma_map_xfer(struct atmel_spi *as, struct spi_transfer *xfer)
+{
+	struct device	*dev = &as->pdev->dev;
+
+	xfer->tx_dma = xfer->rx_dma = INVALID_DMA_ADDRESS;
+	if (xfer->tx_buf) {
+		/* tx_buf is a const void* where we need a void * for the dma
+		 * mapping */
+		void *nonconst_tx = (void *)xfer->tx_buf;
+
+		xfer->tx_dma = dma_map_single(dev,
+				nonconst_tx, xfer->len,
+				DMA_TO_DEVICE);
+		if (dma_mapping_error(dev, xfer->tx_dma))
+			return -ENOMEM;
+	}
+	if (xfer->rx_buf) {
+		xfer->rx_dma = dma_map_single(dev,
+				xfer->rx_buf, xfer->len,
+				DMA_FROM_DEVICE);
+		if (dma_mapping_error(dev, xfer->rx_dma)) {
+			if (xfer->tx_buf)
+				dma_unmap_single(dev,
+						xfer->tx_dma, xfer->len,
+						DMA_TO_DEVICE);
+			return -ENOMEM;
+		}
+	}
+	return 0;
+}
+
+static void atmel_spi_dma_unmap_xfer(struct spi_master *master,
+				     struct spi_transfer *xfer)
+{
+	if (xfer->tx_dma != INVALID_DMA_ADDRESS)
+		dma_unmap_single(master->dev.parent, xfer->tx_dma,
+				 xfer->len, DMA_TO_DEVICE);
+	if (xfer->rx_dma != INVALID_DMA_ADDRESS)
+		dma_unmap_single(master->dev.parent, xfer->rx_dma,
+				 xfer->len, DMA_FROM_DEVICE);
+}
+
+static void atmel_spi_disable_pdc_transfer(struct atmel_spi *as)
+{
+	spi_writel(as, PTCR, SPI_BIT(RXTDIS) | SPI_BIT(TXTDIS));
+}
+
+static void
+atmel_spi_pump_single_data(struct atmel_spi *as, struct spi_transfer *xfer)
+{
+	u8		*rxp;
+	u16		*rxp16;
+	unsigned long	xfer_pos = xfer->len - as->current_remaining_bytes;
+
+	if (xfer->bits_per_word > 8) {
+		rxp16 = (u16 *)(((u8 *)xfer->rx_buf) + xfer_pos);
+		*rxp16 = spi_readl(as, RDR);
+	} else {
+		rxp = ((u8 *)xfer->rx_buf) + xfer_pos;
+		*rxp = spi_readl(as, RDR);
+	}
+	if (xfer->bits_per_word > 8) {
+		if (as->current_remaining_bytes > 2)
+			as->current_remaining_bytes -= 2;
+		else
+			as->current_remaining_bytes = 0;
+	} else {
+		as->current_remaining_bytes--;
+	}
+}
+
+static void
+atmel_spi_pump_fifo_data(struct atmel_spi *as, struct spi_transfer *xfer)
+{
+	u32 fifolr = spi_readl(as, FLR);
+	u32 num_bytes, num_data = SPI_BFEXT(RXFL, fifolr);
+	u32 offset = xfer->len - as->current_remaining_bytes;
+	u16 *words = (u16 *)((u8 *)xfer->rx_buf + offset);
+	u8  *bytes = (u8  *)((u8 *)xfer->rx_buf + offset);
+	u16 rd; /* RD field is the lowest 16 bits of RDR */
+
+	/* Update the number of remaining bytes to transfer */
+	num_bytes = ((xfer->bits_per_word > 8) ?
+		     (num_data << 1) :
+		     num_data);
+
+	if (as->current_remaining_bytes > num_bytes)
+		as->current_remaining_bytes -= num_bytes;
+	else
+		as->current_remaining_bytes = 0;
+
+	/* Handle odd number of bytes when data are more than 8bit width */
+	if (xfer->bits_per_word > 8)
+		as->current_remaining_bytes &= ~0x1;
+
+	/* Read data */
+	while (num_data) {
+		rd = spi_readl(as, RDR);
+		if (xfer->bits_per_word > 8)
+			*words++ = rd;
+		else
+			*bytes++ = rd;
+		num_data--;
+	}
+}
+
+/* Called from IRQ
+ *
+ * Must update "current_remaining_bytes" to keep track of data
+ * to transfer.
+ */
+static void
+atmel_spi_pump_pio_data(struct atmel_spi *as, struct spi_transfer *xfer)
+{
+	if (as->fifo_size)
+		atmel_spi_pump_fifo_data(as, xfer);
+	else
+		atmel_spi_pump_single_data(as, xfer);
+}
+
+/* Interrupt
+ *
+ * No need for locking in this Interrupt handler: done_status is the
+ * only information modified.
+ */
+static irqreturn_t
+atmel_spi_pio_interrupt(int irq, void *dev_id)
+{
+	struct spi_master	*master = dev_id;
+	struct atmel_spi	*as = spi_master_get_devdata(master);
+	u32			status, pending, imr;
+	struct spi_transfer	*xfer;
+	int			ret = IRQ_NONE;
+
+	imr = spi_readl(as, IMR);
+	status = spi_readl(as, SR);
+	pending = status & imr;
+
+	if (pending & SPI_BIT(OVRES)) {
+		ret = IRQ_HANDLED;
+		spi_writel(as, IDR, SPI_BIT(OVRES));
+		dev_warn(master->dev.parent, "overrun\n");
+
+		/*
+		 * When we get an overrun, we disregard the current
+		 * transfer. Data will not be copied back from any
+		 * bounce buffer and msg->actual_len will not be
+		 * updated with the last xfer.
+		 *
+		 * We will also not process any remaning transfers in
+		 * the message.
+		 */
+		as->done_status = -EIO;
+		smp_wmb();
+
+		/* Clear any overrun happening while cleaning up */
+		spi_readl(as, SR);
+
+		complete(&as->xfer_completion);
+
+	} else if (pending & (SPI_BIT(RDRF) | SPI_BIT(RXFTHF))) {
+		atmel_spi_lock(as);
+
+		if (as->current_remaining_bytes) {
+			ret = IRQ_HANDLED;
+			xfer = as->current_transfer;
+			atmel_spi_pump_pio_data(as, xfer);
+			if (!as->current_remaining_bytes)
+				spi_writel(as, IDR, pending);
+
+			complete(&as->xfer_completion);
+		}
+
+		atmel_spi_unlock(as);
+	} else {
+		WARN_ONCE(pending, "IRQ not handled, pending = %x\n", pending);
+		ret = IRQ_HANDLED;
+		spi_writel(as, IDR, pending);
+	}
+
+	return ret;
+}
+
+static irqreturn_t
+atmel_spi_pdc_interrupt(int irq, void *dev_id)
+{
+	struct spi_master	*master = dev_id;
+	struct atmel_spi	*as = spi_master_get_devdata(master);
+	u32			status, pending, imr;
+	int			ret = IRQ_NONE;
+
+	imr = spi_readl(as, IMR);
+	status = spi_readl(as, SR);
+	pending = status & imr;
+
+	if (pending & SPI_BIT(OVRES)) {
+
+		ret = IRQ_HANDLED;
+
+		spi_writel(as, IDR, (SPI_BIT(RXBUFF) | SPI_BIT(ENDRX)
+				     | SPI_BIT(OVRES)));
+
+		/* Clear any overrun happening while cleaning up */
+		spi_readl(as, SR);
+
+		as->done_status = -EIO;
+
+		complete(&as->xfer_completion);
+
+	} else if (pending & (SPI_BIT(RXBUFF) | SPI_BIT(ENDRX))) {
+		ret = IRQ_HANDLED;
+
+		spi_writel(as, IDR, pending);
+
+		complete(&as->xfer_completion);
+	}
+
+	return ret;
+}
+
+static int atmel_spi_setup(struct spi_device *spi)
+{
+	struct atmel_spi	*as;
+	struct atmel_spi_device	*asd;
+	u32			csr;
+	unsigned int		bits = spi->bits_per_word;
+	unsigned int		npcs_pin;
+
+	as = spi_master_get_devdata(spi->master);
+
+	/* see notes above re chipselect */
+	if (!atmel_spi_is_v2(as)
+			&& spi->chip_select == 0
+			&& (spi->mode & SPI_CS_HIGH)) {
+		dev_dbg(&spi->dev, "setup: can't be active-high\n");
+		return -EINVAL;
+	}
+
+	csr = SPI_BF(BITS, bits - 8);
+	if (spi->mode & SPI_CPOL)
+		csr |= SPI_BIT(CPOL);
+	if (!(spi->mode & SPI_CPHA))
+		csr |= SPI_BIT(NCPHA);
+	if (!as->use_cs_gpios)
+		csr |= SPI_BIT(CSAAT);
+
+	/* DLYBS is mostly irrelevant since we manage chipselect using GPIOs.
+	 *
+	 * DLYBCT would add delays between words, slowing down transfers.
+	 * It could potentially be useful to cope with DMA bottlenecks, but
+	 * in those cases it's probably best to just use a lower bitrate.
+	 */
+	csr |= SPI_BF(DLYBS, 0);
+	csr |= SPI_BF(DLYBCT, 0);
+
+	/* chipselect must have been muxed as GPIO (e.g. in board setup) */
+	npcs_pin = (unsigned long)spi->controller_data;
+
+	if (!as->use_cs_gpios)
+		npcs_pin = spi->chip_select;
+	else if (gpio_is_valid(spi->cs_gpio))
+		npcs_pin = spi->cs_gpio;
+
+	asd = spi->controller_state;
+	if (!asd) {
+		asd = kzalloc(sizeof(struct atmel_spi_device), GFP_KERNEL);
+		if (!asd)
+			return -ENOMEM;
+
+		if (as->use_cs_gpios)
+			gpio_direction_output(npcs_pin,
+					      !(spi->mode & SPI_CS_HIGH));
+
+		asd->npcs_pin = npcs_pin;
+		spi->controller_state = asd;
+	}
+
+	asd->csr = csr;
+
+	dev_dbg(&spi->dev,
+		"setup: bpw %u mode 0x%x -> csr%d %08x\n",
+		bits, spi->mode, spi->chip_select, csr);
+
+	if (!atmel_spi_is_v2(as))
+		spi_writel(as, CSR0 + 4 * spi->chip_select, csr);
+
+	return 0;
+}
+
+static int atmel_spi_one_transfer(struct spi_master *master,
+					struct spi_message *msg,
+					struct spi_transfer *xfer)
+{
+	struct atmel_spi	*as;
+	struct spi_device	*spi = msg->spi;
+	u8			bits;
+	u32			len;
+	struct atmel_spi_device	*asd;
+	int			timeout;
+	int			ret;
+	unsigned long		dma_timeout;
+
+	as = spi_master_get_devdata(master);
+
+	if (!(xfer->tx_buf || xfer->rx_buf) && xfer->len) {
+		dev_dbg(&spi->dev, "missing rx or tx buf\n");
+		return -EINVAL;
+	}
+
+	asd = spi->controller_state;
+	bits = (asd->csr >> 4) & 0xf;
+	if (bits != xfer->bits_per_word - 8) {
+		dev_dbg(&spi->dev,
+			"you can't yet change bits_per_word in transfers\n");
+		return -ENOPROTOOPT;
+	}
+
+	/*
+	 * DMA map early, for performance (empties dcache ASAP) and
+	 * better fault reporting.
+	 */
+	if ((!msg->is_dma_mapped)
+		&& as->use_pdc) {
+		if (atmel_spi_dma_map_xfer(as, xfer) < 0)
+			return -ENOMEM;
+	}
+
+	atmel_spi_set_xfer_speed(as, msg->spi, xfer);
+
+	as->done_status = 0;
+	as->current_transfer = xfer;
+	as->current_remaining_bytes = xfer->len;
+	while (as->current_remaining_bytes) {
+		reinit_completion(&as->xfer_completion);
+
+		if (as->use_pdc) {
+			atmel_spi_pdc_next_xfer(master, msg, xfer);
+		} else if (atmel_spi_use_dma(as, xfer)) {
+			len = as->current_remaining_bytes;
+			ret = atmel_spi_next_xfer_dma_submit(master,
+								xfer, &len);
+			if (ret) {
+				dev_err(&spi->dev,
+					"unable to use DMA, fallback to PIO\n");
+				atmel_spi_next_xfer_pio(master, xfer);
+			} else {
+				as->current_remaining_bytes -= len;
+				if (as->current_remaining_bytes < 0)
+					as->current_remaining_bytes = 0;
+			}
+		} else {
+			atmel_spi_next_xfer_pio(master, xfer);
+		}
+
+		/* interrupts are disabled, so free the lock for schedule */
+		atmel_spi_unlock(as);
+		dma_timeout = wait_for_completion_timeout(&as->xfer_completion,
+							  SPI_DMA_TIMEOUT);
+		atmel_spi_lock(as);
+		if (WARN_ON(dma_timeout == 0)) {
+			dev_err(&spi->dev, "spi transfer timeout\n");
+			as->done_status = -EIO;
+		}
+
+		if (as->done_status)
+			break;
+	}
+
+	if (as->done_status) {
+		if (as->use_pdc) {
+			dev_warn(master->dev.parent,
+				"overrun (%u/%u remaining)\n",
+				spi_readl(as, TCR), spi_readl(as, RCR));
+
+			/*
+			 * Clean up DMA registers and make sure the data
+			 * registers are empty.
+			 */
+			spi_writel(as, RNCR, 0);
+			spi_writel(as, TNCR, 0);
+			spi_writel(as, RCR, 0);
+			spi_writel(as, TCR, 0);
+			for (timeout = 1000; timeout; timeout--)
+				if (spi_readl(as, SR) & SPI_BIT(TXEMPTY))
+					break;
+			if (!timeout)
+				dev_warn(master->dev.parent,
+					 "timeout waiting for TXEMPTY");
+			while (spi_readl(as, SR) & SPI_BIT(RDRF))
+				spi_readl(as, RDR);
+
+			/* Clear any overrun happening while cleaning up */
+			spi_readl(as, SR);
+
+		} else if (atmel_spi_use_dma(as, xfer)) {
+			atmel_spi_stop_dma(master);
+		}
+
+		if (!msg->is_dma_mapped
+			&& as->use_pdc)
+			atmel_spi_dma_unmap_xfer(master, xfer);
+
+		return 0;
+
+	} else {
+		/* only update length if no error */
+		msg->actual_length += xfer->len;
+	}
+
+	if (!msg->is_dma_mapped
+		&& as->use_pdc)
+		atmel_spi_dma_unmap_xfer(master, xfer);
+
+	if (xfer->delay_usecs)
+		udelay(xfer->delay_usecs);
+
+	if (xfer->cs_change) {
+		if (list_is_last(&xfer->transfer_list,
+				 &msg->transfers)) {
+			as->keep_cs = true;
+		} else {
+			as->cs_active = !as->cs_active;
+			if (as->cs_active)
+				cs_activate(as, msg->spi);
+			else
+				cs_deactivate(as, msg->spi);
+		}
+	}
+
+	return 0;
+}
+
+static int atmel_spi_transfer_one_message(struct spi_master *master,
+						struct spi_message *msg)
+{
+	struct atmel_spi *as;
+	struct spi_transfer *xfer;
+	struct spi_device *spi = msg->spi;
+	int ret = 0;
+
+	as = spi_master_get_devdata(master);
+
+	dev_dbg(&spi->dev, "new message %p submitted for %s\n",
+					msg, dev_name(&spi->dev));
+
+	atmel_spi_lock(as);
+	cs_activate(as, spi);
+
+	as->cs_active = true;
+	as->keep_cs = false;
+
+	msg->status = 0;
+	msg->actual_length = 0;
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		ret = atmel_spi_one_transfer(master, msg, xfer);
+		if (ret)
+			goto msg_done;
+	}
+
+	if (as->use_pdc)
+		atmel_spi_disable_pdc_transfer(as);
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		dev_dbg(&spi->dev,
+			"  xfer %p: len %u tx %p/%pad rx %p/%pad\n",
+			xfer, xfer->len,
+			xfer->tx_buf, &xfer->tx_dma,
+			xfer->rx_buf, &xfer->rx_dma);
+	}
+
+msg_done:
+	if (!as->keep_cs)
+		cs_deactivate(as, msg->spi);
+
+	atmel_spi_unlock(as);
+
+	msg->status = as->done_status;
+	spi_finalize_current_message(spi->master);
+
+	return ret;
+}
+
+static void atmel_spi_cleanup(struct spi_device *spi)
+{
+	struct atmel_spi_device	*asd = spi->controller_state;
+
+	if (!asd)
+		return;
+
+	spi->controller_state = NULL;
+	kfree(asd);
+}
+
+static inline unsigned int atmel_get_version(struct atmel_spi *as)
+{
+	return spi_readl(as, VERSION) & 0x00000fff;
+}
+
+static void atmel_get_caps(struct atmel_spi *as)
+{
+	unsigned int version;
+
+	version = atmel_get_version(as);
+
+	as->caps.is_spi2 = version > 0x121;
+	as->caps.has_wdrbt = version >= 0x210;
+	as->caps.has_dma_support = version >= 0x212;
+	as->caps.has_pdc_support = version < 0x212;
+}
+
+/*-------------------------------------------------------------------------*/
+static int atmel_spi_gpio_cs(struct platform_device *pdev)
+{
+	struct spi_master	*master = platform_get_drvdata(pdev);
+	struct atmel_spi	*as = spi_master_get_devdata(master);
+	struct device_node	*np = master->dev.of_node;
+	int			i;
+	int			ret = 0;
+	int			nb = 0;
+
+	if (!as->use_cs_gpios)
+		return 0;
+
+	if (!np)
+		return 0;
+
+	nb = of_gpio_named_count(np, "cs-gpios");
+	for (i = 0; i < nb; i++) {
+		int cs_gpio = of_get_named_gpio(pdev->dev.of_node,
+						"cs-gpios", i);
+
+		if (cs_gpio == -EPROBE_DEFER)
+			return cs_gpio;
+
+		if (gpio_is_valid(cs_gpio)) {
+			ret = devm_gpio_request(&pdev->dev, cs_gpio,
+						dev_name(&pdev->dev));
+			if (ret)
+				return ret;
+		}
+	}
+
+	return 0;
+}
+
+static void atmel_spi_init(struct atmel_spi *as)
+{
+	spi_writel(as, CR, SPI_BIT(SWRST));
+	spi_writel(as, CR, SPI_BIT(SWRST)); /* AT91SAM9263 Rev B workaround */
+
+	/* It is recommended to enable FIFOs first thing after reset */
+	if (as->fifo_size)
+		spi_writel(as, CR, SPI_BIT(FIFOEN));
+
+	if (as->caps.has_wdrbt) {
+		spi_writel(as, MR, SPI_BIT(WDRBT) | SPI_BIT(MODFDIS)
+				| SPI_BIT(MSTR));
+	} else {
+		spi_writel(as, MR, SPI_BIT(MSTR) | SPI_BIT(MODFDIS));
+	}
+
+	if (as->use_pdc)
+		spi_writel(as, PTCR, SPI_BIT(RXTDIS) | SPI_BIT(TXTDIS));
+	spi_writel(as, CR, SPI_BIT(SPIEN));
+}
+
+static int atmel_spi_probe(struct platform_device *pdev)
+{
+	struct resource		*regs;
+	int			irq;
+	struct clk		*clk;
+	int			ret;
+	struct spi_master	*master;
+	struct atmel_spi	*as;
+
+	/* Select default pin state */
+	pinctrl_pm_select_default_state(&pdev->dev);
+
+	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!regs)
+		return -ENXIO;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0)
+		return irq;
+
+	clk = devm_clk_get(&pdev->dev, "spi_clk");
+	if (IS_ERR(clk))
+		return PTR_ERR(clk);
+
+	/* setup spi core then atmel-specific driver state */
+	ret = -ENOMEM;
+	master = spi_alloc_master(&pdev->dev, sizeof(*as));
+	if (!master)
+		goto out_free;
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(8, 16);
+	master->dev.of_node = pdev->dev.of_node;
+	master->bus_num = pdev->id;
+	master->num_chipselect = master->dev.of_node ? 0 : 4;
+	master->setup = atmel_spi_setup;
+	master->flags = (SPI_MASTER_MUST_RX | SPI_MASTER_MUST_TX);
+	master->transfer_one_message = atmel_spi_transfer_one_message;
+	master->cleanup = atmel_spi_cleanup;
+	master->auto_runtime_pm = true;
+	master->max_dma_len = SPI_MAX_DMA_XFER;
+	master->can_dma = atmel_spi_can_dma;
+	platform_set_drvdata(pdev, master);
+
+	as = spi_master_get_devdata(master);
+
+	spin_lock_init(&as->lock);
+
+	as->pdev = pdev;
+	as->regs = devm_ioremap_resource(&pdev->dev, regs);
+	if (IS_ERR(as->regs)) {
+		ret = PTR_ERR(as->regs);
+		goto out_unmap_regs;
+	}
+	as->phybase = regs->start;
+	as->irq = irq;
+	as->clk = clk;
+
+	init_completion(&as->xfer_completion);
+
+	atmel_get_caps(as);
+
+	as->use_cs_gpios = true;
+	if (atmel_spi_is_v2(as) &&
+	    pdev->dev.of_node &&
+	    !of_get_property(pdev->dev.of_node, "cs-gpios", NULL)) {
+		as->use_cs_gpios = false;
+		master->num_chipselect = 4;
+	}
+
+	ret = atmel_spi_gpio_cs(pdev);
+	if (ret)
+		goto out_unmap_regs;
+
+	as->use_dma = false;
+	as->use_pdc = false;
+	if (as->caps.has_dma_support) {
+		ret = atmel_spi_configure_dma(master, as);
+		if (ret == 0) {
+			as->use_dma = true;
+		} else if (ret == -EPROBE_DEFER) {
+			return ret;
+		}
+	} else if (as->caps.has_pdc_support) {
+		as->use_pdc = true;
+	}
+
+	if (IS_ENABLED(CONFIG_SOC_SAM_V4_V5)) {
+		as->addr_rx_bbuf = dma_alloc_coherent(&pdev->dev,
+						      SPI_MAX_DMA_XFER,
+						      &as->dma_addr_rx_bbuf,
+						      GFP_KERNEL | GFP_DMA);
+		if (!as->addr_rx_bbuf) {
+			as->use_dma = false;
+		} else {
+			as->addr_tx_bbuf = dma_alloc_coherent(&pdev->dev,
+					SPI_MAX_DMA_XFER,
+					&as->dma_addr_tx_bbuf,
+					GFP_KERNEL | GFP_DMA);
+			if (!as->addr_tx_bbuf) {
+				as->use_dma = false;
+				dma_free_coherent(&pdev->dev, SPI_MAX_DMA_XFER,
+						  as->addr_rx_bbuf,
+						  as->dma_addr_rx_bbuf);
+			}
+		}
+		if (!as->use_dma)
+			dev_info(master->dev.parent,
+				 "  can not allocate dma coherent memory\n");
+	}
+
+	if (as->caps.has_dma_support && !as->use_dma)
+		dev_info(&pdev->dev, "Atmel SPI Controller using PIO only\n");
+
+	if (as->use_pdc) {
+		ret = devm_request_irq(&pdev->dev, irq, atmel_spi_pdc_interrupt,
+					0, dev_name(&pdev->dev), master);
+	} else {
+		ret = devm_request_irq(&pdev->dev, irq, atmel_spi_pio_interrupt,
+					0, dev_name(&pdev->dev), master);
+	}
+	if (ret)
+		goto out_unmap_regs;
+
+	/* Initialize the hardware */
+	ret = clk_prepare_enable(clk);
+	if (ret)
+		goto out_free_irq;
+
+	as->spi_clk = clk_get_rate(clk);
+
+	as->fifo_size = 0;
+	if (!of_property_read_u32(pdev->dev.of_node, "atmel,fifo-size",
+				  &as->fifo_size)) {
+		dev_info(&pdev->dev, "Using FIFO (%u data)\n", as->fifo_size);
+	}
+
+	atmel_spi_init(as);
+
+	pm_runtime_set_autosuspend_delay(&pdev->dev, AUTOSUSPEND_TIMEOUT);
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret)
+		goto out_free_dma;
+
+	/* go! */
+	dev_info(&pdev->dev, "Atmel SPI Controller version 0x%x at 0x%08lx (irq %d)\n",
+			atmel_get_version(as), (unsigned long)regs->start,
+			irq);
+
+	return 0;
+
+out_free_dma:
+	pm_runtime_disable(&pdev->dev);
+	pm_runtime_set_suspended(&pdev->dev);
+
+	if (as->use_dma)
+		atmel_spi_release_dma(master);
+
+	spi_writel(as, CR, SPI_BIT(SWRST));
+	spi_writel(as, CR, SPI_BIT(SWRST)); /* AT91SAM9263 Rev B workaround */
+	clk_disable_unprepare(clk);
+out_free_irq:
+out_unmap_regs:
+out_free:
+	spi_master_put(master);
+	return ret;
+}
+
+static int atmel_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master	*master = platform_get_drvdata(pdev);
+	struct atmel_spi	*as = spi_master_get_devdata(master);
+
+	pm_runtime_get_sync(&pdev->dev);
+
+	/* reset the hardware and block queue progress */
+	if (as->use_dma) {
+		atmel_spi_stop_dma(master);
+		atmel_spi_release_dma(master);
+		if (IS_ENABLED(CONFIG_SOC_SAM_V4_V5)) {
+			dma_free_coherent(&pdev->dev, SPI_MAX_DMA_XFER,
+					  as->addr_tx_bbuf,
+					  as->dma_addr_tx_bbuf);
+			dma_free_coherent(&pdev->dev, SPI_MAX_DMA_XFER,
+					  as->addr_rx_bbuf,
+					  as->dma_addr_rx_bbuf);
+		}
+	}
+
+	spin_lock_irq(&as->lock);
+	spi_writel(as, CR, SPI_BIT(SWRST));
+	spi_writel(as, CR, SPI_BIT(SWRST)); /* AT91SAM9263 Rev B workaround */
+	spi_readl(as, SR);
+	spin_unlock_irq(&as->lock);
+
+	clk_disable_unprepare(as->clk);
+
+	pm_runtime_put_noidle(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int atmel_spi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct atmel_spi *as = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(as->clk);
+	pinctrl_pm_select_sleep_state(dev);
+
+	return 0;
+}
+
+static int atmel_spi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct atmel_spi *as = spi_master_get_devdata(master);
+
+	pinctrl_pm_select_default_state(dev);
+
+	return clk_prepare_enable(as->clk);
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int atmel_spi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	int ret;
+
+	/* Stop the queue running */
+	ret = spi_master_suspend(master);
+	if (ret) {
+		dev_warn(dev, "cannot suspend master\n");
+		return ret;
+	}
+
+	if (!pm_runtime_suspended(dev))
+		atmel_spi_runtime_suspend(dev);
+
+	return 0;
+}
+
+static int atmel_spi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct atmel_spi *as = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(as->clk);
+	if (ret)
+		return ret;
+
+	atmel_spi_init(as);
+
+	clk_disable_unprepare(as->clk);
+
+	if (!pm_runtime_suspended(dev)) {
+		ret = atmel_spi_runtime_resume(dev);
+		if (ret)
+			return ret;
+	}
+
+	/* Start the queue running */
+	ret = spi_master_resume(master);
+	if (ret)
+		dev_err(dev, "problem starting queue (%d)\n", ret);
+
+	return ret;
+}
+#endif
+
+static const struct dev_pm_ops atmel_spi_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(atmel_spi_suspend, atmel_spi_resume)
+	SET_RUNTIME_PM_OPS(atmel_spi_runtime_suspend,
+			   atmel_spi_runtime_resume, NULL)
+};
+#define ATMEL_SPI_PM_OPS	(&atmel_spi_pm_ops)
+#else
+#define ATMEL_SPI_PM_OPS	NULL
+#endif
+
+#if defined(CONFIG_OF)
+static const struct of_device_id atmel_spi_dt_ids[] = {
+	{ .compatible = "atmel,at91rm9200-spi" },
+	{ /* sentinel */ }
+};
+
+MODULE_DEVICE_TABLE(of, atmel_spi_dt_ids);
+#endif
+
+static struct platform_driver atmel_spi_driver = {
+	.driver		= {
+		.name	= "atmel_spi",
+		.pm	= ATMEL_SPI_PM_OPS,
+		.of_match_table	= of_match_ptr(atmel_spi_dt_ids),
+	},
+	.probe		= atmel_spi_probe,
+	.remove		= atmel_spi_remove,
+};
+module_platform_driver(atmel_spi_driver);
+
+MODULE_DESCRIPTION("Atmel AT32/AT91 SPI Controller driver");
+MODULE_AUTHOR("Haavard Skinnemoen (Atmel)");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:atmel_spi");
diff --git a/drivers/spi/spi-au1550.c b/drivers/spi/spi-au1550.c
new file mode 100644
index 0000000..afd239d
--- /dev/null
+++ b/drivers/spi/spi-au1550.c
@@ -0,0 +1,1001 @@
+/*
+ * au1550 psc spi controller driver
+ * may work also with au1200, au1210, au1250
+ * will not work on au1000, au1100 and au1500 (no full spi controller there)
+ *
+ * Copyright (c) 2006 ATRON electronic GmbH
+ * Author: Jan Nikitenko <jan.nikitenko@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/slab.h>
+#include <linux/errno.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/platform_device.h>
+#include <linux/resource.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/dma-mapping.h>
+#include <linux/completion.h>
+#include <asm/mach-au1x00/au1000.h>
+#include <asm/mach-au1x00/au1xxx_psc.h>
+#include <asm/mach-au1x00/au1xxx_dbdma.h>
+
+#include <asm/mach-au1x00/au1550_spi.h>
+
+static unsigned usedma = 1;
+module_param(usedma, uint, 0644);
+
+/*
+#define AU1550_SPI_DEBUG_LOOPBACK
+*/
+
+
+#define AU1550_SPI_DBDMA_DESCRIPTORS 1
+#define AU1550_SPI_DMA_RXTMP_MINSIZE 2048U
+
+struct au1550_spi {
+	struct spi_bitbang bitbang;
+
+	volatile psc_spi_t __iomem *regs;
+	int irq;
+
+	unsigned len;
+	unsigned tx_count;
+	unsigned rx_count;
+	const u8 *tx;
+	u8 *rx;
+
+	void (*rx_word)(struct au1550_spi *hw);
+	void (*tx_word)(struct au1550_spi *hw);
+	int (*txrx_bufs)(struct spi_device *spi, struct spi_transfer *t);
+	irqreturn_t (*irq_callback)(struct au1550_spi *hw);
+
+	struct completion master_done;
+
+	unsigned usedma;
+	u32 dma_tx_id;
+	u32 dma_rx_id;
+	u32 dma_tx_ch;
+	u32 dma_rx_ch;
+
+	u8 *dma_rx_tmpbuf;
+	unsigned dma_rx_tmpbuf_size;
+	u32 dma_rx_tmpbuf_addr;
+
+	struct spi_master *master;
+	struct device *dev;
+	struct au1550_spi_info *pdata;
+	struct resource *ioarea;
+};
+
+
+/* we use an 8-bit memory device for dma transfers to/from spi fifo */
+static dbdev_tab_t au1550_spi_mem_dbdev =
+{
+	.dev_id			= DBDMA_MEM_CHAN,
+	.dev_flags		= DEV_FLAGS_ANYUSE|DEV_FLAGS_SYNC,
+	.dev_tsize		= 0,
+	.dev_devwidth		= 8,
+	.dev_physaddr		= 0x00000000,
+	.dev_intlevel		= 0,
+	.dev_intpolarity	= 0
+};
+
+static int ddma_memid;	/* id to above mem dma device */
+
+static void au1550_spi_bits_handlers_set(struct au1550_spi *hw, int bpw);
+
+
+/*
+ *  compute BRG and DIV bits to setup spi clock based on main input clock rate
+ *  that was specified in platform data structure
+ *  according to au1550 datasheet:
+ *    psc_tempclk = psc_mainclk / (2 << DIV)
+ *    spiclk = psc_tempclk / (2 * (BRG + 1))
+ *    BRG valid range is 4..63
+ *    DIV valid range is 0..3
+ */
+static u32 au1550_spi_baudcfg(struct au1550_spi *hw, unsigned speed_hz)
+{
+	u32 mainclk_hz = hw->pdata->mainclk_hz;
+	u32 div, brg;
+
+	for (div = 0; div < 4; div++) {
+		brg = mainclk_hz / speed_hz / (4 << div);
+		/* now we have BRG+1 in brg, so count with that */
+		if (brg < (4 + 1)) {
+			brg = (4 + 1);	/* speed_hz too big */
+			break;		/* set lowest brg (div is == 0) */
+		}
+		if (brg <= (63 + 1))
+			break;		/* we have valid brg and div */
+	}
+	if (div == 4) {
+		div = 3;		/* speed_hz too small */
+		brg = (63 + 1);		/* set highest brg and div */
+	}
+	brg--;
+	return PSC_SPICFG_SET_BAUD(brg) | PSC_SPICFG_SET_DIV(div);
+}
+
+static inline void au1550_spi_mask_ack_all(struct au1550_spi *hw)
+{
+	hw->regs->psc_spimsk =
+		  PSC_SPIMSK_MM | PSC_SPIMSK_RR | PSC_SPIMSK_RO
+		| PSC_SPIMSK_RU | PSC_SPIMSK_TR | PSC_SPIMSK_TO
+		| PSC_SPIMSK_TU | PSC_SPIMSK_SD | PSC_SPIMSK_MD;
+	wmb(); /* drain writebuffer */
+
+	hw->regs->psc_spievent =
+		  PSC_SPIEVNT_MM | PSC_SPIEVNT_RR | PSC_SPIEVNT_RO
+		| PSC_SPIEVNT_RU | PSC_SPIEVNT_TR | PSC_SPIEVNT_TO
+		| PSC_SPIEVNT_TU | PSC_SPIEVNT_SD | PSC_SPIEVNT_MD;
+	wmb(); /* drain writebuffer */
+}
+
+static void au1550_spi_reset_fifos(struct au1550_spi *hw)
+{
+	u32 pcr;
+
+	hw->regs->psc_spipcr = PSC_SPIPCR_RC | PSC_SPIPCR_TC;
+	wmb(); /* drain writebuffer */
+	do {
+		pcr = hw->regs->psc_spipcr;
+		wmb(); /* drain writebuffer */
+	} while (pcr != 0);
+}
+
+/*
+ * dma transfers are used for the most common spi word size of 8-bits
+ * we cannot easily change already set up dma channels' width, so if we wanted
+ * dma support for more than 8-bit words (up to 24 bits), we would need to
+ * setup dma channels from scratch on each spi transfer, based on bits_per_word
+ * instead we have pre set up 8 bit dma channels supporting spi 4 to 8 bits
+ * transfers, and 9 to 24 bits spi transfers will be done in pio irq based mode
+ * callbacks to handle dma or pio are set up in au1550_spi_bits_handlers_set()
+ */
+static void au1550_spi_chipsel(struct spi_device *spi, int value)
+{
+	struct au1550_spi *hw = spi_master_get_devdata(spi->master);
+	unsigned cspol = spi->mode & SPI_CS_HIGH ? 1 : 0;
+	u32 cfg, stat;
+
+	switch (value) {
+	case BITBANG_CS_INACTIVE:
+		if (hw->pdata->deactivate_cs)
+			hw->pdata->deactivate_cs(hw->pdata, spi->chip_select,
+					cspol);
+		break;
+
+	case BITBANG_CS_ACTIVE:
+		au1550_spi_bits_handlers_set(hw, spi->bits_per_word);
+
+		cfg = hw->regs->psc_spicfg;
+		wmb(); /* drain writebuffer */
+		hw->regs->psc_spicfg = cfg & ~PSC_SPICFG_DE_ENABLE;
+		wmb(); /* drain writebuffer */
+
+		if (spi->mode & SPI_CPOL)
+			cfg |= PSC_SPICFG_BI;
+		else
+			cfg &= ~PSC_SPICFG_BI;
+		if (spi->mode & SPI_CPHA)
+			cfg &= ~PSC_SPICFG_CDE;
+		else
+			cfg |= PSC_SPICFG_CDE;
+
+		if (spi->mode & SPI_LSB_FIRST)
+			cfg |= PSC_SPICFG_MLF;
+		else
+			cfg &= ~PSC_SPICFG_MLF;
+
+		if (hw->usedma && spi->bits_per_word <= 8)
+			cfg &= ~PSC_SPICFG_DD_DISABLE;
+		else
+			cfg |= PSC_SPICFG_DD_DISABLE;
+		cfg = PSC_SPICFG_CLR_LEN(cfg);
+		cfg |= PSC_SPICFG_SET_LEN(spi->bits_per_word);
+
+		cfg = PSC_SPICFG_CLR_BAUD(cfg);
+		cfg &= ~PSC_SPICFG_SET_DIV(3);
+		cfg |= au1550_spi_baudcfg(hw, spi->max_speed_hz);
+
+		hw->regs->psc_spicfg = cfg | PSC_SPICFG_DE_ENABLE;
+		wmb(); /* drain writebuffer */
+		do {
+			stat = hw->regs->psc_spistat;
+			wmb(); /* drain writebuffer */
+		} while ((stat & PSC_SPISTAT_DR) == 0);
+
+		if (hw->pdata->activate_cs)
+			hw->pdata->activate_cs(hw->pdata, spi->chip_select,
+					cspol);
+		break;
+	}
+}
+
+static int au1550_spi_setupxfer(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct au1550_spi *hw = spi_master_get_devdata(spi->master);
+	unsigned bpw, hz;
+	u32 cfg, stat;
+
+	if (t) {
+		bpw = t->bits_per_word;
+		hz = t->speed_hz;
+	} else {
+		bpw = spi->bits_per_word;
+		hz = spi->max_speed_hz;
+	}
+
+	if (!hz)
+		return -EINVAL;
+
+	au1550_spi_bits_handlers_set(hw, spi->bits_per_word);
+
+	cfg = hw->regs->psc_spicfg;
+	wmb(); /* drain writebuffer */
+	hw->regs->psc_spicfg = cfg & ~PSC_SPICFG_DE_ENABLE;
+	wmb(); /* drain writebuffer */
+
+	if (hw->usedma && bpw <= 8)
+		cfg &= ~PSC_SPICFG_DD_DISABLE;
+	else
+		cfg |= PSC_SPICFG_DD_DISABLE;
+	cfg = PSC_SPICFG_CLR_LEN(cfg);
+	cfg |= PSC_SPICFG_SET_LEN(bpw);
+
+	cfg = PSC_SPICFG_CLR_BAUD(cfg);
+	cfg &= ~PSC_SPICFG_SET_DIV(3);
+	cfg |= au1550_spi_baudcfg(hw, hz);
+
+	hw->regs->psc_spicfg = cfg;
+	wmb(); /* drain writebuffer */
+
+	if (cfg & PSC_SPICFG_DE_ENABLE) {
+		do {
+			stat = hw->regs->psc_spistat;
+			wmb(); /* drain writebuffer */
+		} while ((stat & PSC_SPISTAT_DR) == 0);
+	}
+
+	au1550_spi_reset_fifos(hw);
+	au1550_spi_mask_ack_all(hw);
+	return 0;
+}
+
+/*
+ * for dma spi transfers, we have to setup rx channel, otherwise there is
+ * no reliable way how to recognize that spi transfer is done
+ * dma complete callbacks are called before real spi transfer is finished
+ * and if only tx dma channel is set up (and rx fifo overflow event masked)
+ * spi master done event irq is not generated unless rx fifo is empty (emptied)
+ * so we need rx tmp buffer to use for rx dma if user does not provide one
+ */
+static int au1550_spi_dma_rxtmp_alloc(struct au1550_spi *hw, unsigned size)
+{
+	hw->dma_rx_tmpbuf = kmalloc(size, GFP_KERNEL);
+	if (!hw->dma_rx_tmpbuf)
+		return -ENOMEM;
+	hw->dma_rx_tmpbuf_size = size;
+	hw->dma_rx_tmpbuf_addr = dma_map_single(hw->dev, hw->dma_rx_tmpbuf,
+			size, DMA_FROM_DEVICE);
+	if (dma_mapping_error(hw->dev, hw->dma_rx_tmpbuf_addr)) {
+		kfree(hw->dma_rx_tmpbuf);
+		hw->dma_rx_tmpbuf = 0;
+		hw->dma_rx_tmpbuf_size = 0;
+		return -EFAULT;
+	}
+	return 0;
+}
+
+static void au1550_spi_dma_rxtmp_free(struct au1550_spi *hw)
+{
+	dma_unmap_single(hw->dev, hw->dma_rx_tmpbuf_addr,
+			hw->dma_rx_tmpbuf_size, DMA_FROM_DEVICE);
+	kfree(hw->dma_rx_tmpbuf);
+	hw->dma_rx_tmpbuf = 0;
+	hw->dma_rx_tmpbuf_size = 0;
+}
+
+static int au1550_spi_dma_txrxb(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct au1550_spi *hw = spi_master_get_devdata(spi->master);
+	dma_addr_t dma_tx_addr;
+	dma_addr_t dma_rx_addr;
+	u32 res;
+
+	hw->len = t->len;
+	hw->tx_count = 0;
+	hw->rx_count = 0;
+
+	hw->tx = t->tx_buf;
+	hw->rx = t->rx_buf;
+	dma_tx_addr = t->tx_dma;
+	dma_rx_addr = t->rx_dma;
+
+	/*
+	 * check if buffers are already dma mapped, map them otherwise:
+	 * - first map the TX buffer, so cache data gets written to memory
+	 * - then map the RX buffer, so that cache entries (with
+	 *   soon-to-be-stale data) get removed
+	 * use rx buffer in place of tx if tx buffer was not provided
+	 * use temp rx buffer (preallocated or realloc to fit) for rx dma
+	 */
+	if (t->tx_buf) {
+		if (t->tx_dma == 0) {	/* if DMA_ADDR_INVALID, map it */
+			dma_tx_addr = dma_map_single(hw->dev,
+					(void *)t->tx_buf,
+					t->len, DMA_TO_DEVICE);
+			if (dma_mapping_error(hw->dev, dma_tx_addr))
+				dev_err(hw->dev, "tx dma map error\n");
+		}
+	}
+
+	if (t->rx_buf) {
+		if (t->rx_dma == 0) {	/* if DMA_ADDR_INVALID, map it */
+			dma_rx_addr = dma_map_single(hw->dev,
+					(void *)t->rx_buf,
+					t->len, DMA_FROM_DEVICE);
+			if (dma_mapping_error(hw->dev, dma_rx_addr))
+				dev_err(hw->dev, "rx dma map error\n");
+		}
+	} else {
+		if (t->len > hw->dma_rx_tmpbuf_size) {
+			int ret;
+
+			au1550_spi_dma_rxtmp_free(hw);
+			ret = au1550_spi_dma_rxtmp_alloc(hw, max(t->len,
+					AU1550_SPI_DMA_RXTMP_MINSIZE));
+			if (ret < 0)
+				return ret;
+		}
+		hw->rx = hw->dma_rx_tmpbuf;
+		dma_rx_addr = hw->dma_rx_tmpbuf_addr;
+		dma_sync_single_for_device(hw->dev, dma_rx_addr,
+			t->len, DMA_FROM_DEVICE);
+	}
+
+	if (!t->tx_buf) {
+		dma_sync_single_for_device(hw->dev, dma_rx_addr,
+				t->len, DMA_BIDIRECTIONAL);
+		hw->tx = hw->rx;
+	}
+
+	/* put buffers on the ring */
+	res = au1xxx_dbdma_put_dest(hw->dma_rx_ch, virt_to_phys(hw->rx),
+				    t->len, DDMA_FLAGS_IE);
+	if (!res)
+		dev_err(hw->dev, "rx dma put dest error\n");
+
+	res = au1xxx_dbdma_put_source(hw->dma_tx_ch, virt_to_phys(hw->tx),
+				      t->len, DDMA_FLAGS_IE);
+	if (!res)
+		dev_err(hw->dev, "tx dma put source error\n");
+
+	au1xxx_dbdma_start(hw->dma_rx_ch);
+	au1xxx_dbdma_start(hw->dma_tx_ch);
+
+	/* by default enable nearly all events interrupt */
+	hw->regs->psc_spimsk = PSC_SPIMSK_SD;
+	wmb(); /* drain writebuffer */
+
+	/* start the transfer */
+	hw->regs->psc_spipcr = PSC_SPIPCR_MS;
+	wmb(); /* drain writebuffer */
+
+	wait_for_completion(&hw->master_done);
+
+	au1xxx_dbdma_stop(hw->dma_tx_ch);
+	au1xxx_dbdma_stop(hw->dma_rx_ch);
+
+	if (!t->rx_buf) {
+		/* using the temporal preallocated and premapped buffer */
+		dma_sync_single_for_cpu(hw->dev, dma_rx_addr, t->len,
+			DMA_FROM_DEVICE);
+	}
+	/* unmap buffers if mapped above */
+	if (t->rx_buf && t->rx_dma == 0 )
+		dma_unmap_single(hw->dev, dma_rx_addr, t->len,
+			DMA_FROM_DEVICE);
+	if (t->tx_buf && t->tx_dma == 0 )
+		dma_unmap_single(hw->dev, dma_tx_addr, t->len,
+			DMA_TO_DEVICE);
+
+	return hw->rx_count < hw->tx_count ? hw->rx_count : hw->tx_count;
+}
+
+static irqreturn_t au1550_spi_dma_irq_callback(struct au1550_spi *hw)
+{
+	u32 stat, evnt;
+
+	stat = hw->regs->psc_spistat;
+	evnt = hw->regs->psc_spievent;
+	wmb(); /* drain writebuffer */
+	if ((stat & PSC_SPISTAT_DI) == 0) {
+		dev_err(hw->dev, "Unexpected IRQ!\n");
+		return IRQ_NONE;
+	}
+
+	if ((evnt & (PSC_SPIEVNT_MM | PSC_SPIEVNT_RO
+				| PSC_SPIEVNT_RU | PSC_SPIEVNT_TO
+				| PSC_SPIEVNT_TU | PSC_SPIEVNT_SD))
+			!= 0) {
+		/*
+		 * due to an spi error we consider transfer as done,
+		 * so mask all events until before next transfer start
+		 * and stop the possibly running dma immediately
+		 */
+		au1550_spi_mask_ack_all(hw);
+		au1xxx_dbdma_stop(hw->dma_rx_ch);
+		au1xxx_dbdma_stop(hw->dma_tx_ch);
+
+		/* get number of transferred bytes */
+		hw->rx_count = hw->len - au1xxx_get_dma_residue(hw->dma_rx_ch);
+		hw->tx_count = hw->len - au1xxx_get_dma_residue(hw->dma_tx_ch);
+
+		au1xxx_dbdma_reset(hw->dma_rx_ch);
+		au1xxx_dbdma_reset(hw->dma_tx_ch);
+		au1550_spi_reset_fifos(hw);
+
+		if (evnt == PSC_SPIEVNT_RO)
+			dev_err(hw->dev,
+				"dma transfer: receive FIFO overflow!\n");
+		else
+			dev_err(hw->dev,
+				"dma transfer: unexpected SPI error "
+				"(event=0x%x stat=0x%x)!\n", evnt, stat);
+
+		complete(&hw->master_done);
+		return IRQ_HANDLED;
+	}
+
+	if ((evnt & PSC_SPIEVNT_MD) != 0) {
+		/* transfer completed successfully */
+		au1550_spi_mask_ack_all(hw);
+		hw->rx_count = hw->len;
+		hw->tx_count = hw->len;
+		complete(&hw->master_done);
+	}
+	return IRQ_HANDLED;
+}
+
+
+/* routines to handle different word sizes in pio mode */
+#define AU1550_SPI_RX_WORD(size, mask)					\
+static void au1550_spi_rx_word_##size(struct au1550_spi *hw)		\
+{									\
+	u32 fifoword = hw->regs->psc_spitxrx & (u32)(mask);		\
+	wmb(); /* drain writebuffer */					\
+	if (hw->rx) {							\
+		*(u##size *)hw->rx = (u##size)fifoword;			\
+		hw->rx += (size) / 8;					\
+	}								\
+	hw->rx_count += (size) / 8;					\
+}
+
+#define AU1550_SPI_TX_WORD(size, mask)					\
+static void au1550_spi_tx_word_##size(struct au1550_spi *hw)		\
+{									\
+	u32 fifoword = 0;						\
+	if (hw->tx) {							\
+		fifoword = *(u##size *)hw->tx & (u32)(mask);		\
+		hw->tx += (size) / 8;					\
+	}								\
+	hw->tx_count += (size) / 8;					\
+	if (hw->tx_count >= hw->len)					\
+		fifoword |= PSC_SPITXRX_LC;				\
+	hw->regs->psc_spitxrx = fifoword;				\
+	wmb(); /* drain writebuffer */					\
+}
+
+AU1550_SPI_RX_WORD(8,0xff)
+AU1550_SPI_RX_WORD(16,0xffff)
+AU1550_SPI_RX_WORD(32,0xffffff)
+AU1550_SPI_TX_WORD(8,0xff)
+AU1550_SPI_TX_WORD(16,0xffff)
+AU1550_SPI_TX_WORD(32,0xffffff)
+
+static int au1550_spi_pio_txrxb(struct spi_device *spi, struct spi_transfer *t)
+{
+	u32 stat, mask;
+	struct au1550_spi *hw = spi_master_get_devdata(spi->master);
+
+	hw->tx = t->tx_buf;
+	hw->rx = t->rx_buf;
+	hw->len = t->len;
+	hw->tx_count = 0;
+	hw->rx_count = 0;
+
+	/* by default enable nearly all events after filling tx fifo */
+	mask = PSC_SPIMSK_SD;
+
+	/* fill the transmit FIFO */
+	while (hw->tx_count < hw->len) {
+
+		hw->tx_word(hw);
+
+		if (hw->tx_count >= hw->len) {
+			/* mask tx fifo request interrupt as we are done */
+			mask |= PSC_SPIMSK_TR;
+		}
+
+		stat = hw->regs->psc_spistat;
+		wmb(); /* drain writebuffer */
+		if (stat & PSC_SPISTAT_TF)
+			break;
+	}
+
+	/* enable event interrupts */
+	hw->regs->psc_spimsk = mask;
+	wmb(); /* drain writebuffer */
+
+	/* start the transfer */
+	hw->regs->psc_spipcr = PSC_SPIPCR_MS;
+	wmb(); /* drain writebuffer */
+
+	wait_for_completion(&hw->master_done);
+
+	return hw->rx_count < hw->tx_count ? hw->rx_count : hw->tx_count;
+}
+
+static irqreturn_t au1550_spi_pio_irq_callback(struct au1550_spi *hw)
+{
+	int busy;
+	u32 stat, evnt;
+
+	stat = hw->regs->psc_spistat;
+	evnt = hw->regs->psc_spievent;
+	wmb(); /* drain writebuffer */
+	if ((stat & PSC_SPISTAT_DI) == 0) {
+		dev_err(hw->dev, "Unexpected IRQ!\n");
+		return IRQ_NONE;
+	}
+
+	if ((evnt & (PSC_SPIEVNT_MM | PSC_SPIEVNT_RO
+				| PSC_SPIEVNT_RU | PSC_SPIEVNT_TO
+				| PSC_SPIEVNT_SD))
+			!= 0) {
+		/*
+		 * due to an error we consider transfer as done,
+		 * so mask all events until before next transfer start
+		 */
+		au1550_spi_mask_ack_all(hw);
+		au1550_spi_reset_fifos(hw);
+		dev_err(hw->dev,
+			"pio transfer: unexpected SPI error "
+			"(event=0x%x stat=0x%x)!\n", evnt, stat);
+		complete(&hw->master_done);
+		return IRQ_HANDLED;
+	}
+
+	/*
+	 * while there is something to read from rx fifo
+	 * or there is a space to write to tx fifo:
+	 */
+	do {
+		busy = 0;
+		stat = hw->regs->psc_spistat;
+		wmb(); /* drain writebuffer */
+
+		/*
+		 * Take care to not let the Rx FIFO overflow.
+		 *
+		 * We only write a byte if we have read one at least. Initially,
+		 * the write fifo is full, so we should read from the read fifo
+		 * first.
+		 * In case we miss a word from the read fifo, we should get a
+		 * RO event and should back out.
+		 */
+		if (!(stat & PSC_SPISTAT_RE) && hw->rx_count < hw->len) {
+			hw->rx_word(hw);
+			busy = 1;
+
+			if (!(stat & PSC_SPISTAT_TF) && hw->tx_count < hw->len)
+				hw->tx_word(hw);
+		}
+	} while (busy);
+
+	hw->regs->psc_spievent = PSC_SPIEVNT_RR | PSC_SPIEVNT_TR;
+	wmb(); /* drain writebuffer */
+
+	/*
+	 * Restart the SPI transmission in case of a transmit underflow.
+	 * This seems to work despite the notes in the Au1550 data book
+	 * of Figure 8-4 with flowchart for SPI master operation:
+	 *
+	 * """Note 1: An XFR Error Interrupt occurs, unless masked,
+	 * for any of the following events: Tx FIFO Underflow,
+	 * Rx FIFO Overflow, or Multiple-master Error
+	 *    Note 2: In case of a Tx Underflow Error, all zeroes are
+	 * transmitted."""
+	 *
+	 * By simply restarting the spi transfer on Tx Underflow Error,
+	 * we assume that spi transfer was paused instead of zeroes
+	 * transmittion mentioned in the Note 2 of Au1550 data book.
+	 */
+	if (evnt & PSC_SPIEVNT_TU) {
+		hw->regs->psc_spievent = PSC_SPIEVNT_TU | PSC_SPIEVNT_MD;
+		wmb(); /* drain writebuffer */
+		hw->regs->psc_spipcr = PSC_SPIPCR_MS;
+		wmb(); /* drain writebuffer */
+	}
+
+	if (hw->rx_count >= hw->len) {
+		/* transfer completed successfully */
+		au1550_spi_mask_ack_all(hw);
+		complete(&hw->master_done);
+	}
+	return IRQ_HANDLED;
+}
+
+static int au1550_spi_txrx_bufs(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct au1550_spi *hw = spi_master_get_devdata(spi->master);
+	return hw->txrx_bufs(spi, t);
+}
+
+static irqreturn_t au1550_spi_irq(int irq, void *dev)
+{
+	struct au1550_spi *hw = dev;
+	return hw->irq_callback(hw);
+}
+
+static void au1550_spi_bits_handlers_set(struct au1550_spi *hw, int bpw)
+{
+	if (bpw <= 8) {
+		if (hw->usedma) {
+			hw->txrx_bufs = &au1550_spi_dma_txrxb;
+			hw->irq_callback = &au1550_spi_dma_irq_callback;
+		} else {
+			hw->rx_word = &au1550_spi_rx_word_8;
+			hw->tx_word = &au1550_spi_tx_word_8;
+			hw->txrx_bufs = &au1550_spi_pio_txrxb;
+			hw->irq_callback = &au1550_spi_pio_irq_callback;
+		}
+	} else if (bpw <= 16) {
+		hw->rx_word = &au1550_spi_rx_word_16;
+		hw->tx_word = &au1550_spi_tx_word_16;
+		hw->txrx_bufs = &au1550_spi_pio_txrxb;
+		hw->irq_callback = &au1550_spi_pio_irq_callback;
+	} else {
+		hw->rx_word = &au1550_spi_rx_word_32;
+		hw->tx_word = &au1550_spi_tx_word_32;
+		hw->txrx_bufs = &au1550_spi_pio_txrxb;
+		hw->irq_callback = &au1550_spi_pio_irq_callback;
+	}
+}
+
+static void au1550_spi_setup_psc_as_spi(struct au1550_spi *hw)
+{
+	u32 stat, cfg;
+
+	/* set up the PSC for SPI mode */
+	hw->regs->psc_ctrl = PSC_CTRL_DISABLE;
+	wmb(); /* drain writebuffer */
+	hw->regs->psc_sel = PSC_SEL_PS_SPIMODE;
+	wmb(); /* drain writebuffer */
+
+	hw->regs->psc_spicfg = 0;
+	wmb(); /* drain writebuffer */
+
+	hw->regs->psc_ctrl = PSC_CTRL_ENABLE;
+	wmb(); /* drain writebuffer */
+
+	do {
+		stat = hw->regs->psc_spistat;
+		wmb(); /* drain writebuffer */
+	} while ((stat & PSC_SPISTAT_SR) == 0);
+
+
+	cfg = hw->usedma ? 0 : PSC_SPICFG_DD_DISABLE;
+	cfg |= PSC_SPICFG_SET_LEN(8);
+	cfg |= PSC_SPICFG_RT_FIFO8 | PSC_SPICFG_TT_FIFO8;
+	/* use minimal allowed brg and div values as initial setting: */
+	cfg |= PSC_SPICFG_SET_BAUD(4) | PSC_SPICFG_SET_DIV(0);
+
+#ifdef AU1550_SPI_DEBUG_LOOPBACK
+	cfg |= PSC_SPICFG_LB;
+#endif
+
+	hw->regs->psc_spicfg = cfg;
+	wmb(); /* drain writebuffer */
+
+	au1550_spi_mask_ack_all(hw);
+
+	hw->regs->psc_spicfg |= PSC_SPICFG_DE_ENABLE;
+	wmb(); /* drain writebuffer */
+
+	do {
+		stat = hw->regs->psc_spistat;
+		wmb(); /* drain writebuffer */
+	} while ((stat & PSC_SPISTAT_DR) == 0);
+
+	au1550_spi_reset_fifos(hw);
+}
+
+
+static int au1550_spi_probe(struct platform_device *pdev)
+{
+	struct au1550_spi *hw;
+	struct spi_master *master;
+	struct resource *r;
+	int err = 0;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct au1550_spi));
+	if (master == NULL) {
+		dev_err(&pdev->dev, "No memory for spi_master\n");
+		err = -ENOMEM;
+		goto err_nomem;
+	}
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LSB_FIRST;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 24);
+
+	hw = spi_master_get_devdata(master);
+
+	hw->master = master;
+	hw->pdata = dev_get_platdata(&pdev->dev);
+	hw->dev = &pdev->dev;
+
+	if (hw->pdata == NULL) {
+		dev_err(&pdev->dev, "No platform data supplied\n");
+		err = -ENOENT;
+		goto err_no_pdata;
+	}
+
+	r = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
+	if (!r) {
+		dev_err(&pdev->dev, "no IRQ\n");
+		err = -ENODEV;
+		goto err_no_iores;
+	}
+	hw->irq = r->start;
+
+	hw->usedma = 0;
+	r = platform_get_resource(pdev, IORESOURCE_DMA, 0);
+	if (r) {
+		hw->dma_tx_id = r->start;
+		r = platform_get_resource(pdev, IORESOURCE_DMA, 1);
+		if (r) {
+			hw->dma_rx_id = r->start;
+			if (usedma && ddma_memid) {
+				if (pdev->dev.dma_mask == NULL)
+					dev_warn(&pdev->dev, "no dma mask\n");
+				else
+					hw->usedma = 1;
+			}
+		}
+	}
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!r) {
+		dev_err(&pdev->dev, "no mmio resource\n");
+		err = -ENODEV;
+		goto err_no_iores;
+	}
+
+	hw->ioarea = request_mem_region(r->start, sizeof(psc_spi_t),
+					pdev->name);
+	if (!hw->ioarea) {
+		dev_err(&pdev->dev, "Cannot reserve iomem region\n");
+		err = -ENXIO;
+		goto err_no_iores;
+	}
+
+	hw->regs = (psc_spi_t __iomem *)ioremap(r->start, sizeof(psc_spi_t));
+	if (!hw->regs) {
+		dev_err(&pdev->dev, "cannot ioremap\n");
+		err = -ENXIO;
+		goto err_ioremap;
+	}
+
+	platform_set_drvdata(pdev, hw);
+
+	init_completion(&hw->master_done);
+
+	hw->bitbang.master = hw->master;
+	hw->bitbang.setup_transfer = au1550_spi_setupxfer;
+	hw->bitbang.chipselect = au1550_spi_chipsel;
+	hw->bitbang.txrx_bufs = au1550_spi_txrx_bufs;
+
+	if (hw->usedma) {
+		hw->dma_tx_ch = au1xxx_dbdma_chan_alloc(ddma_memid,
+			hw->dma_tx_id, NULL, (void *)hw);
+		if (hw->dma_tx_ch == 0) {
+			dev_err(&pdev->dev,
+				"Cannot allocate tx dma channel\n");
+			err = -ENXIO;
+			goto err_no_txdma;
+		}
+		au1xxx_dbdma_set_devwidth(hw->dma_tx_ch, 8);
+		if (au1xxx_dbdma_ring_alloc(hw->dma_tx_ch,
+			AU1550_SPI_DBDMA_DESCRIPTORS) == 0) {
+			dev_err(&pdev->dev,
+				"Cannot allocate tx dma descriptors\n");
+			err = -ENXIO;
+			goto err_no_txdma_descr;
+		}
+
+
+		hw->dma_rx_ch = au1xxx_dbdma_chan_alloc(hw->dma_rx_id,
+			ddma_memid, NULL, (void *)hw);
+		if (hw->dma_rx_ch == 0) {
+			dev_err(&pdev->dev,
+				"Cannot allocate rx dma channel\n");
+			err = -ENXIO;
+			goto err_no_rxdma;
+		}
+		au1xxx_dbdma_set_devwidth(hw->dma_rx_ch, 8);
+		if (au1xxx_dbdma_ring_alloc(hw->dma_rx_ch,
+			AU1550_SPI_DBDMA_DESCRIPTORS) == 0) {
+			dev_err(&pdev->dev,
+				"Cannot allocate rx dma descriptors\n");
+			err = -ENXIO;
+			goto err_no_rxdma_descr;
+		}
+
+		err = au1550_spi_dma_rxtmp_alloc(hw,
+			AU1550_SPI_DMA_RXTMP_MINSIZE);
+		if (err < 0) {
+			dev_err(&pdev->dev,
+				"Cannot allocate initial rx dma tmp buffer\n");
+			goto err_dma_rxtmp_alloc;
+		}
+	}
+
+	au1550_spi_bits_handlers_set(hw, 8);
+
+	err = request_irq(hw->irq, au1550_spi_irq, 0, pdev->name, hw);
+	if (err) {
+		dev_err(&pdev->dev, "Cannot claim IRQ\n");
+		goto err_no_irq;
+	}
+
+	master->bus_num = pdev->id;
+	master->num_chipselect = hw->pdata->num_chipselect;
+
+	/*
+	 *  precompute valid range for spi freq - from au1550 datasheet:
+	 *    psc_tempclk = psc_mainclk / (2 << DIV)
+	 *    spiclk = psc_tempclk / (2 * (BRG + 1))
+	 *    BRG valid range is 4..63
+	 *    DIV valid range is 0..3
+	 *  round the min and max frequencies to values that would still
+	 *  produce valid brg and div
+	 */
+	{
+		int min_div = (2 << 0) * (2 * (4 + 1));
+		int max_div = (2 << 3) * (2 * (63 + 1));
+		master->max_speed_hz = hw->pdata->mainclk_hz / min_div;
+		master->min_speed_hz =
+				hw->pdata->mainclk_hz / (max_div + 1) + 1;
+	}
+
+	au1550_spi_setup_psc_as_spi(hw);
+
+	err = spi_bitbang_start(&hw->bitbang);
+	if (err) {
+		dev_err(&pdev->dev, "Failed to register SPI master\n");
+		goto err_register;
+	}
+
+	dev_info(&pdev->dev,
+		"spi master registered: bus_num=%d num_chipselect=%d\n",
+		master->bus_num, master->num_chipselect);
+
+	return 0;
+
+err_register:
+	free_irq(hw->irq, hw);
+
+err_no_irq:
+	au1550_spi_dma_rxtmp_free(hw);
+
+err_dma_rxtmp_alloc:
+err_no_rxdma_descr:
+	if (hw->usedma)
+		au1xxx_dbdma_chan_free(hw->dma_rx_ch);
+
+err_no_rxdma:
+err_no_txdma_descr:
+	if (hw->usedma)
+		au1xxx_dbdma_chan_free(hw->dma_tx_ch);
+
+err_no_txdma:
+	iounmap((void __iomem *)hw->regs);
+
+err_ioremap:
+	release_mem_region(r->start, sizeof(psc_spi_t));
+
+err_no_iores:
+err_no_pdata:
+	spi_master_put(hw->master);
+
+err_nomem:
+	return err;
+}
+
+static int au1550_spi_remove(struct platform_device *pdev)
+{
+	struct au1550_spi *hw = platform_get_drvdata(pdev);
+
+	dev_info(&pdev->dev, "spi master remove: bus_num=%d\n",
+		hw->master->bus_num);
+
+	spi_bitbang_stop(&hw->bitbang);
+	free_irq(hw->irq, hw);
+	iounmap((void __iomem *)hw->regs);
+	release_mem_region(hw->ioarea->start, sizeof(psc_spi_t));
+
+	if (hw->usedma) {
+		au1550_spi_dma_rxtmp_free(hw);
+		au1xxx_dbdma_chan_free(hw->dma_rx_ch);
+		au1xxx_dbdma_chan_free(hw->dma_tx_ch);
+	}
+
+	spi_master_put(hw->master);
+	return 0;
+}
+
+/* work with hotplug and coldplug */
+MODULE_ALIAS("platform:au1550-spi");
+
+static struct platform_driver au1550_spi_drv = {
+	.probe = au1550_spi_probe,
+	.remove = au1550_spi_remove,
+	.driver = {
+		.name = "au1550-spi",
+	},
+};
+
+static int __init au1550_spi_init(void)
+{
+	/*
+	 * create memory device with 8 bits dev_devwidth
+	 * needed for proper byte ordering to spi fifo
+	 */
+	switch (alchemy_get_cputype()) {
+	case ALCHEMY_CPU_AU1550:
+	case ALCHEMY_CPU_AU1200:
+	case ALCHEMY_CPU_AU1300:
+		break;
+	default:
+		return -ENODEV;
+	}
+
+	if (usedma) {
+		ddma_memid = au1xxx_ddma_add_device(&au1550_spi_mem_dbdev);
+		if (!ddma_memid)
+			printk(KERN_ERR "au1550-spi: cannot add memory"
+					"dbdma device\n");
+	}
+	return platform_driver_register(&au1550_spi_drv);
+}
+module_init(au1550_spi_init);
+
+static void __exit au1550_spi_exit(void)
+{
+	if (usedma && ddma_memid)
+		au1xxx_ddma_del_device(ddma_memid);
+	platform_driver_unregister(&au1550_spi_drv);
+}
+module_exit(au1550_spi_exit);
+
+MODULE_DESCRIPTION("Au1550 PSC SPI Driver");
+MODULE_AUTHOR("Jan Nikitenko <jan.nikitenko@gmail.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-axi-spi-engine.c b/drivers/spi/spi-axi-spi-engine.c
new file mode 100644
index 0000000..68cfc35
--- /dev/null
+++ b/drivers/spi/spi-axi-spi-engine.c
@@ -0,0 +1,594 @@
+/*
+ * SPI-Engine SPI controller driver
+ * Copyright 2015 Analog Devices Inc.
+ *  Author: Lars-Peter Clausen <lars@metafoo.de>
+ *
+ * Licensed under the GPL-2.
+ */
+
+#include <linux/clk.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/of.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+
+#define SPI_ENGINE_VERSION_MAJOR(x)	((x >> 16) & 0xff)
+#define SPI_ENGINE_VERSION_MINOR(x)	((x >> 8) & 0xff)
+#define SPI_ENGINE_VERSION_PATCH(x)	(x & 0xff)
+
+#define SPI_ENGINE_REG_VERSION			0x00
+
+#define SPI_ENGINE_REG_RESET			0x40
+
+#define SPI_ENGINE_REG_INT_ENABLE		0x80
+#define SPI_ENGINE_REG_INT_PENDING		0x84
+#define SPI_ENGINE_REG_INT_SOURCE		0x88
+
+#define SPI_ENGINE_REG_SYNC_ID			0xc0
+
+#define SPI_ENGINE_REG_CMD_FIFO_ROOM		0xd0
+#define SPI_ENGINE_REG_SDO_FIFO_ROOM		0xd4
+#define SPI_ENGINE_REG_SDI_FIFO_LEVEL		0xd8
+
+#define SPI_ENGINE_REG_CMD_FIFO			0xe0
+#define SPI_ENGINE_REG_SDO_DATA_FIFO		0xe4
+#define SPI_ENGINE_REG_SDI_DATA_FIFO		0xe8
+#define SPI_ENGINE_REG_SDI_DATA_FIFO_PEEK	0xec
+
+#define SPI_ENGINE_INT_CMD_ALMOST_EMPTY		BIT(0)
+#define SPI_ENGINE_INT_SDO_ALMOST_EMPTY		BIT(1)
+#define SPI_ENGINE_INT_SDI_ALMOST_FULL		BIT(2)
+#define SPI_ENGINE_INT_SYNC			BIT(3)
+
+#define SPI_ENGINE_CONFIG_CPHA			BIT(0)
+#define SPI_ENGINE_CONFIG_CPOL			BIT(1)
+#define SPI_ENGINE_CONFIG_3WIRE			BIT(2)
+
+#define SPI_ENGINE_INST_TRANSFER		0x0
+#define SPI_ENGINE_INST_ASSERT			0x1
+#define SPI_ENGINE_INST_WRITE			0x2
+#define SPI_ENGINE_INST_MISC			0x3
+
+#define SPI_ENGINE_CMD_REG_CLK_DIV		0x0
+#define SPI_ENGINE_CMD_REG_CONFIG		0x1
+
+#define SPI_ENGINE_MISC_SYNC			0x0
+#define SPI_ENGINE_MISC_SLEEP			0x1
+
+#define SPI_ENGINE_TRANSFER_WRITE		0x1
+#define SPI_ENGINE_TRANSFER_READ		0x2
+
+#define SPI_ENGINE_CMD(inst, arg1, arg2) \
+	(((inst) << 12) | ((arg1) << 8) | (arg2))
+
+#define SPI_ENGINE_CMD_TRANSFER(flags, n) \
+	SPI_ENGINE_CMD(SPI_ENGINE_INST_TRANSFER, (flags), (n))
+#define SPI_ENGINE_CMD_ASSERT(delay, cs) \
+	SPI_ENGINE_CMD(SPI_ENGINE_INST_ASSERT, (delay), (cs))
+#define SPI_ENGINE_CMD_WRITE(reg, val) \
+	SPI_ENGINE_CMD(SPI_ENGINE_INST_WRITE, (reg), (val))
+#define SPI_ENGINE_CMD_SLEEP(delay) \
+	SPI_ENGINE_CMD(SPI_ENGINE_INST_MISC, SPI_ENGINE_MISC_SLEEP, (delay))
+#define SPI_ENGINE_CMD_SYNC(id) \
+	SPI_ENGINE_CMD(SPI_ENGINE_INST_MISC, SPI_ENGINE_MISC_SYNC, (id))
+
+struct spi_engine_program {
+	unsigned int length;
+	uint16_t instructions[];
+};
+
+struct spi_engine {
+	struct clk *clk;
+	struct clk *ref_clk;
+
+	spinlock_t lock;
+
+	void __iomem *base;
+
+	struct spi_message *msg;
+	struct spi_engine_program *p;
+	unsigned cmd_length;
+	const uint16_t *cmd_buf;
+
+	struct spi_transfer *tx_xfer;
+	unsigned int tx_length;
+	const uint8_t *tx_buf;
+
+	struct spi_transfer *rx_xfer;
+	unsigned int rx_length;
+	uint8_t *rx_buf;
+
+	unsigned int sync_id;
+	unsigned int completed_id;
+
+	unsigned int int_enable;
+};
+
+static void spi_engine_program_add_cmd(struct spi_engine_program *p,
+	bool dry, uint16_t cmd)
+{
+	if (!dry)
+		p->instructions[p->length] = cmd;
+	p->length++;
+}
+
+static unsigned int spi_engine_get_config(struct spi_device *spi)
+{
+	unsigned int config = 0;
+
+	if (spi->mode & SPI_CPOL)
+		config |= SPI_ENGINE_CONFIG_CPOL;
+	if (spi->mode & SPI_CPHA)
+		config |= SPI_ENGINE_CONFIG_CPHA;
+	if (spi->mode & SPI_3WIRE)
+		config |= SPI_ENGINE_CONFIG_3WIRE;
+
+	return config;
+}
+
+static unsigned int spi_engine_get_clk_div(struct spi_engine *spi_engine,
+	struct spi_device *spi, struct spi_transfer *xfer)
+{
+	unsigned int clk_div;
+
+	clk_div = DIV_ROUND_UP(clk_get_rate(spi_engine->ref_clk),
+		xfer->speed_hz * 2);
+	if (clk_div > 255)
+		clk_div = 255;
+	else if (clk_div > 0)
+		clk_div -= 1;
+
+	return clk_div;
+}
+
+static void spi_engine_gen_xfer(struct spi_engine_program *p, bool dry,
+	struct spi_transfer *xfer)
+{
+	unsigned int len = xfer->len;
+
+	while (len) {
+		unsigned int n = min(len, 256U);
+		unsigned int flags = 0;
+
+		if (xfer->tx_buf)
+			flags |= SPI_ENGINE_TRANSFER_WRITE;
+		if (xfer->rx_buf)
+			flags |= SPI_ENGINE_TRANSFER_READ;
+
+		spi_engine_program_add_cmd(p, dry,
+			SPI_ENGINE_CMD_TRANSFER(flags, n - 1));
+		len -= n;
+	}
+}
+
+static void spi_engine_gen_sleep(struct spi_engine_program *p, bool dry,
+	struct spi_engine *spi_engine, unsigned int clk_div, unsigned int delay)
+{
+	unsigned int spi_clk = clk_get_rate(spi_engine->ref_clk);
+	unsigned int t;
+
+	if (delay == 0)
+		return;
+
+	t = DIV_ROUND_UP(delay * spi_clk, (clk_div + 1) * 2);
+	while (t) {
+		unsigned int n = min(t, 256U);
+
+		spi_engine_program_add_cmd(p, dry, SPI_ENGINE_CMD_SLEEP(n - 1));
+		t -= n;
+	}
+}
+
+static void spi_engine_gen_cs(struct spi_engine_program *p, bool dry,
+		struct spi_device *spi, bool assert)
+{
+	unsigned int mask = 0xff;
+
+	if (assert)
+		mask ^= BIT(spi->chip_select);
+
+	spi_engine_program_add_cmd(p, dry, SPI_ENGINE_CMD_ASSERT(1, mask));
+}
+
+static int spi_engine_compile_message(struct spi_engine *spi_engine,
+	struct spi_message *msg, bool dry, struct spi_engine_program *p)
+{
+	struct spi_device *spi = msg->spi;
+	struct spi_transfer *xfer;
+	int clk_div, new_clk_div;
+	bool cs_change = true;
+
+	clk_div = -1;
+
+	spi_engine_program_add_cmd(p, dry,
+		SPI_ENGINE_CMD_WRITE(SPI_ENGINE_CMD_REG_CONFIG,
+			spi_engine_get_config(spi)));
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		new_clk_div = spi_engine_get_clk_div(spi_engine, spi, xfer);
+		if (new_clk_div != clk_div) {
+			clk_div = new_clk_div;
+			spi_engine_program_add_cmd(p, dry,
+				SPI_ENGINE_CMD_WRITE(SPI_ENGINE_CMD_REG_CLK_DIV,
+					clk_div));
+		}
+
+		if (cs_change)
+			spi_engine_gen_cs(p, dry, spi, true);
+
+		spi_engine_gen_xfer(p, dry, xfer);
+		spi_engine_gen_sleep(p, dry, spi_engine, clk_div,
+			xfer->delay_usecs);
+
+		cs_change = xfer->cs_change;
+		if (list_is_last(&xfer->transfer_list, &msg->transfers))
+			cs_change = !cs_change;
+
+		if (cs_change)
+			spi_engine_gen_cs(p, dry, spi, false);
+	}
+
+	return 0;
+}
+
+static void spi_engine_xfer_next(struct spi_engine *spi_engine,
+	struct spi_transfer **_xfer)
+{
+	struct spi_message *msg = spi_engine->msg;
+	struct spi_transfer *xfer = *_xfer;
+
+	if (!xfer) {
+		xfer = list_first_entry(&msg->transfers,
+			struct spi_transfer, transfer_list);
+	} else if (list_is_last(&xfer->transfer_list, &msg->transfers)) {
+		xfer = NULL;
+	} else {
+		xfer = list_next_entry(xfer, transfer_list);
+	}
+
+	*_xfer = xfer;
+}
+
+static void spi_engine_tx_next(struct spi_engine *spi_engine)
+{
+	struct spi_transfer *xfer = spi_engine->tx_xfer;
+
+	do {
+		spi_engine_xfer_next(spi_engine, &xfer);
+	} while (xfer && !xfer->tx_buf);
+
+	spi_engine->tx_xfer = xfer;
+	if (xfer) {
+		spi_engine->tx_length = xfer->len;
+		spi_engine->tx_buf = xfer->tx_buf;
+	} else {
+		spi_engine->tx_buf = NULL;
+	}
+}
+
+static void spi_engine_rx_next(struct spi_engine *spi_engine)
+{
+	struct spi_transfer *xfer = spi_engine->rx_xfer;
+
+	do {
+		spi_engine_xfer_next(spi_engine, &xfer);
+	} while (xfer && !xfer->rx_buf);
+
+	spi_engine->rx_xfer = xfer;
+	if (xfer) {
+		spi_engine->rx_length = xfer->len;
+		spi_engine->rx_buf = xfer->rx_buf;
+	} else {
+		spi_engine->rx_buf = NULL;
+	}
+}
+
+static bool spi_engine_write_cmd_fifo(struct spi_engine *spi_engine)
+{
+	void __iomem *addr = spi_engine->base + SPI_ENGINE_REG_CMD_FIFO;
+	unsigned int n, m, i;
+	const uint16_t *buf;
+
+	n = readl_relaxed(spi_engine->base + SPI_ENGINE_REG_CMD_FIFO_ROOM);
+	while (n && spi_engine->cmd_length) {
+		m = min(n, spi_engine->cmd_length);
+		buf = spi_engine->cmd_buf;
+		for (i = 0; i < m; i++)
+			writel_relaxed(buf[i], addr);
+		spi_engine->cmd_buf += m;
+		spi_engine->cmd_length -= m;
+		n -= m;
+	}
+
+	return spi_engine->cmd_length != 0;
+}
+
+static bool spi_engine_write_tx_fifo(struct spi_engine *spi_engine)
+{
+	void __iomem *addr = spi_engine->base + SPI_ENGINE_REG_SDO_DATA_FIFO;
+	unsigned int n, m, i;
+	const uint8_t *buf;
+
+	n = readl_relaxed(spi_engine->base + SPI_ENGINE_REG_SDO_FIFO_ROOM);
+	while (n && spi_engine->tx_length) {
+		m = min(n, spi_engine->tx_length);
+		buf = spi_engine->tx_buf;
+		for (i = 0; i < m; i++)
+			writel_relaxed(buf[i], addr);
+		spi_engine->tx_buf += m;
+		spi_engine->tx_length -= m;
+		n -= m;
+		if (spi_engine->tx_length == 0)
+			spi_engine_tx_next(spi_engine);
+	}
+
+	return spi_engine->tx_length != 0;
+}
+
+static bool spi_engine_read_rx_fifo(struct spi_engine *spi_engine)
+{
+	void __iomem *addr = spi_engine->base + SPI_ENGINE_REG_SDI_DATA_FIFO;
+	unsigned int n, m, i;
+	uint8_t *buf;
+
+	n = readl_relaxed(spi_engine->base + SPI_ENGINE_REG_SDI_FIFO_LEVEL);
+	while (n && spi_engine->rx_length) {
+		m = min(n, spi_engine->rx_length);
+		buf = spi_engine->rx_buf;
+		for (i = 0; i < m; i++)
+			buf[i] = readl_relaxed(addr);
+		spi_engine->rx_buf += m;
+		spi_engine->rx_length -= m;
+		n -= m;
+		if (spi_engine->rx_length == 0)
+			spi_engine_rx_next(spi_engine);
+	}
+
+	return spi_engine->rx_length != 0;
+}
+
+static irqreturn_t spi_engine_irq(int irq, void *devid)
+{
+	struct spi_master *master = devid;
+	struct spi_engine *spi_engine = spi_master_get_devdata(master);
+	unsigned int disable_int = 0;
+	unsigned int pending;
+
+	pending = readl_relaxed(spi_engine->base + SPI_ENGINE_REG_INT_PENDING);
+
+	if (pending & SPI_ENGINE_INT_SYNC) {
+		writel_relaxed(SPI_ENGINE_INT_SYNC,
+			spi_engine->base + SPI_ENGINE_REG_INT_PENDING);
+		spi_engine->completed_id = readl_relaxed(
+			spi_engine->base + SPI_ENGINE_REG_SYNC_ID);
+	}
+
+	spin_lock(&spi_engine->lock);
+
+	if (pending & SPI_ENGINE_INT_CMD_ALMOST_EMPTY) {
+		if (!spi_engine_write_cmd_fifo(spi_engine))
+			disable_int |= SPI_ENGINE_INT_CMD_ALMOST_EMPTY;
+	}
+
+	if (pending & SPI_ENGINE_INT_SDO_ALMOST_EMPTY) {
+		if (!spi_engine_write_tx_fifo(spi_engine))
+			disable_int |= SPI_ENGINE_INT_SDO_ALMOST_EMPTY;
+	}
+
+	if (pending & (SPI_ENGINE_INT_SDI_ALMOST_FULL | SPI_ENGINE_INT_SYNC)) {
+		if (!spi_engine_read_rx_fifo(spi_engine))
+			disable_int |= SPI_ENGINE_INT_SDI_ALMOST_FULL;
+	}
+
+	if (pending & SPI_ENGINE_INT_SYNC) {
+		if (spi_engine->msg &&
+		    spi_engine->completed_id == spi_engine->sync_id) {
+			struct spi_message *msg = spi_engine->msg;
+
+			kfree(spi_engine->p);
+			msg->status = 0;
+			msg->actual_length = msg->frame_length;
+			spi_engine->msg = NULL;
+			spi_finalize_current_message(master);
+			disable_int |= SPI_ENGINE_INT_SYNC;
+		}
+	}
+
+	if (disable_int) {
+		spi_engine->int_enable &= ~disable_int;
+		writel_relaxed(spi_engine->int_enable,
+			spi_engine->base + SPI_ENGINE_REG_INT_ENABLE);
+	}
+
+	spin_unlock(&spi_engine->lock);
+
+	return IRQ_HANDLED;
+}
+
+static int spi_engine_transfer_one_message(struct spi_master *master,
+	struct spi_message *msg)
+{
+	struct spi_engine_program p_dry, *p;
+	struct spi_engine *spi_engine = spi_master_get_devdata(master);
+	unsigned int int_enable = 0;
+	unsigned long flags;
+	size_t size;
+
+	p_dry.length = 0;
+	spi_engine_compile_message(spi_engine, msg, true, &p_dry);
+
+	size = sizeof(*p->instructions) * (p_dry.length + 1);
+	p = kzalloc(sizeof(*p) + size, GFP_KERNEL);
+	if (!p)
+		return -ENOMEM;
+	spi_engine_compile_message(spi_engine, msg, false, p);
+
+	spin_lock_irqsave(&spi_engine->lock, flags);
+	spi_engine->sync_id = (spi_engine->sync_id + 1) & 0xff;
+	spi_engine_program_add_cmd(p, false,
+		SPI_ENGINE_CMD_SYNC(spi_engine->sync_id));
+
+	spi_engine->msg = msg;
+	spi_engine->p = p;
+
+	spi_engine->cmd_buf = p->instructions;
+	spi_engine->cmd_length = p->length;
+	if (spi_engine_write_cmd_fifo(spi_engine))
+		int_enable |= SPI_ENGINE_INT_CMD_ALMOST_EMPTY;
+
+	spi_engine_tx_next(spi_engine);
+	if (spi_engine_write_tx_fifo(spi_engine))
+		int_enable |= SPI_ENGINE_INT_SDO_ALMOST_EMPTY;
+
+	spi_engine_rx_next(spi_engine);
+	if (spi_engine->rx_length != 0)
+		int_enable |= SPI_ENGINE_INT_SDI_ALMOST_FULL;
+
+	int_enable |= SPI_ENGINE_INT_SYNC;
+
+	writel_relaxed(int_enable,
+		spi_engine->base + SPI_ENGINE_REG_INT_ENABLE);
+	spi_engine->int_enable = int_enable;
+	spin_unlock_irqrestore(&spi_engine->lock, flags);
+
+	return 0;
+}
+
+static int spi_engine_probe(struct platform_device *pdev)
+{
+	struct spi_engine *spi_engine;
+	struct spi_master *master;
+	unsigned int version;
+	struct resource *res;
+	int irq;
+	int ret;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq <= 0)
+		return -ENXIO;
+
+	spi_engine = devm_kzalloc(&pdev->dev, sizeof(*spi_engine), GFP_KERNEL);
+	if (!spi_engine)
+		return -ENOMEM;
+
+	master = spi_alloc_master(&pdev->dev, 0);
+	if (!master)
+		return -ENOMEM;
+
+	spi_master_set_devdata(master, spi_engine);
+
+	spin_lock_init(&spi_engine->lock);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	spi_engine->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(spi_engine->base)) {
+		ret = PTR_ERR(spi_engine->base);
+		goto err_put_master;
+	}
+
+	version = readl(spi_engine->base + SPI_ENGINE_REG_VERSION);
+	if (SPI_ENGINE_VERSION_MAJOR(version) != 1) {
+		dev_err(&pdev->dev, "Unsupported peripheral version %u.%u.%c\n",
+			SPI_ENGINE_VERSION_MAJOR(version),
+			SPI_ENGINE_VERSION_MINOR(version),
+			SPI_ENGINE_VERSION_PATCH(version));
+		ret = -ENODEV;
+		goto err_put_master;
+	}
+
+	spi_engine->clk = devm_clk_get(&pdev->dev, "s_axi_aclk");
+	if (IS_ERR(spi_engine->clk)) {
+		ret = PTR_ERR(spi_engine->clk);
+		goto err_put_master;
+	}
+
+	spi_engine->ref_clk = devm_clk_get(&pdev->dev, "spi_clk");
+	if (IS_ERR(spi_engine->ref_clk)) {
+		ret = PTR_ERR(spi_engine->ref_clk);
+		goto err_put_master;
+	}
+
+	ret = clk_prepare_enable(spi_engine->clk);
+	if (ret)
+		goto err_put_master;
+
+	ret = clk_prepare_enable(spi_engine->ref_clk);
+	if (ret)
+		goto err_clk_disable;
+
+	writel_relaxed(0x00, spi_engine->base + SPI_ENGINE_REG_RESET);
+	writel_relaxed(0xff, spi_engine->base + SPI_ENGINE_REG_INT_PENDING);
+	writel_relaxed(0x00, spi_engine->base + SPI_ENGINE_REG_INT_ENABLE);
+
+	ret = request_irq(irq, spi_engine_irq, 0, pdev->name, master);
+	if (ret)
+		goto err_ref_clk_disable;
+
+	master->dev.of_node = pdev->dev.of_node;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_3WIRE;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->max_speed_hz = clk_get_rate(spi_engine->ref_clk) / 2;
+	master->transfer_one_message = spi_engine_transfer_one_message;
+	master->num_chipselect = 8;
+
+	ret = spi_register_master(master);
+	if (ret)
+		goto err_free_irq;
+
+	platform_set_drvdata(pdev, master);
+
+	return 0;
+err_free_irq:
+	free_irq(irq, master);
+err_ref_clk_disable:
+	clk_disable_unprepare(spi_engine->ref_clk);
+err_clk_disable:
+	clk_disable_unprepare(spi_engine->clk);
+err_put_master:
+	spi_master_put(master);
+	return ret;
+}
+
+static int spi_engine_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = spi_master_get(platform_get_drvdata(pdev));
+	struct spi_engine *spi_engine = spi_master_get_devdata(master);
+	int irq = platform_get_irq(pdev, 0);
+
+	spi_unregister_master(master);
+
+	free_irq(irq, master);
+
+	spi_master_put(master);
+
+	writel_relaxed(0xff, spi_engine->base + SPI_ENGINE_REG_INT_PENDING);
+	writel_relaxed(0x00, spi_engine->base + SPI_ENGINE_REG_INT_ENABLE);
+	writel_relaxed(0x01, spi_engine->base + SPI_ENGINE_REG_RESET);
+
+	clk_disable_unprepare(spi_engine->ref_clk);
+	clk_disable_unprepare(spi_engine->clk);
+
+	return 0;
+}
+
+static const struct of_device_id spi_engine_match_table[] = {
+	{ .compatible = "adi,axi-spi-engine-1.00.a" },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, spi_engine_match_table);
+
+static struct platform_driver spi_engine_driver = {
+	.probe = spi_engine_probe,
+	.remove = spi_engine_remove,
+	.driver = {
+		.name = "spi-engine",
+		.of_match_table = spi_engine_match_table,
+	},
+};
+module_platform_driver(spi_engine_driver);
+
+MODULE_AUTHOR("Lars-Peter Clausen <lars@metafoo.de>");
+MODULE_DESCRIPTION("Analog Devices SPI engine peripheral driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-bcm-qspi.c b/drivers/spi/spi-bcm-qspi.c
new file mode 100644
index 0000000..584bcb0
--- /dev/null
+++ b/drivers/spi/spi-bcm-qspi.c
@@ -0,0 +1,1449 @@
+/*
+ * Driver for Broadcom BRCMSTB, NSP,  NS2, Cygnus SPI Controllers
+ *
+ * Copyright 2016 Broadcom
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation (the "GPL").
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License version 2 (GPLv2) for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * version 2 (GPLv2) along with this source code.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/ioport.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_irq.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi-mem.h>
+#include <linux/sysfs.h>
+#include <linux/types.h>
+#include "spi-bcm-qspi.h"
+
+#define DRIVER_NAME "bcm_qspi"
+
+
+/* BSPI register offsets */
+#define BSPI_REVISION_ID			0x000
+#define BSPI_SCRATCH				0x004
+#define BSPI_MAST_N_BOOT_CTRL			0x008
+#define BSPI_BUSY_STATUS			0x00c
+#define BSPI_INTR_STATUS			0x010
+#define BSPI_B0_STATUS				0x014
+#define BSPI_B0_CTRL				0x018
+#define BSPI_B1_STATUS				0x01c
+#define BSPI_B1_CTRL				0x020
+#define BSPI_STRAP_OVERRIDE_CTRL		0x024
+#define BSPI_FLEX_MODE_ENABLE			0x028
+#define BSPI_BITS_PER_CYCLE			0x02c
+#define BSPI_BITS_PER_PHASE			0x030
+#define BSPI_CMD_AND_MODE_BYTE			0x034
+#define BSPI_BSPI_FLASH_UPPER_ADDR_BYTE	0x038
+#define BSPI_BSPI_XOR_VALUE			0x03c
+#define BSPI_BSPI_XOR_ENABLE			0x040
+#define BSPI_BSPI_PIO_MODE_ENABLE		0x044
+#define BSPI_BSPI_PIO_IODIR			0x048
+#define BSPI_BSPI_PIO_DATA			0x04c
+
+/* RAF register offsets */
+#define BSPI_RAF_START_ADDR			0x100
+#define BSPI_RAF_NUM_WORDS			0x104
+#define BSPI_RAF_CTRL				0x108
+#define BSPI_RAF_FULLNESS			0x10c
+#define BSPI_RAF_WATERMARK			0x110
+#define BSPI_RAF_STATUS			0x114
+#define BSPI_RAF_READ_DATA			0x118
+#define BSPI_RAF_WORD_CNT			0x11c
+#define BSPI_RAF_CURR_ADDR			0x120
+
+/* Override mode masks */
+#define BSPI_STRAP_OVERRIDE_CTRL_OVERRIDE	BIT(0)
+#define BSPI_STRAP_OVERRIDE_CTRL_DATA_DUAL	BIT(1)
+#define BSPI_STRAP_OVERRIDE_CTRL_ADDR_4BYTE	BIT(2)
+#define BSPI_STRAP_OVERRIDE_CTRL_DATA_QUAD	BIT(3)
+#define BSPI_STRAP_OVERRIDE_CTRL_ENDAIN_MODE	BIT(4)
+
+#define BSPI_ADDRLEN_3BYTES			3
+#define BSPI_ADDRLEN_4BYTES			4
+
+#define BSPI_RAF_STATUS_FIFO_EMPTY_MASK	BIT(1)
+
+#define BSPI_RAF_CTRL_START_MASK		BIT(0)
+#define BSPI_RAF_CTRL_CLEAR_MASK		BIT(1)
+
+#define BSPI_BPP_MODE_SELECT_MASK		BIT(8)
+#define BSPI_BPP_ADDR_SELECT_MASK		BIT(16)
+
+#define BSPI_READ_LENGTH			256
+
+/* MSPI register offsets */
+#define MSPI_SPCR0_LSB				0x000
+#define MSPI_SPCR0_MSB				0x004
+#define MSPI_SPCR1_LSB				0x008
+#define MSPI_SPCR1_MSB				0x00c
+#define MSPI_NEWQP				0x010
+#define MSPI_ENDQP				0x014
+#define MSPI_SPCR2				0x018
+#define MSPI_MSPI_STATUS			0x020
+#define MSPI_CPTQP				0x024
+#define MSPI_SPCR3				0x028
+#define MSPI_TXRAM				0x040
+#define MSPI_RXRAM				0x0c0
+#define MSPI_CDRAM				0x140
+#define MSPI_WRITE_LOCK			0x180
+
+#define MSPI_MASTER_BIT			BIT(7)
+
+#define MSPI_NUM_CDRAM				16
+#define MSPI_CDRAM_CONT_BIT			BIT(7)
+#define MSPI_CDRAM_BITSE_BIT			BIT(6)
+#define MSPI_CDRAM_PCS				0xf
+
+#define MSPI_SPCR2_SPE				BIT(6)
+#define MSPI_SPCR2_CONT_AFTER_CMD		BIT(7)
+
+#define MSPI_MSPI_STATUS_SPIF			BIT(0)
+
+#define INTR_BASE_BIT_SHIFT			0x02
+#define INTR_COUNT				0x07
+
+#define NUM_CHIPSELECT				4
+#define QSPI_SPBR_MIN				8U
+#define QSPI_SPBR_MAX				255U
+
+#define OPCODE_DIOR				0xBB
+#define OPCODE_QIOR				0xEB
+#define OPCODE_DIOR_4B				0xBC
+#define OPCODE_QIOR_4B				0xEC
+
+#define MAX_CMD_SIZE				6
+
+#define ADDR_4MB_MASK				GENMASK(22, 0)
+
+/* stop at end of transfer, no other reason */
+#define TRANS_STATUS_BREAK_NONE		0
+/* stop at end of spi_message */
+#define TRANS_STATUS_BREAK_EOM			1
+/* stop at end of spi_transfer if delay */
+#define TRANS_STATUS_BREAK_DELAY		2
+/* stop at end of spi_transfer if cs_change */
+#define TRANS_STATUS_BREAK_CS_CHANGE		4
+/* stop if we run out of bytes */
+#define TRANS_STATUS_BREAK_NO_BYTES		8
+
+/* events that make us stop filling TX slots */
+#define TRANS_STATUS_BREAK_TX (TRANS_STATUS_BREAK_EOM |		\
+			       TRANS_STATUS_BREAK_DELAY |		\
+			       TRANS_STATUS_BREAK_CS_CHANGE)
+
+/* events that make us deassert CS */
+#define TRANS_STATUS_BREAK_DESELECT (TRANS_STATUS_BREAK_EOM |		\
+				     TRANS_STATUS_BREAK_CS_CHANGE)
+
+struct bcm_qspi_parms {
+	u32 speed_hz;
+	u8 mode;
+	u8 bits_per_word;
+};
+
+struct bcm_xfer_mode {
+	bool flex_mode;
+	unsigned int width;
+	unsigned int addrlen;
+	unsigned int hp;
+};
+
+enum base_type {
+	MSPI,
+	BSPI,
+	CHIP_SELECT,
+	BASEMAX,
+};
+
+enum irq_source {
+	SINGLE_L2,
+	MUXED_L1,
+};
+
+struct bcm_qspi_irq {
+	const char *irq_name;
+	const irq_handler_t irq_handler;
+	int irq_source;
+	u32 mask;
+};
+
+struct bcm_qspi_dev_id {
+	const struct bcm_qspi_irq *irqp;
+	void *dev;
+};
+
+
+struct qspi_trans {
+	struct spi_transfer *trans;
+	int byte;
+	bool mspi_last_trans;
+};
+
+struct bcm_qspi {
+	struct platform_device *pdev;
+	struct spi_master *master;
+	struct clk *clk;
+	u32 base_clk;
+	u32 max_speed_hz;
+	void __iomem *base[BASEMAX];
+
+	/* Some SoCs provide custom interrupt status register(s) */
+	struct bcm_qspi_soc_intc	*soc_intc;
+
+	struct bcm_qspi_parms last_parms;
+	struct qspi_trans  trans_pos;
+	int curr_cs;
+	int bspi_maj_rev;
+	int bspi_min_rev;
+	int bspi_enabled;
+	const struct spi_mem_op *bspi_rf_op;
+	u32 bspi_rf_op_idx;
+	u32 bspi_rf_op_len;
+	u32 bspi_rf_op_status;
+	struct bcm_xfer_mode xfer_mode;
+	u32 s3_strap_override_ctrl;
+	bool bspi_mode;
+	bool big_endian;
+	int num_irqs;
+	struct bcm_qspi_dev_id *dev_ids;
+	struct completion mspi_done;
+	struct completion bspi_done;
+};
+
+static inline bool has_bspi(struct bcm_qspi *qspi)
+{
+	return qspi->bspi_mode;
+}
+
+/* Read qspi controller register*/
+static inline u32 bcm_qspi_read(struct bcm_qspi *qspi, enum base_type type,
+				unsigned int offset)
+{
+	return bcm_qspi_readl(qspi->big_endian, qspi->base[type] + offset);
+}
+
+/* Write qspi controller register*/
+static inline void bcm_qspi_write(struct bcm_qspi *qspi, enum base_type type,
+				  unsigned int offset, unsigned int data)
+{
+	bcm_qspi_writel(qspi->big_endian, data, qspi->base[type] + offset);
+}
+
+/* BSPI helpers */
+static int bcm_qspi_bspi_busy_poll(struct bcm_qspi *qspi)
+{
+	int i;
+
+	/* this should normally finish within 10us */
+	for (i = 0; i < 1000; i++) {
+		if (!(bcm_qspi_read(qspi, BSPI, BSPI_BUSY_STATUS) & 1))
+			return 0;
+		udelay(1);
+	}
+	dev_warn(&qspi->pdev->dev, "timeout waiting for !busy_status\n");
+	return -EIO;
+}
+
+static inline bool bcm_qspi_bspi_ver_three(struct bcm_qspi *qspi)
+{
+	if (qspi->bspi_maj_rev < 4)
+		return true;
+	return false;
+}
+
+static void bcm_qspi_bspi_flush_prefetch_buffers(struct bcm_qspi *qspi)
+{
+	bcm_qspi_bspi_busy_poll(qspi);
+	/* Force rising edge for the b0/b1 'flush' field */
+	bcm_qspi_write(qspi, BSPI, BSPI_B0_CTRL, 1);
+	bcm_qspi_write(qspi, BSPI, BSPI_B1_CTRL, 1);
+	bcm_qspi_write(qspi, BSPI, BSPI_B0_CTRL, 0);
+	bcm_qspi_write(qspi, BSPI, BSPI_B1_CTRL, 0);
+}
+
+static int bcm_qspi_bspi_lr_is_fifo_empty(struct bcm_qspi *qspi)
+{
+	return (bcm_qspi_read(qspi, BSPI, BSPI_RAF_STATUS) &
+				BSPI_RAF_STATUS_FIFO_EMPTY_MASK);
+}
+
+static inline u32 bcm_qspi_bspi_lr_read_fifo(struct bcm_qspi *qspi)
+{
+	u32 data = bcm_qspi_read(qspi, BSPI, BSPI_RAF_READ_DATA);
+
+	/* BSPI v3 LR is LE only, convert data to host endianness */
+	if (bcm_qspi_bspi_ver_three(qspi))
+		data = le32_to_cpu(data);
+
+	return data;
+}
+
+static inline void bcm_qspi_bspi_lr_start(struct bcm_qspi *qspi)
+{
+	bcm_qspi_bspi_busy_poll(qspi);
+	bcm_qspi_write(qspi, BSPI, BSPI_RAF_CTRL,
+		       BSPI_RAF_CTRL_START_MASK);
+}
+
+static inline void bcm_qspi_bspi_lr_clear(struct bcm_qspi *qspi)
+{
+	bcm_qspi_write(qspi, BSPI, BSPI_RAF_CTRL,
+		       BSPI_RAF_CTRL_CLEAR_MASK);
+	bcm_qspi_bspi_flush_prefetch_buffers(qspi);
+}
+
+static void bcm_qspi_bspi_lr_data_read(struct bcm_qspi *qspi)
+{
+	u32 *buf = (u32 *)qspi->bspi_rf_op->data.buf.in;
+	u32 data = 0;
+
+	dev_dbg(&qspi->pdev->dev, "xfer %p rx %p rxlen %d\n", qspi->bspi_rf_op,
+		qspi->bspi_rf_op->data.buf.in, qspi->bspi_rf_op_len);
+	while (!bcm_qspi_bspi_lr_is_fifo_empty(qspi)) {
+		data = bcm_qspi_bspi_lr_read_fifo(qspi);
+		if (likely(qspi->bspi_rf_op_len >= 4) &&
+		    IS_ALIGNED((uintptr_t)buf, 4)) {
+			buf[qspi->bspi_rf_op_idx++] = data;
+			qspi->bspi_rf_op_len -= 4;
+		} else {
+			/* Read out remaining bytes, make sure*/
+			u8 *cbuf = (u8 *)&buf[qspi->bspi_rf_op_idx];
+
+			data = cpu_to_le32(data);
+			while (qspi->bspi_rf_op_len) {
+				*cbuf++ = (u8)data;
+				data >>= 8;
+				qspi->bspi_rf_op_len--;
+			}
+		}
+	}
+}
+
+static void bcm_qspi_bspi_set_xfer_params(struct bcm_qspi *qspi, u8 cmd_byte,
+					  int bpp, int bpc, int flex_mode)
+{
+	bcm_qspi_write(qspi, BSPI, BSPI_FLEX_MODE_ENABLE, 0);
+	bcm_qspi_write(qspi, BSPI, BSPI_BITS_PER_CYCLE, bpc);
+	bcm_qspi_write(qspi, BSPI, BSPI_BITS_PER_PHASE, bpp);
+	bcm_qspi_write(qspi, BSPI, BSPI_CMD_AND_MODE_BYTE, cmd_byte);
+	bcm_qspi_write(qspi, BSPI, BSPI_FLEX_MODE_ENABLE, flex_mode);
+}
+
+static int bcm_qspi_bspi_set_flex_mode(struct bcm_qspi *qspi,
+				       const struct spi_mem_op *op, int hp)
+{
+	int bpc = 0, bpp = 0;
+	u8 command = op->cmd.opcode;
+	int width  = op->cmd.buswidth ? op->cmd.buswidth : SPI_NBITS_SINGLE;
+	int addrlen = op->addr.nbytes;
+	int flex_mode = 1;
+
+	dev_dbg(&qspi->pdev->dev, "set flex mode w %x addrlen %x hp %d\n",
+		width, addrlen, hp);
+
+	if (addrlen == BSPI_ADDRLEN_4BYTES)
+		bpp = BSPI_BPP_ADDR_SELECT_MASK;
+
+	bpp |= (op->dummy.nbytes * 8) / op->dummy.buswidth;
+
+	switch (width) {
+	case SPI_NBITS_SINGLE:
+		if (addrlen == BSPI_ADDRLEN_3BYTES)
+			/* default mode, does not need flex_cmd */
+			flex_mode = 0;
+		break;
+	case SPI_NBITS_DUAL:
+		bpc = 0x00000001;
+		if (hp) {
+			bpc |= 0x00010100; /* address and mode are 2-bit */
+			bpp = BSPI_BPP_MODE_SELECT_MASK;
+		}
+		break;
+	case SPI_NBITS_QUAD:
+		bpc = 0x00000002;
+		if (hp) {
+			bpc |= 0x00020200; /* address and mode are 4-bit */
+			bpp |= BSPI_BPP_MODE_SELECT_MASK;
+		}
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	bcm_qspi_bspi_set_xfer_params(qspi, command, bpp, bpc, flex_mode);
+
+	return 0;
+}
+
+static int bcm_qspi_bspi_set_override(struct bcm_qspi *qspi,
+				      const struct spi_mem_op *op, int hp)
+{
+	int width = op->data.buswidth ? op->data.buswidth : SPI_NBITS_SINGLE;
+	int addrlen = op->addr.nbytes;
+	u32 data = bcm_qspi_read(qspi, BSPI, BSPI_STRAP_OVERRIDE_CTRL);
+
+	dev_dbg(&qspi->pdev->dev, "set override mode w %x addrlen %x hp %d\n",
+		width, addrlen, hp);
+
+	switch (width) {
+	case SPI_NBITS_SINGLE:
+		/* clear quad/dual mode */
+		data &= ~(BSPI_STRAP_OVERRIDE_CTRL_DATA_QUAD |
+			  BSPI_STRAP_OVERRIDE_CTRL_DATA_DUAL);
+		break;
+	case SPI_NBITS_QUAD:
+		/* clear dual mode and set quad mode */
+		data &= ~BSPI_STRAP_OVERRIDE_CTRL_DATA_DUAL;
+		data |= BSPI_STRAP_OVERRIDE_CTRL_DATA_QUAD;
+		break;
+	case SPI_NBITS_DUAL:
+		/* clear quad mode set dual mode */
+		data &= ~BSPI_STRAP_OVERRIDE_CTRL_DATA_QUAD;
+		data |= BSPI_STRAP_OVERRIDE_CTRL_DATA_DUAL;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (addrlen == BSPI_ADDRLEN_4BYTES)
+		/* set 4byte mode*/
+		data |= BSPI_STRAP_OVERRIDE_CTRL_ADDR_4BYTE;
+	else
+		/* clear 4 byte mode */
+		data &= ~BSPI_STRAP_OVERRIDE_CTRL_ADDR_4BYTE;
+
+	/* set the override mode */
+	data |=	BSPI_STRAP_OVERRIDE_CTRL_OVERRIDE;
+	bcm_qspi_write(qspi, BSPI, BSPI_STRAP_OVERRIDE_CTRL, data);
+	bcm_qspi_bspi_set_xfer_params(qspi, op->cmd.opcode, 0, 0, 0);
+
+	return 0;
+}
+
+static int bcm_qspi_bspi_set_mode(struct bcm_qspi *qspi,
+				  const struct spi_mem_op *op, int hp)
+{
+	int error = 0;
+	int width = op->data.buswidth ? op->data.buswidth : SPI_NBITS_SINGLE;
+	int addrlen = op->addr.nbytes;
+
+	/* default mode */
+	qspi->xfer_mode.flex_mode = true;
+
+	if (!bcm_qspi_bspi_ver_three(qspi)) {
+		u32 val, mask;
+
+		val = bcm_qspi_read(qspi, BSPI, BSPI_STRAP_OVERRIDE_CTRL);
+		mask = BSPI_STRAP_OVERRIDE_CTRL_OVERRIDE;
+		if (val & mask || qspi->s3_strap_override_ctrl & mask) {
+			qspi->xfer_mode.flex_mode = false;
+			bcm_qspi_write(qspi, BSPI, BSPI_FLEX_MODE_ENABLE, 0);
+			error = bcm_qspi_bspi_set_override(qspi, op, hp);
+		}
+	}
+
+	if (qspi->xfer_mode.flex_mode)
+		error = bcm_qspi_bspi_set_flex_mode(qspi, op, hp);
+
+	if (error) {
+		dev_warn(&qspi->pdev->dev,
+			 "INVALID COMBINATION: width=%d addrlen=%d hp=%d\n",
+			 width, addrlen, hp);
+	} else if (qspi->xfer_mode.width != width ||
+		   qspi->xfer_mode.addrlen != addrlen ||
+		   qspi->xfer_mode.hp != hp) {
+		qspi->xfer_mode.width = width;
+		qspi->xfer_mode.addrlen = addrlen;
+		qspi->xfer_mode.hp = hp;
+		dev_dbg(&qspi->pdev->dev,
+			"cs:%d %d-lane output, %d-byte address%s\n",
+			qspi->curr_cs,
+			qspi->xfer_mode.width,
+			qspi->xfer_mode.addrlen,
+			qspi->xfer_mode.hp != -1 ? ", hp mode" : "");
+	}
+
+	return error;
+}
+
+static void bcm_qspi_enable_bspi(struct bcm_qspi *qspi)
+{
+	if (!has_bspi(qspi))
+		return;
+
+	qspi->bspi_enabled = 1;
+	if ((bcm_qspi_read(qspi, BSPI, BSPI_MAST_N_BOOT_CTRL) & 1) == 0)
+		return;
+
+	bcm_qspi_bspi_flush_prefetch_buffers(qspi);
+	udelay(1);
+	bcm_qspi_write(qspi, BSPI, BSPI_MAST_N_BOOT_CTRL, 0);
+	udelay(1);
+}
+
+static void bcm_qspi_disable_bspi(struct bcm_qspi *qspi)
+{
+	if (!has_bspi(qspi))
+		return;
+
+	qspi->bspi_enabled = 0;
+	if ((bcm_qspi_read(qspi, BSPI, BSPI_MAST_N_BOOT_CTRL) & 1))
+		return;
+
+	bcm_qspi_bspi_busy_poll(qspi);
+	bcm_qspi_write(qspi, BSPI, BSPI_MAST_N_BOOT_CTRL, 1);
+	udelay(1);
+}
+
+static void bcm_qspi_chip_select(struct bcm_qspi *qspi, int cs)
+{
+	u32 rd = 0;
+	u32 wr = 0;
+
+	if (qspi->base[CHIP_SELECT]) {
+		rd = bcm_qspi_read(qspi, CHIP_SELECT, 0);
+		wr = (rd & ~0xff) | (1 << cs);
+		if (rd == wr)
+			return;
+		bcm_qspi_write(qspi, CHIP_SELECT, 0, wr);
+		usleep_range(10, 20);
+	}
+
+	dev_dbg(&qspi->pdev->dev, "using cs:%d\n", cs);
+	qspi->curr_cs = cs;
+}
+
+/* MSPI helpers */
+static void bcm_qspi_hw_set_parms(struct bcm_qspi *qspi,
+				  const struct bcm_qspi_parms *xp)
+{
+	u32 spcr, spbr = 0;
+
+	if (xp->speed_hz)
+		spbr = qspi->base_clk / (2 * xp->speed_hz);
+
+	spcr = clamp_val(spbr, QSPI_SPBR_MIN, QSPI_SPBR_MAX);
+	bcm_qspi_write(qspi, MSPI, MSPI_SPCR0_LSB, spcr);
+
+	spcr = MSPI_MASTER_BIT;
+	/* for 16 bit the data should be zero */
+	if (xp->bits_per_word != 16)
+		spcr |= xp->bits_per_word << 2;
+	spcr |= xp->mode & 3;
+	bcm_qspi_write(qspi, MSPI, MSPI_SPCR0_MSB, spcr);
+
+	qspi->last_parms = *xp;
+}
+
+static void bcm_qspi_update_parms(struct bcm_qspi *qspi,
+				  struct spi_device *spi,
+				  struct spi_transfer *trans)
+{
+	struct bcm_qspi_parms xp;
+
+	xp.speed_hz = trans->speed_hz;
+	xp.bits_per_word = trans->bits_per_word;
+	xp.mode = spi->mode;
+
+	bcm_qspi_hw_set_parms(qspi, &xp);
+}
+
+static int bcm_qspi_setup(struct spi_device *spi)
+{
+	struct bcm_qspi_parms *xp;
+
+	if (spi->bits_per_word > 16)
+		return -EINVAL;
+
+	xp = spi_get_ctldata(spi);
+	if (!xp) {
+		xp = kzalloc(sizeof(*xp), GFP_KERNEL);
+		if (!xp)
+			return -ENOMEM;
+		spi_set_ctldata(spi, xp);
+	}
+	xp->speed_hz = spi->max_speed_hz;
+	xp->mode = spi->mode;
+
+	if (spi->bits_per_word)
+		xp->bits_per_word = spi->bits_per_word;
+	else
+		xp->bits_per_word = 8;
+
+	return 0;
+}
+
+static bool bcm_qspi_mspi_transfer_is_last(struct bcm_qspi *qspi,
+					   struct qspi_trans *qt)
+{
+	if (qt->mspi_last_trans &&
+	    spi_transfer_is_last(qspi->master, qt->trans))
+		return true;
+	else
+		return false;
+}
+
+static int update_qspi_trans_byte_count(struct bcm_qspi *qspi,
+					struct qspi_trans *qt, int flags)
+{
+	int ret = TRANS_STATUS_BREAK_NONE;
+
+	/* count the last transferred bytes */
+	if (qt->trans->bits_per_word <= 8)
+		qt->byte++;
+	else
+		qt->byte += 2;
+
+	if (qt->byte >= qt->trans->len) {
+		/* we're at the end of the spi_transfer */
+		/* in TX mode, need to pause for a delay or CS change */
+		if (qt->trans->delay_usecs &&
+		    (flags & TRANS_STATUS_BREAK_DELAY))
+			ret |= TRANS_STATUS_BREAK_DELAY;
+		if (qt->trans->cs_change &&
+		    (flags & TRANS_STATUS_BREAK_CS_CHANGE))
+			ret |= TRANS_STATUS_BREAK_CS_CHANGE;
+		if (ret)
+			goto done;
+
+		dev_dbg(&qspi->pdev->dev, "advance msg exit\n");
+		if (bcm_qspi_mspi_transfer_is_last(qspi, qt))
+			ret = TRANS_STATUS_BREAK_EOM;
+		else
+			ret = TRANS_STATUS_BREAK_NO_BYTES;
+
+		qt->trans = NULL;
+	}
+
+done:
+	dev_dbg(&qspi->pdev->dev, "trans %p len %d byte %d ret %x\n",
+		qt->trans, qt->trans ? qt->trans->len : 0, qt->byte, ret);
+	return ret;
+}
+
+static inline u8 read_rxram_slot_u8(struct bcm_qspi *qspi, int slot)
+{
+	u32 slot_offset = MSPI_RXRAM + (slot << 3) + 0x4;
+
+	/* mask out reserved bits */
+	return bcm_qspi_read(qspi, MSPI, slot_offset) & 0xff;
+}
+
+static inline u16 read_rxram_slot_u16(struct bcm_qspi *qspi, int slot)
+{
+	u32 reg_offset = MSPI_RXRAM;
+	u32 lsb_offset = reg_offset + (slot << 3) + 0x4;
+	u32 msb_offset = reg_offset + (slot << 3);
+
+	return (bcm_qspi_read(qspi, MSPI, lsb_offset) & 0xff) |
+		((bcm_qspi_read(qspi, MSPI, msb_offset) & 0xff) << 8);
+}
+
+static void read_from_hw(struct bcm_qspi *qspi, int slots)
+{
+	struct qspi_trans tp;
+	int slot;
+
+	bcm_qspi_disable_bspi(qspi);
+
+	if (slots > MSPI_NUM_CDRAM) {
+		/* should never happen */
+		dev_err(&qspi->pdev->dev, "%s: too many slots!\n", __func__);
+		return;
+	}
+
+	tp = qspi->trans_pos;
+
+	for (slot = 0; slot < slots; slot++) {
+		if (tp.trans->bits_per_word <= 8) {
+			u8 *buf = tp.trans->rx_buf;
+
+			if (buf)
+				buf[tp.byte] = read_rxram_slot_u8(qspi, slot);
+			dev_dbg(&qspi->pdev->dev, "RD %02x\n",
+				buf ? buf[tp.byte] : 0xff);
+		} else {
+			u16 *buf = tp.trans->rx_buf;
+
+			if (buf)
+				buf[tp.byte / 2] = read_rxram_slot_u16(qspi,
+								      slot);
+			dev_dbg(&qspi->pdev->dev, "RD %04x\n",
+				buf ? buf[tp.byte] : 0xffff);
+		}
+
+		update_qspi_trans_byte_count(qspi, &tp,
+					     TRANS_STATUS_BREAK_NONE);
+	}
+
+	qspi->trans_pos = tp;
+}
+
+static inline void write_txram_slot_u8(struct bcm_qspi *qspi, int slot,
+				       u8 val)
+{
+	u32 reg_offset = MSPI_TXRAM + (slot << 3);
+
+	/* mask out reserved bits */
+	bcm_qspi_write(qspi, MSPI, reg_offset, val);
+}
+
+static inline void write_txram_slot_u16(struct bcm_qspi *qspi, int slot,
+					u16 val)
+{
+	u32 reg_offset = MSPI_TXRAM;
+	u32 msb_offset = reg_offset + (slot << 3);
+	u32 lsb_offset = reg_offset + (slot << 3) + 0x4;
+
+	bcm_qspi_write(qspi, MSPI, msb_offset, (val >> 8));
+	bcm_qspi_write(qspi, MSPI, lsb_offset, (val & 0xff));
+}
+
+static inline u32 read_cdram_slot(struct bcm_qspi *qspi, int slot)
+{
+	return bcm_qspi_read(qspi, MSPI, MSPI_CDRAM + (slot << 2));
+}
+
+static inline void write_cdram_slot(struct bcm_qspi *qspi, int slot, u32 val)
+{
+	bcm_qspi_write(qspi, MSPI, (MSPI_CDRAM + (slot << 2)), val);
+}
+
+/* Return number of slots written */
+static int write_to_hw(struct bcm_qspi *qspi, struct spi_device *spi)
+{
+	struct qspi_trans tp;
+	int slot = 0, tstatus = 0;
+	u32 mspi_cdram = 0;
+
+	bcm_qspi_disable_bspi(qspi);
+	tp = qspi->trans_pos;
+	bcm_qspi_update_parms(qspi, spi, tp.trans);
+
+	/* Run until end of transfer or reached the max data */
+	while (!tstatus && slot < MSPI_NUM_CDRAM) {
+		if (tp.trans->bits_per_word <= 8) {
+			const u8 *buf = tp.trans->tx_buf;
+			u8 val = buf ? buf[tp.byte] : 0xff;
+
+			write_txram_slot_u8(qspi, slot, val);
+			dev_dbg(&qspi->pdev->dev, "WR %02x\n", val);
+		} else {
+			const u16 *buf = tp.trans->tx_buf;
+			u16 val = buf ? buf[tp.byte / 2] : 0xffff;
+
+			write_txram_slot_u16(qspi, slot, val);
+			dev_dbg(&qspi->pdev->dev, "WR %04x\n", val);
+		}
+		mspi_cdram = MSPI_CDRAM_CONT_BIT;
+
+		if (has_bspi(qspi))
+			mspi_cdram &= ~1;
+		else
+			mspi_cdram |= (~(1 << spi->chip_select) &
+				       MSPI_CDRAM_PCS);
+
+		mspi_cdram |= ((tp.trans->bits_per_word <= 8) ? 0 :
+				MSPI_CDRAM_BITSE_BIT);
+
+		write_cdram_slot(qspi, slot, mspi_cdram);
+
+		tstatus = update_qspi_trans_byte_count(qspi, &tp,
+						       TRANS_STATUS_BREAK_TX);
+		slot++;
+	}
+
+	if (!slot) {
+		dev_err(&qspi->pdev->dev, "%s: no data to send?", __func__);
+		goto done;
+	}
+
+	dev_dbg(&qspi->pdev->dev, "submitting %d slots\n", slot);
+	bcm_qspi_write(qspi, MSPI, MSPI_NEWQP, 0);
+	bcm_qspi_write(qspi, MSPI, MSPI_ENDQP, slot - 1);
+
+	if (tstatus & TRANS_STATUS_BREAK_DESELECT) {
+		mspi_cdram = read_cdram_slot(qspi, slot - 1) &
+			~MSPI_CDRAM_CONT_BIT;
+		write_cdram_slot(qspi, slot - 1, mspi_cdram);
+	}
+
+	if (has_bspi(qspi))
+		bcm_qspi_write(qspi, MSPI, MSPI_WRITE_LOCK, 1);
+
+	/* Must flush previous writes before starting MSPI operation */
+	mb();
+	/* Set cont | spe | spifie */
+	bcm_qspi_write(qspi, MSPI, MSPI_SPCR2, 0xe0);
+
+done:
+	return slot;
+}
+
+static int bcm_qspi_bspi_exec_mem_op(struct spi_device *spi,
+				     const struct spi_mem_op *op)
+{
+	struct bcm_qspi *qspi = spi_master_get_devdata(spi->master);
+	u32 addr = 0, len, rdlen, len_words, from = 0;
+	int ret = 0;
+	unsigned long timeo = msecs_to_jiffies(100);
+	struct bcm_qspi_soc_intc *soc_intc = qspi->soc_intc;
+
+	if (bcm_qspi_bspi_ver_three(qspi))
+		if (op->addr.nbytes == BSPI_ADDRLEN_4BYTES)
+			return -EIO;
+
+	from = op->addr.val;
+	bcm_qspi_chip_select(qspi, spi->chip_select);
+	bcm_qspi_write(qspi, MSPI, MSPI_WRITE_LOCK, 0);
+
+	/*
+	 * when using flex mode we need to send
+	 * the upper address byte to bspi
+	 */
+	if (bcm_qspi_bspi_ver_three(qspi) == false) {
+		addr = from & 0xff000000;
+		bcm_qspi_write(qspi, BSPI,
+			       BSPI_BSPI_FLASH_UPPER_ADDR_BYTE, addr);
+	}
+
+	if (!qspi->xfer_mode.flex_mode)
+		addr = from;
+	else
+		addr = from & 0x00ffffff;
+
+	if (bcm_qspi_bspi_ver_three(qspi) == true)
+		addr = (addr + 0xc00000) & 0xffffff;
+
+	/*
+	 * read into the entire buffer by breaking the reads
+	 * into RAF buffer read lengths
+	 */
+	len = op->data.nbytes;
+	qspi->bspi_rf_op_idx = 0;
+
+	do {
+		if (len > BSPI_READ_LENGTH)
+			rdlen = BSPI_READ_LENGTH;
+		else
+			rdlen = len;
+
+		reinit_completion(&qspi->bspi_done);
+		bcm_qspi_enable_bspi(qspi);
+		len_words = (rdlen + 3) >> 2;
+		qspi->bspi_rf_op = op;
+		qspi->bspi_rf_op_status = 0;
+		qspi->bspi_rf_op_len = rdlen;
+		dev_dbg(&qspi->pdev->dev,
+			"bspi xfr addr 0x%x len 0x%x", addr, rdlen);
+		bcm_qspi_write(qspi, BSPI, BSPI_RAF_START_ADDR, addr);
+		bcm_qspi_write(qspi, BSPI, BSPI_RAF_NUM_WORDS, len_words);
+		bcm_qspi_write(qspi, BSPI, BSPI_RAF_WATERMARK, 0);
+		if (qspi->soc_intc) {
+			/*
+			 * clear soc MSPI and BSPI interrupts and enable
+			 * BSPI interrupts.
+			 */
+			soc_intc->bcm_qspi_int_ack(soc_intc, MSPI_BSPI_DONE);
+			soc_intc->bcm_qspi_int_set(soc_intc, BSPI_DONE, true);
+		}
+
+		/* Must flush previous writes before starting BSPI operation */
+		mb();
+		bcm_qspi_bspi_lr_start(qspi);
+		if (!wait_for_completion_timeout(&qspi->bspi_done, timeo)) {
+			dev_err(&qspi->pdev->dev, "timeout waiting for BSPI\n");
+			ret = -ETIMEDOUT;
+			break;
+		}
+
+		/* set msg return length */
+		addr += rdlen;
+		len -= rdlen;
+	} while (len);
+
+	return ret;
+}
+
+static int bcm_qspi_transfer_one(struct spi_master *master,
+				 struct spi_device *spi,
+				 struct spi_transfer *trans)
+{
+	struct bcm_qspi *qspi = spi_master_get_devdata(master);
+	int slots;
+	unsigned long timeo = msecs_to_jiffies(100);
+
+	bcm_qspi_chip_select(qspi, spi->chip_select);
+	qspi->trans_pos.trans = trans;
+	qspi->trans_pos.byte = 0;
+
+	while (qspi->trans_pos.byte < trans->len) {
+		reinit_completion(&qspi->mspi_done);
+
+		slots = write_to_hw(qspi, spi);
+		if (!wait_for_completion_timeout(&qspi->mspi_done, timeo)) {
+			dev_err(&qspi->pdev->dev, "timeout waiting for MSPI\n");
+			return -ETIMEDOUT;
+		}
+
+		read_from_hw(qspi, slots);
+	}
+
+	return 0;
+}
+
+static int bcm_qspi_mspi_exec_mem_op(struct spi_device *spi,
+				     const struct spi_mem_op *op)
+{
+	struct spi_master *master = spi->master;
+	struct bcm_qspi *qspi = spi_master_get_devdata(master);
+	struct spi_transfer t[2];
+	u8 cmd[6] = { };
+	int ret, i;
+
+	memset(cmd, 0, sizeof(cmd));
+	memset(t, 0, sizeof(t));
+
+	/* tx */
+	/* opcode is in cmd[0] */
+	cmd[0] = op->cmd.opcode;
+	for (i = 0; i < op->addr.nbytes; i++)
+		cmd[1 + i] = op->addr.val >> (8 * (op->addr.nbytes - i - 1));
+
+	t[0].tx_buf = cmd;
+	t[0].len = op->addr.nbytes + op->dummy.nbytes + 1;
+	t[0].bits_per_word = spi->bits_per_word;
+	t[0].tx_nbits = op->cmd.buswidth;
+	/* lets mspi know that this is not last transfer */
+	qspi->trans_pos.mspi_last_trans = false;
+	ret = bcm_qspi_transfer_one(master, spi, &t[0]);
+
+	/* rx */
+	qspi->trans_pos.mspi_last_trans = true;
+	if (!ret) {
+		/* rx */
+		t[1].rx_buf = op->data.buf.in;
+		t[1].len = op->data.nbytes;
+		t[1].rx_nbits =  op->data.buswidth;
+		t[1].bits_per_word = spi->bits_per_word;
+		ret = bcm_qspi_transfer_one(master, spi, &t[1]);
+	}
+
+	return ret;
+}
+
+static int bcm_qspi_exec_mem_op(struct spi_mem *mem,
+				const struct spi_mem_op *op)
+{
+	struct spi_device *spi = mem->spi;
+	struct bcm_qspi *qspi = spi_master_get_devdata(spi->master);
+	int ret = 0;
+	bool mspi_read = false;
+	u32 addr = 0, len;
+	u_char *buf;
+
+	if (!op->data.nbytes || !op->addr.nbytes || op->addr.nbytes > 4 ||
+	    op->data.dir != SPI_MEM_DATA_IN)
+		return -ENOTSUPP;
+
+	buf = op->data.buf.in;
+	addr = op->addr.val;
+	len = op->data.nbytes;
+
+	if (bcm_qspi_bspi_ver_three(qspi) == true) {
+		/*
+		 * The address coming into this function is a raw flash offset.
+		 * But for BSPI <= V3, we need to convert it to a remapped BSPI
+		 * address. If it crosses a 4MB boundary, just revert back to
+		 * using MSPI.
+		 */
+		addr = (addr + 0xc00000) & 0xffffff;
+
+		if ((~ADDR_4MB_MASK & addr) ^
+		    (~ADDR_4MB_MASK & (addr + len - 1)))
+			mspi_read = true;
+	}
+
+	/* non-aligned and very short transfers are handled by MSPI */
+	if (!IS_ALIGNED((uintptr_t)addr, 4) || !IS_ALIGNED((uintptr_t)buf, 4) ||
+	    len < 4)
+		mspi_read = true;
+
+	if (mspi_read)
+		return bcm_qspi_mspi_exec_mem_op(spi, op);
+
+	ret = bcm_qspi_bspi_set_mode(qspi, op, -1);
+
+	if (!ret)
+		ret = bcm_qspi_bspi_exec_mem_op(spi, op);
+
+	return ret;
+}
+
+static void bcm_qspi_cleanup(struct spi_device *spi)
+{
+	struct bcm_qspi_parms *xp = spi_get_ctldata(spi);
+
+	kfree(xp);
+}
+
+static irqreturn_t bcm_qspi_mspi_l2_isr(int irq, void *dev_id)
+{
+	struct bcm_qspi_dev_id *qspi_dev_id = dev_id;
+	struct bcm_qspi *qspi = qspi_dev_id->dev;
+	u32 status = bcm_qspi_read(qspi, MSPI, MSPI_MSPI_STATUS);
+
+	if (status & MSPI_MSPI_STATUS_SPIF) {
+		struct bcm_qspi_soc_intc *soc_intc = qspi->soc_intc;
+		/* clear interrupt */
+		status &= ~MSPI_MSPI_STATUS_SPIF;
+		bcm_qspi_write(qspi, MSPI, MSPI_MSPI_STATUS, status);
+		if (qspi->soc_intc)
+			soc_intc->bcm_qspi_int_ack(soc_intc, MSPI_DONE);
+		complete(&qspi->mspi_done);
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_NONE;
+}
+
+static irqreturn_t bcm_qspi_bspi_lr_l2_isr(int irq, void *dev_id)
+{
+	struct bcm_qspi_dev_id *qspi_dev_id = dev_id;
+	struct bcm_qspi *qspi = qspi_dev_id->dev;
+	struct bcm_qspi_soc_intc *soc_intc = qspi->soc_intc;
+	u32 status = qspi_dev_id->irqp->mask;
+
+	if (qspi->bspi_enabled && qspi->bspi_rf_op) {
+		bcm_qspi_bspi_lr_data_read(qspi);
+		if (qspi->bspi_rf_op_len == 0) {
+			qspi->bspi_rf_op = NULL;
+			if (qspi->soc_intc) {
+				/* disable soc BSPI interrupt */
+				soc_intc->bcm_qspi_int_set(soc_intc, BSPI_DONE,
+							   false);
+				/* indicate done */
+				status = INTR_BSPI_LR_SESSION_DONE_MASK;
+			}
+
+			if (qspi->bspi_rf_op_status)
+				bcm_qspi_bspi_lr_clear(qspi);
+			else
+				bcm_qspi_bspi_flush_prefetch_buffers(qspi);
+		}
+
+		if (qspi->soc_intc)
+			/* clear soc BSPI interrupt */
+			soc_intc->bcm_qspi_int_ack(soc_intc, BSPI_DONE);
+	}
+
+	status &= INTR_BSPI_LR_SESSION_DONE_MASK;
+	if (qspi->bspi_enabled && status && qspi->bspi_rf_op_len == 0)
+		complete(&qspi->bspi_done);
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t bcm_qspi_bspi_lr_err_l2_isr(int irq, void *dev_id)
+{
+	struct bcm_qspi_dev_id *qspi_dev_id = dev_id;
+	struct bcm_qspi *qspi = qspi_dev_id->dev;
+	struct bcm_qspi_soc_intc *soc_intc = qspi->soc_intc;
+
+	dev_err(&qspi->pdev->dev, "BSPI INT error\n");
+	qspi->bspi_rf_op_status = -EIO;
+	if (qspi->soc_intc)
+		/* clear soc interrupt */
+		soc_intc->bcm_qspi_int_ack(soc_intc, BSPI_ERR);
+
+	complete(&qspi->bspi_done);
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t bcm_qspi_l1_isr(int irq, void *dev_id)
+{
+	struct bcm_qspi_dev_id *qspi_dev_id = dev_id;
+	struct bcm_qspi *qspi = qspi_dev_id->dev;
+	struct bcm_qspi_soc_intc *soc_intc = qspi->soc_intc;
+	irqreturn_t ret = IRQ_NONE;
+
+	if (soc_intc) {
+		u32 status = soc_intc->bcm_qspi_get_int_status(soc_intc);
+
+		if (status & MSPI_DONE)
+			ret = bcm_qspi_mspi_l2_isr(irq, dev_id);
+		else if (status & BSPI_DONE)
+			ret = bcm_qspi_bspi_lr_l2_isr(irq, dev_id);
+		else if (status & BSPI_ERR)
+			ret = bcm_qspi_bspi_lr_err_l2_isr(irq, dev_id);
+	}
+
+	return ret;
+}
+
+static const struct bcm_qspi_irq qspi_irq_tab[] = {
+	{
+		.irq_name = "spi_lr_fullness_reached",
+		.irq_handler = bcm_qspi_bspi_lr_l2_isr,
+		.mask = INTR_BSPI_LR_FULLNESS_REACHED_MASK,
+	},
+	{
+		.irq_name = "spi_lr_session_aborted",
+		.irq_handler = bcm_qspi_bspi_lr_err_l2_isr,
+		.mask = INTR_BSPI_LR_SESSION_ABORTED_MASK,
+	},
+	{
+		.irq_name = "spi_lr_impatient",
+		.irq_handler = bcm_qspi_bspi_lr_err_l2_isr,
+		.mask = INTR_BSPI_LR_IMPATIENT_MASK,
+	},
+	{
+		.irq_name = "spi_lr_session_done",
+		.irq_handler = bcm_qspi_bspi_lr_l2_isr,
+		.mask = INTR_BSPI_LR_SESSION_DONE_MASK,
+	},
+#ifdef QSPI_INT_DEBUG
+	/* this interrupt is for debug purposes only, dont request irq */
+	{
+		.irq_name = "spi_lr_overread",
+		.irq_handler = bcm_qspi_bspi_lr_err_l2_isr,
+		.mask = INTR_BSPI_LR_OVERREAD_MASK,
+	},
+#endif
+	{
+		.irq_name = "mspi_done",
+		.irq_handler = bcm_qspi_mspi_l2_isr,
+		.mask = INTR_MSPI_DONE_MASK,
+	},
+	{
+		.irq_name = "mspi_halted",
+		.irq_handler = bcm_qspi_mspi_l2_isr,
+		.mask = INTR_MSPI_HALTED_MASK,
+	},
+	{
+		/* single muxed L1 interrupt source */
+		.irq_name = "spi_l1_intr",
+		.irq_handler = bcm_qspi_l1_isr,
+		.irq_source = MUXED_L1,
+		.mask = QSPI_INTERRUPTS_ALL,
+	},
+};
+
+static void bcm_qspi_bspi_init(struct bcm_qspi *qspi)
+{
+	u32 val = 0;
+
+	val = bcm_qspi_read(qspi, BSPI, BSPI_REVISION_ID);
+	qspi->bspi_maj_rev = (val >> 8) & 0xff;
+	qspi->bspi_min_rev = val & 0xff;
+	if (!(bcm_qspi_bspi_ver_three(qspi))) {
+		/* Force mapping of BSPI address -> flash offset */
+		bcm_qspi_write(qspi, BSPI, BSPI_BSPI_XOR_VALUE, 0);
+		bcm_qspi_write(qspi, BSPI, BSPI_BSPI_XOR_ENABLE, 1);
+	}
+	qspi->bspi_enabled = 1;
+	bcm_qspi_disable_bspi(qspi);
+	bcm_qspi_write(qspi, BSPI, BSPI_B0_CTRL, 0);
+	bcm_qspi_write(qspi, BSPI, BSPI_B1_CTRL, 0);
+}
+
+static void bcm_qspi_hw_init(struct bcm_qspi *qspi)
+{
+	struct bcm_qspi_parms parms;
+
+	bcm_qspi_write(qspi, MSPI, MSPI_SPCR1_LSB, 0);
+	bcm_qspi_write(qspi, MSPI, MSPI_SPCR1_MSB, 0);
+	bcm_qspi_write(qspi, MSPI, MSPI_NEWQP, 0);
+	bcm_qspi_write(qspi, MSPI, MSPI_ENDQP, 0);
+	bcm_qspi_write(qspi, MSPI, MSPI_SPCR2, 0x20);
+
+	parms.mode = SPI_MODE_3;
+	parms.bits_per_word = 8;
+	parms.speed_hz = qspi->max_speed_hz;
+	bcm_qspi_hw_set_parms(qspi, &parms);
+
+	if (has_bspi(qspi))
+		bcm_qspi_bspi_init(qspi);
+}
+
+static void bcm_qspi_hw_uninit(struct bcm_qspi *qspi)
+{
+	bcm_qspi_write(qspi, MSPI, MSPI_SPCR2, 0);
+	if (has_bspi(qspi))
+		bcm_qspi_write(qspi, MSPI, MSPI_WRITE_LOCK, 0);
+
+}
+
+static const struct spi_controller_mem_ops bcm_qspi_mem_ops = {
+	.exec_op = bcm_qspi_exec_mem_op,
+};
+
+static const struct of_device_id bcm_qspi_of_match[] = {
+	{ .compatible = "brcm,spi-bcm-qspi" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, bcm_qspi_of_match);
+
+int bcm_qspi_probe(struct platform_device *pdev,
+		   struct bcm_qspi_soc_intc *soc_intc)
+{
+	struct device *dev = &pdev->dev;
+	struct bcm_qspi *qspi;
+	struct spi_master *master;
+	struct resource *res;
+	int irq, ret = 0, num_ints = 0;
+	u32 val;
+	const char *name = NULL;
+	int num_irqs = ARRAY_SIZE(qspi_irq_tab);
+
+	/* We only support device-tree instantiation */
+	if (!dev->of_node)
+		return -ENODEV;
+
+	if (!of_match_node(bcm_qspi_of_match, dev->of_node))
+		return -ENODEV;
+
+	master = spi_alloc_master(dev, sizeof(struct bcm_qspi));
+	if (!master) {
+		dev_err(dev, "error allocating spi_master\n");
+		return -ENOMEM;
+	}
+
+	qspi = spi_master_get_devdata(master);
+	qspi->pdev = pdev;
+	qspi->trans_pos.trans = NULL;
+	qspi->trans_pos.byte = 0;
+	qspi->trans_pos.mspi_last_trans = true;
+	qspi->master = master;
+
+	master->bus_num = -1;
+	master->mode_bits = SPI_CPHA | SPI_CPOL | SPI_RX_DUAL | SPI_RX_QUAD;
+	master->setup = bcm_qspi_setup;
+	master->transfer_one = bcm_qspi_transfer_one;
+	master->mem_ops = &bcm_qspi_mem_ops;
+	master->cleanup = bcm_qspi_cleanup;
+	master->dev.of_node = dev->of_node;
+	master->num_chipselect = NUM_CHIPSELECT;
+
+	qspi->big_endian = of_device_is_big_endian(dev->of_node);
+
+	if (!of_property_read_u32(dev->of_node, "num-cs", &val))
+		master->num_chipselect = val;
+
+	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "hif_mspi");
+	if (!res)
+		res = platform_get_resource_byname(pdev, IORESOURCE_MEM,
+						   "mspi");
+
+	if (res) {
+		qspi->base[MSPI]  = devm_ioremap_resource(dev, res);
+		if (IS_ERR(qspi->base[MSPI])) {
+			ret = PTR_ERR(qspi->base[MSPI]);
+			goto qspi_resource_err;
+		}
+	} else {
+		goto qspi_resource_err;
+	}
+
+	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "bspi");
+	if (res) {
+		qspi->base[BSPI]  = devm_ioremap_resource(dev, res);
+		if (IS_ERR(qspi->base[BSPI])) {
+			ret = PTR_ERR(qspi->base[BSPI]);
+			goto qspi_resource_err;
+		}
+		qspi->bspi_mode = true;
+	} else {
+		qspi->bspi_mode = false;
+	}
+
+	dev_info(dev, "using %smspi mode\n", qspi->bspi_mode ? "bspi-" : "");
+
+	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "cs_reg");
+	if (res) {
+		qspi->base[CHIP_SELECT]  = devm_ioremap_resource(dev, res);
+		if (IS_ERR(qspi->base[CHIP_SELECT])) {
+			ret = PTR_ERR(qspi->base[CHIP_SELECT]);
+			goto qspi_resource_err;
+		}
+	}
+
+	qspi->dev_ids = kcalloc(num_irqs, sizeof(struct bcm_qspi_dev_id),
+				GFP_KERNEL);
+	if (!qspi->dev_ids) {
+		ret = -ENOMEM;
+		goto qspi_resource_err;
+	}
+
+	for (val = 0; val < num_irqs; val++) {
+		irq = -1;
+		name = qspi_irq_tab[val].irq_name;
+		if (qspi_irq_tab[val].irq_source == SINGLE_L2) {
+			/* get the l2 interrupts */
+			irq = platform_get_irq_byname(pdev, name);
+		} else if (!num_ints && soc_intc) {
+			/* all mspi, bspi intrs muxed to one L1 intr */
+			irq = platform_get_irq(pdev, 0);
+		}
+
+		if (irq  >= 0) {
+			ret = devm_request_irq(&pdev->dev, irq,
+					       qspi_irq_tab[val].irq_handler, 0,
+					       name,
+					       &qspi->dev_ids[val]);
+			if (ret < 0) {
+				dev_err(&pdev->dev, "IRQ %s not found\n", name);
+				goto qspi_probe_err;
+			}
+
+			qspi->dev_ids[val].dev = qspi;
+			qspi->dev_ids[val].irqp = &qspi_irq_tab[val];
+			num_ints++;
+			dev_dbg(&pdev->dev, "registered IRQ %s %d\n",
+				qspi_irq_tab[val].irq_name,
+				irq);
+		}
+	}
+
+	if (!num_ints) {
+		dev_err(&pdev->dev, "no IRQs registered, cannot init driver\n");
+		ret = -EINVAL;
+		goto qspi_probe_err;
+	}
+
+	/*
+	 * Some SoCs integrate spi controller (e.g., its interrupt bits)
+	 * in specific ways
+	 */
+	if (soc_intc) {
+		qspi->soc_intc = soc_intc;
+		soc_intc->bcm_qspi_int_set(soc_intc, MSPI_DONE, true);
+	} else {
+		qspi->soc_intc = NULL;
+	}
+
+	qspi->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(qspi->clk)) {
+		dev_warn(dev, "unable to get clock\n");
+		ret = PTR_ERR(qspi->clk);
+		goto qspi_probe_err;
+	}
+
+	ret = clk_prepare_enable(qspi->clk);
+	if (ret) {
+		dev_err(dev, "failed to prepare clock\n");
+		goto qspi_probe_err;
+	}
+
+	qspi->base_clk = clk_get_rate(qspi->clk);
+	qspi->max_speed_hz = qspi->base_clk / (QSPI_SPBR_MIN * 2);
+
+	bcm_qspi_hw_init(qspi);
+	init_completion(&qspi->mspi_done);
+	init_completion(&qspi->bspi_done);
+	qspi->curr_cs = -1;
+
+	platform_set_drvdata(pdev, qspi);
+
+	qspi->xfer_mode.width = -1;
+	qspi->xfer_mode.addrlen = -1;
+	qspi->xfer_mode.hp = -1;
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret < 0) {
+		dev_err(dev, "can't register master\n");
+		goto qspi_reg_err;
+	}
+
+	return 0;
+
+qspi_reg_err:
+	bcm_qspi_hw_uninit(qspi);
+	clk_disable_unprepare(qspi->clk);
+qspi_probe_err:
+	kfree(qspi->dev_ids);
+qspi_resource_err:
+	spi_master_put(master);
+	return ret;
+}
+/* probe function to be called by SoC specific platform driver probe */
+EXPORT_SYMBOL_GPL(bcm_qspi_probe);
+
+int bcm_qspi_remove(struct platform_device *pdev)
+{
+	struct bcm_qspi *qspi = platform_get_drvdata(pdev);
+
+	bcm_qspi_hw_uninit(qspi);
+	clk_disable_unprepare(qspi->clk);
+	kfree(qspi->dev_ids);
+	spi_unregister_master(qspi->master);
+
+	return 0;
+}
+/* function to be called by SoC specific platform driver remove() */
+EXPORT_SYMBOL_GPL(bcm_qspi_remove);
+
+static int __maybe_unused bcm_qspi_suspend(struct device *dev)
+{
+	struct bcm_qspi *qspi = dev_get_drvdata(dev);
+
+	/* store the override strap value */
+	if (!bcm_qspi_bspi_ver_three(qspi))
+		qspi->s3_strap_override_ctrl =
+			bcm_qspi_read(qspi, BSPI, BSPI_STRAP_OVERRIDE_CTRL);
+
+	spi_master_suspend(qspi->master);
+	clk_disable(qspi->clk);
+	bcm_qspi_hw_uninit(qspi);
+
+	return 0;
+};
+
+static int __maybe_unused bcm_qspi_resume(struct device *dev)
+{
+	struct bcm_qspi *qspi = dev_get_drvdata(dev);
+	int ret = 0;
+
+	bcm_qspi_hw_init(qspi);
+	bcm_qspi_chip_select(qspi, qspi->curr_cs);
+	if (qspi->soc_intc)
+		/* enable MSPI interrupt */
+		qspi->soc_intc->bcm_qspi_int_set(qspi->soc_intc, MSPI_DONE,
+						 true);
+
+	ret = clk_enable(qspi->clk);
+	if (!ret)
+		spi_master_resume(qspi->master);
+
+	return ret;
+}
+
+SIMPLE_DEV_PM_OPS(bcm_qspi_pm_ops, bcm_qspi_suspend, bcm_qspi_resume);
+
+/* pm_ops to be called by SoC specific platform driver */
+EXPORT_SYMBOL_GPL(bcm_qspi_pm_ops);
+
+MODULE_AUTHOR("Kamal Dasu");
+MODULE_DESCRIPTION("Broadcom QSPI driver");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:" DRIVER_NAME);
diff --git a/drivers/spi/spi-bcm-qspi.h b/drivers/spi/spi-bcm-qspi.h
new file mode 100644
index 0000000..7abfc75
--- /dev/null
+++ b/drivers/spi/spi-bcm-qspi.h
@@ -0,0 +1,115 @@
+/*
+ * Copyright 2016 Broadcom
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation (the "GPL").
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License version 2 (GPLv2) for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * version 2 (GPLv2) along with this source code.
+ */
+
+#ifndef __SPI_BCM_QSPI_H__
+#define __SPI_BCM_QSPI_H__
+
+#include <linux/types.h>
+#include <linux/io.h>
+
+/* BSPI interrupt masks */
+#define INTR_BSPI_LR_OVERREAD_MASK		BIT(4)
+#define INTR_BSPI_LR_SESSION_DONE_MASK		BIT(3)
+#define INTR_BSPI_LR_IMPATIENT_MASK		BIT(2)
+#define INTR_BSPI_LR_SESSION_ABORTED_MASK	BIT(1)
+#define INTR_BSPI_LR_FULLNESS_REACHED_MASK	BIT(0)
+
+#define BSPI_LR_INTERRUPTS_DATA		       \
+	(INTR_BSPI_LR_SESSION_DONE_MASK |	       \
+	 INTR_BSPI_LR_FULLNESS_REACHED_MASK)
+
+#define BSPI_LR_INTERRUPTS_ERROR               \
+	(INTR_BSPI_LR_OVERREAD_MASK |	       \
+	 INTR_BSPI_LR_IMPATIENT_MASK |	       \
+	 INTR_BSPI_LR_SESSION_ABORTED_MASK)
+
+#define BSPI_LR_INTERRUPTS_ALL                 \
+	(BSPI_LR_INTERRUPTS_ERROR |	       \
+	 BSPI_LR_INTERRUPTS_DATA)
+
+/* MSPI Interrupt masks */
+#define INTR_MSPI_HALTED_MASK			BIT(6)
+#define INTR_MSPI_DONE_MASK			BIT(5)
+
+#define MSPI_INTERRUPTS_ALL		       \
+	(INTR_MSPI_DONE_MASK |		       \
+	 INTR_MSPI_HALTED_MASK)
+
+#define QSPI_INTERRUPTS_ALL                    \
+	(MSPI_INTERRUPTS_ALL |		       \
+	 BSPI_LR_INTERRUPTS_ALL)
+
+struct platform_device;
+struct dev_pm_ops;
+
+enum {
+	MSPI_DONE = 0x1,
+	BSPI_DONE = 0x2,
+	BSPI_ERR = 0x4,
+	MSPI_BSPI_DONE = 0x7
+};
+
+struct bcm_qspi_soc_intc {
+	void (*bcm_qspi_int_ack)(struct bcm_qspi_soc_intc *soc_intc, int type);
+	void (*bcm_qspi_int_set)(struct bcm_qspi_soc_intc *soc_intc, int type,
+				 bool en);
+	u32 (*bcm_qspi_get_int_status)(struct bcm_qspi_soc_intc *soc_intc);
+};
+
+/* Read controller register*/
+static inline u32 bcm_qspi_readl(bool be, void __iomem *addr)
+{
+	if (be)
+		return ioread32be(addr);
+	else
+		return readl_relaxed(addr);
+}
+
+/* Write controller register*/
+static inline void bcm_qspi_writel(bool be,
+				   unsigned int data, void __iomem *addr)
+{
+	if (be)
+		iowrite32be(data, addr);
+	else
+		writel_relaxed(data, addr);
+}
+
+static inline u32 get_qspi_mask(int type)
+{
+	switch (type) {
+	case MSPI_DONE:
+		return INTR_MSPI_DONE_MASK;
+	case BSPI_DONE:
+		return BSPI_LR_INTERRUPTS_ALL;
+	case MSPI_BSPI_DONE:
+		return QSPI_INTERRUPTS_ALL;
+	case BSPI_ERR:
+		return BSPI_LR_INTERRUPTS_ERROR;
+	}
+
+	return 0;
+}
+
+/* The common driver functions to be called by the SoC platform driver */
+int bcm_qspi_probe(struct platform_device *pdev,
+		   struct bcm_qspi_soc_intc *soc_intc);
+int bcm_qspi_remove(struct platform_device *pdev);
+
+/* pm_ops used by the SoC platform driver called on PM suspend/resume */
+extern const struct dev_pm_ops bcm_qspi_pm_ops;
+
+#endif /* __SPI_BCM_QSPI_H__ */
diff --git a/drivers/spi/spi-bcm2835.c b/drivers/spi/spi-bcm2835.c
new file mode 100644
index 0000000..f35cc10
--- /dev/null
+++ b/drivers/spi/spi-bcm2835.c
@@ -0,0 +1,846 @@
+/*
+ * Driver for Broadcom BCM2835 SPI Controllers
+ *
+ * Copyright (C) 2012 Chris Boot
+ * Copyright (C) 2013 Stephen Warren
+ * Copyright (C) 2015 Martin Sperl
+ *
+ * This driver is inspired by:
+ * spi-ath79.c, Copyright (C) 2009-2011 Gabor Juhos <juhosg@openwrt.org>
+ * spi-atmel.c, Copyright (C) 2006 Atmel Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <asm/page.h>
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_device.h>
+#include <linux/of_gpio.h>
+#include <linux/of_irq.h>
+#include <linux/spi/spi.h>
+
+/* SPI register offsets */
+#define BCM2835_SPI_CS			0x00
+#define BCM2835_SPI_FIFO		0x04
+#define BCM2835_SPI_CLK			0x08
+#define BCM2835_SPI_DLEN		0x0c
+#define BCM2835_SPI_LTOH		0x10
+#define BCM2835_SPI_DC			0x14
+
+/* Bitfields in CS */
+#define BCM2835_SPI_CS_LEN_LONG		0x02000000
+#define BCM2835_SPI_CS_DMA_LEN		0x01000000
+#define BCM2835_SPI_CS_CSPOL2		0x00800000
+#define BCM2835_SPI_CS_CSPOL1		0x00400000
+#define BCM2835_SPI_CS_CSPOL0		0x00200000
+#define BCM2835_SPI_CS_RXF		0x00100000
+#define BCM2835_SPI_CS_RXR		0x00080000
+#define BCM2835_SPI_CS_TXD		0x00040000
+#define BCM2835_SPI_CS_RXD		0x00020000
+#define BCM2835_SPI_CS_DONE		0x00010000
+#define BCM2835_SPI_CS_LEN		0x00002000
+#define BCM2835_SPI_CS_REN		0x00001000
+#define BCM2835_SPI_CS_ADCS		0x00000800
+#define BCM2835_SPI_CS_INTR		0x00000400
+#define BCM2835_SPI_CS_INTD		0x00000200
+#define BCM2835_SPI_CS_DMAEN		0x00000100
+#define BCM2835_SPI_CS_TA		0x00000080
+#define BCM2835_SPI_CS_CSPOL		0x00000040
+#define BCM2835_SPI_CS_CLEAR_RX		0x00000020
+#define BCM2835_SPI_CS_CLEAR_TX		0x00000010
+#define BCM2835_SPI_CS_CPOL		0x00000008
+#define BCM2835_SPI_CS_CPHA		0x00000004
+#define BCM2835_SPI_CS_CS_10		0x00000002
+#define BCM2835_SPI_CS_CS_01		0x00000001
+
+#define BCM2835_SPI_POLLING_LIMIT_US	30
+#define BCM2835_SPI_POLLING_JIFFIES	2
+#define BCM2835_SPI_DMA_MIN_LENGTH	96
+#define BCM2835_SPI_MODE_BITS	(SPI_CPOL | SPI_CPHA | SPI_CS_HIGH \
+				| SPI_NO_CS | SPI_3WIRE)
+
+#define DRV_NAME	"spi-bcm2835"
+
+struct bcm2835_spi {
+	void __iomem *regs;
+	struct clk *clk;
+	int irq;
+	const u8 *tx_buf;
+	u8 *rx_buf;
+	int tx_len;
+	int rx_len;
+	bool dma_pending;
+};
+
+static inline u32 bcm2835_rd(struct bcm2835_spi *bs, unsigned reg)
+{
+	return readl(bs->regs + reg);
+}
+
+static inline void bcm2835_wr(struct bcm2835_spi *bs, unsigned reg, u32 val)
+{
+	writel(val, bs->regs + reg);
+}
+
+static inline void bcm2835_rd_fifo(struct bcm2835_spi *bs)
+{
+	u8 byte;
+
+	while ((bs->rx_len) &&
+	       (bcm2835_rd(bs, BCM2835_SPI_CS) & BCM2835_SPI_CS_RXD)) {
+		byte = bcm2835_rd(bs, BCM2835_SPI_FIFO);
+		if (bs->rx_buf)
+			*bs->rx_buf++ = byte;
+		bs->rx_len--;
+	}
+}
+
+static inline void bcm2835_wr_fifo(struct bcm2835_spi *bs)
+{
+	u8 byte;
+
+	while ((bs->tx_len) &&
+	       (bcm2835_rd(bs, BCM2835_SPI_CS) & BCM2835_SPI_CS_TXD)) {
+		byte = bs->tx_buf ? *bs->tx_buf++ : 0;
+		bcm2835_wr(bs, BCM2835_SPI_FIFO, byte);
+		bs->tx_len--;
+	}
+}
+
+static void bcm2835_spi_reset_hw(struct spi_master *master)
+{
+	struct bcm2835_spi *bs = spi_master_get_devdata(master);
+	u32 cs = bcm2835_rd(bs, BCM2835_SPI_CS);
+
+	/* Disable SPI interrupts and transfer */
+	cs &= ~(BCM2835_SPI_CS_INTR |
+		BCM2835_SPI_CS_INTD |
+		BCM2835_SPI_CS_DMAEN |
+		BCM2835_SPI_CS_TA);
+	/* and reset RX/TX FIFOS */
+	cs |= BCM2835_SPI_CS_CLEAR_RX | BCM2835_SPI_CS_CLEAR_TX;
+
+	/* and reset the SPI_HW */
+	bcm2835_wr(bs, BCM2835_SPI_CS, cs);
+	/* as well as DLEN */
+	bcm2835_wr(bs, BCM2835_SPI_DLEN, 0);
+}
+
+static irqreturn_t bcm2835_spi_interrupt(int irq, void *dev_id)
+{
+	struct spi_master *master = dev_id;
+	struct bcm2835_spi *bs = spi_master_get_devdata(master);
+
+	/* Read as many bytes as possible from FIFO */
+	bcm2835_rd_fifo(bs);
+	/* Write as many bytes as possible to FIFO */
+	bcm2835_wr_fifo(bs);
+
+	/* based on flags decide if we can finish the transfer */
+	if (bcm2835_rd(bs, BCM2835_SPI_CS) & BCM2835_SPI_CS_DONE) {
+		/* Transfer complete - reset SPI HW */
+		bcm2835_spi_reset_hw(master);
+		/* wake up the framework */
+		complete(&master->xfer_completion);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int bcm2835_spi_transfer_one_irq(struct spi_master *master,
+					struct spi_device *spi,
+					struct spi_transfer *tfr,
+					u32 cs)
+{
+	struct bcm2835_spi *bs = spi_master_get_devdata(master);
+
+	/* fill in fifo if we have gpio-cs
+	 * note that there have been rare events where the native-CS
+	 * flapped for <1us which may change the behaviour
+	 * with gpio-cs this does not happen, so it is implemented
+	 * only for this case
+	 */
+	if (gpio_is_valid(spi->cs_gpio)) {
+		/* enable HW block, but without interrupts enabled
+		 * this would triggern an immediate interrupt
+		 */
+		bcm2835_wr(bs, BCM2835_SPI_CS,
+			   cs | BCM2835_SPI_CS_TA);
+		/* fill in tx fifo as much as possible */
+		bcm2835_wr_fifo(bs);
+	}
+
+	/*
+	 * Enable the HW block. This will immediately trigger a DONE (TX
+	 * empty) interrupt, upon which we will fill the TX FIFO with the
+	 * first TX bytes. Pre-filling the TX FIFO here to avoid the
+	 * interrupt doesn't work:-(
+	 */
+	cs |= BCM2835_SPI_CS_INTR | BCM2835_SPI_CS_INTD | BCM2835_SPI_CS_TA;
+	bcm2835_wr(bs, BCM2835_SPI_CS, cs);
+
+	/* signal that we need to wait for completion */
+	return 1;
+}
+
+/*
+ * DMA support
+ *
+ * this implementation has currently a few issues in so far as it does
+ * not work arrount limitations of the HW.
+ *
+ * the main one being that DMA transfers are limited to 16 bit
+ * (so 0 to 65535 bytes) by the SPI HW due to BCM2835_SPI_DLEN
+ *
+ * also we currently assume that the scatter-gather fragments are
+ * all multiple of 4 (except the last) - otherwise we would need
+ * to reset the FIFO before subsequent transfers...
+ * this also means that tx/rx transfers sg's need to be of equal size!
+ *
+ * there may be a few more border-cases we may need to address as well
+ * but unfortunately this would mean splitting up the scatter-gather
+ * list making it slightly unpractical...
+ */
+static void bcm2835_spi_dma_done(void *data)
+{
+	struct spi_master *master = data;
+	struct bcm2835_spi *bs = spi_master_get_devdata(master);
+
+	/* reset fifo and HW */
+	bcm2835_spi_reset_hw(master);
+
+	/* and terminate tx-dma as we do not have an irq for it
+	 * because when the rx dma will terminate and this callback
+	 * is called the tx-dma must have finished - can't get to this
+	 * situation otherwise...
+	 */
+	dmaengine_terminate_all(master->dma_tx);
+
+	/* mark as no longer pending */
+	bs->dma_pending = 0;
+
+	/* and mark as completed */;
+	complete(&master->xfer_completion);
+}
+
+static int bcm2835_spi_prepare_sg(struct spi_master *master,
+				  struct spi_transfer *tfr,
+				  bool is_tx)
+{
+	struct dma_chan *chan;
+	struct scatterlist *sgl;
+	unsigned int nents;
+	enum dma_transfer_direction dir;
+	unsigned long flags;
+
+	struct dma_async_tx_descriptor *desc;
+	dma_cookie_t cookie;
+
+	if (is_tx) {
+		dir   = DMA_MEM_TO_DEV;
+		chan  = master->dma_tx;
+		nents = tfr->tx_sg.nents;
+		sgl   = tfr->tx_sg.sgl;
+		flags = 0 /* no  tx interrupt */;
+
+	} else {
+		dir   = DMA_DEV_TO_MEM;
+		chan  = master->dma_rx;
+		nents = tfr->rx_sg.nents;
+		sgl   = tfr->rx_sg.sgl;
+		flags = DMA_PREP_INTERRUPT;
+	}
+	/* prepare the channel */
+	desc = dmaengine_prep_slave_sg(chan, sgl, nents, dir, flags);
+	if (!desc)
+		return -EINVAL;
+
+	/* set callback for rx */
+	if (!is_tx) {
+		desc->callback = bcm2835_spi_dma_done;
+		desc->callback_param = master;
+	}
+
+	/* submit it to DMA-engine */
+	cookie = dmaengine_submit(desc);
+
+	return dma_submit_error(cookie);
+}
+
+static inline int bcm2835_check_sg_length(struct sg_table *sgt)
+{
+	int i;
+	struct scatterlist *sgl;
+
+	/* check that the sg entries are word-sized (except for last) */
+	for_each_sg(sgt->sgl, sgl, (int)sgt->nents - 1, i) {
+		if (sg_dma_len(sgl) % 4)
+			return -EFAULT;
+	}
+
+	return 0;
+}
+
+static int bcm2835_spi_transfer_one_dma(struct spi_master *master,
+					struct spi_device *spi,
+					struct spi_transfer *tfr,
+					u32 cs)
+{
+	struct bcm2835_spi *bs = spi_master_get_devdata(master);
+	int ret;
+
+	/* check that the scatter gather segments are all a multiple of 4 */
+	if (bcm2835_check_sg_length(&tfr->tx_sg) ||
+	    bcm2835_check_sg_length(&tfr->rx_sg)) {
+		dev_warn_once(&spi->dev,
+			      "scatter gather segment length is not a multiple of 4 - falling back to interrupt mode\n");
+		return bcm2835_spi_transfer_one_irq(master, spi, tfr, cs);
+	}
+
+	/* setup tx-DMA */
+	ret = bcm2835_spi_prepare_sg(master, tfr, true);
+	if (ret)
+		return ret;
+
+	/* start TX early */
+	dma_async_issue_pending(master->dma_tx);
+
+	/* mark as dma pending */
+	bs->dma_pending = 1;
+
+	/* set the DMA length */
+	bcm2835_wr(bs, BCM2835_SPI_DLEN, tfr->len);
+
+	/* start the HW */
+	bcm2835_wr(bs, BCM2835_SPI_CS,
+		   cs | BCM2835_SPI_CS_TA | BCM2835_SPI_CS_DMAEN);
+
+	/* setup rx-DMA late - to run transfers while
+	 * mapping of the rx buffers still takes place
+	 * this saves 10us or more.
+	 */
+	ret = bcm2835_spi_prepare_sg(master, tfr, false);
+	if (ret) {
+		/* need to reset on errors */
+		dmaengine_terminate_all(master->dma_tx);
+		bcm2835_spi_reset_hw(master);
+		return ret;
+	}
+
+	/* start rx dma late */
+	dma_async_issue_pending(master->dma_rx);
+
+	/* wait for wakeup in framework */
+	return 1;
+}
+
+static bool bcm2835_spi_can_dma(struct spi_master *master,
+				struct spi_device *spi,
+				struct spi_transfer *tfr)
+{
+	/* only run for gpio_cs */
+	if (!gpio_is_valid(spi->cs_gpio))
+		return false;
+
+	/* we start DMA efforts only on bigger transfers */
+	if (tfr->len < BCM2835_SPI_DMA_MIN_LENGTH)
+		return false;
+
+	/* BCM2835_SPI_DLEN has defined a max transfer size as
+	 * 16 bit, so max is 65535
+	 * we can revisit this by using an alternative transfer
+	 * method - ideally this would get done without any more
+	 * interaction...
+	 */
+	if (tfr->len > 65535) {
+		dev_warn_once(&spi->dev,
+			      "transfer size of %d too big for dma-transfer\n",
+			      tfr->len);
+		return false;
+	}
+
+	/* if we run rx/tx_buf with word aligned addresses then we are OK */
+	if ((((size_t)tfr->rx_buf & 3) == 0) &&
+	    (((size_t)tfr->tx_buf & 3) == 0))
+		return true;
+
+	/* otherwise we only allow transfers within the same page
+	 * to avoid wasting time on dma_mapping when it is not practical
+	 */
+	if (((size_t)tfr->tx_buf & (PAGE_SIZE - 1)) + tfr->len > PAGE_SIZE) {
+		dev_warn_once(&spi->dev,
+			      "Unaligned spi tx-transfer bridging page\n");
+		return false;
+	}
+	if (((size_t)tfr->rx_buf & (PAGE_SIZE - 1)) + tfr->len > PAGE_SIZE) {
+		dev_warn_once(&spi->dev,
+			      "Unaligned spi rx-transfer bridging page\n");
+		return false;
+	}
+
+	/* return OK */
+	return true;
+}
+
+static void bcm2835_dma_release(struct spi_master *master)
+{
+	if (master->dma_tx) {
+		dmaengine_terminate_all(master->dma_tx);
+		dma_release_channel(master->dma_tx);
+		master->dma_tx = NULL;
+	}
+	if (master->dma_rx) {
+		dmaengine_terminate_all(master->dma_rx);
+		dma_release_channel(master->dma_rx);
+		master->dma_rx = NULL;
+	}
+}
+
+static void bcm2835_dma_init(struct spi_master *master, struct device *dev)
+{
+	struct dma_slave_config slave_config;
+	const __be32 *addr;
+	dma_addr_t dma_reg_base;
+	int ret;
+
+	/* base address in dma-space */
+	addr = of_get_address(master->dev.of_node, 0, NULL, NULL);
+	if (!addr) {
+		dev_err(dev, "could not get DMA-register address - not using dma mode\n");
+		goto err;
+	}
+	dma_reg_base = be32_to_cpup(addr);
+
+	/* get tx/rx dma */
+	master->dma_tx = dma_request_slave_channel(dev, "tx");
+	if (!master->dma_tx) {
+		dev_err(dev, "no tx-dma configuration found - not using dma mode\n");
+		goto err;
+	}
+	master->dma_rx = dma_request_slave_channel(dev, "rx");
+	if (!master->dma_rx) {
+		dev_err(dev, "no rx-dma configuration found - not using dma mode\n");
+		goto err_release;
+	}
+
+	/* configure DMAs */
+	slave_config.direction = DMA_MEM_TO_DEV;
+	slave_config.dst_addr = (u32)(dma_reg_base + BCM2835_SPI_FIFO);
+	slave_config.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+
+	ret = dmaengine_slave_config(master->dma_tx, &slave_config);
+	if (ret)
+		goto err_config;
+
+	slave_config.direction = DMA_DEV_TO_MEM;
+	slave_config.src_addr = (u32)(dma_reg_base + BCM2835_SPI_FIFO);
+	slave_config.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+
+	ret = dmaengine_slave_config(master->dma_rx, &slave_config);
+	if (ret)
+		goto err_config;
+
+	/* all went well, so set can_dma */
+	master->can_dma = bcm2835_spi_can_dma;
+	master->max_dma_len = 65535; /* limitation by BCM2835_SPI_DLEN */
+	/* need to do TX AND RX DMA, so we need dummy buffers */
+	master->flags = SPI_MASTER_MUST_RX | SPI_MASTER_MUST_TX;
+
+	return;
+
+err_config:
+	dev_err(dev, "issue configuring dma: %d - not using DMA mode\n",
+		ret);
+err_release:
+	bcm2835_dma_release(master);
+err:
+	return;
+}
+
+static int bcm2835_spi_transfer_one_poll(struct spi_master *master,
+					 struct spi_device *spi,
+					 struct spi_transfer *tfr,
+					 u32 cs,
+					 unsigned long long xfer_time_us)
+{
+	struct bcm2835_spi *bs = spi_master_get_devdata(master);
+	unsigned long timeout;
+
+	/* enable HW block without interrupts */
+	bcm2835_wr(bs, BCM2835_SPI_CS, cs | BCM2835_SPI_CS_TA);
+
+	/* fill in the fifo before timeout calculations
+	 * if we are interrupted here, then the data is
+	 * getting transferred by the HW while we are interrupted
+	 */
+	bcm2835_wr_fifo(bs);
+
+	/* set the timeout */
+	timeout = jiffies + BCM2835_SPI_POLLING_JIFFIES;
+
+	/* loop until finished the transfer */
+	while (bs->rx_len) {
+		/* fill in tx fifo with remaining data */
+		bcm2835_wr_fifo(bs);
+
+		/* read from fifo as much as possible */
+		bcm2835_rd_fifo(bs);
+
+		/* if there is still data pending to read
+		 * then check the timeout
+		 */
+		if (bs->rx_len && time_after(jiffies, timeout)) {
+			dev_dbg_ratelimited(&spi->dev,
+					    "timeout period reached: jiffies: %lu remaining tx/rx: %d/%d - falling back to interrupt mode\n",
+					    jiffies - timeout,
+					    bs->tx_len, bs->rx_len);
+			/* fall back to interrupt mode */
+			return bcm2835_spi_transfer_one_irq(master, spi,
+							    tfr, cs);
+		}
+	}
+
+	/* Transfer complete - reset SPI HW */
+	bcm2835_spi_reset_hw(master);
+	/* and return without waiting for completion */
+	return 0;
+}
+
+static int bcm2835_spi_transfer_one(struct spi_master *master,
+				    struct spi_device *spi,
+				    struct spi_transfer *tfr)
+{
+	struct bcm2835_spi *bs = spi_master_get_devdata(master);
+	unsigned long spi_hz, clk_hz, cdiv;
+	unsigned long spi_used_hz;
+	unsigned long long xfer_time_us;
+	u32 cs = bcm2835_rd(bs, BCM2835_SPI_CS);
+
+	/* set clock */
+	spi_hz = tfr->speed_hz;
+	clk_hz = clk_get_rate(bs->clk);
+
+	if (spi_hz >= clk_hz / 2) {
+		cdiv = 2; /* clk_hz/2 is the fastest we can go */
+	} else if (spi_hz) {
+		/* CDIV must be a multiple of two */
+		cdiv = DIV_ROUND_UP(clk_hz, spi_hz);
+		cdiv += (cdiv % 2);
+
+		if (cdiv >= 65536)
+			cdiv = 0; /* 0 is the slowest we can go */
+	} else {
+		cdiv = 0; /* 0 is the slowest we can go */
+	}
+	spi_used_hz = cdiv ? (clk_hz / cdiv) : (clk_hz / 65536);
+	bcm2835_wr(bs, BCM2835_SPI_CLK, cdiv);
+
+	/* handle all the 3-wire mode */
+	if ((spi->mode & SPI_3WIRE) && (tfr->rx_buf))
+		cs |= BCM2835_SPI_CS_REN;
+	else
+		cs &= ~BCM2835_SPI_CS_REN;
+
+	/* for gpio_cs set dummy CS so that no HW-CS get changed
+	 * we can not run this in bcm2835_spi_set_cs, as it does
+	 * not get called for cs_gpio cases, so we need to do it here
+	 */
+	if (gpio_is_valid(spi->cs_gpio) || (spi->mode & SPI_NO_CS))
+		cs |= BCM2835_SPI_CS_CS_10 | BCM2835_SPI_CS_CS_01;
+
+	/* set transmit buffers and length */
+	bs->tx_buf = tfr->tx_buf;
+	bs->rx_buf = tfr->rx_buf;
+	bs->tx_len = tfr->len;
+	bs->rx_len = tfr->len;
+
+	/* calculate the estimated time in us the transfer runs */
+	xfer_time_us = (unsigned long long)tfr->len
+		* 9 /* clocks/byte - SPI-HW waits 1 clock after each byte */
+		* 1000000;
+	do_div(xfer_time_us, spi_used_hz);
+
+	/* for short requests run polling*/
+	if (xfer_time_us <= BCM2835_SPI_POLLING_LIMIT_US)
+		return bcm2835_spi_transfer_one_poll(master, spi, tfr,
+						     cs, xfer_time_us);
+
+	/* run in dma mode if conditions are right */
+	if (master->can_dma && bcm2835_spi_can_dma(master, spi, tfr))
+		return bcm2835_spi_transfer_one_dma(master, spi, tfr, cs);
+
+	/* run in interrupt-mode */
+	return bcm2835_spi_transfer_one_irq(master, spi, tfr, cs);
+}
+
+static int bcm2835_spi_prepare_message(struct spi_master *master,
+				       struct spi_message *msg)
+{
+	struct spi_device *spi = msg->spi;
+	struct bcm2835_spi *bs = spi_master_get_devdata(master);
+	u32 cs = bcm2835_rd(bs, BCM2835_SPI_CS);
+
+	cs &= ~(BCM2835_SPI_CS_CPOL | BCM2835_SPI_CS_CPHA);
+
+	if (spi->mode & SPI_CPOL)
+		cs |= BCM2835_SPI_CS_CPOL;
+	if (spi->mode & SPI_CPHA)
+		cs |= BCM2835_SPI_CS_CPHA;
+
+	bcm2835_wr(bs, BCM2835_SPI_CS, cs);
+
+	return 0;
+}
+
+static void bcm2835_spi_handle_err(struct spi_master *master,
+				   struct spi_message *msg)
+{
+	struct bcm2835_spi *bs = spi_master_get_devdata(master);
+
+	/* if an error occurred and we have an active dma, then terminate */
+	if (bs->dma_pending) {
+		dmaengine_terminate_all(master->dma_tx);
+		dmaengine_terminate_all(master->dma_rx);
+		bs->dma_pending = 0;
+	}
+	/* and reset */
+	bcm2835_spi_reset_hw(master);
+}
+
+static void bcm2835_spi_set_cs(struct spi_device *spi, bool gpio_level)
+{
+	/*
+	 * we can assume that we are "native" as per spi_set_cs
+	 *   calling us ONLY when cs_gpio is not set
+	 * we can also assume that we are CS < 3 as per bcm2835_spi_setup
+	 *   we would not get called because of error handling there.
+	 * the level passed is the electrical level not enabled/disabled
+	 *   so it has to get translated back to enable/disable
+	 *   see spi_set_cs in spi.c for the implementation
+	 */
+
+	struct spi_master *master = spi->master;
+	struct bcm2835_spi *bs = spi_master_get_devdata(master);
+	u32 cs = bcm2835_rd(bs, BCM2835_SPI_CS);
+	bool enable;
+
+	/* calculate the enable flag from the passed gpio_level */
+	enable = (spi->mode & SPI_CS_HIGH) ? gpio_level : !gpio_level;
+
+	/* set flags for "reverse" polarity in the registers */
+	if (spi->mode & SPI_CS_HIGH) {
+		/* set the correct CS-bits */
+		cs |= BCM2835_SPI_CS_CSPOL;
+		cs |= BCM2835_SPI_CS_CSPOL0 << spi->chip_select;
+	} else {
+		/* clean the CS-bits */
+		cs &= ~BCM2835_SPI_CS_CSPOL;
+		cs &= ~(BCM2835_SPI_CS_CSPOL0 << spi->chip_select);
+	}
+
+	/* select the correct chip_select depending on disabled/enabled */
+	if (enable) {
+		/* set cs correctly */
+		if (spi->mode & SPI_NO_CS) {
+			/* use the "undefined" chip-select */
+			cs |= BCM2835_SPI_CS_CS_10 | BCM2835_SPI_CS_CS_01;
+		} else {
+			/* set the chip select */
+			cs &= ~(BCM2835_SPI_CS_CS_10 | BCM2835_SPI_CS_CS_01);
+			cs |= spi->chip_select;
+		}
+	} else {
+		/* disable CSPOL which puts HW-CS into deselected state */
+		cs &= ~BCM2835_SPI_CS_CSPOL;
+		/* use the "undefined" chip-select as precaution */
+		cs |= BCM2835_SPI_CS_CS_10 | BCM2835_SPI_CS_CS_01;
+	}
+
+	/* finally set the calculated flags in SPI_CS */
+	bcm2835_wr(bs, BCM2835_SPI_CS, cs);
+}
+
+static int chip_match_name(struct gpio_chip *chip, void *data)
+{
+	return !strcmp(chip->label, data);
+}
+
+static int bcm2835_spi_setup(struct spi_device *spi)
+{
+	int err;
+	struct gpio_chip *chip;
+	/*
+	 * sanity checking the native-chipselects
+	 */
+	if (spi->mode & SPI_NO_CS)
+		return 0;
+	if (gpio_is_valid(spi->cs_gpio))
+		return 0;
+	if (spi->chip_select > 1) {
+		/* error in the case of native CS requested with CS > 1
+		 * officially there is a CS2, but it is not documented
+		 * which GPIO is connected with that...
+		 */
+		dev_err(&spi->dev,
+			"setup: only two native chip-selects are supported\n");
+		return -EINVAL;
+	}
+	/* now translate native cs to GPIO */
+
+	/* get the gpio chip for the base */
+	chip = gpiochip_find("pinctrl-bcm2835", chip_match_name);
+	if (!chip)
+		return 0;
+
+	/* and calculate the real CS */
+	spi->cs_gpio = chip->base + 8 - spi->chip_select;
+
+	/* and set up the "mode" and level */
+	dev_info(&spi->dev, "setting up native-CS%i as GPIO %i\n",
+		 spi->chip_select, spi->cs_gpio);
+
+	/* set up GPIO as output and pull to the correct level */
+	err = gpio_direction_output(spi->cs_gpio,
+				    (spi->mode & SPI_CS_HIGH) ? 0 : 1);
+	if (err) {
+		dev_err(&spi->dev,
+			"could not set CS%i gpio %i as output: %i",
+			spi->chip_select, spi->cs_gpio, err);
+		return err;
+	}
+
+	return 0;
+}
+
+static int bcm2835_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct bcm2835_spi *bs;
+	struct resource *res;
+	int err;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*bs));
+	if (!master) {
+		dev_err(&pdev->dev, "spi_alloc_master() failed\n");
+		return -ENOMEM;
+	}
+
+	platform_set_drvdata(pdev, master);
+
+	master->mode_bits = BCM2835_SPI_MODE_BITS;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->num_chipselect = 3;
+	master->setup = bcm2835_spi_setup;
+	master->set_cs = bcm2835_spi_set_cs;
+	master->transfer_one = bcm2835_spi_transfer_one;
+	master->handle_err = bcm2835_spi_handle_err;
+	master->prepare_message = bcm2835_spi_prepare_message;
+	master->dev.of_node = pdev->dev.of_node;
+
+	bs = spi_master_get_devdata(master);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	bs->regs = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(bs->regs)) {
+		err = PTR_ERR(bs->regs);
+		goto out_master_put;
+	}
+
+	bs->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(bs->clk)) {
+		err = PTR_ERR(bs->clk);
+		dev_err(&pdev->dev, "could not get clk: %d\n", err);
+		goto out_master_put;
+	}
+
+	bs->irq = platform_get_irq(pdev, 0);
+	if (bs->irq <= 0) {
+		dev_err(&pdev->dev, "could not get IRQ: %d\n", bs->irq);
+		err = bs->irq ? bs->irq : -ENODEV;
+		goto out_master_put;
+	}
+
+	clk_prepare_enable(bs->clk);
+
+	bcm2835_dma_init(master, &pdev->dev);
+
+	/* initialise the hardware with the default polarities */
+	bcm2835_wr(bs, BCM2835_SPI_CS,
+		   BCM2835_SPI_CS_CLEAR_RX | BCM2835_SPI_CS_CLEAR_TX);
+
+	err = devm_request_irq(&pdev->dev, bs->irq, bcm2835_spi_interrupt, 0,
+			       dev_name(&pdev->dev), master);
+	if (err) {
+		dev_err(&pdev->dev, "could not request IRQ: %d\n", err);
+		goto out_clk_disable;
+	}
+
+	err = devm_spi_register_master(&pdev->dev, master);
+	if (err) {
+		dev_err(&pdev->dev, "could not register SPI master: %d\n", err);
+		goto out_clk_disable;
+	}
+
+	return 0;
+
+out_clk_disable:
+	clk_disable_unprepare(bs->clk);
+out_master_put:
+	spi_master_put(master);
+	return err;
+}
+
+static int bcm2835_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct bcm2835_spi *bs = spi_master_get_devdata(master);
+
+	/* Clear FIFOs, and disable the HW block */
+	bcm2835_wr(bs, BCM2835_SPI_CS,
+		   BCM2835_SPI_CS_CLEAR_RX | BCM2835_SPI_CS_CLEAR_TX);
+
+	clk_disable_unprepare(bs->clk);
+
+	bcm2835_dma_release(master);
+
+	return 0;
+}
+
+static const struct of_device_id bcm2835_spi_match[] = {
+	{ .compatible = "brcm,bcm2835-spi", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, bcm2835_spi_match);
+
+static struct platform_driver bcm2835_spi_driver = {
+	.driver		= {
+		.name		= DRV_NAME,
+		.of_match_table	= bcm2835_spi_match,
+	},
+	.probe		= bcm2835_spi_probe,
+	.remove		= bcm2835_spi_remove,
+};
+module_platform_driver(bcm2835_spi_driver);
+
+MODULE_DESCRIPTION("SPI controller driver for Broadcom BCM2835");
+MODULE_AUTHOR("Chris Boot <bootc@bootc.net>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-bcm2835aux.c b/drivers/spi/spi-bcm2835aux.c
new file mode 100644
index 0000000..3094d81
--- /dev/null
+++ b/drivers/spi/spi-bcm2835aux.c
@@ -0,0 +1,545 @@
+/*
+ * Driver for Broadcom BCM2835 auxiliary SPI Controllers
+ *
+ * the driver does not rely on the native chipselects at all
+ * but only uses the gpio type chipselects
+ *
+ * Based on: spi-bcm2835.c
+ *
+ * Copyright (C) 2015 Martin Sperl
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_device.h>
+#include <linux/of_gpio.h>
+#include <linux/of_irq.h>
+#include <linux/regmap.h>
+#include <linux/spi/spi.h>
+#include <linux/spinlock.h>
+
+/*
+ * spi register defines
+ *
+ * note there is garbage in the "official" documentation,
+ * so some data is taken from the file:
+ *   brcm_usrlib/dag/vmcsx/vcinclude/bcm2708_chip/aux_io.h
+ * inside of:
+ *   http://www.broadcom.com/docs/support/videocore/Brcm_Android_ICS_Graphics_Stack.tar.gz
+ */
+
+/* SPI register offsets */
+#define BCM2835_AUX_SPI_CNTL0	0x00
+#define BCM2835_AUX_SPI_CNTL1	0x04
+#define BCM2835_AUX_SPI_STAT	0x08
+#define BCM2835_AUX_SPI_PEEK	0x0C
+#define BCM2835_AUX_SPI_IO	0x20
+#define BCM2835_AUX_SPI_TXHOLD	0x30
+
+/* Bitfields in CNTL0 */
+#define BCM2835_AUX_SPI_CNTL0_SPEED	0xFFF00000
+#define BCM2835_AUX_SPI_CNTL0_SPEED_MAX	0xFFF
+#define BCM2835_AUX_SPI_CNTL0_SPEED_SHIFT	20
+#define BCM2835_AUX_SPI_CNTL0_CS	0x000E0000
+#define BCM2835_AUX_SPI_CNTL0_POSTINPUT	0x00010000
+#define BCM2835_AUX_SPI_CNTL0_VAR_CS	0x00008000
+#define BCM2835_AUX_SPI_CNTL0_VAR_WIDTH	0x00004000
+#define BCM2835_AUX_SPI_CNTL0_DOUTHOLD	0x00003000
+#define BCM2835_AUX_SPI_CNTL0_ENABLE	0x00000800
+#define BCM2835_AUX_SPI_CNTL0_IN_RISING	0x00000400
+#define BCM2835_AUX_SPI_CNTL0_CLEARFIFO	0x00000200
+#define BCM2835_AUX_SPI_CNTL0_OUT_RISING	0x00000100
+#define BCM2835_AUX_SPI_CNTL0_CPOL	0x00000080
+#define BCM2835_AUX_SPI_CNTL0_MSBF_OUT	0x00000040
+#define BCM2835_AUX_SPI_CNTL0_SHIFTLEN	0x0000003F
+
+/* Bitfields in CNTL1 */
+#define BCM2835_AUX_SPI_CNTL1_CSHIGH	0x00000700
+#define BCM2835_AUX_SPI_CNTL1_TXEMPTY	0x00000080
+#define BCM2835_AUX_SPI_CNTL1_IDLE	0x00000040
+#define BCM2835_AUX_SPI_CNTL1_MSBF_IN	0x00000002
+#define BCM2835_AUX_SPI_CNTL1_KEEP_IN	0x00000001
+
+/* Bitfields in STAT */
+#define BCM2835_AUX_SPI_STAT_TX_LVL	0xFF000000
+#define BCM2835_AUX_SPI_STAT_RX_LVL	0x00FF0000
+#define BCM2835_AUX_SPI_STAT_TX_FULL	0x00000400
+#define BCM2835_AUX_SPI_STAT_TX_EMPTY	0x00000200
+#define BCM2835_AUX_SPI_STAT_RX_FULL	0x00000100
+#define BCM2835_AUX_SPI_STAT_RX_EMPTY	0x00000080
+#define BCM2835_AUX_SPI_STAT_BUSY	0x00000040
+#define BCM2835_AUX_SPI_STAT_BITCOUNT	0x0000003F
+
+/* timeout values */
+#define BCM2835_AUX_SPI_POLLING_LIMIT_US	30
+#define BCM2835_AUX_SPI_POLLING_JIFFIES		2
+
+struct bcm2835aux_spi {
+	void __iomem *regs;
+	struct clk *clk;
+	int irq;
+	u32 cntl[2];
+	const u8 *tx_buf;
+	u8 *rx_buf;
+	int tx_len;
+	int rx_len;
+	int pending;
+};
+
+static inline u32 bcm2835aux_rd(struct bcm2835aux_spi *bs, unsigned reg)
+{
+	return readl(bs->regs + reg);
+}
+
+static inline void bcm2835aux_wr(struct bcm2835aux_spi *bs, unsigned reg,
+				 u32 val)
+{
+	writel(val, bs->regs + reg);
+}
+
+static inline void bcm2835aux_rd_fifo(struct bcm2835aux_spi *bs)
+{
+	u32 data;
+	int count = min(bs->rx_len, 3);
+
+	data = bcm2835aux_rd(bs, BCM2835_AUX_SPI_IO);
+	if (bs->rx_buf) {
+		switch (count) {
+		case 4:
+			*bs->rx_buf++ = (data >> 24) & 0xff;
+			/* fallthrough */
+		case 3:
+			*bs->rx_buf++ = (data >> 16) & 0xff;
+			/* fallthrough */
+		case 2:
+			*bs->rx_buf++ = (data >> 8) & 0xff;
+			/* fallthrough */
+		case 1:
+			*bs->rx_buf++ = (data >> 0) & 0xff;
+			/* fallthrough - no default */
+		}
+	}
+	bs->rx_len -= count;
+	bs->pending -= count;
+}
+
+static inline void bcm2835aux_wr_fifo(struct bcm2835aux_spi *bs)
+{
+	u32 data;
+	u8 byte;
+	int count;
+	int i;
+
+	/* gather up to 3 bytes to write to the FIFO */
+	count = min(bs->tx_len, 3);
+	data = 0;
+	for (i = 0; i < count; i++) {
+		byte = bs->tx_buf ? *bs->tx_buf++ : 0;
+		data |= byte << (8 * (2 - i));
+	}
+
+	/* and set the variable bit-length */
+	data |= (count * 8) << 24;
+
+	/* and decrement length */
+	bs->tx_len -= count;
+	bs->pending += count;
+
+	/* write to the correct TX-register */
+	if (bs->tx_len)
+		bcm2835aux_wr(bs, BCM2835_AUX_SPI_TXHOLD, data);
+	else
+		bcm2835aux_wr(bs, BCM2835_AUX_SPI_IO, data);
+}
+
+static void bcm2835aux_spi_reset_hw(struct bcm2835aux_spi *bs)
+{
+	/* disable spi clearing fifo and interrupts */
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL1, 0);
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL0,
+		      BCM2835_AUX_SPI_CNTL0_CLEARFIFO);
+}
+
+static irqreturn_t bcm2835aux_spi_interrupt(int irq, void *dev_id)
+{
+	struct spi_master *master = dev_id;
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+	irqreturn_t ret = IRQ_NONE;
+
+	/* IRQ may be shared, so return if our interrupts are disabled */
+	if (!(bcm2835aux_rd(bs, BCM2835_AUX_SPI_CNTL1) &
+	      (BCM2835_AUX_SPI_CNTL1_TXEMPTY | BCM2835_AUX_SPI_CNTL1_IDLE)))
+		return ret;
+
+	/* check if we have data to read */
+	while (bs->rx_len &&
+	       (!(bcm2835aux_rd(bs, BCM2835_AUX_SPI_STAT) &
+		  BCM2835_AUX_SPI_STAT_RX_EMPTY))) {
+		bcm2835aux_rd_fifo(bs);
+		ret = IRQ_HANDLED;
+	}
+
+	/* check if we have data to write */
+	while (bs->tx_len &&
+	       (bs->pending < 12) &&
+	       (!(bcm2835aux_rd(bs, BCM2835_AUX_SPI_STAT) &
+		  BCM2835_AUX_SPI_STAT_TX_FULL))) {
+		bcm2835aux_wr_fifo(bs);
+		ret = IRQ_HANDLED;
+	}
+
+	/* and check if we have reached "done" */
+	while (bs->rx_len &&
+	       (!(bcm2835aux_rd(bs, BCM2835_AUX_SPI_STAT) &
+		  BCM2835_AUX_SPI_STAT_BUSY))) {
+		bcm2835aux_rd_fifo(bs);
+		ret = IRQ_HANDLED;
+	}
+
+	if (!bs->tx_len) {
+		/* disable tx fifo empty interrupt */
+		bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL1, bs->cntl[1] |
+			BCM2835_AUX_SPI_CNTL1_IDLE);
+	}
+
+	/* and if rx_len is 0 then disable interrupts and wake up completion */
+	if (!bs->rx_len) {
+		bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL1, bs->cntl[1]);
+		complete(&master->xfer_completion);
+	}
+
+	/* and return */
+	return ret;
+}
+
+static int __bcm2835aux_spi_transfer_one_irq(struct spi_master *master,
+					     struct spi_device *spi,
+					     struct spi_transfer *tfr)
+{
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+
+	/* enable interrupts */
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL1, bs->cntl[1] |
+		BCM2835_AUX_SPI_CNTL1_TXEMPTY |
+		BCM2835_AUX_SPI_CNTL1_IDLE);
+
+	/* and wait for finish... */
+	return 1;
+}
+
+static int bcm2835aux_spi_transfer_one_irq(struct spi_master *master,
+					   struct spi_device *spi,
+					   struct spi_transfer *tfr)
+{
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+
+	/* fill in registers and fifos before enabling interrupts */
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL1, bs->cntl[1]);
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL0, bs->cntl[0]);
+
+	/* fill in tx fifo with data before enabling interrupts */
+	while ((bs->tx_len) &&
+	       (bs->pending < 12) &&
+	       (!(bcm2835aux_rd(bs, BCM2835_AUX_SPI_STAT) &
+		  BCM2835_AUX_SPI_STAT_TX_FULL))) {
+		bcm2835aux_wr_fifo(bs);
+	}
+
+	/* now run the interrupt mode */
+	return __bcm2835aux_spi_transfer_one_irq(master, spi, tfr);
+}
+
+static int bcm2835aux_spi_transfer_one_poll(struct spi_master *master,
+					    struct spi_device *spi,
+					struct spi_transfer *tfr)
+{
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+	unsigned long timeout;
+	u32 stat;
+
+	/* configure spi */
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL1, bs->cntl[1]);
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL0, bs->cntl[0]);
+
+	/* set the timeout */
+	timeout = jiffies + BCM2835_AUX_SPI_POLLING_JIFFIES;
+
+	/* loop until finished the transfer */
+	while (bs->rx_len) {
+		/* read status */
+		stat = bcm2835aux_rd(bs, BCM2835_AUX_SPI_STAT);
+
+		/* fill in tx fifo with remaining data */
+		if ((bs->tx_len) && (!(stat & BCM2835_AUX_SPI_STAT_TX_FULL))) {
+			bcm2835aux_wr_fifo(bs);
+			continue;
+		}
+
+		/* read data from fifo for both cases */
+		if (!(stat & BCM2835_AUX_SPI_STAT_RX_EMPTY)) {
+			bcm2835aux_rd_fifo(bs);
+			continue;
+		}
+		if (!(stat & BCM2835_AUX_SPI_STAT_BUSY)) {
+			bcm2835aux_rd_fifo(bs);
+			continue;
+		}
+
+		/* there is still data pending to read check the timeout */
+		if (bs->rx_len && time_after(jiffies, timeout)) {
+			dev_dbg_ratelimited(&spi->dev,
+					    "timeout period reached: jiffies: %lu remaining tx/rx: %d/%d - falling back to interrupt mode\n",
+					    jiffies - timeout,
+					    bs->tx_len, bs->rx_len);
+			/* forward to interrupt handler */
+			return __bcm2835aux_spi_transfer_one_irq(master,
+							       spi, tfr);
+		}
+	}
+
+	/* and return without waiting for completion */
+	return 0;
+}
+
+static int bcm2835aux_spi_transfer_one(struct spi_master *master,
+				       struct spi_device *spi,
+				       struct spi_transfer *tfr)
+{
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+	unsigned long spi_hz, clk_hz, speed;
+	unsigned long spi_used_hz;
+
+	/* calculate the registers to handle
+	 *
+	 * note that we use the variable data mode, which
+	 * is not optimal for longer transfers as we waste registers
+	 * resulting (potentially) in more interrupts when transferring
+	 * more than 12 bytes
+	 */
+
+	/* set clock */
+	spi_hz = tfr->speed_hz;
+	clk_hz = clk_get_rate(bs->clk);
+
+	if (spi_hz >= clk_hz / 2) {
+		speed = 0;
+	} else if (spi_hz) {
+		speed = DIV_ROUND_UP(clk_hz, 2 * spi_hz) - 1;
+		if (speed >  BCM2835_AUX_SPI_CNTL0_SPEED_MAX)
+			speed = BCM2835_AUX_SPI_CNTL0_SPEED_MAX;
+	} else { /* the slowest we can go */
+		speed = BCM2835_AUX_SPI_CNTL0_SPEED_MAX;
+	}
+	/* mask out old speed from previous spi_transfer */
+	bs->cntl[0] &= ~(BCM2835_AUX_SPI_CNTL0_SPEED);
+	/* set the new speed */
+	bs->cntl[0] |= speed << BCM2835_AUX_SPI_CNTL0_SPEED_SHIFT;
+
+	spi_used_hz = clk_hz / (2 * (speed + 1));
+
+	/* set transmit buffers and length */
+	bs->tx_buf = tfr->tx_buf;
+	bs->rx_buf = tfr->rx_buf;
+	bs->tx_len = tfr->len;
+	bs->rx_len = tfr->len;
+	bs->pending = 0;
+
+	/* Calculate the estimated time in us the transfer runs.  Note that
+	 * there are are 2 idle clocks cycles after each chunk getting
+	 * transferred - in our case the chunk size is 3 bytes, so we
+	 * approximate this by 9 cycles/byte.  This is used to find the number
+	 * of Hz per byte per polling limit.  E.g., we can transfer 1 byte in
+	 * 30 µs per 300,000 Hz of bus clock.
+	 */
+#define HZ_PER_BYTE ((9 * 1000000) / BCM2835_AUX_SPI_POLLING_LIMIT_US)
+	/* run in polling mode for short transfers */
+	if (tfr->len < spi_used_hz / HZ_PER_BYTE)
+		return bcm2835aux_spi_transfer_one_poll(master, spi, tfr);
+
+	/* run in interrupt mode for all others */
+	return bcm2835aux_spi_transfer_one_irq(master, spi, tfr);
+#undef HZ_PER_BYTE
+}
+
+static int bcm2835aux_spi_prepare_message(struct spi_master *master,
+					  struct spi_message *msg)
+{
+	struct spi_device *spi = msg->spi;
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+
+	bs->cntl[0] = BCM2835_AUX_SPI_CNTL0_ENABLE |
+		      BCM2835_AUX_SPI_CNTL0_VAR_WIDTH |
+		      BCM2835_AUX_SPI_CNTL0_MSBF_OUT;
+	bs->cntl[1] = BCM2835_AUX_SPI_CNTL1_MSBF_IN;
+
+	/* handle all the modes */
+	if (spi->mode & SPI_CPOL) {
+		bs->cntl[0] |= BCM2835_AUX_SPI_CNTL0_CPOL;
+		bs->cntl[0] |= BCM2835_AUX_SPI_CNTL0_OUT_RISING;
+	} else {
+		bs->cntl[0] |= BCM2835_AUX_SPI_CNTL0_IN_RISING;
+	}
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL1, bs->cntl[1]);
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL0, bs->cntl[0]);
+
+	return 0;
+}
+
+static int bcm2835aux_spi_unprepare_message(struct spi_master *master,
+					    struct spi_message *msg)
+{
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+
+	bcm2835aux_spi_reset_hw(bs);
+
+	return 0;
+}
+
+static void bcm2835aux_spi_handle_err(struct spi_master *master,
+				      struct spi_message *msg)
+{
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+
+	bcm2835aux_spi_reset_hw(bs);
+}
+
+static int bcm2835aux_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct bcm2835aux_spi *bs;
+	struct resource *res;
+	unsigned long clk_hz;
+	int err;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*bs));
+	if (!master) {
+		dev_err(&pdev->dev, "spi_alloc_master() failed\n");
+		return -ENOMEM;
+	}
+
+	platform_set_drvdata(pdev, master);
+	master->mode_bits = (SPI_CPOL | SPI_CS_HIGH | SPI_NO_CS);
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->num_chipselect = -1;
+	master->transfer_one = bcm2835aux_spi_transfer_one;
+	master->handle_err = bcm2835aux_spi_handle_err;
+	master->prepare_message = bcm2835aux_spi_prepare_message;
+	master->unprepare_message = bcm2835aux_spi_unprepare_message;
+	master->dev.of_node = pdev->dev.of_node;
+
+	bs = spi_master_get_devdata(master);
+
+	/* the main area */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	bs->regs = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(bs->regs)) {
+		err = PTR_ERR(bs->regs);
+		goto out_master_put;
+	}
+
+	bs->clk = devm_clk_get(&pdev->dev, NULL);
+	if ((!bs->clk) || (IS_ERR(bs->clk))) {
+		err = PTR_ERR(bs->clk);
+		dev_err(&pdev->dev, "could not get clk: %d\n", err);
+		goto out_master_put;
+	}
+
+	bs->irq = platform_get_irq(pdev, 0);
+	if (bs->irq <= 0) {
+		dev_err(&pdev->dev, "could not get IRQ: %d\n", bs->irq);
+		err = bs->irq ? bs->irq : -ENODEV;
+		goto out_master_put;
+	}
+
+	/* this also enables the HW block */
+	err = clk_prepare_enable(bs->clk);
+	if (err) {
+		dev_err(&pdev->dev, "could not prepare clock: %d\n", err);
+		goto out_master_put;
+	}
+
+	/* just checking if the clock returns a sane value */
+	clk_hz = clk_get_rate(bs->clk);
+	if (!clk_hz) {
+		dev_err(&pdev->dev, "clock returns 0 Hz\n");
+		err = -ENODEV;
+		goto out_clk_disable;
+	}
+
+	/* reset SPI-HW block */
+	bcm2835aux_spi_reset_hw(bs);
+
+	err = devm_request_irq(&pdev->dev, bs->irq,
+			       bcm2835aux_spi_interrupt,
+			       IRQF_SHARED,
+			       dev_name(&pdev->dev), master);
+	if (err) {
+		dev_err(&pdev->dev, "could not request IRQ: %d\n", err);
+		goto out_clk_disable;
+	}
+
+	err = devm_spi_register_master(&pdev->dev, master);
+	if (err) {
+		dev_err(&pdev->dev, "could not register SPI master: %d\n", err);
+		goto out_clk_disable;
+	}
+
+	return 0;
+
+out_clk_disable:
+	clk_disable_unprepare(bs->clk);
+out_master_put:
+	spi_master_put(master);
+	return err;
+}
+
+static int bcm2835aux_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+
+	bcm2835aux_spi_reset_hw(bs);
+
+	/* disable the HW block by releasing the clock */
+	clk_disable_unprepare(bs->clk);
+
+	return 0;
+}
+
+static const struct of_device_id bcm2835aux_spi_match[] = {
+	{ .compatible = "brcm,bcm2835-aux-spi", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, bcm2835aux_spi_match);
+
+static struct platform_driver bcm2835aux_spi_driver = {
+	.driver		= {
+		.name		= "spi-bcm2835aux",
+		.of_match_table	= bcm2835aux_spi_match,
+	},
+	.probe		= bcm2835aux_spi_probe,
+	.remove		= bcm2835aux_spi_remove,
+};
+module_platform_driver(bcm2835aux_spi_driver);
+
+MODULE_DESCRIPTION("SPI controller driver for Broadcom BCM2835 aux");
+MODULE_AUTHOR("Martin Sperl <kernel@martin.sperl.org>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-bcm63xx-hsspi.c b/drivers/spi/spi-bcm63xx-hsspi.c
new file mode 100644
index 0000000..c23849f
--- /dev/null
+++ b/drivers/spi/spi-bcm63xx-hsspi.c
@@ -0,0 +1,513 @@
+/*
+ * Broadcom BCM63XX High Speed SPI Controller driver
+ *
+ * Copyright 2000-2010 Broadcom Corporation
+ * Copyright 2012-2013 Jonas Gorski <jogo@openwrt.org>
+ *
+ * Licensed under the GNU/GPL. See COPYING for details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/clk.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/spi/spi.h>
+#include <linux/mutex.h>
+#include <linux/of.h>
+
+#define HSSPI_GLOBAL_CTRL_REG			0x0
+#define GLOBAL_CTRL_CS_POLARITY_SHIFT		0
+#define GLOBAL_CTRL_CS_POLARITY_MASK		0x000000ff
+#define GLOBAL_CTRL_PLL_CLK_CTRL_SHIFT		8
+#define GLOBAL_CTRL_PLL_CLK_CTRL_MASK		0x0000ff00
+#define GLOBAL_CTRL_CLK_GATE_SSOFF		BIT(16)
+#define GLOBAL_CTRL_CLK_POLARITY		BIT(17)
+#define GLOBAL_CTRL_MOSI_IDLE			BIT(18)
+
+#define HSSPI_GLOBAL_EXT_TRIGGER_REG		0x4
+
+#define HSSPI_INT_STATUS_REG			0x8
+#define HSSPI_INT_STATUS_MASKED_REG		0xc
+#define HSSPI_INT_MASK_REG			0x10
+
+#define HSSPI_PINGx_CMD_DONE(i)			BIT((i * 8) + 0)
+#define HSSPI_PINGx_RX_OVER(i)			BIT((i * 8) + 1)
+#define HSSPI_PINGx_TX_UNDER(i)			BIT((i * 8) + 2)
+#define HSSPI_PINGx_POLL_TIMEOUT(i)		BIT((i * 8) + 3)
+#define HSSPI_PINGx_CTRL_INVAL(i)		BIT((i * 8) + 4)
+
+#define HSSPI_INT_CLEAR_ALL			0xff001f1f
+
+#define HSSPI_PINGPONG_COMMAND_REG(x)		(0x80 + (x) * 0x40)
+#define PINGPONG_CMD_COMMAND_MASK		0xf
+#define PINGPONG_COMMAND_NOOP			0
+#define PINGPONG_COMMAND_START_NOW		1
+#define PINGPONG_COMMAND_START_TRIGGER		2
+#define PINGPONG_COMMAND_HALT			3
+#define PINGPONG_COMMAND_FLUSH			4
+#define PINGPONG_CMD_PROFILE_SHIFT		8
+#define PINGPONG_CMD_SS_SHIFT			12
+
+#define HSSPI_PINGPONG_STATUS_REG(x)		(0x84 + (x) * 0x40)
+
+#define HSSPI_PROFILE_CLK_CTRL_REG(x)		(0x100 + (x) * 0x20)
+#define CLK_CTRL_FREQ_CTRL_MASK			0x0000ffff
+#define CLK_CTRL_SPI_CLK_2X_SEL			BIT(14)
+#define CLK_CTRL_ACCUM_RST_ON_LOOP		BIT(15)
+
+#define HSSPI_PROFILE_SIGNAL_CTRL_REG(x)	(0x104 + (x) * 0x20)
+#define SIGNAL_CTRL_LATCH_RISING		BIT(12)
+#define SIGNAL_CTRL_LAUNCH_RISING		BIT(13)
+#define SIGNAL_CTRL_ASYNC_INPUT_PATH		BIT(16)
+
+#define HSSPI_PROFILE_MODE_CTRL_REG(x)		(0x108 + (x) * 0x20)
+#define MODE_CTRL_MULTIDATA_RD_STRT_SHIFT	8
+#define MODE_CTRL_MULTIDATA_WR_STRT_SHIFT	12
+#define MODE_CTRL_MULTIDATA_RD_SIZE_SHIFT	16
+#define MODE_CTRL_MULTIDATA_WR_SIZE_SHIFT	18
+#define MODE_CTRL_MODE_3WIRE			BIT(20)
+#define MODE_CTRL_PREPENDBYTE_CNT_SHIFT		24
+
+#define HSSPI_FIFO_REG(x)			(0x200 + (x) * 0x200)
+
+
+#define HSSPI_OP_MULTIBIT			BIT(11)
+#define HSSPI_OP_CODE_SHIFT			13
+#define HSSPI_OP_SLEEP				(0 << HSSPI_OP_CODE_SHIFT)
+#define HSSPI_OP_READ_WRITE			(1 << HSSPI_OP_CODE_SHIFT)
+#define HSSPI_OP_WRITE				(2 << HSSPI_OP_CODE_SHIFT)
+#define HSSPI_OP_READ				(3 << HSSPI_OP_CODE_SHIFT)
+#define HSSPI_OP_SETIRQ				(4 << HSSPI_OP_CODE_SHIFT)
+
+#define HSSPI_BUFFER_LEN			512
+#define HSSPI_OPCODE_LEN			2
+
+#define HSSPI_MAX_PREPEND_LEN			15
+
+#define HSSPI_MAX_SYNC_CLOCK			30000000
+
+#define HSSPI_SPI_MAX_CS			8
+#define HSSPI_BUS_NUM				1 /* 0 is legacy SPI */
+
+struct bcm63xx_hsspi {
+	struct completion done;
+	struct mutex bus_mutex;
+
+	struct platform_device *pdev;
+	struct clk *clk;
+	void __iomem *regs;
+	u8 __iomem *fifo;
+
+	u32 speed_hz;
+	u8 cs_polarity;
+};
+
+static void bcm63xx_hsspi_set_cs(struct bcm63xx_hsspi *bs, unsigned int cs,
+				 bool active)
+{
+	u32 reg;
+
+	mutex_lock(&bs->bus_mutex);
+	reg = __raw_readl(bs->regs + HSSPI_GLOBAL_CTRL_REG);
+
+	reg &= ~BIT(cs);
+	if (active == !(bs->cs_polarity & BIT(cs)))
+		reg |= BIT(cs);
+
+	__raw_writel(reg, bs->regs + HSSPI_GLOBAL_CTRL_REG);
+	mutex_unlock(&bs->bus_mutex);
+}
+
+static void bcm63xx_hsspi_set_clk(struct bcm63xx_hsspi *bs,
+				  struct spi_device *spi, int hz)
+{
+	unsigned int profile = spi->chip_select;
+	u32 reg;
+
+	reg = DIV_ROUND_UP(2048, DIV_ROUND_UP(bs->speed_hz, hz));
+	__raw_writel(CLK_CTRL_ACCUM_RST_ON_LOOP | reg,
+		     bs->regs + HSSPI_PROFILE_CLK_CTRL_REG(profile));
+
+	reg = __raw_readl(bs->regs + HSSPI_PROFILE_SIGNAL_CTRL_REG(profile));
+	if (hz > HSSPI_MAX_SYNC_CLOCK)
+		reg |= SIGNAL_CTRL_ASYNC_INPUT_PATH;
+	else
+		reg &= ~SIGNAL_CTRL_ASYNC_INPUT_PATH;
+	__raw_writel(reg, bs->regs + HSSPI_PROFILE_SIGNAL_CTRL_REG(profile));
+
+	mutex_lock(&bs->bus_mutex);
+	/* setup clock polarity */
+	reg = __raw_readl(bs->regs + HSSPI_GLOBAL_CTRL_REG);
+	reg &= ~GLOBAL_CTRL_CLK_POLARITY;
+	if (spi->mode & SPI_CPOL)
+		reg |= GLOBAL_CTRL_CLK_POLARITY;
+	__raw_writel(reg, bs->regs + HSSPI_GLOBAL_CTRL_REG);
+	mutex_unlock(&bs->bus_mutex);
+}
+
+static int bcm63xx_hsspi_do_txrx(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct bcm63xx_hsspi *bs = spi_master_get_devdata(spi->master);
+	unsigned int chip_select = spi->chip_select;
+	u16 opcode = 0;
+	int pending = t->len;
+	int step_size = HSSPI_BUFFER_LEN;
+	const u8 *tx = t->tx_buf;
+	u8 *rx = t->rx_buf;
+
+	bcm63xx_hsspi_set_clk(bs, spi, t->speed_hz);
+	bcm63xx_hsspi_set_cs(bs, spi->chip_select, true);
+
+	if (tx && rx)
+		opcode = HSSPI_OP_READ_WRITE;
+	else if (tx)
+		opcode = HSSPI_OP_WRITE;
+	else if (rx)
+		opcode = HSSPI_OP_READ;
+
+	if (opcode != HSSPI_OP_READ)
+		step_size -= HSSPI_OPCODE_LEN;
+
+	if ((opcode == HSSPI_OP_READ && t->rx_nbits == SPI_NBITS_DUAL) ||
+	    (opcode == HSSPI_OP_WRITE && t->tx_nbits == SPI_NBITS_DUAL))
+		opcode |= HSSPI_OP_MULTIBIT;
+
+	__raw_writel(1 << MODE_CTRL_MULTIDATA_WR_SIZE_SHIFT |
+		     1 << MODE_CTRL_MULTIDATA_RD_SIZE_SHIFT | 0xff,
+		     bs->regs + HSSPI_PROFILE_MODE_CTRL_REG(chip_select));
+
+	while (pending > 0) {
+		int curr_step = min_t(int, step_size, pending);
+
+		reinit_completion(&bs->done);
+		if (tx) {
+			memcpy_toio(bs->fifo + HSSPI_OPCODE_LEN, tx, curr_step);
+			tx += curr_step;
+		}
+
+		__raw_writew(opcode | curr_step, bs->fifo);
+
+		/* enable interrupt */
+		__raw_writel(HSSPI_PINGx_CMD_DONE(0),
+			     bs->regs + HSSPI_INT_MASK_REG);
+
+		/* start the transfer */
+		__raw_writel(!chip_select << PINGPONG_CMD_SS_SHIFT |
+			     chip_select << PINGPONG_CMD_PROFILE_SHIFT |
+			     PINGPONG_COMMAND_START_NOW,
+			     bs->regs + HSSPI_PINGPONG_COMMAND_REG(0));
+
+		if (wait_for_completion_timeout(&bs->done, HZ) == 0) {
+			dev_err(&bs->pdev->dev, "transfer timed out!\n");
+			return -ETIMEDOUT;
+		}
+
+		if (rx) {
+			memcpy_fromio(rx, bs->fifo, curr_step);
+			rx += curr_step;
+		}
+
+		pending -= curr_step;
+	}
+
+	return 0;
+}
+
+static int bcm63xx_hsspi_setup(struct spi_device *spi)
+{
+	struct bcm63xx_hsspi *bs = spi_master_get_devdata(spi->master);
+	u32 reg;
+
+	reg = __raw_readl(bs->regs +
+			  HSSPI_PROFILE_SIGNAL_CTRL_REG(spi->chip_select));
+	reg &= ~(SIGNAL_CTRL_LAUNCH_RISING | SIGNAL_CTRL_LATCH_RISING);
+	if (spi->mode & SPI_CPHA)
+		reg |= SIGNAL_CTRL_LAUNCH_RISING;
+	else
+		reg |= SIGNAL_CTRL_LATCH_RISING;
+	__raw_writel(reg, bs->regs +
+		     HSSPI_PROFILE_SIGNAL_CTRL_REG(spi->chip_select));
+
+	mutex_lock(&bs->bus_mutex);
+	reg = __raw_readl(bs->regs + HSSPI_GLOBAL_CTRL_REG);
+
+	/* only change actual polarities if there is no transfer */
+	if ((reg & GLOBAL_CTRL_CS_POLARITY_MASK) == bs->cs_polarity) {
+		if (spi->mode & SPI_CS_HIGH)
+			reg |= BIT(spi->chip_select);
+		else
+			reg &= ~BIT(spi->chip_select);
+		__raw_writel(reg, bs->regs + HSSPI_GLOBAL_CTRL_REG);
+	}
+
+	if (spi->mode & SPI_CS_HIGH)
+		bs->cs_polarity |= BIT(spi->chip_select);
+	else
+		bs->cs_polarity &= ~BIT(spi->chip_select);
+
+	mutex_unlock(&bs->bus_mutex);
+
+	return 0;
+}
+
+static int bcm63xx_hsspi_transfer_one(struct spi_master *master,
+				      struct spi_message *msg)
+{
+	struct bcm63xx_hsspi *bs = spi_master_get_devdata(master);
+	struct spi_transfer *t;
+	struct spi_device *spi = msg->spi;
+	int status = -EINVAL;
+	int dummy_cs;
+	u32 reg;
+
+	/* This controller does not support keeping CS active during idle.
+	 * To work around this, we use the following ugly hack:
+	 *
+	 * a. Invert the target chip select's polarity so it will be active.
+	 * b. Select a "dummy" chip select to use as the hardware target.
+	 * c. Invert the dummy chip select's polarity so it will be inactive
+	 *    during the actual transfers.
+	 * d. Tell the hardware to send to the dummy chip select. Thanks to
+	 *    the multiplexed nature of SPI the actual target will receive
+	 *    the transfer and we see its response.
+	 *
+	 * e. At the end restore the polarities again to their default values.
+	 */
+
+	dummy_cs = !spi->chip_select;
+	bcm63xx_hsspi_set_cs(bs, dummy_cs, true);
+
+	list_for_each_entry(t, &msg->transfers, transfer_list) {
+		status = bcm63xx_hsspi_do_txrx(spi, t);
+		if (status)
+			break;
+
+		msg->actual_length += t->len;
+
+		if (t->delay_usecs)
+			udelay(t->delay_usecs);
+
+		if (t->cs_change)
+			bcm63xx_hsspi_set_cs(bs, spi->chip_select, false);
+	}
+
+	mutex_lock(&bs->bus_mutex);
+	reg = __raw_readl(bs->regs + HSSPI_GLOBAL_CTRL_REG);
+	reg &= ~GLOBAL_CTRL_CS_POLARITY_MASK;
+	reg |= bs->cs_polarity;
+	__raw_writel(reg, bs->regs + HSSPI_GLOBAL_CTRL_REG);
+	mutex_unlock(&bs->bus_mutex);
+
+	msg->status = status;
+	spi_finalize_current_message(master);
+
+	return 0;
+}
+
+static irqreturn_t bcm63xx_hsspi_interrupt(int irq, void *dev_id)
+{
+	struct bcm63xx_hsspi *bs = (struct bcm63xx_hsspi *)dev_id;
+
+	if (__raw_readl(bs->regs + HSSPI_INT_STATUS_MASKED_REG) == 0)
+		return IRQ_NONE;
+
+	__raw_writel(HSSPI_INT_CLEAR_ALL, bs->regs + HSSPI_INT_STATUS_REG);
+	__raw_writel(0, bs->regs + HSSPI_INT_MASK_REG);
+
+	complete(&bs->done);
+
+	return IRQ_HANDLED;
+}
+
+static int bcm63xx_hsspi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct bcm63xx_hsspi *bs;
+	struct resource *res_mem;
+	void __iomem *regs;
+	struct device *dev = &pdev->dev;
+	struct clk *clk;
+	int irq, ret;
+	u32 reg, rate, num_cs = HSSPI_SPI_MAX_CS;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(dev, "no irq: %d\n", irq);
+		return irq;
+	}
+
+	res_mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	regs = devm_ioremap_resource(dev, res_mem);
+	if (IS_ERR(regs))
+		return PTR_ERR(regs);
+
+	clk = devm_clk_get(dev, "hsspi");
+
+	if (IS_ERR(clk))
+		return PTR_ERR(clk);
+
+	ret = clk_prepare_enable(clk);
+	if (ret)
+		return ret;
+
+	rate = clk_get_rate(clk);
+	if (!rate) {
+		struct clk *pll_clk = devm_clk_get(dev, "pll");
+
+		if (IS_ERR(pll_clk)) {
+			ret = PTR_ERR(pll_clk);
+			goto out_disable_clk;
+		}
+
+		ret = clk_prepare_enable(pll_clk);
+		if (ret)
+			goto out_disable_clk;
+
+		rate = clk_get_rate(pll_clk);
+		clk_disable_unprepare(pll_clk);
+		if (!rate) {
+			ret = -EINVAL;
+			goto out_disable_clk;
+		}
+	}
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*bs));
+	if (!master) {
+		ret = -ENOMEM;
+		goto out_disable_clk;
+	}
+
+	bs = spi_master_get_devdata(master);
+	bs->pdev = pdev;
+	bs->clk = clk;
+	bs->regs = regs;
+	bs->speed_hz = rate;
+	bs->fifo = (u8 __iomem *)(bs->regs + HSSPI_FIFO_REG(0));
+
+	mutex_init(&bs->bus_mutex);
+	init_completion(&bs->done);
+
+	master->dev.of_node = dev->of_node;
+	if (!dev->of_node)
+		master->bus_num = HSSPI_BUS_NUM;
+
+	of_property_read_u32(dev->of_node, "num-cs", &num_cs);
+	if (num_cs > 8) {
+		dev_warn(dev, "unsupported number of cs (%i), reducing to 8\n",
+			 num_cs);
+		num_cs = HSSPI_SPI_MAX_CS;
+	}
+	master->num_chipselect = num_cs;
+	master->setup = bcm63xx_hsspi_setup;
+	master->transfer_one_message = bcm63xx_hsspi_transfer_one;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH |
+			    SPI_RX_DUAL | SPI_TX_DUAL;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->auto_runtime_pm = true;
+
+	platform_set_drvdata(pdev, master);
+
+	/* Initialize the hardware */
+	__raw_writel(0, bs->regs + HSSPI_INT_MASK_REG);
+
+	/* clean up any pending interrupts */
+	__raw_writel(HSSPI_INT_CLEAR_ALL, bs->regs + HSSPI_INT_STATUS_REG);
+
+	/* read out default CS polarities */
+	reg = __raw_readl(bs->regs + HSSPI_GLOBAL_CTRL_REG);
+	bs->cs_polarity = reg & GLOBAL_CTRL_CS_POLARITY_MASK;
+	__raw_writel(reg | GLOBAL_CTRL_CLK_GATE_SSOFF,
+		     bs->regs + HSSPI_GLOBAL_CTRL_REG);
+
+	ret = devm_request_irq(dev, irq, bcm63xx_hsspi_interrupt, IRQF_SHARED,
+			       pdev->name, bs);
+
+	if (ret)
+		goto out_put_master;
+
+	/* register and we are done */
+	ret = devm_spi_register_master(dev, master);
+	if (ret)
+		goto out_put_master;
+
+	return 0;
+
+out_put_master:
+	spi_master_put(master);
+out_disable_clk:
+	clk_disable_unprepare(clk);
+	return ret;
+}
+
+
+static int bcm63xx_hsspi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct bcm63xx_hsspi *bs = spi_master_get_devdata(master);
+
+	/* reset the hardware and block queue progress */
+	__raw_writel(0, bs->regs + HSSPI_INT_MASK_REG);
+	clk_disable_unprepare(bs->clk);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int bcm63xx_hsspi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct bcm63xx_hsspi *bs = spi_master_get_devdata(master);
+
+	spi_master_suspend(master);
+	clk_disable_unprepare(bs->clk);
+
+	return 0;
+}
+
+static int bcm63xx_hsspi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct bcm63xx_hsspi *bs = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(bs->clk);
+	if (ret)
+		return ret;
+
+	spi_master_resume(master);
+
+	return 0;
+}
+#endif
+
+static SIMPLE_DEV_PM_OPS(bcm63xx_hsspi_pm_ops, bcm63xx_hsspi_suspend,
+			 bcm63xx_hsspi_resume);
+
+static const struct of_device_id bcm63xx_hsspi_of_match[] = {
+	{ .compatible = "brcm,bcm6328-hsspi", },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, bcm63xx_hsspi_of_match);
+
+static struct platform_driver bcm63xx_hsspi_driver = {
+	.driver = {
+		.name	= "bcm63xx-hsspi",
+		.pm	= &bcm63xx_hsspi_pm_ops,
+		.of_match_table = bcm63xx_hsspi_of_match,
+	},
+	.probe		= bcm63xx_hsspi_probe,
+	.remove		= bcm63xx_hsspi_remove,
+};
+
+module_platform_driver(bcm63xx_hsspi_driver);
+
+MODULE_ALIAS("platform:bcm63xx_hsspi");
+MODULE_DESCRIPTION("Broadcom BCM63xx High Speed SPI Controller driver");
+MODULE_AUTHOR("Jonas Gorski <jogo@openwrt.org>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-bcm63xx.c b/drivers/spi/spi-bcm63xx.c
new file mode 100644
index 0000000..bfe5754
--- /dev/null
+++ b/drivers/spi/spi-bcm63xx.c
@@ -0,0 +1,678 @@
+/*
+ * Broadcom BCM63xx SPI controller support
+ *
+ * Copyright (C) 2009-2012 Florian Fainelli <florian@openwrt.org>
+ * Copyright (C) 2010 Tanguy Bouzeloc <tanguy.bouzeloc@efixo.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/clk.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/spi/spi.h>
+#include <linux/completion.h>
+#include <linux/err.h>
+#include <linux/pm_runtime.h>
+#include <linux/of.h>
+
+/* BCM 6338/6348 SPI core */
+#define SPI_6348_RSET_SIZE		64
+#define SPI_6348_CMD			0x00	/* 16-bits register */
+#define SPI_6348_INT_STATUS		0x02
+#define SPI_6348_INT_MASK_ST		0x03
+#define SPI_6348_INT_MASK		0x04
+#define SPI_6348_ST			0x05
+#define SPI_6348_CLK_CFG		0x06
+#define SPI_6348_FILL_BYTE		0x07
+#define SPI_6348_MSG_TAIL		0x09
+#define SPI_6348_RX_TAIL		0x0b
+#define SPI_6348_MSG_CTL		0x40	/* 8-bits register */
+#define SPI_6348_MSG_CTL_WIDTH		8
+#define SPI_6348_MSG_DATA		0x41
+#define SPI_6348_MSG_DATA_SIZE		0x3f
+#define SPI_6348_RX_DATA		0x80
+#define SPI_6348_RX_DATA_SIZE		0x3f
+
+/* BCM 3368/6358/6262/6368 SPI core */
+#define SPI_6358_RSET_SIZE		1804
+#define SPI_6358_MSG_CTL		0x00	/* 16-bits register */
+#define SPI_6358_MSG_CTL_WIDTH		16
+#define SPI_6358_MSG_DATA		0x02
+#define SPI_6358_MSG_DATA_SIZE		0x21e
+#define SPI_6358_RX_DATA		0x400
+#define SPI_6358_RX_DATA_SIZE		0x220
+#define SPI_6358_CMD			0x700	/* 16-bits register */
+#define SPI_6358_INT_STATUS		0x702
+#define SPI_6358_INT_MASK_ST		0x703
+#define SPI_6358_INT_MASK		0x704
+#define SPI_6358_ST			0x705
+#define SPI_6358_CLK_CFG		0x706
+#define SPI_6358_FILL_BYTE		0x707
+#define SPI_6358_MSG_TAIL		0x709
+#define SPI_6358_RX_TAIL		0x70B
+
+/* Shared SPI definitions */
+
+/* Message configuration */
+#define SPI_FD_RW			0x00
+#define SPI_HD_W			0x01
+#define SPI_HD_R			0x02
+#define SPI_BYTE_CNT_SHIFT		0
+#define SPI_6348_MSG_TYPE_SHIFT		6
+#define SPI_6358_MSG_TYPE_SHIFT		14
+
+/* Command */
+#define SPI_CMD_NOOP			0x00
+#define SPI_CMD_SOFT_RESET		0x01
+#define SPI_CMD_HARD_RESET		0x02
+#define SPI_CMD_START_IMMEDIATE		0x03
+#define SPI_CMD_COMMAND_SHIFT		0
+#define SPI_CMD_COMMAND_MASK		0x000f
+#define SPI_CMD_DEVICE_ID_SHIFT		4
+#define SPI_CMD_PREPEND_BYTE_CNT_SHIFT	8
+#define SPI_CMD_ONE_BYTE_SHIFT		11
+#define SPI_CMD_ONE_WIRE_SHIFT		12
+#define SPI_DEV_ID_0			0
+#define SPI_DEV_ID_1			1
+#define SPI_DEV_ID_2			2
+#define SPI_DEV_ID_3			3
+
+/* Interrupt mask */
+#define SPI_INTR_CMD_DONE		0x01
+#define SPI_INTR_RX_OVERFLOW		0x02
+#define SPI_INTR_TX_UNDERFLOW		0x04
+#define SPI_INTR_TX_OVERFLOW		0x08
+#define SPI_INTR_RX_UNDERFLOW		0x10
+#define SPI_INTR_CLEAR_ALL		0x1f
+
+/* Status */
+#define SPI_RX_EMPTY			0x02
+#define SPI_CMD_BUSY			0x04
+#define SPI_SERIAL_BUSY			0x08
+
+/* Clock configuration */
+#define SPI_CLK_20MHZ			0x00
+#define SPI_CLK_0_391MHZ		0x01
+#define SPI_CLK_0_781MHZ		0x02	/* default */
+#define SPI_CLK_1_563MHZ		0x03
+#define SPI_CLK_3_125MHZ		0x04
+#define SPI_CLK_6_250MHZ		0x05
+#define SPI_CLK_12_50MHZ		0x06
+#define SPI_CLK_MASK			0x07
+#define SPI_SSOFFTIME_MASK		0x38
+#define SPI_SSOFFTIME_SHIFT		3
+#define SPI_BYTE_SWAP			0x80
+
+enum bcm63xx_regs_spi {
+	SPI_CMD,
+	SPI_INT_STATUS,
+	SPI_INT_MASK_ST,
+	SPI_INT_MASK,
+	SPI_ST,
+	SPI_CLK_CFG,
+	SPI_FILL_BYTE,
+	SPI_MSG_TAIL,
+	SPI_RX_TAIL,
+	SPI_MSG_CTL,
+	SPI_MSG_DATA,
+	SPI_RX_DATA,
+	SPI_MSG_TYPE_SHIFT,
+	SPI_MSG_CTL_WIDTH,
+	SPI_MSG_DATA_SIZE,
+};
+
+#define BCM63XX_SPI_MAX_PREPEND		15
+
+#define BCM63XX_SPI_MAX_CS		8
+#define BCM63XX_SPI_BUS_NUM		0
+
+struct bcm63xx_spi {
+	struct completion	done;
+
+	void __iomem		*regs;
+	int			irq;
+
+	/* Platform data */
+	const unsigned long	*reg_offsets;
+	unsigned int		fifo_size;
+	unsigned int		msg_type_shift;
+	unsigned int		msg_ctl_width;
+
+	/* data iomem */
+	u8 __iomem		*tx_io;
+	const u8 __iomem	*rx_io;
+
+	struct clk		*clk;
+	struct platform_device	*pdev;
+};
+
+static inline u8 bcm_spi_readb(struct bcm63xx_spi *bs,
+			       unsigned int offset)
+{
+	return readb(bs->regs + bs->reg_offsets[offset]);
+}
+
+static inline u16 bcm_spi_readw(struct bcm63xx_spi *bs,
+				unsigned int offset)
+{
+#ifdef CONFIG_CPU_BIG_ENDIAN
+	return ioread16be(bs->regs + bs->reg_offsets[offset]);
+#else
+	return readw(bs->regs + bs->reg_offsets[offset]);
+#endif
+}
+
+static inline void bcm_spi_writeb(struct bcm63xx_spi *bs,
+				  u8 value, unsigned int offset)
+{
+	writeb(value, bs->regs + bs->reg_offsets[offset]);
+}
+
+static inline void bcm_spi_writew(struct bcm63xx_spi *bs,
+				  u16 value, unsigned int offset)
+{
+#ifdef CONFIG_CPU_BIG_ENDIAN
+	iowrite16be(value, bs->regs + bs->reg_offsets[offset]);
+#else
+	writew(value, bs->regs + bs->reg_offsets[offset]);
+#endif
+}
+
+static const unsigned int bcm63xx_spi_freq_table[SPI_CLK_MASK][2] = {
+	{ 20000000, SPI_CLK_20MHZ },
+	{ 12500000, SPI_CLK_12_50MHZ },
+	{  6250000, SPI_CLK_6_250MHZ },
+	{  3125000, SPI_CLK_3_125MHZ },
+	{  1563000, SPI_CLK_1_563MHZ },
+	{   781000, SPI_CLK_0_781MHZ },
+	{   391000, SPI_CLK_0_391MHZ }
+};
+
+static void bcm63xx_spi_setup_transfer(struct spi_device *spi,
+				      struct spi_transfer *t)
+{
+	struct bcm63xx_spi *bs = spi_master_get_devdata(spi->master);
+	u8 clk_cfg, reg;
+	int i;
+
+	/* Default to lowest clock configuration */
+	clk_cfg = SPI_CLK_0_391MHZ;
+
+	/* Find the closest clock configuration */
+	for (i = 0; i < SPI_CLK_MASK; i++) {
+		if (t->speed_hz >= bcm63xx_spi_freq_table[i][0]) {
+			clk_cfg = bcm63xx_spi_freq_table[i][1];
+			break;
+		}
+	}
+
+	/* clear existing clock configuration bits of the register */
+	reg = bcm_spi_readb(bs, SPI_CLK_CFG);
+	reg &= ~SPI_CLK_MASK;
+	reg |= clk_cfg;
+
+	bcm_spi_writeb(bs, reg, SPI_CLK_CFG);
+	dev_dbg(&spi->dev, "Setting clock register to %02x (hz %d)\n",
+		clk_cfg, t->speed_hz);
+}
+
+/* the spi->mode bits understood by this driver: */
+#define MODEBITS (SPI_CPOL | SPI_CPHA)
+
+static int bcm63xx_txrx_bufs(struct spi_device *spi, struct spi_transfer *first,
+				unsigned int num_transfers)
+{
+	struct bcm63xx_spi *bs = spi_master_get_devdata(spi->master);
+	u16 msg_ctl;
+	u16 cmd;
+	unsigned int i, timeout = 0, prepend_len = 0, len = 0;
+	struct spi_transfer *t = first;
+	bool do_rx = false;
+	bool do_tx = false;
+
+	/* Disable the CMD_DONE interrupt */
+	bcm_spi_writeb(bs, 0, SPI_INT_MASK);
+
+	dev_dbg(&spi->dev, "txrx: tx %p, rx %p, len %d\n",
+		t->tx_buf, t->rx_buf, t->len);
+
+	if (num_transfers > 1 && t->tx_buf && t->len <= BCM63XX_SPI_MAX_PREPEND)
+		prepend_len = t->len;
+
+	/* prepare the buffer */
+	for (i = 0; i < num_transfers; i++) {
+		if (t->tx_buf) {
+			do_tx = true;
+			memcpy_toio(bs->tx_io + len, t->tx_buf, t->len);
+
+			/* don't prepend more than one tx */
+			if (t != first)
+				prepend_len = 0;
+		}
+
+		if (t->rx_buf) {
+			do_rx = true;
+			/* prepend is half-duplex write only */
+			if (t == first)
+				prepend_len = 0;
+		}
+
+		len += t->len;
+
+		t = list_entry(t->transfer_list.next, struct spi_transfer,
+			       transfer_list);
+	}
+
+	reinit_completion(&bs->done);
+
+	/* Fill in the Message control register */
+	msg_ctl = (len << SPI_BYTE_CNT_SHIFT);
+
+	if (do_rx && do_tx && prepend_len == 0)
+		msg_ctl |= (SPI_FD_RW << bs->msg_type_shift);
+	else if (do_rx)
+		msg_ctl |= (SPI_HD_R << bs->msg_type_shift);
+	else if (do_tx)
+		msg_ctl |= (SPI_HD_W << bs->msg_type_shift);
+
+	switch (bs->msg_ctl_width) {
+	case 8:
+		bcm_spi_writeb(bs, msg_ctl, SPI_MSG_CTL);
+		break;
+	case 16:
+		bcm_spi_writew(bs, msg_ctl, SPI_MSG_CTL);
+		break;
+	}
+
+	/* Issue the transfer */
+	cmd = SPI_CMD_START_IMMEDIATE;
+	cmd |= (prepend_len << SPI_CMD_PREPEND_BYTE_CNT_SHIFT);
+	cmd |= (spi->chip_select << SPI_CMD_DEVICE_ID_SHIFT);
+	bcm_spi_writew(bs, cmd, SPI_CMD);
+
+	/* Enable the CMD_DONE interrupt */
+	bcm_spi_writeb(bs, SPI_INTR_CMD_DONE, SPI_INT_MASK);
+
+	timeout = wait_for_completion_timeout(&bs->done, HZ);
+	if (!timeout)
+		return -ETIMEDOUT;
+
+	if (!do_rx)
+		return 0;
+
+	len = 0;
+	t = first;
+	/* Read out all the data */
+	for (i = 0; i < num_transfers; i++) {
+		if (t->rx_buf)
+			memcpy_fromio(t->rx_buf, bs->rx_io + len, t->len);
+
+		if (t != first || prepend_len == 0)
+			len += t->len;
+
+		t = list_entry(t->transfer_list.next, struct spi_transfer,
+			       transfer_list);
+	}
+
+	return 0;
+}
+
+static int bcm63xx_spi_transfer_one(struct spi_master *master,
+					struct spi_message *m)
+{
+	struct bcm63xx_spi *bs = spi_master_get_devdata(master);
+	struct spi_transfer *t, *first = NULL;
+	struct spi_device *spi = m->spi;
+	int status = 0;
+	unsigned int n_transfers = 0, total_len = 0;
+	bool can_use_prepend = false;
+
+	/*
+	 * This SPI controller does not support keeping CS active after a
+	 * transfer.
+	 * Work around this by merging as many transfers we can into one big
+	 * full-duplex transfers.
+	 */
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if (!first)
+			first = t;
+
+		n_transfers++;
+		total_len += t->len;
+
+		if (n_transfers == 2 && !first->rx_buf && !t->tx_buf &&
+		    first->len <= BCM63XX_SPI_MAX_PREPEND)
+			can_use_prepend = true;
+		else if (can_use_prepend && t->tx_buf)
+			can_use_prepend = false;
+
+		/* we can only transfer one fifo worth of data */
+		if ((can_use_prepend &&
+		     total_len > (bs->fifo_size + BCM63XX_SPI_MAX_PREPEND)) ||
+		    (!can_use_prepend && total_len > bs->fifo_size)) {
+			dev_err(&spi->dev, "unable to do transfers larger than FIFO size (%i > %i)\n",
+				total_len, bs->fifo_size);
+			status = -EINVAL;
+			goto exit;
+		}
+
+		/* all combined transfers have to have the same speed */
+		if (t->speed_hz != first->speed_hz) {
+			dev_err(&spi->dev, "unable to change speed between transfers\n");
+			status = -EINVAL;
+			goto exit;
+		}
+
+		/* CS will be deasserted directly after transfer */
+		if (t->delay_usecs) {
+			dev_err(&spi->dev, "unable to keep CS asserted after transfer\n");
+			status = -EINVAL;
+			goto exit;
+		}
+
+		if (t->cs_change ||
+		    list_is_last(&t->transfer_list, &m->transfers)) {
+			/* configure adapter for a new transfer */
+			bcm63xx_spi_setup_transfer(spi, first);
+
+			/* send the data */
+			status = bcm63xx_txrx_bufs(spi, first, n_transfers);
+			if (status)
+				goto exit;
+
+			m->actual_length += total_len;
+
+			first = NULL;
+			n_transfers = 0;
+			total_len = 0;
+			can_use_prepend = false;
+		}
+	}
+exit:
+	m->status = status;
+	spi_finalize_current_message(master);
+
+	return 0;
+}
+
+/* This driver supports single master mode only. Hence
+ * CMD_DONE is the only interrupt we care about
+ */
+static irqreturn_t bcm63xx_spi_interrupt(int irq, void *dev_id)
+{
+	struct spi_master *master = (struct spi_master *)dev_id;
+	struct bcm63xx_spi *bs = spi_master_get_devdata(master);
+	u8 intr;
+
+	/* Read interupts and clear them immediately */
+	intr = bcm_spi_readb(bs, SPI_INT_STATUS);
+	bcm_spi_writeb(bs, SPI_INTR_CLEAR_ALL, SPI_INT_STATUS);
+	bcm_spi_writeb(bs, 0, SPI_INT_MASK);
+
+	/* A transfer completed */
+	if (intr & SPI_INTR_CMD_DONE)
+		complete(&bs->done);
+
+	return IRQ_HANDLED;
+}
+
+static size_t bcm63xx_spi_max_length(struct spi_device *spi)
+{
+	struct bcm63xx_spi *bs = spi_master_get_devdata(spi->master);
+
+	return bs->fifo_size;
+}
+
+static const unsigned long bcm6348_spi_reg_offsets[] = {
+	[SPI_CMD]		= SPI_6348_CMD,
+	[SPI_INT_STATUS]	= SPI_6348_INT_STATUS,
+	[SPI_INT_MASK_ST]	= SPI_6348_INT_MASK_ST,
+	[SPI_INT_MASK]		= SPI_6348_INT_MASK,
+	[SPI_ST]		= SPI_6348_ST,
+	[SPI_CLK_CFG]		= SPI_6348_CLK_CFG,
+	[SPI_FILL_BYTE]		= SPI_6348_FILL_BYTE,
+	[SPI_MSG_TAIL]		= SPI_6348_MSG_TAIL,
+	[SPI_RX_TAIL]		= SPI_6348_RX_TAIL,
+	[SPI_MSG_CTL]		= SPI_6348_MSG_CTL,
+	[SPI_MSG_DATA]		= SPI_6348_MSG_DATA,
+	[SPI_RX_DATA]		= SPI_6348_RX_DATA,
+	[SPI_MSG_TYPE_SHIFT]	= SPI_6348_MSG_TYPE_SHIFT,
+	[SPI_MSG_CTL_WIDTH]	= SPI_6348_MSG_CTL_WIDTH,
+	[SPI_MSG_DATA_SIZE]	= SPI_6348_MSG_DATA_SIZE,
+};
+
+static const unsigned long bcm6358_spi_reg_offsets[] = {
+	[SPI_CMD]		= SPI_6358_CMD,
+	[SPI_INT_STATUS]	= SPI_6358_INT_STATUS,
+	[SPI_INT_MASK_ST]	= SPI_6358_INT_MASK_ST,
+	[SPI_INT_MASK]		= SPI_6358_INT_MASK,
+	[SPI_ST]		= SPI_6358_ST,
+	[SPI_CLK_CFG]		= SPI_6358_CLK_CFG,
+	[SPI_FILL_BYTE]		= SPI_6358_FILL_BYTE,
+	[SPI_MSG_TAIL]		= SPI_6358_MSG_TAIL,
+	[SPI_RX_TAIL]		= SPI_6358_RX_TAIL,
+	[SPI_MSG_CTL]		= SPI_6358_MSG_CTL,
+	[SPI_MSG_DATA]		= SPI_6358_MSG_DATA,
+	[SPI_RX_DATA]		= SPI_6358_RX_DATA,
+	[SPI_MSG_TYPE_SHIFT]	= SPI_6358_MSG_TYPE_SHIFT,
+	[SPI_MSG_CTL_WIDTH]	= SPI_6358_MSG_CTL_WIDTH,
+	[SPI_MSG_DATA_SIZE]	= SPI_6358_MSG_DATA_SIZE,
+};
+
+static const struct platform_device_id bcm63xx_spi_dev_match[] = {
+	{
+		.name = "bcm6348-spi",
+		.driver_data = (unsigned long)bcm6348_spi_reg_offsets,
+	},
+	{
+		.name = "bcm6358-spi",
+		.driver_data = (unsigned long)bcm6358_spi_reg_offsets,
+	},
+	{
+	},
+};
+
+static const struct of_device_id bcm63xx_spi_of_match[] = {
+	{ .compatible = "brcm,bcm6348-spi", .data = &bcm6348_spi_reg_offsets },
+	{ .compatible = "brcm,bcm6358-spi", .data = &bcm6358_spi_reg_offsets },
+	{ },
+};
+
+static int bcm63xx_spi_probe(struct platform_device *pdev)
+{
+	struct resource *r;
+	const unsigned long *bcm63xx_spireg;
+	struct device *dev = &pdev->dev;
+	int irq, bus_num;
+	struct spi_master *master;
+	struct clk *clk;
+	struct bcm63xx_spi *bs;
+	int ret;
+	u32 num_cs = BCM63XX_SPI_MAX_CS;
+
+	if (dev->of_node) {
+		const struct of_device_id *match;
+
+		match = of_match_node(bcm63xx_spi_of_match, dev->of_node);
+		if (!match)
+			return -EINVAL;
+		bcm63xx_spireg = match->data;
+
+		of_property_read_u32(dev->of_node, "num-cs", &num_cs);
+		if (num_cs > BCM63XX_SPI_MAX_CS) {
+			dev_warn(dev, "unsupported number of cs (%i), reducing to 8\n",
+				 num_cs);
+			num_cs = BCM63XX_SPI_MAX_CS;
+		}
+
+		bus_num = -1;
+	} else if (pdev->id_entry->driver_data) {
+		const struct platform_device_id *match = pdev->id_entry;
+
+		bcm63xx_spireg = (const unsigned long *)match->driver_data;
+		bus_num = BCM63XX_SPI_BUS_NUM;
+	} else {
+		return -EINVAL;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(dev, "no irq: %d\n", irq);
+		return irq;
+	}
+
+	clk = devm_clk_get(dev, "spi");
+	if (IS_ERR(clk)) {
+		dev_err(dev, "no clock for device\n");
+		return PTR_ERR(clk);
+	}
+
+	master = spi_alloc_master(dev, sizeof(*bs));
+	if (!master) {
+		dev_err(dev, "out of memory\n");
+		return -ENOMEM;
+	}
+
+	bs = spi_master_get_devdata(master);
+	init_completion(&bs->done);
+
+	platform_set_drvdata(pdev, master);
+	bs->pdev = pdev;
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	bs->regs = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(bs->regs)) {
+		ret = PTR_ERR(bs->regs);
+		goto out_err;
+	}
+
+	bs->irq = irq;
+	bs->clk = clk;
+	bs->reg_offsets = bcm63xx_spireg;
+	bs->fifo_size = bs->reg_offsets[SPI_MSG_DATA_SIZE];
+
+	ret = devm_request_irq(&pdev->dev, irq, bcm63xx_spi_interrupt, 0,
+							pdev->name, master);
+	if (ret) {
+		dev_err(dev, "unable to request irq\n");
+		goto out_err;
+	}
+
+	master->dev.of_node = dev->of_node;
+	master->bus_num = bus_num;
+	master->num_chipselect = num_cs;
+	master->transfer_one_message = bcm63xx_spi_transfer_one;
+	master->mode_bits = MODEBITS;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->max_transfer_size = bcm63xx_spi_max_length;
+	master->max_message_size = bcm63xx_spi_max_length;
+	master->auto_runtime_pm = true;
+	bs->msg_type_shift = bs->reg_offsets[SPI_MSG_TYPE_SHIFT];
+	bs->msg_ctl_width = bs->reg_offsets[SPI_MSG_CTL_WIDTH];
+	bs->tx_io = (u8 *)(bs->regs + bs->reg_offsets[SPI_MSG_DATA]);
+	bs->rx_io = (const u8 *)(bs->regs + bs->reg_offsets[SPI_RX_DATA]);
+
+	/* Initialize hardware */
+	ret = clk_prepare_enable(bs->clk);
+	if (ret)
+		goto out_err;
+
+	bcm_spi_writeb(bs, SPI_INTR_CLEAR_ALL, SPI_INT_STATUS);
+
+	/* register and we are done */
+	ret = devm_spi_register_master(dev, master);
+	if (ret) {
+		dev_err(dev, "spi register failed\n");
+		goto out_clk_disable;
+	}
+
+	dev_info(dev, "at %pr (irq %d, FIFOs size %d)\n",
+		 r, irq, bs->fifo_size);
+
+	return 0;
+
+out_clk_disable:
+	clk_disable_unprepare(clk);
+out_err:
+	spi_master_put(master);
+	return ret;
+}
+
+static int bcm63xx_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct bcm63xx_spi *bs = spi_master_get_devdata(master);
+
+	/* reset spi block */
+	bcm_spi_writeb(bs, 0, SPI_INT_MASK);
+
+	/* HW shutdown */
+	clk_disable_unprepare(bs->clk);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int bcm63xx_spi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct bcm63xx_spi *bs = spi_master_get_devdata(master);
+
+	spi_master_suspend(master);
+
+	clk_disable_unprepare(bs->clk);
+
+	return 0;
+}
+
+static int bcm63xx_spi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct bcm63xx_spi *bs = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(bs->clk);
+	if (ret)
+		return ret;
+
+	spi_master_resume(master);
+
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops bcm63xx_spi_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(bcm63xx_spi_suspend, bcm63xx_spi_resume)
+};
+
+static struct platform_driver bcm63xx_spi_driver = {
+	.driver = {
+		.name	= "bcm63xx-spi",
+		.pm	= &bcm63xx_spi_pm_ops,
+		.of_match_table = bcm63xx_spi_of_match,
+	},
+	.id_table	= bcm63xx_spi_dev_match,
+	.probe		= bcm63xx_spi_probe,
+	.remove		= bcm63xx_spi_remove,
+};
+
+module_platform_driver(bcm63xx_spi_driver);
+
+MODULE_ALIAS("platform:bcm63xx_spi");
+MODULE_AUTHOR("Florian Fainelli <florian@openwrt.org>");
+MODULE_AUTHOR("Tanguy Bouzeloc <tanguy.bouzeloc@efixo.com>");
+MODULE_DESCRIPTION("Broadcom BCM63xx SPI Controller driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-bitbang-txrx.h b/drivers/spi/spi-bitbang-txrx.h
new file mode 100644
index 0000000..ae61d72
--- /dev/null
+++ b/drivers/spi/spi-bitbang-txrx.h
@@ -0,0 +1,108 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Mix this utility code with some glue code to get one of several types of
+ * simple SPI master driver.  Two do polled word-at-a-time I/O:
+ *
+ *   -	GPIO/parport bitbangers.  Provide chipselect() and txrx_word[](),
+ *	expanding the per-word routines from the inline templates below.
+ *
+ *   -	Drivers for controllers resembling bare shift registers.  Provide
+ *	chipselect() and txrx_word[](), with custom setup()/cleanup() methods
+ *	that use your controller's clock and chipselect registers.
+ *
+ * Some hardware works well with requests at spi_transfer scope:
+ *
+ *   -	Drivers leveraging smarter hardware, with fifos or DMA; or for half
+ *	duplex (MicroWire) controllers.  Provide chipselect() and txrx_bufs(),
+ *	and custom setup()/cleanup() methods.
+ */
+
+/*
+ * The code that knows what GPIO pins do what should have declared four
+ * functions, ideally as inlines, before including this header:
+ *
+ *  void setsck(struct spi_device *, int is_on);
+ *  void setmosi(struct spi_device *, int is_on);
+ *  int getmiso(struct spi_device *);
+ *  void spidelay(unsigned);
+ *
+ * setsck()'s is_on parameter is a zero/nonzero boolean.
+ *
+ * setmosi()'s is_on parameter is a zero/nonzero boolean.
+ *
+ * getmiso() is required to return 0 or 1 only. Any other value is invalid
+ * and will result in improper operation.
+ *
+ * A non-inlined routine would call bitbang_txrx_*() routines.  The
+ * main loop could easily compile down to a handful of instructions,
+ * especially if the delay is a NOP (to run at peak speed).
+ *
+ * Since this is software, the timings may not be exactly what your board's
+ * chips need ... there may be several reasons you'd need to tweak timings
+ * in these routines, not just to make it faster or slower to match a
+ * particular CPU clock rate.
+ */
+
+static inline u32
+bitbang_txrx_be_cpha0(struct spi_device *spi,
+		unsigned nsecs, unsigned cpol, unsigned flags,
+		u32 word, u8 bits)
+{
+	/* if (cpol == 0) this is SPI_MODE_0; else this is SPI_MODE_2 */
+
+	u32 oldbit = (!(word & (1<<(bits-1)))) << 31;
+	/* clock starts at inactive polarity */
+	for (word <<= (32 - bits); likely(bits); bits--) {
+
+		/* setup MSB (to slave) on trailing edge */
+		if ((flags & SPI_MASTER_NO_TX) == 0) {
+			if ((word & (1 << 31)) != oldbit) {
+				setmosi(spi, word & (1 << 31));
+				oldbit = word & (1 << 31);
+			}
+		}
+		spidelay(nsecs);	/* T(setup) */
+
+		setsck(spi, !cpol);
+		spidelay(nsecs);
+
+		/* sample MSB (from slave) on leading edge */
+		word <<= 1;
+		if ((flags & SPI_MASTER_NO_RX) == 0)
+			word |= getmiso(spi);
+		setsck(spi, cpol);
+	}
+	return word;
+}
+
+static inline u32
+bitbang_txrx_be_cpha1(struct spi_device *spi,
+		unsigned nsecs, unsigned cpol, unsigned flags,
+		u32 word, u8 bits)
+{
+	/* if (cpol == 0) this is SPI_MODE_1; else this is SPI_MODE_3 */
+
+	u32 oldbit = (!(word & (1<<(bits-1)))) << 31;
+	/* clock starts at inactive polarity */
+	for (word <<= (32 - bits); likely(bits); bits--) {
+
+		/* setup MSB (to slave) on leading edge */
+		setsck(spi, !cpol);
+		if ((flags & SPI_MASTER_NO_TX) == 0) {
+			if ((word & (1 << 31)) != oldbit) {
+				setmosi(spi, word & (1 << 31));
+				oldbit = word & (1 << 31);
+			}
+		}
+		spidelay(nsecs); /* T(setup) */
+
+		setsck(spi, cpol);
+		spidelay(nsecs);
+
+		/* sample MSB (from slave) on trailing edge */
+		word <<= 1;
+		if ((flags & SPI_MASTER_NO_RX) == 0)
+			word |= getmiso(spi);
+	}
+	return word;
+}
diff --git a/drivers/spi/spi-bitbang.c b/drivers/spi/spi-bitbang.c
new file mode 100644
index 0000000..f291760
--- /dev/null
+++ b/drivers/spi/spi-bitbang.c
@@ -0,0 +1,433 @@
+/*
+ * polling/bitbanging SPI master controller driver utilities
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/spinlock.h>
+#include <linux/workqueue.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/delay.h>
+#include <linux/errno.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+
+#define SPI_BITBANG_CS_DELAY	100
+
+
+/*----------------------------------------------------------------------*/
+
+/*
+ * FIRST PART (OPTIONAL):  word-at-a-time spi_transfer support.
+ * Use this for GPIO or shift-register level hardware APIs.
+ *
+ * spi_bitbang_cs is in spi_device->controller_state, which is unavailable
+ * to glue code.  These bitbang setup() and cleanup() routines are always
+ * used, though maybe they're called from controller-aware code.
+ *
+ * chipselect() and friends may use spi_device->controller_data and
+ * controller registers as appropriate.
+ *
+ *
+ * NOTE:  SPI controller pins can often be used as GPIO pins instead,
+ * which means you could use a bitbang driver either to get hardware
+ * working quickly, or testing for differences that aren't speed related.
+ */
+
+struct spi_bitbang_cs {
+	unsigned	nsecs;	/* (clock cycle time)/2 */
+	u32		(*txrx_word)(struct spi_device *spi, unsigned nsecs,
+					u32 word, u8 bits, unsigned flags);
+	unsigned	(*txrx_bufs)(struct spi_device *,
+					u32 (*txrx_word)(
+						struct spi_device *spi,
+						unsigned nsecs,
+						u32 word, u8 bits,
+						unsigned flags),
+					unsigned, struct spi_transfer *,
+					unsigned);
+};
+
+static unsigned bitbang_txrx_8(
+	struct spi_device	*spi,
+	u32			(*txrx_word)(struct spi_device *spi,
+					unsigned nsecs,
+					u32 word, u8 bits,
+					unsigned flags),
+	unsigned		ns,
+	struct spi_transfer	*t,
+	unsigned flags
+) {
+	unsigned		bits = t->bits_per_word;
+	unsigned		count = t->len;
+	const u8		*tx = t->tx_buf;
+	u8			*rx = t->rx_buf;
+
+	while (likely(count > 0)) {
+		u8		word = 0;
+
+		if (tx)
+			word = *tx++;
+		word = txrx_word(spi, ns, word, bits, flags);
+		if (rx)
+			*rx++ = word;
+		count -= 1;
+	}
+	return t->len - count;
+}
+
+static unsigned bitbang_txrx_16(
+	struct spi_device	*spi,
+	u32			(*txrx_word)(struct spi_device *spi,
+					unsigned nsecs,
+					u32 word, u8 bits,
+					unsigned flags),
+	unsigned		ns,
+	struct spi_transfer	*t,
+	unsigned flags
+) {
+	unsigned		bits = t->bits_per_word;
+	unsigned		count = t->len;
+	const u16		*tx = t->tx_buf;
+	u16			*rx = t->rx_buf;
+
+	while (likely(count > 1)) {
+		u16		word = 0;
+
+		if (tx)
+			word = *tx++;
+		word = txrx_word(spi, ns, word, bits, flags);
+		if (rx)
+			*rx++ = word;
+		count -= 2;
+	}
+	return t->len - count;
+}
+
+static unsigned bitbang_txrx_32(
+	struct spi_device	*spi,
+	u32			(*txrx_word)(struct spi_device *spi,
+					unsigned nsecs,
+					u32 word, u8 bits,
+					unsigned flags),
+	unsigned		ns,
+	struct spi_transfer	*t,
+	unsigned flags
+) {
+	unsigned		bits = t->bits_per_word;
+	unsigned		count = t->len;
+	const u32		*tx = t->tx_buf;
+	u32			*rx = t->rx_buf;
+
+	while (likely(count > 3)) {
+		u32		word = 0;
+
+		if (tx)
+			word = *tx++;
+		word = txrx_word(spi, ns, word, bits, flags);
+		if (rx)
+			*rx++ = word;
+		count -= 4;
+	}
+	return t->len - count;
+}
+
+int spi_bitbang_setup_transfer(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct spi_bitbang_cs	*cs = spi->controller_state;
+	u8			bits_per_word;
+	u32			hz;
+
+	if (t) {
+		bits_per_word = t->bits_per_word;
+		hz = t->speed_hz;
+	} else {
+		bits_per_word = 0;
+		hz = 0;
+	}
+
+	/* spi_transfer level calls that work per-word */
+	if (!bits_per_word)
+		bits_per_word = spi->bits_per_word;
+	if (bits_per_word <= 8)
+		cs->txrx_bufs = bitbang_txrx_8;
+	else if (bits_per_word <= 16)
+		cs->txrx_bufs = bitbang_txrx_16;
+	else if (bits_per_word <= 32)
+		cs->txrx_bufs = bitbang_txrx_32;
+	else
+		return -EINVAL;
+
+	/* nsecs = (clock period)/2 */
+	if (!hz)
+		hz = spi->max_speed_hz;
+	if (hz) {
+		cs->nsecs = (1000000000/2) / hz;
+		if (cs->nsecs > (MAX_UDELAY_MS * 1000 * 1000))
+			return -EINVAL;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(spi_bitbang_setup_transfer);
+
+/**
+ * spi_bitbang_setup - default setup for per-word I/O loops
+ */
+int spi_bitbang_setup(struct spi_device *spi)
+{
+	struct spi_bitbang_cs	*cs = spi->controller_state;
+	struct spi_bitbang	*bitbang;
+
+	bitbang = spi_master_get_devdata(spi->master);
+
+	if (!cs) {
+		cs = kzalloc(sizeof(*cs), GFP_KERNEL);
+		if (!cs)
+			return -ENOMEM;
+		spi->controller_state = cs;
+	}
+
+	/* per-word shift register access, in hardware or bitbanging */
+	cs->txrx_word = bitbang->txrx_word[spi->mode & (SPI_CPOL|SPI_CPHA)];
+	if (!cs->txrx_word)
+		return -EINVAL;
+
+	if (bitbang->setup_transfer) {
+		int retval = bitbang->setup_transfer(spi, NULL);
+		if (retval < 0)
+			return retval;
+	}
+
+	dev_dbg(&spi->dev, "%s, %u nsec/bit\n", __func__, 2 * cs->nsecs);
+
+	/* NOTE we _need_ to call chipselect() early, ideally with adapter
+	 * setup, unless the hardware defaults cooperate to avoid confusion
+	 * between normal (active low) and inverted chipselects.
+	 */
+
+	/* deselect chip (low or high) */
+	mutex_lock(&bitbang->lock);
+	if (!bitbang->busy) {
+		bitbang->chipselect(spi, BITBANG_CS_INACTIVE);
+		ndelay(cs->nsecs);
+	}
+	mutex_unlock(&bitbang->lock);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(spi_bitbang_setup);
+
+/**
+ * spi_bitbang_cleanup - default cleanup for per-word I/O loops
+ */
+void spi_bitbang_cleanup(struct spi_device *spi)
+{
+	kfree(spi->controller_state);
+}
+EXPORT_SYMBOL_GPL(spi_bitbang_cleanup);
+
+static int spi_bitbang_bufs(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct spi_bitbang_cs	*cs = spi->controller_state;
+	unsigned		nsecs = cs->nsecs;
+	struct spi_bitbang	*bitbang;
+
+	bitbang = spi_master_get_devdata(spi->master);
+	if (bitbang->set_line_direction) {
+		int err;
+
+		err = bitbang->set_line_direction(spi, !!(t->tx_buf));
+		if (err < 0)
+			return err;
+	}
+
+	if (spi->mode & SPI_3WIRE) {
+		unsigned flags;
+
+		flags = t->tx_buf ? SPI_MASTER_NO_RX : SPI_MASTER_NO_TX;
+		return cs->txrx_bufs(spi, cs->txrx_word, nsecs, t, flags);
+	}
+	return cs->txrx_bufs(spi, cs->txrx_word, nsecs, t, 0);
+}
+
+/*----------------------------------------------------------------------*/
+
+/*
+ * SECOND PART ... simple transfer queue runner.
+ *
+ * This costs a task context per controller, running the queue by
+ * performing each transfer in sequence.  Smarter hardware can queue
+ * several DMA transfers at once, and process several controller queues
+ * in parallel; this driver doesn't match such hardware very well.
+ *
+ * Drivers can provide word-at-a-time i/o primitives, or provide
+ * transfer-at-a-time ones to leverage dma or fifo hardware.
+ */
+
+static int spi_bitbang_prepare_hardware(struct spi_master *spi)
+{
+	struct spi_bitbang	*bitbang;
+
+	bitbang = spi_master_get_devdata(spi);
+
+	mutex_lock(&bitbang->lock);
+	bitbang->busy = 1;
+	mutex_unlock(&bitbang->lock);
+
+	return 0;
+}
+
+static int spi_bitbang_transfer_one(struct spi_master *master,
+				    struct spi_device *spi,
+				    struct spi_transfer *transfer)
+{
+	struct spi_bitbang *bitbang = spi_master_get_devdata(master);
+	int status = 0;
+
+	if (bitbang->setup_transfer) {
+		status = bitbang->setup_transfer(spi, transfer);
+		if (status < 0)
+			goto out;
+	}
+
+	if (transfer->len)
+		status = bitbang->txrx_bufs(spi, transfer);
+
+	if (status == transfer->len)
+		status = 0;
+	else if (status >= 0)
+		status = -EREMOTEIO;
+
+out:
+	spi_finalize_current_transfer(master);
+
+	return status;
+}
+
+static int spi_bitbang_unprepare_hardware(struct spi_master *spi)
+{
+	struct spi_bitbang	*bitbang;
+
+	bitbang = spi_master_get_devdata(spi);
+
+	mutex_lock(&bitbang->lock);
+	bitbang->busy = 0;
+	mutex_unlock(&bitbang->lock);
+
+	return 0;
+}
+
+static void spi_bitbang_set_cs(struct spi_device *spi, bool enable)
+{
+	struct spi_bitbang *bitbang = spi_master_get_devdata(spi->master);
+
+	/* SPI core provides CS high / low, but bitbang driver
+	 * expects CS active
+	 * spi device driver takes care of handling SPI_CS_HIGH
+	 */
+	enable = (!!(spi->mode & SPI_CS_HIGH) == enable);
+
+	ndelay(SPI_BITBANG_CS_DELAY);
+	bitbang->chipselect(spi, enable ? BITBANG_CS_ACTIVE :
+			    BITBANG_CS_INACTIVE);
+	ndelay(SPI_BITBANG_CS_DELAY);
+}
+
+/*----------------------------------------------------------------------*/
+
+/**
+ * spi_bitbang_start - start up a polled/bitbanging SPI master driver
+ * @bitbang: driver handle
+ *
+ * Caller should have zero-initialized all parts of the structure, and then
+ * provided callbacks for chip selection and I/O loops.  If the master has
+ * a transfer method, its final step should call spi_bitbang_transfer; or,
+ * that's the default if the transfer routine is not initialized.  It should
+ * also set up the bus number and number of chipselects.
+ *
+ * For i/o loops, provide callbacks either per-word (for bitbanging, or for
+ * hardware that basically exposes a shift register) or per-spi_transfer
+ * (which takes better advantage of hardware like fifos or DMA engines).
+ *
+ * Drivers using per-word I/O loops should use (or call) spi_bitbang_setup,
+ * spi_bitbang_cleanup and spi_bitbang_setup_transfer to handle those spi
+ * master methods.  Those methods are the defaults if the bitbang->txrx_bufs
+ * routine isn't initialized.
+ *
+ * This routine registers the spi_master, which will process requests in a
+ * dedicated task, keeping IRQs unblocked most of the time.  To stop
+ * processing those requests, call spi_bitbang_stop().
+ *
+ * On success, this routine will take a reference to master. The caller is
+ * responsible for calling spi_bitbang_stop() to decrement the reference and
+ * spi_master_put() as counterpart of spi_alloc_master() to prevent a memory
+ * leak.
+ */
+int spi_bitbang_start(struct spi_bitbang *bitbang)
+{
+	struct spi_master *master = bitbang->master;
+	int ret;
+
+	if (!master || !bitbang->chipselect)
+		return -EINVAL;
+
+	mutex_init(&bitbang->lock);
+
+	if (!master->mode_bits)
+		master->mode_bits = SPI_CPOL | SPI_CPHA | bitbang->flags;
+
+	if (master->transfer || master->transfer_one_message)
+		return -EINVAL;
+
+	master->prepare_transfer_hardware = spi_bitbang_prepare_hardware;
+	master->unprepare_transfer_hardware = spi_bitbang_unprepare_hardware;
+	master->transfer_one = spi_bitbang_transfer_one;
+	master->set_cs = spi_bitbang_set_cs;
+
+	if (!bitbang->txrx_bufs) {
+		bitbang->use_dma = 0;
+		bitbang->txrx_bufs = spi_bitbang_bufs;
+		if (!master->setup) {
+			if (!bitbang->setup_transfer)
+				bitbang->setup_transfer =
+					 spi_bitbang_setup_transfer;
+			master->setup = spi_bitbang_setup;
+			master->cleanup = spi_bitbang_cleanup;
+		}
+	}
+
+	/* driver may get busy before register() returns, especially
+	 * if someone registered boardinfo for devices
+	 */
+	ret = spi_register_master(spi_master_get(master));
+	if (ret)
+		spi_master_put(master);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(spi_bitbang_start);
+
+/**
+ * spi_bitbang_stop - stops the task providing spi communication
+ */
+void spi_bitbang_stop(struct spi_bitbang *bitbang)
+{
+	spi_unregister_master(bitbang->master);
+}
+EXPORT_SYMBOL_GPL(spi_bitbang_stop);
+
+MODULE_LICENSE("GPL");
+
diff --git a/drivers/spi/spi-brcmstb-qspi.c b/drivers/spi/spi-brcmstb-qspi.c
new file mode 100644
index 0000000..c7df92e
--- /dev/null
+++ b/drivers/spi/spi-brcmstb-qspi.c
@@ -0,0 +1,53 @@
+/*
+ * Copyright 2016 Broadcom
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation (the "GPL").
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License version 2 (GPLv2) for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * version 2 (GPLv2) along with this source code.
+ */
+
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/of.h>
+#include "spi-bcm-qspi.h"
+
+static const struct of_device_id brcmstb_qspi_of_match[] = {
+	{ .compatible = "brcm,spi-brcmstb-qspi" },
+	{ .compatible = "brcm,spi-brcmstb-mspi" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, brcmstb_qspi_of_match);
+
+static int brcmstb_qspi_probe(struct platform_device *pdev)
+{
+	return bcm_qspi_probe(pdev, NULL);
+}
+
+static int brcmstb_qspi_remove(struct platform_device *pdev)
+{
+	return bcm_qspi_remove(pdev);
+}
+
+static struct platform_driver brcmstb_qspi_driver = {
+	.probe			= brcmstb_qspi_probe,
+	.remove			= brcmstb_qspi_remove,
+	.driver = {
+		.name		= "brcmstb_qspi",
+		.pm		= &bcm_qspi_pm_ops,
+		.of_match_table = brcmstb_qspi_of_match,
+	}
+};
+module_platform_driver(brcmstb_qspi_driver);
+
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Kamal Dasu");
+MODULE_DESCRIPTION("Broadcom SPI driver for settop SoC");
diff --git a/drivers/spi/spi-butterfly.c b/drivers/spi/spi-butterfly.c
new file mode 100644
index 0000000..1a35102
--- /dev/null
+++ b/drivers/spi/spi-butterfly.c
@@ -0,0 +1,343 @@
+/*
+ * parport-to-butterfly adapter
+ *
+ * Copyright (C) 2005 David Brownell
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/parport.h>
+
+#include <linux/sched.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/spi/flash.h>
+
+#include <linux/mtd/partitions.h>
+
+/*
+ * This uses SPI to talk with an "AVR Butterfly", which is a $US20 card
+ * with a battery powered AVR microcontroller and lots of goodies.  You
+ * can use GCC to develop firmware for this.
+ *
+ * See Documentation/spi/butterfly for information about how to build
+ * and use this custom parallel port cable.
+ */
+
+/* DATA output bits (pins 2..9 == D0..D7) */
+#define	butterfly_nreset (1 << 1)		/* pin 3 */
+
+#define	spi_sck_bit	(1 << 0)		/* pin 2 */
+#define	spi_mosi_bit	(1 << 7)		/* pin 9 */
+
+#define	vcc_bits	((1 << 6) | (1 << 5))	/* pins 7, 8 */
+
+/* STATUS input bits */
+#define	spi_miso_bit	PARPORT_STATUS_BUSY	/* pin 11 */
+
+/* CONTROL output bits */
+#define	spi_cs_bit	PARPORT_CONTROL_SELECT	/* pin 17 */
+
+static inline struct butterfly *spidev_to_pp(struct spi_device *spi)
+{
+	return spi->controller_data;
+}
+
+struct butterfly {
+	/* REVISIT ... for now, this must be first */
+	struct spi_bitbang	bitbang;
+
+	struct parport		*port;
+	struct pardevice	*pd;
+
+	u8			lastbyte;
+
+	struct spi_device	*dataflash;
+	struct spi_device	*butterfly;
+	struct spi_board_info	info[2];
+
+};
+
+/*----------------------------------------------------------------------*/
+
+static inline void
+setsck(struct spi_device *spi, int is_on)
+{
+	struct butterfly	*pp = spidev_to_pp(spi);
+	u8			bit, byte = pp->lastbyte;
+
+	bit = spi_sck_bit;
+
+	if (is_on)
+		byte |= bit;
+	else
+		byte &= ~bit;
+	parport_write_data(pp->port, byte);
+	pp->lastbyte = byte;
+}
+
+static inline void
+setmosi(struct spi_device *spi, int is_on)
+{
+	struct butterfly	*pp = spidev_to_pp(spi);
+	u8			bit, byte = pp->lastbyte;
+
+	bit = spi_mosi_bit;
+
+	if (is_on)
+		byte |= bit;
+	else
+		byte &= ~bit;
+	parport_write_data(pp->port, byte);
+	pp->lastbyte = byte;
+}
+
+static inline int getmiso(struct spi_device *spi)
+{
+	struct butterfly	*pp = spidev_to_pp(spi);
+	int			value;
+	u8			bit;
+
+	bit = spi_miso_bit;
+
+	/* only STATUS_BUSY is NOT negated */
+	value = !(parport_read_status(pp->port) & bit);
+	return (bit == PARPORT_STATUS_BUSY) ? value : !value;
+}
+
+static void butterfly_chipselect(struct spi_device *spi, int value)
+{
+	struct butterfly	*pp = spidev_to_pp(spi);
+
+	/* set default clock polarity */
+	if (value != BITBANG_CS_INACTIVE)
+		setsck(spi, spi->mode & SPI_CPOL);
+
+	/* here, value == "activate or not";
+	 * most PARPORT_CONTROL_* bits are negated, so we must
+	 * morph it to value == "bit value to write in control register"
+	 */
+	if (spi_cs_bit == PARPORT_CONTROL_INIT)
+		value = !value;
+
+	parport_frob_control(pp->port, spi_cs_bit, value ? spi_cs_bit : 0);
+}
+
+/* we only needed to implement one mode here, and choose SPI_MODE_0 */
+
+#define spidelay(X)	do { } while (0)
+/* #define spidelay	ndelay */
+
+#include "spi-bitbang-txrx.h"
+
+static u32
+butterfly_txrx_word_mode0(struct spi_device *spi, unsigned nsecs, u32 word,
+			  u8 bits, unsigned flags)
+{
+	return bitbang_txrx_be_cpha0(spi, nsecs, 0, flags, word, bits);
+}
+
+/*----------------------------------------------------------------------*/
+
+/* override default partitioning with cmdlinepart */
+static struct mtd_partition partitions[] = { {
+	/* JFFS2 wants partitions of 4*N blocks for this device,
+	 * so sectors 0 and 1 can't be partitions by themselves.
+	 */
+
+	/* sector 0 = 8 pages * 264 bytes/page (1 block)
+	 * sector 1 = 248 pages * 264 bytes/page
+	 */
+	.name		= "bookkeeping",	/* 66 KB */
+	.offset		= 0,
+	.size		= (8 + 248) * 264,
+	/* .mask_flags	= MTD_WRITEABLE, */
+}, {
+	/* sector 2 = 256 pages * 264 bytes/page
+	 * sectors 3-5 = 512 pages * 264 bytes/page
+	 */
+	.name		= "filesystem",		/* 462 KB */
+	.offset		= MTDPART_OFS_APPEND,
+	.size		= MTDPART_SIZ_FULL,
+} };
+
+static struct flash_platform_data flash = {
+	.name		= "butterflash",
+	.parts		= partitions,
+	.nr_parts	= ARRAY_SIZE(partitions),
+};
+
+/* REVISIT remove this ugly global and its "only one" limitation */
+static struct butterfly *butterfly;
+
+static void butterfly_attach(struct parport *p)
+{
+	struct pardevice	*pd;
+	int			status;
+	struct butterfly	*pp;
+	struct spi_master	*master;
+	struct device		*dev = p->physport->dev;
+	struct pardev_cb	butterfly_cb;
+
+	if (butterfly || !dev)
+		return;
+
+	/* REVISIT:  this just _assumes_ a butterfly is there ... no probe,
+	 * and no way to be selective about what it binds to.
+	 */
+
+	master = spi_alloc_master(dev, sizeof(*pp));
+	if (!master) {
+		status = -ENOMEM;
+		goto done;
+	}
+	pp = spi_master_get_devdata(master);
+
+	/*
+	 * SPI and bitbang hookup
+	 *
+	 * use default setup(), cleanup(), and transfer() methods; and
+	 * only bother implementing mode 0.  Start it later.
+	 */
+	master->bus_num = 42;
+	master->num_chipselect = 2;
+
+	pp->bitbang.master = master;
+	pp->bitbang.chipselect = butterfly_chipselect;
+	pp->bitbang.txrx_word[SPI_MODE_0] = butterfly_txrx_word_mode0;
+
+	/*
+	 * parport hookup
+	 */
+	pp->port = p;
+	memset(&butterfly_cb, 0, sizeof(butterfly_cb));
+	butterfly_cb.private = pp;
+	pd = parport_register_dev_model(p, "spi_butterfly", &butterfly_cb, 0);
+	if (!pd) {
+		status = -ENOMEM;
+		goto clean0;
+	}
+	pp->pd = pd;
+
+	status = parport_claim(pd);
+	if (status < 0)
+		goto clean1;
+
+	/*
+	 * Butterfly reset, powerup, run firmware
+	 */
+	pr_debug("%s: powerup/reset Butterfly\n", p->name);
+
+	/* nCS for dataflash (this bit is inverted on output) */
+	parport_frob_control(pp->port, spi_cs_bit, 0);
+
+	/* stabilize power with chip in reset (nRESET), and
+	 * spi_sck_bit clear (CPOL=0)
+	 */
+	pp->lastbyte |= vcc_bits;
+	parport_write_data(pp->port, pp->lastbyte);
+	msleep(5);
+
+	/* take it out of reset; assume long reset delay */
+	pp->lastbyte |= butterfly_nreset;
+	parport_write_data(pp->port, pp->lastbyte);
+	msleep(100);
+
+	/*
+	 * Start SPI ... for now, hide that we're two physical busses.
+	 */
+	status = spi_bitbang_start(&pp->bitbang);
+	if (status < 0)
+		goto clean2;
+
+	/* Bus 1 lets us talk to at45db041b (firmware disables AVR SPI), AVR
+	 * (firmware resets at45, acts as spi slave) or neither (we ignore
+	 * both, AVR uses AT45).  Here we expect firmware for the first option.
+	 */
+
+	pp->info[0].max_speed_hz = 15 * 1000 * 1000;
+	strcpy(pp->info[0].modalias, "mtd_dataflash");
+	pp->info[0].platform_data = &flash;
+	pp->info[0].chip_select = 1;
+	pp->info[0].controller_data = pp;
+	pp->dataflash = spi_new_device(pp->bitbang.master, &pp->info[0]);
+	if (pp->dataflash)
+		pr_debug("%s: dataflash at %s\n", p->name,
+			 dev_name(&pp->dataflash->dev));
+
+	pr_info("%s: AVR Butterfly\n", p->name);
+	butterfly = pp;
+	return;
+
+clean2:
+	/* turn off VCC */
+	parport_write_data(pp->port, 0);
+
+	parport_release(pp->pd);
+clean1:
+	parport_unregister_device(pd);
+clean0:
+	spi_master_put(pp->bitbang.master);
+done:
+	pr_debug("%s: butterfly probe, fail %d\n", p->name, status);
+}
+
+static void butterfly_detach(struct parport *p)
+{
+	struct butterfly	*pp;
+
+	/* FIXME this global is ugly ... but, how to quickly get from
+	 * the parport to the "struct butterfly" associated with it?
+	 * "old school" driver-internal device lists?
+	 */
+	if (!butterfly || butterfly->port != p)
+		return;
+	pp = butterfly;
+	butterfly = NULL;
+
+	/* stop() unregisters child devices too */
+	spi_bitbang_stop(&pp->bitbang);
+
+	/* turn off VCC */
+	parport_write_data(pp->port, 0);
+	msleep(10);
+
+	parport_release(pp->pd);
+	parport_unregister_device(pp->pd);
+
+	spi_master_put(pp->bitbang.master);
+}
+
+static struct parport_driver butterfly_driver = {
+	.name =		"spi_butterfly",
+	.match_port =	butterfly_attach,
+	.detach =	butterfly_detach,
+	.devmodel = true,
+};
+
+static int __init butterfly_init(void)
+{
+	return parport_register_driver(&butterfly_driver);
+}
+device_initcall(butterfly_init);
+
+static void __exit butterfly_exit(void)
+{
+	parport_unregister_driver(&butterfly_driver);
+}
+module_exit(butterfly_exit);
+
+MODULE_DESCRIPTION("Parport Adapter driver for AVR Butterfly");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-cadence.c b/drivers/spi/spi-cadence.c
new file mode 100644
index 0000000..7c88f74
--- /dev/null
+++ b/drivers/spi/spi-cadence.c
@@ -0,0 +1,795 @@
+/*
+ * Cadence SPI controller driver (master mode only)
+ *
+ * Copyright (C) 2008 - 2014 Xilinx, Inc.
+ *
+ * based on Blackfin On-Chip SPI Driver (spi_bfin5xx.c)
+ *
+ * This program is free software; you can redistribute it and/or modify it under
+ * the terms of the GNU General Public License version 2 as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/gpio.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of_irq.h>
+#include <linux/of_address.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/spi/spi.h>
+
+/* Name of this driver */
+#define CDNS_SPI_NAME		"cdns-spi"
+
+/* Register offset definitions */
+#define CDNS_SPI_CR	0x00 /* Configuration  Register, RW */
+#define CDNS_SPI_ISR	0x04 /* Interrupt Status Register, RO */
+#define CDNS_SPI_IER	0x08 /* Interrupt Enable Register, WO */
+#define CDNS_SPI_IDR	0x0c /* Interrupt Disable Register, WO */
+#define CDNS_SPI_IMR	0x10 /* Interrupt Enabled Mask Register, RO */
+#define CDNS_SPI_ER	0x14 /* Enable/Disable Register, RW */
+#define CDNS_SPI_DR	0x18 /* Delay Register, RW */
+#define CDNS_SPI_TXD	0x1C /* Data Transmit Register, WO */
+#define CDNS_SPI_RXD	0x20 /* Data Receive Register, RO */
+#define CDNS_SPI_SICR	0x24 /* Slave Idle Count Register, RW */
+#define CDNS_SPI_THLD	0x28 /* Transmit FIFO Watermark Register,RW */
+
+#define SPI_AUTOSUSPEND_TIMEOUT		3000
+/*
+ * SPI Configuration Register bit Masks
+ *
+ * This register contains various control bits that affect the operation
+ * of the SPI controller
+ */
+#define CDNS_SPI_CR_MANSTRT	0x00010000 /* Manual TX Start */
+#define CDNS_SPI_CR_CPHA		0x00000004 /* Clock Phase Control */
+#define CDNS_SPI_CR_CPOL		0x00000002 /* Clock Polarity Control */
+#define CDNS_SPI_CR_SSCTRL		0x00003C00 /* Slave Select Mask */
+#define CDNS_SPI_CR_PERI_SEL	0x00000200 /* Peripheral Select Decode */
+#define CDNS_SPI_CR_BAUD_DIV	0x00000038 /* Baud Rate Divisor Mask */
+#define CDNS_SPI_CR_MSTREN		0x00000001 /* Master Enable Mask */
+#define CDNS_SPI_CR_MANSTRTEN	0x00008000 /* Manual TX Enable Mask */
+#define CDNS_SPI_CR_SSFORCE	0x00004000 /* Manual SS Enable Mask */
+#define CDNS_SPI_CR_BAUD_DIV_4	0x00000008 /* Default Baud Div Mask */
+#define CDNS_SPI_CR_DEFAULT	(CDNS_SPI_CR_MSTREN | \
+					CDNS_SPI_CR_SSCTRL | \
+					CDNS_SPI_CR_SSFORCE | \
+					CDNS_SPI_CR_BAUD_DIV_4)
+
+/*
+ * SPI Configuration Register - Baud rate and slave select
+ *
+ * These are the values used in the calculation of baud rate divisor and
+ * setting the slave select.
+ */
+
+#define CDNS_SPI_BAUD_DIV_MAX		7 /* Baud rate divisor maximum */
+#define CDNS_SPI_BAUD_DIV_MIN		1 /* Baud rate divisor minimum */
+#define CDNS_SPI_BAUD_DIV_SHIFT		3 /* Baud rate divisor shift in CR */
+#define CDNS_SPI_SS_SHIFT		10 /* Slave Select field shift in CR */
+#define CDNS_SPI_SS0			0x1 /* Slave Select zero */
+
+/*
+ * SPI Interrupt Registers bit Masks
+ *
+ * All the four interrupt registers (Status/Mask/Enable/Disable) have the same
+ * bit definitions.
+ */
+#define CDNS_SPI_IXR_TXOW	0x00000004 /* SPI TX FIFO Overwater */
+#define CDNS_SPI_IXR_MODF	0x00000002 /* SPI Mode Fault */
+#define CDNS_SPI_IXR_RXNEMTY 0x00000010 /* SPI RX FIFO Not Empty */
+#define CDNS_SPI_IXR_DEFAULT	(CDNS_SPI_IXR_TXOW | \
+					CDNS_SPI_IXR_MODF)
+#define CDNS_SPI_IXR_TXFULL	0x00000008 /* SPI TX Full */
+#define CDNS_SPI_IXR_ALL	0x0000007F /* SPI all interrupts */
+
+/*
+ * SPI Enable Register bit Masks
+ *
+ * This register is used to enable or disable the SPI controller
+ */
+#define CDNS_SPI_ER_ENABLE	0x00000001 /* SPI Enable Bit Mask */
+#define CDNS_SPI_ER_DISABLE	0x0 /* SPI Disable Bit Mask */
+
+/* SPI FIFO depth in bytes */
+#define CDNS_SPI_FIFO_DEPTH	128
+
+/* Default number of chip select lines */
+#define CDNS_SPI_DEFAULT_NUM_CS		4
+
+/**
+ * struct cdns_spi - This definition defines spi driver instance
+ * @regs:		Virtual address of the SPI controller registers
+ * @ref_clk:		Pointer to the peripheral clock
+ * @pclk:		Pointer to the APB clock
+ * @speed_hz:		Current SPI bus clock speed in Hz
+ * @txbuf:		Pointer	to the TX buffer
+ * @rxbuf:		Pointer to the RX buffer
+ * @tx_bytes:		Number of bytes left to transfer
+ * @rx_bytes:		Number of bytes requested
+ * @dev_busy:		Device busy flag
+ * @is_decoded_cs:	Flag for decoder property set or not
+ */
+struct cdns_spi {
+	void __iomem *regs;
+	struct clk *ref_clk;
+	struct clk *pclk;
+	u32 speed_hz;
+	const u8 *txbuf;
+	u8 *rxbuf;
+	int tx_bytes;
+	int rx_bytes;
+	u8 dev_busy;
+	u32 is_decoded_cs;
+};
+
+struct cdns_spi_device_data {
+	bool gpio_requested;
+};
+
+/* Macros for the SPI controller read/write */
+static inline u32 cdns_spi_read(struct cdns_spi *xspi, u32 offset)
+{
+	return readl_relaxed(xspi->regs + offset);
+}
+
+static inline void cdns_spi_write(struct cdns_spi *xspi, u32 offset, u32 val)
+{
+	writel_relaxed(val, xspi->regs + offset);
+}
+
+/**
+ * cdns_spi_init_hw - Initialize the hardware and configure the SPI controller
+ * @xspi:	Pointer to the cdns_spi structure
+ *
+ * On reset the SPI controller is configured to be in master mode, baud rate
+ * divisor is set to 4, threshold value for TX FIFO not full interrupt is set
+ * to 1 and size of the word to be transferred as 8 bit.
+ * This function initializes the SPI controller to disable and clear all the
+ * interrupts, enable manual slave select and manual start, deselect all the
+ * chip select lines, and enable the SPI controller.
+ */
+static void cdns_spi_init_hw(struct cdns_spi *xspi)
+{
+	u32 ctrl_reg = CDNS_SPI_CR_DEFAULT;
+
+	if (xspi->is_decoded_cs)
+		ctrl_reg |= CDNS_SPI_CR_PERI_SEL;
+
+	cdns_spi_write(xspi, CDNS_SPI_ER, CDNS_SPI_ER_DISABLE);
+	cdns_spi_write(xspi, CDNS_SPI_IDR, CDNS_SPI_IXR_ALL);
+
+	/* Clear the RX FIFO */
+	while (cdns_spi_read(xspi, CDNS_SPI_ISR) & CDNS_SPI_IXR_RXNEMTY)
+		cdns_spi_read(xspi, CDNS_SPI_RXD);
+
+	cdns_spi_write(xspi, CDNS_SPI_ISR, CDNS_SPI_IXR_ALL);
+	cdns_spi_write(xspi, CDNS_SPI_CR, ctrl_reg);
+	cdns_spi_write(xspi, CDNS_SPI_ER, CDNS_SPI_ER_ENABLE);
+}
+
+/**
+ * cdns_spi_chipselect - Select or deselect the chip select line
+ * @spi:	Pointer to the spi_device structure
+ * @is_high:	Select(0) or deselect (1) the chip select line
+ */
+static void cdns_spi_chipselect(struct spi_device *spi, bool is_high)
+{
+	struct cdns_spi *xspi = spi_master_get_devdata(spi->master);
+	u32 ctrl_reg;
+
+	ctrl_reg = cdns_spi_read(xspi, CDNS_SPI_CR);
+
+	if (is_high) {
+		/* Deselect the slave */
+		ctrl_reg |= CDNS_SPI_CR_SSCTRL;
+	} else {
+		/* Select the slave */
+		ctrl_reg &= ~CDNS_SPI_CR_SSCTRL;
+		if (!(xspi->is_decoded_cs))
+			ctrl_reg |= ((~(CDNS_SPI_SS0 << spi->chip_select)) <<
+				     CDNS_SPI_SS_SHIFT) &
+				     CDNS_SPI_CR_SSCTRL;
+		else
+			ctrl_reg |= (spi->chip_select << CDNS_SPI_SS_SHIFT) &
+				     CDNS_SPI_CR_SSCTRL;
+	}
+
+	cdns_spi_write(xspi, CDNS_SPI_CR, ctrl_reg);
+}
+
+/**
+ * cdns_spi_config_clock_mode - Sets clock polarity and phase
+ * @spi:	Pointer to the spi_device structure
+ *
+ * Sets the requested clock polarity and phase.
+ */
+static void cdns_spi_config_clock_mode(struct spi_device *spi)
+{
+	struct cdns_spi *xspi = spi_master_get_devdata(spi->master);
+	u32 ctrl_reg, new_ctrl_reg;
+
+	new_ctrl_reg = cdns_spi_read(xspi, CDNS_SPI_CR);
+	ctrl_reg = new_ctrl_reg;
+
+	/* Set the SPI clock phase and clock polarity */
+	new_ctrl_reg &= ~(CDNS_SPI_CR_CPHA | CDNS_SPI_CR_CPOL);
+	if (spi->mode & SPI_CPHA)
+		new_ctrl_reg |= CDNS_SPI_CR_CPHA;
+	if (spi->mode & SPI_CPOL)
+		new_ctrl_reg |= CDNS_SPI_CR_CPOL;
+
+	if (new_ctrl_reg != ctrl_reg) {
+		/*
+		 * Just writing the CR register does not seem to apply the clock
+		 * setting changes. This is problematic when changing the clock
+		 * polarity as it will cause the SPI slave to see spurious clock
+		 * transitions. To workaround the issue toggle the ER register.
+		 */
+		cdns_spi_write(xspi, CDNS_SPI_ER, CDNS_SPI_ER_DISABLE);
+		cdns_spi_write(xspi, CDNS_SPI_CR, new_ctrl_reg);
+		cdns_spi_write(xspi, CDNS_SPI_ER, CDNS_SPI_ER_ENABLE);
+	}
+}
+
+/**
+ * cdns_spi_config_clock_freq - Sets clock frequency
+ * @spi:	Pointer to the spi_device structure
+ * @transfer:	Pointer to the spi_transfer structure which provides
+ *		information about next transfer setup parameters
+ *
+ * Sets the requested clock frequency.
+ * Note: If the requested frequency is not an exact match with what can be
+ * obtained using the prescalar value the driver sets the clock frequency which
+ * is lower than the requested frequency (maximum lower) for the transfer. If
+ * the requested frequency is higher or lower than that is supported by the SPI
+ * controller the driver will set the highest or lowest frequency supported by
+ * controller.
+ */
+static void cdns_spi_config_clock_freq(struct spi_device *spi,
+				       struct spi_transfer *transfer)
+{
+	struct cdns_spi *xspi = spi_master_get_devdata(spi->master);
+	u32 ctrl_reg, baud_rate_val;
+	unsigned long frequency;
+
+	frequency = clk_get_rate(xspi->ref_clk);
+
+	ctrl_reg = cdns_spi_read(xspi, CDNS_SPI_CR);
+
+	/* Set the clock frequency */
+	if (xspi->speed_hz != transfer->speed_hz) {
+		/* first valid value is 1 */
+		baud_rate_val = CDNS_SPI_BAUD_DIV_MIN;
+		while ((baud_rate_val < CDNS_SPI_BAUD_DIV_MAX) &&
+		       (frequency / (2 << baud_rate_val)) > transfer->speed_hz)
+			baud_rate_val++;
+
+		ctrl_reg &= ~CDNS_SPI_CR_BAUD_DIV;
+		ctrl_reg |= baud_rate_val << CDNS_SPI_BAUD_DIV_SHIFT;
+
+		xspi->speed_hz = frequency / (2 << baud_rate_val);
+	}
+	cdns_spi_write(xspi, CDNS_SPI_CR, ctrl_reg);
+}
+
+/**
+ * cdns_spi_setup_transfer - Configure SPI controller for specified transfer
+ * @spi:	Pointer to the spi_device structure
+ * @transfer:	Pointer to the spi_transfer structure which provides
+ *		information about next transfer setup parameters
+ *
+ * Sets the operational mode of SPI controller for the next SPI transfer and
+ * sets the requested clock frequency.
+ *
+ * Return:	Always 0
+ */
+static int cdns_spi_setup_transfer(struct spi_device *spi,
+				   struct spi_transfer *transfer)
+{
+	struct cdns_spi *xspi = spi_master_get_devdata(spi->master);
+
+	cdns_spi_config_clock_freq(spi, transfer);
+
+	dev_dbg(&spi->dev, "%s, mode %d, %u bits/w, %u clock speed\n",
+		__func__, spi->mode, spi->bits_per_word,
+		xspi->speed_hz);
+
+	return 0;
+}
+
+/**
+ * cdns_spi_fill_tx_fifo - Fills the TX FIFO with as many bytes as possible
+ * @xspi:	Pointer to the cdns_spi structure
+ */
+static void cdns_spi_fill_tx_fifo(struct cdns_spi *xspi)
+{
+	unsigned long trans_cnt = 0;
+
+	while ((trans_cnt < CDNS_SPI_FIFO_DEPTH) &&
+	       (xspi->tx_bytes > 0)) {
+
+		/* When xspi in busy condition, bytes may send failed,
+		 * then spi control did't work thoroughly, add one byte delay
+		 */
+		if (cdns_spi_read(xspi, CDNS_SPI_ISR) &
+		    CDNS_SPI_IXR_TXFULL)
+			udelay(10);
+
+		if (xspi->txbuf)
+			cdns_spi_write(xspi, CDNS_SPI_TXD, *xspi->txbuf++);
+		else
+			cdns_spi_write(xspi, CDNS_SPI_TXD, 0);
+
+		xspi->tx_bytes--;
+		trans_cnt++;
+	}
+}
+
+/**
+ * cdns_spi_irq - Interrupt service routine of the SPI controller
+ * @irq:	IRQ number
+ * @dev_id:	Pointer to the xspi structure
+ *
+ * This function handles TX empty and Mode Fault interrupts only.
+ * On TX empty interrupt this function reads the received data from RX FIFO and
+ * fills the TX FIFO if there is any data remaining to be transferred.
+ * On Mode Fault interrupt this function indicates that transfer is completed,
+ * the SPI subsystem will identify the error as the remaining bytes to be
+ * transferred is non-zero.
+ *
+ * Return:	IRQ_HANDLED when handled; IRQ_NONE otherwise.
+ */
+static irqreturn_t cdns_spi_irq(int irq, void *dev_id)
+{
+	struct spi_master *master = dev_id;
+	struct cdns_spi *xspi = spi_master_get_devdata(master);
+	u32 intr_status, status;
+
+	status = IRQ_NONE;
+	intr_status = cdns_spi_read(xspi, CDNS_SPI_ISR);
+	cdns_spi_write(xspi, CDNS_SPI_ISR, intr_status);
+
+	if (intr_status & CDNS_SPI_IXR_MODF) {
+		/* Indicate that transfer is completed, the SPI subsystem will
+		 * identify the error as the remaining bytes to be
+		 * transferred is non-zero
+		 */
+		cdns_spi_write(xspi, CDNS_SPI_IDR, CDNS_SPI_IXR_DEFAULT);
+		spi_finalize_current_transfer(master);
+		status = IRQ_HANDLED;
+	} else if (intr_status & CDNS_SPI_IXR_TXOW) {
+		unsigned long trans_cnt;
+
+		trans_cnt = xspi->rx_bytes - xspi->tx_bytes;
+
+		/* Read out the data from the RX FIFO */
+		while (trans_cnt) {
+			u8 data;
+
+			data = cdns_spi_read(xspi, CDNS_SPI_RXD);
+			if (xspi->rxbuf)
+				*xspi->rxbuf++ = data;
+
+			xspi->rx_bytes--;
+			trans_cnt--;
+		}
+
+		if (xspi->tx_bytes) {
+			/* There is more data to send */
+			cdns_spi_fill_tx_fifo(xspi);
+		} else {
+			/* Transfer is completed */
+			cdns_spi_write(xspi, CDNS_SPI_IDR,
+				       CDNS_SPI_IXR_DEFAULT);
+			spi_finalize_current_transfer(master);
+		}
+		status = IRQ_HANDLED;
+	}
+
+	return status;
+}
+
+static int cdns_prepare_message(struct spi_master *master,
+				struct spi_message *msg)
+{
+	cdns_spi_config_clock_mode(msg->spi);
+	return 0;
+}
+
+/**
+ * cdns_transfer_one - Initiates the SPI transfer
+ * @master:	Pointer to spi_master structure
+ * @spi:	Pointer to the spi_device structure
+ * @transfer:	Pointer to the spi_transfer structure which provides
+ *		information about next transfer parameters
+ *
+ * This function fills the TX FIFO, starts the SPI transfer and
+ * returns a positive transfer count so that core will wait for completion.
+ *
+ * Return:	Number of bytes transferred in the last transfer
+ */
+static int cdns_transfer_one(struct spi_master *master,
+			     struct spi_device *spi,
+			     struct spi_transfer *transfer)
+{
+	struct cdns_spi *xspi = spi_master_get_devdata(master);
+
+	xspi->txbuf = transfer->tx_buf;
+	xspi->rxbuf = transfer->rx_buf;
+	xspi->tx_bytes = transfer->len;
+	xspi->rx_bytes = transfer->len;
+
+	cdns_spi_setup_transfer(spi, transfer);
+
+	cdns_spi_fill_tx_fifo(xspi);
+
+	cdns_spi_write(xspi, CDNS_SPI_IER, CDNS_SPI_IXR_DEFAULT);
+	return transfer->len;
+}
+
+/**
+ * cdns_prepare_transfer_hardware - Prepares hardware for transfer.
+ * @master:	Pointer to the spi_master structure which provides
+ *		information about the controller.
+ *
+ * This function enables SPI master controller.
+ *
+ * Return:	0 always
+ */
+static int cdns_prepare_transfer_hardware(struct spi_master *master)
+{
+	struct cdns_spi *xspi = spi_master_get_devdata(master);
+
+	cdns_spi_write(xspi, CDNS_SPI_ER, CDNS_SPI_ER_ENABLE);
+
+	return 0;
+}
+
+/**
+ * cdns_unprepare_transfer_hardware - Relaxes hardware after transfer
+ * @master:	Pointer to the spi_master structure which provides
+ *		information about the controller.
+ *
+ * This function disables the SPI master controller.
+ *
+ * Return:	0 always
+ */
+static int cdns_unprepare_transfer_hardware(struct spi_master *master)
+{
+	struct cdns_spi *xspi = spi_master_get_devdata(master);
+
+	cdns_spi_write(xspi, CDNS_SPI_ER, CDNS_SPI_ER_DISABLE);
+
+	return 0;
+}
+
+static int cdns_spi_setup(struct spi_device *spi)
+{
+
+	int ret = -EINVAL;
+	struct cdns_spi_device_data *cdns_spi_data = spi_get_ctldata(spi);
+
+	/* this is a pin managed by the controller, leave it alone */
+	if (spi->cs_gpio == -ENOENT)
+		return 0;
+
+	/* this seems to be the first time we're here */
+	if (!cdns_spi_data) {
+		cdns_spi_data = kzalloc(sizeof(*cdns_spi_data), GFP_KERNEL);
+		if (!cdns_spi_data)
+			return -ENOMEM;
+		cdns_spi_data->gpio_requested = false;
+		spi_set_ctldata(spi, cdns_spi_data);
+	}
+
+	/* if we haven't done so, grab the gpio */
+	if (!cdns_spi_data->gpio_requested && gpio_is_valid(spi->cs_gpio)) {
+		ret = gpio_request_one(spi->cs_gpio,
+				       (spi->mode & SPI_CS_HIGH) ?
+				       GPIOF_OUT_INIT_LOW : GPIOF_OUT_INIT_HIGH,
+				       dev_name(&spi->dev));
+		if (ret)
+			dev_err(&spi->dev, "can't request chipselect gpio %d\n",
+				spi->cs_gpio);
+		else
+			cdns_spi_data->gpio_requested = true;
+	} else {
+		if (gpio_is_valid(spi->cs_gpio)) {
+			int mode = ((spi->mode & SPI_CS_HIGH) ?
+				    GPIOF_OUT_INIT_LOW : GPIOF_OUT_INIT_HIGH);
+
+			ret = gpio_direction_output(spi->cs_gpio, mode);
+			if (ret)
+				dev_err(&spi->dev, "chipselect gpio %d setup failed (%d)\n",
+					spi->cs_gpio, ret);
+		}
+	}
+
+	return ret;
+}
+
+static void cdns_spi_cleanup(struct spi_device *spi)
+{
+	struct cdns_spi_device_data *cdns_spi_data = spi_get_ctldata(spi);
+
+	if (cdns_spi_data) {
+		if (cdns_spi_data->gpio_requested)
+			gpio_free(spi->cs_gpio);
+		kfree(cdns_spi_data);
+		spi_set_ctldata(spi, NULL);
+	}
+
+}
+
+/**
+ * cdns_spi_probe - Probe method for the SPI driver
+ * @pdev:	Pointer to the platform_device structure
+ *
+ * This function initializes the driver data structures and the hardware.
+ *
+ * Return:	0 on success and error value on error
+ */
+static int cdns_spi_probe(struct platform_device *pdev)
+{
+	int ret = 0, irq;
+	struct spi_master *master;
+	struct cdns_spi *xspi;
+	struct resource *res;
+	u32 num_cs;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*xspi));
+	if (!master)
+		return -ENOMEM;
+
+	xspi = spi_master_get_devdata(master);
+	master->dev.of_node = pdev->dev.of_node;
+	platform_set_drvdata(pdev, master);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	xspi->regs = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(xspi->regs)) {
+		ret = PTR_ERR(xspi->regs);
+		goto remove_master;
+	}
+
+	xspi->pclk = devm_clk_get(&pdev->dev, "pclk");
+	if (IS_ERR(xspi->pclk)) {
+		dev_err(&pdev->dev, "pclk clock not found.\n");
+		ret = PTR_ERR(xspi->pclk);
+		goto remove_master;
+	}
+
+	xspi->ref_clk = devm_clk_get(&pdev->dev, "ref_clk");
+	if (IS_ERR(xspi->ref_clk)) {
+		dev_err(&pdev->dev, "ref_clk clock not found.\n");
+		ret = PTR_ERR(xspi->ref_clk);
+		goto remove_master;
+	}
+
+	ret = clk_prepare_enable(xspi->pclk);
+	if (ret) {
+		dev_err(&pdev->dev, "Unable to enable APB clock.\n");
+		goto remove_master;
+	}
+
+	ret = clk_prepare_enable(xspi->ref_clk);
+	if (ret) {
+		dev_err(&pdev->dev, "Unable to enable device clock.\n");
+		goto clk_dis_apb;
+	}
+
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_runtime_set_autosuspend_delay(&pdev->dev, SPI_AUTOSUSPEND_TIMEOUT);
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+
+	ret = of_property_read_u32(pdev->dev.of_node, "num-cs", &num_cs);
+	if (ret < 0)
+		master->num_chipselect = CDNS_SPI_DEFAULT_NUM_CS;
+	else
+		master->num_chipselect = num_cs;
+
+	ret = of_property_read_u32(pdev->dev.of_node, "is-decoded-cs",
+				   &xspi->is_decoded_cs);
+	if (ret < 0)
+		xspi->is_decoded_cs = 0;
+
+	/* SPI controller initializations */
+	cdns_spi_init_hw(xspi);
+
+	pm_runtime_mark_last_busy(&pdev->dev);
+	pm_runtime_put_autosuspend(&pdev->dev);
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq <= 0) {
+		ret = -ENXIO;
+		dev_err(&pdev->dev, "irq number is invalid\n");
+		goto clk_dis_all;
+	}
+
+	ret = devm_request_irq(&pdev->dev, irq, cdns_spi_irq,
+			       0, pdev->name, master);
+	if (ret != 0) {
+		ret = -ENXIO;
+		dev_err(&pdev->dev, "request_irq failed\n");
+		goto clk_dis_all;
+	}
+
+	master->prepare_transfer_hardware = cdns_prepare_transfer_hardware;
+	master->prepare_message = cdns_prepare_message;
+	master->transfer_one = cdns_transfer_one;
+	master->unprepare_transfer_hardware = cdns_unprepare_transfer_hardware;
+	master->set_cs = cdns_spi_chipselect;
+	master->setup = cdns_spi_setup;
+	master->cleanup = cdns_spi_cleanup;
+	master->auto_runtime_pm = true;
+	master->mode_bits = SPI_CPOL | SPI_CPHA;
+
+	/* Set to default valid value */
+	master->max_speed_hz = clk_get_rate(xspi->ref_clk) / 4;
+	xspi->speed_hz = master->max_speed_hz;
+
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+
+	ret = spi_register_master(master);
+	if (ret) {
+		dev_err(&pdev->dev, "spi_register_master failed\n");
+		goto clk_dis_all;
+	}
+
+	return ret;
+
+clk_dis_all:
+	pm_runtime_set_suspended(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+	clk_disable_unprepare(xspi->ref_clk);
+clk_dis_apb:
+	clk_disable_unprepare(xspi->pclk);
+remove_master:
+	spi_master_put(master);
+	return ret;
+}
+
+/**
+ * cdns_spi_remove - Remove method for the SPI driver
+ * @pdev:	Pointer to the platform_device structure
+ *
+ * This function is called if a device is physically removed from the system or
+ * if the driver module is being unloaded. It frees all resources allocated to
+ * the device.
+ *
+ * Return:	0 on success and error value on error
+ */
+static int cdns_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct cdns_spi *xspi = spi_master_get_devdata(master);
+
+	cdns_spi_write(xspi, CDNS_SPI_ER, CDNS_SPI_ER_DISABLE);
+
+	clk_disable_unprepare(xspi->ref_clk);
+	clk_disable_unprepare(xspi->pclk);
+	pm_runtime_set_suspended(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+
+	spi_unregister_master(master);
+
+	return 0;
+}
+
+/**
+ * cdns_spi_suspend - Suspend method for the SPI driver
+ * @dev:	Address of the platform_device structure
+ *
+ * This function disables the SPI controller and
+ * changes the driver state to "suspend"
+ *
+ * Return:	0 on success and error value on error
+ */
+static int __maybe_unused cdns_spi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+
+	return spi_master_suspend(master);
+}
+
+/**
+ * cdns_spi_resume - Resume method for the SPI driver
+ * @dev:	Address of the platform_device structure
+ *
+ * This function changes the driver state to "ready"
+ *
+ * Return:	0 on success and error value on error
+ */
+static int __maybe_unused cdns_spi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct cdns_spi *xspi = spi_master_get_devdata(master);
+
+	cdns_spi_init_hw(xspi);
+	return spi_master_resume(master);
+}
+
+/**
+ * cdns_spi_runtime_resume - Runtime resume method for the SPI driver
+ * @dev:	Address of the platform_device structure
+ *
+ * This function enables the clocks
+ *
+ * Return:	0 on success and error value on error
+ */
+static int __maybe_unused cnds_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct cdns_spi *xspi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(xspi->pclk);
+	if (ret) {
+		dev_err(dev, "Cannot enable APB clock.\n");
+		return ret;
+	}
+
+	ret = clk_prepare_enable(xspi->ref_clk);
+	if (ret) {
+		dev_err(dev, "Cannot enable device clock.\n");
+		clk_disable_unprepare(xspi->pclk);
+		return ret;
+	}
+	return 0;
+}
+
+/**
+ * cdns_spi_runtime_suspend - Runtime suspend method for the SPI driver
+ * @dev:	Address of the platform_device structure
+ *
+ * This function disables the clocks
+ *
+ * Return:	Always 0
+ */
+static int __maybe_unused cnds_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct cdns_spi *xspi = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(xspi->ref_clk);
+	clk_disable_unprepare(xspi->pclk);
+
+	return 0;
+}
+
+static const struct dev_pm_ops cdns_spi_dev_pm_ops = {
+	SET_RUNTIME_PM_OPS(cnds_runtime_suspend,
+			   cnds_runtime_resume, NULL)
+	SET_SYSTEM_SLEEP_PM_OPS(cdns_spi_suspend, cdns_spi_resume)
+};
+
+static const struct of_device_id cdns_spi_of_match[] = {
+	{ .compatible = "xlnx,zynq-spi-r1p6" },
+	{ .compatible = "cdns,spi-r1p6" },
+	{ /* end of table */ }
+};
+MODULE_DEVICE_TABLE(of, cdns_spi_of_match);
+
+/* cdns_spi_driver - This structure defines the SPI subsystem platform driver */
+static struct platform_driver cdns_spi_driver = {
+	.probe	= cdns_spi_probe,
+	.remove	= cdns_spi_remove,
+	.driver = {
+		.name = CDNS_SPI_NAME,
+		.of_match_table = cdns_spi_of_match,
+		.pm = &cdns_spi_dev_pm_ops,
+	},
+};
+
+module_platform_driver(cdns_spi_driver);
+
+MODULE_AUTHOR("Xilinx, Inc.");
+MODULE_DESCRIPTION("Cadence SPI driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-cavium-octeon.c b/drivers/spi/spi-cavium-octeon.c
new file mode 100644
index 0000000..ee4703e
--- /dev/null
+++ b/drivers/spi/spi-cavium-octeon.c
@@ -0,0 +1,104 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2011, 2012 Cavium, Inc.
+ */
+
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/module.h>
+#include <linux/io.h>
+#include <linux/of.h>
+
+#include <asm/octeon/octeon.h>
+
+#include "spi-cavium.h"
+
+static int octeon_spi_probe(struct platform_device *pdev)
+{
+	struct resource *res_mem;
+	void __iomem *reg_base;
+	struct spi_master *master;
+	struct octeon_spi *p;
+	int err = -ENOENT;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct octeon_spi));
+	if (!master)
+		return -ENOMEM;
+	p = spi_master_get_devdata(master);
+	platform_set_drvdata(pdev, master);
+
+	res_mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	reg_base = devm_ioremap_resource(&pdev->dev, res_mem);
+	if (IS_ERR(reg_base)) {
+		err = PTR_ERR(reg_base);
+		goto fail;
+	}
+
+	p->register_base = reg_base;
+	p->sys_freq = octeon_get_io_clock_rate();
+
+	p->regs.config = 0;
+	p->regs.status = 0x08;
+	p->regs.tx = 0x10;
+	p->regs.data = 0x80;
+
+	master->num_chipselect = 4;
+	master->mode_bits = SPI_CPHA |
+			    SPI_CPOL |
+			    SPI_CS_HIGH |
+			    SPI_LSB_FIRST |
+			    SPI_3WIRE;
+
+	master->transfer_one_message = octeon_spi_transfer_one_message;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->max_speed_hz = OCTEON_SPI_MAX_CLOCK_HZ;
+
+	master->dev.of_node = pdev->dev.of_node;
+	err = devm_spi_register_master(&pdev->dev, master);
+	if (err) {
+		dev_err(&pdev->dev, "register master failed: %d\n", err);
+		goto fail;
+	}
+
+	dev_info(&pdev->dev, "OCTEON SPI bus driver\n");
+
+	return 0;
+fail:
+	spi_master_put(master);
+	return err;
+}
+
+static int octeon_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct octeon_spi *p = spi_master_get_devdata(master);
+
+	/* Clear the CSENA* and put everything in a known state. */
+	writeq(0, p->register_base + OCTEON_SPI_CFG(p));
+
+	return 0;
+}
+
+static const struct of_device_id octeon_spi_match[] = {
+	{ .compatible = "cavium,octeon-3010-spi", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, octeon_spi_match);
+
+static struct platform_driver octeon_spi_driver = {
+	.driver = {
+		.name		= "spi-octeon",
+		.of_match_table = octeon_spi_match,
+	},
+	.probe		= octeon_spi_probe,
+	.remove		= octeon_spi_remove,
+};
+
+module_platform_driver(octeon_spi_driver);
+
+MODULE_DESCRIPTION("Cavium, Inc. OCTEON SPI bus driver");
+MODULE_AUTHOR("David Daney");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-cavium-thunderx.c b/drivers/spi/spi-cavium-thunderx.c
new file mode 100644
index 0000000..8779377
--- /dev/null
+++ b/drivers/spi/spi-cavium-thunderx.c
@@ -0,0 +1,120 @@
+/*
+ * Cavium ThunderX SPI driver.
+ *
+ * Copyright (C) 2016 Cavium Inc.
+ * Authors: Jan Glauber <jglauber@cavium.com>
+ */
+
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/spi/spi.h>
+
+#include "spi-cavium.h"
+
+#define DRV_NAME "spi-thunderx"
+
+#define SYS_FREQ_DEFAULT 700000000 /* 700 Mhz */
+
+static int thunderx_spi_probe(struct pci_dev *pdev,
+			      const struct pci_device_id *ent)
+{
+	struct device *dev = &pdev->dev;
+	struct spi_master *master;
+	struct octeon_spi *p;
+	int ret;
+
+	master = spi_alloc_master(dev, sizeof(struct octeon_spi));
+	if (!master)
+		return -ENOMEM;
+
+	p = spi_master_get_devdata(master);
+
+	ret = pcim_enable_device(pdev);
+	if (ret)
+		goto error;
+
+	ret = pci_request_regions(pdev, DRV_NAME);
+	if (ret)
+		goto error;
+
+	p->register_base = pcim_iomap(pdev, 0, pci_resource_len(pdev, 0));
+	if (!p->register_base) {
+		ret = -EINVAL;
+		goto error;
+	}
+
+	p->regs.config = 0x1000;
+	p->regs.status = 0x1008;
+	p->regs.tx = 0x1010;
+	p->regs.data = 0x1080;
+
+	p->clk = devm_clk_get(dev, NULL);
+	if (IS_ERR(p->clk)) {
+		ret = PTR_ERR(p->clk);
+		goto error;
+	}
+
+	ret = clk_prepare_enable(p->clk);
+	if (ret)
+		goto error;
+
+	p->sys_freq = clk_get_rate(p->clk);
+	if (!p->sys_freq)
+		p->sys_freq = SYS_FREQ_DEFAULT;
+	dev_info(dev, "Set system clock to %u\n", p->sys_freq);
+
+	master->num_chipselect = 4;
+	master->mode_bits = SPI_CPHA | SPI_CPOL | SPI_CS_HIGH |
+			    SPI_LSB_FIRST | SPI_3WIRE;
+	master->transfer_one_message = octeon_spi_transfer_one_message;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->max_speed_hz = OCTEON_SPI_MAX_CLOCK_HZ;
+	master->dev.of_node = pdev->dev.of_node;
+
+	pci_set_drvdata(pdev, master);
+
+	ret = devm_spi_register_master(dev, master);
+	if (ret)
+		goto error;
+
+	return 0;
+
+error:
+	clk_disable_unprepare(p->clk);
+	spi_master_put(master);
+	return ret;
+}
+
+static void thunderx_spi_remove(struct pci_dev *pdev)
+{
+	struct spi_master *master = pci_get_drvdata(pdev);
+	struct octeon_spi *p;
+
+	p = spi_master_get_devdata(master);
+	if (!p)
+		return;
+
+	clk_disable_unprepare(p->clk);
+	/* Put everything in a known state. */
+	writeq(0, p->register_base + OCTEON_SPI_CFG(p));
+}
+
+static const struct pci_device_id thunderx_spi_pci_id_table[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_CAVIUM, 0xa00b) },
+	{ 0, }
+};
+
+MODULE_DEVICE_TABLE(pci, thunderx_spi_pci_id_table);
+
+static struct pci_driver thunderx_spi_driver = {
+	.name		= DRV_NAME,
+	.id_table	= thunderx_spi_pci_id_table,
+	.probe		= thunderx_spi_probe,
+	.remove		= thunderx_spi_remove,
+};
+
+module_pci_driver(thunderx_spi_driver);
+
+MODULE_DESCRIPTION("Cavium, Inc. ThunderX SPI bus driver");
+MODULE_AUTHOR("Jan Glauber");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-cavium.c b/drivers/spi/spi-cavium.c
new file mode 100644
index 0000000..5aaf215
--- /dev/null
+++ b/drivers/spi/spi-cavium.c
@@ -0,0 +1,151 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2011, 2012 Cavium, Inc.
+ */
+
+#include <linux/spi/spi.h>
+#include <linux/module.h>
+#include <linux/delay.h>
+#include <linux/io.h>
+
+#include "spi-cavium.h"
+
+static void octeon_spi_wait_ready(struct octeon_spi *p)
+{
+	union cvmx_mpi_sts mpi_sts;
+	unsigned int loops = 0;
+
+	do {
+		if (loops++)
+			__delay(500);
+		mpi_sts.u64 = readq(p->register_base + OCTEON_SPI_STS(p));
+	} while (mpi_sts.s.busy);
+}
+
+static int octeon_spi_do_transfer(struct octeon_spi *p,
+				  struct spi_message *msg,
+				  struct spi_transfer *xfer,
+				  bool last_xfer)
+{
+	struct spi_device *spi = msg->spi;
+	union cvmx_mpi_cfg mpi_cfg;
+	union cvmx_mpi_tx mpi_tx;
+	unsigned int clkdiv;
+	int mode;
+	bool cpha, cpol;
+	const u8 *tx_buf;
+	u8 *rx_buf;
+	int len;
+	int i;
+
+	mode = spi->mode;
+	cpha = mode & SPI_CPHA;
+	cpol = mode & SPI_CPOL;
+
+	clkdiv = p->sys_freq / (2 * xfer->speed_hz);
+
+	mpi_cfg.u64 = 0;
+
+	mpi_cfg.s.clkdiv = clkdiv;
+	mpi_cfg.s.cshi = (mode & SPI_CS_HIGH) ? 1 : 0;
+	mpi_cfg.s.lsbfirst = (mode & SPI_LSB_FIRST) ? 1 : 0;
+	mpi_cfg.s.wireor = (mode & SPI_3WIRE) ? 1 : 0;
+	mpi_cfg.s.idlelo = cpha != cpol;
+	mpi_cfg.s.cslate = cpha ? 1 : 0;
+	mpi_cfg.s.enable = 1;
+
+	if (spi->chip_select < 4)
+		p->cs_enax |= 1ull << (12 + spi->chip_select);
+	mpi_cfg.u64 |= p->cs_enax;
+
+	if (mpi_cfg.u64 != p->last_cfg) {
+		p->last_cfg = mpi_cfg.u64;
+		writeq(mpi_cfg.u64, p->register_base + OCTEON_SPI_CFG(p));
+	}
+	tx_buf = xfer->tx_buf;
+	rx_buf = xfer->rx_buf;
+	len = xfer->len;
+	while (len > OCTEON_SPI_MAX_BYTES) {
+		for (i = 0; i < OCTEON_SPI_MAX_BYTES; i++) {
+			u8 d;
+			if (tx_buf)
+				d = *tx_buf++;
+			else
+				d = 0;
+			writeq(d, p->register_base + OCTEON_SPI_DAT0(p) + (8 * i));
+		}
+		mpi_tx.u64 = 0;
+		mpi_tx.s.csid = spi->chip_select;
+		mpi_tx.s.leavecs = 1;
+		mpi_tx.s.txnum = tx_buf ? OCTEON_SPI_MAX_BYTES : 0;
+		mpi_tx.s.totnum = OCTEON_SPI_MAX_BYTES;
+		writeq(mpi_tx.u64, p->register_base + OCTEON_SPI_TX(p));
+
+		octeon_spi_wait_ready(p);
+		if (rx_buf)
+			for (i = 0; i < OCTEON_SPI_MAX_BYTES; i++) {
+				u64 v = readq(p->register_base + OCTEON_SPI_DAT0(p) + (8 * i));
+				*rx_buf++ = (u8)v;
+			}
+		len -= OCTEON_SPI_MAX_BYTES;
+	}
+
+	for (i = 0; i < len; i++) {
+		u8 d;
+		if (tx_buf)
+			d = *tx_buf++;
+		else
+			d = 0;
+		writeq(d, p->register_base + OCTEON_SPI_DAT0(p) + (8 * i));
+	}
+
+	mpi_tx.u64 = 0;
+	mpi_tx.s.csid = spi->chip_select;
+	if (last_xfer)
+		mpi_tx.s.leavecs = xfer->cs_change;
+	else
+		mpi_tx.s.leavecs = !xfer->cs_change;
+	mpi_tx.s.txnum = tx_buf ? len : 0;
+	mpi_tx.s.totnum = len;
+	writeq(mpi_tx.u64, p->register_base + OCTEON_SPI_TX(p));
+
+	octeon_spi_wait_ready(p);
+	if (rx_buf)
+		for (i = 0; i < len; i++) {
+			u64 v = readq(p->register_base + OCTEON_SPI_DAT0(p) + (8 * i));
+			*rx_buf++ = (u8)v;
+		}
+
+	if (xfer->delay_usecs)
+		udelay(xfer->delay_usecs);
+
+	return xfer->len;
+}
+
+int octeon_spi_transfer_one_message(struct spi_master *master,
+				    struct spi_message *msg)
+{
+	struct octeon_spi *p = spi_master_get_devdata(master);
+	unsigned int total_len = 0;
+	int status = 0;
+	struct spi_transfer *xfer;
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		bool last_xfer = list_is_last(&xfer->transfer_list,
+					      &msg->transfers);
+		int r = octeon_spi_do_transfer(p, msg, xfer, last_xfer);
+		if (r < 0) {
+			status = r;
+			goto err;
+		}
+		total_len += r;
+	}
+err:
+	msg->status = status;
+	msg->actual_length = total_len;
+	spi_finalize_current_message(master);
+	return status;
+}
diff --git a/drivers/spi/spi-cavium.h b/drivers/spi/spi-cavium.h
new file mode 100644
index 0000000..1f3ac46
--- /dev/null
+++ b/drivers/spi/spi-cavium.h
@@ -0,0 +1,333 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __SPI_CAVIUM_H
+#define __SPI_CAVIUM_H
+
+#include <linux/clk.h>
+
+#define OCTEON_SPI_MAX_BYTES 9
+#define OCTEON_SPI_MAX_CLOCK_HZ 16000000
+
+struct octeon_spi_regs {
+	int config;
+	int status;
+	int tx;
+	int data;
+};
+
+struct octeon_spi {
+	void __iomem *register_base;
+	u64 last_cfg;
+	u64 cs_enax;
+	int sys_freq;
+	struct octeon_spi_regs regs;
+	struct clk *clk;
+};
+
+#define OCTEON_SPI_CFG(x)	(x->regs.config)
+#define OCTEON_SPI_STS(x)	(x->regs.status)
+#define OCTEON_SPI_TX(x)	(x->regs.tx)
+#define OCTEON_SPI_DAT0(x)	(x->regs.data)
+
+int octeon_spi_transfer_one_message(struct spi_master *master,
+				    struct spi_message *msg);
+
+/* MPI register descriptions */
+
+#define CVMX_MPI_CFG (CVMX_ADD_IO_SEG(0x0001070000001000ull))
+#define CVMX_MPI_DATX(offset) (CVMX_ADD_IO_SEG(0x0001070000001080ull) + ((offset) & 15) * 8)
+#define CVMX_MPI_STS (CVMX_ADD_IO_SEG(0x0001070000001008ull))
+#define CVMX_MPI_TX (CVMX_ADD_IO_SEG(0x0001070000001010ull))
+
+union cvmx_mpi_cfg {
+	uint64_t u64;
+	struct cvmx_mpi_cfg_s {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t reserved_29_63:35;
+		uint64_t clkdiv:13;
+		uint64_t csena3:1;
+		uint64_t csena2:1;
+		uint64_t csena1:1;
+		uint64_t csena0:1;
+		uint64_t cslate:1;
+		uint64_t tritx:1;
+		uint64_t idleclks:2;
+		uint64_t cshi:1;
+		uint64_t csena:1;
+		uint64_t int_ena:1;
+		uint64_t lsbfirst:1;
+		uint64_t wireor:1;
+		uint64_t clk_cont:1;
+		uint64_t idlelo:1;
+		uint64_t enable:1;
+#else
+		uint64_t enable:1;
+		uint64_t idlelo:1;
+		uint64_t clk_cont:1;
+		uint64_t wireor:1;
+		uint64_t lsbfirst:1;
+		uint64_t int_ena:1;
+		uint64_t csena:1;
+		uint64_t cshi:1;
+		uint64_t idleclks:2;
+		uint64_t tritx:1;
+		uint64_t cslate:1;
+		uint64_t csena0:1;
+		uint64_t csena1:1;
+		uint64_t csena2:1;
+		uint64_t csena3:1;
+		uint64_t clkdiv:13;
+		uint64_t reserved_29_63:35;
+#endif
+	} s;
+	struct cvmx_mpi_cfg_cn30xx {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t reserved_29_63:35;
+		uint64_t clkdiv:13;
+		uint64_t reserved_12_15:4;
+		uint64_t cslate:1;
+		uint64_t tritx:1;
+		uint64_t idleclks:2;
+		uint64_t cshi:1;
+		uint64_t csena:1;
+		uint64_t int_ena:1;
+		uint64_t lsbfirst:1;
+		uint64_t wireor:1;
+		uint64_t clk_cont:1;
+		uint64_t idlelo:1;
+		uint64_t enable:1;
+#else
+		uint64_t enable:1;
+		uint64_t idlelo:1;
+		uint64_t clk_cont:1;
+		uint64_t wireor:1;
+		uint64_t lsbfirst:1;
+		uint64_t int_ena:1;
+		uint64_t csena:1;
+		uint64_t cshi:1;
+		uint64_t idleclks:2;
+		uint64_t tritx:1;
+		uint64_t cslate:1;
+		uint64_t reserved_12_15:4;
+		uint64_t clkdiv:13;
+		uint64_t reserved_29_63:35;
+#endif
+	} cn30xx;
+	struct cvmx_mpi_cfg_cn31xx {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t reserved_29_63:35;
+		uint64_t clkdiv:13;
+		uint64_t reserved_11_15:5;
+		uint64_t tritx:1;
+		uint64_t idleclks:2;
+		uint64_t cshi:1;
+		uint64_t csena:1;
+		uint64_t int_ena:1;
+		uint64_t lsbfirst:1;
+		uint64_t wireor:1;
+		uint64_t clk_cont:1;
+		uint64_t idlelo:1;
+		uint64_t enable:1;
+#else
+		uint64_t enable:1;
+		uint64_t idlelo:1;
+		uint64_t clk_cont:1;
+		uint64_t wireor:1;
+		uint64_t lsbfirst:1;
+		uint64_t int_ena:1;
+		uint64_t csena:1;
+		uint64_t cshi:1;
+		uint64_t idleclks:2;
+		uint64_t tritx:1;
+		uint64_t reserved_11_15:5;
+		uint64_t clkdiv:13;
+		uint64_t reserved_29_63:35;
+#endif
+	} cn31xx;
+	struct cvmx_mpi_cfg_cn30xx cn50xx;
+	struct cvmx_mpi_cfg_cn61xx {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t reserved_29_63:35;
+		uint64_t clkdiv:13;
+		uint64_t reserved_14_15:2;
+		uint64_t csena1:1;
+		uint64_t csena0:1;
+		uint64_t cslate:1;
+		uint64_t tritx:1;
+		uint64_t idleclks:2;
+		uint64_t cshi:1;
+		uint64_t reserved_6_6:1;
+		uint64_t int_ena:1;
+		uint64_t lsbfirst:1;
+		uint64_t wireor:1;
+		uint64_t clk_cont:1;
+		uint64_t idlelo:1;
+		uint64_t enable:1;
+#else
+		uint64_t enable:1;
+		uint64_t idlelo:1;
+		uint64_t clk_cont:1;
+		uint64_t wireor:1;
+		uint64_t lsbfirst:1;
+		uint64_t int_ena:1;
+		uint64_t reserved_6_6:1;
+		uint64_t cshi:1;
+		uint64_t idleclks:2;
+		uint64_t tritx:1;
+		uint64_t cslate:1;
+		uint64_t csena0:1;
+		uint64_t csena1:1;
+		uint64_t reserved_14_15:2;
+		uint64_t clkdiv:13;
+		uint64_t reserved_29_63:35;
+#endif
+	} cn61xx;
+	struct cvmx_mpi_cfg_cn66xx {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t reserved_29_63:35;
+		uint64_t clkdiv:13;
+		uint64_t csena3:1;
+		uint64_t csena2:1;
+		uint64_t reserved_12_13:2;
+		uint64_t cslate:1;
+		uint64_t tritx:1;
+		uint64_t idleclks:2;
+		uint64_t cshi:1;
+		uint64_t reserved_6_6:1;
+		uint64_t int_ena:1;
+		uint64_t lsbfirst:1;
+		uint64_t wireor:1;
+		uint64_t clk_cont:1;
+		uint64_t idlelo:1;
+		uint64_t enable:1;
+#else
+		uint64_t enable:1;
+		uint64_t idlelo:1;
+		uint64_t clk_cont:1;
+		uint64_t wireor:1;
+		uint64_t lsbfirst:1;
+		uint64_t int_ena:1;
+		uint64_t reserved_6_6:1;
+		uint64_t cshi:1;
+		uint64_t idleclks:2;
+		uint64_t tritx:1;
+		uint64_t cslate:1;
+		uint64_t reserved_12_13:2;
+		uint64_t csena2:1;
+		uint64_t csena3:1;
+		uint64_t clkdiv:13;
+		uint64_t reserved_29_63:35;
+#endif
+	} cn66xx;
+	struct cvmx_mpi_cfg_cn61xx cnf71xx;
+};
+
+union cvmx_mpi_datx {
+	uint64_t u64;
+	struct cvmx_mpi_datx_s {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t reserved_8_63:56;
+		uint64_t data:8;
+#else
+		uint64_t data:8;
+		uint64_t reserved_8_63:56;
+#endif
+	} s;
+	struct cvmx_mpi_datx_s cn30xx;
+	struct cvmx_mpi_datx_s cn31xx;
+	struct cvmx_mpi_datx_s cn50xx;
+	struct cvmx_mpi_datx_s cn61xx;
+	struct cvmx_mpi_datx_s cn66xx;
+	struct cvmx_mpi_datx_s cnf71xx;
+};
+
+union cvmx_mpi_sts {
+	uint64_t u64;
+	struct cvmx_mpi_sts_s {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t reserved_13_63:51;
+		uint64_t rxnum:5;
+		uint64_t reserved_1_7:7;
+		uint64_t busy:1;
+#else
+		uint64_t busy:1;
+		uint64_t reserved_1_7:7;
+		uint64_t rxnum:5;
+		uint64_t reserved_13_63:51;
+#endif
+	} s;
+	struct cvmx_mpi_sts_s cn30xx;
+	struct cvmx_mpi_sts_s cn31xx;
+	struct cvmx_mpi_sts_s cn50xx;
+	struct cvmx_mpi_sts_s cn61xx;
+	struct cvmx_mpi_sts_s cn66xx;
+	struct cvmx_mpi_sts_s cnf71xx;
+};
+
+union cvmx_mpi_tx {
+	uint64_t u64;
+	struct cvmx_mpi_tx_s {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t reserved_22_63:42;
+		uint64_t csid:2;
+		uint64_t reserved_17_19:3;
+		uint64_t leavecs:1;
+		uint64_t reserved_13_15:3;
+		uint64_t txnum:5;
+		uint64_t reserved_5_7:3;
+		uint64_t totnum:5;
+#else
+		uint64_t totnum:5;
+		uint64_t reserved_5_7:3;
+		uint64_t txnum:5;
+		uint64_t reserved_13_15:3;
+		uint64_t leavecs:1;
+		uint64_t reserved_17_19:3;
+		uint64_t csid:2;
+		uint64_t reserved_22_63:42;
+#endif
+	} s;
+	struct cvmx_mpi_tx_cn30xx {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t reserved_17_63:47;
+		uint64_t leavecs:1;
+		uint64_t reserved_13_15:3;
+		uint64_t txnum:5;
+		uint64_t reserved_5_7:3;
+		uint64_t totnum:5;
+#else
+		uint64_t totnum:5;
+		uint64_t reserved_5_7:3;
+		uint64_t txnum:5;
+		uint64_t reserved_13_15:3;
+		uint64_t leavecs:1;
+		uint64_t reserved_17_63:47;
+#endif
+	} cn30xx;
+	struct cvmx_mpi_tx_cn30xx cn31xx;
+	struct cvmx_mpi_tx_cn30xx cn50xx;
+	struct cvmx_mpi_tx_cn61xx {
+#ifdef __BIG_ENDIAN_BITFIELD
+		uint64_t reserved_21_63:43;
+		uint64_t csid:1;
+		uint64_t reserved_17_19:3;
+		uint64_t leavecs:1;
+		uint64_t reserved_13_15:3;
+		uint64_t txnum:5;
+		uint64_t reserved_5_7:3;
+		uint64_t totnum:5;
+#else
+		uint64_t totnum:5;
+		uint64_t reserved_5_7:3;
+		uint64_t txnum:5;
+		uint64_t reserved_13_15:3;
+		uint64_t leavecs:1;
+		uint64_t reserved_17_19:3;
+		uint64_t csid:1;
+		uint64_t reserved_21_63:43;
+#endif
+	} cn61xx;
+	struct cvmx_mpi_tx_s cn66xx;
+	struct cvmx_mpi_tx_cn61xx cnf71xx;
+};
+
+#endif /* __SPI_CAVIUM_H */
diff --git a/drivers/spi/spi-clps711x.c b/drivers/spi/spi-clps711x.c
new file mode 100644
index 0000000..18193df
--- /dev/null
+++ b/drivers/spi/spi-clps711x.c
@@ -0,0 +1,197 @@
+/*
+ *  CLPS711X SPI bus driver
+ *
+ *  Copyright (C) 2012-2016 Alexander Shiyan <shc_work@mail.ru>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/io.h>
+#include <linux/clk.h>
+#include <linux/gpio.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/regmap.h>
+#include <linux/mfd/syscon.h>
+#include <linux/mfd/syscon/clps711x.h>
+#include <linux/spi/spi.h>
+
+#define DRIVER_NAME		"clps711x-spi"
+
+#define SYNCIO_FRMLEN(x)	((x) << 8)
+#define SYNCIO_TXFRMEN		(1 << 14)
+
+struct spi_clps711x_data {
+	void __iomem		*syncio;
+	struct regmap		*syscon;
+	struct clk		*spi_clk;
+
+	u8			*tx_buf;
+	u8			*rx_buf;
+	unsigned int		bpw;
+	int			len;
+};
+
+static int spi_clps711x_setup(struct spi_device *spi)
+{
+	if (!spi->controller_state) {
+		int ret;
+
+		ret = devm_gpio_request(&spi->master->dev, spi->cs_gpio,
+					dev_name(&spi->master->dev));
+		if (ret)
+			return ret;
+
+		spi->controller_state = spi;
+	}
+
+	/* We are expect that SPI-device is not selected */
+	gpio_direction_output(spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
+
+	return 0;
+}
+
+static int spi_clps711x_prepare_message(struct spi_master *master,
+					struct spi_message *msg)
+{
+	struct spi_clps711x_data *hw = spi_master_get_devdata(master);
+	struct spi_device *spi = msg->spi;
+
+	/* Setup mode for transfer */
+	return regmap_update_bits(hw->syscon, SYSCON_OFFSET, SYSCON3_ADCCKNSEN,
+				  (spi->mode & SPI_CPHA) ?
+				  SYSCON3_ADCCKNSEN : 0);
+}
+
+static int spi_clps711x_transfer_one(struct spi_master *master,
+				     struct spi_device *spi,
+				     struct spi_transfer *xfer)
+{
+	struct spi_clps711x_data *hw = spi_master_get_devdata(master);
+	u8 data;
+
+	clk_set_rate(hw->spi_clk, xfer->speed_hz ? : spi->max_speed_hz);
+
+	hw->len = xfer->len;
+	hw->bpw = xfer->bits_per_word;
+	hw->tx_buf = (u8 *)xfer->tx_buf;
+	hw->rx_buf = (u8 *)xfer->rx_buf;
+
+	/* Initiate transfer */
+	data = hw->tx_buf ? *hw->tx_buf++ : 0;
+	writel(data | SYNCIO_FRMLEN(hw->bpw) | SYNCIO_TXFRMEN, hw->syncio);
+
+	return 1;
+}
+
+static irqreturn_t spi_clps711x_isr(int irq, void *dev_id)
+{
+	struct spi_master *master = dev_id;
+	struct spi_clps711x_data *hw = spi_master_get_devdata(master);
+	u8 data;
+
+	/* Handle RX */
+	data = readb(hw->syncio);
+	if (hw->rx_buf)
+		*hw->rx_buf++ = data;
+
+	/* Handle TX */
+	if (--hw->len > 0) {
+		data = hw->tx_buf ? *hw->tx_buf++ : 0;
+		writel(data | SYNCIO_FRMLEN(hw->bpw) | SYNCIO_TXFRMEN,
+		       hw->syncio);
+	} else
+		spi_finalize_current_transfer(master);
+
+	return IRQ_HANDLED;
+}
+
+static int spi_clps711x_probe(struct platform_device *pdev)
+{
+	struct spi_clps711x_data *hw;
+	struct spi_master *master;
+	struct resource *res;
+	int irq, ret;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0)
+		return irq;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*hw));
+	if (!master)
+		return -ENOMEM;
+
+	master->bus_num = -1;
+	master->mode_bits = SPI_CPHA | SPI_CS_HIGH;
+	master->bits_per_word_mask =  SPI_BPW_RANGE_MASK(1, 8);
+	master->dev.of_node = pdev->dev.of_node;
+	master->setup = spi_clps711x_setup;
+	master->prepare_message = spi_clps711x_prepare_message;
+	master->transfer_one = spi_clps711x_transfer_one;
+
+	hw = spi_master_get_devdata(master);
+
+	hw->spi_clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(hw->spi_clk)) {
+		ret = PTR_ERR(hw->spi_clk);
+		goto err_out;
+	}
+
+	hw->syscon =
+		syscon_regmap_lookup_by_compatible("cirrus,ep7209-syscon3");
+	if (IS_ERR(hw->syscon)) {
+		ret = PTR_ERR(hw->syscon);
+		goto err_out;
+	}
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	hw->syncio = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(hw->syncio)) {
+		ret = PTR_ERR(hw->syncio);
+		goto err_out;
+	}
+
+	/* Disable extended mode due hardware problems */
+	regmap_update_bits(hw->syscon, SYSCON_OFFSET, SYSCON3_ADCCON, 0);
+
+	/* Clear possible pending interrupt */
+	readl(hw->syncio);
+
+	ret = devm_request_irq(&pdev->dev, irq, spi_clps711x_isr, 0,
+			       dev_name(&pdev->dev), master);
+	if (ret)
+		goto err_out;
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (!ret)
+		return 0;
+
+err_out:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static const struct of_device_id clps711x_spi_dt_ids[] = {
+	{ .compatible = "cirrus,ep7209-spi", },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, clps711x_spi_dt_ids);
+
+static struct platform_driver clps711x_spi_driver = {
+	.driver	= {
+		.name	= DRIVER_NAME,
+		.of_match_table = clps711x_spi_dt_ids,
+	},
+	.probe	= spi_clps711x_probe,
+};
+module_platform_driver(clps711x_spi_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Alexander Shiyan <shc_work@mail.ru>");
+MODULE_DESCRIPTION("CLPS711X SPI bus driver");
+MODULE_ALIAS("platform:" DRIVER_NAME);
diff --git a/drivers/spi/spi-coldfire-qspi.c b/drivers/spi/spi-coldfire-qspi.c
new file mode 100644
index 0000000..23f6fff
--- /dev/null
+++ b/drivers/spi/spi-coldfire-qspi.c
@@ -0,0 +1,530 @@
+/*
+ * Freescale/Motorola Coldfire Queued SPI driver
+ *
+ * Copyright 2010 Steven King <sfking@fdwdc.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+*/
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/errno.h>
+#include <linux/platform_device.h>
+#include <linux/sched.h>
+#include <linux/delay.h>
+#include <linux/io.h>
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/spi/spi.h>
+#include <linux/pm_runtime.h>
+
+#include <asm/coldfire.h>
+#include <asm/mcfsim.h>
+#include <asm/mcfqspi.h>
+
+#define	DRIVER_NAME "mcfqspi"
+
+#define	MCFQSPI_BUSCLK			(MCF_BUSCLK / 2)
+
+#define	MCFQSPI_QMR			0x00
+#define		MCFQSPI_QMR_MSTR	0x8000
+#define		MCFQSPI_QMR_CPOL	0x0200
+#define		MCFQSPI_QMR_CPHA	0x0100
+#define	MCFQSPI_QDLYR			0x04
+#define		MCFQSPI_QDLYR_SPE	0x8000
+#define	MCFQSPI_QWR			0x08
+#define		MCFQSPI_QWR_HALT	0x8000
+#define		MCFQSPI_QWR_WREN	0x4000
+#define		MCFQSPI_QWR_CSIV	0x1000
+#define	MCFQSPI_QIR			0x0C
+#define		MCFQSPI_QIR_WCEFB	0x8000
+#define		MCFQSPI_QIR_ABRTB	0x4000
+#define		MCFQSPI_QIR_ABRTL	0x1000
+#define		MCFQSPI_QIR_WCEFE	0x0800
+#define		MCFQSPI_QIR_ABRTE	0x0400
+#define		MCFQSPI_QIR_SPIFE	0x0100
+#define		MCFQSPI_QIR_WCEF	0x0008
+#define		MCFQSPI_QIR_ABRT	0x0004
+#define		MCFQSPI_QIR_SPIF	0x0001
+#define	MCFQSPI_QAR			0x010
+#define		MCFQSPI_QAR_TXBUF	0x00
+#define		MCFQSPI_QAR_RXBUF	0x10
+#define		MCFQSPI_QAR_CMDBUF	0x20
+#define	MCFQSPI_QDR			0x014
+#define	MCFQSPI_QCR			0x014
+#define		MCFQSPI_QCR_CONT	0x8000
+#define		MCFQSPI_QCR_BITSE	0x4000
+#define		MCFQSPI_QCR_DT		0x2000
+
+struct mcfqspi {
+	void __iomem *iobase;
+	int irq;
+	struct clk *clk;
+	struct mcfqspi_cs_control *cs_control;
+
+	wait_queue_head_t waitq;
+};
+
+static void mcfqspi_wr_qmr(struct mcfqspi *mcfqspi, u16 val)
+{
+	writew(val, mcfqspi->iobase + MCFQSPI_QMR);
+}
+
+static void mcfqspi_wr_qdlyr(struct mcfqspi *mcfqspi, u16 val)
+{
+	writew(val, mcfqspi->iobase + MCFQSPI_QDLYR);
+}
+
+static u16 mcfqspi_rd_qdlyr(struct mcfqspi *mcfqspi)
+{
+	return readw(mcfqspi->iobase + MCFQSPI_QDLYR);
+}
+
+static void mcfqspi_wr_qwr(struct mcfqspi *mcfqspi, u16 val)
+{
+	writew(val, mcfqspi->iobase + MCFQSPI_QWR);
+}
+
+static void mcfqspi_wr_qir(struct mcfqspi *mcfqspi, u16 val)
+{
+	writew(val, mcfqspi->iobase + MCFQSPI_QIR);
+}
+
+static void mcfqspi_wr_qar(struct mcfqspi *mcfqspi, u16 val)
+{
+	writew(val, mcfqspi->iobase + MCFQSPI_QAR);
+}
+
+static void mcfqspi_wr_qdr(struct mcfqspi *mcfqspi, u16 val)
+{
+	writew(val, mcfqspi->iobase + MCFQSPI_QDR);
+}
+
+static u16 mcfqspi_rd_qdr(struct mcfqspi *mcfqspi)
+{
+	return readw(mcfqspi->iobase + MCFQSPI_QDR);
+}
+
+static void mcfqspi_cs_select(struct mcfqspi *mcfqspi, u8 chip_select,
+			    bool cs_high)
+{
+	mcfqspi->cs_control->select(mcfqspi->cs_control, chip_select, cs_high);
+}
+
+static void mcfqspi_cs_deselect(struct mcfqspi *mcfqspi, u8 chip_select,
+				bool cs_high)
+{
+	mcfqspi->cs_control->deselect(mcfqspi->cs_control, chip_select, cs_high);
+}
+
+static int mcfqspi_cs_setup(struct mcfqspi *mcfqspi)
+{
+	return (mcfqspi->cs_control->setup) ?
+		mcfqspi->cs_control->setup(mcfqspi->cs_control) : 0;
+}
+
+static void mcfqspi_cs_teardown(struct mcfqspi *mcfqspi)
+{
+	if (mcfqspi->cs_control->teardown)
+		mcfqspi->cs_control->teardown(mcfqspi->cs_control);
+}
+
+static u8 mcfqspi_qmr_baud(u32 speed_hz)
+{
+	return clamp((MCFQSPI_BUSCLK + speed_hz - 1) / speed_hz, 2u, 255u);
+}
+
+static bool mcfqspi_qdlyr_spe(struct mcfqspi *mcfqspi)
+{
+	return mcfqspi_rd_qdlyr(mcfqspi) & MCFQSPI_QDLYR_SPE;
+}
+
+static irqreturn_t mcfqspi_irq_handler(int this_irq, void *dev_id)
+{
+	struct mcfqspi *mcfqspi = dev_id;
+
+	/* clear interrupt */
+	mcfqspi_wr_qir(mcfqspi, MCFQSPI_QIR_SPIFE | MCFQSPI_QIR_SPIF);
+	wake_up(&mcfqspi->waitq);
+
+	return IRQ_HANDLED;
+}
+
+static void mcfqspi_transfer_msg8(struct mcfqspi *mcfqspi, unsigned count,
+				  const u8 *txbuf, u8 *rxbuf)
+{
+	unsigned i, n, offset = 0;
+
+	n = min(count, 16u);
+
+	mcfqspi_wr_qar(mcfqspi, MCFQSPI_QAR_CMDBUF);
+	for (i = 0; i < n; ++i)
+		mcfqspi_wr_qdr(mcfqspi, MCFQSPI_QCR_BITSE);
+
+	mcfqspi_wr_qar(mcfqspi, MCFQSPI_QAR_TXBUF);
+	if (txbuf)
+		for (i = 0; i < n; ++i)
+			mcfqspi_wr_qdr(mcfqspi, *txbuf++);
+	else
+		for (i = 0; i < count; ++i)
+			mcfqspi_wr_qdr(mcfqspi, 0);
+
+	count -= n;
+	if (count) {
+		u16 qwr = 0xf08;
+		mcfqspi_wr_qwr(mcfqspi, 0x700);
+		mcfqspi_wr_qdlyr(mcfqspi, MCFQSPI_QDLYR_SPE);
+
+		do {
+			wait_event(mcfqspi->waitq, !mcfqspi_qdlyr_spe(mcfqspi));
+			mcfqspi_wr_qwr(mcfqspi, qwr);
+			mcfqspi_wr_qdlyr(mcfqspi, MCFQSPI_QDLYR_SPE);
+			if (rxbuf) {
+				mcfqspi_wr_qar(mcfqspi,
+					       MCFQSPI_QAR_RXBUF + offset);
+				for (i = 0; i < 8; ++i)
+					*rxbuf++ = mcfqspi_rd_qdr(mcfqspi);
+			}
+			n = min(count, 8u);
+			if (txbuf) {
+				mcfqspi_wr_qar(mcfqspi,
+					       MCFQSPI_QAR_TXBUF + offset);
+				for (i = 0; i < n; ++i)
+					mcfqspi_wr_qdr(mcfqspi, *txbuf++);
+			}
+			qwr = (offset ? 0x808 : 0) + ((n - 1) << 8);
+			offset ^= 8;
+			count -= n;
+		} while (count);
+		wait_event(mcfqspi->waitq, !mcfqspi_qdlyr_spe(mcfqspi));
+		mcfqspi_wr_qwr(mcfqspi, qwr);
+		mcfqspi_wr_qdlyr(mcfqspi, MCFQSPI_QDLYR_SPE);
+		if (rxbuf) {
+			mcfqspi_wr_qar(mcfqspi, MCFQSPI_QAR_RXBUF + offset);
+			for (i = 0; i < 8; ++i)
+				*rxbuf++ = mcfqspi_rd_qdr(mcfqspi);
+			offset ^= 8;
+		}
+	} else {
+		mcfqspi_wr_qwr(mcfqspi, (n - 1) << 8);
+		mcfqspi_wr_qdlyr(mcfqspi, MCFQSPI_QDLYR_SPE);
+	}
+	wait_event(mcfqspi->waitq, !mcfqspi_qdlyr_spe(mcfqspi));
+	if (rxbuf) {
+		mcfqspi_wr_qar(mcfqspi, MCFQSPI_QAR_RXBUF + offset);
+		for (i = 0; i < n; ++i)
+			*rxbuf++ = mcfqspi_rd_qdr(mcfqspi);
+	}
+}
+
+static void mcfqspi_transfer_msg16(struct mcfqspi *mcfqspi, unsigned count,
+				   const u16 *txbuf, u16 *rxbuf)
+{
+	unsigned i, n, offset = 0;
+
+	n = min(count, 16u);
+
+	mcfqspi_wr_qar(mcfqspi, MCFQSPI_QAR_CMDBUF);
+	for (i = 0; i < n; ++i)
+		mcfqspi_wr_qdr(mcfqspi, MCFQSPI_QCR_BITSE);
+
+	mcfqspi_wr_qar(mcfqspi, MCFQSPI_QAR_TXBUF);
+	if (txbuf)
+		for (i = 0; i < n; ++i)
+			mcfqspi_wr_qdr(mcfqspi, *txbuf++);
+	else
+		for (i = 0; i < count; ++i)
+			mcfqspi_wr_qdr(mcfqspi, 0);
+
+	count -= n;
+	if (count) {
+		u16 qwr = 0xf08;
+		mcfqspi_wr_qwr(mcfqspi, 0x700);
+		mcfqspi_wr_qdlyr(mcfqspi, MCFQSPI_QDLYR_SPE);
+
+		do {
+			wait_event(mcfqspi->waitq, !mcfqspi_qdlyr_spe(mcfqspi));
+			mcfqspi_wr_qwr(mcfqspi, qwr);
+			mcfqspi_wr_qdlyr(mcfqspi, MCFQSPI_QDLYR_SPE);
+			if (rxbuf) {
+				mcfqspi_wr_qar(mcfqspi,
+					       MCFQSPI_QAR_RXBUF + offset);
+				for (i = 0; i < 8; ++i)
+					*rxbuf++ = mcfqspi_rd_qdr(mcfqspi);
+			}
+			n = min(count, 8u);
+			if (txbuf) {
+				mcfqspi_wr_qar(mcfqspi,
+					       MCFQSPI_QAR_TXBUF + offset);
+				for (i = 0; i < n; ++i)
+					mcfqspi_wr_qdr(mcfqspi, *txbuf++);
+			}
+			qwr = (offset ? 0x808 : 0x000) + ((n - 1) << 8);
+			offset ^= 8;
+			count -= n;
+		} while (count);
+		wait_event(mcfqspi->waitq, !mcfqspi_qdlyr_spe(mcfqspi));
+		mcfqspi_wr_qwr(mcfqspi, qwr);
+		mcfqspi_wr_qdlyr(mcfqspi, MCFQSPI_QDLYR_SPE);
+		if (rxbuf) {
+			mcfqspi_wr_qar(mcfqspi, MCFQSPI_QAR_RXBUF + offset);
+			for (i = 0; i < 8; ++i)
+				*rxbuf++ = mcfqspi_rd_qdr(mcfqspi);
+			offset ^= 8;
+		}
+	} else {
+		mcfqspi_wr_qwr(mcfqspi, (n - 1) << 8);
+		mcfqspi_wr_qdlyr(mcfqspi, MCFQSPI_QDLYR_SPE);
+	}
+	wait_event(mcfqspi->waitq, !mcfqspi_qdlyr_spe(mcfqspi));
+	if (rxbuf) {
+		mcfqspi_wr_qar(mcfqspi, MCFQSPI_QAR_RXBUF + offset);
+		for (i = 0; i < n; ++i)
+			*rxbuf++ = mcfqspi_rd_qdr(mcfqspi);
+	}
+}
+
+static void mcfqspi_set_cs(struct spi_device *spi, bool enable)
+{
+	struct mcfqspi *mcfqspi = spi_master_get_devdata(spi->master);
+	bool cs_high = spi->mode & SPI_CS_HIGH;
+
+	if (enable)
+		mcfqspi_cs_select(mcfqspi, spi->chip_select, cs_high);
+	else
+		mcfqspi_cs_deselect(mcfqspi, spi->chip_select, cs_high);
+}
+
+static int mcfqspi_transfer_one(struct spi_master *master,
+				struct spi_device *spi,
+				struct spi_transfer *t)
+{
+	struct mcfqspi *mcfqspi = spi_master_get_devdata(master);
+	u16 qmr = MCFQSPI_QMR_MSTR;
+
+	qmr |= t->bits_per_word << 10;
+	if (spi->mode & SPI_CPHA)
+		qmr |= MCFQSPI_QMR_CPHA;
+	if (spi->mode & SPI_CPOL)
+		qmr |= MCFQSPI_QMR_CPOL;
+	qmr |= mcfqspi_qmr_baud(t->speed_hz);
+	mcfqspi_wr_qmr(mcfqspi, qmr);
+
+	mcfqspi_wr_qir(mcfqspi, MCFQSPI_QIR_SPIFE);
+	if (t->bits_per_word == 8)
+		mcfqspi_transfer_msg8(mcfqspi, t->len, t->tx_buf, t->rx_buf);
+	else
+		mcfqspi_transfer_msg16(mcfqspi, t->len / 2, t->tx_buf,
+				       t->rx_buf);
+	mcfqspi_wr_qir(mcfqspi, 0);
+
+	return 0;
+}
+
+static int mcfqspi_setup(struct spi_device *spi)
+{
+	mcfqspi_cs_deselect(spi_master_get_devdata(spi->master),
+			    spi->chip_select, spi->mode & SPI_CS_HIGH);
+
+	dev_dbg(&spi->dev,
+			"bits per word %d, chip select %d, speed %d KHz\n",
+			spi->bits_per_word, spi->chip_select,
+			(MCFQSPI_BUSCLK / mcfqspi_qmr_baud(spi->max_speed_hz))
+			/ 1000);
+
+	return 0;
+}
+
+static int mcfqspi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct mcfqspi *mcfqspi;
+	struct resource *res;
+	struct mcfqspi_platform_data *pdata;
+	int status;
+
+	pdata = dev_get_platdata(&pdev->dev);
+	if (!pdata) {
+		dev_dbg(&pdev->dev, "platform data is missing\n");
+		return -ENOENT;
+	}
+
+	if (!pdata->cs_control) {
+		dev_dbg(&pdev->dev, "pdata->cs_control is NULL\n");
+		return -EINVAL;
+	}
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*mcfqspi));
+	if (master == NULL) {
+		dev_dbg(&pdev->dev, "spi_alloc_master failed\n");
+		return -ENOMEM;
+	}
+
+	mcfqspi = spi_master_get_devdata(master);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	mcfqspi->iobase = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(mcfqspi->iobase)) {
+		status = PTR_ERR(mcfqspi->iobase);
+		goto fail0;
+	}
+
+	mcfqspi->irq = platform_get_irq(pdev, 0);
+	if (mcfqspi->irq < 0) {
+		dev_dbg(&pdev->dev, "platform_get_irq failed\n");
+		status = -ENXIO;
+		goto fail0;
+	}
+
+	status = devm_request_irq(&pdev->dev, mcfqspi->irq, mcfqspi_irq_handler,
+				0, pdev->name, mcfqspi);
+	if (status) {
+		dev_dbg(&pdev->dev, "request_irq failed\n");
+		goto fail0;
+	}
+
+	mcfqspi->clk = devm_clk_get(&pdev->dev, "qspi_clk");
+	if (IS_ERR(mcfqspi->clk)) {
+		dev_dbg(&pdev->dev, "clk_get failed\n");
+		status = PTR_ERR(mcfqspi->clk);
+		goto fail0;
+	}
+	clk_enable(mcfqspi->clk);
+
+	master->bus_num = pdata->bus_num;
+	master->num_chipselect = pdata->num_chipselect;
+
+	mcfqspi->cs_control = pdata->cs_control;
+	status = mcfqspi_cs_setup(mcfqspi);
+	if (status) {
+		dev_dbg(&pdev->dev, "error initializing cs_control\n");
+		goto fail1;
+	}
+
+	init_waitqueue_head(&mcfqspi->waitq);
+
+	master->mode_bits = SPI_CS_HIGH | SPI_CPOL | SPI_CPHA;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(8, 16);
+	master->setup = mcfqspi_setup;
+	master->set_cs = mcfqspi_set_cs;
+	master->transfer_one = mcfqspi_transfer_one;
+	master->auto_runtime_pm = true;
+
+	platform_set_drvdata(pdev, master);
+	pm_runtime_enable(&pdev->dev);
+
+	status = devm_spi_register_master(&pdev->dev, master);
+	if (status) {
+		dev_dbg(&pdev->dev, "spi_register_master failed\n");
+		goto fail2;
+	}
+
+	dev_info(&pdev->dev, "Coldfire QSPI bus driver\n");
+
+	return 0;
+
+fail2:
+	pm_runtime_disable(&pdev->dev);
+	mcfqspi_cs_teardown(mcfqspi);
+fail1:
+	clk_disable(mcfqspi->clk);
+fail0:
+	spi_master_put(master);
+
+	dev_dbg(&pdev->dev, "Coldfire QSPI probe failed\n");
+
+	return status;
+}
+
+static int mcfqspi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct mcfqspi *mcfqspi = spi_master_get_devdata(master);
+
+	pm_runtime_disable(&pdev->dev);
+	/* disable the hardware (set the baud rate to 0) */
+	mcfqspi_wr_qmr(mcfqspi, MCFQSPI_QMR_MSTR);
+
+	mcfqspi_cs_teardown(mcfqspi);
+	clk_disable(mcfqspi->clk);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int mcfqspi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mcfqspi *mcfqspi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = spi_master_suspend(master);
+	if (ret)
+		return ret;
+
+	clk_disable(mcfqspi->clk);
+
+	return 0;
+}
+
+static int mcfqspi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mcfqspi *mcfqspi = spi_master_get_devdata(master);
+
+	clk_enable(mcfqspi->clk);
+
+	return spi_master_resume(master);
+}
+#endif
+
+#ifdef CONFIG_PM
+static int mcfqspi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mcfqspi *mcfqspi = spi_master_get_devdata(master);
+
+	clk_disable(mcfqspi->clk);
+
+	return 0;
+}
+
+static int mcfqspi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mcfqspi *mcfqspi = spi_master_get_devdata(master);
+
+	clk_enable(mcfqspi->clk);
+
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops mcfqspi_pm = {
+	SET_SYSTEM_SLEEP_PM_OPS(mcfqspi_suspend, mcfqspi_resume)
+	SET_RUNTIME_PM_OPS(mcfqspi_runtime_suspend, mcfqspi_runtime_resume,
+			NULL)
+};
+
+static struct platform_driver mcfqspi_driver = {
+	.driver.name	= DRIVER_NAME,
+	.driver.owner	= THIS_MODULE,
+	.driver.pm	= &mcfqspi_pm,
+	.probe		= mcfqspi_probe,
+	.remove		= mcfqspi_remove,
+};
+module_platform_driver(mcfqspi_driver);
+
+MODULE_AUTHOR("Steven King <sfking@fdwdc.com>");
+MODULE_DESCRIPTION("Coldfire QSPI Controller Driver");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:" DRIVER_NAME);
diff --git a/drivers/spi/spi-davinci.c b/drivers/spi/spi-davinci.c
new file mode 100644
index 0000000..a02099c
--- /dev/null
+++ b/drivers/spi/spi-davinci.c
@@ -0,0 +1,1110 @@
+/*
+ * Copyright (C) 2009 Texas Instruments.
+ * Copyright (C) 2010 EF Johnson Technologies
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/gpio.h>
+#include <linux/module.h>
+#include <linux/delay.h>
+#include <linux/platform_device.h>
+#include <linux/err.h>
+#include <linux/clk.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/of_gpio.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/slab.h>
+
+#include <linux/platform_data/spi-davinci.h>
+
+#define CS_DEFAULT	0xFF
+
+#define SPIFMT_PHASE_MASK	BIT(16)
+#define SPIFMT_POLARITY_MASK	BIT(17)
+#define SPIFMT_DISTIMER_MASK	BIT(18)
+#define SPIFMT_SHIFTDIR_MASK	BIT(20)
+#define SPIFMT_WAITENA_MASK	BIT(21)
+#define SPIFMT_PARITYENA_MASK	BIT(22)
+#define SPIFMT_ODD_PARITY_MASK	BIT(23)
+#define SPIFMT_WDELAY_MASK	0x3f000000u
+#define SPIFMT_WDELAY_SHIFT	24
+#define SPIFMT_PRESCALE_SHIFT	8
+
+/* SPIPC0 */
+#define SPIPC0_DIFUN_MASK	BIT(11)		/* MISO */
+#define SPIPC0_DOFUN_MASK	BIT(10)		/* MOSI */
+#define SPIPC0_CLKFUN_MASK	BIT(9)		/* CLK */
+#define SPIPC0_SPIENA_MASK	BIT(8)		/* nREADY */
+
+#define SPIINT_MASKALL		0x0101035F
+#define SPIINT_MASKINT		0x0000015F
+#define SPI_INTLVL_1		0x000001FF
+#define SPI_INTLVL_0		0x00000000
+
+/* SPIDAT1 (upper 16 bit defines) */
+#define SPIDAT1_CSHOLD_MASK	BIT(12)
+#define SPIDAT1_WDEL		BIT(10)
+
+/* SPIGCR1 */
+#define SPIGCR1_CLKMOD_MASK	BIT(1)
+#define SPIGCR1_MASTER_MASK     BIT(0)
+#define SPIGCR1_POWERDOWN_MASK	BIT(8)
+#define SPIGCR1_LOOPBACK_MASK	BIT(16)
+#define SPIGCR1_SPIENA_MASK	BIT(24)
+
+/* SPIBUF */
+#define SPIBUF_TXFULL_MASK	BIT(29)
+#define SPIBUF_RXEMPTY_MASK	BIT(31)
+
+/* SPIDELAY */
+#define SPIDELAY_C2TDELAY_SHIFT 24
+#define SPIDELAY_C2TDELAY_MASK  (0xFF << SPIDELAY_C2TDELAY_SHIFT)
+#define SPIDELAY_T2CDELAY_SHIFT 16
+#define SPIDELAY_T2CDELAY_MASK  (0xFF << SPIDELAY_T2CDELAY_SHIFT)
+#define SPIDELAY_T2EDELAY_SHIFT 8
+#define SPIDELAY_T2EDELAY_MASK  (0xFF << SPIDELAY_T2EDELAY_SHIFT)
+#define SPIDELAY_C2EDELAY_SHIFT 0
+#define SPIDELAY_C2EDELAY_MASK  0xFF
+
+/* Error Masks */
+#define SPIFLG_DLEN_ERR_MASK		BIT(0)
+#define SPIFLG_TIMEOUT_MASK		BIT(1)
+#define SPIFLG_PARERR_MASK		BIT(2)
+#define SPIFLG_DESYNC_MASK		BIT(3)
+#define SPIFLG_BITERR_MASK		BIT(4)
+#define SPIFLG_OVRRUN_MASK		BIT(6)
+#define SPIFLG_BUF_INIT_ACTIVE_MASK	BIT(24)
+#define SPIFLG_ERROR_MASK		(SPIFLG_DLEN_ERR_MASK \
+				| SPIFLG_TIMEOUT_MASK | SPIFLG_PARERR_MASK \
+				| SPIFLG_DESYNC_MASK | SPIFLG_BITERR_MASK \
+				| SPIFLG_OVRRUN_MASK)
+
+#define SPIINT_DMA_REQ_EN	BIT(16)
+
+/* SPI Controller registers */
+#define SPIGCR0		0x00
+#define SPIGCR1		0x04
+#define SPIINT		0x08
+#define SPILVL		0x0c
+#define SPIFLG		0x10
+#define SPIPC0		0x14
+#define SPIDAT1		0x3c
+#define SPIBUF		0x40
+#define SPIDELAY	0x48
+#define SPIDEF		0x4c
+#define SPIFMT0		0x50
+
+#define DMA_MIN_BYTES	16
+
+/* SPI Controller driver's private data. */
+struct davinci_spi {
+	struct spi_bitbang	bitbang;
+	struct clk		*clk;
+
+	u8			version;
+	resource_size_t		pbase;
+	void __iomem		*base;
+	u32			irq;
+	struct completion	done;
+
+	const void		*tx;
+	void			*rx;
+	int			rcount;
+	int			wcount;
+
+	struct dma_chan		*dma_rx;
+	struct dma_chan		*dma_tx;
+
+	struct davinci_spi_platform_data pdata;
+
+	void			(*get_rx)(u32 rx_data, struct davinci_spi *);
+	u32			(*get_tx)(struct davinci_spi *);
+
+	u8			*bytes_per_word;
+
+	u8			prescaler_limit;
+};
+
+static struct davinci_spi_config davinci_spi_default_cfg;
+
+static void davinci_spi_rx_buf_u8(u32 data, struct davinci_spi *dspi)
+{
+	if (dspi->rx) {
+		u8 *rx = dspi->rx;
+		*rx++ = (u8)data;
+		dspi->rx = rx;
+	}
+}
+
+static void davinci_spi_rx_buf_u16(u32 data, struct davinci_spi *dspi)
+{
+	if (dspi->rx) {
+		u16 *rx = dspi->rx;
+		*rx++ = (u16)data;
+		dspi->rx = rx;
+	}
+}
+
+static u32 davinci_spi_tx_buf_u8(struct davinci_spi *dspi)
+{
+	u32 data = 0;
+
+	if (dspi->tx) {
+		const u8 *tx = dspi->tx;
+
+		data = *tx++;
+		dspi->tx = tx;
+	}
+	return data;
+}
+
+static u32 davinci_spi_tx_buf_u16(struct davinci_spi *dspi)
+{
+	u32 data = 0;
+
+	if (dspi->tx) {
+		const u16 *tx = dspi->tx;
+
+		data = *tx++;
+		dspi->tx = tx;
+	}
+	return data;
+}
+
+static inline void set_io_bits(void __iomem *addr, u32 bits)
+{
+	u32 v = ioread32(addr);
+
+	v |= bits;
+	iowrite32(v, addr);
+}
+
+static inline void clear_io_bits(void __iomem *addr, u32 bits)
+{
+	u32 v = ioread32(addr);
+
+	v &= ~bits;
+	iowrite32(v, addr);
+}
+
+/*
+ * Interface to control the chip select signal
+ */
+static void davinci_spi_chipselect(struct spi_device *spi, int value)
+{
+	struct davinci_spi *dspi;
+	struct davinci_spi_platform_data *pdata;
+	struct davinci_spi_config *spicfg = spi->controller_data;
+	u8 chip_sel = spi->chip_select;
+	u16 spidat1 = CS_DEFAULT;
+
+	dspi = spi_master_get_devdata(spi->master);
+	pdata = &dspi->pdata;
+
+	/* program delay transfers if tx_delay is non zero */
+	if (spicfg && spicfg->wdelay)
+		spidat1 |= SPIDAT1_WDEL;
+
+	/*
+	 * Board specific chip select logic decides the polarity and cs
+	 * line for the controller
+	 */
+	if (spi->cs_gpio >= 0) {
+		if (value == BITBANG_CS_ACTIVE)
+			gpio_set_value(spi->cs_gpio, spi->mode & SPI_CS_HIGH);
+		else
+			gpio_set_value(spi->cs_gpio,
+				!(spi->mode & SPI_CS_HIGH));
+	} else {
+		if (value == BITBANG_CS_ACTIVE) {
+			spidat1 |= SPIDAT1_CSHOLD_MASK;
+			spidat1 &= ~(0x1 << chip_sel);
+		}
+	}
+
+	iowrite16(spidat1, dspi->base + SPIDAT1 + 2);
+}
+
+/**
+ * davinci_spi_get_prescale - Calculates the correct prescale value
+ * @maxspeed_hz: the maximum rate the SPI clock can run at
+ *
+ * This function calculates the prescale value that generates a clock rate
+ * less than or equal to the specified maximum.
+ *
+ * Returns: calculated prescale value for easy programming into SPI registers
+ * or negative error number if valid prescalar cannot be updated.
+ */
+static inline int davinci_spi_get_prescale(struct davinci_spi *dspi,
+							u32 max_speed_hz)
+{
+	int ret;
+
+	/* Subtract 1 to match what will be programmed into SPI register. */
+	ret = DIV_ROUND_UP(clk_get_rate(dspi->clk), max_speed_hz) - 1;
+
+	if (ret < dspi->prescaler_limit || ret > 255)
+		return -EINVAL;
+
+	return ret;
+}
+
+/**
+ * davinci_spi_setup_transfer - This functions will determine transfer method
+ * @spi: spi device on which data transfer to be done
+ * @t: spi transfer in which transfer info is filled
+ *
+ * This function determines data transfer method (8/16/32 bit transfer).
+ * It will also set the SPI Clock Control register according to
+ * SPI slave device freq.
+ */
+static int davinci_spi_setup_transfer(struct spi_device *spi,
+		struct spi_transfer *t)
+{
+
+	struct davinci_spi *dspi;
+	struct davinci_spi_config *spicfg;
+	u8 bits_per_word = 0;
+	u32 hz = 0, spifmt = 0;
+	int prescale;
+
+	dspi = spi_master_get_devdata(spi->master);
+	spicfg = spi->controller_data;
+	if (!spicfg)
+		spicfg = &davinci_spi_default_cfg;
+
+	if (t) {
+		bits_per_word = t->bits_per_word;
+		hz = t->speed_hz;
+	}
+
+	/* if bits_per_word is not set then set it default */
+	if (!bits_per_word)
+		bits_per_word = spi->bits_per_word;
+
+	/*
+	 * Assign function pointer to appropriate transfer method
+	 * 8bit, 16bit or 32bit transfer
+	 */
+	if (bits_per_word <= 8) {
+		dspi->get_rx = davinci_spi_rx_buf_u8;
+		dspi->get_tx = davinci_spi_tx_buf_u8;
+		dspi->bytes_per_word[spi->chip_select] = 1;
+	} else {
+		dspi->get_rx = davinci_spi_rx_buf_u16;
+		dspi->get_tx = davinci_spi_tx_buf_u16;
+		dspi->bytes_per_word[spi->chip_select] = 2;
+	}
+
+	if (!hz)
+		hz = spi->max_speed_hz;
+
+	/* Set up SPIFMTn register, unique to this chipselect. */
+
+	prescale = davinci_spi_get_prescale(dspi, hz);
+	if (prescale < 0)
+		return prescale;
+
+	spifmt = (prescale << SPIFMT_PRESCALE_SHIFT) | (bits_per_word & 0x1f);
+
+	if (spi->mode & SPI_LSB_FIRST)
+		spifmt |= SPIFMT_SHIFTDIR_MASK;
+
+	if (spi->mode & SPI_CPOL)
+		spifmt |= SPIFMT_POLARITY_MASK;
+
+	if (!(spi->mode & SPI_CPHA))
+		spifmt |= SPIFMT_PHASE_MASK;
+
+	/*
+	* Assume wdelay is used only on SPI peripherals that has this field
+	* in SPIFMTn register and when it's configured from board file or DT.
+	*/
+	if (spicfg->wdelay)
+		spifmt |= ((spicfg->wdelay << SPIFMT_WDELAY_SHIFT)
+				& SPIFMT_WDELAY_MASK);
+
+	/*
+	 * Version 1 hardware supports two basic SPI modes:
+	 *  - Standard SPI mode uses 4 pins, with chipselect
+	 *  - 3 pin SPI is a 4 pin variant without CS (SPI_NO_CS)
+	 *	(distinct from SPI_3WIRE, with just one data wire;
+	 *	or similar variants without MOSI or without MISO)
+	 *
+	 * Version 2 hardware supports an optional handshaking signal,
+	 * so it can support two more modes:
+	 *  - 5 pin SPI variant is standard SPI plus SPI_READY
+	 *  - 4 pin with enable is (SPI_READY | SPI_NO_CS)
+	 */
+
+	if (dspi->version == SPI_VERSION_2) {
+
+		u32 delay = 0;
+
+		if (spicfg->odd_parity)
+			spifmt |= SPIFMT_ODD_PARITY_MASK;
+
+		if (spicfg->parity_enable)
+			spifmt |= SPIFMT_PARITYENA_MASK;
+
+		if (spicfg->timer_disable) {
+			spifmt |= SPIFMT_DISTIMER_MASK;
+		} else {
+			delay |= (spicfg->c2tdelay << SPIDELAY_C2TDELAY_SHIFT)
+						& SPIDELAY_C2TDELAY_MASK;
+			delay |= (spicfg->t2cdelay << SPIDELAY_T2CDELAY_SHIFT)
+						& SPIDELAY_T2CDELAY_MASK;
+		}
+
+		if (spi->mode & SPI_READY) {
+			spifmt |= SPIFMT_WAITENA_MASK;
+			delay |= (spicfg->t2edelay << SPIDELAY_T2EDELAY_SHIFT)
+						& SPIDELAY_T2EDELAY_MASK;
+			delay |= (spicfg->c2edelay << SPIDELAY_C2EDELAY_SHIFT)
+						& SPIDELAY_C2EDELAY_MASK;
+		}
+
+		iowrite32(delay, dspi->base + SPIDELAY);
+	}
+
+	iowrite32(spifmt, dspi->base + SPIFMT0);
+
+	return 0;
+}
+
+static int davinci_spi_of_setup(struct spi_device *spi)
+{
+	struct davinci_spi_config *spicfg = spi->controller_data;
+	struct device_node *np = spi->dev.of_node;
+	struct davinci_spi *dspi = spi_master_get_devdata(spi->master);
+	u32 prop;
+
+	if (spicfg == NULL && np) {
+		spicfg = kzalloc(sizeof(*spicfg), GFP_KERNEL);
+		if (!spicfg)
+			return -ENOMEM;
+		*spicfg = davinci_spi_default_cfg;
+		/* override with dt configured values */
+		if (!of_property_read_u32(np, "ti,spi-wdelay", &prop))
+			spicfg->wdelay = (u8)prop;
+		spi->controller_data = spicfg;
+
+		if (dspi->dma_rx && dspi->dma_tx)
+			spicfg->io_type = SPI_IO_TYPE_DMA;
+	}
+
+	return 0;
+}
+
+/**
+ * davinci_spi_setup - This functions will set default transfer method
+ * @spi: spi device on which data transfer to be done
+ *
+ * This functions sets the default transfer method.
+ */
+static int davinci_spi_setup(struct spi_device *spi)
+{
+	int retval = 0;
+	struct davinci_spi *dspi;
+	struct davinci_spi_platform_data *pdata;
+	struct spi_master *master = spi->master;
+	struct device_node *np = spi->dev.of_node;
+	bool internal_cs = true;
+
+	dspi = spi_master_get_devdata(spi->master);
+	pdata = &dspi->pdata;
+
+	if (!(spi->mode & SPI_NO_CS)) {
+		if (np && (master->cs_gpios != NULL) && (spi->cs_gpio >= 0)) {
+			retval = gpio_direction_output(
+				      spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
+			internal_cs = false;
+		} else if (pdata->chip_sel &&
+			   spi->chip_select < pdata->num_chipselect &&
+			   pdata->chip_sel[spi->chip_select] != SPI_INTERN_CS) {
+			spi->cs_gpio = pdata->chip_sel[spi->chip_select];
+			retval = gpio_direction_output(
+				      spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
+			internal_cs = false;
+		}
+
+		if (retval) {
+			dev_err(&spi->dev, "GPIO %d setup failed (%d)\n",
+				spi->cs_gpio, retval);
+			return retval;
+		}
+
+		if (internal_cs)
+			set_io_bits(dspi->base + SPIPC0, 1 << spi->chip_select);
+	}
+
+	if (spi->mode & SPI_READY)
+		set_io_bits(dspi->base + SPIPC0, SPIPC0_SPIENA_MASK);
+
+	if (spi->mode & SPI_LOOP)
+		set_io_bits(dspi->base + SPIGCR1, SPIGCR1_LOOPBACK_MASK);
+	else
+		clear_io_bits(dspi->base + SPIGCR1, SPIGCR1_LOOPBACK_MASK);
+
+	return davinci_spi_of_setup(spi);
+}
+
+static void davinci_spi_cleanup(struct spi_device *spi)
+{
+	struct davinci_spi_config *spicfg = spi->controller_data;
+
+	spi->controller_data = NULL;
+	if (spi->dev.of_node)
+		kfree(spicfg);
+}
+
+static bool davinci_spi_can_dma(struct spi_master *master,
+				struct spi_device *spi,
+				struct spi_transfer *xfer)
+{
+	struct davinci_spi_config *spicfg = spi->controller_data;
+	bool can_dma = false;
+
+	if (spicfg)
+		can_dma = (spicfg->io_type == SPI_IO_TYPE_DMA) &&
+			(xfer->len >= DMA_MIN_BYTES) &&
+			!is_vmalloc_addr(xfer->rx_buf) &&
+			!is_vmalloc_addr(xfer->tx_buf);
+
+	return can_dma;
+}
+
+static int davinci_spi_check_error(struct davinci_spi *dspi, int int_status)
+{
+	struct device *sdev = dspi->bitbang.master->dev.parent;
+
+	if (int_status & SPIFLG_TIMEOUT_MASK) {
+		dev_err(sdev, "SPI Time-out Error\n");
+		return -ETIMEDOUT;
+	}
+	if (int_status & SPIFLG_DESYNC_MASK) {
+		dev_err(sdev, "SPI Desynchronization Error\n");
+		return -EIO;
+	}
+	if (int_status & SPIFLG_BITERR_MASK) {
+		dev_err(sdev, "SPI Bit error\n");
+		return -EIO;
+	}
+
+	if (dspi->version == SPI_VERSION_2) {
+		if (int_status & SPIFLG_DLEN_ERR_MASK) {
+			dev_err(sdev, "SPI Data Length Error\n");
+			return -EIO;
+		}
+		if (int_status & SPIFLG_PARERR_MASK) {
+			dev_err(sdev, "SPI Parity Error\n");
+			return -EIO;
+		}
+		if (int_status & SPIFLG_OVRRUN_MASK) {
+			dev_err(sdev, "SPI Data Overrun error\n");
+			return -EIO;
+		}
+		if (int_status & SPIFLG_BUF_INIT_ACTIVE_MASK) {
+			dev_err(sdev, "SPI Buffer Init Active\n");
+			return -EBUSY;
+		}
+	}
+
+	return 0;
+}
+
+/**
+ * davinci_spi_process_events - check for and handle any SPI controller events
+ * @dspi: the controller data
+ *
+ * This function will check the SPIFLG register and handle any events that are
+ * detected there
+ */
+static int davinci_spi_process_events(struct davinci_spi *dspi)
+{
+	u32 buf, status, errors = 0, spidat1;
+
+	buf = ioread32(dspi->base + SPIBUF);
+
+	if (dspi->rcount > 0 && !(buf & SPIBUF_RXEMPTY_MASK)) {
+		dspi->get_rx(buf & 0xFFFF, dspi);
+		dspi->rcount--;
+	}
+
+	status = ioread32(dspi->base + SPIFLG);
+
+	if (unlikely(status & SPIFLG_ERROR_MASK)) {
+		errors = status & SPIFLG_ERROR_MASK;
+		goto out;
+	}
+
+	if (dspi->wcount > 0 && !(buf & SPIBUF_TXFULL_MASK)) {
+		spidat1 = ioread32(dspi->base + SPIDAT1);
+		dspi->wcount--;
+		spidat1 &= ~0xFFFF;
+		spidat1 |= 0xFFFF & dspi->get_tx(dspi);
+		iowrite32(spidat1, dspi->base + SPIDAT1);
+	}
+
+out:
+	return errors;
+}
+
+static void davinci_spi_dma_rx_callback(void *data)
+{
+	struct davinci_spi *dspi = (struct davinci_spi *)data;
+
+	dspi->rcount = 0;
+
+	if (!dspi->wcount && !dspi->rcount)
+		complete(&dspi->done);
+}
+
+static void davinci_spi_dma_tx_callback(void *data)
+{
+	struct davinci_spi *dspi = (struct davinci_spi *)data;
+
+	dspi->wcount = 0;
+
+	if (!dspi->wcount && !dspi->rcount)
+		complete(&dspi->done);
+}
+
+/**
+ * davinci_spi_bufs - functions which will handle transfer data
+ * @spi: spi device on which data transfer to be done
+ * @t: spi transfer in which transfer info is filled
+ *
+ * This function will put data to be transferred into data register
+ * of SPI controller and then wait until the completion will be marked
+ * by the IRQ Handler.
+ */
+static int davinci_spi_bufs(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct davinci_spi *dspi;
+	int data_type, ret = -ENOMEM;
+	u32 tx_data, spidat1;
+	u32 errors = 0;
+	struct davinci_spi_config *spicfg;
+	struct davinci_spi_platform_data *pdata;
+	unsigned uninitialized_var(rx_buf_count);
+
+	dspi = spi_master_get_devdata(spi->master);
+	pdata = &dspi->pdata;
+	spicfg = (struct davinci_spi_config *)spi->controller_data;
+	if (!spicfg)
+		spicfg = &davinci_spi_default_cfg;
+
+	/* convert len to words based on bits_per_word */
+	data_type = dspi->bytes_per_word[spi->chip_select];
+
+	dspi->tx = t->tx_buf;
+	dspi->rx = t->rx_buf;
+	dspi->wcount = t->len / data_type;
+	dspi->rcount = dspi->wcount;
+
+	spidat1 = ioread32(dspi->base + SPIDAT1);
+
+	clear_io_bits(dspi->base + SPIGCR1, SPIGCR1_POWERDOWN_MASK);
+	set_io_bits(dspi->base + SPIGCR1, SPIGCR1_SPIENA_MASK);
+
+	reinit_completion(&dspi->done);
+
+	if (!davinci_spi_can_dma(spi->master, spi, t)) {
+		if (spicfg->io_type != SPI_IO_TYPE_POLL)
+			set_io_bits(dspi->base + SPIINT, SPIINT_MASKINT);
+		/* start the transfer */
+		dspi->wcount--;
+		tx_data = dspi->get_tx(dspi);
+		spidat1 &= 0xFFFF0000;
+		spidat1 |= tx_data & 0xFFFF;
+		iowrite32(spidat1, dspi->base + SPIDAT1);
+	} else {
+		struct dma_slave_config dma_rx_conf = {
+			.direction = DMA_DEV_TO_MEM,
+			.src_addr = (unsigned long)dspi->pbase + SPIBUF,
+			.src_addr_width = data_type,
+			.src_maxburst = 1,
+		};
+		struct dma_slave_config dma_tx_conf = {
+			.direction = DMA_MEM_TO_DEV,
+			.dst_addr = (unsigned long)dspi->pbase + SPIDAT1,
+			.dst_addr_width = data_type,
+			.dst_maxburst = 1,
+		};
+		struct dma_async_tx_descriptor *rxdesc;
+		struct dma_async_tx_descriptor *txdesc;
+
+		dmaengine_slave_config(dspi->dma_rx, &dma_rx_conf);
+		dmaengine_slave_config(dspi->dma_tx, &dma_tx_conf);
+
+		rxdesc = dmaengine_prep_slave_sg(dspi->dma_rx,
+				t->rx_sg.sgl, t->rx_sg.nents, DMA_DEV_TO_MEM,
+				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!rxdesc)
+			goto err_desc;
+
+		if (!t->tx_buf) {
+			/* To avoid errors when doing rx-only transfers with
+			 * many SG entries (> 20), use the rx buffer as the
+			 * dummy tx buffer so that dma reloads are done at the
+			 * same time for rx and tx.
+			 */
+			t->tx_sg.sgl = t->rx_sg.sgl;
+			t->tx_sg.nents = t->rx_sg.nents;
+		}
+
+		txdesc = dmaengine_prep_slave_sg(dspi->dma_tx,
+				t->tx_sg.sgl, t->tx_sg.nents, DMA_MEM_TO_DEV,
+				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!txdesc)
+			goto err_desc;
+
+		rxdesc->callback = davinci_spi_dma_rx_callback;
+		rxdesc->callback_param = (void *)dspi;
+		txdesc->callback = davinci_spi_dma_tx_callback;
+		txdesc->callback_param = (void *)dspi;
+
+		if (pdata->cshold_bug)
+			iowrite16(spidat1 >> 16, dspi->base + SPIDAT1 + 2);
+
+		dmaengine_submit(rxdesc);
+		dmaengine_submit(txdesc);
+
+		dma_async_issue_pending(dspi->dma_rx);
+		dma_async_issue_pending(dspi->dma_tx);
+
+		set_io_bits(dspi->base + SPIINT, SPIINT_DMA_REQ_EN);
+	}
+
+	/* Wait for the transfer to complete */
+	if (spicfg->io_type != SPI_IO_TYPE_POLL) {
+		if (wait_for_completion_timeout(&dspi->done, HZ) == 0)
+			errors = SPIFLG_TIMEOUT_MASK;
+	} else {
+		while (dspi->rcount > 0 || dspi->wcount > 0) {
+			errors = davinci_spi_process_events(dspi);
+			if (errors)
+				break;
+			cpu_relax();
+		}
+	}
+
+	clear_io_bits(dspi->base + SPIINT, SPIINT_MASKALL);
+	if (davinci_spi_can_dma(spi->master, spi, t))
+		clear_io_bits(dspi->base + SPIINT, SPIINT_DMA_REQ_EN);
+
+	clear_io_bits(dspi->base + SPIGCR1, SPIGCR1_SPIENA_MASK);
+	set_io_bits(dspi->base + SPIGCR1, SPIGCR1_POWERDOWN_MASK);
+
+	/*
+	 * Check for bit error, desync error,parity error,timeout error and
+	 * receive overflow errors
+	 */
+	if (errors) {
+		ret = davinci_spi_check_error(dspi, errors);
+		WARN(!ret, "%s: error reported but no error found!\n",
+							dev_name(&spi->dev));
+		return ret;
+	}
+
+	if (dspi->rcount != 0 || dspi->wcount != 0) {
+		dev_err(&spi->dev, "SPI data transfer error\n");
+		return -EIO;
+	}
+
+	return t->len;
+
+err_desc:
+	return ret;
+}
+
+/**
+ * dummy_thread_fn - dummy thread function
+ * @irq: IRQ number for this SPI Master
+ * @context_data: structure for SPI Master controller davinci_spi
+ *
+ * This is to satisfy the request_threaded_irq() API so that the irq
+ * handler is called in interrupt context.
+ */
+static irqreturn_t dummy_thread_fn(s32 irq, void *data)
+{
+	return IRQ_HANDLED;
+}
+
+/**
+ * davinci_spi_irq - Interrupt handler for SPI Master Controller
+ * @irq: IRQ number for this SPI Master
+ * @context_data: structure for SPI Master controller davinci_spi
+ *
+ * ISR will determine that interrupt arrives either for READ or WRITE command.
+ * According to command it will do the appropriate action. It will check
+ * transfer length and if it is not zero then dispatch transfer command again.
+ * If transfer length is zero then it will indicate the COMPLETION so that
+ * davinci_spi_bufs function can go ahead.
+ */
+static irqreturn_t davinci_spi_irq(s32 irq, void *data)
+{
+	struct davinci_spi *dspi = data;
+	int status;
+
+	status = davinci_spi_process_events(dspi);
+	if (unlikely(status != 0))
+		clear_io_bits(dspi->base + SPIINT, SPIINT_MASKINT);
+
+	if ((!dspi->rcount && !dspi->wcount) || status)
+		complete(&dspi->done);
+
+	return IRQ_HANDLED;
+}
+
+static int davinci_spi_request_dma(struct davinci_spi *dspi)
+{
+	struct device *sdev = dspi->bitbang.master->dev.parent;
+
+	dspi->dma_rx = dma_request_chan(sdev, "rx");
+	if (IS_ERR(dspi->dma_rx))
+		return PTR_ERR(dspi->dma_rx);
+
+	dspi->dma_tx = dma_request_chan(sdev, "tx");
+	if (IS_ERR(dspi->dma_tx)) {
+		dma_release_channel(dspi->dma_rx);
+		return PTR_ERR(dspi->dma_tx);
+	}
+
+	return 0;
+}
+
+#if defined(CONFIG_OF)
+
+/* OF SPI data structure */
+struct davinci_spi_of_data {
+	u8	version;
+	u8	prescaler_limit;
+};
+
+static const struct davinci_spi_of_data dm6441_spi_data = {
+	.version = SPI_VERSION_1,
+	.prescaler_limit = 2,
+};
+
+static const struct davinci_spi_of_data da830_spi_data = {
+	.version = SPI_VERSION_2,
+	.prescaler_limit = 2,
+};
+
+static const struct davinci_spi_of_data keystone_spi_data = {
+	.version = SPI_VERSION_1,
+	.prescaler_limit = 0,
+};
+
+static const struct of_device_id davinci_spi_of_match[] = {
+	{
+		.compatible = "ti,dm6441-spi",
+		.data = &dm6441_spi_data,
+	},
+	{
+		.compatible = "ti,da830-spi",
+		.data = &da830_spi_data,
+	},
+	{
+		.compatible = "ti,keystone-spi",
+		.data = &keystone_spi_data,
+	},
+	{ },
+};
+MODULE_DEVICE_TABLE(of, davinci_spi_of_match);
+
+/**
+ * spi_davinci_get_pdata - Get platform data from DTS binding
+ * @pdev: ptr to platform data
+ * @dspi: ptr to driver data
+ *
+ * Parses and populates pdata in dspi from device tree bindings.
+ *
+ * NOTE: Not all platform data params are supported currently.
+ */
+static int spi_davinci_get_pdata(struct platform_device *pdev,
+			struct davinci_spi *dspi)
+{
+	struct device_node *node = pdev->dev.of_node;
+	struct davinci_spi_of_data *spi_data;
+	struct davinci_spi_platform_data *pdata;
+	unsigned int num_cs, intr_line = 0;
+	const struct of_device_id *match;
+
+	pdata = &dspi->pdata;
+
+	match = of_match_device(davinci_spi_of_match, &pdev->dev);
+	if (!match)
+		return -ENODEV;
+
+	spi_data = (struct davinci_spi_of_data *)match->data;
+
+	pdata->version = spi_data->version;
+	pdata->prescaler_limit = spi_data->prescaler_limit;
+	/*
+	 * default num_cs is 1 and all chipsel are internal to the chip
+	 * indicated by chip_sel being NULL or cs_gpios being NULL or
+	 * set to -ENOENT. num-cs includes internal as well as gpios.
+	 * indicated by chip_sel being NULL. GPIO based CS is not
+	 * supported yet in DT bindings.
+	 */
+	num_cs = 1;
+	of_property_read_u32(node, "num-cs", &num_cs);
+	pdata->num_chipselect = num_cs;
+	of_property_read_u32(node, "ti,davinci-spi-intr-line", &intr_line);
+	pdata->intr_line = intr_line;
+	return 0;
+}
+#else
+static int spi_davinci_get_pdata(struct platform_device *pdev,
+			struct davinci_spi *dspi)
+{
+	return -ENODEV;
+}
+#endif
+
+/**
+ * davinci_spi_probe - probe function for SPI Master Controller
+ * @pdev: platform_device structure which contains plateform specific data
+ *
+ * According to Linux Device Model this function will be invoked by Linux
+ * with platform_device struct which contains the device specific info.
+ * This function will map the SPI controller's memory, register IRQ,
+ * Reset SPI controller and setting its registers to default value.
+ * It will invoke spi_bitbang_start to create work queue so that client driver
+ * can register transfer method to work queue.
+ */
+static int davinci_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct davinci_spi *dspi;
+	struct davinci_spi_platform_data *pdata;
+	struct resource *r;
+	int ret = 0;
+	u32 spipc0;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct davinci_spi));
+	if (master == NULL) {
+		ret = -ENOMEM;
+		goto err;
+	}
+
+	platform_set_drvdata(pdev, master);
+
+	dspi = spi_master_get_devdata(master);
+
+	if (dev_get_platdata(&pdev->dev)) {
+		pdata = dev_get_platdata(&pdev->dev);
+		dspi->pdata = *pdata;
+	} else {
+		/* update dspi pdata with that from the DT */
+		ret = spi_davinci_get_pdata(pdev, dspi);
+		if (ret < 0)
+			goto free_master;
+	}
+
+	/* pdata in dspi is now updated and point pdata to that */
+	pdata = &dspi->pdata;
+
+	dspi->bytes_per_word = devm_kcalloc(&pdev->dev,
+					    pdata->num_chipselect,
+					    sizeof(*dspi->bytes_per_word),
+					    GFP_KERNEL);
+	if (dspi->bytes_per_word == NULL) {
+		ret = -ENOMEM;
+		goto free_master;
+	}
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (r == NULL) {
+		ret = -ENOENT;
+		goto free_master;
+	}
+
+	dspi->pbase = r->start;
+
+	dspi->base = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(dspi->base)) {
+		ret = PTR_ERR(dspi->base);
+		goto free_master;
+	}
+
+	init_completion(&dspi->done);
+
+	ret = platform_get_irq(pdev, 0);
+	if (ret == 0)
+		ret = -EINVAL;
+	if (ret < 0)
+		goto free_master;
+	dspi->irq = ret;
+
+	ret = devm_request_threaded_irq(&pdev->dev, dspi->irq, davinci_spi_irq,
+				dummy_thread_fn, 0, dev_name(&pdev->dev), dspi);
+	if (ret)
+		goto free_master;
+
+	dspi->bitbang.master = master;
+
+	dspi->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(dspi->clk)) {
+		ret = -ENODEV;
+		goto free_master;
+	}
+	ret = clk_prepare_enable(dspi->clk);
+	if (ret)
+		goto free_master;
+
+	master->dev.of_node = pdev->dev.of_node;
+	master->bus_num = pdev->id;
+	master->num_chipselect = pdata->num_chipselect;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(2, 16);
+	master->flags = SPI_MASTER_MUST_RX;
+	master->setup = davinci_spi_setup;
+	master->cleanup = davinci_spi_cleanup;
+	master->can_dma = davinci_spi_can_dma;
+
+	dspi->bitbang.chipselect = davinci_spi_chipselect;
+	dspi->bitbang.setup_transfer = davinci_spi_setup_transfer;
+	dspi->prescaler_limit = pdata->prescaler_limit;
+	dspi->version = pdata->version;
+
+	dspi->bitbang.flags = SPI_NO_CS | SPI_LSB_FIRST | SPI_LOOP;
+	if (dspi->version == SPI_VERSION_2)
+		dspi->bitbang.flags |= SPI_READY;
+
+	if (pdev->dev.of_node) {
+		int i;
+
+		for (i = 0; i < pdata->num_chipselect; i++) {
+			int cs_gpio = of_get_named_gpio(pdev->dev.of_node,
+							"cs-gpios", i);
+
+			if (cs_gpio == -EPROBE_DEFER) {
+				ret = cs_gpio;
+				goto free_clk;
+			}
+
+			if (gpio_is_valid(cs_gpio)) {
+				ret = devm_gpio_request(&pdev->dev, cs_gpio,
+							dev_name(&pdev->dev));
+				if (ret)
+					goto free_clk;
+			}
+		}
+	}
+
+	dspi->bitbang.txrx_bufs = davinci_spi_bufs;
+
+	ret = davinci_spi_request_dma(dspi);
+	if (ret == -EPROBE_DEFER) {
+		goto free_clk;
+	} else if (ret) {
+		dev_info(&pdev->dev, "DMA is not supported (%d)\n", ret);
+		dspi->dma_rx = NULL;
+		dspi->dma_tx = NULL;
+	}
+
+	dspi->get_rx = davinci_spi_rx_buf_u8;
+	dspi->get_tx = davinci_spi_tx_buf_u8;
+
+	/* Reset In/OUT SPI module */
+	iowrite32(0, dspi->base + SPIGCR0);
+	udelay(100);
+	iowrite32(1, dspi->base + SPIGCR0);
+
+	/* Set up SPIPC0.  CS and ENA init is done in davinci_spi_setup */
+	spipc0 = SPIPC0_DIFUN_MASK | SPIPC0_DOFUN_MASK | SPIPC0_CLKFUN_MASK;
+	iowrite32(spipc0, dspi->base + SPIPC0);
+
+	if (pdata->intr_line)
+		iowrite32(SPI_INTLVL_1, dspi->base + SPILVL);
+	else
+		iowrite32(SPI_INTLVL_0, dspi->base + SPILVL);
+
+	iowrite32(CS_DEFAULT, dspi->base + SPIDEF);
+
+	/* master mode default */
+	set_io_bits(dspi->base + SPIGCR1, SPIGCR1_CLKMOD_MASK);
+	set_io_bits(dspi->base + SPIGCR1, SPIGCR1_MASTER_MASK);
+	set_io_bits(dspi->base + SPIGCR1, SPIGCR1_POWERDOWN_MASK);
+
+	ret = spi_bitbang_start(&dspi->bitbang);
+	if (ret)
+		goto free_dma;
+
+	dev_info(&pdev->dev, "Controller at 0x%p\n", dspi->base);
+
+	return ret;
+
+free_dma:
+	if (dspi->dma_rx) {
+		dma_release_channel(dspi->dma_rx);
+		dma_release_channel(dspi->dma_tx);
+	}
+free_clk:
+	clk_disable_unprepare(dspi->clk);
+free_master:
+	spi_master_put(master);
+err:
+	return ret;
+}
+
+/**
+ * davinci_spi_remove - remove function for SPI Master Controller
+ * @pdev: platform_device structure which contains plateform specific data
+ *
+ * This function will do the reverse action of davinci_spi_probe function
+ * It will free the IRQ and SPI controller's memory region.
+ * It will also call spi_bitbang_stop to destroy the work queue which was
+ * created by spi_bitbang_start.
+ */
+static int davinci_spi_remove(struct platform_device *pdev)
+{
+	struct davinci_spi *dspi;
+	struct spi_master *master;
+
+	master = platform_get_drvdata(pdev);
+	dspi = spi_master_get_devdata(master);
+
+	spi_bitbang_stop(&dspi->bitbang);
+
+	clk_disable_unprepare(dspi->clk);
+	spi_master_put(master);
+
+	if (dspi->dma_rx) {
+		dma_release_channel(dspi->dma_rx);
+		dma_release_channel(dspi->dma_tx);
+	}
+
+	return 0;
+}
+
+static struct platform_driver davinci_spi_driver = {
+	.driver = {
+		.name = "spi_davinci",
+		.of_match_table = of_match_ptr(davinci_spi_of_match),
+	},
+	.probe = davinci_spi_probe,
+	.remove = davinci_spi_remove,
+};
+module_platform_driver(davinci_spi_driver);
+
+MODULE_DESCRIPTION("TI DaVinci SPI Master Controller Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-dln2.c b/drivers/spi/spi-dln2.c
new file mode 100644
index 0000000..b62a99c
--- /dev/null
+++ b/drivers/spi/spi-dln2.c
@@ -0,0 +1,883 @@
+/*
+ * Driver for the Diolan DLN-2 USB-SPI adapter
+ *
+ * Copyright (c) 2014 Intel Corporation
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation, version 2.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/mfd/dln2.h>
+#include <linux/spi/spi.h>
+#include <linux/pm_runtime.h>
+#include <asm/unaligned.h>
+
+#define DLN2_SPI_MODULE_ID		0x02
+#define DLN2_SPI_CMD(cmd)		DLN2_CMD(cmd, DLN2_SPI_MODULE_ID)
+
+/* SPI commands */
+#define DLN2_SPI_GET_PORT_COUNT			DLN2_SPI_CMD(0x00)
+#define DLN2_SPI_ENABLE				DLN2_SPI_CMD(0x11)
+#define DLN2_SPI_DISABLE			DLN2_SPI_CMD(0x12)
+#define DLN2_SPI_IS_ENABLED			DLN2_SPI_CMD(0x13)
+#define DLN2_SPI_SET_MODE			DLN2_SPI_CMD(0x14)
+#define DLN2_SPI_GET_MODE			DLN2_SPI_CMD(0x15)
+#define DLN2_SPI_SET_FRAME_SIZE			DLN2_SPI_CMD(0x16)
+#define DLN2_SPI_GET_FRAME_SIZE			DLN2_SPI_CMD(0x17)
+#define DLN2_SPI_SET_FREQUENCY			DLN2_SPI_CMD(0x18)
+#define DLN2_SPI_GET_FREQUENCY			DLN2_SPI_CMD(0x19)
+#define DLN2_SPI_READ_WRITE			DLN2_SPI_CMD(0x1A)
+#define DLN2_SPI_READ				DLN2_SPI_CMD(0x1B)
+#define DLN2_SPI_WRITE				DLN2_SPI_CMD(0x1C)
+#define DLN2_SPI_SET_DELAY_BETWEEN_SS		DLN2_SPI_CMD(0x20)
+#define DLN2_SPI_GET_DELAY_BETWEEN_SS		DLN2_SPI_CMD(0x21)
+#define DLN2_SPI_SET_DELAY_AFTER_SS		DLN2_SPI_CMD(0x22)
+#define DLN2_SPI_GET_DELAY_AFTER_SS		DLN2_SPI_CMD(0x23)
+#define DLN2_SPI_SET_DELAY_BETWEEN_FRAMES	DLN2_SPI_CMD(0x24)
+#define DLN2_SPI_GET_DELAY_BETWEEN_FRAMES	DLN2_SPI_CMD(0x25)
+#define DLN2_SPI_SET_SS				DLN2_SPI_CMD(0x26)
+#define DLN2_SPI_GET_SS				DLN2_SPI_CMD(0x27)
+#define DLN2_SPI_RELEASE_SS			DLN2_SPI_CMD(0x28)
+#define DLN2_SPI_SS_VARIABLE_ENABLE		DLN2_SPI_CMD(0x2B)
+#define DLN2_SPI_SS_VARIABLE_DISABLE		DLN2_SPI_CMD(0x2C)
+#define DLN2_SPI_SS_VARIABLE_IS_ENABLED		DLN2_SPI_CMD(0x2D)
+#define DLN2_SPI_SS_AAT_ENABLE			DLN2_SPI_CMD(0x2E)
+#define DLN2_SPI_SS_AAT_DISABLE			DLN2_SPI_CMD(0x2F)
+#define DLN2_SPI_SS_AAT_IS_ENABLED		DLN2_SPI_CMD(0x30)
+#define DLN2_SPI_SS_BETWEEN_FRAMES_ENABLE	DLN2_SPI_CMD(0x31)
+#define DLN2_SPI_SS_BETWEEN_FRAMES_DISABLE	DLN2_SPI_CMD(0x32)
+#define DLN2_SPI_SS_BETWEEN_FRAMES_IS_ENABLED	DLN2_SPI_CMD(0x33)
+#define DLN2_SPI_SET_CPHA			DLN2_SPI_CMD(0x34)
+#define DLN2_SPI_GET_CPHA			DLN2_SPI_CMD(0x35)
+#define DLN2_SPI_SET_CPOL			DLN2_SPI_CMD(0x36)
+#define DLN2_SPI_GET_CPOL			DLN2_SPI_CMD(0x37)
+#define DLN2_SPI_SS_MULTI_ENABLE		DLN2_SPI_CMD(0x38)
+#define DLN2_SPI_SS_MULTI_DISABLE		DLN2_SPI_CMD(0x39)
+#define DLN2_SPI_SS_MULTI_IS_ENABLED		DLN2_SPI_CMD(0x3A)
+#define DLN2_SPI_GET_SUPPORTED_MODES		DLN2_SPI_CMD(0x40)
+#define DLN2_SPI_GET_SUPPORTED_CPHA_VALUES	DLN2_SPI_CMD(0x41)
+#define DLN2_SPI_GET_SUPPORTED_CPOL_VALUES	DLN2_SPI_CMD(0x42)
+#define DLN2_SPI_GET_SUPPORTED_FRAME_SIZES	DLN2_SPI_CMD(0x43)
+#define DLN2_SPI_GET_SS_COUNT			DLN2_SPI_CMD(0x44)
+#define DLN2_SPI_GET_MIN_FREQUENCY		DLN2_SPI_CMD(0x45)
+#define DLN2_SPI_GET_MAX_FREQUENCY		DLN2_SPI_CMD(0x46)
+#define DLN2_SPI_GET_MIN_DELAY_BETWEEN_SS	DLN2_SPI_CMD(0x47)
+#define DLN2_SPI_GET_MAX_DELAY_BETWEEN_SS	DLN2_SPI_CMD(0x48)
+#define DLN2_SPI_GET_MIN_DELAY_AFTER_SS		DLN2_SPI_CMD(0x49)
+#define DLN2_SPI_GET_MAX_DELAY_AFTER_SS		DLN2_SPI_CMD(0x4A)
+#define DLN2_SPI_GET_MIN_DELAY_BETWEEN_FRAMES	DLN2_SPI_CMD(0x4B)
+#define DLN2_SPI_GET_MAX_DELAY_BETWEEN_FRAMES	DLN2_SPI_CMD(0x4C)
+
+#define DLN2_SPI_MAX_XFER_SIZE			256
+#define DLN2_SPI_BUF_SIZE			(DLN2_SPI_MAX_XFER_SIZE + 16)
+#define DLN2_SPI_ATTR_LEAVE_SS_LOW		BIT(0)
+#define DLN2_TRANSFERS_WAIT_COMPLETE		1
+#define DLN2_TRANSFERS_CANCEL			0
+#define DLN2_RPM_AUTOSUSPEND_TIMEOUT		2000
+
+struct dln2_spi {
+	struct platform_device *pdev;
+	struct spi_master *master;
+	u8 port;
+
+	/*
+	 * This buffer will be used mainly for read/write operations. Since
+	 * they're quite large, we cannot use the stack. Protection is not
+	 * needed because all SPI communication is serialized by the SPI core.
+	 */
+	void *buf;
+
+	u8 bpw;
+	u32 speed;
+	u16 mode;
+	u8 cs;
+};
+
+/*
+ * Enable/Disable SPI module. The disable command will wait for transfers to
+ * complete first.
+ */
+static int dln2_spi_enable(struct dln2_spi *dln2, bool enable)
+{
+	u16 cmd;
+	struct {
+		u8 port;
+		u8 wait_for_completion;
+	} tx;
+	unsigned len = sizeof(tx);
+
+	tx.port = dln2->port;
+
+	if (enable) {
+		cmd = DLN2_SPI_ENABLE;
+		len -= sizeof(tx.wait_for_completion);
+	} else {
+		tx.wait_for_completion = DLN2_TRANSFERS_WAIT_COMPLETE;
+		cmd = DLN2_SPI_DISABLE;
+	}
+
+	return dln2_transfer_tx(dln2->pdev, cmd, &tx, len);
+}
+
+/*
+ * Select/unselect multiple CS lines. The selected lines will be automatically
+ * toggled LOW/HIGH by the board firmware during transfers, provided they're
+ * enabled first.
+ *
+ * Ex: cs_mask = 0x03 -> CS0 & CS1 will be selected and the next WR/RD operation
+ *                       will toggle the lines LOW/HIGH automatically.
+ */
+static int dln2_spi_cs_set(struct dln2_spi *dln2, u8 cs_mask)
+{
+	struct {
+		u8 port;
+		u8 cs;
+	} tx;
+
+	tx.port = dln2->port;
+
+	/*
+	 * According to Diolan docs, "a slave device can be selected by changing
+	 * the corresponding bit value to 0". The rest must be set to 1. Hence
+	 * the bitwise NOT in front.
+	 */
+	tx.cs = ~cs_mask;
+
+	return dln2_transfer_tx(dln2->pdev, DLN2_SPI_SET_SS, &tx, sizeof(tx));
+}
+
+/*
+ * Select one CS line. The other lines will be un-selected.
+ */
+static int dln2_spi_cs_set_one(struct dln2_spi *dln2, u8 cs)
+{
+	return dln2_spi_cs_set(dln2, BIT(cs));
+}
+
+/*
+ * Enable/disable CS lines for usage. The module has to be disabled first.
+ */
+static int dln2_spi_cs_enable(struct dln2_spi *dln2, u8 cs_mask, bool enable)
+{
+	struct {
+		u8 port;
+		u8 cs;
+	} tx;
+	u16 cmd;
+
+	tx.port = dln2->port;
+	tx.cs = cs_mask;
+	cmd = enable ? DLN2_SPI_SS_MULTI_ENABLE : DLN2_SPI_SS_MULTI_DISABLE;
+
+	return dln2_transfer_tx(dln2->pdev, cmd, &tx, sizeof(tx));
+}
+
+static int dln2_spi_cs_enable_all(struct dln2_spi *dln2, bool enable)
+{
+	u8 cs_mask = GENMASK(dln2->master->num_chipselect - 1, 0);
+
+	return dln2_spi_cs_enable(dln2, cs_mask, enable);
+}
+
+static int dln2_spi_get_cs_num(struct dln2_spi *dln2, u16 *cs_num)
+{
+	int ret;
+	struct {
+		u8 port;
+	} tx;
+	struct {
+		__le16 cs_count;
+	} rx;
+	unsigned rx_len = sizeof(rx);
+
+	tx.port = dln2->port;
+	ret = dln2_transfer(dln2->pdev, DLN2_SPI_GET_SS_COUNT, &tx, sizeof(tx),
+			    &rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(rx))
+		return -EPROTO;
+
+	*cs_num = le16_to_cpu(rx.cs_count);
+
+	dev_dbg(&dln2->pdev->dev, "cs_num = %d\n", *cs_num);
+
+	return 0;
+}
+
+static int dln2_spi_get_speed(struct dln2_spi *dln2, u16 cmd, u32 *freq)
+{
+	int ret;
+	struct {
+		u8 port;
+	} tx;
+	struct {
+		__le32 speed;
+	} rx;
+	unsigned rx_len = sizeof(rx);
+
+	tx.port = dln2->port;
+
+	ret = dln2_transfer(dln2->pdev, cmd, &tx, sizeof(tx), &rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(rx))
+		return -EPROTO;
+
+	*freq = le32_to_cpu(rx.speed);
+
+	return 0;
+}
+
+/*
+ * Get bus min/max frequencies.
+ */
+static int dln2_spi_get_speed_range(struct dln2_spi *dln2, u32 *fmin, u32 *fmax)
+{
+	int ret;
+
+	ret = dln2_spi_get_speed(dln2, DLN2_SPI_GET_MIN_FREQUENCY, fmin);
+	if (ret < 0)
+		return ret;
+
+	ret = dln2_spi_get_speed(dln2, DLN2_SPI_GET_MAX_FREQUENCY, fmax);
+	if (ret < 0)
+		return ret;
+
+	dev_dbg(&dln2->pdev->dev, "freq_min = %d, freq_max = %d\n",
+		*fmin, *fmax);
+
+	return 0;
+}
+
+/*
+ * Set the bus speed. The module will automatically round down to the closest
+ * available frequency and returns it. The module has to be disabled first.
+ */
+static int dln2_spi_set_speed(struct dln2_spi *dln2, u32 speed)
+{
+	int ret;
+	struct {
+		u8 port;
+		__le32 speed;
+	} __packed tx;
+	struct {
+		__le32 speed;
+	} rx;
+	int rx_len = sizeof(rx);
+
+	tx.port = dln2->port;
+	tx.speed = cpu_to_le32(speed);
+
+	ret = dln2_transfer(dln2->pdev, DLN2_SPI_SET_FREQUENCY, &tx, sizeof(tx),
+			    &rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(rx))
+		return -EPROTO;
+
+	return 0;
+}
+
+/*
+ * Change CPOL & CPHA. The module has to be disabled first.
+ */
+static int dln2_spi_set_mode(struct dln2_spi *dln2, u8 mode)
+{
+	struct {
+		u8 port;
+		u8 mode;
+	} tx;
+
+	tx.port = dln2->port;
+	tx.mode = mode;
+
+	return dln2_transfer_tx(dln2->pdev, DLN2_SPI_SET_MODE, &tx, sizeof(tx));
+}
+
+/*
+ * Change frame size. The module has to be disabled first.
+ */
+static int dln2_spi_set_bpw(struct dln2_spi *dln2, u8 bpw)
+{
+	struct {
+		u8 port;
+		u8 bpw;
+	} tx;
+
+	tx.port = dln2->port;
+	tx.bpw = bpw;
+
+	return dln2_transfer_tx(dln2->pdev, DLN2_SPI_SET_FRAME_SIZE,
+				&tx, sizeof(tx));
+}
+
+static int dln2_spi_get_supported_frame_sizes(struct dln2_spi *dln2,
+					      u32 *bpw_mask)
+{
+	int ret;
+	struct {
+		u8 port;
+	} tx;
+	struct {
+		u8 count;
+		u8 frame_sizes[36];
+	} *rx = dln2->buf;
+	unsigned rx_len = sizeof(*rx);
+	int i;
+
+	tx.port = dln2->port;
+
+	ret = dln2_transfer(dln2->pdev, DLN2_SPI_GET_SUPPORTED_FRAME_SIZES,
+			    &tx, sizeof(tx), rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(*rx))
+		return -EPROTO;
+	if (rx->count > ARRAY_SIZE(rx->frame_sizes))
+		return -EPROTO;
+
+	*bpw_mask = 0;
+	for (i = 0; i < rx->count; i++)
+		*bpw_mask |= BIT(rx->frame_sizes[i] - 1);
+
+	dev_dbg(&dln2->pdev->dev, "bpw_mask = 0x%X\n", *bpw_mask);
+
+	return 0;
+}
+
+/*
+ * Copy the data to DLN2 buffer and change the byte order to LE, requested by
+ * DLN2 module. SPI core makes sure that the data length is a multiple of word
+ * size.
+ */
+static int dln2_spi_copy_to_buf(u8 *dln2_buf, const u8 *src, u16 len, u8 bpw)
+{
+#ifdef __LITTLE_ENDIAN
+	memcpy(dln2_buf, src, len);
+#else
+	if (bpw <= 8) {
+		memcpy(dln2_buf, src, len);
+	} else if (bpw <= 16) {
+		__le16 *d = (__le16 *)dln2_buf;
+		u16 *s = (u16 *)src;
+
+		len = len / 2;
+		while (len--)
+			*d++ = cpu_to_le16p(s++);
+	} else {
+		__le32 *d = (__le32 *)dln2_buf;
+		u32 *s = (u32 *)src;
+
+		len = len / 4;
+		while (len--)
+			*d++ = cpu_to_le32p(s++);
+	}
+#endif
+
+	return 0;
+}
+
+/*
+ * Copy the data from DLN2 buffer and convert to CPU byte order since the DLN2
+ * buffer is LE ordered. SPI core makes sure that the data length is a multiple
+ * of word size. The RX dln2_buf is 2 byte aligned so, for BE, we have to make
+ * sure we avoid unaligned accesses for 32 bit case.
+ */
+static int dln2_spi_copy_from_buf(u8 *dest, const u8 *dln2_buf, u16 len, u8 bpw)
+{
+#ifdef __LITTLE_ENDIAN
+	memcpy(dest, dln2_buf, len);
+#else
+	if (bpw <= 8) {
+		memcpy(dest, dln2_buf, len);
+	} else if (bpw <= 16) {
+		u16 *d = (u16 *)dest;
+		__le16 *s = (__le16 *)dln2_buf;
+
+		len = len / 2;
+		while (len--)
+			*d++ = le16_to_cpup(s++);
+	} else {
+		u32 *d = (u32 *)dest;
+		__le32 *s = (__le32 *)dln2_buf;
+
+		len = len / 4;
+		while (len--)
+			*d++ = get_unaligned_le32(s++);
+	}
+#endif
+
+	return 0;
+}
+
+/*
+ * Perform one write operation.
+ */
+static int dln2_spi_write_one(struct dln2_spi *dln2, const u8 *data,
+			      u16 data_len, u8 attr)
+{
+	struct {
+		u8 port;
+		__le16 size;
+		u8 attr;
+		u8 buf[DLN2_SPI_MAX_XFER_SIZE];
+	} __packed *tx = dln2->buf;
+	unsigned tx_len;
+
+	BUILD_BUG_ON(sizeof(*tx) > DLN2_SPI_BUF_SIZE);
+
+	if (data_len > DLN2_SPI_MAX_XFER_SIZE)
+		return -EINVAL;
+
+	tx->port = dln2->port;
+	tx->size = cpu_to_le16(data_len);
+	tx->attr = attr;
+
+	dln2_spi_copy_to_buf(tx->buf, data, data_len, dln2->bpw);
+
+	tx_len = sizeof(*tx) + data_len - DLN2_SPI_MAX_XFER_SIZE;
+	return dln2_transfer_tx(dln2->pdev, DLN2_SPI_WRITE, tx, tx_len);
+}
+
+/*
+ * Perform one read operation.
+ */
+static int dln2_spi_read_one(struct dln2_spi *dln2, u8 *data,
+			     u16 data_len, u8 attr)
+{
+	int ret;
+	struct {
+		u8 port;
+		__le16 size;
+		u8 attr;
+	} __packed tx;
+	struct {
+		__le16 size;
+		u8 buf[DLN2_SPI_MAX_XFER_SIZE];
+	} __packed *rx = dln2->buf;
+	unsigned rx_len = sizeof(*rx);
+
+	BUILD_BUG_ON(sizeof(*rx) > DLN2_SPI_BUF_SIZE);
+
+	if (data_len > DLN2_SPI_MAX_XFER_SIZE)
+		return -EINVAL;
+
+	tx.port = dln2->port;
+	tx.size = cpu_to_le16(data_len);
+	tx.attr = attr;
+
+	ret = dln2_transfer(dln2->pdev, DLN2_SPI_READ, &tx, sizeof(tx),
+			    rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(rx->size) + data_len)
+		return -EPROTO;
+	if (le16_to_cpu(rx->size) != data_len)
+		return -EPROTO;
+
+	dln2_spi_copy_from_buf(data, rx->buf, data_len, dln2->bpw);
+
+	return 0;
+}
+
+/*
+ * Perform one write & read operation.
+ */
+static int dln2_spi_read_write_one(struct dln2_spi *dln2, const u8 *tx_data,
+				   u8 *rx_data, u16 data_len, u8 attr)
+{
+	int ret;
+	struct {
+		u8 port;
+		__le16 size;
+		u8 attr;
+		u8 buf[DLN2_SPI_MAX_XFER_SIZE];
+	} __packed *tx;
+	struct {
+		__le16 size;
+		u8 buf[DLN2_SPI_MAX_XFER_SIZE];
+	} __packed *rx;
+	unsigned tx_len, rx_len;
+
+	BUILD_BUG_ON(sizeof(*tx) > DLN2_SPI_BUF_SIZE ||
+		     sizeof(*rx) > DLN2_SPI_BUF_SIZE);
+
+	if (data_len > DLN2_SPI_MAX_XFER_SIZE)
+		return -EINVAL;
+
+	/*
+	 * Since this is a pseudo full-duplex communication, we're perfectly
+	 * safe to use the same buffer for both tx and rx. When DLN2 sends the
+	 * response back, with the rx data, we don't need the tx buffer anymore.
+	 */
+	tx = dln2->buf;
+	rx = dln2->buf;
+
+	tx->port = dln2->port;
+	tx->size = cpu_to_le16(data_len);
+	tx->attr = attr;
+
+	dln2_spi_copy_to_buf(tx->buf, tx_data, data_len, dln2->bpw);
+
+	tx_len = sizeof(*tx) + data_len - DLN2_SPI_MAX_XFER_SIZE;
+	rx_len = sizeof(*rx);
+
+	ret = dln2_transfer(dln2->pdev, DLN2_SPI_READ_WRITE, tx, tx_len,
+			    rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(rx->size) + data_len)
+		return -EPROTO;
+	if (le16_to_cpu(rx->size) != data_len)
+		return -EPROTO;
+
+	dln2_spi_copy_from_buf(rx_data, rx->buf, data_len, dln2->bpw);
+
+	return 0;
+}
+
+/*
+ * Read/Write wrapper. It will automatically split an operation into multiple
+ * single ones due to device buffer constraints.
+ */
+static int dln2_spi_rdwr(struct dln2_spi *dln2, const u8 *tx_data,
+			 u8 *rx_data, u16 data_len, u8 attr) {
+	int ret;
+	u16 len;
+	u8 temp_attr;
+	u16 remaining = data_len;
+	u16 offset;
+
+	do {
+		if (remaining > DLN2_SPI_MAX_XFER_SIZE) {
+			len = DLN2_SPI_MAX_XFER_SIZE;
+			temp_attr = DLN2_SPI_ATTR_LEAVE_SS_LOW;
+		} else {
+			len = remaining;
+			temp_attr = attr;
+		}
+
+		offset = data_len - remaining;
+
+		if (tx_data && rx_data) {
+			ret = dln2_spi_read_write_one(dln2,
+						      tx_data + offset,
+						      rx_data + offset,
+						      len, temp_attr);
+		} else if (tx_data) {
+			ret = dln2_spi_write_one(dln2,
+						 tx_data + offset,
+						 len, temp_attr);
+		} else if (rx_data) {
+			ret = dln2_spi_read_one(dln2,
+						rx_data + offset,
+						len, temp_attr);
+		 } else {
+			return -EINVAL;
+		 }
+
+		if (ret < 0)
+			return ret;
+
+		remaining -= len;
+	} while (remaining);
+
+	return 0;
+}
+
+static int dln2_spi_prepare_message(struct spi_master *master,
+				    struct spi_message *message)
+{
+	int ret;
+	struct dln2_spi *dln2 = spi_master_get_devdata(master);
+	struct spi_device *spi = message->spi;
+
+	if (dln2->cs != spi->chip_select) {
+		ret = dln2_spi_cs_set_one(dln2, spi->chip_select);
+		if (ret < 0)
+			return ret;
+
+		dln2->cs = spi->chip_select;
+	}
+
+	return 0;
+}
+
+static int dln2_spi_transfer_setup(struct dln2_spi *dln2, u32 speed,
+				   u8 bpw, u8 mode)
+{
+	int ret;
+	bool bus_setup_change;
+
+	bus_setup_change = dln2->speed != speed || dln2->mode != mode ||
+			   dln2->bpw != bpw;
+
+	if (!bus_setup_change)
+		return 0;
+
+	ret = dln2_spi_enable(dln2, false);
+	if (ret < 0)
+		return ret;
+
+	if (dln2->speed != speed) {
+		ret = dln2_spi_set_speed(dln2, speed);
+		if (ret < 0)
+			return ret;
+
+		dln2->speed = speed;
+	}
+
+	if (dln2->mode != mode) {
+		ret = dln2_spi_set_mode(dln2, mode & 0x3);
+		if (ret < 0)
+			return ret;
+
+		dln2->mode = mode;
+	}
+
+	if (dln2->bpw != bpw) {
+		ret = dln2_spi_set_bpw(dln2, bpw);
+		if (ret < 0)
+			return ret;
+
+		dln2->bpw = bpw;
+	}
+
+	return dln2_spi_enable(dln2, true);
+}
+
+static int dln2_spi_transfer_one(struct spi_master *master,
+				 struct spi_device *spi,
+				 struct spi_transfer *xfer)
+{
+	struct dln2_spi *dln2 = spi_master_get_devdata(master);
+	int status;
+	u8 attr = 0;
+
+	status = dln2_spi_transfer_setup(dln2, xfer->speed_hz,
+					 xfer->bits_per_word,
+					 spi->mode);
+	if (status < 0) {
+		dev_err(&dln2->pdev->dev, "Cannot setup transfer\n");
+		return status;
+	}
+
+	if (!xfer->cs_change && !spi_transfer_is_last(master, xfer))
+		attr = DLN2_SPI_ATTR_LEAVE_SS_LOW;
+
+	status = dln2_spi_rdwr(dln2, xfer->tx_buf, xfer->rx_buf,
+			       xfer->len, attr);
+	if (status < 0)
+		dev_err(&dln2->pdev->dev, "write/read failed!\n");
+
+	return status;
+}
+
+static int dln2_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct dln2_spi *dln2;
+	struct dln2_platform_data *pdata = dev_get_platdata(&pdev->dev);
+	struct device *dev = &pdev->dev;
+	int ret;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*dln2));
+	if (!master)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, master);
+
+	dln2 = spi_master_get_devdata(master);
+
+	dln2->buf = devm_kmalloc(&pdev->dev, DLN2_SPI_BUF_SIZE, GFP_KERNEL);
+	if (!dln2->buf) {
+		ret = -ENOMEM;
+		goto exit_free_master;
+	}
+
+	dln2->master = master;
+	dln2->master->dev.of_node = dev->of_node;
+	dln2->pdev = pdev;
+	dln2->port = pdata->port;
+	/* cs/mode can never be 0xff, so the first transfer will set them */
+	dln2->cs = 0xff;
+	dln2->mode = 0xff;
+
+	/* disable SPI module before continuing with the setup */
+	ret = dln2_spi_enable(dln2, false);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to disable SPI module\n");
+		goto exit_free_master;
+	}
+
+	ret = dln2_spi_get_cs_num(dln2, &master->num_chipselect);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to get number of CS pins\n");
+		goto exit_free_master;
+	}
+
+	ret = dln2_spi_get_speed_range(dln2,
+				       &master->min_speed_hz,
+				       &master->max_speed_hz);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to read bus min/max freqs\n");
+		goto exit_free_master;
+	}
+
+	ret = dln2_spi_get_supported_frame_sizes(dln2,
+						 &master->bits_per_word_mask);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to read supported frame sizes\n");
+		goto exit_free_master;
+	}
+
+	ret = dln2_spi_cs_enable_all(dln2, true);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to enable CS pins\n");
+		goto exit_free_master;
+	}
+
+	master->bus_num = -1;
+	master->mode_bits = SPI_CPOL | SPI_CPHA;
+	master->prepare_message = dln2_spi_prepare_message;
+	master->transfer_one = dln2_spi_transfer_one;
+	master->auto_runtime_pm = true;
+
+	/* enable SPI module, we're good to go */
+	ret = dln2_spi_enable(dln2, true);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to enable SPI module\n");
+		goto exit_free_master;
+	}
+
+	pm_runtime_set_autosuspend_delay(&pdev->dev,
+					 DLN2_RPM_AUTOSUSPEND_TIMEOUT);
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to register master\n");
+		goto exit_register;
+	}
+
+	return ret;
+
+exit_register:
+	pm_runtime_disable(&pdev->dev);
+	pm_runtime_set_suspended(&pdev->dev);
+
+	if (dln2_spi_enable(dln2, false) < 0)
+		dev_err(&pdev->dev, "Failed to disable SPI module\n");
+exit_free_master:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int dln2_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = spi_master_get(platform_get_drvdata(pdev));
+	struct dln2_spi *dln2 = spi_master_get_devdata(master);
+
+	pm_runtime_disable(&pdev->dev);
+
+	if (dln2_spi_enable(dln2, false) < 0)
+		dev_err(&pdev->dev, "Failed to disable SPI module\n");
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int dln2_spi_suspend(struct device *dev)
+{
+	int ret;
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct dln2_spi *dln2 = spi_master_get_devdata(master);
+
+	ret = spi_master_suspend(master);
+	if (ret < 0)
+		return ret;
+
+	if (!pm_runtime_suspended(dev)) {
+		ret = dln2_spi_enable(dln2, false);
+		if (ret < 0)
+			return ret;
+	}
+
+	/*
+	 * USB power may be cut off during sleep. Resetting the following
+	 * parameters will force the board to be set up before first transfer.
+	 */
+	dln2->cs = 0xff;
+	dln2->speed = 0;
+	dln2->bpw = 0;
+	dln2->mode = 0xff;
+
+	return 0;
+}
+
+static int dln2_spi_resume(struct device *dev)
+{
+	int ret;
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct dln2_spi *dln2 = spi_master_get_devdata(master);
+
+	if (!pm_runtime_suspended(dev)) {
+		ret = dln2_spi_cs_enable_all(dln2, true);
+		if (ret < 0)
+			return ret;
+
+		ret = dln2_spi_enable(dln2, true);
+		if (ret < 0)
+			return ret;
+	}
+
+	return spi_master_resume(master);
+}
+#endif /* CONFIG_PM_SLEEP */
+
+#ifdef CONFIG_PM
+static int dln2_spi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct dln2_spi *dln2 = spi_master_get_devdata(master);
+
+	return dln2_spi_enable(dln2, false);
+}
+
+static int dln2_spi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct dln2_spi *dln2 = spi_master_get_devdata(master);
+
+	return  dln2_spi_enable(dln2, true);
+}
+#endif /* CONFIG_PM */
+
+static const struct dev_pm_ops dln2_spi_pm = {
+	SET_SYSTEM_SLEEP_PM_OPS(dln2_spi_suspend, dln2_spi_resume)
+	SET_RUNTIME_PM_OPS(dln2_spi_runtime_suspend,
+			   dln2_spi_runtime_resume, NULL)
+};
+
+static struct platform_driver spi_dln2_driver = {
+	.driver = {
+		.name	= "dln2-spi",
+		.pm	= &dln2_spi_pm,
+	},
+	.probe		= dln2_spi_probe,
+	.remove		= dln2_spi_remove,
+};
+module_platform_driver(spi_dln2_driver);
+
+MODULE_DESCRIPTION("Driver for the Diolan DLN2 SPI master interface");
+MODULE_AUTHOR("Laurentiu Palcu <laurentiu.palcu@intel.com>");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:dln2-spi");
diff --git a/drivers/spi/spi-dw-mid.c b/drivers/spi/spi-dw-mid.c
new file mode 100644
index 0000000..3db905f
--- /dev/null
+++ b/drivers/spi/spi-dw-mid.c
@@ -0,0 +1,330 @@
+/*
+ * Special handling for DW core on Intel MID platform
+ *
+ * Copyright (c) 2009, 2014 Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ */
+
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/interrupt.h>
+#include <linux/slab.h>
+#include <linux/spi/spi.h>
+#include <linux/types.h>
+
+#include "spi-dw.h"
+
+#ifdef CONFIG_SPI_DW_MID_DMA
+#include <linux/pci.h>
+#include <linux/platform_data/dma-dw.h>
+
+#define RX_BUSY		0
+#define TX_BUSY		1
+
+static struct dw_dma_slave mid_dma_tx = { .dst_id = 1 };
+static struct dw_dma_slave mid_dma_rx = { .src_id = 0 };
+
+static bool mid_spi_dma_chan_filter(struct dma_chan *chan, void *param)
+{
+	struct dw_dma_slave *s = param;
+
+	if (s->dma_dev != chan->device->dev)
+		return false;
+
+	chan->private = s;
+	return true;
+}
+
+static int mid_spi_dma_init(struct dw_spi *dws)
+{
+	struct pci_dev *dma_dev;
+	struct dw_dma_slave *tx = dws->dma_tx;
+	struct dw_dma_slave *rx = dws->dma_rx;
+	dma_cap_mask_t mask;
+
+	/*
+	 * Get pci device for DMA controller, currently it could only
+	 * be the DMA controller of Medfield
+	 */
+	dma_dev = pci_get_device(PCI_VENDOR_ID_INTEL, 0x0827, NULL);
+	if (!dma_dev)
+		return -ENODEV;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	/* 1. Init rx channel */
+	rx->dma_dev = &dma_dev->dev;
+	dws->rxchan = dma_request_channel(mask, mid_spi_dma_chan_filter, rx);
+	if (!dws->rxchan)
+		goto err_exit;
+	dws->master->dma_rx = dws->rxchan;
+
+	/* 2. Init tx channel */
+	tx->dma_dev = &dma_dev->dev;
+	dws->txchan = dma_request_channel(mask, mid_spi_dma_chan_filter, tx);
+	if (!dws->txchan)
+		goto free_rxchan;
+	dws->master->dma_tx = dws->txchan;
+
+	dws->dma_inited = 1;
+	return 0;
+
+free_rxchan:
+	dma_release_channel(dws->rxchan);
+err_exit:
+	return -EBUSY;
+}
+
+static void mid_spi_dma_exit(struct dw_spi *dws)
+{
+	if (!dws->dma_inited)
+		return;
+
+	dmaengine_terminate_sync(dws->txchan);
+	dma_release_channel(dws->txchan);
+
+	dmaengine_terminate_sync(dws->rxchan);
+	dma_release_channel(dws->rxchan);
+}
+
+static irqreturn_t dma_transfer(struct dw_spi *dws)
+{
+	u16 irq_status = dw_readl(dws, DW_SPI_ISR);
+
+	if (!irq_status)
+		return IRQ_NONE;
+
+	dw_readl(dws, DW_SPI_ICR);
+	spi_reset_chip(dws);
+
+	dev_err(&dws->master->dev, "%s: FIFO overrun/underrun\n", __func__);
+	dws->master->cur_msg->status = -EIO;
+	spi_finalize_current_transfer(dws->master);
+	return IRQ_HANDLED;
+}
+
+static bool mid_spi_can_dma(struct spi_controller *master,
+		struct spi_device *spi, struct spi_transfer *xfer)
+{
+	struct dw_spi *dws = spi_controller_get_devdata(master);
+
+	if (!dws->dma_inited)
+		return false;
+
+	return xfer->len > dws->fifo_len;
+}
+
+static enum dma_slave_buswidth convert_dma_width(u32 dma_width) {
+	if (dma_width == 1)
+		return DMA_SLAVE_BUSWIDTH_1_BYTE;
+	else if (dma_width == 2)
+		return DMA_SLAVE_BUSWIDTH_2_BYTES;
+
+	return DMA_SLAVE_BUSWIDTH_UNDEFINED;
+}
+
+/*
+ * dws->dma_chan_busy is set before the dma transfer starts, callback for tx
+ * channel will clear a corresponding bit.
+ */
+static void dw_spi_dma_tx_done(void *arg)
+{
+	struct dw_spi *dws = arg;
+
+	clear_bit(TX_BUSY, &dws->dma_chan_busy);
+	if (test_bit(RX_BUSY, &dws->dma_chan_busy))
+		return;
+	spi_finalize_current_transfer(dws->master);
+}
+
+static struct dma_async_tx_descriptor *dw_spi_dma_prepare_tx(struct dw_spi *dws,
+		struct spi_transfer *xfer)
+{
+	struct dma_slave_config txconf;
+	struct dma_async_tx_descriptor *txdesc;
+
+	if (!xfer->tx_buf)
+		return NULL;
+
+	txconf.direction = DMA_MEM_TO_DEV;
+	txconf.dst_addr = dws->dma_addr;
+	txconf.dst_maxburst = 16;
+	txconf.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+	txconf.dst_addr_width = convert_dma_width(dws->dma_width);
+	txconf.device_fc = false;
+
+	dmaengine_slave_config(dws->txchan, &txconf);
+
+	txdesc = dmaengine_prep_slave_sg(dws->txchan,
+				xfer->tx_sg.sgl,
+				xfer->tx_sg.nents,
+				DMA_MEM_TO_DEV,
+				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!txdesc)
+		return NULL;
+
+	txdesc->callback = dw_spi_dma_tx_done;
+	txdesc->callback_param = dws;
+
+	return txdesc;
+}
+
+/*
+ * dws->dma_chan_busy is set before the dma transfer starts, callback for rx
+ * channel will clear a corresponding bit.
+ */
+static void dw_spi_dma_rx_done(void *arg)
+{
+	struct dw_spi *dws = arg;
+
+	clear_bit(RX_BUSY, &dws->dma_chan_busy);
+	if (test_bit(TX_BUSY, &dws->dma_chan_busy))
+		return;
+	spi_finalize_current_transfer(dws->master);
+}
+
+static struct dma_async_tx_descriptor *dw_spi_dma_prepare_rx(struct dw_spi *dws,
+		struct spi_transfer *xfer)
+{
+	struct dma_slave_config rxconf;
+	struct dma_async_tx_descriptor *rxdesc;
+
+	if (!xfer->rx_buf)
+		return NULL;
+
+	rxconf.direction = DMA_DEV_TO_MEM;
+	rxconf.src_addr = dws->dma_addr;
+	rxconf.src_maxburst = 16;
+	rxconf.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+	rxconf.src_addr_width = convert_dma_width(dws->dma_width);
+	rxconf.device_fc = false;
+
+	dmaengine_slave_config(dws->rxchan, &rxconf);
+
+	rxdesc = dmaengine_prep_slave_sg(dws->rxchan,
+				xfer->rx_sg.sgl,
+				xfer->rx_sg.nents,
+				DMA_DEV_TO_MEM,
+				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!rxdesc)
+		return NULL;
+
+	rxdesc->callback = dw_spi_dma_rx_done;
+	rxdesc->callback_param = dws;
+
+	return rxdesc;
+}
+
+static int mid_spi_dma_setup(struct dw_spi *dws, struct spi_transfer *xfer)
+{
+	u16 dma_ctrl = 0;
+
+	dw_writel(dws, DW_SPI_DMARDLR, 0xf);
+	dw_writel(dws, DW_SPI_DMATDLR, 0x10);
+
+	if (xfer->tx_buf)
+		dma_ctrl |= SPI_DMA_TDMAE;
+	if (xfer->rx_buf)
+		dma_ctrl |= SPI_DMA_RDMAE;
+	dw_writel(dws, DW_SPI_DMACR, dma_ctrl);
+
+	/* Set the interrupt mask */
+	spi_umask_intr(dws, SPI_INT_TXOI | SPI_INT_RXUI | SPI_INT_RXOI);
+
+	dws->transfer_handler = dma_transfer;
+
+	return 0;
+}
+
+static int mid_spi_dma_transfer(struct dw_spi *dws, struct spi_transfer *xfer)
+{
+	struct dma_async_tx_descriptor *txdesc, *rxdesc;
+
+	/* Prepare the TX dma transfer */
+	txdesc = dw_spi_dma_prepare_tx(dws, xfer);
+
+	/* Prepare the RX dma transfer */
+	rxdesc = dw_spi_dma_prepare_rx(dws, xfer);
+
+	/* rx must be started before tx due to spi instinct */
+	if (rxdesc) {
+		set_bit(RX_BUSY, &dws->dma_chan_busy);
+		dmaengine_submit(rxdesc);
+		dma_async_issue_pending(dws->rxchan);
+	}
+
+	if (txdesc) {
+		set_bit(TX_BUSY, &dws->dma_chan_busy);
+		dmaengine_submit(txdesc);
+		dma_async_issue_pending(dws->txchan);
+	}
+
+	return 0;
+}
+
+static void mid_spi_dma_stop(struct dw_spi *dws)
+{
+	if (test_bit(TX_BUSY, &dws->dma_chan_busy)) {
+		dmaengine_terminate_sync(dws->txchan);
+		clear_bit(TX_BUSY, &dws->dma_chan_busy);
+	}
+	if (test_bit(RX_BUSY, &dws->dma_chan_busy)) {
+		dmaengine_terminate_sync(dws->rxchan);
+		clear_bit(RX_BUSY, &dws->dma_chan_busy);
+	}
+}
+
+static const struct dw_spi_dma_ops mid_dma_ops = {
+	.dma_init	= mid_spi_dma_init,
+	.dma_exit	= mid_spi_dma_exit,
+	.dma_setup	= mid_spi_dma_setup,
+	.can_dma	= mid_spi_can_dma,
+	.dma_transfer	= mid_spi_dma_transfer,
+	.dma_stop	= mid_spi_dma_stop,
+};
+#endif
+
+/* Some specific info for SPI0 controller on Intel MID */
+
+/* HW info for MRST Clk Control Unit, 32b reg per controller */
+#define MRST_SPI_CLK_BASE	100000000	/* 100m */
+#define MRST_CLK_SPI_REG	0xff11d86c
+#define CLK_SPI_BDIV_OFFSET	0
+#define CLK_SPI_BDIV_MASK	0x00000007
+#define CLK_SPI_CDIV_OFFSET	9
+#define CLK_SPI_CDIV_MASK	0x00000e00
+#define CLK_SPI_DISABLE_OFFSET	8
+
+int dw_spi_mid_init(struct dw_spi *dws)
+{
+	void __iomem *clk_reg;
+	u32 clk_cdiv;
+
+	clk_reg = ioremap_nocache(MRST_CLK_SPI_REG, 16);
+	if (!clk_reg)
+		return -ENOMEM;
+
+	/* Get SPI controller operating freq info */
+	clk_cdiv = readl(clk_reg + dws->bus_num * sizeof(u32));
+	clk_cdiv &= CLK_SPI_CDIV_MASK;
+	clk_cdiv >>= CLK_SPI_CDIV_OFFSET;
+	dws->max_freq = MRST_SPI_CLK_BASE / (clk_cdiv + 1);
+
+	iounmap(clk_reg);
+
+#ifdef CONFIG_SPI_DW_MID_DMA
+	dws->dma_tx = &mid_dma_tx;
+	dws->dma_rx = &mid_dma_rx;
+	dws->dma_ops = &mid_dma_ops;
+#endif
+	return 0;
+}
diff --git a/drivers/spi/spi-dw-mmio.c b/drivers/spi/spi-dw-mmio.c
new file mode 100644
index 0000000..e80f60e
--- /dev/null
+++ b/drivers/spi/spi-dw-mmio.c
@@ -0,0 +1,232 @@
+/*
+ * Memory-mapped interface driver for DW SPI Core
+ *
+ * Copyright (c) 2010, Octasic semiconductor.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/spi/spi.h>
+#include <linux/scatterlist.h>
+#include <linux/mfd/syscon.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+#include <linux/of_platform.h>
+#include <linux/property.h>
+#include <linux/regmap.h>
+
+#include "spi-dw.h"
+
+#define DRIVER_NAME "dw_spi_mmio"
+
+struct dw_spi_mmio {
+	struct dw_spi  dws;
+	struct clk     *clk;
+	void           *priv;
+};
+
+#define MSCC_CPU_SYSTEM_CTRL_GENERAL_CTRL	0x24
+#define OCELOT_IF_SI_OWNER_MASK			GENMASK(5, 4)
+#define OCELOT_IF_SI_OWNER_OFFSET		4
+#define MSCC_IF_SI_OWNER_SISL			0
+#define MSCC_IF_SI_OWNER_SIBM			1
+#define MSCC_IF_SI_OWNER_SIMC			2
+
+#define MSCC_SPI_MST_SW_MODE			0x14
+#define MSCC_SPI_MST_SW_MODE_SW_PIN_CTRL_MODE	BIT(13)
+#define MSCC_SPI_MST_SW_MODE_SW_SPI_CS(x)	(x << 5)
+
+struct dw_spi_mscc {
+	struct regmap       *syscon;
+	void __iomem        *spi_mst;
+};
+
+/*
+ * The Designware SPI controller (referred to as master in the documentation)
+ * automatically deasserts chip select when the tx fifo is empty. The chip
+ * selects then needs to be either driven as GPIOs or, for the first 4 using the
+ * the SPI boot controller registers. the final chip select is an OR gate
+ * between the Designware SPI controller and the SPI boot controller.
+ */
+static void dw_spi_mscc_set_cs(struct spi_device *spi, bool enable)
+{
+	struct dw_spi *dws = spi_master_get_devdata(spi->master);
+	struct dw_spi_mmio *dwsmmio = container_of(dws, struct dw_spi_mmio, dws);
+	struct dw_spi_mscc *dwsmscc = dwsmmio->priv;
+	u32 cs = spi->chip_select;
+
+	if (cs < 4) {
+		u32 sw_mode = MSCC_SPI_MST_SW_MODE_SW_PIN_CTRL_MODE;
+
+		if (!enable)
+			sw_mode |= MSCC_SPI_MST_SW_MODE_SW_SPI_CS(BIT(cs));
+
+		writel(sw_mode, dwsmscc->spi_mst + MSCC_SPI_MST_SW_MODE);
+	}
+
+	dw_spi_set_cs(spi, enable);
+}
+
+static int dw_spi_mscc_init(struct platform_device *pdev,
+			    struct dw_spi_mmio *dwsmmio)
+{
+	struct dw_spi_mscc *dwsmscc;
+	struct resource *res;
+
+	dwsmscc = devm_kzalloc(&pdev->dev, sizeof(*dwsmscc), GFP_KERNEL);
+	if (!dwsmscc)
+		return -ENOMEM;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	dwsmscc->spi_mst = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(dwsmscc->spi_mst)) {
+		dev_err(&pdev->dev, "SPI_MST region map failed\n");
+		return PTR_ERR(dwsmscc->spi_mst);
+	}
+
+	dwsmscc->syscon = syscon_regmap_lookup_by_compatible("mscc,ocelot-cpu-syscon");
+	if (IS_ERR(dwsmscc->syscon))
+		return PTR_ERR(dwsmscc->syscon);
+
+	/* Deassert all CS */
+	writel(0, dwsmscc->spi_mst + MSCC_SPI_MST_SW_MODE);
+
+	/* Select the owner of the SI interface */
+	regmap_update_bits(dwsmscc->syscon, MSCC_CPU_SYSTEM_CTRL_GENERAL_CTRL,
+			   OCELOT_IF_SI_OWNER_MASK,
+			   MSCC_IF_SI_OWNER_SIMC << OCELOT_IF_SI_OWNER_OFFSET);
+
+	dwsmmio->dws.set_cs = dw_spi_mscc_set_cs;
+	dwsmmio->priv = dwsmscc;
+
+	return 0;
+}
+
+static int dw_spi_mmio_probe(struct platform_device *pdev)
+{
+	int (*init_func)(struct platform_device *pdev,
+			 struct dw_spi_mmio *dwsmmio);
+	struct dw_spi_mmio *dwsmmio;
+	struct dw_spi *dws;
+	struct resource *mem;
+	int ret;
+	int num_cs;
+
+	dwsmmio = devm_kzalloc(&pdev->dev, sizeof(struct dw_spi_mmio),
+			GFP_KERNEL);
+	if (!dwsmmio)
+		return -ENOMEM;
+
+	dws = &dwsmmio->dws;
+
+	/* Get basic io resource and map it */
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	dws->regs = devm_ioremap_resource(&pdev->dev, mem);
+	if (IS_ERR(dws->regs)) {
+		dev_err(&pdev->dev, "SPI region map failed\n");
+		return PTR_ERR(dws->regs);
+	}
+
+	dws->irq = platform_get_irq(pdev, 0);
+	if (dws->irq < 0) {
+		dev_err(&pdev->dev, "no irq resource?\n");
+		return dws->irq; /* -ENXIO */
+	}
+
+	dwsmmio->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(dwsmmio->clk))
+		return PTR_ERR(dwsmmio->clk);
+	ret = clk_prepare_enable(dwsmmio->clk);
+	if (ret)
+		return ret;
+
+	dws->bus_num = pdev->id;
+
+	dws->max_freq = clk_get_rate(dwsmmio->clk);
+
+	device_property_read_u32(&pdev->dev, "reg-io-width", &dws->reg_io_width);
+
+	num_cs = 4;
+
+	device_property_read_u32(&pdev->dev, "num-cs", &num_cs);
+
+	dws->num_cs = num_cs;
+
+	if (pdev->dev.of_node) {
+		int i;
+
+		for (i = 0; i < dws->num_cs; i++) {
+			int cs_gpio = of_get_named_gpio(pdev->dev.of_node,
+					"cs-gpios", i);
+
+			if (cs_gpio == -EPROBE_DEFER) {
+				ret = cs_gpio;
+				goto out;
+			}
+
+			if (gpio_is_valid(cs_gpio)) {
+				ret = devm_gpio_request(&pdev->dev, cs_gpio,
+						dev_name(&pdev->dev));
+				if (ret)
+					goto out;
+			}
+		}
+	}
+
+	init_func = device_get_match_data(&pdev->dev);
+	if (init_func) {
+		ret = init_func(pdev, dwsmmio);
+		if (ret)
+			goto out;
+	}
+
+	ret = dw_spi_add_host(&pdev->dev, dws);
+	if (ret)
+		goto out;
+
+	platform_set_drvdata(pdev, dwsmmio);
+	return 0;
+
+out:
+	clk_disable_unprepare(dwsmmio->clk);
+	return ret;
+}
+
+static int dw_spi_mmio_remove(struct platform_device *pdev)
+{
+	struct dw_spi_mmio *dwsmmio = platform_get_drvdata(pdev);
+
+	dw_spi_remove_host(&dwsmmio->dws);
+	clk_disable_unprepare(dwsmmio->clk);
+
+	return 0;
+}
+
+static const struct of_device_id dw_spi_mmio_of_match[] = {
+	{ .compatible = "snps,dw-apb-ssi", },
+	{ .compatible = "mscc,ocelot-spi", .data = dw_spi_mscc_init},
+	{ /* end of table */}
+};
+MODULE_DEVICE_TABLE(of, dw_spi_mmio_of_match);
+
+static struct platform_driver dw_spi_mmio_driver = {
+	.probe		= dw_spi_mmio_probe,
+	.remove		= dw_spi_mmio_remove,
+	.driver		= {
+		.name	= DRIVER_NAME,
+		.of_match_table = dw_spi_mmio_of_match,
+	},
+};
+module_platform_driver(dw_spi_mmio_driver);
+
+MODULE_AUTHOR("Jean-Hugues Deschenes <jean-hugues.deschenes@octasic.com>");
+MODULE_DESCRIPTION("Memory-mapped I/O interface driver for DW SPI Core");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-dw-pci.c b/drivers/spi/spi-dw-pci.c
new file mode 100644
index 0000000..ef7db75
--- /dev/null
+++ b/drivers/spi/spi-dw-pci.c
@@ -0,0 +1,153 @@
+/*
+ * PCI interface driver for DW SPI Core
+ *
+ * Copyright (c) 2009, 2014 Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ */
+
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include <linux/slab.h>
+#include <linux/spi/spi.h>
+#include <linux/module.h>
+
+#include "spi-dw.h"
+
+#define DRIVER_NAME "dw_spi_pci"
+
+struct spi_pci_desc {
+	int	(*setup)(struct dw_spi *);
+	u16	num_cs;
+	u16	bus_num;
+};
+
+static struct spi_pci_desc spi_pci_mid_desc_1 = {
+	.setup = dw_spi_mid_init,
+	.num_cs = 5,
+	.bus_num = 0,
+};
+
+static struct spi_pci_desc spi_pci_mid_desc_2 = {
+	.setup = dw_spi_mid_init,
+	.num_cs = 2,
+	.bus_num = 1,
+};
+
+static int spi_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	struct dw_spi *dws;
+	struct spi_pci_desc *desc = (struct spi_pci_desc *)ent->driver_data;
+	int pci_bar = 0;
+	int ret;
+
+	ret = pcim_enable_device(pdev);
+	if (ret)
+		return ret;
+
+	dws = devm_kzalloc(&pdev->dev, sizeof(*dws), GFP_KERNEL);
+	if (!dws)
+		return -ENOMEM;
+
+	/* Get basic io resource and map it */
+	dws->paddr = pci_resource_start(pdev, pci_bar);
+
+	ret = pcim_iomap_regions(pdev, 1 << pci_bar, pci_name(pdev));
+	if (ret)
+		return ret;
+
+	dws->regs = pcim_iomap_table(pdev)[pci_bar];
+	dws->irq = pdev->irq;
+
+	/*
+	 * Specific handling for platforms, like dma setup,
+	 * clock rate, FIFO depth.
+	 */
+	if (desc) {
+		dws->num_cs = desc->num_cs;
+		dws->bus_num = desc->bus_num;
+
+		if (desc->setup) {
+			ret = desc->setup(dws);
+			if (ret)
+				return ret;
+		}
+	} else {
+		return -ENODEV;
+	}
+
+	ret = dw_spi_add_host(&pdev->dev, dws);
+	if (ret)
+		return ret;
+
+	/* PCI hook and SPI hook use the same drv data */
+	pci_set_drvdata(pdev, dws);
+
+	dev_info(&pdev->dev, "found PCI SPI controller(ID: %04x:%04x)\n",
+		pdev->vendor, pdev->device);
+
+	return 0;
+}
+
+static void spi_pci_remove(struct pci_dev *pdev)
+{
+	struct dw_spi *dws = pci_get_drvdata(pdev);
+
+	dw_spi_remove_host(dws);
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int spi_suspend(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct dw_spi *dws = pci_get_drvdata(pdev);
+
+	return dw_spi_suspend_host(dws);
+}
+
+static int spi_resume(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct dw_spi *dws = pci_get_drvdata(pdev);
+
+	return dw_spi_resume_host(dws);
+}
+#endif
+
+static SIMPLE_DEV_PM_OPS(dw_spi_pm_ops, spi_suspend, spi_resume);
+
+static const struct pci_device_id pci_ids[] = {
+	/* Intel MID platform SPI controller 0 */
+	/*
+	 * The access to the device 8086:0801 is disabled by HW, since it's
+	 * exclusively used by SCU to communicate with MSIC.
+	 */
+	/* Intel MID platform SPI controller 1 */
+	{ PCI_VDEVICE(INTEL, 0x0800), (kernel_ulong_t)&spi_pci_mid_desc_1},
+	/* Intel MID platform SPI controller 2 */
+	{ PCI_VDEVICE(INTEL, 0x0812), (kernel_ulong_t)&spi_pci_mid_desc_2},
+	{},
+};
+
+static struct pci_driver dw_spi_driver = {
+	.name =		DRIVER_NAME,
+	.id_table =	pci_ids,
+	.probe =	spi_pci_probe,
+	.remove =	spi_pci_remove,
+	.driver         = {
+		.pm     = &dw_spi_pm_ops,
+	},
+};
+
+module_pci_driver(dw_spi_driver);
+
+MODULE_AUTHOR("Feng Tang <feng.tang@intel.com>");
+MODULE_DESCRIPTION("PCI interface driver for DW SPI Core");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-dw.c b/drivers/spi/spi-dw.c
new file mode 100644
index 0000000..ac2eb89
--- /dev/null
+++ b/drivers/spi/spi-dw.c
@@ -0,0 +1,587 @@
+/*
+ * Designware SPI core controller driver (refer pxa2xx_spi.c)
+ *
+ * Copyright (c) 2009, Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ */
+
+#include <linux/dma-mapping.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/highmem.h>
+#include <linux/delay.h>
+#include <linux/slab.h>
+#include <linux/spi/spi.h>
+#include <linux/gpio.h>
+
+#include "spi-dw.h"
+
+#ifdef CONFIG_DEBUG_FS
+#include <linux/debugfs.h>
+#endif
+
+/* Slave spi_dev related */
+struct chip_data {
+	u8 tmode;		/* TR/TO/RO/EEPROM */
+	u8 type;		/* SPI/SSP/MicroWire */
+
+	u8 poll_mode;		/* 1 means use poll mode */
+
+	u16 clk_div;		/* baud rate divider */
+	u32 speed_hz;		/* baud rate */
+	void (*cs_control)(u32 command);
+};
+
+#ifdef CONFIG_DEBUG_FS
+#define SPI_REGS_BUFSIZE	1024
+static ssize_t dw_spi_show_regs(struct file *file, char __user *user_buf,
+		size_t count, loff_t *ppos)
+{
+	struct dw_spi *dws = file->private_data;
+	char *buf;
+	u32 len = 0;
+	ssize_t ret;
+
+	buf = kzalloc(SPI_REGS_BUFSIZE, GFP_KERNEL);
+	if (!buf)
+		return 0;
+
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"%s registers:\n", dev_name(&dws->master->dev));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"=================================\n");
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"CTRL0: \t\t0x%08x\n", dw_readl(dws, DW_SPI_CTRL0));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"CTRL1: \t\t0x%08x\n", dw_readl(dws, DW_SPI_CTRL1));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"SSIENR: \t0x%08x\n", dw_readl(dws, DW_SPI_SSIENR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"SER: \t\t0x%08x\n", dw_readl(dws, DW_SPI_SER));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"BAUDR: \t\t0x%08x\n", dw_readl(dws, DW_SPI_BAUDR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"TXFTLR: \t0x%08x\n", dw_readl(dws, DW_SPI_TXFLTR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"RXFTLR: \t0x%08x\n", dw_readl(dws, DW_SPI_RXFLTR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"TXFLR: \t\t0x%08x\n", dw_readl(dws, DW_SPI_TXFLR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"RXFLR: \t\t0x%08x\n", dw_readl(dws, DW_SPI_RXFLR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"SR: \t\t0x%08x\n", dw_readl(dws, DW_SPI_SR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"IMR: \t\t0x%08x\n", dw_readl(dws, DW_SPI_IMR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"ISR: \t\t0x%08x\n", dw_readl(dws, DW_SPI_ISR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"DMACR: \t\t0x%08x\n", dw_readl(dws, DW_SPI_DMACR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"DMATDLR: \t0x%08x\n", dw_readl(dws, DW_SPI_DMATDLR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"DMARDLR: \t0x%08x\n", dw_readl(dws, DW_SPI_DMARDLR));
+	len += snprintf(buf + len, SPI_REGS_BUFSIZE - len,
+			"=================================\n");
+
+	ret = simple_read_from_buffer(user_buf, count, ppos, buf, len);
+	kfree(buf);
+	return ret;
+}
+
+static const struct file_operations dw_spi_regs_ops = {
+	.owner		= THIS_MODULE,
+	.open		= simple_open,
+	.read		= dw_spi_show_regs,
+	.llseek		= default_llseek,
+};
+
+static int dw_spi_debugfs_init(struct dw_spi *dws)
+{
+	char name[32];
+
+	snprintf(name, 32, "dw_spi%d", dws->master->bus_num);
+	dws->debugfs = debugfs_create_dir(name, NULL);
+	if (!dws->debugfs)
+		return -ENOMEM;
+
+	debugfs_create_file("registers", S_IFREG | S_IRUGO,
+		dws->debugfs, (void *)dws, &dw_spi_regs_ops);
+	return 0;
+}
+
+static void dw_spi_debugfs_remove(struct dw_spi *dws)
+{
+	debugfs_remove_recursive(dws->debugfs);
+}
+
+#else
+static inline int dw_spi_debugfs_init(struct dw_spi *dws)
+{
+	return 0;
+}
+
+static inline void dw_spi_debugfs_remove(struct dw_spi *dws)
+{
+}
+#endif /* CONFIG_DEBUG_FS */
+
+void dw_spi_set_cs(struct spi_device *spi, bool enable)
+{
+	struct dw_spi *dws = spi_controller_get_devdata(spi->controller);
+	struct chip_data *chip = spi_get_ctldata(spi);
+
+	/* Chip select logic is inverted from spi_set_cs() */
+	if (chip && chip->cs_control)
+		chip->cs_control(!enable);
+
+	if (!enable)
+		dw_writel(dws, DW_SPI_SER, BIT(spi->chip_select));
+}
+EXPORT_SYMBOL_GPL(dw_spi_set_cs);
+
+/* Return the max entries we can fill into tx fifo */
+static inline u32 tx_max(struct dw_spi *dws)
+{
+	u32 tx_left, tx_room, rxtx_gap;
+
+	tx_left = (dws->tx_end - dws->tx) / dws->n_bytes;
+	tx_room = dws->fifo_len - dw_readl(dws, DW_SPI_TXFLR);
+
+	/*
+	 * Another concern is about the tx/rx mismatch, we
+	 * though to use (dws->fifo_len - rxflr - txflr) as
+	 * one maximum value for tx, but it doesn't cover the
+	 * data which is out of tx/rx fifo and inside the
+	 * shift registers. So a control from sw point of
+	 * view is taken.
+	 */
+	rxtx_gap =  ((dws->rx_end - dws->rx) - (dws->tx_end - dws->tx))
+			/ dws->n_bytes;
+
+	return min3(tx_left, tx_room, (u32) (dws->fifo_len - rxtx_gap));
+}
+
+/* Return the max entries we should read out of rx fifo */
+static inline u32 rx_max(struct dw_spi *dws)
+{
+	u32 rx_left = (dws->rx_end - dws->rx) / dws->n_bytes;
+
+	return min_t(u32, rx_left, dw_readl(dws, DW_SPI_RXFLR));
+}
+
+static void dw_writer(struct dw_spi *dws)
+{
+	u32 max = tx_max(dws);
+	u16 txw = 0;
+
+	while (max--) {
+		/* Set the tx word if the transfer's original "tx" is not null */
+		if (dws->tx_end - dws->len) {
+			if (dws->n_bytes == 1)
+				txw = *(u8 *)(dws->tx);
+			else
+				txw = *(u16 *)(dws->tx);
+		}
+		dw_write_io_reg(dws, DW_SPI_DR, txw);
+		dws->tx += dws->n_bytes;
+	}
+}
+
+static void dw_reader(struct dw_spi *dws)
+{
+	u32 max = rx_max(dws);
+	u16 rxw;
+
+	while (max--) {
+		rxw = dw_read_io_reg(dws, DW_SPI_DR);
+		/* Care rx only if the transfer's original "rx" is not null */
+		if (dws->rx_end - dws->len) {
+			if (dws->n_bytes == 1)
+				*(u8 *)(dws->rx) = rxw;
+			else
+				*(u16 *)(dws->rx) = rxw;
+		}
+		dws->rx += dws->n_bytes;
+	}
+}
+
+static void int_error_stop(struct dw_spi *dws, const char *msg)
+{
+	spi_reset_chip(dws);
+
+	dev_err(&dws->master->dev, "%s\n", msg);
+	dws->master->cur_msg->status = -EIO;
+	spi_finalize_current_transfer(dws->master);
+}
+
+static irqreturn_t interrupt_transfer(struct dw_spi *dws)
+{
+	u16 irq_status = dw_readl(dws, DW_SPI_ISR);
+
+	/* Error handling */
+	if (irq_status & (SPI_INT_TXOI | SPI_INT_RXOI | SPI_INT_RXUI)) {
+		dw_readl(dws, DW_SPI_ICR);
+		int_error_stop(dws, "interrupt_transfer: fifo overrun/underrun");
+		return IRQ_HANDLED;
+	}
+
+	dw_reader(dws);
+	if (dws->rx_end == dws->rx) {
+		spi_mask_intr(dws, SPI_INT_TXEI);
+		spi_finalize_current_transfer(dws->master);
+		return IRQ_HANDLED;
+	}
+	if (irq_status & SPI_INT_TXEI) {
+		spi_mask_intr(dws, SPI_INT_TXEI);
+		dw_writer(dws);
+		/* Enable TX irq always, it will be disabled when RX finished */
+		spi_umask_intr(dws, SPI_INT_TXEI);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t dw_spi_irq(int irq, void *dev_id)
+{
+	struct spi_controller *master = dev_id;
+	struct dw_spi *dws = spi_controller_get_devdata(master);
+	u16 irq_status = dw_readl(dws, DW_SPI_ISR) & 0x3f;
+
+	if (!irq_status)
+		return IRQ_NONE;
+
+	if (!master->cur_msg) {
+		spi_mask_intr(dws, SPI_INT_TXEI);
+		return IRQ_HANDLED;
+	}
+
+	return dws->transfer_handler(dws);
+}
+
+/* Must be called inside pump_transfers() */
+static int poll_transfer(struct dw_spi *dws)
+{
+	do {
+		dw_writer(dws);
+		dw_reader(dws);
+		cpu_relax();
+	} while (dws->rx_end > dws->rx);
+
+	return 0;
+}
+
+static int dw_spi_transfer_one(struct spi_controller *master,
+		struct spi_device *spi, struct spi_transfer *transfer)
+{
+	struct dw_spi *dws = spi_controller_get_devdata(master);
+	struct chip_data *chip = spi_get_ctldata(spi);
+	u8 imask = 0;
+	u16 txlevel = 0;
+	u32 cr0;
+	int ret;
+
+	dws->dma_mapped = 0;
+
+	dws->tx = (void *)transfer->tx_buf;
+	dws->tx_end = dws->tx + transfer->len;
+	dws->rx = transfer->rx_buf;
+	dws->rx_end = dws->rx + transfer->len;
+	dws->len = transfer->len;
+
+	spi_enable_chip(dws, 0);
+
+	/* Handle per transfer options for bpw and speed */
+	if (transfer->speed_hz != dws->current_freq) {
+		if (transfer->speed_hz != chip->speed_hz) {
+			/* clk_div doesn't support odd number */
+			chip->clk_div = (DIV_ROUND_UP(dws->max_freq, transfer->speed_hz) + 1) & 0xfffe;
+			chip->speed_hz = transfer->speed_hz;
+		}
+		dws->current_freq = transfer->speed_hz;
+		spi_set_clk(dws, chip->clk_div);
+	}
+	if (transfer->bits_per_word == 8) {
+		dws->n_bytes = 1;
+		dws->dma_width = 1;
+	} else if (transfer->bits_per_word == 16) {
+		dws->n_bytes = 2;
+		dws->dma_width = 2;
+	} else {
+		return -EINVAL;
+	}
+	/* Default SPI mode is SCPOL = 0, SCPH = 0 */
+	cr0 = (transfer->bits_per_word - 1)
+		| (chip->type << SPI_FRF_OFFSET)
+		| (spi->mode << SPI_MODE_OFFSET)
+		| (chip->tmode << SPI_TMOD_OFFSET);
+
+	/*
+	 * Adjust transfer mode if necessary. Requires platform dependent
+	 * chipselect mechanism.
+	 */
+	if (chip->cs_control) {
+		if (dws->rx && dws->tx)
+			chip->tmode = SPI_TMOD_TR;
+		else if (dws->rx)
+			chip->tmode = SPI_TMOD_RO;
+		else
+			chip->tmode = SPI_TMOD_TO;
+
+		cr0 &= ~SPI_TMOD_MASK;
+		cr0 |= (chip->tmode << SPI_TMOD_OFFSET);
+	}
+
+	dw_writel(dws, DW_SPI_CTRL0, cr0);
+
+	/* Check if current transfer is a DMA transaction */
+	if (master->can_dma && master->can_dma(master, spi, transfer))
+		dws->dma_mapped = master->cur_msg_mapped;
+
+	/* For poll mode just disable all interrupts */
+	spi_mask_intr(dws, 0xff);
+
+	/*
+	 * Interrupt mode
+	 * we only need set the TXEI IRQ, as TX/RX always happen syncronizely
+	 */
+	if (dws->dma_mapped) {
+		ret = dws->dma_ops->dma_setup(dws, transfer);
+		if (ret < 0) {
+			spi_enable_chip(dws, 1);
+			return ret;
+		}
+	} else if (!chip->poll_mode) {
+		txlevel = min_t(u16, dws->fifo_len / 2, dws->len / dws->n_bytes);
+		dw_writel(dws, DW_SPI_TXFLTR, txlevel);
+
+		/* Set the interrupt mask */
+		imask |= SPI_INT_TXEI | SPI_INT_TXOI |
+			 SPI_INT_RXUI | SPI_INT_RXOI;
+		spi_umask_intr(dws, imask);
+
+		dws->transfer_handler = interrupt_transfer;
+	}
+
+	spi_enable_chip(dws, 1);
+
+	if (dws->dma_mapped) {
+		ret = dws->dma_ops->dma_transfer(dws, transfer);
+		if (ret < 0)
+			return ret;
+	}
+
+	if (chip->poll_mode)
+		return poll_transfer(dws);
+
+	return 1;
+}
+
+static void dw_spi_handle_err(struct spi_controller *master,
+		struct spi_message *msg)
+{
+	struct dw_spi *dws = spi_controller_get_devdata(master);
+
+	if (dws->dma_mapped)
+		dws->dma_ops->dma_stop(dws);
+
+	spi_reset_chip(dws);
+}
+
+/* This may be called twice for each spi dev */
+static int dw_spi_setup(struct spi_device *spi)
+{
+	struct dw_spi_chip *chip_info = NULL;
+	struct chip_data *chip;
+	int ret;
+
+	/* Only alloc on first setup */
+	chip = spi_get_ctldata(spi);
+	if (!chip) {
+		chip = kzalloc(sizeof(struct chip_data), GFP_KERNEL);
+		if (!chip)
+			return -ENOMEM;
+		spi_set_ctldata(spi, chip);
+	}
+
+	/*
+	 * Protocol drivers may change the chip settings, so...
+	 * if chip_info exists, use it
+	 */
+	chip_info = spi->controller_data;
+
+	/* chip_info doesn't always exist */
+	if (chip_info) {
+		if (chip_info->cs_control)
+			chip->cs_control = chip_info->cs_control;
+
+		chip->poll_mode = chip_info->poll_mode;
+		chip->type = chip_info->type;
+	}
+
+	chip->tmode = SPI_TMOD_TR;
+
+	if (gpio_is_valid(spi->cs_gpio)) {
+		ret = gpio_direction_output(spi->cs_gpio,
+				!(spi->mode & SPI_CS_HIGH));
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static void dw_spi_cleanup(struct spi_device *spi)
+{
+	struct chip_data *chip = spi_get_ctldata(spi);
+
+	kfree(chip);
+	spi_set_ctldata(spi, NULL);
+}
+
+/* Restart the controller, disable all interrupts, clean rx fifo */
+static void spi_hw_init(struct device *dev, struct dw_spi *dws)
+{
+	spi_reset_chip(dws);
+
+	/*
+	 * Try to detect the FIFO depth if not set by interface driver,
+	 * the depth could be from 2 to 256 from HW spec
+	 */
+	if (!dws->fifo_len) {
+		u32 fifo;
+
+		for (fifo = 1; fifo < 256; fifo++) {
+			dw_writel(dws, DW_SPI_TXFLTR, fifo);
+			if (fifo != dw_readl(dws, DW_SPI_TXFLTR))
+				break;
+		}
+		dw_writel(dws, DW_SPI_TXFLTR, 0);
+
+		dws->fifo_len = (fifo == 1) ? 0 : fifo;
+		dev_dbg(dev, "Detected FIFO size: %u bytes\n", dws->fifo_len);
+	}
+}
+
+int dw_spi_add_host(struct device *dev, struct dw_spi *dws)
+{
+	struct spi_controller *master;
+	int ret;
+
+	BUG_ON(dws == NULL);
+
+	master = spi_alloc_master(dev, 0);
+	if (!master)
+		return -ENOMEM;
+
+	dws->master = master;
+	dws->type = SSI_MOTO_SPI;
+	dws->dma_inited = 0;
+	dws->dma_addr = (dma_addr_t)(dws->paddr + DW_SPI_DR);
+
+	spi_controller_set_devdata(master, dws);
+
+	ret = request_irq(dws->irq, dw_spi_irq, IRQF_SHARED, dev_name(dev),
+			  master);
+	if (ret < 0) {
+		dev_err(dev, "can not get IRQ\n");
+		goto err_free_master;
+	}
+
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_LOOP;
+	master->bits_per_word_mask = SPI_BPW_MASK(8) | SPI_BPW_MASK(16);
+	master->bus_num = dws->bus_num;
+	master->num_chipselect = dws->num_cs;
+	master->setup = dw_spi_setup;
+	master->cleanup = dw_spi_cleanup;
+	master->set_cs = dw_spi_set_cs;
+	master->transfer_one = dw_spi_transfer_one;
+	master->handle_err = dw_spi_handle_err;
+	master->max_speed_hz = dws->max_freq;
+	master->dev.of_node = dev->of_node;
+	master->flags = SPI_MASTER_GPIO_SS;
+
+	if (dws->set_cs)
+		master->set_cs = dws->set_cs;
+
+	/* Basic HW init */
+	spi_hw_init(dev, dws);
+
+	if (dws->dma_ops && dws->dma_ops->dma_init) {
+		ret = dws->dma_ops->dma_init(dws);
+		if (ret) {
+			dev_warn(dev, "DMA init failed\n");
+			dws->dma_inited = 0;
+		} else {
+			master->can_dma = dws->dma_ops->can_dma;
+		}
+	}
+
+	ret = devm_spi_register_controller(dev, master);
+	if (ret) {
+		dev_err(&master->dev, "problem registering spi master\n");
+		goto err_dma_exit;
+	}
+
+	dw_spi_debugfs_init(dws);
+	return 0;
+
+err_dma_exit:
+	if (dws->dma_ops && dws->dma_ops->dma_exit)
+		dws->dma_ops->dma_exit(dws);
+	spi_enable_chip(dws, 0);
+	free_irq(dws->irq, master);
+err_free_master:
+	spi_controller_put(master);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(dw_spi_add_host);
+
+void dw_spi_remove_host(struct dw_spi *dws)
+{
+	dw_spi_debugfs_remove(dws);
+
+	if (dws->dma_ops && dws->dma_ops->dma_exit)
+		dws->dma_ops->dma_exit(dws);
+
+	spi_shutdown_chip(dws);
+
+	free_irq(dws->irq, dws->master);
+}
+EXPORT_SYMBOL_GPL(dw_spi_remove_host);
+
+int dw_spi_suspend_host(struct dw_spi *dws)
+{
+	int ret;
+
+	ret = spi_controller_suspend(dws->master);
+	if (ret)
+		return ret;
+
+	spi_shutdown_chip(dws);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(dw_spi_suspend_host);
+
+int dw_spi_resume_host(struct dw_spi *dws)
+{
+	int ret;
+
+	spi_hw_init(&dws->master->dev, dws);
+	ret = spi_controller_resume(dws->master);
+	if (ret)
+		dev_err(&dws->master->dev, "fail to start queue (%d)\n", ret);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(dw_spi_resume_host);
+
+MODULE_AUTHOR("Feng Tang <feng.tang@intel.com>");
+MODULE_DESCRIPTION("Driver for DesignWare SPI controller core");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-dw.h b/drivers/spi/spi-dw.h
new file mode 100644
index 0000000..0168b08
--- /dev/null
+++ b/drivers/spi/spi-dw.h
@@ -0,0 +1,256 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef DW_SPI_HEADER_H
+#define DW_SPI_HEADER_H
+
+#include <linux/io.h>
+#include <linux/scatterlist.h>
+#include <linux/gpio.h>
+
+/* Register offsets */
+#define DW_SPI_CTRL0			0x00
+#define DW_SPI_CTRL1			0x04
+#define DW_SPI_SSIENR			0x08
+#define DW_SPI_MWCR			0x0c
+#define DW_SPI_SER			0x10
+#define DW_SPI_BAUDR			0x14
+#define DW_SPI_TXFLTR			0x18
+#define DW_SPI_RXFLTR			0x1c
+#define DW_SPI_TXFLR			0x20
+#define DW_SPI_RXFLR			0x24
+#define DW_SPI_SR			0x28
+#define DW_SPI_IMR			0x2c
+#define DW_SPI_ISR			0x30
+#define DW_SPI_RISR			0x34
+#define DW_SPI_TXOICR			0x38
+#define DW_SPI_RXOICR			0x3c
+#define DW_SPI_RXUICR			0x40
+#define DW_SPI_MSTICR			0x44
+#define DW_SPI_ICR			0x48
+#define DW_SPI_DMACR			0x4c
+#define DW_SPI_DMATDLR			0x50
+#define DW_SPI_DMARDLR			0x54
+#define DW_SPI_IDR			0x58
+#define DW_SPI_VERSION			0x5c
+#define DW_SPI_DR			0x60
+
+/* Bit fields in CTRLR0 */
+#define SPI_DFS_OFFSET			0
+
+#define SPI_FRF_OFFSET			4
+#define SPI_FRF_SPI			0x0
+#define SPI_FRF_SSP			0x1
+#define SPI_FRF_MICROWIRE		0x2
+#define SPI_FRF_RESV			0x3
+
+#define SPI_MODE_OFFSET			6
+#define SPI_SCPH_OFFSET			6
+#define SPI_SCOL_OFFSET			7
+
+#define SPI_TMOD_OFFSET			8
+#define SPI_TMOD_MASK			(0x3 << SPI_TMOD_OFFSET)
+#define	SPI_TMOD_TR			0x0		/* xmit & recv */
+#define SPI_TMOD_TO			0x1		/* xmit only */
+#define SPI_TMOD_RO			0x2		/* recv only */
+#define SPI_TMOD_EPROMREAD		0x3		/* eeprom read mode */
+
+#define SPI_SLVOE_OFFSET		10
+#define SPI_SRL_OFFSET			11
+#define SPI_CFS_OFFSET			12
+
+/* Bit fields in SR, 7 bits */
+#define SR_MASK				0x7f		/* cover 7 bits */
+#define SR_BUSY				(1 << 0)
+#define SR_TF_NOT_FULL			(1 << 1)
+#define SR_TF_EMPT			(1 << 2)
+#define SR_RF_NOT_EMPT			(1 << 3)
+#define SR_RF_FULL			(1 << 4)
+#define SR_TX_ERR			(1 << 5)
+#define SR_DCOL				(1 << 6)
+
+/* Bit fields in ISR, IMR, RISR, 7 bits */
+#define SPI_INT_TXEI			(1 << 0)
+#define SPI_INT_TXOI			(1 << 1)
+#define SPI_INT_RXUI			(1 << 2)
+#define SPI_INT_RXOI			(1 << 3)
+#define SPI_INT_RXFI			(1 << 4)
+#define SPI_INT_MSTI			(1 << 5)
+
+/* Bit fields in DMACR */
+#define SPI_DMA_RDMAE			(1 << 0)
+#define SPI_DMA_TDMAE			(1 << 1)
+
+/* TX RX interrupt level threshold, max can be 256 */
+#define SPI_INT_THRESHOLD		32
+
+enum dw_ssi_type {
+	SSI_MOTO_SPI = 0,
+	SSI_TI_SSP,
+	SSI_NS_MICROWIRE,
+};
+
+struct dw_spi;
+struct dw_spi_dma_ops {
+	int (*dma_init)(struct dw_spi *dws);
+	void (*dma_exit)(struct dw_spi *dws);
+	int (*dma_setup)(struct dw_spi *dws, struct spi_transfer *xfer);
+	bool (*can_dma)(struct spi_controller *master, struct spi_device *spi,
+			struct spi_transfer *xfer);
+	int (*dma_transfer)(struct dw_spi *dws, struct spi_transfer *xfer);
+	void (*dma_stop)(struct dw_spi *dws);
+};
+
+struct dw_spi {
+	struct spi_controller	*master;
+	enum dw_ssi_type	type;
+
+	void __iomem		*regs;
+	unsigned long		paddr;
+	int			irq;
+	u32			fifo_len;	/* depth of the FIFO buffer */
+	u32			max_freq;	/* max bus freq supported */
+
+	u32			reg_io_width;	/* DR I/O width in bytes */
+	u16			bus_num;
+	u16			num_cs;		/* supported slave numbers */
+	void (*set_cs)(struct spi_device *spi, bool enable);
+
+	/* Current message transfer state info */
+	size_t			len;
+	void			*tx;
+	void			*tx_end;
+	void			*rx;
+	void			*rx_end;
+	int			dma_mapped;
+	u8			n_bytes;	/* current is a 1/2 bytes op */
+	u32			dma_width;
+	irqreturn_t		(*transfer_handler)(struct dw_spi *dws);
+	u32			current_freq;	/* frequency in hz */
+
+	/* DMA info */
+	int			dma_inited;
+	struct dma_chan		*txchan;
+	struct dma_chan		*rxchan;
+	unsigned long		dma_chan_busy;
+	dma_addr_t		dma_addr; /* phy address of the Data register */
+	const struct dw_spi_dma_ops *dma_ops;
+	void			*dma_tx;
+	void			*dma_rx;
+
+	/* Bus interface info */
+	void			*priv;
+#ifdef CONFIG_DEBUG_FS
+	struct dentry *debugfs;
+#endif
+};
+
+static inline u32 dw_readl(struct dw_spi *dws, u32 offset)
+{
+	return __raw_readl(dws->regs + offset);
+}
+
+static inline u16 dw_readw(struct dw_spi *dws, u32 offset)
+{
+	return __raw_readw(dws->regs + offset);
+}
+
+static inline void dw_writel(struct dw_spi *dws, u32 offset, u32 val)
+{
+	__raw_writel(val, dws->regs + offset);
+}
+
+static inline void dw_writew(struct dw_spi *dws, u32 offset, u16 val)
+{
+	__raw_writew(val, dws->regs + offset);
+}
+
+static inline u32 dw_read_io_reg(struct dw_spi *dws, u32 offset)
+{
+	switch (dws->reg_io_width) {
+	case 2:
+		return dw_readw(dws, offset);
+	case 4:
+	default:
+		return dw_readl(dws, offset);
+	}
+}
+
+static inline void dw_write_io_reg(struct dw_spi *dws, u32 offset, u32 val)
+{
+	switch (dws->reg_io_width) {
+	case 2:
+		dw_writew(dws, offset, val);
+		break;
+	case 4:
+	default:
+		dw_writel(dws, offset, val);
+		break;
+	}
+}
+
+static inline void spi_enable_chip(struct dw_spi *dws, int enable)
+{
+	dw_writel(dws, DW_SPI_SSIENR, (enable ? 1 : 0));
+}
+
+static inline void spi_set_clk(struct dw_spi *dws, u16 div)
+{
+	dw_writel(dws, DW_SPI_BAUDR, div);
+}
+
+/* Disable IRQ bits */
+static inline void spi_mask_intr(struct dw_spi *dws, u32 mask)
+{
+	u32 new_mask;
+
+	new_mask = dw_readl(dws, DW_SPI_IMR) & ~mask;
+	dw_writel(dws, DW_SPI_IMR, new_mask);
+}
+
+/* Enable IRQ bits */
+static inline void spi_umask_intr(struct dw_spi *dws, u32 mask)
+{
+	u32 new_mask;
+
+	new_mask = dw_readl(dws, DW_SPI_IMR) | mask;
+	dw_writel(dws, DW_SPI_IMR, new_mask);
+}
+
+/*
+ * This does disable the SPI controller, interrupts, and re-enable the
+ * controller back. Transmit and receive FIFO buffers are cleared when the
+ * device is disabled.
+ */
+static inline void spi_reset_chip(struct dw_spi *dws)
+{
+	spi_enable_chip(dws, 0);
+	spi_mask_intr(dws, 0xff);
+	spi_enable_chip(dws, 1);
+}
+
+static inline void spi_shutdown_chip(struct dw_spi *dws)
+{
+	spi_enable_chip(dws, 0);
+	spi_set_clk(dws, 0);
+}
+
+/*
+ * Each SPI slave device to work with dw_api controller should
+ * has such a structure claiming its working mode (poll or PIO/DMA),
+ * which can be save in the "controller_data" member of the
+ * struct spi_device.
+ */
+struct dw_spi_chip {
+	u8 poll_mode;	/* 1 for controller polling mode */
+	u8 type;	/* SPI/SSP/MicroWire */
+	void (*cs_control)(u32 command);
+};
+
+extern void dw_spi_set_cs(struct spi_device *spi, bool enable);
+extern int dw_spi_add_host(struct device *dev, struct dw_spi *dws);
+extern void dw_spi_remove_host(struct dw_spi *dws);
+extern int dw_spi_suspend_host(struct dw_spi *dws);
+extern int dw_spi_resume_host(struct dw_spi *dws);
+
+/* platform related setup */
+extern int dw_spi_mid_init(struct dw_spi *dws); /* Intel MID platforms */
+#endif /* DW_SPI_HEADER_H */
diff --git a/drivers/spi/spi-efm32.c b/drivers/spi/spi-efm32.c
new file mode 100644
index 0000000..065fe87
--- /dev/null
+++ b/drivers/spi/spi-efm32.c
@@ -0,0 +1,505 @@
+/*
+ * Copyright (C) 2012-2013 Uwe Kleine-Koenig for Pengutronix
+ *
+ * This program is free software; you can redistribute it and/or modify it under
+ * the terms of the GNU General Public License version 2 as published by the
+ * Free Software Foundation.
+ */
+#include <linux/kernel.h>
+#include <linux/io.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/gpio.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/module.h>
+#include <linux/of_gpio.h>
+#include <linux/platform_data/efm32-spi.h>
+
+#define DRIVER_NAME "efm32-spi"
+
+#define MASK_VAL(mask, val)		((val << __ffs(mask)) & mask)
+
+#define REG_CTRL		0x00
+#define REG_CTRL_SYNC			0x0001
+#define REG_CTRL_CLKPOL			0x0100
+#define REG_CTRL_CLKPHA			0x0200
+#define REG_CTRL_MSBF			0x0400
+#define REG_CTRL_TXBIL			0x1000
+
+#define REG_FRAME		0x04
+#define REG_FRAME_DATABITS__MASK	0x000f
+#define REG_FRAME_DATABITS(n)		((n) - 3)
+
+#define REG_CMD			0x0c
+#define REG_CMD_RXEN			0x0001
+#define REG_CMD_RXDIS			0x0002
+#define REG_CMD_TXEN			0x0004
+#define REG_CMD_TXDIS			0x0008
+#define REG_CMD_MASTEREN		0x0010
+
+#define REG_STATUS		0x10
+#define REG_STATUS_TXENS		0x0002
+#define REG_STATUS_TXC			0x0020
+#define REG_STATUS_TXBL			0x0040
+#define REG_STATUS_RXDATAV		0x0080
+
+#define REG_CLKDIV		0x14
+
+#define REG_RXDATAX		0x18
+#define REG_RXDATAX_RXDATA__MASK	0x01ff
+#define REG_RXDATAX_PERR		0x4000
+#define REG_RXDATAX_FERR		0x8000
+
+#define REG_TXDATA		0x34
+
+#define REG_IF		0x40
+#define REG_IF_TXBL			0x0002
+#define REG_IF_RXDATAV			0x0004
+
+#define REG_IFS		0x44
+#define REG_IFC		0x48
+#define REG_IEN		0x4c
+
+#define REG_ROUTE		0x54
+#define REG_ROUTE_RXPEN			0x0001
+#define REG_ROUTE_TXPEN			0x0002
+#define REG_ROUTE_CLKPEN		0x0008
+#define REG_ROUTE_LOCATION__MASK	0x0700
+#define REG_ROUTE_LOCATION(n)		MASK_VAL(REG_ROUTE_LOCATION__MASK, (n))
+
+struct efm32_spi_ddata {
+	struct spi_bitbang bitbang;
+
+	spinlock_t lock;
+
+	struct clk *clk;
+	void __iomem *base;
+	unsigned int rxirq, txirq;
+	struct efm32_spi_pdata pdata;
+
+	/* irq data */
+	struct completion done;
+	const u8 *tx_buf;
+	u8 *rx_buf;
+	unsigned tx_len, rx_len;
+
+	/* chip selects */
+	unsigned csgpio[];
+};
+
+#define ddata_to_dev(ddata)	(&(ddata->bitbang.master->dev))
+#define efm32_spi_vdbg(ddata, format, arg...)	\
+	dev_vdbg(ddata_to_dev(ddata), format, ##arg)
+
+static void efm32_spi_write32(struct efm32_spi_ddata *ddata,
+		u32 value, unsigned offset)
+{
+	writel_relaxed(value, ddata->base + offset);
+}
+
+static u32 efm32_spi_read32(struct efm32_spi_ddata *ddata, unsigned offset)
+{
+	return readl_relaxed(ddata->base + offset);
+}
+
+static void efm32_spi_chipselect(struct spi_device *spi, int is_on)
+{
+	struct efm32_spi_ddata *ddata = spi_master_get_devdata(spi->master);
+	int value = !(spi->mode & SPI_CS_HIGH) == !(is_on == BITBANG_CS_ACTIVE);
+
+	gpio_set_value(ddata->csgpio[spi->chip_select], value);
+}
+
+static int efm32_spi_setup_transfer(struct spi_device *spi,
+		struct spi_transfer *t)
+{
+	struct efm32_spi_ddata *ddata = spi_master_get_devdata(spi->master);
+
+	unsigned bpw = t->bits_per_word ?: spi->bits_per_word;
+	unsigned speed = t->speed_hz ?: spi->max_speed_hz;
+	unsigned long clkfreq = clk_get_rate(ddata->clk);
+	u32 clkdiv;
+
+	efm32_spi_write32(ddata, REG_CTRL_SYNC | REG_CTRL_MSBF |
+			(spi->mode & SPI_CPHA ? REG_CTRL_CLKPHA : 0) |
+			(spi->mode & SPI_CPOL ? REG_CTRL_CLKPOL : 0), REG_CTRL);
+
+	efm32_spi_write32(ddata,
+			REG_FRAME_DATABITS(bpw), REG_FRAME);
+
+	if (2 * speed >= clkfreq)
+		clkdiv = 0;
+	else
+		clkdiv = 64 * (DIV_ROUND_UP(2 * clkfreq, speed) - 4);
+
+	if (clkdiv > (1U << 21))
+		return -EINVAL;
+
+	efm32_spi_write32(ddata, clkdiv, REG_CLKDIV);
+	efm32_spi_write32(ddata, REG_CMD_MASTEREN, REG_CMD);
+	efm32_spi_write32(ddata, REG_CMD_RXEN | REG_CMD_TXEN, REG_CMD);
+
+	return 0;
+}
+
+static void efm32_spi_tx_u8(struct efm32_spi_ddata *ddata)
+{
+	u8 val = 0;
+
+	if (ddata->tx_buf) {
+		val = *ddata->tx_buf;
+		ddata->tx_buf++;
+	}
+
+	ddata->tx_len--;
+	efm32_spi_write32(ddata, val, REG_TXDATA);
+	efm32_spi_vdbg(ddata, "%s: tx 0x%x\n", __func__, val);
+}
+
+static void efm32_spi_rx_u8(struct efm32_spi_ddata *ddata)
+{
+	u32 rxdata = efm32_spi_read32(ddata, REG_RXDATAX);
+	efm32_spi_vdbg(ddata, "%s: rx 0x%x\n", __func__, rxdata);
+
+	if (ddata->rx_buf) {
+		*ddata->rx_buf = rxdata;
+		ddata->rx_buf++;
+	}
+
+	ddata->rx_len--;
+}
+
+static void efm32_spi_filltx(struct efm32_spi_ddata *ddata)
+{
+	while (ddata->tx_len &&
+			ddata->tx_len + 2 > ddata->rx_len &&
+			efm32_spi_read32(ddata, REG_STATUS) & REG_STATUS_TXBL) {
+		efm32_spi_tx_u8(ddata);
+	}
+}
+
+static int efm32_spi_txrx_bufs(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct efm32_spi_ddata *ddata = spi_master_get_devdata(spi->master);
+	int ret = -EBUSY;
+
+	spin_lock_irq(&ddata->lock);
+
+	if (ddata->tx_buf || ddata->rx_buf)
+		goto out_unlock;
+
+	ddata->tx_buf = t->tx_buf;
+	ddata->rx_buf = t->rx_buf;
+	ddata->tx_len = ddata->rx_len =
+		t->len * DIV_ROUND_UP(t->bits_per_word, 8);
+
+	efm32_spi_filltx(ddata);
+
+	reinit_completion(&ddata->done);
+
+	efm32_spi_write32(ddata, REG_IF_TXBL | REG_IF_RXDATAV, REG_IEN);
+
+	spin_unlock_irq(&ddata->lock);
+
+	wait_for_completion(&ddata->done);
+
+	spin_lock_irq(&ddata->lock);
+
+	ret = t->len - max(ddata->tx_len, ddata->rx_len);
+
+	efm32_spi_write32(ddata, 0, REG_IEN);
+	ddata->tx_buf = ddata->rx_buf = NULL;
+
+out_unlock:
+	spin_unlock_irq(&ddata->lock);
+
+	return ret;
+}
+
+static irqreturn_t efm32_spi_rxirq(int irq, void *data)
+{
+	struct efm32_spi_ddata *ddata = data;
+	irqreturn_t ret = IRQ_NONE;
+
+	spin_lock(&ddata->lock);
+
+	while (ddata->rx_len > 0 &&
+			efm32_spi_read32(ddata, REG_STATUS) &
+			REG_STATUS_RXDATAV) {
+		efm32_spi_rx_u8(ddata);
+
+		ret = IRQ_HANDLED;
+	}
+
+	if (!ddata->rx_len) {
+		u32 ien = efm32_spi_read32(ddata, REG_IEN);
+
+		ien &= ~REG_IF_RXDATAV;
+
+		efm32_spi_write32(ddata, ien, REG_IEN);
+
+		complete(&ddata->done);
+	}
+
+	spin_unlock(&ddata->lock);
+
+	return ret;
+}
+
+static irqreturn_t efm32_spi_txirq(int irq, void *data)
+{
+	struct efm32_spi_ddata *ddata = data;
+
+	efm32_spi_vdbg(ddata,
+			"%s: txlen = %u, rxlen = %u, if=0x%08x, stat=0x%08x\n",
+			__func__, ddata->tx_len, ddata->rx_len,
+			efm32_spi_read32(ddata, REG_IF),
+			efm32_spi_read32(ddata, REG_STATUS));
+
+	spin_lock(&ddata->lock);
+
+	efm32_spi_filltx(ddata);
+
+	efm32_spi_vdbg(ddata, "%s: txlen = %u, rxlen = %u\n",
+			__func__, ddata->tx_len, ddata->rx_len);
+
+	if (!ddata->tx_len) {
+		u32 ien = efm32_spi_read32(ddata, REG_IEN);
+
+		ien &= ~REG_IF_TXBL;
+
+		efm32_spi_write32(ddata, ien, REG_IEN);
+		efm32_spi_vdbg(ddata, "disable TXBL\n");
+	}
+
+	spin_unlock(&ddata->lock);
+
+	return IRQ_HANDLED;
+}
+
+static u32 efm32_spi_get_configured_location(struct efm32_spi_ddata *ddata)
+{
+	u32 reg = efm32_spi_read32(ddata, REG_ROUTE);
+
+	return (reg & REG_ROUTE_LOCATION__MASK) >> __ffs(REG_ROUTE_LOCATION__MASK);
+}
+
+static void efm32_spi_probe_dt(struct platform_device *pdev,
+		struct spi_master *master, struct efm32_spi_ddata *ddata)
+{
+	struct device_node *np = pdev->dev.of_node;
+	u32 location;
+	int ret;
+
+	ret = of_property_read_u32(np, "energymicro,location", &location);
+
+	if (ret)
+		/* fall back to wrongly namespaced property */
+		ret = of_property_read_u32(np, "efm32,location", &location);
+
+	if (ret)
+		/* fall back to old and (wrongly) generic property "location" */
+		ret = of_property_read_u32(np, "location", &location);
+
+	if (!ret) {
+		dev_dbg(&pdev->dev, "using location %u\n", location);
+	} else {
+		/* default to location configured in hardware */
+		location = efm32_spi_get_configured_location(ddata);
+
+		dev_info(&pdev->dev, "fall back to location %u\n", location);
+	}
+
+	ddata->pdata.location = location;
+}
+
+static int efm32_spi_probe(struct platform_device *pdev)
+{
+	struct efm32_spi_ddata *ddata;
+	struct resource *res;
+	int ret;
+	struct spi_master *master;
+	struct device_node *np = pdev->dev.of_node;
+	int num_cs, i;
+
+	if (!np)
+		return -EINVAL;
+
+	num_cs = of_gpio_named_count(np, "cs-gpios");
+	if (num_cs < 0)
+		return num_cs;
+
+	master = spi_alloc_master(&pdev->dev,
+			sizeof(*ddata) + num_cs * sizeof(unsigned));
+	if (!master) {
+		dev_dbg(&pdev->dev,
+				"failed to allocate spi master controller\n");
+		return -ENOMEM;
+	}
+	platform_set_drvdata(pdev, master);
+
+	master->dev.of_node = pdev->dev.of_node;
+
+	master->num_chipselect = num_cs;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 16);
+
+	ddata = spi_master_get_devdata(master);
+
+	ddata->bitbang.master = master;
+	ddata->bitbang.chipselect = efm32_spi_chipselect;
+	ddata->bitbang.setup_transfer = efm32_spi_setup_transfer;
+	ddata->bitbang.txrx_bufs = efm32_spi_txrx_bufs;
+
+	spin_lock_init(&ddata->lock);
+	init_completion(&ddata->done);
+
+	ddata->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(ddata->clk)) {
+		ret = PTR_ERR(ddata->clk);
+		dev_err(&pdev->dev, "failed to get clock: %d\n", ret);
+		goto err;
+	}
+
+	for (i = 0; i < num_cs; ++i) {
+		ret = of_get_named_gpio(np, "cs-gpios", i);
+		if (ret < 0) {
+			dev_err(&pdev->dev, "failed to get csgpio#%u (%d)\n",
+					i, ret);
+			goto err;
+		}
+		ddata->csgpio[i] = ret;
+		dev_dbg(&pdev->dev, "csgpio#%u = %u\n", i, ddata->csgpio[i]);
+		ret = devm_gpio_request_one(&pdev->dev, ddata->csgpio[i],
+				GPIOF_OUT_INIT_LOW, DRIVER_NAME);
+		if (ret < 0) {
+			dev_err(&pdev->dev,
+					"failed to configure csgpio#%u (%d)\n",
+					i, ret);
+			goto err;
+		}
+	}
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res) {
+		ret = -ENODEV;
+		dev_err(&pdev->dev, "failed to determine base address\n");
+		goto err;
+	}
+
+	if (resource_size(res) < 0x60) {
+		ret = -EINVAL;
+		dev_err(&pdev->dev, "memory resource too small\n");
+		goto err;
+	}
+
+	ddata->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(ddata->base)) {
+		ret = PTR_ERR(ddata->base);
+		goto err;
+	}
+
+	ret = platform_get_irq(pdev, 0);
+	if (ret <= 0) {
+		dev_err(&pdev->dev, "failed to get rx irq (%d)\n", ret);
+		goto err;
+	}
+
+	ddata->rxirq = ret;
+
+	ret = platform_get_irq(pdev, 1);
+	if (ret <= 0)
+		ret = ddata->rxirq + 1;
+
+	ddata->txirq = ret;
+
+	ret = clk_prepare_enable(ddata->clk);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to enable clock (%d)\n", ret);
+		goto err;
+	}
+
+	efm32_spi_probe_dt(pdev, master, ddata);
+
+	efm32_spi_write32(ddata, 0, REG_IEN);
+	efm32_spi_write32(ddata, REG_ROUTE_TXPEN | REG_ROUTE_RXPEN |
+			REG_ROUTE_CLKPEN |
+			REG_ROUTE_LOCATION(ddata->pdata.location), REG_ROUTE);
+
+	ret = request_irq(ddata->rxirq, efm32_spi_rxirq,
+			0, DRIVER_NAME " rx", ddata);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to register rxirq (%d)\n", ret);
+		goto err_disable_clk;
+	}
+
+	ret = request_irq(ddata->txirq, efm32_spi_txirq,
+			0, DRIVER_NAME " tx", ddata);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to register txirq (%d)\n", ret);
+		goto err_free_rx_irq;
+	}
+
+	ret = spi_bitbang_start(&ddata->bitbang);
+	if (ret) {
+		dev_err(&pdev->dev, "spi_bitbang_start failed (%d)\n", ret);
+
+		free_irq(ddata->txirq, ddata);
+err_free_rx_irq:
+		free_irq(ddata->rxirq, ddata);
+err_disable_clk:
+		clk_disable_unprepare(ddata->clk);
+err:
+		spi_master_put(master);
+	}
+
+	return ret;
+}
+
+static int efm32_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct efm32_spi_ddata *ddata = spi_master_get_devdata(master);
+
+	spi_bitbang_stop(&ddata->bitbang);
+
+	efm32_spi_write32(ddata, 0, REG_IEN);
+
+	free_irq(ddata->txirq, ddata);
+	free_irq(ddata->rxirq, ddata);
+	clk_disable_unprepare(ddata->clk);
+	spi_master_put(master);
+
+	return 0;
+}
+
+static const struct of_device_id efm32_spi_dt_ids[] = {
+	{
+		.compatible = "energymicro,efm32-spi",
+	}, {
+		/* doesn't follow the "vendor,device" scheme, don't use */
+		.compatible = "efm32,spi",
+	}, {
+		/* sentinel */
+	}
+};
+MODULE_DEVICE_TABLE(of, efm32_spi_dt_ids);
+
+static struct platform_driver efm32_spi_driver = {
+	.probe = efm32_spi_probe,
+	.remove = efm32_spi_remove,
+
+	.driver = {
+		.name = DRIVER_NAME,
+		.of_match_table = efm32_spi_dt_ids,
+	},
+};
+module_platform_driver(efm32_spi_driver);
+
+MODULE_AUTHOR("Uwe Kleine-Koenig <u.kleine-koenig@pengutronix.de>");
+MODULE_DESCRIPTION("EFM32 SPI driver");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:" DRIVER_NAME);
diff --git a/drivers/spi/spi-ep93xx.c b/drivers/spi/spi-ep93xx.c
new file mode 100644
index 0000000..79fc394
--- /dev/null
+++ b/drivers/spi/spi-ep93xx.c
@@ -0,0 +1,792 @@
+/*
+ * Driver for Cirrus Logic EP93xx SPI controller.
+ *
+ * Copyright (C) 2010-2011 Mika Westerberg
+ *
+ * Explicit FIFO handling code was inspired by amba-pl022 driver.
+ *
+ * Chip select support using other than built-in GPIOs by H. Hartley Sweeten.
+ *
+ * For more information about the SPI controller see documentation on Cirrus
+ * Logic web site:
+ *     http://www.cirrus.com/en/pubs/manual/EP93xx_Users_Guide_UM1.pdf
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/io.h>
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/dmaengine.h>
+#include <linux/bitops.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/sched.h>
+#include <linux/scatterlist.h>
+#include <linux/gpio.h>
+#include <linux/spi/spi.h>
+
+#include <linux/platform_data/dma-ep93xx.h>
+#include <linux/platform_data/spi-ep93xx.h>
+
+#define SSPCR0			0x0000
+#define SSPCR0_MODE_SHIFT	6
+#define SSPCR0_SCR_SHIFT	8
+
+#define SSPCR1			0x0004
+#define SSPCR1_RIE		BIT(0)
+#define SSPCR1_TIE		BIT(1)
+#define SSPCR1_RORIE		BIT(2)
+#define SSPCR1_LBM		BIT(3)
+#define SSPCR1_SSE		BIT(4)
+#define SSPCR1_MS		BIT(5)
+#define SSPCR1_SOD		BIT(6)
+
+#define SSPDR			0x0008
+
+#define SSPSR			0x000c
+#define SSPSR_TFE		BIT(0)
+#define SSPSR_TNF		BIT(1)
+#define SSPSR_RNE		BIT(2)
+#define SSPSR_RFF		BIT(3)
+#define SSPSR_BSY		BIT(4)
+#define SSPCPSR			0x0010
+
+#define SSPIIR			0x0014
+#define SSPIIR_RIS		BIT(0)
+#define SSPIIR_TIS		BIT(1)
+#define SSPIIR_RORIS		BIT(2)
+#define SSPICR			SSPIIR
+
+/* timeout in milliseconds */
+#define SPI_TIMEOUT		5
+/* maximum depth of RX/TX FIFO */
+#define SPI_FIFO_SIZE		8
+
+/**
+ * struct ep93xx_spi - EP93xx SPI controller structure
+ * @clk: clock for the controller
+ * @mmio: pointer to ioremap()'d registers
+ * @sspdr_phys: physical address of the SSPDR register
+ * @tx: current byte in transfer to transmit
+ * @rx: current byte in transfer to receive
+ * @fifo_level: how full is FIFO (%0..%SPI_FIFO_SIZE - %1). Receiving one
+ *              frame decreases this level and sending one frame increases it.
+ * @dma_rx: RX DMA channel
+ * @dma_tx: TX DMA channel
+ * @dma_rx_data: RX parameters passed to the DMA engine
+ * @dma_tx_data: TX parameters passed to the DMA engine
+ * @rx_sgt: sg table for RX transfers
+ * @tx_sgt: sg table for TX transfers
+ * @zeropage: dummy page used as RX buffer when only TX buffer is passed in by
+ *            the client
+ */
+struct ep93xx_spi {
+	struct clk			*clk;
+	void __iomem			*mmio;
+	unsigned long			sspdr_phys;
+	size_t				tx;
+	size_t				rx;
+	size_t				fifo_level;
+	struct dma_chan			*dma_rx;
+	struct dma_chan			*dma_tx;
+	struct ep93xx_dma_data		dma_rx_data;
+	struct ep93xx_dma_data		dma_tx_data;
+	struct sg_table			rx_sgt;
+	struct sg_table			tx_sgt;
+	void				*zeropage;
+};
+
+/* converts bits per word to CR0.DSS value */
+#define bits_per_word_to_dss(bpw)	((bpw) - 1)
+
+/**
+ * ep93xx_spi_calc_divisors() - calculates SPI clock divisors
+ * @master: SPI master
+ * @rate: desired SPI output clock rate
+ * @div_cpsr: pointer to return the cpsr (pre-scaler) divider
+ * @div_scr: pointer to return the scr divider
+ */
+static int ep93xx_spi_calc_divisors(struct spi_master *master,
+				    u32 rate, u8 *div_cpsr, u8 *div_scr)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	unsigned long spi_clk_rate = clk_get_rate(espi->clk);
+	int cpsr, scr;
+
+	/*
+	 * Make sure that max value is between values supported by the
+	 * controller.
+	 */
+	rate = clamp(rate, master->min_speed_hz, master->max_speed_hz);
+
+	/*
+	 * Calculate divisors so that we can get speed according the
+	 * following formula:
+	 *	rate = spi_clock_rate / (cpsr * (1 + scr))
+	 *
+	 * cpsr must be even number and starts from 2, scr can be any number
+	 * between 0 and 255.
+	 */
+	for (cpsr = 2; cpsr <= 254; cpsr += 2) {
+		for (scr = 0; scr <= 255; scr++) {
+			if ((spi_clk_rate / (cpsr * (scr + 1))) <= rate) {
+				*div_scr = (u8)scr;
+				*div_cpsr = (u8)cpsr;
+				return 0;
+			}
+		}
+	}
+
+	return -EINVAL;
+}
+
+static int ep93xx_spi_chip_setup(struct spi_master *master,
+				 struct spi_device *spi,
+				 struct spi_transfer *xfer)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	u8 dss = bits_per_word_to_dss(xfer->bits_per_word);
+	u8 div_cpsr = 0;
+	u8 div_scr = 0;
+	u16 cr0;
+	int err;
+
+	err = ep93xx_spi_calc_divisors(master, xfer->speed_hz,
+				       &div_cpsr, &div_scr);
+	if (err)
+		return err;
+
+	cr0 = div_scr << SSPCR0_SCR_SHIFT;
+	cr0 |= (spi->mode & (SPI_CPHA | SPI_CPOL)) << SSPCR0_MODE_SHIFT;
+	cr0 |= dss;
+
+	dev_dbg(&master->dev, "setup: mode %d, cpsr %d, scr %d, dss %d\n",
+		spi->mode, div_cpsr, div_scr, dss);
+	dev_dbg(&master->dev, "setup: cr0 %#x\n", cr0);
+
+	writel(div_cpsr, espi->mmio + SSPCPSR);
+	writel(cr0, espi->mmio + SSPCR0);
+
+	return 0;
+}
+
+static void ep93xx_do_write(struct spi_master *master)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	struct spi_transfer *xfer = master->cur_msg->state;
+	u32 val = 0;
+
+	if (xfer->bits_per_word > 8) {
+		if (xfer->tx_buf)
+			val = ((u16 *)xfer->tx_buf)[espi->tx];
+		espi->tx += 2;
+	} else {
+		if (xfer->tx_buf)
+			val = ((u8 *)xfer->tx_buf)[espi->tx];
+		espi->tx += 1;
+	}
+	writel(val, espi->mmio + SSPDR);
+}
+
+static void ep93xx_do_read(struct spi_master *master)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	struct spi_transfer *xfer = master->cur_msg->state;
+	u32 val;
+
+	val = readl(espi->mmio + SSPDR);
+	if (xfer->bits_per_word > 8) {
+		if (xfer->rx_buf)
+			((u16 *)xfer->rx_buf)[espi->rx] = val;
+		espi->rx += 2;
+	} else {
+		if (xfer->rx_buf)
+			((u8 *)xfer->rx_buf)[espi->rx] = val;
+		espi->rx += 1;
+	}
+}
+
+/**
+ * ep93xx_spi_read_write() - perform next RX/TX transfer
+ * @espi: ep93xx SPI controller struct
+ *
+ * This function transfers next bytes (or half-words) to/from RX/TX FIFOs. If
+ * called several times, the whole transfer will be completed. Returns
+ * %-EINPROGRESS when current transfer was not yet completed otherwise %0.
+ *
+ * When this function is finished, RX FIFO should be empty and TX FIFO should be
+ * full.
+ */
+static int ep93xx_spi_read_write(struct spi_master *master)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	struct spi_transfer *xfer = master->cur_msg->state;
+
+	/* read as long as RX FIFO has frames in it */
+	while ((readl(espi->mmio + SSPSR) & SSPSR_RNE)) {
+		ep93xx_do_read(master);
+		espi->fifo_level--;
+	}
+
+	/* write as long as TX FIFO has room */
+	while (espi->fifo_level < SPI_FIFO_SIZE && espi->tx < xfer->len) {
+		ep93xx_do_write(master);
+		espi->fifo_level++;
+	}
+
+	if (espi->rx == xfer->len)
+		return 0;
+
+	return -EINPROGRESS;
+}
+
+static enum dma_transfer_direction
+ep93xx_dma_data_to_trans_dir(enum dma_data_direction dir)
+{
+	switch (dir) {
+	case DMA_TO_DEVICE:
+		return DMA_MEM_TO_DEV;
+	case DMA_FROM_DEVICE:
+		return DMA_DEV_TO_MEM;
+	default:
+		return DMA_TRANS_NONE;
+	}
+}
+
+/**
+ * ep93xx_spi_dma_prepare() - prepares a DMA transfer
+ * @master: SPI master
+ * @dir: DMA transfer direction
+ *
+ * Function configures the DMA, maps the buffer and prepares the DMA
+ * descriptor. Returns a valid DMA descriptor in case of success and ERR_PTR
+ * in case of failure.
+ */
+static struct dma_async_tx_descriptor *
+ep93xx_spi_dma_prepare(struct spi_master *master,
+		       enum dma_data_direction dir)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	struct spi_transfer *xfer = master->cur_msg->state;
+	struct dma_async_tx_descriptor *txd;
+	enum dma_slave_buswidth buswidth;
+	struct dma_slave_config conf;
+	struct scatterlist *sg;
+	struct sg_table *sgt;
+	struct dma_chan *chan;
+	const void *buf, *pbuf;
+	size_t len = xfer->len;
+	int i, ret, nents;
+
+	if (xfer->bits_per_word > 8)
+		buswidth = DMA_SLAVE_BUSWIDTH_2_BYTES;
+	else
+		buswidth = DMA_SLAVE_BUSWIDTH_1_BYTE;
+
+	memset(&conf, 0, sizeof(conf));
+	conf.direction = ep93xx_dma_data_to_trans_dir(dir);
+
+	if (dir == DMA_FROM_DEVICE) {
+		chan = espi->dma_rx;
+		buf = xfer->rx_buf;
+		sgt = &espi->rx_sgt;
+
+		conf.src_addr = espi->sspdr_phys;
+		conf.src_addr_width = buswidth;
+	} else {
+		chan = espi->dma_tx;
+		buf = xfer->tx_buf;
+		sgt = &espi->tx_sgt;
+
+		conf.dst_addr = espi->sspdr_phys;
+		conf.dst_addr_width = buswidth;
+	}
+
+	ret = dmaengine_slave_config(chan, &conf);
+	if (ret)
+		return ERR_PTR(ret);
+
+	/*
+	 * We need to split the transfer into PAGE_SIZE'd chunks. This is
+	 * because we are using @espi->zeropage to provide a zero RX buffer
+	 * for the TX transfers and we have only allocated one page for that.
+	 *
+	 * For performance reasons we allocate a new sg_table only when
+	 * needed. Otherwise we will re-use the current one. Eventually the
+	 * last sg_table is released in ep93xx_spi_release_dma().
+	 */
+
+	nents = DIV_ROUND_UP(len, PAGE_SIZE);
+	if (nents != sgt->nents) {
+		sg_free_table(sgt);
+
+		ret = sg_alloc_table(sgt, nents, GFP_KERNEL);
+		if (ret)
+			return ERR_PTR(ret);
+	}
+
+	pbuf = buf;
+	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
+		size_t bytes = min_t(size_t, len, PAGE_SIZE);
+
+		if (buf) {
+			sg_set_page(sg, virt_to_page(pbuf), bytes,
+				    offset_in_page(pbuf));
+		} else {
+			sg_set_page(sg, virt_to_page(espi->zeropage),
+				    bytes, 0);
+		}
+
+		pbuf += bytes;
+		len -= bytes;
+	}
+
+	if (WARN_ON(len)) {
+		dev_warn(&master->dev, "len = %zu expected 0!\n", len);
+		return ERR_PTR(-EINVAL);
+	}
+
+	nents = dma_map_sg(chan->device->dev, sgt->sgl, sgt->nents, dir);
+	if (!nents)
+		return ERR_PTR(-ENOMEM);
+
+	txd = dmaengine_prep_slave_sg(chan, sgt->sgl, nents, conf.direction,
+				      DMA_CTRL_ACK);
+	if (!txd) {
+		dma_unmap_sg(chan->device->dev, sgt->sgl, sgt->nents, dir);
+		return ERR_PTR(-ENOMEM);
+	}
+	return txd;
+}
+
+/**
+ * ep93xx_spi_dma_finish() - finishes with a DMA transfer
+ * @master: SPI master
+ * @dir: DMA transfer direction
+ *
+ * Function finishes with the DMA transfer. After this, the DMA buffer is
+ * unmapped.
+ */
+static void ep93xx_spi_dma_finish(struct spi_master *master,
+				  enum dma_data_direction dir)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	struct dma_chan *chan;
+	struct sg_table *sgt;
+
+	if (dir == DMA_FROM_DEVICE) {
+		chan = espi->dma_rx;
+		sgt = &espi->rx_sgt;
+	} else {
+		chan = espi->dma_tx;
+		sgt = &espi->tx_sgt;
+	}
+
+	dma_unmap_sg(chan->device->dev, sgt->sgl, sgt->nents, dir);
+}
+
+static void ep93xx_spi_dma_callback(void *callback_param)
+{
+	struct spi_master *master = callback_param;
+
+	ep93xx_spi_dma_finish(master, DMA_TO_DEVICE);
+	ep93xx_spi_dma_finish(master, DMA_FROM_DEVICE);
+
+	spi_finalize_current_transfer(master);
+}
+
+static int ep93xx_spi_dma_transfer(struct spi_master *master)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	struct dma_async_tx_descriptor *rxd, *txd;
+
+	rxd = ep93xx_spi_dma_prepare(master, DMA_FROM_DEVICE);
+	if (IS_ERR(rxd)) {
+		dev_err(&master->dev, "DMA RX failed: %ld\n", PTR_ERR(rxd));
+		return PTR_ERR(rxd);
+	}
+
+	txd = ep93xx_spi_dma_prepare(master, DMA_TO_DEVICE);
+	if (IS_ERR(txd)) {
+		ep93xx_spi_dma_finish(master, DMA_FROM_DEVICE);
+		dev_err(&master->dev, "DMA TX failed: %ld\n", PTR_ERR(txd));
+		return PTR_ERR(txd);
+	}
+
+	/* We are ready when RX is done */
+	rxd->callback = ep93xx_spi_dma_callback;
+	rxd->callback_param = master;
+
+	/* Now submit both descriptors and start DMA */
+	dmaengine_submit(rxd);
+	dmaengine_submit(txd);
+
+	dma_async_issue_pending(espi->dma_rx);
+	dma_async_issue_pending(espi->dma_tx);
+
+	/* signal that we need to wait for completion */
+	return 1;
+}
+
+static irqreturn_t ep93xx_spi_interrupt(int irq, void *dev_id)
+{
+	struct spi_master *master = dev_id;
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	u32 val;
+
+	/*
+	 * If we got ROR (receive overrun) interrupt we know that something is
+	 * wrong. Just abort the message.
+	 */
+	if (readl(espi->mmio + SSPIIR) & SSPIIR_RORIS) {
+		/* clear the overrun interrupt */
+		writel(0, espi->mmio + SSPICR);
+		dev_warn(&master->dev,
+			 "receive overrun, aborting the message\n");
+		master->cur_msg->status = -EIO;
+	} else {
+		/*
+		 * Interrupt is either RX (RIS) or TX (TIS). For both cases we
+		 * simply execute next data transfer.
+		 */
+		if (ep93xx_spi_read_write(master)) {
+			/*
+			 * In normal case, there still is some processing left
+			 * for current transfer. Let's wait for the next
+			 * interrupt then.
+			 */
+			return IRQ_HANDLED;
+		}
+	}
+
+	/*
+	 * Current transfer is finished, either with error or with success. In
+	 * any case we disable interrupts and notify the worker to handle
+	 * any post-processing of the message.
+	 */
+	val = readl(espi->mmio + SSPCR1);
+	val &= ~(SSPCR1_RORIE | SSPCR1_TIE | SSPCR1_RIE);
+	writel(val, espi->mmio + SSPCR1);
+
+	spi_finalize_current_transfer(master);
+
+	return IRQ_HANDLED;
+}
+
+static int ep93xx_spi_transfer_one(struct spi_master *master,
+				   struct spi_device *spi,
+				   struct spi_transfer *xfer)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	u32 val;
+	int ret;
+
+	ret = ep93xx_spi_chip_setup(master, spi, xfer);
+	if (ret) {
+		dev_err(&master->dev, "failed to setup chip for transfer\n");
+		return ret;
+	}
+
+	master->cur_msg->state = xfer;
+	espi->rx = 0;
+	espi->tx = 0;
+
+	/*
+	 * There is no point of setting up DMA for the transfers which will
+	 * fit into the FIFO and can be transferred with a single interrupt.
+	 * So in these cases we will be using PIO and don't bother for DMA.
+	 */
+	if (espi->dma_rx && xfer->len > SPI_FIFO_SIZE)
+		return ep93xx_spi_dma_transfer(master);
+
+	/* Using PIO so prime the TX FIFO and enable interrupts */
+	ep93xx_spi_read_write(master);
+
+	val = readl(espi->mmio + SSPCR1);
+	val |= (SSPCR1_RORIE | SSPCR1_TIE | SSPCR1_RIE);
+	writel(val, espi->mmio + SSPCR1);
+
+	/* signal that we need to wait for completion */
+	return 1;
+}
+
+static int ep93xx_spi_prepare_message(struct spi_master *master,
+				      struct spi_message *msg)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	unsigned long timeout;
+
+	/*
+	 * Just to be sure: flush any data from RX FIFO.
+	 */
+	timeout = jiffies + msecs_to_jiffies(SPI_TIMEOUT);
+	while (readl(espi->mmio + SSPSR) & SSPSR_RNE) {
+		if (time_after(jiffies, timeout)) {
+			dev_warn(&master->dev,
+				 "timeout while flushing RX FIFO\n");
+			return -ETIMEDOUT;
+		}
+		readl(espi->mmio + SSPDR);
+	}
+
+	/*
+	 * We explicitly handle FIFO level. This way we don't have to check TX
+	 * FIFO status using %SSPSR_TNF bit which may cause RX FIFO overruns.
+	 */
+	espi->fifo_level = 0;
+
+	return 0;
+}
+
+static int ep93xx_spi_prepare_hardware(struct spi_master *master)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	u32 val;
+	int ret;
+
+	ret = clk_enable(espi->clk);
+	if (ret)
+		return ret;
+
+	val = readl(espi->mmio + SSPCR1);
+	val |= SSPCR1_SSE;
+	writel(val, espi->mmio + SSPCR1);
+
+	return 0;
+}
+
+static int ep93xx_spi_unprepare_hardware(struct spi_master *master)
+{
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+	u32 val;
+
+	val = readl(espi->mmio + SSPCR1);
+	val &= ~SSPCR1_SSE;
+	writel(val, espi->mmio + SSPCR1);
+
+	clk_disable(espi->clk);
+
+	return 0;
+}
+
+static bool ep93xx_spi_dma_filter(struct dma_chan *chan, void *filter_param)
+{
+	if (ep93xx_dma_chan_is_m2p(chan))
+		return false;
+
+	chan->private = filter_param;
+	return true;
+}
+
+static int ep93xx_spi_setup_dma(struct ep93xx_spi *espi)
+{
+	dma_cap_mask_t mask;
+	int ret;
+
+	espi->zeropage = (void *)get_zeroed_page(GFP_KERNEL);
+	if (!espi->zeropage)
+		return -ENOMEM;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	espi->dma_rx_data.port = EP93XX_DMA_SSP;
+	espi->dma_rx_data.direction = DMA_DEV_TO_MEM;
+	espi->dma_rx_data.name = "ep93xx-spi-rx";
+
+	espi->dma_rx = dma_request_channel(mask, ep93xx_spi_dma_filter,
+					   &espi->dma_rx_data);
+	if (!espi->dma_rx) {
+		ret = -ENODEV;
+		goto fail_free_page;
+	}
+
+	espi->dma_tx_data.port = EP93XX_DMA_SSP;
+	espi->dma_tx_data.direction = DMA_MEM_TO_DEV;
+	espi->dma_tx_data.name = "ep93xx-spi-tx";
+
+	espi->dma_tx = dma_request_channel(mask, ep93xx_spi_dma_filter,
+					   &espi->dma_tx_data);
+	if (!espi->dma_tx) {
+		ret = -ENODEV;
+		goto fail_release_rx;
+	}
+
+	return 0;
+
+fail_release_rx:
+	dma_release_channel(espi->dma_rx);
+	espi->dma_rx = NULL;
+fail_free_page:
+	free_page((unsigned long)espi->zeropage);
+
+	return ret;
+}
+
+static void ep93xx_spi_release_dma(struct ep93xx_spi *espi)
+{
+	if (espi->dma_rx) {
+		dma_release_channel(espi->dma_rx);
+		sg_free_table(&espi->rx_sgt);
+	}
+	if (espi->dma_tx) {
+		dma_release_channel(espi->dma_tx);
+		sg_free_table(&espi->tx_sgt);
+	}
+
+	if (espi->zeropage)
+		free_page((unsigned long)espi->zeropage);
+}
+
+static int ep93xx_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct ep93xx_spi_info *info;
+	struct ep93xx_spi *espi;
+	struct resource *res;
+	int irq;
+	int error;
+	int i;
+
+	info = dev_get_platdata(&pdev->dev);
+	if (!info) {
+		dev_err(&pdev->dev, "missing platform data\n");
+		return -EINVAL;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "failed to get irq resources\n");
+		return -EBUSY;
+	}
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res) {
+		dev_err(&pdev->dev, "unable to get iomem resource\n");
+		return -ENODEV;
+	}
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*espi));
+	if (!master)
+		return -ENOMEM;
+
+	master->prepare_transfer_hardware = ep93xx_spi_prepare_hardware;
+	master->unprepare_transfer_hardware = ep93xx_spi_unprepare_hardware;
+	master->prepare_message = ep93xx_spi_prepare_message;
+	master->transfer_one = ep93xx_spi_transfer_one;
+	master->bus_num = pdev->id;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 16);
+
+	master->num_chipselect = info->num_chipselect;
+	master->cs_gpios = devm_kcalloc(&master->dev,
+					master->num_chipselect, sizeof(int),
+					GFP_KERNEL);
+	if (!master->cs_gpios) {
+		error = -ENOMEM;
+		goto fail_release_master;
+	}
+
+	for (i = 0; i < master->num_chipselect; i++) {
+		master->cs_gpios[i] = info->chipselect[i];
+
+		if (!gpio_is_valid(master->cs_gpios[i]))
+			continue;
+
+		error = devm_gpio_request_one(&pdev->dev, master->cs_gpios[i],
+					      GPIOF_OUT_INIT_HIGH,
+					      "ep93xx-spi");
+		if (error) {
+			dev_err(&pdev->dev, "could not request cs gpio %d\n",
+				master->cs_gpios[i]);
+			goto fail_release_master;
+		}
+	}
+
+	platform_set_drvdata(pdev, master);
+
+	espi = spi_master_get_devdata(master);
+
+	espi->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(espi->clk)) {
+		dev_err(&pdev->dev, "unable to get spi clock\n");
+		error = PTR_ERR(espi->clk);
+		goto fail_release_master;
+	}
+
+	/*
+	 * Calculate maximum and minimum supported clock rates
+	 * for the controller.
+	 */
+	master->max_speed_hz = clk_get_rate(espi->clk) / 2;
+	master->min_speed_hz = clk_get_rate(espi->clk) / (254 * 256);
+
+	espi->sspdr_phys = res->start + SSPDR;
+
+	espi->mmio = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(espi->mmio)) {
+		error = PTR_ERR(espi->mmio);
+		goto fail_release_master;
+	}
+
+	error = devm_request_irq(&pdev->dev, irq, ep93xx_spi_interrupt,
+				0, "ep93xx-spi", master);
+	if (error) {
+		dev_err(&pdev->dev, "failed to request irq\n");
+		goto fail_release_master;
+	}
+
+	if (info->use_dma && ep93xx_spi_setup_dma(espi))
+		dev_warn(&pdev->dev, "DMA setup failed. Falling back to PIO\n");
+
+	/* make sure that the hardware is disabled */
+	writel(0, espi->mmio + SSPCR1);
+
+	error = devm_spi_register_master(&pdev->dev, master);
+	if (error) {
+		dev_err(&pdev->dev, "failed to register SPI master\n");
+		goto fail_free_dma;
+	}
+
+	dev_info(&pdev->dev, "EP93xx SPI Controller at 0x%08lx irq %d\n",
+		 (unsigned long)res->start, irq);
+
+	return 0;
+
+fail_free_dma:
+	ep93xx_spi_release_dma(espi);
+fail_release_master:
+	spi_master_put(master);
+
+	return error;
+}
+
+static int ep93xx_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct ep93xx_spi *espi = spi_master_get_devdata(master);
+
+	ep93xx_spi_release_dma(espi);
+
+	return 0;
+}
+
+static struct platform_driver ep93xx_spi_driver = {
+	.driver		= {
+		.name	= "ep93xx-spi",
+	},
+	.probe		= ep93xx_spi_probe,
+	.remove		= ep93xx_spi_remove,
+};
+module_platform_driver(ep93xx_spi_driver);
+
+MODULE_DESCRIPTION("EP93xx SPI Controller driver");
+MODULE_AUTHOR("Mika Westerberg <mika.westerberg@iki.fi>");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:ep93xx-spi");
diff --git a/drivers/spi/spi-falcon.c b/drivers/spi/spi-falcon.c
new file mode 100644
index 0000000..f8638e8
--- /dev/null
+++ b/drivers/spi/spi-falcon.c
@@ -0,0 +1,434 @@
+/*
+ *  This program is free software; you can redistribute it and/or modify it
+ *  under the terms of the GNU General Public License version 2 as published
+ *  by the Free Software Foundation.
+ *
+ *  Copyright (C) 2012 Thomas Langer <thomas.langer@lantiq.com>
+ */
+
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/delay.h>
+#include <linux/of.h>
+#include <linux/of_platform.h>
+
+#include <lantiq_soc.h>
+
+#define DRV_NAME		"sflash-falcon"
+
+#define FALCON_SPI_XFER_BEGIN	(1 << 0)
+#define FALCON_SPI_XFER_END	(1 << 1)
+
+/* Bus Read Configuration Register0 */
+#define BUSRCON0		0x00000010
+/* Bus Write Configuration Register0 */
+#define BUSWCON0		0x00000018
+/* Serial Flash Configuration Register */
+#define SFCON			0x00000080
+/* Serial Flash Time Register */
+#define SFTIME			0x00000084
+/* Serial Flash Status Register */
+#define SFSTAT			0x00000088
+/* Serial Flash Command Register */
+#define SFCMD			0x0000008C
+/* Serial Flash Address Register */
+#define SFADDR			0x00000090
+/* Serial Flash Data Register */
+#define SFDATA			0x00000094
+/* Serial Flash I/O Control Register */
+#define SFIO			0x00000098
+/* EBU Clock Control Register */
+#define EBUCC			0x000000C4
+
+/* Dummy Phase Length */
+#define SFCMD_DUMLEN_OFFSET	16
+#define SFCMD_DUMLEN_MASK	0x000F0000
+/* Chip Select */
+#define SFCMD_CS_OFFSET		24
+#define SFCMD_CS_MASK		0x07000000
+/* field offset */
+#define SFCMD_ALEN_OFFSET	20
+#define SFCMD_ALEN_MASK		0x00700000
+/* SCK Rise-edge Position */
+#define SFTIME_SCKR_POS_OFFSET	8
+#define SFTIME_SCKR_POS_MASK	0x00000F00
+/* SCK Period */
+#define SFTIME_SCK_PER_OFFSET	0
+#define SFTIME_SCK_PER_MASK	0x0000000F
+/* SCK Fall-edge Position */
+#define SFTIME_SCKF_POS_OFFSET	12
+#define SFTIME_SCKF_POS_MASK	0x0000F000
+/* Device Size */
+#define SFCON_DEV_SIZE_A23_0	0x03000000
+#define SFCON_DEV_SIZE_MASK	0x0F000000
+/* Read Data Position */
+#define SFTIME_RD_POS_MASK	0x000F0000
+/* Data Output */
+#define SFIO_UNUSED_WD_MASK	0x0000000F
+/* Command Opcode mask */
+#define SFCMD_OPC_MASK		0x000000FF
+/* dlen bytes of data to write */
+#define SFCMD_DIR_WRITE		0x00000100
+/* Data Length offset */
+#define SFCMD_DLEN_OFFSET	9
+/* Command Error */
+#define SFSTAT_CMD_ERR		0x20000000
+/* Access Command Pending */
+#define SFSTAT_CMD_PEND		0x00400000
+/* Frequency set to 100MHz. */
+#define EBUCC_EBUDIV_SELF100	0x00000001
+/* Serial Flash */
+#define BUSRCON0_AGEN_SERIAL_FLASH	0xF0000000
+/* 8-bit multiplexed */
+#define BUSRCON0_PORTW_8_BIT_MUX	0x00000000
+/* Serial Flash */
+#define BUSWCON0_AGEN_SERIAL_FLASH	0xF0000000
+/* Chip Select after opcode */
+#define SFCMD_KEEP_CS_KEEP_SELECTED	0x00008000
+
+#define CLOCK_100M	100000000
+#define CLOCK_50M	50000000
+
+struct falcon_sflash {
+	u32 sfcmd; /* for caching of opcode, direction, ... */
+	struct spi_master *master;
+};
+
+int falcon_sflash_xfer(struct spi_device *spi, struct spi_transfer *t,
+		unsigned long flags)
+{
+	struct device *dev = &spi->dev;
+	struct falcon_sflash *priv = spi_master_get_devdata(spi->master);
+	const u8 *txp = t->tx_buf;
+	u8 *rxp = t->rx_buf;
+	unsigned int bytelen = ((8 * t->len + 7) / 8);
+	unsigned int len, alen, dumlen;
+	u32 val;
+	enum {
+		state_init,
+		state_command_prepare,
+		state_write,
+		state_read,
+		state_disable_cs,
+		state_end
+	} state = state_init;
+
+	do {
+		switch (state) {
+		case state_init: /* detect phase of upper layer sequence */
+		{
+			/* initial write ? */
+			if (flags & FALCON_SPI_XFER_BEGIN) {
+				if (!txp) {
+					dev_err(dev,
+						"BEGIN without tx data!\n");
+					return -ENODATA;
+				}
+				/*
+				 * Prepare the parts of the sfcmd register,
+				 * which should not change during a sequence!
+				 * Only exception are the length fields,
+				 * especially alen and dumlen.
+				 */
+
+				priv->sfcmd = ((spi->chip_select
+						<< SFCMD_CS_OFFSET)
+					       & SFCMD_CS_MASK);
+				priv->sfcmd |= SFCMD_KEEP_CS_KEEP_SELECTED;
+				priv->sfcmd |= *txp;
+				txp++;
+				bytelen--;
+				if (bytelen) {
+					/*
+					 * more data:
+					 * maybe address and/or dummy
+					 */
+					state = state_command_prepare;
+					break;
+				} else {
+					dev_dbg(dev, "write cmd %02X\n",
+						priv->sfcmd & SFCMD_OPC_MASK);
+				}
+			}
+			/* continued write ? */
+			if (txp && bytelen) {
+				state = state_write;
+				break;
+			}
+			/* read data? */
+			if (rxp && bytelen) {
+				state = state_read;
+				break;
+			}
+			/* end of sequence? */
+			if (flags & FALCON_SPI_XFER_END)
+				state = state_disable_cs;
+			else
+				state = state_end;
+			break;
+		}
+		/* collect tx data for address and dummy phase */
+		case state_command_prepare:
+		{
+			/* txp is valid, already checked */
+			val = 0;
+			alen = 0;
+			dumlen = 0;
+			while (bytelen > 0) {
+				if (alen < 3) {
+					val = (val << 8) | (*txp++);
+					alen++;
+				} else if ((dumlen < 15) && (*txp == 0)) {
+					/*
+					 * assume dummy bytes are set to 0
+					 * from upper layer
+					 */
+					dumlen++;
+					txp++;
+				} else {
+					break;
+				}
+				bytelen--;
+			}
+			priv->sfcmd &= ~(SFCMD_ALEN_MASK | SFCMD_DUMLEN_MASK);
+			priv->sfcmd |= (alen << SFCMD_ALEN_OFFSET) |
+					 (dumlen << SFCMD_DUMLEN_OFFSET);
+			if (alen > 0)
+				ltq_ebu_w32(val, SFADDR);
+
+			dev_dbg(dev, "wr %02X, alen=%d (addr=%06X) dlen=%d\n",
+				priv->sfcmd & SFCMD_OPC_MASK,
+				alen, val, dumlen);
+
+			if (bytelen > 0) {
+				/* continue with write */
+				state = state_write;
+			} else if (flags & FALCON_SPI_XFER_END) {
+				/* end of sequence? */
+				state = state_disable_cs;
+			} else {
+				/*
+				 * go to end and expect another
+				 * call (read or write)
+				 */
+				state = state_end;
+			}
+			break;
+		}
+		case state_write:
+		{
+			/* txp still valid */
+			priv->sfcmd |= SFCMD_DIR_WRITE;
+			len = 0;
+			val = 0;
+			do {
+				if (bytelen--)
+					val |= (*txp++) << (8 * len++);
+				if ((flags & FALCON_SPI_XFER_END)
+				    && (bytelen == 0)) {
+					priv->sfcmd &=
+						~SFCMD_KEEP_CS_KEEP_SELECTED;
+				}
+				if ((len == 4) || (bytelen == 0)) {
+					ltq_ebu_w32(val, SFDATA);
+					ltq_ebu_w32(priv->sfcmd
+						| (len<<SFCMD_DLEN_OFFSET),
+						SFCMD);
+					len = 0;
+					val = 0;
+					priv->sfcmd &= ~(SFCMD_ALEN_MASK
+							 | SFCMD_DUMLEN_MASK);
+				}
+			} while (bytelen);
+			state = state_end;
+			break;
+		}
+		case state_read:
+		{
+			/* read data */
+			priv->sfcmd &= ~SFCMD_DIR_WRITE;
+			do {
+				if ((flags & FALCON_SPI_XFER_END)
+				    && (bytelen <= 4)) {
+					priv->sfcmd &=
+						~SFCMD_KEEP_CS_KEEP_SELECTED;
+				}
+				len = (bytelen > 4) ? 4 : bytelen;
+				bytelen -= len;
+				ltq_ebu_w32(priv->sfcmd
+					| (len << SFCMD_DLEN_OFFSET), SFCMD);
+				priv->sfcmd &= ~(SFCMD_ALEN_MASK
+						 | SFCMD_DUMLEN_MASK);
+				do {
+					val = ltq_ebu_r32(SFSTAT);
+					if (val & SFSTAT_CMD_ERR) {
+						/* reset error status */
+						dev_err(dev, "SFSTAT: CMD_ERR");
+						dev_err(dev, " (%x)\n", val);
+						ltq_ebu_w32(SFSTAT_CMD_ERR,
+							SFSTAT);
+						return -EBADE;
+					}
+				} while (val & SFSTAT_CMD_PEND);
+				val = ltq_ebu_r32(SFDATA);
+				do {
+					*rxp = (val & 0xFF);
+					rxp++;
+					val >>= 8;
+					len--;
+				} while (len);
+			} while (bytelen);
+			state = state_end;
+			break;
+		}
+		case state_disable_cs:
+		{
+			priv->sfcmd &= ~SFCMD_KEEP_CS_KEEP_SELECTED;
+			ltq_ebu_w32(priv->sfcmd | (0 << SFCMD_DLEN_OFFSET),
+				SFCMD);
+			val = ltq_ebu_r32(SFSTAT);
+			if (val & SFSTAT_CMD_ERR) {
+				/* reset error status */
+				dev_err(dev, "SFSTAT: CMD_ERR (%x)\n", val);
+				ltq_ebu_w32(SFSTAT_CMD_ERR, SFSTAT);
+				return -EBADE;
+			}
+			state = state_end;
+			break;
+		}
+		case state_end:
+			break;
+		}
+	} while (state != state_end);
+
+	return 0;
+}
+
+static int falcon_sflash_setup(struct spi_device *spi)
+{
+	unsigned int i;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ebu_lock, flags);
+
+	if (spi->max_speed_hz >= CLOCK_100M) {
+		/* set EBU clock to 100 MHz */
+		ltq_sys1_w32_mask(0, EBUCC_EBUDIV_SELF100, EBUCC);
+		i = 1; /* divider */
+	} else {
+		/* set EBU clock to 50 MHz */
+		ltq_sys1_w32_mask(EBUCC_EBUDIV_SELF100, 0, EBUCC);
+
+		/* search for suitable divider */
+		for (i = 1; i < 7; i++) {
+			if (CLOCK_50M / i <= spi->max_speed_hz)
+				break;
+		}
+	}
+
+	/* setup period of serial clock */
+	ltq_ebu_w32_mask(SFTIME_SCKF_POS_MASK
+		     | SFTIME_SCKR_POS_MASK
+		     | SFTIME_SCK_PER_MASK,
+		     (i << SFTIME_SCKR_POS_OFFSET)
+		     | (i << (SFTIME_SCK_PER_OFFSET + 1)),
+		     SFTIME);
+
+	/*
+	 * set some bits of unused_wd, to not trigger HOLD/WP
+	 * signals on non QUAD flashes
+	 */
+	ltq_ebu_w32((SFIO_UNUSED_WD_MASK & (0x8 | 0x4)), SFIO);
+
+	ltq_ebu_w32(BUSRCON0_AGEN_SERIAL_FLASH | BUSRCON0_PORTW_8_BIT_MUX,
+			BUSRCON0);
+	ltq_ebu_w32(BUSWCON0_AGEN_SERIAL_FLASH, BUSWCON0);
+	/* set address wrap around to maximum for 24-bit addresses */
+	ltq_ebu_w32_mask(SFCON_DEV_SIZE_MASK, SFCON_DEV_SIZE_A23_0, SFCON);
+
+	spin_unlock_irqrestore(&ebu_lock, flags);
+
+	return 0;
+}
+
+static int falcon_sflash_xfer_one(struct spi_master *master,
+					struct spi_message *m)
+{
+	struct falcon_sflash *priv = spi_master_get_devdata(master);
+	struct spi_transfer *t;
+	unsigned long spi_flags;
+	unsigned long flags;
+	int ret = 0;
+
+	priv->sfcmd = 0;
+	m->actual_length = 0;
+
+	spi_flags = FALCON_SPI_XFER_BEGIN;
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if (list_is_last(&t->transfer_list, &m->transfers))
+			spi_flags |= FALCON_SPI_XFER_END;
+
+		spin_lock_irqsave(&ebu_lock, flags);
+		ret = falcon_sflash_xfer(m->spi, t, spi_flags);
+		spin_unlock_irqrestore(&ebu_lock, flags);
+
+		if (ret)
+			break;
+
+		m->actual_length += t->len;
+
+		WARN_ON(t->delay_usecs || t->cs_change);
+		spi_flags = 0;
+	}
+
+	m->status = ret;
+	spi_finalize_current_message(master);
+
+	return 0;
+}
+
+static int falcon_sflash_probe(struct platform_device *pdev)
+{
+	struct falcon_sflash *priv;
+	struct spi_master *master;
+	int ret;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*priv));
+	if (!master)
+		return -ENOMEM;
+
+	priv = spi_master_get_devdata(master);
+	priv->master = master;
+
+	master->mode_bits = SPI_MODE_3;
+	master->flags = SPI_MASTER_HALF_DUPLEX;
+	master->setup = falcon_sflash_setup;
+	master->transfer_one_message = falcon_sflash_xfer_one;
+	master->dev.of_node = pdev->dev.of_node;
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret)
+		spi_master_put(master);
+	return ret;
+}
+
+static const struct of_device_id falcon_sflash_match[] = {
+	{ .compatible = "lantiq,sflash-falcon" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, falcon_sflash_match);
+
+static struct platform_driver falcon_sflash_driver = {
+	.probe	= falcon_sflash_probe,
+	.driver = {
+		.name	= DRV_NAME,
+		.of_match_table = falcon_sflash_match,
+	}
+};
+
+module_platform_driver(falcon_sflash_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Lantiq Falcon SPI/SFLASH controller driver");
diff --git a/drivers/spi/spi-fsl-cpm.c b/drivers/spi/spi-fsl-cpm.c
new file mode 100644
index 0000000..8f7b26e
--- /dev/null
+++ b/drivers/spi/spi-fsl-cpm.c
@@ -0,0 +1,406 @@
+/*
+ * Freescale SPI controller driver cpm functions.
+ *
+ * Maintainer: Kumar Gala
+ *
+ * Copyright (C) 2006 Polycom, Inc.
+ * Copyright 2010 Freescale Semiconductor, Inc.
+ *
+ * CPM SPI and QE buffer descriptors mode support:
+ * Copyright (c) 2009  MontaVista Software, Inc.
+ * Author: Anton Vorontsov <avorontsov@ru.mvista.com>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+#include <asm/cpm.h>
+#include <soc/fsl/qe/qe.h>
+#include <linux/dma-mapping.h>
+#include <linux/fsl_devices.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/spi/spi.h>
+#include <linux/types.h>
+#include <linux/platform_device.h>
+
+#include "spi-fsl-cpm.h"
+#include "spi-fsl-lib.h"
+#include "spi-fsl-spi.h"
+
+/* CPM1 and CPM2 are mutually exclusive. */
+#ifdef CONFIG_CPM1
+#include <asm/cpm1.h>
+#define CPM_SPI_CMD mk_cr_cmd(CPM_CR_CH_SPI, 0)
+#else
+#include <asm/cpm2.h>
+#define CPM_SPI_CMD mk_cr_cmd(CPM_CR_SPI_PAGE, CPM_CR_SPI_SBLOCK, 0, 0)
+#endif
+
+#define	SPIE_TXB	0x00000200	/* Last char is written to tx fifo */
+#define	SPIE_RXB	0x00000100	/* Last char is written to rx buf */
+
+/* SPCOM register values */
+#define	SPCOM_STR	(1 << 23)	/* Start transmit */
+
+#define	SPI_PRAM_SIZE	0x100
+#define	SPI_MRBLR	((unsigned int)PAGE_SIZE)
+
+static void *fsl_dummy_rx;
+static DEFINE_MUTEX(fsl_dummy_rx_lock);
+static int fsl_dummy_rx_refcnt;
+
+void fsl_spi_cpm_reinit_txrx(struct mpc8xxx_spi *mspi)
+{
+	if (mspi->flags & SPI_QE) {
+		qe_issue_cmd(QE_INIT_TX_RX, mspi->subblock,
+			     QE_CR_PROTOCOL_UNSPECIFIED, 0);
+	} else {
+		if (mspi->flags & SPI_CPM1) {
+			out_be32(&mspi->pram->rstate, 0);
+			out_be16(&mspi->pram->rbptr,
+				 in_be16(&mspi->pram->rbase));
+			out_be32(&mspi->pram->tstate, 0);
+			out_be16(&mspi->pram->tbptr,
+				 in_be16(&mspi->pram->tbase));
+		} else {
+			cpm_command(CPM_SPI_CMD, CPM_CR_INIT_TRX);
+		}
+	}
+}
+EXPORT_SYMBOL_GPL(fsl_spi_cpm_reinit_txrx);
+
+static void fsl_spi_cpm_bufs_start(struct mpc8xxx_spi *mspi)
+{
+	struct cpm_buf_desc __iomem *tx_bd = mspi->tx_bd;
+	struct cpm_buf_desc __iomem *rx_bd = mspi->rx_bd;
+	unsigned int xfer_len = min(mspi->count, SPI_MRBLR);
+	unsigned int xfer_ofs;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
+
+	xfer_ofs = mspi->xfer_in_progress->len - mspi->count;
+
+	if (mspi->rx_dma == mspi->dma_dummy_rx)
+		out_be32(&rx_bd->cbd_bufaddr, mspi->rx_dma);
+	else
+		out_be32(&rx_bd->cbd_bufaddr, mspi->rx_dma + xfer_ofs);
+	out_be16(&rx_bd->cbd_datlen, 0);
+	out_be16(&rx_bd->cbd_sc, BD_SC_EMPTY | BD_SC_INTRPT | BD_SC_WRAP);
+
+	if (mspi->tx_dma == mspi->dma_dummy_tx)
+		out_be32(&tx_bd->cbd_bufaddr, mspi->tx_dma);
+	else
+		out_be32(&tx_bd->cbd_bufaddr, mspi->tx_dma + xfer_ofs);
+	out_be16(&tx_bd->cbd_datlen, xfer_len);
+	out_be16(&tx_bd->cbd_sc, BD_SC_READY | BD_SC_INTRPT | BD_SC_WRAP |
+				 BD_SC_LAST);
+
+	/* start transfer */
+	mpc8xxx_spi_write_reg(&reg_base->command, SPCOM_STR);
+}
+
+int fsl_spi_cpm_bufs(struct mpc8xxx_spi *mspi,
+		     struct spi_transfer *t, bool is_dma_mapped)
+{
+	struct device *dev = mspi->dev;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
+
+	if (is_dma_mapped) {
+		mspi->map_tx_dma = 0;
+		mspi->map_rx_dma = 0;
+	} else {
+		mspi->map_tx_dma = 1;
+		mspi->map_rx_dma = 1;
+	}
+
+	if (!t->tx_buf) {
+		mspi->tx_dma = mspi->dma_dummy_tx;
+		mspi->map_tx_dma = 0;
+	}
+
+	if (!t->rx_buf) {
+		mspi->rx_dma = mspi->dma_dummy_rx;
+		mspi->map_rx_dma = 0;
+	}
+
+	if (mspi->map_tx_dma) {
+		void *nonconst_tx = (void *)mspi->tx; /* shut up gcc */
+
+		mspi->tx_dma = dma_map_single(dev, nonconst_tx, t->len,
+					      DMA_TO_DEVICE);
+		if (dma_mapping_error(dev, mspi->tx_dma)) {
+			dev_err(dev, "unable to map tx dma\n");
+			return -ENOMEM;
+		}
+	} else if (t->tx_buf) {
+		mspi->tx_dma = t->tx_dma;
+	}
+
+	if (mspi->map_rx_dma) {
+		mspi->rx_dma = dma_map_single(dev, mspi->rx, t->len,
+					      DMA_FROM_DEVICE);
+		if (dma_mapping_error(dev, mspi->rx_dma)) {
+			dev_err(dev, "unable to map rx dma\n");
+			goto err_rx_dma;
+		}
+	} else if (t->rx_buf) {
+		mspi->rx_dma = t->rx_dma;
+	}
+
+	/* enable rx ints */
+	mpc8xxx_spi_write_reg(&reg_base->mask, SPIE_RXB);
+
+	mspi->xfer_in_progress = t;
+	mspi->count = t->len;
+
+	/* start CPM transfers */
+	fsl_spi_cpm_bufs_start(mspi);
+
+	return 0;
+
+err_rx_dma:
+	if (mspi->map_tx_dma)
+		dma_unmap_single(dev, mspi->tx_dma, t->len, DMA_TO_DEVICE);
+	return -ENOMEM;
+}
+EXPORT_SYMBOL_GPL(fsl_spi_cpm_bufs);
+
+void fsl_spi_cpm_bufs_complete(struct mpc8xxx_spi *mspi)
+{
+	struct device *dev = mspi->dev;
+	struct spi_transfer *t = mspi->xfer_in_progress;
+
+	if (mspi->map_tx_dma)
+		dma_unmap_single(dev, mspi->tx_dma, t->len, DMA_TO_DEVICE);
+	if (mspi->map_rx_dma)
+		dma_unmap_single(dev, mspi->rx_dma, t->len, DMA_FROM_DEVICE);
+	mspi->xfer_in_progress = NULL;
+}
+EXPORT_SYMBOL_GPL(fsl_spi_cpm_bufs_complete);
+
+void fsl_spi_cpm_irq(struct mpc8xxx_spi *mspi, u32 events)
+{
+	u16 len;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
+
+	dev_dbg(mspi->dev, "%s: bd datlen %d, count %d\n", __func__,
+		in_be16(&mspi->rx_bd->cbd_datlen), mspi->count);
+
+	len = in_be16(&mspi->rx_bd->cbd_datlen);
+	if (len > mspi->count) {
+		WARN_ON(1);
+		len = mspi->count;
+	}
+
+	/* Clear the events */
+	mpc8xxx_spi_write_reg(&reg_base->event, events);
+
+	mspi->count -= len;
+	if (mspi->count)
+		fsl_spi_cpm_bufs_start(mspi);
+	else
+		complete(&mspi->done);
+}
+EXPORT_SYMBOL_GPL(fsl_spi_cpm_irq);
+
+static void *fsl_spi_alloc_dummy_rx(void)
+{
+	mutex_lock(&fsl_dummy_rx_lock);
+
+	if (!fsl_dummy_rx)
+		fsl_dummy_rx = kmalloc(SPI_MRBLR, GFP_KERNEL);
+	if (fsl_dummy_rx)
+		fsl_dummy_rx_refcnt++;
+
+	mutex_unlock(&fsl_dummy_rx_lock);
+
+	return fsl_dummy_rx;
+}
+
+static void fsl_spi_free_dummy_rx(void)
+{
+	mutex_lock(&fsl_dummy_rx_lock);
+
+	switch (fsl_dummy_rx_refcnt) {
+	case 0:
+		WARN_ON(1);
+		break;
+	case 1:
+		kfree(fsl_dummy_rx);
+		fsl_dummy_rx = NULL;
+		/* fall through */
+	default:
+		fsl_dummy_rx_refcnt--;
+		break;
+	}
+
+	mutex_unlock(&fsl_dummy_rx_lock);
+}
+
+static unsigned long fsl_spi_cpm_get_pram(struct mpc8xxx_spi *mspi)
+{
+	struct device *dev = mspi->dev;
+	struct device_node *np = dev->of_node;
+	const u32 *iprop;
+	int size;
+	void __iomem *spi_base;
+	unsigned long pram_ofs = -ENOMEM;
+
+	/* Can't use of_address_to_resource(), QE muram isn't at 0. */
+	iprop = of_get_property(np, "reg", &size);
+
+	/* QE with a fixed pram location? */
+	if (mspi->flags & SPI_QE && iprop && size == sizeof(*iprop) * 4)
+		return cpm_muram_alloc_fixed(iprop[2], SPI_PRAM_SIZE);
+
+	/* QE but with a dynamic pram location? */
+	if (mspi->flags & SPI_QE) {
+		pram_ofs = cpm_muram_alloc(SPI_PRAM_SIZE, 64);
+		qe_issue_cmd(QE_ASSIGN_PAGE_TO_DEVICE, mspi->subblock,
+			     QE_CR_PROTOCOL_UNSPECIFIED, pram_ofs);
+		return pram_ofs;
+	}
+
+	spi_base = of_iomap(np, 1);
+	if (spi_base == NULL)
+		return -EINVAL;
+
+	if (mspi->flags & SPI_CPM2) {
+		pram_ofs = cpm_muram_alloc(SPI_PRAM_SIZE, 64);
+		out_be16(spi_base, pram_ofs);
+	}
+
+	iounmap(spi_base);
+	return pram_ofs;
+}
+
+int fsl_spi_cpm_init(struct mpc8xxx_spi *mspi)
+{
+	struct device *dev = mspi->dev;
+	struct device_node *np = dev->of_node;
+	const u32 *iprop;
+	int size;
+	unsigned long bds_ofs;
+
+	if (!(mspi->flags & SPI_CPM_MODE))
+		return 0;
+
+	if (!fsl_spi_alloc_dummy_rx())
+		return -ENOMEM;
+
+	if (mspi->flags & SPI_QE) {
+		iprop = of_get_property(np, "cell-index", &size);
+		if (iprop && size == sizeof(*iprop))
+			mspi->subblock = *iprop;
+
+		switch (mspi->subblock) {
+		default:
+			dev_warn(dev, "cell-index unspecified, assuming SPI1\n");
+			/* fall through */
+		case 0:
+			mspi->subblock = QE_CR_SUBBLOCK_SPI1;
+			break;
+		case 1:
+			mspi->subblock = QE_CR_SUBBLOCK_SPI2;
+			break;
+		}
+	}
+
+	if (mspi->flags & SPI_CPM1) {
+		struct resource *res;
+		void *pram;
+
+		res = platform_get_resource(to_platform_device(dev),
+					    IORESOURCE_MEM, 1);
+		pram = devm_ioremap_resource(dev, res);
+		if (IS_ERR(pram))
+			mspi->pram = NULL;
+		else
+			mspi->pram = pram;
+	} else {
+		unsigned long pram_ofs = fsl_spi_cpm_get_pram(mspi);
+
+		if (IS_ERR_VALUE(pram_ofs))
+			mspi->pram = NULL;
+		else
+			mspi->pram = cpm_muram_addr(pram_ofs);
+	}
+	if (mspi->pram == NULL) {
+		dev_err(dev, "can't allocate spi parameter ram\n");
+		goto err_pram;
+	}
+
+	bds_ofs = cpm_muram_alloc(sizeof(*mspi->tx_bd) +
+				  sizeof(*mspi->rx_bd), 8);
+	if (IS_ERR_VALUE(bds_ofs)) {
+		dev_err(dev, "can't allocate bds\n");
+		goto err_bds;
+	}
+
+	mspi->dma_dummy_tx = dma_map_single(dev, empty_zero_page, PAGE_SIZE,
+					    DMA_TO_DEVICE);
+	if (dma_mapping_error(dev, mspi->dma_dummy_tx)) {
+		dev_err(dev, "unable to map dummy tx buffer\n");
+		goto err_dummy_tx;
+	}
+
+	mspi->dma_dummy_rx = dma_map_single(dev, fsl_dummy_rx, SPI_MRBLR,
+					    DMA_FROM_DEVICE);
+	if (dma_mapping_error(dev, mspi->dma_dummy_rx)) {
+		dev_err(dev, "unable to map dummy rx buffer\n");
+		goto err_dummy_rx;
+	}
+
+	mspi->tx_bd = cpm_muram_addr(bds_ofs);
+	mspi->rx_bd = cpm_muram_addr(bds_ofs + sizeof(*mspi->tx_bd));
+
+	/* Initialize parameter ram. */
+	out_be16(&mspi->pram->tbase, cpm_muram_offset(mspi->tx_bd));
+	out_be16(&mspi->pram->rbase, cpm_muram_offset(mspi->rx_bd));
+	out_8(&mspi->pram->tfcr, CPMFCR_EB | CPMFCR_GBL);
+	out_8(&mspi->pram->rfcr, CPMFCR_EB | CPMFCR_GBL);
+	out_be16(&mspi->pram->mrblr, SPI_MRBLR);
+	out_be32(&mspi->pram->rstate, 0);
+	out_be32(&mspi->pram->rdp, 0);
+	out_be16(&mspi->pram->rbptr, 0);
+	out_be16(&mspi->pram->rbc, 0);
+	out_be32(&mspi->pram->rxtmp, 0);
+	out_be32(&mspi->pram->tstate, 0);
+	out_be32(&mspi->pram->tdp, 0);
+	out_be16(&mspi->pram->tbptr, 0);
+	out_be16(&mspi->pram->tbc, 0);
+	out_be32(&mspi->pram->txtmp, 0);
+
+	return 0;
+
+err_dummy_rx:
+	dma_unmap_single(dev, mspi->dma_dummy_tx, PAGE_SIZE, DMA_TO_DEVICE);
+err_dummy_tx:
+	cpm_muram_free(bds_ofs);
+err_bds:
+	if (!(mspi->flags & SPI_CPM1))
+		cpm_muram_free(cpm_muram_offset(mspi->pram));
+err_pram:
+	fsl_spi_free_dummy_rx();
+	return -ENOMEM;
+}
+EXPORT_SYMBOL_GPL(fsl_spi_cpm_init);
+
+void fsl_spi_cpm_free(struct mpc8xxx_spi *mspi)
+{
+	struct device *dev = mspi->dev;
+
+	if (!(mspi->flags & SPI_CPM_MODE))
+		return;
+
+	dma_unmap_single(dev, mspi->dma_dummy_rx, SPI_MRBLR, DMA_FROM_DEVICE);
+	dma_unmap_single(dev, mspi->dma_dummy_tx, PAGE_SIZE, DMA_TO_DEVICE);
+	cpm_muram_free(cpm_muram_offset(mspi->tx_bd));
+	cpm_muram_free(cpm_muram_offset(mspi->pram));
+	fsl_spi_free_dummy_rx();
+}
+EXPORT_SYMBOL_GPL(fsl_spi_cpm_free);
+
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-fsl-cpm.h b/drivers/spi/spi-fsl-cpm.h
new file mode 100644
index 0000000..c711158
--- /dev/null
+++ b/drivers/spi/spi-fsl-cpm.h
@@ -0,0 +1,43 @@
+/*
+ * Freescale SPI controller driver cpm functions.
+ *
+ * Maintainer: Kumar Gala
+ *
+ * Copyright (C) 2006 Polycom, Inc.
+ * Copyright 2010 Freescale Semiconductor, Inc.
+ *
+ * CPM SPI and QE buffer descriptors mode support:
+ * Copyright (c) 2009  MontaVista Software, Inc.
+ * Author: Anton Vorontsov <avorontsov@ru.mvista.com>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+
+#ifndef __SPI_FSL_CPM_H__
+#define __SPI_FSL_CPM_H__
+
+#include "spi-fsl-lib.h"
+
+#ifdef CONFIG_FSL_SOC
+extern void fsl_spi_cpm_reinit_txrx(struct mpc8xxx_spi *mspi);
+extern int fsl_spi_cpm_bufs(struct mpc8xxx_spi *mspi,
+			    struct spi_transfer *t, bool is_dma_mapped);
+extern void fsl_spi_cpm_bufs_complete(struct mpc8xxx_spi *mspi);
+extern void fsl_spi_cpm_irq(struct mpc8xxx_spi *mspi, u32 events);
+extern int fsl_spi_cpm_init(struct mpc8xxx_spi *mspi);
+extern void fsl_spi_cpm_free(struct mpc8xxx_spi *mspi);
+#else
+static inline void fsl_spi_cpm_reinit_txrx(struct mpc8xxx_spi *mspi) { }
+static inline int fsl_spi_cpm_bufs(struct mpc8xxx_spi *mspi,
+				   struct spi_transfer *t,
+				   bool is_dma_mapped) { return 0; }
+static inline void fsl_spi_cpm_bufs_complete(struct mpc8xxx_spi *mspi) { }
+static inline void fsl_spi_cpm_irq(struct mpc8xxx_spi *mspi, u32 events) { }
+static inline int fsl_spi_cpm_init(struct mpc8xxx_spi *mspi) { return 0; }
+static inline void fsl_spi_cpm_free(struct mpc8xxx_spi *mspi) { }
+#endif
+
+#endif /* __SPI_FSL_CPM_H__ */
diff --git a/drivers/spi/spi-fsl-dspi.c b/drivers/spi/spi-fsl-dspi.c
new file mode 100644
index 0000000..3082e72
--- /dev/null
+++ b/drivers/spi/spi-fsl-dspi.c
@@ -0,0 +1,1155 @@
+// SPDX-License-Identifier: GPL-2.0+
+//
+// Copyright 2013 Freescale Semiconductor, Inc.
+//
+// Freescale DSPI driver
+// This file contains a driver for the Freescale DSPI
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/err.h>
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/math64.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/regmap.h>
+#include <linux/sched.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi-fsl-dspi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/time.h>
+
+#define DRIVER_NAME "fsl-dspi"
+
+#ifdef CONFIG_M5441x
+#define DSPI_FIFO_SIZE			16
+#else
+#define DSPI_FIFO_SIZE			4
+#endif
+#define DSPI_DMA_BUFSIZE		(DSPI_FIFO_SIZE * 1024)
+
+#define SPI_MCR		0x00
+#define SPI_MCR_MASTER		(1 << 31)
+#define SPI_MCR_PCSIS		(0x3F << 16)
+#define SPI_MCR_CLR_TXF	(1 << 11)
+#define SPI_MCR_CLR_RXF	(1 << 10)
+#define SPI_MCR_XSPI		(1 << 3)
+
+#define SPI_TCR			0x08
+#define SPI_TCR_GET_TCNT(x)	(((x) & 0xffff0000) >> 16)
+
+#define SPI_CTAR(x)		(0x0c + (((x) & 0x3) * 4))
+#define SPI_CTAR_FMSZ(x)	(((x) & 0x0000000f) << 27)
+#define SPI_CTAR_CPOL(x)	((x) << 26)
+#define SPI_CTAR_CPHA(x)	((x) << 25)
+#define SPI_CTAR_LSBFE(x)	((x) << 24)
+#define SPI_CTAR_PCSSCK(x)	(((x) & 0x00000003) << 22)
+#define SPI_CTAR_PASC(x)	(((x) & 0x00000003) << 20)
+#define SPI_CTAR_PDT(x)	(((x) & 0x00000003) << 18)
+#define SPI_CTAR_PBR(x)	(((x) & 0x00000003) << 16)
+#define SPI_CTAR_CSSCK(x)	(((x) & 0x0000000f) << 12)
+#define SPI_CTAR_ASC(x)	(((x) & 0x0000000f) << 8)
+#define SPI_CTAR_DT(x)		(((x) & 0x0000000f) << 4)
+#define SPI_CTAR_BR(x)		((x) & 0x0000000f)
+#define SPI_CTAR_SCALE_BITS	0xf
+
+#define SPI_CTAR0_SLAVE	0x0c
+
+#define SPI_SR			0x2c
+#define SPI_SR_EOQF		0x10000000
+#define SPI_SR_TCFQF		0x80000000
+#define SPI_SR_CLEAR		0xdaad0000
+
+#define SPI_RSER_TFFFE		BIT(25)
+#define SPI_RSER_TFFFD		BIT(24)
+#define SPI_RSER_RFDFE		BIT(17)
+#define SPI_RSER_RFDFD		BIT(16)
+
+#define SPI_RSER		0x30
+#define SPI_RSER_EOQFE		0x10000000
+#define SPI_RSER_TCFQE		0x80000000
+
+#define SPI_PUSHR		0x34
+#define SPI_PUSHR_CMD_CONT	(1 << 15)
+#define SPI_PUSHR_CONT		(SPI_PUSHR_CMD_CONT << 16)
+#define SPI_PUSHR_CMD_CTAS(x)	(((x) & 0x0003) << 12)
+#define SPI_PUSHR_CTAS(x)	(SPI_PUSHR_CMD_CTAS(x) << 16)
+#define SPI_PUSHR_CMD_EOQ	(1 << 11)
+#define SPI_PUSHR_EOQ		(SPI_PUSHR_CMD_EOQ << 16)
+#define SPI_PUSHR_CMD_CTCNT	(1 << 10)
+#define SPI_PUSHR_CTCNT		(SPI_PUSHR_CMD_CTCNT << 16)
+#define SPI_PUSHR_CMD_PCS(x)	((1 << x) & 0x003f)
+#define SPI_PUSHR_PCS(x)	(SPI_PUSHR_CMD_PCS(x) << 16)
+#define SPI_PUSHR_TXDATA(x)	((x) & 0x0000ffff)
+
+#define SPI_PUSHR_SLAVE	0x34
+
+#define SPI_POPR		0x38
+#define SPI_POPR_RXDATA(x)	((x) & 0x0000ffff)
+
+#define SPI_TXFR0		0x3c
+#define SPI_TXFR1		0x40
+#define SPI_TXFR2		0x44
+#define SPI_TXFR3		0x48
+#define SPI_RXFR0		0x7c
+#define SPI_RXFR1		0x80
+#define SPI_RXFR2		0x84
+#define SPI_RXFR3		0x88
+
+#define SPI_CTARE(x)		(0x11c + (((x) & 0x3) * 4))
+#define SPI_CTARE_FMSZE(x)	(((x) & 0x1) << 16)
+#define SPI_CTARE_DTCP(x)	((x) & 0x7ff)
+
+#define SPI_SREX		0x13c
+
+#define SPI_FRAME_BITS(bits)	SPI_CTAR_FMSZ((bits) - 1)
+#define SPI_FRAME_BITS_MASK	SPI_CTAR_FMSZ(0xf)
+#define SPI_FRAME_BITS_16	SPI_CTAR_FMSZ(0xf)
+#define SPI_FRAME_BITS_8	SPI_CTAR_FMSZ(0x7)
+
+#define SPI_FRAME_EBITS(bits)	SPI_CTARE_FMSZE(((bits) - 1) >> 4)
+#define SPI_FRAME_EBITS_MASK	SPI_CTARE_FMSZE(1)
+
+/* Register offsets for regmap_pushr */
+#define PUSHR_CMD		0x0
+#define PUSHR_TX		0x2
+
+#define SPI_CS_INIT		0x01
+#define SPI_CS_ASSERT		0x02
+#define SPI_CS_DROP		0x04
+
+#define DMA_COMPLETION_TIMEOUT	msecs_to_jiffies(3000)
+
+struct chip_data {
+	u32 ctar_val;
+	u16 void_write_data;
+};
+
+enum dspi_trans_mode {
+	DSPI_EOQ_MODE = 0,
+	DSPI_TCFQ_MODE,
+	DSPI_DMA_MODE,
+};
+
+struct fsl_dspi_devtype_data {
+	enum dspi_trans_mode trans_mode;
+	u8 max_clock_factor;
+	bool xspi_mode;
+};
+
+static const struct fsl_dspi_devtype_data vf610_data = {
+	.trans_mode = DSPI_DMA_MODE,
+	.max_clock_factor = 2,
+};
+
+static const struct fsl_dspi_devtype_data ls1021a_v1_data = {
+	.trans_mode = DSPI_TCFQ_MODE,
+	.max_clock_factor = 8,
+	.xspi_mode = true,
+};
+
+static const struct fsl_dspi_devtype_data ls2085a_data = {
+	.trans_mode = DSPI_TCFQ_MODE,
+	.max_clock_factor = 8,
+};
+
+static const struct fsl_dspi_devtype_data coldfire_data = {
+	.trans_mode = DSPI_EOQ_MODE,
+	.max_clock_factor = 8,
+};
+
+struct fsl_dspi_dma {
+	/* Length of transfer in words of DSPI_FIFO_SIZE */
+	u32 curr_xfer_len;
+
+	u32 *tx_dma_buf;
+	struct dma_chan *chan_tx;
+	dma_addr_t tx_dma_phys;
+	struct completion cmd_tx_complete;
+	struct dma_async_tx_descriptor *tx_desc;
+
+	u32 *rx_dma_buf;
+	struct dma_chan *chan_rx;
+	dma_addr_t rx_dma_phys;
+	struct completion cmd_rx_complete;
+	struct dma_async_tx_descriptor *rx_desc;
+};
+
+struct fsl_dspi {
+	struct spi_master	*master;
+	struct platform_device	*pdev;
+
+	struct regmap		*regmap;
+	struct regmap		*regmap_pushr;
+	int			irq;
+	struct clk		*clk;
+
+	struct spi_transfer	*cur_transfer;
+	struct spi_message	*cur_msg;
+	struct chip_data	*cur_chip;
+	size_t			len;
+	const void		*tx;
+	void			*rx;
+	void			*rx_end;
+	u16			void_write_data;
+	u16			tx_cmd;
+	u8			bits_per_word;
+	u8			bytes_per_word;
+	const struct fsl_dspi_devtype_data *devtype_data;
+
+	wait_queue_head_t	waitq;
+	u32			waitflags;
+
+	struct fsl_dspi_dma	*dma;
+};
+
+static u32 dspi_pop_tx(struct fsl_dspi *dspi)
+{
+	u32 txdata = 0;
+
+	if (dspi->tx) {
+		if (dspi->bytes_per_word == 1)
+			txdata = *(u8 *)dspi->tx;
+		else if (dspi->bytes_per_word == 2)
+			txdata = *(u16 *)dspi->tx;
+		else  /* dspi->bytes_per_word == 4 */
+			txdata = *(u32 *)dspi->tx;
+		dspi->tx += dspi->bytes_per_word;
+	}
+	dspi->len -= dspi->bytes_per_word;
+	return txdata;
+}
+
+static u32 dspi_pop_tx_pushr(struct fsl_dspi *dspi)
+{
+	u16 cmd = dspi->tx_cmd, data = dspi_pop_tx(dspi);
+
+	if (dspi->len > 0)
+		cmd |= SPI_PUSHR_CMD_CONT;
+	return cmd << 16 | data;
+}
+
+static void dspi_push_rx(struct fsl_dspi *dspi, u32 rxdata)
+{
+	if (!dspi->rx)
+		return;
+
+	/* Mask of undefined bits */
+	rxdata &= (1 << dspi->bits_per_word) - 1;
+
+	if (dspi->bytes_per_word == 1)
+		*(u8 *)dspi->rx = rxdata;
+	else if (dspi->bytes_per_word == 2)
+		*(u16 *)dspi->rx = rxdata;
+	else /* dspi->bytes_per_word == 4 */
+		*(u32 *)dspi->rx = rxdata;
+	dspi->rx += dspi->bytes_per_word;
+}
+
+static void dspi_tx_dma_callback(void *arg)
+{
+	struct fsl_dspi *dspi = arg;
+	struct fsl_dspi_dma *dma = dspi->dma;
+
+	complete(&dma->cmd_tx_complete);
+}
+
+static void dspi_rx_dma_callback(void *arg)
+{
+	struct fsl_dspi *dspi = arg;
+	struct fsl_dspi_dma *dma = dspi->dma;
+	int i;
+
+	if (dspi->rx) {
+		for (i = 0; i < dma->curr_xfer_len; i++)
+			dspi_push_rx(dspi, dspi->dma->rx_dma_buf[i]);
+	}
+
+	complete(&dma->cmd_rx_complete);
+}
+
+static int dspi_next_xfer_dma_submit(struct fsl_dspi *dspi)
+{
+	struct fsl_dspi_dma *dma = dspi->dma;
+	struct device *dev = &dspi->pdev->dev;
+	int time_left;
+	int i;
+
+	for (i = 0; i < dma->curr_xfer_len; i++)
+		dspi->dma->tx_dma_buf[i] = dspi_pop_tx_pushr(dspi);
+
+	dma->tx_desc = dmaengine_prep_slave_single(dma->chan_tx,
+					dma->tx_dma_phys,
+					dma->curr_xfer_len *
+					DMA_SLAVE_BUSWIDTH_4_BYTES,
+					DMA_MEM_TO_DEV,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!dma->tx_desc) {
+		dev_err(dev, "Not able to get desc for DMA xfer\n");
+		return -EIO;
+	}
+
+	dma->tx_desc->callback = dspi_tx_dma_callback;
+	dma->tx_desc->callback_param = dspi;
+	if (dma_submit_error(dmaengine_submit(dma->tx_desc))) {
+		dev_err(dev, "DMA submit failed\n");
+		return -EINVAL;
+	}
+
+	dma->rx_desc = dmaengine_prep_slave_single(dma->chan_rx,
+					dma->rx_dma_phys,
+					dma->curr_xfer_len *
+					DMA_SLAVE_BUSWIDTH_4_BYTES,
+					DMA_DEV_TO_MEM,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!dma->rx_desc) {
+		dev_err(dev, "Not able to get desc for DMA xfer\n");
+		return -EIO;
+	}
+
+	dma->rx_desc->callback = dspi_rx_dma_callback;
+	dma->rx_desc->callback_param = dspi;
+	if (dma_submit_error(dmaengine_submit(dma->rx_desc))) {
+		dev_err(dev, "DMA submit failed\n");
+		return -EINVAL;
+	}
+
+	reinit_completion(&dspi->dma->cmd_rx_complete);
+	reinit_completion(&dspi->dma->cmd_tx_complete);
+
+	dma_async_issue_pending(dma->chan_rx);
+	dma_async_issue_pending(dma->chan_tx);
+
+	time_left = wait_for_completion_timeout(&dspi->dma->cmd_tx_complete,
+					DMA_COMPLETION_TIMEOUT);
+	if (time_left == 0) {
+		dev_err(dev, "DMA tx timeout\n");
+		dmaengine_terminate_all(dma->chan_tx);
+		dmaengine_terminate_all(dma->chan_rx);
+		return -ETIMEDOUT;
+	}
+
+	time_left = wait_for_completion_timeout(&dspi->dma->cmd_rx_complete,
+					DMA_COMPLETION_TIMEOUT);
+	if (time_left == 0) {
+		dev_err(dev, "DMA rx timeout\n");
+		dmaengine_terminate_all(dma->chan_tx);
+		dmaengine_terminate_all(dma->chan_rx);
+		return -ETIMEDOUT;
+	}
+
+	return 0;
+}
+
+static int dspi_dma_xfer(struct fsl_dspi *dspi)
+{
+	struct fsl_dspi_dma *dma = dspi->dma;
+	struct device *dev = &dspi->pdev->dev;
+	struct spi_message *message = dspi->cur_msg;
+	int curr_remaining_bytes;
+	int bytes_per_buffer;
+	int ret = 0;
+
+	curr_remaining_bytes = dspi->len;
+	bytes_per_buffer = DSPI_DMA_BUFSIZE / DSPI_FIFO_SIZE;
+	while (curr_remaining_bytes) {
+		/* Check if current transfer fits the DMA buffer */
+		dma->curr_xfer_len = curr_remaining_bytes
+			/ dspi->bytes_per_word;
+		if (dma->curr_xfer_len > bytes_per_buffer)
+			dma->curr_xfer_len = bytes_per_buffer;
+
+		ret = dspi_next_xfer_dma_submit(dspi);
+		if (ret) {
+			dev_err(dev, "DMA transfer failed\n");
+			goto exit;
+
+		} else {
+			const int len =
+				dma->curr_xfer_len * dspi->bytes_per_word;
+			curr_remaining_bytes -= len;
+			message->actual_length += len;
+			if (curr_remaining_bytes < 0)
+				curr_remaining_bytes = 0;
+		}
+	}
+
+exit:
+	return ret;
+}
+
+static int dspi_request_dma(struct fsl_dspi *dspi, phys_addr_t phy_addr)
+{
+	struct fsl_dspi_dma *dma;
+	struct dma_slave_config cfg;
+	struct device *dev = &dspi->pdev->dev;
+	int ret;
+
+	dma = devm_kzalloc(dev, sizeof(*dma), GFP_KERNEL);
+	if (!dma)
+		return -ENOMEM;
+
+	dma->chan_rx = dma_request_slave_channel(dev, "rx");
+	if (!dma->chan_rx) {
+		dev_err(dev, "rx dma channel not available\n");
+		ret = -ENODEV;
+		return ret;
+	}
+
+	dma->chan_tx = dma_request_slave_channel(dev, "tx");
+	if (!dma->chan_tx) {
+		dev_err(dev, "tx dma channel not available\n");
+		ret = -ENODEV;
+		goto err_tx_channel;
+	}
+
+	dma->tx_dma_buf = dma_alloc_coherent(dev, DSPI_DMA_BUFSIZE,
+					&dma->tx_dma_phys, GFP_KERNEL);
+	if (!dma->tx_dma_buf) {
+		ret = -ENOMEM;
+		goto err_tx_dma_buf;
+	}
+
+	dma->rx_dma_buf = dma_alloc_coherent(dev, DSPI_DMA_BUFSIZE,
+					&dma->rx_dma_phys, GFP_KERNEL);
+	if (!dma->rx_dma_buf) {
+		ret = -ENOMEM;
+		goto err_rx_dma_buf;
+	}
+
+	cfg.src_addr = phy_addr + SPI_POPR;
+	cfg.dst_addr = phy_addr + SPI_PUSHR;
+	cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+	cfg.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+	cfg.src_maxburst = 1;
+	cfg.dst_maxburst = 1;
+
+	cfg.direction = DMA_DEV_TO_MEM;
+	ret = dmaengine_slave_config(dma->chan_rx, &cfg);
+	if (ret) {
+		dev_err(dev, "can't configure rx dma channel\n");
+		ret = -EINVAL;
+		goto err_slave_config;
+	}
+
+	cfg.direction = DMA_MEM_TO_DEV;
+	ret = dmaengine_slave_config(dma->chan_tx, &cfg);
+	if (ret) {
+		dev_err(dev, "can't configure tx dma channel\n");
+		ret = -EINVAL;
+		goto err_slave_config;
+	}
+
+	dspi->dma = dma;
+	init_completion(&dma->cmd_tx_complete);
+	init_completion(&dma->cmd_rx_complete);
+
+	return 0;
+
+err_slave_config:
+	dma_free_coherent(dev, DSPI_DMA_BUFSIZE,
+			dma->rx_dma_buf, dma->rx_dma_phys);
+err_rx_dma_buf:
+	dma_free_coherent(dev, DSPI_DMA_BUFSIZE,
+			dma->tx_dma_buf, dma->tx_dma_phys);
+err_tx_dma_buf:
+	dma_release_channel(dma->chan_tx);
+err_tx_channel:
+	dma_release_channel(dma->chan_rx);
+
+	devm_kfree(dev, dma);
+	dspi->dma = NULL;
+
+	return ret;
+}
+
+static void dspi_release_dma(struct fsl_dspi *dspi)
+{
+	struct fsl_dspi_dma *dma = dspi->dma;
+	struct device *dev = &dspi->pdev->dev;
+
+	if (dma) {
+		if (dma->chan_tx) {
+			dma_unmap_single(dev, dma->tx_dma_phys,
+					DSPI_DMA_BUFSIZE, DMA_TO_DEVICE);
+			dma_release_channel(dma->chan_tx);
+		}
+
+		if (dma->chan_rx) {
+			dma_unmap_single(dev, dma->rx_dma_phys,
+					DSPI_DMA_BUFSIZE, DMA_FROM_DEVICE);
+			dma_release_channel(dma->chan_rx);
+		}
+	}
+}
+
+static void hz_to_spi_baud(char *pbr, char *br, int speed_hz,
+		unsigned long clkrate)
+{
+	/* Valid baud rate pre-scaler values */
+	int pbr_tbl[4] = {2, 3, 5, 7};
+	int brs[16] = {	2,	4,	6,	8,
+		16,	32,	64,	128,
+		256,	512,	1024,	2048,
+		4096,	8192,	16384,	32768 };
+	int scale_needed, scale, minscale = INT_MAX;
+	int i, j;
+
+	scale_needed = clkrate / speed_hz;
+	if (clkrate % speed_hz)
+		scale_needed++;
+
+	for (i = 0; i < ARRAY_SIZE(brs); i++)
+		for (j = 0; j < ARRAY_SIZE(pbr_tbl); j++) {
+			scale = brs[i] * pbr_tbl[j];
+			if (scale >= scale_needed) {
+				if (scale < minscale) {
+					minscale = scale;
+					*br = i;
+					*pbr = j;
+				}
+				break;
+			}
+		}
+
+	if (minscale == INT_MAX) {
+		pr_warn("Can not find valid baud rate,speed_hz is %d,clkrate is %ld, we use the max prescaler value.\n",
+			speed_hz, clkrate);
+		*pbr = ARRAY_SIZE(pbr_tbl) - 1;
+		*br =  ARRAY_SIZE(brs) - 1;
+	}
+}
+
+static void ns_delay_scale(char *psc, char *sc, int delay_ns,
+		unsigned long clkrate)
+{
+	int pscale_tbl[4] = {1, 3, 5, 7};
+	int scale_needed, scale, minscale = INT_MAX;
+	int i, j;
+	u32 remainder;
+
+	scale_needed = div_u64_rem((u64)delay_ns * clkrate, NSEC_PER_SEC,
+			&remainder);
+	if (remainder)
+		scale_needed++;
+
+	for (i = 0; i < ARRAY_SIZE(pscale_tbl); i++)
+		for (j = 0; j <= SPI_CTAR_SCALE_BITS; j++) {
+			scale = pscale_tbl[i] * (2 << j);
+			if (scale >= scale_needed) {
+				if (scale < minscale) {
+					minscale = scale;
+					*psc = i;
+					*sc = j;
+				}
+				break;
+			}
+		}
+
+	if (minscale == INT_MAX) {
+		pr_warn("Cannot find correct scale values for %dns delay at clkrate %ld, using max prescaler value",
+			delay_ns, clkrate);
+		*psc = ARRAY_SIZE(pscale_tbl) - 1;
+		*sc = SPI_CTAR_SCALE_BITS;
+	}
+}
+
+static void fifo_write(struct fsl_dspi *dspi)
+{
+	regmap_write(dspi->regmap, SPI_PUSHR, dspi_pop_tx_pushr(dspi));
+}
+
+static void cmd_fifo_write(struct fsl_dspi *dspi)
+{
+	u16 cmd = dspi->tx_cmd;
+
+	if (dspi->len > 0)
+		cmd |= SPI_PUSHR_CMD_CONT;
+	regmap_write(dspi->regmap_pushr, PUSHR_CMD, cmd);
+}
+
+static void tx_fifo_write(struct fsl_dspi *dspi, u16 txdata)
+{
+	regmap_write(dspi->regmap_pushr, PUSHR_TX, txdata);
+}
+
+static void dspi_tcfq_write(struct fsl_dspi *dspi)
+{
+	/* Clear transfer count */
+	dspi->tx_cmd |= SPI_PUSHR_CMD_CTCNT;
+
+	if (dspi->devtype_data->xspi_mode && dspi->bits_per_word > 16) {
+		/* Write two TX FIFO entries first, and then the corresponding
+		 * CMD FIFO entry.
+		 */
+		u32 data = dspi_pop_tx(dspi);
+
+		if (dspi->cur_chip->ctar_val & SPI_CTAR_LSBFE(1)) {
+			/* LSB */
+			tx_fifo_write(dspi, data & 0xFFFF);
+			tx_fifo_write(dspi, data >> 16);
+		} else {
+			/* MSB */
+			tx_fifo_write(dspi, data >> 16);
+			tx_fifo_write(dspi, data & 0xFFFF);
+		}
+		cmd_fifo_write(dspi);
+	} else {
+		/* Write one entry to both TX FIFO and CMD FIFO
+		 * simultaneously.
+		 */
+		fifo_write(dspi);
+	}
+}
+
+static u32 fifo_read(struct fsl_dspi *dspi)
+{
+	u32 rxdata = 0;
+
+	regmap_read(dspi->regmap, SPI_POPR, &rxdata);
+	return rxdata;
+}
+
+static void dspi_tcfq_read(struct fsl_dspi *dspi)
+{
+	dspi_push_rx(dspi, fifo_read(dspi));
+}
+
+static void dspi_eoq_write(struct fsl_dspi *dspi)
+{
+	int fifo_size = DSPI_FIFO_SIZE;
+	u16 xfer_cmd = dspi->tx_cmd;
+
+	/* Fill TX FIFO with as many transfers as possible */
+	while (dspi->len && fifo_size--) {
+		dspi->tx_cmd = xfer_cmd;
+		/* Request EOQF for last transfer in FIFO */
+		if (dspi->len == dspi->bytes_per_word || fifo_size == 0)
+			dspi->tx_cmd |= SPI_PUSHR_CMD_EOQ;
+		/* Clear transfer count for first transfer in FIFO */
+		if (fifo_size == (DSPI_FIFO_SIZE - 1))
+			dspi->tx_cmd |= SPI_PUSHR_CMD_CTCNT;
+		/* Write combined TX FIFO and CMD FIFO entry */
+		fifo_write(dspi);
+	}
+}
+
+static void dspi_eoq_read(struct fsl_dspi *dspi)
+{
+	int fifo_size = DSPI_FIFO_SIZE;
+
+	/* Read one FIFO entry at and push to rx buffer */
+	while ((dspi->rx < dspi->rx_end) && fifo_size--)
+		dspi_push_rx(dspi, fifo_read(dspi));
+}
+
+static int dspi_transfer_one_message(struct spi_master *master,
+		struct spi_message *message)
+{
+	struct fsl_dspi *dspi = spi_master_get_devdata(master);
+	struct spi_device *spi = message->spi;
+	struct spi_transfer *transfer;
+	int status = 0;
+	enum dspi_trans_mode trans_mode;
+
+	message->actual_length = 0;
+
+	list_for_each_entry(transfer, &message->transfers, transfer_list) {
+		dspi->cur_transfer = transfer;
+		dspi->cur_msg = message;
+		dspi->cur_chip = spi_get_ctldata(spi);
+		/* Prepare command word for CMD FIFO */
+		dspi->tx_cmd = SPI_PUSHR_CMD_CTAS(0) |
+			SPI_PUSHR_CMD_PCS(spi->chip_select);
+		if (list_is_last(&dspi->cur_transfer->transfer_list,
+				 &dspi->cur_msg->transfers)) {
+			/* Leave PCS activated after last transfer when
+			 * cs_change is set.
+			 */
+			if (transfer->cs_change)
+				dspi->tx_cmd |= SPI_PUSHR_CMD_CONT;
+		} else {
+			/* Keep PCS active between transfers in same message
+			 * when cs_change is not set, and de-activate PCS
+			 * between transfers in the same message when
+			 * cs_change is set.
+			 */
+			if (!transfer->cs_change)
+				dspi->tx_cmd |= SPI_PUSHR_CMD_CONT;
+		}
+
+		dspi->void_write_data = dspi->cur_chip->void_write_data;
+
+		dspi->tx = transfer->tx_buf;
+		dspi->rx = transfer->rx_buf;
+		dspi->rx_end = dspi->rx + transfer->len;
+		dspi->len = transfer->len;
+		/* Validated transfer specific frame size (defaults applied) */
+		dspi->bits_per_word = transfer->bits_per_word;
+		if (transfer->bits_per_word <= 8)
+			dspi->bytes_per_word = 1;
+		else if (transfer->bits_per_word <= 16)
+			dspi->bytes_per_word = 2;
+		else
+			dspi->bytes_per_word = 4;
+
+		regmap_update_bits(dspi->regmap, SPI_MCR,
+				   SPI_MCR_CLR_TXF | SPI_MCR_CLR_RXF,
+				   SPI_MCR_CLR_TXF | SPI_MCR_CLR_RXF);
+		regmap_write(dspi->regmap, SPI_CTAR(0),
+			     dspi->cur_chip->ctar_val |
+			     SPI_FRAME_BITS(transfer->bits_per_word));
+		if (dspi->devtype_data->xspi_mode)
+			regmap_write(dspi->regmap, SPI_CTARE(0),
+				     SPI_FRAME_EBITS(transfer->bits_per_word)
+				     | SPI_CTARE_DTCP(1));
+
+		trans_mode = dspi->devtype_data->trans_mode;
+		switch (trans_mode) {
+		case DSPI_EOQ_MODE:
+			regmap_write(dspi->regmap, SPI_RSER, SPI_RSER_EOQFE);
+			dspi_eoq_write(dspi);
+			break;
+		case DSPI_TCFQ_MODE:
+			regmap_write(dspi->regmap, SPI_RSER, SPI_RSER_TCFQE);
+			dspi_tcfq_write(dspi);
+			break;
+		case DSPI_DMA_MODE:
+			regmap_write(dspi->regmap, SPI_RSER,
+				SPI_RSER_TFFFE | SPI_RSER_TFFFD |
+				SPI_RSER_RFDFE | SPI_RSER_RFDFD);
+			status = dspi_dma_xfer(dspi);
+			break;
+		default:
+			dev_err(&dspi->pdev->dev, "unsupported trans_mode %u\n",
+				trans_mode);
+			status = -EINVAL;
+			goto out;
+		}
+
+		if (trans_mode != DSPI_DMA_MODE) {
+			if (wait_event_interruptible(dspi->waitq,
+						dspi->waitflags))
+				dev_err(&dspi->pdev->dev,
+					"wait transfer complete fail!\n");
+			dspi->waitflags = 0;
+		}
+
+		if (transfer->delay_usecs)
+			udelay(transfer->delay_usecs);
+	}
+
+out:
+	message->status = status;
+	spi_finalize_current_message(master);
+
+	return status;
+}
+
+static int dspi_setup(struct spi_device *spi)
+{
+	struct chip_data *chip;
+	struct fsl_dspi *dspi = spi_master_get_devdata(spi->master);
+	struct fsl_dspi_platform_data *pdata;
+	u32 cs_sck_delay = 0, sck_cs_delay = 0;
+	unsigned char br = 0, pbr = 0, pcssck = 0, cssck = 0;
+	unsigned char pasc = 0, asc = 0;
+	unsigned long clkrate;
+
+	/* Only alloc on first setup */
+	chip = spi_get_ctldata(spi);
+	if (chip == NULL) {
+		chip = kzalloc(sizeof(struct chip_data), GFP_KERNEL);
+		if (!chip)
+			return -ENOMEM;
+	}
+
+	pdata = dev_get_platdata(&dspi->pdev->dev);
+
+	if (!pdata) {
+		of_property_read_u32(spi->dev.of_node, "fsl,spi-cs-sck-delay",
+				&cs_sck_delay);
+
+		of_property_read_u32(spi->dev.of_node, "fsl,spi-sck-cs-delay",
+				&sck_cs_delay);
+	} else {
+		cs_sck_delay = pdata->cs_sck_delay;
+		sck_cs_delay = pdata->sck_cs_delay;
+	}
+
+	chip->void_write_data = 0;
+
+	clkrate = clk_get_rate(dspi->clk);
+	hz_to_spi_baud(&pbr, &br, spi->max_speed_hz, clkrate);
+
+	/* Set PCS to SCK delay scale values */
+	ns_delay_scale(&pcssck, &cssck, cs_sck_delay, clkrate);
+
+	/* Set After SCK delay scale values */
+	ns_delay_scale(&pasc, &asc, sck_cs_delay, clkrate);
+
+	chip->ctar_val = SPI_CTAR_CPOL(spi->mode & SPI_CPOL ? 1 : 0)
+		| SPI_CTAR_CPHA(spi->mode & SPI_CPHA ? 1 : 0)
+		| SPI_CTAR_LSBFE(spi->mode & SPI_LSB_FIRST ? 1 : 0)
+		| SPI_CTAR_PCSSCK(pcssck)
+		| SPI_CTAR_CSSCK(cssck)
+		| SPI_CTAR_PASC(pasc)
+		| SPI_CTAR_ASC(asc)
+		| SPI_CTAR_PBR(pbr)
+		| SPI_CTAR_BR(br);
+
+	spi_set_ctldata(spi, chip);
+
+	return 0;
+}
+
+static void dspi_cleanup(struct spi_device *spi)
+{
+	struct chip_data *chip = spi_get_ctldata((struct spi_device *)spi);
+
+	dev_dbg(&spi->dev, "spi_device %u.%u cleanup\n",
+			spi->master->bus_num, spi->chip_select);
+
+	kfree(chip);
+}
+
+static irqreturn_t dspi_interrupt(int irq, void *dev_id)
+{
+	struct fsl_dspi *dspi = (struct fsl_dspi *)dev_id;
+	struct spi_message *msg = dspi->cur_msg;
+	enum dspi_trans_mode trans_mode;
+	u32 spi_sr, spi_tcr;
+	u16 spi_tcnt;
+
+	regmap_read(dspi->regmap, SPI_SR, &spi_sr);
+	regmap_write(dspi->regmap, SPI_SR, spi_sr);
+
+
+	if (spi_sr & (SPI_SR_EOQF | SPI_SR_TCFQF)) {
+		/* Get transfer counter (in number of SPI transfers). It was
+		 * reset to 0 when transfer(s) were started.
+		 */
+		regmap_read(dspi->regmap, SPI_TCR, &spi_tcr);
+		spi_tcnt = SPI_TCR_GET_TCNT(spi_tcr);
+		/* Update total number of bytes that were transferred */
+		msg->actual_length += spi_tcnt * dspi->bytes_per_word;
+
+		trans_mode = dspi->devtype_data->trans_mode;
+		switch (trans_mode) {
+		case DSPI_EOQ_MODE:
+			dspi_eoq_read(dspi);
+			break;
+		case DSPI_TCFQ_MODE:
+			dspi_tcfq_read(dspi);
+			break;
+		default:
+			dev_err(&dspi->pdev->dev, "unsupported trans_mode %u\n",
+				trans_mode);
+				return IRQ_HANDLED;
+		}
+
+		if (!dspi->len) {
+			dspi->waitflags = 1;
+			wake_up_interruptible(&dspi->waitq);
+		} else {
+			switch (trans_mode) {
+			case DSPI_EOQ_MODE:
+				dspi_eoq_write(dspi);
+				break;
+			case DSPI_TCFQ_MODE:
+				dspi_tcfq_write(dspi);
+				break;
+			default:
+				dev_err(&dspi->pdev->dev,
+					"unsupported trans_mode %u\n",
+					trans_mode);
+			}
+		}
+	}
+
+	return IRQ_HANDLED;
+}
+
+static const struct of_device_id fsl_dspi_dt_ids[] = {
+	{ .compatible = "fsl,vf610-dspi", .data = &vf610_data, },
+	{ .compatible = "fsl,ls1021a-v1.0-dspi", .data = &ls1021a_v1_data, },
+	{ .compatible = "fsl,ls2085a-dspi", .data = &ls2085a_data, },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, fsl_dspi_dt_ids);
+
+#ifdef CONFIG_PM_SLEEP
+static int dspi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct fsl_dspi *dspi = spi_master_get_devdata(master);
+
+	spi_master_suspend(master);
+	clk_disable_unprepare(dspi->clk);
+
+	pinctrl_pm_select_sleep_state(dev);
+
+	return 0;
+}
+
+static int dspi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct fsl_dspi *dspi = spi_master_get_devdata(master);
+	int ret;
+
+	pinctrl_pm_select_default_state(dev);
+
+	ret = clk_prepare_enable(dspi->clk);
+	if (ret)
+		return ret;
+	spi_master_resume(master);
+
+	return 0;
+}
+#endif /* CONFIG_PM_SLEEP */
+
+static SIMPLE_DEV_PM_OPS(dspi_pm, dspi_suspend, dspi_resume);
+
+static const struct regmap_range dspi_volatile_ranges[] = {
+	regmap_reg_range(SPI_MCR, SPI_TCR),
+	regmap_reg_range(SPI_SR, SPI_SR),
+	regmap_reg_range(SPI_PUSHR, SPI_RXFR3),
+};
+
+static const struct regmap_access_table dspi_volatile_table = {
+	.yes_ranges     = dspi_volatile_ranges,
+	.n_yes_ranges   = ARRAY_SIZE(dspi_volatile_ranges),
+};
+
+static const struct regmap_config dspi_regmap_config = {
+	.reg_bits = 32,
+	.val_bits = 32,
+	.reg_stride = 4,
+	.max_register = 0x88,
+	.volatile_table = &dspi_volatile_table,
+};
+
+static const struct regmap_range dspi_xspi_volatile_ranges[] = {
+	regmap_reg_range(SPI_MCR, SPI_TCR),
+	regmap_reg_range(SPI_SR, SPI_SR),
+	regmap_reg_range(SPI_PUSHR, SPI_RXFR3),
+	regmap_reg_range(SPI_SREX, SPI_SREX),
+};
+
+static const struct regmap_access_table dspi_xspi_volatile_table = {
+	.yes_ranges     = dspi_xspi_volatile_ranges,
+	.n_yes_ranges   = ARRAY_SIZE(dspi_xspi_volatile_ranges),
+};
+
+static const struct regmap_config dspi_xspi_regmap_config[] = {
+	{
+		.reg_bits = 32,
+		.val_bits = 32,
+		.reg_stride = 4,
+		.max_register = 0x13c,
+		.volatile_table = &dspi_xspi_volatile_table,
+	},
+	{
+		.name = "pushr",
+		.reg_bits = 16,
+		.val_bits = 16,
+		.reg_stride = 2,
+		.max_register = 0x2,
+	},
+};
+
+static void dspi_init(struct fsl_dspi *dspi)
+{
+	regmap_write(dspi->regmap, SPI_MCR, SPI_MCR_MASTER | SPI_MCR_PCSIS |
+		     (dspi->devtype_data->xspi_mode ? SPI_MCR_XSPI : 0));
+	regmap_write(dspi->regmap, SPI_SR, SPI_SR_CLEAR);
+	if (dspi->devtype_data->xspi_mode)
+		regmap_write(dspi->regmap, SPI_CTARE(0),
+			     SPI_CTARE_FMSZE(0) | SPI_CTARE_DTCP(1));
+}
+
+static int dspi_probe(struct platform_device *pdev)
+{
+	struct device_node *np = pdev->dev.of_node;
+	struct spi_master *master;
+	struct fsl_dspi *dspi;
+	struct resource *res;
+	const struct regmap_config *regmap_config;
+	void __iomem *base;
+	struct fsl_dspi_platform_data *pdata;
+	int ret = 0, cs_num, bus_num;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct fsl_dspi));
+	if (!master)
+		return -ENOMEM;
+
+	dspi = spi_master_get_devdata(master);
+	dspi->pdev = pdev;
+	dspi->master = master;
+
+	master->transfer = NULL;
+	master->setup = dspi_setup;
+	master->transfer_one_message = dspi_transfer_one_message;
+	master->dev.of_node = pdev->dev.of_node;
+
+	master->cleanup = dspi_cleanup;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_LSB_FIRST;
+
+	pdata = dev_get_platdata(&pdev->dev);
+	if (pdata) {
+		master->num_chipselect = pdata->cs_num;
+		master->bus_num = pdata->bus_num;
+
+		dspi->devtype_data = &coldfire_data;
+	} else {
+
+		ret = of_property_read_u32(np, "spi-num-chipselects", &cs_num);
+		if (ret < 0) {
+			dev_err(&pdev->dev, "can't get spi-num-chipselects\n");
+			goto out_master_put;
+		}
+		master->num_chipselect = cs_num;
+
+		ret = of_property_read_u32(np, "bus-num", &bus_num);
+		if (ret < 0) {
+			dev_err(&pdev->dev, "can't get bus-num\n");
+			goto out_master_put;
+		}
+		master->bus_num = bus_num;
+
+		dspi->devtype_data = of_device_get_match_data(&pdev->dev);
+		if (!dspi->devtype_data) {
+			dev_err(&pdev->dev, "can't get devtype_data\n");
+			ret = -EFAULT;
+			goto out_master_put;
+		}
+	}
+
+	if (dspi->devtype_data->xspi_mode)
+		master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 32);
+	else
+		master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 16);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(base)) {
+		ret = PTR_ERR(base);
+		goto out_master_put;
+	}
+
+	if (dspi->devtype_data->xspi_mode)
+		regmap_config = &dspi_xspi_regmap_config[0];
+	else
+		regmap_config = &dspi_regmap_config;
+	dspi->regmap = devm_regmap_init_mmio(&pdev->dev, base, regmap_config);
+	if (IS_ERR(dspi->regmap)) {
+		dev_err(&pdev->dev, "failed to init regmap: %ld\n",
+				PTR_ERR(dspi->regmap));
+		ret = PTR_ERR(dspi->regmap);
+		goto out_master_put;
+	}
+
+	if (dspi->devtype_data->xspi_mode) {
+		dspi->regmap_pushr = devm_regmap_init_mmio(
+			&pdev->dev, base + SPI_PUSHR,
+			&dspi_xspi_regmap_config[1]);
+		if (IS_ERR(dspi->regmap_pushr)) {
+			dev_err(&pdev->dev,
+				"failed to init pushr regmap: %ld\n",
+				PTR_ERR(dspi->regmap_pushr));
+			ret = PTR_ERR(dspi->regmap_pushr);
+			goto out_master_put;
+		}
+	}
+
+	dspi->clk = devm_clk_get(&pdev->dev, "dspi");
+	if (IS_ERR(dspi->clk)) {
+		ret = PTR_ERR(dspi->clk);
+		dev_err(&pdev->dev, "unable to get clock\n");
+		goto out_master_put;
+	}
+	ret = clk_prepare_enable(dspi->clk);
+	if (ret)
+		goto out_master_put;
+
+	dspi_init(dspi);
+	dspi->irq = platform_get_irq(pdev, 0);
+	if (dspi->irq < 0) {
+		dev_err(&pdev->dev, "can't get platform irq\n");
+		ret = dspi->irq;
+		goto out_clk_put;
+	}
+
+	ret = devm_request_irq(&pdev->dev, dspi->irq, dspi_interrupt, 0,
+			pdev->name, dspi);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Unable to attach DSPI interrupt\n");
+		goto out_clk_put;
+	}
+
+	if (dspi->devtype_data->trans_mode == DSPI_DMA_MODE) {
+		ret = dspi_request_dma(dspi, res->start);
+		if (ret < 0) {
+			dev_err(&pdev->dev, "can't get dma channels\n");
+			goto out_clk_put;
+		}
+	}
+
+	master->max_speed_hz =
+		clk_get_rate(dspi->clk) / dspi->devtype_data->max_clock_factor;
+
+	init_waitqueue_head(&dspi->waitq);
+	platform_set_drvdata(pdev, master);
+
+	ret = spi_register_master(master);
+	if (ret != 0) {
+		dev_err(&pdev->dev, "Problem registering DSPI master\n");
+		goto out_clk_put;
+	}
+
+	return ret;
+
+out_clk_put:
+	clk_disable_unprepare(dspi->clk);
+out_master_put:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int dspi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct fsl_dspi *dspi = spi_master_get_devdata(master);
+
+	/* Disconnect from the SPI framework */
+	dspi_release_dma(dspi);
+	clk_disable_unprepare(dspi->clk);
+	spi_unregister_master(dspi->master);
+
+	return 0;
+}
+
+static struct platform_driver fsl_dspi_driver = {
+	.driver.name    = DRIVER_NAME,
+	.driver.of_match_table = fsl_dspi_dt_ids,
+	.driver.owner   = THIS_MODULE,
+	.driver.pm = &dspi_pm,
+	.probe          = dspi_probe,
+	.remove		= dspi_remove,
+};
+module_platform_driver(fsl_dspi_driver);
+
+MODULE_DESCRIPTION("Freescale DSPI Controller Driver");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:" DRIVER_NAME);
diff --git a/drivers/spi/spi-fsl-espi.c b/drivers/spi/spi-fsl-espi.c
new file mode 100644
index 0000000..1e8ff62
--- /dev/null
+++ b/drivers/spi/spi-fsl-espi.c
@@ -0,0 +1,849 @@
+/*
+ * Freescale eSPI controller driver.
+ *
+ * Copyright 2010 Freescale Semiconductor, Inc.
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/fsl_devices.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/mm.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/pm_runtime.h>
+#include <sysdev/fsl_soc.h>
+
+/* eSPI Controller registers */
+#define ESPI_SPMODE	0x00	/* eSPI mode register */
+#define ESPI_SPIE	0x04	/* eSPI event register */
+#define ESPI_SPIM	0x08	/* eSPI mask register */
+#define ESPI_SPCOM	0x0c	/* eSPI command register */
+#define ESPI_SPITF	0x10	/* eSPI transmit FIFO access register*/
+#define ESPI_SPIRF	0x14	/* eSPI receive FIFO access register*/
+#define ESPI_SPMODE0	0x20	/* eSPI cs0 mode register */
+
+#define ESPI_SPMODEx(x)	(ESPI_SPMODE0 + (x) * 4)
+
+/* eSPI Controller mode register definitions */
+#define SPMODE_ENABLE		BIT(31)
+#define SPMODE_LOOP		BIT(30)
+#define SPMODE_TXTHR(x)		((x) << 8)
+#define SPMODE_RXTHR(x)		((x) << 0)
+
+/* eSPI Controller CS mode register definitions */
+#define CSMODE_CI_INACTIVEHIGH	BIT(31)
+#define CSMODE_CP_BEGIN_EDGECLK	BIT(30)
+#define CSMODE_REV		BIT(29)
+#define CSMODE_DIV16		BIT(28)
+#define CSMODE_PM(x)		((x) << 24)
+#define CSMODE_POL_1		BIT(20)
+#define CSMODE_LEN(x)		((x) << 16)
+#define CSMODE_BEF(x)		((x) << 12)
+#define CSMODE_AFT(x)		((x) << 8)
+#define CSMODE_CG(x)		((x) << 3)
+
+#define FSL_ESPI_FIFO_SIZE	32
+#define FSL_ESPI_RXTHR		15
+
+/* Default mode/csmode for eSPI controller */
+#define SPMODE_INIT_VAL (SPMODE_TXTHR(4) | SPMODE_RXTHR(FSL_ESPI_RXTHR))
+#define CSMODE_INIT_VAL (CSMODE_POL_1 | CSMODE_BEF(0) \
+		| CSMODE_AFT(0) | CSMODE_CG(1))
+
+/* SPIE register values */
+#define SPIE_RXCNT(reg)     ((reg >> 24) & 0x3F)
+#define SPIE_TXCNT(reg)     ((reg >> 16) & 0x3F)
+#define	SPIE_TXE		BIT(15)	/* TX FIFO empty */
+#define	SPIE_DON		BIT(14)	/* TX done */
+#define	SPIE_RXT		BIT(13)	/* RX FIFO threshold */
+#define	SPIE_RXF		BIT(12)	/* RX FIFO full */
+#define	SPIE_TXT		BIT(11)	/* TX FIFO threshold*/
+#define	SPIE_RNE		BIT(9)	/* RX FIFO not empty */
+#define	SPIE_TNF		BIT(8)	/* TX FIFO not full */
+
+/* SPIM register values */
+#define	SPIM_TXE		BIT(15)	/* TX FIFO empty */
+#define	SPIM_DON		BIT(14)	/* TX done */
+#define	SPIM_RXT		BIT(13)	/* RX FIFO threshold */
+#define	SPIM_RXF		BIT(12)	/* RX FIFO full */
+#define	SPIM_TXT		BIT(11)	/* TX FIFO threshold*/
+#define	SPIM_RNE		BIT(9)	/* RX FIFO not empty */
+#define	SPIM_TNF		BIT(8)	/* TX FIFO not full */
+
+/* SPCOM register values */
+#define SPCOM_CS(x)		((x) << 30)
+#define SPCOM_DO		BIT(28) /* Dual output */
+#define SPCOM_TO		BIT(27) /* TX only */
+#define SPCOM_RXSKIP(x)		((x) << 16)
+#define SPCOM_TRANLEN(x)	((x) << 0)
+
+#define	SPCOM_TRANLEN_MAX	0x10000	/* Max transaction length */
+
+#define AUTOSUSPEND_TIMEOUT 2000
+
+struct fsl_espi {
+	struct device *dev;
+	void __iomem *reg_base;
+
+	struct list_head *m_transfers;
+	struct spi_transfer *tx_t;
+	unsigned int tx_pos;
+	bool tx_done;
+	struct spi_transfer *rx_t;
+	unsigned int rx_pos;
+	bool rx_done;
+
+	bool swab;
+	unsigned int rxskip;
+
+	spinlock_t lock;
+
+	u32 spibrg;             /* SPIBRG input clock */
+
+	struct completion done;
+};
+
+struct fsl_espi_cs {
+	u32 hw_mode;
+};
+
+static inline u32 fsl_espi_read_reg(struct fsl_espi *espi, int offset)
+{
+	return ioread32be(espi->reg_base + offset);
+}
+
+static inline u16 fsl_espi_read_reg16(struct fsl_espi *espi, int offset)
+{
+	return ioread16be(espi->reg_base + offset);
+}
+
+static inline u8 fsl_espi_read_reg8(struct fsl_espi *espi, int offset)
+{
+	return ioread8(espi->reg_base + offset);
+}
+
+static inline void fsl_espi_write_reg(struct fsl_espi *espi, int offset,
+				      u32 val)
+{
+	iowrite32be(val, espi->reg_base + offset);
+}
+
+static inline void fsl_espi_write_reg16(struct fsl_espi *espi, int offset,
+					u16 val)
+{
+	iowrite16be(val, espi->reg_base + offset);
+}
+
+static inline void fsl_espi_write_reg8(struct fsl_espi *espi, int offset,
+				       u8 val)
+{
+	iowrite8(val, espi->reg_base + offset);
+}
+
+static int fsl_espi_check_message(struct spi_message *m)
+{
+	struct fsl_espi *espi = spi_master_get_devdata(m->spi->master);
+	struct spi_transfer *t, *first;
+
+	if (m->frame_length > SPCOM_TRANLEN_MAX) {
+		dev_err(espi->dev, "message too long, size is %u bytes\n",
+			m->frame_length);
+		return -EMSGSIZE;
+	}
+
+	first = list_first_entry(&m->transfers, struct spi_transfer,
+				 transfer_list);
+
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if (first->bits_per_word != t->bits_per_word ||
+		    first->speed_hz != t->speed_hz) {
+			dev_err(espi->dev, "bits_per_word/speed_hz should be the same for all transfers\n");
+			return -EINVAL;
+		}
+	}
+
+	/* ESPI supports MSB-first transfers for word size 8 / 16 only */
+	if (!(m->spi->mode & SPI_LSB_FIRST) && first->bits_per_word != 8 &&
+	    first->bits_per_word != 16) {
+		dev_err(espi->dev,
+			"MSB-first transfer not supported for wordsize %u\n",
+			first->bits_per_word);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static unsigned int fsl_espi_check_rxskip_mode(struct spi_message *m)
+{
+	struct spi_transfer *t;
+	unsigned int i = 0, rxskip = 0;
+
+	/*
+	 * prerequisites for ESPI rxskip mode:
+	 * - message has two transfers
+	 * - first transfer is a write and second is a read
+	 *
+	 * In addition the current low-level transfer mechanism requires
+	 * that the rxskip bytes fit into the TX FIFO. Else the transfer
+	 * would hang because after the first FSL_ESPI_FIFO_SIZE bytes
+	 * the TX FIFO isn't re-filled.
+	 */
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if (i == 0) {
+			if (!t->tx_buf || t->rx_buf ||
+			    t->len > FSL_ESPI_FIFO_SIZE)
+				return 0;
+			rxskip = t->len;
+		} else if (i == 1) {
+			if (t->tx_buf || !t->rx_buf)
+				return 0;
+		}
+		i++;
+	}
+
+	return i == 2 ? rxskip : 0;
+}
+
+static void fsl_espi_fill_tx_fifo(struct fsl_espi *espi, u32 events)
+{
+	u32 tx_fifo_avail;
+	unsigned int tx_left;
+	const void *tx_buf;
+
+	/* if events is zero transfer has not started and tx fifo is empty */
+	tx_fifo_avail = events ? SPIE_TXCNT(events) :  FSL_ESPI_FIFO_SIZE;
+start:
+	tx_left = espi->tx_t->len - espi->tx_pos;
+	tx_buf = espi->tx_t->tx_buf;
+	while (tx_fifo_avail >= min(4U, tx_left) && tx_left) {
+		if (tx_left >= 4) {
+			if (!tx_buf)
+				fsl_espi_write_reg(espi, ESPI_SPITF, 0);
+			else if (espi->swab)
+				fsl_espi_write_reg(espi, ESPI_SPITF,
+					swahb32p(tx_buf + espi->tx_pos));
+			else
+				fsl_espi_write_reg(espi, ESPI_SPITF,
+					*(u32 *)(tx_buf + espi->tx_pos));
+			espi->tx_pos += 4;
+			tx_left -= 4;
+			tx_fifo_avail -= 4;
+		} else if (tx_left >= 2 && tx_buf && espi->swab) {
+			fsl_espi_write_reg16(espi, ESPI_SPITF,
+					swab16p(tx_buf + espi->tx_pos));
+			espi->tx_pos += 2;
+			tx_left -= 2;
+			tx_fifo_avail -= 2;
+		} else {
+			if (!tx_buf)
+				fsl_espi_write_reg8(espi, ESPI_SPITF, 0);
+			else
+				fsl_espi_write_reg8(espi, ESPI_SPITF,
+					*(u8 *)(tx_buf + espi->tx_pos));
+			espi->tx_pos += 1;
+			tx_left -= 1;
+			tx_fifo_avail -= 1;
+		}
+	}
+
+	if (!tx_left) {
+		/* Last transfer finished, in rxskip mode only one is needed */
+		if (list_is_last(&espi->tx_t->transfer_list,
+		    espi->m_transfers) || espi->rxskip) {
+			espi->tx_done = true;
+			return;
+		}
+		espi->tx_t = list_next_entry(espi->tx_t, transfer_list);
+		espi->tx_pos = 0;
+		/* continue with next transfer if tx fifo is not full */
+		if (tx_fifo_avail)
+			goto start;
+	}
+}
+
+static void fsl_espi_read_rx_fifo(struct fsl_espi *espi, u32 events)
+{
+	u32 rx_fifo_avail = SPIE_RXCNT(events);
+	unsigned int rx_left;
+	void *rx_buf;
+
+start:
+	rx_left = espi->rx_t->len - espi->rx_pos;
+	rx_buf = espi->rx_t->rx_buf;
+	while (rx_fifo_avail >= min(4U, rx_left) && rx_left) {
+		if (rx_left >= 4) {
+			u32 val = fsl_espi_read_reg(espi, ESPI_SPIRF);
+
+			if (rx_buf && espi->swab)
+				*(u32 *)(rx_buf + espi->rx_pos) = swahb32(val);
+			else if (rx_buf)
+				*(u32 *)(rx_buf + espi->rx_pos) = val;
+			espi->rx_pos += 4;
+			rx_left -= 4;
+			rx_fifo_avail -= 4;
+		} else if (rx_left >= 2 && rx_buf && espi->swab) {
+			u16 val = fsl_espi_read_reg16(espi, ESPI_SPIRF);
+
+			*(u16 *)(rx_buf + espi->rx_pos) = swab16(val);
+			espi->rx_pos += 2;
+			rx_left -= 2;
+			rx_fifo_avail -= 2;
+		} else {
+			u8 val = fsl_espi_read_reg8(espi, ESPI_SPIRF);
+
+			if (rx_buf)
+				*(u8 *)(rx_buf + espi->rx_pos) = val;
+			espi->rx_pos += 1;
+			rx_left -= 1;
+			rx_fifo_avail -= 1;
+		}
+	}
+
+	if (!rx_left) {
+		if (list_is_last(&espi->rx_t->transfer_list,
+		    espi->m_transfers)) {
+			espi->rx_done = true;
+			return;
+		}
+		espi->rx_t = list_next_entry(espi->rx_t, transfer_list);
+		espi->rx_pos = 0;
+		/* continue with next transfer if rx fifo is not empty */
+		if (rx_fifo_avail)
+			goto start;
+	}
+}
+
+static void fsl_espi_setup_transfer(struct spi_device *spi,
+					struct spi_transfer *t)
+{
+	struct fsl_espi *espi = spi_master_get_devdata(spi->master);
+	int bits_per_word = t ? t->bits_per_word : spi->bits_per_word;
+	u32 pm, hz = t ? t->speed_hz : spi->max_speed_hz;
+	struct fsl_espi_cs *cs = spi_get_ctldata(spi);
+	u32 hw_mode_old = cs->hw_mode;
+
+	/* mask out bits we are going to set */
+	cs->hw_mode &= ~(CSMODE_LEN(0xF) | CSMODE_DIV16 | CSMODE_PM(0xF));
+
+	cs->hw_mode |= CSMODE_LEN(bits_per_word - 1);
+
+	pm = DIV_ROUND_UP(espi->spibrg, hz * 4) - 1;
+
+	if (pm > 15) {
+		cs->hw_mode |= CSMODE_DIV16;
+		pm = DIV_ROUND_UP(espi->spibrg, hz * 16 * 4) - 1;
+	}
+
+	cs->hw_mode |= CSMODE_PM(pm);
+
+	/* don't write the mode register if the mode doesn't change */
+	if (cs->hw_mode != hw_mode_old)
+		fsl_espi_write_reg(espi, ESPI_SPMODEx(spi->chip_select),
+				   cs->hw_mode);
+}
+
+static int fsl_espi_bufs(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct fsl_espi *espi = spi_master_get_devdata(spi->master);
+	unsigned int rx_len = t->len;
+	u32 mask, spcom;
+	int ret;
+
+	reinit_completion(&espi->done);
+
+	/* Set SPCOM[CS] and SPCOM[TRANLEN] field */
+	spcom = SPCOM_CS(spi->chip_select);
+	spcom |= SPCOM_TRANLEN(t->len - 1);
+
+	/* configure RXSKIP mode */
+	if (espi->rxskip) {
+		spcom |= SPCOM_RXSKIP(espi->rxskip);
+		rx_len = t->len - espi->rxskip;
+		if (t->rx_nbits == SPI_NBITS_DUAL)
+			spcom |= SPCOM_DO;
+	}
+
+	fsl_espi_write_reg(espi, ESPI_SPCOM, spcom);
+
+	/* enable interrupts */
+	mask = SPIM_DON;
+	if (rx_len > FSL_ESPI_FIFO_SIZE)
+		mask |= SPIM_RXT;
+	fsl_espi_write_reg(espi, ESPI_SPIM, mask);
+
+	/* Prevent filling the fifo from getting interrupted */
+	spin_lock_irq(&espi->lock);
+	fsl_espi_fill_tx_fifo(espi, 0);
+	spin_unlock_irq(&espi->lock);
+
+	/* Won't hang up forever, SPI bus sometimes got lost interrupts... */
+	ret = wait_for_completion_timeout(&espi->done, 2 * HZ);
+	if (ret == 0)
+		dev_err(espi->dev, "Transfer timed out!\n");
+
+	/* disable rx ints */
+	fsl_espi_write_reg(espi, ESPI_SPIM, 0);
+
+	return ret == 0 ? -ETIMEDOUT : 0;
+}
+
+static int fsl_espi_trans(struct spi_message *m, struct spi_transfer *trans)
+{
+	struct fsl_espi *espi = spi_master_get_devdata(m->spi->master);
+	struct spi_device *spi = m->spi;
+	int ret;
+
+	/* In case of LSB-first and bits_per_word > 8 byte-swap all words */
+	espi->swab = spi->mode & SPI_LSB_FIRST && trans->bits_per_word > 8;
+
+	espi->m_transfers = &m->transfers;
+	espi->tx_t = list_first_entry(&m->transfers, struct spi_transfer,
+				      transfer_list);
+	espi->tx_pos = 0;
+	espi->tx_done = false;
+	espi->rx_t = list_first_entry(&m->transfers, struct spi_transfer,
+				      transfer_list);
+	espi->rx_pos = 0;
+	espi->rx_done = false;
+
+	espi->rxskip = fsl_espi_check_rxskip_mode(m);
+	if (trans->rx_nbits == SPI_NBITS_DUAL && !espi->rxskip) {
+		dev_err(espi->dev, "Dual output mode requires RXSKIP mode!\n");
+		return -EINVAL;
+	}
+
+	/* In RXSKIP mode skip first transfer for reads */
+	if (espi->rxskip)
+		espi->rx_t = list_next_entry(espi->rx_t, transfer_list);
+
+	fsl_espi_setup_transfer(spi, trans);
+
+	ret = fsl_espi_bufs(spi, trans);
+
+	if (trans->delay_usecs)
+		udelay(trans->delay_usecs);
+
+	return ret;
+}
+
+static int fsl_espi_do_one_msg(struct spi_master *master,
+			       struct spi_message *m)
+{
+	unsigned int delay_usecs = 0, rx_nbits = 0;
+	struct spi_transfer *t, trans = {};
+	int ret;
+
+	ret = fsl_espi_check_message(m);
+	if (ret)
+		goto out;
+
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if (t->delay_usecs > delay_usecs)
+			delay_usecs = t->delay_usecs;
+		if (t->rx_nbits > rx_nbits)
+			rx_nbits = t->rx_nbits;
+	}
+
+	t = list_first_entry(&m->transfers, struct spi_transfer,
+			     transfer_list);
+
+	trans.len = m->frame_length;
+	trans.speed_hz = t->speed_hz;
+	trans.bits_per_word = t->bits_per_word;
+	trans.delay_usecs = delay_usecs;
+	trans.rx_nbits = rx_nbits;
+
+	if (trans.len)
+		ret = fsl_espi_trans(m, &trans);
+
+	m->actual_length = ret ? 0 : trans.len;
+out:
+	if (m->status == -EINPROGRESS)
+		m->status = ret;
+
+	spi_finalize_current_message(master);
+
+	return ret;
+}
+
+static int fsl_espi_setup(struct spi_device *spi)
+{
+	struct fsl_espi *espi;
+	u32 loop_mode;
+	struct fsl_espi_cs *cs = spi_get_ctldata(spi);
+
+	if (!cs) {
+		cs = kzalloc(sizeof(*cs), GFP_KERNEL);
+		if (!cs)
+			return -ENOMEM;
+		spi_set_ctldata(spi, cs);
+	}
+
+	espi = spi_master_get_devdata(spi->master);
+
+	pm_runtime_get_sync(espi->dev);
+
+	cs->hw_mode = fsl_espi_read_reg(espi, ESPI_SPMODEx(spi->chip_select));
+	/* mask out bits we are going to set */
+	cs->hw_mode &= ~(CSMODE_CP_BEGIN_EDGECLK | CSMODE_CI_INACTIVEHIGH
+			 | CSMODE_REV);
+
+	if (spi->mode & SPI_CPHA)
+		cs->hw_mode |= CSMODE_CP_BEGIN_EDGECLK;
+	if (spi->mode & SPI_CPOL)
+		cs->hw_mode |= CSMODE_CI_INACTIVEHIGH;
+	if (!(spi->mode & SPI_LSB_FIRST))
+		cs->hw_mode |= CSMODE_REV;
+
+	/* Handle the loop mode */
+	loop_mode = fsl_espi_read_reg(espi, ESPI_SPMODE);
+	loop_mode &= ~SPMODE_LOOP;
+	if (spi->mode & SPI_LOOP)
+		loop_mode |= SPMODE_LOOP;
+	fsl_espi_write_reg(espi, ESPI_SPMODE, loop_mode);
+
+	fsl_espi_setup_transfer(spi, NULL);
+
+	pm_runtime_mark_last_busy(espi->dev);
+	pm_runtime_put_autosuspend(espi->dev);
+
+	return 0;
+}
+
+static void fsl_espi_cleanup(struct spi_device *spi)
+{
+	struct fsl_espi_cs *cs = spi_get_ctldata(spi);
+
+	kfree(cs);
+	spi_set_ctldata(spi, NULL);
+}
+
+static void fsl_espi_cpu_irq(struct fsl_espi *espi, u32 events)
+{
+	if (!espi->rx_done)
+		fsl_espi_read_rx_fifo(espi, events);
+
+	if (!espi->tx_done)
+		fsl_espi_fill_tx_fifo(espi, events);
+
+	if (!espi->tx_done || !espi->rx_done)
+		return;
+
+	/* we're done, but check for errors before returning */
+	events = fsl_espi_read_reg(espi, ESPI_SPIE);
+
+	if (!(events & SPIE_DON))
+		dev_err(espi->dev,
+			"Transfer done but SPIE_DON isn't set!\n");
+
+	if (SPIE_RXCNT(events) || SPIE_TXCNT(events) != FSL_ESPI_FIFO_SIZE) {
+		dev_err(espi->dev, "Transfer done but rx/tx fifo's aren't empty!\n");
+		dev_err(espi->dev, "SPIE_RXCNT = %d, SPIE_TXCNT = %d\n",
+			SPIE_RXCNT(events), SPIE_TXCNT(events));
+	}
+
+	complete(&espi->done);
+}
+
+static irqreturn_t fsl_espi_irq(s32 irq, void *context_data)
+{
+	struct fsl_espi *espi = context_data;
+	u32 events;
+
+	spin_lock(&espi->lock);
+
+	/* Get interrupt events(tx/rx) */
+	events = fsl_espi_read_reg(espi, ESPI_SPIE);
+	if (!events) {
+		spin_unlock(&espi->lock);
+		return IRQ_NONE;
+	}
+
+	dev_vdbg(espi->dev, "%s: events %x\n", __func__, events);
+
+	fsl_espi_cpu_irq(espi, events);
+
+	/* Clear the events */
+	fsl_espi_write_reg(espi, ESPI_SPIE, events);
+
+	spin_unlock(&espi->lock);
+
+	return IRQ_HANDLED;
+}
+
+#ifdef CONFIG_PM
+static int fsl_espi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct fsl_espi *espi = spi_master_get_devdata(master);
+	u32 regval;
+
+	regval = fsl_espi_read_reg(espi, ESPI_SPMODE);
+	regval &= ~SPMODE_ENABLE;
+	fsl_espi_write_reg(espi, ESPI_SPMODE, regval);
+
+	return 0;
+}
+
+static int fsl_espi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct fsl_espi *espi = spi_master_get_devdata(master);
+	u32 regval;
+
+	regval = fsl_espi_read_reg(espi, ESPI_SPMODE);
+	regval |= SPMODE_ENABLE;
+	fsl_espi_write_reg(espi, ESPI_SPMODE, regval);
+
+	return 0;
+}
+#endif
+
+static size_t fsl_espi_max_message_size(struct spi_device *spi)
+{
+	return SPCOM_TRANLEN_MAX;
+}
+
+static void fsl_espi_init_regs(struct device *dev, bool initial)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct fsl_espi *espi = spi_master_get_devdata(master);
+	struct device_node *nc;
+	u32 csmode, cs, prop;
+	int ret;
+
+	/* SPI controller initializations */
+	fsl_espi_write_reg(espi, ESPI_SPMODE, 0);
+	fsl_espi_write_reg(espi, ESPI_SPIM, 0);
+	fsl_espi_write_reg(espi, ESPI_SPCOM, 0);
+	fsl_espi_write_reg(espi, ESPI_SPIE, 0xffffffff);
+
+	/* Init eSPI CS mode register */
+	for_each_available_child_of_node(master->dev.of_node, nc) {
+		/* get chip select */
+		ret = of_property_read_u32(nc, "reg", &cs);
+		if (ret || cs >= master->num_chipselect)
+			continue;
+
+		csmode = CSMODE_INIT_VAL;
+
+		/* check if CSBEF is set in device tree */
+		ret = of_property_read_u32(nc, "fsl,csbef", &prop);
+		if (!ret) {
+			csmode &= ~(CSMODE_BEF(0xf));
+			csmode |= CSMODE_BEF(prop);
+		}
+
+		/* check if CSAFT is set in device tree */
+		ret = of_property_read_u32(nc, "fsl,csaft", &prop);
+		if (!ret) {
+			csmode &= ~(CSMODE_AFT(0xf));
+			csmode |= CSMODE_AFT(prop);
+		}
+
+		fsl_espi_write_reg(espi, ESPI_SPMODEx(cs), csmode);
+
+		if (initial)
+			dev_info(dev, "cs=%u, init_csmode=0x%x\n", cs, csmode);
+	}
+
+	/* Enable SPI interface */
+	fsl_espi_write_reg(espi, ESPI_SPMODE, SPMODE_INIT_VAL | SPMODE_ENABLE);
+}
+
+static int fsl_espi_probe(struct device *dev, struct resource *mem,
+			  unsigned int irq, unsigned int num_cs)
+{
+	struct spi_master *master;
+	struct fsl_espi *espi;
+	int ret;
+
+	master = spi_alloc_master(dev, sizeof(struct fsl_espi));
+	if (!master)
+		return -ENOMEM;
+
+	dev_set_drvdata(dev, master);
+
+	master->mode_bits = SPI_RX_DUAL | SPI_CPOL | SPI_CPHA | SPI_CS_HIGH |
+			    SPI_LSB_FIRST | SPI_LOOP;
+	master->dev.of_node = dev->of_node;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 16);
+	master->setup = fsl_espi_setup;
+	master->cleanup = fsl_espi_cleanup;
+	master->transfer_one_message = fsl_espi_do_one_msg;
+	master->auto_runtime_pm = true;
+	master->max_message_size = fsl_espi_max_message_size;
+	master->num_chipselect = num_cs;
+
+	espi = spi_master_get_devdata(master);
+	spin_lock_init(&espi->lock);
+
+	espi->dev = dev;
+	espi->spibrg = fsl_get_sys_freq();
+	if (espi->spibrg == -1) {
+		dev_err(dev, "Can't get sys frequency!\n");
+		ret = -EINVAL;
+		goto err_probe;
+	}
+	/* determined by clock divider fields DIV16/PM in register SPMODEx */
+	master->min_speed_hz = DIV_ROUND_UP(espi->spibrg, 4 * 16 * 16);
+	master->max_speed_hz = DIV_ROUND_UP(espi->spibrg, 4);
+
+	init_completion(&espi->done);
+
+	espi->reg_base = devm_ioremap_resource(dev, mem);
+	if (IS_ERR(espi->reg_base)) {
+		ret = PTR_ERR(espi->reg_base);
+		goto err_probe;
+	}
+
+	/* Register for SPI Interrupt */
+	ret = devm_request_irq(dev, irq, fsl_espi_irq, 0, "fsl_espi", espi);
+	if (ret)
+		goto err_probe;
+
+	fsl_espi_init_regs(dev, true);
+
+	pm_runtime_set_autosuspend_delay(dev, AUTOSUSPEND_TIMEOUT);
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_set_active(dev);
+	pm_runtime_enable(dev);
+	pm_runtime_get_sync(dev);
+
+	ret = devm_spi_register_master(dev, master);
+	if (ret < 0)
+		goto err_pm;
+
+	dev_info(dev, "at 0x%p (irq = %u)\n", espi->reg_base, irq);
+
+	pm_runtime_mark_last_busy(dev);
+	pm_runtime_put_autosuspend(dev);
+
+	return 0;
+
+err_pm:
+	pm_runtime_put_noidle(dev);
+	pm_runtime_disable(dev);
+	pm_runtime_set_suspended(dev);
+err_probe:
+	spi_master_put(master);
+	return ret;
+}
+
+static int of_fsl_espi_get_chipselects(struct device *dev)
+{
+	struct device_node *np = dev->of_node;
+	u32 num_cs;
+	int ret;
+
+	ret = of_property_read_u32(np, "fsl,espi-num-chipselects", &num_cs);
+	if (ret) {
+		dev_err(dev, "No 'fsl,espi-num-chipselects' property\n");
+		return 0;
+	}
+
+	return num_cs;
+}
+
+static int of_fsl_espi_probe(struct platform_device *ofdev)
+{
+	struct device *dev = &ofdev->dev;
+	struct device_node *np = ofdev->dev.of_node;
+	struct resource mem;
+	unsigned int irq, num_cs;
+	int ret;
+
+	if (of_property_read_bool(np, "mode")) {
+		dev_err(dev, "mode property is not supported on ESPI!\n");
+		return -EINVAL;
+	}
+
+	num_cs = of_fsl_espi_get_chipselects(dev);
+	if (!num_cs)
+		return -EINVAL;
+
+	ret = of_address_to_resource(np, 0, &mem);
+	if (ret)
+		return ret;
+
+	irq = irq_of_parse_and_map(np, 0);
+	if (!irq)
+		return -EINVAL;
+
+	return fsl_espi_probe(dev, &mem, irq, num_cs);
+}
+
+static int of_fsl_espi_remove(struct platform_device *dev)
+{
+	pm_runtime_disable(&dev->dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int of_fsl_espi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	int ret;
+
+	ret = spi_master_suspend(master);
+	if (ret) {
+		dev_warn(dev, "cannot suspend master\n");
+		return ret;
+	}
+
+	return pm_runtime_force_suspend(dev);
+}
+
+static int of_fsl_espi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	int ret;
+
+	fsl_espi_init_regs(dev, false);
+
+	ret = pm_runtime_force_resume(dev);
+	if (ret < 0)
+		return ret;
+
+	return spi_master_resume(master);
+}
+#endif /* CONFIG_PM_SLEEP */
+
+static const struct dev_pm_ops espi_pm = {
+	SET_RUNTIME_PM_OPS(fsl_espi_runtime_suspend,
+			   fsl_espi_runtime_resume, NULL)
+	SET_SYSTEM_SLEEP_PM_OPS(of_fsl_espi_suspend, of_fsl_espi_resume)
+};
+
+static const struct of_device_id of_fsl_espi_match[] = {
+	{ .compatible = "fsl,mpc8536-espi" },
+	{}
+};
+MODULE_DEVICE_TABLE(of, of_fsl_espi_match);
+
+static struct platform_driver fsl_espi_driver = {
+	.driver = {
+		.name = "fsl_espi",
+		.of_match_table = of_fsl_espi_match,
+		.pm = &espi_pm,
+	},
+	.probe		= of_fsl_espi_probe,
+	.remove		= of_fsl_espi_remove,
+};
+module_platform_driver(fsl_espi_driver);
+
+MODULE_AUTHOR("Mingkai Hu");
+MODULE_DESCRIPTION("Enhanced Freescale SPI Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-fsl-lib.c b/drivers/spi/spi-fsl-lib.c
new file mode 100644
index 0000000..1e43412
--- /dev/null
+++ b/drivers/spi/spi-fsl-lib.c
@@ -0,0 +1,164 @@
+/*
+ * Freescale SPI/eSPI controller driver library.
+ *
+ * Maintainer: Kumar Gala
+ *
+ * Copyright (C) 2006 Polycom, Inc.
+ *
+ * CPM SPI and QE buffer descriptors mode support:
+ * Copyright (c) 2009  MontaVista Software, Inc.
+ * Author: Anton Vorontsov <avorontsov@ru.mvista.com>
+ *
+ * Copyright 2010 Freescale Semiconductor, Inc.
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+#include <linux/dma-mapping.h>
+#include <linux/fsl_devices.h>
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/module.h>
+#include <linux/of_platform.h>
+#include <linux/spi/spi.h>
+#ifdef CONFIG_FSL_SOC
+#include <sysdev/fsl_soc.h>
+#endif
+
+#include "spi-fsl-lib.h"
+
+#define MPC8XXX_SPI_RX_BUF(type) 					  \
+void mpc8xxx_spi_rx_buf_##type(u32 data, struct mpc8xxx_spi *mpc8xxx_spi) \
+{									  \
+	type *rx = mpc8xxx_spi->rx;					  \
+	*rx++ = (type)(data >> mpc8xxx_spi->rx_shift);			  \
+	mpc8xxx_spi->rx = rx;						  \
+}									  \
+EXPORT_SYMBOL_GPL(mpc8xxx_spi_rx_buf_##type);
+
+#define MPC8XXX_SPI_TX_BUF(type)				\
+u32 mpc8xxx_spi_tx_buf_##type(struct mpc8xxx_spi *mpc8xxx_spi)	\
+{								\
+	u32 data;						\
+	const type *tx = mpc8xxx_spi->tx;			\
+	if (!tx)						\
+		return 0;					\
+	data = *tx++ << mpc8xxx_spi->tx_shift;			\
+	mpc8xxx_spi->tx = tx;					\
+	return data;						\
+}								\
+EXPORT_SYMBOL_GPL(mpc8xxx_spi_tx_buf_##type);
+
+MPC8XXX_SPI_RX_BUF(u8)
+MPC8XXX_SPI_RX_BUF(u16)
+MPC8XXX_SPI_RX_BUF(u32)
+MPC8XXX_SPI_TX_BUF(u8)
+MPC8XXX_SPI_TX_BUF(u16)
+MPC8XXX_SPI_TX_BUF(u32)
+
+struct mpc8xxx_spi_probe_info *to_of_pinfo(struct fsl_spi_platform_data *pdata)
+{
+	return container_of(pdata, struct mpc8xxx_spi_probe_info, pdata);
+}
+EXPORT_SYMBOL_GPL(to_of_pinfo);
+
+const char *mpc8xxx_spi_strmode(unsigned int flags)
+{
+	if (flags & SPI_QE_CPU_MODE) {
+		return "QE CPU";
+	} else if (flags & SPI_CPM_MODE) {
+		if (flags & SPI_QE)
+			return "QE";
+		else if (flags & SPI_CPM2)
+			return "CPM2";
+		else
+			return "CPM1";
+	}
+	return "CPU";
+}
+EXPORT_SYMBOL_GPL(mpc8xxx_spi_strmode);
+
+void mpc8xxx_spi_probe(struct device *dev, struct resource *mem,
+			unsigned int irq)
+{
+	struct fsl_spi_platform_data *pdata = dev_get_platdata(dev);
+	struct spi_master *master;
+	struct mpc8xxx_spi *mpc8xxx_spi;
+
+	master = dev_get_drvdata(dev);
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH
+			| SPI_LSB_FIRST | SPI_LOOP;
+
+	master->dev.of_node = dev->of_node;
+
+	mpc8xxx_spi = spi_master_get_devdata(master);
+	mpc8xxx_spi->dev = dev;
+	mpc8xxx_spi->get_rx = mpc8xxx_spi_rx_buf_u8;
+	mpc8xxx_spi->get_tx = mpc8xxx_spi_tx_buf_u8;
+	mpc8xxx_spi->flags = pdata->flags;
+	mpc8xxx_spi->spibrg = pdata->sysclk;
+	mpc8xxx_spi->irq = irq;
+
+	mpc8xxx_spi->rx_shift = 0;
+	mpc8xxx_spi->tx_shift = 0;
+
+	master->bus_num = pdata->bus_num;
+	master->num_chipselect = pdata->max_chipselect;
+
+	init_completion(&mpc8xxx_spi->done);
+}
+EXPORT_SYMBOL_GPL(mpc8xxx_spi_probe);
+
+int of_mpc8xxx_spi_probe(struct platform_device *ofdev)
+{
+	struct device *dev = &ofdev->dev;
+	struct device_node *np = ofdev->dev.of_node;
+	struct mpc8xxx_spi_probe_info *pinfo;
+	struct fsl_spi_platform_data *pdata;
+	const void *prop;
+	int ret = -ENOMEM;
+
+	pinfo = devm_kzalloc(&ofdev->dev, sizeof(*pinfo), GFP_KERNEL);
+	if (!pinfo)
+		return ret;
+
+	pdata = &pinfo->pdata;
+	dev->platform_data = pdata;
+
+	/* Allocate bus num dynamically. */
+	pdata->bus_num = -1;
+
+#ifdef CONFIG_FSL_SOC
+	/* SPI controller is either clocked from QE or SoC clock. */
+	pdata->sysclk = get_brgfreq();
+	if (pdata->sysclk == -1) {
+		pdata->sysclk = fsl_get_sys_freq();
+		if (pdata->sysclk == -1)
+			return -ENODEV;
+	}
+#else
+	ret = of_property_read_u32(np, "clock-frequency", &pdata->sysclk);
+	if (ret)
+		return ret;
+#endif
+
+	prop = of_get_property(np, "mode", NULL);
+	if (prop && !strcmp(prop, "cpu-qe"))
+		pdata->flags = SPI_QE_CPU_MODE;
+	else if (prop && !strcmp(prop, "qe"))
+		pdata->flags = SPI_CPM_MODE | SPI_QE;
+	else if (of_device_is_compatible(np, "fsl,cpm2-spi"))
+		pdata->flags = SPI_CPM_MODE | SPI_CPM2;
+	else if (of_device_is_compatible(np, "fsl,cpm1-spi"))
+		pdata->flags = SPI_CPM_MODE | SPI_CPM1;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(of_mpc8xxx_spi_probe);
+
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-fsl-lib.h b/drivers/spi/spi-fsl-lib.h
new file mode 100644
index 0000000..f303f30
--- /dev/null
+++ b/drivers/spi/spi-fsl-lib.h
@@ -0,0 +1,119 @@
+/*
+ * Freescale SPI/eSPI controller driver library.
+ *
+ * Maintainer: Kumar Gala
+ *
+ * Copyright 2010 Freescale Semiconductor, Inc.
+ * Copyright (C) 2006 Polycom, Inc.
+ *
+ * CPM SPI and QE buffer descriptors mode support:
+ * Copyright (c) 2009  MontaVista Software, Inc.
+ * Author: Anton Vorontsov <avorontsov@ru.mvista.com>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+#ifndef __SPI_FSL_LIB_H__
+#define __SPI_FSL_LIB_H__
+
+#include <asm/io.h>
+
+/* SPI/eSPI Controller driver's private data. */
+struct mpc8xxx_spi {
+	struct device *dev;
+	void __iomem *reg_base;
+
+	/* rx & tx bufs from the spi_transfer */
+	const void *tx;
+	void *rx;
+
+	int subblock;
+	struct spi_pram __iomem *pram;
+#ifdef CONFIG_FSL_SOC
+	struct cpm_buf_desc __iomem *tx_bd;
+	struct cpm_buf_desc __iomem *rx_bd;
+#endif
+
+	struct spi_transfer *xfer_in_progress;
+
+	/* dma addresses for CPM transfers */
+	dma_addr_t tx_dma;
+	dma_addr_t rx_dma;
+	bool map_tx_dma;
+	bool map_rx_dma;
+
+	dma_addr_t dma_dummy_tx;
+	dma_addr_t dma_dummy_rx;
+
+	/* functions to deal with different sized buffers */
+	void (*get_rx) (u32 rx_data, struct mpc8xxx_spi *);
+	u32(*get_tx) (struct mpc8xxx_spi *);
+
+	unsigned int count;
+	unsigned int irq;
+
+	unsigned nsecs;		/* (clock cycle time)/2 */
+
+	u32 spibrg;		/* SPIBRG input clock */
+	u32 rx_shift;		/* RX data reg shift when in qe mode */
+	u32 tx_shift;		/* TX data reg shift when in qe mode */
+
+	unsigned int flags;
+
+#if IS_ENABLED(CONFIG_SPI_FSL_SPI)
+	int type;
+	int native_chipselects;
+	u8 max_bits_per_word;
+
+	void (*set_shifts)(u32 *rx_shift, u32 *tx_shift,
+			   int bits_per_word, int msb_first);
+#endif
+
+	struct completion done;
+};
+
+struct spi_mpc8xxx_cs {
+	/* functions to deal with different sized buffers */
+	void (*get_rx) (u32 rx_data, struct mpc8xxx_spi *);
+	u32 (*get_tx) (struct mpc8xxx_spi *);
+	u32 rx_shift;		/* RX data reg shift when in qe mode */
+	u32 tx_shift;		/* TX data reg shift when in qe mode */
+	u32 hw_mode;		/* Holds HW mode register settings */
+};
+
+static inline void mpc8xxx_spi_write_reg(__be32 __iomem *reg, u32 val)
+{
+	iowrite32be(val, reg);
+}
+
+static inline u32 mpc8xxx_spi_read_reg(__be32 __iomem *reg)
+{
+	return ioread32be(reg);
+}
+
+struct mpc8xxx_spi_probe_info {
+	struct fsl_spi_platform_data pdata;
+	int *gpios;
+	bool *alow_flags;
+};
+
+extern u32 mpc8xxx_spi_tx_buf_u8(struct mpc8xxx_spi *mpc8xxx_spi);
+extern u32 mpc8xxx_spi_tx_buf_u16(struct mpc8xxx_spi *mpc8xxx_spi);
+extern u32 mpc8xxx_spi_tx_buf_u32(struct mpc8xxx_spi *mpc8xxx_spi);
+extern void mpc8xxx_spi_rx_buf_u8(u32 data, struct mpc8xxx_spi *mpc8xxx_spi);
+extern void mpc8xxx_spi_rx_buf_u16(u32 data, struct mpc8xxx_spi *mpc8xxx_spi);
+extern void mpc8xxx_spi_rx_buf_u32(u32 data, struct mpc8xxx_spi *mpc8xxx_spi);
+
+extern struct mpc8xxx_spi_probe_info *to_of_pinfo(
+		struct fsl_spi_platform_data *pdata);
+extern int mpc8xxx_spi_bufs(struct mpc8xxx_spi *mspi,
+		struct spi_transfer *t, unsigned int len);
+extern const char *mpc8xxx_spi_strmode(unsigned int flags);
+extern void mpc8xxx_spi_probe(struct device *dev, struct resource *mem,
+		unsigned int irq);
+extern int mpc8xxx_spi_remove(struct device *dev);
+extern int of_mpc8xxx_spi_probe(struct platform_device *ofdev);
+
+#endif /* __SPI_FSL_LIB_H__ */
diff --git a/drivers/spi/spi-fsl-lpspi.c b/drivers/spi/spi-fsl-lpspi.c
new file mode 100644
index 0000000..e6d5cc6
--- /dev/null
+++ b/drivers/spi/spi-fsl-lpspi.c
@@ -0,0 +1,514 @@
+// SPDX-License-Identifier: GPL-2.0+
+//
+// Freescale i.MX7ULP LPSPI driver
+//
+// Copyright 2016 Freescale Semiconductor, Inc.
+
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/types.h>
+
+#define DRIVER_NAME "fsl_lpspi"
+
+/* i.MX7ULP LPSPI registers */
+#define IMX7ULP_VERID	0x0
+#define IMX7ULP_PARAM	0x4
+#define IMX7ULP_CR	0x10
+#define IMX7ULP_SR	0x14
+#define IMX7ULP_IER	0x18
+#define IMX7ULP_DER	0x1c
+#define IMX7ULP_CFGR0	0x20
+#define IMX7ULP_CFGR1	0x24
+#define IMX7ULP_DMR0	0x30
+#define IMX7ULP_DMR1	0x34
+#define IMX7ULP_CCR	0x40
+#define IMX7ULP_FCR	0x58
+#define IMX7ULP_FSR	0x5c
+#define IMX7ULP_TCR	0x60
+#define IMX7ULP_TDR	0x64
+#define IMX7ULP_RSR	0x70
+#define IMX7ULP_RDR	0x74
+
+/* General control register field define */
+#define CR_RRF		BIT(9)
+#define CR_RTF		BIT(8)
+#define CR_RST		BIT(1)
+#define CR_MEN		BIT(0)
+#define SR_TCF		BIT(10)
+#define SR_RDF		BIT(1)
+#define SR_TDF		BIT(0)
+#define IER_TCIE	BIT(10)
+#define IER_RDIE	BIT(1)
+#define IER_TDIE	BIT(0)
+#define CFGR1_PCSCFG	BIT(27)
+#define CFGR1_PCSPOL	BIT(8)
+#define CFGR1_NOSTALL	BIT(3)
+#define CFGR1_MASTER	BIT(0)
+#define RSR_RXEMPTY	BIT(1)
+#define TCR_CPOL	BIT(31)
+#define TCR_CPHA	BIT(30)
+#define TCR_CONT	BIT(21)
+#define TCR_CONTC	BIT(20)
+#define TCR_RXMSK	BIT(19)
+#define TCR_TXMSK	BIT(18)
+
+static int clkdivs[] = {1, 2, 4, 8, 16, 32, 64, 128};
+
+struct lpspi_config {
+	u8 bpw;
+	u8 chip_select;
+	u8 prescale;
+	u16 mode;
+	u32 speed_hz;
+};
+
+struct fsl_lpspi_data {
+	struct device *dev;
+	void __iomem *base;
+	struct clk *clk;
+
+	void *rx_buf;
+	const void *tx_buf;
+	void (*tx)(struct fsl_lpspi_data *);
+	void (*rx)(struct fsl_lpspi_data *);
+
+	u32 remain;
+	u8 txfifosize;
+	u8 rxfifosize;
+
+	struct lpspi_config config;
+	struct completion xfer_done;
+};
+
+static const struct of_device_id fsl_lpspi_dt_ids[] = {
+	{ .compatible = "fsl,imx7ulp-spi", },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, fsl_lpspi_dt_ids);
+
+#define LPSPI_BUF_RX(type)						\
+static void fsl_lpspi_buf_rx_##type(struct fsl_lpspi_data *fsl_lpspi)	\
+{									\
+	unsigned int val = readl(fsl_lpspi->base + IMX7ULP_RDR);	\
+									\
+	if (fsl_lpspi->rx_buf) {					\
+		*(type *)fsl_lpspi->rx_buf = val;			\
+		fsl_lpspi->rx_buf += sizeof(type);                      \
+	}								\
+}
+
+#define LPSPI_BUF_TX(type)						\
+static void fsl_lpspi_buf_tx_##type(struct fsl_lpspi_data *fsl_lpspi)	\
+{									\
+	type val = 0;							\
+									\
+	if (fsl_lpspi->tx_buf) {					\
+		val = *(type *)fsl_lpspi->tx_buf;			\
+		fsl_lpspi->tx_buf += sizeof(type);			\
+	}								\
+									\
+	fsl_lpspi->remain -= sizeof(type);				\
+	writel(val, fsl_lpspi->base + IMX7ULP_TDR);			\
+}
+
+LPSPI_BUF_RX(u8)
+LPSPI_BUF_TX(u8)
+LPSPI_BUF_RX(u16)
+LPSPI_BUF_TX(u16)
+LPSPI_BUF_RX(u32)
+LPSPI_BUF_TX(u32)
+
+static void fsl_lpspi_intctrl(struct fsl_lpspi_data *fsl_lpspi,
+			      unsigned int enable)
+{
+	writel(enable, fsl_lpspi->base + IMX7ULP_IER);
+}
+
+static int lpspi_prepare_xfer_hardware(struct spi_master *master)
+{
+	struct fsl_lpspi_data *fsl_lpspi = spi_master_get_devdata(master);
+
+	return clk_prepare_enable(fsl_lpspi->clk);
+}
+
+static int lpspi_unprepare_xfer_hardware(struct spi_master *master)
+{
+	struct fsl_lpspi_data *fsl_lpspi = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(fsl_lpspi->clk);
+
+	return 0;
+}
+
+static int fsl_lpspi_txfifo_empty(struct fsl_lpspi_data *fsl_lpspi)
+{
+	u32 txcnt;
+	unsigned long orig_jiffies = jiffies;
+
+	do {
+		txcnt = readl(fsl_lpspi->base + IMX7ULP_FSR) & 0xff;
+
+		if (time_after(jiffies, orig_jiffies + msecs_to_jiffies(500))) {
+			dev_dbg(fsl_lpspi->dev, "txfifo empty timeout\n");
+			return -ETIMEDOUT;
+		}
+		cond_resched();
+
+	} while (txcnt);
+
+	return 0;
+}
+
+static void fsl_lpspi_write_tx_fifo(struct fsl_lpspi_data *fsl_lpspi)
+{
+	u8 txfifo_cnt;
+
+	txfifo_cnt = readl(fsl_lpspi->base + IMX7ULP_FSR) & 0xff;
+
+	while (txfifo_cnt < fsl_lpspi->txfifosize) {
+		if (!fsl_lpspi->remain)
+			break;
+		fsl_lpspi->tx(fsl_lpspi);
+		txfifo_cnt++;
+	}
+
+	if (!fsl_lpspi->remain && (txfifo_cnt < fsl_lpspi->txfifosize))
+		writel(0, fsl_lpspi->base + IMX7ULP_TDR);
+	else
+		fsl_lpspi_intctrl(fsl_lpspi, IER_TDIE);
+}
+
+static void fsl_lpspi_read_rx_fifo(struct fsl_lpspi_data *fsl_lpspi)
+{
+	while (!(readl(fsl_lpspi->base + IMX7ULP_RSR) & RSR_RXEMPTY))
+		fsl_lpspi->rx(fsl_lpspi);
+}
+
+static void fsl_lpspi_set_cmd(struct fsl_lpspi_data *fsl_lpspi,
+			      bool is_first_xfer)
+{
+	u32 temp = 0;
+
+	temp |= fsl_lpspi->config.bpw - 1;
+	temp |= fsl_lpspi->config.prescale << 27;
+	temp |= (fsl_lpspi->config.mode & 0x3) << 30;
+	temp |= (fsl_lpspi->config.chip_select & 0x3) << 24;
+
+	/*
+	 * Set TCR_CONT will keep SS asserted after current transfer.
+	 * For the first transfer, clear TCR_CONTC to assert SS.
+	 * For subsequent transfer, set TCR_CONTC to keep SS asserted.
+	 */
+	temp |= TCR_CONT;
+	if (is_first_xfer)
+		temp &= ~TCR_CONTC;
+	else
+		temp |= TCR_CONTC;
+
+	writel(temp, fsl_lpspi->base + IMX7ULP_TCR);
+
+	dev_dbg(fsl_lpspi->dev, "TCR=0x%x\n", temp);
+}
+
+static void fsl_lpspi_set_watermark(struct fsl_lpspi_data *fsl_lpspi)
+{
+	u32 temp;
+
+	temp = fsl_lpspi->txfifosize >> 1 | (fsl_lpspi->rxfifosize >> 1) << 16;
+
+	writel(temp, fsl_lpspi->base + IMX7ULP_FCR);
+
+	dev_dbg(fsl_lpspi->dev, "FCR=0x%x\n", temp);
+}
+
+static int fsl_lpspi_set_bitrate(struct fsl_lpspi_data *fsl_lpspi)
+{
+	struct lpspi_config config = fsl_lpspi->config;
+	unsigned int perclk_rate, scldiv;
+	u8 prescale;
+
+	perclk_rate = clk_get_rate(fsl_lpspi->clk);
+	for (prescale = 0; prescale < 8; prescale++) {
+		scldiv = perclk_rate /
+			 (clkdivs[prescale] * config.speed_hz) - 2;
+		if (scldiv < 256) {
+			fsl_lpspi->config.prescale = prescale;
+			break;
+		}
+	}
+
+	if (prescale == 8 && scldiv >= 256)
+		return -EINVAL;
+
+	writel(scldiv, fsl_lpspi->base + IMX7ULP_CCR);
+
+	dev_dbg(fsl_lpspi->dev, "perclk=%d, speed=%d, prescale =%d, scldiv=%d\n",
+		perclk_rate, config.speed_hz, prescale, scldiv);
+
+	return 0;
+}
+
+static int fsl_lpspi_config(struct fsl_lpspi_data *fsl_lpspi)
+{
+	u32 temp;
+	int ret;
+
+	temp = CR_RST;
+	writel(temp, fsl_lpspi->base + IMX7ULP_CR);
+	writel(0, fsl_lpspi->base + IMX7ULP_CR);
+
+	ret = fsl_lpspi_set_bitrate(fsl_lpspi);
+	if (ret)
+		return ret;
+
+	fsl_lpspi_set_watermark(fsl_lpspi);
+
+	temp = CFGR1_PCSCFG | CFGR1_MASTER | CFGR1_NOSTALL;
+	if (fsl_lpspi->config.mode & SPI_CS_HIGH)
+		temp |= CFGR1_PCSPOL;
+	writel(temp, fsl_lpspi->base + IMX7ULP_CFGR1);
+
+	temp = readl(fsl_lpspi->base + IMX7ULP_CR);
+	temp |= CR_RRF | CR_RTF | CR_MEN;
+	writel(temp, fsl_lpspi->base + IMX7ULP_CR);
+
+	return 0;
+}
+
+static void fsl_lpspi_setup_transfer(struct spi_device *spi,
+				     struct spi_transfer *t)
+{
+	struct fsl_lpspi_data *fsl_lpspi = spi_master_get_devdata(spi->master);
+
+	fsl_lpspi->config.mode = spi->mode;
+	fsl_lpspi->config.bpw = t ? t->bits_per_word : spi->bits_per_word;
+	fsl_lpspi->config.speed_hz = t ? t->speed_hz : spi->max_speed_hz;
+	fsl_lpspi->config.chip_select = spi->chip_select;
+
+	if (!fsl_lpspi->config.speed_hz)
+		fsl_lpspi->config.speed_hz = spi->max_speed_hz;
+	if (!fsl_lpspi->config.bpw)
+		fsl_lpspi->config.bpw = spi->bits_per_word;
+
+	/* Initialize the functions for transfer */
+	if (fsl_lpspi->config.bpw <= 8) {
+		fsl_lpspi->rx = fsl_lpspi_buf_rx_u8;
+		fsl_lpspi->tx = fsl_lpspi_buf_tx_u8;
+	} else if (fsl_lpspi->config.bpw <= 16) {
+		fsl_lpspi->rx = fsl_lpspi_buf_rx_u16;
+		fsl_lpspi->tx = fsl_lpspi_buf_tx_u16;
+	} else {
+		fsl_lpspi->rx = fsl_lpspi_buf_rx_u32;
+		fsl_lpspi->tx = fsl_lpspi_buf_tx_u32;
+	}
+
+	fsl_lpspi_config(fsl_lpspi);
+}
+
+static int fsl_lpspi_transfer_one(struct spi_master *master,
+				  struct spi_device *spi,
+				  struct spi_transfer *t)
+{
+	struct fsl_lpspi_data *fsl_lpspi = spi_master_get_devdata(master);
+	int ret;
+
+	fsl_lpspi->tx_buf = t->tx_buf;
+	fsl_lpspi->rx_buf = t->rx_buf;
+	fsl_lpspi->remain = t->len;
+
+	reinit_completion(&fsl_lpspi->xfer_done);
+	fsl_lpspi_write_tx_fifo(fsl_lpspi);
+
+	ret = wait_for_completion_timeout(&fsl_lpspi->xfer_done, HZ);
+	if (!ret) {
+		dev_dbg(fsl_lpspi->dev, "wait for completion timeout\n");
+		return -ETIMEDOUT;
+	}
+
+	ret = fsl_lpspi_txfifo_empty(fsl_lpspi);
+	if (ret)
+		return ret;
+
+	fsl_lpspi_read_rx_fifo(fsl_lpspi);
+
+	return 0;
+}
+
+static int fsl_lpspi_transfer_one_msg(struct spi_master *master,
+				      struct spi_message *msg)
+{
+	struct fsl_lpspi_data *fsl_lpspi = spi_master_get_devdata(master);
+	struct spi_device *spi = msg->spi;
+	struct spi_transfer *xfer;
+	bool is_first_xfer = true;
+	u32 temp;
+	int ret = 0;
+
+	msg->status = 0;
+	msg->actual_length = 0;
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		fsl_lpspi_setup_transfer(spi, xfer);
+		fsl_lpspi_set_cmd(fsl_lpspi, is_first_xfer);
+
+		is_first_xfer = false;
+
+		ret = fsl_lpspi_transfer_one(master, spi, xfer);
+		if (ret < 0)
+			goto complete;
+
+		msg->actual_length += xfer->len;
+	}
+
+complete:
+	/* de-assert SS, then finalize current message */
+	temp = readl(fsl_lpspi->base + IMX7ULP_TCR);
+	temp &= ~TCR_CONTC;
+	writel(temp, fsl_lpspi->base + IMX7ULP_TCR);
+
+	msg->status = ret;
+	spi_finalize_current_message(master);
+
+	return ret;
+}
+
+static irqreturn_t fsl_lpspi_isr(int irq, void *dev_id)
+{
+	struct fsl_lpspi_data *fsl_lpspi = dev_id;
+	u32 temp;
+
+	fsl_lpspi_intctrl(fsl_lpspi, 0);
+	temp = readl(fsl_lpspi->base + IMX7ULP_SR);
+
+	fsl_lpspi_read_rx_fifo(fsl_lpspi);
+
+	if (temp & SR_TDF) {
+		fsl_lpspi_write_tx_fifo(fsl_lpspi);
+
+		if (!fsl_lpspi->remain)
+			complete(&fsl_lpspi->xfer_done);
+
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_NONE;
+}
+
+static int fsl_lpspi_probe(struct platform_device *pdev)
+{
+	struct fsl_lpspi_data *fsl_lpspi;
+	struct spi_master *master;
+	struct resource *res;
+	int ret, irq;
+	u32 temp;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct fsl_lpspi_data));
+	if (!master)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, master);
+
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(8, 32);
+	master->bus_num = pdev->id;
+
+	fsl_lpspi = spi_master_get_devdata(master);
+	fsl_lpspi->dev = &pdev->dev;
+
+	master->transfer_one_message = fsl_lpspi_transfer_one_msg;
+	master->prepare_transfer_hardware = lpspi_prepare_xfer_hardware;
+	master->unprepare_transfer_hardware = lpspi_unprepare_xfer_hardware;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->flags = SPI_MASTER_MUST_RX | SPI_MASTER_MUST_TX;
+	master->dev.of_node = pdev->dev.of_node;
+	master->bus_num = pdev->id;
+
+	init_completion(&fsl_lpspi->xfer_done);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	fsl_lpspi->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(fsl_lpspi->base)) {
+		ret = PTR_ERR(fsl_lpspi->base);
+		goto out_master_put;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		ret = irq;
+		goto out_master_put;
+	}
+
+	ret = devm_request_irq(&pdev->dev, irq, fsl_lpspi_isr, 0,
+			       dev_name(&pdev->dev), fsl_lpspi);
+	if (ret) {
+		dev_err(&pdev->dev, "can't get irq%d: %d\n", irq, ret);
+		goto out_master_put;
+	}
+
+	fsl_lpspi->clk = devm_clk_get(&pdev->dev, "ipg");
+	if (IS_ERR(fsl_lpspi->clk)) {
+		ret = PTR_ERR(fsl_lpspi->clk);
+		goto out_master_put;
+	}
+
+	ret = clk_prepare_enable(fsl_lpspi->clk);
+	if (ret) {
+		dev_err(&pdev->dev, "can't enable lpspi clock, ret=%d\n", ret);
+		goto out_master_put;
+	}
+
+	temp = readl(fsl_lpspi->base + IMX7ULP_PARAM);
+	fsl_lpspi->txfifosize = 1 << (temp & 0x0f);
+	fsl_lpspi->rxfifosize = 1 << ((temp >> 8) & 0x0f);
+
+	clk_disable_unprepare(fsl_lpspi->clk);
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "spi_register_master error.\n");
+		goto out_master_put;
+	}
+
+	return 0;
+
+out_master_put:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int fsl_lpspi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct fsl_lpspi_data *fsl_lpspi = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(fsl_lpspi->clk);
+
+	return 0;
+}
+
+static struct platform_driver fsl_lpspi_driver = {
+	.driver = {
+		.name = DRIVER_NAME,
+		.of_match_table = fsl_lpspi_dt_ids,
+	},
+	.probe = fsl_lpspi_probe,
+	.remove = fsl_lpspi_remove,
+};
+module_platform_driver(fsl_lpspi_driver);
+
+MODULE_DESCRIPTION("LPSPI Master Controller driver");
+MODULE_AUTHOR("Gao Pan <pandy.gao@nxp.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-fsl-spi.c b/drivers/spi/spi-fsl-spi.c
new file mode 100644
index 0000000..8f2e978
--- /dev/null
+++ b/drivers/spi/spi-fsl-spi.c
@@ -0,0 +1,958 @@
+/*
+ * Freescale SPI controller driver.
+ *
+ * Maintainer: Kumar Gala
+ *
+ * Copyright (C) 2006 Polycom, Inc.
+ * Copyright 2010 Freescale Semiconductor, Inc.
+ *
+ * CPM SPI and QE buffer descriptors mode support:
+ * Copyright (c) 2009  MontaVista Software, Inc.
+ * Author: Anton Vorontsov <avorontsov@ru.mvista.com>
+ *
+ * GRLIB support:
+ * Copyright (c) 2012 Aeroflex Gaisler AB.
+ * Author: Andreas Larsson <andreas@gaisler.com>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/fsl_devices.h>
+#include <linux/gpio.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/of_gpio.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/types.h>
+
+#include "spi-fsl-lib.h"
+#include "spi-fsl-cpm.h"
+#include "spi-fsl-spi.h"
+
+#define TYPE_FSL	0
+#define TYPE_GRLIB	1
+
+struct fsl_spi_match_data {
+	int type;
+};
+
+static struct fsl_spi_match_data of_fsl_spi_fsl_config = {
+	.type = TYPE_FSL,
+};
+
+static struct fsl_spi_match_data of_fsl_spi_grlib_config = {
+	.type = TYPE_GRLIB,
+};
+
+static const struct of_device_id of_fsl_spi_match[] = {
+	{
+		.compatible = "fsl,spi",
+		.data = &of_fsl_spi_fsl_config,
+	},
+	{
+		.compatible = "aeroflexgaisler,spictrl",
+		.data = &of_fsl_spi_grlib_config,
+	},
+	{}
+};
+MODULE_DEVICE_TABLE(of, of_fsl_spi_match);
+
+static int fsl_spi_get_type(struct device *dev)
+{
+	const struct of_device_id *match;
+
+	if (dev->of_node) {
+		match = of_match_node(of_fsl_spi_match, dev->of_node);
+		if (match && match->data)
+			return ((struct fsl_spi_match_data *)match->data)->type;
+	}
+	return TYPE_FSL;
+}
+
+static void fsl_spi_change_mode(struct spi_device *spi)
+{
+	struct mpc8xxx_spi *mspi = spi_master_get_devdata(spi->master);
+	struct spi_mpc8xxx_cs *cs = spi->controller_state;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
+	__be32 __iomem *mode = &reg_base->mode;
+	unsigned long flags;
+
+	if (cs->hw_mode == mpc8xxx_spi_read_reg(mode))
+		return;
+
+	/* Turn off IRQs locally to minimize time that SPI is disabled. */
+	local_irq_save(flags);
+
+	/* Turn off SPI unit prior changing mode */
+	mpc8xxx_spi_write_reg(mode, cs->hw_mode & ~SPMODE_ENABLE);
+
+	/* When in CPM mode, we need to reinit tx and rx. */
+	if (mspi->flags & SPI_CPM_MODE) {
+		fsl_spi_cpm_reinit_txrx(mspi);
+	}
+	mpc8xxx_spi_write_reg(mode, cs->hw_mode);
+	local_irq_restore(flags);
+}
+
+static void fsl_spi_chipselect(struct spi_device *spi, int value)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(spi->master);
+	struct fsl_spi_platform_data *pdata;
+	bool pol = spi->mode & SPI_CS_HIGH;
+	struct spi_mpc8xxx_cs	*cs = spi->controller_state;
+
+	pdata = spi->dev.parent->parent->platform_data;
+
+	if (value == BITBANG_CS_INACTIVE) {
+		if (pdata->cs_control)
+			pdata->cs_control(spi, !pol);
+	}
+
+	if (value == BITBANG_CS_ACTIVE) {
+		mpc8xxx_spi->rx_shift = cs->rx_shift;
+		mpc8xxx_spi->tx_shift = cs->tx_shift;
+		mpc8xxx_spi->get_rx = cs->get_rx;
+		mpc8xxx_spi->get_tx = cs->get_tx;
+
+		fsl_spi_change_mode(spi);
+
+		if (pdata->cs_control)
+			pdata->cs_control(spi, pol);
+	}
+}
+
+static void fsl_spi_qe_cpu_set_shifts(u32 *rx_shift, u32 *tx_shift,
+				      int bits_per_word, int msb_first)
+{
+	*rx_shift = 0;
+	*tx_shift = 0;
+	if (msb_first) {
+		if (bits_per_word <= 8) {
+			*rx_shift = 16;
+			*tx_shift = 24;
+		} else if (bits_per_word <= 16) {
+			*rx_shift = 16;
+			*tx_shift = 16;
+		}
+	} else {
+		if (bits_per_word <= 8)
+			*rx_shift = 8;
+	}
+}
+
+static void fsl_spi_grlib_set_shifts(u32 *rx_shift, u32 *tx_shift,
+				     int bits_per_word, int msb_first)
+{
+	*rx_shift = 0;
+	*tx_shift = 0;
+	if (bits_per_word <= 16) {
+		if (msb_first) {
+			*rx_shift = 16; /* LSB in bit 16 */
+			*tx_shift = 32 - bits_per_word; /* MSB in bit 31 */
+		} else {
+			*rx_shift = 16 - bits_per_word; /* MSB in bit 15 */
+		}
+	}
+}
+
+static int mspi_apply_cpu_mode_quirks(struct spi_mpc8xxx_cs *cs,
+				struct spi_device *spi,
+				struct mpc8xxx_spi *mpc8xxx_spi,
+				int bits_per_word)
+{
+	cs->rx_shift = 0;
+	cs->tx_shift = 0;
+	if (bits_per_word <= 8) {
+		cs->get_rx = mpc8xxx_spi_rx_buf_u8;
+		cs->get_tx = mpc8xxx_spi_tx_buf_u8;
+	} else if (bits_per_word <= 16) {
+		cs->get_rx = mpc8xxx_spi_rx_buf_u16;
+		cs->get_tx = mpc8xxx_spi_tx_buf_u16;
+	} else if (bits_per_word <= 32) {
+		cs->get_rx = mpc8xxx_spi_rx_buf_u32;
+		cs->get_tx = mpc8xxx_spi_tx_buf_u32;
+	} else
+		return -EINVAL;
+
+	if (mpc8xxx_spi->set_shifts)
+		mpc8xxx_spi->set_shifts(&cs->rx_shift, &cs->tx_shift,
+					bits_per_word,
+					!(spi->mode & SPI_LSB_FIRST));
+
+	mpc8xxx_spi->rx_shift = cs->rx_shift;
+	mpc8xxx_spi->tx_shift = cs->tx_shift;
+	mpc8xxx_spi->get_rx = cs->get_rx;
+	mpc8xxx_spi->get_tx = cs->get_tx;
+
+	return bits_per_word;
+}
+
+static int mspi_apply_qe_mode_quirks(struct spi_mpc8xxx_cs *cs,
+				struct spi_device *spi,
+				int bits_per_word)
+{
+	/* QE uses Little Endian for words > 8
+	 * so transform all words > 8 into 8 bits
+	 * Unfortnatly that doesn't work for LSB so
+	 * reject these for now */
+	/* Note: 32 bits word, LSB works iff
+	 * tfcr/rfcr is set to CPMFCR_GBL */
+	if (spi->mode & SPI_LSB_FIRST &&
+	    bits_per_word > 8)
+		return -EINVAL;
+	if (bits_per_word > 8)
+		return 8; /* pretend its 8 bits */
+	return bits_per_word;
+}
+
+static int fsl_spi_setup_transfer(struct spi_device *spi,
+					struct spi_transfer *t)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi;
+	int bits_per_word = 0;
+	u8 pm;
+	u32 hz = 0;
+	struct spi_mpc8xxx_cs	*cs = spi->controller_state;
+
+	mpc8xxx_spi = spi_master_get_devdata(spi->master);
+
+	if (t) {
+		bits_per_word = t->bits_per_word;
+		hz = t->speed_hz;
+	}
+
+	/* spi_transfer level calls that work per-word */
+	if (!bits_per_word)
+		bits_per_word = spi->bits_per_word;
+
+	if (!hz)
+		hz = spi->max_speed_hz;
+
+	if (!(mpc8xxx_spi->flags & SPI_CPM_MODE))
+		bits_per_word = mspi_apply_cpu_mode_quirks(cs, spi,
+							   mpc8xxx_spi,
+							   bits_per_word);
+	else if (mpc8xxx_spi->flags & SPI_QE)
+		bits_per_word = mspi_apply_qe_mode_quirks(cs, spi,
+							  bits_per_word);
+
+	if (bits_per_word < 0)
+		return bits_per_word;
+
+	if (bits_per_word == 32)
+		bits_per_word = 0;
+	else
+		bits_per_word = bits_per_word - 1;
+
+	/* mask out bits we are going to set */
+	cs->hw_mode &= ~(SPMODE_LEN(0xF) | SPMODE_DIV16
+				  | SPMODE_PM(0xF));
+
+	cs->hw_mode |= SPMODE_LEN(bits_per_word);
+
+	if ((mpc8xxx_spi->spibrg / hz) > 64) {
+		cs->hw_mode |= SPMODE_DIV16;
+		pm = (mpc8xxx_spi->spibrg - 1) / (hz * 64) + 1;
+		WARN_ONCE(pm > 16,
+			  "%s: Requested speed is too low: %d Hz. Will use %d Hz instead.\n",
+			  dev_name(&spi->dev), hz, mpc8xxx_spi->spibrg / 1024);
+		if (pm > 16)
+			pm = 16;
+	} else {
+		pm = (mpc8xxx_spi->spibrg - 1) / (hz * 4) + 1;
+	}
+	if (pm)
+		pm--;
+
+	cs->hw_mode |= SPMODE_PM(pm);
+
+	fsl_spi_change_mode(spi);
+	return 0;
+}
+
+static int fsl_spi_cpu_bufs(struct mpc8xxx_spi *mspi,
+				struct spi_transfer *t, unsigned int len)
+{
+	u32 word;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
+
+	mspi->count = len;
+
+	/* enable rx ints */
+	mpc8xxx_spi_write_reg(&reg_base->mask, SPIM_NE);
+
+	/* transmit word */
+	word = mspi->get_tx(mspi);
+	mpc8xxx_spi_write_reg(&reg_base->transmit, word);
+
+	return 0;
+}
+
+static int fsl_spi_bufs(struct spi_device *spi, struct spi_transfer *t,
+			    bool is_dma_mapped)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(spi->master);
+	struct fsl_spi_reg *reg_base;
+	unsigned int len = t->len;
+	u8 bits_per_word;
+	int ret;
+
+	reg_base = mpc8xxx_spi->reg_base;
+	bits_per_word = spi->bits_per_word;
+	if (t->bits_per_word)
+		bits_per_word = t->bits_per_word;
+
+	if (bits_per_word > 8) {
+		/* invalid length? */
+		if (len & 1)
+			return -EINVAL;
+		len /= 2;
+	}
+	if (bits_per_word > 16) {
+		/* invalid length? */
+		if (len & 1)
+			return -EINVAL;
+		len /= 2;
+	}
+
+	mpc8xxx_spi->tx = t->tx_buf;
+	mpc8xxx_spi->rx = t->rx_buf;
+
+	reinit_completion(&mpc8xxx_spi->done);
+
+	if (mpc8xxx_spi->flags & SPI_CPM_MODE)
+		ret = fsl_spi_cpm_bufs(mpc8xxx_spi, t, is_dma_mapped);
+	else
+		ret = fsl_spi_cpu_bufs(mpc8xxx_spi, t, len);
+	if (ret)
+		return ret;
+
+	wait_for_completion(&mpc8xxx_spi->done);
+
+	/* disable rx ints */
+	mpc8xxx_spi_write_reg(&reg_base->mask, 0);
+
+	if (mpc8xxx_spi->flags & SPI_CPM_MODE)
+		fsl_spi_cpm_bufs_complete(mpc8xxx_spi);
+
+	return mpc8xxx_spi->count;
+}
+
+static int fsl_spi_do_one_msg(struct spi_master *master,
+			      struct spi_message *m)
+{
+	struct spi_device *spi = m->spi;
+	struct spi_transfer *t, *first;
+	unsigned int cs_change;
+	const int nsecs = 50;
+	int status;
+
+	/* Don't allow changes if CS is active */
+	first = list_first_entry(&m->transfers, struct spi_transfer,
+			transfer_list);
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if ((first->bits_per_word != t->bits_per_word) ||
+			(first->speed_hz != t->speed_hz)) {
+			dev_err(&spi->dev,
+				"bits_per_word/speed_hz should be same for the same SPI transfer\n");
+			return -EINVAL;
+		}
+	}
+
+	cs_change = 1;
+	status = -EINVAL;
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if (t->bits_per_word || t->speed_hz) {
+			if (cs_change)
+				status = fsl_spi_setup_transfer(spi, t);
+			if (status < 0)
+				break;
+		}
+
+		if (cs_change) {
+			fsl_spi_chipselect(spi, BITBANG_CS_ACTIVE);
+			ndelay(nsecs);
+		}
+		cs_change = t->cs_change;
+		if (t->len)
+			status = fsl_spi_bufs(spi, t, m->is_dma_mapped);
+		if (status) {
+			status = -EMSGSIZE;
+			break;
+		}
+		m->actual_length += t->len;
+
+		if (t->delay_usecs)
+			udelay(t->delay_usecs);
+
+		if (cs_change) {
+			ndelay(nsecs);
+			fsl_spi_chipselect(spi, BITBANG_CS_INACTIVE);
+			ndelay(nsecs);
+		}
+	}
+
+	m->status = status;
+	spi_finalize_current_message(master);
+
+	if (status || !cs_change) {
+		ndelay(nsecs);
+		fsl_spi_chipselect(spi, BITBANG_CS_INACTIVE);
+	}
+
+	fsl_spi_setup_transfer(spi, NULL);
+	return 0;
+}
+
+static int fsl_spi_setup(struct spi_device *spi)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi;
+	struct fsl_spi_reg *reg_base;
+	int retval;
+	u32 hw_mode;
+	struct spi_mpc8xxx_cs *cs = spi_get_ctldata(spi);
+
+	if (!spi->max_speed_hz)
+		return -EINVAL;
+
+	if (!cs) {
+		cs = kzalloc(sizeof(*cs), GFP_KERNEL);
+		if (!cs)
+			return -ENOMEM;
+		spi_set_ctldata(spi, cs);
+	}
+	mpc8xxx_spi = spi_master_get_devdata(spi->master);
+
+	reg_base = mpc8xxx_spi->reg_base;
+
+	hw_mode = cs->hw_mode; /* Save original settings */
+	cs->hw_mode = mpc8xxx_spi_read_reg(&reg_base->mode);
+	/* mask out bits we are going to set */
+	cs->hw_mode &= ~(SPMODE_CP_BEGIN_EDGECLK | SPMODE_CI_INACTIVEHIGH
+			 | SPMODE_REV | SPMODE_LOOP);
+
+	if (spi->mode & SPI_CPHA)
+		cs->hw_mode |= SPMODE_CP_BEGIN_EDGECLK;
+	if (spi->mode & SPI_CPOL)
+		cs->hw_mode |= SPMODE_CI_INACTIVEHIGH;
+	if (!(spi->mode & SPI_LSB_FIRST))
+		cs->hw_mode |= SPMODE_REV;
+	if (spi->mode & SPI_LOOP)
+		cs->hw_mode |= SPMODE_LOOP;
+
+	retval = fsl_spi_setup_transfer(spi, NULL);
+	if (retval < 0) {
+		cs->hw_mode = hw_mode; /* Restore settings */
+		return retval;
+	}
+
+	if (mpc8xxx_spi->type == TYPE_GRLIB) {
+		if (gpio_is_valid(spi->cs_gpio)) {
+			int desel;
+
+			retval = gpio_request(spi->cs_gpio,
+					      dev_name(&spi->dev));
+			if (retval)
+				return retval;
+
+			desel = !(spi->mode & SPI_CS_HIGH);
+			retval = gpio_direction_output(spi->cs_gpio, desel);
+			if (retval) {
+				gpio_free(spi->cs_gpio);
+				return retval;
+			}
+		} else if (spi->cs_gpio != -ENOENT) {
+			if (spi->cs_gpio < 0)
+				return spi->cs_gpio;
+			return -EINVAL;
+		}
+		/* When spi->cs_gpio == -ENOENT, a hole in the phandle list
+		 * indicates to use native chipselect if present, or allow for
+		 * an always selected chip
+		 */
+	}
+
+	/* Initialize chipselect - might be active for SPI_CS_HIGH mode */
+	fsl_spi_chipselect(spi, BITBANG_CS_INACTIVE);
+
+	return 0;
+}
+
+static void fsl_spi_cleanup(struct spi_device *spi)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(spi->master);
+	struct spi_mpc8xxx_cs *cs = spi_get_ctldata(spi);
+
+	if (mpc8xxx_spi->type == TYPE_GRLIB && gpio_is_valid(spi->cs_gpio))
+		gpio_free(spi->cs_gpio);
+
+	kfree(cs);
+	spi_set_ctldata(spi, NULL);
+}
+
+static void fsl_spi_cpu_irq(struct mpc8xxx_spi *mspi, u32 events)
+{
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
+
+	/* We need handle RX first */
+	if (events & SPIE_NE) {
+		u32 rx_data = mpc8xxx_spi_read_reg(&reg_base->receive);
+
+		if (mspi->rx)
+			mspi->get_rx(rx_data, mspi);
+	}
+
+	if ((events & SPIE_NF) == 0)
+		/* spin until TX is done */
+		while (((events =
+			mpc8xxx_spi_read_reg(&reg_base->event)) &
+						SPIE_NF) == 0)
+			cpu_relax();
+
+	/* Clear the events */
+	mpc8xxx_spi_write_reg(&reg_base->event, events);
+
+	mspi->count -= 1;
+	if (mspi->count) {
+		u32 word = mspi->get_tx(mspi);
+
+		mpc8xxx_spi_write_reg(&reg_base->transmit, word);
+	} else {
+		complete(&mspi->done);
+	}
+}
+
+static irqreturn_t fsl_spi_irq(s32 irq, void *context_data)
+{
+	struct mpc8xxx_spi *mspi = context_data;
+	irqreturn_t ret = IRQ_NONE;
+	u32 events;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
+
+	/* Get interrupt events(tx/rx) */
+	events = mpc8xxx_spi_read_reg(&reg_base->event);
+	if (events)
+		ret = IRQ_HANDLED;
+
+	dev_dbg(mspi->dev, "%s: events %x\n", __func__, events);
+
+	if (mspi->flags & SPI_CPM_MODE)
+		fsl_spi_cpm_irq(mspi, events);
+	else
+		fsl_spi_cpu_irq(mspi, events);
+
+	return ret;
+}
+
+static void fsl_spi_grlib_cs_control(struct spi_device *spi, bool on)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(spi->master);
+	struct fsl_spi_reg *reg_base = mpc8xxx_spi->reg_base;
+	u32 slvsel;
+	u16 cs = spi->chip_select;
+
+	if (gpio_is_valid(spi->cs_gpio)) {
+		gpio_set_value(spi->cs_gpio, on);
+	} else if (cs < mpc8xxx_spi->native_chipselects) {
+		slvsel = mpc8xxx_spi_read_reg(&reg_base->slvsel);
+		slvsel = on ? (slvsel | (1 << cs)) : (slvsel & ~(1 << cs));
+		mpc8xxx_spi_write_reg(&reg_base->slvsel, slvsel);
+	}
+}
+
+static void fsl_spi_grlib_probe(struct device *dev)
+{
+	struct fsl_spi_platform_data *pdata = dev_get_platdata(dev);
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(master);
+	struct fsl_spi_reg *reg_base = mpc8xxx_spi->reg_base;
+	int mbits;
+	u32 capabilities;
+
+	capabilities = mpc8xxx_spi_read_reg(&reg_base->cap);
+
+	mpc8xxx_spi->set_shifts = fsl_spi_grlib_set_shifts;
+	mbits = SPCAP_MAXWLEN(capabilities);
+	if (mbits)
+		mpc8xxx_spi->max_bits_per_word = mbits + 1;
+
+	mpc8xxx_spi->native_chipselects = 0;
+	if (SPCAP_SSEN(capabilities)) {
+		mpc8xxx_spi->native_chipselects = SPCAP_SSSZ(capabilities);
+		mpc8xxx_spi_write_reg(&reg_base->slvsel, 0xffffffff);
+	}
+	master->num_chipselect = mpc8xxx_spi->native_chipselects;
+	pdata->cs_control = fsl_spi_grlib_cs_control;
+}
+
+static struct spi_master * fsl_spi_probe(struct device *dev,
+		struct resource *mem, unsigned int irq)
+{
+	struct fsl_spi_platform_data *pdata = dev_get_platdata(dev);
+	struct spi_master *master;
+	struct mpc8xxx_spi *mpc8xxx_spi;
+	struct fsl_spi_reg *reg_base;
+	u32 regval;
+	int ret = 0;
+
+	master = spi_alloc_master(dev, sizeof(struct mpc8xxx_spi));
+	if (master == NULL) {
+		ret = -ENOMEM;
+		goto err;
+	}
+
+	dev_set_drvdata(dev, master);
+
+	mpc8xxx_spi_probe(dev, mem, irq);
+
+	master->setup = fsl_spi_setup;
+	master->cleanup = fsl_spi_cleanup;
+	master->transfer_one_message = fsl_spi_do_one_msg;
+
+	mpc8xxx_spi = spi_master_get_devdata(master);
+	mpc8xxx_spi->max_bits_per_word = 32;
+	mpc8xxx_spi->type = fsl_spi_get_type(dev);
+
+	ret = fsl_spi_cpm_init(mpc8xxx_spi);
+	if (ret)
+		goto err_cpm_init;
+
+	mpc8xxx_spi->reg_base = devm_ioremap_resource(dev, mem);
+	if (IS_ERR(mpc8xxx_spi->reg_base)) {
+		ret = PTR_ERR(mpc8xxx_spi->reg_base);
+		goto err_probe;
+	}
+
+	if (mpc8xxx_spi->type == TYPE_GRLIB)
+		fsl_spi_grlib_probe(dev);
+
+	master->bits_per_word_mask =
+		(SPI_BPW_RANGE_MASK(4, 16) | SPI_BPW_MASK(32)) &
+		SPI_BPW_RANGE_MASK(1, mpc8xxx_spi->max_bits_per_word);
+
+	if (mpc8xxx_spi->flags & SPI_QE_CPU_MODE)
+		mpc8xxx_spi->set_shifts = fsl_spi_qe_cpu_set_shifts;
+
+	if (mpc8xxx_spi->set_shifts)
+		/* 8 bits per word and MSB first */
+		mpc8xxx_spi->set_shifts(&mpc8xxx_spi->rx_shift,
+					&mpc8xxx_spi->tx_shift, 8, 1);
+
+	/* Register for SPI Interrupt */
+	ret = devm_request_irq(dev, mpc8xxx_spi->irq, fsl_spi_irq,
+			       0, "fsl_spi", mpc8xxx_spi);
+
+	if (ret != 0)
+		goto err_probe;
+
+	reg_base = mpc8xxx_spi->reg_base;
+
+	/* SPI controller initializations */
+	mpc8xxx_spi_write_reg(&reg_base->mode, 0);
+	mpc8xxx_spi_write_reg(&reg_base->mask, 0);
+	mpc8xxx_spi_write_reg(&reg_base->command, 0);
+	mpc8xxx_spi_write_reg(&reg_base->event, 0xffffffff);
+
+	/* Enable SPI interface */
+	regval = pdata->initial_spmode | SPMODE_INIT_VAL | SPMODE_ENABLE;
+	if (mpc8xxx_spi->max_bits_per_word < 8) {
+		regval &= ~SPMODE_LEN(0xF);
+		regval |= SPMODE_LEN(mpc8xxx_spi->max_bits_per_word - 1);
+	}
+	if (mpc8xxx_spi->flags & SPI_QE_CPU_MODE)
+		regval |= SPMODE_OP;
+
+	mpc8xxx_spi_write_reg(&reg_base->mode, regval);
+
+	ret = devm_spi_register_master(dev, master);
+	if (ret < 0)
+		goto err_probe;
+
+	dev_info(dev, "at 0x%p (irq = %d), %s mode\n", reg_base,
+		 mpc8xxx_spi->irq, mpc8xxx_spi_strmode(mpc8xxx_spi->flags));
+
+	return master;
+
+err_probe:
+	fsl_spi_cpm_free(mpc8xxx_spi);
+err_cpm_init:
+	spi_master_put(master);
+err:
+	return ERR_PTR(ret);
+}
+
+static void fsl_spi_cs_control(struct spi_device *spi, bool on)
+{
+	struct device *dev = spi->dev.parent->parent;
+	struct fsl_spi_platform_data *pdata = dev_get_platdata(dev);
+	struct mpc8xxx_spi_probe_info *pinfo = to_of_pinfo(pdata);
+	u16 cs = spi->chip_select;
+	int gpio = pinfo->gpios[cs];
+	bool alow = pinfo->alow_flags[cs];
+
+	gpio_set_value(gpio, on ^ alow);
+}
+
+static int of_fsl_spi_get_chipselects(struct device *dev)
+{
+	struct device_node *np = dev->of_node;
+	struct fsl_spi_platform_data *pdata = dev_get_platdata(dev);
+	struct mpc8xxx_spi_probe_info *pinfo = to_of_pinfo(pdata);
+	int ngpios;
+	int i = 0;
+	int ret;
+
+	ngpios = of_gpio_count(np);
+	if (ngpios <= 0) {
+		/*
+		 * SPI w/o chip-select line. One SPI device is still permitted
+		 * though.
+		 */
+		pdata->max_chipselect = 1;
+		return 0;
+	}
+
+	pinfo->gpios = kmalloc_array(ngpios, sizeof(*pinfo->gpios),
+				     GFP_KERNEL);
+	if (!pinfo->gpios)
+		return -ENOMEM;
+	memset(pinfo->gpios, -1, ngpios * sizeof(*pinfo->gpios));
+
+	pinfo->alow_flags = kcalloc(ngpios, sizeof(*pinfo->alow_flags),
+				    GFP_KERNEL);
+	if (!pinfo->alow_flags) {
+		ret = -ENOMEM;
+		goto err_alloc_flags;
+	}
+
+	for (; i < ngpios; i++) {
+		int gpio;
+		enum of_gpio_flags flags;
+
+		gpio = of_get_gpio_flags(np, i, &flags);
+		if (!gpio_is_valid(gpio)) {
+			dev_err(dev, "invalid gpio #%d: %d\n", i, gpio);
+			ret = gpio;
+			goto err_loop;
+		}
+
+		ret = gpio_request(gpio, dev_name(dev));
+		if (ret) {
+			dev_err(dev, "can't request gpio #%d: %d\n", i, ret);
+			goto err_loop;
+		}
+
+		pinfo->gpios[i] = gpio;
+		pinfo->alow_flags[i] = flags & OF_GPIO_ACTIVE_LOW;
+
+		ret = gpio_direction_output(pinfo->gpios[i],
+					    pinfo->alow_flags[i]);
+		if (ret) {
+			dev_err(dev,
+				"can't set output direction for gpio #%d: %d\n",
+				i, ret);
+			goto err_loop;
+		}
+	}
+
+	pdata->max_chipselect = ngpios;
+	pdata->cs_control = fsl_spi_cs_control;
+
+	return 0;
+
+err_loop:
+	while (i >= 0) {
+		if (gpio_is_valid(pinfo->gpios[i]))
+			gpio_free(pinfo->gpios[i]);
+		i--;
+	}
+
+	kfree(pinfo->alow_flags);
+	pinfo->alow_flags = NULL;
+err_alloc_flags:
+	kfree(pinfo->gpios);
+	pinfo->gpios = NULL;
+	return ret;
+}
+
+static int of_fsl_spi_free_chipselects(struct device *dev)
+{
+	struct fsl_spi_platform_data *pdata = dev_get_platdata(dev);
+	struct mpc8xxx_spi_probe_info *pinfo = to_of_pinfo(pdata);
+	int i;
+
+	if (!pinfo->gpios)
+		return 0;
+
+	for (i = 0; i < pdata->max_chipselect; i++) {
+		if (gpio_is_valid(pinfo->gpios[i]))
+			gpio_free(pinfo->gpios[i]);
+	}
+
+	kfree(pinfo->gpios);
+	kfree(pinfo->alow_flags);
+	return 0;
+}
+
+static int of_fsl_spi_probe(struct platform_device *ofdev)
+{
+	struct device *dev = &ofdev->dev;
+	struct device_node *np = ofdev->dev.of_node;
+	struct spi_master *master;
+	struct resource mem;
+	int irq = 0, type;
+	int ret = -ENOMEM;
+
+	ret = of_mpc8xxx_spi_probe(ofdev);
+	if (ret)
+		return ret;
+
+	type = fsl_spi_get_type(&ofdev->dev);
+	if (type == TYPE_FSL) {
+		ret = of_fsl_spi_get_chipselects(dev);
+		if (ret)
+			goto err;
+	}
+
+	ret = of_address_to_resource(np, 0, &mem);
+	if (ret)
+		goto err;
+
+	irq = irq_of_parse_and_map(np, 0);
+	if (!irq) {
+		ret = -EINVAL;
+		goto err;
+	}
+
+	master = fsl_spi_probe(dev, &mem, irq);
+	if (IS_ERR(master)) {
+		ret = PTR_ERR(master);
+		goto err;
+	}
+
+	return 0;
+
+err:
+	irq_dispose_mapping(irq);
+	if (type == TYPE_FSL)
+		of_fsl_spi_free_chipselects(dev);
+	return ret;
+}
+
+static int of_fsl_spi_remove(struct platform_device *ofdev)
+{
+	struct spi_master *master = platform_get_drvdata(ofdev);
+	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(master);
+
+	fsl_spi_cpm_free(mpc8xxx_spi);
+	if (mpc8xxx_spi->type == TYPE_FSL)
+		of_fsl_spi_free_chipselects(&ofdev->dev);
+	return 0;
+}
+
+static struct platform_driver of_fsl_spi_driver = {
+	.driver = {
+		.name = "fsl_spi",
+		.of_match_table = of_fsl_spi_match,
+	},
+	.probe		= of_fsl_spi_probe,
+	.remove		= of_fsl_spi_remove,
+};
+
+#ifdef CONFIG_MPC832x_RDB
+/*
+ * XXX XXX XXX
+ * This is "legacy" platform driver, was used by the MPC8323E-RDB boards
+ * only. The driver should go away soon, since newer MPC8323E-RDB's device
+ * tree can work with OpenFirmware driver. But for now we support old trees
+ * as well.
+ */
+static int plat_mpc8xxx_spi_probe(struct platform_device *pdev)
+{
+	struct resource *mem;
+	int irq;
+	struct spi_master *master;
+
+	if (!dev_get_platdata(&pdev->dev))
+		return -EINVAL;
+
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!mem)
+		return -EINVAL;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq <= 0)
+		return -EINVAL;
+
+	master = fsl_spi_probe(&pdev->dev, mem, irq);
+	return PTR_ERR_OR_ZERO(master);
+}
+
+static int plat_mpc8xxx_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(master);
+
+	fsl_spi_cpm_free(mpc8xxx_spi);
+
+	return 0;
+}
+
+MODULE_ALIAS("platform:mpc8xxx_spi");
+static struct platform_driver mpc8xxx_spi_driver = {
+	.probe = plat_mpc8xxx_spi_probe,
+	.remove = plat_mpc8xxx_spi_remove,
+	.driver = {
+		.name = "mpc8xxx_spi",
+	},
+};
+
+static bool legacy_driver_failed;
+
+static void __init legacy_driver_register(void)
+{
+	legacy_driver_failed = platform_driver_register(&mpc8xxx_spi_driver);
+}
+
+static void __exit legacy_driver_unregister(void)
+{
+	if (legacy_driver_failed)
+		return;
+	platform_driver_unregister(&mpc8xxx_spi_driver);
+}
+#else
+static void __init legacy_driver_register(void) {}
+static void __exit legacy_driver_unregister(void) {}
+#endif /* CONFIG_MPC832x_RDB */
+
+static int __init fsl_spi_init(void)
+{
+	legacy_driver_register();
+	return platform_driver_register(&of_fsl_spi_driver);
+}
+module_init(fsl_spi_init);
+
+static void __exit fsl_spi_exit(void)
+{
+	platform_driver_unregister(&of_fsl_spi_driver);
+	legacy_driver_unregister();
+}
+module_exit(fsl_spi_exit);
+
+MODULE_AUTHOR("Kumar Gala");
+MODULE_DESCRIPTION("Simple Freescale SPI Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-fsl-spi.h b/drivers/spi/spi-fsl-spi.h
new file mode 100644
index 0000000..9a6dae0
--- /dev/null
+++ b/drivers/spi/spi-fsl-spi.h
@@ -0,0 +1,72 @@
+/*
+ * Freescale SPI controller driver.
+ *
+ * Maintainer: Kumar Gala
+ *
+ * Copyright (C) 2006 Polycom, Inc.
+ * Copyright 2010 Freescale Semiconductor, Inc.
+ *
+ * CPM SPI and QE buffer descriptors mode support:
+ * Copyright (c) 2009  MontaVista Software, Inc.
+ * Author: Anton Vorontsov <avorontsov@ru.mvista.com>
+ *
+ * GRLIB support:
+ * Copyright (c) 2012 Aeroflex Gaisler AB.
+ * Author: Andreas Larsson <andreas@gaisler.com>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+
+#ifndef __SPI_FSL_SPI_H__
+#define __SPI_FSL_SPI_H__
+
+/* SPI Controller registers */
+struct fsl_spi_reg {
+	__be32 cap; /* TYPE_GRLIB specific */
+	u8 res1[0x1C];
+	__be32 mode;
+	__be32 event;
+	__be32 mask;
+	__be32 command;
+	__be32 transmit;
+	__be32 receive;
+	__be32 slvsel; /* TYPE_GRLIB specific */
+};
+
+/* SPI Controller mode register definitions */
+#define	SPMODE_LOOP		(1 << 30)
+#define	SPMODE_CI_INACTIVEHIGH	(1 << 29)
+#define	SPMODE_CP_BEGIN_EDGECLK	(1 << 28)
+#define	SPMODE_DIV16		(1 << 27)
+#define	SPMODE_REV		(1 << 26)
+#define	SPMODE_MS		(1 << 25)
+#define	SPMODE_ENABLE		(1 << 24)
+#define	SPMODE_LEN(x)		((x) << 20)
+#define	SPMODE_PM(x)		((x) << 16)
+#define	SPMODE_OP		(1 << 14)
+#define	SPMODE_CG(x)		((x) << 7)
+
+/* TYPE_GRLIB SPI Controller capability register definitions */
+#define SPCAP_SSEN(x)		(((x) >> 16) & 0x1)
+#define SPCAP_SSSZ(x)		(((x) >> 24) & 0xff)
+#define SPCAP_MAXWLEN(x)	(((x) >> 20) & 0xf)
+
+/*
+ * Default for SPI Mode:
+ *	SPI MODE 0 (inactive low, phase middle, MSB, 8-bit length, slow clk
+ */
+#define	SPMODE_INIT_VAL (SPMODE_CI_INACTIVEHIGH | SPMODE_DIV16 | SPMODE_REV | \
+			 SPMODE_MS | SPMODE_LEN(7) | SPMODE_PM(0xf))
+
+/* SPIE register values */
+#define	SPIE_NE		0x00000200	/* Not empty */
+#define	SPIE_NF		0x00000100	/* Not full */
+
+/* SPIM register values */
+#define	SPIM_NE		0x00000200	/* Not empty */
+#define	SPIM_NF		0x00000100	/* Not full */
+
+#endif /* __SPI_FSL_SPI_H__ */
diff --git a/drivers/spi/spi-gpio.c b/drivers/spi/spi-gpio.c
new file mode 100644
index 0000000..088772e
--- /dev/null
+++ b/drivers/spi/spi-gpio.c
@@ -0,0 +1,479 @@
+/*
+ * SPI master driver using generic bitbanged GPIO
+ *
+ * Copyright (C) 2006,2008 David Brownell
+ * Copyright (C) 2017 Linus Walleij
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/gpio/consumer.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/spi/spi_gpio.h>
+
+
+/*
+ * This bitbanging SPI master driver should help make systems usable
+ * when a native hardware SPI engine is not available, perhaps because
+ * its driver isn't yet working or because the I/O pins it requires
+ * are used for other purposes.
+ *
+ * platform_device->driver_data ... points to spi_gpio
+ *
+ * spi->controller_state ... reserved for bitbang framework code
+ * spi->controller_data ... holds chipselect GPIO
+ *
+ * spi->master->dev.driver_data ... points to spi_gpio->bitbang
+ */
+
+struct spi_gpio {
+	struct spi_bitbang		bitbang;
+	struct spi_gpio_platform_data	pdata;
+	struct platform_device		*pdev;
+	struct gpio_desc		*sck;
+	struct gpio_desc		*miso;
+	struct gpio_desc		*mosi;
+	struct gpio_desc		**cs_gpios;
+	bool				has_cs;
+};
+
+/*----------------------------------------------------------------------*/
+
+/*
+ * Because the overhead of going through four GPIO procedure calls
+ * per transferred bit can make performance a problem, this code
+ * is set up so that you can use it in either of two ways:
+ *
+ *   - The slow generic way:  set up platform_data to hold the GPIO
+ *     numbers used for MISO/MOSI/SCK, and issue procedure calls for
+ *     each of them.  This driver can handle several such busses.
+ *
+ *   - The quicker inlined way:  only helps with platform GPIO code
+ *     that inlines operations for constant GPIOs.  This can give
+ *     you tight (fast!) inner loops, but each such bus needs a
+ *     new driver.  You'll define a new C file, with Makefile and
+ *     Kconfig support; the C code can be a total of six lines:
+ *
+ *		#define DRIVER_NAME	"myboard_spi2"
+ *		#define	SPI_MISO_GPIO	119
+ *		#define	SPI_MOSI_GPIO	120
+ *		#define	SPI_SCK_GPIO	121
+ *		#define	SPI_N_CHIPSEL	4
+ *		#include "spi-gpio.c"
+ */
+
+#ifndef DRIVER_NAME
+#define DRIVER_NAME	"spi_gpio"
+
+#define GENERIC_BITBANG	/* vs tight inlines */
+
+#endif
+
+/*----------------------------------------------------------------------*/
+
+static inline struct spi_gpio *__pure
+spi_to_spi_gpio(const struct spi_device *spi)
+{
+	const struct spi_bitbang	*bang;
+	struct spi_gpio			*spi_gpio;
+
+	bang = spi_master_get_devdata(spi->master);
+	spi_gpio = container_of(bang, struct spi_gpio, bitbang);
+	return spi_gpio;
+}
+
+static inline struct spi_gpio_platform_data *__pure
+spi_to_pdata(const struct spi_device *spi)
+{
+	return &spi_to_spi_gpio(spi)->pdata;
+}
+
+/* These helpers are in turn called by the bitbang inlines */
+static inline void setsck(const struct spi_device *spi, int is_on)
+{
+	struct spi_gpio *spi_gpio = spi_to_spi_gpio(spi);
+
+	gpiod_set_value_cansleep(spi_gpio->sck, is_on);
+}
+
+static inline void setmosi(const struct spi_device *spi, int is_on)
+{
+	struct spi_gpio *spi_gpio = spi_to_spi_gpio(spi);
+
+	gpiod_set_value_cansleep(spi_gpio->mosi, is_on);
+}
+
+static inline int getmiso(const struct spi_device *spi)
+{
+	struct spi_gpio *spi_gpio = spi_to_spi_gpio(spi);
+
+	if (spi->mode & SPI_3WIRE)
+		return !!gpiod_get_value_cansleep(spi_gpio->mosi);
+	else
+		return !!gpiod_get_value_cansleep(spi_gpio->miso);
+}
+
+/*
+ * NOTE:  this clocks "as fast as we can".  It "should" be a function of the
+ * requested device clock.  Software overhead means we usually have trouble
+ * reaching even one Mbit/sec (except when we can inline bitops), so for now
+ * we'll just assume we never need additional per-bit slowdowns.
+ */
+#define spidelay(nsecs)	do {} while (0)
+
+#include "spi-bitbang-txrx.h"
+
+/*
+ * These functions can leverage inline expansion of GPIO calls to shrink
+ * costs for a txrx bit, often by factors of around ten (by instruction
+ * count).  That is particularly visible for larger word sizes, but helps
+ * even with default 8-bit words.
+ *
+ * REVISIT overheads calling these functions for each word also have
+ * significant performance costs.  Having txrx_bufs() calls that inline
+ * the txrx_word() logic would help performance, e.g. on larger blocks
+ * used with flash storage or MMC/SD.  There should also be ways to make
+ * GCC be less stupid about reloading registers inside the I/O loops,
+ * even without inlined GPIO calls; __attribute__((hot)) on GCC 4.3?
+ */
+
+static u32 spi_gpio_txrx_word_mode0(struct spi_device *spi,
+		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+{
+	return bitbang_txrx_be_cpha0(spi, nsecs, 0, flags, word, bits);
+}
+
+static u32 spi_gpio_txrx_word_mode1(struct spi_device *spi,
+		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+{
+	return bitbang_txrx_be_cpha1(spi, nsecs, 0, flags, word, bits);
+}
+
+static u32 spi_gpio_txrx_word_mode2(struct spi_device *spi,
+		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+{
+	return bitbang_txrx_be_cpha0(spi, nsecs, 1, flags, word, bits);
+}
+
+static u32 spi_gpio_txrx_word_mode3(struct spi_device *spi,
+		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+{
+	return bitbang_txrx_be_cpha1(spi, nsecs, 1, flags, word, bits);
+}
+
+/*
+ * These functions do not call setmosi or getmiso if respective flag
+ * (SPI_MASTER_NO_RX or SPI_MASTER_NO_TX) is set, so they are safe to
+ * call when such pin is not present or defined in the controller.
+ * A separate set of callbacks is defined to get highest possible
+ * speed in the generic case (when both MISO and MOSI lines are
+ * available), as optimiser will remove the checks when argument is
+ * constant.
+ */
+
+static u32 spi_gpio_spec_txrx_word_mode0(struct spi_device *spi,
+		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+{
+	flags = spi->master->flags;
+	return bitbang_txrx_be_cpha0(spi, nsecs, 0, flags, word, bits);
+}
+
+static u32 spi_gpio_spec_txrx_word_mode1(struct spi_device *spi,
+		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+{
+	flags = spi->master->flags;
+	return bitbang_txrx_be_cpha1(spi, nsecs, 0, flags, word, bits);
+}
+
+static u32 spi_gpio_spec_txrx_word_mode2(struct spi_device *spi,
+		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+{
+	flags = spi->master->flags;
+	return bitbang_txrx_be_cpha0(spi, nsecs, 1, flags, word, bits);
+}
+
+static u32 spi_gpio_spec_txrx_word_mode3(struct spi_device *spi,
+		unsigned nsecs, u32 word, u8 bits, unsigned flags)
+{
+	flags = spi->master->flags;
+	return bitbang_txrx_be_cpha1(spi, nsecs, 1, flags, word, bits);
+}
+
+/*----------------------------------------------------------------------*/
+
+static void spi_gpio_chipselect(struct spi_device *spi, int is_active)
+{
+	struct spi_gpio *spi_gpio = spi_to_spi_gpio(spi);
+
+	/* set initial clock line level */
+	if (is_active)
+		gpiod_set_value_cansleep(spi_gpio->sck, spi->mode & SPI_CPOL);
+
+	/* Drive chip select line, if we have one */
+	if (spi_gpio->has_cs) {
+		struct gpio_desc *cs = spi_gpio->cs_gpios[spi->chip_select];
+
+		/* SPI chip selects are normally active-low */
+		gpiod_set_value_cansleep(cs, (spi->mode & SPI_CS_HIGH) ? is_active : !is_active);
+	}
+}
+
+static int spi_gpio_setup(struct spi_device *spi)
+{
+	struct gpio_desc	*cs;
+	int			status = 0;
+	struct spi_gpio		*spi_gpio = spi_to_spi_gpio(spi);
+
+	/*
+	 * The CS GPIOs have already been
+	 * initialized from the descriptor lookup.
+	 */
+	cs = spi_gpio->cs_gpios[spi->chip_select];
+	if (!spi->controller_state && cs)
+		status = gpiod_direction_output(cs,
+						!(spi->mode & SPI_CS_HIGH));
+
+	if (!status)
+		status = spi_bitbang_setup(spi);
+
+	return status;
+}
+
+static int spi_gpio_set_direction(struct spi_device *spi, bool output)
+{
+	struct spi_gpio *spi_gpio = spi_to_spi_gpio(spi);
+
+	if (output)
+		return gpiod_direction_output(spi_gpio->mosi, 1);
+	else
+		return gpiod_direction_input(spi_gpio->mosi);
+}
+
+static void spi_gpio_cleanup(struct spi_device *spi)
+{
+	spi_bitbang_cleanup(spi);
+}
+
+/*
+ * It can be convenient to use this driver with pins that have alternate
+ * functions associated with a "native" SPI controller if a driver for that
+ * controller is not available, or is missing important functionality.
+ *
+ * On platforms which can do so, configure MISO with a weak pullup unless
+ * there's an external pullup on that signal.  That saves power by avoiding
+ * floating signals.  (A weak pulldown would save power too, but many
+ * drivers expect to see all-ones data as the no slave "response".)
+ */
+static int spi_gpio_request(struct device *dev,
+			    struct spi_gpio *spi_gpio,
+			    unsigned int num_chipselects,
+			    u16 *mflags)
+{
+	int i;
+
+	spi_gpio->mosi = devm_gpiod_get_optional(dev, "mosi", GPIOD_OUT_LOW);
+	if (IS_ERR(spi_gpio->mosi))
+		return PTR_ERR(spi_gpio->mosi);
+	if (!spi_gpio->mosi)
+		/* HW configuration without MOSI pin */
+		*mflags |= SPI_MASTER_NO_TX;
+
+	spi_gpio->miso = devm_gpiod_get_optional(dev, "miso", GPIOD_IN);
+	if (IS_ERR(spi_gpio->miso))
+		return PTR_ERR(spi_gpio->miso);
+	/*
+	 * No setting SPI_MASTER_NO_RX here - if there is only a MOSI
+	 * pin connected the host can still do RX by changing the
+	 * direction of the line.
+	 */
+
+	spi_gpio->sck = devm_gpiod_get(dev, "sck", GPIOD_OUT_LOW);
+	if (IS_ERR(spi_gpio->sck))
+		return PTR_ERR(spi_gpio->sck);
+
+	for (i = 0; i < num_chipselects; i++) {
+		spi_gpio->cs_gpios[i] = devm_gpiod_get_index(dev, "cs",
+							     i, GPIOD_OUT_HIGH);
+		if (IS_ERR(spi_gpio->cs_gpios[i]))
+			return PTR_ERR(spi_gpio->cs_gpios[i]);
+	}
+
+	return 0;
+}
+
+#ifdef CONFIG_OF
+static const struct of_device_id spi_gpio_dt_ids[] = {
+	{ .compatible = "spi-gpio" },
+	{}
+};
+MODULE_DEVICE_TABLE(of, spi_gpio_dt_ids);
+
+static int spi_gpio_probe_dt(struct platform_device *pdev)
+{
+	int ret;
+	u32 tmp;
+	struct spi_gpio_platform_data	*pdata;
+	struct device_node *np = pdev->dev.of_node;
+	const struct of_device_id *of_id =
+			of_match_device(spi_gpio_dt_ids, &pdev->dev);
+
+	if (!of_id)
+		return 0;
+
+	pdata = devm_kzalloc(&pdev->dev, sizeof(*pdata), GFP_KERNEL);
+	if (!pdata)
+		return -ENOMEM;
+
+
+	ret = of_property_read_u32(np, "num-chipselects", &tmp);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "num-chipselects property not found\n");
+		goto error_free;
+	}
+
+	pdata->num_chipselect = tmp;
+	pdev->dev.platform_data = pdata;
+
+	return 1;
+
+error_free:
+	devm_kfree(&pdev->dev, pdata);
+	return ret;
+}
+#else
+static inline int spi_gpio_probe_dt(struct platform_device *pdev)
+{
+	return 0;
+}
+#endif
+
+static int spi_gpio_probe(struct platform_device *pdev)
+{
+	int				status;
+	struct spi_master		*master;
+	struct spi_gpio			*spi_gpio;
+	struct spi_gpio_platform_data	*pdata;
+	u16 master_flags = 0;
+	bool use_of = 0;
+
+	status = spi_gpio_probe_dt(pdev);
+	if (status < 0)
+		return status;
+	if (status > 0)
+		use_of = 1;
+
+	pdata = dev_get_platdata(&pdev->dev);
+#ifdef GENERIC_BITBANG
+	if (!pdata || (!use_of && !pdata->num_chipselect))
+		return -ENODEV;
+#endif
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*spi_gpio));
+	if (!master)
+		return -ENOMEM;
+
+	spi_gpio = spi_master_get_devdata(master);
+
+	spi_gpio->cs_gpios = devm_kcalloc(&pdev->dev,
+				pdata->num_chipselect,
+				sizeof(*spi_gpio->cs_gpios),
+				GFP_KERNEL);
+	if (!spi_gpio->cs_gpios)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, spi_gpio);
+
+	/* Determine if we have chip selects connected */
+	spi_gpio->has_cs = !!pdata->num_chipselect;
+
+	spi_gpio->pdev = pdev;
+	if (pdata)
+		spi_gpio->pdata = *pdata;
+
+	status = spi_gpio_request(&pdev->dev, spi_gpio,
+				  pdata->num_chipselect, &master_flags);
+	if (status)
+		return status;
+
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(1, 32);
+	master->mode_bits = SPI_3WIRE | SPI_CPHA | SPI_CPOL;
+	master->flags = master_flags;
+	master->bus_num = pdev->id;
+	/* The master needs to think there is a chipselect even if not connected */
+	master->num_chipselect = spi_gpio->has_cs ? pdata->num_chipselect : 1;
+	master->setup = spi_gpio_setup;
+	master->cleanup = spi_gpio_cleanup;
+#ifdef CONFIG_OF
+	master->dev.of_node = pdev->dev.of_node;
+#endif
+
+	spi_gpio->bitbang.master = master;
+	spi_gpio->bitbang.chipselect = spi_gpio_chipselect;
+	spi_gpio->bitbang.set_line_direction = spi_gpio_set_direction;
+
+	if ((master_flags & SPI_MASTER_NO_TX) == 0) {
+		spi_gpio->bitbang.txrx_word[SPI_MODE_0] = spi_gpio_txrx_word_mode0;
+		spi_gpio->bitbang.txrx_word[SPI_MODE_1] = spi_gpio_txrx_word_mode1;
+		spi_gpio->bitbang.txrx_word[SPI_MODE_2] = spi_gpio_txrx_word_mode2;
+		spi_gpio->bitbang.txrx_word[SPI_MODE_3] = spi_gpio_txrx_word_mode3;
+	} else {
+		spi_gpio->bitbang.txrx_word[SPI_MODE_0] = spi_gpio_spec_txrx_word_mode0;
+		spi_gpio->bitbang.txrx_word[SPI_MODE_1] = spi_gpio_spec_txrx_word_mode1;
+		spi_gpio->bitbang.txrx_word[SPI_MODE_2] = spi_gpio_spec_txrx_word_mode2;
+		spi_gpio->bitbang.txrx_word[SPI_MODE_3] = spi_gpio_spec_txrx_word_mode3;
+	}
+	spi_gpio->bitbang.setup_transfer = spi_bitbang_setup_transfer;
+	spi_gpio->bitbang.flags = SPI_CS_HIGH;
+
+	status = spi_bitbang_start(&spi_gpio->bitbang);
+	if (status)
+		spi_master_put(master);
+
+	return status;
+}
+
+static int spi_gpio_remove(struct platform_device *pdev)
+{
+	struct spi_gpio			*spi_gpio;
+	struct spi_gpio_platform_data	*pdata;
+
+	spi_gpio = platform_get_drvdata(pdev);
+	pdata = dev_get_platdata(&pdev->dev);
+
+	/* stop() unregisters child devices too */
+	spi_bitbang_stop(&spi_gpio->bitbang);
+
+	spi_master_put(spi_gpio->bitbang.master);
+
+	return 0;
+}
+
+MODULE_ALIAS("platform:" DRIVER_NAME);
+
+static struct platform_driver spi_gpio_driver = {
+	.driver = {
+		.name	= DRIVER_NAME,
+		.of_match_table = of_match_ptr(spi_gpio_dt_ids),
+	},
+	.probe		= spi_gpio_probe,
+	.remove		= spi_gpio_remove,
+};
+module_platform_driver(spi_gpio_driver);
+
+MODULE_DESCRIPTION("SPI master driver using generic bitbanged GPIO ");
+MODULE_AUTHOR("David Brownell");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-img-spfi.c b/drivers/spi/spi-img-spfi.c
new file mode 100644
index 0000000..e6eb979
--- /dev/null
+++ b/drivers/spi/spi-img-spfi.c
@@ -0,0 +1,809 @@
+/*
+ * IMG SPFI controller driver
+ *
+ * Copyright (C) 2007,2008,2013 Imagination Technologies Ltd.
+ * Copyright (C) 2014 Google, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/dmaengine.h>
+#include <linux/gpio.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/scatterlist.h>
+#include <linux/slab.h>
+#include <linux/spi/spi.h>
+#include <linux/spinlock.h>
+
+#define SPFI_DEVICE_PARAMETER(x)		(0x00 + 0x4 * (x))
+#define SPFI_DEVICE_PARAMETER_BITCLK_SHIFT	24
+#define SPFI_DEVICE_PARAMETER_BITCLK_MASK	0xff
+#define SPFI_DEVICE_PARAMETER_CSSETUP_SHIFT	16
+#define SPFI_DEVICE_PARAMETER_CSSETUP_MASK	0xff
+#define SPFI_DEVICE_PARAMETER_CSHOLD_SHIFT	8
+#define SPFI_DEVICE_PARAMETER_CSHOLD_MASK	0xff
+#define SPFI_DEVICE_PARAMETER_CSDELAY_SHIFT	0
+#define SPFI_DEVICE_PARAMETER_CSDELAY_MASK	0xff
+
+#define SPFI_CONTROL				0x14
+#define SPFI_CONTROL_CONTINUE			BIT(12)
+#define SPFI_CONTROL_SOFT_RESET			BIT(11)
+#define SPFI_CONTROL_SEND_DMA			BIT(10)
+#define SPFI_CONTROL_GET_DMA			BIT(9)
+#define SPFI_CONTROL_SE			BIT(8)
+#define SPFI_CONTROL_TMODE_SHIFT		5
+#define SPFI_CONTROL_TMODE_MASK			0x7
+#define SPFI_CONTROL_TMODE_SINGLE		0
+#define SPFI_CONTROL_TMODE_DUAL			1
+#define SPFI_CONTROL_TMODE_QUAD			2
+#define SPFI_CONTROL_SPFI_EN			BIT(0)
+
+#define SPFI_TRANSACTION			0x18
+#define SPFI_TRANSACTION_TSIZE_SHIFT		16
+#define SPFI_TRANSACTION_TSIZE_MASK		0xffff
+
+#define SPFI_PORT_STATE				0x1c
+#define SPFI_PORT_STATE_DEV_SEL_SHIFT		20
+#define SPFI_PORT_STATE_DEV_SEL_MASK		0x7
+#define SPFI_PORT_STATE_CK_POL(x)		BIT(19 - (x))
+#define SPFI_PORT_STATE_CK_PHASE(x)		BIT(14 - (x))
+
+#define SPFI_TX_32BIT_VALID_DATA		0x20
+#define SPFI_TX_8BIT_VALID_DATA			0x24
+#define SPFI_RX_32BIT_VALID_DATA		0x28
+#define SPFI_RX_8BIT_VALID_DATA			0x2c
+
+#define SPFI_INTERRUPT_STATUS			0x30
+#define SPFI_INTERRUPT_ENABLE			0x34
+#define SPFI_INTERRUPT_CLEAR			0x38
+#define SPFI_INTERRUPT_IACCESS			BIT(12)
+#define SPFI_INTERRUPT_GDEX8BIT			BIT(11)
+#define SPFI_INTERRUPT_ALLDONETRIG		BIT(9)
+#define SPFI_INTERRUPT_GDFUL			BIT(8)
+#define SPFI_INTERRUPT_GDHF			BIT(7)
+#define SPFI_INTERRUPT_GDEX32BIT		BIT(6)
+#define SPFI_INTERRUPT_GDTRIG			BIT(5)
+#define SPFI_INTERRUPT_SDFUL			BIT(3)
+#define SPFI_INTERRUPT_SDHF			BIT(2)
+#define SPFI_INTERRUPT_SDE			BIT(1)
+#define SPFI_INTERRUPT_SDTRIG			BIT(0)
+
+/*
+ * There are four parallel FIFOs of 16 bytes each.  The word buffer
+ * (*_32BIT_VALID_DATA) accesses all four FIFOs at once, resulting in an
+ * effective FIFO size of 64 bytes.  The byte buffer (*_8BIT_VALID_DATA)
+ * accesses only a single FIFO, resulting in an effective FIFO size of
+ * 16 bytes.
+ */
+#define SPFI_32BIT_FIFO_SIZE			64
+#define SPFI_8BIT_FIFO_SIZE			16
+
+struct img_spfi {
+	struct device *dev;
+	struct spi_master *master;
+	spinlock_t lock;
+
+	void __iomem *regs;
+	phys_addr_t phys;
+	int irq;
+	struct clk *spfi_clk;
+	struct clk *sys_clk;
+
+	struct dma_chan *rx_ch;
+	struct dma_chan *tx_ch;
+	bool tx_dma_busy;
+	bool rx_dma_busy;
+};
+
+struct img_spfi_device_data {
+	bool gpio_requested;
+};
+
+static inline u32 spfi_readl(struct img_spfi *spfi, u32 reg)
+{
+	return readl(spfi->regs + reg);
+}
+
+static inline void spfi_writel(struct img_spfi *spfi, u32 val, u32 reg)
+{
+	writel(val, spfi->regs + reg);
+}
+
+static inline void spfi_start(struct img_spfi *spfi)
+{
+	u32 val;
+
+	val = spfi_readl(spfi, SPFI_CONTROL);
+	val |= SPFI_CONTROL_SPFI_EN;
+	spfi_writel(spfi, val, SPFI_CONTROL);
+}
+
+static inline void spfi_reset(struct img_spfi *spfi)
+{
+	spfi_writel(spfi, SPFI_CONTROL_SOFT_RESET, SPFI_CONTROL);
+	spfi_writel(spfi, 0, SPFI_CONTROL);
+}
+
+static int spfi_wait_all_done(struct img_spfi *spfi)
+{
+	unsigned long timeout = jiffies + msecs_to_jiffies(50);
+
+	while (time_before(jiffies, timeout)) {
+		u32 status = spfi_readl(spfi, SPFI_INTERRUPT_STATUS);
+
+		if (status & SPFI_INTERRUPT_ALLDONETRIG) {
+			spfi_writel(spfi, SPFI_INTERRUPT_ALLDONETRIG,
+				    SPFI_INTERRUPT_CLEAR);
+			return 0;
+		}
+		cpu_relax();
+	}
+
+	dev_err(spfi->dev, "Timed out waiting for transaction to complete\n");
+	spfi_reset(spfi);
+
+	return -ETIMEDOUT;
+}
+
+static unsigned int spfi_pio_write32(struct img_spfi *spfi, const u32 *buf,
+				     unsigned int max)
+{
+	unsigned int count = 0;
+	u32 status;
+
+	while (count < max / 4) {
+		spfi_writel(spfi, SPFI_INTERRUPT_SDFUL, SPFI_INTERRUPT_CLEAR);
+		status = spfi_readl(spfi, SPFI_INTERRUPT_STATUS);
+		if (status & SPFI_INTERRUPT_SDFUL)
+			break;
+		spfi_writel(spfi, buf[count], SPFI_TX_32BIT_VALID_DATA);
+		count++;
+	}
+
+	return count * 4;
+}
+
+static unsigned int spfi_pio_write8(struct img_spfi *spfi, const u8 *buf,
+				    unsigned int max)
+{
+	unsigned int count = 0;
+	u32 status;
+
+	while (count < max) {
+		spfi_writel(spfi, SPFI_INTERRUPT_SDFUL, SPFI_INTERRUPT_CLEAR);
+		status = spfi_readl(spfi, SPFI_INTERRUPT_STATUS);
+		if (status & SPFI_INTERRUPT_SDFUL)
+			break;
+		spfi_writel(spfi, buf[count], SPFI_TX_8BIT_VALID_DATA);
+		count++;
+	}
+
+	return count;
+}
+
+static unsigned int spfi_pio_read32(struct img_spfi *spfi, u32 *buf,
+				    unsigned int max)
+{
+	unsigned int count = 0;
+	u32 status;
+
+	while (count < max / 4) {
+		spfi_writel(spfi, SPFI_INTERRUPT_GDEX32BIT,
+			    SPFI_INTERRUPT_CLEAR);
+		status = spfi_readl(spfi, SPFI_INTERRUPT_STATUS);
+		if (!(status & SPFI_INTERRUPT_GDEX32BIT))
+			break;
+		buf[count] = spfi_readl(spfi, SPFI_RX_32BIT_VALID_DATA);
+		count++;
+	}
+
+	return count * 4;
+}
+
+static unsigned int spfi_pio_read8(struct img_spfi *spfi, u8 *buf,
+				   unsigned int max)
+{
+	unsigned int count = 0;
+	u32 status;
+
+	while (count < max) {
+		spfi_writel(spfi, SPFI_INTERRUPT_GDEX8BIT,
+			    SPFI_INTERRUPT_CLEAR);
+		status = spfi_readl(spfi, SPFI_INTERRUPT_STATUS);
+		if (!(status & SPFI_INTERRUPT_GDEX8BIT))
+			break;
+		buf[count] = spfi_readl(spfi, SPFI_RX_8BIT_VALID_DATA);
+		count++;
+	}
+
+	return count;
+}
+
+static int img_spfi_start_pio(struct spi_master *master,
+			       struct spi_device *spi,
+			       struct spi_transfer *xfer)
+{
+	struct img_spfi *spfi = spi_master_get_devdata(spi->master);
+	unsigned int tx_bytes = 0, rx_bytes = 0;
+	const void *tx_buf = xfer->tx_buf;
+	void *rx_buf = xfer->rx_buf;
+	unsigned long timeout;
+	int ret;
+
+	if (tx_buf)
+		tx_bytes = xfer->len;
+	if (rx_buf)
+		rx_bytes = xfer->len;
+
+	spfi_start(spfi);
+
+	timeout = jiffies +
+		msecs_to_jiffies(xfer->len * 8 * 1000 / xfer->speed_hz + 100);
+	while ((tx_bytes > 0 || rx_bytes > 0) &&
+	       time_before(jiffies, timeout)) {
+		unsigned int tx_count, rx_count;
+
+		if (tx_bytes >= 4)
+			tx_count = spfi_pio_write32(spfi, tx_buf, tx_bytes);
+		else
+			tx_count = spfi_pio_write8(spfi, tx_buf, tx_bytes);
+
+		if (rx_bytes >= 4)
+			rx_count = spfi_pio_read32(spfi, rx_buf, rx_bytes);
+		else
+			rx_count = spfi_pio_read8(spfi, rx_buf, rx_bytes);
+
+		tx_buf += tx_count;
+		rx_buf += rx_count;
+		tx_bytes -= tx_count;
+		rx_bytes -= rx_count;
+
+		cpu_relax();
+	}
+
+	if (rx_bytes > 0 || tx_bytes > 0) {
+		dev_err(spfi->dev, "PIO transfer timed out\n");
+		return -ETIMEDOUT;
+	}
+
+	ret = spfi_wait_all_done(spfi);
+	if (ret < 0)
+		return ret;
+
+	return 0;
+}
+
+static void img_spfi_dma_rx_cb(void *data)
+{
+	struct img_spfi *spfi = data;
+	unsigned long flags;
+
+	spfi_wait_all_done(spfi);
+
+	spin_lock_irqsave(&spfi->lock, flags);
+	spfi->rx_dma_busy = false;
+	if (!spfi->tx_dma_busy)
+		spi_finalize_current_transfer(spfi->master);
+	spin_unlock_irqrestore(&spfi->lock, flags);
+}
+
+static void img_spfi_dma_tx_cb(void *data)
+{
+	struct img_spfi *spfi = data;
+	unsigned long flags;
+
+	spfi_wait_all_done(spfi);
+
+	spin_lock_irqsave(&spfi->lock, flags);
+	spfi->tx_dma_busy = false;
+	if (!spfi->rx_dma_busy)
+		spi_finalize_current_transfer(spfi->master);
+	spin_unlock_irqrestore(&spfi->lock, flags);
+}
+
+static int img_spfi_start_dma(struct spi_master *master,
+			      struct spi_device *spi,
+			      struct spi_transfer *xfer)
+{
+	struct img_spfi *spfi = spi_master_get_devdata(spi->master);
+	struct dma_async_tx_descriptor *rxdesc = NULL, *txdesc = NULL;
+	struct dma_slave_config rxconf, txconf;
+
+	spfi->rx_dma_busy = false;
+	spfi->tx_dma_busy = false;
+
+	if (xfer->rx_buf) {
+		rxconf.direction = DMA_DEV_TO_MEM;
+		if (xfer->len % 4 == 0) {
+			rxconf.src_addr = spfi->phys + SPFI_RX_32BIT_VALID_DATA;
+			rxconf.src_addr_width = 4;
+			rxconf.src_maxburst = 4;
+		} else {
+			rxconf.src_addr = spfi->phys + SPFI_RX_8BIT_VALID_DATA;
+			rxconf.src_addr_width = 1;
+			rxconf.src_maxburst = 4;
+		}
+		dmaengine_slave_config(spfi->rx_ch, &rxconf);
+
+		rxdesc = dmaengine_prep_slave_sg(spfi->rx_ch, xfer->rx_sg.sgl,
+						 xfer->rx_sg.nents,
+						 DMA_DEV_TO_MEM,
+						 DMA_PREP_INTERRUPT);
+		if (!rxdesc)
+			goto stop_dma;
+
+		rxdesc->callback = img_spfi_dma_rx_cb;
+		rxdesc->callback_param = spfi;
+	}
+
+	if (xfer->tx_buf) {
+		txconf.direction = DMA_MEM_TO_DEV;
+		if (xfer->len % 4 == 0) {
+			txconf.dst_addr = spfi->phys + SPFI_TX_32BIT_VALID_DATA;
+			txconf.dst_addr_width = 4;
+			txconf.dst_maxburst = 4;
+		} else {
+			txconf.dst_addr = spfi->phys + SPFI_TX_8BIT_VALID_DATA;
+			txconf.dst_addr_width = 1;
+			txconf.dst_maxburst = 4;
+		}
+		dmaengine_slave_config(spfi->tx_ch, &txconf);
+
+		txdesc = dmaengine_prep_slave_sg(spfi->tx_ch, xfer->tx_sg.sgl,
+						 xfer->tx_sg.nents,
+						 DMA_MEM_TO_DEV,
+						 DMA_PREP_INTERRUPT);
+		if (!txdesc)
+			goto stop_dma;
+
+		txdesc->callback = img_spfi_dma_tx_cb;
+		txdesc->callback_param = spfi;
+	}
+
+	if (xfer->rx_buf) {
+		spfi->rx_dma_busy = true;
+		dmaengine_submit(rxdesc);
+		dma_async_issue_pending(spfi->rx_ch);
+	}
+
+	spfi_start(spfi);
+
+	if (xfer->tx_buf) {
+		spfi->tx_dma_busy = true;
+		dmaengine_submit(txdesc);
+		dma_async_issue_pending(spfi->tx_ch);
+	}
+
+	return 1;
+
+stop_dma:
+	dmaengine_terminate_all(spfi->rx_ch);
+	dmaengine_terminate_all(spfi->tx_ch);
+	return -EIO;
+}
+
+static void img_spfi_handle_err(struct spi_master *master,
+				struct spi_message *msg)
+{
+	struct img_spfi *spfi = spi_master_get_devdata(master);
+	unsigned long flags;
+
+	/*
+	 * Stop all DMA and reset the controller if the previous transaction
+	 * timed-out and never completed it's DMA.
+	 */
+	spin_lock_irqsave(&spfi->lock, flags);
+	if (spfi->tx_dma_busy || spfi->rx_dma_busy) {
+		spfi->tx_dma_busy = false;
+		spfi->rx_dma_busy = false;
+
+		dmaengine_terminate_all(spfi->tx_ch);
+		dmaengine_terminate_all(spfi->rx_ch);
+	}
+	spin_unlock_irqrestore(&spfi->lock, flags);
+}
+
+static int img_spfi_prepare(struct spi_master *master, struct spi_message *msg)
+{
+	struct img_spfi *spfi = spi_master_get_devdata(master);
+	u32 val;
+
+	val = spfi_readl(spfi, SPFI_PORT_STATE);
+	val &= ~(SPFI_PORT_STATE_DEV_SEL_MASK <<
+		 SPFI_PORT_STATE_DEV_SEL_SHIFT);
+	val |= msg->spi->chip_select << SPFI_PORT_STATE_DEV_SEL_SHIFT;
+	if (msg->spi->mode & SPI_CPHA)
+		val |= SPFI_PORT_STATE_CK_PHASE(msg->spi->chip_select);
+	else
+		val &= ~SPFI_PORT_STATE_CK_PHASE(msg->spi->chip_select);
+	if (msg->spi->mode & SPI_CPOL)
+		val |= SPFI_PORT_STATE_CK_POL(msg->spi->chip_select);
+	else
+		val &= ~SPFI_PORT_STATE_CK_POL(msg->spi->chip_select);
+	spfi_writel(spfi, val, SPFI_PORT_STATE);
+
+	return 0;
+}
+
+static int img_spfi_unprepare(struct spi_master *master,
+			      struct spi_message *msg)
+{
+	struct img_spfi *spfi = spi_master_get_devdata(master);
+
+	spfi_reset(spfi);
+
+	return 0;
+}
+
+static int img_spfi_setup(struct spi_device *spi)
+{
+	int ret = -EINVAL;
+	struct img_spfi_device_data *spfi_data = spi_get_ctldata(spi);
+
+	if (!spfi_data) {
+		spfi_data = kzalloc(sizeof(*spfi_data), GFP_KERNEL);
+		if (!spfi_data)
+			return -ENOMEM;
+		spfi_data->gpio_requested = false;
+		spi_set_ctldata(spi, spfi_data);
+	}
+	if (!spfi_data->gpio_requested) {
+		ret = gpio_request_one(spi->cs_gpio,
+				       (spi->mode & SPI_CS_HIGH) ?
+				       GPIOF_OUT_INIT_LOW : GPIOF_OUT_INIT_HIGH,
+				       dev_name(&spi->dev));
+		if (ret)
+			dev_err(&spi->dev, "can't request chipselect gpio %d\n",
+				spi->cs_gpio);
+		else
+			spfi_data->gpio_requested = true;
+	} else {
+		if (gpio_is_valid(spi->cs_gpio)) {
+			int mode = ((spi->mode & SPI_CS_HIGH) ?
+				    GPIOF_OUT_INIT_LOW : GPIOF_OUT_INIT_HIGH);
+
+			ret = gpio_direction_output(spi->cs_gpio, mode);
+			if (ret)
+				dev_err(&spi->dev, "chipselect gpio %d setup failed (%d)\n",
+					spi->cs_gpio, ret);
+		}
+	}
+	return ret;
+}
+
+static void img_spfi_cleanup(struct spi_device *spi)
+{
+	struct img_spfi_device_data *spfi_data = spi_get_ctldata(spi);
+
+	if (spfi_data) {
+		if (spfi_data->gpio_requested)
+			gpio_free(spi->cs_gpio);
+		kfree(spfi_data);
+		spi_set_ctldata(spi, NULL);
+	}
+}
+
+static void img_spfi_config(struct spi_master *master, struct spi_device *spi,
+			    struct spi_transfer *xfer)
+{
+	struct img_spfi *spfi = spi_master_get_devdata(spi->master);
+	u32 val, div;
+
+	/*
+	 * output = spfi_clk * (BITCLK / 512), where BITCLK must be a
+	 * power of 2 up to 128
+	 */
+	div = DIV_ROUND_UP(clk_get_rate(spfi->spfi_clk), xfer->speed_hz);
+	div = clamp(512 / (1 << get_count_order(div)), 1, 128);
+
+	val = spfi_readl(spfi, SPFI_DEVICE_PARAMETER(spi->chip_select));
+	val &= ~(SPFI_DEVICE_PARAMETER_BITCLK_MASK <<
+		 SPFI_DEVICE_PARAMETER_BITCLK_SHIFT);
+	val |= div << SPFI_DEVICE_PARAMETER_BITCLK_SHIFT;
+	spfi_writel(spfi, val, SPFI_DEVICE_PARAMETER(spi->chip_select));
+
+	spfi_writel(spfi, xfer->len << SPFI_TRANSACTION_TSIZE_SHIFT,
+		    SPFI_TRANSACTION);
+
+	val = spfi_readl(spfi, SPFI_CONTROL);
+	val &= ~(SPFI_CONTROL_SEND_DMA | SPFI_CONTROL_GET_DMA);
+	if (xfer->tx_buf)
+		val |= SPFI_CONTROL_SEND_DMA;
+	if (xfer->rx_buf)
+		val |= SPFI_CONTROL_GET_DMA;
+	val &= ~(SPFI_CONTROL_TMODE_MASK << SPFI_CONTROL_TMODE_SHIFT);
+	if (xfer->tx_nbits == SPI_NBITS_DUAL &&
+	    xfer->rx_nbits == SPI_NBITS_DUAL)
+		val |= SPFI_CONTROL_TMODE_DUAL << SPFI_CONTROL_TMODE_SHIFT;
+	else if (xfer->tx_nbits == SPI_NBITS_QUAD &&
+		 xfer->rx_nbits == SPI_NBITS_QUAD)
+		val |= SPFI_CONTROL_TMODE_QUAD << SPFI_CONTROL_TMODE_SHIFT;
+	val |= SPFI_CONTROL_SE;
+	spfi_writel(spfi, val, SPFI_CONTROL);
+}
+
+static int img_spfi_transfer_one(struct spi_master *master,
+				 struct spi_device *spi,
+				 struct spi_transfer *xfer)
+{
+	struct img_spfi *spfi = spi_master_get_devdata(spi->master);
+	int ret;
+
+	if (xfer->len > SPFI_TRANSACTION_TSIZE_MASK) {
+		dev_err(spfi->dev,
+			"Transfer length (%d) is greater than the max supported (%d)",
+			xfer->len, SPFI_TRANSACTION_TSIZE_MASK);
+		return -EINVAL;
+	}
+
+	img_spfi_config(master, spi, xfer);
+	if (master->can_dma && master->can_dma(master, spi, xfer))
+		ret = img_spfi_start_dma(master, spi, xfer);
+	else
+		ret = img_spfi_start_pio(master, spi, xfer);
+
+	return ret;
+}
+
+static bool img_spfi_can_dma(struct spi_master *master, struct spi_device *spi,
+			     struct spi_transfer *xfer)
+{
+	if (xfer->len > SPFI_32BIT_FIFO_SIZE)
+		return true;
+	return false;
+}
+
+static irqreturn_t img_spfi_irq(int irq, void *dev_id)
+{
+	struct img_spfi *spfi = (struct img_spfi *)dev_id;
+	u32 status;
+
+	status = spfi_readl(spfi, SPFI_INTERRUPT_STATUS);
+	if (status & SPFI_INTERRUPT_IACCESS) {
+		spfi_writel(spfi, SPFI_INTERRUPT_IACCESS, SPFI_INTERRUPT_CLEAR);
+		dev_err(spfi->dev, "Illegal access interrupt");
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_NONE;
+}
+
+static int img_spfi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct img_spfi *spfi;
+	struct resource *res;
+	int ret;
+	u32 max_speed_hz;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*spfi));
+	if (!master)
+		return -ENOMEM;
+	platform_set_drvdata(pdev, master);
+
+	spfi = spi_master_get_devdata(master);
+	spfi->dev = &pdev->dev;
+	spfi->master = master;
+	spin_lock_init(&spfi->lock);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	spfi->regs = devm_ioremap_resource(spfi->dev, res);
+	if (IS_ERR(spfi->regs)) {
+		ret = PTR_ERR(spfi->regs);
+		goto put_spi;
+	}
+	spfi->phys = res->start;
+
+	spfi->irq = platform_get_irq(pdev, 0);
+	if (spfi->irq < 0) {
+		ret = spfi->irq;
+		goto put_spi;
+	}
+	ret = devm_request_irq(spfi->dev, spfi->irq, img_spfi_irq,
+			       IRQ_TYPE_LEVEL_HIGH, dev_name(spfi->dev), spfi);
+	if (ret)
+		goto put_spi;
+
+	spfi->sys_clk = devm_clk_get(spfi->dev, "sys");
+	if (IS_ERR(spfi->sys_clk)) {
+		ret = PTR_ERR(spfi->sys_clk);
+		goto put_spi;
+	}
+	spfi->spfi_clk = devm_clk_get(spfi->dev, "spfi");
+	if (IS_ERR(spfi->spfi_clk)) {
+		ret = PTR_ERR(spfi->spfi_clk);
+		goto put_spi;
+	}
+
+	ret = clk_prepare_enable(spfi->sys_clk);
+	if (ret)
+		goto put_spi;
+	ret = clk_prepare_enable(spfi->spfi_clk);
+	if (ret)
+		goto disable_pclk;
+
+	spfi_reset(spfi);
+	/*
+	 * Only enable the error (IACCESS) interrupt.  In PIO mode we'll
+	 * poll the status of the FIFOs.
+	 */
+	spfi_writel(spfi, SPFI_INTERRUPT_IACCESS, SPFI_INTERRUPT_ENABLE);
+
+	master->auto_runtime_pm = true;
+	master->bus_num = pdev->id;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_TX_DUAL | SPI_RX_DUAL;
+	if (of_property_read_bool(spfi->dev->of_node, "img,supports-quad-mode"))
+		master->mode_bits |= SPI_TX_QUAD | SPI_RX_QUAD;
+	master->dev.of_node = pdev->dev.of_node;
+	master->bits_per_word_mask = SPI_BPW_MASK(32) | SPI_BPW_MASK(8);
+	master->max_speed_hz = clk_get_rate(spfi->spfi_clk) / 4;
+	master->min_speed_hz = clk_get_rate(spfi->spfi_clk) / 512;
+
+	/*
+	 * Maximum speed supported by spfi is limited to the lower value
+	 * between 1/4 of the SPFI clock or to "spfi-max-frequency"
+	 * defined in the device tree.
+	 * If no value is defined in the device tree assume the maximum
+	 * speed supported to be 1/4 of the SPFI clock.
+	 */
+	if (!of_property_read_u32(spfi->dev->of_node, "spfi-max-frequency",
+				  &max_speed_hz)) {
+		if (master->max_speed_hz > max_speed_hz)
+			master->max_speed_hz = max_speed_hz;
+	}
+
+	master->setup = img_spfi_setup;
+	master->cleanup = img_spfi_cleanup;
+	master->transfer_one = img_spfi_transfer_one;
+	master->prepare_message = img_spfi_prepare;
+	master->unprepare_message = img_spfi_unprepare;
+	master->handle_err = img_spfi_handle_err;
+
+	spfi->tx_ch = dma_request_slave_channel(spfi->dev, "tx");
+	spfi->rx_ch = dma_request_slave_channel(spfi->dev, "rx");
+	if (!spfi->tx_ch || !spfi->rx_ch) {
+		if (spfi->tx_ch)
+			dma_release_channel(spfi->tx_ch);
+		if (spfi->rx_ch)
+			dma_release_channel(spfi->rx_ch);
+		dev_warn(spfi->dev, "Failed to get DMA channels, falling back to PIO mode\n");
+	} else {
+		master->dma_tx = spfi->tx_ch;
+		master->dma_rx = spfi->rx_ch;
+		master->can_dma = img_spfi_can_dma;
+	}
+
+	pm_runtime_set_active(spfi->dev);
+	pm_runtime_enable(spfi->dev);
+
+	ret = devm_spi_register_master(spfi->dev, master);
+	if (ret)
+		goto disable_pm;
+
+	return 0;
+
+disable_pm:
+	pm_runtime_disable(spfi->dev);
+	if (spfi->rx_ch)
+		dma_release_channel(spfi->rx_ch);
+	if (spfi->tx_ch)
+		dma_release_channel(spfi->tx_ch);
+	clk_disable_unprepare(spfi->spfi_clk);
+disable_pclk:
+	clk_disable_unprepare(spfi->sys_clk);
+put_spi:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int img_spfi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct img_spfi *spfi = spi_master_get_devdata(master);
+
+	if (spfi->tx_ch)
+		dma_release_channel(spfi->tx_ch);
+	if (spfi->rx_ch)
+		dma_release_channel(spfi->rx_ch);
+
+	pm_runtime_disable(spfi->dev);
+	if (!pm_runtime_status_suspended(spfi->dev)) {
+		clk_disable_unprepare(spfi->spfi_clk);
+		clk_disable_unprepare(spfi->sys_clk);
+	}
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int img_spfi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct img_spfi *spfi = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(spfi->spfi_clk);
+	clk_disable_unprepare(spfi->sys_clk);
+
+	return 0;
+}
+
+static int img_spfi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct img_spfi *spfi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(spfi->sys_clk);
+	if (ret)
+		return ret;
+	ret = clk_prepare_enable(spfi->spfi_clk);
+	if (ret) {
+		clk_disable_unprepare(spfi->sys_clk);
+		return ret;
+	}
+
+	return 0;
+}
+#endif /* CONFIG_PM */
+
+#ifdef CONFIG_PM_SLEEP
+static int img_spfi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+
+	return spi_master_suspend(master);
+}
+
+static int img_spfi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct img_spfi *spfi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = pm_runtime_get_sync(dev);
+	if (ret)
+		return ret;
+	spfi_reset(spfi);
+	pm_runtime_put(dev);
+
+	return spi_master_resume(master);
+}
+#endif /* CONFIG_PM_SLEEP */
+
+static const struct dev_pm_ops img_spfi_pm_ops = {
+	SET_RUNTIME_PM_OPS(img_spfi_runtime_suspend, img_spfi_runtime_resume,
+			   NULL)
+	SET_SYSTEM_SLEEP_PM_OPS(img_spfi_suspend, img_spfi_resume)
+};
+
+static const struct of_device_id img_spfi_of_match[] = {
+	{ .compatible = "img,spfi", },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, img_spfi_of_match);
+
+static struct platform_driver img_spfi_driver = {
+	.driver = {
+		.name = "img-spfi",
+		.pm = &img_spfi_pm_ops,
+		.of_match_table = of_match_ptr(img_spfi_of_match),
+	},
+	.probe = img_spfi_probe,
+	.remove = img_spfi_remove,
+};
+module_platform_driver(img_spfi_driver);
+
+MODULE_DESCRIPTION("IMG SPFI controller driver");
+MODULE_AUTHOR("Andrew Bresticker <abrestic@chromium.org>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-imx.c b/drivers/spi/spi-imx.c
new file mode 100644
index 0000000..08dd3a3
--- /dev/null
+++ b/drivers/spi/spi-imx.c
@@ -0,0 +1,1732 @@
+// SPDX-License-Identifier: GPL-2.0+
+// Copyright 2004-2007 Freescale Semiconductor, Inc. All Rights Reserved.
+// Copyright (C) 2008 Juergen Beisert
+
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/err.h>
+#include <linux/gpio.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/types.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/of_gpio.h>
+
+#include <linux/platform_data/dma-imx.h>
+#include <linux/platform_data/spi-imx.h>
+
+#define DRIVER_NAME "spi_imx"
+
+#define MXC_CSPIRXDATA		0x00
+#define MXC_CSPITXDATA		0x04
+#define MXC_CSPICTRL		0x08
+#define MXC_CSPIINT		0x0c
+#define MXC_RESET		0x1c
+
+/* generic defines to abstract from the different register layouts */
+#define MXC_INT_RR	(1 << 0) /* Receive data ready interrupt */
+#define MXC_INT_TE	(1 << 1) /* Transmit FIFO empty interrupt */
+#define MXC_INT_RDR	BIT(4) /* Receive date threshold interrupt */
+
+/* The maximum  bytes that a sdma BD can transfer.*/
+#define MAX_SDMA_BD_BYTES  (1 << 15)
+#define MX51_ECSPI_CTRL_MAX_BURST	512
+/* The maximum bytes that IMX53_ECSPI can transfer in slave mode.*/
+#define MX53_MAX_TRANSFER_BYTES		512
+
+enum spi_imx_devtype {
+	IMX1_CSPI,
+	IMX21_CSPI,
+	IMX27_CSPI,
+	IMX31_CSPI,
+	IMX35_CSPI,	/* CSPI on all i.mx except above */
+	IMX51_ECSPI,	/* ECSPI on i.mx51 */
+	IMX53_ECSPI,	/* ECSPI on i.mx53 and later */
+};
+
+struct spi_imx_data;
+
+struct spi_imx_devtype_data {
+	void (*intctrl)(struct spi_imx_data *, int);
+	int (*config)(struct spi_device *);
+	void (*trigger)(struct spi_imx_data *);
+	int (*rx_available)(struct spi_imx_data *);
+	void (*reset)(struct spi_imx_data *);
+	void (*disable)(struct spi_imx_data *);
+	bool has_dmamode;
+	bool has_slavemode;
+	unsigned int fifo_size;
+	bool dynamic_burst;
+	enum spi_imx_devtype devtype;
+};
+
+struct spi_imx_data {
+	struct spi_bitbang bitbang;
+	struct device *dev;
+
+	struct completion xfer_done;
+	void __iomem *base;
+	unsigned long base_phys;
+
+	struct clk *clk_per;
+	struct clk *clk_ipg;
+	unsigned long spi_clk;
+	unsigned int spi_bus_clk;
+
+	unsigned int speed_hz;
+	unsigned int bits_per_word;
+	unsigned int spi_drctl;
+
+	unsigned int count, remainder;
+	void (*tx)(struct spi_imx_data *);
+	void (*rx)(struct spi_imx_data *);
+	void *rx_buf;
+	const void *tx_buf;
+	unsigned int txfifo; /* number of words pushed in tx FIFO */
+	unsigned int dynamic_burst;
+
+	/* Slave mode */
+	bool slave_mode;
+	bool slave_aborted;
+	unsigned int slave_burst;
+
+	/* DMA */
+	bool usedma;
+	u32 wml;
+	struct completion dma_rx_completion;
+	struct completion dma_tx_completion;
+
+	const struct spi_imx_devtype_data *devtype_data;
+};
+
+static inline int is_imx27_cspi(struct spi_imx_data *d)
+{
+	return d->devtype_data->devtype == IMX27_CSPI;
+}
+
+static inline int is_imx35_cspi(struct spi_imx_data *d)
+{
+	return d->devtype_data->devtype == IMX35_CSPI;
+}
+
+static inline int is_imx51_ecspi(struct spi_imx_data *d)
+{
+	return d->devtype_data->devtype == IMX51_ECSPI;
+}
+
+static inline int is_imx53_ecspi(struct spi_imx_data *d)
+{
+	return d->devtype_data->devtype == IMX53_ECSPI;
+}
+
+#define MXC_SPI_BUF_RX(type)						\
+static void spi_imx_buf_rx_##type(struct spi_imx_data *spi_imx)		\
+{									\
+	unsigned int val = readl(spi_imx->base + MXC_CSPIRXDATA);	\
+									\
+	if (spi_imx->rx_buf) {						\
+		*(type *)spi_imx->rx_buf = val;				\
+		spi_imx->rx_buf += sizeof(type);			\
+	}								\
+									\
+	spi_imx->remainder -= sizeof(type);				\
+}
+
+#define MXC_SPI_BUF_TX(type)						\
+static void spi_imx_buf_tx_##type(struct spi_imx_data *spi_imx)		\
+{									\
+	type val = 0;							\
+									\
+	if (spi_imx->tx_buf) {						\
+		val = *(type *)spi_imx->tx_buf;				\
+		spi_imx->tx_buf += sizeof(type);			\
+	}								\
+									\
+	spi_imx->count -= sizeof(type);					\
+									\
+	writel(val, spi_imx->base + MXC_CSPITXDATA);			\
+}
+
+MXC_SPI_BUF_RX(u8)
+MXC_SPI_BUF_TX(u8)
+MXC_SPI_BUF_RX(u16)
+MXC_SPI_BUF_TX(u16)
+MXC_SPI_BUF_RX(u32)
+MXC_SPI_BUF_TX(u32)
+
+/* First entry is reserved, second entry is valid only if SDHC_SPIEN is set
+ * (which is currently not the case in this driver)
+ */
+static int mxc_clkdivs[] = {0, 3, 4, 6, 8, 12, 16, 24, 32, 48, 64, 96, 128, 192,
+	256, 384, 512, 768, 1024};
+
+/* MX21, MX27 */
+static unsigned int spi_imx_clkdiv_1(unsigned int fin,
+		unsigned int fspi, unsigned int max, unsigned int *fres)
+{
+	int i;
+
+	for (i = 2; i < max; i++)
+		if (fspi * mxc_clkdivs[i] >= fin)
+			break;
+
+	*fres = fin / mxc_clkdivs[i];
+	return i;
+}
+
+/* MX1, MX31, MX35, MX51 CSPI */
+static unsigned int spi_imx_clkdiv_2(unsigned int fin,
+		unsigned int fspi, unsigned int *fres)
+{
+	int i, div = 4;
+
+	for (i = 0; i < 7; i++) {
+		if (fspi * div >= fin)
+			goto out;
+		div <<= 1;
+	}
+
+out:
+	*fres = fin / div;
+	return i;
+}
+
+static int spi_imx_bytes_per_word(const int bits_per_word)
+{
+	if (bits_per_word <= 8)
+		return 1;
+	else if (bits_per_word <= 16)
+		return 2;
+	else
+		return 4;
+}
+
+static bool spi_imx_can_dma(struct spi_master *master, struct spi_device *spi,
+			 struct spi_transfer *transfer)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(master);
+	unsigned int bytes_per_word, i;
+
+	if (!master->dma_rx)
+		return false;
+
+	if (spi_imx->slave_mode)
+		return false;
+
+	bytes_per_word = spi_imx_bytes_per_word(transfer->bits_per_word);
+
+	for (i = spi_imx->devtype_data->fifo_size / 2; i > 0; i--) {
+		if (!(transfer->len % (i * bytes_per_word)))
+			break;
+	}
+
+	spi_imx->wml = i;
+	spi_imx->dynamic_burst = 0;
+
+	return true;
+}
+
+#define MX51_ECSPI_CTRL		0x08
+#define MX51_ECSPI_CTRL_ENABLE		(1 <<  0)
+#define MX51_ECSPI_CTRL_XCH		(1 <<  2)
+#define MX51_ECSPI_CTRL_SMC		(1 << 3)
+#define MX51_ECSPI_CTRL_MODE_MASK	(0xf << 4)
+#define MX51_ECSPI_CTRL_DRCTL(drctl)	((drctl) << 16)
+#define MX51_ECSPI_CTRL_POSTDIV_OFFSET	8
+#define MX51_ECSPI_CTRL_PREDIV_OFFSET	12
+#define MX51_ECSPI_CTRL_CS(cs)		((cs) << 18)
+#define MX51_ECSPI_CTRL_BL_OFFSET	20
+#define MX51_ECSPI_CTRL_BL_MASK		(0xfff << 20)
+
+#define MX51_ECSPI_CONFIG	0x0c
+#define MX51_ECSPI_CONFIG_SCLKPHA(cs)	(1 << ((cs) +  0))
+#define MX51_ECSPI_CONFIG_SCLKPOL(cs)	(1 << ((cs) +  4))
+#define MX51_ECSPI_CONFIG_SBBCTRL(cs)	(1 << ((cs) +  8))
+#define MX51_ECSPI_CONFIG_SSBPOL(cs)	(1 << ((cs) + 12))
+#define MX51_ECSPI_CONFIG_SCLKCTL(cs)	(1 << ((cs) + 20))
+
+#define MX51_ECSPI_INT		0x10
+#define MX51_ECSPI_INT_TEEN		(1 <<  0)
+#define MX51_ECSPI_INT_RREN		(1 <<  3)
+#define MX51_ECSPI_INT_RDREN		(1 <<  4)
+
+#define MX51_ECSPI_DMA      0x14
+#define MX51_ECSPI_DMA_TX_WML(wml)	((wml) & 0x3f)
+#define MX51_ECSPI_DMA_RX_WML(wml)	(((wml) & 0x3f) << 16)
+#define MX51_ECSPI_DMA_RXT_WML(wml)	(((wml) & 0x3f) << 24)
+
+#define MX51_ECSPI_DMA_TEDEN		(1 << 7)
+#define MX51_ECSPI_DMA_RXDEN		(1 << 23)
+#define MX51_ECSPI_DMA_RXTDEN		(1 << 31)
+
+#define MX51_ECSPI_STAT		0x18
+#define MX51_ECSPI_STAT_RR		(1 <<  3)
+
+#define MX51_ECSPI_TESTREG	0x20
+#define MX51_ECSPI_TESTREG_LBC	BIT(31)
+
+static void spi_imx_buf_rx_swap_u32(struct spi_imx_data *spi_imx)
+{
+	unsigned int val = readl(spi_imx->base + MXC_CSPIRXDATA);
+#ifdef __LITTLE_ENDIAN
+	unsigned int bytes_per_word;
+#endif
+
+	if (spi_imx->rx_buf) {
+#ifdef __LITTLE_ENDIAN
+		bytes_per_word = spi_imx_bytes_per_word(spi_imx->bits_per_word);
+		if (bytes_per_word == 1)
+			val = cpu_to_be32(val);
+		else if (bytes_per_word == 2)
+			val = (val << 16) | (val >> 16);
+#endif
+		*(u32 *)spi_imx->rx_buf = val;
+		spi_imx->rx_buf += sizeof(u32);
+	}
+
+	spi_imx->remainder -= sizeof(u32);
+}
+
+static void spi_imx_buf_rx_swap(struct spi_imx_data *spi_imx)
+{
+	int unaligned;
+	u32 val;
+
+	unaligned = spi_imx->remainder % 4;
+
+	if (!unaligned) {
+		spi_imx_buf_rx_swap_u32(spi_imx);
+		return;
+	}
+
+	if (spi_imx_bytes_per_word(spi_imx->bits_per_word) == 2) {
+		spi_imx_buf_rx_u16(spi_imx);
+		return;
+	}
+
+	val = readl(spi_imx->base + MXC_CSPIRXDATA);
+
+	while (unaligned--) {
+		if (spi_imx->rx_buf) {
+			*(u8 *)spi_imx->rx_buf = (val >> (8 * unaligned)) & 0xff;
+			spi_imx->rx_buf++;
+		}
+		spi_imx->remainder--;
+	}
+}
+
+static void spi_imx_buf_tx_swap_u32(struct spi_imx_data *spi_imx)
+{
+	u32 val = 0;
+#ifdef __LITTLE_ENDIAN
+	unsigned int bytes_per_word;
+#endif
+
+	if (spi_imx->tx_buf) {
+		val = *(u32 *)spi_imx->tx_buf;
+		spi_imx->tx_buf += sizeof(u32);
+	}
+
+	spi_imx->count -= sizeof(u32);
+#ifdef __LITTLE_ENDIAN
+	bytes_per_word = spi_imx_bytes_per_word(spi_imx->bits_per_word);
+
+	if (bytes_per_word == 1)
+		val = cpu_to_be32(val);
+	else if (bytes_per_word == 2)
+		val = (val << 16) | (val >> 16);
+#endif
+	writel(val, spi_imx->base + MXC_CSPITXDATA);
+}
+
+static void spi_imx_buf_tx_swap(struct spi_imx_data *spi_imx)
+{
+	int unaligned;
+	u32 val = 0;
+
+	unaligned = spi_imx->count % 4;
+
+	if (!unaligned) {
+		spi_imx_buf_tx_swap_u32(spi_imx);
+		return;
+	}
+
+	if (spi_imx_bytes_per_word(spi_imx->bits_per_word) == 2) {
+		spi_imx_buf_tx_u16(spi_imx);
+		return;
+	}
+
+	while (unaligned--) {
+		if (spi_imx->tx_buf) {
+			val |= *(u8 *)spi_imx->tx_buf << (8 * unaligned);
+			spi_imx->tx_buf++;
+		}
+		spi_imx->count--;
+	}
+
+	writel(val, spi_imx->base + MXC_CSPITXDATA);
+}
+
+static void mx53_ecspi_rx_slave(struct spi_imx_data *spi_imx)
+{
+	u32 val = be32_to_cpu(readl(spi_imx->base + MXC_CSPIRXDATA));
+
+	if (spi_imx->rx_buf) {
+		int n_bytes = spi_imx->slave_burst % sizeof(val);
+
+		if (!n_bytes)
+			n_bytes = sizeof(val);
+
+		memcpy(spi_imx->rx_buf,
+		       ((u8 *)&val) + sizeof(val) - n_bytes, n_bytes);
+
+		spi_imx->rx_buf += n_bytes;
+		spi_imx->slave_burst -= n_bytes;
+	}
+
+	spi_imx->remainder -= sizeof(u32);
+}
+
+static void mx53_ecspi_tx_slave(struct spi_imx_data *spi_imx)
+{
+	u32 val = 0;
+	int n_bytes = spi_imx->count % sizeof(val);
+
+	if (!n_bytes)
+		n_bytes = sizeof(val);
+
+	if (spi_imx->tx_buf) {
+		memcpy(((u8 *)&val) + sizeof(val) - n_bytes,
+		       spi_imx->tx_buf, n_bytes);
+		val = cpu_to_be32(val);
+		spi_imx->tx_buf += n_bytes;
+	}
+
+	spi_imx->count -= n_bytes;
+
+	writel(val, spi_imx->base + MXC_CSPITXDATA);
+}
+
+/* MX51 eCSPI */
+static unsigned int mx51_ecspi_clkdiv(struct spi_imx_data *spi_imx,
+				      unsigned int fspi, unsigned int *fres)
+{
+	/*
+	 * there are two 4-bit dividers, the pre-divider divides by
+	 * $pre, the post-divider by 2^$post
+	 */
+	unsigned int pre, post;
+	unsigned int fin = spi_imx->spi_clk;
+
+	if (unlikely(fspi > fin))
+		return 0;
+
+	post = fls(fin) - fls(fspi);
+	if (fin > fspi << post)
+		post++;
+
+	/* now we have: (fin <= fspi << post) with post being minimal */
+
+	post = max(4U, post) - 4;
+	if (unlikely(post > 0xf)) {
+		dev_err(spi_imx->dev, "cannot set clock freq: %u (base freq: %u)\n",
+				fspi, fin);
+		return 0xff;
+	}
+
+	pre = DIV_ROUND_UP(fin, fspi << post) - 1;
+
+	dev_dbg(spi_imx->dev, "%s: fin: %u, fspi: %u, post: %u, pre: %u\n",
+			__func__, fin, fspi, post, pre);
+
+	/* Resulting frequency for the SCLK line. */
+	*fres = (fin / (pre + 1)) >> post;
+
+	return (pre << MX51_ECSPI_CTRL_PREDIV_OFFSET) |
+		(post << MX51_ECSPI_CTRL_POSTDIV_OFFSET);
+}
+
+static void mx51_ecspi_intctrl(struct spi_imx_data *spi_imx, int enable)
+{
+	unsigned val = 0;
+
+	if (enable & MXC_INT_TE)
+		val |= MX51_ECSPI_INT_TEEN;
+
+	if (enable & MXC_INT_RR)
+		val |= MX51_ECSPI_INT_RREN;
+
+	if (enable & MXC_INT_RDR)
+		val |= MX51_ECSPI_INT_RDREN;
+
+	writel(val, spi_imx->base + MX51_ECSPI_INT);
+}
+
+static void mx51_ecspi_trigger(struct spi_imx_data *spi_imx)
+{
+	u32 reg;
+
+	reg = readl(spi_imx->base + MX51_ECSPI_CTRL);
+	reg |= MX51_ECSPI_CTRL_XCH;
+	writel(reg, spi_imx->base + MX51_ECSPI_CTRL);
+}
+
+static void mx51_ecspi_disable(struct spi_imx_data *spi_imx)
+{
+	u32 ctrl;
+
+	ctrl = readl(spi_imx->base + MX51_ECSPI_CTRL);
+	ctrl &= ~MX51_ECSPI_CTRL_ENABLE;
+	writel(ctrl, spi_imx->base + MX51_ECSPI_CTRL);
+}
+
+static int mx51_ecspi_config(struct spi_device *spi)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(spi->master);
+	u32 ctrl = MX51_ECSPI_CTRL_ENABLE;
+	u32 clk = spi_imx->speed_hz, delay, reg;
+	u32 cfg = readl(spi_imx->base + MX51_ECSPI_CONFIG);
+
+	/* set Master or Slave mode */
+	if (spi_imx->slave_mode)
+		ctrl &= ~MX51_ECSPI_CTRL_MODE_MASK;
+	else
+		ctrl |= MX51_ECSPI_CTRL_MODE_MASK;
+
+	/*
+	 * Enable SPI_RDY handling (falling edge/level triggered).
+	 */
+	if (spi->mode & SPI_READY)
+		ctrl |= MX51_ECSPI_CTRL_DRCTL(spi_imx->spi_drctl);
+
+	/* set clock speed */
+	ctrl |= mx51_ecspi_clkdiv(spi_imx, spi_imx->speed_hz, &clk);
+	spi_imx->spi_bus_clk = clk;
+
+	/* set chip select to use */
+	ctrl |= MX51_ECSPI_CTRL_CS(spi->chip_select);
+
+	if (spi_imx->slave_mode && is_imx53_ecspi(spi_imx))
+		ctrl |= (spi_imx->slave_burst * 8 - 1)
+			<< MX51_ECSPI_CTRL_BL_OFFSET;
+	else
+		ctrl |= (spi_imx->bits_per_word - 1)
+			<< MX51_ECSPI_CTRL_BL_OFFSET;
+
+	/*
+	 * eCSPI burst completion by Chip Select signal in Slave mode
+	 * is not functional for imx53 Soc, config SPI burst completed when
+	 * BURST_LENGTH + 1 bits are received
+	 */
+	if (spi_imx->slave_mode && is_imx53_ecspi(spi_imx))
+		cfg &= ~MX51_ECSPI_CONFIG_SBBCTRL(spi->chip_select);
+	else
+		cfg |= MX51_ECSPI_CONFIG_SBBCTRL(spi->chip_select);
+
+	if (spi->mode & SPI_CPHA)
+		cfg |= MX51_ECSPI_CONFIG_SCLKPHA(spi->chip_select);
+	else
+		cfg &= ~MX51_ECSPI_CONFIG_SCLKPHA(spi->chip_select);
+
+	if (spi->mode & SPI_CPOL) {
+		cfg |= MX51_ECSPI_CONFIG_SCLKPOL(spi->chip_select);
+		cfg |= MX51_ECSPI_CONFIG_SCLKCTL(spi->chip_select);
+	} else {
+		cfg &= ~MX51_ECSPI_CONFIG_SCLKPOL(spi->chip_select);
+		cfg &= ~MX51_ECSPI_CONFIG_SCLKCTL(spi->chip_select);
+	}
+	if (spi->mode & SPI_CS_HIGH)
+		cfg |= MX51_ECSPI_CONFIG_SSBPOL(spi->chip_select);
+	else
+		cfg &= ~MX51_ECSPI_CONFIG_SSBPOL(spi->chip_select);
+
+	if (spi_imx->usedma)
+		ctrl |= MX51_ECSPI_CTRL_SMC;
+
+	/* CTRL register always go first to bring out controller from reset */
+	writel(ctrl, spi_imx->base + MX51_ECSPI_CTRL);
+
+	reg = readl(spi_imx->base + MX51_ECSPI_TESTREG);
+	if (spi->mode & SPI_LOOP)
+		reg |= MX51_ECSPI_TESTREG_LBC;
+	else
+		reg &= ~MX51_ECSPI_TESTREG_LBC;
+	writel(reg, spi_imx->base + MX51_ECSPI_TESTREG);
+
+	writel(cfg, spi_imx->base + MX51_ECSPI_CONFIG);
+
+	/*
+	 * Wait until the changes in the configuration register CONFIGREG
+	 * propagate into the hardware. It takes exactly one tick of the
+	 * SCLK clock, but we will wait two SCLK clock just to be sure. The
+	 * effect of the delay it takes for the hardware to apply changes
+	 * is noticable if the SCLK clock run very slow. In such a case, if
+	 * the polarity of SCLK should be inverted, the GPIO ChipSelect might
+	 * be asserted before the SCLK polarity changes, which would disrupt
+	 * the SPI communication as the device on the other end would consider
+	 * the change of SCLK polarity as a clock tick already.
+	 */
+	delay = (2 * 1000000) / clk;
+	if (likely(delay < 10))	/* SCLK is faster than 100 kHz */
+		udelay(delay);
+	else			/* SCLK is _very_ slow */
+		usleep_range(delay, delay + 10);
+
+	/*
+	 * Configure the DMA register: setup the watermark
+	 * and enable DMA request.
+	 */
+
+	writel(MX51_ECSPI_DMA_RX_WML(spi_imx->wml) |
+		MX51_ECSPI_DMA_TX_WML(spi_imx->wml) |
+		MX51_ECSPI_DMA_RXT_WML(spi_imx->wml) |
+		MX51_ECSPI_DMA_TEDEN | MX51_ECSPI_DMA_RXDEN |
+		MX51_ECSPI_DMA_RXTDEN, spi_imx->base + MX51_ECSPI_DMA);
+
+	return 0;
+}
+
+static int mx51_ecspi_rx_available(struct spi_imx_data *spi_imx)
+{
+	return readl(spi_imx->base + MX51_ECSPI_STAT) & MX51_ECSPI_STAT_RR;
+}
+
+static void mx51_ecspi_reset(struct spi_imx_data *spi_imx)
+{
+	/* drain receive buffer */
+	while (mx51_ecspi_rx_available(spi_imx))
+		readl(spi_imx->base + MXC_CSPIRXDATA);
+}
+
+#define MX31_INTREG_TEEN	(1 << 0)
+#define MX31_INTREG_RREN	(1 << 3)
+
+#define MX31_CSPICTRL_ENABLE	(1 << 0)
+#define MX31_CSPICTRL_MASTER	(1 << 1)
+#define MX31_CSPICTRL_XCH	(1 << 2)
+#define MX31_CSPICTRL_SMC	(1 << 3)
+#define MX31_CSPICTRL_POL	(1 << 4)
+#define MX31_CSPICTRL_PHA	(1 << 5)
+#define MX31_CSPICTRL_SSCTL	(1 << 6)
+#define MX31_CSPICTRL_SSPOL	(1 << 7)
+#define MX31_CSPICTRL_BC_SHIFT	8
+#define MX35_CSPICTRL_BL_SHIFT	20
+#define MX31_CSPICTRL_CS_SHIFT	24
+#define MX35_CSPICTRL_CS_SHIFT	12
+#define MX31_CSPICTRL_DR_SHIFT	16
+
+#define MX31_CSPI_DMAREG	0x10
+#define MX31_DMAREG_RH_DEN	(1<<4)
+#define MX31_DMAREG_TH_DEN	(1<<1)
+
+#define MX31_CSPISTATUS		0x14
+#define MX31_STATUS_RR		(1 << 3)
+
+#define MX31_CSPI_TESTREG	0x1C
+#define MX31_TEST_LBC		(1 << 14)
+
+/* These functions also work for the i.MX35, but be aware that
+ * the i.MX35 has a slightly different register layout for bits
+ * we do not use here.
+ */
+static void mx31_intctrl(struct spi_imx_data *spi_imx, int enable)
+{
+	unsigned int val = 0;
+
+	if (enable & MXC_INT_TE)
+		val |= MX31_INTREG_TEEN;
+	if (enable & MXC_INT_RR)
+		val |= MX31_INTREG_RREN;
+
+	writel(val, spi_imx->base + MXC_CSPIINT);
+}
+
+static void mx31_trigger(struct spi_imx_data *spi_imx)
+{
+	unsigned int reg;
+
+	reg = readl(spi_imx->base + MXC_CSPICTRL);
+	reg |= MX31_CSPICTRL_XCH;
+	writel(reg, spi_imx->base + MXC_CSPICTRL);
+}
+
+static int mx31_config(struct spi_device *spi)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(spi->master);
+	unsigned int reg = MX31_CSPICTRL_ENABLE | MX31_CSPICTRL_MASTER;
+	unsigned int clk;
+
+	reg |= spi_imx_clkdiv_2(spi_imx->spi_clk, spi_imx->speed_hz, &clk) <<
+		MX31_CSPICTRL_DR_SHIFT;
+	spi_imx->spi_bus_clk = clk;
+
+	if (is_imx35_cspi(spi_imx)) {
+		reg |= (spi_imx->bits_per_word - 1) << MX35_CSPICTRL_BL_SHIFT;
+		reg |= MX31_CSPICTRL_SSCTL;
+	} else {
+		reg |= (spi_imx->bits_per_word - 1) << MX31_CSPICTRL_BC_SHIFT;
+	}
+
+	if (spi->mode & SPI_CPHA)
+		reg |= MX31_CSPICTRL_PHA;
+	if (spi->mode & SPI_CPOL)
+		reg |= MX31_CSPICTRL_POL;
+	if (spi->mode & SPI_CS_HIGH)
+		reg |= MX31_CSPICTRL_SSPOL;
+	if (!gpio_is_valid(spi->cs_gpio))
+		reg |= (spi->chip_select) <<
+			(is_imx35_cspi(spi_imx) ? MX35_CSPICTRL_CS_SHIFT :
+						  MX31_CSPICTRL_CS_SHIFT);
+
+	if (spi_imx->usedma)
+		reg |= MX31_CSPICTRL_SMC;
+
+	writel(reg, spi_imx->base + MXC_CSPICTRL);
+
+	reg = readl(spi_imx->base + MX31_CSPI_TESTREG);
+	if (spi->mode & SPI_LOOP)
+		reg |= MX31_TEST_LBC;
+	else
+		reg &= ~MX31_TEST_LBC;
+	writel(reg, spi_imx->base + MX31_CSPI_TESTREG);
+
+	if (spi_imx->usedma) {
+		/* configure DMA requests when RXFIFO is half full and
+		   when TXFIFO is half empty */
+		writel(MX31_DMAREG_RH_DEN | MX31_DMAREG_TH_DEN,
+			spi_imx->base + MX31_CSPI_DMAREG);
+	}
+
+	return 0;
+}
+
+static int mx31_rx_available(struct spi_imx_data *spi_imx)
+{
+	return readl(spi_imx->base + MX31_CSPISTATUS) & MX31_STATUS_RR;
+}
+
+static void mx31_reset(struct spi_imx_data *spi_imx)
+{
+	/* drain receive buffer */
+	while (readl(spi_imx->base + MX31_CSPISTATUS) & MX31_STATUS_RR)
+		readl(spi_imx->base + MXC_CSPIRXDATA);
+}
+
+#define MX21_INTREG_RR		(1 << 4)
+#define MX21_INTREG_TEEN	(1 << 9)
+#define MX21_INTREG_RREN	(1 << 13)
+
+#define MX21_CSPICTRL_POL	(1 << 5)
+#define MX21_CSPICTRL_PHA	(1 << 6)
+#define MX21_CSPICTRL_SSPOL	(1 << 8)
+#define MX21_CSPICTRL_XCH	(1 << 9)
+#define MX21_CSPICTRL_ENABLE	(1 << 10)
+#define MX21_CSPICTRL_MASTER	(1 << 11)
+#define MX21_CSPICTRL_DR_SHIFT	14
+#define MX21_CSPICTRL_CS_SHIFT	19
+
+static void mx21_intctrl(struct spi_imx_data *spi_imx, int enable)
+{
+	unsigned int val = 0;
+
+	if (enable & MXC_INT_TE)
+		val |= MX21_INTREG_TEEN;
+	if (enable & MXC_INT_RR)
+		val |= MX21_INTREG_RREN;
+
+	writel(val, spi_imx->base + MXC_CSPIINT);
+}
+
+static void mx21_trigger(struct spi_imx_data *spi_imx)
+{
+	unsigned int reg;
+
+	reg = readl(spi_imx->base + MXC_CSPICTRL);
+	reg |= MX21_CSPICTRL_XCH;
+	writel(reg, spi_imx->base + MXC_CSPICTRL);
+}
+
+static int mx21_config(struct spi_device *spi)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(spi->master);
+	unsigned int reg = MX21_CSPICTRL_ENABLE | MX21_CSPICTRL_MASTER;
+	unsigned int max = is_imx27_cspi(spi_imx) ? 16 : 18;
+	unsigned int clk;
+
+	reg |= spi_imx_clkdiv_1(spi_imx->spi_clk, spi_imx->speed_hz, max, &clk)
+		<< MX21_CSPICTRL_DR_SHIFT;
+	spi_imx->spi_bus_clk = clk;
+
+	reg |= spi_imx->bits_per_word - 1;
+
+	if (spi->mode & SPI_CPHA)
+		reg |= MX21_CSPICTRL_PHA;
+	if (spi->mode & SPI_CPOL)
+		reg |= MX21_CSPICTRL_POL;
+	if (spi->mode & SPI_CS_HIGH)
+		reg |= MX21_CSPICTRL_SSPOL;
+	if (!gpio_is_valid(spi->cs_gpio))
+		reg |= spi->chip_select << MX21_CSPICTRL_CS_SHIFT;
+
+	writel(reg, spi_imx->base + MXC_CSPICTRL);
+
+	return 0;
+}
+
+static int mx21_rx_available(struct spi_imx_data *spi_imx)
+{
+	return readl(spi_imx->base + MXC_CSPIINT) & MX21_INTREG_RR;
+}
+
+static void mx21_reset(struct spi_imx_data *spi_imx)
+{
+	writel(1, spi_imx->base + MXC_RESET);
+}
+
+#define MX1_INTREG_RR		(1 << 3)
+#define MX1_INTREG_TEEN		(1 << 8)
+#define MX1_INTREG_RREN		(1 << 11)
+
+#define MX1_CSPICTRL_POL	(1 << 4)
+#define MX1_CSPICTRL_PHA	(1 << 5)
+#define MX1_CSPICTRL_XCH	(1 << 8)
+#define MX1_CSPICTRL_ENABLE	(1 << 9)
+#define MX1_CSPICTRL_MASTER	(1 << 10)
+#define MX1_CSPICTRL_DR_SHIFT	13
+
+static void mx1_intctrl(struct spi_imx_data *spi_imx, int enable)
+{
+	unsigned int val = 0;
+
+	if (enable & MXC_INT_TE)
+		val |= MX1_INTREG_TEEN;
+	if (enable & MXC_INT_RR)
+		val |= MX1_INTREG_RREN;
+
+	writel(val, spi_imx->base + MXC_CSPIINT);
+}
+
+static void mx1_trigger(struct spi_imx_data *spi_imx)
+{
+	unsigned int reg;
+
+	reg = readl(spi_imx->base + MXC_CSPICTRL);
+	reg |= MX1_CSPICTRL_XCH;
+	writel(reg, spi_imx->base + MXC_CSPICTRL);
+}
+
+static int mx1_config(struct spi_device *spi)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(spi->master);
+	unsigned int reg = MX1_CSPICTRL_ENABLE | MX1_CSPICTRL_MASTER;
+	unsigned int clk;
+
+	reg |= spi_imx_clkdiv_2(spi_imx->spi_clk, spi_imx->speed_hz, &clk) <<
+		MX1_CSPICTRL_DR_SHIFT;
+	spi_imx->spi_bus_clk = clk;
+
+	reg |= spi_imx->bits_per_word - 1;
+
+	if (spi->mode & SPI_CPHA)
+		reg |= MX1_CSPICTRL_PHA;
+	if (spi->mode & SPI_CPOL)
+		reg |= MX1_CSPICTRL_POL;
+
+	writel(reg, spi_imx->base + MXC_CSPICTRL);
+
+	return 0;
+}
+
+static int mx1_rx_available(struct spi_imx_data *spi_imx)
+{
+	return readl(spi_imx->base + MXC_CSPIINT) & MX1_INTREG_RR;
+}
+
+static void mx1_reset(struct spi_imx_data *spi_imx)
+{
+	writel(1, spi_imx->base + MXC_RESET);
+}
+
+static struct spi_imx_devtype_data imx1_cspi_devtype_data = {
+	.intctrl = mx1_intctrl,
+	.config = mx1_config,
+	.trigger = mx1_trigger,
+	.rx_available = mx1_rx_available,
+	.reset = mx1_reset,
+	.fifo_size = 8,
+	.has_dmamode = false,
+	.dynamic_burst = false,
+	.has_slavemode = false,
+	.devtype = IMX1_CSPI,
+};
+
+static struct spi_imx_devtype_data imx21_cspi_devtype_data = {
+	.intctrl = mx21_intctrl,
+	.config = mx21_config,
+	.trigger = mx21_trigger,
+	.rx_available = mx21_rx_available,
+	.reset = mx21_reset,
+	.fifo_size = 8,
+	.has_dmamode = false,
+	.dynamic_burst = false,
+	.has_slavemode = false,
+	.devtype = IMX21_CSPI,
+};
+
+static struct spi_imx_devtype_data imx27_cspi_devtype_data = {
+	/* i.mx27 cspi shares the functions with i.mx21 one */
+	.intctrl = mx21_intctrl,
+	.config = mx21_config,
+	.trigger = mx21_trigger,
+	.rx_available = mx21_rx_available,
+	.reset = mx21_reset,
+	.fifo_size = 8,
+	.has_dmamode = false,
+	.dynamic_burst = false,
+	.has_slavemode = false,
+	.devtype = IMX27_CSPI,
+};
+
+static struct spi_imx_devtype_data imx31_cspi_devtype_data = {
+	.intctrl = mx31_intctrl,
+	.config = mx31_config,
+	.trigger = mx31_trigger,
+	.rx_available = mx31_rx_available,
+	.reset = mx31_reset,
+	.fifo_size = 8,
+	.has_dmamode = false,
+	.dynamic_burst = false,
+	.has_slavemode = false,
+	.devtype = IMX31_CSPI,
+};
+
+static struct spi_imx_devtype_data imx35_cspi_devtype_data = {
+	/* i.mx35 and later cspi shares the functions with i.mx31 one */
+	.intctrl = mx31_intctrl,
+	.config = mx31_config,
+	.trigger = mx31_trigger,
+	.rx_available = mx31_rx_available,
+	.reset = mx31_reset,
+	.fifo_size = 8,
+	.has_dmamode = true,
+	.dynamic_burst = false,
+	.has_slavemode = false,
+	.devtype = IMX35_CSPI,
+};
+
+static struct spi_imx_devtype_data imx51_ecspi_devtype_data = {
+	.intctrl = mx51_ecspi_intctrl,
+	.config = mx51_ecspi_config,
+	.trigger = mx51_ecspi_trigger,
+	.rx_available = mx51_ecspi_rx_available,
+	.reset = mx51_ecspi_reset,
+	.fifo_size = 64,
+	.has_dmamode = true,
+	.dynamic_burst = true,
+	.has_slavemode = true,
+	.disable = mx51_ecspi_disable,
+	.devtype = IMX51_ECSPI,
+};
+
+static struct spi_imx_devtype_data imx53_ecspi_devtype_data = {
+	.intctrl = mx51_ecspi_intctrl,
+	.config = mx51_ecspi_config,
+	.trigger = mx51_ecspi_trigger,
+	.rx_available = mx51_ecspi_rx_available,
+	.reset = mx51_ecspi_reset,
+	.fifo_size = 64,
+	.has_dmamode = true,
+	.has_slavemode = true,
+	.disable = mx51_ecspi_disable,
+	.devtype = IMX53_ECSPI,
+};
+
+static const struct platform_device_id spi_imx_devtype[] = {
+	{
+		.name = "imx1-cspi",
+		.driver_data = (kernel_ulong_t) &imx1_cspi_devtype_data,
+	}, {
+		.name = "imx21-cspi",
+		.driver_data = (kernel_ulong_t) &imx21_cspi_devtype_data,
+	}, {
+		.name = "imx27-cspi",
+		.driver_data = (kernel_ulong_t) &imx27_cspi_devtype_data,
+	}, {
+		.name = "imx31-cspi",
+		.driver_data = (kernel_ulong_t) &imx31_cspi_devtype_data,
+	}, {
+		.name = "imx35-cspi",
+		.driver_data = (kernel_ulong_t) &imx35_cspi_devtype_data,
+	}, {
+		.name = "imx51-ecspi",
+		.driver_data = (kernel_ulong_t) &imx51_ecspi_devtype_data,
+	}, {
+		.name = "imx53-ecspi",
+		.driver_data = (kernel_ulong_t) &imx53_ecspi_devtype_data,
+	}, {
+		/* sentinel */
+	}
+};
+
+static const struct of_device_id spi_imx_dt_ids[] = {
+	{ .compatible = "fsl,imx1-cspi", .data = &imx1_cspi_devtype_data, },
+	{ .compatible = "fsl,imx21-cspi", .data = &imx21_cspi_devtype_data, },
+	{ .compatible = "fsl,imx27-cspi", .data = &imx27_cspi_devtype_data, },
+	{ .compatible = "fsl,imx31-cspi", .data = &imx31_cspi_devtype_data, },
+	{ .compatible = "fsl,imx35-cspi", .data = &imx35_cspi_devtype_data, },
+	{ .compatible = "fsl,imx51-ecspi", .data = &imx51_ecspi_devtype_data, },
+	{ .compatible = "fsl,imx53-ecspi", .data = &imx53_ecspi_devtype_data, },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, spi_imx_dt_ids);
+
+static void spi_imx_chipselect(struct spi_device *spi, int is_active)
+{
+	int active = is_active != BITBANG_CS_INACTIVE;
+	int dev_is_lowactive = !(spi->mode & SPI_CS_HIGH);
+
+	if (spi->mode & SPI_NO_CS)
+		return;
+
+	if (!gpio_is_valid(spi->cs_gpio))
+		return;
+
+	gpio_set_value(spi->cs_gpio, dev_is_lowactive ^ active);
+}
+
+static void spi_imx_set_burst_len(struct spi_imx_data *spi_imx, int n_bits)
+{
+	u32 ctrl;
+
+	ctrl = readl(spi_imx->base + MX51_ECSPI_CTRL);
+	ctrl &= ~MX51_ECSPI_CTRL_BL_MASK;
+	ctrl |= ((n_bits - 1) << MX51_ECSPI_CTRL_BL_OFFSET);
+	writel(ctrl, spi_imx->base + MX51_ECSPI_CTRL);
+}
+
+static void spi_imx_push(struct spi_imx_data *spi_imx)
+{
+	unsigned int burst_len, fifo_words;
+
+	if (spi_imx->dynamic_burst)
+		fifo_words = 4;
+	else
+		fifo_words = spi_imx_bytes_per_word(spi_imx->bits_per_word);
+	/*
+	 * Reload the FIFO when the remaining bytes to be transferred in the
+	 * current burst is 0. This only applies when bits_per_word is a
+	 * multiple of 8.
+	 */
+	if (!spi_imx->remainder) {
+		if (spi_imx->dynamic_burst) {
+
+			/* We need to deal unaligned data first */
+			burst_len = spi_imx->count % MX51_ECSPI_CTRL_MAX_BURST;
+
+			if (!burst_len)
+				burst_len = MX51_ECSPI_CTRL_MAX_BURST;
+
+			spi_imx_set_burst_len(spi_imx, burst_len * 8);
+
+			spi_imx->remainder = burst_len;
+		} else {
+			spi_imx->remainder = fifo_words;
+		}
+	}
+
+	while (spi_imx->txfifo < spi_imx->devtype_data->fifo_size) {
+		if (!spi_imx->count)
+			break;
+		if (spi_imx->dynamic_burst &&
+		    spi_imx->txfifo >=  DIV_ROUND_UP(spi_imx->remainder,
+						     fifo_words))
+			break;
+		spi_imx->tx(spi_imx);
+		spi_imx->txfifo++;
+	}
+
+	if (!spi_imx->slave_mode)
+		spi_imx->devtype_data->trigger(spi_imx);
+}
+
+static irqreturn_t spi_imx_isr(int irq, void *dev_id)
+{
+	struct spi_imx_data *spi_imx = dev_id;
+
+	while (spi_imx->txfifo &&
+	       spi_imx->devtype_data->rx_available(spi_imx)) {
+		spi_imx->rx(spi_imx);
+		spi_imx->txfifo--;
+	}
+
+	if (spi_imx->count) {
+		spi_imx_push(spi_imx);
+		return IRQ_HANDLED;
+	}
+
+	if (spi_imx->txfifo) {
+		/* No data left to push, but still waiting for rx data,
+		 * enable receive data available interrupt.
+		 */
+		spi_imx->devtype_data->intctrl(
+				spi_imx, MXC_INT_RR);
+		return IRQ_HANDLED;
+	}
+
+	spi_imx->devtype_data->intctrl(spi_imx, 0);
+	complete(&spi_imx->xfer_done);
+
+	return IRQ_HANDLED;
+}
+
+static int spi_imx_dma_configure(struct spi_master *master)
+{
+	int ret;
+	enum dma_slave_buswidth buswidth;
+	struct dma_slave_config rx = {}, tx = {};
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(master);
+
+	switch (spi_imx_bytes_per_word(spi_imx->bits_per_word)) {
+	case 4:
+		buswidth = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		break;
+	case 2:
+		buswidth = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		break;
+	case 1:
+		buswidth = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	tx.direction = DMA_MEM_TO_DEV;
+	tx.dst_addr = spi_imx->base_phys + MXC_CSPITXDATA;
+	tx.dst_addr_width = buswidth;
+	tx.dst_maxburst = spi_imx->wml;
+	ret = dmaengine_slave_config(master->dma_tx, &tx);
+	if (ret) {
+		dev_err(spi_imx->dev, "TX dma configuration failed with %d\n", ret);
+		return ret;
+	}
+
+	rx.direction = DMA_DEV_TO_MEM;
+	rx.src_addr = spi_imx->base_phys + MXC_CSPIRXDATA;
+	rx.src_addr_width = buswidth;
+	rx.src_maxburst = spi_imx->wml;
+	ret = dmaengine_slave_config(master->dma_rx, &rx);
+	if (ret) {
+		dev_err(spi_imx->dev, "RX dma configuration failed with %d\n", ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static int spi_imx_setupxfer(struct spi_device *spi,
+				 struct spi_transfer *t)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(spi->master);
+	int ret;
+
+	if (!t)
+		return 0;
+
+	spi_imx->bits_per_word = t->bits_per_word;
+	spi_imx->speed_hz  = t->speed_hz;
+
+	/*
+	 * Initialize the functions for transfer. To transfer non byte-aligned
+	 * words, we have to use multiple word-size bursts, we can't use
+	 * dynamic_burst in that case.
+	 */
+	if (spi_imx->devtype_data->dynamic_burst && !spi_imx->slave_mode &&
+	    (spi_imx->bits_per_word == 8 ||
+	    spi_imx->bits_per_word == 16 ||
+	    spi_imx->bits_per_word == 32)) {
+
+		spi_imx->rx = spi_imx_buf_rx_swap;
+		spi_imx->tx = spi_imx_buf_tx_swap;
+		spi_imx->dynamic_burst = 1;
+
+	} else {
+		if (spi_imx->bits_per_word <= 8) {
+			spi_imx->rx = spi_imx_buf_rx_u8;
+			spi_imx->tx = spi_imx_buf_tx_u8;
+		} else if (spi_imx->bits_per_word <= 16) {
+			spi_imx->rx = spi_imx_buf_rx_u16;
+			spi_imx->tx = spi_imx_buf_tx_u16;
+		} else {
+			spi_imx->rx = spi_imx_buf_rx_u32;
+			spi_imx->tx = spi_imx_buf_tx_u32;
+		}
+		spi_imx->dynamic_burst = 0;
+	}
+
+	if (spi_imx_can_dma(spi_imx->bitbang.master, spi, t))
+		spi_imx->usedma = 1;
+	else
+		spi_imx->usedma = 0;
+
+	if (spi_imx->usedma) {
+		ret = spi_imx_dma_configure(spi->master);
+		if (ret)
+			return ret;
+	}
+
+	if (is_imx53_ecspi(spi_imx) && spi_imx->slave_mode) {
+		spi_imx->rx = mx53_ecspi_rx_slave;
+		spi_imx->tx = mx53_ecspi_tx_slave;
+		spi_imx->slave_burst = t->len;
+	}
+
+	spi_imx->devtype_data->config(spi);
+
+	return 0;
+}
+
+static void spi_imx_sdma_exit(struct spi_imx_data *spi_imx)
+{
+	struct spi_master *master = spi_imx->bitbang.master;
+
+	if (master->dma_rx) {
+		dma_release_channel(master->dma_rx);
+		master->dma_rx = NULL;
+	}
+
+	if (master->dma_tx) {
+		dma_release_channel(master->dma_tx);
+		master->dma_tx = NULL;
+	}
+}
+
+static int spi_imx_sdma_init(struct device *dev, struct spi_imx_data *spi_imx,
+			     struct spi_master *master)
+{
+	int ret;
+
+	/* use pio mode for i.mx6dl chip TKT238285 */
+	if (of_machine_is_compatible("fsl,imx6dl"))
+		return 0;
+
+	spi_imx->wml = spi_imx->devtype_data->fifo_size / 2;
+
+	/* Prepare for TX DMA: */
+	master->dma_tx = dma_request_slave_channel_reason(dev, "tx");
+	if (IS_ERR(master->dma_tx)) {
+		ret = PTR_ERR(master->dma_tx);
+		dev_dbg(dev, "can't get the TX DMA channel, error %d!\n", ret);
+		master->dma_tx = NULL;
+		goto err;
+	}
+
+	/* Prepare for RX : */
+	master->dma_rx = dma_request_slave_channel_reason(dev, "rx");
+	if (IS_ERR(master->dma_rx)) {
+		ret = PTR_ERR(master->dma_rx);
+		dev_dbg(dev, "can't get the RX DMA channel, error %d\n", ret);
+		master->dma_rx = NULL;
+		goto err;
+	}
+
+	init_completion(&spi_imx->dma_rx_completion);
+	init_completion(&spi_imx->dma_tx_completion);
+	master->can_dma = spi_imx_can_dma;
+	master->max_dma_len = MAX_SDMA_BD_BYTES;
+	spi_imx->bitbang.master->flags = SPI_MASTER_MUST_RX |
+					 SPI_MASTER_MUST_TX;
+
+	return 0;
+err:
+	spi_imx_sdma_exit(spi_imx);
+	return ret;
+}
+
+static void spi_imx_dma_rx_callback(void *cookie)
+{
+	struct spi_imx_data *spi_imx = (struct spi_imx_data *)cookie;
+
+	complete(&spi_imx->dma_rx_completion);
+}
+
+static void spi_imx_dma_tx_callback(void *cookie)
+{
+	struct spi_imx_data *spi_imx = (struct spi_imx_data *)cookie;
+
+	complete(&spi_imx->dma_tx_completion);
+}
+
+static int spi_imx_calculate_timeout(struct spi_imx_data *spi_imx, int size)
+{
+	unsigned long timeout = 0;
+
+	/* Time with actual data transfer and CS change delay related to HW */
+	timeout = (8 + 4) * size / spi_imx->spi_bus_clk;
+
+	/* Add extra second for scheduler related activities */
+	timeout += 1;
+
+	/* Double calculated timeout */
+	return msecs_to_jiffies(2 * timeout * MSEC_PER_SEC);
+}
+
+static int spi_imx_dma_transfer(struct spi_imx_data *spi_imx,
+				struct spi_transfer *transfer)
+{
+	struct dma_async_tx_descriptor *desc_tx, *desc_rx;
+	unsigned long transfer_timeout;
+	unsigned long timeout;
+	struct spi_master *master = spi_imx->bitbang.master;
+	struct sg_table *tx = &transfer->tx_sg, *rx = &transfer->rx_sg;
+
+	/*
+	 * The TX DMA setup starts the transfer, so make sure RX is configured
+	 * before TX.
+	 */
+	desc_rx = dmaengine_prep_slave_sg(master->dma_rx,
+				rx->sgl, rx->nents, DMA_DEV_TO_MEM,
+				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!desc_rx)
+		return -EINVAL;
+
+	desc_rx->callback = spi_imx_dma_rx_callback;
+	desc_rx->callback_param = (void *)spi_imx;
+	dmaengine_submit(desc_rx);
+	reinit_completion(&spi_imx->dma_rx_completion);
+	dma_async_issue_pending(master->dma_rx);
+
+	desc_tx = dmaengine_prep_slave_sg(master->dma_tx,
+				tx->sgl, tx->nents, DMA_MEM_TO_DEV,
+				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!desc_tx) {
+		dmaengine_terminate_all(master->dma_tx);
+		return -EINVAL;
+	}
+
+	desc_tx->callback = spi_imx_dma_tx_callback;
+	desc_tx->callback_param = (void *)spi_imx;
+	dmaengine_submit(desc_tx);
+	reinit_completion(&spi_imx->dma_tx_completion);
+	dma_async_issue_pending(master->dma_tx);
+
+	transfer_timeout = spi_imx_calculate_timeout(spi_imx, transfer->len);
+
+	/* Wait SDMA to finish the data transfer.*/
+	timeout = wait_for_completion_timeout(&spi_imx->dma_tx_completion,
+						transfer_timeout);
+	if (!timeout) {
+		dev_err(spi_imx->dev, "I/O Error in DMA TX\n");
+		dmaengine_terminate_all(master->dma_tx);
+		dmaengine_terminate_all(master->dma_rx);
+		return -ETIMEDOUT;
+	}
+
+	timeout = wait_for_completion_timeout(&spi_imx->dma_rx_completion,
+					      transfer_timeout);
+	if (!timeout) {
+		dev_err(&master->dev, "I/O Error in DMA RX\n");
+		spi_imx->devtype_data->reset(spi_imx);
+		dmaengine_terminate_all(master->dma_rx);
+		return -ETIMEDOUT;
+	}
+
+	return transfer->len;
+}
+
+static int spi_imx_pio_transfer(struct spi_device *spi,
+				struct spi_transfer *transfer)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(spi->master);
+	unsigned long transfer_timeout;
+	unsigned long timeout;
+
+	spi_imx->tx_buf = transfer->tx_buf;
+	spi_imx->rx_buf = transfer->rx_buf;
+	spi_imx->count = transfer->len;
+	spi_imx->txfifo = 0;
+	spi_imx->remainder = 0;
+
+	reinit_completion(&spi_imx->xfer_done);
+
+	spi_imx_push(spi_imx);
+
+	spi_imx->devtype_data->intctrl(spi_imx, MXC_INT_TE);
+
+	transfer_timeout = spi_imx_calculate_timeout(spi_imx, transfer->len);
+
+	timeout = wait_for_completion_timeout(&spi_imx->xfer_done,
+					      transfer_timeout);
+	if (!timeout) {
+		dev_err(&spi->dev, "I/O Error in PIO\n");
+		spi_imx->devtype_data->reset(spi_imx);
+		return -ETIMEDOUT;
+	}
+
+	return transfer->len;
+}
+
+static int spi_imx_pio_transfer_slave(struct spi_device *spi,
+				      struct spi_transfer *transfer)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(spi->master);
+	int ret = transfer->len;
+
+	if (is_imx53_ecspi(spi_imx) &&
+	    transfer->len > MX53_MAX_TRANSFER_BYTES) {
+		dev_err(&spi->dev, "Transaction too big, max size is %d bytes\n",
+			MX53_MAX_TRANSFER_BYTES);
+		return -EMSGSIZE;
+	}
+
+	spi_imx->tx_buf = transfer->tx_buf;
+	spi_imx->rx_buf = transfer->rx_buf;
+	spi_imx->count = transfer->len;
+	spi_imx->txfifo = 0;
+	spi_imx->remainder = 0;
+
+	reinit_completion(&spi_imx->xfer_done);
+	spi_imx->slave_aborted = false;
+
+	spi_imx_push(spi_imx);
+
+	spi_imx->devtype_data->intctrl(spi_imx, MXC_INT_TE | MXC_INT_RDR);
+
+	if (wait_for_completion_interruptible(&spi_imx->xfer_done) ||
+	    spi_imx->slave_aborted) {
+		dev_dbg(&spi->dev, "interrupted\n");
+		ret = -EINTR;
+	}
+
+	/* ecspi has a HW issue when works in Slave mode,
+	 * after 64 words writtern to TXFIFO, even TXFIFO becomes empty,
+	 * ECSPI_TXDATA keeps shift out the last word data,
+	 * so we have to disable ECSPI when in slave mode after the
+	 * transfer completes
+	 */
+	if (spi_imx->devtype_data->disable)
+		spi_imx->devtype_data->disable(spi_imx);
+
+	return ret;
+}
+
+static int spi_imx_transfer(struct spi_device *spi,
+				struct spi_transfer *transfer)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(spi->master);
+
+	/* flush rxfifo before transfer */
+	while (spi_imx->devtype_data->rx_available(spi_imx))
+		spi_imx->rx(spi_imx);
+
+	if (spi_imx->slave_mode)
+		return spi_imx_pio_transfer_slave(spi, transfer);
+
+	if (spi_imx->usedma)
+		return spi_imx_dma_transfer(spi_imx, transfer);
+	else
+		return spi_imx_pio_transfer(spi, transfer);
+}
+
+static int spi_imx_setup(struct spi_device *spi)
+{
+	dev_dbg(&spi->dev, "%s: mode %d, %u bpw, %d hz\n", __func__,
+		 spi->mode, spi->bits_per_word, spi->max_speed_hz);
+
+	if (spi->mode & SPI_NO_CS)
+		return 0;
+
+	if (gpio_is_valid(spi->cs_gpio))
+		gpio_direction_output(spi->cs_gpio,
+				      spi->mode & SPI_CS_HIGH ? 0 : 1);
+
+	spi_imx_chipselect(spi, BITBANG_CS_INACTIVE);
+
+	return 0;
+}
+
+static void spi_imx_cleanup(struct spi_device *spi)
+{
+}
+
+static int
+spi_imx_prepare_message(struct spi_master *master, struct spi_message *msg)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_enable(spi_imx->clk_per);
+	if (ret)
+		return ret;
+
+	ret = clk_enable(spi_imx->clk_ipg);
+	if (ret) {
+		clk_disable(spi_imx->clk_per);
+		return ret;
+	}
+
+	return 0;
+}
+
+static int
+spi_imx_unprepare_message(struct spi_master *master, struct spi_message *msg)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(master);
+
+	clk_disable(spi_imx->clk_ipg);
+	clk_disable(spi_imx->clk_per);
+	return 0;
+}
+
+static int spi_imx_slave_abort(struct spi_master *master)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(master);
+
+	spi_imx->slave_aborted = true;
+	complete(&spi_imx->xfer_done);
+
+	return 0;
+}
+
+static int spi_imx_probe(struct platform_device *pdev)
+{
+	struct device_node *np = pdev->dev.of_node;
+	const struct of_device_id *of_id =
+			of_match_device(spi_imx_dt_ids, &pdev->dev);
+	struct spi_imx_master *mxc_platform_info =
+			dev_get_platdata(&pdev->dev);
+	struct spi_master *master;
+	struct spi_imx_data *spi_imx;
+	struct resource *res;
+	int i, ret, irq, spi_drctl;
+	const struct spi_imx_devtype_data *devtype_data = of_id ? of_id->data :
+		(struct spi_imx_devtype_data *)pdev->id_entry->driver_data;
+	bool slave_mode;
+
+	if (!np && !mxc_platform_info) {
+		dev_err(&pdev->dev, "can't get the platform data\n");
+		return -EINVAL;
+	}
+
+	slave_mode = devtype_data->has_slavemode &&
+			of_property_read_bool(np, "spi-slave");
+	if (slave_mode)
+		master = spi_alloc_slave(&pdev->dev,
+					 sizeof(struct spi_imx_data));
+	else
+		master = spi_alloc_master(&pdev->dev,
+					  sizeof(struct spi_imx_data));
+	if (!master)
+		return -ENOMEM;
+
+	ret = of_property_read_u32(np, "fsl,spi-rdy-drctl", &spi_drctl);
+	if ((ret < 0) || (spi_drctl >= 0x3)) {
+		/* '11' is reserved */
+		spi_drctl = 0;
+	}
+
+	platform_set_drvdata(pdev, master);
+
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(1, 32);
+	master->bus_num = np ? -1 : pdev->id;
+
+	spi_imx = spi_master_get_devdata(master);
+	spi_imx->bitbang.master = master;
+	spi_imx->dev = &pdev->dev;
+	spi_imx->slave_mode = slave_mode;
+
+	spi_imx->devtype_data = devtype_data;
+
+	/* Get number of chip selects, either platform data or OF */
+	if (mxc_platform_info) {
+		master->num_chipselect = mxc_platform_info->num_chipselect;
+		if (mxc_platform_info->chipselect) {
+			master->cs_gpios = devm_kcalloc(&master->dev,
+				master->num_chipselect, sizeof(int),
+				GFP_KERNEL);
+			if (!master->cs_gpios)
+				return -ENOMEM;
+
+			for (i = 0; i < master->num_chipselect; i++)
+				master->cs_gpios[i] = mxc_platform_info->chipselect[i];
+		}
+	} else {
+		u32 num_cs;
+
+		if (!of_property_read_u32(np, "num-cs", &num_cs))
+			master->num_chipselect = num_cs;
+		/* If not preset, default value of 1 is used */
+	}
+
+	spi_imx->bitbang.chipselect = spi_imx_chipselect;
+	spi_imx->bitbang.setup_transfer = spi_imx_setupxfer;
+	spi_imx->bitbang.txrx_bufs = spi_imx_transfer;
+	spi_imx->bitbang.master->setup = spi_imx_setup;
+	spi_imx->bitbang.master->cleanup = spi_imx_cleanup;
+	spi_imx->bitbang.master->prepare_message = spi_imx_prepare_message;
+	spi_imx->bitbang.master->unprepare_message = spi_imx_unprepare_message;
+	spi_imx->bitbang.master->slave_abort = spi_imx_slave_abort;
+	spi_imx->bitbang.master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH \
+					     | SPI_NO_CS;
+	if (is_imx35_cspi(spi_imx) || is_imx51_ecspi(spi_imx) ||
+	    is_imx53_ecspi(spi_imx))
+		spi_imx->bitbang.master->mode_bits |= SPI_LOOP | SPI_READY;
+
+	spi_imx->spi_drctl = spi_drctl;
+
+	init_completion(&spi_imx->xfer_done);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	spi_imx->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(spi_imx->base)) {
+		ret = PTR_ERR(spi_imx->base);
+		goto out_master_put;
+	}
+	spi_imx->base_phys = res->start;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		ret = irq;
+		goto out_master_put;
+	}
+
+	ret = devm_request_irq(&pdev->dev, irq, spi_imx_isr, 0,
+			       dev_name(&pdev->dev), spi_imx);
+	if (ret) {
+		dev_err(&pdev->dev, "can't get irq%d: %d\n", irq, ret);
+		goto out_master_put;
+	}
+
+	spi_imx->clk_ipg = devm_clk_get(&pdev->dev, "ipg");
+	if (IS_ERR(spi_imx->clk_ipg)) {
+		ret = PTR_ERR(spi_imx->clk_ipg);
+		goto out_master_put;
+	}
+
+	spi_imx->clk_per = devm_clk_get(&pdev->dev, "per");
+	if (IS_ERR(spi_imx->clk_per)) {
+		ret = PTR_ERR(spi_imx->clk_per);
+		goto out_master_put;
+	}
+
+	ret = clk_prepare_enable(spi_imx->clk_per);
+	if (ret)
+		goto out_master_put;
+
+	ret = clk_prepare_enable(spi_imx->clk_ipg);
+	if (ret)
+		goto out_put_per;
+
+	spi_imx->spi_clk = clk_get_rate(spi_imx->clk_per);
+	/*
+	 * Only validated on i.mx35 and i.mx6 now, can remove the constraint
+	 * if validated on other chips.
+	 */
+	if (spi_imx->devtype_data->has_dmamode) {
+		ret = spi_imx_sdma_init(&pdev->dev, spi_imx, master);
+		if (ret == -EPROBE_DEFER)
+			goto out_clk_put;
+
+		if (ret < 0)
+			dev_err(&pdev->dev, "dma setup error %d, use pio\n",
+				ret);
+	}
+
+	spi_imx->devtype_data->reset(spi_imx);
+
+	spi_imx->devtype_data->intctrl(spi_imx, 0);
+
+	master->dev.of_node = pdev->dev.of_node;
+	ret = spi_bitbang_start(&spi_imx->bitbang);
+	if (ret) {
+		dev_err(&pdev->dev, "bitbang start failed with %d\n", ret);
+		goto out_clk_put;
+	}
+
+	/* Request GPIO CS lines, if any */
+	if (!spi_imx->slave_mode && master->cs_gpios) {
+		for (i = 0; i < master->num_chipselect; i++) {
+			if (!gpio_is_valid(master->cs_gpios[i]))
+				continue;
+
+			ret = devm_gpio_request(&pdev->dev,
+						master->cs_gpios[i],
+						DRIVER_NAME);
+			if (ret) {
+				dev_err(&pdev->dev, "Can't get CS GPIO %i\n",
+					master->cs_gpios[i]);
+				goto out_spi_bitbang;
+			}
+		}
+	}
+
+	dev_info(&pdev->dev, "probed\n");
+
+	clk_disable(spi_imx->clk_ipg);
+	clk_disable(spi_imx->clk_per);
+	return ret;
+
+out_spi_bitbang:
+	spi_bitbang_stop(&spi_imx->bitbang);
+out_clk_put:
+	clk_disable_unprepare(spi_imx->clk_ipg);
+out_put_per:
+	clk_disable_unprepare(spi_imx->clk_per);
+out_master_put:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int spi_imx_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(master);
+	int ret;
+
+	spi_bitbang_stop(&spi_imx->bitbang);
+
+	ret = clk_enable(spi_imx->clk_per);
+	if (ret)
+		return ret;
+
+	ret = clk_enable(spi_imx->clk_ipg);
+	if (ret) {
+		clk_disable(spi_imx->clk_per);
+		return ret;
+	}
+
+	writel(0, spi_imx->base + MXC_CSPICTRL);
+	clk_disable_unprepare(spi_imx->clk_ipg);
+	clk_disable_unprepare(spi_imx->clk_per);
+	spi_imx_sdma_exit(spi_imx);
+	spi_master_put(master);
+
+	return 0;
+}
+
+static struct platform_driver spi_imx_driver = {
+	.driver = {
+		   .name = DRIVER_NAME,
+		   .of_match_table = spi_imx_dt_ids,
+		   },
+	.id_table = spi_imx_devtype,
+	.probe = spi_imx_probe,
+	.remove = spi_imx_remove,
+};
+module_platform_driver(spi_imx_driver);
+
+MODULE_DESCRIPTION("SPI Controller driver");
+MODULE_AUTHOR("Sascha Hauer, Pengutronix");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:" DRIVER_NAME);
diff --git a/drivers/spi/spi-iproc-qspi.c b/drivers/spi/spi-iproc-qspi.c
new file mode 100644
index 0000000..be6ccb2
--- /dev/null
+++ b/drivers/spi/spi-iproc-qspi.c
@@ -0,0 +1,163 @@
+/*
+ * Copyright 2016 Broadcom Limited
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/device.h>
+#include <linux/io.h>
+#include <linux/ioport.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+
+#include "spi-bcm-qspi.h"
+
+#define INTR_BASE_BIT_SHIFT			0x02
+#define INTR_COUNT				0x07
+
+struct bcm_iproc_intc {
+	struct bcm_qspi_soc_intc soc_intc;
+	struct platform_device *pdev;
+	void __iomem *int_reg;
+	void __iomem *int_status_reg;
+	spinlock_t soclock;
+	bool big_endian;
+};
+
+static u32 bcm_iproc_qspi_get_l2_int_status(struct bcm_qspi_soc_intc *soc_intc)
+{
+	struct bcm_iproc_intc *priv =
+			container_of(soc_intc, struct bcm_iproc_intc, soc_intc);
+	void __iomem *mmio = priv->int_status_reg;
+	int i;
+	u32 val = 0, sts = 0;
+
+	for (i = 0; i < INTR_COUNT; i++) {
+		if (bcm_qspi_readl(priv->big_endian, mmio + (i * 4)))
+			val |= 1UL << i;
+	}
+
+	if (val & INTR_MSPI_DONE_MASK)
+		sts |= MSPI_DONE;
+
+	if (val & BSPI_LR_INTERRUPTS_ALL)
+		sts |= BSPI_DONE;
+
+	if (val & BSPI_LR_INTERRUPTS_ERROR)
+		sts |= BSPI_ERR;
+
+	return sts;
+}
+
+static void bcm_iproc_qspi_int_ack(struct bcm_qspi_soc_intc *soc_intc, int type)
+{
+	struct bcm_iproc_intc *priv =
+			container_of(soc_intc, struct bcm_iproc_intc, soc_intc);
+	void __iomem *mmio = priv->int_status_reg;
+	u32 mask = get_qspi_mask(type);
+	int i;
+
+	for (i = 0; i < INTR_COUNT; i++) {
+		if (mask & (1UL << i))
+			bcm_qspi_writel(priv->big_endian, 1, mmio + (i * 4));
+	}
+}
+
+static void bcm_iproc_qspi_int_set(struct bcm_qspi_soc_intc *soc_intc, int type,
+				   bool en)
+{
+	struct bcm_iproc_intc *priv =
+			container_of(soc_intc, struct bcm_iproc_intc, soc_intc);
+	void __iomem *mmio = priv->int_reg;
+	u32 mask = get_qspi_mask(type);
+	u32 val;
+	unsigned long flags;
+
+	spin_lock_irqsave(&priv->soclock, flags);
+
+	val = bcm_qspi_readl(priv->big_endian, mmio);
+
+	if (en)
+		val = val | (mask << INTR_BASE_BIT_SHIFT);
+	else
+		val = val & ~(mask << INTR_BASE_BIT_SHIFT);
+
+	bcm_qspi_writel(priv->big_endian, val, mmio);
+
+	spin_unlock_irqrestore(&priv->soclock, flags);
+}
+
+static int bcm_iproc_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct bcm_iproc_intc *priv;
+	struct bcm_qspi_soc_intc *soc_intc;
+	struct resource *res;
+
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+	soc_intc = &priv->soc_intc;
+	priv->pdev = pdev;
+
+	spin_lock_init(&priv->soclock);
+
+	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "intr_regs");
+	priv->int_reg = devm_ioremap_resource(dev, res);
+	if (IS_ERR(priv->int_reg))
+		return PTR_ERR(priv->int_reg);
+
+	res = platform_get_resource_byname(pdev, IORESOURCE_MEM,
+					   "intr_status_reg");
+	priv->int_status_reg = devm_ioremap_resource(dev, res);
+	if (IS_ERR(priv->int_status_reg))
+		return PTR_ERR(priv->int_status_reg);
+
+	priv->big_endian = of_device_is_big_endian(dev->of_node);
+
+	bcm_iproc_qspi_int_ack(soc_intc, MSPI_BSPI_DONE);
+	bcm_iproc_qspi_int_set(soc_intc, MSPI_BSPI_DONE, false);
+
+	soc_intc->bcm_qspi_int_ack = bcm_iproc_qspi_int_ack;
+	soc_intc->bcm_qspi_int_set = bcm_iproc_qspi_int_set;
+	soc_intc->bcm_qspi_get_int_status = bcm_iproc_qspi_get_l2_int_status;
+
+	return bcm_qspi_probe(pdev, soc_intc);
+}
+
+static int bcm_iproc_remove(struct platform_device *pdev)
+{
+	return bcm_qspi_remove(pdev);
+}
+
+static const struct of_device_id bcm_iproc_of_match[] = {
+	{ .compatible = "brcm,spi-nsp-qspi" },
+	{ .compatible = "brcm,spi-ns2-qspi" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, bcm_iproc_of_match);
+
+static struct platform_driver bcm_iproc_driver = {
+	.probe			= bcm_iproc_probe,
+	.remove			= bcm_iproc_remove,
+	.driver = {
+		.name		= "bcm_iproc",
+		.pm		= &bcm_qspi_pm_ops,
+		.of_match_table = bcm_iproc_of_match,
+	}
+};
+module_platform_driver(bcm_iproc_driver);
+
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Kamal Dasu");
+MODULE_DESCRIPTION("SPI flash driver for Broadcom iProc SoCs");
diff --git a/drivers/spi/spi-jcore.c b/drivers/spi/spi-jcore.c
new file mode 100644
index 0000000..702fe57
--- /dev/null
+++ b/drivers/spi/spi-jcore.c
@@ -0,0 +1,233 @@
+/*
+ * J-Core SPI controller driver
+ *
+ * Copyright (C) 2012-2016 Smart Energy Instruments, Inc.
+ *
+ * Current version by Rich Felker
+ * Based loosely on initial version by Oleksandr G Zhadan
+ *
+ */
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/errno.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/io.h>
+#include <linux/of.h>
+#include <linux/delay.h>
+
+#define DRV_NAME	"jcore_spi"
+
+#define CTRL_REG	0x0
+#define DATA_REG	0x4
+
+#define JCORE_SPI_CTRL_XMIT		0x02
+#define JCORE_SPI_STAT_BUSY		0x02
+#define JCORE_SPI_CTRL_LOOP		0x08
+#define JCORE_SPI_CTRL_CS_BITS		0x15
+
+#define JCORE_SPI_WAIT_RDY_MAX_LOOP	2000000
+
+struct jcore_spi {
+	struct spi_master *master;
+	void __iomem *base;
+	unsigned int cs_reg;
+	unsigned int speed_reg;
+	unsigned int speed_hz;
+	unsigned int clock_freq;
+};
+
+static int jcore_spi_wait(void __iomem *ctrl_reg)
+{
+	unsigned timeout = JCORE_SPI_WAIT_RDY_MAX_LOOP;
+
+	do {
+		if (!(readl(ctrl_reg) & JCORE_SPI_STAT_BUSY))
+			return 0;
+		cpu_relax();
+	} while (--timeout);
+
+	return -EBUSY;
+}
+
+static void jcore_spi_program(struct jcore_spi *hw)
+{
+	void __iomem *ctrl_reg = hw->base + CTRL_REG;
+
+	if (jcore_spi_wait(ctrl_reg))
+		dev_err(hw->master->dev.parent,
+			"timeout waiting to program ctrl reg.\n");
+
+	writel(hw->cs_reg | hw->speed_reg, ctrl_reg);
+}
+
+static void jcore_spi_chipsel(struct spi_device *spi, bool value)
+{
+	struct jcore_spi *hw = spi_master_get_devdata(spi->master);
+	u32 csbit = 1U << (2 * spi->chip_select);
+
+	dev_dbg(hw->master->dev.parent, "chipselect %d\n", spi->chip_select);
+
+	if (value)
+		hw->cs_reg |= csbit;
+	else
+		hw->cs_reg &= ~csbit;
+
+	jcore_spi_program(hw);
+}
+
+static void jcore_spi_baudrate(struct jcore_spi *hw, int speed)
+{
+	if (speed == hw->speed_hz) return;
+	hw->speed_hz = speed;
+	if (speed >= hw->clock_freq / 2)
+		hw->speed_reg = 0;
+	else
+		hw->speed_reg = ((hw->clock_freq / 2 / speed) - 1) << 27;
+	jcore_spi_program(hw);
+	dev_dbg(hw->master->dev.parent, "speed=%d reg=0x%x\n",
+		speed, hw->speed_reg);
+}
+
+static int jcore_spi_txrx(struct spi_master *master, struct spi_device *spi,
+			  struct spi_transfer *t)
+{
+	struct jcore_spi *hw = spi_master_get_devdata(master);
+
+	void __iomem *ctrl_reg = hw->base + CTRL_REG;
+	void __iomem *data_reg = hw->base + DATA_REG;
+	u32 xmit;
+
+	/* data buffers */
+	const unsigned char *tx;
+	unsigned char *rx;
+	unsigned int len;
+	unsigned int count;
+
+	jcore_spi_baudrate(hw, t->speed_hz);
+
+	xmit = hw->cs_reg | hw->speed_reg | JCORE_SPI_CTRL_XMIT;
+	tx = t->tx_buf;
+	rx = t->rx_buf;
+	len = t->len;
+
+	for (count = 0; count < len; count++) {
+		if (jcore_spi_wait(ctrl_reg))
+			break;
+
+		writel(tx ? *tx++ : 0, data_reg);
+		writel(xmit, ctrl_reg);
+
+		if (jcore_spi_wait(ctrl_reg))
+			break;
+
+		if (rx)
+			*rx++ = readl(data_reg);
+	}
+
+	spi_finalize_current_transfer(master);
+
+	if (count < len)
+		return -EREMOTEIO;
+
+	return 0;
+}
+
+static int jcore_spi_probe(struct platform_device *pdev)
+{
+	struct device_node *node = pdev->dev.of_node;
+	struct jcore_spi *hw;
+	struct spi_master *master;
+	struct resource *res;
+	u32 clock_freq;
+	struct clk *clk;
+	int err = -ENODEV;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct jcore_spi));
+	if (!master)
+		return err;
+
+	/* Setup the master state. */
+	master->num_chipselect = 3;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->transfer_one = jcore_spi_txrx;
+	master->set_cs = jcore_spi_chipsel;
+	master->dev.of_node = node;
+	master->bus_num = pdev->id;
+
+	hw = spi_master_get_devdata(master);
+	hw->master = master;
+	platform_set_drvdata(pdev, hw);
+
+	/* Find and map our resources */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res)
+		goto exit_busy;
+	if (!devm_request_mem_region(&pdev->dev, res->start,
+				     resource_size(res), pdev->name))
+		goto exit_busy;
+	hw->base = devm_ioremap_nocache(&pdev->dev, res->start,
+					resource_size(res));
+	if (!hw->base)
+		goto exit_busy;
+
+	/*
+	 * The SPI clock rate controlled via a configurable clock divider
+	 * which is applied to the reference clock. A 50 MHz reference is
+	 * most suitable for obtaining standard SPI clock rates, but some
+	 * designs may have a different reference clock, and the DT must
+	 * make the driver aware so that it can properly program the
+	 * requested rate. If the clock is omitted, 50 MHz is assumed.
+	 */
+	clock_freq = 50000000;
+	clk = devm_clk_get(&pdev->dev, "ref_clk");
+	if (!IS_ERR(clk)) {
+		if (clk_prepare_enable(clk) == 0) {
+			clock_freq = clk_get_rate(clk);
+			clk_disable_unprepare(clk);
+		} else
+			dev_warn(&pdev->dev, "could not enable ref_clk\n");
+	}
+	hw->clock_freq = clock_freq;
+
+	/* Initialize all CS bits to high. */
+	hw->cs_reg = JCORE_SPI_CTRL_CS_BITS;
+	jcore_spi_baudrate(hw, 400000);
+
+	/* Register our spi controller */
+	err = devm_spi_register_master(&pdev->dev, master);
+	if (err)
+		goto exit;
+
+	return 0;
+
+exit_busy:
+	err = -EBUSY;
+exit:
+	spi_master_put(master);
+	return err;
+}
+
+static const struct of_device_id jcore_spi_of_match[] = {
+	{ .compatible = "jcore,spi2" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, jcore_spi_of_match);
+
+static struct platform_driver jcore_spi_driver = {
+	.probe = jcore_spi_probe,
+	.driver = {
+		.name = DRV_NAME,
+		.of_match_table = jcore_spi_of_match,
+	},
+};
+
+module_platform_driver(jcore_spi_driver);
+
+MODULE_DESCRIPTION("J-Core SPI driver");
+MODULE_AUTHOR("Rich Felker <dalias@libc.org>");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:" DRV_NAME);
diff --git a/drivers/spi/spi-lantiq-ssc.c b/drivers/spi/spi-lantiq-ssc.c
new file mode 100644
index 0000000..d597661
--- /dev/null
+++ b/drivers/spi/spi-lantiq-ssc.c
@@ -0,0 +1,989 @@
+/*
+ * Copyright (C) 2011-2015 Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
+ * Copyright (C) 2016 Hauke Mehrtens <hauke@hauke-m.de>
+ *
+ * This program is free software; you can distribute it and/or modify it
+ * under the terms of the GNU General Public License (Version 2) as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/clk.h>
+#include <linux/io.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/sched.h>
+#include <linux/completion.h>
+#include <linux/spinlock.h>
+#include <linux/err.h>
+#include <linux/gpio.h>
+#include <linux/pm_runtime.h>
+#include <linux/spi/spi.h>
+
+#ifdef CONFIG_LANTIQ
+#include <lantiq_soc.h>
+#endif
+
+#define LTQ_SPI_RX_IRQ_NAME	"spi_rx"
+#define LTQ_SPI_TX_IRQ_NAME	"spi_tx"
+#define LTQ_SPI_ERR_IRQ_NAME	"spi_err"
+#define LTQ_SPI_FRM_IRQ_NAME	"spi_frm"
+
+#define LTQ_SPI_CLC		0x00
+#define LTQ_SPI_PISEL		0x04
+#define LTQ_SPI_ID		0x08
+#define LTQ_SPI_CON		0x10
+#define LTQ_SPI_STAT		0x14
+#define LTQ_SPI_WHBSTATE	0x18
+#define LTQ_SPI_TB		0x20
+#define LTQ_SPI_RB		0x24
+#define LTQ_SPI_RXFCON		0x30
+#define LTQ_SPI_TXFCON		0x34
+#define LTQ_SPI_FSTAT		0x38
+#define LTQ_SPI_BRT		0x40
+#define LTQ_SPI_BRSTAT		0x44
+#define LTQ_SPI_SFCON		0x60
+#define LTQ_SPI_SFSTAT		0x64
+#define LTQ_SPI_GPOCON		0x70
+#define LTQ_SPI_GPOSTAT		0x74
+#define LTQ_SPI_FPGO		0x78
+#define LTQ_SPI_RXREQ		0x80
+#define LTQ_SPI_RXCNT		0x84
+#define LTQ_SPI_DMACON		0xec
+#define LTQ_SPI_IRNEN		0xf4
+#define LTQ_SPI_IRNICR		0xf8
+#define LTQ_SPI_IRNCR		0xfc
+
+#define LTQ_SPI_CLC_SMC_S	16	/* Clock divider for sleep mode */
+#define LTQ_SPI_CLC_SMC_M	(0xFF << LTQ_SPI_CLC_SMC_S)
+#define LTQ_SPI_CLC_RMC_S	8	/* Clock divider for normal run mode */
+#define LTQ_SPI_CLC_RMC_M	(0xFF << LTQ_SPI_CLC_RMC_S)
+#define LTQ_SPI_CLC_DISS	BIT(1)	/* Disable status bit */
+#define LTQ_SPI_CLC_DISR	BIT(0)	/* Disable request bit */
+
+#define LTQ_SPI_ID_TXFS_S	24	/* Implemented TX FIFO size */
+#define LTQ_SPI_ID_TXFS_M	(0x3F << LTQ_SPI_ID_TXFS_S)
+#define LTQ_SPI_ID_RXFS_S	16	/* Implemented RX FIFO size */
+#define LTQ_SPI_ID_RXFS_M	(0x3F << LTQ_SPI_ID_RXFS_S)
+#define LTQ_SPI_ID_MOD_S	8	/* Module ID */
+#define LTQ_SPI_ID_MOD_M	(0xff << LTQ_SPI_ID_MOD_S)
+#define LTQ_SPI_ID_CFG_S	5	/* DMA interface support */
+#define LTQ_SPI_ID_CFG_M	(1 << LTQ_SPI_ID_CFG_S)
+#define LTQ_SPI_ID_REV_M	0x1F	/* Hardware revision number */
+
+#define LTQ_SPI_CON_BM_S	16	/* Data width selection */
+#define LTQ_SPI_CON_BM_M	(0x1F << LTQ_SPI_CON_BM_S)
+#define LTQ_SPI_CON_EM		BIT(24)	/* Echo mode */
+#define LTQ_SPI_CON_IDLE	BIT(23)	/* Idle bit value */
+#define LTQ_SPI_CON_ENBV	BIT(22)	/* Enable byte valid control */
+#define LTQ_SPI_CON_RUEN	BIT(12)	/* Receive underflow error enable */
+#define LTQ_SPI_CON_TUEN	BIT(11)	/* Transmit underflow error enable */
+#define LTQ_SPI_CON_AEN		BIT(10)	/* Abort error enable */
+#define LTQ_SPI_CON_REN		BIT(9)	/* Receive overflow error enable */
+#define LTQ_SPI_CON_TEN		BIT(8)	/* Transmit overflow error enable */
+#define LTQ_SPI_CON_LB		BIT(7)	/* Loopback control */
+#define LTQ_SPI_CON_PO		BIT(6)	/* Clock polarity control */
+#define LTQ_SPI_CON_PH		BIT(5)	/* Clock phase control */
+#define LTQ_SPI_CON_HB		BIT(4)	/* Heading control */
+#define LTQ_SPI_CON_RXOFF	BIT(1)	/* Switch receiver off */
+#define LTQ_SPI_CON_TXOFF	BIT(0)	/* Switch transmitter off */
+
+#define LTQ_SPI_STAT_RXBV_S	28
+#define LTQ_SPI_STAT_RXBV_M	(0x7 << LTQ_SPI_STAT_RXBV_S)
+#define LTQ_SPI_STAT_BSY	BIT(13)	/* Busy flag */
+#define LTQ_SPI_STAT_RUE	BIT(12)	/* Receive underflow error flag */
+#define LTQ_SPI_STAT_TUE	BIT(11)	/* Transmit underflow error flag */
+#define LTQ_SPI_STAT_AE		BIT(10)	/* Abort error flag */
+#define LTQ_SPI_STAT_RE		BIT(9)	/* Receive error flag */
+#define LTQ_SPI_STAT_TE		BIT(8)	/* Transmit error flag */
+#define LTQ_SPI_STAT_ME		BIT(7)	/* Mode error flag */
+#define LTQ_SPI_STAT_MS		BIT(1)	/* Master/slave select bit */
+#define LTQ_SPI_STAT_EN		BIT(0)	/* Enable bit */
+#define LTQ_SPI_STAT_ERRORS	(LTQ_SPI_STAT_ME | LTQ_SPI_STAT_TE | \
+				 LTQ_SPI_STAT_RE | LTQ_SPI_STAT_AE | \
+				 LTQ_SPI_STAT_TUE | LTQ_SPI_STAT_RUE)
+
+#define LTQ_SPI_WHBSTATE_SETTUE	BIT(15)	/* Set transmit underflow error flag */
+#define LTQ_SPI_WHBSTATE_SETAE	BIT(14)	/* Set abort error flag */
+#define LTQ_SPI_WHBSTATE_SETRE	BIT(13)	/* Set receive error flag */
+#define LTQ_SPI_WHBSTATE_SETTE	BIT(12)	/* Set transmit error flag */
+#define LTQ_SPI_WHBSTATE_CLRTUE	BIT(11)	/* Clear transmit underflow error flag */
+#define LTQ_SPI_WHBSTATE_CLRAE	BIT(10)	/* Clear abort error flag */
+#define LTQ_SPI_WHBSTATE_CLRRE	BIT(9)	/* Clear receive error flag */
+#define LTQ_SPI_WHBSTATE_CLRTE	BIT(8)	/* Clear transmit error flag */
+#define LTQ_SPI_WHBSTATE_SETME	BIT(7)	/* Set mode error flag */
+#define LTQ_SPI_WHBSTATE_CLRME	BIT(6)	/* Clear mode error flag */
+#define LTQ_SPI_WHBSTATE_SETRUE	BIT(5)	/* Set receive underflow error flag */
+#define LTQ_SPI_WHBSTATE_CLRRUE	BIT(4)	/* Clear receive underflow error flag */
+#define LTQ_SPI_WHBSTATE_SETMS	BIT(3)	/* Set master select bit */
+#define LTQ_SPI_WHBSTATE_CLRMS	BIT(2)	/* Clear master select bit */
+#define LTQ_SPI_WHBSTATE_SETEN	BIT(1)	/* Set enable bit (operational mode) */
+#define LTQ_SPI_WHBSTATE_CLREN	BIT(0)	/* Clear enable bit (config mode */
+#define LTQ_SPI_WHBSTATE_CLR_ERRORS	(LTQ_SPI_WHBSTATE_CLRRUE | \
+					 LTQ_SPI_WHBSTATE_CLRME | \
+					 LTQ_SPI_WHBSTATE_CLRTE | \
+					 LTQ_SPI_WHBSTATE_CLRRE | \
+					 LTQ_SPI_WHBSTATE_CLRAE | \
+					 LTQ_SPI_WHBSTATE_CLRTUE)
+
+#define LTQ_SPI_RXFCON_RXFITL_S	8	/* FIFO interrupt trigger level */
+#define LTQ_SPI_RXFCON_RXFITL_M	(0x3F << LTQ_SPI_RXFCON_RXFITL_S)
+#define LTQ_SPI_RXFCON_RXFLU	BIT(1)	/* FIFO flush */
+#define LTQ_SPI_RXFCON_RXFEN	BIT(0)	/* FIFO enable */
+
+#define LTQ_SPI_TXFCON_TXFITL_S	8	/* FIFO interrupt trigger level */
+#define LTQ_SPI_TXFCON_TXFITL_M	(0x3F << LTQ_SPI_TXFCON_TXFITL_S)
+#define LTQ_SPI_TXFCON_TXFLU	BIT(1)	/* FIFO flush */
+#define LTQ_SPI_TXFCON_TXFEN	BIT(0)	/* FIFO enable */
+
+#define LTQ_SPI_FSTAT_RXFFL_S	0
+#define LTQ_SPI_FSTAT_RXFFL_M	(0x3f << LTQ_SPI_FSTAT_RXFFL_S)
+#define LTQ_SPI_FSTAT_TXFFL_S	8
+#define LTQ_SPI_FSTAT_TXFFL_M	(0x3f << LTQ_SPI_FSTAT_TXFFL_S)
+
+#define LTQ_SPI_GPOCON_ISCSBN_S	8
+#define LTQ_SPI_GPOCON_INVOUTN_S	0
+
+#define LTQ_SPI_FGPO_SETOUTN_S	8
+#define LTQ_SPI_FGPO_CLROUTN_S	0
+
+#define LTQ_SPI_RXREQ_RXCNT_M	0xFFFF	/* Receive count value */
+#define LTQ_SPI_RXCNT_TODO_M	0xFFFF	/* Recevie to-do value */
+
+#define LTQ_SPI_IRNEN_TFI	BIT(4)	/* TX finished interrupt */
+#define LTQ_SPI_IRNEN_F		BIT(3)	/* Frame end interrupt request */
+#define LTQ_SPI_IRNEN_E		BIT(2)	/* Error end interrupt request */
+#define LTQ_SPI_IRNEN_T_XWAY	BIT(1)	/* Transmit end interrupt request */
+#define LTQ_SPI_IRNEN_R_XWAY	BIT(0)	/* Receive end interrupt request */
+#define LTQ_SPI_IRNEN_R_XRX	BIT(1)	/* Transmit end interrupt request */
+#define LTQ_SPI_IRNEN_T_XRX	BIT(0)	/* Receive end interrupt request */
+#define LTQ_SPI_IRNEN_ALL	0x1F
+
+struct lantiq_ssc_hwcfg {
+	unsigned int irnen_r;
+	unsigned int irnen_t;
+};
+
+struct lantiq_ssc_spi {
+	struct spi_master		*master;
+	struct device			*dev;
+	void __iomem			*regbase;
+	struct clk			*spi_clk;
+	struct clk			*fpi_clk;
+	const struct lantiq_ssc_hwcfg	*hwcfg;
+
+	spinlock_t			lock;
+	struct workqueue_struct		*wq;
+	struct work_struct		work;
+
+	const u8			*tx;
+	u8				*rx;
+	unsigned int			tx_todo;
+	unsigned int			rx_todo;
+	unsigned int			bits_per_word;
+	unsigned int			speed_hz;
+	unsigned int			tx_fifo_size;
+	unsigned int			rx_fifo_size;
+	unsigned int			base_cs;
+};
+
+static u32 lantiq_ssc_readl(const struct lantiq_ssc_spi *spi, u32 reg)
+{
+	return __raw_readl(spi->regbase + reg);
+}
+
+static void lantiq_ssc_writel(const struct lantiq_ssc_spi *spi, u32 val,
+			      u32 reg)
+{
+	__raw_writel(val, spi->regbase + reg);
+}
+
+static void lantiq_ssc_maskl(const struct lantiq_ssc_spi *spi, u32 clr,
+			     u32 set, u32 reg)
+{
+	u32 val = __raw_readl(spi->regbase + reg);
+
+	val &= ~clr;
+	val |= set;
+	__raw_writel(val, spi->regbase + reg);
+}
+
+static unsigned int tx_fifo_level(const struct lantiq_ssc_spi *spi)
+{
+	u32 fstat = lantiq_ssc_readl(spi, LTQ_SPI_FSTAT);
+
+	return (fstat & LTQ_SPI_FSTAT_TXFFL_M) >> LTQ_SPI_FSTAT_TXFFL_S;
+}
+
+static unsigned int rx_fifo_level(const struct lantiq_ssc_spi *spi)
+{
+	u32 fstat = lantiq_ssc_readl(spi, LTQ_SPI_FSTAT);
+
+	return fstat & LTQ_SPI_FSTAT_RXFFL_M;
+}
+
+static unsigned int tx_fifo_free(const struct lantiq_ssc_spi *spi)
+{
+	return spi->tx_fifo_size - tx_fifo_level(spi);
+}
+
+static void rx_fifo_reset(const struct lantiq_ssc_spi *spi)
+{
+	u32 val = spi->rx_fifo_size << LTQ_SPI_RXFCON_RXFITL_S;
+
+	val |= LTQ_SPI_RXFCON_RXFEN | LTQ_SPI_RXFCON_RXFLU;
+	lantiq_ssc_writel(spi, val, LTQ_SPI_RXFCON);
+}
+
+static void tx_fifo_reset(const struct lantiq_ssc_spi *spi)
+{
+	u32 val = 1 << LTQ_SPI_TXFCON_TXFITL_S;
+
+	val |= LTQ_SPI_TXFCON_TXFEN | LTQ_SPI_TXFCON_TXFLU;
+	lantiq_ssc_writel(spi, val, LTQ_SPI_TXFCON);
+}
+
+static void rx_fifo_flush(const struct lantiq_ssc_spi *spi)
+{
+	lantiq_ssc_maskl(spi, 0, LTQ_SPI_RXFCON_RXFLU, LTQ_SPI_RXFCON);
+}
+
+static void tx_fifo_flush(const struct lantiq_ssc_spi *spi)
+{
+	lantiq_ssc_maskl(spi, 0, LTQ_SPI_TXFCON_TXFLU, LTQ_SPI_TXFCON);
+}
+
+static void hw_enter_config_mode(const struct lantiq_ssc_spi *spi)
+{
+	lantiq_ssc_writel(spi, LTQ_SPI_WHBSTATE_CLREN, LTQ_SPI_WHBSTATE);
+}
+
+static void hw_enter_active_mode(const struct lantiq_ssc_spi *spi)
+{
+	lantiq_ssc_writel(spi, LTQ_SPI_WHBSTATE_SETEN, LTQ_SPI_WHBSTATE);
+}
+
+static void hw_setup_speed_hz(const struct lantiq_ssc_spi *spi,
+			      unsigned int max_speed_hz)
+{
+	u32 spi_clk, brt;
+
+	/*
+	 * SPI module clock is derived from FPI bus clock dependent on
+	 * divider value in CLC.RMS which is always set to 1.
+	 *
+	 *                 f_SPI
+	 * baudrate = --------------
+	 *             2 * (BR + 1)
+	 */
+	spi_clk = clk_get_rate(spi->fpi_clk) / 2;
+
+	if (max_speed_hz > spi_clk)
+		brt = 0;
+	else
+		brt = spi_clk / max_speed_hz - 1;
+
+	if (brt > 0xFFFF)
+		brt = 0xFFFF;
+
+	dev_dbg(spi->dev, "spi_clk %u, max_speed_hz %u, brt %u\n",
+		spi_clk, max_speed_hz, brt);
+
+	lantiq_ssc_writel(spi, brt, LTQ_SPI_BRT);
+}
+
+static void hw_setup_bits_per_word(const struct lantiq_ssc_spi *spi,
+				   unsigned int bits_per_word)
+{
+	u32 bm;
+
+	/* CON.BM value = bits_per_word - 1 */
+	bm = (bits_per_word - 1) << LTQ_SPI_CON_BM_S;
+
+	lantiq_ssc_maskl(spi, LTQ_SPI_CON_BM_M, bm, LTQ_SPI_CON);
+}
+
+static void hw_setup_clock_mode(const struct lantiq_ssc_spi *spi,
+				unsigned int mode)
+{
+	u32 con_set = 0, con_clr = 0;
+
+	/*
+	 * SPI mode mapping in CON register:
+	 * Mode CPOL CPHA CON.PO CON.PH
+	 *  0    0    0      0      1
+	 *  1    0    1      0      0
+	 *  2    1    0      1      1
+	 *  3    1    1      1      0
+	 */
+	if (mode & SPI_CPHA)
+		con_clr |= LTQ_SPI_CON_PH;
+	else
+		con_set |= LTQ_SPI_CON_PH;
+
+	if (mode & SPI_CPOL)
+		con_set |= LTQ_SPI_CON_PO | LTQ_SPI_CON_IDLE;
+	else
+		con_clr |= LTQ_SPI_CON_PO | LTQ_SPI_CON_IDLE;
+
+	/* Set heading control */
+	if (mode & SPI_LSB_FIRST)
+		con_clr |= LTQ_SPI_CON_HB;
+	else
+		con_set |= LTQ_SPI_CON_HB;
+
+	/* Set loopback mode */
+	if (mode & SPI_LOOP)
+		con_set |= LTQ_SPI_CON_LB;
+	else
+		con_clr |= LTQ_SPI_CON_LB;
+
+	lantiq_ssc_maskl(spi, con_clr, con_set, LTQ_SPI_CON);
+}
+
+static void lantiq_ssc_hw_init(const struct lantiq_ssc_spi *spi)
+{
+	const struct lantiq_ssc_hwcfg *hwcfg = spi->hwcfg;
+
+	/*
+	 * Set clock divider for run mode to 1 to
+	 * run at same frequency as FPI bus
+	 */
+	lantiq_ssc_writel(spi, 1 << LTQ_SPI_CLC_RMC_S, LTQ_SPI_CLC);
+
+	/* Put controller into config mode */
+	hw_enter_config_mode(spi);
+
+	/* Clear error flags */
+	lantiq_ssc_maskl(spi, 0, LTQ_SPI_WHBSTATE_CLR_ERRORS, LTQ_SPI_WHBSTATE);
+
+	/* Enable error checking, disable TX/RX */
+	lantiq_ssc_writel(spi, LTQ_SPI_CON_RUEN | LTQ_SPI_CON_AEN |
+		LTQ_SPI_CON_TEN | LTQ_SPI_CON_REN | LTQ_SPI_CON_TXOFF |
+		LTQ_SPI_CON_RXOFF, LTQ_SPI_CON);
+
+	/* Setup default SPI mode */
+	hw_setup_bits_per_word(spi, spi->bits_per_word);
+	hw_setup_clock_mode(spi, SPI_MODE_0);
+
+	/* Enable master mode and clear error flags */
+	lantiq_ssc_writel(spi, LTQ_SPI_WHBSTATE_SETMS |
+			       LTQ_SPI_WHBSTATE_CLR_ERRORS,
+			       LTQ_SPI_WHBSTATE);
+
+	/* Reset GPIO/CS registers */
+	lantiq_ssc_writel(spi, 0, LTQ_SPI_GPOCON);
+	lantiq_ssc_writel(spi, 0xFF00, LTQ_SPI_FPGO);
+
+	/* Enable and flush FIFOs */
+	rx_fifo_reset(spi);
+	tx_fifo_reset(spi);
+
+	/* Enable interrupts */
+	lantiq_ssc_writel(spi, hwcfg->irnen_t | hwcfg->irnen_r |
+			  LTQ_SPI_IRNEN_E, LTQ_SPI_IRNEN);
+}
+
+static int lantiq_ssc_setup(struct spi_device *spidev)
+{
+	struct spi_master *master = spidev->master;
+	struct lantiq_ssc_spi *spi = spi_master_get_devdata(master);
+	unsigned int cs = spidev->chip_select;
+	u32 gpocon;
+
+	/* GPIOs are used for CS */
+	if (gpio_is_valid(spidev->cs_gpio))
+		return 0;
+
+	dev_dbg(spi->dev, "using internal chipselect %u\n", cs);
+
+	if (cs < spi->base_cs) {
+		dev_err(spi->dev,
+			"chipselect %i too small (min %i)\n", cs, spi->base_cs);
+		return -EINVAL;
+	}
+
+	/* set GPO pin to CS mode */
+	gpocon = 1 << ((cs - spi->base_cs) + LTQ_SPI_GPOCON_ISCSBN_S);
+
+	/* invert GPO pin */
+	if (spidev->mode & SPI_CS_HIGH)
+		gpocon |= 1 << (cs - spi->base_cs);
+
+	lantiq_ssc_maskl(spi, 0, gpocon, LTQ_SPI_GPOCON);
+
+	return 0;
+}
+
+static int lantiq_ssc_prepare_message(struct spi_master *master,
+				      struct spi_message *message)
+{
+	struct lantiq_ssc_spi *spi = spi_master_get_devdata(master);
+
+	hw_enter_config_mode(spi);
+	hw_setup_clock_mode(spi, message->spi->mode);
+	hw_enter_active_mode(spi);
+
+	return 0;
+}
+
+static void hw_setup_transfer(struct lantiq_ssc_spi *spi,
+			      struct spi_device *spidev, struct spi_transfer *t)
+{
+	unsigned int speed_hz = t->speed_hz;
+	unsigned int bits_per_word = t->bits_per_word;
+	u32 con;
+
+	if (bits_per_word != spi->bits_per_word ||
+		speed_hz != spi->speed_hz) {
+		hw_enter_config_mode(spi);
+		hw_setup_speed_hz(spi, speed_hz);
+		hw_setup_bits_per_word(spi, bits_per_word);
+		hw_enter_active_mode(spi);
+
+		spi->speed_hz = speed_hz;
+		spi->bits_per_word = bits_per_word;
+	}
+
+	/* Configure transmitter and receiver */
+	con = lantiq_ssc_readl(spi, LTQ_SPI_CON);
+	if (t->tx_buf)
+		con &= ~LTQ_SPI_CON_TXOFF;
+	else
+		con |= LTQ_SPI_CON_TXOFF;
+
+	if (t->rx_buf)
+		con &= ~LTQ_SPI_CON_RXOFF;
+	else
+		con |= LTQ_SPI_CON_RXOFF;
+
+	lantiq_ssc_writel(spi, con, LTQ_SPI_CON);
+}
+
+static int lantiq_ssc_unprepare_message(struct spi_master *master,
+					struct spi_message *message)
+{
+	struct lantiq_ssc_spi *spi = spi_master_get_devdata(master);
+
+	flush_workqueue(spi->wq);
+
+	/* Disable transmitter and receiver while idle */
+	lantiq_ssc_maskl(spi, 0, LTQ_SPI_CON_TXOFF | LTQ_SPI_CON_RXOFF,
+			 LTQ_SPI_CON);
+
+	return 0;
+}
+
+static void tx_fifo_write(struct lantiq_ssc_spi *spi)
+{
+	const u8 *tx8;
+	const u16 *tx16;
+	const u32 *tx32;
+	u32 data;
+	unsigned int tx_free = tx_fifo_free(spi);
+
+	while (spi->tx_todo && tx_free) {
+		switch (spi->bits_per_word) {
+		case 2 ... 8:
+			tx8 = spi->tx;
+			data = *tx8;
+			spi->tx_todo--;
+			spi->tx++;
+			break;
+		case 16:
+			tx16 = (u16 *) spi->tx;
+			data = *tx16;
+			spi->tx_todo -= 2;
+			spi->tx += 2;
+			break;
+		case 32:
+			tx32 = (u32 *) spi->tx;
+			data = *tx32;
+			spi->tx_todo -= 4;
+			spi->tx += 4;
+			break;
+		default:
+			WARN_ON(1);
+			data = 0;
+			break;
+		}
+
+		lantiq_ssc_writel(spi, data, LTQ_SPI_TB);
+		tx_free--;
+	}
+}
+
+static void rx_fifo_read_full_duplex(struct lantiq_ssc_spi *spi)
+{
+	u8 *rx8;
+	u16 *rx16;
+	u32 *rx32;
+	u32 data;
+	unsigned int rx_fill = rx_fifo_level(spi);
+
+	while (rx_fill) {
+		data = lantiq_ssc_readl(spi, LTQ_SPI_RB);
+
+		switch (spi->bits_per_word) {
+		case 2 ... 8:
+			rx8 = spi->rx;
+			*rx8 = data;
+			spi->rx_todo--;
+			spi->rx++;
+			break;
+		case 16:
+			rx16 = (u16 *) spi->rx;
+			*rx16 = data;
+			spi->rx_todo -= 2;
+			spi->rx += 2;
+			break;
+		case 32:
+			rx32 = (u32 *) spi->rx;
+			*rx32 = data;
+			spi->rx_todo -= 4;
+			spi->rx += 4;
+			break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+
+		rx_fill--;
+	}
+}
+
+static void rx_fifo_read_half_duplex(struct lantiq_ssc_spi *spi)
+{
+	u32 data, *rx32;
+	u8 *rx8;
+	unsigned int rxbv, shift;
+	unsigned int rx_fill = rx_fifo_level(spi);
+
+	/*
+	 * In RX-only mode the bits per word value is ignored by HW. A value
+	 * of 32 is used instead. Thus all 4 bytes per FIFO must be read.
+	 * If remaining RX bytes are less than 4, the FIFO must be read
+	 * differently. The amount of received and valid bytes is indicated
+	 * by STAT.RXBV register value.
+	 */
+	while (rx_fill) {
+		if (spi->rx_todo < 4)  {
+			rxbv = (lantiq_ssc_readl(spi, LTQ_SPI_STAT) &
+				LTQ_SPI_STAT_RXBV_M) >> LTQ_SPI_STAT_RXBV_S;
+			data = lantiq_ssc_readl(spi, LTQ_SPI_RB);
+
+			shift = (rxbv - 1) * 8;
+			rx8 = spi->rx;
+
+			while (rxbv) {
+				*rx8++ = (data >> shift) & 0xFF;
+				rxbv--;
+				shift -= 8;
+				spi->rx_todo--;
+				spi->rx++;
+			}
+		} else {
+			data = lantiq_ssc_readl(spi, LTQ_SPI_RB);
+			rx32 = (u32 *) spi->rx;
+
+			*rx32++ = data;
+			spi->rx_todo -= 4;
+			spi->rx += 4;
+		}
+		rx_fill--;
+	}
+}
+
+static void rx_request(struct lantiq_ssc_spi *spi)
+{
+	unsigned int rxreq, rxreq_max;
+
+	/*
+	 * To avoid receive overflows at high clocks it is better to request
+	 * only the amount of bytes that fits into all FIFOs. This value
+	 * depends on the FIFO size implemented in hardware.
+	 */
+	rxreq = spi->rx_todo;
+	rxreq_max = spi->rx_fifo_size * 4;
+	if (rxreq > rxreq_max)
+		rxreq = rxreq_max;
+
+	lantiq_ssc_writel(spi, rxreq, LTQ_SPI_RXREQ);
+}
+
+static irqreturn_t lantiq_ssc_xmit_interrupt(int irq, void *data)
+{
+	struct lantiq_ssc_spi *spi = data;
+
+	if (spi->tx) {
+		if (spi->rx && spi->rx_todo)
+			rx_fifo_read_full_duplex(spi);
+
+		if (spi->tx_todo)
+			tx_fifo_write(spi);
+		else if (!tx_fifo_level(spi))
+			goto completed;
+	} else if (spi->rx) {
+		if (spi->rx_todo) {
+			rx_fifo_read_half_duplex(spi);
+
+			if (spi->rx_todo)
+				rx_request(spi);
+			else
+				goto completed;
+		} else {
+			goto completed;
+		}
+	}
+
+	return IRQ_HANDLED;
+
+completed:
+	queue_work(spi->wq, &spi->work);
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t lantiq_ssc_err_interrupt(int irq, void *data)
+{
+	struct lantiq_ssc_spi *spi = data;
+	u32 stat = lantiq_ssc_readl(spi, LTQ_SPI_STAT);
+
+	if (!(stat & LTQ_SPI_STAT_ERRORS))
+		return IRQ_NONE;
+
+	if (stat & LTQ_SPI_STAT_RUE)
+		dev_err(spi->dev, "receive underflow error\n");
+	if (stat & LTQ_SPI_STAT_TUE)
+		dev_err(spi->dev, "transmit underflow error\n");
+	if (stat & LTQ_SPI_STAT_AE)
+		dev_err(spi->dev, "abort error\n");
+	if (stat & LTQ_SPI_STAT_RE)
+		dev_err(spi->dev, "receive overflow error\n");
+	if (stat & LTQ_SPI_STAT_TE)
+		dev_err(spi->dev, "transmit overflow error\n");
+	if (stat & LTQ_SPI_STAT_ME)
+		dev_err(spi->dev, "mode error\n");
+
+	/* Clear error flags */
+	lantiq_ssc_maskl(spi, 0, LTQ_SPI_WHBSTATE_CLR_ERRORS, LTQ_SPI_WHBSTATE);
+
+	/* set bad status so it can be retried */
+	if (spi->master->cur_msg)
+		spi->master->cur_msg->status = -EIO;
+	queue_work(spi->wq, &spi->work);
+
+	return IRQ_HANDLED;
+}
+
+static int transfer_start(struct lantiq_ssc_spi *spi, struct spi_device *spidev,
+			  struct spi_transfer *t)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&spi->lock, flags);
+
+	spi->tx = t->tx_buf;
+	spi->rx = t->rx_buf;
+
+	if (t->tx_buf) {
+		spi->tx_todo = t->len;
+
+		/* initially fill TX FIFO */
+		tx_fifo_write(spi);
+	}
+
+	if (spi->rx) {
+		spi->rx_todo = t->len;
+
+		/* start shift clock in RX-only mode */
+		if (!spi->tx)
+			rx_request(spi);
+	}
+
+	spin_unlock_irqrestore(&spi->lock, flags);
+
+	return t->len;
+}
+
+/*
+ * The driver only gets an interrupt when the FIFO is empty, but there
+ * is an additional shift register from which the data is written to
+ * the wire. We get the last interrupt when the controller starts to
+ * write the last word to the wire, not when it is finished. Do busy
+ * waiting till it finishes.
+ */
+static void lantiq_ssc_bussy_work(struct work_struct *work)
+{
+	struct lantiq_ssc_spi *spi;
+	unsigned long long timeout = 8LL * 1000LL;
+	unsigned long end;
+
+	spi = container_of(work, typeof(*spi), work);
+
+	do_div(timeout, spi->speed_hz);
+	timeout += timeout + 100; /* some tolerance */
+
+	end = jiffies + msecs_to_jiffies(timeout);
+	do {
+		u32 stat = lantiq_ssc_readl(spi, LTQ_SPI_STAT);
+
+		if (!(stat & LTQ_SPI_STAT_BSY)) {
+			spi_finalize_current_transfer(spi->master);
+			return;
+		}
+
+		cond_resched();
+	} while (!time_after_eq(jiffies, end));
+
+	if (spi->master->cur_msg)
+		spi->master->cur_msg->status = -EIO;
+	spi_finalize_current_transfer(spi->master);
+}
+
+static void lantiq_ssc_handle_err(struct spi_master *master,
+				  struct spi_message *message)
+{
+	struct lantiq_ssc_spi *spi = spi_master_get_devdata(master);
+
+	/* flush FIFOs on timeout */
+	rx_fifo_flush(spi);
+	tx_fifo_flush(spi);
+}
+
+static void lantiq_ssc_set_cs(struct spi_device *spidev, bool enable)
+{
+	struct lantiq_ssc_spi *spi = spi_master_get_devdata(spidev->master);
+	unsigned int cs = spidev->chip_select;
+	u32 fgpo;
+
+	if (!!(spidev->mode & SPI_CS_HIGH) == enable)
+		fgpo = (1 << (cs - spi->base_cs));
+	else
+		fgpo = (1 << (cs - spi->base_cs + LTQ_SPI_FGPO_SETOUTN_S));
+
+	lantiq_ssc_writel(spi, fgpo, LTQ_SPI_FPGO);
+}
+
+static int lantiq_ssc_transfer_one(struct spi_master *master,
+				   struct spi_device *spidev,
+				   struct spi_transfer *t)
+{
+	struct lantiq_ssc_spi *spi = spi_master_get_devdata(master);
+
+	hw_setup_transfer(spi, spidev, t);
+
+	return transfer_start(spi, spidev, t);
+}
+
+static const struct lantiq_ssc_hwcfg lantiq_ssc_xway = {
+	.irnen_r = LTQ_SPI_IRNEN_R_XWAY,
+	.irnen_t = LTQ_SPI_IRNEN_T_XWAY,
+};
+
+static const struct lantiq_ssc_hwcfg lantiq_ssc_xrx = {
+	.irnen_r = LTQ_SPI_IRNEN_R_XRX,
+	.irnen_t = LTQ_SPI_IRNEN_T_XRX,
+};
+
+static const struct of_device_id lantiq_ssc_match[] = {
+	{ .compatible = "lantiq,ase-spi", .data = &lantiq_ssc_xway, },
+	{ .compatible = "lantiq,falcon-spi", .data = &lantiq_ssc_xrx, },
+	{ .compatible = "lantiq,xrx100-spi", .data = &lantiq_ssc_xrx, },
+	{},
+};
+MODULE_DEVICE_TABLE(of, lantiq_ssc_match);
+
+static int lantiq_ssc_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct spi_master *master;
+	struct resource *res;
+	struct lantiq_ssc_spi *spi;
+	const struct lantiq_ssc_hwcfg *hwcfg;
+	const struct of_device_id *match;
+	int err, rx_irq, tx_irq, err_irq;
+	u32 id, supports_dma, revision;
+	unsigned int num_cs;
+
+	match = of_match_device(lantiq_ssc_match, dev);
+	if (!match) {
+		dev_err(dev, "no device match\n");
+		return -EINVAL;
+	}
+	hwcfg = match->data;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res) {
+		dev_err(dev, "failed to get resources\n");
+		return -ENXIO;
+	}
+
+	rx_irq = platform_get_irq_byname(pdev, LTQ_SPI_RX_IRQ_NAME);
+	if (rx_irq < 0) {
+		dev_err(dev, "failed to get %s\n", LTQ_SPI_RX_IRQ_NAME);
+		return -ENXIO;
+	}
+
+	tx_irq = platform_get_irq_byname(pdev, LTQ_SPI_TX_IRQ_NAME);
+	if (tx_irq < 0) {
+		dev_err(dev, "failed to get %s\n", LTQ_SPI_TX_IRQ_NAME);
+		return -ENXIO;
+	}
+
+	err_irq = platform_get_irq_byname(pdev, LTQ_SPI_ERR_IRQ_NAME);
+	if (err_irq < 0) {
+		dev_err(dev, "failed to get %s\n", LTQ_SPI_ERR_IRQ_NAME);
+		return -ENXIO;
+	}
+
+	master = spi_alloc_master(dev, sizeof(struct lantiq_ssc_spi));
+	if (!master)
+		return -ENOMEM;
+
+	spi = spi_master_get_devdata(master);
+	spi->master = master;
+	spi->dev = dev;
+	spi->hwcfg = hwcfg;
+	platform_set_drvdata(pdev, spi);
+
+	spi->regbase = devm_ioremap_resource(dev, res);
+	if (IS_ERR(spi->regbase)) {
+		err = PTR_ERR(spi->regbase);
+		goto err_master_put;
+	}
+
+	err = devm_request_irq(dev, rx_irq, lantiq_ssc_xmit_interrupt,
+			       0, LTQ_SPI_RX_IRQ_NAME, spi);
+	if (err)
+		goto err_master_put;
+
+	err = devm_request_irq(dev, tx_irq, lantiq_ssc_xmit_interrupt,
+			       0, LTQ_SPI_TX_IRQ_NAME, spi);
+	if (err)
+		goto err_master_put;
+
+	err = devm_request_irq(dev, err_irq, lantiq_ssc_err_interrupt,
+			       0, LTQ_SPI_ERR_IRQ_NAME, spi);
+	if (err)
+		goto err_master_put;
+
+	spi->spi_clk = devm_clk_get(dev, "gate");
+	if (IS_ERR(spi->spi_clk)) {
+		err = PTR_ERR(spi->spi_clk);
+		goto err_master_put;
+	}
+	err = clk_prepare_enable(spi->spi_clk);
+	if (err)
+		goto err_master_put;
+
+	/*
+	 * Use the old clk_get_fpi() function on Lantiq platform, till it
+	 * supports common clk.
+	 */
+#if defined(CONFIG_LANTIQ) && !defined(CONFIG_COMMON_CLK)
+	spi->fpi_clk = clk_get_fpi();
+#else
+	spi->fpi_clk = clk_get(dev, "freq");
+#endif
+	if (IS_ERR(spi->fpi_clk)) {
+		err = PTR_ERR(spi->fpi_clk);
+		goto err_clk_disable;
+	}
+
+	num_cs = 8;
+	of_property_read_u32(pdev->dev.of_node, "num-cs", &num_cs);
+
+	spi->base_cs = 1;
+	of_property_read_u32(pdev->dev.of_node, "base-cs", &spi->base_cs);
+
+	spin_lock_init(&spi->lock);
+	spi->bits_per_word = 8;
+	spi->speed_hz = 0;
+
+	master->dev.of_node = pdev->dev.of_node;
+	master->num_chipselect = num_cs;
+	master->setup = lantiq_ssc_setup;
+	master->set_cs = lantiq_ssc_set_cs;
+	master->handle_err = lantiq_ssc_handle_err;
+	master->prepare_message = lantiq_ssc_prepare_message;
+	master->unprepare_message = lantiq_ssc_unprepare_message;
+	master->transfer_one = lantiq_ssc_transfer_one;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_LSB_FIRST | SPI_CS_HIGH |
+				SPI_LOOP;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(2, 8) |
+				     SPI_BPW_MASK(16) | SPI_BPW_MASK(32);
+
+	spi->wq = alloc_ordered_workqueue(dev_name(dev), 0);
+	if (!spi->wq) {
+		err = -ENOMEM;
+		goto err_clk_put;
+	}
+	INIT_WORK(&spi->work, lantiq_ssc_bussy_work);
+
+	id = lantiq_ssc_readl(spi, LTQ_SPI_ID);
+	spi->tx_fifo_size = (id & LTQ_SPI_ID_TXFS_M) >> LTQ_SPI_ID_TXFS_S;
+	spi->rx_fifo_size = (id & LTQ_SPI_ID_RXFS_M) >> LTQ_SPI_ID_RXFS_S;
+	supports_dma = (id & LTQ_SPI_ID_CFG_M) >> LTQ_SPI_ID_CFG_S;
+	revision = id & LTQ_SPI_ID_REV_M;
+
+	lantiq_ssc_hw_init(spi);
+
+	dev_info(dev,
+		"Lantiq SSC SPI controller (Rev %i, TXFS %u, RXFS %u, DMA %u)\n",
+		revision, spi->tx_fifo_size, spi->rx_fifo_size, supports_dma);
+
+	err = devm_spi_register_master(dev, master);
+	if (err) {
+		dev_err(dev, "failed to register spi_master\n");
+		goto err_wq_destroy;
+	}
+
+	return 0;
+
+err_wq_destroy:
+	destroy_workqueue(spi->wq);
+err_clk_put:
+	clk_put(spi->fpi_clk);
+err_clk_disable:
+	clk_disable_unprepare(spi->spi_clk);
+err_master_put:
+	spi_master_put(master);
+
+	return err;
+}
+
+static int lantiq_ssc_remove(struct platform_device *pdev)
+{
+	struct lantiq_ssc_spi *spi = platform_get_drvdata(pdev);
+
+	lantiq_ssc_writel(spi, 0, LTQ_SPI_IRNEN);
+	lantiq_ssc_writel(spi, 0, LTQ_SPI_CLC);
+	rx_fifo_flush(spi);
+	tx_fifo_flush(spi);
+	hw_enter_config_mode(spi);
+
+	destroy_workqueue(spi->wq);
+	clk_disable_unprepare(spi->spi_clk);
+	clk_put(spi->fpi_clk);
+
+	return 0;
+}
+
+static struct platform_driver lantiq_ssc_driver = {
+	.probe = lantiq_ssc_probe,
+	.remove = lantiq_ssc_remove,
+	.driver = {
+		.name = "spi-lantiq-ssc",
+		.of_match_table = lantiq_ssc_match,
+	},
+};
+module_platform_driver(lantiq_ssc_driver);
+
+MODULE_DESCRIPTION("Lantiq SSC SPI controller driver");
+MODULE_AUTHOR("Daniel Schwierzeck <daniel.schwierzeck@gmail.com>");
+MODULE_AUTHOR("Hauke Mehrtens <hauke@hauke-m.de>");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:spi-lantiq-ssc");
diff --git a/drivers/spi/spi-lm70llp.c b/drivers/spi/spi-lm70llp.c
new file mode 100644
index 0000000..4549efd
--- /dev/null
+++ b/drivers/spi/spi-lm70llp.c
@@ -0,0 +1,348 @@
+/*
+ * Driver for LM70EVAL-LLP board for the LM70 sensor
+ *
+ * Copyright (C) 2006 Kaiwan N Billimoria <kaiwan@designergraphix.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/parport.h>
+#include <linux/sysfs.h>
+#include <linux/workqueue.h>
+
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+
+/*
+ * The LM70 communicates with a host processor using a 3-wire variant of
+ * the SPI/Microwire bus interface. This driver specifically supports an
+ * NS LM70 LLP Evaluation Board, interfacing to a PC using its parallel
+ * port to bitbang an SPI-parport bridge.  Accordingly, this is an SPI
+ * master controller driver.  The hwmon/lm70 driver is a "SPI protocol
+ * driver", layered on top of this one and usable without the lm70llp.
+ *
+ * Datasheet and Schematic:
+ * The LM70 is a temperature sensor chip from National Semiconductor; its
+ * datasheet is available at http://www.national.com/pf/LM/LM70.html
+ * The schematic for this particular board (the LM70EVAL-LLP) is
+ * available (on page 4) here:
+ *  http://www.national.com/appinfo/tempsensors/files/LM70LLPEVALmanual.pdf
+ *
+ * Also see Documentation/spi/spi-lm70llp.  The SPI<->parport code here is
+ * (heavily) based on spi-butterfly by David Brownell.
+ *
+ * The LM70 LLP connects to the PC parallel port in the following manner:
+ *
+ *   Parallel                 LM70 LLP
+ *     Port      Direction   JP2 Header
+ *  -----------  ---------  ------------
+ *      D0    2      -         -
+ *      D1    3     -->      V+   5
+ *      D2    4     -->      V+   5
+ *      D3    5     -->      V+   5
+ *      D4    6     -->      V+   5
+ *      D5    7     -->      nCS  8
+ *      D6    8     -->      SCLK 3
+ *      D7    9     -->      SI/O 5
+ *     GND   25      -       GND  7
+ *    Select 13     <--      SI/O 1
+ *
+ * Note that parport pin 13 actually gets inverted by the transistor
+ * arrangement which lets either the parport or the LM70 drive the
+ * SI/SO signal (see the schematic for details).
+ */
+
+#define DRVNAME		"spi-lm70llp"
+
+#define lm70_INIT	0xBE
+#define SIO		0x10
+#define nCS		0x20
+#define SCLK		0x40
+
+/*-------------------------------------------------------------------------*/
+
+struct spi_lm70llp {
+	struct spi_bitbang	bitbang;
+	struct parport		*port;
+	struct pardevice	*pd;
+	struct spi_device	*spidev_lm70;
+	struct spi_board_info	info;
+	//struct device		*dev;
+};
+
+/* REVISIT : ugly global ; provides "exclusive open" facility */
+static struct spi_lm70llp *lm70llp;
+
+/*-------------------------------------------------------------------*/
+
+static inline struct spi_lm70llp *spidev_to_pp(struct spi_device *spi)
+{
+	return spi->controller_data;
+}
+
+/*---------------------- LM70 LLP eval board-specific inlines follow */
+
+/* NOTE:  we don't actually need to reread the output values, since they'll
+ * still be what we wrote before.  Plus, going through parport builds in
+ * a ~1ms/operation delay; these SPI transfers could easily be faster.
+ */
+
+static inline void deassertCS(struct spi_lm70llp *pp)
+{
+	u8 data = parport_read_data(pp->port);
+
+	data &= ~0x80;		/* pull D7/SI-out low while de-asserted */
+	parport_write_data(pp->port, data | nCS);
+}
+
+static inline void assertCS(struct spi_lm70llp *pp)
+{
+	u8 data = parport_read_data(pp->port);
+
+	data |= 0x80;		/* pull D7/SI-out high so lm70 drives SO-in */
+	parport_write_data(pp->port, data & ~nCS);
+}
+
+static inline void clkHigh(struct spi_lm70llp *pp)
+{
+	u8 data = parport_read_data(pp->port);
+
+	parport_write_data(pp->port, data | SCLK);
+}
+
+static inline void clkLow(struct spi_lm70llp *pp)
+{
+	u8 data = parport_read_data(pp->port);
+
+	parport_write_data(pp->port, data & ~SCLK);
+}
+
+/*------------------------- SPI-LM70-specific inlines ----------------------*/
+
+static inline void spidelay(unsigned d)
+{
+	udelay(d);
+}
+
+static inline void setsck(struct spi_device *s, int is_on)
+{
+	struct spi_lm70llp *pp = spidev_to_pp(s);
+
+	if (is_on)
+		clkHigh(pp);
+	else
+		clkLow(pp);
+}
+
+static inline void setmosi(struct spi_device *s, int is_on)
+{
+	/* FIXME update D7 ... this way we can put the chip
+	 * into shutdown mode and read the manufacturer ID,
+	 * but we can't put it back into operational mode.
+	 */
+}
+
+/*
+ * getmiso:
+ * Why do we return 0 when the SIO line is high and vice-versa?
+ * The fact is, the lm70 eval board from NS (which this driver drives),
+ * is wired in just such a way : when the lm70's SIO goes high, a transistor
+ * switches it to low reflecting this on the parport (pin 13), and vice-versa.
+ */
+static inline int getmiso(struct spi_device *s)
+{
+	struct spi_lm70llp *pp = spidev_to_pp(s);
+
+	return ((SIO == (parport_read_status(pp->port) & SIO)) ? 0 : 1);
+}
+
+/*--------------------------------------------------------------------*/
+
+#include "spi-bitbang-txrx.h"
+
+static void lm70_chipselect(struct spi_device *spi, int value)
+{
+	struct spi_lm70llp *pp = spidev_to_pp(spi);
+
+	if (value)
+		assertCS(pp);
+	else
+		deassertCS(pp);
+}
+
+/*
+ * Our actual bitbanger routine.
+ */
+static u32 lm70_txrx(struct spi_device *spi, unsigned nsecs, u32 word, u8 bits,
+		     unsigned flags)
+{
+	return bitbang_txrx_be_cpha0(spi, nsecs, 0, flags, word, bits);
+}
+
+static void spi_lm70llp_attach(struct parport *p)
+{
+	struct pardevice	*pd;
+	struct spi_lm70llp	*pp;
+	struct spi_master	*master;
+	int			status;
+	struct pardev_cb	lm70llp_cb;
+
+	if (lm70llp) {
+		pr_warn("spi_lm70llp instance already loaded. Aborting.\n");
+		return;
+	}
+
+	/* TODO:  this just _assumes_ a lm70 is there ... no probe;
+	 * the lm70 driver could verify it, reading the manf ID.
+	 */
+
+	master = spi_alloc_master(p->physport->dev, sizeof *pp);
+	if (!master) {
+		status = -ENOMEM;
+		goto out_fail;
+	}
+	pp = spi_master_get_devdata(master);
+
+	/*
+	 * SPI and bitbang hookup.
+	 */
+	pp->bitbang.master = master;
+	pp->bitbang.chipselect = lm70_chipselect;
+	pp->bitbang.txrx_word[SPI_MODE_0] = lm70_txrx;
+	pp->bitbang.flags = SPI_3WIRE;
+
+	/*
+	 * Parport hookup
+	 */
+	pp->port = p;
+	memset(&lm70llp_cb, 0, sizeof(lm70llp_cb));
+	lm70llp_cb.private = pp;
+	lm70llp_cb.flags = PARPORT_FLAG_EXCL;
+	pd = parport_register_dev_model(p, DRVNAME, &lm70llp_cb, 0);
+
+	if (!pd) {
+		status = -ENOMEM;
+		goto out_free_master;
+	}
+	pp->pd = pd;
+
+	status = parport_claim(pd);
+	if (status < 0)
+		goto out_parport_unreg;
+
+	/*
+	 * Start SPI ...
+	 */
+	status = spi_bitbang_start(&pp->bitbang);
+	if (status < 0) {
+		dev_warn(&pd->dev, "spi_bitbang_start failed with status %d\n",
+			 status);
+		goto out_off_and_release;
+	}
+
+	/*
+	 * The modalias name MUST match the device_driver name
+	 * for the bus glue code to match and subsequently bind them.
+	 * We are binding to the generic drivers/hwmon/lm70.c device
+	 * driver.
+	 */
+	strcpy(pp->info.modalias, "lm70");
+	pp->info.max_speed_hz = 6 * 1000 * 1000;
+	pp->info.chip_select = 0;
+	pp->info.mode = SPI_3WIRE | SPI_MODE_0;
+
+	/* power up the chip, and let the LM70 control SI/SO */
+	parport_write_data(pp->port, lm70_INIT);
+
+	/* Enable access to our primary data structure via
+	 * the board info's (void *)controller_data.
+	 */
+	pp->info.controller_data = pp;
+	pp->spidev_lm70 = spi_new_device(pp->bitbang.master, &pp->info);
+	if (pp->spidev_lm70)
+		dev_dbg(&pp->spidev_lm70->dev, "spidev_lm70 at %s\n",
+			dev_name(&pp->spidev_lm70->dev));
+	else {
+		dev_warn(&pd->dev, "spi_new_device failed\n");
+		status = -ENODEV;
+		goto out_bitbang_stop;
+	}
+	pp->spidev_lm70->bits_per_word = 8;
+
+	lm70llp = pp;
+	return;
+
+out_bitbang_stop:
+	spi_bitbang_stop(&pp->bitbang);
+out_off_and_release:
+	/* power down */
+	parport_write_data(pp->port, 0);
+	mdelay(10);
+	parport_release(pp->pd);
+out_parport_unreg:
+	parport_unregister_device(pd);
+out_free_master:
+	spi_master_put(master);
+out_fail:
+	pr_info("spi_lm70llp probe fail, status %d\n", status);
+}
+
+static void spi_lm70llp_detach(struct parport *p)
+{
+	struct spi_lm70llp		*pp;
+
+	if (!lm70llp || lm70llp->port != p)
+		return;
+
+	pp = lm70llp;
+	spi_bitbang_stop(&pp->bitbang);
+
+	/* power down */
+	parport_write_data(pp->port, 0);
+
+	parport_release(pp->pd);
+	parport_unregister_device(pp->pd);
+
+	spi_master_put(pp->bitbang.master);
+
+	lm70llp = NULL;
+}
+
+static struct parport_driver spi_lm70llp_drv = {
+	.name =		DRVNAME,
+	.match_port =	spi_lm70llp_attach,
+	.detach =	spi_lm70llp_detach,
+	.devmodel =	true,
+};
+
+static int __init init_spi_lm70llp(void)
+{
+	return parport_register_driver(&spi_lm70llp_drv);
+}
+module_init(init_spi_lm70llp);
+
+static void __exit cleanup_spi_lm70llp(void)
+{
+	parport_unregister_driver(&spi_lm70llp_drv);
+}
+module_exit(cleanup_spi_lm70llp);
+
+MODULE_AUTHOR("Kaiwan N Billimoria <kaiwan@designergraphix.com>");
+MODULE_DESCRIPTION(
+	"Parport adapter for the National Semiconductor LM70 LLP eval board");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-loopback-test.c b/drivers/spi/spi-loopback-test.c
new file mode 100644
index 0000000..bed7403
--- /dev/null
+++ b/drivers/spi/spi-loopback-test.c
@@ -0,0 +1,1086 @@
+/*
+ *  linux/drivers/spi/spi-loopback-test.c
+ *
+ *  (c) Martin Sperl <kernel@martin.sperl.org>
+ *
+ *  Loopback test driver to test several typical spi_message conditions
+ *  that a spi_master driver may encounter
+ *  this can also get used for regression testing
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ */
+
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/ktime.h>
+#include <linux/list.h>
+#include <linux/list_sort.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/printk.h>
+#include <linux/vmalloc.h>
+#include <linux/spi/spi.h>
+
+#include "spi-test.h"
+
+/* flag to only simulate transfers */
+static int simulate_only;
+module_param(simulate_only, int, 0);
+MODULE_PARM_DESC(simulate_only, "if not 0 do not execute the spi message");
+
+/* dump spi messages */
+static int dump_messages;
+module_param(dump_messages, int, 0);
+MODULE_PARM_DESC(dump_messages,
+		 "=1 dump the basic spi_message_structure, " \
+		 "=2 dump the spi_message_structure including data, " \
+		 "=3 dump the spi_message structure before and after execution");
+/* the device is jumpered for loopback - enabling some rx_buf tests */
+static int loopback;
+module_param(loopback, int, 0);
+MODULE_PARM_DESC(loopback,
+		 "if set enable loopback mode, where the rx_buf "	\
+		 "is checked to match tx_buf after the spi_message "	\
+		 "is executed");
+
+static int loop_req;
+module_param(loop_req, int, 0);
+MODULE_PARM_DESC(loop_req,
+		 "if set controller will be asked to enable test loop mode. " \
+		 "If controller supported it, MISO and MOSI will be connected");
+
+static int no_cs;
+module_param(no_cs, int, 0);
+MODULE_PARM_DESC(no_cs,
+		 "if set Chip Select (CS) will not be used");
+
+/* run only a specific test */
+static int run_only_test = -1;
+module_param(run_only_test, int, 0);
+MODULE_PARM_DESC(run_only_test,
+		 "only run the test with this number (0-based !)");
+
+/* use vmalloc'ed buffers */
+static int use_vmalloc;
+module_param(use_vmalloc, int, 0644);
+MODULE_PARM_DESC(use_vmalloc,
+		 "use vmalloc'ed buffers instead of kmalloc'ed");
+
+/* check rx ranges */
+static int check_ranges = 1;
+module_param(check_ranges, int, 0644);
+MODULE_PARM_DESC(check_ranges,
+		 "checks rx_buffer pattern are valid");
+
+/* the actual tests to execute */
+static struct spi_test spi_tests[] = {
+	{
+		.description	= "tx/rx-transfer - start of page",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_rx_align = ITERATE_ALIGN,
+		.transfer_count = 1,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "tx/rx-transfer - crossing PAGE_SIZE",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_rx_align = ITERATE_ALIGN,
+		.transfer_count = 1,
+		.transfers		= {
+			{
+				.tx_buf = TX(PAGE_SIZE - 4),
+				.rx_buf = RX(PAGE_SIZE - 4),
+			},
+		},
+	},
+	{
+		.description	= "tx-transfer - only",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.transfer_count = 1,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+			},
+		},
+	},
+	{
+		.description	= "rx-transfer - only",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_rx_align = ITERATE_ALIGN,
+		.transfer_count = 1,
+		.transfers		= {
+			{
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx-transfers - alter both",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0) | BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+			},
+			{
+				/* this is why we cant use ITERATE_MAX_LEN */
+				.tx_buf = TX(SPI_TEST_MAX_SIZE_HALF),
+			},
+		},
+	},
+	{
+		.description	= "two tx-transfers - alter first",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(64),
+			},
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx-transfers - alter second",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.len = 16,
+				.tx_buf = TX(0),
+			},
+			{
+				.tx_buf = TX(64),
+			},
+		},
+	},
+	{
+		.description	= "two transfers tx then rx - alter both",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0) | BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+			},
+			{
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two transfers tx then rx - alter tx",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+			},
+			{
+				.len = 1,
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two transfers tx then rx - alter rx",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+			},
+			{
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx+rx transfers - alter both",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0) | BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+			{
+				/* making sure we align without overwrite
+				 * the reason we can not use ITERATE_MAX_LEN
+				 */
+				.tx_buf = TX(SPI_TEST_MAX_SIZE_HALF),
+				.rx_buf = RX(SPI_TEST_MAX_SIZE_HALF),
+			},
+		},
+	},
+	{
+		.description	= "two tx+rx transfers - alter first",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				/* making sure we align without overwrite */
+				.tx_buf = TX(1024),
+				.rx_buf = RX(1024),
+			},
+			{
+				.len = 1,
+				/* making sure we align without overwrite */
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx+rx transfers - alter second",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+			{
+				/* making sure we align without overwrite */
+				.tx_buf = TX(1024),
+				.rx_buf = RX(1024),
+			},
+		},
+	},
+	{
+		.description	= "two tx+rx transfers - delay after transfer",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_transfer_mask = BIT(0) | BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+				.delay_usecs = 1000,
+			},
+			{
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+				.delay_usecs = 1000,
+			},
+		},
+	},
+
+	{ /* end of tests sequence */ }
+};
+
+static int spi_loopback_test_probe(struct spi_device *spi)
+{
+	int ret;
+
+	if (loop_req || no_cs) {
+		spi->mode |= loop_req ? SPI_LOOP : 0;
+		spi->mode |= no_cs ? SPI_NO_CS : 0;
+		ret = spi_setup(spi);
+		if (ret) {
+			dev_err(&spi->dev, "SPI setup with SPI_LOOP or SPI_NO_CS failed (%d)\n",
+				ret);
+			return ret;
+		}
+	}
+
+	dev_info(&spi->dev, "Executing spi-loopback-tests\n");
+
+	ret = spi_test_run_tests(spi, spi_tests);
+
+	dev_info(&spi->dev, "Finished spi-loopback-tests with return: %i\n",
+		 ret);
+
+	return ret;
+}
+
+/* non const match table to permit to change via a module parameter */
+static struct of_device_id spi_loopback_test_of_match[] = {
+	{ .compatible	= "linux,spi-loopback-test", },
+	{ }
+};
+
+/* allow to override the compatible string via a module_parameter */
+module_param_string(compatible, spi_loopback_test_of_match[0].compatible,
+		    sizeof(spi_loopback_test_of_match[0].compatible),
+		    0000);
+
+MODULE_DEVICE_TABLE(of, spi_loopback_test_of_match);
+
+static struct spi_driver spi_loopback_test_driver = {
+	.driver = {
+		.name = "spi-loopback-test",
+		.owner = THIS_MODULE,
+		.of_match_table = spi_loopback_test_of_match,
+	},
+	.probe = spi_loopback_test_probe,
+};
+
+module_spi_driver(spi_loopback_test_driver);
+
+MODULE_AUTHOR("Martin Sperl <kernel@martin.sperl.org>");
+MODULE_DESCRIPTION("test spi_driver to check core functionality");
+MODULE_LICENSE("GPL");
+
+/*-------------------------------------------------------------------------*/
+
+/* spi_test implementation */
+
+#define RANGE_CHECK(ptr, plen, start, slen) \
+	((ptr >= start) && (ptr + plen <= start + slen))
+
+/* we allocate one page more, to allow for offsets */
+#define SPI_TEST_MAX_SIZE_PLUS (SPI_TEST_MAX_SIZE + PAGE_SIZE)
+
+static void spi_test_print_hex_dump(char *pre, const void *ptr, size_t len)
+{
+	/* limit the hex_dump */
+	if (len < 1024) {
+		print_hex_dump(KERN_INFO, pre,
+			       DUMP_PREFIX_OFFSET, 16, 1,
+			       ptr, len, 0);
+		return;
+	}
+	/* print head */
+	print_hex_dump(KERN_INFO, pre,
+		       DUMP_PREFIX_OFFSET, 16, 1,
+		       ptr, 512, 0);
+	/* print tail */
+	pr_info("%s truncated - continuing at offset %04zx\n",
+		pre, len - 512);
+	print_hex_dump(KERN_INFO, pre,
+		       DUMP_PREFIX_OFFSET, 16, 1,
+		       ptr + (len - 512), 512, 0);
+}
+
+static void spi_test_dump_message(struct spi_device *spi,
+				  struct spi_message *msg,
+				  bool dump_data)
+{
+	struct spi_transfer *xfer;
+	int i;
+	u8 b;
+
+	dev_info(&spi->dev, "  spi_msg@%pK\n", msg);
+	if (msg->status)
+		dev_info(&spi->dev, "    status:        %i\n",
+			 msg->status);
+	dev_info(&spi->dev, "    frame_length:  %i\n",
+		 msg->frame_length);
+	dev_info(&spi->dev, "    actual_length: %i\n",
+		 msg->actual_length);
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		dev_info(&spi->dev, "    spi_transfer@%pK\n", xfer);
+		dev_info(&spi->dev, "      len:    %i\n", xfer->len);
+		dev_info(&spi->dev, "      tx_buf: %pK\n", xfer->tx_buf);
+		if (dump_data && xfer->tx_buf)
+			spi_test_print_hex_dump("          TX: ",
+						xfer->tx_buf,
+						xfer->len);
+
+		dev_info(&spi->dev, "      rx_buf: %pK\n", xfer->rx_buf);
+		if (dump_data && xfer->rx_buf)
+			spi_test_print_hex_dump("          RX: ",
+						xfer->rx_buf,
+						xfer->len);
+		/* check for unwritten test pattern on rx_buf */
+		if (xfer->rx_buf) {
+			for (i = 0 ; i < xfer->len ; i++) {
+				b = ((u8 *)xfer->rx_buf)[xfer->len - 1 - i];
+				if (b != SPI_TEST_PATTERN_UNWRITTEN)
+					break;
+			}
+			if (i)
+				dev_info(&spi->dev,
+					 "      rx_buf filled with %02x starts at offset: %i\n",
+					 SPI_TEST_PATTERN_UNWRITTEN,
+					 xfer->len - i);
+		}
+	}
+}
+
+struct rx_ranges {
+	struct list_head list;
+	u8 *start;
+	u8 *end;
+};
+
+static int rx_ranges_cmp(void *priv, struct list_head *a, struct list_head *b)
+{
+	struct rx_ranges *rx_a = list_entry(a, struct rx_ranges, list);
+	struct rx_ranges *rx_b = list_entry(b, struct rx_ranges, list);
+
+	if (rx_a->start > rx_b->start)
+		return 1;
+	if (rx_a->start < rx_b->start)
+		return -1;
+	return 0;
+}
+
+static int spi_check_rx_ranges(struct spi_device *spi,
+			       struct spi_message *msg,
+			       void *rx)
+{
+	struct spi_transfer *xfer;
+	struct rx_ranges ranges[SPI_TEST_MAX_TRANSFERS], *r;
+	int i = 0;
+	LIST_HEAD(ranges_list);
+	u8 *addr;
+	int ret = 0;
+
+	/* loop over all transfers to fill in the rx_ranges */
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		/* if there is no rx, then no check is needed */
+		if (!xfer->rx_buf)
+			continue;
+		/* fill in the rx_range */
+		if (RANGE_CHECK(xfer->rx_buf, xfer->len,
+				rx, SPI_TEST_MAX_SIZE_PLUS)) {
+			ranges[i].start = xfer->rx_buf;
+			ranges[i].end = xfer->rx_buf + xfer->len;
+			list_add(&ranges[i].list, &ranges_list);
+			i++;
+		}
+	}
+
+	/* if no ranges, then we can return and avoid the checks...*/
+	if (!i)
+		return 0;
+
+	/* sort the list */
+	list_sort(NULL, &ranges_list, rx_ranges_cmp);
+
+	/* and iterate over all the rx addresses */
+	for (addr = rx; addr < (u8 *)rx + SPI_TEST_MAX_SIZE_PLUS; addr++) {
+		/* if we are the DO not write pattern,
+		 * then continue with the loop...
+		 */
+		if (*addr == SPI_TEST_PATTERN_DO_NOT_WRITE)
+			continue;
+
+		/* check if we are inside a range */
+		list_for_each_entry(r, &ranges_list, list) {
+			/* if so then set to end... */
+			if ((addr >= r->start) && (addr < r->end))
+				addr = r->end;
+		}
+		/* second test after a (hopefull) translation */
+		if (*addr == SPI_TEST_PATTERN_DO_NOT_WRITE)
+			continue;
+
+		/* if still not found then something has modified too much */
+		/* we could list the "closest" transfer here... */
+		dev_err(&spi->dev,
+			"loopback strangeness - rx changed outside of allowed range at: %pK\n",
+			addr);
+		/* do not return, only set ret,
+		 * so that we list all addresses
+		 */
+		ret = -ERANGE;
+	}
+
+	return ret;
+}
+
+static int spi_test_check_elapsed_time(struct spi_device *spi,
+				       struct spi_test *test)
+{
+	int i;
+	unsigned long long estimated_time = 0;
+	unsigned long long delay_usecs = 0;
+
+	for (i = 0; i < test->transfer_count; i++) {
+		struct spi_transfer *xfer = test->transfers + i;
+		unsigned long long nbits = (unsigned long long)BITS_PER_BYTE *
+					   xfer->len;
+
+		delay_usecs += xfer->delay_usecs;
+		if (!xfer->speed_hz)
+			continue;
+		estimated_time += div_u64(nbits * NSEC_PER_SEC, xfer->speed_hz);
+	}
+
+	estimated_time += delay_usecs * NSEC_PER_USEC;
+	if (test->elapsed_time < estimated_time) {
+		dev_err(&spi->dev,
+			"elapsed time %lld ns is shorter than minimum estimated time %lld ns\n",
+			test->elapsed_time, estimated_time);
+
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int spi_test_check_loopback_result(struct spi_device *spi,
+					  struct spi_message *msg,
+					  void *tx, void *rx)
+{
+	struct spi_transfer *xfer;
+	u8 rxb, txb;
+	size_t i;
+	int ret;
+
+	/* checks rx_buffer pattern are valid with loopback or without */
+	if (check_ranges) {
+		ret = spi_check_rx_ranges(spi, msg, rx);
+		if (ret)
+			return ret;
+	}
+
+	/* if we run without loopback, then return now */
+	if (!loopback)
+		return 0;
+
+	/* if applicable to transfer check that rx_buf is equal to tx_buf */
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		/* if there is no rx, then no check is needed */
+		if (!xfer->len || !xfer->rx_buf)
+			continue;
+		/* so depending on tx_buf we need to handle things */
+		if (xfer->tx_buf) {
+			for (i = 0; i < xfer->len; i++) {
+				txb = ((u8 *)xfer->tx_buf)[i];
+				rxb = ((u8 *)xfer->rx_buf)[i];
+				if (txb != rxb)
+					goto mismatch_error;
+			}
+		} else {
+			/* first byte received */
+			txb = ((u8 *)xfer->rx_buf)[0];
+			/* first byte may be 0 or xff */
+			if (!((txb == 0) || (txb == 0xff))) {
+				dev_err(&spi->dev,
+					"loopback strangeness - we expect 0x00 or 0xff, but not 0x%02x\n",
+					txb);
+				return -EINVAL;
+			}
+			/* check that all bytes are identical */
+			for (i = 1; i < xfer->len; i++) {
+				rxb = ((u8 *)xfer->rx_buf)[i];
+				if (rxb != txb)
+					goto mismatch_error;
+			}
+		}
+	}
+
+	return 0;
+
+mismatch_error:
+	dev_err(&spi->dev,
+		"loopback strangeness - transfer mismatch on byte %04zx - expected 0x%02x, but got 0x%02x\n",
+		i, txb, rxb);
+
+	return -EINVAL;
+}
+
+static int spi_test_translate(struct spi_device *spi,
+			      void **ptr, size_t len,
+			      void *tx, void *rx)
+{
+	size_t off;
+
+	/* return on null */
+	if (!*ptr)
+		return 0;
+
+	/* in the MAX_SIZE_HALF case modify the pointer */
+	if (((size_t)*ptr) & SPI_TEST_MAX_SIZE_HALF)
+		/* move the pointer to the correct range */
+		*ptr += (SPI_TEST_MAX_SIZE_PLUS / 2) -
+			SPI_TEST_MAX_SIZE_HALF;
+
+	/* RX range
+	 * - we check against MAX_SIZE_PLUS to allow for automated alignment
+	 */
+	if (RANGE_CHECK(*ptr, len,  RX(0), SPI_TEST_MAX_SIZE_PLUS)) {
+		off = *ptr - RX(0);
+		*ptr = rx + off;
+
+		return 0;
+	}
+
+	/* TX range */
+	if (RANGE_CHECK(*ptr, len,  TX(0), SPI_TEST_MAX_SIZE_PLUS)) {
+		off = *ptr - TX(0);
+		*ptr = tx + off;
+
+		return 0;
+	}
+
+	dev_err(&spi->dev,
+		"PointerRange [%pK:%pK[ not in range [%pK:%pK[ or [%pK:%pK[\n",
+		*ptr, *ptr + len,
+		RX(0), RX(SPI_TEST_MAX_SIZE),
+		TX(0), TX(SPI_TEST_MAX_SIZE));
+
+	return -EINVAL;
+}
+
+static int spi_test_fill_pattern(struct spi_device *spi,
+				 struct spi_test *test)
+{
+	struct spi_transfer *xfers = test->transfers;
+	u8 *tx_buf;
+	size_t count = 0;
+	int i, j;
+
+#ifdef __BIG_ENDIAN
+#define GET_VALUE_BYTE(value, index, bytes) \
+	(value >> (8 * (bytes - 1 - count % bytes)))
+#else
+#define GET_VALUE_BYTE(value, index, bytes) \
+	(value >> (8 * (count % bytes)))
+#endif
+
+	/* fill all transfers with the pattern requested */
+	for (i = 0; i < test->transfer_count; i++) {
+		/* fill rx_buf with SPI_TEST_PATTERN_UNWRITTEN */
+		if (xfers[i].rx_buf)
+			memset(xfers[i].rx_buf, SPI_TEST_PATTERN_UNWRITTEN,
+			       xfers[i].len);
+		/* if tx_buf is NULL then skip */
+		tx_buf = (u8 *)xfers[i].tx_buf;
+		if (!tx_buf)
+			continue;
+		/* modify all the transfers */
+		for (j = 0; j < xfers[i].len; j++, tx_buf++, count++) {
+			/* fill tx */
+			switch (test->fill_option) {
+			case FILL_MEMSET_8:
+				*tx_buf = test->fill_pattern;
+				break;
+			case FILL_MEMSET_16:
+				*tx_buf = GET_VALUE_BYTE(test->fill_pattern,
+							 count, 2);
+				break;
+			case FILL_MEMSET_24:
+				*tx_buf = GET_VALUE_BYTE(test->fill_pattern,
+							 count, 3);
+				break;
+			case FILL_MEMSET_32:
+				*tx_buf = GET_VALUE_BYTE(test->fill_pattern,
+							 count, 4);
+				break;
+			case FILL_COUNT_8:
+				*tx_buf = count;
+				break;
+			case FILL_COUNT_16:
+				*tx_buf = GET_VALUE_BYTE(count, count, 2);
+				break;
+			case FILL_COUNT_24:
+				*tx_buf = GET_VALUE_BYTE(count, count, 3);
+				break;
+			case FILL_COUNT_32:
+				*tx_buf = GET_VALUE_BYTE(count, count, 4);
+				break;
+			case FILL_TRANSFER_BYTE_8:
+				*tx_buf = j;
+				break;
+			case FILL_TRANSFER_BYTE_16:
+				*tx_buf = GET_VALUE_BYTE(j, j, 2);
+				break;
+			case FILL_TRANSFER_BYTE_24:
+				*tx_buf = GET_VALUE_BYTE(j, j, 3);
+				break;
+			case FILL_TRANSFER_BYTE_32:
+				*tx_buf = GET_VALUE_BYTE(j, j, 4);
+				break;
+			case FILL_TRANSFER_NUM:
+				*tx_buf = i;
+				break;
+			default:
+				dev_err(&spi->dev,
+					"unsupported fill_option: %i\n",
+					test->fill_option);
+				return -EINVAL;
+			}
+		}
+	}
+
+	return 0;
+}
+
+static int _spi_test_run_iter(struct spi_device *spi,
+			      struct spi_test *test,
+			      void *tx, void *rx)
+{
+	struct spi_message *msg = &test->msg;
+	struct spi_transfer *x;
+	int i, ret;
+
+	/* initialize message - zero-filled via static initialization */
+	spi_message_init_no_memset(msg);
+
+	/* fill rx with the DO_NOT_WRITE pattern */
+	memset(rx, SPI_TEST_PATTERN_DO_NOT_WRITE, SPI_TEST_MAX_SIZE_PLUS);
+
+	/* add the individual transfers */
+	for (i = 0; i < test->transfer_count; i++) {
+		x = &test->transfers[i];
+
+		/* patch the values of tx_buf */
+		ret = spi_test_translate(spi, (void **)&x->tx_buf, x->len,
+					 (void *)tx, rx);
+		if (ret)
+			return ret;
+
+		/* patch the values of rx_buf */
+		ret = spi_test_translate(spi, &x->rx_buf, x->len,
+					 (void *)tx, rx);
+		if (ret)
+			return ret;
+
+		/* and add it to the list */
+		spi_message_add_tail(x, msg);
+	}
+
+	/* fill in the transfer buffers with pattern */
+	ret = spi_test_fill_pattern(spi, test);
+	if (ret)
+		return ret;
+
+	/* and execute */
+	if (test->execute_msg)
+		ret = test->execute_msg(spi, test, tx, rx);
+	else
+		ret = spi_test_execute_msg(spi, test, tx, rx);
+
+	/* handle result */
+	if (ret == test->expected_return)
+		return 0;
+
+	dev_err(&spi->dev,
+		"test failed - test returned %i, but we expect %i\n",
+		ret, test->expected_return);
+
+	if (ret)
+		return ret;
+
+	/* if it is 0, as we expected something else,
+	 * then return something special
+	 */
+	return -EFAULT;
+}
+
+static int spi_test_run_iter(struct spi_device *spi,
+			     const struct spi_test *testtemplate,
+			     void *tx, void *rx,
+			     size_t len,
+			     size_t tx_off,
+			     size_t rx_off
+	)
+{
+	struct spi_test test;
+	int i, tx_count, rx_count;
+
+	/* copy the test template to test */
+	memcpy(&test, testtemplate, sizeof(test));
+
+	/* if iterate_transfer_mask is not set,
+	 * then set it to first transfer only
+	 */
+	if (!(test.iterate_transfer_mask & (BIT(test.transfer_count) - 1)))
+		test.iterate_transfer_mask = 1;
+
+	/* count number of transfers with tx/rx_buf != NULL */
+	rx_count = tx_count = 0;
+	for (i = 0; i < test.transfer_count; i++) {
+		if (test.transfers[i].tx_buf)
+			tx_count++;
+		if (test.transfers[i].rx_buf)
+			rx_count++;
+	}
+
+	/* in some iteration cases warn and exit early,
+	 * as there is nothing to do, that has not been tested already...
+	 */
+	if (tx_off && (!tx_count)) {
+		dev_warn_once(&spi->dev,
+			      "%s: iterate_tx_off configured with tx_buf==NULL - ignoring\n",
+			      test.description);
+		return 0;
+	}
+	if (rx_off && (!rx_count)) {
+		dev_warn_once(&spi->dev,
+			      "%s: iterate_rx_off configured with rx_buf==NULL - ignoring\n",
+			      test.description);
+		return 0;
+	}
+
+	/* write out info */
+	if (!(len || tx_off || rx_off)) {
+		dev_info(&spi->dev, "Running test %s\n", test.description);
+	} else {
+		dev_info(&spi->dev,
+			 "  with iteration values: len = %zu, tx_off = %zu, rx_off = %zu\n",
+			 len, tx_off, rx_off);
+	}
+
+	/* update in the values from iteration values */
+	for (i = 0; i < test.transfer_count; i++) {
+		/* only when bit in transfer mask is set */
+		if (!(test.iterate_transfer_mask & BIT(i)))
+			continue;
+		test.transfers[i].len = len;
+		if (test.transfers[i].tx_buf)
+			test.transfers[i].tx_buf += tx_off;
+		if (test.transfers[i].tx_buf)
+			test.transfers[i].rx_buf += rx_off;
+	}
+
+	/* and execute */
+	return _spi_test_run_iter(spi, &test, tx, rx);
+}
+
+/**
+ * spi_test_execute_msg - default implementation to run a test
+ *
+ * spi: @spi_device on which to run the @spi_message
+ * test: the test to execute, which already contains @msg
+ * tx:   the tx buffer allocated for the test sequence
+ * rx:   the rx buffer allocated for the test sequence
+ *
+ * Returns: error code of spi_sync as well as basic error checking
+ */
+int spi_test_execute_msg(struct spi_device *spi, struct spi_test *test,
+			 void *tx, void *rx)
+{
+	struct spi_message *msg = &test->msg;
+	int ret = 0;
+	int i;
+
+	/* only if we do not simulate */
+	if (!simulate_only) {
+		ktime_t start;
+
+		/* dump the complete message before and after the transfer */
+		if (dump_messages == 3)
+			spi_test_dump_message(spi, msg, true);
+
+		start = ktime_get();
+		/* run spi message */
+		ret = spi_sync(spi, msg);
+		test->elapsed_time = ktime_to_ns(ktime_sub(ktime_get(), start));
+		if (ret == -ETIMEDOUT) {
+			dev_info(&spi->dev,
+				 "spi-message timed out - rerunning...\n");
+			/* rerun after a few explicit schedules */
+			for (i = 0; i < 16; i++)
+				schedule();
+			ret = spi_sync(spi, msg);
+		}
+		if (ret) {
+			dev_err(&spi->dev,
+				"Failed to execute spi_message: %i\n",
+				ret);
+			goto exit;
+		}
+
+		/* do some extra error checks */
+		if (msg->frame_length != msg->actual_length) {
+			dev_err(&spi->dev,
+				"actual length differs from expected\n");
+			ret = -EIO;
+			goto exit;
+		}
+
+		/* run rx-buffer tests */
+		ret = spi_test_check_loopback_result(spi, msg, tx, rx);
+		if (ret)
+			goto exit;
+
+		ret = spi_test_check_elapsed_time(spi, test);
+	}
+
+	/* if requested or on error dump message (including data) */
+exit:
+	if (dump_messages || ret)
+		spi_test_dump_message(spi, msg,
+				      (dump_messages >= 2) || (ret));
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(spi_test_execute_msg);
+
+/**
+ * spi_test_run_test - run an individual spi_test
+ *                     including all the relevant iterations on:
+ *                     length and buffer alignment
+ *
+ * spi:  the spi_device to send the messages to
+ * test: the test which we need to execute
+ * tx:   the tx buffer allocated for the test sequence
+ * rx:   the rx buffer allocated for the test sequence
+ *
+ * Returns: status code of spi_sync or other failures
+ */
+
+int spi_test_run_test(struct spi_device *spi, const struct spi_test *test,
+		      void *tx, void *rx)
+{
+	int idx_len;
+	size_t len;
+	size_t tx_align, rx_align;
+	int ret;
+
+	/* test for transfer limits */
+	if (test->transfer_count >= SPI_TEST_MAX_TRANSFERS) {
+		dev_err(&spi->dev,
+			"%s: Exceeded max number of transfers with %i\n",
+			test->description, test->transfer_count);
+		return -E2BIG;
+	}
+
+	/* setting up some values in spi_message
+	 * based on some settings in spi_master
+	 * some of this can also get done in the run() method
+	 */
+
+	/* iterate over all the iterable values using macros
+	 * (to make it a bit more readable...
+	 */
+#define FOR_EACH_ALIGNMENT(var)						\
+	for (var = 0;							\
+	    var < (test->iterate_##var ?				\
+			(spi->master->dma_alignment ?			\
+			 spi->master->dma_alignment :			\
+			 test->iterate_##var) :				\
+			1);						\
+	    var++)
+
+	for (idx_len = 0; idx_len < SPI_TEST_MAX_ITERATE &&
+	     (len = test->iterate_len[idx_len]) != -1; idx_len++) {
+		FOR_EACH_ALIGNMENT(tx_align) {
+			FOR_EACH_ALIGNMENT(rx_align) {
+				/* and run the iteration */
+				ret = spi_test_run_iter(spi, test,
+							tx, rx,
+							len,
+							tx_align,
+							rx_align);
+				if (ret)
+					return ret;
+			}
+		}
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(spi_test_run_test);
+
+/**
+ * spi_test_run_tests - run an array of spi_messages tests
+ * @spi: the spi device on which to run the tests
+ * @tests: NULL-terminated array of @spi_test
+ *
+ * Returns: status errors as per @spi_test_run_test()
+ */
+
+int spi_test_run_tests(struct spi_device *spi,
+		       struct spi_test *tests)
+{
+	char *rx = NULL, *tx = NULL;
+	int ret = 0, count = 0;
+	struct spi_test *test;
+
+	/* allocate rx/tx buffers of 128kB size without devm
+	 * in the hope that is on a page boundary
+	 */
+	if (use_vmalloc)
+		rx = vmalloc(SPI_TEST_MAX_SIZE_PLUS);
+	else
+		rx = kzalloc(SPI_TEST_MAX_SIZE_PLUS, GFP_KERNEL);
+	if (!rx)
+		return -ENOMEM;
+
+
+	if (use_vmalloc)
+		tx = vmalloc(SPI_TEST_MAX_SIZE_PLUS);
+	else
+		tx = kzalloc(SPI_TEST_MAX_SIZE_PLUS, GFP_KERNEL);
+	if (!tx) {
+		ret = -ENOMEM;
+		goto err_tx;
+	}
+
+	/* now run the individual tests in the table */
+	for (test = tests, count = 0; test->description[0];
+	     test++, count++) {
+		/* only run test if requested */
+		if ((run_only_test > -1) && (count != run_only_test))
+			continue;
+		/* run custom implementation */
+		if (test->run_test)
+			ret = test->run_test(spi, test, tx, rx);
+		else
+			ret = spi_test_run_test(spi, test, tx, rx);
+		if (ret)
+			goto out;
+		/* add some delays so that we can easily
+		 * detect the individual tests when using a logic analyzer
+		 * we also add scheduling to avoid potential spi_timeouts...
+		 */
+		mdelay(100);
+		schedule();
+	}
+
+out:
+	kvfree(tx);
+err_tx:
+	kvfree(rx);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(spi_test_run_tests);
diff --git a/drivers/spi/spi-lp8841-rtc.c b/drivers/spi/spi-lp8841-rtc.c
new file mode 100644
index 0000000..faa577d
--- /dev/null
+++ b/drivers/spi/spi-lp8841-rtc.c
@@ -0,0 +1,256 @@
+/*
+ * SPI master driver for ICP DAS LP-8841 RTC
+ *
+ * Copyright (C) 2016 Sergei Ianovich
+ *
+ * based on
+ *
+ * Dallas DS1302 RTC Support
+ * Copyright (C) 2002 David McCullough
+ * Copyright (C) 2003 - 2007 Paul Mundt
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/spi/spi.h>
+
+#define DRIVER_NAME	"spi_lp8841_rtc"
+
+#define SPI_LP8841_RTC_CE	0x01
+#define SPI_LP8841_RTC_CLK	0x02
+#define SPI_LP8841_RTC_nWE	0x04
+#define SPI_LP8841_RTC_MOSI	0x08
+#define SPI_LP8841_RTC_MISO	0x01
+
+/*
+ * REVISIT If there is support for SPI_3WIRE and SPI_LSB_FIRST in SPI
+ * GPIO driver, this SPI driver can be replaced by a simple GPIO driver
+ * providing 3 GPIO pins.
+ */
+
+struct spi_lp8841_rtc {
+	void		*iomem;
+	unsigned long	state;
+};
+
+static inline void
+setsck(struct spi_lp8841_rtc *data, int is_on)
+{
+	if (is_on)
+		data->state |= SPI_LP8841_RTC_CLK;
+	else
+		data->state &= ~SPI_LP8841_RTC_CLK;
+	writeb(data->state, data->iomem);
+}
+
+static inline void
+setmosi(struct spi_lp8841_rtc *data, int is_on)
+{
+	if (is_on)
+		data->state |= SPI_LP8841_RTC_MOSI;
+	else
+		data->state &= ~SPI_LP8841_RTC_MOSI;
+	writeb(data->state, data->iomem);
+}
+
+static inline int
+getmiso(struct spi_lp8841_rtc *data)
+{
+	return ioread8(data->iomem) & SPI_LP8841_RTC_MISO;
+}
+
+static inline u32
+bitbang_txrx_be_cpha0_lsb(struct spi_lp8841_rtc *data,
+		unsigned usecs, unsigned cpol, unsigned flags,
+		u32 word, u8 bits)
+{
+	/* if (cpol == 0) this is SPI_MODE_0; else this is SPI_MODE_2 */
+
+	u32 shift = 32 - bits;
+	/* clock starts at inactive polarity */
+	for (; likely(bits); bits--) {
+
+		/* setup LSB (to slave) on leading edge */
+		if ((flags & SPI_MASTER_NO_TX) == 0)
+			setmosi(data, (word & 1));
+
+		usleep_range(usecs, usecs + 1);	/* T(setup) */
+
+		/* sample LSB (from slave) on trailing edge */
+		word >>= 1;
+		if ((flags & SPI_MASTER_NO_RX) == 0)
+			word |= (getmiso(data) << 31);
+
+		setsck(data, !cpol);
+		usleep_range(usecs, usecs + 1);
+
+		setsck(data, cpol);
+	}
+
+	word >>= shift;
+	return word;
+}
+
+static int
+spi_lp8841_rtc_transfer_one(struct spi_master *master,
+			    struct spi_device *spi,
+			    struct spi_transfer *t)
+{
+	struct spi_lp8841_rtc	*data = spi_master_get_devdata(master);
+	unsigned		count = t->len;
+	const u8		*tx = t->tx_buf;
+	u8			*rx = t->rx_buf;
+	u8			word = 0;
+	int			ret = 0;
+
+	if (tx) {
+		data->state &= ~SPI_LP8841_RTC_nWE;
+		writeb(data->state, data->iomem);
+		while (likely(count > 0)) {
+			word = *tx++;
+			bitbang_txrx_be_cpha0_lsb(data, 1, 0,
+					SPI_MASTER_NO_RX, word, 8);
+			count--;
+		}
+	} else if (rx) {
+		data->state |= SPI_LP8841_RTC_nWE;
+		writeb(data->state, data->iomem);
+		while (likely(count > 0)) {
+			word = bitbang_txrx_be_cpha0_lsb(data, 1, 0,
+					SPI_MASTER_NO_TX, word, 8);
+			*rx++ = word;
+			count--;
+		}
+	} else {
+		ret = -EINVAL;
+	}
+
+	spi_finalize_current_transfer(master);
+
+	return ret;
+}
+
+static void
+spi_lp8841_rtc_set_cs(struct spi_device *spi, bool enable)
+{
+	struct spi_lp8841_rtc *data = spi_master_get_devdata(spi->master);
+
+	data->state = 0;
+	writeb(data->state, data->iomem);
+	if (enable) {
+		usleep_range(4, 5);
+		data->state |= SPI_LP8841_RTC_CE;
+		writeb(data->state, data->iomem);
+		usleep_range(4, 5);
+	}
+}
+
+static int
+spi_lp8841_rtc_setup(struct spi_device *spi)
+{
+	if ((spi->mode & SPI_CS_HIGH) == 0) {
+		dev_err(&spi->dev, "unsupported active low chip select\n");
+		return -EINVAL;
+	}
+
+	if ((spi->mode & SPI_LSB_FIRST) == 0) {
+		dev_err(&spi->dev, "unsupported MSB first mode\n");
+		return -EINVAL;
+	}
+
+	if ((spi->mode & SPI_3WIRE) == 0) {
+		dev_err(&spi->dev, "unsupported wiring. 3 wires required\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+#ifdef CONFIG_OF
+static const struct of_device_id spi_lp8841_rtc_dt_ids[] = {
+	{ .compatible = "icpdas,lp8841-spi-rtc" },
+	{ }
+};
+
+MODULE_DEVICE_TABLE(of, spi_lp8841_rtc_dt_ids);
+#endif
+
+static int
+spi_lp8841_rtc_probe(struct platform_device *pdev)
+{
+	int				ret;
+	struct spi_master		*master;
+	struct spi_lp8841_rtc		*data;
+	void				*iomem;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*data));
+	if (!master)
+		return -ENOMEM;
+	platform_set_drvdata(pdev, master);
+
+	master->flags = SPI_MASTER_HALF_DUPLEX;
+	master->mode_bits = SPI_CS_HIGH | SPI_3WIRE | SPI_LSB_FIRST;
+
+	master->bus_num = pdev->id;
+	master->num_chipselect = 1;
+	master->setup = spi_lp8841_rtc_setup;
+	master->set_cs = spi_lp8841_rtc_set_cs;
+	master->transfer_one = spi_lp8841_rtc_transfer_one;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+#ifdef CONFIG_OF
+	master->dev.of_node = pdev->dev.of_node;
+#endif
+
+	data = spi_master_get_devdata(master);
+
+	iomem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	data->iomem = devm_ioremap_resource(&pdev->dev, iomem);
+	ret = PTR_ERR_OR_ZERO(data->iomem);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to get IO address\n");
+		goto err_put_master;
+	}
+
+	/* register with the SPI framework */
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret) {
+		dev_err(&pdev->dev, "cannot register spi master\n");
+		goto err_put_master;
+	}
+
+	return ret;
+
+
+err_put_master:
+	spi_master_put(master);
+
+	return ret;
+}
+
+MODULE_ALIAS("platform:" DRIVER_NAME);
+
+static struct platform_driver spi_lp8841_rtc_driver = {
+	.driver = {
+		.name	= DRIVER_NAME,
+		.of_match_table = of_match_ptr(spi_lp8841_rtc_dt_ids),
+	},
+	.probe		= spi_lp8841_rtc_probe,
+};
+module_platform_driver(spi_lp8841_rtc_driver);
+
+MODULE_DESCRIPTION("SPI master driver for ICP DAS LP-8841 RTC");
+MODULE_AUTHOR("Sergei Ianovich");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-mem.c b/drivers/spi/spi-mem.c
new file mode 100644
index 0000000..eb72dba
--- /dev/null
+++ b/drivers/spi/spi-mem.c
@@ -0,0 +1,453 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * Copyright (C) 2018 Exceet Electronics GmbH
+ * Copyright (C) 2018 Bootlin
+ *
+ * Author: Boris Brezillon <boris.brezillon@bootlin.com>
+ */
+#include <linux/dmaengine.h>
+#include <linux/pm_runtime.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi-mem.h>
+
+#include "internals.h"
+
+/**
+ * spi_controller_dma_map_mem_op_data() - DMA-map the buffer attached to a
+ *					  memory operation
+ * @ctlr: the SPI controller requesting this dma_map()
+ * @op: the memory operation containing the buffer to map
+ * @sgt: a pointer to a non-initialized sg_table that will be filled by this
+ *	 function
+ *
+ * Some controllers might want to do DMA on the data buffer embedded in @op.
+ * This helper prepares everything for you and provides a ready-to-use
+ * sg_table. This function is not intended to be called from spi drivers.
+ * Only SPI controller drivers should use it.
+ * Note that the caller must ensure the memory region pointed by
+ * op->data.buf.{in,out} is DMA-able before calling this function.
+ *
+ * Return: 0 in case of success, a negative error code otherwise.
+ */
+int spi_controller_dma_map_mem_op_data(struct spi_controller *ctlr,
+				       const struct spi_mem_op *op,
+				       struct sg_table *sgt)
+{
+	struct device *dmadev;
+
+	if (!op->data.nbytes)
+		return -EINVAL;
+
+	if (op->data.dir == SPI_MEM_DATA_OUT && ctlr->dma_tx)
+		dmadev = ctlr->dma_tx->device->dev;
+	else if (op->data.dir == SPI_MEM_DATA_IN && ctlr->dma_rx)
+		dmadev = ctlr->dma_rx->device->dev;
+	else
+		dmadev = ctlr->dev.parent;
+
+	if (!dmadev)
+		return -EINVAL;
+
+	return spi_map_buf(ctlr, dmadev, sgt, op->data.buf.in, op->data.nbytes,
+			   op->data.dir == SPI_MEM_DATA_IN ?
+			   DMA_FROM_DEVICE : DMA_TO_DEVICE);
+}
+EXPORT_SYMBOL_GPL(spi_controller_dma_map_mem_op_data);
+
+/**
+ * spi_controller_dma_unmap_mem_op_data() - DMA-unmap the buffer attached to a
+ *					    memory operation
+ * @ctlr: the SPI controller requesting this dma_unmap()
+ * @op: the memory operation containing the buffer to unmap
+ * @sgt: a pointer to an sg_table previously initialized by
+ *	 spi_controller_dma_map_mem_op_data()
+ *
+ * Some controllers might want to do DMA on the data buffer embedded in @op.
+ * This helper prepares things so that the CPU can access the
+ * op->data.buf.{in,out} buffer again.
+ *
+ * This function is not intended to be called from SPI drivers. Only SPI
+ * controller drivers should use it.
+ *
+ * This function should be called after the DMA operation has finished and is
+ * only valid if the previous spi_controller_dma_map_mem_op_data() call
+ * returned 0.
+ *
+ * Return: 0 in case of success, a negative error code otherwise.
+ */
+void spi_controller_dma_unmap_mem_op_data(struct spi_controller *ctlr,
+					  const struct spi_mem_op *op,
+					  struct sg_table *sgt)
+{
+	struct device *dmadev;
+
+	if (!op->data.nbytes)
+		return;
+
+	if (op->data.dir == SPI_MEM_DATA_OUT && ctlr->dma_tx)
+		dmadev = ctlr->dma_tx->device->dev;
+	else if (op->data.dir == SPI_MEM_DATA_IN && ctlr->dma_rx)
+		dmadev = ctlr->dma_rx->device->dev;
+	else
+		dmadev = ctlr->dev.parent;
+
+	spi_unmap_buf(ctlr, dmadev, sgt,
+		      op->data.dir == SPI_MEM_DATA_IN ?
+		      DMA_FROM_DEVICE : DMA_TO_DEVICE);
+}
+EXPORT_SYMBOL_GPL(spi_controller_dma_unmap_mem_op_data);
+
+static int spi_check_buswidth_req(struct spi_mem *mem, u8 buswidth, bool tx)
+{
+	u32 mode = mem->spi->mode;
+
+	switch (buswidth) {
+	case 1:
+		return 0;
+
+	case 2:
+		if ((tx && (mode & (SPI_TX_DUAL | SPI_TX_QUAD))) ||
+		    (!tx && (mode & (SPI_RX_DUAL | SPI_RX_QUAD))))
+			return 0;
+
+		break;
+
+	case 4:
+		if ((tx && (mode & SPI_TX_QUAD)) ||
+		    (!tx && (mode & SPI_RX_QUAD)))
+			return 0;
+
+		break;
+
+	default:
+		break;
+	}
+
+	return -ENOTSUPP;
+}
+
+static bool spi_mem_default_supports_op(struct spi_mem *mem,
+					const struct spi_mem_op *op)
+{
+	if (spi_check_buswidth_req(mem, op->cmd.buswidth, true))
+		return false;
+
+	if (op->addr.nbytes &&
+	    spi_check_buswidth_req(mem, op->addr.buswidth, true))
+		return false;
+
+	if (op->dummy.nbytes &&
+	    spi_check_buswidth_req(mem, op->dummy.buswidth, true))
+		return false;
+
+	if (op->data.nbytes &&
+	    spi_check_buswidth_req(mem, op->data.buswidth,
+				   op->data.dir == SPI_MEM_DATA_OUT))
+		return false;
+
+	return true;
+}
+EXPORT_SYMBOL_GPL(spi_mem_default_supports_op);
+
+/**
+ * spi_mem_supports_op() - Check if a memory device and the controller it is
+ *			   connected to support a specific memory operation
+ * @mem: the SPI memory
+ * @op: the memory operation to check
+ *
+ * Some controllers are only supporting Single or Dual IOs, others might only
+ * support specific opcodes, or it can even be that the controller and device
+ * both support Quad IOs but the hardware prevents you from using it because
+ * only 2 IO lines are connected.
+ *
+ * This function checks whether a specific operation is supported.
+ *
+ * Return: true if @op is supported, false otherwise.
+ */
+bool spi_mem_supports_op(struct spi_mem *mem, const struct spi_mem_op *op)
+{
+	struct spi_controller *ctlr = mem->spi->controller;
+
+	if (ctlr->mem_ops && ctlr->mem_ops->supports_op)
+		return ctlr->mem_ops->supports_op(mem, op);
+
+	return spi_mem_default_supports_op(mem, op);
+}
+EXPORT_SYMBOL_GPL(spi_mem_supports_op);
+
+/**
+ * spi_mem_exec_op() - Execute a memory operation
+ * @mem: the SPI memory
+ * @op: the memory operation to execute
+ *
+ * Executes a memory operation.
+ *
+ * This function first checks that @op is supported and then tries to execute
+ * it.
+ *
+ * Return: 0 in case of success, a negative error code otherwise.
+ */
+int spi_mem_exec_op(struct spi_mem *mem, const struct spi_mem_op *op)
+{
+	unsigned int tmpbufsize, xferpos = 0, totalxferlen = 0;
+	struct spi_controller *ctlr = mem->spi->controller;
+	struct spi_transfer xfers[4] = { };
+	struct spi_message msg;
+	u8 *tmpbuf;
+	int ret;
+
+	if (!spi_mem_supports_op(mem, op))
+		return -ENOTSUPP;
+
+	if (ctlr->mem_ops) {
+		/*
+		 * Flush the message queue before executing our SPI memory
+		 * operation to prevent preemption of regular SPI transfers.
+		 */
+		spi_flush_queue(ctlr);
+
+		if (ctlr->auto_runtime_pm) {
+			ret = pm_runtime_get_sync(ctlr->dev.parent);
+			if (ret < 0) {
+				dev_err(&ctlr->dev,
+					"Failed to power device: %d\n",
+					ret);
+				return ret;
+			}
+		}
+
+		mutex_lock(&ctlr->bus_lock_mutex);
+		mutex_lock(&ctlr->io_mutex);
+		ret = ctlr->mem_ops->exec_op(mem, op);
+		mutex_unlock(&ctlr->io_mutex);
+		mutex_unlock(&ctlr->bus_lock_mutex);
+
+		if (ctlr->auto_runtime_pm)
+			pm_runtime_put(ctlr->dev.parent);
+
+		/*
+		 * Some controllers only optimize specific paths (typically the
+		 * read path) and expect the core to use the regular SPI
+		 * interface in other cases.
+		 */
+		if (!ret || ret != -ENOTSUPP)
+			return ret;
+	}
+
+	tmpbufsize = sizeof(op->cmd.opcode) + op->addr.nbytes +
+		     op->dummy.nbytes;
+
+	/*
+	 * Allocate a buffer to transmit the CMD, ADDR cycles with kmalloc() so
+	 * we're guaranteed that this buffer is DMA-able, as required by the
+	 * SPI layer.
+	 */
+	tmpbuf = kzalloc(tmpbufsize, GFP_KERNEL | GFP_DMA);
+	if (!tmpbuf)
+		return -ENOMEM;
+
+	spi_message_init(&msg);
+
+	tmpbuf[0] = op->cmd.opcode;
+	xfers[xferpos].tx_buf = tmpbuf;
+	xfers[xferpos].len = sizeof(op->cmd.opcode);
+	xfers[xferpos].tx_nbits = op->cmd.buswidth;
+	spi_message_add_tail(&xfers[xferpos], &msg);
+	xferpos++;
+	totalxferlen++;
+
+	if (op->addr.nbytes) {
+		int i;
+
+		for (i = 0; i < op->addr.nbytes; i++)
+			tmpbuf[i + 1] = op->addr.val >>
+					(8 * (op->addr.nbytes - i - 1));
+
+		xfers[xferpos].tx_buf = tmpbuf + 1;
+		xfers[xferpos].len = op->addr.nbytes;
+		xfers[xferpos].tx_nbits = op->addr.buswidth;
+		spi_message_add_tail(&xfers[xferpos], &msg);
+		xferpos++;
+		totalxferlen += op->addr.nbytes;
+	}
+
+	if (op->dummy.nbytes) {
+		memset(tmpbuf + op->addr.nbytes + 1, 0xff, op->dummy.nbytes);
+		xfers[xferpos].tx_buf = tmpbuf + op->addr.nbytes + 1;
+		xfers[xferpos].len = op->dummy.nbytes;
+		xfers[xferpos].tx_nbits = op->dummy.buswidth;
+		spi_message_add_tail(&xfers[xferpos], &msg);
+		xferpos++;
+		totalxferlen += op->dummy.nbytes;
+	}
+
+	if (op->data.nbytes) {
+		if (op->data.dir == SPI_MEM_DATA_IN) {
+			xfers[xferpos].rx_buf = op->data.buf.in;
+			xfers[xferpos].rx_nbits = op->data.buswidth;
+		} else {
+			xfers[xferpos].tx_buf = op->data.buf.out;
+			xfers[xferpos].tx_nbits = op->data.buswidth;
+		}
+
+		xfers[xferpos].len = op->data.nbytes;
+		spi_message_add_tail(&xfers[xferpos], &msg);
+		xferpos++;
+		totalxferlen += op->data.nbytes;
+	}
+
+	ret = spi_sync(mem->spi, &msg);
+
+	kfree(tmpbuf);
+
+	if (ret)
+		return ret;
+
+	if (msg.actual_length != totalxferlen)
+		return -EIO;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(spi_mem_exec_op);
+
+/**
+ * spi_mem_get_name() - Return the SPI mem device name to be used by the
+ *			upper layer if necessary
+ * @mem: the SPI memory
+ *
+ * This function allows SPI mem users to retrieve the SPI mem device name.
+ * It is useful if the upper layer needs to expose a custom name for
+ * compatibility reasons.
+ *
+ * Return: a string containing the name of the memory device to be used
+ *	   by the SPI mem user
+ */
+const char *spi_mem_get_name(struct spi_mem *mem)
+{
+	return mem->name;
+}
+EXPORT_SYMBOL_GPL(spi_mem_get_name);
+
+/**
+ * spi_mem_adjust_op_size() - Adjust the data size of a SPI mem operation to
+ *			      match controller limitations
+ * @mem: the SPI memory
+ * @op: the operation to adjust
+ *
+ * Some controllers have FIFO limitations and must split a data transfer
+ * operation into multiple ones, others require a specific alignment for
+ * optimized accesses. This function allows SPI mem drivers to split a single
+ * operation into multiple sub-operations when required.
+ *
+ * Return: a negative error code if the controller can't properly adjust @op,
+ *	   0 otherwise. Note that @op->data.nbytes will be updated if @op
+ *	   can't be handled in a single step.
+ */
+int spi_mem_adjust_op_size(struct spi_mem *mem, struct spi_mem_op *op)
+{
+	struct spi_controller *ctlr = mem->spi->controller;
+	size_t len;
+
+	len = sizeof(op->cmd.opcode) + op->addr.nbytes + op->dummy.nbytes;
+
+	if (ctlr->mem_ops && ctlr->mem_ops->adjust_op_size)
+		return ctlr->mem_ops->adjust_op_size(mem, op);
+
+	if (!ctlr->mem_ops || !ctlr->mem_ops->exec_op) {
+		if (len > spi_max_transfer_size(mem->spi))
+			return -EINVAL;
+
+		op->data.nbytes = min3((size_t)op->data.nbytes,
+				       spi_max_transfer_size(mem->spi),
+				       spi_max_message_size(mem->spi) -
+				       len);
+		if (!op->data.nbytes)
+			return -EINVAL;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(spi_mem_adjust_op_size);
+
+static inline struct spi_mem_driver *to_spi_mem_drv(struct device_driver *drv)
+{
+	return container_of(drv, struct spi_mem_driver, spidrv.driver);
+}
+
+static int spi_mem_probe(struct spi_device *spi)
+{
+	struct spi_mem_driver *memdrv = to_spi_mem_drv(spi->dev.driver);
+	struct spi_controller *ctlr = spi->controller;
+	struct spi_mem *mem;
+
+	mem = devm_kzalloc(&spi->dev, sizeof(*mem), GFP_KERNEL);
+	if (!mem)
+		return -ENOMEM;
+
+	mem->spi = spi;
+
+	if (ctlr->mem_ops && ctlr->mem_ops->get_name)
+		mem->name = ctlr->mem_ops->get_name(mem);
+	else
+		mem->name = dev_name(&spi->dev);
+
+	if (IS_ERR_OR_NULL(mem->name))
+		return PTR_ERR(mem->name);
+
+	spi_set_drvdata(spi, mem);
+
+	return memdrv->probe(mem);
+}
+
+static int spi_mem_remove(struct spi_device *spi)
+{
+	struct spi_mem_driver *memdrv = to_spi_mem_drv(spi->dev.driver);
+	struct spi_mem *mem = spi_get_drvdata(spi);
+
+	if (memdrv->remove)
+		return memdrv->remove(mem);
+
+	return 0;
+}
+
+static void spi_mem_shutdown(struct spi_device *spi)
+{
+	struct spi_mem_driver *memdrv = to_spi_mem_drv(spi->dev.driver);
+	struct spi_mem *mem = spi_get_drvdata(spi);
+
+	if (memdrv->shutdown)
+		memdrv->shutdown(mem);
+}
+
+/**
+ * spi_mem_driver_register_with_owner() - Register a SPI memory driver
+ * @memdrv: the SPI memory driver to register
+ * @owner: the owner of this driver
+ *
+ * Registers a SPI memory driver.
+ *
+ * Return: 0 in case of success, a negative error core otherwise.
+ */
+
+int spi_mem_driver_register_with_owner(struct spi_mem_driver *memdrv,
+				       struct module *owner)
+{
+	memdrv->spidrv.probe = spi_mem_probe;
+	memdrv->spidrv.remove = spi_mem_remove;
+	memdrv->spidrv.shutdown = spi_mem_shutdown;
+
+	return __spi_register_driver(owner, &memdrv->spidrv);
+}
+EXPORT_SYMBOL_GPL(spi_mem_driver_register_with_owner);
+
+/**
+ * spi_mem_driver_unregister_with_owner() - Unregister a SPI memory driver
+ * @memdrv: the SPI memory driver to unregister
+ *
+ * Unregisters a SPI memory driver.
+ */
+void spi_mem_driver_unregister(struct spi_mem_driver *memdrv)
+{
+	spi_unregister_driver(&memdrv->spidrv);
+}
+EXPORT_SYMBOL_GPL(spi_mem_driver_unregister);
diff --git a/drivers/spi/spi-meson-spicc.c b/drivers/spi/spi-meson-spicc.c
new file mode 100644
index 0000000..7fe4488
--- /dev/null
+++ b/drivers/spi/spi-meson-spicc.c
@@ -0,0 +1,625 @@
+/*
+ * Driver for Amlogic Meson SPI communication controller (SPICC)
+ *
+ * Copyright (C) BayLibre, SAS
+ * Author: Neil Armstrong <narmstrong@baylibre.com>
+ *
+ * SPDX-License-Identifier: GPL-2.0+
+ */
+
+#include <linux/bitfield.h>
+#include <linux/clk.h>
+#include <linux/device.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/types.h>
+#include <linux/interrupt.h>
+#include <linux/reset.h>
+#include <linux/gpio.h>
+
+/*
+ * The Meson SPICC controller could support DMA based transfers, but is not
+ * implemented by the vendor code, and while having the registers documentation
+ * it has never worked on the GXL Hardware.
+ * The PIO mode is the only mode implemented, and due to badly designed HW :
+ * - all transfers are cutted in 16 words burst because the FIFO hangs on
+ *   TX underflow, and there is no TX "Half-Empty" interrupt, so we go by
+ *   FIFO max size chunk only
+ * - CS management is dumb, and goes UP between every burst, so is really a
+ *   "Data Valid" signal than a Chip Select, GPIO link should be used instead
+ *   to have a CS go down over the full transfer
+ */
+
+#define SPICC_MAX_FREQ	30000000
+#define SPICC_MAX_BURST	128
+
+/* Register Map */
+#define SPICC_RXDATA	0x00
+
+#define SPICC_TXDATA	0x04
+
+#define SPICC_CONREG	0x08
+#define SPICC_ENABLE		BIT(0)
+#define SPICC_MODE_MASTER	BIT(1)
+#define SPICC_XCH		BIT(2)
+#define SPICC_SMC		BIT(3)
+#define SPICC_POL		BIT(4)
+#define SPICC_PHA		BIT(5)
+#define SPICC_SSCTL		BIT(6)
+#define SPICC_SSPOL		BIT(7)
+#define SPICC_DRCTL_MASK	GENMASK(9, 8)
+#define SPICC_DRCTL_IGNORE	0
+#define SPICC_DRCTL_FALLING	1
+#define SPICC_DRCTL_LOWLEVEL	2
+#define SPICC_CS_MASK		GENMASK(13, 12)
+#define SPICC_DATARATE_MASK	GENMASK(18, 16)
+#define SPICC_DATARATE_DIV4	0
+#define SPICC_DATARATE_DIV8	1
+#define SPICC_DATARATE_DIV16	2
+#define SPICC_DATARATE_DIV32	3
+#define SPICC_BITLENGTH_MASK	GENMASK(24, 19)
+#define SPICC_BURSTLENGTH_MASK	GENMASK(31, 25)
+
+#define SPICC_INTREG	0x0c
+#define SPICC_TE_EN	BIT(0) /* TX FIFO Empty Interrupt */
+#define SPICC_TH_EN	BIT(1) /* TX FIFO Half-Full Interrupt */
+#define SPICC_TF_EN	BIT(2) /* TX FIFO Full Interrupt */
+#define SPICC_RR_EN	BIT(3) /* RX FIFO Ready Interrupt */
+#define SPICC_RH_EN	BIT(4) /* RX FIFO Half-Full Interrupt */
+#define SPICC_RF_EN	BIT(5) /* RX FIFO Full Interrupt */
+#define SPICC_RO_EN	BIT(6) /* RX FIFO Overflow Interrupt */
+#define SPICC_TC_EN	BIT(7) /* Transfert Complete Interrupt */
+
+#define SPICC_DMAREG	0x10
+#define SPICC_DMA_ENABLE		BIT(0)
+#define SPICC_TXFIFO_THRESHOLD_MASK	GENMASK(5, 1)
+#define SPICC_RXFIFO_THRESHOLD_MASK	GENMASK(10, 6)
+#define SPICC_READ_BURST_MASK		GENMASK(14, 11)
+#define SPICC_WRITE_BURST_MASK		GENMASK(18, 15)
+#define SPICC_DMA_URGENT		BIT(19)
+#define SPICC_DMA_THREADID_MASK		GENMASK(25, 20)
+#define SPICC_DMA_BURSTNUM_MASK		GENMASK(31, 26)
+
+#define SPICC_STATREG	0x14
+#define SPICC_TE	BIT(0) /* TX FIFO Empty Interrupt */
+#define SPICC_TH	BIT(1) /* TX FIFO Half-Full Interrupt */
+#define SPICC_TF	BIT(2) /* TX FIFO Full Interrupt */
+#define SPICC_RR	BIT(3) /* RX FIFO Ready Interrupt */
+#define SPICC_RH	BIT(4) /* RX FIFO Half-Full Interrupt */
+#define SPICC_RF	BIT(5) /* RX FIFO Full Interrupt */
+#define SPICC_RO	BIT(6) /* RX FIFO Overflow Interrupt */
+#define SPICC_TC	BIT(7) /* Transfert Complete Interrupt */
+
+#define SPICC_PERIODREG	0x18
+#define SPICC_PERIOD	GENMASK(14, 0)	/* Wait cycles */
+
+#define SPICC_TESTREG	0x1c
+#define SPICC_TXCNT_MASK	GENMASK(4, 0)	/* TX FIFO Counter */
+#define SPICC_RXCNT_MASK	GENMASK(9, 5)	/* RX FIFO Counter */
+#define SPICC_SMSTATUS_MASK	GENMASK(12, 10)	/* State Machine Status */
+#define SPICC_LBC_RO		BIT(13)	/* Loop Back Control Read-Only */
+#define SPICC_LBC_W1		BIT(14) /* Loop Back Control Write-Only */
+#define SPICC_SWAP_RO		BIT(14) /* RX FIFO Data Swap Read-Only */
+#define SPICC_SWAP_W1		BIT(15) /* RX FIFO Data Swap Write-Only */
+#define SPICC_DLYCTL_RO_MASK	GENMASK(20, 15) /* Delay Control Read-Only */
+#define SPICC_DLYCTL_W1_MASK	GENMASK(21, 16) /* Delay Control Write-Only */
+#define SPICC_FIFORST_RO_MASK	GENMASK(22, 21) /* FIFO Softreset Read-Only */
+#define SPICC_FIFORST_W1_MASK	GENMASK(23, 22) /* FIFO Softreset Write-Only */
+
+#define SPICC_DRADDR	0x20	/* Read Address of DMA */
+
+#define SPICC_DWADDR	0x24	/* Write Address of DMA */
+
+#define writel_bits_relaxed(mask, val, addr) \
+	writel_relaxed((readl_relaxed(addr) & ~(mask)) | (val), addr)
+
+#define SPICC_BURST_MAX	16
+#define SPICC_FIFO_HALF 10
+
+struct meson_spicc_device {
+	struct spi_master		*master;
+	struct platform_device		*pdev;
+	void __iomem			*base;
+	struct clk			*core;
+	struct spi_message		*message;
+	struct spi_transfer		*xfer;
+	u8				*tx_buf;
+	u8				*rx_buf;
+	unsigned int			bytes_per_word;
+	unsigned long			tx_remain;
+	unsigned long			txb_remain;
+	unsigned long			rx_remain;
+	unsigned long			rxb_remain;
+	unsigned long			xfer_remain;
+	bool				is_burst_end;
+	bool				is_last_burst;
+};
+
+static inline bool meson_spicc_txfull(struct meson_spicc_device *spicc)
+{
+	return !!FIELD_GET(SPICC_TF,
+			   readl_relaxed(spicc->base + SPICC_STATREG));
+}
+
+static inline bool meson_spicc_rxready(struct meson_spicc_device *spicc)
+{
+	return FIELD_GET(SPICC_RH | SPICC_RR | SPICC_RF_EN,
+			 readl_relaxed(spicc->base + SPICC_STATREG));
+}
+
+static inline u32 meson_spicc_pull_data(struct meson_spicc_device *spicc)
+{
+	unsigned int bytes = spicc->bytes_per_word;
+	unsigned int byte_shift = 0;
+	u32 data = 0;
+	u8 byte;
+
+	while (bytes--) {
+		byte = *spicc->tx_buf++;
+		data |= (byte & 0xff) << byte_shift;
+		byte_shift += 8;
+	}
+
+	spicc->tx_remain--;
+	return data;
+}
+
+static inline void meson_spicc_push_data(struct meson_spicc_device *spicc,
+					 u32 data)
+{
+	unsigned int bytes = spicc->bytes_per_word;
+	unsigned int byte_shift = 0;
+	u8 byte;
+
+	while (bytes--) {
+		byte = (data >> byte_shift) & 0xff;
+		*spicc->rx_buf++ = byte;
+		byte_shift += 8;
+	}
+
+	spicc->rx_remain--;
+}
+
+static inline void meson_spicc_rx(struct meson_spicc_device *spicc)
+{
+	/* Empty RX FIFO */
+	while (spicc->rx_remain &&
+	       meson_spicc_rxready(spicc))
+		meson_spicc_push_data(spicc,
+				readl_relaxed(spicc->base + SPICC_RXDATA));
+}
+
+static inline void meson_spicc_tx(struct meson_spicc_device *spicc)
+{
+	/* Fill Up TX FIFO */
+	while (spicc->tx_remain &&
+	       !meson_spicc_txfull(spicc))
+		writel_relaxed(meson_spicc_pull_data(spicc),
+			       spicc->base + SPICC_TXDATA);
+}
+
+static inline u32 meson_spicc_setup_rx_irq(struct meson_spicc_device *spicc,
+					   u32 irq_ctrl)
+{
+	if (spicc->rx_remain > SPICC_FIFO_HALF)
+		irq_ctrl |= SPICC_RH_EN;
+	else
+		irq_ctrl |= SPICC_RR_EN;
+
+	return irq_ctrl;
+}
+
+static inline void meson_spicc_setup_burst(struct meson_spicc_device *spicc,
+					   unsigned int burst_len)
+{
+	/* Setup Xfer variables */
+	spicc->tx_remain = burst_len;
+	spicc->rx_remain = burst_len;
+	spicc->xfer_remain -= burst_len * spicc->bytes_per_word;
+	spicc->is_burst_end = false;
+	if (burst_len < SPICC_BURST_MAX || !spicc->xfer_remain)
+		spicc->is_last_burst = true;
+	else
+		spicc->is_last_burst = false;
+
+	/* Setup burst length */
+	writel_bits_relaxed(SPICC_BURSTLENGTH_MASK,
+			FIELD_PREP(SPICC_BURSTLENGTH_MASK,
+				burst_len),
+			spicc->base + SPICC_CONREG);
+
+	/* Fill TX FIFO */
+	meson_spicc_tx(spicc);
+}
+
+static irqreturn_t meson_spicc_irq(int irq, void *data)
+{
+	struct meson_spicc_device *spicc = (void *) data;
+	u32 ctrl = readl_relaxed(spicc->base + SPICC_INTREG);
+	u32 stat = readl_relaxed(spicc->base + SPICC_STATREG) & ctrl;
+
+	ctrl &= ~(SPICC_RH_EN | SPICC_RR_EN);
+
+	/* Empty RX FIFO */
+	meson_spicc_rx(spicc);
+
+	/* Enable TC interrupt since we transferred everything */
+	if (!spicc->tx_remain && !spicc->rx_remain) {
+		spicc->is_burst_end = true;
+
+		/* Enable TC interrupt */
+		ctrl |= SPICC_TC_EN;
+
+		/* Reload IRQ status */
+		stat = readl_relaxed(spicc->base + SPICC_STATREG) & ctrl;
+	}
+
+	/* Check transfer complete */
+	if ((stat & SPICC_TC) && spicc->is_burst_end) {
+		unsigned int burst_len;
+
+		/* Clear TC bit */
+		writel_relaxed(SPICC_TC, spicc->base + SPICC_STATREG);
+
+		/* Disable TC interrupt */
+		ctrl &= ~SPICC_TC_EN;
+
+		if (spicc->is_last_burst) {
+			/* Disable all IRQs */
+			writel(0, spicc->base + SPICC_INTREG);
+
+			spi_finalize_current_transfer(spicc->master);
+
+			return IRQ_HANDLED;
+		}
+
+		burst_len = min_t(unsigned int,
+				  spicc->xfer_remain / spicc->bytes_per_word,
+				  SPICC_BURST_MAX);
+
+		/* Setup burst */
+		meson_spicc_setup_burst(spicc, burst_len);
+
+		/* Restart burst */
+		writel_bits_relaxed(SPICC_XCH, SPICC_XCH,
+				    spicc->base + SPICC_CONREG);
+	}
+
+	/* Setup RX interrupt trigger */
+	ctrl = meson_spicc_setup_rx_irq(spicc, ctrl);
+
+	/* Reconfigure interrupts */
+	writel(ctrl, spicc->base + SPICC_INTREG);
+
+	return IRQ_HANDLED;
+}
+
+static u32 meson_spicc_setup_speed(struct meson_spicc_device *spicc, u32 conf,
+				   u32 speed)
+{
+	unsigned long parent, value;
+	unsigned int i, div;
+
+	parent = clk_get_rate(spicc->core);
+
+	/* Find closest inferior/equal possible speed */
+	for (i = 0 ; i < 7 ; ++i) {
+		/* 2^(data_rate+2) */
+		value = parent >> (i + 2);
+
+		if (value <= speed)
+			break;
+	}
+
+	/* If provided speed it lower than max divider, use max divider */
+	if (i > 7) {
+		div = 7;
+		dev_warn_once(&spicc->pdev->dev, "unable to get close to speed %u\n",
+			      speed);
+	} else
+		div = i;
+
+	dev_dbg(&spicc->pdev->dev, "parent %lu, speed %u -> %lu (%u)\n",
+		parent, speed, value, div);
+
+	conf &= ~SPICC_DATARATE_MASK;
+	conf |= FIELD_PREP(SPICC_DATARATE_MASK, div);
+
+	return conf;
+}
+
+static void meson_spicc_setup_xfer(struct meson_spicc_device *spicc,
+				   struct spi_transfer *xfer)
+{
+	u32 conf, conf_orig;
+
+	/* Read original configuration */
+	conf = conf_orig = readl_relaxed(spicc->base + SPICC_CONREG);
+
+	/* Select closest divider */
+	conf = meson_spicc_setup_speed(spicc, conf, xfer->speed_hz);
+
+	/* Setup word width */
+	conf &= ~SPICC_BITLENGTH_MASK;
+	conf |= FIELD_PREP(SPICC_BITLENGTH_MASK,
+			   (spicc->bytes_per_word << 3) - 1);
+
+	/* Ignore if unchanged */
+	if (conf != conf_orig)
+		writel_relaxed(conf, spicc->base + SPICC_CONREG);
+}
+
+static int meson_spicc_transfer_one(struct spi_master *master,
+				    struct spi_device *spi,
+				    struct spi_transfer *xfer)
+{
+	struct meson_spicc_device *spicc = spi_master_get_devdata(master);
+	unsigned int burst_len;
+	u32 irq = 0;
+
+	/* Store current transfer */
+	spicc->xfer = xfer;
+
+	/* Setup transfer parameters */
+	spicc->tx_buf = (u8 *)xfer->tx_buf;
+	spicc->rx_buf = (u8 *)xfer->rx_buf;
+	spicc->xfer_remain = xfer->len;
+
+	/* Pre-calculate word size */
+	spicc->bytes_per_word =
+	   DIV_ROUND_UP(spicc->xfer->bits_per_word, 8);
+
+	/* Setup transfer parameters */
+	meson_spicc_setup_xfer(spicc, xfer);
+
+	burst_len = min_t(unsigned int,
+			  spicc->xfer_remain / spicc->bytes_per_word,
+			  SPICC_BURST_MAX);
+
+	meson_spicc_setup_burst(spicc, burst_len);
+
+	irq = meson_spicc_setup_rx_irq(spicc, irq);
+
+	/* Start burst */
+	writel_bits_relaxed(SPICC_XCH, SPICC_XCH, spicc->base + SPICC_CONREG);
+
+	/* Enable interrupts */
+	writel_relaxed(irq, spicc->base + SPICC_INTREG);
+
+	return 1;
+}
+
+static int meson_spicc_prepare_message(struct spi_master *master,
+				       struct spi_message *message)
+{
+	struct meson_spicc_device *spicc = spi_master_get_devdata(master);
+	struct spi_device *spi = message->spi;
+	u32 conf = 0;
+
+	/* Store current message */
+	spicc->message = message;
+
+	/* Enable Master */
+	conf |= SPICC_ENABLE;
+	conf |= SPICC_MODE_MASTER;
+
+	/* SMC = 0 */
+
+	/* Setup transfer mode */
+	if (spi->mode & SPI_CPOL)
+		conf |= SPICC_POL;
+	else
+		conf &= ~SPICC_POL;
+
+	if (spi->mode & SPI_CPHA)
+		conf |= SPICC_PHA;
+	else
+		conf &= ~SPICC_PHA;
+
+	/* SSCTL = 0 */
+
+	if (spi->mode & SPI_CS_HIGH)
+		conf |= SPICC_SSPOL;
+	else
+		conf &= ~SPICC_SSPOL;
+
+	if (spi->mode & SPI_READY)
+		conf |= FIELD_PREP(SPICC_DRCTL_MASK, SPICC_DRCTL_LOWLEVEL);
+	else
+		conf |= FIELD_PREP(SPICC_DRCTL_MASK, SPICC_DRCTL_IGNORE);
+
+	/* Select CS */
+	conf |= FIELD_PREP(SPICC_CS_MASK, spi->chip_select);
+
+	/* Default Clock rate core/4 */
+
+	/* Default 8bit word */
+	conf |= FIELD_PREP(SPICC_BITLENGTH_MASK, 8 - 1);
+
+	writel_relaxed(conf, spicc->base + SPICC_CONREG);
+
+	/* Setup no wait cycles by default */
+	writel_relaxed(0, spicc->base + SPICC_PERIODREG);
+
+	writel_bits_relaxed(BIT(24), BIT(24), spicc->base + SPICC_TESTREG);
+
+	return 0;
+}
+
+static int meson_spicc_unprepare_transfer(struct spi_master *master)
+{
+	struct meson_spicc_device *spicc = spi_master_get_devdata(master);
+
+	/* Disable all IRQs */
+	writel(0, spicc->base + SPICC_INTREG);
+
+	/* Disable controller */
+	writel_bits_relaxed(SPICC_ENABLE, 0, spicc->base + SPICC_CONREG);
+
+	device_reset_optional(&spicc->pdev->dev);
+
+	return 0;
+}
+
+static int meson_spicc_setup(struct spi_device *spi)
+{
+	int ret = 0;
+
+	if (!spi->controller_state)
+		spi->controller_state = spi_master_get_devdata(spi->master);
+	else if (gpio_is_valid(spi->cs_gpio))
+		goto out_gpio;
+	else if (spi->cs_gpio == -ENOENT)
+		return 0;
+
+	if (gpio_is_valid(spi->cs_gpio)) {
+		ret = gpio_request(spi->cs_gpio, dev_name(&spi->dev));
+		if (ret) {
+			dev_err(&spi->dev, "failed to request cs gpio\n");
+			return ret;
+		}
+	}
+
+out_gpio:
+	ret = gpio_direction_output(spi->cs_gpio,
+			!(spi->mode & SPI_CS_HIGH));
+
+	return ret;
+}
+
+static void meson_spicc_cleanup(struct spi_device *spi)
+{
+	if (gpio_is_valid(spi->cs_gpio))
+		gpio_free(spi->cs_gpio);
+
+	spi->controller_state = NULL;
+}
+
+static int meson_spicc_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct meson_spicc_device *spicc;
+	struct resource *res;
+	int ret, irq, rate;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*spicc));
+	if (!master) {
+		dev_err(&pdev->dev, "master allocation failed\n");
+		return -ENOMEM;
+	}
+	spicc = spi_master_get_devdata(master);
+	spicc->master = master;
+
+	spicc->pdev = pdev;
+	platform_set_drvdata(pdev, spicc);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	spicc->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(spicc->base)) {
+		dev_err(&pdev->dev, "io resource mapping failed\n");
+		ret = PTR_ERR(spicc->base);
+		goto out_master;
+	}
+
+	/* Disable all IRQs */
+	writel_relaxed(0, spicc->base + SPICC_INTREG);
+
+	irq = platform_get_irq(pdev, 0);
+	ret = devm_request_irq(&pdev->dev, irq, meson_spicc_irq,
+			       0, NULL, spicc);
+	if (ret) {
+		dev_err(&pdev->dev, "irq request failed\n");
+		goto out_master;
+	}
+
+	spicc->core = devm_clk_get(&pdev->dev, "core");
+	if (IS_ERR(spicc->core)) {
+		dev_err(&pdev->dev, "core clock request failed\n");
+		ret = PTR_ERR(spicc->core);
+		goto out_master;
+	}
+
+	ret = clk_prepare_enable(spicc->core);
+	if (ret) {
+		dev_err(&pdev->dev, "core clock enable failed\n");
+		goto out_master;
+	}
+	rate = clk_get_rate(spicc->core);
+
+	device_reset_optional(&pdev->dev);
+
+	master->num_chipselect = 4;
+	master->dev.of_node = pdev->dev.of_node;
+	master->mode_bits = SPI_CPHA | SPI_CPOL | SPI_CS_HIGH;
+	master->bits_per_word_mask = SPI_BPW_MASK(32) |
+				     SPI_BPW_MASK(24) |
+				     SPI_BPW_MASK(16) |
+				     SPI_BPW_MASK(8);
+	master->flags = (SPI_MASTER_MUST_RX | SPI_MASTER_MUST_TX);
+	master->min_speed_hz = rate >> 9;
+	master->setup = meson_spicc_setup;
+	master->cleanup = meson_spicc_cleanup;
+	master->prepare_message = meson_spicc_prepare_message;
+	master->unprepare_transfer_hardware = meson_spicc_unprepare_transfer;
+	master->transfer_one = meson_spicc_transfer_one;
+
+	/* Setup max rate according to the Meson GX datasheet */
+	if ((rate >> 2) > SPICC_MAX_FREQ)
+		master->max_speed_hz = SPICC_MAX_FREQ;
+	else
+		master->max_speed_hz = rate >> 2;
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret) {
+		dev_err(&pdev->dev, "spi master registration failed\n");
+		goto out_clk;
+	}
+
+	return 0;
+
+out_clk:
+	clk_disable_unprepare(spicc->core);
+
+out_master:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int meson_spicc_remove(struct platform_device *pdev)
+{
+	struct meson_spicc_device *spicc = platform_get_drvdata(pdev);
+
+	/* Disable SPI */
+	writel(0, spicc->base + SPICC_CONREG);
+
+	clk_disable_unprepare(spicc->core);
+
+	return 0;
+}
+
+static const struct of_device_id meson_spicc_of_match[] = {
+	{ .compatible = "amlogic,meson-gx-spicc", },
+	{ .compatible = "amlogic,meson-axg-spicc", },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, meson_spicc_of_match);
+
+static struct platform_driver meson_spicc_driver = {
+	.probe   = meson_spicc_probe,
+	.remove  = meson_spicc_remove,
+	.driver  = {
+		.name = "meson-spicc",
+		.of_match_table = of_match_ptr(meson_spicc_of_match),
+	},
+};
+
+module_platform_driver(meson_spicc_driver);
+
+MODULE_DESCRIPTION("Meson SPI Communication Controller driver");
+MODULE_AUTHOR("Neil Armstrong <narmstrong@baylibre.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-meson-spifc.c b/drivers/spi/spi-meson-spifc.c
new file mode 100644
index 0000000..616566e
--- /dev/null
+++ b/drivers/spi/spi-meson-spifc.c
@@ -0,0 +1,464 @@
+/*
+ * Driver for Amlogic Meson SPI flash controller (SPIFC)
+ *
+ * Copyright (C) 2014 Beniamino Galvani <b.galvani@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * version 2 as published by the Free Software Foundation.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/regmap.h>
+#include <linux/spi/spi.h>
+#include <linux/types.h>
+
+/* register map */
+#define REG_CMD			0x00
+#define REG_ADDR		0x04
+#define REG_CTRL		0x08
+#define REG_CTRL1		0x0c
+#define REG_STATUS		0x10
+#define REG_CTRL2		0x14
+#define REG_CLOCK		0x18
+#define REG_USER		0x1c
+#define REG_USER1		0x20
+#define REG_USER2		0x24
+#define REG_USER3		0x28
+#define REG_USER4		0x2c
+#define REG_SLAVE		0x30
+#define REG_SLAVE1		0x34
+#define REG_SLAVE2		0x38
+#define REG_SLAVE3		0x3c
+#define REG_C0			0x40
+#define REG_B8			0x60
+#define REG_MAX			0x7c
+
+/* register fields */
+#define CMD_USER		BIT(18)
+#define CTRL_ENABLE_AHB		BIT(17)
+#define CLOCK_SOURCE		BIT(31)
+#define CLOCK_DIV_SHIFT		12
+#define CLOCK_DIV_MASK		(0x3f << CLOCK_DIV_SHIFT)
+#define CLOCK_CNT_HIGH_SHIFT	6
+#define CLOCK_CNT_HIGH_MASK	(0x3f << CLOCK_CNT_HIGH_SHIFT)
+#define CLOCK_CNT_LOW_SHIFT	0
+#define CLOCK_CNT_LOW_MASK	(0x3f << CLOCK_CNT_LOW_SHIFT)
+#define USER_DIN_EN_MS		BIT(0)
+#define USER_CMP_MODE		BIT(2)
+#define USER_UC_DOUT_SEL	BIT(27)
+#define USER_UC_DIN_SEL		BIT(28)
+#define USER_UC_MASK		((BIT(5) - 1) << 27)
+#define USER1_BN_UC_DOUT_SHIFT	17
+#define USER1_BN_UC_DOUT_MASK	(0xff << 16)
+#define USER1_BN_UC_DIN_SHIFT	8
+#define USER1_BN_UC_DIN_MASK	(0xff << 8)
+#define USER4_CS_ACT		BIT(30)
+#define SLAVE_TRST_DONE		BIT(4)
+#define SLAVE_OP_MODE		BIT(30)
+#define SLAVE_SW_RST		BIT(31)
+
+#define SPIFC_BUFFER_SIZE	64
+
+/**
+ * struct meson_spifc
+ * @master:	the SPI master
+ * @regmap:	regmap for device registers
+ * @clk:	input clock of the built-in baud rate generator
+ * @device:	the device structure
+ */
+struct meson_spifc {
+	struct spi_master *master;
+	struct regmap *regmap;
+	struct clk *clk;
+	struct device *dev;
+};
+
+static const struct regmap_config spifc_regmap_config = {
+	.reg_bits = 32,
+	.val_bits = 32,
+	.reg_stride = 4,
+	.max_register = REG_MAX,
+};
+
+/**
+ * meson_spifc_wait_ready() - wait for the current operation to terminate
+ * @spifc:	the Meson SPI device
+ * Return:	0 on success, a negative value on error
+ */
+static int meson_spifc_wait_ready(struct meson_spifc *spifc)
+{
+	unsigned long deadline = jiffies + msecs_to_jiffies(5);
+	u32 data;
+
+	do {
+		regmap_read(spifc->regmap, REG_SLAVE, &data);
+		if (data & SLAVE_TRST_DONE)
+			return 0;
+		cond_resched();
+	} while (!time_after(jiffies, deadline));
+
+	return -ETIMEDOUT;
+}
+
+/**
+ * meson_spifc_drain_buffer() - copy data from device buffer to memory
+ * @spifc:	the Meson SPI device
+ * @buf:	the destination buffer
+ * @len:	number of bytes to copy
+ */
+static void meson_spifc_drain_buffer(struct meson_spifc *spifc, u8 *buf,
+				     int len)
+{
+	u32 data;
+	int i = 0;
+
+	while (i < len) {
+		regmap_read(spifc->regmap, REG_C0 + i, &data);
+
+		if (len - i >= 4) {
+			*((u32 *)buf) = data;
+			buf += 4;
+		} else {
+			memcpy(buf, &data, len - i);
+			break;
+		}
+		i += 4;
+	}
+}
+
+/**
+ * meson_spifc_fill_buffer() - copy data from memory to device buffer
+ * @spifc:	the Meson SPI device
+ * @buf:	the source buffer
+ * @len:	number of bytes to copy
+ */
+static void meson_spifc_fill_buffer(struct meson_spifc *spifc, const u8 *buf,
+				    int len)
+{
+	u32 data;
+	int i = 0;
+
+	while (i < len) {
+		if (len - i >= 4)
+			data = *(u32 *)buf;
+		else
+			memcpy(&data, buf, len - i);
+
+		regmap_write(spifc->regmap, REG_C0 + i, data);
+
+		buf += 4;
+		i += 4;
+	}
+}
+
+/**
+ * meson_spifc_setup_speed() - program the clock divider
+ * @spifc:	the Meson SPI device
+ * @speed:	desired speed in Hz
+ */
+static void meson_spifc_setup_speed(struct meson_spifc *spifc, u32 speed)
+{
+	unsigned long parent, value;
+	int n;
+
+	parent = clk_get_rate(spifc->clk);
+	n = max_t(int, parent / speed - 1, 1);
+
+	dev_dbg(spifc->dev, "parent %lu, speed %u, n %d\n", parent,
+		speed, n);
+
+	value = (n << CLOCK_DIV_SHIFT) & CLOCK_DIV_MASK;
+	value |= (n << CLOCK_CNT_LOW_SHIFT) & CLOCK_CNT_LOW_MASK;
+	value |= (((n + 1) / 2 - 1) << CLOCK_CNT_HIGH_SHIFT) &
+		CLOCK_CNT_HIGH_MASK;
+
+	regmap_write(spifc->regmap, REG_CLOCK, value);
+}
+
+/**
+ * meson_spifc_txrx() - transfer a chunk of data
+ * @spifc:	the Meson SPI device
+ * @xfer:	the current SPI transfer
+ * @offset:	offset of the data to transfer
+ * @len:	length of the data to transfer
+ * @last_xfer:	whether this is the last transfer of the message
+ * @last_chunk:	whether this is the last chunk of the transfer
+ * Return:	0 on success, a negative value on error
+ */
+static int meson_spifc_txrx(struct meson_spifc *spifc,
+			    struct spi_transfer *xfer,
+			    int offset, int len, bool last_xfer,
+			    bool last_chunk)
+{
+	bool keep_cs = true;
+	int ret;
+
+	if (xfer->tx_buf)
+		meson_spifc_fill_buffer(spifc, xfer->tx_buf + offset, len);
+
+	/* enable DOUT stage */
+	regmap_update_bits(spifc->regmap, REG_USER, USER_UC_MASK,
+			   USER_UC_DOUT_SEL);
+	regmap_write(spifc->regmap, REG_USER1,
+		     (8 * len - 1) << USER1_BN_UC_DOUT_SHIFT);
+
+	/* enable data input during DOUT */
+	regmap_update_bits(spifc->regmap, REG_USER, USER_DIN_EN_MS,
+			   USER_DIN_EN_MS);
+
+	if (last_chunk) {
+		if (last_xfer)
+			keep_cs = xfer->cs_change;
+		else
+			keep_cs = !xfer->cs_change;
+	}
+
+	regmap_update_bits(spifc->regmap, REG_USER4, USER4_CS_ACT,
+			   keep_cs ? USER4_CS_ACT : 0);
+
+	/* clear transition done bit */
+	regmap_update_bits(spifc->regmap, REG_SLAVE, SLAVE_TRST_DONE, 0);
+	/* start transfer */
+	regmap_update_bits(spifc->regmap, REG_CMD, CMD_USER, CMD_USER);
+
+	ret = meson_spifc_wait_ready(spifc);
+
+	if (!ret && xfer->rx_buf)
+		meson_spifc_drain_buffer(spifc, xfer->rx_buf + offset, len);
+
+	return ret;
+}
+
+/**
+ * meson_spifc_transfer_one() - perform a single transfer
+ * @master:	the SPI master
+ * @spi:	the SPI device
+ * @xfer:	the current SPI transfer
+ * Return:	0 on success, a negative value on error
+ */
+static int meson_spifc_transfer_one(struct spi_master *master,
+				    struct spi_device *spi,
+				    struct spi_transfer *xfer)
+{
+	struct meson_spifc *spifc = spi_master_get_devdata(master);
+	int len, done = 0, ret = 0;
+
+	meson_spifc_setup_speed(spifc, xfer->speed_hz);
+
+	regmap_update_bits(spifc->regmap, REG_CTRL, CTRL_ENABLE_AHB, 0);
+
+	while (done < xfer->len && !ret) {
+		len = min_t(int, xfer->len - done, SPIFC_BUFFER_SIZE);
+		ret = meson_spifc_txrx(spifc, xfer, done, len,
+				       spi_transfer_is_last(master, xfer),
+				       done + len >= xfer->len);
+		done += len;
+	}
+
+	regmap_update_bits(spifc->regmap, REG_CTRL, CTRL_ENABLE_AHB,
+			   CTRL_ENABLE_AHB);
+
+	return ret;
+}
+
+/**
+ * meson_spifc_hw_init() - reset and initialize the SPI controller
+ * @spifc:	the Meson SPI device
+ */
+static void meson_spifc_hw_init(struct meson_spifc *spifc)
+{
+	/* reset device */
+	regmap_update_bits(spifc->regmap, REG_SLAVE, SLAVE_SW_RST,
+			   SLAVE_SW_RST);
+	/* disable compatible mode */
+	regmap_update_bits(spifc->regmap, REG_USER, USER_CMP_MODE, 0);
+	/* set master mode */
+	regmap_update_bits(spifc->regmap, REG_SLAVE, SLAVE_OP_MODE, 0);
+}
+
+static int meson_spifc_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct meson_spifc *spifc;
+	struct resource *res;
+	void __iomem *base;
+	unsigned int rate;
+	int ret = 0;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct meson_spifc));
+	if (!master)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, master);
+
+	spifc = spi_master_get_devdata(master);
+	spifc->dev = &pdev->dev;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	base = devm_ioremap_resource(spifc->dev, res);
+	if (IS_ERR(base)) {
+		ret = PTR_ERR(base);
+		goto out_err;
+	}
+
+	spifc->regmap = devm_regmap_init_mmio(spifc->dev, base,
+					      &spifc_regmap_config);
+	if (IS_ERR(spifc->regmap)) {
+		ret = PTR_ERR(spifc->regmap);
+		goto out_err;
+	}
+
+	spifc->clk = devm_clk_get(spifc->dev, NULL);
+	if (IS_ERR(spifc->clk)) {
+		dev_err(spifc->dev, "missing clock\n");
+		ret = PTR_ERR(spifc->clk);
+		goto out_err;
+	}
+
+	ret = clk_prepare_enable(spifc->clk);
+	if (ret) {
+		dev_err(spifc->dev, "can't prepare clock\n");
+		goto out_err;
+	}
+
+	rate = clk_get_rate(spifc->clk);
+
+	master->num_chipselect = 1;
+	master->dev.of_node = pdev->dev.of_node;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->auto_runtime_pm = true;
+	master->transfer_one = meson_spifc_transfer_one;
+	master->min_speed_hz = rate >> 6;
+	master->max_speed_hz = rate >> 1;
+
+	meson_spifc_hw_init(spifc);
+
+	pm_runtime_set_active(spifc->dev);
+	pm_runtime_enable(spifc->dev);
+
+	ret = devm_spi_register_master(spifc->dev, master);
+	if (ret) {
+		dev_err(spifc->dev, "failed to register spi master\n");
+		goto out_clk;
+	}
+
+	return 0;
+out_clk:
+	clk_disable_unprepare(spifc->clk);
+out_err:
+	spi_master_put(master);
+	return ret;
+}
+
+static int meson_spifc_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct meson_spifc *spifc = spi_master_get_devdata(master);
+
+	pm_runtime_get_sync(&pdev->dev);
+	clk_disable_unprepare(spifc->clk);
+	pm_runtime_disable(&pdev->dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int meson_spifc_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct meson_spifc *spifc = spi_master_get_devdata(master);
+	int ret;
+
+	ret = spi_master_suspend(master);
+	if (ret)
+		return ret;
+
+	if (!pm_runtime_suspended(dev))
+		clk_disable_unprepare(spifc->clk);
+
+	return 0;
+}
+
+static int meson_spifc_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct meson_spifc *spifc = spi_master_get_devdata(master);
+	int ret;
+
+	if (!pm_runtime_suspended(dev)) {
+		ret = clk_prepare_enable(spifc->clk);
+		if (ret)
+			return ret;
+	}
+
+	meson_spifc_hw_init(spifc);
+
+	ret = spi_master_resume(master);
+	if (ret)
+		clk_disable_unprepare(spifc->clk);
+
+	return ret;
+}
+#endif /* CONFIG_PM_SLEEP */
+
+#ifdef CONFIG_PM
+static int meson_spifc_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct meson_spifc *spifc = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(spifc->clk);
+
+	return 0;
+}
+
+static int meson_spifc_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct meson_spifc *spifc = spi_master_get_devdata(master);
+
+	return clk_prepare_enable(spifc->clk);
+}
+#endif /* CONFIG_PM */
+
+static const struct dev_pm_ops meson_spifc_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(meson_spifc_suspend, meson_spifc_resume)
+	SET_RUNTIME_PM_OPS(meson_spifc_runtime_suspend,
+			   meson_spifc_runtime_resume,
+			   NULL)
+};
+
+static const struct of_device_id meson_spifc_dt_match[] = {
+	{ .compatible = "amlogic,meson6-spifc", },
+	{ .compatible = "amlogic,meson-gxbb-spifc", },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, meson_spifc_dt_match);
+
+static struct platform_driver meson_spifc_driver = {
+	.probe	= meson_spifc_probe,
+	.remove	= meson_spifc_remove,
+	.driver	= {
+		.name		= "meson-spifc",
+		.of_match_table	= of_match_ptr(meson_spifc_dt_match),
+		.pm		= &meson_spifc_pm_ops,
+	},
+};
+
+module_platform_driver(meson_spifc_driver);
+
+MODULE_AUTHOR("Beniamino Galvani <b.galvani@gmail.com>");
+MODULE_DESCRIPTION("Amlogic Meson SPIFC driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-mpc512x-psc.c b/drivers/spi/spi-mpc512x-psc.c
new file mode 100644
index 0000000..c3ec46c
--- /dev/null
+++ b/drivers/spi/spi-mpc512x-psc.c
@@ -0,0 +1,632 @@
+/*
+ * MPC512x PSC in SPI mode driver.
+ *
+ * Copyright (C) 2007,2008 Freescale Semiconductor Inc.
+ * Original port from 52xx driver:
+ *	Hongjun Chen <hong-jun.chen@freescale.com>
+ *
+ * Fork of mpc52xx_psc_spi.c:
+ *	Copyright (C) 2006 TOPTICA Photonics AG., Dragos Carp
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/of_platform.h>
+#include <linux/completion.h>
+#include <linux/io.h>
+#include <linux/delay.h>
+#include <linux/clk.h>
+#include <linux/spi/spi.h>
+#include <linux/fsl_devices.h>
+#include <linux/gpio.h>
+#include <asm/mpc52xx_psc.h>
+
+enum {
+	TYPE_MPC5121,
+	TYPE_MPC5125,
+};
+
+/*
+ * This macro abstracts the differences in the PSC register layout between
+ * MPC5121 (which uses a struct mpc52xx_psc) and MPC5125 (using mpc5125_psc).
+ */
+#define psc_addr(mps, regname) ({					\
+	void *__ret = NULL;						\
+	switch (mps->type) {						\
+	case TYPE_MPC5121: {						\
+			struct mpc52xx_psc __iomem *psc = mps->psc;	\
+			__ret = &psc->regname;				\
+		};							\
+		break;							\
+	case TYPE_MPC5125: {						\
+			struct mpc5125_psc __iomem *psc = mps->psc;	\
+			__ret = &psc->regname;				\
+		};							\
+		break;							\
+	}								\
+	__ret; })
+
+struct mpc512x_psc_spi {
+	void (*cs_control)(struct spi_device *spi, bool on);
+
+	/* driver internal data */
+	int type;
+	void __iomem *psc;
+	struct mpc512x_psc_fifo __iomem *fifo;
+	unsigned int irq;
+	u8 bits_per_word;
+	struct clk *clk_mclk;
+	struct clk *clk_ipg;
+	u32 mclk_rate;
+
+	struct completion txisrdone;
+};
+
+/* controller state */
+struct mpc512x_psc_spi_cs {
+	int bits_per_word;
+	int speed_hz;
+};
+
+/* set clock freq, clock ramp, bits per work
+ * if t is NULL then reset the values to the default values
+ */
+static int mpc512x_psc_spi_transfer_setup(struct spi_device *spi,
+					  struct spi_transfer *t)
+{
+	struct mpc512x_psc_spi_cs *cs = spi->controller_state;
+
+	cs->speed_hz = (t && t->speed_hz)
+	    ? t->speed_hz : spi->max_speed_hz;
+	cs->bits_per_word = (t && t->bits_per_word)
+	    ? t->bits_per_word : spi->bits_per_word;
+	cs->bits_per_word = ((cs->bits_per_word + 7) / 8) * 8;
+	return 0;
+}
+
+static void mpc512x_psc_spi_activate_cs(struct spi_device *spi)
+{
+	struct mpc512x_psc_spi_cs *cs = spi->controller_state;
+	struct mpc512x_psc_spi *mps = spi_master_get_devdata(spi->master);
+	u32 sicr;
+	u32 ccr;
+	int speed;
+	u16 bclkdiv;
+
+	sicr = in_be32(psc_addr(mps, sicr));
+
+	/* Set clock phase and polarity */
+	if (spi->mode & SPI_CPHA)
+		sicr |= 0x00001000;
+	else
+		sicr &= ~0x00001000;
+
+	if (spi->mode & SPI_CPOL)
+		sicr |= 0x00002000;
+	else
+		sicr &= ~0x00002000;
+
+	if (spi->mode & SPI_LSB_FIRST)
+		sicr |= 0x10000000;
+	else
+		sicr &= ~0x10000000;
+	out_be32(psc_addr(mps, sicr), sicr);
+
+	ccr = in_be32(psc_addr(mps, ccr));
+	ccr &= 0xFF000000;
+	speed = cs->speed_hz;
+	if (!speed)
+		speed = 1000000;	/* default 1MHz */
+	bclkdiv = (mps->mclk_rate / speed) - 1;
+
+	ccr |= (((bclkdiv & 0xff) << 16) | (((bclkdiv >> 8) & 0xff) << 8));
+	out_be32(psc_addr(mps, ccr), ccr);
+	mps->bits_per_word = cs->bits_per_word;
+
+	if (mps->cs_control && gpio_is_valid(spi->cs_gpio))
+		mps->cs_control(spi, (spi->mode & SPI_CS_HIGH) ? 1 : 0);
+}
+
+static void mpc512x_psc_spi_deactivate_cs(struct spi_device *spi)
+{
+	struct mpc512x_psc_spi *mps = spi_master_get_devdata(spi->master);
+
+	if (mps->cs_control && gpio_is_valid(spi->cs_gpio))
+		mps->cs_control(spi, (spi->mode & SPI_CS_HIGH) ? 0 : 1);
+
+}
+
+/* extract and scale size field in txsz or rxsz */
+#define MPC512x_PSC_FIFO_SZ(sz) ((sz & 0x7ff) << 2);
+
+#define EOFBYTE 1
+
+static int mpc512x_psc_spi_transfer_rxtx(struct spi_device *spi,
+					 struct spi_transfer *t)
+{
+	struct mpc512x_psc_spi *mps = spi_master_get_devdata(spi->master);
+	struct mpc512x_psc_fifo __iomem *fifo = mps->fifo;
+	size_t tx_len = t->len;
+	size_t rx_len = t->len;
+	u8 *tx_buf = (u8 *)t->tx_buf;
+	u8 *rx_buf = (u8 *)t->rx_buf;
+
+	if (!tx_buf && !rx_buf && t->len)
+		return -EINVAL;
+
+	while (rx_len || tx_len) {
+		size_t txcount;
+		u8 data;
+		size_t fifosz;
+		size_t rxcount;
+		int rxtries;
+
+		/*
+		 * send the TX bytes in as large a chunk as possible
+		 * but neither exceed the TX nor the RX FIFOs
+		 */
+		fifosz = MPC512x_PSC_FIFO_SZ(in_be32(&fifo->txsz));
+		txcount = min(fifosz, tx_len);
+		fifosz = MPC512x_PSC_FIFO_SZ(in_be32(&fifo->rxsz));
+		fifosz -= in_be32(&fifo->rxcnt) + 1;
+		txcount = min(fifosz, txcount);
+		if (txcount) {
+
+			/* fill the TX FIFO */
+			while (txcount-- > 0) {
+				data = tx_buf ? *tx_buf++ : 0;
+				if (tx_len == EOFBYTE && t->cs_change)
+					setbits32(&fifo->txcmd,
+						  MPC512x_PSC_FIFO_EOF);
+				out_8(&fifo->txdata_8, data);
+				tx_len--;
+			}
+
+			/* have the ISR trigger when the TX FIFO is empty */
+			reinit_completion(&mps->txisrdone);
+			out_be32(&fifo->txisr, MPC512x_PSC_FIFO_EMPTY);
+			out_be32(&fifo->tximr, MPC512x_PSC_FIFO_EMPTY);
+			wait_for_completion(&mps->txisrdone);
+		}
+
+		/*
+		 * consume as much RX data as the FIFO holds, while we
+		 * iterate over the transfer's TX data length
+		 *
+		 * only insist in draining all the remaining RX bytes
+		 * when the TX bytes were exhausted (that's at the very
+		 * end of this transfer, not when still iterating over
+		 * the transfer's chunks)
+		 */
+		rxtries = 50;
+		do {
+
+			/*
+			 * grab whatever was in the FIFO when we started
+			 * looking, don't bother fetching what was added to
+			 * the FIFO while we read from it -- we'll return
+			 * here eventually and prefer sending out remaining
+			 * TX data
+			 */
+			fifosz = in_be32(&fifo->rxcnt);
+			rxcount = min(fifosz, rx_len);
+			while (rxcount-- > 0) {
+				data = in_8(&fifo->rxdata_8);
+				if (rx_buf)
+					*rx_buf++ = data;
+				rx_len--;
+			}
+
+			/*
+			 * come back later if there still is TX data to send,
+			 * bail out of the RX drain loop if all of the TX data
+			 * was sent and all of the RX data was received (i.e.
+			 * when the transmission has completed)
+			 */
+			if (tx_len)
+				break;
+			if (!rx_len)
+				break;
+
+			/*
+			 * TX data transmission has completed while RX data
+			 * is still pending -- that's a transient situation
+			 * which depends on wire speed and specific
+			 * hardware implementation details (buffering) yet
+			 * should resolve very quickly
+			 *
+			 * just yield for a moment to not hog the CPU for
+			 * too long when running SPI at low speed
+			 *
+			 * the timeout range is rather arbitrary and tries
+			 * to balance throughput against system load; the
+			 * chosen values result in a minimal timeout of 50
+			 * times 10us and thus work at speeds as low as
+			 * some 20kbps, while the maximum timeout at the
+			 * transfer's end could be 5ms _if_ nothing else
+			 * ticks in the system _and_ RX data still wasn't
+			 * received, which only occurs in situations that
+			 * are exceptional; removing the unpredictability
+			 * of the timeout either decreases throughput
+			 * (longer timeouts), or puts more load on the
+			 * system (fixed short timeouts) or requires the
+			 * use of a timeout API instead of a counter and an
+			 * unknown inner delay
+			 */
+			usleep_range(10, 100);
+
+		} while (--rxtries > 0);
+		if (!tx_len && rx_len && !rxtries) {
+			/*
+			 * not enough RX bytes even after several retries
+			 * and the resulting rather long timeout?
+			 */
+			rxcount = in_be32(&fifo->rxcnt);
+			dev_warn(&spi->dev,
+				 "short xfer, missing %zd RX bytes, FIFO level %zd\n",
+				 rx_len, rxcount);
+		}
+
+		/*
+		 * drain and drop RX data which "should not be there" in
+		 * the first place, for undisturbed transmission this turns
+		 * into a NOP (except for the FIFO level fetch)
+		 */
+		if (!tx_len && !rx_len) {
+			while (in_be32(&fifo->rxcnt))
+				in_8(&fifo->rxdata_8);
+		}
+
+	}
+	return 0;
+}
+
+static int mpc512x_psc_spi_msg_xfer(struct spi_master *master,
+				    struct spi_message *m)
+{
+	struct spi_device *spi;
+	unsigned cs_change;
+	int status;
+	struct spi_transfer *t;
+
+	spi = m->spi;
+	cs_change = 1;
+	status = 0;
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		status = mpc512x_psc_spi_transfer_setup(spi, t);
+		if (status < 0)
+			break;
+
+		if (cs_change)
+			mpc512x_psc_spi_activate_cs(spi);
+		cs_change = t->cs_change;
+
+		status = mpc512x_psc_spi_transfer_rxtx(spi, t);
+		if (status)
+			break;
+		m->actual_length += t->len;
+
+		if (t->delay_usecs)
+			udelay(t->delay_usecs);
+
+		if (cs_change)
+			mpc512x_psc_spi_deactivate_cs(spi);
+	}
+
+	m->status = status;
+	if (m->complete)
+		m->complete(m->context);
+
+	if (status || !cs_change)
+		mpc512x_psc_spi_deactivate_cs(spi);
+
+	mpc512x_psc_spi_transfer_setup(spi, NULL);
+
+	spi_finalize_current_message(master);
+	return status;
+}
+
+static int mpc512x_psc_spi_prep_xfer_hw(struct spi_master *master)
+{
+	struct mpc512x_psc_spi *mps = spi_master_get_devdata(master);
+
+	dev_dbg(&master->dev, "%s()\n", __func__);
+
+	/* Zero MR2 */
+	in_8(psc_addr(mps, mr2));
+	out_8(psc_addr(mps, mr2), 0x0);
+
+	/* enable transmitter/receiver */
+	out_8(psc_addr(mps, command), MPC52xx_PSC_TX_ENABLE | MPC52xx_PSC_RX_ENABLE);
+
+	return 0;
+}
+
+static int mpc512x_psc_spi_unprep_xfer_hw(struct spi_master *master)
+{
+	struct mpc512x_psc_spi *mps = spi_master_get_devdata(master);
+	struct mpc512x_psc_fifo __iomem *fifo = mps->fifo;
+
+	dev_dbg(&master->dev, "%s()\n", __func__);
+
+	/* disable transmitter/receiver and fifo interrupt */
+	out_8(psc_addr(mps, command), MPC52xx_PSC_TX_DISABLE | MPC52xx_PSC_RX_DISABLE);
+	out_be32(&fifo->tximr, 0);
+
+	return 0;
+}
+
+static int mpc512x_psc_spi_setup(struct spi_device *spi)
+{
+	struct mpc512x_psc_spi_cs *cs = spi->controller_state;
+	int ret;
+
+	if (spi->bits_per_word % 8)
+		return -EINVAL;
+
+	if (!cs) {
+		cs = kzalloc(sizeof *cs, GFP_KERNEL);
+		if (!cs)
+			return -ENOMEM;
+
+		if (gpio_is_valid(spi->cs_gpio)) {
+			ret = gpio_request(spi->cs_gpio, dev_name(&spi->dev));
+			if (ret) {
+				dev_err(&spi->dev, "can't get CS gpio: %d\n",
+					ret);
+				kfree(cs);
+				return ret;
+			}
+			gpio_direction_output(spi->cs_gpio,
+					spi->mode & SPI_CS_HIGH ? 0 : 1);
+		}
+
+		spi->controller_state = cs;
+	}
+
+	cs->bits_per_word = spi->bits_per_word;
+	cs->speed_hz = spi->max_speed_hz;
+
+	return 0;
+}
+
+static void mpc512x_psc_spi_cleanup(struct spi_device *spi)
+{
+	if (gpio_is_valid(spi->cs_gpio))
+		gpio_free(spi->cs_gpio);
+	kfree(spi->controller_state);
+}
+
+static int mpc512x_psc_spi_port_config(struct spi_master *master,
+				       struct mpc512x_psc_spi *mps)
+{
+	struct mpc512x_psc_fifo __iomem *fifo = mps->fifo;
+	u32 sicr;
+	u32 ccr;
+	int speed;
+	u16 bclkdiv;
+
+	/* Reset the PSC into a known state */
+	out_8(psc_addr(mps, command), MPC52xx_PSC_RST_RX);
+	out_8(psc_addr(mps, command), MPC52xx_PSC_RST_TX);
+	out_8(psc_addr(mps, command), MPC52xx_PSC_TX_DISABLE | MPC52xx_PSC_RX_DISABLE);
+
+	/* Disable psc interrupts all useful interrupts are in fifo */
+	out_be16(psc_addr(mps, isr_imr.imr), 0);
+
+	/* Disable fifo interrupts, will be enabled later */
+	out_be32(&fifo->tximr, 0);
+	out_be32(&fifo->rximr, 0);
+
+	/* Setup fifo slice address and size */
+	/*out_be32(&fifo->txsz, 0x0fe00004);*/
+	/*out_be32(&fifo->rxsz, 0x0ff00004);*/
+
+	sicr =	0x01000000 |	/* SIM = 0001 -- 8 bit */
+		0x00800000 |	/* GenClk = 1 -- internal clk */
+		0x00008000 |	/* SPI = 1 */
+		0x00004000 |	/* MSTR = 1   -- SPI master */
+		0x00000800;	/* UseEOF = 1 -- SS low until EOF */
+
+	out_be32(psc_addr(mps, sicr), sicr);
+
+	ccr = in_be32(psc_addr(mps, ccr));
+	ccr &= 0xFF000000;
+	speed = 1000000;	/* default 1MHz */
+	bclkdiv = (mps->mclk_rate / speed) - 1;
+	ccr |= (((bclkdiv & 0xff) << 16) | (((bclkdiv >> 8) & 0xff) << 8));
+	out_be32(psc_addr(mps, ccr), ccr);
+
+	/* Set 2ms DTL delay */
+	out_8(psc_addr(mps, ctur), 0x00);
+	out_8(psc_addr(mps, ctlr), 0x82);
+
+	/* we don't use the alarms */
+	out_be32(&fifo->rxalarm, 0xfff);
+	out_be32(&fifo->txalarm, 0);
+
+	/* Enable FIFO slices for Rx/Tx */
+	out_be32(&fifo->rxcmd,
+		 MPC512x_PSC_FIFO_ENABLE_SLICE | MPC512x_PSC_FIFO_ENABLE_DMA);
+	out_be32(&fifo->txcmd,
+		 MPC512x_PSC_FIFO_ENABLE_SLICE | MPC512x_PSC_FIFO_ENABLE_DMA);
+
+	mps->bits_per_word = 8;
+
+	return 0;
+}
+
+static irqreturn_t mpc512x_psc_spi_isr(int irq, void *dev_id)
+{
+	struct mpc512x_psc_spi *mps = (struct mpc512x_psc_spi *)dev_id;
+	struct mpc512x_psc_fifo __iomem *fifo = mps->fifo;
+
+	/* clear interrupt and wake up the rx/tx routine */
+	if (in_be32(&fifo->txisr) &
+	    in_be32(&fifo->tximr) & MPC512x_PSC_FIFO_EMPTY) {
+		out_be32(&fifo->txisr, MPC512x_PSC_FIFO_EMPTY);
+		out_be32(&fifo->tximr, 0);
+		complete(&mps->txisrdone);
+		return IRQ_HANDLED;
+	}
+	return IRQ_NONE;
+}
+
+static void mpc512x_spi_cs_control(struct spi_device *spi, bool onoff)
+{
+	gpio_set_value(spi->cs_gpio, onoff);
+}
+
+static int mpc512x_psc_spi_do_probe(struct device *dev, u32 regaddr,
+					      u32 size, unsigned int irq)
+{
+	struct fsl_spi_platform_data *pdata = dev_get_platdata(dev);
+	struct mpc512x_psc_spi *mps;
+	struct spi_master *master;
+	int ret;
+	void *tempp;
+	struct clk *clk;
+
+	master = spi_alloc_master(dev, sizeof *mps);
+	if (master == NULL)
+		return -ENOMEM;
+
+	dev_set_drvdata(dev, master);
+	mps = spi_master_get_devdata(master);
+	mps->type = (int)of_device_get_match_data(dev);
+	mps->irq = irq;
+
+	if (pdata == NULL) {
+		mps->cs_control = mpc512x_spi_cs_control;
+	} else {
+		mps->cs_control = pdata->cs_control;
+		master->bus_num = pdata->bus_num;
+		master->num_chipselect = pdata->max_chipselect;
+	}
+
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LSB_FIRST;
+	master->setup = mpc512x_psc_spi_setup;
+	master->prepare_transfer_hardware = mpc512x_psc_spi_prep_xfer_hw;
+	master->transfer_one_message = mpc512x_psc_spi_msg_xfer;
+	master->unprepare_transfer_hardware = mpc512x_psc_spi_unprep_xfer_hw;
+	master->cleanup = mpc512x_psc_spi_cleanup;
+	master->dev.of_node = dev->of_node;
+
+	tempp = devm_ioremap(dev, regaddr, size);
+	if (!tempp) {
+		dev_err(dev, "could not ioremap I/O port range\n");
+		ret = -EFAULT;
+		goto free_master;
+	}
+	mps->psc = tempp;
+	mps->fifo =
+		(struct mpc512x_psc_fifo *)(tempp + sizeof(struct mpc52xx_psc));
+	ret = devm_request_irq(dev, mps->irq, mpc512x_psc_spi_isr, IRQF_SHARED,
+				"mpc512x-psc-spi", mps);
+	if (ret)
+		goto free_master;
+	init_completion(&mps->txisrdone);
+
+	clk = devm_clk_get(dev, "mclk");
+	if (IS_ERR(clk)) {
+		ret = PTR_ERR(clk);
+		goto free_master;
+	}
+	ret = clk_prepare_enable(clk);
+	if (ret)
+		goto free_master;
+	mps->clk_mclk = clk;
+	mps->mclk_rate = clk_get_rate(clk);
+
+	clk = devm_clk_get(dev, "ipg");
+	if (IS_ERR(clk)) {
+		ret = PTR_ERR(clk);
+		goto free_mclk_clock;
+	}
+	ret = clk_prepare_enable(clk);
+	if (ret)
+		goto free_mclk_clock;
+	mps->clk_ipg = clk;
+
+	ret = mpc512x_psc_spi_port_config(master, mps);
+	if (ret < 0)
+		goto free_ipg_clock;
+
+	ret = devm_spi_register_master(dev, master);
+	if (ret < 0)
+		goto free_ipg_clock;
+
+	return ret;
+
+free_ipg_clock:
+	clk_disable_unprepare(mps->clk_ipg);
+free_mclk_clock:
+	clk_disable_unprepare(mps->clk_mclk);
+free_master:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int mpc512x_psc_spi_do_remove(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mpc512x_psc_spi *mps = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(mps->clk_mclk);
+	clk_disable_unprepare(mps->clk_ipg);
+
+	return 0;
+}
+
+static int mpc512x_psc_spi_of_probe(struct platform_device *op)
+{
+	const u32 *regaddr_p;
+	u64 regaddr64, size64;
+
+	regaddr_p = of_get_address(op->dev.of_node, 0, &size64, NULL);
+	if (!regaddr_p) {
+		dev_err(&op->dev, "Invalid PSC address\n");
+		return -EINVAL;
+	}
+	regaddr64 = of_translate_address(op->dev.of_node, regaddr_p);
+
+	return mpc512x_psc_spi_do_probe(&op->dev, (u32) regaddr64, (u32) size64,
+				irq_of_parse_and_map(op->dev.of_node, 0));
+}
+
+static int mpc512x_psc_spi_of_remove(struct platform_device *op)
+{
+	return mpc512x_psc_spi_do_remove(&op->dev);
+}
+
+static const struct of_device_id mpc512x_psc_spi_of_match[] = {
+	{ .compatible = "fsl,mpc5121-psc-spi", .data = (void *)TYPE_MPC5121 },
+	{ .compatible = "fsl,mpc5125-psc-spi", .data = (void *)TYPE_MPC5125 },
+	{},
+};
+
+MODULE_DEVICE_TABLE(of, mpc512x_psc_spi_of_match);
+
+static struct platform_driver mpc512x_psc_spi_of_driver = {
+	.probe = mpc512x_psc_spi_of_probe,
+	.remove = mpc512x_psc_spi_of_remove,
+	.driver = {
+		.name = "mpc512x-psc-spi",
+		.of_match_table = mpc512x_psc_spi_of_match,
+	},
+};
+module_platform_driver(mpc512x_psc_spi_of_driver);
+
+MODULE_AUTHOR("John Rigby");
+MODULE_DESCRIPTION("MPC512x PSC SPI Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-mpc52xx-psc.c b/drivers/spi/spi-mpc52xx-psc.c
new file mode 100644
index 0000000..42a8b85
--- /dev/null
+++ b/drivers/spi/spi-mpc52xx-psc.c
@@ -0,0 +1,507 @@
+/*
+ * MPC52xx PSC in SPI mode driver.
+ *
+ * Maintainer: Dragos Carp
+ *
+ * Copyright (C) 2006 TOPTICA Photonics AG.
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/of_address.h>
+#include <linux/of_platform.h>
+#include <linux/workqueue.h>
+#include <linux/completion.h>
+#include <linux/io.h>
+#include <linux/delay.h>
+#include <linux/spi/spi.h>
+#include <linux/fsl_devices.h>
+#include <linux/slab.h>
+
+#include <asm/mpc52xx.h>
+#include <asm/mpc52xx_psc.h>
+
+#define MCLK 20000000 /* PSC port MClk in hz */
+
+struct mpc52xx_psc_spi {
+	/* fsl_spi_platform data */
+	void (*cs_control)(struct spi_device *spi, bool on);
+	u32 sysclk;
+
+	/* driver internal data */
+	struct mpc52xx_psc __iomem *psc;
+	struct mpc52xx_psc_fifo __iomem *fifo;
+	unsigned int irq;
+	u8 bits_per_word;
+	u8 busy;
+
+	struct work_struct work;
+
+	struct list_head queue;
+	spinlock_t lock;
+
+	struct completion done;
+};
+
+/* controller state */
+struct mpc52xx_psc_spi_cs {
+	int bits_per_word;
+	int speed_hz;
+};
+
+/* set clock freq, clock ramp, bits per work
+ * if t is NULL then reset the values to the default values
+ */
+static int mpc52xx_psc_spi_transfer_setup(struct spi_device *spi,
+		struct spi_transfer *t)
+{
+	struct mpc52xx_psc_spi_cs *cs = spi->controller_state;
+
+	cs->speed_hz = (t && t->speed_hz)
+			? t->speed_hz : spi->max_speed_hz;
+	cs->bits_per_word = (t && t->bits_per_word)
+			? t->bits_per_word : spi->bits_per_word;
+	cs->bits_per_word = ((cs->bits_per_word + 7) / 8) * 8;
+	return 0;
+}
+
+static void mpc52xx_psc_spi_activate_cs(struct spi_device *spi)
+{
+	struct mpc52xx_psc_spi_cs *cs = spi->controller_state;
+	struct mpc52xx_psc_spi *mps = spi_master_get_devdata(spi->master);
+	struct mpc52xx_psc __iomem *psc = mps->psc;
+	u32 sicr;
+	u16 ccr;
+
+	sicr = in_be32(&psc->sicr);
+
+	/* Set clock phase and polarity */
+	if (spi->mode & SPI_CPHA)
+		sicr |= 0x00001000;
+	else
+		sicr &= ~0x00001000;
+	if (spi->mode & SPI_CPOL)
+		sicr |= 0x00002000;
+	else
+		sicr &= ~0x00002000;
+
+	if (spi->mode & SPI_LSB_FIRST)
+		sicr |= 0x10000000;
+	else
+		sicr &= ~0x10000000;
+	out_be32(&psc->sicr, sicr);
+
+	/* Set clock frequency and bits per word
+	 * Because psc->ccr is defined as 16bit register instead of 32bit
+	 * just set the lower byte of BitClkDiv
+	 */
+	ccr = in_be16((u16 __iomem *)&psc->ccr);
+	ccr &= 0xFF00;
+	if (cs->speed_hz)
+		ccr |= (MCLK / cs->speed_hz - 1) & 0xFF;
+	else /* by default SPI Clk 1MHz */
+		ccr |= (MCLK / 1000000 - 1) & 0xFF;
+	out_be16((u16 __iomem *)&psc->ccr, ccr);
+	mps->bits_per_word = cs->bits_per_word;
+
+	if (mps->cs_control)
+		mps->cs_control(spi, (spi->mode & SPI_CS_HIGH) ? 1 : 0);
+}
+
+static void mpc52xx_psc_spi_deactivate_cs(struct spi_device *spi)
+{
+	struct mpc52xx_psc_spi *mps = spi_master_get_devdata(spi->master);
+
+	if (mps->cs_control)
+		mps->cs_control(spi, (spi->mode & SPI_CS_HIGH) ? 0 : 1);
+}
+
+#define MPC52xx_PSC_BUFSIZE (MPC52xx_PSC_RFNUM_MASK + 1)
+/* wake up when 80% fifo full */
+#define MPC52xx_PSC_RFALARM (MPC52xx_PSC_BUFSIZE * 20 / 100)
+
+static int mpc52xx_psc_spi_transfer_rxtx(struct spi_device *spi,
+						struct spi_transfer *t)
+{
+	struct mpc52xx_psc_spi *mps = spi_master_get_devdata(spi->master);
+	struct mpc52xx_psc __iomem *psc = mps->psc;
+	struct mpc52xx_psc_fifo __iomem *fifo = mps->fifo;
+	unsigned rb = 0;	/* number of bytes receieved */
+	unsigned sb = 0;	/* number of bytes sent */
+	unsigned char *rx_buf = (unsigned char *)t->rx_buf;
+	unsigned char *tx_buf = (unsigned char *)t->tx_buf;
+	unsigned rfalarm;
+	unsigned send_at_once = MPC52xx_PSC_BUFSIZE;
+	unsigned recv_at_once;
+	int last_block = 0;
+
+	if (!t->tx_buf && !t->rx_buf && t->len)
+		return -EINVAL;
+
+	/* enable transmiter/receiver */
+	out_8(&psc->command, MPC52xx_PSC_TX_ENABLE | MPC52xx_PSC_RX_ENABLE);
+	while (rb < t->len) {
+		if (t->len - rb > MPC52xx_PSC_BUFSIZE) {
+			rfalarm = MPC52xx_PSC_RFALARM;
+			last_block = 0;
+		} else {
+			send_at_once = t->len - sb;
+			rfalarm = MPC52xx_PSC_BUFSIZE - (t->len - rb);
+			last_block = 1;
+		}
+
+		dev_dbg(&spi->dev, "send %d bytes...\n", send_at_once);
+		for (; send_at_once; sb++, send_at_once--) {
+			/* set EOF flag before the last word is sent */
+			if (send_at_once == 1 && last_block)
+				out_8(&psc->ircr2, 0x01);
+
+			if (tx_buf)
+				out_8(&psc->mpc52xx_psc_buffer_8, tx_buf[sb]);
+			else
+				out_8(&psc->mpc52xx_psc_buffer_8, 0);
+		}
+
+
+		/* enable interrupts and wait for wake up
+		 * if just one byte is expected the Rx FIFO genererates no
+		 * FFULL interrupt, so activate the RxRDY interrupt
+		 */
+		out_8(&psc->command, MPC52xx_PSC_SEL_MODE_REG_1);
+		if (t->len - rb == 1) {
+			out_8(&psc->mode, 0);
+		} else {
+			out_8(&psc->mode, MPC52xx_PSC_MODE_FFULL);
+			out_be16(&fifo->rfalarm, rfalarm);
+		}
+		out_be16(&psc->mpc52xx_psc_imr, MPC52xx_PSC_IMR_RXRDY);
+		wait_for_completion(&mps->done);
+		recv_at_once = in_be16(&fifo->rfnum);
+		dev_dbg(&spi->dev, "%d bytes received\n", recv_at_once);
+
+		send_at_once = recv_at_once;
+		if (rx_buf) {
+			for (; recv_at_once; rb++, recv_at_once--)
+				rx_buf[rb] = in_8(&psc->mpc52xx_psc_buffer_8);
+		} else {
+			for (; recv_at_once; rb++, recv_at_once--)
+				in_8(&psc->mpc52xx_psc_buffer_8);
+		}
+	}
+	/* disable transmiter/receiver */
+	out_8(&psc->command, MPC52xx_PSC_TX_DISABLE | MPC52xx_PSC_RX_DISABLE);
+
+	return 0;
+}
+
+static void mpc52xx_psc_spi_work(struct work_struct *work)
+{
+	struct mpc52xx_psc_spi *mps =
+		container_of(work, struct mpc52xx_psc_spi, work);
+
+	spin_lock_irq(&mps->lock);
+	mps->busy = 1;
+	while (!list_empty(&mps->queue)) {
+		struct spi_message *m;
+		struct spi_device *spi;
+		struct spi_transfer *t = NULL;
+		unsigned cs_change;
+		int status;
+
+		m = container_of(mps->queue.next, struct spi_message, queue);
+		list_del_init(&m->queue);
+		spin_unlock_irq(&mps->lock);
+
+		spi = m->spi;
+		cs_change = 1;
+		status = 0;
+		list_for_each_entry (t, &m->transfers, transfer_list) {
+			if (t->bits_per_word || t->speed_hz) {
+				status = mpc52xx_psc_spi_transfer_setup(spi, t);
+				if (status < 0)
+					break;
+			}
+
+			if (cs_change)
+				mpc52xx_psc_spi_activate_cs(spi);
+			cs_change = t->cs_change;
+
+			status = mpc52xx_psc_spi_transfer_rxtx(spi, t);
+			if (status)
+				break;
+			m->actual_length += t->len;
+
+			if (t->delay_usecs)
+				udelay(t->delay_usecs);
+
+			if (cs_change)
+				mpc52xx_psc_spi_deactivate_cs(spi);
+		}
+
+		m->status = status;
+		if (m->complete)
+			m->complete(m->context);
+
+		if (status || !cs_change)
+			mpc52xx_psc_spi_deactivate_cs(spi);
+
+		mpc52xx_psc_spi_transfer_setup(spi, NULL);
+
+		spin_lock_irq(&mps->lock);
+	}
+	mps->busy = 0;
+	spin_unlock_irq(&mps->lock);
+}
+
+static int mpc52xx_psc_spi_setup(struct spi_device *spi)
+{
+	struct mpc52xx_psc_spi *mps = spi_master_get_devdata(spi->master);
+	struct mpc52xx_psc_spi_cs *cs = spi->controller_state;
+	unsigned long flags;
+
+	if (spi->bits_per_word%8)
+		return -EINVAL;
+
+	if (!cs) {
+		cs = kzalloc(sizeof *cs, GFP_KERNEL);
+		if (!cs)
+			return -ENOMEM;
+		spi->controller_state = cs;
+	}
+
+	cs->bits_per_word = spi->bits_per_word;
+	cs->speed_hz = spi->max_speed_hz;
+
+	spin_lock_irqsave(&mps->lock, flags);
+	if (!mps->busy)
+		mpc52xx_psc_spi_deactivate_cs(spi);
+	spin_unlock_irqrestore(&mps->lock, flags);
+
+	return 0;
+}
+
+static int mpc52xx_psc_spi_transfer(struct spi_device *spi,
+		struct spi_message *m)
+{
+	struct mpc52xx_psc_spi *mps = spi_master_get_devdata(spi->master);
+	unsigned long flags;
+
+	m->actual_length = 0;
+	m->status = -EINPROGRESS;
+
+	spin_lock_irqsave(&mps->lock, flags);
+	list_add_tail(&m->queue, &mps->queue);
+	schedule_work(&mps->work);
+	spin_unlock_irqrestore(&mps->lock, flags);
+
+	return 0;
+}
+
+static void mpc52xx_psc_spi_cleanup(struct spi_device *spi)
+{
+	kfree(spi->controller_state);
+}
+
+static int mpc52xx_psc_spi_port_config(int psc_id, struct mpc52xx_psc_spi *mps)
+{
+	struct mpc52xx_psc __iomem *psc = mps->psc;
+	struct mpc52xx_psc_fifo __iomem *fifo = mps->fifo;
+	u32 mclken_div;
+	int ret;
+
+	/* default sysclk is 512MHz */
+	mclken_div = (mps->sysclk ? mps->sysclk : 512000000) / MCLK;
+	ret = mpc52xx_set_psc_clkdiv(psc_id, mclken_div);
+	if (ret)
+		return ret;
+
+	/* Reset the PSC into a known state */
+	out_8(&psc->command, MPC52xx_PSC_RST_RX);
+	out_8(&psc->command, MPC52xx_PSC_RST_TX);
+	out_8(&psc->command, MPC52xx_PSC_TX_DISABLE | MPC52xx_PSC_RX_DISABLE);
+
+	/* Disable interrupts, interrupts are based on alarm level */
+	out_be16(&psc->mpc52xx_psc_imr, 0);
+	out_8(&psc->command, MPC52xx_PSC_SEL_MODE_REG_1);
+	out_8(&fifo->rfcntl, 0);
+	out_8(&psc->mode, MPC52xx_PSC_MODE_FFULL);
+
+	/* Configure 8bit codec mode as a SPI master and use EOF flags */
+	/* SICR_SIM_CODEC8|SICR_GENCLK|SICR_SPI|SICR_MSTR|SICR_USEEOF */
+	out_be32(&psc->sicr, 0x0180C800);
+	out_be16((u16 __iomem *)&psc->ccr, 0x070F); /* default SPI Clk 1MHz */
+
+	/* Set 2ms DTL delay */
+	out_8(&psc->ctur, 0x00);
+	out_8(&psc->ctlr, 0x84);
+
+	mps->bits_per_word = 8;
+
+	return 0;
+}
+
+static irqreturn_t mpc52xx_psc_spi_isr(int irq, void *dev_id)
+{
+	struct mpc52xx_psc_spi *mps = (struct mpc52xx_psc_spi *)dev_id;
+	struct mpc52xx_psc __iomem *psc = mps->psc;
+
+	/* disable interrupt and wake up the work queue */
+	if (in_be16(&psc->mpc52xx_psc_isr) & MPC52xx_PSC_IMR_RXRDY) {
+		out_be16(&psc->mpc52xx_psc_imr, 0);
+		complete(&mps->done);
+		return IRQ_HANDLED;
+	}
+	return IRQ_NONE;
+}
+
+/* bus_num is used only for the case dev->platform_data == NULL */
+static int mpc52xx_psc_spi_do_probe(struct device *dev, u32 regaddr,
+				u32 size, unsigned int irq, s16 bus_num)
+{
+	struct fsl_spi_platform_data *pdata = dev_get_platdata(dev);
+	struct mpc52xx_psc_spi *mps;
+	struct spi_master *master;
+	int ret;
+
+	master = spi_alloc_master(dev, sizeof *mps);
+	if (master == NULL)
+		return -ENOMEM;
+
+	dev_set_drvdata(dev, master);
+	mps = spi_master_get_devdata(master);
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LSB_FIRST;
+
+	mps->irq = irq;
+	if (pdata == NULL) {
+		dev_warn(dev,
+			 "probe called without platform data, no cs_control function will be called\n");
+		mps->cs_control = NULL;
+		mps->sysclk = 0;
+		master->bus_num = bus_num;
+		master->num_chipselect = 255;
+	} else {
+		mps->cs_control = pdata->cs_control;
+		mps->sysclk = pdata->sysclk;
+		master->bus_num = pdata->bus_num;
+		master->num_chipselect = pdata->max_chipselect;
+	}
+	master->setup = mpc52xx_psc_spi_setup;
+	master->transfer = mpc52xx_psc_spi_transfer;
+	master->cleanup = mpc52xx_psc_spi_cleanup;
+	master->dev.of_node = dev->of_node;
+
+	mps->psc = ioremap(regaddr, size);
+	if (!mps->psc) {
+		dev_err(dev, "could not ioremap I/O port range\n");
+		ret = -EFAULT;
+		goto free_master;
+	}
+	/* On the 5200, fifo regs are immediately ajacent to the psc regs */
+	mps->fifo = ((void __iomem *)mps->psc) + sizeof(struct mpc52xx_psc);
+
+	ret = request_irq(mps->irq, mpc52xx_psc_spi_isr, 0, "mpc52xx-psc-spi",
+				mps);
+	if (ret)
+		goto free_master;
+
+	ret = mpc52xx_psc_spi_port_config(master->bus_num, mps);
+	if (ret < 0) {
+		dev_err(dev, "can't configure PSC! Is it capable of SPI?\n");
+		goto free_irq;
+	}
+
+	spin_lock_init(&mps->lock);
+	init_completion(&mps->done);
+	INIT_WORK(&mps->work, mpc52xx_psc_spi_work);
+	INIT_LIST_HEAD(&mps->queue);
+
+	ret = spi_register_master(master);
+	if (ret < 0)
+		goto free_irq;
+
+	return ret;
+
+free_irq:
+	free_irq(mps->irq, mps);
+free_master:
+	if (mps->psc)
+		iounmap(mps->psc);
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int mpc52xx_psc_spi_of_probe(struct platform_device *op)
+{
+	const u32 *regaddr_p;
+	u64 regaddr64, size64;
+	s16 id = -1;
+
+	regaddr_p = of_get_address(op->dev.of_node, 0, &size64, NULL);
+	if (!regaddr_p) {
+		dev_err(&op->dev, "Invalid PSC address\n");
+		return -EINVAL;
+	}
+	regaddr64 = of_translate_address(op->dev.of_node, regaddr_p);
+
+	/* get PSC id (1..6, used by port_config) */
+	if (op->dev.platform_data == NULL) {
+		const u32 *psc_nump;
+
+		psc_nump = of_get_property(op->dev.of_node, "cell-index", NULL);
+		if (!psc_nump || *psc_nump > 5) {
+			dev_err(&op->dev, "Invalid cell-index property\n");
+			return -EINVAL;
+		}
+		id = *psc_nump + 1;
+	}
+
+	return mpc52xx_psc_spi_do_probe(&op->dev, (u32)regaddr64, (u32)size64,
+				irq_of_parse_and_map(op->dev.of_node, 0), id);
+}
+
+static int mpc52xx_psc_spi_of_remove(struct platform_device *op)
+{
+	struct spi_master *master = spi_master_get(platform_get_drvdata(op));
+	struct mpc52xx_psc_spi *mps = spi_master_get_devdata(master);
+
+	flush_work(&mps->work);
+	spi_unregister_master(master);
+	free_irq(mps->irq, mps);
+	if (mps->psc)
+		iounmap(mps->psc);
+	spi_master_put(master);
+
+	return 0;
+}
+
+static const struct of_device_id mpc52xx_psc_spi_of_match[] = {
+	{ .compatible = "fsl,mpc5200-psc-spi", },
+	{ .compatible = "mpc5200-psc-spi", }, /* old */
+	{}
+};
+
+MODULE_DEVICE_TABLE(of, mpc52xx_psc_spi_of_match);
+
+static struct platform_driver mpc52xx_psc_spi_of_driver = {
+	.probe = mpc52xx_psc_spi_of_probe,
+	.remove = mpc52xx_psc_spi_of_remove,
+	.driver = {
+		.name = "mpc52xx-psc-spi",
+		.of_match_table = mpc52xx_psc_spi_of_match,
+	},
+};
+module_platform_driver(mpc52xx_psc_spi_of_driver);
+
+MODULE_AUTHOR("Dragos Carp");
+MODULE_DESCRIPTION("MPC52xx PSC SPI Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-mpc52xx.c b/drivers/spi/spi-mpc52xx.c
new file mode 100644
index 0000000..0e55784
--- /dev/null
+++ b/drivers/spi/spi-mpc52xx.c
@@ -0,0 +1,551 @@
+/*
+ * MPC52xx SPI bus driver.
+ *
+ * Copyright (C) 2008 Secret Lab Technologies Ltd.
+ *
+ * This file is released under the GPLv2
+ *
+ * This is the driver for the MPC5200's dedicated SPI controller.
+ *
+ * Note: this driver does not support the MPC5200 PSC in SPI mode.  For
+ * that driver see drivers/spi/mpc52xx_psc_spi.c
+ */
+
+#include <linux/module.h>
+#include <linux/errno.h>
+#include <linux/of_platform.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/spi/spi.h>
+#include <linux/io.h>
+#include <linux/of_gpio.h>
+#include <linux/slab.h>
+#include <asm/time.h>
+#include <asm/mpc52xx.h>
+
+MODULE_AUTHOR("Grant Likely <grant.likely@secretlab.ca>");
+MODULE_DESCRIPTION("MPC52xx SPI (non-PSC) Driver");
+MODULE_LICENSE("GPL");
+
+/* Register offsets */
+#define SPI_CTRL1	0x00
+#define SPI_CTRL1_SPIE		(1 << 7)
+#define SPI_CTRL1_SPE		(1 << 6)
+#define SPI_CTRL1_MSTR		(1 << 4)
+#define SPI_CTRL1_CPOL		(1 << 3)
+#define SPI_CTRL1_CPHA		(1 << 2)
+#define SPI_CTRL1_SSOE		(1 << 1)
+#define SPI_CTRL1_LSBFE		(1 << 0)
+
+#define SPI_CTRL2	0x01
+#define SPI_BRR		0x04
+
+#define SPI_STATUS	0x05
+#define SPI_STATUS_SPIF		(1 << 7)
+#define SPI_STATUS_WCOL		(1 << 6)
+#define SPI_STATUS_MODF		(1 << 4)
+
+#define SPI_DATA	0x09
+#define SPI_PORTDATA	0x0d
+#define SPI_DATADIR	0x10
+
+/* FSM state return values */
+#define FSM_STOP	0	/* Nothing more for the state machine to */
+				/* do.  If something interesting happens */
+				/* then an IRQ will be received */
+#define FSM_POLL	1	/* need to poll for completion, an IRQ is */
+				/* not expected */
+#define FSM_CONTINUE	2	/* Keep iterating the state machine */
+
+/* Driver internal data */
+struct mpc52xx_spi {
+	struct spi_master *master;
+	void __iomem *regs;
+	int irq0;	/* MODF irq */
+	int irq1;	/* SPIF irq */
+	unsigned int ipb_freq;
+
+	/* Statistics; not used now, but will be reintroduced for debugfs */
+	int msg_count;
+	int wcol_count;
+	int wcol_ticks;
+	u32 wcol_tx_timestamp;
+	int modf_count;
+	int byte_count;
+
+	struct list_head queue;		/* queue of pending messages */
+	spinlock_t lock;
+	struct work_struct work;
+
+	/* Details of current transfer (length, and buffer pointers) */
+	struct spi_message *message;	/* current message */
+	struct spi_transfer *transfer;	/* current transfer */
+	int (*state)(int irq, struct mpc52xx_spi *ms, u8 status, u8 data);
+	int len;
+	int timestamp;
+	u8 *rx_buf;
+	const u8 *tx_buf;
+	int cs_change;
+	int gpio_cs_count;
+	unsigned int *gpio_cs;
+};
+
+/*
+ * CS control function
+ */
+static void mpc52xx_spi_chipsel(struct mpc52xx_spi *ms, int value)
+{
+	int cs;
+
+	if (ms->gpio_cs_count > 0) {
+		cs = ms->message->spi->chip_select;
+		gpio_set_value(ms->gpio_cs[cs], value ? 0 : 1);
+	} else
+		out_8(ms->regs + SPI_PORTDATA, value ? 0 : 0x08);
+}
+
+/*
+ * Start a new transfer.  This is called both by the idle state
+ * for the first transfer in a message, and by the wait state when the
+ * previous transfer in a message is complete.
+ */
+static void mpc52xx_spi_start_transfer(struct mpc52xx_spi *ms)
+{
+	ms->rx_buf = ms->transfer->rx_buf;
+	ms->tx_buf = ms->transfer->tx_buf;
+	ms->len = ms->transfer->len;
+
+	/* Activate the chip select */
+	if (ms->cs_change)
+		mpc52xx_spi_chipsel(ms, 1);
+	ms->cs_change = ms->transfer->cs_change;
+
+	/* Write out the first byte */
+	ms->wcol_tx_timestamp = get_tbl();
+	if (ms->tx_buf)
+		out_8(ms->regs + SPI_DATA, *ms->tx_buf++);
+	else
+		out_8(ms->regs + SPI_DATA, 0);
+}
+
+/* Forward declaration of state handlers */
+static int mpc52xx_spi_fsmstate_transfer(int irq, struct mpc52xx_spi *ms,
+					 u8 status, u8 data);
+static int mpc52xx_spi_fsmstate_wait(int irq, struct mpc52xx_spi *ms,
+				     u8 status, u8 data);
+
+/*
+ * IDLE state
+ *
+ * No transfers are in progress; if another transfer is pending then retrieve
+ * it and kick it off.  Otherwise, stop processing the state machine
+ */
+static int
+mpc52xx_spi_fsmstate_idle(int irq, struct mpc52xx_spi *ms, u8 status, u8 data)
+{
+	struct spi_device *spi;
+	int spr, sppr;
+	u8 ctrl1;
+
+	if (status && (irq != NO_IRQ))
+		dev_err(&ms->master->dev, "spurious irq, status=0x%.2x\n",
+			status);
+
+	/* Check if there is another transfer waiting. */
+	if (list_empty(&ms->queue))
+		return FSM_STOP;
+
+	/* get the head of the queue */
+	ms->message = list_first_entry(&ms->queue, struct spi_message, queue);
+	list_del_init(&ms->message->queue);
+
+	/* Setup the controller parameters */
+	ctrl1 = SPI_CTRL1_SPIE | SPI_CTRL1_SPE | SPI_CTRL1_MSTR;
+	spi = ms->message->spi;
+	if (spi->mode & SPI_CPHA)
+		ctrl1 |= SPI_CTRL1_CPHA;
+	if (spi->mode & SPI_CPOL)
+		ctrl1 |= SPI_CTRL1_CPOL;
+	if (spi->mode & SPI_LSB_FIRST)
+		ctrl1 |= SPI_CTRL1_LSBFE;
+	out_8(ms->regs + SPI_CTRL1, ctrl1);
+
+	/* Setup the controller speed */
+	/* minimum divider is '2'.  Also, add '1' to force rounding the
+	 * divider up. */
+	sppr = ((ms->ipb_freq / ms->message->spi->max_speed_hz) + 1) >> 1;
+	spr = 0;
+	if (sppr < 1)
+		sppr = 1;
+	while (((sppr - 1) & ~0x7) != 0) {
+		sppr = (sppr + 1) >> 1; /* add '1' to force rounding up */
+		spr++;
+	}
+	sppr--;		/* sppr quantity in register is offset by 1 */
+	if (spr > 7) {
+		/* Don't overrun limits of SPI baudrate register */
+		spr = 7;
+		sppr = 7;
+	}
+	out_8(ms->regs + SPI_BRR, sppr << 4 | spr); /* Set speed */
+
+	ms->cs_change = 1;
+	ms->transfer = container_of(ms->message->transfers.next,
+				    struct spi_transfer, transfer_list);
+
+	mpc52xx_spi_start_transfer(ms);
+	ms->state = mpc52xx_spi_fsmstate_transfer;
+
+	return FSM_CONTINUE;
+}
+
+/*
+ * TRANSFER state
+ *
+ * In the middle of a transfer.  If the SPI core has completed processing
+ * a byte, then read out the received data and write out the next byte
+ * (unless this transfer is finished; in which case go on to the wait
+ * state)
+ */
+static int mpc52xx_spi_fsmstate_transfer(int irq, struct mpc52xx_spi *ms,
+					 u8 status, u8 data)
+{
+	if (!status)
+		return ms->irq0 ? FSM_STOP : FSM_POLL;
+
+	if (status & SPI_STATUS_WCOL) {
+		/* The SPI controller is stoopid.  At slower speeds, it may
+		 * raise the SPIF flag before the state machine is actually
+		 * finished, which causes a collision (internal to the state
+		 * machine only).  The manual recommends inserting a delay
+		 * between receiving the interrupt and sending the next byte,
+		 * but it can also be worked around simply by retrying the
+		 * transfer which is what we do here. */
+		ms->wcol_count++;
+		ms->wcol_ticks += get_tbl() - ms->wcol_tx_timestamp;
+		ms->wcol_tx_timestamp = get_tbl();
+		data = 0;
+		if (ms->tx_buf)
+			data = *(ms->tx_buf - 1);
+		out_8(ms->regs + SPI_DATA, data); /* try again */
+		return FSM_CONTINUE;
+	} else if (status & SPI_STATUS_MODF) {
+		ms->modf_count++;
+		dev_err(&ms->master->dev, "mode fault\n");
+		mpc52xx_spi_chipsel(ms, 0);
+		ms->message->status = -EIO;
+		if (ms->message->complete)
+			ms->message->complete(ms->message->context);
+		ms->state = mpc52xx_spi_fsmstate_idle;
+		return FSM_CONTINUE;
+	}
+
+	/* Read data out of the spi device */
+	ms->byte_count++;
+	if (ms->rx_buf)
+		*ms->rx_buf++ = data;
+
+	/* Is the transfer complete? */
+	ms->len--;
+	if (ms->len == 0) {
+		ms->timestamp = get_tbl();
+		ms->timestamp += ms->transfer->delay_usecs * tb_ticks_per_usec;
+		ms->state = mpc52xx_spi_fsmstate_wait;
+		return FSM_CONTINUE;
+	}
+
+	/* Write out the next byte */
+	ms->wcol_tx_timestamp = get_tbl();
+	if (ms->tx_buf)
+		out_8(ms->regs + SPI_DATA, *ms->tx_buf++);
+	else
+		out_8(ms->regs + SPI_DATA, 0);
+
+	return FSM_CONTINUE;
+}
+
+/*
+ * WAIT state
+ *
+ * A transfer has completed; need to wait for the delay period to complete
+ * before starting the next transfer
+ */
+static int
+mpc52xx_spi_fsmstate_wait(int irq, struct mpc52xx_spi *ms, u8 status, u8 data)
+{
+	if (status && irq)
+		dev_err(&ms->master->dev, "spurious irq, status=0x%.2x\n",
+			status);
+
+	if (((int)get_tbl()) - ms->timestamp < 0)
+		return FSM_POLL;
+
+	ms->message->actual_length += ms->transfer->len;
+
+	/* Check if there is another transfer in this message.  If there
+	 * aren't then deactivate CS, notify sender, and drop back to idle
+	 * to start the next message. */
+	if (ms->transfer->transfer_list.next == &ms->message->transfers) {
+		ms->msg_count++;
+		mpc52xx_spi_chipsel(ms, 0);
+		ms->message->status = 0;
+		if (ms->message->complete)
+			ms->message->complete(ms->message->context);
+		ms->state = mpc52xx_spi_fsmstate_idle;
+		return FSM_CONTINUE;
+	}
+
+	/* There is another transfer; kick it off */
+
+	if (ms->cs_change)
+		mpc52xx_spi_chipsel(ms, 0);
+
+	ms->transfer = container_of(ms->transfer->transfer_list.next,
+				    struct spi_transfer, transfer_list);
+	mpc52xx_spi_start_transfer(ms);
+	ms->state = mpc52xx_spi_fsmstate_transfer;
+	return FSM_CONTINUE;
+}
+
+/**
+ * mpc52xx_spi_fsm_process - Finite State Machine iteration function
+ * @irq: irq number that triggered the FSM or 0 for polling
+ * @ms: pointer to mpc52xx_spi driver data
+ */
+static void mpc52xx_spi_fsm_process(int irq, struct mpc52xx_spi *ms)
+{
+	int rc = FSM_CONTINUE;
+	u8 status, data;
+
+	while (rc == FSM_CONTINUE) {
+		/* Interrupt cleared by read of STATUS followed by
+		 * read of DATA registers */
+		status = in_8(ms->regs + SPI_STATUS);
+		data = in_8(ms->regs + SPI_DATA);
+		rc = ms->state(irq, ms, status, data);
+	}
+
+	if (rc == FSM_POLL)
+		schedule_work(&ms->work);
+}
+
+/**
+ * mpc52xx_spi_irq - IRQ handler
+ */
+static irqreturn_t mpc52xx_spi_irq(int irq, void *_ms)
+{
+	struct mpc52xx_spi *ms = _ms;
+	spin_lock(&ms->lock);
+	mpc52xx_spi_fsm_process(irq, ms);
+	spin_unlock(&ms->lock);
+	return IRQ_HANDLED;
+}
+
+/**
+ * mpc52xx_spi_wq - Workqueue function for polling the state machine
+ */
+static void mpc52xx_spi_wq(struct work_struct *work)
+{
+	struct mpc52xx_spi *ms = container_of(work, struct mpc52xx_spi, work);
+	unsigned long flags;
+
+	spin_lock_irqsave(&ms->lock, flags);
+	mpc52xx_spi_fsm_process(0, ms);
+	spin_unlock_irqrestore(&ms->lock, flags);
+}
+
+/*
+ * spi_master ops
+ */
+
+static int mpc52xx_spi_transfer(struct spi_device *spi, struct spi_message *m)
+{
+	struct mpc52xx_spi *ms = spi_master_get_devdata(spi->master);
+	unsigned long flags;
+
+	m->actual_length = 0;
+	m->status = -EINPROGRESS;
+
+	spin_lock_irqsave(&ms->lock, flags);
+	list_add_tail(&m->queue, &ms->queue);
+	spin_unlock_irqrestore(&ms->lock, flags);
+	schedule_work(&ms->work);
+
+	return 0;
+}
+
+/*
+ * OF Platform Bus Binding
+ */
+static int mpc52xx_spi_probe(struct platform_device *op)
+{
+	struct spi_master *master;
+	struct mpc52xx_spi *ms;
+	void __iomem *regs;
+	u8 ctrl1;
+	int rc, i = 0;
+	int gpio_cs;
+
+	/* MMIO registers */
+	dev_dbg(&op->dev, "probing mpc5200 SPI device\n");
+	regs = of_iomap(op->dev.of_node, 0);
+	if (!regs)
+		return -ENODEV;
+
+	/* initialize the device */
+	ctrl1 = SPI_CTRL1_SPIE | SPI_CTRL1_SPE | SPI_CTRL1_MSTR;
+	out_8(regs + SPI_CTRL1, ctrl1);
+	out_8(regs + SPI_CTRL2, 0x0);
+	out_8(regs + SPI_DATADIR, 0xe);	/* Set output pins */
+	out_8(regs + SPI_PORTDATA, 0x8);	/* Deassert /SS signal */
+
+	/* Clear the status register and re-read it to check for a MODF
+	 * failure.  This driver cannot currently handle multiple masters
+	 * on the SPI bus.  This fault will also occur if the SPI signals
+	 * are not connected to any pins (port_config setting) */
+	in_8(regs + SPI_STATUS);
+	out_8(regs + SPI_CTRL1, ctrl1);
+
+	in_8(regs + SPI_DATA);
+	if (in_8(regs + SPI_STATUS) & SPI_STATUS_MODF) {
+		dev_err(&op->dev, "mode fault; is port_config correct?\n");
+		rc = -EIO;
+		goto err_init;
+	}
+
+	dev_dbg(&op->dev, "allocating spi_master struct\n");
+	master = spi_alloc_master(&op->dev, sizeof *ms);
+	if (!master) {
+		rc = -ENOMEM;
+		goto err_alloc;
+	}
+
+	master->transfer = mpc52xx_spi_transfer;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_LSB_FIRST;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->dev.of_node = op->dev.of_node;
+
+	platform_set_drvdata(op, master);
+
+	ms = spi_master_get_devdata(master);
+	ms->master = master;
+	ms->regs = regs;
+	ms->irq0 = irq_of_parse_and_map(op->dev.of_node, 0);
+	ms->irq1 = irq_of_parse_and_map(op->dev.of_node, 1);
+	ms->state = mpc52xx_spi_fsmstate_idle;
+	ms->ipb_freq = mpc5xxx_get_bus_frequency(op->dev.of_node);
+	ms->gpio_cs_count = of_gpio_count(op->dev.of_node);
+	if (ms->gpio_cs_count > 0) {
+		master->num_chipselect = ms->gpio_cs_count;
+		ms->gpio_cs = kmalloc_array(ms->gpio_cs_count,
+					    sizeof(*ms->gpio_cs),
+					    GFP_KERNEL);
+		if (!ms->gpio_cs) {
+			rc = -ENOMEM;
+			goto err_alloc_gpio;
+		}
+
+		for (i = 0; i < ms->gpio_cs_count; i++) {
+			gpio_cs = of_get_gpio(op->dev.of_node, i);
+			if (!gpio_is_valid(gpio_cs)) {
+				dev_err(&op->dev,
+					"could not parse the gpio field in oftree\n");
+				rc = -ENODEV;
+				goto err_gpio;
+			}
+
+			rc = gpio_request(gpio_cs, dev_name(&op->dev));
+			if (rc) {
+				dev_err(&op->dev,
+					"can't request spi cs gpio #%d on gpio line %d\n",
+					i, gpio_cs);
+				goto err_gpio;
+			}
+
+			gpio_direction_output(gpio_cs, 1);
+			ms->gpio_cs[i] = gpio_cs;
+		}
+	}
+
+	spin_lock_init(&ms->lock);
+	INIT_LIST_HEAD(&ms->queue);
+	INIT_WORK(&ms->work, mpc52xx_spi_wq);
+
+	/* Decide if interrupts can be used */
+	if (ms->irq0 && ms->irq1) {
+		rc = request_irq(ms->irq0, mpc52xx_spi_irq, 0,
+				  "mpc5200-spi-modf", ms);
+		rc |= request_irq(ms->irq1, mpc52xx_spi_irq, 0,
+				  "mpc5200-spi-spif", ms);
+		if (rc) {
+			free_irq(ms->irq0, ms);
+			free_irq(ms->irq1, ms);
+			ms->irq0 = ms->irq1 = 0;
+		}
+	} else {
+		/* operate in polled mode */
+		ms->irq0 = ms->irq1 = 0;
+	}
+
+	if (!ms->irq0)
+		dev_info(&op->dev, "using polled mode\n");
+
+	dev_dbg(&op->dev, "registering spi_master struct\n");
+	rc = spi_register_master(master);
+	if (rc)
+		goto err_register;
+
+	dev_info(&ms->master->dev, "registered MPC5200 SPI bus\n");
+
+	return rc;
+
+ err_register:
+	dev_err(&ms->master->dev, "initialization failed\n");
+ err_gpio:
+	while (i-- > 0)
+		gpio_free(ms->gpio_cs[i]);
+
+	kfree(ms->gpio_cs);
+ err_alloc_gpio:
+	spi_master_put(master);
+ err_alloc:
+ err_init:
+	iounmap(regs);
+	return rc;
+}
+
+static int mpc52xx_spi_remove(struct platform_device *op)
+{
+	struct spi_master *master = spi_master_get(platform_get_drvdata(op));
+	struct mpc52xx_spi *ms = spi_master_get_devdata(master);
+	int i;
+
+	free_irq(ms->irq0, ms);
+	free_irq(ms->irq1, ms);
+
+	for (i = 0; i < ms->gpio_cs_count; i++)
+		gpio_free(ms->gpio_cs[i]);
+
+	kfree(ms->gpio_cs);
+	spi_unregister_master(master);
+	iounmap(ms->regs);
+	spi_master_put(master);
+
+	return 0;
+}
+
+static const struct of_device_id mpc52xx_spi_match[] = {
+	{ .compatible = "fsl,mpc5200-spi", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, mpc52xx_spi_match);
+
+static struct platform_driver mpc52xx_spi_of_driver = {
+	.driver = {
+		.name = "mpc52xx-spi",
+		.of_match_table = mpc52xx_spi_match,
+	},
+	.probe = mpc52xx_spi_probe,
+	.remove = mpc52xx_spi_remove,
+};
+module_platform_driver(mpc52xx_spi_of_driver);
diff --git a/drivers/spi/spi-mt65xx.c b/drivers/spi/spi-mt65xx.c
new file mode 100644
index 0000000..86bf456
--- /dev/null
+++ b/drivers/spi/spi-mt65xx.c
@@ -0,0 +1,859 @@
+/*
+ * Copyright (c) 2015 MediaTek Inc.
+ * Author: Leilk Liu <leilk.liu@mediatek.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/clk.h>
+#include <linux/device.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/ioport.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+#include <linux/platform_device.h>
+#include <linux/platform_data/spi-mt65xx.h>
+#include <linux/pm_runtime.h>
+#include <linux/spi/spi.h>
+
+#define SPI_CFG0_REG                      0x0000
+#define SPI_CFG1_REG                      0x0004
+#define SPI_TX_SRC_REG                    0x0008
+#define SPI_RX_DST_REG                    0x000c
+#define SPI_TX_DATA_REG                   0x0010
+#define SPI_RX_DATA_REG                   0x0014
+#define SPI_CMD_REG                       0x0018
+#define SPI_STATUS0_REG                   0x001c
+#define SPI_PAD_SEL_REG                   0x0024
+#define SPI_CFG2_REG                      0x0028
+
+#define SPI_CFG0_SCK_HIGH_OFFSET          0
+#define SPI_CFG0_SCK_LOW_OFFSET           8
+#define SPI_CFG0_CS_HOLD_OFFSET           16
+#define SPI_CFG0_CS_SETUP_OFFSET          24
+#define SPI_ADJUST_CFG0_SCK_LOW_OFFSET    16
+#define SPI_ADJUST_CFG0_CS_HOLD_OFFSET    0
+#define SPI_ADJUST_CFG0_CS_SETUP_OFFSET   16
+
+#define SPI_CFG1_CS_IDLE_OFFSET           0
+#define SPI_CFG1_PACKET_LOOP_OFFSET       8
+#define SPI_CFG1_PACKET_LENGTH_OFFSET     16
+#define SPI_CFG1_GET_TICK_DLY_OFFSET      30
+
+#define SPI_CFG1_CS_IDLE_MASK             0xff
+#define SPI_CFG1_PACKET_LOOP_MASK         0xff00
+#define SPI_CFG1_PACKET_LENGTH_MASK       0x3ff0000
+
+#define SPI_CMD_ACT                  BIT(0)
+#define SPI_CMD_RESUME               BIT(1)
+#define SPI_CMD_RST                  BIT(2)
+#define SPI_CMD_PAUSE_EN             BIT(4)
+#define SPI_CMD_DEASSERT             BIT(5)
+#define SPI_CMD_SAMPLE_SEL           BIT(6)
+#define SPI_CMD_CS_POL               BIT(7)
+#define SPI_CMD_CPHA                 BIT(8)
+#define SPI_CMD_CPOL                 BIT(9)
+#define SPI_CMD_RX_DMA               BIT(10)
+#define SPI_CMD_TX_DMA               BIT(11)
+#define SPI_CMD_TXMSBF               BIT(12)
+#define SPI_CMD_RXMSBF               BIT(13)
+#define SPI_CMD_RX_ENDIAN            BIT(14)
+#define SPI_CMD_TX_ENDIAN            BIT(15)
+#define SPI_CMD_FINISH_IE            BIT(16)
+#define SPI_CMD_PAUSE_IE             BIT(17)
+
+#define MT8173_SPI_MAX_PAD_SEL 3
+
+#define MTK_SPI_PAUSE_INT_STATUS 0x2
+
+#define MTK_SPI_IDLE 0
+#define MTK_SPI_PAUSED 1
+
+#define MTK_SPI_MAX_FIFO_SIZE 32U
+#define MTK_SPI_PACKET_SIZE 1024
+
+struct mtk_spi_compatible {
+	bool need_pad_sel;
+	/* Must explicitly send dummy Tx bytes to do Rx only transfer */
+	bool must_tx;
+	/* some IC design adjust cfg register to enhance time accuracy */
+	bool enhance_timing;
+};
+
+struct mtk_spi {
+	void __iomem *base;
+	u32 state;
+	int pad_num;
+	u32 *pad_sel;
+	struct clk *parent_clk, *sel_clk, *spi_clk;
+	struct spi_transfer *cur_transfer;
+	u32 xfer_len;
+	struct scatterlist *tx_sgl, *rx_sgl;
+	u32 tx_sgl_len, rx_sgl_len;
+	const struct mtk_spi_compatible *dev_comp;
+};
+
+static const struct mtk_spi_compatible mtk_common_compat;
+
+static const struct mtk_spi_compatible mt2712_compat = {
+	.must_tx = true,
+};
+
+static const struct mtk_spi_compatible mt7622_compat = {
+	.must_tx = true,
+	.enhance_timing = true,
+};
+
+static const struct mtk_spi_compatible mt8173_compat = {
+	.need_pad_sel = true,
+	.must_tx = true,
+};
+
+/*
+ * A piece of default chip info unless the platform
+ * supplies it.
+ */
+static const struct mtk_chip_config mtk_default_chip_info = {
+	.rx_mlsb = 1,
+	.tx_mlsb = 1,
+	.cs_pol = 0,
+	.sample_sel = 0,
+};
+
+static const struct of_device_id mtk_spi_of_match[] = {
+	{ .compatible = "mediatek,mt2701-spi",
+		.data = (void *)&mtk_common_compat,
+	},
+	{ .compatible = "mediatek,mt2712-spi",
+		.data = (void *)&mt2712_compat,
+	},
+	{ .compatible = "mediatek,mt6589-spi",
+		.data = (void *)&mtk_common_compat,
+	},
+	{ .compatible = "mediatek,mt7622-spi",
+		.data = (void *)&mt7622_compat,
+	},
+	{ .compatible = "mediatek,mt8135-spi",
+		.data = (void *)&mtk_common_compat,
+	},
+	{ .compatible = "mediatek,mt8173-spi",
+		.data = (void *)&mt8173_compat,
+	},
+	{}
+};
+MODULE_DEVICE_TABLE(of, mtk_spi_of_match);
+
+static void mtk_spi_reset(struct mtk_spi *mdata)
+{
+	u32 reg_val;
+
+	/* set the software reset bit in SPI_CMD_REG. */
+	reg_val = readl(mdata->base + SPI_CMD_REG);
+	reg_val |= SPI_CMD_RST;
+	writel(reg_val, mdata->base + SPI_CMD_REG);
+
+	reg_val = readl(mdata->base + SPI_CMD_REG);
+	reg_val &= ~SPI_CMD_RST;
+	writel(reg_val, mdata->base + SPI_CMD_REG);
+}
+
+static int mtk_spi_prepare_message(struct spi_master *master,
+				   struct spi_message *msg)
+{
+	u16 cpha, cpol;
+	u32 reg_val;
+	struct spi_device *spi = msg->spi;
+	struct mtk_chip_config *chip_config = spi->controller_data;
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	cpha = spi->mode & SPI_CPHA ? 1 : 0;
+	cpol = spi->mode & SPI_CPOL ? 1 : 0;
+
+	reg_val = readl(mdata->base + SPI_CMD_REG);
+	if (cpha)
+		reg_val |= SPI_CMD_CPHA;
+	else
+		reg_val &= ~SPI_CMD_CPHA;
+	if (cpol)
+		reg_val |= SPI_CMD_CPOL;
+	else
+		reg_val &= ~SPI_CMD_CPOL;
+
+	/* set the mlsbx and mlsbtx */
+	if (chip_config->tx_mlsb)
+		reg_val |= SPI_CMD_TXMSBF;
+	else
+		reg_val &= ~SPI_CMD_TXMSBF;
+	if (chip_config->rx_mlsb)
+		reg_val |= SPI_CMD_RXMSBF;
+	else
+		reg_val &= ~SPI_CMD_RXMSBF;
+
+	/* set the tx/rx endian */
+#ifdef __LITTLE_ENDIAN
+	reg_val &= ~SPI_CMD_TX_ENDIAN;
+	reg_val &= ~SPI_CMD_RX_ENDIAN;
+#else
+	reg_val |= SPI_CMD_TX_ENDIAN;
+	reg_val |= SPI_CMD_RX_ENDIAN;
+#endif
+
+	if (mdata->dev_comp->enhance_timing) {
+		if (chip_config->cs_pol)
+			reg_val |= SPI_CMD_CS_POL;
+		else
+			reg_val &= ~SPI_CMD_CS_POL;
+		if (chip_config->sample_sel)
+			reg_val |= SPI_CMD_SAMPLE_SEL;
+		else
+			reg_val &= ~SPI_CMD_SAMPLE_SEL;
+	}
+
+	/* set finish and pause interrupt always enable */
+	reg_val |= SPI_CMD_FINISH_IE | SPI_CMD_PAUSE_IE;
+
+	/* disable dma mode */
+	reg_val &= ~(SPI_CMD_TX_DMA | SPI_CMD_RX_DMA);
+
+	/* disable deassert mode */
+	reg_val &= ~SPI_CMD_DEASSERT;
+
+	writel(reg_val, mdata->base + SPI_CMD_REG);
+
+	/* pad select */
+	if (mdata->dev_comp->need_pad_sel)
+		writel(mdata->pad_sel[spi->chip_select],
+		       mdata->base + SPI_PAD_SEL_REG);
+
+	return 0;
+}
+
+static void mtk_spi_set_cs(struct spi_device *spi, bool enable)
+{
+	u32 reg_val;
+	struct mtk_spi *mdata = spi_master_get_devdata(spi->master);
+
+	reg_val = readl(mdata->base + SPI_CMD_REG);
+	if (!enable) {
+		reg_val |= SPI_CMD_PAUSE_EN;
+		writel(reg_val, mdata->base + SPI_CMD_REG);
+	} else {
+		reg_val &= ~SPI_CMD_PAUSE_EN;
+		writel(reg_val, mdata->base + SPI_CMD_REG);
+		mdata->state = MTK_SPI_IDLE;
+		mtk_spi_reset(mdata);
+	}
+}
+
+static void mtk_spi_prepare_transfer(struct spi_master *master,
+				     struct spi_transfer *xfer)
+{
+	u32 spi_clk_hz, div, sck_time, cs_time, reg_val = 0;
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	spi_clk_hz = clk_get_rate(mdata->spi_clk);
+	if (xfer->speed_hz < spi_clk_hz / 2)
+		div = DIV_ROUND_UP(spi_clk_hz, xfer->speed_hz);
+	else
+		div = 1;
+
+	sck_time = (div + 1) / 2;
+	cs_time = sck_time * 2;
+
+	if (mdata->dev_comp->enhance_timing) {
+		reg_val |= (((sck_time - 1) & 0xffff)
+			   << SPI_CFG0_SCK_HIGH_OFFSET);
+		reg_val |= (((sck_time - 1) & 0xffff)
+			   << SPI_ADJUST_CFG0_SCK_LOW_OFFSET);
+		writel(reg_val, mdata->base + SPI_CFG2_REG);
+		reg_val |= (((cs_time - 1) & 0xffff)
+			   << SPI_ADJUST_CFG0_CS_HOLD_OFFSET);
+		reg_val |= (((cs_time - 1) & 0xffff)
+			   << SPI_ADJUST_CFG0_CS_SETUP_OFFSET);
+		writel(reg_val, mdata->base + SPI_CFG0_REG);
+	} else {
+		reg_val |= (((sck_time - 1) & 0xff)
+			   << SPI_CFG0_SCK_HIGH_OFFSET);
+		reg_val |= (((sck_time - 1) & 0xff) << SPI_CFG0_SCK_LOW_OFFSET);
+		reg_val |= (((cs_time - 1) & 0xff) << SPI_CFG0_CS_HOLD_OFFSET);
+		reg_val |= (((cs_time - 1) & 0xff) << SPI_CFG0_CS_SETUP_OFFSET);
+		writel(reg_val, mdata->base + SPI_CFG0_REG);
+	}
+
+	reg_val = readl(mdata->base + SPI_CFG1_REG);
+	reg_val &= ~SPI_CFG1_CS_IDLE_MASK;
+	reg_val |= (((cs_time - 1) & 0xff) << SPI_CFG1_CS_IDLE_OFFSET);
+	writel(reg_val, mdata->base + SPI_CFG1_REG);
+}
+
+static void mtk_spi_setup_packet(struct spi_master *master)
+{
+	u32 packet_size, packet_loop, reg_val;
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	packet_size = min_t(u32, mdata->xfer_len, MTK_SPI_PACKET_SIZE);
+	packet_loop = mdata->xfer_len / packet_size;
+
+	reg_val = readl(mdata->base + SPI_CFG1_REG);
+	reg_val &= ~(SPI_CFG1_PACKET_LENGTH_MASK | SPI_CFG1_PACKET_LOOP_MASK);
+	reg_val |= (packet_size - 1) << SPI_CFG1_PACKET_LENGTH_OFFSET;
+	reg_val |= (packet_loop - 1) << SPI_CFG1_PACKET_LOOP_OFFSET;
+	writel(reg_val, mdata->base + SPI_CFG1_REG);
+}
+
+static void mtk_spi_enable_transfer(struct spi_master *master)
+{
+	u32 cmd;
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	cmd = readl(mdata->base + SPI_CMD_REG);
+	if (mdata->state == MTK_SPI_IDLE)
+		cmd |= SPI_CMD_ACT;
+	else
+		cmd |= SPI_CMD_RESUME;
+	writel(cmd, mdata->base + SPI_CMD_REG);
+}
+
+static int mtk_spi_get_mult_delta(u32 xfer_len)
+{
+	u32 mult_delta;
+
+	if (xfer_len > MTK_SPI_PACKET_SIZE)
+		mult_delta = xfer_len % MTK_SPI_PACKET_SIZE;
+	else
+		mult_delta = 0;
+
+	return mult_delta;
+}
+
+static void mtk_spi_update_mdata_len(struct spi_master *master)
+{
+	int mult_delta;
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	if (mdata->tx_sgl_len && mdata->rx_sgl_len) {
+		if (mdata->tx_sgl_len > mdata->rx_sgl_len) {
+			mult_delta = mtk_spi_get_mult_delta(mdata->rx_sgl_len);
+			mdata->xfer_len = mdata->rx_sgl_len - mult_delta;
+			mdata->rx_sgl_len = mult_delta;
+			mdata->tx_sgl_len -= mdata->xfer_len;
+		} else {
+			mult_delta = mtk_spi_get_mult_delta(mdata->tx_sgl_len);
+			mdata->xfer_len = mdata->tx_sgl_len - mult_delta;
+			mdata->tx_sgl_len = mult_delta;
+			mdata->rx_sgl_len -= mdata->xfer_len;
+		}
+	} else if (mdata->tx_sgl_len) {
+		mult_delta = mtk_spi_get_mult_delta(mdata->tx_sgl_len);
+		mdata->xfer_len = mdata->tx_sgl_len - mult_delta;
+		mdata->tx_sgl_len = mult_delta;
+	} else if (mdata->rx_sgl_len) {
+		mult_delta = mtk_spi_get_mult_delta(mdata->rx_sgl_len);
+		mdata->xfer_len = mdata->rx_sgl_len - mult_delta;
+		mdata->rx_sgl_len = mult_delta;
+	}
+}
+
+static void mtk_spi_setup_dma_addr(struct spi_master *master,
+				   struct spi_transfer *xfer)
+{
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	if (mdata->tx_sgl)
+		writel(xfer->tx_dma, mdata->base + SPI_TX_SRC_REG);
+	if (mdata->rx_sgl)
+		writel(xfer->rx_dma, mdata->base + SPI_RX_DST_REG);
+}
+
+static int mtk_spi_fifo_transfer(struct spi_master *master,
+				 struct spi_device *spi,
+				 struct spi_transfer *xfer)
+{
+	int cnt, remainder;
+	u32 reg_val;
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	mdata->cur_transfer = xfer;
+	mdata->xfer_len = min(MTK_SPI_MAX_FIFO_SIZE, xfer->len);
+	mtk_spi_prepare_transfer(master, xfer);
+	mtk_spi_setup_packet(master);
+
+	cnt = xfer->len / 4;
+	iowrite32_rep(mdata->base + SPI_TX_DATA_REG, xfer->tx_buf, cnt);
+
+	remainder = xfer->len % 4;
+	if (remainder > 0) {
+		reg_val = 0;
+		memcpy(&reg_val, xfer->tx_buf + (cnt * 4), remainder);
+		writel(reg_val, mdata->base + SPI_TX_DATA_REG);
+	}
+
+	mtk_spi_enable_transfer(master);
+
+	return 1;
+}
+
+static int mtk_spi_dma_transfer(struct spi_master *master,
+				struct spi_device *spi,
+				struct spi_transfer *xfer)
+{
+	int cmd;
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	mdata->tx_sgl = NULL;
+	mdata->rx_sgl = NULL;
+	mdata->tx_sgl_len = 0;
+	mdata->rx_sgl_len = 0;
+	mdata->cur_transfer = xfer;
+
+	mtk_spi_prepare_transfer(master, xfer);
+
+	cmd = readl(mdata->base + SPI_CMD_REG);
+	if (xfer->tx_buf)
+		cmd |= SPI_CMD_TX_DMA;
+	if (xfer->rx_buf)
+		cmd |= SPI_CMD_RX_DMA;
+	writel(cmd, mdata->base + SPI_CMD_REG);
+
+	if (xfer->tx_buf)
+		mdata->tx_sgl = xfer->tx_sg.sgl;
+	if (xfer->rx_buf)
+		mdata->rx_sgl = xfer->rx_sg.sgl;
+
+	if (mdata->tx_sgl) {
+		xfer->tx_dma = sg_dma_address(mdata->tx_sgl);
+		mdata->tx_sgl_len = sg_dma_len(mdata->tx_sgl);
+	}
+	if (mdata->rx_sgl) {
+		xfer->rx_dma = sg_dma_address(mdata->rx_sgl);
+		mdata->rx_sgl_len = sg_dma_len(mdata->rx_sgl);
+	}
+
+	mtk_spi_update_mdata_len(master);
+	mtk_spi_setup_packet(master);
+	mtk_spi_setup_dma_addr(master, xfer);
+	mtk_spi_enable_transfer(master);
+
+	return 1;
+}
+
+static int mtk_spi_transfer_one(struct spi_master *master,
+				struct spi_device *spi,
+				struct spi_transfer *xfer)
+{
+	if (master->can_dma(master, spi, xfer))
+		return mtk_spi_dma_transfer(master, spi, xfer);
+	else
+		return mtk_spi_fifo_transfer(master, spi, xfer);
+}
+
+static bool mtk_spi_can_dma(struct spi_master *master,
+			    struct spi_device *spi,
+			    struct spi_transfer *xfer)
+{
+	/* Buffers for DMA transactions must be 4-byte aligned */
+	return (xfer->len > MTK_SPI_MAX_FIFO_SIZE &&
+		(unsigned long)xfer->tx_buf % 4 == 0 &&
+		(unsigned long)xfer->rx_buf % 4 == 0);
+}
+
+static int mtk_spi_setup(struct spi_device *spi)
+{
+	struct mtk_spi *mdata = spi_master_get_devdata(spi->master);
+
+	if (!spi->controller_data)
+		spi->controller_data = (void *)&mtk_default_chip_info;
+
+	if (mdata->dev_comp->need_pad_sel && gpio_is_valid(spi->cs_gpio))
+		gpio_direction_output(spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
+
+	return 0;
+}
+
+static irqreturn_t mtk_spi_interrupt(int irq, void *dev_id)
+{
+	u32 cmd, reg_val, cnt, remainder;
+	struct spi_master *master = dev_id;
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+	struct spi_transfer *trans = mdata->cur_transfer;
+
+	reg_val = readl(mdata->base + SPI_STATUS0_REG);
+	if (reg_val & MTK_SPI_PAUSE_INT_STATUS)
+		mdata->state = MTK_SPI_PAUSED;
+	else
+		mdata->state = MTK_SPI_IDLE;
+
+	if (!master->can_dma(master, master->cur_msg->spi, trans)) {
+		if (trans->rx_buf) {
+			cnt = mdata->xfer_len / 4;
+			ioread32_rep(mdata->base + SPI_RX_DATA_REG,
+				     trans->rx_buf, cnt);
+			remainder = mdata->xfer_len % 4;
+			if (remainder > 0) {
+				reg_val = readl(mdata->base + SPI_RX_DATA_REG);
+				memcpy(trans->rx_buf + (cnt * 4),
+					&reg_val, remainder);
+			}
+		}
+
+		trans->len -= mdata->xfer_len;
+		if (!trans->len) {
+			spi_finalize_current_transfer(master);
+			return IRQ_HANDLED;
+		}
+
+		if (trans->tx_buf)
+			trans->tx_buf += mdata->xfer_len;
+		if (trans->rx_buf)
+			trans->rx_buf += mdata->xfer_len;
+
+		mdata->xfer_len = min(MTK_SPI_MAX_FIFO_SIZE, trans->len);
+		mtk_spi_setup_packet(master);
+
+		cnt = trans->len / 4;
+		iowrite32_rep(mdata->base + SPI_TX_DATA_REG, trans->tx_buf, cnt);
+
+		remainder = trans->len % 4;
+		if (remainder > 0) {
+			reg_val = 0;
+			memcpy(&reg_val, trans->tx_buf + (cnt * 4), remainder);
+			writel(reg_val, mdata->base + SPI_TX_DATA_REG);
+		}
+
+		mtk_spi_enable_transfer(master);
+
+		return IRQ_HANDLED;
+	}
+
+	if (mdata->tx_sgl)
+		trans->tx_dma += mdata->xfer_len;
+	if (mdata->rx_sgl)
+		trans->rx_dma += mdata->xfer_len;
+
+	if (mdata->tx_sgl && (mdata->tx_sgl_len == 0)) {
+		mdata->tx_sgl = sg_next(mdata->tx_sgl);
+		if (mdata->tx_sgl) {
+			trans->tx_dma = sg_dma_address(mdata->tx_sgl);
+			mdata->tx_sgl_len = sg_dma_len(mdata->tx_sgl);
+		}
+	}
+	if (mdata->rx_sgl && (mdata->rx_sgl_len == 0)) {
+		mdata->rx_sgl = sg_next(mdata->rx_sgl);
+		if (mdata->rx_sgl) {
+			trans->rx_dma = sg_dma_address(mdata->rx_sgl);
+			mdata->rx_sgl_len = sg_dma_len(mdata->rx_sgl);
+		}
+	}
+
+	if (!mdata->tx_sgl && !mdata->rx_sgl) {
+		/* spi disable dma */
+		cmd = readl(mdata->base + SPI_CMD_REG);
+		cmd &= ~SPI_CMD_TX_DMA;
+		cmd &= ~SPI_CMD_RX_DMA;
+		writel(cmd, mdata->base + SPI_CMD_REG);
+
+		spi_finalize_current_transfer(master);
+		return IRQ_HANDLED;
+	}
+
+	mtk_spi_update_mdata_len(master);
+	mtk_spi_setup_packet(master);
+	mtk_spi_setup_dma_addr(master, trans);
+	mtk_spi_enable_transfer(master);
+
+	return IRQ_HANDLED;
+}
+
+static int mtk_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct mtk_spi *mdata;
+	const struct of_device_id *of_id;
+	struct resource *res;
+	int i, irq, ret;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*mdata));
+	if (!master) {
+		dev_err(&pdev->dev, "failed to alloc spi master\n");
+		return -ENOMEM;
+	}
+
+	master->auto_runtime_pm = true;
+	master->dev.of_node = pdev->dev.of_node;
+	master->mode_bits = SPI_CPOL | SPI_CPHA;
+
+	master->set_cs = mtk_spi_set_cs;
+	master->prepare_message = mtk_spi_prepare_message;
+	master->transfer_one = mtk_spi_transfer_one;
+	master->can_dma = mtk_spi_can_dma;
+	master->setup = mtk_spi_setup;
+
+	of_id = of_match_node(mtk_spi_of_match, pdev->dev.of_node);
+	if (!of_id) {
+		dev_err(&pdev->dev, "failed to probe of_node\n");
+		ret = -EINVAL;
+		goto err_put_master;
+	}
+
+	mdata = spi_master_get_devdata(master);
+	mdata->dev_comp = of_id->data;
+	if (mdata->dev_comp->must_tx)
+		master->flags = SPI_MASTER_MUST_TX;
+
+	if (mdata->dev_comp->need_pad_sel) {
+		mdata->pad_num = of_property_count_u32_elems(
+			pdev->dev.of_node,
+			"mediatek,pad-select");
+		if (mdata->pad_num < 0) {
+			dev_err(&pdev->dev,
+				"No 'mediatek,pad-select' property\n");
+			ret = -EINVAL;
+			goto err_put_master;
+		}
+
+		mdata->pad_sel = devm_kmalloc_array(&pdev->dev, mdata->pad_num,
+						    sizeof(u32), GFP_KERNEL);
+		if (!mdata->pad_sel) {
+			ret = -ENOMEM;
+			goto err_put_master;
+		}
+
+		for (i = 0; i < mdata->pad_num; i++) {
+			of_property_read_u32_index(pdev->dev.of_node,
+						   "mediatek,pad-select",
+						   i, &mdata->pad_sel[i]);
+			if (mdata->pad_sel[i] > MT8173_SPI_MAX_PAD_SEL) {
+				dev_err(&pdev->dev, "wrong pad-sel[%d]: %u\n",
+					i, mdata->pad_sel[i]);
+				ret = -EINVAL;
+				goto err_put_master;
+			}
+		}
+	}
+
+	platform_set_drvdata(pdev, master);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res) {
+		ret = -ENODEV;
+		dev_err(&pdev->dev, "failed to determine base address\n");
+		goto err_put_master;
+	}
+
+	mdata->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(mdata->base)) {
+		ret = PTR_ERR(mdata->base);
+		goto err_put_master;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "failed to get irq (%d)\n", irq);
+		ret = irq;
+		goto err_put_master;
+	}
+
+	if (!pdev->dev.dma_mask)
+		pdev->dev.dma_mask = &pdev->dev.coherent_dma_mask;
+
+	ret = devm_request_irq(&pdev->dev, irq, mtk_spi_interrupt,
+			       IRQF_TRIGGER_NONE, dev_name(&pdev->dev), master);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to register irq (%d)\n", ret);
+		goto err_put_master;
+	}
+
+	mdata->parent_clk = devm_clk_get(&pdev->dev, "parent-clk");
+	if (IS_ERR(mdata->parent_clk)) {
+		ret = PTR_ERR(mdata->parent_clk);
+		dev_err(&pdev->dev, "failed to get parent-clk: %d\n", ret);
+		goto err_put_master;
+	}
+
+	mdata->sel_clk = devm_clk_get(&pdev->dev, "sel-clk");
+	if (IS_ERR(mdata->sel_clk)) {
+		ret = PTR_ERR(mdata->sel_clk);
+		dev_err(&pdev->dev, "failed to get sel-clk: %d\n", ret);
+		goto err_put_master;
+	}
+
+	mdata->spi_clk = devm_clk_get(&pdev->dev, "spi-clk");
+	if (IS_ERR(mdata->spi_clk)) {
+		ret = PTR_ERR(mdata->spi_clk);
+		dev_err(&pdev->dev, "failed to get spi-clk: %d\n", ret);
+		goto err_put_master;
+	}
+
+	ret = clk_prepare_enable(mdata->spi_clk);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to enable spi_clk (%d)\n", ret);
+		goto err_put_master;
+	}
+
+	ret = clk_set_parent(mdata->sel_clk, mdata->parent_clk);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to clk_set_parent (%d)\n", ret);
+		clk_disable_unprepare(mdata->spi_clk);
+		goto err_put_master;
+	}
+
+	clk_disable_unprepare(mdata->spi_clk);
+
+	pm_runtime_enable(&pdev->dev);
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to register master (%d)\n", ret);
+		goto err_disable_runtime_pm;
+	}
+
+	if (mdata->dev_comp->need_pad_sel) {
+		if (mdata->pad_num != master->num_chipselect) {
+			dev_err(&pdev->dev,
+				"pad_num does not match num_chipselect(%d != %d)\n",
+				mdata->pad_num, master->num_chipselect);
+			ret = -EINVAL;
+			goto err_disable_runtime_pm;
+		}
+
+		if (!master->cs_gpios && master->num_chipselect > 1) {
+			dev_err(&pdev->dev,
+				"cs_gpios not specified and num_chipselect > 1\n");
+			ret = -EINVAL;
+			goto err_disable_runtime_pm;
+		}
+
+		if (master->cs_gpios) {
+			for (i = 0; i < master->num_chipselect; i++) {
+				ret = devm_gpio_request(&pdev->dev,
+							master->cs_gpios[i],
+							dev_name(&pdev->dev));
+				if (ret) {
+					dev_err(&pdev->dev,
+						"can't get CS GPIO %i\n", i);
+					goto err_disable_runtime_pm;
+				}
+			}
+		}
+	}
+
+	return 0;
+
+err_disable_runtime_pm:
+	pm_runtime_disable(&pdev->dev);
+err_put_master:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int mtk_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	pm_runtime_disable(&pdev->dev);
+
+	mtk_spi_reset(mdata);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int mtk_spi_suspend(struct device *dev)
+{
+	int ret;
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	ret = spi_master_suspend(master);
+	if (ret)
+		return ret;
+
+	if (!pm_runtime_suspended(dev))
+		clk_disable_unprepare(mdata->spi_clk);
+
+	return ret;
+}
+
+static int mtk_spi_resume(struct device *dev)
+{
+	int ret;
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	if (!pm_runtime_suspended(dev)) {
+		ret = clk_prepare_enable(mdata->spi_clk);
+		if (ret < 0) {
+			dev_err(dev, "failed to enable spi_clk (%d)\n", ret);
+			return ret;
+		}
+	}
+
+	ret = spi_master_resume(master);
+	if (ret < 0)
+		clk_disable_unprepare(mdata->spi_clk);
+
+	return ret;
+}
+#endif /* CONFIG_PM_SLEEP */
+
+#ifdef CONFIG_PM
+static int mtk_spi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(mdata->spi_clk);
+
+	return 0;
+}
+
+static int mtk_spi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(mdata->spi_clk);
+	if (ret < 0) {
+		dev_err(dev, "failed to enable spi_clk (%d)\n", ret);
+		return ret;
+	}
+
+	return 0;
+}
+#endif /* CONFIG_PM */
+
+static const struct dev_pm_ops mtk_spi_pm = {
+	SET_SYSTEM_SLEEP_PM_OPS(mtk_spi_suspend, mtk_spi_resume)
+	SET_RUNTIME_PM_OPS(mtk_spi_runtime_suspend,
+			   mtk_spi_runtime_resume, NULL)
+};
+
+static struct platform_driver mtk_spi_driver = {
+	.driver = {
+		.name = "mtk-spi",
+		.pm	= &mtk_spi_pm,
+		.of_match_table = mtk_spi_of_match,
+	},
+	.probe = mtk_spi_probe,
+	.remove = mtk_spi_remove,
+};
+
+module_platform_driver(mtk_spi_driver);
+
+MODULE_DESCRIPTION("MTK SPI Controller driver");
+MODULE_AUTHOR("Leilk Liu <leilk.liu@mediatek.com>");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:mtk-spi");
diff --git a/drivers/spi/spi-mxs.c b/drivers/spi/spi-mxs.c
new file mode 100644
index 0000000..6ac95a2
--- /dev/null
+++ b/drivers/spi/spi-mxs.c
@@ -0,0 +1,673 @@
+// SPDX-License-Identifier: GPL-2.0+
+//
+// Freescale MXS SPI master driver
+//
+// Copyright 2012 DENX Software Engineering, GmbH.
+// Copyright 2012 Freescale Semiconductor, Inc.
+// Copyright 2008 Embedded Alley Solutions, Inc All Rights Reserved.
+//
+// Rework and transition to new API by:
+// Marek Vasut <marex@denx.de>
+//
+// Based on previous attempt by:
+// Fabio Estevam <fabio.estevam@freescale.com>
+//
+// Based on code from U-Boot bootloader by:
+// Marek Vasut <marex@denx.de>
+//
+// Based on spi-stmp.c, which is:
+// Author: Dmitry Pervushin <dimka@embeddedalley.com>
+
+#include <linux/kernel.h>
+#include <linux/ioport.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/of_gpio.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/highmem.h>
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/completion.h>
+#include <linux/gpio.h>
+#include <linux/regulator/consumer.h>
+#include <linux/pm_runtime.h>
+#include <linux/module.h>
+#include <linux/stmp_device.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/mxs-spi.h>
+
+#define DRIVER_NAME		"mxs-spi"
+
+/* Use 10S timeout for very long transfers, it should suffice. */
+#define SSP_TIMEOUT		10000
+
+#define SG_MAXLEN		0xff00
+
+/*
+ * Flags for txrx functions.  More efficient that using an argument register for
+ * each one.
+ */
+#define TXRX_WRITE		(1<<0)	/* This is a write */
+#define TXRX_DEASSERT_CS	(1<<1)	/* De-assert CS at end of txrx */
+
+struct mxs_spi {
+	struct mxs_ssp		ssp;
+	struct completion	c;
+	unsigned int		sck;	/* Rate requested (vs actual) */
+};
+
+static int mxs_spi_setup_transfer(struct spi_device *dev,
+				  const struct spi_transfer *t)
+{
+	struct mxs_spi *spi = spi_master_get_devdata(dev->master);
+	struct mxs_ssp *ssp = &spi->ssp;
+	const unsigned int hz = min(dev->max_speed_hz, t->speed_hz);
+
+	if (hz == 0) {
+		dev_err(&dev->dev, "SPI clock rate of zero not allowed\n");
+		return -EINVAL;
+	}
+
+	if (hz != spi->sck) {
+		mxs_ssp_set_clk_rate(ssp, hz);
+		/*
+		 * Save requested rate, hz, rather than the actual rate,
+		 * ssp->clk_rate.  Otherwise we would set the rate every transfer
+		 * when the actual rate is not quite the same as requested rate.
+		 */
+		spi->sck = hz;
+		/*
+		 * Perhaps we should return an error if the actual clock is
+		 * nowhere close to what was requested?
+		 */
+	}
+
+	writel(BM_SSP_CTRL0_LOCK_CS,
+		ssp->base + HW_SSP_CTRL0 + STMP_OFFSET_REG_SET);
+
+	writel(BF_SSP_CTRL1_SSP_MODE(BV_SSP_CTRL1_SSP_MODE__SPI) |
+	       BF_SSP_CTRL1_WORD_LENGTH(BV_SSP_CTRL1_WORD_LENGTH__EIGHT_BITS) |
+	       ((dev->mode & SPI_CPOL) ? BM_SSP_CTRL1_POLARITY : 0) |
+	       ((dev->mode & SPI_CPHA) ? BM_SSP_CTRL1_PHASE : 0),
+	       ssp->base + HW_SSP_CTRL1(ssp));
+
+	writel(0x0, ssp->base + HW_SSP_CMD0);
+	writel(0x0, ssp->base + HW_SSP_CMD1);
+
+	return 0;
+}
+
+static u32 mxs_spi_cs_to_reg(unsigned cs)
+{
+	u32 select = 0;
+
+	/*
+	 * i.MX28 Datasheet: 17.10.1: HW_SSP_CTRL0
+	 *
+	 * The bits BM_SSP_CTRL0_WAIT_FOR_CMD and BM_SSP_CTRL0_WAIT_FOR_IRQ
+	 * in HW_SSP_CTRL0 register do have multiple usage, please refer to
+	 * the datasheet for further details. In SPI mode, they are used to
+	 * toggle the chip-select lines (nCS pins).
+	 */
+	if (cs & 1)
+		select |= BM_SSP_CTRL0_WAIT_FOR_CMD;
+	if (cs & 2)
+		select |= BM_SSP_CTRL0_WAIT_FOR_IRQ;
+
+	return select;
+}
+
+static int mxs_ssp_wait(struct mxs_spi *spi, int offset, int mask, bool set)
+{
+	const unsigned long timeout = jiffies + msecs_to_jiffies(SSP_TIMEOUT);
+	struct mxs_ssp *ssp = &spi->ssp;
+	u32 reg;
+
+	do {
+		reg = readl_relaxed(ssp->base + offset);
+
+		if (!set)
+			reg = ~reg;
+
+		reg &= mask;
+
+		if (reg == mask)
+			return 0;
+	} while (time_before(jiffies, timeout));
+
+	return -ETIMEDOUT;
+}
+
+static void mxs_ssp_dma_irq_callback(void *param)
+{
+	struct mxs_spi *spi = param;
+
+	complete(&spi->c);
+}
+
+static irqreturn_t mxs_ssp_irq_handler(int irq, void *dev_id)
+{
+	struct mxs_ssp *ssp = dev_id;
+
+	dev_err(ssp->dev, "%s[%i] CTRL1=%08x STATUS=%08x\n",
+		__func__, __LINE__,
+		readl(ssp->base + HW_SSP_CTRL1(ssp)),
+		readl(ssp->base + HW_SSP_STATUS(ssp)));
+	return IRQ_HANDLED;
+}
+
+static int mxs_spi_txrx_dma(struct mxs_spi *spi,
+			    unsigned char *buf, int len,
+			    unsigned int flags)
+{
+	struct mxs_ssp *ssp = &spi->ssp;
+	struct dma_async_tx_descriptor *desc = NULL;
+	const bool vmalloced_buf = is_vmalloc_addr(buf);
+	const int desc_len = vmalloced_buf ? PAGE_SIZE : SG_MAXLEN;
+	const int sgs = DIV_ROUND_UP(len, desc_len);
+	int sg_count;
+	int min, ret;
+	u32 ctrl0;
+	struct page *vm_page;
+	struct {
+		u32			pio[4];
+		struct scatterlist	sg;
+	} *dma_xfer;
+
+	if (!len)
+		return -EINVAL;
+
+	dma_xfer = kcalloc(sgs, sizeof(*dma_xfer), GFP_KERNEL);
+	if (!dma_xfer)
+		return -ENOMEM;
+
+	reinit_completion(&spi->c);
+
+	/* Chip select was already programmed into CTRL0 */
+	ctrl0 = readl(ssp->base + HW_SSP_CTRL0);
+	ctrl0 &= ~(BM_SSP_CTRL0_XFER_COUNT | BM_SSP_CTRL0_IGNORE_CRC |
+		 BM_SSP_CTRL0_READ);
+	ctrl0 |= BM_SSP_CTRL0_DATA_XFER;
+
+	if (!(flags & TXRX_WRITE))
+		ctrl0 |= BM_SSP_CTRL0_READ;
+
+	/* Queue the DMA data transfer. */
+	for (sg_count = 0; sg_count < sgs; sg_count++) {
+		/* Prepare the transfer descriptor. */
+		min = min(len, desc_len);
+
+		/*
+		 * De-assert CS on last segment if flag is set (i.e., no more
+		 * transfers will follow)
+		 */
+		if ((sg_count + 1 == sgs) && (flags & TXRX_DEASSERT_CS))
+			ctrl0 |= BM_SSP_CTRL0_IGNORE_CRC;
+
+		if (ssp->devid == IMX23_SSP) {
+			ctrl0 &= ~BM_SSP_CTRL0_XFER_COUNT;
+			ctrl0 |= min;
+		}
+
+		dma_xfer[sg_count].pio[0] = ctrl0;
+		dma_xfer[sg_count].pio[3] = min;
+
+		if (vmalloced_buf) {
+			vm_page = vmalloc_to_page(buf);
+			if (!vm_page) {
+				ret = -ENOMEM;
+				goto err_vmalloc;
+			}
+
+			sg_init_table(&dma_xfer[sg_count].sg, 1);
+			sg_set_page(&dma_xfer[sg_count].sg, vm_page,
+				    min, offset_in_page(buf));
+		} else {
+			sg_init_one(&dma_xfer[sg_count].sg, buf, min);
+		}
+
+		ret = dma_map_sg(ssp->dev, &dma_xfer[sg_count].sg, 1,
+			(flags & TXRX_WRITE) ? DMA_TO_DEVICE : DMA_FROM_DEVICE);
+
+		len -= min;
+		buf += min;
+
+		/* Queue the PIO register write transfer. */
+		desc = dmaengine_prep_slave_sg(ssp->dmach,
+				(struct scatterlist *)dma_xfer[sg_count].pio,
+				(ssp->devid == IMX23_SSP) ? 1 : 4,
+				DMA_TRANS_NONE,
+				sg_count ? DMA_PREP_INTERRUPT : 0);
+		if (!desc) {
+			dev_err(ssp->dev,
+				"Failed to get PIO reg. write descriptor.\n");
+			ret = -EINVAL;
+			goto err_mapped;
+		}
+
+		desc = dmaengine_prep_slave_sg(ssp->dmach,
+				&dma_xfer[sg_count].sg, 1,
+				(flags & TXRX_WRITE) ? DMA_MEM_TO_DEV : DMA_DEV_TO_MEM,
+				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+
+		if (!desc) {
+			dev_err(ssp->dev,
+				"Failed to get DMA data write descriptor.\n");
+			ret = -EINVAL;
+			goto err_mapped;
+		}
+	}
+
+	/*
+	 * The last descriptor must have this callback,
+	 * to finish the DMA transaction.
+	 */
+	desc->callback = mxs_ssp_dma_irq_callback;
+	desc->callback_param = spi;
+
+	/* Start the transfer. */
+	dmaengine_submit(desc);
+	dma_async_issue_pending(ssp->dmach);
+
+	if (!wait_for_completion_timeout(&spi->c,
+					 msecs_to_jiffies(SSP_TIMEOUT))) {
+		dev_err(ssp->dev, "DMA transfer timeout\n");
+		ret = -ETIMEDOUT;
+		dmaengine_terminate_all(ssp->dmach);
+		goto err_vmalloc;
+	}
+
+	ret = 0;
+
+err_vmalloc:
+	while (--sg_count >= 0) {
+err_mapped:
+		dma_unmap_sg(ssp->dev, &dma_xfer[sg_count].sg, 1,
+			(flags & TXRX_WRITE) ? DMA_TO_DEVICE : DMA_FROM_DEVICE);
+	}
+
+	kfree(dma_xfer);
+
+	return ret;
+}
+
+static int mxs_spi_txrx_pio(struct mxs_spi *spi,
+			    unsigned char *buf, int len,
+			    unsigned int flags)
+{
+	struct mxs_ssp *ssp = &spi->ssp;
+
+	writel(BM_SSP_CTRL0_IGNORE_CRC,
+	       ssp->base + HW_SSP_CTRL0 + STMP_OFFSET_REG_CLR);
+
+	while (len--) {
+		if (len == 0 && (flags & TXRX_DEASSERT_CS))
+			writel(BM_SSP_CTRL0_IGNORE_CRC,
+			       ssp->base + HW_SSP_CTRL0 + STMP_OFFSET_REG_SET);
+
+		if (ssp->devid == IMX23_SSP) {
+			writel(BM_SSP_CTRL0_XFER_COUNT,
+				ssp->base + HW_SSP_CTRL0 + STMP_OFFSET_REG_CLR);
+			writel(1,
+				ssp->base + HW_SSP_CTRL0 + STMP_OFFSET_REG_SET);
+		} else {
+			writel(1, ssp->base + HW_SSP_XFER_SIZE);
+		}
+
+		if (flags & TXRX_WRITE)
+			writel(BM_SSP_CTRL0_READ,
+				ssp->base + HW_SSP_CTRL0 + STMP_OFFSET_REG_CLR);
+		else
+			writel(BM_SSP_CTRL0_READ,
+				ssp->base + HW_SSP_CTRL0 + STMP_OFFSET_REG_SET);
+
+		writel(BM_SSP_CTRL0_RUN,
+				ssp->base + HW_SSP_CTRL0 + STMP_OFFSET_REG_SET);
+
+		if (mxs_ssp_wait(spi, HW_SSP_CTRL0, BM_SSP_CTRL0_RUN, 1))
+			return -ETIMEDOUT;
+
+		if (flags & TXRX_WRITE)
+			writel(*buf, ssp->base + HW_SSP_DATA(ssp));
+
+		writel(BM_SSP_CTRL0_DATA_XFER,
+			     ssp->base + HW_SSP_CTRL0 + STMP_OFFSET_REG_SET);
+
+		if (!(flags & TXRX_WRITE)) {
+			if (mxs_ssp_wait(spi, HW_SSP_STATUS(ssp),
+						BM_SSP_STATUS_FIFO_EMPTY, 0))
+				return -ETIMEDOUT;
+
+			*buf = (readl(ssp->base + HW_SSP_DATA(ssp)) & 0xff);
+		}
+
+		if (mxs_ssp_wait(spi, HW_SSP_CTRL0, BM_SSP_CTRL0_RUN, 0))
+			return -ETIMEDOUT;
+
+		buf++;
+	}
+
+	if (len <= 0)
+		return 0;
+
+	return -ETIMEDOUT;
+}
+
+static int mxs_spi_transfer_one(struct spi_master *master,
+				struct spi_message *m)
+{
+	struct mxs_spi *spi = spi_master_get_devdata(master);
+	struct mxs_ssp *ssp = &spi->ssp;
+	struct spi_transfer *t;
+	unsigned int flag;
+	int status = 0;
+
+	/* Program CS register bits here, it will be used for all transfers. */
+	writel(BM_SSP_CTRL0_WAIT_FOR_CMD | BM_SSP_CTRL0_WAIT_FOR_IRQ,
+	       ssp->base + HW_SSP_CTRL0 + STMP_OFFSET_REG_CLR);
+	writel(mxs_spi_cs_to_reg(m->spi->chip_select),
+	       ssp->base + HW_SSP_CTRL0 + STMP_OFFSET_REG_SET);
+
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+
+		status = mxs_spi_setup_transfer(m->spi, t);
+		if (status)
+			break;
+
+		/* De-assert on last transfer, inverted by cs_change flag */
+		flag = (&t->transfer_list == m->transfers.prev) ^ t->cs_change ?
+		       TXRX_DEASSERT_CS : 0;
+
+		/*
+		 * Small blocks can be transfered via PIO.
+		 * Measured by empiric means:
+		 *
+		 * dd if=/dev/mtdblock0 of=/dev/null bs=1024k count=1
+		 *
+		 * DMA only: 2.164808 seconds, 473.0KB/s
+		 * Combined: 1.676276 seconds, 610.9KB/s
+		 */
+		if (t->len < 32) {
+			writel(BM_SSP_CTRL1_DMA_ENABLE,
+				ssp->base + HW_SSP_CTRL1(ssp) +
+				STMP_OFFSET_REG_CLR);
+
+			if (t->tx_buf)
+				status = mxs_spi_txrx_pio(spi,
+						(void *)t->tx_buf,
+						t->len, flag | TXRX_WRITE);
+			if (t->rx_buf)
+				status = mxs_spi_txrx_pio(spi,
+						t->rx_buf, t->len,
+						flag);
+		} else {
+			writel(BM_SSP_CTRL1_DMA_ENABLE,
+				ssp->base + HW_SSP_CTRL1(ssp) +
+				STMP_OFFSET_REG_SET);
+
+			if (t->tx_buf)
+				status = mxs_spi_txrx_dma(spi,
+						(void *)t->tx_buf, t->len,
+						flag | TXRX_WRITE);
+			if (t->rx_buf)
+				status = mxs_spi_txrx_dma(spi,
+						t->rx_buf, t->len,
+						flag);
+		}
+
+		if (status) {
+			stmp_reset_block(ssp->base);
+			break;
+		}
+
+		m->actual_length += t->len;
+	}
+
+	m->status = status;
+	spi_finalize_current_message(master);
+
+	return status;
+}
+
+static int mxs_spi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mxs_spi *spi = spi_master_get_devdata(master);
+	struct mxs_ssp *ssp = &spi->ssp;
+	int ret;
+
+	clk_disable_unprepare(ssp->clk);
+
+	ret = pinctrl_pm_select_idle_state(dev);
+	if (ret) {
+		int ret2 = clk_prepare_enable(ssp->clk);
+
+		if (ret2)
+			dev_warn(dev, "Failed to reenable clock after failing pinctrl request (pinctrl: %d, clk: %d)\n",
+				 ret, ret2);
+	}
+
+	return ret;
+}
+
+static int mxs_spi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct mxs_spi *spi = spi_master_get_devdata(master);
+	struct mxs_ssp *ssp = &spi->ssp;
+	int ret;
+
+	ret = pinctrl_pm_select_default_state(dev);
+	if (ret)
+		return ret;
+
+	ret = clk_prepare_enable(ssp->clk);
+	if (ret)
+		pinctrl_pm_select_idle_state(dev);
+
+	return ret;
+}
+
+static int __maybe_unused mxs_spi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	int ret;
+
+	ret = spi_master_suspend(master);
+	if (ret)
+		return ret;
+
+	if (!pm_runtime_suspended(dev))
+		return mxs_spi_runtime_suspend(dev);
+	else
+		return 0;
+}
+
+static int __maybe_unused mxs_spi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	int ret;
+
+	if (!pm_runtime_suspended(dev))
+		ret = mxs_spi_runtime_resume(dev);
+	else
+		ret = 0;
+	if (ret)
+		return ret;
+
+	ret = spi_master_resume(master);
+	if (ret < 0 && !pm_runtime_suspended(dev))
+		mxs_spi_runtime_suspend(dev);
+
+	return ret;
+}
+
+static const struct dev_pm_ops mxs_spi_pm = {
+	SET_RUNTIME_PM_OPS(mxs_spi_runtime_suspend,
+			   mxs_spi_runtime_resume, NULL)
+	SET_SYSTEM_SLEEP_PM_OPS(mxs_spi_suspend, mxs_spi_resume)
+};
+
+static const struct of_device_id mxs_spi_dt_ids[] = {
+	{ .compatible = "fsl,imx23-spi", .data = (void *) IMX23_SSP, },
+	{ .compatible = "fsl,imx28-spi", .data = (void *) IMX28_SSP, },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, mxs_spi_dt_ids);
+
+static int mxs_spi_probe(struct platform_device *pdev)
+{
+	const struct of_device_id *of_id =
+			of_match_device(mxs_spi_dt_ids, &pdev->dev);
+	struct device_node *np = pdev->dev.of_node;
+	struct spi_master *master;
+	struct mxs_spi *spi;
+	struct mxs_ssp *ssp;
+	struct resource *iores;
+	struct clk *clk;
+	void __iomem *base;
+	int devid, clk_freq;
+	int ret = 0, irq_err;
+
+	/*
+	 * Default clock speed for the SPI core. 160MHz seems to
+	 * work reasonably well with most SPI flashes, so use this
+	 * as a default. Override with "clock-frequency" DT prop.
+	 */
+	const int clk_freq_default = 160000000;
+
+	iores = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	irq_err = platform_get_irq(pdev, 0);
+	if (irq_err < 0)
+		return irq_err;
+
+	base = devm_ioremap_resource(&pdev->dev, iores);
+	if (IS_ERR(base))
+		return PTR_ERR(base);
+
+	clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(clk))
+		return PTR_ERR(clk);
+
+	devid = (enum mxs_ssp_id) of_id->data;
+	ret = of_property_read_u32(np, "clock-frequency",
+				   &clk_freq);
+	if (ret)
+		clk_freq = clk_freq_default;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*spi));
+	if (!master)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, master);
+
+	master->transfer_one_message = mxs_spi_transfer_one;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->mode_bits = SPI_CPOL | SPI_CPHA;
+	master->num_chipselect = 3;
+	master->dev.of_node = np;
+	master->flags = SPI_MASTER_HALF_DUPLEX;
+	master->auto_runtime_pm = true;
+
+	spi = spi_master_get_devdata(master);
+	ssp = &spi->ssp;
+	ssp->dev = &pdev->dev;
+	ssp->clk = clk;
+	ssp->base = base;
+	ssp->devid = devid;
+
+	init_completion(&spi->c);
+
+	ret = devm_request_irq(&pdev->dev, irq_err, mxs_ssp_irq_handler, 0,
+			       dev_name(&pdev->dev), ssp);
+	if (ret)
+		goto out_master_free;
+
+	ssp->dmach = dma_request_slave_channel(&pdev->dev, "rx-tx");
+	if (!ssp->dmach) {
+		dev_err(ssp->dev, "Failed to request DMA\n");
+		ret = -ENODEV;
+		goto out_master_free;
+	}
+
+	pm_runtime_enable(ssp->dev);
+	if (!pm_runtime_enabled(ssp->dev)) {
+		ret = mxs_spi_runtime_resume(ssp->dev);
+		if (ret < 0) {
+			dev_err(ssp->dev, "runtime resume failed\n");
+			goto out_dma_release;
+		}
+	}
+
+	ret = pm_runtime_get_sync(ssp->dev);
+	if (ret < 0) {
+		dev_err(ssp->dev, "runtime_get_sync failed\n");
+		goto out_pm_runtime_disable;
+	}
+
+	clk_set_rate(ssp->clk, clk_freq);
+
+	ret = stmp_reset_block(ssp->base);
+	if (ret)
+		goto out_pm_runtime_put;
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret) {
+		dev_err(&pdev->dev, "Cannot register SPI master, %d\n", ret);
+		goto out_pm_runtime_put;
+	}
+
+	pm_runtime_put(ssp->dev);
+
+	return 0;
+
+out_pm_runtime_put:
+	pm_runtime_put(ssp->dev);
+out_pm_runtime_disable:
+	pm_runtime_disable(ssp->dev);
+out_dma_release:
+	dma_release_channel(ssp->dmach);
+out_master_free:
+	spi_master_put(master);
+	return ret;
+}
+
+static int mxs_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct mxs_spi *spi;
+	struct mxs_ssp *ssp;
+
+	master = platform_get_drvdata(pdev);
+	spi = spi_master_get_devdata(master);
+	ssp = &spi->ssp;
+
+	pm_runtime_disable(&pdev->dev);
+	if (!pm_runtime_status_suspended(&pdev->dev))
+		mxs_spi_runtime_suspend(&pdev->dev);
+
+	dma_release_channel(ssp->dmach);
+
+	return 0;
+}
+
+static struct platform_driver mxs_spi_driver = {
+	.probe	= mxs_spi_probe,
+	.remove	= mxs_spi_remove,
+	.driver	= {
+		.name	= DRIVER_NAME,
+		.of_match_table = mxs_spi_dt_ids,
+		.pm = &mxs_spi_pm,
+	},
+};
+
+module_platform_driver(mxs_spi_driver);
+
+MODULE_AUTHOR("Marek Vasut <marex@denx.de>");
+MODULE_DESCRIPTION("MXS SPI master driver");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:mxs-spi");
diff --git a/drivers/spi/spi-nuc900.c b/drivers/spi/spi-nuc900.c
new file mode 100644
index 0000000..f51a058
--- /dev/null
+++ b/drivers/spi/spi-nuc900.c
@@ -0,0 +1,433 @@
+/*
+ * Copyright (c) 2009 Nuvoton technology.
+ * Wan ZongShun <mcuos.com@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/spinlock.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/errno.h>
+#include <linux/err.h>
+#include <linux/clk.h>
+#include <linux/device.h>
+#include <linux/platform_device.h>
+#include <linux/gpio.h>
+#include <linux/io.h>
+#include <linux/slab.h>
+
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+
+#include <linux/platform_data/spi-nuc900.h>
+
+/* usi registers offset */
+#define USI_CNT		0x00
+#define USI_DIV		0x04
+#define USI_SSR		0x08
+#define USI_RX0		0x10
+#define USI_TX0		0x10
+
+/* usi register bit */
+#define ENINT		(0x01 << 17)
+#define ENFLG		(0x01 << 16)
+#define SLEEP		(0x0f << 12)
+#define TXNUM		(0x03 << 8)
+#define TXBITLEN	(0x1f << 3)
+#define TXNEG		(0x01 << 2)
+#define RXNEG		(0x01 << 1)
+#define LSB		(0x01 << 10)
+#define SELECTLEV	(0x01 << 2)
+#define SELECTPOL	(0x01 << 31)
+#define SELECTSLAVE	0x01
+#define GOBUSY		0x01
+
+struct nuc900_spi {
+	struct spi_bitbang	 bitbang;
+	struct completion	 done;
+	void __iomem		*regs;
+	int			 irq;
+	int			 len;
+	int			 count;
+	const unsigned char	*tx;
+	unsigned char		*rx;
+	struct clk		*clk;
+	struct spi_master	*master;
+	struct nuc900_spi_info *pdata;
+	spinlock_t		lock;
+};
+
+static inline struct nuc900_spi *to_hw(struct spi_device *sdev)
+{
+	return spi_master_get_devdata(sdev->master);
+}
+
+static void nuc900_slave_select(struct spi_device *spi, unsigned int ssr)
+{
+	struct nuc900_spi *hw = to_hw(spi);
+	unsigned int val;
+	unsigned int cs = spi->mode & SPI_CS_HIGH ? 1 : 0;
+	unsigned int cpol = spi->mode & SPI_CPOL ? 1 : 0;
+	unsigned long flags;
+
+	spin_lock_irqsave(&hw->lock, flags);
+
+	val = __raw_readl(hw->regs + USI_SSR);
+
+	if (!cs)
+		val &= ~SELECTLEV;
+	else
+		val |= SELECTLEV;
+
+	if (!ssr)
+		val &= ~SELECTSLAVE;
+	else
+		val |= SELECTSLAVE;
+
+	__raw_writel(val, hw->regs + USI_SSR);
+
+	val = __raw_readl(hw->regs + USI_CNT);
+
+	if (!cpol)
+		val &= ~SELECTPOL;
+	else
+		val |= SELECTPOL;
+
+	__raw_writel(val, hw->regs + USI_CNT);
+
+	spin_unlock_irqrestore(&hw->lock, flags);
+}
+
+static void nuc900_spi_chipsel(struct spi_device *spi, int value)
+{
+	switch (value) {
+	case BITBANG_CS_INACTIVE:
+		nuc900_slave_select(spi, 0);
+		break;
+
+	case BITBANG_CS_ACTIVE:
+		nuc900_slave_select(spi, 1);
+		break;
+	}
+}
+
+static void nuc900_spi_setup_txnum(struct nuc900_spi *hw, unsigned int txnum)
+{
+	unsigned int val;
+	unsigned long flags;
+
+	spin_lock_irqsave(&hw->lock, flags);
+
+	val = __raw_readl(hw->regs + USI_CNT) & ~TXNUM;
+
+	if (txnum)
+		val |= txnum << 0x08;
+
+	__raw_writel(val, hw->regs + USI_CNT);
+
+	spin_unlock_irqrestore(&hw->lock, flags);
+
+}
+
+static void nuc900_spi_setup_txbitlen(struct nuc900_spi *hw,
+							unsigned int txbitlen)
+{
+	unsigned int val;
+	unsigned long flags;
+
+	spin_lock_irqsave(&hw->lock, flags);
+
+	val = __raw_readl(hw->regs + USI_CNT) & ~TXBITLEN;
+
+	val |= (txbitlen << 0x03);
+
+	__raw_writel(val, hw->regs + USI_CNT);
+
+	spin_unlock_irqrestore(&hw->lock, flags);
+}
+
+static void nuc900_spi_gobusy(struct nuc900_spi *hw)
+{
+	unsigned int val;
+	unsigned long flags;
+
+	spin_lock_irqsave(&hw->lock, flags);
+
+	val = __raw_readl(hw->regs + USI_CNT);
+
+	val |= GOBUSY;
+
+	__raw_writel(val, hw->regs + USI_CNT);
+
+	spin_unlock_irqrestore(&hw->lock, flags);
+}
+
+static inline unsigned int hw_txbyte(struct nuc900_spi *hw, int count)
+{
+	return hw->tx ? hw->tx[count] : 0;
+}
+
+static int nuc900_spi_txrx(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct nuc900_spi *hw = to_hw(spi);
+
+	hw->tx = t->tx_buf;
+	hw->rx = t->rx_buf;
+	hw->len = t->len;
+	hw->count = 0;
+
+	__raw_writel(hw_txbyte(hw, 0x0), hw->regs + USI_TX0);
+
+	nuc900_spi_gobusy(hw);
+
+	wait_for_completion(&hw->done);
+
+	return hw->count;
+}
+
+static irqreturn_t nuc900_spi_irq(int irq, void *dev)
+{
+	struct nuc900_spi *hw = dev;
+	unsigned int status;
+	unsigned int count = hw->count;
+
+	status = __raw_readl(hw->regs + USI_CNT);
+	__raw_writel(status, hw->regs + USI_CNT);
+
+	if (status & ENFLG) {
+		hw->count++;
+
+		if (hw->rx)
+			hw->rx[count] = __raw_readl(hw->regs + USI_RX0);
+		count++;
+
+		if (count < hw->len) {
+			__raw_writel(hw_txbyte(hw, count), hw->regs + USI_TX0);
+			nuc900_spi_gobusy(hw);
+		} else {
+			complete(&hw->done);
+		}
+
+		return IRQ_HANDLED;
+	}
+
+	complete(&hw->done);
+	return IRQ_HANDLED;
+}
+
+static void nuc900_tx_edge(struct nuc900_spi *hw, unsigned int edge)
+{
+	unsigned int val;
+	unsigned long flags;
+
+	spin_lock_irqsave(&hw->lock, flags);
+
+	val = __raw_readl(hw->regs + USI_CNT);
+
+	if (edge)
+		val |= TXNEG;
+	else
+		val &= ~TXNEG;
+	__raw_writel(val, hw->regs + USI_CNT);
+
+	spin_unlock_irqrestore(&hw->lock, flags);
+}
+
+static void nuc900_rx_edge(struct nuc900_spi *hw, unsigned int edge)
+{
+	unsigned int val;
+	unsigned long flags;
+
+	spin_lock_irqsave(&hw->lock, flags);
+
+	val = __raw_readl(hw->regs + USI_CNT);
+
+	if (edge)
+		val |= RXNEG;
+	else
+		val &= ~RXNEG;
+	__raw_writel(val, hw->regs + USI_CNT);
+
+	spin_unlock_irqrestore(&hw->lock, flags);
+}
+
+static void nuc900_send_first(struct nuc900_spi *hw, unsigned int lsb)
+{
+	unsigned int val;
+	unsigned long flags;
+
+	spin_lock_irqsave(&hw->lock, flags);
+
+	val = __raw_readl(hw->regs + USI_CNT);
+
+	if (lsb)
+		val |= LSB;
+	else
+		val &= ~LSB;
+	__raw_writel(val, hw->regs + USI_CNT);
+
+	spin_unlock_irqrestore(&hw->lock, flags);
+}
+
+static void nuc900_set_sleep(struct nuc900_spi *hw, unsigned int sleep)
+{
+	unsigned int val;
+	unsigned long flags;
+
+	spin_lock_irqsave(&hw->lock, flags);
+
+	val = __raw_readl(hw->regs + USI_CNT) & ~SLEEP;
+
+	if (sleep)
+		val |= (sleep << 12);
+
+	__raw_writel(val, hw->regs + USI_CNT);
+
+	spin_unlock_irqrestore(&hw->lock, flags);
+}
+
+static void nuc900_enable_int(struct nuc900_spi *hw)
+{
+	unsigned int val;
+	unsigned long flags;
+
+	spin_lock_irqsave(&hw->lock, flags);
+
+	val = __raw_readl(hw->regs + USI_CNT);
+
+	val |= ENINT;
+
+	__raw_writel(val, hw->regs + USI_CNT);
+
+	spin_unlock_irqrestore(&hw->lock, flags);
+}
+
+static void nuc900_set_divider(struct nuc900_spi *hw)
+{
+	__raw_writel(hw->pdata->divider, hw->regs + USI_DIV);
+}
+
+static void nuc900_init_spi(struct nuc900_spi *hw)
+{
+	clk_enable(hw->clk);
+	spin_lock_init(&hw->lock);
+
+	nuc900_tx_edge(hw, hw->pdata->txneg);
+	nuc900_rx_edge(hw, hw->pdata->rxneg);
+	nuc900_send_first(hw, hw->pdata->lsb);
+	nuc900_set_sleep(hw, hw->pdata->sleep);
+	nuc900_spi_setup_txbitlen(hw, hw->pdata->txbitlen);
+	nuc900_spi_setup_txnum(hw, hw->pdata->txnum);
+	nuc900_set_divider(hw);
+	nuc900_enable_int(hw);
+}
+
+static int nuc900_spi_probe(struct platform_device *pdev)
+{
+	struct nuc900_spi *hw;
+	struct spi_master *master;
+	struct resource *res;
+	int err = 0;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct nuc900_spi));
+	if (master == NULL) {
+		dev_err(&pdev->dev, "No memory for spi_master\n");
+		return -ENOMEM;
+	}
+
+	hw = spi_master_get_devdata(master);
+	hw->master = master;
+	hw->pdata  = dev_get_platdata(&pdev->dev);
+
+	if (hw->pdata == NULL) {
+		dev_err(&pdev->dev, "No platform data supplied\n");
+		err = -ENOENT;
+		goto err_pdata;
+	}
+
+	platform_set_drvdata(pdev, hw);
+	init_completion(&hw->done);
+
+	master->mode_bits          = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	if (hw->pdata->lsb)
+		master->mode_bits |= SPI_LSB_FIRST;
+	master->num_chipselect     = hw->pdata->num_cs;
+	master->bus_num            = hw->pdata->bus_num;
+	hw->bitbang.master         = hw->master;
+	hw->bitbang.chipselect     = nuc900_spi_chipsel;
+	hw->bitbang.txrx_bufs      = nuc900_spi_txrx;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	hw->regs = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(hw->regs)) {
+		err = PTR_ERR(hw->regs);
+		goto err_pdata;
+	}
+
+	hw->irq = platform_get_irq(pdev, 0);
+	if (hw->irq < 0) {
+		dev_err(&pdev->dev, "No IRQ specified\n");
+		err = -ENOENT;
+		goto err_pdata;
+	}
+
+	err = devm_request_irq(&pdev->dev, hw->irq, nuc900_spi_irq, 0,
+				pdev->name, hw);
+	if (err) {
+		dev_err(&pdev->dev, "Cannot claim IRQ\n");
+		goto err_pdata;
+	}
+
+	hw->clk = devm_clk_get(&pdev->dev, "spi");
+	if (IS_ERR(hw->clk)) {
+		dev_err(&pdev->dev, "No clock for device\n");
+		err = PTR_ERR(hw->clk);
+		goto err_pdata;
+	}
+
+	mfp_set_groupg(&pdev->dev, NULL);
+	nuc900_init_spi(hw);
+
+	err = spi_bitbang_start(&hw->bitbang);
+	if (err) {
+		dev_err(&pdev->dev, "Failed to register SPI master\n");
+		goto err_register;
+	}
+
+	return 0;
+
+err_register:
+	clk_disable(hw->clk);
+err_pdata:
+	spi_master_put(hw->master);
+	return err;
+}
+
+static int nuc900_spi_remove(struct platform_device *dev)
+{
+	struct nuc900_spi *hw = platform_get_drvdata(dev);
+
+	spi_bitbang_stop(&hw->bitbang);
+	clk_disable(hw->clk);
+	spi_master_put(hw->master);
+	return 0;
+}
+
+static struct platform_driver nuc900_spi_driver = {
+	.probe		= nuc900_spi_probe,
+	.remove		= nuc900_spi_remove,
+	.driver		= {
+		.name	= "nuc900-spi",
+	},
+};
+module_platform_driver(nuc900_spi_driver);
+
+MODULE_AUTHOR("Wan ZongShun <mcuos.com@gmail.com>");
+MODULE_DESCRIPTION("nuc900 spi driver!");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:nuc900-spi");
diff --git a/drivers/spi/spi-oc-tiny.c b/drivers/spi/spi-oc-tiny.c
new file mode 100644
index 0000000..085f580
--- /dev/null
+++ b/drivers/spi/spi-oc-tiny.c
@@ -0,0 +1,359 @@
+/*
+ * OpenCores tiny SPI master driver
+ *
+ * http://opencores.org/project,tiny_spi
+ *
+ * Copyright (C) 2011 Thomas Chou <thomas@wytron.com.tw>
+ *
+ * Based on spi_s3c24xx.c, which is:
+ * Copyright (c) 2006 Ben Dooks
+ * Copyright (c) 2006 Simtec Electronics
+ *	Ben Dooks <ben@simtec.co.uk>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/interrupt.h>
+#include <linux/errno.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/spi/spi_oc_tiny.h>
+#include <linux/io.h>
+#include <linux/gpio.h>
+#include <linux/of.h>
+
+#define DRV_NAME "spi_oc_tiny"
+
+#define TINY_SPI_RXDATA 0
+#define TINY_SPI_TXDATA 4
+#define TINY_SPI_STATUS 8
+#define TINY_SPI_CONTROL 12
+#define TINY_SPI_BAUD 16
+
+#define TINY_SPI_STATUS_TXE 0x1
+#define TINY_SPI_STATUS_TXR 0x2
+
+struct tiny_spi {
+	/* bitbang has to be first */
+	struct spi_bitbang bitbang;
+	struct completion done;
+
+	void __iomem *base;
+	int irq;
+	unsigned int freq;
+	unsigned int baudwidth;
+	unsigned int baud;
+	unsigned int speed_hz;
+	unsigned int mode;
+	unsigned int len;
+	unsigned int txc, rxc;
+	const u8 *txp;
+	u8 *rxp;
+	int gpio_cs_count;
+	int *gpio_cs;
+};
+
+static inline struct tiny_spi *tiny_spi_to_hw(struct spi_device *sdev)
+{
+	return spi_master_get_devdata(sdev->master);
+}
+
+static unsigned int tiny_spi_baud(struct spi_device *spi, unsigned int hz)
+{
+	struct tiny_spi *hw = tiny_spi_to_hw(spi);
+
+	return min(DIV_ROUND_UP(hw->freq, hz * 2), (1U << hw->baudwidth)) - 1;
+}
+
+static void tiny_spi_chipselect(struct spi_device *spi, int is_active)
+{
+	struct tiny_spi *hw = tiny_spi_to_hw(spi);
+
+	if (hw->gpio_cs_count > 0) {
+		gpio_set_value(hw->gpio_cs[spi->chip_select],
+			(spi->mode & SPI_CS_HIGH) ? is_active : !is_active);
+	}
+}
+
+static int tiny_spi_setup_transfer(struct spi_device *spi,
+				   struct spi_transfer *t)
+{
+	struct tiny_spi *hw = tiny_spi_to_hw(spi);
+	unsigned int baud = hw->baud;
+
+	if (t) {
+		if (t->speed_hz && t->speed_hz != hw->speed_hz)
+			baud = tiny_spi_baud(spi, t->speed_hz);
+	}
+	writel(baud, hw->base + TINY_SPI_BAUD);
+	writel(hw->mode, hw->base + TINY_SPI_CONTROL);
+	return 0;
+}
+
+static int tiny_spi_setup(struct spi_device *spi)
+{
+	struct tiny_spi *hw = tiny_spi_to_hw(spi);
+
+	if (spi->max_speed_hz != hw->speed_hz) {
+		hw->speed_hz = spi->max_speed_hz;
+		hw->baud = tiny_spi_baud(spi, hw->speed_hz);
+	}
+	hw->mode = spi->mode & (SPI_CPOL | SPI_CPHA);
+	return 0;
+}
+
+static inline void tiny_spi_wait_txr(struct tiny_spi *hw)
+{
+	while (!(readb(hw->base + TINY_SPI_STATUS) &
+		 TINY_SPI_STATUS_TXR))
+		cpu_relax();
+}
+
+static inline void tiny_spi_wait_txe(struct tiny_spi *hw)
+{
+	while (!(readb(hw->base + TINY_SPI_STATUS) &
+		 TINY_SPI_STATUS_TXE))
+		cpu_relax();
+}
+
+static int tiny_spi_txrx_bufs(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct tiny_spi *hw = tiny_spi_to_hw(spi);
+	const u8 *txp = t->tx_buf;
+	u8 *rxp = t->rx_buf;
+	unsigned int i;
+
+	if (hw->irq >= 0) {
+		/* use interrupt driven data transfer */
+		hw->len = t->len;
+		hw->txp = t->tx_buf;
+		hw->rxp = t->rx_buf;
+		hw->txc = 0;
+		hw->rxc = 0;
+
+		/* send the first byte */
+		if (t->len > 1) {
+			writeb(hw->txp ? *hw->txp++ : 0,
+			       hw->base + TINY_SPI_TXDATA);
+			hw->txc++;
+			writeb(hw->txp ? *hw->txp++ : 0,
+			       hw->base + TINY_SPI_TXDATA);
+			hw->txc++;
+			writeb(TINY_SPI_STATUS_TXR, hw->base + TINY_SPI_STATUS);
+		} else {
+			writeb(hw->txp ? *hw->txp++ : 0,
+			       hw->base + TINY_SPI_TXDATA);
+			hw->txc++;
+			writeb(TINY_SPI_STATUS_TXE, hw->base + TINY_SPI_STATUS);
+		}
+
+		wait_for_completion(&hw->done);
+	} else {
+		/* we need to tighten the transfer loop */
+		writeb(txp ? *txp++ : 0, hw->base + TINY_SPI_TXDATA);
+		for (i = 1; i < t->len; i++) {
+			writeb(txp ? *txp++ : 0, hw->base + TINY_SPI_TXDATA);
+
+			if (rxp || (i != t->len - 1))
+				tiny_spi_wait_txr(hw);
+			if (rxp)
+				*rxp++ = readb(hw->base + TINY_SPI_TXDATA);
+		}
+		tiny_spi_wait_txe(hw);
+		if (rxp)
+			*rxp++ = readb(hw->base + TINY_SPI_RXDATA);
+	}
+
+	return t->len;
+}
+
+static irqreturn_t tiny_spi_irq(int irq, void *dev)
+{
+	struct tiny_spi *hw = dev;
+
+	writeb(0, hw->base + TINY_SPI_STATUS);
+	if (hw->rxc + 1 == hw->len) {
+		if (hw->rxp)
+			*hw->rxp++ = readb(hw->base + TINY_SPI_RXDATA);
+		hw->rxc++;
+		complete(&hw->done);
+	} else {
+		if (hw->rxp)
+			*hw->rxp++ = readb(hw->base + TINY_SPI_TXDATA);
+		hw->rxc++;
+		if (hw->txc < hw->len) {
+			writeb(hw->txp ? *hw->txp++ : 0,
+			       hw->base + TINY_SPI_TXDATA);
+			hw->txc++;
+			writeb(TINY_SPI_STATUS_TXR,
+			       hw->base + TINY_SPI_STATUS);
+		} else {
+			writeb(TINY_SPI_STATUS_TXE,
+			       hw->base + TINY_SPI_STATUS);
+		}
+	}
+	return IRQ_HANDLED;
+}
+
+#ifdef CONFIG_OF
+#include <linux/of_gpio.h>
+
+static int tiny_spi_of_probe(struct platform_device *pdev)
+{
+	struct tiny_spi *hw = platform_get_drvdata(pdev);
+	struct device_node *np = pdev->dev.of_node;
+	unsigned int i;
+	u32 val;
+
+	if (!np)
+		return 0;
+	hw->gpio_cs_count = of_gpio_count(np);
+	if (hw->gpio_cs_count > 0) {
+		hw->gpio_cs = devm_kcalloc(&pdev->dev,
+				hw->gpio_cs_count, sizeof(unsigned int),
+				GFP_KERNEL);
+		if (!hw->gpio_cs)
+			return -ENOMEM;
+	}
+	for (i = 0; i < hw->gpio_cs_count; i++) {
+		hw->gpio_cs[i] = of_get_gpio_flags(np, i, NULL);
+		if (hw->gpio_cs[i] < 0)
+			return -ENODEV;
+	}
+	hw->bitbang.master->dev.of_node = pdev->dev.of_node;
+	if (!of_property_read_u32(np, "clock-frequency", &val))
+		hw->freq = val;
+	if (!of_property_read_u32(np, "baud-width", &val))
+		hw->baudwidth = val;
+	return 0;
+}
+#else /* !CONFIG_OF */
+static int tiny_spi_of_probe(struct platform_device *pdev)
+{
+	return 0;
+}
+#endif /* CONFIG_OF */
+
+static int tiny_spi_probe(struct platform_device *pdev)
+{
+	struct tiny_spi_platform_data *platp = dev_get_platdata(&pdev->dev);
+	struct tiny_spi *hw;
+	struct spi_master *master;
+	struct resource *res;
+	unsigned int i;
+	int err = -ENODEV;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct tiny_spi));
+	if (!master)
+		return err;
+
+	/* setup the master state. */
+	master->bus_num = pdev->id;
+	master->num_chipselect = 255;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->setup = tiny_spi_setup;
+
+	hw = spi_master_get_devdata(master);
+	platform_set_drvdata(pdev, hw);
+
+	/* setup the state for the bitbang driver */
+	hw->bitbang.master = master;
+	hw->bitbang.setup_transfer = tiny_spi_setup_transfer;
+	hw->bitbang.chipselect = tiny_spi_chipselect;
+	hw->bitbang.txrx_bufs = tiny_spi_txrx_bufs;
+
+	/* find and map our resources */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	hw->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(hw->base)) {
+		err = PTR_ERR(hw->base);
+		goto exit;
+	}
+	/* irq is optional */
+	hw->irq = platform_get_irq(pdev, 0);
+	if (hw->irq >= 0) {
+		init_completion(&hw->done);
+		err = devm_request_irq(&pdev->dev, hw->irq, tiny_spi_irq, 0,
+				       pdev->name, hw);
+		if (err)
+			goto exit;
+	}
+	/* find platform data */
+	if (platp) {
+		hw->gpio_cs_count = platp->gpio_cs_count;
+		hw->gpio_cs = platp->gpio_cs;
+		if (platp->gpio_cs_count && !platp->gpio_cs) {
+			err = -EBUSY;
+			goto exit;
+		}
+		hw->freq = platp->freq;
+		hw->baudwidth = platp->baudwidth;
+	} else {
+		err = tiny_spi_of_probe(pdev);
+		if (err)
+			goto exit;
+	}
+	for (i = 0; i < hw->gpio_cs_count; i++) {
+		err = gpio_request(hw->gpio_cs[i], dev_name(&pdev->dev));
+		if (err)
+			goto exit_gpio;
+		gpio_direction_output(hw->gpio_cs[i], 1);
+	}
+	hw->bitbang.master->num_chipselect = max(1, hw->gpio_cs_count);
+
+	/* register our spi controller */
+	err = spi_bitbang_start(&hw->bitbang);
+	if (err)
+		goto exit;
+	dev_info(&pdev->dev, "base %p, irq %d\n", hw->base, hw->irq);
+
+	return 0;
+
+exit_gpio:
+	while (i-- > 0)
+		gpio_free(hw->gpio_cs[i]);
+exit:
+	spi_master_put(master);
+	return err;
+}
+
+static int tiny_spi_remove(struct platform_device *pdev)
+{
+	struct tiny_spi *hw = platform_get_drvdata(pdev);
+	struct spi_master *master = hw->bitbang.master;
+	unsigned int i;
+
+	spi_bitbang_stop(&hw->bitbang);
+	for (i = 0; i < hw->gpio_cs_count; i++)
+		gpio_free(hw->gpio_cs[i]);
+	spi_master_put(master);
+	return 0;
+}
+
+#ifdef CONFIG_OF
+static const struct of_device_id tiny_spi_match[] = {
+	{ .compatible = "opencores,tiny-spi-rtlsvn2", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, tiny_spi_match);
+#endif /* CONFIG_OF */
+
+static struct platform_driver tiny_spi_driver = {
+	.probe = tiny_spi_probe,
+	.remove = tiny_spi_remove,
+	.driver = {
+		.name = DRV_NAME,
+		.pm = NULL,
+		.of_match_table = of_match_ptr(tiny_spi_match),
+	},
+};
+module_platform_driver(tiny_spi_driver);
+
+MODULE_DESCRIPTION("OpenCores tiny SPI driver");
+MODULE_AUTHOR("Thomas Chou <thomas@wytron.com.tw>");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:" DRV_NAME);
diff --git a/drivers/spi/spi-omap-100k.c b/drivers/spi/spi-omap-100k.c
new file mode 100644
index 0000000..76a8425
--- /dev/null
+++ b/drivers/spi/spi-omap-100k.c
@@ -0,0 +1,502 @@
+/*
+ * OMAP7xx SPI 100k controller driver
+ * Author: Fabrice Crohas <fcrohas@gmail.com>
+ * from original omap1_mcspi driver
+ *
+ * Copyright (C) 2005, 2006 Nokia Corporation
+ * Author:      Samuel Ortiz <samuel.ortiz@nokia.com> and
+ *              Juha Yrj�l� <juha.yrjola@nokia.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/delay.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/err.h>
+#include <linux/clk.h>
+#include <linux/io.h>
+#include <linux/gpio.h>
+#include <linux/slab.h>
+
+#include <linux/spi/spi.h>
+
+#define OMAP1_SPI100K_MAX_FREQ          48000000
+
+#define ICR_SPITAS      (OMAP7XX_ICR_BASE + 0x12)
+
+#define SPI_SETUP1      0x00
+#define SPI_SETUP2      0x02
+#define SPI_CTRL        0x04
+#define SPI_STATUS      0x06
+#define SPI_TX_LSB      0x08
+#define SPI_TX_MSB      0x0a
+#define SPI_RX_LSB      0x0c
+#define SPI_RX_MSB      0x0e
+
+#define SPI_SETUP1_INT_READ_ENABLE      (1UL << 5)
+#define SPI_SETUP1_INT_WRITE_ENABLE     (1UL << 4)
+#define SPI_SETUP1_CLOCK_DIVISOR(x)     ((x) << 1)
+#define SPI_SETUP1_CLOCK_ENABLE         (1UL << 0)
+
+#define SPI_SETUP2_ACTIVE_EDGE_FALLING  (0UL << 0)
+#define SPI_SETUP2_ACTIVE_EDGE_RISING   (1UL << 0)
+#define SPI_SETUP2_NEGATIVE_LEVEL       (0UL << 5)
+#define SPI_SETUP2_POSITIVE_LEVEL       (1UL << 5)
+#define SPI_SETUP2_LEVEL_TRIGGER        (0UL << 10)
+#define SPI_SETUP2_EDGE_TRIGGER         (1UL << 10)
+
+#define SPI_CTRL_SEN(x)                 ((x) << 7)
+#define SPI_CTRL_WORD_SIZE(x)           (((x) - 1) << 2)
+#define SPI_CTRL_WR                     (1UL << 1)
+#define SPI_CTRL_RD                     (1UL << 0)
+
+#define SPI_STATUS_WE                   (1UL << 1)
+#define SPI_STATUS_RD                   (1UL << 0)
+
+/* use PIO for small transfers, avoiding DMA setup/teardown overhead and
+ * cache operations; better heuristics consider wordsize and bitrate.
+ */
+#define DMA_MIN_BYTES                   8
+
+#define SPI_RUNNING	0
+#define SPI_SHUTDOWN	1
+
+struct omap1_spi100k {
+	struct clk              *ick;
+	struct clk              *fck;
+
+	/* Virtual base address of the controller */
+	void __iomem            *base;
+};
+
+struct omap1_spi100k_cs {
+	void __iomem            *base;
+	int                     word_len;
+};
+
+static void spi100k_enable_clock(struct spi_master *master)
+{
+	unsigned int val;
+	struct omap1_spi100k *spi100k = spi_master_get_devdata(master);
+
+	/* enable SPI */
+	val = readw(spi100k->base + SPI_SETUP1);
+	val |= SPI_SETUP1_CLOCK_ENABLE;
+	writew(val, spi100k->base + SPI_SETUP1);
+}
+
+static void spi100k_disable_clock(struct spi_master *master)
+{
+	unsigned int val;
+	struct omap1_spi100k *spi100k = spi_master_get_devdata(master);
+
+	/* disable SPI */
+	val = readw(spi100k->base + SPI_SETUP1);
+	val &= ~SPI_SETUP1_CLOCK_ENABLE;
+	writew(val, spi100k->base + SPI_SETUP1);
+}
+
+static void spi100k_write_data(struct spi_master *master, int len, int data)
+{
+	struct omap1_spi100k *spi100k = spi_master_get_devdata(master);
+
+	/* write 16-bit word, shifting 8-bit data if necessary */
+	if (len <= 8) {
+		data <<= 8;
+		len = 16;
+	}
+
+	spi100k_enable_clock(master);
+	writew(data , spi100k->base + SPI_TX_MSB);
+
+	writew(SPI_CTRL_SEN(0) |
+	       SPI_CTRL_WORD_SIZE(len) |
+	       SPI_CTRL_WR,
+	       spi100k->base + SPI_CTRL);
+
+	/* Wait for bit ack send change */
+	while ((readw(spi100k->base + SPI_STATUS) & SPI_STATUS_WE) != SPI_STATUS_WE)
+		;
+	udelay(1000);
+
+	spi100k_disable_clock(master);
+}
+
+static int spi100k_read_data(struct spi_master *master, int len)
+{
+	int dataH, dataL;
+	struct omap1_spi100k *spi100k = spi_master_get_devdata(master);
+
+	/* Always do at least 16 bits */
+	if (len <= 8)
+		len = 16;
+
+	spi100k_enable_clock(master);
+	writew(SPI_CTRL_SEN(0) |
+	       SPI_CTRL_WORD_SIZE(len) |
+	       SPI_CTRL_RD,
+	       spi100k->base + SPI_CTRL);
+
+	while ((readw(spi100k->base + SPI_STATUS) & SPI_STATUS_RD) != SPI_STATUS_RD)
+		;
+	udelay(1000);
+
+	dataL = readw(spi100k->base + SPI_RX_LSB);
+	dataH = readw(spi100k->base + SPI_RX_MSB);
+	spi100k_disable_clock(master);
+
+	return dataL;
+}
+
+static void spi100k_open(struct spi_master *master)
+{
+	/* get control of SPI */
+	struct omap1_spi100k *spi100k = spi_master_get_devdata(master);
+
+	writew(SPI_SETUP1_INT_READ_ENABLE |
+	       SPI_SETUP1_INT_WRITE_ENABLE |
+	       SPI_SETUP1_CLOCK_DIVISOR(0), spi100k->base + SPI_SETUP1);
+
+	/* configure clock and interrupts */
+	writew(SPI_SETUP2_ACTIVE_EDGE_FALLING |
+	       SPI_SETUP2_NEGATIVE_LEVEL |
+	       SPI_SETUP2_LEVEL_TRIGGER, spi100k->base + SPI_SETUP2);
+}
+
+static void omap1_spi100k_force_cs(struct omap1_spi100k *spi100k, int enable)
+{
+	if (enable)
+		writew(0x05fc, spi100k->base + SPI_CTRL);
+	else
+		writew(0x05fd, spi100k->base + SPI_CTRL);
+}
+
+static unsigned
+omap1_spi100k_txrx_pio(struct spi_device *spi, struct spi_transfer *xfer)
+{
+	struct omap1_spi100k_cs *cs = spi->controller_state;
+	unsigned int            count, c;
+	int                     word_len;
+
+	count = xfer->len;
+	c = count;
+	word_len = cs->word_len;
+
+	if (word_len <= 8) {
+		u8              *rx;
+		const u8        *tx;
+
+		rx = xfer->rx_buf;
+		tx = xfer->tx_buf;
+		do {
+			c -= 1;
+			if (xfer->tx_buf != NULL)
+				spi100k_write_data(spi->master, word_len, *tx++);
+			if (xfer->rx_buf != NULL)
+				*rx++ = spi100k_read_data(spi->master, word_len);
+		} while (c);
+	} else if (word_len <= 16) {
+		u16             *rx;
+		const u16       *tx;
+
+		rx = xfer->rx_buf;
+		tx = xfer->tx_buf;
+		do {
+			c -= 2;
+			if (xfer->tx_buf != NULL)
+				spi100k_write_data(spi->master, word_len, *tx++);
+			if (xfer->rx_buf != NULL)
+				*rx++ = spi100k_read_data(spi->master, word_len);
+		} while (c);
+	} else if (word_len <= 32) {
+		u32             *rx;
+		const u32       *tx;
+
+		rx = xfer->rx_buf;
+		tx = xfer->tx_buf;
+		do {
+			c -= 4;
+			if (xfer->tx_buf != NULL)
+				spi100k_write_data(spi->master, word_len, *tx);
+			if (xfer->rx_buf != NULL)
+				*rx = spi100k_read_data(spi->master, word_len);
+		} while (c);
+	}
+	return count - c;
+}
+
+/* called only when no transfer is active to this device */
+static int omap1_spi100k_setup_transfer(struct spi_device *spi,
+		struct spi_transfer *t)
+{
+	struct omap1_spi100k *spi100k = spi_master_get_devdata(spi->master);
+	struct omap1_spi100k_cs *cs = spi->controller_state;
+	u8 word_len;
+
+	if (t != NULL)
+		word_len = t->bits_per_word;
+	else
+		word_len = spi->bits_per_word;
+
+	if (spi->bits_per_word > 32)
+		return -EINVAL;
+	cs->word_len = word_len;
+
+	/* SPI init before transfer */
+	writew(0x3e , spi100k->base + SPI_SETUP1);
+	writew(0x00 , spi100k->base + SPI_STATUS);
+	writew(0x3e , spi100k->base + SPI_CTRL);
+
+	return 0;
+}
+
+/* the spi->mode bits understood by this driver: */
+#define MODEBITS (SPI_CPOL | SPI_CPHA | SPI_CS_HIGH)
+
+static int omap1_spi100k_setup(struct spi_device *spi)
+{
+	int                     ret;
+	struct omap1_spi100k    *spi100k;
+	struct omap1_spi100k_cs *cs = spi->controller_state;
+
+	spi100k = spi_master_get_devdata(spi->master);
+
+	if (!cs) {
+		cs = devm_kzalloc(&spi->dev, sizeof(*cs), GFP_KERNEL);
+		if (!cs)
+			return -ENOMEM;
+		cs->base = spi100k->base + spi->chip_select * 0x14;
+		spi->controller_state = cs;
+	}
+
+	spi100k_open(spi->master);
+
+	clk_prepare_enable(spi100k->ick);
+	clk_prepare_enable(spi100k->fck);
+
+	ret = omap1_spi100k_setup_transfer(spi, NULL);
+
+	clk_disable_unprepare(spi100k->ick);
+	clk_disable_unprepare(spi100k->fck);
+
+	return ret;
+}
+
+static int omap1_spi100k_transfer_one_message(struct spi_master *master,
+					      struct spi_message *m)
+{
+	struct omap1_spi100k *spi100k = spi_master_get_devdata(master);
+	struct spi_device *spi = m->spi;
+	struct spi_transfer *t = NULL;
+	int cs_active = 0;
+	int status = 0;
+
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if (t->tx_buf == NULL && t->rx_buf == NULL && t->len) {
+			status = -EINVAL;
+			break;
+		}
+		status = omap1_spi100k_setup_transfer(spi, t);
+		if (status < 0)
+			break;
+
+		if (!cs_active) {
+			omap1_spi100k_force_cs(spi100k, 1);
+			cs_active = 1;
+		}
+
+		if (t->len) {
+			unsigned count;
+
+			count = omap1_spi100k_txrx_pio(spi, t);
+			m->actual_length += count;
+
+			if (count != t->len) {
+				status = -EIO;
+				break;
+			}
+		}
+
+		if (t->delay_usecs)
+			udelay(t->delay_usecs);
+
+		/* ignore the "leave it on after last xfer" hint */
+
+		if (t->cs_change) {
+			omap1_spi100k_force_cs(spi100k, 0);
+			cs_active = 0;
+		}
+	}
+
+	status = omap1_spi100k_setup_transfer(spi, NULL);
+
+	if (cs_active)
+		omap1_spi100k_force_cs(spi100k, 0);
+
+	m->status = status;
+
+	spi_finalize_current_message(master);
+
+	return status;
+}
+
+static int omap1_spi100k_probe(struct platform_device *pdev)
+{
+	struct spi_master       *master;
+	struct omap1_spi100k    *spi100k;
+	int                     status = 0;
+
+	if (!pdev->id)
+		return -EINVAL;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*spi100k));
+	if (master == NULL) {
+		dev_dbg(&pdev->dev, "master allocation failed\n");
+		return -ENOMEM;
+	}
+
+	if (pdev->id != -1)
+		master->bus_num = pdev->id;
+
+	master->setup = omap1_spi100k_setup;
+	master->transfer_one_message = omap1_spi100k_transfer_one_message;
+	master->num_chipselect = 2;
+	master->mode_bits = MODEBITS;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 32);
+	master->min_speed_hz = OMAP1_SPI100K_MAX_FREQ/(1<<16);
+	master->max_speed_hz = OMAP1_SPI100K_MAX_FREQ;
+	master->auto_runtime_pm = true;
+
+	spi100k = spi_master_get_devdata(master);
+
+	/*
+	 * The memory region base address is taken as the platform_data.
+	 * You should allocate this with ioremap() before initializing
+	 * the SPI.
+	 */
+	spi100k->base = (void __iomem *)dev_get_platdata(&pdev->dev);
+
+	spi100k->ick = devm_clk_get(&pdev->dev, "ick");
+	if (IS_ERR(spi100k->ick)) {
+		dev_dbg(&pdev->dev, "can't get spi100k_ick\n");
+		status = PTR_ERR(spi100k->ick);
+		goto err;
+	}
+
+	spi100k->fck = devm_clk_get(&pdev->dev, "fck");
+	if (IS_ERR(spi100k->fck)) {
+		dev_dbg(&pdev->dev, "can't get spi100k_fck\n");
+		status = PTR_ERR(spi100k->fck);
+		goto err;
+	}
+
+	status = clk_prepare_enable(spi100k->ick);
+	if (status != 0) {
+		dev_err(&pdev->dev, "failed to enable ick: %d\n", status);
+		goto err;
+	}
+
+	status = clk_prepare_enable(spi100k->fck);
+	if (status != 0) {
+		dev_err(&pdev->dev, "failed to enable fck: %d\n", status);
+		goto err_ick;
+	}
+
+	pm_runtime_enable(&pdev->dev);
+	pm_runtime_set_active(&pdev->dev);
+
+	status = devm_spi_register_master(&pdev->dev, master);
+	if (status < 0)
+		goto err_fck;
+
+	return status;
+
+err_fck:
+	clk_disable_unprepare(spi100k->fck);
+err_ick:
+	clk_disable_unprepare(spi100k->ick);
+err:
+	spi_master_put(master);
+	return status;
+}
+
+static int omap1_spi100k_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = spi_master_get(platform_get_drvdata(pdev));
+	struct omap1_spi100k *spi100k = spi_master_get_devdata(master);
+
+	pm_runtime_disable(&pdev->dev);
+
+	clk_disable_unprepare(spi100k->fck);
+	clk_disable_unprepare(spi100k->ick);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int omap1_spi100k_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = spi_master_get(dev_get_drvdata(dev));
+	struct omap1_spi100k *spi100k = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(spi100k->ick);
+	clk_disable_unprepare(spi100k->fck);
+
+	return 0;
+}
+
+static int omap1_spi100k_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = spi_master_get(dev_get_drvdata(dev));
+	struct omap1_spi100k *spi100k = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(spi100k->ick);
+	if (ret != 0) {
+		dev_err(dev, "Failed to enable ick: %d\n", ret);
+		return ret;
+	}
+
+	ret = clk_prepare_enable(spi100k->fck);
+	if (ret != 0) {
+		dev_err(dev, "Failed to enable fck: %d\n", ret);
+		clk_disable_unprepare(spi100k->ick);
+		return ret;
+	}
+
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops omap1_spi100k_pm = {
+	SET_RUNTIME_PM_OPS(omap1_spi100k_runtime_suspend,
+			   omap1_spi100k_runtime_resume, NULL)
+};
+
+static struct platform_driver omap1_spi100k_driver = {
+	.driver = {
+		.name		= "omap1_spi100k",
+		.pm		= &omap1_spi100k_pm,
+	},
+	.probe		= omap1_spi100k_probe,
+	.remove		= omap1_spi100k_remove,
+};
+
+module_platform_driver(omap1_spi100k_driver);
+
+MODULE_DESCRIPTION("OMAP7xx SPI 100k controller driver");
+MODULE_AUTHOR("Fabrice Crohas <fcrohas@gmail.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-omap-uwire.c b/drivers/spi/spi-omap-uwire.c
new file mode 100644
index 0000000..ce8dbdb
--- /dev/null
+++ b/drivers/spi/spi-omap-uwire.c
@@ -0,0 +1,562 @@
+/*
+ * MicroWire interface driver for OMAP
+ *
+ * Copyright 2003 MontaVista Software Inc. <source@mvista.com>
+ *
+ * Ported to 2.6 OMAP uwire interface.
+ * Copyright (C) 2004 Texas Instruments.
+ *
+ * Generalization patches by Juha Yrjola <juha.yrjola@nokia.com>
+ *
+ * Copyright (C) 2005 David Brownell (ported to 2.6 SPI interface)
+ * Copyright (C) 2006 Nokia
+ *
+ * Many updates by Imre Deak <imre.deak@nokia.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ *
+ * THIS SOFTWARE IS PROVIDED "AS IS" AND ANY EXPRESS OR IMPLIED
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
+ * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
+ * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
+ * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/platform_device.h>
+#include <linux/interrupt.h>
+#include <linux/err.h>
+#include <linux/clk.h>
+#include <linux/slab.h>
+#include <linux/device.h>
+
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/module.h>
+#include <linux/io.h>
+
+#include <mach/hardware.h>
+#include <asm/mach-types.h>
+
+#include <mach/mux.h>
+
+#include <mach/omap7xx.h>	/* OMAP7XX_IO_CONF registers */
+
+
+/* FIXME address is now a platform device resource,
+ * and irqs should show there too...
+ */
+#define UWIRE_BASE_PHYS		0xFFFB3000
+
+/* uWire Registers: */
+#define UWIRE_IO_SIZE 0x20
+#define UWIRE_TDR     0x00
+#define UWIRE_RDR     0x00
+#define UWIRE_CSR     0x01
+#define UWIRE_SR1     0x02
+#define UWIRE_SR2     0x03
+#define UWIRE_SR3     0x04
+#define UWIRE_SR4     0x05
+#define UWIRE_SR5     0x06
+
+/* CSR bits */
+#define	RDRB	(1 << 15)
+#define	CSRB	(1 << 14)
+#define	START	(1 << 13)
+#define	CS_CMD	(1 << 12)
+
+/* SR1 or SR2 bits */
+#define UWIRE_READ_FALLING_EDGE		0x0001
+#define UWIRE_READ_RISING_EDGE		0x0000
+#define UWIRE_WRITE_FALLING_EDGE	0x0000
+#define UWIRE_WRITE_RISING_EDGE		0x0002
+#define UWIRE_CS_ACTIVE_LOW		0x0000
+#define UWIRE_CS_ACTIVE_HIGH		0x0004
+#define UWIRE_FREQ_DIV_2		0x0000
+#define UWIRE_FREQ_DIV_4		0x0008
+#define UWIRE_FREQ_DIV_8		0x0010
+#define UWIRE_CHK_READY			0x0020
+#define UWIRE_CLK_INVERTED		0x0040
+
+
+struct uwire_spi {
+	struct spi_bitbang	bitbang;
+	struct clk		*ck;
+};
+
+struct uwire_state {
+	unsigned	div1_idx;
+};
+
+/* REVISIT compile time constant for idx_shift? */
+/*
+ * Or, put it in a structure which is used throughout the driver;
+ * that avoids having to issue two loads for each bit of static data.
+ */
+static unsigned int uwire_idx_shift;
+static void __iomem *uwire_base;
+
+static inline void uwire_write_reg(int idx, u16 val)
+{
+	__raw_writew(val, uwire_base + (idx << uwire_idx_shift));
+}
+
+static inline u16 uwire_read_reg(int idx)
+{
+	return __raw_readw(uwire_base + (idx << uwire_idx_shift));
+}
+
+static inline void omap_uwire_configure_mode(u8 cs, unsigned long flags)
+{
+	u16	w, val = 0;
+	int	shift, reg;
+
+	if (flags & UWIRE_CLK_INVERTED)
+		val ^= 0x03;
+	val = flags & 0x3f;
+	if (cs & 1)
+		shift = 6;
+	else
+		shift = 0;
+	if (cs <= 1)
+		reg = UWIRE_SR1;
+	else
+		reg = UWIRE_SR2;
+
+	w = uwire_read_reg(reg);
+	w &= ~(0x3f << shift);
+	w |= val << shift;
+	uwire_write_reg(reg, w);
+}
+
+static int wait_uwire_csr_flag(u16 mask, u16 val, int might_not_catch)
+{
+	u16 w;
+	int c = 0;
+	unsigned long max_jiffies = jiffies + HZ;
+
+	for (;;) {
+		w = uwire_read_reg(UWIRE_CSR);
+		if ((w & mask) == val)
+			break;
+		if (time_after(jiffies, max_jiffies)) {
+			printk(KERN_ERR "%s: timeout. reg=%#06x "
+					"mask=%#06x val=%#06x\n",
+			       __func__, w, mask, val);
+			return -1;
+		}
+		c++;
+		if (might_not_catch && c > 64)
+			break;
+	}
+	return 0;
+}
+
+static void uwire_set_clk1_div(int div1_idx)
+{
+	u16 w;
+
+	w = uwire_read_reg(UWIRE_SR3);
+	w &= ~(0x03 << 1);
+	w |= div1_idx << 1;
+	uwire_write_reg(UWIRE_SR3, w);
+}
+
+static void uwire_chipselect(struct spi_device *spi, int value)
+{
+	struct	uwire_state *ust = spi->controller_state;
+	u16	w;
+	int	old_cs;
+
+
+	BUG_ON(wait_uwire_csr_flag(CSRB, 0, 0));
+
+	w = uwire_read_reg(UWIRE_CSR);
+	old_cs = (w >> 10) & 0x03;
+	if (value == BITBANG_CS_INACTIVE || old_cs != spi->chip_select) {
+		/* Deselect this CS, or the previous CS */
+		w &= ~CS_CMD;
+		uwire_write_reg(UWIRE_CSR, w);
+	}
+	/* activate specfied chipselect */
+	if (value == BITBANG_CS_ACTIVE) {
+		uwire_set_clk1_div(ust->div1_idx);
+		/* invert clock? */
+		if (spi->mode & SPI_CPOL)
+			uwire_write_reg(UWIRE_SR4, 1);
+		else
+			uwire_write_reg(UWIRE_SR4, 0);
+
+		w = spi->chip_select << 10;
+		w |= CS_CMD;
+		uwire_write_reg(UWIRE_CSR, w);
+	}
+}
+
+static int uwire_txrx(struct spi_device *spi, struct spi_transfer *t)
+{
+	unsigned	len = t->len;
+	unsigned	bits = t->bits_per_word;
+	unsigned	bytes;
+	u16		val, w;
+	int		status = 0;
+
+	if (!t->tx_buf && !t->rx_buf)
+		return 0;
+
+	w = spi->chip_select << 10;
+	w |= CS_CMD;
+
+	if (t->tx_buf) {
+		const u8	*buf = t->tx_buf;
+
+		/* NOTE:  DMA could be used for TX transfers */
+
+		/* write one or two bytes at a time */
+		while (len >= 1) {
+			/* tx bit 15 is first sent; we byteswap multibyte words
+			 * (msb-first) on the way out from memory.
+			 */
+			val = *buf++;
+			if (bits > 8) {
+				bytes = 2;
+				val |= *buf++ << 8;
+			} else
+				bytes = 1;
+			val <<= 16 - bits;
+
+#ifdef	VERBOSE
+			pr_debug("%s: write-%d =%04x\n",
+					dev_name(&spi->dev), bits, val);
+#endif
+			if (wait_uwire_csr_flag(CSRB, 0, 0))
+				goto eio;
+
+			uwire_write_reg(UWIRE_TDR, val);
+
+			/* start write */
+			val = START | w | (bits << 5);
+
+			uwire_write_reg(UWIRE_CSR, val);
+			len -= bytes;
+
+			/* Wait till write actually starts.
+			 * This is needed with MPU clock 60+ MHz.
+			 * REVISIT: we may not have time to catch it...
+			 */
+			if (wait_uwire_csr_flag(CSRB, CSRB, 1))
+				goto eio;
+
+			status += bytes;
+		}
+
+		/* REVISIT:  save this for later to get more i/o overlap */
+		if (wait_uwire_csr_flag(CSRB, 0, 0))
+			goto eio;
+
+	} else if (t->rx_buf) {
+		u8		*buf = t->rx_buf;
+
+		/* read one or two bytes at a time */
+		while (len) {
+			if (bits > 8) {
+				bytes = 2;
+			} else
+				bytes = 1;
+
+			/* start read */
+			val = START | w | (bits << 0);
+			uwire_write_reg(UWIRE_CSR, val);
+			len -= bytes;
+
+			/* Wait till read actually starts */
+			(void) wait_uwire_csr_flag(CSRB, CSRB, 1);
+
+			if (wait_uwire_csr_flag(RDRB | CSRB,
+						RDRB, 0))
+				goto eio;
+
+			/* rx bit 0 is last received; multibyte words will
+			 * be properly byteswapped on the way to memory.
+			 */
+			val = uwire_read_reg(UWIRE_RDR);
+			val &= (1 << bits) - 1;
+			*buf++ = (u8) val;
+			if (bytes == 2)
+				*buf++ = val >> 8;
+			status += bytes;
+#ifdef	VERBOSE
+			pr_debug("%s: read-%d =%04x\n",
+					dev_name(&spi->dev), bits, val);
+#endif
+
+		}
+	}
+	return status;
+eio:
+	return -EIO;
+}
+
+static int uwire_setup_transfer(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct uwire_state	*ust = spi->controller_state;
+	struct uwire_spi	*uwire;
+	unsigned		flags = 0;
+	unsigned		hz;
+	unsigned long		rate;
+	int			div1_idx;
+	int			div1;
+	int			div2;
+	int			status;
+
+	uwire = spi_master_get_devdata(spi->master);
+
+	/* mode 0..3, clock inverted separately;
+	 * standard nCS signaling;
+	 * don't treat DI=high as "not ready"
+	 */
+	if (spi->mode & SPI_CS_HIGH)
+		flags |= UWIRE_CS_ACTIVE_HIGH;
+
+	if (spi->mode & SPI_CPOL)
+		flags |= UWIRE_CLK_INVERTED;
+
+	switch (spi->mode & (SPI_CPOL | SPI_CPHA)) {
+	case SPI_MODE_0:
+	case SPI_MODE_3:
+		flags |= UWIRE_WRITE_FALLING_EDGE | UWIRE_READ_RISING_EDGE;
+		break;
+	case SPI_MODE_1:
+	case SPI_MODE_2:
+		flags |= UWIRE_WRITE_RISING_EDGE | UWIRE_READ_FALLING_EDGE;
+		break;
+	}
+
+	/* assume it's already enabled */
+	rate = clk_get_rate(uwire->ck);
+
+	if (t != NULL)
+		hz = t->speed_hz;
+	else
+		hz = spi->max_speed_hz;
+
+	if (!hz) {
+		pr_debug("%s: zero speed?\n", dev_name(&spi->dev));
+		status = -EINVAL;
+		goto done;
+	}
+
+	/* F_INT = mpu_xor_clk / DIV1 */
+	for (div1_idx = 0; div1_idx < 4; div1_idx++) {
+		switch (div1_idx) {
+		case 0:
+			div1 = 2;
+			break;
+		case 1:
+			div1 = 4;
+			break;
+		case 2:
+			div1 = 7;
+			break;
+		default:
+		case 3:
+			div1 = 10;
+			break;
+		}
+		div2 = (rate / div1 + hz - 1) / hz;
+		if (div2 <= 8)
+			break;
+	}
+	if (div1_idx == 4) {
+		pr_debug("%s: lowest clock %ld, need %d\n",
+			dev_name(&spi->dev), rate / 10 / 8, hz);
+		status = -EDOM;
+		goto done;
+	}
+
+	/* we have to cache this and reset in uwire_chipselect as this is a
+	 * global parameter and another uwire device can change it under
+	 * us */
+	ust->div1_idx = div1_idx;
+	uwire_set_clk1_div(div1_idx);
+
+	rate /= div1;
+
+	switch (div2) {
+	case 0:
+	case 1:
+	case 2:
+		flags |= UWIRE_FREQ_DIV_2;
+		rate /= 2;
+		break;
+	case 3:
+	case 4:
+		flags |= UWIRE_FREQ_DIV_4;
+		rate /= 4;
+		break;
+	case 5:
+	case 6:
+	case 7:
+	case 8:
+		flags |= UWIRE_FREQ_DIV_8;
+		rate /= 8;
+		break;
+	}
+	omap_uwire_configure_mode(spi->chip_select, flags);
+	pr_debug("%s: uwire flags %02x, armxor %lu KHz, SCK %lu KHz\n",
+			__func__, flags,
+			clk_get_rate(uwire->ck) / 1000,
+			rate / 1000);
+	status = 0;
+done:
+	return status;
+}
+
+static int uwire_setup(struct spi_device *spi)
+{
+	struct uwire_state *ust = spi->controller_state;
+
+	if (ust == NULL) {
+		ust = kzalloc(sizeof(*ust), GFP_KERNEL);
+		if (ust == NULL)
+			return -ENOMEM;
+		spi->controller_state = ust;
+	}
+
+	return uwire_setup_transfer(spi, NULL);
+}
+
+static void uwire_cleanup(struct spi_device *spi)
+{
+	kfree(spi->controller_state);
+}
+
+static void uwire_off(struct uwire_spi *uwire)
+{
+	uwire_write_reg(UWIRE_SR3, 0);
+	clk_disable(uwire->ck);
+	spi_master_put(uwire->bitbang.master);
+}
+
+static int uwire_probe(struct platform_device *pdev)
+{
+	struct spi_master	*master;
+	struct uwire_spi	*uwire;
+	int			status;
+
+	master = spi_alloc_master(&pdev->dev, sizeof *uwire);
+	if (!master)
+		return -ENODEV;
+
+	uwire = spi_master_get_devdata(master);
+
+	uwire_base = devm_ioremap(&pdev->dev, UWIRE_BASE_PHYS, UWIRE_IO_SIZE);
+	if (!uwire_base) {
+		dev_dbg(&pdev->dev, "can't ioremap UWIRE\n");
+		spi_master_put(master);
+		return -ENOMEM;
+	}
+
+	platform_set_drvdata(pdev, uwire);
+
+	uwire->ck = devm_clk_get(&pdev->dev, "fck");
+	if (IS_ERR(uwire->ck)) {
+		status = PTR_ERR(uwire->ck);
+		dev_dbg(&pdev->dev, "no functional clock?\n");
+		spi_master_put(master);
+		return status;
+	}
+	clk_enable(uwire->ck);
+
+	if (cpu_is_omap7xx())
+		uwire_idx_shift = 1;
+	else
+		uwire_idx_shift = 2;
+
+	uwire_write_reg(UWIRE_SR3, 1);
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(1, 16);
+	master->flags = SPI_MASTER_HALF_DUPLEX;
+
+	master->bus_num = 2;	/* "official" */
+	master->num_chipselect = 4;
+	master->setup = uwire_setup;
+	master->cleanup = uwire_cleanup;
+
+	uwire->bitbang.master = master;
+	uwire->bitbang.chipselect = uwire_chipselect;
+	uwire->bitbang.setup_transfer = uwire_setup_transfer;
+	uwire->bitbang.txrx_bufs = uwire_txrx;
+
+	status = spi_bitbang_start(&uwire->bitbang);
+	if (status < 0) {
+		uwire_off(uwire);
+	}
+	return status;
+}
+
+static int uwire_remove(struct platform_device *pdev)
+{
+	struct uwire_spi	*uwire = platform_get_drvdata(pdev);
+
+	// FIXME remove all child devices, somewhere ...
+
+	spi_bitbang_stop(&uwire->bitbang);
+	uwire_off(uwire);
+	return 0;
+}
+
+/* work with hotplug and coldplug */
+MODULE_ALIAS("platform:omap_uwire");
+
+static struct platform_driver uwire_driver = {
+	.driver = {
+		.name		= "omap_uwire",
+	},
+	.probe = uwire_probe,
+	.remove = uwire_remove,
+	// suspend ... unuse ck
+	// resume ... use ck
+};
+
+static int __init omap_uwire_init(void)
+{
+	/* FIXME move these into the relevant board init code. also, include
+	 * H3 support; it uses tsc2101 like H2 (on a different chipselect).
+	 */
+
+	if (machine_is_omap_h2()) {
+		/* defaults: W21 SDO, U18 SDI, V19 SCL */
+		omap_cfg_reg(N14_1610_UWIRE_CS0);
+		omap_cfg_reg(N15_1610_UWIRE_CS1);
+	}
+	if (machine_is_omap_perseus2()) {
+		/* configure pins: MPU_UW_nSCS1, MPU_UW_SDO, MPU_UW_SCLK */
+		int val = omap_readl(OMAP7XX_IO_CONF_9) & ~0x00EEE000;
+		omap_writel(val | 0x00AAA000, OMAP7XX_IO_CONF_9);
+	}
+
+	return platform_driver_register(&uwire_driver);
+}
+
+static void __exit omap_uwire_exit(void)
+{
+	platform_driver_unregister(&uwire_driver);
+}
+
+subsys_initcall(omap_uwire_init);
+module_exit(omap_uwire_exit);
+
+MODULE_LICENSE("GPL");
+
diff --git a/drivers/spi/spi-omap2-mcspi.c b/drivers/spi/spi-omap2-mcspi.c
new file mode 100644
index 0000000..e2be7da
--- /dev/null
+++ b/drivers/spi/spi-omap2-mcspi.c
@@ -0,0 +1,1513 @@
+/*
+ * OMAP2 McSPI controller driver
+ *
+ * Copyright (C) 2005, 2006 Nokia Corporation
+ * Author:	Samuel Ortiz <samuel.ortiz@nokia.com> and
+ *		Juha Yrj�l� <juha.yrjola@nokia.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/platform_device.h>
+#include <linux/err.h>
+#include <linux/clk.h>
+#include <linux/io.h>
+#include <linux/slab.h>
+#include <linux/pm_runtime.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/gcd.h>
+
+#include <linux/spi/spi.h>
+#include <linux/gpio.h>
+
+#include <linux/platform_data/spi-omap2-mcspi.h>
+
+#define OMAP2_MCSPI_MAX_FREQ		48000000
+#define OMAP2_MCSPI_MAX_DIVIDER		4096
+#define OMAP2_MCSPI_MAX_FIFODEPTH	64
+#define OMAP2_MCSPI_MAX_FIFOWCNT	0xFFFF
+#define SPI_AUTOSUSPEND_TIMEOUT		2000
+
+#define OMAP2_MCSPI_REVISION		0x00
+#define OMAP2_MCSPI_SYSSTATUS		0x14
+#define OMAP2_MCSPI_IRQSTATUS		0x18
+#define OMAP2_MCSPI_IRQENABLE		0x1c
+#define OMAP2_MCSPI_WAKEUPENABLE	0x20
+#define OMAP2_MCSPI_SYST		0x24
+#define OMAP2_MCSPI_MODULCTRL		0x28
+#define OMAP2_MCSPI_XFERLEVEL		0x7c
+
+/* per-channel banks, 0x14 bytes each, first is: */
+#define OMAP2_MCSPI_CHCONF0		0x2c
+#define OMAP2_MCSPI_CHSTAT0		0x30
+#define OMAP2_MCSPI_CHCTRL0		0x34
+#define OMAP2_MCSPI_TX0			0x38
+#define OMAP2_MCSPI_RX0			0x3c
+
+/* per-register bitmasks: */
+#define OMAP2_MCSPI_IRQSTATUS_EOW	BIT(17)
+
+#define OMAP2_MCSPI_MODULCTRL_SINGLE	BIT(0)
+#define OMAP2_MCSPI_MODULCTRL_MS	BIT(2)
+#define OMAP2_MCSPI_MODULCTRL_STEST	BIT(3)
+
+#define OMAP2_MCSPI_CHCONF_PHA		BIT(0)
+#define OMAP2_MCSPI_CHCONF_POL		BIT(1)
+#define OMAP2_MCSPI_CHCONF_CLKD_MASK	(0x0f << 2)
+#define OMAP2_MCSPI_CHCONF_EPOL		BIT(6)
+#define OMAP2_MCSPI_CHCONF_WL_MASK	(0x1f << 7)
+#define OMAP2_MCSPI_CHCONF_TRM_RX_ONLY	BIT(12)
+#define OMAP2_MCSPI_CHCONF_TRM_TX_ONLY	BIT(13)
+#define OMAP2_MCSPI_CHCONF_TRM_MASK	(0x03 << 12)
+#define OMAP2_MCSPI_CHCONF_DMAW		BIT(14)
+#define OMAP2_MCSPI_CHCONF_DMAR		BIT(15)
+#define OMAP2_MCSPI_CHCONF_DPE0		BIT(16)
+#define OMAP2_MCSPI_CHCONF_DPE1		BIT(17)
+#define OMAP2_MCSPI_CHCONF_IS		BIT(18)
+#define OMAP2_MCSPI_CHCONF_TURBO	BIT(19)
+#define OMAP2_MCSPI_CHCONF_FORCE	BIT(20)
+#define OMAP2_MCSPI_CHCONF_FFET		BIT(27)
+#define OMAP2_MCSPI_CHCONF_FFER		BIT(28)
+#define OMAP2_MCSPI_CHCONF_CLKG		BIT(29)
+
+#define OMAP2_MCSPI_CHSTAT_RXS		BIT(0)
+#define OMAP2_MCSPI_CHSTAT_TXS		BIT(1)
+#define OMAP2_MCSPI_CHSTAT_EOT		BIT(2)
+#define OMAP2_MCSPI_CHSTAT_TXFFE	BIT(3)
+
+#define OMAP2_MCSPI_CHCTRL_EN		BIT(0)
+#define OMAP2_MCSPI_CHCTRL_EXTCLK_MASK	(0xff << 8)
+
+#define OMAP2_MCSPI_WAKEUPENABLE_WKEN	BIT(0)
+
+/* We have 2 DMA channels per CS, one for RX and one for TX */
+struct omap2_mcspi_dma {
+	struct dma_chan *dma_tx;
+	struct dma_chan *dma_rx;
+
+	struct completion dma_tx_completion;
+	struct completion dma_rx_completion;
+
+	char dma_rx_ch_name[14];
+	char dma_tx_ch_name[14];
+};
+
+/* use PIO for small transfers, avoiding DMA setup/teardown overhead and
+ * cache operations; better heuristics consider wordsize and bitrate.
+ */
+#define DMA_MIN_BYTES			160
+
+
+/*
+ * Used for context save and restore, structure members to be updated whenever
+ * corresponding registers are modified.
+ */
+struct omap2_mcspi_regs {
+	u32 modulctrl;
+	u32 wakeupenable;
+	struct list_head cs;
+};
+
+struct omap2_mcspi {
+	struct spi_master	*master;
+	/* Virtual base address of the controller */
+	void __iomem		*base;
+	unsigned long		phys;
+	/* SPI1 has 4 channels, while SPI2 has 2 */
+	struct omap2_mcspi_dma	*dma_channels;
+	struct device		*dev;
+	struct omap2_mcspi_regs ctx;
+	int			fifo_depth;
+	unsigned int		pin_dir:1;
+};
+
+struct omap2_mcspi_cs {
+	void __iomem		*base;
+	unsigned long		phys;
+	int			word_len;
+	u16			mode;
+	struct list_head	node;
+	/* Context save and restore shadow register */
+	u32			chconf0, chctrl0;
+};
+
+static inline void mcspi_write_reg(struct spi_master *master,
+		int idx, u32 val)
+{
+	struct omap2_mcspi *mcspi = spi_master_get_devdata(master);
+
+	writel_relaxed(val, mcspi->base + idx);
+}
+
+static inline u32 mcspi_read_reg(struct spi_master *master, int idx)
+{
+	struct omap2_mcspi *mcspi = spi_master_get_devdata(master);
+
+	return readl_relaxed(mcspi->base + idx);
+}
+
+static inline void mcspi_write_cs_reg(const struct spi_device *spi,
+		int idx, u32 val)
+{
+	struct omap2_mcspi_cs	*cs = spi->controller_state;
+
+	writel_relaxed(val, cs->base +  idx);
+}
+
+static inline u32 mcspi_read_cs_reg(const struct spi_device *spi, int idx)
+{
+	struct omap2_mcspi_cs	*cs = spi->controller_state;
+
+	return readl_relaxed(cs->base + idx);
+}
+
+static inline u32 mcspi_cached_chconf0(const struct spi_device *spi)
+{
+	struct omap2_mcspi_cs *cs = spi->controller_state;
+
+	return cs->chconf0;
+}
+
+static inline void mcspi_write_chconf0(const struct spi_device *spi, u32 val)
+{
+	struct omap2_mcspi_cs *cs = spi->controller_state;
+
+	cs->chconf0 = val;
+	mcspi_write_cs_reg(spi, OMAP2_MCSPI_CHCONF0, val);
+	mcspi_read_cs_reg(spi, OMAP2_MCSPI_CHCONF0);
+}
+
+static inline int mcspi_bytes_per_word(int word_len)
+{
+	if (word_len <= 8)
+		return 1;
+	else if (word_len <= 16)
+		return 2;
+	else /* word_len <= 32 */
+		return 4;
+}
+
+static void omap2_mcspi_set_dma_req(const struct spi_device *spi,
+		int is_read, int enable)
+{
+	u32 l, rw;
+
+	l = mcspi_cached_chconf0(spi);
+
+	if (is_read) /* 1 is read, 0 write */
+		rw = OMAP2_MCSPI_CHCONF_DMAR;
+	else
+		rw = OMAP2_MCSPI_CHCONF_DMAW;
+
+	if (enable)
+		l |= rw;
+	else
+		l &= ~rw;
+
+	mcspi_write_chconf0(spi, l);
+}
+
+static void omap2_mcspi_set_enable(const struct spi_device *spi, int enable)
+{
+	struct omap2_mcspi_cs *cs = spi->controller_state;
+	u32 l;
+
+	l = cs->chctrl0;
+	if (enable)
+		l |= OMAP2_MCSPI_CHCTRL_EN;
+	else
+		l &= ~OMAP2_MCSPI_CHCTRL_EN;
+	cs->chctrl0 = l;
+	mcspi_write_cs_reg(spi, OMAP2_MCSPI_CHCTRL0, cs->chctrl0);
+	/* Flash post-writes */
+	mcspi_read_cs_reg(spi, OMAP2_MCSPI_CHCTRL0);
+}
+
+static void omap2_mcspi_set_cs(struct spi_device *spi, bool enable)
+{
+	struct omap2_mcspi *mcspi = spi_master_get_devdata(spi->master);
+	u32 l;
+
+	/* The controller handles the inverted chip selects
+	 * using the OMAP2_MCSPI_CHCONF_EPOL bit so revert
+	 * the inversion from the core spi_set_cs function.
+	 */
+	if (spi->mode & SPI_CS_HIGH)
+		enable = !enable;
+
+	if (spi->controller_state) {
+		int err = pm_runtime_get_sync(mcspi->dev);
+		if (err < 0) {
+			pm_runtime_put_noidle(mcspi->dev);
+			dev_err(mcspi->dev, "failed to get sync: %d\n", err);
+			return;
+		}
+
+		l = mcspi_cached_chconf0(spi);
+
+		if (enable)
+			l &= ~OMAP2_MCSPI_CHCONF_FORCE;
+		else
+			l |= OMAP2_MCSPI_CHCONF_FORCE;
+
+		mcspi_write_chconf0(spi, l);
+
+		pm_runtime_mark_last_busy(mcspi->dev);
+		pm_runtime_put_autosuspend(mcspi->dev);
+	}
+}
+
+static void omap2_mcspi_set_master_mode(struct spi_master *master)
+{
+	struct omap2_mcspi	*mcspi = spi_master_get_devdata(master);
+	struct omap2_mcspi_regs	*ctx = &mcspi->ctx;
+	u32 l;
+
+	/*
+	 * Setup when switching from (reset default) slave mode
+	 * to single-channel master mode
+	 */
+	l = mcspi_read_reg(master, OMAP2_MCSPI_MODULCTRL);
+	l &= ~(OMAP2_MCSPI_MODULCTRL_STEST | OMAP2_MCSPI_MODULCTRL_MS);
+	l |= OMAP2_MCSPI_MODULCTRL_SINGLE;
+	mcspi_write_reg(master, OMAP2_MCSPI_MODULCTRL, l);
+
+	ctx->modulctrl = l;
+}
+
+static void omap2_mcspi_set_fifo(const struct spi_device *spi,
+				struct spi_transfer *t, int enable)
+{
+	struct spi_master *master = spi->master;
+	struct omap2_mcspi_cs *cs = spi->controller_state;
+	struct omap2_mcspi *mcspi;
+	unsigned int wcnt;
+	int max_fifo_depth, fifo_depth, bytes_per_word;
+	u32 chconf, xferlevel;
+
+	mcspi = spi_master_get_devdata(master);
+
+	chconf = mcspi_cached_chconf0(spi);
+	if (enable) {
+		bytes_per_word = mcspi_bytes_per_word(cs->word_len);
+		if (t->len % bytes_per_word != 0)
+			goto disable_fifo;
+
+		if (t->rx_buf != NULL && t->tx_buf != NULL)
+			max_fifo_depth = OMAP2_MCSPI_MAX_FIFODEPTH / 2;
+		else
+			max_fifo_depth = OMAP2_MCSPI_MAX_FIFODEPTH;
+
+		fifo_depth = gcd(t->len, max_fifo_depth);
+		if (fifo_depth < 2 || fifo_depth % bytes_per_word != 0)
+			goto disable_fifo;
+
+		wcnt = t->len / bytes_per_word;
+		if (wcnt > OMAP2_MCSPI_MAX_FIFOWCNT)
+			goto disable_fifo;
+
+		xferlevel = wcnt << 16;
+		if (t->rx_buf != NULL) {
+			chconf |= OMAP2_MCSPI_CHCONF_FFER;
+			xferlevel |= (fifo_depth - 1) << 8;
+		}
+		if (t->tx_buf != NULL) {
+			chconf |= OMAP2_MCSPI_CHCONF_FFET;
+			xferlevel |= fifo_depth - 1;
+		}
+
+		mcspi_write_reg(master, OMAP2_MCSPI_XFERLEVEL, xferlevel);
+		mcspi_write_chconf0(spi, chconf);
+		mcspi->fifo_depth = fifo_depth;
+
+		return;
+	}
+
+disable_fifo:
+	if (t->rx_buf != NULL)
+		chconf &= ~OMAP2_MCSPI_CHCONF_FFER;
+
+	if (t->tx_buf != NULL)
+		chconf &= ~OMAP2_MCSPI_CHCONF_FFET;
+
+	mcspi_write_chconf0(spi, chconf);
+	mcspi->fifo_depth = 0;
+}
+
+static int mcspi_wait_for_reg_bit(void __iomem *reg, unsigned long bit)
+{
+	unsigned long timeout;
+
+	timeout = jiffies + msecs_to_jiffies(1000);
+	while (!(readl_relaxed(reg) & bit)) {
+		if (time_after(jiffies, timeout)) {
+			if (!(readl_relaxed(reg) & bit))
+				return -ETIMEDOUT;
+			else
+				return 0;
+		}
+		cpu_relax();
+	}
+	return 0;
+}
+
+static void omap2_mcspi_rx_callback(void *data)
+{
+	struct spi_device *spi = data;
+	struct omap2_mcspi *mcspi = spi_master_get_devdata(spi->master);
+	struct omap2_mcspi_dma *mcspi_dma = &mcspi->dma_channels[spi->chip_select];
+
+	/* We must disable the DMA RX request */
+	omap2_mcspi_set_dma_req(spi, 1, 0);
+
+	complete(&mcspi_dma->dma_rx_completion);
+}
+
+static void omap2_mcspi_tx_callback(void *data)
+{
+	struct spi_device *spi = data;
+	struct omap2_mcspi *mcspi = spi_master_get_devdata(spi->master);
+	struct omap2_mcspi_dma *mcspi_dma = &mcspi->dma_channels[spi->chip_select];
+
+	/* We must disable the DMA TX request */
+	omap2_mcspi_set_dma_req(spi, 0, 0);
+
+	complete(&mcspi_dma->dma_tx_completion);
+}
+
+static void omap2_mcspi_tx_dma(struct spi_device *spi,
+				struct spi_transfer *xfer,
+				struct dma_slave_config cfg)
+{
+	struct omap2_mcspi	*mcspi;
+	struct omap2_mcspi_dma  *mcspi_dma;
+
+	mcspi = spi_master_get_devdata(spi->master);
+	mcspi_dma = &mcspi->dma_channels[spi->chip_select];
+
+	if (mcspi_dma->dma_tx) {
+		struct dma_async_tx_descriptor *tx;
+
+		dmaengine_slave_config(mcspi_dma->dma_tx, &cfg);
+
+		tx = dmaengine_prep_slave_sg(mcspi_dma->dma_tx, xfer->tx_sg.sgl,
+					     xfer->tx_sg.nents,
+					     DMA_MEM_TO_DEV,
+					     DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (tx) {
+			tx->callback = omap2_mcspi_tx_callback;
+			tx->callback_param = spi;
+			dmaengine_submit(tx);
+		} else {
+			/* FIXME: fall back to PIO? */
+		}
+	}
+	dma_async_issue_pending(mcspi_dma->dma_tx);
+	omap2_mcspi_set_dma_req(spi, 0, 1);
+
+}
+
+static unsigned
+omap2_mcspi_rx_dma(struct spi_device *spi, struct spi_transfer *xfer,
+				struct dma_slave_config cfg,
+				unsigned es)
+{
+	struct omap2_mcspi	*mcspi;
+	struct omap2_mcspi_dma  *mcspi_dma;
+	unsigned int		count, transfer_reduction = 0;
+	struct scatterlist	*sg_out[2];
+	int			nb_sizes = 0, out_mapped_nents[2], ret, x;
+	size_t			sizes[2];
+	u32			l;
+	int			elements = 0;
+	int			word_len, element_count;
+	struct omap2_mcspi_cs	*cs = spi->controller_state;
+	void __iomem		*chstat_reg = cs->base + OMAP2_MCSPI_CHSTAT0;
+
+	mcspi = spi_master_get_devdata(spi->master);
+	mcspi_dma = &mcspi->dma_channels[spi->chip_select];
+	count = xfer->len;
+
+	/*
+	 *  In the "End-of-Transfer Procedure" section for DMA RX in OMAP35x TRM
+	 *  it mentions reducing DMA transfer length by one element in master
+	 *  normal mode.
+	 */
+	if (mcspi->fifo_depth == 0)
+		transfer_reduction = es;
+
+	word_len = cs->word_len;
+	l = mcspi_cached_chconf0(spi);
+
+	if (word_len <= 8)
+		element_count = count;
+	else if (word_len <= 16)
+		element_count = count >> 1;
+	else /* word_len <= 32 */
+		element_count = count >> 2;
+
+	if (mcspi_dma->dma_rx) {
+		struct dma_async_tx_descriptor *tx;
+
+		dmaengine_slave_config(mcspi_dma->dma_rx, &cfg);
+
+		/*
+		 *  Reduce DMA transfer length by one more if McSPI is
+		 *  configured in turbo mode.
+		 */
+		if ((l & OMAP2_MCSPI_CHCONF_TURBO) && mcspi->fifo_depth == 0)
+			transfer_reduction += es;
+
+		if (transfer_reduction) {
+			/* Split sgl into two. The second sgl won't be used. */
+			sizes[0] = count - transfer_reduction;
+			sizes[1] = transfer_reduction;
+			nb_sizes = 2;
+		} else {
+			/*
+			 * Don't bother splitting the sgl. This essentially
+			 * clones the original sgl.
+			 */
+			sizes[0] = count;
+			nb_sizes = 1;
+		}
+
+		ret = sg_split(xfer->rx_sg.sgl, xfer->rx_sg.nents,
+			       0, nb_sizes,
+			       sizes,
+			       sg_out, out_mapped_nents,
+			       GFP_KERNEL);
+
+		if (ret < 0) {
+			dev_err(&spi->dev, "sg_split failed\n");
+			return 0;
+		}
+
+		tx = dmaengine_prep_slave_sg(mcspi_dma->dma_rx,
+					     sg_out[0],
+					     out_mapped_nents[0],
+					     DMA_DEV_TO_MEM,
+					     DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (tx) {
+			tx->callback = omap2_mcspi_rx_callback;
+			tx->callback_param = spi;
+			dmaengine_submit(tx);
+		} else {
+				/* FIXME: fall back to PIO? */
+		}
+	}
+
+	dma_async_issue_pending(mcspi_dma->dma_rx);
+	omap2_mcspi_set_dma_req(spi, 1, 1);
+
+	wait_for_completion(&mcspi_dma->dma_rx_completion);
+
+	for (x = 0; x < nb_sizes; x++)
+		kfree(sg_out[x]);
+
+	if (mcspi->fifo_depth > 0)
+		return count;
+
+	/*
+	 *  Due to the DMA transfer length reduction the missing bytes must
+	 *  be read manually to receive all of the expected data.
+	 */
+	omap2_mcspi_set_enable(spi, 0);
+
+	elements = element_count - 1;
+
+	if (l & OMAP2_MCSPI_CHCONF_TURBO) {
+		elements--;
+
+		if (!mcspi_wait_for_reg_bit(chstat_reg,
+					    OMAP2_MCSPI_CHSTAT_RXS)) {
+			u32 w;
+
+			w = mcspi_read_cs_reg(spi, OMAP2_MCSPI_RX0);
+			if (word_len <= 8)
+				((u8 *)xfer->rx_buf)[elements++] = w;
+			else if (word_len <= 16)
+				((u16 *)xfer->rx_buf)[elements++] = w;
+			else /* word_len <= 32 */
+				((u32 *)xfer->rx_buf)[elements++] = w;
+		} else {
+			int bytes_per_word = mcspi_bytes_per_word(word_len);
+			dev_err(&spi->dev, "DMA RX penultimate word empty\n");
+			count -= (bytes_per_word << 1);
+			omap2_mcspi_set_enable(spi, 1);
+			return count;
+		}
+	}
+	if (!mcspi_wait_for_reg_bit(chstat_reg, OMAP2_MCSPI_CHSTAT_RXS)) {
+		u32 w;
+
+		w = mcspi_read_cs_reg(spi, OMAP2_MCSPI_RX0);
+		if (word_len <= 8)
+			((u8 *)xfer->rx_buf)[elements] = w;
+		else if (word_len <= 16)
+			((u16 *)xfer->rx_buf)[elements] = w;
+		else /* word_len <= 32 */
+			((u32 *)xfer->rx_buf)[elements] = w;
+	} else {
+		dev_err(&spi->dev, "DMA RX last word empty\n");
+		count -= mcspi_bytes_per_word(word_len);
+	}
+	omap2_mcspi_set_enable(spi, 1);
+	return count;
+}
+
+static unsigned
+omap2_mcspi_txrx_dma(struct spi_device *spi, struct spi_transfer *xfer)
+{
+	struct omap2_mcspi	*mcspi;
+	struct omap2_mcspi_cs	*cs = spi->controller_state;
+	struct omap2_mcspi_dma  *mcspi_dma;
+	unsigned int		count;
+	u8			*rx;
+	const u8		*tx;
+	struct dma_slave_config	cfg;
+	enum dma_slave_buswidth width;
+	unsigned es;
+	u32			burst;
+	void __iomem		*chstat_reg;
+	void __iomem            *irqstat_reg;
+	int			wait_res;
+
+	mcspi = spi_master_get_devdata(spi->master);
+	mcspi_dma = &mcspi->dma_channels[spi->chip_select];
+
+	if (cs->word_len <= 8) {
+		width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		es = 1;
+	} else if (cs->word_len <= 16) {
+		width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		es = 2;
+	} else {
+		width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		es = 4;
+	}
+
+	count = xfer->len;
+	burst = 1;
+
+	if (mcspi->fifo_depth > 0) {
+		if (count > mcspi->fifo_depth)
+			burst = mcspi->fifo_depth / es;
+		else
+			burst = count / es;
+	}
+
+	memset(&cfg, 0, sizeof(cfg));
+	cfg.src_addr = cs->phys + OMAP2_MCSPI_RX0;
+	cfg.dst_addr = cs->phys + OMAP2_MCSPI_TX0;
+	cfg.src_addr_width = width;
+	cfg.dst_addr_width = width;
+	cfg.src_maxburst = burst;
+	cfg.dst_maxburst = burst;
+
+	rx = xfer->rx_buf;
+	tx = xfer->tx_buf;
+
+	if (tx != NULL)
+		omap2_mcspi_tx_dma(spi, xfer, cfg);
+
+	if (rx != NULL)
+		count = omap2_mcspi_rx_dma(spi, xfer, cfg, es);
+
+	if (tx != NULL) {
+		wait_for_completion(&mcspi_dma->dma_tx_completion);
+
+		if (mcspi->fifo_depth > 0) {
+			irqstat_reg = mcspi->base + OMAP2_MCSPI_IRQSTATUS;
+
+			if (mcspi_wait_for_reg_bit(irqstat_reg,
+						OMAP2_MCSPI_IRQSTATUS_EOW) < 0)
+				dev_err(&spi->dev, "EOW timed out\n");
+
+			mcspi_write_reg(mcspi->master, OMAP2_MCSPI_IRQSTATUS,
+					OMAP2_MCSPI_IRQSTATUS_EOW);
+		}
+
+		/* for TX_ONLY mode, be sure all words have shifted out */
+		if (rx == NULL) {
+			chstat_reg = cs->base + OMAP2_MCSPI_CHSTAT0;
+			if (mcspi->fifo_depth > 0) {
+				wait_res = mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_TXFFE);
+				if (wait_res < 0)
+					dev_err(&spi->dev, "TXFFE timed out\n");
+			} else {
+				wait_res = mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_TXS);
+				if (wait_res < 0)
+					dev_err(&spi->dev, "TXS timed out\n");
+			}
+			if (wait_res >= 0 &&
+				(mcspi_wait_for_reg_bit(chstat_reg,
+					OMAP2_MCSPI_CHSTAT_EOT) < 0))
+				dev_err(&spi->dev, "EOT timed out\n");
+		}
+	}
+	return count;
+}
+
+static unsigned
+omap2_mcspi_txrx_pio(struct spi_device *spi, struct spi_transfer *xfer)
+{
+	struct omap2_mcspi_cs	*cs = spi->controller_state;
+	unsigned int		count, c;
+	u32			l;
+	void __iomem		*base = cs->base;
+	void __iomem		*tx_reg;
+	void __iomem		*rx_reg;
+	void __iomem		*chstat_reg;
+	int			word_len;
+
+	count = xfer->len;
+	c = count;
+	word_len = cs->word_len;
+
+	l = mcspi_cached_chconf0(spi);
+
+	/* We store the pre-calculated register addresses on stack to speed
+	 * up the transfer loop. */
+	tx_reg		= base + OMAP2_MCSPI_TX0;
+	rx_reg		= base + OMAP2_MCSPI_RX0;
+	chstat_reg	= base + OMAP2_MCSPI_CHSTAT0;
+
+	if (c < (word_len>>3))
+		return 0;
+
+	if (word_len <= 8) {
+		u8		*rx;
+		const u8	*tx;
+
+		rx = xfer->rx_buf;
+		tx = xfer->tx_buf;
+
+		do {
+			c -= 1;
+			if (tx != NULL) {
+				if (mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_TXS) < 0) {
+					dev_err(&spi->dev, "TXS timed out\n");
+					goto out;
+				}
+				dev_vdbg(&spi->dev, "write-%d %02x\n",
+						word_len, *tx);
+				writel_relaxed(*tx++, tx_reg);
+			}
+			if (rx != NULL) {
+				if (mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_RXS) < 0) {
+					dev_err(&spi->dev, "RXS timed out\n");
+					goto out;
+				}
+
+				if (c == 1 && tx == NULL &&
+				    (l & OMAP2_MCSPI_CHCONF_TURBO)) {
+					omap2_mcspi_set_enable(spi, 0);
+					*rx++ = readl_relaxed(rx_reg);
+					dev_vdbg(&spi->dev, "read-%d %02x\n",
+						    word_len, *(rx - 1));
+					if (mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_RXS) < 0) {
+						dev_err(&spi->dev,
+							"RXS timed out\n");
+						goto out;
+					}
+					c = 0;
+				} else if (c == 0 && tx == NULL) {
+					omap2_mcspi_set_enable(spi, 0);
+				}
+
+				*rx++ = readl_relaxed(rx_reg);
+				dev_vdbg(&spi->dev, "read-%d %02x\n",
+						word_len, *(rx - 1));
+			}
+		} while (c);
+	} else if (word_len <= 16) {
+		u16		*rx;
+		const u16	*tx;
+
+		rx = xfer->rx_buf;
+		tx = xfer->tx_buf;
+		do {
+			c -= 2;
+			if (tx != NULL) {
+				if (mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_TXS) < 0) {
+					dev_err(&spi->dev, "TXS timed out\n");
+					goto out;
+				}
+				dev_vdbg(&spi->dev, "write-%d %04x\n",
+						word_len, *tx);
+				writel_relaxed(*tx++, tx_reg);
+			}
+			if (rx != NULL) {
+				if (mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_RXS) < 0) {
+					dev_err(&spi->dev, "RXS timed out\n");
+					goto out;
+				}
+
+				if (c == 2 && tx == NULL &&
+				    (l & OMAP2_MCSPI_CHCONF_TURBO)) {
+					omap2_mcspi_set_enable(spi, 0);
+					*rx++ = readl_relaxed(rx_reg);
+					dev_vdbg(&spi->dev, "read-%d %04x\n",
+						    word_len, *(rx - 1));
+					if (mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_RXS) < 0) {
+						dev_err(&spi->dev,
+							"RXS timed out\n");
+						goto out;
+					}
+					c = 0;
+				} else if (c == 0 && tx == NULL) {
+					omap2_mcspi_set_enable(spi, 0);
+				}
+
+				*rx++ = readl_relaxed(rx_reg);
+				dev_vdbg(&spi->dev, "read-%d %04x\n",
+						word_len, *(rx - 1));
+			}
+		} while (c >= 2);
+	} else if (word_len <= 32) {
+		u32		*rx;
+		const u32	*tx;
+
+		rx = xfer->rx_buf;
+		tx = xfer->tx_buf;
+		do {
+			c -= 4;
+			if (tx != NULL) {
+				if (mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_TXS) < 0) {
+					dev_err(&spi->dev, "TXS timed out\n");
+					goto out;
+				}
+				dev_vdbg(&spi->dev, "write-%d %08x\n",
+						word_len, *tx);
+				writel_relaxed(*tx++, tx_reg);
+			}
+			if (rx != NULL) {
+				if (mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_RXS) < 0) {
+					dev_err(&spi->dev, "RXS timed out\n");
+					goto out;
+				}
+
+				if (c == 4 && tx == NULL &&
+				    (l & OMAP2_MCSPI_CHCONF_TURBO)) {
+					omap2_mcspi_set_enable(spi, 0);
+					*rx++ = readl_relaxed(rx_reg);
+					dev_vdbg(&spi->dev, "read-%d %08x\n",
+						    word_len, *(rx - 1));
+					if (mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_RXS) < 0) {
+						dev_err(&spi->dev,
+							"RXS timed out\n");
+						goto out;
+					}
+					c = 0;
+				} else if (c == 0 && tx == NULL) {
+					omap2_mcspi_set_enable(spi, 0);
+				}
+
+				*rx++ = readl_relaxed(rx_reg);
+				dev_vdbg(&spi->dev, "read-%d %08x\n",
+						word_len, *(rx - 1));
+			}
+		} while (c >= 4);
+	}
+
+	/* for TX_ONLY mode, be sure all words have shifted out */
+	if (xfer->rx_buf == NULL) {
+		if (mcspi_wait_for_reg_bit(chstat_reg,
+				OMAP2_MCSPI_CHSTAT_TXS) < 0) {
+			dev_err(&spi->dev, "TXS timed out\n");
+		} else if (mcspi_wait_for_reg_bit(chstat_reg,
+				OMAP2_MCSPI_CHSTAT_EOT) < 0)
+			dev_err(&spi->dev, "EOT timed out\n");
+
+		/* disable chan to purge rx datas received in TX_ONLY transfer,
+		 * otherwise these rx datas will affect the direct following
+		 * RX_ONLY transfer.
+		 */
+		omap2_mcspi_set_enable(spi, 0);
+	}
+out:
+	omap2_mcspi_set_enable(spi, 1);
+	return count - c;
+}
+
+static u32 omap2_mcspi_calc_divisor(u32 speed_hz)
+{
+	u32 div;
+
+	for (div = 0; div < 15; div++)
+		if (speed_hz >= (OMAP2_MCSPI_MAX_FREQ >> div))
+			return div;
+
+	return 15;
+}
+
+/* called only when no transfer is active to this device */
+static int omap2_mcspi_setup_transfer(struct spi_device *spi,
+		struct spi_transfer *t)
+{
+	struct omap2_mcspi_cs *cs = spi->controller_state;
+	struct omap2_mcspi *mcspi;
+	u32 l = 0, clkd = 0, div, extclk = 0, clkg = 0;
+	u8 word_len = spi->bits_per_word;
+	u32 speed_hz = spi->max_speed_hz;
+
+	mcspi = spi_master_get_devdata(spi->master);
+
+	if (t != NULL && t->bits_per_word)
+		word_len = t->bits_per_word;
+
+	cs->word_len = word_len;
+
+	if (t && t->speed_hz)
+		speed_hz = t->speed_hz;
+
+	speed_hz = min_t(u32, speed_hz, OMAP2_MCSPI_MAX_FREQ);
+	if (speed_hz < (OMAP2_MCSPI_MAX_FREQ / OMAP2_MCSPI_MAX_DIVIDER)) {
+		clkd = omap2_mcspi_calc_divisor(speed_hz);
+		speed_hz = OMAP2_MCSPI_MAX_FREQ >> clkd;
+		clkg = 0;
+	} else {
+		div = (OMAP2_MCSPI_MAX_FREQ + speed_hz - 1) / speed_hz;
+		speed_hz = OMAP2_MCSPI_MAX_FREQ / div;
+		clkd = (div - 1) & 0xf;
+		extclk = (div - 1) >> 4;
+		clkg = OMAP2_MCSPI_CHCONF_CLKG;
+	}
+
+	l = mcspi_cached_chconf0(spi);
+
+	/* standard 4-wire master mode:  SCK, MOSI/out, MISO/in, nCS
+	 * REVISIT: this controller could support SPI_3WIRE mode.
+	 */
+	if (mcspi->pin_dir == MCSPI_PINDIR_D0_IN_D1_OUT) {
+		l &= ~OMAP2_MCSPI_CHCONF_IS;
+		l &= ~OMAP2_MCSPI_CHCONF_DPE1;
+		l |= OMAP2_MCSPI_CHCONF_DPE0;
+	} else {
+		l |= OMAP2_MCSPI_CHCONF_IS;
+		l |= OMAP2_MCSPI_CHCONF_DPE1;
+		l &= ~OMAP2_MCSPI_CHCONF_DPE0;
+	}
+
+	/* wordlength */
+	l &= ~OMAP2_MCSPI_CHCONF_WL_MASK;
+	l |= (word_len - 1) << 7;
+
+	/* set chipselect polarity; manage with FORCE */
+	if (!(spi->mode & SPI_CS_HIGH))
+		l |= OMAP2_MCSPI_CHCONF_EPOL;	/* active-low; normal */
+	else
+		l &= ~OMAP2_MCSPI_CHCONF_EPOL;
+
+	/* set clock divisor */
+	l &= ~OMAP2_MCSPI_CHCONF_CLKD_MASK;
+	l |= clkd << 2;
+
+	/* set clock granularity */
+	l &= ~OMAP2_MCSPI_CHCONF_CLKG;
+	l |= clkg;
+	if (clkg) {
+		cs->chctrl0 &= ~OMAP2_MCSPI_CHCTRL_EXTCLK_MASK;
+		cs->chctrl0 |= extclk << 8;
+		mcspi_write_cs_reg(spi, OMAP2_MCSPI_CHCTRL0, cs->chctrl0);
+	}
+
+	/* set SPI mode 0..3 */
+	if (spi->mode & SPI_CPOL)
+		l |= OMAP2_MCSPI_CHCONF_POL;
+	else
+		l &= ~OMAP2_MCSPI_CHCONF_POL;
+	if (spi->mode & SPI_CPHA)
+		l |= OMAP2_MCSPI_CHCONF_PHA;
+	else
+		l &= ~OMAP2_MCSPI_CHCONF_PHA;
+
+	mcspi_write_chconf0(spi, l);
+
+	cs->mode = spi->mode;
+
+	dev_dbg(&spi->dev, "setup: speed %d, sample %s edge, clk %s\n",
+			speed_hz,
+			(spi->mode & SPI_CPHA) ? "trailing" : "leading",
+			(spi->mode & SPI_CPOL) ? "inverted" : "normal");
+
+	return 0;
+}
+
+/*
+ * Note that we currently allow DMA only if we get a channel
+ * for both rx and tx. Otherwise we'll do PIO for both rx and tx.
+ */
+static int omap2_mcspi_request_dma(struct spi_device *spi)
+{
+	struct spi_master	*master = spi->master;
+	struct omap2_mcspi	*mcspi;
+	struct omap2_mcspi_dma	*mcspi_dma;
+	int ret = 0;
+
+	mcspi = spi_master_get_devdata(master);
+	mcspi_dma = mcspi->dma_channels + spi->chip_select;
+
+	init_completion(&mcspi_dma->dma_rx_completion);
+	init_completion(&mcspi_dma->dma_tx_completion);
+
+	mcspi_dma->dma_rx = dma_request_chan(&master->dev,
+					     mcspi_dma->dma_rx_ch_name);
+	if (IS_ERR(mcspi_dma->dma_rx)) {
+		ret = PTR_ERR(mcspi_dma->dma_rx);
+		mcspi_dma->dma_rx = NULL;
+		goto no_dma;
+	}
+
+	mcspi_dma->dma_tx = dma_request_chan(&master->dev,
+					     mcspi_dma->dma_tx_ch_name);
+	if (IS_ERR(mcspi_dma->dma_tx)) {
+		ret = PTR_ERR(mcspi_dma->dma_tx);
+		mcspi_dma->dma_tx = NULL;
+		dma_release_channel(mcspi_dma->dma_rx);
+		mcspi_dma->dma_rx = NULL;
+	}
+
+no_dma:
+	return ret;
+}
+
+static int omap2_mcspi_setup(struct spi_device *spi)
+{
+	int			ret;
+	struct omap2_mcspi	*mcspi = spi_master_get_devdata(spi->master);
+	struct omap2_mcspi_regs	*ctx = &mcspi->ctx;
+	struct omap2_mcspi_dma	*mcspi_dma;
+	struct omap2_mcspi_cs	*cs = spi->controller_state;
+
+	mcspi_dma = &mcspi->dma_channels[spi->chip_select];
+
+	if (!cs) {
+		cs = kzalloc(sizeof *cs, GFP_KERNEL);
+		if (!cs)
+			return -ENOMEM;
+		cs->base = mcspi->base + spi->chip_select * 0x14;
+		cs->phys = mcspi->phys + spi->chip_select * 0x14;
+		cs->mode = 0;
+		cs->chconf0 = 0;
+		cs->chctrl0 = 0;
+		spi->controller_state = cs;
+		/* Link this to context save list */
+		list_add_tail(&cs->node, &ctx->cs);
+
+		if (gpio_is_valid(spi->cs_gpio)) {
+			ret = gpio_request(spi->cs_gpio, dev_name(&spi->dev));
+			if (ret) {
+				dev_err(&spi->dev, "failed to request gpio\n");
+				return ret;
+			}
+			gpio_direction_output(spi->cs_gpio,
+					 !(spi->mode & SPI_CS_HIGH));
+		}
+	}
+
+	if (!mcspi_dma->dma_rx || !mcspi_dma->dma_tx) {
+		ret = omap2_mcspi_request_dma(spi);
+		if (ret)
+			dev_warn(&spi->dev, "not using DMA for McSPI (%d)\n",
+				 ret);
+	}
+
+	ret = pm_runtime_get_sync(mcspi->dev);
+	if (ret < 0) {
+		pm_runtime_put_noidle(mcspi->dev);
+
+		return ret;
+	}
+
+	ret = omap2_mcspi_setup_transfer(spi, NULL);
+	pm_runtime_mark_last_busy(mcspi->dev);
+	pm_runtime_put_autosuspend(mcspi->dev);
+
+	return ret;
+}
+
+static void omap2_mcspi_cleanup(struct spi_device *spi)
+{
+	struct omap2_mcspi	*mcspi;
+	struct omap2_mcspi_dma	*mcspi_dma;
+	struct omap2_mcspi_cs	*cs;
+
+	mcspi = spi_master_get_devdata(spi->master);
+
+	if (spi->controller_state) {
+		/* Unlink controller state from context save list */
+		cs = spi->controller_state;
+		list_del(&cs->node);
+
+		kfree(cs);
+	}
+
+	if (spi->chip_select < spi->master->num_chipselect) {
+		mcspi_dma = &mcspi->dma_channels[spi->chip_select];
+
+		if (mcspi_dma->dma_rx) {
+			dma_release_channel(mcspi_dma->dma_rx);
+			mcspi_dma->dma_rx = NULL;
+		}
+		if (mcspi_dma->dma_tx) {
+			dma_release_channel(mcspi_dma->dma_tx);
+			mcspi_dma->dma_tx = NULL;
+		}
+	}
+
+	if (gpio_is_valid(spi->cs_gpio))
+		gpio_free(spi->cs_gpio);
+}
+
+static int omap2_mcspi_transfer_one(struct spi_master *master,
+				    struct spi_device *spi,
+				    struct spi_transfer *t)
+{
+
+	/* We only enable one channel at a time -- the one whose message is
+	 * -- although this controller would gladly
+	 * arbitrate among multiple channels.  This corresponds to "single
+	 * channel" master mode.  As a side effect, we need to manage the
+	 * chipselect with the FORCE bit ... CS != channel enable.
+	 */
+
+	struct omap2_mcspi		*mcspi;
+	struct omap2_mcspi_dma		*mcspi_dma;
+	struct omap2_mcspi_cs		*cs;
+	struct omap2_mcspi_device_config *cd;
+	int				par_override = 0;
+	int				status = 0;
+	u32				chconf;
+
+	mcspi = spi_master_get_devdata(master);
+	mcspi_dma = mcspi->dma_channels + spi->chip_select;
+	cs = spi->controller_state;
+	cd = spi->controller_data;
+
+	/*
+	 * The slave driver could have changed spi->mode in which case
+	 * it will be different from cs->mode (the current hardware setup).
+	 * If so, set par_override (even though its not a parity issue) so
+	 * omap2_mcspi_setup_transfer will be called to configure the hardware
+	 * with the correct mode on the first iteration of the loop below.
+	 */
+	if (spi->mode != cs->mode)
+		par_override = 1;
+
+	omap2_mcspi_set_enable(spi, 0);
+
+	if (gpio_is_valid(spi->cs_gpio))
+		omap2_mcspi_set_cs(spi, spi->mode & SPI_CS_HIGH);
+
+	if (par_override ||
+	    (t->speed_hz != spi->max_speed_hz) ||
+	    (t->bits_per_word != spi->bits_per_word)) {
+		par_override = 1;
+		status = omap2_mcspi_setup_transfer(spi, t);
+		if (status < 0)
+			goto out;
+		if (t->speed_hz == spi->max_speed_hz &&
+		    t->bits_per_word == spi->bits_per_word)
+			par_override = 0;
+	}
+	if (cd && cd->cs_per_word) {
+		chconf = mcspi->ctx.modulctrl;
+		chconf &= ~OMAP2_MCSPI_MODULCTRL_SINGLE;
+		mcspi_write_reg(master, OMAP2_MCSPI_MODULCTRL, chconf);
+		mcspi->ctx.modulctrl =
+			mcspi_read_cs_reg(spi, OMAP2_MCSPI_MODULCTRL);
+	}
+
+	chconf = mcspi_cached_chconf0(spi);
+	chconf &= ~OMAP2_MCSPI_CHCONF_TRM_MASK;
+	chconf &= ~OMAP2_MCSPI_CHCONF_TURBO;
+
+	if (t->tx_buf == NULL)
+		chconf |= OMAP2_MCSPI_CHCONF_TRM_RX_ONLY;
+	else if (t->rx_buf == NULL)
+		chconf |= OMAP2_MCSPI_CHCONF_TRM_TX_ONLY;
+
+	if (cd && cd->turbo_mode && t->tx_buf == NULL) {
+		/* Turbo mode is for more than one word */
+		if (t->len > ((cs->word_len + 7) >> 3))
+			chconf |= OMAP2_MCSPI_CHCONF_TURBO;
+	}
+
+	mcspi_write_chconf0(spi, chconf);
+
+	if (t->len) {
+		unsigned	count;
+
+		if ((mcspi_dma->dma_rx && mcspi_dma->dma_tx) &&
+		    master->cur_msg_mapped &&
+		    master->can_dma(master, spi, t))
+			omap2_mcspi_set_fifo(spi, t, 1);
+
+		omap2_mcspi_set_enable(spi, 1);
+
+		/* RX_ONLY mode needs dummy data in TX reg */
+		if (t->tx_buf == NULL)
+			writel_relaxed(0, cs->base
+					+ OMAP2_MCSPI_TX0);
+
+		if ((mcspi_dma->dma_rx && mcspi_dma->dma_tx) &&
+		    master->cur_msg_mapped &&
+		    master->can_dma(master, spi, t))
+			count = omap2_mcspi_txrx_dma(spi, t);
+		else
+			count = omap2_mcspi_txrx_pio(spi, t);
+
+		if (count != t->len) {
+			status = -EIO;
+			goto out;
+		}
+	}
+
+	omap2_mcspi_set_enable(spi, 0);
+
+	if (mcspi->fifo_depth > 0)
+		omap2_mcspi_set_fifo(spi, t, 0);
+
+out:
+	/* Restore defaults if they were overriden */
+	if (par_override) {
+		par_override = 0;
+		status = omap2_mcspi_setup_transfer(spi, NULL);
+	}
+
+	if (cd && cd->cs_per_word) {
+		chconf = mcspi->ctx.modulctrl;
+		chconf |= OMAP2_MCSPI_MODULCTRL_SINGLE;
+		mcspi_write_reg(master, OMAP2_MCSPI_MODULCTRL, chconf);
+		mcspi->ctx.modulctrl =
+			mcspi_read_cs_reg(spi, OMAP2_MCSPI_MODULCTRL);
+	}
+
+	omap2_mcspi_set_enable(spi, 0);
+
+	if (gpio_is_valid(spi->cs_gpio))
+		omap2_mcspi_set_cs(spi, !(spi->mode & SPI_CS_HIGH));
+
+	if (mcspi->fifo_depth > 0 && t)
+		omap2_mcspi_set_fifo(spi, t, 0);
+
+	return status;
+}
+
+static int omap2_mcspi_prepare_message(struct spi_master *master,
+				       struct spi_message *msg)
+{
+	struct omap2_mcspi	*mcspi = spi_master_get_devdata(master);
+	struct omap2_mcspi_regs	*ctx = &mcspi->ctx;
+	struct omap2_mcspi_cs	*cs;
+
+	/* Only a single channel can have the FORCE bit enabled
+	 * in its chconf0 register.
+	 * Scan all channels and disable them except the current one.
+	 * A FORCE can remain from a last transfer having cs_change enabled
+	 */
+	list_for_each_entry(cs, &ctx->cs, node) {
+		if (msg->spi->controller_state == cs)
+			continue;
+
+		if ((cs->chconf0 & OMAP2_MCSPI_CHCONF_FORCE)) {
+			cs->chconf0 &= ~OMAP2_MCSPI_CHCONF_FORCE;
+			writel_relaxed(cs->chconf0,
+					cs->base + OMAP2_MCSPI_CHCONF0);
+			readl_relaxed(cs->base + OMAP2_MCSPI_CHCONF0);
+		}
+	}
+
+	return 0;
+}
+
+static bool omap2_mcspi_can_dma(struct spi_master *master,
+				struct spi_device *spi,
+				struct spi_transfer *xfer)
+{
+	return (xfer->len >= DMA_MIN_BYTES);
+}
+
+static int omap2_mcspi_master_setup(struct omap2_mcspi *mcspi)
+{
+	struct spi_master	*master = mcspi->master;
+	struct omap2_mcspi_regs	*ctx = &mcspi->ctx;
+	int			ret = 0;
+
+	ret = pm_runtime_get_sync(mcspi->dev);
+	if (ret < 0) {
+		pm_runtime_put_noidle(mcspi->dev);
+
+		return ret;
+	}
+
+	mcspi_write_reg(master, OMAP2_MCSPI_WAKEUPENABLE,
+			OMAP2_MCSPI_WAKEUPENABLE_WKEN);
+	ctx->wakeupenable = OMAP2_MCSPI_WAKEUPENABLE_WKEN;
+
+	omap2_mcspi_set_master_mode(master);
+	pm_runtime_mark_last_busy(mcspi->dev);
+	pm_runtime_put_autosuspend(mcspi->dev);
+	return 0;
+}
+
+/*
+ * When SPI wake up from off-mode, CS is in activate state. If it was in
+ * inactive state when driver was suspend, then force it to inactive state at
+ * wake up.
+ */
+static int omap_mcspi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct omap2_mcspi *mcspi = spi_master_get_devdata(master);
+	struct omap2_mcspi_regs *ctx = &mcspi->ctx;
+	struct omap2_mcspi_cs *cs;
+
+	/* McSPI: context restore */
+	mcspi_write_reg(master, OMAP2_MCSPI_MODULCTRL, ctx->modulctrl);
+	mcspi_write_reg(master, OMAP2_MCSPI_WAKEUPENABLE, ctx->wakeupenable);
+
+	list_for_each_entry(cs, &ctx->cs, node) {
+		/*
+		 * We need to toggle CS state for OMAP take this
+		 * change in account.
+		 */
+		if ((cs->chconf0 & OMAP2_MCSPI_CHCONF_FORCE) == 0) {
+			cs->chconf0 |= OMAP2_MCSPI_CHCONF_FORCE;
+			writel_relaxed(cs->chconf0,
+				       cs->base + OMAP2_MCSPI_CHCONF0);
+			cs->chconf0 &= ~OMAP2_MCSPI_CHCONF_FORCE;
+			writel_relaxed(cs->chconf0,
+				       cs->base + OMAP2_MCSPI_CHCONF0);
+		} else {
+			writel_relaxed(cs->chconf0,
+				       cs->base + OMAP2_MCSPI_CHCONF0);
+		}
+	}
+
+	return 0;
+}
+
+static struct omap2_mcspi_platform_config omap2_pdata = {
+	.regs_offset = 0,
+};
+
+static struct omap2_mcspi_platform_config omap4_pdata = {
+	.regs_offset = OMAP4_MCSPI_REG_OFFSET,
+};
+
+static const struct of_device_id omap_mcspi_of_match[] = {
+	{
+		.compatible = "ti,omap2-mcspi",
+		.data = &omap2_pdata,
+	},
+	{
+		.compatible = "ti,omap4-mcspi",
+		.data = &omap4_pdata,
+	},
+	{ },
+};
+MODULE_DEVICE_TABLE(of, omap_mcspi_of_match);
+
+static int omap2_mcspi_probe(struct platform_device *pdev)
+{
+	struct spi_master	*master;
+	const struct omap2_mcspi_platform_config *pdata;
+	struct omap2_mcspi	*mcspi;
+	struct resource		*r;
+	int			status = 0, i;
+	u32			regs_offset = 0;
+	struct device_node	*node = pdev->dev.of_node;
+	const struct of_device_id *match;
+
+	master = spi_alloc_master(&pdev->dev, sizeof *mcspi);
+	if (master == NULL) {
+		dev_dbg(&pdev->dev, "master allocation failed\n");
+		return -ENOMEM;
+	}
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 32);
+	master->setup = omap2_mcspi_setup;
+	master->auto_runtime_pm = true;
+	master->prepare_message = omap2_mcspi_prepare_message;
+	master->can_dma = omap2_mcspi_can_dma;
+	master->transfer_one = omap2_mcspi_transfer_one;
+	master->set_cs = omap2_mcspi_set_cs;
+	master->cleanup = omap2_mcspi_cleanup;
+	master->dev.of_node = node;
+	master->max_speed_hz = OMAP2_MCSPI_MAX_FREQ;
+	master->min_speed_hz = OMAP2_MCSPI_MAX_FREQ >> 15;
+
+	platform_set_drvdata(pdev, master);
+
+	mcspi = spi_master_get_devdata(master);
+	mcspi->master = master;
+
+	match = of_match_device(omap_mcspi_of_match, &pdev->dev);
+	if (match) {
+		u32 num_cs = 1; /* default number of chipselect */
+		pdata = match->data;
+
+		of_property_read_u32(node, "ti,spi-num-cs", &num_cs);
+		master->num_chipselect = num_cs;
+		if (of_get_property(node, "ti,pindir-d0-out-d1-in", NULL))
+			mcspi->pin_dir = MCSPI_PINDIR_D0_OUT_D1_IN;
+	} else {
+		pdata = dev_get_platdata(&pdev->dev);
+		master->num_chipselect = pdata->num_cs;
+		mcspi->pin_dir = pdata->pin_dir;
+	}
+	regs_offset = pdata->regs_offset;
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	mcspi->base = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(mcspi->base)) {
+		status = PTR_ERR(mcspi->base);
+		goto free_master;
+	}
+	mcspi->phys = r->start + regs_offset;
+	mcspi->base += regs_offset;
+
+	mcspi->dev = &pdev->dev;
+
+	INIT_LIST_HEAD(&mcspi->ctx.cs);
+
+	mcspi->dma_channels = devm_kcalloc(&pdev->dev, master->num_chipselect,
+					   sizeof(struct omap2_mcspi_dma),
+					   GFP_KERNEL);
+	if (mcspi->dma_channels == NULL) {
+		status = -ENOMEM;
+		goto free_master;
+	}
+
+	for (i = 0; i < master->num_chipselect; i++) {
+		sprintf(mcspi->dma_channels[i].dma_rx_ch_name, "rx%d", i);
+		sprintf(mcspi->dma_channels[i].dma_tx_ch_name, "tx%d", i);
+	}
+
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_runtime_set_autosuspend_delay(&pdev->dev, SPI_AUTOSUSPEND_TIMEOUT);
+	pm_runtime_enable(&pdev->dev);
+
+	status = omap2_mcspi_master_setup(mcspi);
+	if (status < 0)
+		goto disable_pm;
+
+	status = devm_spi_register_master(&pdev->dev, master);
+	if (status < 0)
+		goto disable_pm;
+
+	return status;
+
+disable_pm:
+	pm_runtime_dont_use_autosuspend(&pdev->dev);
+	pm_runtime_put_sync(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+free_master:
+	spi_master_put(master);
+	return status;
+}
+
+static int omap2_mcspi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct omap2_mcspi *mcspi = spi_master_get_devdata(master);
+
+	pm_runtime_dont_use_autosuspend(mcspi->dev);
+	pm_runtime_put_sync(mcspi->dev);
+	pm_runtime_disable(&pdev->dev);
+
+	return 0;
+}
+
+/* work with hotplug and coldplug */
+MODULE_ALIAS("platform:omap2_mcspi");
+
+static int __maybe_unused omap2_mcspi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct omap2_mcspi *mcspi = spi_master_get_devdata(master);
+	int error;
+
+	error = pinctrl_pm_select_sleep_state(dev);
+	if (error)
+		dev_warn(mcspi->dev, "%s: failed to set pins: %i\n",
+			 __func__, error);
+
+	error = spi_master_suspend(master);
+	if (error)
+		dev_warn(mcspi->dev, "%s: master suspend failed: %i\n",
+			 __func__, error);
+
+	return pm_runtime_force_suspend(dev);
+}
+
+static int __maybe_unused omap2_mcspi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct omap2_mcspi *mcspi = spi_master_get_devdata(master);
+	int error;
+
+	error = pinctrl_pm_select_default_state(dev);
+	if (error)
+		dev_warn(mcspi->dev, "%s: failed to set pins: %i\n",
+			 __func__, error);
+
+	error = spi_master_resume(master);
+	if (error)
+		dev_warn(mcspi->dev, "%s: master resume failed: %i\n",
+			 __func__, error);
+
+	return pm_runtime_force_resume(dev);
+}
+
+static const struct dev_pm_ops omap2_mcspi_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(omap2_mcspi_suspend,
+				omap2_mcspi_resume)
+	.runtime_resume	= omap_mcspi_runtime_resume,
+};
+
+static struct platform_driver omap2_mcspi_driver = {
+	.driver = {
+		.name =		"omap2_mcspi",
+		.pm =		&omap2_mcspi_pm_ops,
+		.of_match_table = omap_mcspi_of_match,
+	},
+	.probe =	omap2_mcspi_probe,
+	.remove =	omap2_mcspi_remove,
+};
+
+module_platform_driver(omap2_mcspi_driver);
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-orion.c b/drivers/spi/spi-orion.c
new file mode 100644
index 0000000..47ef6b1
--- /dev/null
+++ b/drivers/spi/spi-orion.c
@@ -0,0 +1,855 @@
+/*
+ * Marvell Orion SPI controller driver
+ *
+ * Author: Shadi Ammouri <shadi@marvell.com>
+ * Copyright (C) 2007-2008 Marvell Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/platform_device.h>
+#include <linux/err.h>
+#include <linux/io.h>
+#include <linux/spi/spi.h>
+#include <linux/module.h>
+#include <linux/pm_runtime.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_device.h>
+#include <linux/of_gpio.h>
+#include <linux/clk.h>
+#include <linux/sizes.h>
+#include <linux/gpio.h>
+#include <asm/unaligned.h>
+
+#define DRIVER_NAME			"orion_spi"
+
+/* Runtime PM autosuspend timeout: PM is fairly light on this driver */
+#define SPI_AUTOSUSPEND_TIMEOUT		200
+
+/* Some SoCs using this driver support up to 8 chip selects.
+ * It is up to the implementer to only use the chip selects
+ * that are available.
+ */
+#define ORION_NUM_CHIPSELECTS		8
+
+#define ORION_SPI_WAIT_RDY_MAX_LOOP	2000 /* in usec */
+
+#define ORION_SPI_IF_CTRL_REG		0x00
+#define ORION_SPI_IF_CONFIG_REG		0x04
+#define ORION_SPI_IF_RXLSBF		BIT(14)
+#define ORION_SPI_IF_TXLSBF		BIT(13)
+#define ORION_SPI_DATA_OUT_REG		0x08
+#define ORION_SPI_DATA_IN_REG		0x0c
+#define ORION_SPI_INT_CAUSE_REG		0x10
+#define ORION_SPI_TIMING_PARAMS_REG	0x18
+
+/* Register for the "Direct Mode" */
+#define SPI_DIRECT_WRITE_CONFIG_REG	0x20
+
+#define ORION_SPI_TMISO_SAMPLE_MASK	(0x3 << 6)
+#define ORION_SPI_TMISO_SAMPLE_1	(1 << 6)
+#define ORION_SPI_TMISO_SAMPLE_2	(2 << 6)
+
+#define ORION_SPI_MODE_CPOL		(1 << 11)
+#define ORION_SPI_MODE_CPHA		(1 << 12)
+#define ORION_SPI_IF_8_16_BIT_MODE	(1 << 5)
+#define ORION_SPI_CLK_PRESCALE_MASK	0x1F
+#define ARMADA_SPI_CLK_PRESCALE_MASK	0xDF
+#define ORION_SPI_MODE_MASK		(ORION_SPI_MODE_CPOL | \
+					 ORION_SPI_MODE_CPHA)
+#define ORION_SPI_CS_MASK	0x1C
+#define ORION_SPI_CS_SHIFT	2
+#define ORION_SPI_CS(cs)	((cs << ORION_SPI_CS_SHIFT) & \
+					ORION_SPI_CS_MASK)
+
+enum orion_spi_type {
+	ORION_SPI,
+	ARMADA_SPI,
+};
+
+struct orion_spi_dev {
+	enum orion_spi_type	typ;
+	/*
+	 * min_divisor and max_hz should be exclusive, the only we can
+	 * have both is for managing the armada-370-spi case with old
+	 * device tree
+	 */
+	unsigned long		max_hz;
+	unsigned int		min_divisor;
+	unsigned int		max_divisor;
+	u32			prescale_mask;
+	bool			is_errata_50mhz_ac;
+};
+
+struct orion_direct_acc {
+	void __iomem		*vaddr;
+	u32			size;
+};
+
+struct orion_child_options {
+	struct orion_direct_acc direct_access;
+};
+
+struct orion_spi {
+	struct spi_master	*master;
+	void __iomem		*base;
+	struct clk              *clk;
+	struct clk              *axi_clk;
+	const struct orion_spi_dev *devdata;
+	int			unused_hw_gpio;
+
+	struct orion_child_options	child[ORION_NUM_CHIPSELECTS];
+};
+
+static inline void __iomem *spi_reg(struct orion_spi *orion_spi, u32 reg)
+{
+	return orion_spi->base + reg;
+}
+
+static inline void
+orion_spi_setbits(struct orion_spi *orion_spi, u32 reg, u32 mask)
+{
+	void __iomem *reg_addr = spi_reg(orion_spi, reg);
+	u32 val;
+
+	val = readl(reg_addr);
+	val |= mask;
+	writel(val, reg_addr);
+}
+
+static inline void
+orion_spi_clrbits(struct orion_spi *orion_spi, u32 reg, u32 mask)
+{
+	void __iomem *reg_addr = spi_reg(orion_spi, reg);
+	u32 val;
+
+	val = readl(reg_addr);
+	val &= ~mask;
+	writel(val, reg_addr);
+}
+
+static int orion_spi_baudrate_set(struct spi_device *spi, unsigned int speed)
+{
+	u32 tclk_hz;
+	u32 rate;
+	u32 prescale;
+	u32 reg;
+	struct orion_spi *orion_spi;
+	const struct orion_spi_dev *devdata;
+
+	orion_spi = spi_master_get_devdata(spi->master);
+	devdata = orion_spi->devdata;
+
+	tclk_hz = clk_get_rate(orion_spi->clk);
+
+	if (devdata->typ == ARMADA_SPI) {
+		/*
+		 * Given the core_clk (tclk_hz) and the target rate (speed) we
+		 * determine the best values for SPR (in [0 .. 15]) and SPPR (in
+		 * [0..7]) such that
+		 *
+		 * 	core_clk / (SPR * 2 ** SPPR)
+		 *
+		 * is as big as possible but not bigger than speed.
+		 */
+
+		/* best integer divider: */
+		unsigned divider = DIV_ROUND_UP(tclk_hz, speed);
+		unsigned spr, sppr;
+
+		if (divider < 16) {
+			/* This is the easy case, divider is less than 16 */
+			spr = divider;
+			sppr = 0;
+
+		} else {
+			unsigned two_pow_sppr;
+			/*
+			 * Find the highest bit set in divider. This and the
+			 * three next bits define SPR (apart from rounding).
+			 * SPPR is then the number of zero bits that must be
+			 * appended:
+			 */
+			sppr = fls(divider) - 4;
+
+			/*
+			 * As SPR only has 4 bits, we have to round divider up
+			 * to the next multiple of 2 ** sppr.
+			 */
+			two_pow_sppr = 1 << sppr;
+			divider = (divider + two_pow_sppr - 1) & -two_pow_sppr;
+
+			/*
+			 * recalculate sppr as rounding up divider might have
+			 * increased it enough to change the position of the
+			 * highest set bit. In this case the bit that now
+			 * doesn't make it into SPR is 0, so there is no need to
+			 * round again.
+			 */
+			sppr = fls(divider) - 4;
+			spr = divider >> sppr;
+
+			/*
+			 * Now do range checking. SPR is constructed to have a
+			 * width of 4 bits, so this is fine for sure. So we
+			 * still need to check for sppr to fit into 3 bits:
+			 */
+			if (sppr > 7)
+				return -EINVAL;
+		}
+
+		prescale = ((sppr & 0x6) << 5) | ((sppr & 0x1) << 4) | spr;
+	} else {
+		/*
+		 * the supported rates are: 4,6,8...30
+		 * round up as we look for equal or less speed
+		 */
+		rate = DIV_ROUND_UP(tclk_hz, speed);
+		rate = roundup(rate, 2);
+
+		/* check if requested speed is too small */
+		if (rate > 30)
+			return -EINVAL;
+
+		if (rate < 4)
+			rate = 4;
+
+		/* Convert the rate to SPI clock divisor value.	*/
+		prescale = 0x10 + rate/2;
+	}
+
+	reg = readl(spi_reg(orion_spi, ORION_SPI_IF_CONFIG_REG));
+	reg = ((reg & ~devdata->prescale_mask) | prescale);
+	writel(reg, spi_reg(orion_spi, ORION_SPI_IF_CONFIG_REG));
+
+	return 0;
+}
+
+static void
+orion_spi_mode_set(struct spi_device *spi)
+{
+	u32 reg;
+	struct orion_spi *orion_spi;
+
+	orion_spi = spi_master_get_devdata(spi->master);
+
+	reg = readl(spi_reg(orion_spi, ORION_SPI_IF_CONFIG_REG));
+	reg &= ~ORION_SPI_MODE_MASK;
+	if (spi->mode & SPI_CPOL)
+		reg |= ORION_SPI_MODE_CPOL;
+	if (spi->mode & SPI_CPHA)
+		reg |= ORION_SPI_MODE_CPHA;
+	if (spi->mode & SPI_LSB_FIRST)
+		reg |= ORION_SPI_IF_RXLSBF | ORION_SPI_IF_TXLSBF;
+	else
+		reg &= ~(ORION_SPI_IF_RXLSBF | ORION_SPI_IF_TXLSBF);
+
+	writel(reg, spi_reg(orion_spi, ORION_SPI_IF_CONFIG_REG));
+}
+
+static void
+orion_spi_50mhz_ac_timing_erratum(struct spi_device *spi, unsigned int speed)
+{
+	u32 reg;
+	struct orion_spi *orion_spi;
+
+	orion_spi = spi_master_get_devdata(spi->master);
+
+	/*
+	 * Erratum description: (Erratum NO. FE-9144572) The device
+	 * SPI interface supports frequencies of up to 50 MHz.
+	 * However, due to this erratum, when the device core clock is
+	 * 250 MHz and the SPI interfaces is configured for 50MHz SPI
+	 * clock and CPOL=CPHA=1 there might occur data corruption on
+	 * reads from the SPI device.
+	 * Erratum Workaround:
+	 * Work in one of the following configurations:
+	 * 1. Set CPOL=CPHA=0 in "SPI Interface Configuration
+	 * Register".
+	 * 2. Set TMISO_SAMPLE value to 0x2 in "SPI Timing Parameters 1
+	 * Register" before setting the interface.
+	 */
+	reg = readl(spi_reg(orion_spi, ORION_SPI_TIMING_PARAMS_REG));
+	reg &= ~ORION_SPI_TMISO_SAMPLE_MASK;
+
+	if (clk_get_rate(orion_spi->clk) == 250000000 &&
+			speed == 50000000 && spi->mode & SPI_CPOL &&
+			spi->mode & SPI_CPHA)
+		reg |= ORION_SPI_TMISO_SAMPLE_2;
+	else
+		reg |= ORION_SPI_TMISO_SAMPLE_1; /* This is the default value */
+
+	writel(reg, spi_reg(orion_spi, ORION_SPI_TIMING_PARAMS_REG));
+}
+
+/*
+ * called only when no transfer is active on the bus
+ */
+static int
+orion_spi_setup_transfer(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct orion_spi *orion_spi;
+	unsigned int speed = spi->max_speed_hz;
+	unsigned int bits_per_word = spi->bits_per_word;
+	int	rc;
+
+	orion_spi = spi_master_get_devdata(spi->master);
+
+	if ((t != NULL) && t->speed_hz)
+		speed = t->speed_hz;
+
+	if ((t != NULL) && t->bits_per_word)
+		bits_per_word = t->bits_per_word;
+
+	orion_spi_mode_set(spi);
+
+	if (orion_spi->devdata->is_errata_50mhz_ac)
+		orion_spi_50mhz_ac_timing_erratum(spi, speed);
+
+	rc = orion_spi_baudrate_set(spi, speed);
+	if (rc)
+		return rc;
+
+	if (bits_per_word == 16)
+		orion_spi_setbits(orion_spi, ORION_SPI_IF_CONFIG_REG,
+				  ORION_SPI_IF_8_16_BIT_MODE);
+	else
+		orion_spi_clrbits(orion_spi, ORION_SPI_IF_CONFIG_REG,
+				  ORION_SPI_IF_8_16_BIT_MODE);
+
+	return 0;
+}
+
+static void orion_spi_set_cs(struct spi_device *spi, bool enable)
+{
+	struct orion_spi *orion_spi;
+	int cs;
+
+	orion_spi = spi_master_get_devdata(spi->master);
+
+	if (gpio_is_valid(spi->cs_gpio))
+		cs = orion_spi->unused_hw_gpio;
+	else
+		cs = spi->chip_select;
+
+	orion_spi_clrbits(orion_spi, ORION_SPI_IF_CTRL_REG, ORION_SPI_CS_MASK);
+	orion_spi_setbits(orion_spi, ORION_SPI_IF_CTRL_REG,
+				ORION_SPI_CS(cs));
+
+	/* Chip select logic is inverted from spi_set_cs */
+	if (!enable)
+		orion_spi_setbits(orion_spi, ORION_SPI_IF_CTRL_REG, 0x1);
+	else
+		orion_spi_clrbits(orion_spi, ORION_SPI_IF_CTRL_REG, 0x1);
+}
+
+static inline int orion_spi_wait_till_ready(struct orion_spi *orion_spi)
+{
+	int i;
+
+	for (i = 0; i < ORION_SPI_WAIT_RDY_MAX_LOOP; i++) {
+		if (readl(spi_reg(orion_spi, ORION_SPI_INT_CAUSE_REG)))
+			return 1;
+
+		udelay(1);
+	}
+
+	return -1;
+}
+
+static inline int
+orion_spi_write_read_8bit(struct spi_device *spi,
+			  const u8 **tx_buf, u8 **rx_buf)
+{
+	void __iomem *tx_reg, *rx_reg, *int_reg;
+	struct orion_spi *orion_spi;
+
+	orion_spi = spi_master_get_devdata(spi->master);
+	tx_reg = spi_reg(orion_spi, ORION_SPI_DATA_OUT_REG);
+	rx_reg = spi_reg(orion_spi, ORION_SPI_DATA_IN_REG);
+	int_reg = spi_reg(orion_spi, ORION_SPI_INT_CAUSE_REG);
+
+	/* clear the interrupt cause register */
+	writel(0x0, int_reg);
+
+	if (tx_buf && *tx_buf)
+		writel(*(*tx_buf)++, tx_reg);
+	else
+		writel(0, tx_reg);
+
+	if (orion_spi_wait_till_ready(orion_spi) < 0) {
+		dev_err(&spi->dev, "TXS timed out\n");
+		return -1;
+	}
+
+	if (rx_buf && *rx_buf)
+		*(*rx_buf)++ = readl(rx_reg);
+
+	return 1;
+}
+
+static inline int
+orion_spi_write_read_16bit(struct spi_device *spi,
+			   const u16 **tx_buf, u16 **rx_buf)
+{
+	void __iomem *tx_reg, *rx_reg, *int_reg;
+	struct orion_spi *orion_spi;
+
+	orion_spi = spi_master_get_devdata(spi->master);
+	tx_reg = spi_reg(orion_spi, ORION_SPI_DATA_OUT_REG);
+	rx_reg = spi_reg(orion_spi, ORION_SPI_DATA_IN_REG);
+	int_reg = spi_reg(orion_spi, ORION_SPI_INT_CAUSE_REG);
+
+	/* clear the interrupt cause register */
+	writel(0x0, int_reg);
+
+	if (tx_buf && *tx_buf)
+		writel(__cpu_to_le16(get_unaligned((*tx_buf)++)), tx_reg);
+	else
+		writel(0, tx_reg);
+
+	if (orion_spi_wait_till_ready(orion_spi) < 0) {
+		dev_err(&spi->dev, "TXS timed out\n");
+		return -1;
+	}
+
+	if (rx_buf && *rx_buf)
+		put_unaligned(__le16_to_cpu(readl(rx_reg)), (*rx_buf)++);
+
+	return 1;
+}
+
+static unsigned int
+orion_spi_write_read(struct spi_device *spi, struct spi_transfer *xfer)
+{
+	unsigned int count;
+	int word_len;
+	struct orion_spi *orion_spi;
+	int cs = spi->chip_select;
+
+	word_len = spi->bits_per_word;
+	count = xfer->len;
+
+	orion_spi = spi_master_get_devdata(spi->master);
+
+	/*
+	 * Use SPI direct write mode if base address is available. Otherwise
+	 * fall back to PIO mode for this transfer.
+	 */
+	if ((orion_spi->child[cs].direct_access.vaddr) && (xfer->tx_buf) &&
+	    (word_len == 8)) {
+		unsigned int cnt = count / 4;
+		unsigned int rem = count % 4;
+
+		/*
+		 * Send the TX-data to the SPI device via the direct
+		 * mapped address window
+		 */
+		iowrite32_rep(orion_spi->child[cs].direct_access.vaddr,
+			      xfer->tx_buf, cnt);
+		if (rem) {
+			u32 *buf = (u32 *)xfer->tx_buf;
+
+			iowrite8_rep(orion_spi->child[cs].direct_access.vaddr,
+				     &buf[cnt], rem);
+		}
+
+		return count;
+	}
+
+	if (word_len == 8) {
+		const u8 *tx = xfer->tx_buf;
+		u8 *rx = xfer->rx_buf;
+
+		do {
+			if (orion_spi_write_read_8bit(spi, &tx, &rx) < 0)
+				goto out;
+			count--;
+		} while (count);
+	} else if (word_len == 16) {
+		const u16 *tx = xfer->tx_buf;
+		u16 *rx = xfer->rx_buf;
+
+		do {
+			if (orion_spi_write_read_16bit(spi, &tx, &rx) < 0)
+				goto out;
+			count -= 2;
+		} while (count);
+	}
+
+out:
+	return xfer->len - count;
+}
+
+static int orion_spi_transfer_one(struct spi_master *master,
+					struct spi_device *spi,
+					struct spi_transfer *t)
+{
+	int status = 0;
+
+	status = orion_spi_setup_transfer(spi, t);
+	if (status < 0)
+		return status;
+
+	if (t->len)
+		orion_spi_write_read(spi, t);
+
+	return status;
+}
+
+static int orion_spi_setup(struct spi_device *spi)
+{
+	if (gpio_is_valid(spi->cs_gpio)) {
+		gpio_direction_output(spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
+	}
+	return orion_spi_setup_transfer(spi, NULL);
+}
+
+static int orion_spi_reset(struct orion_spi *orion_spi)
+{
+	/* Verify that the CS is deasserted */
+	orion_spi_clrbits(orion_spi, ORION_SPI_IF_CTRL_REG, 0x1);
+
+	/* Don't deassert CS between the direct mapped SPI transfers */
+	writel(0, spi_reg(orion_spi, SPI_DIRECT_WRITE_CONFIG_REG));
+
+	return 0;
+}
+
+static const struct orion_spi_dev orion_spi_dev_data = {
+	.typ = ORION_SPI,
+	.min_divisor = 4,
+	.max_divisor = 30,
+	.prescale_mask = ORION_SPI_CLK_PRESCALE_MASK,
+};
+
+static const struct orion_spi_dev armada_370_spi_dev_data = {
+	.typ = ARMADA_SPI,
+	.min_divisor = 4,
+	.max_divisor = 1920,
+	.max_hz = 50000000,
+	.prescale_mask = ARMADA_SPI_CLK_PRESCALE_MASK,
+};
+
+static const struct orion_spi_dev armada_xp_spi_dev_data = {
+	.typ = ARMADA_SPI,
+	.max_hz = 50000000,
+	.max_divisor = 1920,
+	.prescale_mask = ARMADA_SPI_CLK_PRESCALE_MASK,
+};
+
+static const struct orion_spi_dev armada_375_spi_dev_data = {
+	.typ = ARMADA_SPI,
+	.min_divisor = 15,
+	.max_divisor = 1920,
+	.prescale_mask = ARMADA_SPI_CLK_PRESCALE_MASK,
+};
+
+static const struct orion_spi_dev armada_380_spi_dev_data = {
+	.typ = ARMADA_SPI,
+	.max_hz = 50000000,
+	.max_divisor = 1920,
+	.prescale_mask = ARMADA_SPI_CLK_PRESCALE_MASK,
+	.is_errata_50mhz_ac = true,
+};
+
+static const struct of_device_id orion_spi_of_match_table[] = {
+	{
+		.compatible = "marvell,orion-spi",
+		.data = &orion_spi_dev_data,
+	},
+	{
+		.compatible = "marvell,armada-370-spi",
+		.data = &armada_370_spi_dev_data,
+	},
+	{
+		.compatible = "marvell,armada-375-spi",
+		.data = &armada_375_spi_dev_data,
+	},
+	{
+		.compatible = "marvell,armada-380-spi",
+		.data = &armada_380_spi_dev_data,
+	},
+	{
+		.compatible = "marvell,armada-390-spi",
+		.data = &armada_xp_spi_dev_data,
+	},
+	{
+		.compatible = "marvell,armada-xp-spi",
+		.data = &armada_xp_spi_dev_data,
+	},
+
+	{}
+};
+MODULE_DEVICE_TABLE(of, orion_spi_of_match_table);
+
+static int orion_spi_probe(struct platform_device *pdev)
+{
+	const struct of_device_id *of_id;
+	const struct orion_spi_dev *devdata;
+	struct spi_master *master;
+	struct orion_spi *spi;
+	struct resource *r;
+	unsigned long tclk_hz;
+	int status = 0;
+	struct device_node *np;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*spi));
+	if (master == NULL) {
+		dev_dbg(&pdev->dev, "master allocation failed\n");
+		return -ENOMEM;
+	}
+
+	if (pdev->id != -1)
+		master->bus_num = pdev->id;
+	if (pdev->dev.of_node) {
+		u32 cell_index;
+
+		if (!of_property_read_u32(pdev->dev.of_node, "cell-index",
+					  &cell_index))
+			master->bus_num = cell_index;
+	}
+
+	/* we support all 4 SPI modes and LSB first option */
+	master->mode_bits = SPI_CPHA | SPI_CPOL | SPI_LSB_FIRST;
+	master->set_cs = orion_spi_set_cs;
+	master->transfer_one = orion_spi_transfer_one;
+	master->num_chipselect = ORION_NUM_CHIPSELECTS;
+	master->setup = orion_spi_setup;
+	master->bits_per_word_mask = SPI_BPW_MASK(8) | SPI_BPW_MASK(16);
+	master->auto_runtime_pm = true;
+	master->flags = SPI_MASTER_GPIO_SS;
+
+	platform_set_drvdata(pdev, master);
+
+	spi = spi_master_get_devdata(master);
+	spi->master = master;
+	spi->unused_hw_gpio = -1;
+
+	of_id = of_match_device(orion_spi_of_match_table, &pdev->dev);
+	devdata = (of_id) ? of_id->data : &orion_spi_dev_data;
+	spi->devdata = devdata;
+
+	spi->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(spi->clk)) {
+		status = PTR_ERR(spi->clk);
+		goto out;
+	}
+
+	status = clk_prepare_enable(spi->clk);
+	if (status)
+		goto out;
+
+	/* The following clock is only used by some SoCs */
+	spi->axi_clk = devm_clk_get(&pdev->dev, "axi");
+	if (IS_ERR(spi->axi_clk) &&
+	    PTR_ERR(spi->axi_clk) == -EPROBE_DEFER) {
+		status = -EPROBE_DEFER;
+		goto out_rel_clk;
+	}
+	if (!IS_ERR(spi->axi_clk))
+		clk_prepare_enable(spi->axi_clk);
+
+	tclk_hz = clk_get_rate(spi->clk);
+
+	/*
+	 * With old device tree, armada-370-spi could be used with
+	 * Armada XP, however for this SoC the maximum frequency is
+	 * 50MHz instead of tclk/4. On Armada 370, tclk cannot be
+	 * higher than 200MHz. So, in order to be able to handle both
+	 * SoCs, we can take the minimum of 50MHz and tclk/4.
+	 */
+	if (of_device_is_compatible(pdev->dev.of_node,
+					"marvell,armada-370-spi"))
+		master->max_speed_hz = min(devdata->max_hz,
+				DIV_ROUND_UP(tclk_hz, devdata->min_divisor));
+	else if (devdata->min_divisor)
+		master->max_speed_hz =
+			DIV_ROUND_UP(tclk_hz, devdata->min_divisor);
+	else
+		master->max_speed_hz = devdata->max_hz;
+	master->min_speed_hz = DIV_ROUND_UP(tclk_hz, devdata->max_divisor);
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	spi->base = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(spi->base)) {
+		status = PTR_ERR(spi->base);
+		goto out_rel_axi_clk;
+	}
+
+	for_each_available_child_of_node(pdev->dev.of_node, np) {
+		u32 cs;
+		int cs_gpio;
+
+		/* Get chip-select number from the "reg" property */
+		status = of_property_read_u32(np, "reg", &cs);
+		if (status) {
+			dev_err(&pdev->dev,
+				"%pOF has no valid 'reg' property (%d)\n",
+				np, status);
+			continue;
+		}
+
+		/*
+		 * Initialize the CS GPIO:
+		 * - properly request the actual GPIO signal
+		 * - de-assert the logical signal so that all GPIO CS lines
+		 *   are inactive when probing for slaves
+		 * - find an unused physical CS which will be driven for any
+		 *   slave which uses a CS GPIO
+		 */
+		cs_gpio = of_get_named_gpio(pdev->dev.of_node, "cs-gpios", cs);
+		if (cs_gpio > 0) {
+			char *gpio_name;
+			int cs_flags;
+
+			if (spi->unused_hw_gpio == -1) {
+				dev_info(&pdev->dev,
+					"Selected unused HW CS#%d for any GPIO CSes\n",
+					cs);
+				spi->unused_hw_gpio = cs;
+			}
+
+			gpio_name = devm_kasprintf(&pdev->dev, GFP_KERNEL,
+					"%s-CS%d", dev_name(&pdev->dev), cs);
+			if (!gpio_name) {
+				status = -ENOMEM;
+				goto out_rel_axi_clk;
+			}
+
+			cs_flags = of_property_read_bool(np, "spi-cs-high") ?
+				GPIOF_OUT_INIT_LOW : GPIOF_OUT_INIT_HIGH;
+			status = devm_gpio_request_one(&pdev->dev, cs_gpio,
+					cs_flags, gpio_name);
+			if (status) {
+				dev_err(&pdev->dev,
+					"Can't request GPIO for CS %d\n", cs);
+				goto out_rel_axi_clk;
+			}
+		}
+
+		/*
+		 * Check if an address is configured for this SPI device. If
+		 * not, the MBus mapping via the 'ranges' property in the 'soc'
+		 * node is not configured and this device should not use the
+		 * direct mode. In this case, just continue with the next
+		 * device.
+		 */
+		status = of_address_to_resource(pdev->dev.of_node, cs + 1, r);
+		if (status)
+			continue;
+
+		/*
+		 * Only map one page for direct access. This is enough for the
+		 * simple TX transfer which only writes to the first word.
+		 * This needs to get extended for the direct SPI-NOR / SPI-NAND
+		 * support, once this gets implemented.
+		 */
+		spi->child[cs].direct_access.vaddr = devm_ioremap(&pdev->dev,
+							    r->start,
+							    PAGE_SIZE);
+		if (!spi->child[cs].direct_access.vaddr) {
+			status = -ENOMEM;
+			goto out_rel_axi_clk;
+		}
+		spi->child[cs].direct_access.size = PAGE_SIZE;
+
+		dev_info(&pdev->dev, "CS%d configured for direct access\n", cs);
+	}
+
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_runtime_set_autosuspend_delay(&pdev->dev, SPI_AUTOSUSPEND_TIMEOUT);
+	pm_runtime_enable(&pdev->dev);
+
+	status = orion_spi_reset(spi);
+	if (status < 0)
+		goto out_rel_pm;
+
+	pm_runtime_mark_last_busy(&pdev->dev);
+	pm_runtime_put_autosuspend(&pdev->dev);
+
+	master->dev.of_node = pdev->dev.of_node;
+	status = spi_register_master(master);
+	if (status < 0)
+		goto out_rel_pm;
+
+	return status;
+
+out_rel_pm:
+	pm_runtime_disable(&pdev->dev);
+out_rel_axi_clk:
+	clk_disable_unprepare(spi->axi_clk);
+out_rel_clk:
+	clk_disable_unprepare(spi->clk);
+out:
+	spi_master_put(master);
+	return status;
+}
+
+
+static int orion_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct orion_spi *spi = spi_master_get_devdata(master);
+
+	pm_runtime_get_sync(&pdev->dev);
+	clk_disable_unprepare(spi->axi_clk);
+	clk_disable_unprepare(spi->clk);
+
+	spi_unregister_master(master);
+	pm_runtime_disable(&pdev->dev);
+
+	return 0;
+}
+
+MODULE_ALIAS("platform:" DRIVER_NAME);
+
+#ifdef CONFIG_PM
+static int orion_spi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct orion_spi *spi = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(spi->axi_clk);
+	clk_disable_unprepare(spi->clk);
+	return 0;
+}
+
+static int orion_spi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct orion_spi *spi = spi_master_get_devdata(master);
+
+	if (!IS_ERR(spi->axi_clk))
+		clk_prepare_enable(spi->axi_clk);
+	return clk_prepare_enable(spi->clk);
+}
+#endif
+
+static const struct dev_pm_ops orion_spi_pm_ops = {
+	SET_RUNTIME_PM_OPS(orion_spi_runtime_suspend,
+			   orion_spi_runtime_resume,
+			   NULL)
+};
+
+static struct platform_driver orion_spi_driver = {
+	.driver = {
+		.name	= DRIVER_NAME,
+		.pm	= &orion_spi_pm_ops,
+		.of_match_table = of_match_ptr(orion_spi_of_match_table),
+	},
+	.probe		= orion_spi_probe,
+	.remove		= orion_spi_remove,
+};
+
+module_platform_driver(orion_spi_driver);
+
+MODULE_DESCRIPTION("Orion SPI driver");
+MODULE_AUTHOR("Shadi Ammouri <shadi@marvell.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-pic32-sqi.c b/drivers/spi/spi-pic32-sqi.c
new file mode 100644
index 0000000..bd1c6b5
--- /dev/null
+++ b/drivers/spi/spi-pic32-sqi.c
@@ -0,0 +1,726 @@
+/*
+ * PIC32 Quad SPI controller driver.
+ *
+ * Purna Chandra Mandal <purna.mandal@microchip.com>
+ * Copyright (c) 2016, Microchip Technology Inc.
+ *
+ * This program is free software; you can distribute it and/or modify it
+ * under the terms of the GNU General Public License (Version 2) as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * for more details.
+ */
+
+#include <linux/clk.h>
+#include <linux/dma-mapping.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/iopoll.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/spi/spi.h>
+
+/* SQI registers */
+#define PESQI_XIP_CONF1_REG	0x00
+#define PESQI_XIP_CONF2_REG	0x04
+#define PESQI_CONF_REG		0x08
+#define PESQI_CTRL_REG		0x0C
+#define PESQI_CLK_CTRL_REG	0x10
+#define PESQI_CMD_THRES_REG	0x14
+#define PESQI_INT_THRES_REG	0x18
+#define PESQI_INT_ENABLE_REG	0x1C
+#define PESQI_INT_STAT_REG	0x20
+#define PESQI_TX_DATA_REG	0x24
+#define PESQI_RX_DATA_REG	0x28
+#define PESQI_STAT1_REG		0x2C
+#define PESQI_STAT2_REG		0x30
+#define PESQI_BD_CTRL_REG	0x34
+#define PESQI_BD_CUR_ADDR_REG	0x38
+#define PESQI_BD_BASE_ADDR_REG	0x40
+#define PESQI_BD_STAT_REG	0x44
+#define PESQI_BD_POLL_CTRL_REG	0x48
+#define PESQI_BD_TX_DMA_STAT_REG	0x4C
+#define PESQI_BD_RX_DMA_STAT_REG	0x50
+#define PESQI_THRES_REG		0x54
+#define PESQI_INT_SIGEN_REG	0x58
+
+/* PESQI_CONF_REG fields */
+#define PESQI_MODE		0x7
+#define  PESQI_MODE_BOOT	0
+#define  PESQI_MODE_PIO		1
+#define  PESQI_MODE_DMA		2
+#define  PESQI_MODE_XIP		3
+#define PESQI_MODE_SHIFT	0
+#define PESQI_CPHA		BIT(3)
+#define PESQI_CPOL		BIT(4)
+#define PESQI_LSBF		BIT(5)
+#define PESQI_RXLATCH		BIT(7)
+#define PESQI_SERMODE		BIT(8)
+#define PESQI_WP_EN		BIT(9)
+#define PESQI_HOLD_EN		BIT(10)
+#define PESQI_BURST_EN		BIT(12)
+#define PESQI_CS_CTRL_HW	BIT(15)
+#define PESQI_SOFT_RESET	BIT(16)
+#define PESQI_LANES_SHIFT	20
+#define  PESQI_SINGLE_LANE	0
+#define  PESQI_DUAL_LANE	1
+#define  PESQI_QUAD_LANE	2
+#define PESQI_CSEN_SHIFT	24
+#define PESQI_EN		BIT(23)
+
+/* PESQI_CLK_CTRL_REG fields */
+#define PESQI_CLK_EN		BIT(0)
+#define PESQI_CLK_STABLE	BIT(1)
+#define PESQI_CLKDIV_SHIFT	8
+#define PESQI_CLKDIV		0xff
+
+/* PESQI_INT_THR/CMD_THR_REG */
+#define PESQI_TXTHR_MASK	0x1f
+#define PESQI_TXTHR_SHIFT	8
+#define PESQI_RXTHR_MASK	0x1f
+#define PESQI_RXTHR_SHIFT	0
+
+/* PESQI_INT_EN/INT_STAT/INT_SIG_EN_REG */
+#define PESQI_TXEMPTY		BIT(0)
+#define PESQI_TXFULL		BIT(1)
+#define PESQI_TXTHR		BIT(2)
+#define PESQI_RXEMPTY		BIT(3)
+#define PESQI_RXFULL		BIT(4)
+#define PESQI_RXTHR		BIT(5)
+#define PESQI_BDDONE		BIT(9)  /* BD processing complete */
+#define PESQI_PKTCOMP		BIT(10) /* packet processing complete */
+#define PESQI_DMAERR		BIT(11) /* error */
+
+/* PESQI_BD_CTRL_REG */
+#define PESQI_DMA_EN		BIT(0) /* enable DMA engine */
+#define PESQI_POLL_EN		BIT(1) /* enable polling */
+#define PESQI_BDP_START		BIT(2) /* start BD processor */
+
+/* PESQI controller buffer descriptor */
+struct buf_desc {
+	u32 bd_ctrl;	/* control */
+	u32 bd_status;	/* reserved */
+	u32 bd_addr;	/* DMA buffer addr */
+	u32 bd_nextp;	/* next item in chain */
+};
+
+/* bd_ctrl */
+#define BD_BUFLEN		0x1ff
+#define BD_CBD_INT_EN		BIT(16)	/* Current BD is processed */
+#define BD_PKT_INT_EN		BIT(17) /* All BDs of PKT processed */
+#define BD_LIFM			BIT(18) /* last data of pkt */
+#define BD_LAST			BIT(19) /* end of list */
+#define BD_DATA_RECV		BIT(20) /* receive data */
+#define BD_DDR			BIT(21) /* DDR mode */
+#define BD_DUAL			BIT(22)	/* Dual SPI */
+#define BD_QUAD			BIT(23) /* Quad SPI */
+#define BD_LSBF			BIT(25)	/* LSB First */
+#define BD_STAT_CHECK		BIT(27) /* Status poll */
+#define BD_DEVSEL_SHIFT		28	/* CS */
+#define BD_CS_DEASSERT		BIT(30) /* de-assert CS after current BD */
+#define BD_EN			BIT(31) /* BD owned by H/W */
+
+/**
+ * struct ring_desc - Representation of SQI ring descriptor
+ * @list:	list element to add to free or used list.
+ * @bd:		PESQI controller buffer descriptor
+ * @bd_dma:	DMA address of PESQI controller buffer descriptor
+ * @xfer_len:	transfer length
+ */
+struct ring_desc {
+	struct list_head list;
+	struct buf_desc *bd;
+	dma_addr_t bd_dma;
+	u32 xfer_len;
+};
+
+/* Global constants */
+#define PESQI_BD_BUF_LEN_MAX	256
+#define PESQI_BD_COUNT		256 /* max 64KB data per spi message */
+
+struct pic32_sqi {
+	void __iomem		*regs;
+	struct clk		*sys_clk;
+	struct clk		*base_clk; /* drives spi clock */
+	struct spi_master	*master;
+	int			irq;
+	struct completion	xfer_done;
+	struct ring_desc	*ring;
+	void			*bd;
+	dma_addr_t		bd_dma;
+	struct list_head	bd_list_free; /* free */
+	struct list_head	bd_list_used; /* allocated */
+	struct spi_device	*cur_spi;
+	u32			cur_speed;
+	u8			cur_mode;
+};
+
+static inline void pic32_setbits(void __iomem *reg, u32 set)
+{
+	writel(readl(reg) | set, reg);
+}
+
+static inline void pic32_clrbits(void __iomem *reg, u32 clr)
+{
+	writel(readl(reg) & ~clr, reg);
+}
+
+static int pic32_sqi_set_clk_rate(struct pic32_sqi *sqi, u32 sck)
+{
+	u32 val, div;
+
+	/* div = base_clk / (2 * spi_clk) */
+	div = clk_get_rate(sqi->base_clk) / (2 * sck);
+	div &= PESQI_CLKDIV;
+
+	val = readl(sqi->regs + PESQI_CLK_CTRL_REG);
+	/* apply new divider */
+	val &= ~(PESQI_CLK_STABLE | (PESQI_CLKDIV << PESQI_CLKDIV_SHIFT));
+	val |= div << PESQI_CLKDIV_SHIFT;
+	writel(val, sqi->regs + PESQI_CLK_CTRL_REG);
+
+	/* wait for stability */
+	return readl_poll_timeout(sqi->regs + PESQI_CLK_CTRL_REG, val,
+				  val & PESQI_CLK_STABLE, 1, 5000);
+}
+
+static inline void pic32_sqi_enable_int(struct pic32_sqi *sqi)
+{
+	u32 mask = PESQI_DMAERR | PESQI_BDDONE | PESQI_PKTCOMP;
+
+	writel(mask, sqi->regs + PESQI_INT_ENABLE_REG);
+	/* INT_SIGEN works as interrupt-gate to INTR line */
+	writel(mask, sqi->regs + PESQI_INT_SIGEN_REG);
+}
+
+static inline void pic32_sqi_disable_int(struct pic32_sqi *sqi)
+{
+	writel(0, sqi->regs + PESQI_INT_ENABLE_REG);
+	writel(0, sqi->regs + PESQI_INT_SIGEN_REG);
+}
+
+static irqreturn_t pic32_sqi_isr(int irq, void *dev_id)
+{
+	struct pic32_sqi *sqi = dev_id;
+	u32 enable, status;
+
+	enable = readl(sqi->regs + PESQI_INT_ENABLE_REG);
+	status = readl(sqi->regs + PESQI_INT_STAT_REG);
+
+	/* check spurious interrupt */
+	if (!status)
+		return IRQ_NONE;
+
+	if (status & PESQI_DMAERR) {
+		enable = 0;
+		goto irq_done;
+	}
+
+	if (status & PESQI_TXTHR)
+		enable &= ~(PESQI_TXTHR | PESQI_TXFULL | PESQI_TXEMPTY);
+
+	if (status & PESQI_RXTHR)
+		enable &= ~(PESQI_RXTHR | PESQI_RXFULL | PESQI_RXEMPTY);
+
+	if (status & PESQI_BDDONE)
+		enable &= ~PESQI_BDDONE;
+
+	/* packet processing completed */
+	if (status & PESQI_PKTCOMP) {
+		/* mask all interrupts */
+		enable = 0;
+		/* complete trasaction */
+		complete(&sqi->xfer_done);
+	}
+
+irq_done:
+	/* interrupts are sticky, so mask when handled */
+	writel(enable, sqi->regs + PESQI_INT_ENABLE_REG);
+
+	return IRQ_HANDLED;
+}
+
+static struct ring_desc *ring_desc_get(struct pic32_sqi *sqi)
+{
+	struct ring_desc *rdesc;
+
+	if (list_empty(&sqi->bd_list_free))
+		return NULL;
+
+	rdesc = list_first_entry(&sqi->bd_list_free, struct ring_desc, list);
+	list_move_tail(&rdesc->list, &sqi->bd_list_used);
+	return rdesc;
+}
+
+static void ring_desc_put(struct pic32_sqi *sqi, struct ring_desc *rdesc)
+{
+	list_move(&rdesc->list, &sqi->bd_list_free);
+}
+
+static int pic32_sqi_one_transfer(struct pic32_sqi *sqi,
+				  struct spi_message *mesg,
+				  struct spi_transfer *xfer)
+{
+	struct spi_device *spi = mesg->spi;
+	struct scatterlist *sg, *sgl;
+	struct ring_desc *rdesc;
+	struct buf_desc *bd;
+	int nents, i;
+	u32 bd_ctrl;
+	u32 nbits;
+
+	/* Device selection */
+	bd_ctrl = spi->chip_select << BD_DEVSEL_SHIFT;
+
+	/* half-duplex: select transfer buffer, direction and lane */
+	if (xfer->rx_buf) {
+		bd_ctrl |= BD_DATA_RECV;
+		nbits = xfer->rx_nbits;
+		sgl = xfer->rx_sg.sgl;
+		nents = xfer->rx_sg.nents;
+	} else {
+		nbits = xfer->tx_nbits;
+		sgl = xfer->tx_sg.sgl;
+		nents = xfer->tx_sg.nents;
+	}
+
+	if (nbits & SPI_NBITS_QUAD)
+		bd_ctrl |= BD_QUAD;
+	else if (nbits & SPI_NBITS_DUAL)
+		bd_ctrl |= BD_DUAL;
+
+	/* LSB first */
+	if (spi->mode & SPI_LSB_FIRST)
+		bd_ctrl |= BD_LSBF;
+
+	/* ownership to hardware */
+	bd_ctrl |= BD_EN;
+
+	for_each_sg(sgl, sg, nents, i) {
+		/* get ring descriptor */
+		rdesc = ring_desc_get(sqi);
+		if (!rdesc)
+			break;
+
+		bd = rdesc->bd;
+
+		/* BD CTRL: length */
+		rdesc->xfer_len = sg_dma_len(sg);
+		bd->bd_ctrl = bd_ctrl;
+		bd->bd_ctrl |= rdesc->xfer_len;
+
+		/* BD STAT */
+		bd->bd_status = 0;
+
+		/* BD BUFFER ADDRESS */
+		bd->bd_addr = sg->dma_address;
+	}
+
+	return 0;
+}
+
+static int pic32_sqi_prepare_hardware(struct spi_master *master)
+{
+	struct pic32_sqi *sqi = spi_master_get_devdata(master);
+
+	/* enable spi interface */
+	pic32_setbits(sqi->regs + PESQI_CONF_REG, PESQI_EN);
+	/* enable spi clk */
+	pic32_setbits(sqi->regs + PESQI_CLK_CTRL_REG, PESQI_CLK_EN);
+
+	return 0;
+}
+
+static bool pic32_sqi_can_dma(struct spi_master *master,
+			      struct spi_device *spi,
+			      struct spi_transfer *x)
+{
+	/* Do DMA irrespective of transfer size */
+	return true;
+}
+
+static int pic32_sqi_one_message(struct spi_master *master,
+				 struct spi_message *msg)
+{
+	struct spi_device *spi = msg->spi;
+	struct ring_desc *rdesc, *next;
+	struct spi_transfer *xfer;
+	struct pic32_sqi *sqi;
+	int ret = 0, mode;
+	unsigned long timeout;
+	u32 val;
+
+	sqi = spi_master_get_devdata(master);
+
+	reinit_completion(&sqi->xfer_done);
+	msg->actual_length = 0;
+
+	/* We can't handle spi_transfer specific "speed_hz", "bits_per_word"
+	 * and "delay_usecs". But spi_device specific speed and mode change
+	 * can be handled at best during spi chip-select switch.
+	 */
+	if (sqi->cur_spi != spi) {
+		/* set spi speed */
+		if (sqi->cur_speed != spi->max_speed_hz) {
+			sqi->cur_speed = spi->max_speed_hz;
+			ret = pic32_sqi_set_clk_rate(sqi, spi->max_speed_hz);
+			if (ret)
+				dev_warn(&spi->dev, "set_clk, %d\n", ret);
+		}
+
+		/* set spi mode */
+		mode = spi->mode & (SPI_MODE_3 | SPI_LSB_FIRST);
+		if (sqi->cur_mode != mode) {
+			val = readl(sqi->regs + PESQI_CONF_REG);
+			val &= ~(PESQI_CPOL | PESQI_CPHA | PESQI_LSBF);
+			if (mode & SPI_CPOL)
+				val |= PESQI_CPOL;
+			if (mode & SPI_LSB_FIRST)
+				val |= PESQI_LSBF;
+			val |= PESQI_CPHA;
+			writel(val, sqi->regs + PESQI_CONF_REG);
+
+			sqi->cur_mode = mode;
+		}
+		sqi->cur_spi = spi;
+	}
+
+	/* prepare hardware desc-list(BD) for transfer(s) */
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		ret = pic32_sqi_one_transfer(sqi, msg, xfer);
+		if (ret) {
+			dev_err(&spi->dev, "xfer %p err\n", xfer);
+			goto xfer_out;
+		}
+	}
+
+	/* BDs are prepared and chained. Now mark LAST_BD, CS_DEASSERT at last
+	 * element of the list.
+	 */
+	rdesc = list_last_entry(&sqi->bd_list_used, struct ring_desc, list);
+	rdesc->bd->bd_ctrl |= BD_LAST | BD_CS_DEASSERT |
+			      BD_LIFM | BD_PKT_INT_EN;
+
+	/* set base address BD list for DMA engine */
+	rdesc = list_first_entry(&sqi->bd_list_used, struct ring_desc, list);
+	writel(rdesc->bd_dma, sqi->regs + PESQI_BD_BASE_ADDR_REG);
+
+	/* enable interrupt */
+	pic32_sqi_enable_int(sqi);
+
+	/* enable DMA engine */
+	val = PESQI_DMA_EN | PESQI_POLL_EN | PESQI_BDP_START;
+	writel(val, sqi->regs + PESQI_BD_CTRL_REG);
+
+	/* wait for xfer completion */
+	timeout = wait_for_completion_timeout(&sqi->xfer_done, 5 * HZ);
+	if (timeout == 0) {
+		dev_err(&sqi->master->dev, "wait timedout/interrupted\n");
+		ret = -ETIMEDOUT;
+		msg->status = ret;
+	} else {
+		/* success */
+		msg->status = 0;
+		ret = 0;
+	}
+
+	/* disable DMA */
+	writel(0, sqi->regs + PESQI_BD_CTRL_REG);
+
+	pic32_sqi_disable_int(sqi);
+
+xfer_out:
+	list_for_each_entry_safe_reverse(rdesc, next,
+					 &sqi->bd_list_used, list) {
+		/* Update total byte transferred */
+		msg->actual_length += rdesc->xfer_len;
+		/* release ring descr */
+		ring_desc_put(sqi, rdesc);
+	}
+	spi_finalize_current_message(spi->master);
+
+	return ret;
+}
+
+static int pic32_sqi_unprepare_hardware(struct spi_master *master)
+{
+	struct pic32_sqi *sqi = spi_master_get_devdata(master);
+
+	/* disable clk */
+	pic32_clrbits(sqi->regs + PESQI_CLK_CTRL_REG, PESQI_CLK_EN);
+	/* disable spi */
+	pic32_clrbits(sqi->regs + PESQI_CONF_REG, PESQI_EN);
+
+	return 0;
+}
+
+static int ring_desc_ring_alloc(struct pic32_sqi *sqi)
+{
+	struct ring_desc *rdesc;
+	struct buf_desc *bd;
+	int i;
+
+	/* allocate coherent DMAable memory for hardware buffer descriptors. */
+	sqi->bd = dma_zalloc_coherent(&sqi->master->dev,
+				      sizeof(*bd) * PESQI_BD_COUNT,
+				      &sqi->bd_dma, GFP_DMA32);
+	if (!sqi->bd) {
+		dev_err(&sqi->master->dev, "failed allocating dma buffer\n");
+		return -ENOMEM;
+	}
+
+	/* allocate software ring descriptors */
+	sqi->ring = kcalloc(PESQI_BD_COUNT, sizeof(*rdesc), GFP_KERNEL);
+	if (!sqi->ring) {
+		dma_free_coherent(&sqi->master->dev,
+				  sizeof(*bd) * PESQI_BD_COUNT,
+				  sqi->bd, sqi->bd_dma);
+		return -ENOMEM;
+	}
+
+	bd = (struct buf_desc *)sqi->bd;
+
+	INIT_LIST_HEAD(&sqi->bd_list_free);
+	INIT_LIST_HEAD(&sqi->bd_list_used);
+
+	/* initialize ring-desc */
+	for (i = 0, rdesc = sqi->ring; i < PESQI_BD_COUNT; i++, rdesc++) {
+		INIT_LIST_HEAD(&rdesc->list);
+		rdesc->bd = &bd[i];
+		rdesc->bd_dma = sqi->bd_dma + (void *)&bd[i] - (void *)bd;
+		list_add_tail(&rdesc->list, &sqi->bd_list_free);
+	}
+
+	/* Prepare BD: chain to next BD(s) */
+	for (i = 0, rdesc = sqi->ring; i < PESQI_BD_COUNT - 1; i++)
+		bd[i].bd_nextp = rdesc[i + 1].bd_dma;
+	bd[PESQI_BD_COUNT - 1].bd_nextp = 0;
+
+	return 0;
+}
+
+static void ring_desc_ring_free(struct pic32_sqi *sqi)
+{
+	dma_free_coherent(&sqi->master->dev,
+			  sizeof(struct buf_desc) * PESQI_BD_COUNT,
+			  sqi->bd, sqi->bd_dma);
+	kfree(sqi->ring);
+}
+
+static void pic32_sqi_hw_init(struct pic32_sqi *sqi)
+{
+	unsigned long flags;
+	u32 val;
+
+	/* Soft-reset of PESQI controller triggers interrupt.
+	 * We are not yet ready to handle them so disable CPU
+	 * interrupt for the time being.
+	 */
+	local_irq_save(flags);
+
+	/* assert soft-reset */
+	writel(PESQI_SOFT_RESET, sqi->regs + PESQI_CONF_REG);
+
+	/* wait until clear */
+	readl_poll_timeout_atomic(sqi->regs + PESQI_CONF_REG, val,
+				  !(val & PESQI_SOFT_RESET), 1, 5000);
+
+	/* disable all interrupts */
+	pic32_sqi_disable_int(sqi);
+
+	/* Now it is safe to enable back CPU interrupt */
+	local_irq_restore(flags);
+
+	/* tx and rx fifo interrupt threshold */
+	val = readl(sqi->regs + PESQI_CMD_THRES_REG);
+	val &= ~(PESQI_TXTHR_MASK << PESQI_TXTHR_SHIFT);
+	val &= ~(PESQI_RXTHR_MASK << PESQI_RXTHR_SHIFT);
+	val |= (1U << PESQI_TXTHR_SHIFT) | (1U << PESQI_RXTHR_SHIFT);
+	writel(val, sqi->regs + PESQI_CMD_THRES_REG);
+
+	val = readl(sqi->regs + PESQI_INT_THRES_REG);
+	val &= ~(PESQI_TXTHR_MASK << PESQI_TXTHR_SHIFT);
+	val &= ~(PESQI_RXTHR_MASK << PESQI_RXTHR_SHIFT);
+	val |= (1U << PESQI_TXTHR_SHIFT) | (1U << PESQI_RXTHR_SHIFT);
+	writel(val, sqi->regs + PESQI_INT_THRES_REG);
+
+	/* default configuration */
+	val = readl(sqi->regs + PESQI_CONF_REG);
+
+	/* set mode: DMA */
+	val &= ~PESQI_MODE;
+	val |= PESQI_MODE_DMA << PESQI_MODE_SHIFT;
+	writel(val, sqi->regs + PESQI_CONF_REG);
+
+	/* DATAEN - SQIID0-ID3 */
+	val |= PESQI_QUAD_LANE << PESQI_LANES_SHIFT;
+
+	/* burst/INCR4 enable */
+	val |= PESQI_BURST_EN;
+
+	/* CSEN - all CS */
+	val |= 3U << PESQI_CSEN_SHIFT;
+	writel(val, sqi->regs + PESQI_CONF_REG);
+
+	/* write poll count */
+	writel(0, sqi->regs + PESQI_BD_POLL_CTRL_REG);
+
+	sqi->cur_speed = 0;
+	sqi->cur_mode = -1;
+}
+
+static int pic32_sqi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct pic32_sqi *sqi;
+	struct resource *reg;
+	int ret;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*sqi));
+	if (!master)
+		return -ENOMEM;
+
+	sqi = spi_master_get_devdata(master);
+	sqi->master = master;
+
+	reg = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	sqi->regs = devm_ioremap_resource(&pdev->dev, reg);
+	if (IS_ERR(sqi->regs)) {
+		ret = PTR_ERR(sqi->regs);
+		goto err_free_master;
+	}
+
+	/* irq */
+	sqi->irq = platform_get_irq(pdev, 0);
+	if (sqi->irq < 0) {
+		dev_err(&pdev->dev, "no irq found\n");
+		ret = sqi->irq;
+		goto err_free_master;
+	}
+
+	/* clocks */
+	sqi->sys_clk = devm_clk_get(&pdev->dev, "reg_ck");
+	if (IS_ERR(sqi->sys_clk)) {
+		ret = PTR_ERR(sqi->sys_clk);
+		dev_err(&pdev->dev, "no sys_clk ?\n");
+		goto err_free_master;
+	}
+
+	sqi->base_clk = devm_clk_get(&pdev->dev, "spi_ck");
+	if (IS_ERR(sqi->base_clk)) {
+		ret = PTR_ERR(sqi->base_clk);
+		dev_err(&pdev->dev, "no base clk ?\n");
+		goto err_free_master;
+	}
+
+	ret = clk_prepare_enable(sqi->sys_clk);
+	if (ret) {
+		dev_err(&pdev->dev, "sys clk enable failed\n");
+		goto err_free_master;
+	}
+
+	ret = clk_prepare_enable(sqi->base_clk);
+	if (ret) {
+		dev_err(&pdev->dev, "base clk enable failed\n");
+		clk_disable_unprepare(sqi->sys_clk);
+		goto err_free_master;
+	}
+
+	init_completion(&sqi->xfer_done);
+
+	/* initialize hardware */
+	pic32_sqi_hw_init(sqi);
+
+	/* allocate buffers & descriptors */
+	ret = ring_desc_ring_alloc(sqi);
+	if (ret) {
+		dev_err(&pdev->dev, "ring alloc failed\n");
+		goto err_disable_clk;
+	}
+
+	/* install irq handlers */
+	ret = request_irq(sqi->irq, pic32_sqi_isr, 0,
+			  dev_name(&pdev->dev), sqi);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "request_irq(%d), failed\n", sqi->irq);
+		goto err_free_ring;
+	}
+
+	/* register master */
+	master->num_chipselect	= 2;
+	master->max_speed_hz	= clk_get_rate(sqi->base_clk);
+	master->dma_alignment	= 32;
+	master->max_dma_len	= PESQI_BD_BUF_LEN_MAX;
+	master->dev.of_node	= of_node_get(pdev->dev.of_node);
+	master->mode_bits	= SPI_MODE_3 | SPI_MODE_0 | SPI_TX_DUAL |
+				  SPI_RX_DUAL | SPI_TX_QUAD | SPI_RX_QUAD;
+	master->flags		= SPI_MASTER_HALF_DUPLEX;
+	master->can_dma		= pic32_sqi_can_dma;
+	master->bits_per_word_mask	= SPI_BPW_RANGE_MASK(8, 32);
+	master->transfer_one_message	= pic32_sqi_one_message;
+	master->prepare_transfer_hardware	= pic32_sqi_prepare_hardware;
+	master->unprepare_transfer_hardware	= pic32_sqi_unprepare_hardware;
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret) {
+		dev_err(&master->dev, "failed registering spi master\n");
+		free_irq(sqi->irq, sqi);
+		goto err_free_ring;
+	}
+
+	platform_set_drvdata(pdev, sqi);
+
+	return 0;
+
+err_free_ring:
+	ring_desc_ring_free(sqi);
+
+err_disable_clk:
+	clk_disable_unprepare(sqi->base_clk);
+	clk_disable_unprepare(sqi->sys_clk);
+
+err_free_master:
+	spi_master_put(master);
+	return ret;
+}
+
+static int pic32_sqi_remove(struct platform_device *pdev)
+{
+	struct pic32_sqi *sqi = platform_get_drvdata(pdev);
+
+	/* release resources */
+	free_irq(sqi->irq, sqi);
+	ring_desc_ring_free(sqi);
+
+	/* disable clk */
+	clk_disable_unprepare(sqi->base_clk);
+	clk_disable_unprepare(sqi->sys_clk);
+
+	return 0;
+}
+
+static const struct of_device_id pic32_sqi_of_ids[] = {
+	{.compatible = "microchip,pic32mzda-sqi",},
+	{},
+};
+MODULE_DEVICE_TABLE(of, pic32_sqi_of_ids);
+
+static struct platform_driver pic32_sqi_driver = {
+	.driver = {
+		.name = "sqi-pic32",
+		.of_match_table = of_match_ptr(pic32_sqi_of_ids),
+	},
+	.probe = pic32_sqi_probe,
+	.remove = pic32_sqi_remove,
+};
+
+module_platform_driver(pic32_sqi_driver);
+
+MODULE_AUTHOR("Purna Chandra Mandal <purna.mandal@microchip.com>");
+MODULE_DESCRIPTION("Microchip SPI driver for PIC32 SQI controller.");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-pic32.c b/drivers/spi/spi-pic32.c
new file mode 100644
index 0000000..f8a45af
--- /dev/null
+++ b/drivers/spi/spi-pic32.c
@@ -0,0 +1,879 @@
+/*
+ * Microchip PIC32 SPI controller driver.
+ *
+ * Purna Chandra Mandal <purna.mandal@microchip.com>
+ * Copyright (c) 2016, Microchip Technology Inc.
+ *
+ * This program is free software; you can distribute it and/or modify it
+ * under the terms of the GNU General Public License (Version 2) as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * for more details.
+ */
+
+#include <linux/clk.h>
+#include <linux/clkdev.h>
+#include <linux/delay.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/highmem.h>
+#include <linux/module.h>
+#include <linux/io.h>
+#include <linux/interrupt.h>
+#include <linux/of.h>
+#include <linux/of_irq.h>
+#include <linux/of_gpio.h>
+#include <linux/of_address.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+
+/* SPI controller registers */
+struct pic32_spi_regs {
+	u32 ctrl;
+	u32 ctrl_clr;
+	u32 ctrl_set;
+	u32 ctrl_inv;
+	u32 status;
+	u32 status_clr;
+	u32 status_set;
+	u32 status_inv;
+	u32 buf;
+	u32 dontuse[3];
+	u32 baud;
+	u32 dontuse2[3];
+	u32 ctrl2;
+	u32 ctrl2_clr;
+	u32 ctrl2_set;
+	u32 ctrl2_inv;
+};
+
+/* Bit fields of SPI Control Register */
+#define CTRL_RX_INT_SHIFT	0  /* Rx interrupt generation */
+#define  RX_FIFO_EMPTY		0
+#define  RX_FIFO_NOT_EMPTY	1 /* not empty */
+#define  RX_FIFO_HALF_FULL	2 /* full by half or more */
+#define  RX_FIFO_FULL		3 /* completely full */
+
+#define CTRL_TX_INT_SHIFT	2  /* TX interrupt generation */
+#define  TX_FIFO_ALL_EMPTY	0 /* completely empty */
+#define  TX_FIFO_EMPTY		1 /* empty */
+#define  TX_FIFO_HALF_EMPTY	2 /* empty by half or more */
+#define  TX_FIFO_NOT_FULL	3 /* atleast one empty */
+
+#define CTRL_MSTEN	BIT(5) /* enable master mode */
+#define CTRL_CKP	BIT(6) /* active low */
+#define CTRL_CKE	BIT(8) /* Tx on falling edge */
+#define CTRL_SMP	BIT(9) /* Rx at middle or end of tx */
+#define CTRL_BPW_MASK	0x03   /* bits per word/sample */
+#define CTRL_BPW_SHIFT	10
+#define  PIC32_BPW_8	0
+#define  PIC32_BPW_16	1
+#define  PIC32_BPW_32	2
+#define CTRL_SIDL	BIT(13) /* sleep when idle */
+#define CTRL_ON		BIT(15) /* enable macro */
+#define CTRL_ENHBUF	BIT(16) /* enable enhanced buffering */
+#define CTRL_MCLKSEL	BIT(23) /* select clock source */
+#define CTRL_MSSEN	BIT(28) /* macro driven /SS */
+#define CTRL_FRMEN	BIT(31) /* enable framing mode */
+
+/* Bit fields of SPI Status Register */
+#define STAT_RF_EMPTY	BIT(5) /* RX Fifo empty */
+#define STAT_RX_OV	BIT(6) /* err, s/w needs to clear */
+#define STAT_TX_UR	BIT(8) /* UR in Framed SPI modes */
+#define STAT_FRM_ERR	BIT(12) /* Multiple Frame Sync pulse */
+#define STAT_TF_LVL_MASK	0x1F
+#define STAT_TF_LVL_SHIFT	16
+#define STAT_RF_LVL_MASK	0x1F
+#define STAT_RF_LVL_SHIFT	24
+
+/* Bit fields of SPI Baud Register */
+#define BAUD_MASK		0x1ff
+
+/* Bit fields of SPI Control2 Register */
+#define CTRL2_TX_UR_EN		BIT(10) /* Enable int on Tx under-run */
+#define CTRL2_RX_OV_EN		BIT(11) /* Enable int on Rx over-run */
+#define CTRL2_FRM_ERR_EN	BIT(12) /* Enable frame err int */
+
+/* Minimum DMA transfer size */
+#define PIC32_DMA_LEN_MIN	64
+
+struct pic32_spi {
+	dma_addr_t		dma_base;
+	struct pic32_spi_regs __iomem *regs;
+	int			fault_irq;
+	int			rx_irq;
+	int			tx_irq;
+	u32			fifo_n_byte; /* FIFO depth in bytes */
+	struct clk		*clk;
+	struct spi_master	*master;
+	/* Current controller setting */
+	u32			speed_hz; /* spi-clk rate */
+	u32			mode;
+	u32			bits_per_word;
+	u32			fifo_n_elm; /* FIFO depth in words */
+#define PIC32F_DMA_PREP		0 /* DMA chnls configured */
+	unsigned long		flags;
+	/* Current transfer state */
+	struct completion	xfer_done;
+	/* PIO transfer specific */
+	const void		*tx;
+	const void		*tx_end;
+	const void		*rx;
+	const void		*rx_end;
+	int			len;
+	void (*rx_fifo)(struct pic32_spi *);
+	void (*tx_fifo)(struct pic32_spi *);
+};
+
+static inline void pic32_spi_enable(struct pic32_spi *pic32s)
+{
+	writel(CTRL_ON | CTRL_SIDL, &pic32s->regs->ctrl_set);
+}
+
+static inline void pic32_spi_disable(struct pic32_spi *pic32s)
+{
+	writel(CTRL_ON | CTRL_SIDL, &pic32s->regs->ctrl_clr);
+
+	/* avoid SPI registers read/write at immediate next CPU clock */
+	ndelay(20);
+}
+
+static void pic32_spi_set_clk_rate(struct pic32_spi *pic32s, u32 spi_ck)
+{
+	u32 div;
+
+	/* div = (clk_in / 2 * spi_ck) - 1 */
+	div = DIV_ROUND_CLOSEST(clk_get_rate(pic32s->clk), 2 * spi_ck) - 1;
+
+	writel(div & BAUD_MASK, &pic32s->regs->baud);
+}
+
+static inline u32 pic32_rx_fifo_level(struct pic32_spi *pic32s)
+{
+	u32 sr = readl(&pic32s->regs->status);
+
+	return (sr >> STAT_RF_LVL_SHIFT) & STAT_RF_LVL_MASK;
+}
+
+static inline u32 pic32_tx_fifo_level(struct pic32_spi *pic32s)
+{
+	u32 sr = readl(&pic32s->regs->status);
+
+	return (sr >> STAT_TF_LVL_SHIFT) & STAT_TF_LVL_MASK;
+}
+
+/* Return the max entries we can fill into tx fifo */
+static u32 pic32_tx_max(struct pic32_spi *pic32s, int n_bytes)
+{
+	u32 tx_left, tx_room, rxtx_gap;
+
+	tx_left = (pic32s->tx_end - pic32s->tx) / n_bytes;
+	tx_room = pic32s->fifo_n_elm - pic32_tx_fifo_level(pic32s);
+
+	/*
+	 * Another concern is about the tx/rx mismatch, we
+	 * though to use (pic32s->fifo_n_byte - rxfl - txfl) as
+	 * one maximum value for tx, but it doesn't cover the
+	 * data which is out of tx/rx fifo and inside the
+	 * shift registers. So a ctrl from sw point of
+	 * view is taken.
+	 */
+	rxtx_gap = ((pic32s->rx_end - pic32s->rx) -
+		    (pic32s->tx_end - pic32s->tx)) / n_bytes;
+	return min3(tx_left, tx_room, (u32)(pic32s->fifo_n_elm - rxtx_gap));
+}
+
+/* Return the max entries we should read out of rx fifo */
+static u32 pic32_rx_max(struct pic32_spi *pic32s, int n_bytes)
+{
+	u32 rx_left = (pic32s->rx_end - pic32s->rx) / n_bytes;
+
+	return min_t(u32, rx_left, pic32_rx_fifo_level(pic32s));
+}
+
+#define BUILD_SPI_FIFO_RW(__name, __type, __bwl)		\
+static void pic32_spi_rx_##__name(struct pic32_spi *pic32s)	\
+{								\
+	__type v;						\
+	u32 mx = pic32_rx_max(pic32s, sizeof(__type));		\
+	for (; mx; mx--) {					\
+		v = read##__bwl(&pic32s->regs->buf);		\
+		if (pic32s->rx_end - pic32s->len)		\
+			*(__type *)(pic32s->rx) = v;		\
+		pic32s->rx += sizeof(__type);			\
+	}							\
+}								\
+								\
+static void pic32_spi_tx_##__name(struct pic32_spi *pic32s)	\
+{								\
+	__type v;						\
+	u32 mx = pic32_tx_max(pic32s, sizeof(__type));		\
+	for (; mx ; mx--) {					\
+		v = (__type)~0U;				\
+		if (pic32s->tx_end - pic32s->len)		\
+			v = *(__type *)(pic32s->tx);		\
+		write##__bwl(v, &pic32s->regs->buf);		\
+		pic32s->tx += sizeof(__type);			\
+	}							\
+}
+
+BUILD_SPI_FIFO_RW(byte, u8, b);
+BUILD_SPI_FIFO_RW(word, u16, w);
+BUILD_SPI_FIFO_RW(dword, u32, l);
+
+static void pic32_err_stop(struct pic32_spi *pic32s, const char *msg)
+{
+	/* disable all interrupts */
+	disable_irq_nosync(pic32s->fault_irq);
+	disable_irq_nosync(pic32s->rx_irq);
+	disable_irq_nosync(pic32s->tx_irq);
+
+	/* Show err message and abort xfer with err */
+	dev_err(&pic32s->master->dev, "%s\n", msg);
+	if (pic32s->master->cur_msg)
+		pic32s->master->cur_msg->status = -EIO;
+	complete(&pic32s->xfer_done);
+}
+
+static irqreturn_t pic32_spi_fault_irq(int irq, void *dev_id)
+{
+	struct pic32_spi *pic32s = dev_id;
+	u32 status;
+
+	status = readl(&pic32s->regs->status);
+
+	/* Error handling */
+	if (status & (STAT_RX_OV | STAT_TX_UR)) {
+		writel(STAT_RX_OV, &pic32s->regs->status_clr);
+		writel(STAT_TX_UR, &pic32s->regs->status_clr);
+		pic32_err_stop(pic32s, "err_irq: fifo ov/ur-run\n");
+		return IRQ_HANDLED;
+	}
+
+	if (status & STAT_FRM_ERR) {
+		pic32_err_stop(pic32s, "err_irq: frame error");
+		return IRQ_HANDLED;
+	}
+
+	if (!pic32s->master->cur_msg) {
+		pic32_err_stop(pic32s, "err_irq: no mesg");
+		return IRQ_NONE;
+	}
+
+	return IRQ_NONE;
+}
+
+static irqreturn_t pic32_spi_rx_irq(int irq, void *dev_id)
+{
+	struct pic32_spi *pic32s = dev_id;
+
+	pic32s->rx_fifo(pic32s);
+
+	/* rx complete ? */
+	if (pic32s->rx_end == pic32s->rx) {
+		/* disable all interrupts */
+		disable_irq_nosync(pic32s->fault_irq);
+		disable_irq_nosync(pic32s->rx_irq);
+
+		/* complete current xfer */
+		complete(&pic32s->xfer_done);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t pic32_spi_tx_irq(int irq, void *dev_id)
+{
+	struct pic32_spi *pic32s = dev_id;
+
+	pic32s->tx_fifo(pic32s);
+
+	/* tx complete? disable tx interrupt */
+	if (pic32s->tx_end == pic32s->tx)
+		disable_irq_nosync(pic32s->tx_irq);
+
+	return IRQ_HANDLED;
+}
+
+static void pic32_spi_dma_rx_notify(void *data)
+{
+	struct pic32_spi *pic32s = data;
+
+	complete(&pic32s->xfer_done);
+}
+
+static int pic32_spi_dma_transfer(struct pic32_spi *pic32s,
+				  struct spi_transfer *xfer)
+{
+	struct spi_master *master = pic32s->master;
+	struct dma_async_tx_descriptor *desc_rx;
+	struct dma_async_tx_descriptor *desc_tx;
+	dma_cookie_t cookie;
+	int ret;
+
+	if (!master->dma_rx || !master->dma_tx)
+		return -ENODEV;
+
+	desc_rx = dmaengine_prep_slave_sg(master->dma_rx,
+					  xfer->rx_sg.sgl,
+					  xfer->rx_sg.nents,
+					  DMA_FROM_DEVICE,
+					  DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!desc_rx) {
+		ret = -EINVAL;
+		goto err_dma;
+	}
+
+	desc_tx = dmaengine_prep_slave_sg(master->dma_tx,
+					  xfer->tx_sg.sgl,
+					  xfer->tx_sg.nents,
+					  DMA_TO_DEVICE,
+					  DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!desc_tx) {
+		ret = -EINVAL;
+		goto err_dma;
+	}
+
+	/* Put callback on the RX transfer, that should finish last */
+	desc_rx->callback = pic32_spi_dma_rx_notify;
+	desc_rx->callback_param = pic32s;
+
+	cookie = dmaengine_submit(desc_rx);
+	ret = dma_submit_error(cookie);
+	if (ret)
+		goto err_dma;
+
+	cookie = dmaengine_submit(desc_tx);
+	ret = dma_submit_error(cookie);
+	if (ret)
+		goto err_dma_tx;
+
+	dma_async_issue_pending(master->dma_rx);
+	dma_async_issue_pending(master->dma_tx);
+
+	return 0;
+
+err_dma_tx:
+	dmaengine_terminate_all(master->dma_rx);
+err_dma:
+	return ret;
+}
+
+static int pic32_spi_dma_config(struct pic32_spi *pic32s, u32 dma_width)
+{
+	int buf_offset = offsetof(struct pic32_spi_regs, buf);
+	struct spi_master *master = pic32s->master;
+	struct dma_slave_config cfg;
+	int ret;
+
+	cfg.device_fc = true;
+	cfg.src_addr = pic32s->dma_base + buf_offset;
+	cfg.dst_addr = pic32s->dma_base + buf_offset;
+	cfg.src_maxburst = pic32s->fifo_n_elm / 2; /* fill one-half */
+	cfg.dst_maxburst = pic32s->fifo_n_elm / 2; /* drain one-half */
+	cfg.src_addr_width = dma_width;
+	cfg.dst_addr_width = dma_width;
+	/* tx channel */
+	cfg.slave_id = pic32s->tx_irq;
+	cfg.direction = DMA_MEM_TO_DEV;
+	ret = dmaengine_slave_config(master->dma_tx, &cfg);
+	if (ret) {
+		dev_err(&master->dev, "tx channel setup failed\n");
+		return ret;
+	}
+	/* rx channel */
+	cfg.slave_id = pic32s->rx_irq;
+	cfg.direction = DMA_DEV_TO_MEM;
+	ret = dmaengine_slave_config(master->dma_rx, &cfg);
+	if (ret)
+		dev_err(&master->dev, "rx channel setup failed\n");
+
+	return ret;
+}
+
+static int pic32_spi_set_word_size(struct pic32_spi *pic32s, u8 bits_per_word)
+{
+	enum dma_slave_buswidth dmawidth;
+	u32 buswidth, v;
+
+	switch (bits_per_word) {
+	case 8:
+		pic32s->rx_fifo = pic32_spi_rx_byte;
+		pic32s->tx_fifo = pic32_spi_tx_byte;
+		buswidth = PIC32_BPW_8;
+		dmawidth = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		break;
+	case 16:
+		pic32s->rx_fifo = pic32_spi_rx_word;
+		pic32s->tx_fifo = pic32_spi_tx_word;
+		buswidth = PIC32_BPW_16;
+		dmawidth = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		break;
+	case 32:
+		pic32s->rx_fifo = pic32_spi_rx_dword;
+		pic32s->tx_fifo = pic32_spi_tx_dword;
+		buswidth = PIC32_BPW_32;
+		dmawidth = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		break;
+	default:
+		/* not supported */
+		return -EINVAL;
+	}
+
+	/* calculate maximum number of words fifos can hold */
+	pic32s->fifo_n_elm = DIV_ROUND_UP(pic32s->fifo_n_byte,
+					  bits_per_word / 8);
+	/* set word size */
+	v = readl(&pic32s->regs->ctrl);
+	v &= ~(CTRL_BPW_MASK << CTRL_BPW_SHIFT);
+	v |= buswidth << CTRL_BPW_SHIFT;
+	writel(v, &pic32s->regs->ctrl);
+
+	/* re-configure dma width, if required */
+	if (test_bit(PIC32F_DMA_PREP, &pic32s->flags))
+		pic32_spi_dma_config(pic32s, dmawidth);
+
+	return 0;
+}
+
+static int pic32_spi_prepare_hardware(struct spi_master *master)
+{
+	struct pic32_spi *pic32s = spi_master_get_devdata(master);
+
+	pic32_spi_enable(pic32s);
+
+	return 0;
+}
+
+static int pic32_spi_prepare_message(struct spi_master *master,
+				     struct spi_message *msg)
+{
+	struct pic32_spi *pic32s = spi_master_get_devdata(master);
+	struct spi_device *spi = msg->spi;
+	u32 val;
+
+	/* set device specific bits_per_word */
+	if (pic32s->bits_per_word != spi->bits_per_word) {
+		pic32_spi_set_word_size(pic32s, spi->bits_per_word);
+		pic32s->bits_per_word = spi->bits_per_word;
+	}
+
+	/* device specific speed change */
+	if (pic32s->speed_hz != spi->max_speed_hz) {
+		pic32_spi_set_clk_rate(pic32s, spi->max_speed_hz);
+		pic32s->speed_hz = spi->max_speed_hz;
+	}
+
+	/* device specific mode change */
+	if (pic32s->mode != spi->mode) {
+		val = readl(&pic32s->regs->ctrl);
+		/* active low */
+		if (spi->mode & SPI_CPOL)
+			val |= CTRL_CKP;
+		else
+			val &= ~CTRL_CKP;
+		/* tx on rising edge */
+		if (spi->mode & SPI_CPHA)
+			val &= ~CTRL_CKE;
+		else
+			val |= CTRL_CKE;
+
+		/* rx at end of tx */
+		val |= CTRL_SMP;
+		writel(val, &pic32s->regs->ctrl);
+		pic32s->mode = spi->mode;
+	}
+
+	return 0;
+}
+
+static bool pic32_spi_can_dma(struct spi_master *master,
+			      struct spi_device *spi,
+			      struct spi_transfer *xfer)
+{
+	struct pic32_spi *pic32s = spi_master_get_devdata(master);
+
+	/* skip using DMA on small size transfer to avoid overhead.*/
+	return (xfer->len >= PIC32_DMA_LEN_MIN) &&
+	       test_bit(PIC32F_DMA_PREP, &pic32s->flags);
+}
+
+static int pic32_spi_one_transfer(struct spi_master *master,
+				  struct spi_device *spi,
+				  struct spi_transfer *transfer)
+{
+	struct pic32_spi *pic32s;
+	bool dma_issued = false;
+	unsigned long timeout;
+	int ret;
+
+	pic32s = spi_master_get_devdata(master);
+
+	/* handle transfer specific word size change */
+	if (transfer->bits_per_word &&
+	    (transfer->bits_per_word != pic32s->bits_per_word)) {
+		ret = pic32_spi_set_word_size(pic32s, transfer->bits_per_word);
+		if (ret)
+			return ret;
+		pic32s->bits_per_word = transfer->bits_per_word;
+	}
+
+	/* handle transfer specific speed change */
+	if (transfer->speed_hz && (transfer->speed_hz != pic32s->speed_hz)) {
+		pic32_spi_set_clk_rate(pic32s, transfer->speed_hz);
+		pic32s->speed_hz = transfer->speed_hz;
+	}
+
+	reinit_completion(&pic32s->xfer_done);
+
+	/* transact by DMA mode */
+	if (transfer->rx_sg.nents && transfer->tx_sg.nents) {
+		ret = pic32_spi_dma_transfer(pic32s, transfer);
+		if (ret) {
+			dev_err(&spi->dev, "dma submit error\n");
+			return ret;
+		}
+
+		/* DMA issued */
+		dma_issued = true;
+	} else {
+		/* set current transfer information */
+		pic32s->tx = (const void *)transfer->tx_buf;
+		pic32s->rx = (const void *)transfer->rx_buf;
+		pic32s->tx_end = pic32s->tx + transfer->len;
+		pic32s->rx_end = pic32s->rx + transfer->len;
+		pic32s->len = transfer->len;
+
+		/* transact by interrupt driven PIO */
+		enable_irq(pic32s->fault_irq);
+		enable_irq(pic32s->rx_irq);
+		enable_irq(pic32s->tx_irq);
+	}
+
+	/* wait for completion */
+	timeout = wait_for_completion_timeout(&pic32s->xfer_done, 2 * HZ);
+	if (timeout == 0) {
+		dev_err(&spi->dev, "wait error/timedout\n");
+		if (dma_issued) {
+			dmaengine_terminate_all(master->dma_rx);
+			dmaengine_terminate_all(master->dma_rx);
+		}
+		ret = -ETIMEDOUT;
+	} else {
+		ret = 0;
+	}
+
+	return ret;
+}
+
+static int pic32_spi_unprepare_message(struct spi_master *master,
+				       struct spi_message *msg)
+{
+	/* nothing to do */
+	return 0;
+}
+
+static int pic32_spi_unprepare_hardware(struct spi_master *master)
+{
+	struct pic32_spi *pic32s = spi_master_get_devdata(master);
+
+	pic32_spi_disable(pic32s);
+
+	return 0;
+}
+
+/* This may be called multiple times by same spi dev */
+static int pic32_spi_setup(struct spi_device *spi)
+{
+	if (!spi->max_speed_hz) {
+		dev_err(&spi->dev, "No max speed HZ parameter\n");
+		return -EINVAL;
+	}
+
+	/* PIC32 spi controller can drive /CS during transfer depending
+	 * on tx fifo fill-level. /CS will stay asserted as long as TX
+	 * fifo is non-empty, else will be deasserted indicating
+	 * completion of the ongoing transfer. This might result into
+	 * unreliable/erroneous SPI transactions.
+	 * To avoid that we will always handle /CS by toggling GPIO.
+	 */
+	if (!gpio_is_valid(spi->cs_gpio))
+		return -EINVAL;
+
+	gpio_direction_output(spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
+
+	return 0;
+}
+
+static void pic32_spi_cleanup(struct spi_device *spi)
+{
+	/* de-activate cs-gpio */
+	gpio_direction_output(spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
+}
+
+static void pic32_spi_dma_prep(struct pic32_spi *pic32s, struct device *dev)
+{
+	struct spi_master *master = pic32s->master;
+	dma_cap_mask_t mask;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	master->dma_rx = dma_request_slave_channel_compat(mask, NULL, NULL,
+							  dev, "spi-rx");
+	if (!master->dma_rx) {
+		dev_warn(dev, "RX channel not found.\n");
+		goto out_err;
+	}
+
+	master->dma_tx = dma_request_slave_channel_compat(mask, NULL, NULL,
+							  dev, "spi-tx");
+	if (!master->dma_tx) {
+		dev_warn(dev, "TX channel not found.\n");
+		goto out_err;
+	}
+
+	if (pic32_spi_dma_config(pic32s, DMA_SLAVE_BUSWIDTH_1_BYTE))
+		goto out_err;
+
+	/* DMA chnls allocated and prepared */
+	set_bit(PIC32F_DMA_PREP, &pic32s->flags);
+
+	return;
+
+out_err:
+	if (master->dma_rx)
+		dma_release_channel(master->dma_rx);
+
+	if (master->dma_tx)
+		dma_release_channel(master->dma_tx);
+}
+
+static void pic32_spi_dma_unprep(struct pic32_spi *pic32s)
+{
+	if (!test_bit(PIC32F_DMA_PREP, &pic32s->flags))
+		return;
+
+	clear_bit(PIC32F_DMA_PREP, &pic32s->flags);
+	if (pic32s->master->dma_rx)
+		dma_release_channel(pic32s->master->dma_rx);
+
+	if (pic32s->master->dma_tx)
+		dma_release_channel(pic32s->master->dma_tx);
+}
+
+static void pic32_spi_hw_init(struct pic32_spi *pic32s)
+{
+	u32 ctrl;
+
+	/* disable hardware */
+	pic32_spi_disable(pic32s);
+
+	ctrl = readl(&pic32s->regs->ctrl);
+	/* enable enhanced fifo of 128bit deep */
+	ctrl |= CTRL_ENHBUF;
+	pic32s->fifo_n_byte = 16;
+
+	/* disable framing mode */
+	ctrl &= ~CTRL_FRMEN;
+
+	/* enable master mode while disabled */
+	ctrl |= CTRL_MSTEN;
+
+	/* set tx fifo threshold interrupt */
+	ctrl &= ~(0x3 << CTRL_TX_INT_SHIFT);
+	ctrl |= (TX_FIFO_HALF_EMPTY << CTRL_TX_INT_SHIFT);
+
+	/* set rx fifo threshold interrupt */
+	ctrl &= ~(0x3 << CTRL_RX_INT_SHIFT);
+	ctrl |= (RX_FIFO_NOT_EMPTY << CTRL_RX_INT_SHIFT);
+
+	/* select clk source */
+	ctrl &= ~CTRL_MCLKSEL;
+
+	/* set manual /CS mode */
+	ctrl &= ~CTRL_MSSEN;
+
+	writel(ctrl, &pic32s->regs->ctrl);
+
+	/* enable error reporting */
+	ctrl = CTRL2_TX_UR_EN | CTRL2_RX_OV_EN | CTRL2_FRM_ERR_EN;
+	writel(ctrl, &pic32s->regs->ctrl2_set);
+}
+
+static int pic32_spi_hw_probe(struct platform_device *pdev,
+			      struct pic32_spi *pic32s)
+{
+	struct resource *mem;
+	int ret;
+
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	pic32s->regs = devm_ioremap_resource(&pdev->dev, mem);
+	if (IS_ERR(pic32s->regs))
+		return PTR_ERR(pic32s->regs);
+
+	pic32s->dma_base = mem->start;
+
+	/* get irq resources: err-irq, rx-irq, tx-irq */
+	pic32s->fault_irq = platform_get_irq_byname(pdev, "fault");
+	if (pic32s->fault_irq < 0) {
+		dev_err(&pdev->dev, "fault-irq not found\n");
+		return pic32s->fault_irq;
+	}
+
+	pic32s->rx_irq = platform_get_irq_byname(pdev, "rx");
+	if (pic32s->rx_irq < 0) {
+		dev_err(&pdev->dev, "rx-irq not found\n");
+		return pic32s->rx_irq;
+	}
+
+	pic32s->tx_irq = platform_get_irq_byname(pdev, "tx");
+	if (pic32s->tx_irq < 0) {
+		dev_err(&pdev->dev, "tx-irq not found\n");
+		return pic32s->tx_irq;
+	}
+
+	/* get clock */
+	pic32s->clk = devm_clk_get(&pdev->dev, "mck0");
+	if (IS_ERR(pic32s->clk)) {
+		dev_err(&pdev->dev, "clk not found\n");
+		ret = PTR_ERR(pic32s->clk);
+		goto err_unmap_mem;
+	}
+
+	ret = clk_prepare_enable(pic32s->clk);
+	if (ret)
+		goto err_unmap_mem;
+
+	pic32_spi_hw_init(pic32s);
+
+	return 0;
+
+err_unmap_mem:
+	dev_err(&pdev->dev, "%s failed, err %d\n", __func__, ret);
+	return ret;
+}
+
+static int pic32_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct pic32_spi *pic32s;
+	int ret;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*pic32s));
+	if (!master)
+		return -ENOMEM;
+
+	pic32s = spi_master_get_devdata(master);
+	pic32s->master = master;
+
+	ret = pic32_spi_hw_probe(pdev, pic32s);
+	if (ret)
+		goto err_master;
+
+	master->dev.of_node	= of_node_get(pdev->dev.of_node);
+	master->mode_bits	= SPI_MODE_3 | SPI_MODE_0 | SPI_CS_HIGH;
+	master->num_chipselect	= 1; /* single chip-select */
+	master->max_speed_hz	= clk_get_rate(pic32s->clk);
+	master->setup		= pic32_spi_setup;
+	master->cleanup		= pic32_spi_cleanup;
+	master->flags		= SPI_MASTER_MUST_TX | SPI_MASTER_MUST_RX;
+	master->bits_per_word_mask	= SPI_BPW_MASK(8) | SPI_BPW_MASK(16) |
+					  SPI_BPW_MASK(32);
+	master->transfer_one		= pic32_spi_one_transfer;
+	master->prepare_message		= pic32_spi_prepare_message;
+	master->unprepare_message	= pic32_spi_unprepare_message;
+	master->prepare_transfer_hardware	= pic32_spi_prepare_hardware;
+	master->unprepare_transfer_hardware	= pic32_spi_unprepare_hardware;
+
+	/* optional DMA support */
+	pic32_spi_dma_prep(pic32s, &pdev->dev);
+	if (test_bit(PIC32F_DMA_PREP, &pic32s->flags))
+		master->can_dma	= pic32_spi_can_dma;
+
+	init_completion(&pic32s->xfer_done);
+	pic32s->mode = -1;
+
+	/* install irq handlers (with irq-disabled) */
+	irq_set_status_flags(pic32s->fault_irq, IRQ_NOAUTOEN);
+	ret = devm_request_irq(&pdev->dev, pic32s->fault_irq,
+			       pic32_spi_fault_irq, IRQF_NO_THREAD,
+			       dev_name(&pdev->dev), pic32s);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "request fault-irq %d\n", pic32s->rx_irq);
+		goto err_bailout;
+	}
+
+	/* receive interrupt handler */
+	irq_set_status_flags(pic32s->rx_irq, IRQ_NOAUTOEN);
+	ret = devm_request_irq(&pdev->dev, pic32s->rx_irq,
+			       pic32_spi_rx_irq, IRQF_NO_THREAD,
+			       dev_name(&pdev->dev), pic32s);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "request rx-irq %d\n", pic32s->rx_irq);
+		goto err_bailout;
+	}
+
+	/* transmit interrupt handler */
+	irq_set_status_flags(pic32s->tx_irq, IRQ_NOAUTOEN);
+	ret = devm_request_irq(&pdev->dev, pic32s->tx_irq,
+			       pic32_spi_tx_irq, IRQF_NO_THREAD,
+			       dev_name(&pdev->dev), pic32s);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "request tx-irq %d\n", pic32s->tx_irq);
+		goto err_bailout;
+	}
+
+	/* register master */
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret) {
+		dev_err(&master->dev, "failed registering spi master\n");
+		goto err_bailout;
+	}
+
+	platform_set_drvdata(pdev, pic32s);
+
+	return 0;
+
+err_bailout:
+	clk_disable_unprepare(pic32s->clk);
+err_master:
+	spi_master_put(master);
+	return ret;
+}
+
+static int pic32_spi_remove(struct platform_device *pdev)
+{
+	struct pic32_spi *pic32s;
+
+	pic32s = platform_get_drvdata(pdev);
+	pic32_spi_disable(pic32s);
+	clk_disable_unprepare(pic32s->clk);
+	pic32_spi_dma_unprep(pic32s);
+
+	return 0;
+}
+
+static const struct of_device_id pic32_spi_of_match[] = {
+	{.compatible = "microchip,pic32mzda-spi",},
+	{},
+};
+MODULE_DEVICE_TABLE(of, pic32_spi_of_match);
+
+static struct platform_driver pic32_spi_driver = {
+	.driver = {
+		.name = "spi-pic32",
+		.of_match_table = of_match_ptr(pic32_spi_of_match),
+	},
+	.probe = pic32_spi_probe,
+	.remove = pic32_spi_remove,
+};
+
+module_platform_driver(pic32_spi_driver);
+
+MODULE_AUTHOR("Purna Chandra Mandal <purna.mandal@microchip.com>");
+MODULE_DESCRIPTION("Microchip SPI driver for PIC32 SPI controller.");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-pl022.c b/drivers/spi/spi-pl022.c
new file mode 100644
index 0000000..1af8c96
--- /dev/null
+++ b/drivers/spi/spi-pl022.c
@@ -0,0 +1,2501 @@
+/*
+ * A driver for the ARM PL022 PrimeCell SSP/SPI bus master.
+ *
+ * Copyright (C) 2008-2012 ST-Ericsson AB
+ * Copyright (C) 2006 STMicroelectronics Pvt. Ltd.
+ *
+ * Author: Linus Walleij <linus.walleij@stericsson.com>
+ *
+ * Initial version inspired by:
+ *	linux-2.6.17-rc3-mm1/drivers/spi/pxa2xx_spi.c
+ * Initial adoption to PL022 by:
+ *      Sachin Verma <sachin.verma@st.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/ioport.h>
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/spi/spi.h>
+#include <linux/delay.h>
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/amba/bus.h>
+#include <linux/amba/pl022.h>
+#include <linux/io.h>
+#include <linux/slab.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/scatterlist.h>
+#include <linux/pm_runtime.h>
+#include <linux/gpio.h>
+#include <linux/of_gpio.h>
+#include <linux/pinctrl/consumer.h>
+
+/*
+ * This macro is used to define some register default values.
+ * reg is masked with mask, the OR:ed with an (again masked)
+ * val shifted sb steps to the left.
+ */
+#define SSP_WRITE_BITS(reg, val, mask, sb) \
+ ((reg) = (((reg) & ~(mask)) | (((val)<<(sb)) & (mask))))
+
+/*
+ * This macro is also used to define some default values.
+ * It will just shift val by sb steps to the left and mask
+ * the result with mask.
+ */
+#define GEN_MASK_BITS(val, mask, sb) \
+ (((val)<<(sb)) & (mask))
+
+#define DRIVE_TX		0
+#define DO_NOT_DRIVE_TX		1
+
+#define DO_NOT_QUEUE_DMA	0
+#define QUEUE_DMA		1
+
+#define RX_TRANSFER		1
+#define TX_TRANSFER		2
+
+/*
+ * Macros to access SSP Registers with their offsets
+ */
+#define SSP_CR0(r)	(r + 0x000)
+#define SSP_CR1(r)	(r + 0x004)
+#define SSP_DR(r)	(r + 0x008)
+#define SSP_SR(r)	(r + 0x00C)
+#define SSP_CPSR(r)	(r + 0x010)
+#define SSP_IMSC(r)	(r + 0x014)
+#define SSP_RIS(r)	(r + 0x018)
+#define SSP_MIS(r)	(r + 0x01C)
+#define SSP_ICR(r)	(r + 0x020)
+#define SSP_DMACR(r)	(r + 0x024)
+#define SSP_CSR(r)	(r + 0x030) /* vendor extension */
+#define SSP_ITCR(r)	(r + 0x080)
+#define SSP_ITIP(r)	(r + 0x084)
+#define SSP_ITOP(r)	(r + 0x088)
+#define SSP_TDR(r)	(r + 0x08C)
+
+#define SSP_PID0(r)	(r + 0xFE0)
+#define SSP_PID1(r)	(r + 0xFE4)
+#define SSP_PID2(r)	(r + 0xFE8)
+#define SSP_PID3(r)	(r + 0xFEC)
+
+#define SSP_CID0(r)	(r + 0xFF0)
+#define SSP_CID1(r)	(r + 0xFF4)
+#define SSP_CID2(r)	(r + 0xFF8)
+#define SSP_CID3(r)	(r + 0xFFC)
+
+/*
+ * SSP Control Register 0  - SSP_CR0
+ */
+#define SSP_CR0_MASK_DSS	(0x0FUL << 0)
+#define SSP_CR0_MASK_FRF	(0x3UL << 4)
+#define SSP_CR0_MASK_SPO	(0x1UL << 6)
+#define SSP_CR0_MASK_SPH	(0x1UL << 7)
+#define SSP_CR0_MASK_SCR	(0xFFUL << 8)
+
+/*
+ * The ST version of this block moves som bits
+ * in SSP_CR0 and extends it to 32 bits
+ */
+#define SSP_CR0_MASK_DSS_ST	(0x1FUL << 0)
+#define SSP_CR0_MASK_HALFDUP_ST	(0x1UL << 5)
+#define SSP_CR0_MASK_CSS_ST	(0x1FUL << 16)
+#define SSP_CR0_MASK_FRF_ST	(0x3UL << 21)
+
+/*
+ * SSP Control Register 0  - SSP_CR1
+ */
+#define SSP_CR1_MASK_LBM	(0x1UL << 0)
+#define SSP_CR1_MASK_SSE	(0x1UL << 1)
+#define SSP_CR1_MASK_MS		(0x1UL << 2)
+#define SSP_CR1_MASK_SOD	(0x1UL << 3)
+
+/*
+ * The ST version of this block adds some bits
+ * in SSP_CR1
+ */
+#define SSP_CR1_MASK_RENDN_ST	(0x1UL << 4)
+#define SSP_CR1_MASK_TENDN_ST	(0x1UL << 5)
+#define SSP_CR1_MASK_MWAIT_ST	(0x1UL << 6)
+#define SSP_CR1_MASK_RXIFLSEL_ST (0x7UL << 7)
+#define SSP_CR1_MASK_TXIFLSEL_ST (0x7UL << 10)
+/* This one is only in the PL023 variant */
+#define SSP_CR1_MASK_FBCLKDEL_ST (0x7UL << 13)
+
+/*
+ * SSP Status Register - SSP_SR
+ */
+#define SSP_SR_MASK_TFE		(0x1UL << 0) /* Transmit FIFO empty */
+#define SSP_SR_MASK_TNF		(0x1UL << 1) /* Transmit FIFO not full */
+#define SSP_SR_MASK_RNE		(0x1UL << 2) /* Receive FIFO not empty */
+#define SSP_SR_MASK_RFF		(0x1UL << 3) /* Receive FIFO full */
+#define SSP_SR_MASK_BSY		(0x1UL << 4) /* Busy Flag */
+
+/*
+ * SSP Clock Prescale Register  - SSP_CPSR
+ */
+#define SSP_CPSR_MASK_CPSDVSR	(0xFFUL << 0)
+
+/*
+ * SSP Interrupt Mask Set/Clear Register - SSP_IMSC
+ */
+#define SSP_IMSC_MASK_RORIM (0x1UL << 0) /* Receive Overrun Interrupt mask */
+#define SSP_IMSC_MASK_RTIM  (0x1UL << 1) /* Receive timeout Interrupt mask */
+#define SSP_IMSC_MASK_RXIM  (0x1UL << 2) /* Receive FIFO Interrupt mask */
+#define SSP_IMSC_MASK_TXIM  (0x1UL << 3) /* Transmit FIFO Interrupt mask */
+
+/*
+ * SSP Raw Interrupt Status Register - SSP_RIS
+ */
+/* Receive Overrun Raw Interrupt status */
+#define SSP_RIS_MASK_RORRIS		(0x1UL << 0)
+/* Receive Timeout Raw Interrupt status */
+#define SSP_RIS_MASK_RTRIS		(0x1UL << 1)
+/* Receive FIFO Raw Interrupt status */
+#define SSP_RIS_MASK_RXRIS		(0x1UL << 2)
+/* Transmit FIFO Raw Interrupt status */
+#define SSP_RIS_MASK_TXRIS		(0x1UL << 3)
+
+/*
+ * SSP Masked Interrupt Status Register - SSP_MIS
+ */
+/* Receive Overrun Masked Interrupt status */
+#define SSP_MIS_MASK_RORMIS		(0x1UL << 0)
+/* Receive Timeout Masked Interrupt status */
+#define SSP_MIS_MASK_RTMIS		(0x1UL << 1)
+/* Receive FIFO Masked Interrupt status */
+#define SSP_MIS_MASK_RXMIS		(0x1UL << 2)
+/* Transmit FIFO Masked Interrupt status */
+#define SSP_MIS_MASK_TXMIS		(0x1UL << 3)
+
+/*
+ * SSP Interrupt Clear Register - SSP_ICR
+ */
+/* Receive Overrun Raw Clear Interrupt bit */
+#define SSP_ICR_MASK_RORIC		(0x1UL << 0)
+/* Receive Timeout Clear Interrupt bit */
+#define SSP_ICR_MASK_RTIC		(0x1UL << 1)
+
+/*
+ * SSP DMA Control Register - SSP_DMACR
+ */
+/* Receive DMA Enable bit */
+#define SSP_DMACR_MASK_RXDMAE		(0x1UL << 0)
+/* Transmit DMA Enable bit */
+#define SSP_DMACR_MASK_TXDMAE		(0x1UL << 1)
+
+/*
+ * SSP Chip Select Control Register - SSP_CSR
+ * (vendor extension)
+ */
+#define SSP_CSR_CSVALUE_MASK		(0x1FUL << 0)
+
+/*
+ * SSP Integration Test control Register - SSP_ITCR
+ */
+#define SSP_ITCR_MASK_ITEN		(0x1UL << 0)
+#define SSP_ITCR_MASK_TESTFIFO		(0x1UL << 1)
+
+/*
+ * SSP Integration Test Input Register - SSP_ITIP
+ */
+#define ITIP_MASK_SSPRXD		 (0x1UL << 0)
+#define ITIP_MASK_SSPFSSIN		 (0x1UL << 1)
+#define ITIP_MASK_SSPCLKIN		 (0x1UL << 2)
+#define ITIP_MASK_RXDMAC		 (0x1UL << 3)
+#define ITIP_MASK_TXDMAC		 (0x1UL << 4)
+#define ITIP_MASK_SSPTXDIN		 (0x1UL << 5)
+
+/*
+ * SSP Integration Test output Register - SSP_ITOP
+ */
+#define ITOP_MASK_SSPTXD		 (0x1UL << 0)
+#define ITOP_MASK_SSPFSSOUT		 (0x1UL << 1)
+#define ITOP_MASK_SSPCLKOUT		 (0x1UL << 2)
+#define ITOP_MASK_SSPOEn		 (0x1UL << 3)
+#define ITOP_MASK_SSPCTLOEn		 (0x1UL << 4)
+#define ITOP_MASK_RORINTR		 (0x1UL << 5)
+#define ITOP_MASK_RTINTR		 (0x1UL << 6)
+#define ITOP_MASK_RXINTR		 (0x1UL << 7)
+#define ITOP_MASK_TXINTR		 (0x1UL << 8)
+#define ITOP_MASK_INTR			 (0x1UL << 9)
+#define ITOP_MASK_RXDMABREQ		 (0x1UL << 10)
+#define ITOP_MASK_RXDMASREQ		 (0x1UL << 11)
+#define ITOP_MASK_TXDMABREQ		 (0x1UL << 12)
+#define ITOP_MASK_TXDMASREQ		 (0x1UL << 13)
+
+/*
+ * SSP Test Data Register - SSP_TDR
+ */
+#define TDR_MASK_TESTDATA		(0xFFFFFFFF)
+
+/*
+ * Message State
+ * we use the spi_message.state (void *) pointer to
+ * hold a single state value, that's why all this
+ * (void *) casting is done here.
+ */
+#define STATE_START			((void *) 0)
+#define STATE_RUNNING			((void *) 1)
+#define STATE_DONE			((void *) 2)
+#define STATE_ERROR			((void *) -1)
+
+/*
+ * SSP State - Whether Enabled or Disabled
+ */
+#define SSP_DISABLED			(0)
+#define SSP_ENABLED			(1)
+
+/*
+ * SSP DMA State - Whether DMA Enabled or Disabled
+ */
+#define SSP_DMA_DISABLED		(0)
+#define SSP_DMA_ENABLED			(1)
+
+/*
+ * SSP Clock Defaults
+ */
+#define SSP_DEFAULT_CLKRATE 0x2
+#define SSP_DEFAULT_PRESCALE 0x40
+
+/*
+ * SSP Clock Parameter ranges
+ */
+#define CPSDVR_MIN 0x02
+#define CPSDVR_MAX 0xFE
+#define SCR_MIN 0x00
+#define SCR_MAX 0xFF
+
+/*
+ * SSP Interrupt related Macros
+ */
+#define DEFAULT_SSP_REG_IMSC  0x0UL
+#define DISABLE_ALL_INTERRUPTS DEFAULT_SSP_REG_IMSC
+#define ENABLE_ALL_INTERRUPTS ( \
+	SSP_IMSC_MASK_RORIM | \
+	SSP_IMSC_MASK_RTIM | \
+	SSP_IMSC_MASK_RXIM | \
+	SSP_IMSC_MASK_TXIM \
+)
+
+#define CLEAR_ALL_INTERRUPTS  0x3
+
+#define SPI_POLLING_TIMEOUT 1000
+
+/*
+ * The type of reading going on on this chip
+ */
+enum ssp_reading {
+	READING_NULL,
+	READING_U8,
+	READING_U16,
+	READING_U32
+};
+
+/**
+ * The type of writing going on on this chip
+ */
+enum ssp_writing {
+	WRITING_NULL,
+	WRITING_U8,
+	WRITING_U16,
+	WRITING_U32
+};
+
+/**
+ * struct vendor_data - vendor-specific config parameters
+ * for PL022 derivates
+ * @fifodepth: depth of FIFOs (both)
+ * @max_bpw: maximum number of bits per word
+ * @unidir: supports unidirection transfers
+ * @extended_cr: 32 bit wide control register 0 with extra
+ * features and extra features in CR1 as found in the ST variants
+ * @pl023: supports a subset of the ST extensions called "PL023"
+ * @internal_cs_ctrl: supports chip select control register
+ */
+struct vendor_data {
+	int fifodepth;
+	int max_bpw;
+	bool unidir;
+	bool extended_cr;
+	bool pl023;
+	bool loopback;
+	bool internal_cs_ctrl;
+};
+
+/**
+ * struct pl022 - This is the private SSP driver data structure
+ * @adev: AMBA device model hookup
+ * @vendor: vendor data for the IP block
+ * @phybase: the physical memory where the SSP device resides
+ * @virtbase: the virtual memory where the SSP is mapped
+ * @clk: outgoing clock "SPICLK" for the SPI bus
+ * @master: SPI framework hookup
+ * @master_info: controller-specific data from machine setup
+ * @pump_transfers: Tasklet used in Interrupt Transfer mode
+ * @cur_msg: Pointer to current spi_message being processed
+ * @cur_transfer: Pointer to current spi_transfer
+ * @cur_chip: pointer to current clients chip(assigned from controller_state)
+ * @next_msg_cs_active: the next message in the queue has been examined
+ *  and it was found that it uses the same chip select as the previous
+ *  message, so we left it active after the previous transfer, and it's
+ *  active already.
+ * @tx: current position in TX buffer to be read
+ * @tx_end: end position in TX buffer to be read
+ * @rx: current position in RX buffer to be written
+ * @rx_end: end position in RX buffer to be written
+ * @read: the type of read currently going on
+ * @write: the type of write currently going on
+ * @exp_fifo_level: expected FIFO level
+ * @dma_rx_channel: optional channel for RX DMA
+ * @dma_tx_channel: optional channel for TX DMA
+ * @sgt_rx: scattertable for the RX transfer
+ * @sgt_tx: scattertable for the TX transfer
+ * @dummypage: a dummy page used for driving data on the bus with DMA
+ * @cur_cs: current chip select (gpio)
+ * @chipselects: list of chipselects (gpios)
+ */
+struct pl022 {
+	struct amba_device		*adev;
+	struct vendor_data		*vendor;
+	resource_size_t			phybase;
+	void __iomem			*virtbase;
+	struct clk			*clk;
+	struct spi_master		*master;
+	struct pl022_ssp_controller	*master_info;
+	/* Message per-transfer pump */
+	struct tasklet_struct		pump_transfers;
+	struct spi_message		*cur_msg;
+	struct spi_transfer		*cur_transfer;
+	struct chip_data		*cur_chip;
+	bool				next_msg_cs_active;
+	void				*tx;
+	void				*tx_end;
+	void				*rx;
+	void				*rx_end;
+	enum ssp_reading		read;
+	enum ssp_writing		write;
+	u32				exp_fifo_level;
+	enum ssp_rx_level_trig		rx_lev_trig;
+	enum ssp_tx_level_trig		tx_lev_trig;
+	/* DMA settings */
+#ifdef CONFIG_DMA_ENGINE
+	struct dma_chan			*dma_rx_channel;
+	struct dma_chan			*dma_tx_channel;
+	struct sg_table			sgt_rx;
+	struct sg_table			sgt_tx;
+	char				*dummypage;
+	bool				dma_running;
+#endif
+	int cur_cs;
+	int *chipselects;
+};
+
+/**
+ * struct chip_data - To maintain runtime state of SSP for each client chip
+ * @cr0: Value of control register CR0 of SSP - on later ST variants this
+ *       register is 32 bits wide rather than just 16
+ * @cr1: Value of control register CR1 of SSP
+ * @dmacr: Value of DMA control Register of SSP
+ * @cpsr: Value of Clock prescale register
+ * @n_bytes: how many bytes(power of 2) reqd for a given data width of client
+ * @enable_dma: Whether to enable DMA or not
+ * @read: function ptr to be used to read when doing xfer for this chip
+ * @write: function ptr to be used to write when doing xfer for this chip
+ * @cs_control: chip select callback provided by chip
+ * @xfer_type: polling/interrupt/DMA
+ *
+ * Runtime state of the SSP controller, maintained per chip,
+ * This would be set according to the current message that would be served
+ */
+struct chip_data {
+	u32 cr0;
+	u16 cr1;
+	u16 dmacr;
+	u16 cpsr;
+	u8 n_bytes;
+	bool enable_dma;
+	enum ssp_reading read;
+	enum ssp_writing write;
+	void (*cs_control) (u32 command);
+	int xfer_type;
+};
+
+/**
+ * null_cs_control - Dummy chip select function
+ * @command: select/delect the chip
+ *
+ * If no chip select function is provided by client this is used as dummy
+ * chip select
+ */
+static void null_cs_control(u32 command)
+{
+	pr_debug("pl022: dummy chip select control, CS=0x%x\n", command);
+}
+
+/**
+ * internal_cs_control - Control chip select signals via SSP_CSR.
+ * @pl022: SSP driver private data structure
+ * @command: select/delect the chip
+ *
+ * Used on controller with internal chip select control via SSP_CSR register
+ * (vendor extension). Each of the 5 LSB in the register controls one chip
+ * select signal.
+ */
+static void internal_cs_control(struct pl022 *pl022, u32 command)
+{
+	u32 tmp;
+
+	tmp = readw(SSP_CSR(pl022->virtbase));
+	if (command == SSP_CHIP_SELECT)
+		tmp &= ~BIT(pl022->cur_cs);
+	else
+		tmp |= BIT(pl022->cur_cs);
+	writew(tmp, SSP_CSR(pl022->virtbase));
+}
+
+static void pl022_cs_control(struct pl022 *pl022, u32 command)
+{
+	if (pl022->vendor->internal_cs_ctrl)
+		internal_cs_control(pl022, command);
+	else if (gpio_is_valid(pl022->cur_cs))
+		gpio_set_value(pl022->cur_cs, command);
+	else
+		pl022->cur_chip->cs_control(command);
+}
+
+/**
+ * giveback - current spi_message is over, schedule next message and call
+ * callback of this message. Assumes that caller already
+ * set message->status; dma and pio irqs are blocked
+ * @pl022: SSP driver private data structure
+ */
+static void giveback(struct pl022 *pl022)
+{
+	struct spi_transfer *last_transfer;
+	pl022->next_msg_cs_active = false;
+
+	last_transfer = list_last_entry(&pl022->cur_msg->transfers,
+					struct spi_transfer, transfer_list);
+
+	/* Delay if requested before any change in chip select */
+	if (last_transfer->delay_usecs)
+		/*
+		 * FIXME: This runs in interrupt context.
+		 * Is this really smart?
+		 */
+		udelay(last_transfer->delay_usecs);
+
+	if (!last_transfer->cs_change) {
+		struct spi_message *next_msg;
+
+		/*
+		 * cs_change was not set. We can keep the chip select
+		 * enabled if there is message in the queue and it is
+		 * for the same spi device.
+		 *
+		 * We cannot postpone this until pump_messages, because
+		 * after calling msg->complete (below) the driver that
+		 * sent the current message could be unloaded, which
+		 * could invalidate the cs_control() callback...
+		 */
+		/* get a pointer to the next message, if any */
+		next_msg = spi_get_next_queued_message(pl022->master);
+
+		/*
+		 * see if the next and current messages point
+		 * to the same spi device.
+		 */
+		if (next_msg && next_msg->spi != pl022->cur_msg->spi)
+			next_msg = NULL;
+		if (!next_msg || pl022->cur_msg->state == STATE_ERROR)
+			pl022_cs_control(pl022, SSP_CHIP_DESELECT);
+		else
+			pl022->next_msg_cs_active = true;
+
+	}
+
+	pl022->cur_msg = NULL;
+	pl022->cur_transfer = NULL;
+	pl022->cur_chip = NULL;
+
+	/* disable the SPI/SSP operation */
+	writew((readw(SSP_CR1(pl022->virtbase)) &
+		(~SSP_CR1_MASK_SSE)), SSP_CR1(pl022->virtbase));
+
+	spi_finalize_current_message(pl022->master);
+}
+
+/**
+ * flush - flush the FIFO to reach a clean state
+ * @pl022: SSP driver private data structure
+ */
+static int flush(struct pl022 *pl022)
+{
+	unsigned long limit = loops_per_jiffy << 1;
+
+	dev_dbg(&pl022->adev->dev, "flush\n");
+	do {
+		while (readw(SSP_SR(pl022->virtbase)) & SSP_SR_MASK_RNE)
+			readw(SSP_DR(pl022->virtbase));
+	} while ((readw(SSP_SR(pl022->virtbase)) & SSP_SR_MASK_BSY) && limit--);
+
+	pl022->exp_fifo_level = 0;
+
+	return limit;
+}
+
+/**
+ * restore_state - Load configuration of current chip
+ * @pl022: SSP driver private data structure
+ */
+static void restore_state(struct pl022 *pl022)
+{
+	struct chip_data *chip = pl022->cur_chip;
+
+	if (pl022->vendor->extended_cr)
+		writel(chip->cr0, SSP_CR0(pl022->virtbase));
+	else
+		writew(chip->cr0, SSP_CR0(pl022->virtbase));
+	writew(chip->cr1, SSP_CR1(pl022->virtbase));
+	writew(chip->dmacr, SSP_DMACR(pl022->virtbase));
+	writew(chip->cpsr, SSP_CPSR(pl022->virtbase));
+	writew(DISABLE_ALL_INTERRUPTS, SSP_IMSC(pl022->virtbase));
+	writew(CLEAR_ALL_INTERRUPTS, SSP_ICR(pl022->virtbase));
+}
+
+/*
+ * Default SSP Register Values
+ */
+#define DEFAULT_SSP_REG_CR0 ( \
+	GEN_MASK_BITS(SSP_DATA_BITS_12, SSP_CR0_MASK_DSS, 0)	| \
+	GEN_MASK_BITS(SSP_INTERFACE_MOTOROLA_SPI, SSP_CR0_MASK_FRF, 4) | \
+	GEN_MASK_BITS(SSP_CLK_POL_IDLE_LOW, SSP_CR0_MASK_SPO, 6) | \
+	GEN_MASK_BITS(SSP_CLK_SECOND_EDGE, SSP_CR0_MASK_SPH, 7) | \
+	GEN_MASK_BITS(SSP_DEFAULT_CLKRATE, SSP_CR0_MASK_SCR, 8) \
+)
+
+/* ST versions have slightly different bit layout */
+#define DEFAULT_SSP_REG_CR0_ST ( \
+	GEN_MASK_BITS(SSP_DATA_BITS_12, SSP_CR0_MASK_DSS_ST, 0)	| \
+	GEN_MASK_BITS(SSP_MICROWIRE_CHANNEL_FULL_DUPLEX, SSP_CR0_MASK_HALFDUP_ST, 5) | \
+	GEN_MASK_BITS(SSP_CLK_POL_IDLE_LOW, SSP_CR0_MASK_SPO, 6) | \
+	GEN_MASK_BITS(SSP_CLK_SECOND_EDGE, SSP_CR0_MASK_SPH, 7) | \
+	GEN_MASK_BITS(SSP_DEFAULT_CLKRATE, SSP_CR0_MASK_SCR, 8) | \
+	GEN_MASK_BITS(SSP_BITS_8, SSP_CR0_MASK_CSS_ST, 16)	| \
+	GEN_MASK_BITS(SSP_INTERFACE_MOTOROLA_SPI, SSP_CR0_MASK_FRF_ST, 21) \
+)
+
+/* The PL023 version is slightly different again */
+#define DEFAULT_SSP_REG_CR0_ST_PL023 ( \
+	GEN_MASK_BITS(SSP_DATA_BITS_12, SSP_CR0_MASK_DSS_ST, 0)	| \
+	GEN_MASK_BITS(SSP_CLK_POL_IDLE_LOW, SSP_CR0_MASK_SPO, 6) | \
+	GEN_MASK_BITS(SSP_CLK_SECOND_EDGE, SSP_CR0_MASK_SPH, 7) | \
+	GEN_MASK_BITS(SSP_DEFAULT_CLKRATE, SSP_CR0_MASK_SCR, 8) \
+)
+
+#define DEFAULT_SSP_REG_CR1 ( \
+	GEN_MASK_BITS(LOOPBACK_DISABLED, SSP_CR1_MASK_LBM, 0) | \
+	GEN_MASK_BITS(SSP_DISABLED, SSP_CR1_MASK_SSE, 1) | \
+	GEN_MASK_BITS(SSP_MASTER, SSP_CR1_MASK_MS, 2) | \
+	GEN_MASK_BITS(DO_NOT_DRIVE_TX, SSP_CR1_MASK_SOD, 3) \
+)
+
+/* ST versions extend this register to use all 16 bits */
+#define DEFAULT_SSP_REG_CR1_ST ( \
+	DEFAULT_SSP_REG_CR1 | \
+	GEN_MASK_BITS(SSP_RX_MSB, SSP_CR1_MASK_RENDN_ST, 4) | \
+	GEN_MASK_BITS(SSP_TX_MSB, SSP_CR1_MASK_TENDN_ST, 5) | \
+	GEN_MASK_BITS(SSP_MWIRE_WAIT_ZERO, SSP_CR1_MASK_MWAIT_ST, 6) |\
+	GEN_MASK_BITS(SSP_RX_1_OR_MORE_ELEM, SSP_CR1_MASK_RXIFLSEL_ST, 7) | \
+	GEN_MASK_BITS(SSP_TX_1_OR_MORE_EMPTY_LOC, SSP_CR1_MASK_TXIFLSEL_ST, 10) \
+)
+
+/*
+ * The PL023 variant has further differences: no loopback mode, no microwire
+ * support, and a new clock feedback delay setting.
+ */
+#define DEFAULT_SSP_REG_CR1_ST_PL023 ( \
+	GEN_MASK_BITS(SSP_DISABLED, SSP_CR1_MASK_SSE, 1) | \
+	GEN_MASK_BITS(SSP_MASTER, SSP_CR1_MASK_MS, 2) | \
+	GEN_MASK_BITS(DO_NOT_DRIVE_TX, SSP_CR1_MASK_SOD, 3) | \
+	GEN_MASK_BITS(SSP_RX_MSB, SSP_CR1_MASK_RENDN_ST, 4) | \
+	GEN_MASK_BITS(SSP_TX_MSB, SSP_CR1_MASK_TENDN_ST, 5) | \
+	GEN_MASK_BITS(SSP_RX_1_OR_MORE_ELEM, SSP_CR1_MASK_RXIFLSEL_ST, 7) | \
+	GEN_MASK_BITS(SSP_TX_1_OR_MORE_EMPTY_LOC, SSP_CR1_MASK_TXIFLSEL_ST, 10) | \
+	GEN_MASK_BITS(SSP_FEEDBACK_CLK_DELAY_NONE, SSP_CR1_MASK_FBCLKDEL_ST, 13) \
+)
+
+#define DEFAULT_SSP_REG_CPSR ( \
+	GEN_MASK_BITS(SSP_DEFAULT_PRESCALE, SSP_CPSR_MASK_CPSDVSR, 0) \
+)
+
+#define DEFAULT_SSP_REG_DMACR (\
+	GEN_MASK_BITS(SSP_DMA_DISABLED, SSP_DMACR_MASK_RXDMAE, 0) | \
+	GEN_MASK_BITS(SSP_DMA_DISABLED, SSP_DMACR_MASK_TXDMAE, 1) \
+)
+
+/**
+ * load_ssp_default_config - Load default configuration for SSP
+ * @pl022: SSP driver private data structure
+ */
+static void load_ssp_default_config(struct pl022 *pl022)
+{
+	if (pl022->vendor->pl023) {
+		writel(DEFAULT_SSP_REG_CR0_ST_PL023, SSP_CR0(pl022->virtbase));
+		writew(DEFAULT_SSP_REG_CR1_ST_PL023, SSP_CR1(pl022->virtbase));
+	} else if (pl022->vendor->extended_cr) {
+		writel(DEFAULT_SSP_REG_CR0_ST, SSP_CR0(pl022->virtbase));
+		writew(DEFAULT_SSP_REG_CR1_ST, SSP_CR1(pl022->virtbase));
+	} else {
+		writew(DEFAULT_SSP_REG_CR0, SSP_CR0(pl022->virtbase));
+		writew(DEFAULT_SSP_REG_CR1, SSP_CR1(pl022->virtbase));
+	}
+	writew(DEFAULT_SSP_REG_DMACR, SSP_DMACR(pl022->virtbase));
+	writew(DEFAULT_SSP_REG_CPSR, SSP_CPSR(pl022->virtbase));
+	writew(DISABLE_ALL_INTERRUPTS, SSP_IMSC(pl022->virtbase));
+	writew(CLEAR_ALL_INTERRUPTS, SSP_ICR(pl022->virtbase));
+}
+
+/**
+ * This will write to TX and read from RX according to the parameters
+ * set in pl022.
+ */
+static void readwriter(struct pl022 *pl022)
+{
+
+	/*
+	 * The FIFO depth is different between primecell variants.
+	 * I believe filling in too much in the FIFO might cause
+	 * errons in 8bit wide transfers on ARM variants (just 8 words
+	 * FIFO, means only 8x8 = 64 bits in FIFO) at least.
+	 *
+	 * To prevent this issue, the TX FIFO is only filled to the
+	 * unused RX FIFO fill length, regardless of what the TX
+	 * FIFO status flag indicates.
+	 */
+	dev_dbg(&pl022->adev->dev,
+		"%s, rx: %p, rxend: %p, tx: %p, txend: %p\n",
+		__func__, pl022->rx, pl022->rx_end, pl022->tx, pl022->tx_end);
+
+	/* Read as much as you can */
+	while ((readw(SSP_SR(pl022->virtbase)) & SSP_SR_MASK_RNE)
+	       && (pl022->rx < pl022->rx_end)) {
+		switch (pl022->read) {
+		case READING_NULL:
+			readw(SSP_DR(pl022->virtbase));
+			break;
+		case READING_U8:
+			*(u8 *) (pl022->rx) =
+				readw(SSP_DR(pl022->virtbase)) & 0xFFU;
+			break;
+		case READING_U16:
+			*(u16 *) (pl022->rx) =
+				(u16) readw(SSP_DR(pl022->virtbase));
+			break;
+		case READING_U32:
+			*(u32 *) (pl022->rx) =
+				readl(SSP_DR(pl022->virtbase));
+			break;
+		}
+		pl022->rx += (pl022->cur_chip->n_bytes);
+		pl022->exp_fifo_level--;
+	}
+	/*
+	 * Write as much as possible up to the RX FIFO size
+	 */
+	while ((pl022->exp_fifo_level < pl022->vendor->fifodepth)
+	       && (pl022->tx < pl022->tx_end)) {
+		switch (pl022->write) {
+		case WRITING_NULL:
+			writew(0x0, SSP_DR(pl022->virtbase));
+			break;
+		case WRITING_U8:
+			writew(*(u8 *) (pl022->tx), SSP_DR(pl022->virtbase));
+			break;
+		case WRITING_U16:
+			writew((*(u16 *) (pl022->tx)), SSP_DR(pl022->virtbase));
+			break;
+		case WRITING_U32:
+			writel(*(u32 *) (pl022->tx), SSP_DR(pl022->virtbase));
+			break;
+		}
+		pl022->tx += (pl022->cur_chip->n_bytes);
+		pl022->exp_fifo_level++;
+		/*
+		 * This inner reader takes care of things appearing in the RX
+		 * FIFO as we're transmitting. This will happen a lot since the
+		 * clock starts running when you put things into the TX FIFO,
+		 * and then things are continuously clocked into the RX FIFO.
+		 */
+		while ((readw(SSP_SR(pl022->virtbase)) & SSP_SR_MASK_RNE)
+		       && (pl022->rx < pl022->rx_end)) {
+			switch (pl022->read) {
+			case READING_NULL:
+				readw(SSP_DR(pl022->virtbase));
+				break;
+			case READING_U8:
+				*(u8 *) (pl022->rx) =
+					readw(SSP_DR(pl022->virtbase)) & 0xFFU;
+				break;
+			case READING_U16:
+				*(u16 *) (pl022->rx) =
+					(u16) readw(SSP_DR(pl022->virtbase));
+				break;
+			case READING_U32:
+				*(u32 *) (pl022->rx) =
+					readl(SSP_DR(pl022->virtbase));
+				break;
+			}
+			pl022->rx += (pl022->cur_chip->n_bytes);
+			pl022->exp_fifo_level--;
+		}
+	}
+	/*
+	 * When we exit here the TX FIFO should be full and the RX FIFO
+	 * should be empty
+	 */
+}
+
+/**
+ * next_transfer - Move to the Next transfer in the current spi message
+ * @pl022: SSP driver private data structure
+ *
+ * This function moves though the linked list of spi transfers in the
+ * current spi message and returns with the state of current spi
+ * message i.e whether its last transfer is done(STATE_DONE) or
+ * Next transfer is ready(STATE_RUNNING)
+ */
+static void *next_transfer(struct pl022 *pl022)
+{
+	struct spi_message *msg = pl022->cur_msg;
+	struct spi_transfer *trans = pl022->cur_transfer;
+
+	/* Move to next transfer */
+	if (trans->transfer_list.next != &msg->transfers) {
+		pl022->cur_transfer =
+		    list_entry(trans->transfer_list.next,
+			       struct spi_transfer, transfer_list);
+		return STATE_RUNNING;
+	}
+	return STATE_DONE;
+}
+
+/*
+ * This DMA functionality is only compiled in if we have
+ * access to the generic DMA devices/DMA engine.
+ */
+#ifdef CONFIG_DMA_ENGINE
+static void unmap_free_dma_scatter(struct pl022 *pl022)
+{
+	/* Unmap and free the SG tables */
+	dma_unmap_sg(pl022->dma_tx_channel->device->dev, pl022->sgt_tx.sgl,
+		     pl022->sgt_tx.nents, DMA_TO_DEVICE);
+	dma_unmap_sg(pl022->dma_rx_channel->device->dev, pl022->sgt_rx.sgl,
+		     pl022->sgt_rx.nents, DMA_FROM_DEVICE);
+	sg_free_table(&pl022->sgt_rx);
+	sg_free_table(&pl022->sgt_tx);
+}
+
+static void dma_callback(void *data)
+{
+	struct pl022 *pl022 = data;
+	struct spi_message *msg = pl022->cur_msg;
+
+	BUG_ON(!pl022->sgt_rx.sgl);
+
+#ifdef VERBOSE_DEBUG
+	/*
+	 * Optionally dump out buffers to inspect contents, this is
+	 * good if you want to convince yourself that the loopback
+	 * read/write contents are the same, when adopting to a new
+	 * DMA engine.
+	 */
+	{
+		struct scatterlist *sg;
+		unsigned int i;
+
+		dma_sync_sg_for_cpu(&pl022->adev->dev,
+				    pl022->sgt_rx.sgl,
+				    pl022->sgt_rx.nents,
+				    DMA_FROM_DEVICE);
+
+		for_each_sg(pl022->sgt_rx.sgl, sg, pl022->sgt_rx.nents, i) {
+			dev_dbg(&pl022->adev->dev, "SPI RX SG ENTRY: %d", i);
+			print_hex_dump(KERN_ERR, "SPI RX: ",
+				       DUMP_PREFIX_OFFSET,
+				       16,
+				       1,
+				       sg_virt(sg),
+				       sg_dma_len(sg),
+				       1);
+		}
+		for_each_sg(pl022->sgt_tx.sgl, sg, pl022->sgt_tx.nents, i) {
+			dev_dbg(&pl022->adev->dev, "SPI TX SG ENTRY: %d", i);
+			print_hex_dump(KERN_ERR, "SPI TX: ",
+				       DUMP_PREFIX_OFFSET,
+				       16,
+				       1,
+				       sg_virt(sg),
+				       sg_dma_len(sg),
+				       1);
+		}
+	}
+#endif
+
+	unmap_free_dma_scatter(pl022);
+
+	/* Update total bytes transferred */
+	msg->actual_length += pl022->cur_transfer->len;
+	if (pl022->cur_transfer->cs_change)
+		pl022_cs_control(pl022, SSP_CHIP_DESELECT);
+
+	/* Move to next transfer */
+	msg->state = next_transfer(pl022);
+	tasklet_schedule(&pl022->pump_transfers);
+}
+
+static void setup_dma_scatter(struct pl022 *pl022,
+			      void *buffer,
+			      unsigned int length,
+			      struct sg_table *sgtab)
+{
+	struct scatterlist *sg;
+	int bytesleft = length;
+	void *bufp = buffer;
+	int mapbytes;
+	int i;
+
+	if (buffer) {
+		for_each_sg(sgtab->sgl, sg, sgtab->nents, i) {
+			/*
+			 * If there are less bytes left than what fits
+			 * in the current page (plus page alignment offset)
+			 * we just feed in this, else we stuff in as much
+			 * as we can.
+			 */
+			if (bytesleft < (PAGE_SIZE - offset_in_page(bufp)))
+				mapbytes = bytesleft;
+			else
+				mapbytes = PAGE_SIZE - offset_in_page(bufp);
+			sg_set_page(sg, virt_to_page(bufp),
+				    mapbytes, offset_in_page(bufp));
+			bufp += mapbytes;
+			bytesleft -= mapbytes;
+			dev_dbg(&pl022->adev->dev,
+				"set RX/TX target page @ %p, %d bytes, %d left\n",
+				bufp, mapbytes, bytesleft);
+		}
+	} else {
+		/* Map the dummy buffer on every page */
+		for_each_sg(sgtab->sgl, sg, sgtab->nents, i) {
+			if (bytesleft < PAGE_SIZE)
+				mapbytes = bytesleft;
+			else
+				mapbytes = PAGE_SIZE;
+			sg_set_page(sg, virt_to_page(pl022->dummypage),
+				    mapbytes, 0);
+			bytesleft -= mapbytes;
+			dev_dbg(&pl022->adev->dev,
+				"set RX/TX to dummy page %d bytes, %d left\n",
+				mapbytes, bytesleft);
+
+		}
+	}
+	BUG_ON(bytesleft);
+}
+
+/**
+ * configure_dma - configures the channels for the next transfer
+ * @pl022: SSP driver's private data structure
+ */
+static int configure_dma(struct pl022 *pl022)
+{
+	struct dma_slave_config rx_conf = {
+		.src_addr = SSP_DR(pl022->phybase),
+		.direction = DMA_DEV_TO_MEM,
+		.device_fc = false,
+	};
+	struct dma_slave_config tx_conf = {
+		.dst_addr = SSP_DR(pl022->phybase),
+		.direction = DMA_MEM_TO_DEV,
+		.device_fc = false,
+	};
+	unsigned int pages;
+	int ret;
+	int rx_sglen, tx_sglen;
+	struct dma_chan *rxchan = pl022->dma_rx_channel;
+	struct dma_chan *txchan = pl022->dma_tx_channel;
+	struct dma_async_tx_descriptor *rxdesc;
+	struct dma_async_tx_descriptor *txdesc;
+
+	/* Check that the channels are available */
+	if (!rxchan || !txchan)
+		return -ENODEV;
+
+	/*
+	 * If supplied, the DMA burstsize should equal the FIFO trigger level.
+	 * Notice that the DMA engine uses one-to-one mapping. Since we can
+	 * not trigger on 2 elements this needs explicit mapping rather than
+	 * calculation.
+	 */
+	switch (pl022->rx_lev_trig) {
+	case SSP_RX_1_OR_MORE_ELEM:
+		rx_conf.src_maxburst = 1;
+		break;
+	case SSP_RX_4_OR_MORE_ELEM:
+		rx_conf.src_maxburst = 4;
+		break;
+	case SSP_RX_8_OR_MORE_ELEM:
+		rx_conf.src_maxburst = 8;
+		break;
+	case SSP_RX_16_OR_MORE_ELEM:
+		rx_conf.src_maxburst = 16;
+		break;
+	case SSP_RX_32_OR_MORE_ELEM:
+		rx_conf.src_maxburst = 32;
+		break;
+	default:
+		rx_conf.src_maxburst = pl022->vendor->fifodepth >> 1;
+		break;
+	}
+
+	switch (pl022->tx_lev_trig) {
+	case SSP_TX_1_OR_MORE_EMPTY_LOC:
+		tx_conf.dst_maxburst = 1;
+		break;
+	case SSP_TX_4_OR_MORE_EMPTY_LOC:
+		tx_conf.dst_maxburst = 4;
+		break;
+	case SSP_TX_8_OR_MORE_EMPTY_LOC:
+		tx_conf.dst_maxburst = 8;
+		break;
+	case SSP_TX_16_OR_MORE_EMPTY_LOC:
+		tx_conf.dst_maxburst = 16;
+		break;
+	case SSP_TX_32_OR_MORE_EMPTY_LOC:
+		tx_conf.dst_maxburst = 32;
+		break;
+	default:
+		tx_conf.dst_maxburst = pl022->vendor->fifodepth >> 1;
+		break;
+	}
+
+	switch (pl022->read) {
+	case READING_NULL:
+		/* Use the same as for writing */
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_UNDEFINED;
+		break;
+	case READING_U8:
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		break;
+	case READING_U16:
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		break;
+	case READING_U32:
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		break;
+	}
+
+	switch (pl022->write) {
+	case WRITING_NULL:
+		/* Use the same as for reading */
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_UNDEFINED;
+		break;
+	case WRITING_U8:
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		break;
+	case WRITING_U16:
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		break;
+	case WRITING_U32:
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		break;
+	}
+
+	/* SPI pecularity: we need to read and write the same width */
+	if (rx_conf.src_addr_width == DMA_SLAVE_BUSWIDTH_UNDEFINED)
+		rx_conf.src_addr_width = tx_conf.dst_addr_width;
+	if (tx_conf.dst_addr_width == DMA_SLAVE_BUSWIDTH_UNDEFINED)
+		tx_conf.dst_addr_width = rx_conf.src_addr_width;
+	BUG_ON(rx_conf.src_addr_width != tx_conf.dst_addr_width);
+
+	dmaengine_slave_config(rxchan, &rx_conf);
+	dmaengine_slave_config(txchan, &tx_conf);
+
+	/* Create sglists for the transfers */
+	pages = DIV_ROUND_UP(pl022->cur_transfer->len, PAGE_SIZE);
+	dev_dbg(&pl022->adev->dev, "using %d pages for transfer\n", pages);
+
+	ret = sg_alloc_table(&pl022->sgt_rx, pages, GFP_ATOMIC);
+	if (ret)
+		goto err_alloc_rx_sg;
+
+	ret = sg_alloc_table(&pl022->sgt_tx, pages, GFP_ATOMIC);
+	if (ret)
+		goto err_alloc_tx_sg;
+
+	/* Fill in the scatterlists for the RX+TX buffers */
+	setup_dma_scatter(pl022, pl022->rx,
+			  pl022->cur_transfer->len, &pl022->sgt_rx);
+	setup_dma_scatter(pl022, pl022->tx,
+			  pl022->cur_transfer->len, &pl022->sgt_tx);
+
+	/* Map DMA buffers */
+	rx_sglen = dma_map_sg(rxchan->device->dev, pl022->sgt_rx.sgl,
+			   pl022->sgt_rx.nents, DMA_FROM_DEVICE);
+	if (!rx_sglen)
+		goto err_rx_sgmap;
+
+	tx_sglen = dma_map_sg(txchan->device->dev, pl022->sgt_tx.sgl,
+			   pl022->sgt_tx.nents, DMA_TO_DEVICE);
+	if (!tx_sglen)
+		goto err_tx_sgmap;
+
+	/* Send both scatterlists */
+	rxdesc = dmaengine_prep_slave_sg(rxchan,
+				      pl022->sgt_rx.sgl,
+				      rx_sglen,
+				      DMA_DEV_TO_MEM,
+				      DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!rxdesc)
+		goto err_rxdesc;
+
+	txdesc = dmaengine_prep_slave_sg(txchan,
+				      pl022->sgt_tx.sgl,
+				      tx_sglen,
+				      DMA_MEM_TO_DEV,
+				      DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!txdesc)
+		goto err_txdesc;
+
+	/* Put the callback on the RX transfer only, that should finish last */
+	rxdesc->callback = dma_callback;
+	rxdesc->callback_param = pl022;
+
+	/* Submit and fire RX and TX with TX last so we're ready to read! */
+	dmaengine_submit(rxdesc);
+	dmaengine_submit(txdesc);
+	dma_async_issue_pending(rxchan);
+	dma_async_issue_pending(txchan);
+	pl022->dma_running = true;
+
+	return 0;
+
+err_txdesc:
+	dmaengine_terminate_all(txchan);
+err_rxdesc:
+	dmaengine_terminate_all(rxchan);
+	dma_unmap_sg(txchan->device->dev, pl022->sgt_tx.sgl,
+		     pl022->sgt_tx.nents, DMA_TO_DEVICE);
+err_tx_sgmap:
+	dma_unmap_sg(rxchan->device->dev, pl022->sgt_rx.sgl,
+		     pl022->sgt_rx.nents, DMA_FROM_DEVICE);
+err_rx_sgmap:
+	sg_free_table(&pl022->sgt_tx);
+err_alloc_tx_sg:
+	sg_free_table(&pl022->sgt_rx);
+err_alloc_rx_sg:
+	return -ENOMEM;
+}
+
+static int pl022_dma_probe(struct pl022 *pl022)
+{
+	dma_cap_mask_t mask;
+
+	/* Try to acquire a generic DMA engine slave channel */
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+	/*
+	 * We need both RX and TX channels to do DMA, else do none
+	 * of them.
+	 */
+	pl022->dma_rx_channel = dma_request_channel(mask,
+					    pl022->master_info->dma_filter,
+					    pl022->master_info->dma_rx_param);
+	if (!pl022->dma_rx_channel) {
+		dev_dbg(&pl022->adev->dev, "no RX DMA channel!\n");
+		goto err_no_rxchan;
+	}
+
+	pl022->dma_tx_channel = dma_request_channel(mask,
+					    pl022->master_info->dma_filter,
+					    pl022->master_info->dma_tx_param);
+	if (!pl022->dma_tx_channel) {
+		dev_dbg(&pl022->adev->dev, "no TX DMA channel!\n");
+		goto err_no_txchan;
+	}
+
+	pl022->dummypage = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!pl022->dummypage)
+		goto err_no_dummypage;
+
+	dev_info(&pl022->adev->dev, "setup for DMA on RX %s, TX %s\n",
+		 dma_chan_name(pl022->dma_rx_channel),
+		 dma_chan_name(pl022->dma_tx_channel));
+
+	return 0;
+
+err_no_dummypage:
+	dma_release_channel(pl022->dma_tx_channel);
+err_no_txchan:
+	dma_release_channel(pl022->dma_rx_channel);
+	pl022->dma_rx_channel = NULL;
+err_no_rxchan:
+	dev_err(&pl022->adev->dev,
+			"Failed to work in dma mode, work without dma!\n");
+	return -ENODEV;
+}
+
+static int pl022_dma_autoprobe(struct pl022 *pl022)
+{
+	struct device *dev = &pl022->adev->dev;
+	struct dma_chan *chan;
+	int err;
+
+	/* automatically configure DMA channels from platform, normally using DT */
+	chan = dma_request_slave_channel_reason(dev, "rx");
+	if (IS_ERR(chan)) {
+		err = PTR_ERR(chan);
+		goto err_no_rxchan;
+	}
+
+	pl022->dma_rx_channel = chan;
+
+	chan = dma_request_slave_channel_reason(dev, "tx");
+	if (IS_ERR(chan)) {
+		err = PTR_ERR(chan);
+		goto err_no_txchan;
+	}
+
+	pl022->dma_tx_channel = chan;
+
+	pl022->dummypage = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!pl022->dummypage) {
+		err = -ENOMEM;
+		goto err_no_dummypage;
+	}
+
+	return 0;
+
+err_no_dummypage:
+	dma_release_channel(pl022->dma_tx_channel);
+	pl022->dma_tx_channel = NULL;
+err_no_txchan:
+	dma_release_channel(pl022->dma_rx_channel);
+	pl022->dma_rx_channel = NULL;
+err_no_rxchan:
+	return err;
+}
+		
+static void terminate_dma(struct pl022 *pl022)
+{
+	struct dma_chan *rxchan = pl022->dma_rx_channel;
+	struct dma_chan *txchan = pl022->dma_tx_channel;
+
+	dmaengine_terminate_all(rxchan);
+	dmaengine_terminate_all(txchan);
+	unmap_free_dma_scatter(pl022);
+	pl022->dma_running = false;
+}
+
+static void pl022_dma_remove(struct pl022 *pl022)
+{
+	if (pl022->dma_running)
+		terminate_dma(pl022);
+	if (pl022->dma_tx_channel)
+		dma_release_channel(pl022->dma_tx_channel);
+	if (pl022->dma_rx_channel)
+		dma_release_channel(pl022->dma_rx_channel);
+	kfree(pl022->dummypage);
+}
+
+#else
+static inline int configure_dma(struct pl022 *pl022)
+{
+	return -ENODEV;
+}
+
+static inline int pl022_dma_autoprobe(struct pl022 *pl022)
+{
+	return 0;
+}
+
+static inline int pl022_dma_probe(struct pl022 *pl022)
+{
+	return 0;
+}
+
+static inline void pl022_dma_remove(struct pl022 *pl022)
+{
+}
+#endif
+
+/**
+ * pl022_interrupt_handler - Interrupt handler for SSP controller
+ *
+ * This function handles interrupts generated for an interrupt based transfer.
+ * If a receive overrun (ROR) interrupt is there then we disable SSP, flag the
+ * current message's state as STATE_ERROR and schedule the tasklet
+ * pump_transfers which will do the postprocessing of the current message by
+ * calling giveback(). Otherwise it reads data from RX FIFO till there is no
+ * more data, and writes data in TX FIFO till it is not full. If we complete
+ * the transfer we move to the next transfer and schedule the tasklet.
+ */
+static irqreturn_t pl022_interrupt_handler(int irq, void *dev_id)
+{
+	struct pl022 *pl022 = dev_id;
+	struct spi_message *msg = pl022->cur_msg;
+	u16 irq_status = 0;
+
+	if (unlikely(!msg)) {
+		dev_err(&pl022->adev->dev,
+			"bad message state in interrupt handler");
+		/* Never fail */
+		return IRQ_HANDLED;
+	}
+
+	/* Read the Interrupt Status Register */
+	irq_status = readw(SSP_MIS(pl022->virtbase));
+
+	if (unlikely(!irq_status))
+		return IRQ_NONE;
+
+	/*
+	 * This handles the FIFO interrupts, the timeout
+	 * interrupts are flatly ignored, they cannot be
+	 * trusted.
+	 */
+	if (unlikely(irq_status & SSP_MIS_MASK_RORMIS)) {
+		/*
+		 * Overrun interrupt - bail out since our Data has been
+		 * corrupted
+		 */
+		dev_err(&pl022->adev->dev, "FIFO overrun\n");
+		if (readw(SSP_SR(pl022->virtbase)) & SSP_SR_MASK_RFF)
+			dev_err(&pl022->adev->dev,
+				"RXFIFO is full\n");
+
+		/*
+		 * Disable and clear interrupts, disable SSP,
+		 * mark message with bad status so it can be
+		 * retried.
+		 */
+		writew(DISABLE_ALL_INTERRUPTS,
+		       SSP_IMSC(pl022->virtbase));
+		writew(CLEAR_ALL_INTERRUPTS, SSP_ICR(pl022->virtbase));
+		writew((readw(SSP_CR1(pl022->virtbase)) &
+			(~SSP_CR1_MASK_SSE)), SSP_CR1(pl022->virtbase));
+		msg->state = STATE_ERROR;
+
+		/* Schedule message queue handler */
+		tasklet_schedule(&pl022->pump_transfers);
+		return IRQ_HANDLED;
+	}
+
+	readwriter(pl022);
+
+	if (pl022->tx == pl022->tx_end) {
+		/* Disable Transmit interrupt, enable receive interrupt */
+		writew((readw(SSP_IMSC(pl022->virtbase)) &
+		       ~SSP_IMSC_MASK_TXIM) | SSP_IMSC_MASK_RXIM,
+		       SSP_IMSC(pl022->virtbase));
+	}
+
+	/*
+	 * Since all transactions must write as much as shall be read,
+	 * we can conclude the entire transaction once RX is complete.
+	 * At this point, all TX will always be finished.
+	 */
+	if (pl022->rx >= pl022->rx_end) {
+		writew(DISABLE_ALL_INTERRUPTS,
+		       SSP_IMSC(pl022->virtbase));
+		writew(CLEAR_ALL_INTERRUPTS, SSP_ICR(pl022->virtbase));
+		if (unlikely(pl022->rx > pl022->rx_end)) {
+			dev_warn(&pl022->adev->dev, "read %u surplus "
+				 "bytes (did you request an odd "
+				 "number of bytes on a 16bit bus?)\n",
+				 (u32) (pl022->rx - pl022->rx_end));
+		}
+		/* Update total bytes transferred */
+		msg->actual_length += pl022->cur_transfer->len;
+		if (pl022->cur_transfer->cs_change)
+			pl022_cs_control(pl022, SSP_CHIP_DESELECT);
+		/* Move to next transfer */
+		msg->state = next_transfer(pl022);
+		tasklet_schedule(&pl022->pump_transfers);
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_HANDLED;
+}
+
+/**
+ * This sets up the pointers to memory for the next message to
+ * send out on the SPI bus.
+ */
+static int set_up_next_transfer(struct pl022 *pl022,
+				struct spi_transfer *transfer)
+{
+	int residue;
+
+	/* Sanity check the message for this bus width */
+	residue = pl022->cur_transfer->len % pl022->cur_chip->n_bytes;
+	if (unlikely(residue != 0)) {
+		dev_err(&pl022->adev->dev,
+			"message of %u bytes to transmit but the current "
+			"chip bus has a data width of %u bytes!\n",
+			pl022->cur_transfer->len,
+			pl022->cur_chip->n_bytes);
+		dev_err(&pl022->adev->dev, "skipping this message\n");
+		return -EIO;
+	}
+	pl022->tx = (void *)transfer->tx_buf;
+	pl022->tx_end = pl022->tx + pl022->cur_transfer->len;
+	pl022->rx = (void *)transfer->rx_buf;
+	pl022->rx_end = pl022->rx + pl022->cur_transfer->len;
+	pl022->write =
+	    pl022->tx ? pl022->cur_chip->write : WRITING_NULL;
+	pl022->read = pl022->rx ? pl022->cur_chip->read : READING_NULL;
+	return 0;
+}
+
+/**
+ * pump_transfers - Tasklet function which schedules next transfer
+ * when running in interrupt or DMA transfer mode.
+ * @data: SSP driver private data structure
+ *
+ */
+static void pump_transfers(unsigned long data)
+{
+	struct pl022 *pl022 = (struct pl022 *) data;
+	struct spi_message *message = NULL;
+	struct spi_transfer *transfer = NULL;
+	struct spi_transfer *previous = NULL;
+
+	/* Get current state information */
+	message = pl022->cur_msg;
+	transfer = pl022->cur_transfer;
+
+	/* Handle for abort */
+	if (message->state == STATE_ERROR) {
+		message->status = -EIO;
+		giveback(pl022);
+		return;
+	}
+
+	/* Handle end of message */
+	if (message->state == STATE_DONE) {
+		message->status = 0;
+		giveback(pl022);
+		return;
+	}
+
+	/* Delay if requested at end of transfer before CS change */
+	if (message->state == STATE_RUNNING) {
+		previous = list_entry(transfer->transfer_list.prev,
+					struct spi_transfer,
+					transfer_list);
+		if (previous->delay_usecs)
+			/*
+			 * FIXME: This runs in interrupt context.
+			 * Is this really smart?
+			 */
+			udelay(previous->delay_usecs);
+
+		/* Reselect chip select only if cs_change was requested */
+		if (previous->cs_change)
+			pl022_cs_control(pl022, SSP_CHIP_SELECT);
+	} else {
+		/* STATE_START */
+		message->state = STATE_RUNNING;
+	}
+
+	if (set_up_next_transfer(pl022, transfer)) {
+		message->state = STATE_ERROR;
+		message->status = -EIO;
+		giveback(pl022);
+		return;
+	}
+	/* Flush the FIFOs and let's go! */
+	flush(pl022);
+
+	if (pl022->cur_chip->enable_dma) {
+		if (configure_dma(pl022)) {
+			dev_dbg(&pl022->adev->dev,
+				"configuration of DMA failed, fall back to interrupt mode\n");
+			goto err_config_dma;
+		}
+		return;
+	}
+
+err_config_dma:
+	/* enable all interrupts except RX */
+	writew(ENABLE_ALL_INTERRUPTS & ~SSP_IMSC_MASK_RXIM, SSP_IMSC(pl022->virtbase));
+}
+
+static void do_interrupt_dma_transfer(struct pl022 *pl022)
+{
+	/*
+	 * Default is to enable all interrupts except RX -
+	 * this will be enabled once TX is complete
+	 */
+	u32 irqflags = (u32)(ENABLE_ALL_INTERRUPTS & ~SSP_IMSC_MASK_RXIM);
+
+	/* Enable target chip, if not already active */
+	if (!pl022->next_msg_cs_active)
+		pl022_cs_control(pl022, SSP_CHIP_SELECT);
+
+	if (set_up_next_transfer(pl022, pl022->cur_transfer)) {
+		/* Error path */
+		pl022->cur_msg->state = STATE_ERROR;
+		pl022->cur_msg->status = -EIO;
+		giveback(pl022);
+		return;
+	}
+	/* If we're using DMA, set up DMA here */
+	if (pl022->cur_chip->enable_dma) {
+		/* Configure DMA transfer */
+		if (configure_dma(pl022)) {
+			dev_dbg(&pl022->adev->dev,
+				"configuration of DMA failed, fall back to interrupt mode\n");
+			goto err_config_dma;
+		}
+		/* Disable interrupts in DMA mode, IRQ from DMA controller */
+		irqflags = DISABLE_ALL_INTERRUPTS;
+	}
+err_config_dma:
+	/* Enable SSP, turn on interrupts */
+	writew((readw(SSP_CR1(pl022->virtbase)) | SSP_CR1_MASK_SSE),
+	       SSP_CR1(pl022->virtbase));
+	writew(irqflags, SSP_IMSC(pl022->virtbase));
+}
+
+static void do_polling_transfer(struct pl022 *pl022)
+{
+	struct spi_message *message = NULL;
+	struct spi_transfer *transfer = NULL;
+	struct spi_transfer *previous = NULL;
+	struct chip_data *chip;
+	unsigned long time, timeout;
+
+	chip = pl022->cur_chip;
+	message = pl022->cur_msg;
+
+	while (message->state != STATE_DONE) {
+		/* Handle for abort */
+		if (message->state == STATE_ERROR)
+			break;
+		transfer = pl022->cur_transfer;
+
+		/* Delay if requested at end of transfer */
+		if (message->state == STATE_RUNNING) {
+			previous =
+			    list_entry(transfer->transfer_list.prev,
+				       struct spi_transfer, transfer_list);
+			if (previous->delay_usecs)
+				udelay(previous->delay_usecs);
+			if (previous->cs_change)
+				pl022_cs_control(pl022, SSP_CHIP_SELECT);
+		} else {
+			/* STATE_START */
+			message->state = STATE_RUNNING;
+			if (!pl022->next_msg_cs_active)
+				pl022_cs_control(pl022, SSP_CHIP_SELECT);
+		}
+
+		/* Configuration Changing Per Transfer */
+		if (set_up_next_transfer(pl022, transfer)) {
+			/* Error path */
+			message->state = STATE_ERROR;
+			break;
+		}
+		/* Flush FIFOs and enable SSP */
+		flush(pl022);
+		writew((readw(SSP_CR1(pl022->virtbase)) | SSP_CR1_MASK_SSE),
+		       SSP_CR1(pl022->virtbase));
+
+		dev_dbg(&pl022->adev->dev, "polling transfer ongoing ...\n");
+
+		timeout = jiffies + msecs_to_jiffies(SPI_POLLING_TIMEOUT);
+		while (pl022->tx < pl022->tx_end || pl022->rx < pl022->rx_end) {
+			time = jiffies;
+			readwriter(pl022);
+			if (time_after(time, timeout)) {
+				dev_warn(&pl022->adev->dev,
+				"%s: timeout!\n", __func__);
+				message->state = STATE_ERROR;
+				goto out;
+			}
+			cpu_relax();
+		}
+
+		/* Update total byte transferred */
+		message->actual_length += pl022->cur_transfer->len;
+		if (pl022->cur_transfer->cs_change)
+			pl022_cs_control(pl022, SSP_CHIP_DESELECT);
+		/* Move to next transfer */
+		message->state = next_transfer(pl022);
+	}
+out:
+	/* Handle end of message */
+	if (message->state == STATE_DONE)
+		message->status = 0;
+	else
+		message->status = -EIO;
+
+	giveback(pl022);
+	return;
+}
+
+static int pl022_transfer_one_message(struct spi_master *master,
+				      struct spi_message *msg)
+{
+	struct pl022 *pl022 = spi_master_get_devdata(master);
+
+	/* Initial message state */
+	pl022->cur_msg = msg;
+	msg->state = STATE_START;
+
+	pl022->cur_transfer = list_entry(msg->transfers.next,
+					 struct spi_transfer, transfer_list);
+
+	/* Setup the SPI using the per chip configuration */
+	pl022->cur_chip = spi_get_ctldata(msg->spi);
+	pl022->cur_cs = pl022->chipselects[msg->spi->chip_select];
+
+	restore_state(pl022);
+	flush(pl022);
+
+	if (pl022->cur_chip->xfer_type == POLLING_TRANSFER)
+		do_polling_transfer(pl022);
+	else
+		do_interrupt_dma_transfer(pl022);
+
+	return 0;
+}
+
+static int pl022_unprepare_transfer_hardware(struct spi_master *master)
+{
+	struct pl022 *pl022 = spi_master_get_devdata(master);
+
+	/* nothing more to do - disable spi/ssp and power off */
+	writew((readw(SSP_CR1(pl022->virtbase)) &
+		(~SSP_CR1_MASK_SSE)), SSP_CR1(pl022->virtbase));
+
+	return 0;
+}
+
+static int verify_controller_parameters(struct pl022 *pl022,
+				struct pl022_config_chip const *chip_info)
+{
+	if ((chip_info->iface < SSP_INTERFACE_MOTOROLA_SPI)
+	    || (chip_info->iface > SSP_INTERFACE_UNIDIRECTIONAL)) {
+		dev_err(&pl022->adev->dev,
+			"interface is configured incorrectly\n");
+		return -EINVAL;
+	}
+	if ((chip_info->iface == SSP_INTERFACE_UNIDIRECTIONAL) &&
+	    (!pl022->vendor->unidir)) {
+		dev_err(&pl022->adev->dev,
+			"unidirectional mode not supported in this "
+			"hardware version\n");
+		return -EINVAL;
+	}
+	if ((chip_info->hierarchy != SSP_MASTER)
+	    && (chip_info->hierarchy != SSP_SLAVE)) {
+		dev_err(&pl022->adev->dev,
+			"hierarchy is configured incorrectly\n");
+		return -EINVAL;
+	}
+	if ((chip_info->com_mode != INTERRUPT_TRANSFER)
+	    && (chip_info->com_mode != DMA_TRANSFER)
+	    && (chip_info->com_mode != POLLING_TRANSFER)) {
+		dev_err(&pl022->adev->dev,
+			"Communication mode is configured incorrectly\n");
+		return -EINVAL;
+	}
+	switch (chip_info->rx_lev_trig) {
+	case SSP_RX_1_OR_MORE_ELEM:
+	case SSP_RX_4_OR_MORE_ELEM:
+	case SSP_RX_8_OR_MORE_ELEM:
+		/* These are always OK, all variants can handle this */
+		break;
+	case SSP_RX_16_OR_MORE_ELEM:
+		if (pl022->vendor->fifodepth < 16) {
+			dev_err(&pl022->adev->dev,
+			"RX FIFO Trigger Level is configured incorrectly\n");
+			return -EINVAL;
+		}
+		break;
+	case SSP_RX_32_OR_MORE_ELEM:
+		if (pl022->vendor->fifodepth < 32) {
+			dev_err(&pl022->adev->dev,
+			"RX FIFO Trigger Level is configured incorrectly\n");
+			return -EINVAL;
+		}
+		break;
+	default:
+		dev_err(&pl022->adev->dev,
+			"RX FIFO Trigger Level is configured incorrectly\n");
+		return -EINVAL;
+	}
+	switch (chip_info->tx_lev_trig) {
+	case SSP_TX_1_OR_MORE_EMPTY_LOC:
+	case SSP_TX_4_OR_MORE_EMPTY_LOC:
+	case SSP_TX_8_OR_MORE_EMPTY_LOC:
+		/* These are always OK, all variants can handle this */
+		break;
+	case SSP_TX_16_OR_MORE_EMPTY_LOC:
+		if (pl022->vendor->fifodepth < 16) {
+			dev_err(&pl022->adev->dev,
+			"TX FIFO Trigger Level is configured incorrectly\n");
+			return -EINVAL;
+		}
+		break;
+	case SSP_TX_32_OR_MORE_EMPTY_LOC:
+		if (pl022->vendor->fifodepth < 32) {
+			dev_err(&pl022->adev->dev,
+			"TX FIFO Trigger Level is configured incorrectly\n");
+			return -EINVAL;
+		}
+		break;
+	default:
+		dev_err(&pl022->adev->dev,
+			"TX FIFO Trigger Level is configured incorrectly\n");
+		return -EINVAL;
+	}
+	if (chip_info->iface == SSP_INTERFACE_NATIONAL_MICROWIRE) {
+		if ((chip_info->ctrl_len < SSP_BITS_4)
+		    || (chip_info->ctrl_len > SSP_BITS_32)) {
+			dev_err(&pl022->adev->dev,
+				"CTRL LEN is configured incorrectly\n");
+			return -EINVAL;
+		}
+		if ((chip_info->wait_state != SSP_MWIRE_WAIT_ZERO)
+		    && (chip_info->wait_state != SSP_MWIRE_WAIT_ONE)) {
+			dev_err(&pl022->adev->dev,
+				"Wait State is configured incorrectly\n");
+			return -EINVAL;
+		}
+		/* Half duplex is only available in the ST Micro version */
+		if (pl022->vendor->extended_cr) {
+			if ((chip_info->duplex !=
+			     SSP_MICROWIRE_CHANNEL_FULL_DUPLEX)
+			    && (chip_info->duplex !=
+				SSP_MICROWIRE_CHANNEL_HALF_DUPLEX)) {
+				dev_err(&pl022->adev->dev,
+					"Microwire duplex mode is configured incorrectly\n");
+				return -EINVAL;
+			}
+		} else {
+			if (chip_info->duplex != SSP_MICROWIRE_CHANNEL_FULL_DUPLEX)
+				dev_err(&pl022->adev->dev,
+					"Microwire half duplex mode requested,"
+					" but this is only available in the"
+					" ST version of PL022\n");
+			return -EINVAL;
+		}
+	}
+	return 0;
+}
+
+static inline u32 spi_rate(u32 rate, u16 cpsdvsr, u16 scr)
+{
+	return rate / (cpsdvsr * (1 + scr));
+}
+
+static int calculate_effective_freq(struct pl022 *pl022, int freq, struct
+				    ssp_clock_params * clk_freq)
+{
+	/* Lets calculate the frequency parameters */
+	u16 cpsdvsr = CPSDVR_MIN, scr = SCR_MIN;
+	u32 rate, max_tclk, min_tclk, best_freq = 0, best_cpsdvsr = 0,
+		best_scr = 0, tmp, found = 0;
+
+	rate = clk_get_rate(pl022->clk);
+	/* cpsdvscr = 2 & scr 0 */
+	max_tclk = spi_rate(rate, CPSDVR_MIN, SCR_MIN);
+	/* cpsdvsr = 254 & scr = 255 */
+	min_tclk = spi_rate(rate, CPSDVR_MAX, SCR_MAX);
+
+	if (freq > max_tclk)
+		dev_warn(&pl022->adev->dev,
+			"Max speed that can be programmed is %d Hz, you requested %d\n",
+			max_tclk, freq);
+
+	if (freq < min_tclk) {
+		dev_err(&pl022->adev->dev,
+			"Requested frequency: %d Hz is less than minimum possible %d Hz\n",
+			freq, min_tclk);
+		return -EINVAL;
+	}
+
+	/*
+	 * best_freq will give closest possible available rate (<= requested
+	 * freq) for all values of scr & cpsdvsr.
+	 */
+	while ((cpsdvsr <= CPSDVR_MAX) && !found) {
+		while (scr <= SCR_MAX) {
+			tmp = spi_rate(rate, cpsdvsr, scr);
+
+			if (tmp > freq) {
+				/* we need lower freq */
+				scr++;
+				continue;
+			}
+
+			/*
+			 * If found exact value, mark found and break.
+			 * If found more closer value, update and break.
+			 */
+			if (tmp > best_freq) {
+				best_freq = tmp;
+				best_cpsdvsr = cpsdvsr;
+				best_scr = scr;
+
+				if (tmp == freq)
+					found = 1;
+			}
+			/*
+			 * increased scr will give lower rates, which are not
+			 * required
+			 */
+			break;
+		}
+		cpsdvsr += 2;
+		scr = SCR_MIN;
+	}
+
+	WARN(!best_freq, "pl022: Matching cpsdvsr and scr not found for %d Hz rate \n",
+			freq);
+
+	clk_freq->cpsdvsr = (u8) (best_cpsdvsr & 0xFF);
+	clk_freq->scr = (u8) (best_scr & 0xFF);
+	dev_dbg(&pl022->adev->dev,
+		"SSP Target Frequency is: %u, Effective Frequency is %u\n",
+		freq, best_freq);
+	dev_dbg(&pl022->adev->dev, "SSP cpsdvsr = %d, scr = %d\n",
+		clk_freq->cpsdvsr, clk_freq->scr);
+
+	return 0;
+}
+
+/*
+ * A piece of default chip info unless the platform
+ * supplies it.
+ */
+static const struct pl022_config_chip pl022_default_chip_info = {
+	.com_mode = POLLING_TRANSFER,
+	.iface = SSP_INTERFACE_MOTOROLA_SPI,
+	.hierarchy = SSP_SLAVE,
+	.slave_tx_disable = DO_NOT_DRIVE_TX,
+	.rx_lev_trig = SSP_RX_1_OR_MORE_ELEM,
+	.tx_lev_trig = SSP_TX_1_OR_MORE_EMPTY_LOC,
+	.ctrl_len = SSP_BITS_8,
+	.wait_state = SSP_MWIRE_WAIT_ZERO,
+	.duplex = SSP_MICROWIRE_CHANNEL_FULL_DUPLEX,
+	.cs_control = null_cs_control,
+};
+
+/**
+ * pl022_setup - setup function registered to SPI master framework
+ * @spi: spi device which is requesting setup
+ *
+ * This function is registered to the SPI framework for this SPI master
+ * controller. If it is the first time when setup is called by this device,
+ * this function will initialize the runtime state for this chip and save
+ * the same in the device structure. Else it will update the runtime info
+ * with the updated chip info. Nothing is really being written to the
+ * controller hardware here, that is not done until the actual transfer
+ * commence.
+ */
+static int pl022_setup(struct spi_device *spi)
+{
+	struct pl022_config_chip const *chip_info;
+	struct pl022_config_chip chip_info_dt;
+	struct chip_data *chip;
+	struct ssp_clock_params clk_freq = { .cpsdvsr = 0, .scr = 0};
+	int status = 0;
+	struct pl022 *pl022 = spi_master_get_devdata(spi->master);
+	unsigned int bits = spi->bits_per_word;
+	u32 tmp;
+	struct device_node *np = spi->dev.of_node;
+
+	if (!spi->max_speed_hz)
+		return -EINVAL;
+
+	/* Get controller_state if one is supplied */
+	chip = spi_get_ctldata(spi);
+
+	if (chip == NULL) {
+		chip = kzalloc(sizeof(struct chip_data), GFP_KERNEL);
+		if (!chip)
+			return -ENOMEM;
+		dev_dbg(&spi->dev,
+			"allocated memory for controller's runtime state\n");
+	}
+
+	/* Get controller data if one is supplied */
+	chip_info = spi->controller_data;
+
+	if (chip_info == NULL) {
+		if (np) {
+			chip_info_dt = pl022_default_chip_info;
+
+			chip_info_dt.hierarchy = SSP_MASTER;
+			of_property_read_u32(np, "pl022,interface",
+				&chip_info_dt.iface);
+			of_property_read_u32(np, "pl022,com-mode",
+				&chip_info_dt.com_mode);
+			of_property_read_u32(np, "pl022,rx-level-trig",
+				&chip_info_dt.rx_lev_trig);
+			of_property_read_u32(np, "pl022,tx-level-trig",
+				&chip_info_dt.tx_lev_trig);
+			of_property_read_u32(np, "pl022,ctrl-len",
+				&chip_info_dt.ctrl_len);
+			of_property_read_u32(np, "pl022,wait-state",
+				&chip_info_dt.wait_state);
+			of_property_read_u32(np, "pl022,duplex",
+				&chip_info_dt.duplex);
+
+			chip_info = &chip_info_dt;
+		} else {
+			chip_info = &pl022_default_chip_info;
+			/* spi_board_info.controller_data not is supplied */
+			dev_dbg(&spi->dev,
+				"using default controller_data settings\n");
+		}
+	} else
+		dev_dbg(&spi->dev,
+			"using user supplied controller_data settings\n");
+
+	/*
+	 * We can override with custom divisors, else we use the board
+	 * frequency setting
+	 */
+	if ((0 == chip_info->clk_freq.cpsdvsr)
+	    && (0 == chip_info->clk_freq.scr)) {
+		status = calculate_effective_freq(pl022,
+						  spi->max_speed_hz,
+						  &clk_freq);
+		if (status < 0)
+			goto err_config_params;
+	} else {
+		memcpy(&clk_freq, &chip_info->clk_freq, sizeof(clk_freq));
+		if ((clk_freq.cpsdvsr % 2) != 0)
+			clk_freq.cpsdvsr =
+				clk_freq.cpsdvsr - 1;
+	}
+	if ((clk_freq.cpsdvsr < CPSDVR_MIN)
+	    || (clk_freq.cpsdvsr > CPSDVR_MAX)) {
+		status = -EINVAL;
+		dev_err(&spi->dev,
+			"cpsdvsr is configured incorrectly\n");
+		goto err_config_params;
+	}
+
+	status = verify_controller_parameters(pl022, chip_info);
+	if (status) {
+		dev_err(&spi->dev, "controller data is incorrect");
+		goto err_config_params;
+	}
+
+	pl022->rx_lev_trig = chip_info->rx_lev_trig;
+	pl022->tx_lev_trig = chip_info->tx_lev_trig;
+
+	/* Now set controller state based on controller data */
+	chip->xfer_type = chip_info->com_mode;
+	if (!chip_info->cs_control) {
+		chip->cs_control = null_cs_control;
+		if (!gpio_is_valid(pl022->chipselects[spi->chip_select]))
+			dev_warn(&spi->dev,
+				 "invalid chip select\n");
+	} else
+		chip->cs_control = chip_info->cs_control;
+
+	/* Check bits per word with vendor specific range */
+	if ((bits <= 3) || (bits > pl022->vendor->max_bpw)) {
+		status = -ENOTSUPP;
+		dev_err(&spi->dev, "illegal data size for this controller!\n");
+		dev_err(&spi->dev, "This controller can only handle 4 <= n <= %d bit words\n",
+				pl022->vendor->max_bpw);
+		goto err_config_params;
+	} else if (bits <= 8) {
+		dev_dbg(&spi->dev, "4 <= n <=8 bits per word\n");
+		chip->n_bytes = 1;
+		chip->read = READING_U8;
+		chip->write = WRITING_U8;
+	} else if (bits <= 16) {
+		dev_dbg(&spi->dev, "9 <= n <= 16 bits per word\n");
+		chip->n_bytes = 2;
+		chip->read = READING_U16;
+		chip->write = WRITING_U16;
+	} else {
+		dev_dbg(&spi->dev, "17 <= n <= 32 bits per word\n");
+		chip->n_bytes = 4;
+		chip->read = READING_U32;
+		chip->write = WRITING_U32;
+	}
+
+	/* Now Initialize all register settings required for this chip */
+	chip->cr0 = 0;
+	chip->cr1 = 0;
+	chip->dmacr = 0;
+	chip->cpsr = 0;
+	if ((chip_info->com_mode == DMA_TRANSFER)
+	    && ((pl022->master_info)->enable_dma)) {
+		chip->enable_dma = true;
+		dev_dbg(&spi->dev, "DMA mode set in controller state\n");
+		SSP_WRITE_BITS(chip->dmacr, SSP_DMA_ENABLED,
+			       SSP_DMACR_MASK_RXDMAE, 0);
+		SSP_WRITE_BITS(chip->dmacr, SSP_DMA_ENABLED,
+			       SSP_DMACR_MASK_TXDMAE, 1);
+	} else {
+		chip->enable_dma = false;
+		dev_dbg(&spi->dev, "DMA mode NOT set in controller state\n");
+		SSP_WRITE_BITS(chip->dmacr, SSP_DMA_DISABLED,
+			       SSP_DMACR_MASK_RXDMAE, 0);
+		SSP_WRITE_BITS(chip->dmacr, SSP_DMA_DISABLED,
+			       SSP_DMACR_MASK_TXDMAE, 1);
+	}
+
+	chip->cpsr = clk_freq.cpsdvsr;
+
+	/* Special setup for the ST micro extended control registers */
+	if (pl022->vendor->extended_cr) {
+		u32 etx;
+
+		if (pl022->vendor->pl023) {
+			/* These bits are only in the PL023 */
+			SSP_WRITE_BITS(chip->cr1, chip_info->clkdelay,
+				       SSP_CR1_MASK_FBCLKDEL_ST, 13);
+		} else {
+			/* These bits are in the PL022 but not PL023 */
+			SSP_WRITE_BITS(chip->cr0, chip_info->duplex,
+				       SSP_CR0_MASK_HALFDUP_ST, 5);
+			SSP_WRITE_BITS(chip->cr0, chip_info->ctrl_len,
+				       SSP_CR0_MASK_CSS_ST, 16);
+			SSP_WRITE_BITS(chip->cr0, chip_info->iface,
+				       SSP_CR0_MASK_FRF_ST, 21);
+			SSP_WRITE_BITS(chip->cr1, chip_info->wait_state,
+				       SSP_CR1_MASK_MWAIT_ST, 6);
+		}
+		SSP_WRITE_BITS(chip->cr0, bits - 1,
+			       SSP_CR0_MASK_DSS_ST, 0);
+
+		if (spi->mode & SPI_LSB_FIRST) {
+			tmp = SSP_RX_LSB;
+			etx = SSP_TX_LSB;
+		} else {
+			tmp = SSP_RX_MSB;
+			etx = SSP_TX_MSB;
+		}
+		SSP_WRITE_BITS(chip->cr1, tmp, SSP_CR1_MASK_RENDN_ST, 4);
+		SSP_WRITE_BITS(chip->cr1, etx, SSP_CR1_MASK_TENDN_ST, 5);
+		SSP_WRITE_BITS(chip->cr1, chip_info->rx_lev_trig,
+			       SSP_CR1_MASK_RXIFLSEL_ST, 7);
+		SSP_WRITE_BITS(chip->cr1, chip_info->tx_lev_trig,
+			       SSP_CR1_MASK_TXIFLSEL_ST, 10);
+	} else {
+		SSP_WRITE_BITS(chip->cr0, bits - 1,
+			       SSP_CR0_MASK_DSS, 0);
+		SSP_WRITE_BITS(chip->cr0, chip_info->iface,
+			       SSP_CR0_MASK_FRF, 4);
+	}
+
+	/* Stuff that is common for all versions */
+	if (spi->mode & SPI_CPOL)
+		tmp = SSP_CLK_POL_IDLE_HIGH;
+	else
+		tmp = SSP_CLK_POL_IDLE_LOW;
+	SSP_WRITE_BITS(chip->cr0, tmp, SSP_CR0_MASK_SPO, 6);
+
+	if (spi->mode & SPI_CPHA)
+		tmp = SSP_CLK_SECOND_EDGE;
+	else
+		tmp = SSP_CLK_FIRST_EDGE;
+	SSP_WRITE_BITS(chip->cr0, tmp, SSP_CR0_MASK_SPH, 7);
+
+	SSP_WRITE_BITS(chip->cr0, clk_freq.scr, SSP_CR0_MASK_SCR, 8);
+	/* Loopback is available on all versions except PL023 */
+	if (pl022->vendor->loopback) {
+		if (spi->mode & SPI_LOOP)
+			tmp = LOOPBACK_ENABLED;
+		else
+			tmp = LOOPBACK_DISABLED;
+		SSP_WRITE_BITS(chip->cr1, tmp, SSP_CR1_MASK_LBM, 0);
+	}
+	SSP_WRITE_BITS(chip->cr1, SSP_DISABLED, SSP_CR1_MASK_SSE, 1);
+	SSP_WRITE_BITS(chip->cr1, chip_info->hierarchy, SSP_CR1_MASK_MS, 2);
+	SSP_WRITE_BITS(chip->cr1, chip_info->slave_tx_disable, SSP_CR1_MASK_SOD,
+		3);
+
+	/* Save controller_state */
+	spi_set_ctldata(spi, chip);
+	return status;
+ err_config_params:
+	spi_set_ctldata(spi, NULL);
+	kfree(chip);
+	return status;
+}
+
+/**
+ * pl022_cleanup - cleanup function registered to SPI master framework
+ * @spi: spi device which is requesting cleanup
+ *
+ * This function is registered to the SPI framework for this SPI master
+ * controller. It will free the runtime state of chip.
+ */
+static void pl022_cleanup(struct spi_device *spi)
+{
+	struct chip_data *chip = spi_get_ctldata(spi);
+
+	spi_set_ctldata(spi, NULL);
+	kfree(chip);
+}
+
+static struct pl022_ssp_controller *
+pl022_platform_data_dt_get(struct device *dev)
+{
+	struct device_node *np = dev->of_node;
+	struct pl022_ssp_controller *pd;
+	u32 tmp = 0;
+
+	if (!np) {
+		dev_err(dev, "no dt node defined\n");
+		return NULL;
+	}
+
+	pd = devm_kzalloc(dev, sizeof(struct pl022_ssp_controller), GFP_KERNEL);
+	if (!pd)
+		return NULL;
+
+	pd->bus_id = -1;
+	pd->enable_dma = 1;
+	of_property_read_u32(np, "num-cs", &tmp);
+	pd->num_chipselect = tmp;
+	of_property_read_u32(np, "pl022,autosuspend-delay",
+			     &pd->autosuspend_delay);
+	pd->rt = of_property_read_bool(np, "pl022,rt");
+
+	return pd;
+}
+
+static int pl022_probe(struct amba_device *adev, const struct amba_id *id)
+{
+	struct device *dev = &adev->dev;
+	struct pl022_ssp_controller *platform_info =
+			dev_get_platdata(&adev->dev);
+	struct spi_master *master;
+	struct pl022 *pl022 = NULL;	/*Data for this driver */
+	struct device_node *np = adev->dev.of_node;
+	int status = 0, i, num_cs;
+
+	dev_info(&adev->dev,
+		 "ARM PL022 driver, device ID: 0x%08x\n", adev->periphid);
+	if (!platform_info && IS_ENABLED(CONFIG_OF))
+		platform_info = pl022_platform_data_dt_get(dev);
+
+	if (!platform_info) {
+		dev_err(dev, "probe: no platform data defined\n");
+		return -ENODEV;
+	}
+
+	if (platform_info->num_chipselect) {
+		num_cs = platform_info->num_chipselect;
+	} else {
+		dev_err(dev, "probe: no chip select defined\n");
+		return -ENODEV;
+	}
+
+	/* Allocate master with space for data */
+	master = spi_alloc_master(dev, sizeof(struct pl022));
+	if (master == NULL) {
+		dev_err(&adev->dev, "probe - cannot alloc SPI master\n");
+		return -ENOMEM;
+	}
+
+	pl022 = spi_master_get_devdata(master);
+	pl022->master = master;
+	pl022->master_info = platform_info;
+	pl022->adev = adev;
+	pl022->vendor = id->data;
+	pl022->chipselects = devm_kcalloc(dev, num_cs, sizeof(int),
+					  GFP_KERNEL);
+	if (!pl022->chipselects) {
+		status = -ENOMEM;
+		goto err_no_mem;
+	}
+
+	/*
+	 * Bus Number Which has been Assigned to this SSP controller
+	 * on this board
+	 */
+	master->bus_num = platform_info->bus_id;
+	master->num_chipselect = num_cs;
+	master->cleanup = pl022_cleanup;
+	master->setup = pl022_setup;
+	master->auto_runtime_pm = true;
+	master->transfer_one_message = pl022_transfer_one_message;
+	master->unprepare_transfer_hardware = pl022_unprepare_transfer_hardware;
+	master->rt = platform_info->rt;
+	master->dev.of_node = dev->of_node;
+
+	if (platform_info->num_chipselect && platform_info->chipselects) {
+		for (i = 0; i < num_cs; i++)
+			pl022->chipselects[i] = platform_info->chipselects[i];
+	} else if (pl022->vendor->internal_cs_ctrl) {
+		for (i = 0; i < num_cs; i++)
+			pl022->chipselects[i] = i;
+	} else if (IS_ENABLED(CONFIG_OF)) {
+		for (i = 0; i < num_cs; i++) {
+			int cs_gpio = of_get_named_gpio(np, "cs-gpios", i);
+
+			if (cs_gpio == -EPROBE_DEFER) {
+				status = -EPROBE_DEFER;
+				goto err_no_gpio;
+			}
+
+			pl022->chipselects[i] = cs_gpio;
+
+			if (gpio_is_valid(cs_gpio)) {
+				if (devm_gpio_request(dev, cs_gpio, "ssp-pl022"))
+					dev_err(&adev->dev,
+						"could not request %d gpio\n",
+						cs_gpio);
+				else if (gpio_direction_output(cs_gpio, 1))
+					dev_err(&adev->dev,
+						"could not set gpio %d as output\n",
+						cs_gpio);
+			}
+		}
+	}
+
+	/*
+	 * Supports mode 0-3, loopback, and active low CS. Transfers are
+	 * always MS bit first on the original pl022.
+	 */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LOOP;
+	if (pl022->vendor->extended_cr)
+		master->mode_bits |= SPI_LSB_FIRST;
+
+	dev_dbg(&adev->dev, "BUSNO: %d\n", master->bus_num);
+
+	status = amba_request_regions(adev, NULL);
+	if (status)
+		goto err_no_ioregion;
+
+	pl022->phybase = adev->res.start;
+	pl022->virtbase = devm_ioremap(dev, adev->res.start,
+				       resource_size(&adev->res));
+	if (pl022->virtbase == NULL) {
+		status = -ENOMEM;
+		goto err_no_ioremap;
+	}
+	dev_info(&adev->dev, "mapped registers from %pa to %p\n",
+		&adev->res.start, pl022->virtbase);
+
+	pl022->clk = devm_clk_get(&adev->dev, NULL);
+	if (IS_ERR(pl022->clk)) {
+		status = PTR_ERR(pl022->clk);
+		dev_err(&adev->dev, "could not retrieve SSP/SPI bus clock\n");
+		goto err_no_clk;
+	}
+
+	status = clk_prepare_enable(pl022->clk);
+	if (status) {
+		dev_err(&adev->dev, "could not enable SSP/SPI bus clock\n");
+		goto err_no_clk_en;
+	}
+
+	/* Initialize transfer pump */
+	tasklet_init(&pl022->pump_transfers, pump_transfers,
+		     (unsigned long)pl022);
+
+	/* Disable SSP */
+	writew((readw(SSP_CR1(pl022->virtbase)) & (~SSP_CR1_MASK_SSE)),
+	       SSP_CR1(pl022->virtbase));
+	load_ssp_default_config(pl022);
+
+	status = devm_request_irq(dev, adev->irq[0], pl022_interrupt_handler,
+				  0, "pl022", pl022);
+	if (status < 0) {
+		dev_err(&adev->dev, "probe - cannot get IRQ (%d)\n", status);
+		goto err_no_irq;
+	}
+
+	/* Get DMA channels, try autoconfiguration first */
+	status = pl022_dma_autoprobe(pl022);
+	if (status == -EPROBE_DEFER) {
+		dev_dbg(dev, "deferring probe to get DMA channel\n");
+		goto err_no_irq;
+	}
+
+	/* If that failed, use channels from platform_info */
+	if (status == 0)
+		platform_info->enable_dma = 1;
+	else if (platform_info->enable_dma) {
+		status = pl022_dma_probe(pl022);
+		if (status != 0)
+			platform_info->enable_dma = 0;
+	}
+
+	/* Register with the SPI framework */
+	amba_set_drvdata(adev, pl022);
+	status = devm_spi_register_master(&adev->dev, master);
+	if (status != 0) {
+		dev_err(&adev->dev,
+			"probe - problem registering spi master\n");
+		goto err_spi_register;
+	}
+	dev_dbg(dev, "probe succeeded\n");
+
+	/* let runtime pm put suspend */
+	if (platform_info->autosuspend_delay > 0) {
+		dev_info(&adev->dev,
+			"will use autosuspend for runtime pm, delay %dms\n",
+			platform_info->autosuspend_delay);
+		pm_runtime_set_autosuspend_delay(dev,
+			platform_info->autosuspend_delay);
+		pm_runtime_use_autosuspend(dev);
+	}
+	pm_runtime_put(dev);
+
+	return 0;
+
+ err_spi_register:
+	if (platform_info->enable_dma)
+		pl022_dma_remove(pl022);
+ err_no_irq:
+	clk_disable_unprepare(pl022->clk);
+ err_no_clk_en:
+ err_no_clk:
+ err_no_ioremap:
+	amba_release_regions(adev);
+ err_no_ioregion:
+ err_no_gpio:
+ err_no_mem:
+	spi_master_put(master);
+	return status;
+}
+
+static int
+pl022_remove(struct amba_device *adev)
+{
+	struct pl022 *pl022 = amba_get_drvdata(adev);
+
+	if (!pl022)
+		return 0;
+
+	/*
+	 * undo pm_runtime_put() in probe.  I assume that we're not
+	 * accessing the primecell here.
+	 */
+	pm_runtime_get_noresume(&adev->dev);
+
+	load_ssp_default_config(pl022);
+	if (pl022->master_info->enable_dma)
+		pl022_dma_remove(pl022);
+
+	clk_disable_unprepare(pl022->clk);
+	amba_release_regions(adev);
+	tasklet_disable(&pl022->pump_transfers);
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int pl022_suspend(struct device *dev)
+{
+	struct pl022 *pl022 = dev_get_drvdata(dev);
+	int ret;
+
+	ret = spi_master_suspend(pl022->master);
+	if (ret) {
+		dev_warn(dev, "cannot suspend master\n");
+		return ret;
+	}
+
+	ret = pm_runtime_force_suspend(dev);
+	if (ret) {
+		spi_master_resume(pl022->master);
+		return ret;
+	}
+
+	pinctrl_pm_select_sleep_state(dev);
+
+	dev_dbg(dev, "suspended\n");
+	return 0;
+}
+
+static int pl022_resume(struct device *dev)
+{
+	struct pl022 *pl022 = dev_get_drvdata(dev);
+	int ret;
+
+	ret = pm_runtime_force_resume(dev);
+	if (ret)
+		dev_err(dev, "problem resuming\n");
+
+	/* Start the queue running */
+	ret = spi_master_resume(pl022->master);
+	if (ret)
+		dev_err(dev, "problem starting queue (%d)\n", ret);
+	else
+		dev_dbg(dev, "resumed\n");
+
+	return ret;
+}
+#endif
+
+#ifdef CONFIG_PM
+static int pl022_runtime_suspend(struct device *dev)
+{
+	struct pl022 *pl022 = dev_get_drvdata(dev);
+
+	clk_disable_unprepare(pl022->clk);
+	pinctrl_pm_select_idle_state(dev);
+
+	return 0;
+}
+
+static int pl022_runtime_resume(struct device *dev)
+{
+	struct pl022 *pl022 = dev_get_drvdata(dev);
+
+	pinctrl_pm_select_default_state(dev);
+	clk_prepare_enable(pl022->clk);
+
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops pl022_dev_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(pl022_suspend, pl022_resume)
+	SET_RUNTIME_PM_OPS(pl022_runtime_suspend, pl022_runtime_resume, NULL)
+};
+
+static struct vendor_data vendor_arm = {
+	.fifodepth = 8,
+	.max_bpw = 16,
+	.unidir = false,
+	.extended_cr = false,
+	.pl023 = false,
+	.loopback = true,
+	.internal_cs_ctrl = false,
+};
+
+static struct vendor_data vendor_st = {
+	.fifodepth = 32,
+	.max_bpw = 32,
+	.unidir = false,
+	.extended_cr = true,
+	.pl023 = false,
+	.loopback = true,
+	.internal_cs_ctrl = false,
+};
+
+static struct vendor_data vendor_st_pl023 = {
+	.fifodepth = 32,
+	.max_bpw = 32,
+	.unidir = false,
+	.extended_cr = true,
+	.pl023 = true,
+	.loopback = false,
+	.internal_cs_ctrl = false,
+};
+
+static struct vendor_data vendor_lsi = {
+	.fifodepth = 8,
+	.max_bpw = 16,
+	.unidir = false,
+	.extended_cr = false,
+	.pl023 = false,
+	.loopback = true,
+	.internal_cs_ctrl = true,
+};
+
+static const struct amba_id pl022_ids[] = {
+	{
+		/*
+		 * ARM PL022 variant, this has a 16bit wide
+		 * and 8 locations deep TX/RX FIFO
+		 */
+		.id	= 0x00041022,
+		.mask	= 0x000fffff,
+		.data	= &vendor_arm,
+	},
+	{
+		/*
+		 * ST Micro derivative, this has 32bit wide
+		 * and 32 locations deep TX/RX FIFO
+		 */
+		.id	= 0x01080022,
+		.mask	= 0xffffffff,
+		.data	= &vendor_st,
+	},
+	{
+		/*
+		 * ST-Ericsson derivative "PL023" (this is not
+		 * an official ARM number), this is a PL022 SSP block
+		 * stripped to SPI mode only, it has 32bit wide
+		 * and 32 locations deep TX/RX FIFO but no extended
+		 * CR0/CR1 register
+		 */
+		.id	= 0x00080023,
+		.mask	= 0xffffffff,
+		.data	= &vendor_st_pl023,
+	},
+	{
+		/*
+		 * PL022 variant that has a chip select control register whih
+		 * allows control of 5 output signals nCS[0:4].
+		 */
+		.id	= 0x000b6022,
+		.mask	= 0x000fffff,
+		.data	= &vendor_lsi,
+	},
+	{ 0, 0 },
+};
+
+MODULE_DEVICE_TABLE(amba, pl022_ids);
+
+static struct amba_driver pl022_driver = {
+	.drv = {
+		.name	= "ssp-pl022",
+		.pm	= &pl022_dev_pm_ops,
+	},
+	.id_table	= pl022_ids,
+	.probe		= pl022_probe,
+	.remove		= pl022_remove,
+};
+
+static int __init pl022_init(void)
+{
+	return amba_driver_register(&pl022_driver);
+}
+subsys_initcall(pl022_init);
+
+static void __exit pl022_exit(void)
+{
+	amba_driver_unregister(&pl022_driver);
+}
+module_exit(pl022_exit);
+
+MODULE_AUTHOR("Linus Walleij <linus.walleij@stericsson.com>");
+MODULE_DESCRIPTION("PL022 SSP Controller Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-ppc4xx.c b/drivers/spi/spi-ppc4xx.c
new file mode 100644
index 0000000..967d948
--- /dev/null
+++ b/drivers/spi/spi-ppc4xx.c
@@ -0,0 +1,586 @@
+/*
+ * SPI_PPC4XX SPI controller driver.
+ *
+ * Copyright (C) 2007 Gary Jennejohn <garyj@denx.de>
+ * Copyright 2008 Stefan Roese <sr@denx.de>, DENX Software Engineering
+ * Copyright 2009 Harris Corporation, Steven A. Falco <sfalco@harris.com>
+ *
+ * Based in part on drivers/spi/spi_s3c24xx.c
+ *
+ * Copyright (c) 2006 Ben Dooks
+ * Copyright (c) 2006 Simtec Electronics
+ *	Ben Dooks <ben@simtec.co.uk>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published
+ * by the Free Software Foundation.
+ */
+
+/*
+ * The PPC4xx SPI controller has no FIFO so each sent/received byte will
+ * generate an interrupt to the CPU. This can cause high CPU utilization.
+ * This driver allows platforms to reduce the interrupt load on the CPU
+ * during SPI transfers by setting max_speed_hz via the device tree.
+ */
+
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
+#include <linux/errno.h>
+#include <linux/wait.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/of_platform.h>
+#include <linux/of_gpio.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+
+#include <linux/gpio.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+
+#include <asm/io.h>
+#include <asm/dcr.h>
+#include <asm/dcr-regs.h>
+
+/* bits in mode register - bit 0 is MSb */
+
+/*
+ * SPI_PPC4XX_MODE_SCP = 0 means "data latched on trailing edge of clock"
+ * SPI_PPC4XX_MODE_SCP = 1 means "data latched on leading edge of clock"
+ * Note: This is the inverse of CPHA.
+ */
+#define SPI_PPC4XX_MODE_SCP	(0x80 >> 3)
+
+/* SPI_PPC4XX_MODE_SPE = 1 means "port enabled" */
+#define SPI_PPC4XX_MODE_SPE	(0x80 >> 4)
+
+/*
+ * SPI_PPC4XX_MODE_RD = 0 means "MSB first" - this is the normal mode
+ * SPI_PPC4XX_MODE_RD = 1 means "LSB first" - this is bit-reversed mode
+ * Note: This is identical to SPI_LSB_FIRST.
+ */
+#define SPI_PPC4XX_MODE_RD	(0x80 >> 5)
+
+/*
+ * SPI_PPC4XX_MODE_CI = 0 means "clock idles low"
+ * SPI_PPC4XX_MODE_CI = 1 means "clock idles high"
+ * Note: This is identical to CPOL.
+ */
+#define SPI_PPC4XX_MODE_CI	(0x80 >> 6)
+
+/*
+ * SPI_PPC4XX_MODE_IL = 0 means "loopback disable"
+ * SPI_PPC4XX_MODE_IL = 1 means "loopback enable"
+ */
+#define SPI_PPC4XX_MODE_IL	(0x80 >> 7)
+
+/* bits in control register */
+/* starts a transfer when set */
+#define SPI_PPC4XX_CR_STR	(0x80 >> 7)
+
+/* bits in status register */
+/* port is busy with a transfer */
+#define SPI_PPC4XX_SR_BSY	(0x80 >> 6)
+/* RxD ready */
+#define SPI_PPC4XX_SR_RBR	(0x80 >> 7)
+
+/* clock settings (SCP and CI) for various SPI modes */
+#define SPI_CLK_MODE0	(SPI_PPC4XX_MODE_SCP | 0)
+#define SPI_CLK_MODE1	(0 | 0)
+#define SPI_CLK_MODE2	(SPI_PPC4XX_MODE_SCP | SPI_PPC4XX_MODE_CI)
+#define SPI_CLK_MODE3	(0 | SPI_PPC4XX_MODE_CI)
+
+#define DRIVER_NAME	"spi_ppc4xx_of"
+
+struct spi_ppc4xx_regs {
+	u8 mode;
+	u8 rxd;
+	u8 txd;
+	u8 cr;
+	u8 sr;
+	u8 dummy;
+	/*
+	 * Clock divisor modulus register
+	 * This uses the following formula:
+	 *    SCPClkOut = OPBCLK/(4(CDM + 1))
+	 * or
+	 *    CDM = (OPBCLK/4*SCPClkOut) - 1
+	 * bit 0 is the MSb!
+	 */
+	u8 cdm;
+};
+
+/* SPI Controller driver's private data. */
+struct ppc4xx_spi {
+	/* bitbang has to be first */
+	struct spi_bitbang bitbang;
+	struct completion done;
+
+	u64 mapbase;
+	u64 mapsize;
+	int irqnum;
+	/* need this to set the SPI clock */
+	unsigned int opb_freq;
+
+	/* for transfers */
+	int len;
+	int count;
+	/* data buffers */
+	const unsigned char *tx;
+	unsigned char *rx;
+
+	int *gpios;
+
+	struct spi_ppc4xx_regs __iomem *regs; /* pointer to the registers */
+	struct spi_master *master;
+	struct device *dev;
+};
+
+/* need this so we can set the clock in the chipselect routine */
+struct spi_ppc4xx_cs {
+	u8 mode;
+};
+
+static int spi_ppc4xx_txrx(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct ppc4xx_spi *hw;
+	u8 data;
+
+	dev_dbg(&spi->dev, "txrx: tx %p, rx %p, len %d\n",
+		t->tx_buf, t->rx_buf, t->len);
+
+	hw = spi_master_get_devdata(spi->master);
+
+	hw->tx = t->tx_buf;
+	hw->rx = t->rx_buf;
+	hw->len = t->len;
+	hw->count = 0;
+
+	/* send the first byte */
+	data = hw->tx ? hw->tx[0] : 0;
+	out_8(&hw->regs->txd, data);
+	out_8(&hw->regs->cr, SPI_PPC4XX_CR_STR);
+	wait_for_completion(&hw->done);
+
+	return hw->count;
+}
+
+static int spi_ppc4xx_setupxfer(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct ppc4xx_spi *hw = spi_master_get_devdata(spi->master);
+	struct spi_ppc4xx_cs *cs = spi->controller_state;
+	int scr;
+	u8 cdm = 0;
+	u32 speed;
+	u8 bits_per_word;
+
+	/* Start with the generic configuration for this device. */
+	bits_per_word = spi->bits_per_word;
+	speed = spi->max_speed_hz;
+
+	/*
+	 * Modify the configuration if the transfer overrides it.  Do not allow
+	 * the transfer to overwrite the generic configuration with zeros.
+	 */
+	if (t) {
+		if (t->bits_per_word)
+			bits_per_word = t->bits_per_word;
+
+		if (t->speed_hz)
+			speed = min(t->speed_hz, spi->max_speed_hz);
+	}
+
+	if (!speed || (speed > spi->max_speed_hz)) {
+		dev_err(&spi->dev, "invalid speed_hz (%d)\n", speed);
+		return -EINVAL;
+	}
+
+	/* Write new configuration */
+	out_8(&hw->regs->mode, cs->mode);
+
+	/* Set the clock */
+	/* opb_freq was already divided by 4 */
+	scr = (hw->opb_freq / speed) - 1;
+	if (scr > 0)
+		cdm = min(scr, 0xff);
+
+	dev_dbg(&spi->dev, "setting pre-scaler to %d (hz %d)\n", cdm, speed);
+
+	if (in_8(&hw->regs->cdm) != cdm)
+		out_8(&hw->regs->cdm, cdm);
+
+	mutex_lock(&hw->bitbang.lock);
+	if (!hw->bitbang.busy) {
+		hw->bitbang.chipselect(spi, BITBANG_CS_INACTIVE);
+		/* Need to ndelay here? */
+	}
+	mutex_unlock(&hw->bitbang.lock);
+
+	return 0;
+}
+
+static int spi_ppc4xx_setup(struct spi_device *spi)
+{
+	struct spi_ppc4xx_cs *cs = spi->controller_state;
+
+	if (!spi->max_speed_hz) {
+		dev_err(&spi->dev, "invalid max_speed_hz (must be non-zero)\n");
+		return -EINVAL;
+	}
+
+	if (cs == NULL) {
+		cs = kzalloc(sizeof *cs, GFP_KERNEL);
+		if (!cs)
+			return -ENOMEM;
+		spi->controller_state = cs;
+	}
+
+	/*
+	 * We set all bits of the SPI0_MODE register, so,
+	 * no need to read-modify-write
+	 */
+	cs->mode = SPI_PPC4XX_MODE_SPE;
+
+	switch (spi->mode & (SPI_CPHA | SPI_CPOL)) {
+	case SPI_MODE_0:
+		cs->mode |= SPI_CLK_MODE0;
+		break;
+	case SPI_MODE_1:
+		cs->mode |= SPI_CLK_MODE1;
+		break;
+	case SPI_MODE_2:
+		cs->mode |= SPI_CLK_MODE2;
+		break;
+	case SPI_MODE_3:
+		cs->mode |= SPI_CLK_MODE3;
+		break;
+	}
+
+	if (spi->mode & SPI_LSB_FIRST)
+		cs->mode |= SPI_PPC4XX_MODE_RD;
+
+	return 0;
+}
+
+static void spi_ppc4xx_chipsel(struct spi_device *spi, int value)
+{
+	struct ppc4xx_spi *hw = spi_master_get_devdata(spi->master);
+	unsigned int cs = spi->chip_select;
+	unsigned int cspol;
+
+	/*
+	 * If there are no chip selects at all, or if this is the special
+	 * case of a non-existent (dummy) chip select, do nothing.
+	 */
+
+	if (!hw->master->num_chipselect || hw->gpios[cs] == -EEXIST)
+		return;
+
+	cspol = spi->mode & SPI_CS_HIGH ? 1 : 0;
+	if (value == BITBANG_CS_INACTIVE)
+		cspol = !cspol;
+
+	gpio_set_value(hw->gpios[cs], cspol);
+}
+
+static irqreturn_t spi_ppc4xx_int(int irq, void *dev_id)
+{
+	struct ppc4xx_spi *hw;
+	u8 status;
+	u8 data;
+	unsigned int count;
+
+	hw = (struct ppc4xx_spi *)dev_id;
+
+	status = in_8(&hw->regs->sr);
+	if (!status)
+		return IRQ_NONE;
+
+	/*
+	 * BSY de-asserts one cycle after the transfer is complete.  The
+	 * interrupt is asserted after the transfer is complete.  The exact
+	 * relationship is not documented, hence this code.
+	 */
+
+	if (unlikely(status & SPI_PPC4XX_SR_BSY)) {
+		u8 lstatus;
+		int cnt = 0;
+
+		dev_dbg(hw->dev, "got interrupt but spi still busy?\n");
+		do {
+			ndelay(10);
+			lstatus = in_8(&hw->regs->sr);
+		} while (++cnt < 100 && lstatus & SPI_PPC4XX_SR_BSY);
+
+		if (cnt >= 100) {
+			dev_err(hw->dev, "busywait: too many loops!\n");
+			complete(&hw->done);
+			return IRQ_HANDLED;
+		} else {
+			/* status is always 1 (RBR) here */
+			status = in_8(&hw->regs->sr);
+			dev_dbg(hw->dev, "loops %d status %x\n", cnt, status);
+		}
+	}
+
+	count = hw->count;
+	hw->count++;
+
+	/* RBR triggered this interrupt.  Therefore, data must be ready. */
+	data = in_8(&hw->regs->rxd);
+	if (hw->rx)
+		hw->rx[count] = data;
+
+	count++;
+
+	if (count < hw->len) {
+		data = hw->tx ? hw->tx[count] : 0;
+		out_8(&hw->regs->txd, data);
+		out_8(&hw->regs->cr, SPI_PPC4XX_CR_STR);
+	} else {
+		complete(&hw->done);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static void spi_ppc4xx_cleanup(struct spi_device *spi)
+{
+	kfree(spi->controller_state);
+}
+
+static void spi_ppc4xx_enable(struct ppc4xx_spi *hw)
+{
+	/*
+	 * On all 4xx PPC's the SPI bus is shared/multiplexed with
+	 * the 2nd I2C bus. We need to enable the the SPI bus before
+	 * using it.
+	 */
+
+	/* need to clear bit 14 to enable SPC */
+	dcri_clrset(SDR0, SDR0_PFC1, 0x80000000 >> 14, 0);
+}
+
+static void free_gpios(struct ppc4xx_spi *hw)
+{
+	if (hw->master->num_chipselect) {
+		int i;
+		for (i = 0; i < hw->master->num_chipselect; i++)
+			if (gpio_is_valid(hw->gpios[i]))
+				gpio_free(hw->gpios[i]);
+
+		kfree(hw->gpios);
+		hw->gpios = NULL;
+	}
+}
+
+/*
+ * platform_device layer stuff...
+ */
+static int spi_ppc4xx_of_probe(struct platform_device *op)
+{
+	struct ppc4xx_spi *hw;
+	struct spi_master *master;
+	struct spi_bitbang *bbp;
+	struct resource resource;
+	struct device_node *np = op->dev.of_node;
+	struct device *dev = &op->dev;
+	struct device_node *opbnp;
+	int ret;
+	int num_gpios;
+	const unsigned int *clk;
+
+	master = spi_alloc_master(dev, sizeof *hw);
+	if (master == NULL)
+		return -ENOMEM;
+	master->dev.of_node = np;
+	platform_set_drvdata(op, master);
+	hw = spi_master_get_devdata(master);
+	hw->master = master;
+	hw->dev = dev;
+
+	init_completion(&hw->done);
+
+	/*
+	 * A count of zero implies a single SPI device without any chip-select.
+	 * Note that of_gpio_count counts all gpios assigned to this spi master.
+	 * This includes both "null" gpio's and real ones.
+	 */
+	num_gpios = of_gpio_count(np);
+	if (num_gpios > 0) {
+		int i;
+
+		hw->gpios = kcalloc(num_gpios, sizeof(*hw->gpios), GFP_KERNEL);
+		if (!hw->gpios) {
+			ret = -ENOMEM;
+			goto free_master;
+		}
+
+		for (i = 0; i < num_gpios; i++) {
+			int gpio;
+			enum of_gpio_flags flags;
+
+			gpio = of_get_gpio_flags(np, i, &flags);
+			hw->gpios[i] = gpio;
+
+			if (gpio_is_valid(gpio)) {
+				/* Real CS - set the initial state. */
+				ret = gpio_request(gpio, np->name);
+				if (ret < 0) {
+					dev_err(dev,
+						"can't request gpio #%d: %d\n",
+						i, ret);
+					goto free_gpios;
+				}
+
+				gpio_direction_output(gpio,
+						!!(flags & OF_GPIO_ACTIVE_LOW));
+			} else if (gpio == -EEXIST) {
+				; /* No CS, but that's OK. */
+			} else {
+				dev_err(dev, "invalid gpio #%d: %d\n", i, gpio);
+				ret = -EINVAL;
+				goto free_gpios;
+			}
+		}
+	}
+
+	/* Setup the state for the bitbang driver */
+	bbp = &hw->bitbang;
+	bbp->master = hw->master;
+	bbp->setup_transfer = spi_ppc4xx_setupxfer;
+	bbp->chipselect = spi_ppc4xx_chipsel;
+	bbp->txrx_bufs = spi_ppc4xx_txrx;
+	bbp->use_dma = 0;
+	bbp->master->setup = spi_ppc4xx_setup;
+	bbp->master->cleanup = spi_ppc4xx_cleanup;
+	bbp->master->bits_per_word_mask = SPI_BPW_MASK(8);
+
+	/* the spi->mode bits understood by this driver: */
+	bbp->master->mode_bits =
+		SPI_CPHA | SPI_CPOL | SPI_CS_HIGH | SPI_LSB_FIRST;
+
+	/* this many pins in all GPIO controllers */
+	bbp->master->num_chipselect = num_gpios > 0 ? num_gpios : 0;
+
+	/* Get the clock for the OPB */
+	opbnp = of_find_compatible_node(NULL, NULL, "ibm,opb");
+	if (opbnp == NULL) {
+		dev_err(dev, "OPB: cannot find node\n");
+		ret = -ENODEV;
+		goto free_gpios;
+	}
+	/* Get the clock (Hz) for the OPB */
+	clk = of_get_property(opbnp, "clock-frequency", NULL);
+	if (clk == NULL) {
+		dev_err(dev, "OPB: no clock-frequency property set\n");
+		of_node_put(opbnp);
+		ret = -ENODEV;
+		goto free_gpios;
+	}
+	hw->opb_freq = *clk;
+	hw->opb_freq >>= 2;
+	of_node_put(opbnp);
+
+	ret = of_address_to_resource(np, 0, &resource);
+	if (ret) {
+		dev_err(dev, "error while parsing device node resource\n");
+		goto free_gpios;
+	}
+	hw->mapbase = resource.start;
+	hw->mapsize = resource_size(&resource);
+
+	/* Sanity check */
+	if (hw->mapsize < sizeof(struct spi_ppc4xx_regs)) {
+		dev_err(dev, "too small to map registers\n");
+		ret = -EINVAL;
+		goto free_gpios;
+	}
+
+	/* Request IRQ */
+	hw->irqnum = irq_of_parse_and_map(np, 0);
+	ret = request_irq(hw->irqnum, spi_ppc4xx_int,
+			  0, "spi_ppc4xx_of", (void *)hw);
+	if (ret) {
+		dev_err(dev, "unable to allocate interrupt\n");
+		goto free_gpios;
+	}
+
+	if (!request_mem_region(hw->mapbase, hw->mapsize, DRIVER_NAME)) {
+		dev_err(dev, "resource unavailable\n");
+		ret = -EBUSY;
+		goto request_mem_error;
+	}
+
+	hw->regs = ioremap(hw->mapbase, sizeof(struct spi_ppc4xx_regs));
+
+	if (!hw->regs) {
+		dev_err(dev, "unable to memory map registers\n");
+		ret = -ENXIO;
+		goto map_io_error;
+	}
+
+	spi_ppc4xx_enable(hw);
+
+	/* Finally register our spi controller */
+	dev->dma_mask = 0;
+	ret = spi_bitbang_start(bbp);
+	if (ret) {
+		dev_err(dev, "failed to register SPI master\n");
+		goto unmap_regs;
+	}
+
+	dev_info(dev, "driver initialized\n");
+
+	return 0;
+
+unmap_regs:
+	iounmap(hw->regs);
+map_io_error:
+	release_mem_region(hw->mapbase, hw->mapsize);
+request_mem_error:
+	free_irq(hw->irqnum, hw);
+free_gpios:
+	free_gpios(hw);
+free_master:
+	spi_master_put(master);
+
+	dev_err(dev, "initialization failed\n");
+	return ret;
+}
+
+static int spi_ppc4xx_of_remove(struct platform_device *op)
+{
+	struct spi_master *master = platform_get_drvdata(op);
+	struct ppc4xx_spi *hw = spi_master_get_devdata(master);
+
+	spi_bitbang_stop(&hw->bitbang);
+	release_mem_region(hw->mapbase, hw->mapsize);
+	free_irq(hw->irqnum, hw);
+	iounmap(hw->regs);
+	free_gpios(hw);
+	spi_master_put(master);
+	return 0;
+}
+
+static const struct of_device_id spi_ppc4xx_of_match[] = {
+	{ .compatible = "ibm,ppc4xx-spi", },
+	{},
+};
+
+MODULE_DEVICE_TABLE(of, spi_ppc4xx_of_match);
+
+static struct platform_driver spi_ppc4xx_of_driver = {
+	.probe = spi_ppc4xx_of_probe,
+	.remove = spi_ppc4xx_of_remove,
+	.driver = {
+		.name = DRIVER_NAME,
+		.of_match_table = spi_ppc4xx_of_match,
+	},
+};
+module_platform_driver(spi_ppc4xx_of_driver);
+
+MODULE_AUTHOR("Gary Jennejohn & Stefan Roese");
+MODULE_DESCRIPTION("Simple PPC4xx SPI Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-pxa2xx-dma.c b/drivers/spi/spi-pxa2xx-dma.c
new file mode 100644
index 0000000..2fa7f4b
--- /dev/null
+++ b/drivers/spi/spi-pxa2xx-dma.c
@@ -0,0 +1,253 @@
+/*
+ * PXA2xx SPI DMA engine support.
+ *
+ * Copyright (C) 2013, Intel Corporation
+ * Author: Mika Westerberg <mika.westerberg@linux.intel.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/device.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/pxa2xx_ssp.h>
+#include <linux/scatterlist.h>
+#include <linux/sizes.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/pxa2xx_spi.h>
+
+#include "spi-pxa2xx.h"
+
+static void pxa2xx_spi_dma_transfer_complete(struct driver_data *drv_data,
+					     bool error)
+{
+	struct spi_message *msg = drv_data->master->cur_msg;
+
+	/*
+	 * It is possible that one CPU is handling ROR interrupt and other
+	 * just gets DMA completion. Calling pump_transfers() twice for the
+	 * same transfer leads to problems thus we prevent concurrent calls
+	 * by using ->dma_running.
+	 */
+	if (atomic_dec_and_test(&drv_data->dma_running)) {
+		/*
+		 * If the other CPU is still handling the ROR interrupt we
+		 * might not know about the error yet. So we re-check the
+		 * ROR bit here before we clear the status register.
+		 */
+		if (!error) {
+			u32 status = pxa2xx_spi_read(drv_data, SSSR)
+				     & drv_data->mask_sr;
+			error = status & SSSR_ROR;
+		}
+
+		/* Clear status & disable interrupts */
+		pxa2xx_spi_write(drv_data, SSCR1,
+				 pxa2xx_spi_read(drv_data, SSCR1)
+				 & ~drv_data->dma_cr1);
+		write_SSSR_CS(drv_data, drv_data->clear_sr);
+		if (!pxa25x_ssp_comp(drv_data))
+			pxa2xx_spi_write(drv_data, SSTO, 0);
+
+		if (error) {
+			/* In case we got an error we disable the SSP now */
+			pxa2xx_spi_write(drv_data, SSCR0,
+					 pxa2xx_spi_read(drv_data, SSCR0)
+					 & ~SSCR0_SSE);
+			msg->status = -EIO;
+		}
+
+		spi_finalize_current_transfer(drv_data->master);
+	}
+}
+
+static void pxa2xx_spi_dma_callback(void *data)
+{
+	pxa2xx_spi_dma_transfer_complete(data, false);
+}
+
+static struct dma_async_tx_descriptor *
+pxa2xx_spi_dma_prepare_one(struct driver_data *drv_data,
+			   enum dma_transfer_direction dir,
+			   struct spi_transfer *xfer)
+{
+	struct chip_data *chip =
+		spi_get_ctldata(drv_data->master->cur_msg->spi);
+	enum dma_slave_buswidth width;
+	struct dma_slave_config cfg;
+	struct dma_chan *chan;
+	struct sg_table *sgt;
+	int ret;
+
+	switch (drv_data->n_bytes) {
+	case 1:
+		width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		break;
+	case 2:
+		width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		break;
+	default:
+		width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		break;
+	}
+
+	memset(&cfg, 0, sizeof(cfg));
+	cfg.direction = dir;
+
+	if (dir == DMA_MEM_TO_DEV) {
+		cfg.dst_addr = drv_data->ssdr_physical;
+		cfg.dst_addr_width = width;
+		cfg.dst_maxburst = chip->dma_burst_size;
+
+		sgt = &xfer->tx_sg;
+		chan = drv_data->master->dma_tx;
+	} else {
+		cfg.src_addr = drv_data->ssdr_physical;
+		cfg.src_addr_width = width;
+		cfg.src_maxburst = chip->dma_burst_size;
+
+		sgt = &xfer->rx_sg;
+		chan = drv_data->master->dma_rx;
+	}
+
+	ret = dmaengine_slave_config(chan, &cfg);
+	if (ret) {
+		dev_warn(&drv_data->pdev->dev, "DMA slave config failed\n");
+		return NULL;
+	}
+
+	return dmaengine_prep_slave_sg(chan, sgt->sgl, sgt->nents, dir,
+				       DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+}
+
+irqreturn_t pxa2xx_spi_dma_transfer(struct driver_data *drv_data)
+{
+	u32 status;
+
+	status = pxa2xx_spi_read(drv_data, SSSR) & drv_data->mask_sr;
+	if (status & SSSR_ROR) {
+		dev_err(&drv_data->pdev->dev, "FIFO overrun\n");
+
+		dmaengine_terminate_async(drv_data->master->dma_rx);
+		dmaengine_terminate_async(drv_data->master->dma_tx);
+
+		pxa2xx_spi_dma_transfer_complete(drv_data, true);
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_NONE;
+}
+
+int pxa2xx_spi_dma_prepare(struct driver_data *drv_data,
+			   struct spi_transfer *xfer)
+{
+	struct dma_async_tx_descriptor *tx_desc, *rx_desc;
+	int err;
+
+	tx_desc = pxa2xx_spi_dma_prepare_one(drv_data, DMA_MEM_TO_DEV, xfer);
+	if (!tx_desc) {
+		dev_err(&drv_data->pdev->dev,
+			"failed to get DMA TX descriptor\n");
+		err = -EBUSY;
+		goto err_tx;
+	}
+
+	rx_desc = pxa2xx_spi_dma_prepare_one(drv_data, DMA_DEV_TO_MEM, xfer);
+	if (!rx_desc) {
+		dev_err(&drv_data->pdev->dev,
+			"failed to get DMA RX descriptor\n");
+		err = -EBUSY;
+		goto err_rx;
+	}
+
+	/* We are ready when RX completes */
+	rx_desc->callback = pxa2xx_spi_dma_callback;
+	rx_desc->callback_param = drv_data;
+
+	dmaengine_submit(rx_desc);
+	dmaengine_submit(tx_desc);
+	return 0;
+
+err_rx:
+	dmaengine_terminate_async(drv_data->master->dma_tx);
+err_tx:
+	return err;
+}
+
+void pxa2xx_spi_dma_start(struct driver_data *drv_data)
+{
+	dma_async_issue_pending(drv_data->master->dma_rx);
+	dma_async_issue_pending(drv_data->master->dma_tx);
+
+	atomic_set(&drv_data->dma_running, 1);
+}
+
+void pxa2xx_spi_dma_stop(struct driver_data *drv_data)
+{
+	atomic_set(&drv_data->dma_running, 0);
+	dmaengine_terminate_sync(drv_data->master->dma_rx);
+	dmaengine_terminate_sync(drv_data->master->dma_tx);
+}
+
+int pxa2xx_spi_dma_setup(struct driver_data *drv_data)
+{
+	struct pxa2xx_spi_master *pdata = drv_data->master_info;
+	struct device *dev = &drv_data->pdev->dev;
+	struct spi_controller *master = drv_data->master;
+	dma_cap_mask_t mask;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	master->dma_tx = dma_request_slave_channel_compat(mask,
+				pdata->dma_filter, pdata->tx_param, dev, "tx");
+	if (!master->dma_tx)
+		return -ENODEV;
+
+	master->dma_rx = dma_request_slave_channel_compat(mask,
+				pdata->dma_filter, pdata->rx_param, dev, "rx");
+	if (!master->dma_rx) {
+		dma_release_channel(master->dma_tx);
+		master->dma_tx = NULL;
+		return -ENODEV;
+	}
+
+	return 0;
+}
+
+void pxa2xx_spi_dma_release(struct driver_data *drv_data)
+{
+	struct spi_controller *master = drv_data->master;
+
+	if (master->dma_rx) {
+		dmaengine_terminate_sync(master->dma_rx);
+		dma_release_channel(master->dma_rx);
+		master->dma_rx = NULL;
+	}
+	if (master->dma_tx) {
+		dmaengine_terminate_sync(master->dma_tx);
+		dma_release_channel(master->dma_tx);
+		master->dma_tx = NULL;
+	}
+}
+
+int pxa2xx_spi_set_dma_burst_and_threshold(struct chip_data *chip,
+					   struct spi_device *spi,
+					   u8 bits_per_word, u32 *burst_code,
+					   u32 *threshold)
+{
+	struct pxa2xx_spi_chip *chip_info = spi->controller_data;
+
+	/*
+	 * If the DMA burst size is given in chip_info we use that,
+	 * otherwise we use the default. Also we use the default FIFO
+	 * thresholds for now.
+	 */
+	*burst_code = chip_info ? chip_info->dma_burst_size : 1;
+	*threshold = SSCR1_RxTresh(RX_THRESH_DFLT)
+		   | SSCR1_TxTresh(TX_THRESH_DFLT);
+
+	return 0;
+}
diff --git a/drivers/spi/spi-pxa2xx-pci.c b/drivers/spi/spi-pxa2xx-pci.c
new file mode 100644
index 0000000..869f188
--- /dev/null
+++ b/drivers/spi/spi-pxa2xx-pci.c
@@ -0,0 +1,300 @@
+/*
+ * CE4100's SPI device is more or less the same one as found on PXA
+ *
+ * Copyright (C) 2016, Intel Corporation
+ */
+#include <linux/clk-provider.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/pci.h>
+#include <linux/platform_device.h>
+#include <linux/spi/pxa2xx_spi.h>
+
+#include <linux/dmaengine.h>
+#include <linux/platform_data/dma-dw.h>
+
+enum {
+	PORT_QUARK_X1000,
+	PORT_BYT,
+	PORT_MRFLD,
+	PORT_BSW0,
+	PORT_BSW1,
+	PORT_BSW2,
+	PORT_CE4100,
+	PORT_LPT,
+};
+
+struct pxa_spi_info {
+	enum pxa_ssp_type type;
+	int port_id;
+	int num_chipselect;
+	unsigned long max_clk_rate;
+
+	/* DMA channel request parameters */
+	bool (*dma_filter)(struct dma_chan *chan, void *param);
+	void *tx_param;
+	void *rx_param;
+
+	int (*setup)(struct pci_dev *pdev, struct pxa_spi_info *c);
+};
+
+static struct dw_dma_slave byt_tx_param = { .dst_id = 0 };
+static struct dw_dma_slave byt_rx_param = { .src_id = 1 };
+
+static struct dw_dma_slave mrfld3_tx_param = { .dst_id = 15 };
+static struct dw_dma_slave mrfld3_rx_param = { .src_id = 14 };
+static struct dw_dma_slave mrfld5_tx_param = { .dst_id = 13 };
+static struct dw_dma_slave mrfld5_rx_param = { .src_id = 12 };
+static struct dw_dma_slave mrfld6_tx_param = { .dst_id = 11 };
+static struct dw_dma_slave mrfld6_rx_param = { .src_id = 10 };
+
+static struct dw_dma_slave bsw0_tx_param = { .dst_id = 0 };
+static struct dw_dma_slave bsw0_rx_param = { .src_id = 1 };
+static struct dw_dma_slave bsw1_tx_param = { .dst_id = 6 };
+static struct dw_dma_slave bsw1_rx_param = { .src_id = 7 };
+static struct dw_dma_slave bsw2_tx_param = { .dst_id = 8 };
+static struct dw_dma_slave bsw2_rx_param = { .src_id = 9 };
+
+static struct dw_dma_slave lpt_tx_param = { .dst_id = 0 };
+static struct dw_dma_slave lpt_rx_param = { .src_id = 1 };
+
+static bool lpss_dma_filter(struct dma_chan *chan, void *param)
+{
+	struct dw_dma_slave *dws = param;
+
+	if (dws->dma_dev != chan->device->dev)
+		return false;
+
+	chan->private = dws;
+	return true;
+}
+
+static int lpss_spi_setup(struct pci_dev *dev, struct pxa_spi_info *c)
+{
+	struct pci_dev *dma_dev;
+
+	c->num_chipselect = 1;
+	c->max_clk_rate = 50000000;
+
+	dma_dev = pci_get_slot(dev->bus, PCI_DEVFN(PCI_SLOT(dev->devfn), 0));
+
+	if (c->tx_param) {
+		struct dw_dma_slave *slave = c->tx_param;
+
+		slave->dma_dev = &dma_dev->dev;
+		slave->m_master = 0;
+		slave->p_master = 1;
+	}
+
+	if (c->rx_param) {
+		struct dw_dma_slave *slave = c->rx_param;
+
+		slave->dma_dev = &dma_dev->dev;
+		slave->m_master = 0;
+		slave->p_master = 1;
+	}
+
+	c->dma_filter = lpss_dma_filter;
+	return 0;
+}
+
+static int mrfld_spi_setup(struct pci_dev *dev, struct pxa_spi_info *c)
+{
+	struct pci_dev *dma_dev = pci_get_slot(dev->bus, PCI_DEVFN(21, 0));
+	struct dw_dma_slave *tx, *rx;
+
+	switch (PCI_FUNC(dev->devfn)) {
+	case 0:
+		c->port_id = 3;
+		c->num_chipselect = 1;
+		c->tx_param = &mrfld3_tx_param;
+		c->rx_param = &mrfld3_rx_param;
+		break;
+	case 1:
+		c->port_id = 5;
+		c->num_chipselect = 4;
+		c->tx_param = &mrfld5_tx_param;
+		c->rx_param = &mrfld5_rx_param;
+		break;
+	case 2:
+		c->port_id = 6;
+		c->num_chipselect = 1;
+		c->tx_param = &mrfld6_tx_param;
+		c->rx_param = &mrfld6_rx_param;
+		break;
+	default:
+		return -ENODEV;
+	}
+
+	tx = c->tx_param;
+	tx->dma_dev = &dma_dev->dev;
+
+	rx = c->rx_param;
+	rx->dma_dev = &dma_dev->dev;
+
+	c->dma_filter = lpss_dma_filter;
+	return 0;
+}
+
+static struct pxa_spi_info spi_info_configs[] = {
+	[PORT_CE4100] = {
+		.type = PXA25x_SSP,
+		.port_id =  -1,
+		.num_chipselect = -1,
+		.max_clk_rate = 3686400,
+	},
+	[PORT_BYT] = {
+		.type = LPSS_BYT_SSP,
+		.port_id = 0,
+		.setup = lpss_spi_setup,
+		.tx_param = &byt_tx_param,
+		.rx_param = &byt_rx_param,
+	},
+	[PORT_BSW0] = {
+		.type = LPSS_BSW_SSP,
+		.port_id = 0,
+		.setup = lpss_spi_setup,
+		.tx_param = &bsw0_tx_param,
+		.rx_param = &bsw0_rx_param,
+	},
+	[PORT_BSW1] = {
+		.type = LPSS_BSW_SSP,
+		.port_id = 1,
+		.setup = lpss_spi_setup,
+		.tx_param = &bsw1_tx_param,
+		.rx_param = &bsw1_rx_param,
+	},
+	[PORT_BSW2] = {
+		.type = LPSS_BSW_SSP,
+		.port_id = 2,
+		.setup = lpss_spi_setup,
+		.tx_param = &bsw2_tx_param,
+		.rx_param = &bsw2_rx_param,
+	},
+	[PORT_MRFLD] = {
+		.type = PXA27x_SSP,
+		.max_clk_rate = 25000000,
+		.setup = mrfld_spi_setup,
+	},
+	[PORT_QUARK_X1000] = {
+		.type = QUARK_X1000_SSP,
+		.port_id = -1,
+		.num_chipselect = 1,
+		.max_clk_rate = 50000000,
+	},
+	[PORT_LPT] = {
+		.type = LPSS_LPT_SSP,
+		.port_id = 0,
+		.setup = lpss_spi_setup,
+		.tx_param = &lpt_tx_param,
+		.rx_param = &lpt_rx_param,
+	},
+};
+
+static int pxa2xx_spi_pci_probe(struct pci_dev *dev,
+		const struct pci_device_id *ent)
+{
+	struct platform_device_info pi;
+	int ret;
+	struct platform_device *pdev;
+	struct pxa2xx_spi_master spi_pdata;
+	struct ssp_device *ssp;
+	struct pxa_spi_info *c;
+	char buf[40];
+
+	ret = pcim_enable_device(dev);
+	if (ret)
+		return ret;
+
+	ret = pcim_iomap_regions(dev, 1 << 0, "PXA2xx SPI");
+	if (ret)
+		return ret;
+
+	c = &spi_info_configs[ent->driver_data];
+	if (c->setup) {
+		ret = c->setup(dev, c);
+		if (ret)
+			return ret;
+	}
+
+	memset(&spi_pdata, 0, sizeof(spi_pdata));
+	spi_pdata.num_chipselect = (c->num_chipselect > 0) ? c->num_chipselect : dev->devfn;
+	spi_pdata.dma_filter = c->dma_filter;
+	spi_pdata.tx_param = c->tx_param;
+	spi_pdata.rx_param = c->rx_param;
+	spi_pdata.enable_dma = c->rx_param && c->tx_param;
+
+	ssp = &spi_pdata.ssp;
+	ssp->phys_base = pci_resource_start(dev, 0);
+	ssp->mmio_base = pcim_iomap_table(dev)[0];
+	ssp->port_id = (c->port_id >= 0) ? c->port_id : dev->devfn;
+	ssp->type = c->type;
+
+	pci_set_master(dev);
+
+	ret = pci_alloc_irq_vectors(dev, 1, 1, PCI_IRQ_ALL_TYPES);
+	if (ret < 0)
+		return ret;
+	ssp->irq = pci_irq_vector(dev, 0);
+
+	snprintf(buf, sizeof(buf), "pxa2xx-spi.%d", ssp->port_id);
+	ssp->clk = clk_register_fixed_rate(&dev->dev, buf , NULL, 0,
+					   c->max_clk_rate);
+	 if (IS_ERR(ssp->clk))
+		return PTR_ERR(ssp->clk);
+
+	memset(&pi, 0, sizeof(pi));
+	pi.fwnode = dev->dev.fwnode;
+	pi.parent = &dev->dev;
+	pi.name = "pxa2xx-spi";
+	pi.id = ssp->port_id;
+	pi.data = &spi_pdata;
+	pi.size_data = sizeof(spi_pdata);
+
+	pdev = platform_device_register_full(&pi);
+	if (IS_ERR(pdev)) {
+		clk_unregister(ssp->clk);
+		return PTR_ERR(pdev);
+	}
+
+	pci_set_drvdata(dev, pdev);
+
+	return 0;
+}
+
+static void pxa2xx_spi_pci_remove(struct pci_dev *dev)
+{
+	struct platform_device *pdev = pci_get_drvdata(dev);
+	struct pxa2xx_spi_master *spi_pdata;
+
+	spi_pdata = dev_get_platdata(&pdev->dev);
+
+	platform_device_unregister(pdev);
+	clk_unregister(spi_pdata->ssp.clk);
+}
+
+static const struct pci_device_id pxa2xx_spi_pci_devices[] = {
+	{ PCI_VDEVICE(INTEL, 0x0935), PORT_QUARK_X1000 },
+	{ PCI_VDEVICE(INTEL, 0x0f0e), PORT_BYT },
+	{ PCI_VDEVICE(INTEL, 0x1194), PORT_MRFLD },
+	{ PCI_VDEVICE(INTEL, 0x228e), PORT_BSW0 },
+	{ PCI_VDEVICE(INTEL, 0x2290), PORT_BSW1 },
+	{ PCI_VDEVICE(INTEL, 0x22ac), PORT_BSW2 },
+	{ PCI_VDEVICE(INTEL, 0x2e6a), PORT_CE4100 },
+	{ PCI_VDEVICE(INTEL, 0x9ce6), PORT_LPT },
+	{ },
+};
+MODULE_DEVICE_TABLE(pci, pxa2xx_spi_pci_devices);
+
+static struct pci_driver pxa2xx_spi_pci_driver = {
+	.name           = "pxa2xx_spi_pci",
+	.id_table       = pxa2xx_spi_pci_devices,
+	.probe          = pxa2xx_spi_pci_probe,
+	.remove         = pxa2xx_spi_pci_remove,
+};
+
+module_pci_driver(pxa2xx_spi_pci_driver);
+
+MODULE_DESCRIPTION("CE4100/LPSS PCI-SPI glue code for PXA's driver");
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Sebastian Andrzej Siewior <bigeasy@linutronix.de>");
diff --git a/drivers/spi/spi-pxa2xx.c b/drivers/spi/spi-pxa2xx.c
new file mode 100644
index 0000000..14f4ea5
--- /dev/null
+++ b/drivers/spi/spi-pxa2xx.c
@@ -0,0 +1,1867 @@
+/*
+ * Copyright (C) 2005 Stephen Street / StreetFire Sound Labs
+ * Copyright (C) 2013, Intel Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/bitops.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/ioport.h>
+#include <linux/errno.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/pci.h>
+#include <linux/platform_device.h>
+#include <linux/spi/pxa2xx_spi.h>
+#include <linux/spi/spi.h>
+#include <linux/delay.h>
+#include <linux/gpio.h>
+#include <linux/gpio/consumer.h>
+#include <linux/slab.h>
+#include <linux/clk.h>
+#include <linux/pm_runtime.h>
+#include <linux/acpi.h>
+
+#include "spi-pxa2xx.h"
+
+MODULE_AUTHOR("Stephen Street");
+MODULE_DESCRIPTION("PXA2xx SSP SPI Controller");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:pxa2xx-spi");
+
+#define TIMOUT_DFLT		1000
+
+/*
+ * for testing SSCR1 changes that require SSP restart, basically
+ * everything except the service and interrupt enables, the pxa270 developer
+ * manual says only SSCR1_SCFR, SSCR1_SPH, SSCR1_SPO need to be in this
+ * list, but the PXA255 dev man says all bits without really meaning the
+ * service and interrupt enables
+ */
+#define SSCR1_CHANGE_MASK (SSCR1_TTELP | SSCR1_TTE | SSCR1_SCFR \
+				| SSCR1_ECRA | SSCR1_ECRB | SSCR1_SCLKDIR \
+				| SSCR1_SFRMDIR | SSCR1_RWOT | SSCR1_TRAIL \
+				| SSCR1_IFS | SSCR1_STRF | SSCR1_EFWR \
+				| SSCR1_RFT | SSCR1_TFT | SSCR1_MWDS \
+				| SSCR1_SPH | SSCR1_SPO | SSCR1_LBM)
+
+#define QUARK_X1000_SSCR1_CHANGE_MASK (QUARK_X1000_SSCR1_STRF	\
+				| QUARK_X1000_SSCR1_EFWR	\
+				| QUARK_X1000_SSCR1_RFT		\
+				| QUARK_X1000_SSCR1_TFT		\
+				| SSCR1_SPH | SSCR1_SPO | SSCR1_LBM)
+
+#define CE4100_SSCR1_CHANGE_MASK (SSCR1_TTELP | SSCR1_TTE | SSCR1_SCFR \
+				| SSCR1_ECRA | SSCR1_ECRB | SSCR1_SCLKDIR \
+				| SSCR1_SFRMDIR | SSCR1_RWOT | SSCR1_TRAIL \
+				| SSCR1_IFS | SSCR1_STRF | SSCR1_EFWR \
+				| CE4100_SSCR1_RFT | CE4100_SSCR1_TFT | SSCR1_MWDS \
+				| SSCR1_SPH | SSCR1_SPO | SSCR1_LBM)
+
+#define LPSS_GENERAL_REG_RXTO_HOLDOFF_DISABLE	BIT(24)
+#define LPSS_CS_CONTROL_SW_MODE			BIT(0)
+#define LPSS_CS_CONTROL_CS_HIGH			BIT(1)
+#define LPSS_CAPS_CS_EN_SHIFT			9
+#define LPSS_CAPS_CS_EN_MASK			(0xf << LPSS_CAPS_CS_EN_SHIFT)
+
+struct lpss_config {
+	/* LPSS offset from drv_data->ioaddr */
+	unsigned offset;
+	/* Register offsets from drv_data->lpss_base or -1 */
+	int reg_general;
+	int reg_ssp;
+	int reg_cs_ctrl;
+	int reg_capabilities;
+	/* FIFO thresholds */
+	u32 rx_threshold;
+	u32 tx_threshold_lo;
+	u32 tx_threshold_hi;
+	/* Chip select control */
+	unsigned cs_sel_shift;
+	unsigned cs_sel_mask;
+	unsigned cs_num;
+};
+
+/* Keep these sorted with enum pxa_ssp_type */
+static const struct lpss_config lpss_platforms[] = {
+	{	/* LPSS_LPT_SSP */
+		.offset = 0x800,
+		.reg_general = 0x08,
+		.reg_ssp = 0x0c,
+		.reg_cs_ctrl = 0x18,
+		.reg_capabilities = -1,
+		.rx_threshold = 64,
+		.tx_threshold_lo = 160,
+		.tx_threshold_hi = 224,
+	},
+	{	/* LPSS_BYT_SSP */
+		.offset = 0x400,
+		.reg_general = 0x08,
+		.reg_ssp = 0x0c,
+		.reg_cs_ctrl = 0x18,
+		.reg_capabilities = -1,
+		.rx_threshold = 64,
+		.tx_threshold_lo = 160,
+		.tx_threshold_hi = 224,
+	},
+	{	/* LPSS_BSW_SSP */
+		.offset = 0x400,
+		.reg_general = 0x08,
+		.reg_ssp = 0x0c,
+		.reg_cs_ctrl = 0x18,
+		.reg_capabilities = -1,
+		.rx_threshold = 64,
+		.tx_threshold_lo = 160,
+		.tx_threshold_hi = 224,
+		.cs_sel_shift = 2,
+		.cs_sel_mask = 1 << 2,
+		.cs_num = 2,
+	},
+	{	/* LPSS_SPT_SSP */
+		.offset = 0x200,
+		.reg_general = -1,
+		.reg_ssp = 0x20,
+		.reg_cs_ctrl = 0x24,
+		.reg_capabilities = -1,
+		.rx_threshold = 1,
+		.tx_threshold_lo = 32,
+		.tx_threshold_hi = 56,
+	},
+	{	/* LPSS_BXT_SSP */
+		.offset = 0x200,
+		.reg_general = -1,
+		.reg_ssp = 0x20,
+		.reg_cs_ctrl = 0x24,
+		.reg_capabilities = 0xfc,
+		.rx_threshold = 1,
+		.tx_threshold_lo = 16,
+		.tx_threshold_hi = 48,
+		.cs_sel_shift = 8,
+		.cs_sel_mask = 3 << 8,
+	},
+	{	/* LPSS_CNL_SSP */
+		.offset = 0x200,
+		.reg_general = -1,
+		.reg_ssp = 0x20,
+		.reg_cs_ctrl = 0x24,
+		.reg_capabilities = 0xfc,
+		.rx_threshold = 1,
+		.tx_threshold_lo = 32,
+		.tx_threshold_hi = 56,
+		.cs_sel_shift = 8,
+		.cs_sel_mask = 3 << 8,
+	},
+};
+
+static inline const struct lpss_config
+*lpss_get_config(const struct driver_data *drv_data)
+{
+	return &lpss_platforms[drv_data->ssp_type - LPSS_LPT_SSP];
+}
+
+static bool is_lpss_ssp(const struct driver_data *drv_data)
+{
+	switch (drv_data->ssp_type) {
+	case LPSS_LPT_SSP:
+	case LPSS_BYT_SSP:
+	case LPSS_BSW_SSP:
+	case LPSS_SPT_SSP:
+	case LPSS_BXT_SSP:
+	case LPSS_CNL_SSP:
+		return true;
+	default:
+		return false;
+	}
+}
+
+static bool is_quark_x1000_ssp(const struct driver_data *drv_data)
+{
+	return drv_data->ssp_type == QUARK_X1000_SSP;
+}
+
+static u32 pxa2xx_spi_get_ssrc1_change_mask(const struct driver_data *drv_data)
+{
+	switch (drv_data->ssp_type) {
+	case QUARK_X1000_SSP:
+		return QUARK_X1000_SSCR1_CHANGE_MASK;
+	case CE4100_SSP:
+		return CE4100_SSCR1_CHANGE_MASK;
+	default:
+		return SSCR1_CHANGE_MASK;
+	}
+}
+
+static u32
+pxa2xx_spi_get_rx_default_thre(const struct driver_data *drv_data)
+{
+	switch (drv_data->ssp_type) {
+	case QUARK_X1000_SSP:
+		return RX_THRESH_QUARK_X1000_DFLT;
+	case CE4100_SSP:
+		return RX_THRESH_CE4100_DFLT;
+	default:
+		return RX_THRESH_DFLT;
+	}
+}
+
+static bool pxa2xx_spi_txfifo_full(const struct driver_data *drv_data)
+{
+	u32 mask;
+
+	switch (drv_data->ssp_type) {
+	case QUARK_X1000_SSP:
+		mask = QUARK_X1000_SSSR_TFL_MASK;
+		break;
+	case CE4100_SSP:
+		mask = CE4100_SSSR_TFL_MASK;
+		break;
+	default:
+		mask = SSSR_TFL_MASK;
+		break;
+	}
+
+	return (pxa2xx_spi_read(drv_data, SSSR) & mask) == mask;
+}
+
+static void pxa2xx_spi_clear_rx_thre(const struct driver_data *drv_data,
+				     u32 *sccr1_reg)
+{
+	u32 mask;
+
+	switch (drv_data->ssp_type) {
+	case QUARK_X1000_SSP:
+		mask = QUARK_X1000_SSCR1_RFT;
+		break;
+	case CE4100_SSP:
+		mask = CE4100_SSCR1_RFT;
+		break;
+	default:
+		mask = SSCR1_RFT;
+		break;
+	}
+	*sccr1_reg &= ~mask;
+}
+
+static void pxa2xx_spi_set_rx_thre(const struct driver_data *drv_data,
+				   u32 *sccr1_reg, u32 threshold)
+{
+	switch (drv_data->ssp_type) {
+	case QUARK_X1000_SSP:
+		*sccr1_reg |= QUARK_X1000_SSCR1_RxTresh(threshold);
+		break;
+	case CE4100_SSP:
+		*sccr1_reg |= CE4100_SSCR1_RxTresh(threshold);
+		break;
+	default:
+		*sccr1_reg |= SSCR1_RxTresh(threshold);
+		break;
+	}
+}
+
+static u32 pxa2xx_configure_sscr0(const struct driver_data *drv_data,
+				  u32 clk_div, u8 bits)
+{
+	switch (drv_data->ssp_type) {
+	case QUARK_X1000_SSP:
+		return clk_div
+			| QUARK_X1000_SSCR0_Motorola
+			| QUARK_X1000_SSCR0_DataSize(bits > 32 ? 8 : bits)
+			| SSCR0_SSE;
+	default:
+		return clk_div
+			| SSCR0_Motorola
+			| SSCR0_DataSize(bits > 16 ? bits - 16 : bits)
+			| SSCR0_SSE
+			| (bits > 16 ? SSCR0_EDSS : 0);
+	}
+}
+
+/*
+ * Read and write LPSS SSP private registers. Caller must first check that
+ * is_lpss_ssp() returns true before these can be called.
+ */
+static u32 __lpss_ssp_read_priv(struct driver_data *drv_data, unsigned offset)
+{
+	WARN_ON(!drv_data->lpss_base);
+	return readl(drv_data->lpss_base + offset);
+}
+
+static void __lpss_ssp_write_priv(struct driver_data *drv_data,
+				  unsigned offset, u32 value)
+{
+	WARN_ON(!drv_data->lpss_base);
+	writel(value, drv_data->lpss_base + offset);
+}
+
+/*
+ * lpss_ssp_setup - perform LPSS SSP specific setup
+ * @drv_data: pointer to the driver private data
+ *
+ * Perform LPSS SSP specific setup. This function must be called first if
+ * one is going to use LPSS SSP private registers.
+ */
+static void lpss_ssp_setup(struct driver_data *drv_data)
+{
+	const struct lpss_config *config;
+	u32 value;
+
+	config = lpss_get_config(drv_data);
+	drv_data->lpss_base = drv_data->ioaddr + config->offset;
+
+	/* Enable software chip select control */
+	value = __lpss_ssp_read_priv(drv_data, config->reg_cs_ctrl);
+	value &= ~(LPSS_CS_CONTROL_SW_MODE | LPSS_CS_CONTROL_CS_HIGH);
+	value |= LPSS_CS_CONTROL_SW_MODE | LPSS_CS_CONTROL_CS_HIGH;
+	__lpss_ssp_write_priv(drv_data, config->reg_cs_ctrl, value);
+
+	/* Enable multiblock DMA transfers */
+	if (drv_data->master_info->enable_dma) {
+		__lpss_ssp_write_priv(drv_data, config->reg_ssp, 1);
+
+		if (config->reg_general >= 0) {
+			value = __lpss_ssp_read_priv(drv_data,
+						     config->reg_general);
+			value |= LPSS_GENERAL_REG_RXTO_HOLDOFF_DISABLE;
+			__lpss_ssp_write_priv(drv_data,
+					      config->reg_general, value);
+		}
+	}
+}
+
+static void lpss_ssp_select_cs(struct spi_device *spi,
+			       const struct lpss_config *config)
+{
+	struct driver_data *drv_data =
+		spi_controller_get_devdata(spi->controller);
+	u32 value, cs;
+
+	if (!config->cs_sel_mask)
+		return;
+
+	value = __lpss_ssp_read_priv(drv_data, config->reg_cs_ctrl);
+
+	cs = spi->chip_select;
+	cs <<= config->cs_sel_shift;
+	if (cs != (value & config->cs_sel_mask)) {
+		/*
+		 * When switching another chip select output active the
+		 * output must be selected first and wait 2 ssp_clk cycles
+		 * before changing state to active. Otherwise a short
+		 * glitch will occur on the previous chip select since
+		 * output select is latched but state control is not.
+		 */
+		value &= ~config->cs_sel_mask;
+		value |= cs;
+		__lpss_ssp_write_priv(drv_data,
+				      config->reg_cs_ctrl, value);
+		ndelay(1000000000 /
+		       (drv_data->master->max_speed_hz / 2));
+	}
+}
+
+static void lpss_ssp_cs_control(struct spi_device *spi, bool enable)
+{
+	struct driver_data *drv_data =
+		spi_controller_get_devdata(spi->controller);
+	const struct lpss_config *config;
+	u32 value;
+
+	config = lpss_get_config(drv_data);
+
+	if (enable)
+		lpss_ssp_select_cs(spi, config);
+
+	value = __lpss_ssp_read_priv(drv_data, config->reg_cs_ctrl);
+	if (enable)
+		value &= ~LPSS_CS_CONTROL_CS_HIGH;
+	else
+		value |= LPSS_CS_CONTROL_CS_HIGH;
+	__lpss_ssp_write_priv(drv_data, config->reg_cs_ctrl, value);
+}
+
+static void cs_assert(struct spi_device *spi)
+{
+	struct chip_data *chip = spi_get_ctldata(spi);
+	struct driver_data *drv_data =
+		spi_controller_get_devdata(spi->controller);
+
+	if (drv_data->ssp_type == CE4100_SSP) {
+		pxa2xx_spi_write(drv_data, SSSR, chip->frm);
+		return;
+	}
+
+	if (chip->cs_control) {
+		chip->cs_control(PXA2XX_CS_ASSERT);
+		return;
+	}
+
+	if (chip->gpiod_cs) {
+		gpiod_set_value(chip->gpiod_cs, chip->gpio_cs_inverted);
+		return;
+	}
+
+	if (is_lpss_ssp(drv_data))
+		lpss_ssp_cs_control(spi, true);
+}
+
+static void cs_deassert(struct spi_device *spi)
+{
+	struct chip_data *chip = spi_get_ctldata(spi);
+	struct driver_data *drv_data =
+		spi_controller_get_devdata(spi->controller);
+	unsigned long timeout;
+
+	if (drv_data->ssp_type == CE4100_SSP)
+		return;
+
+	/* Wait until SSP becomes idle before deasserting the CS */
+	timeout = jiffies + msecs_to_jiffies(10);
+	while (pxa2xx_spi_read(drv_data, SSSR) & SSSR_BSY &&
+	       !time_after(jiffies, timeout))
+		cpu_relax();
+
+	if (chip->cs_control) {
+		chip->cs_control(PXA2XX_CS_DEASSERT);
+		return;
+	}
+
+	if (chip->gpiod_cs) {
+		gpiod_set_value(chip->gpiod_cs, !chip->gpio_cs_inverted);
+		return;
+	}
+
+	if (is_lpss_ssp(drv_data))
+		lpss_ssp_cs_control(spi, false);
+}
+
+static void pxa2xx_spi_set_cs(struct spi_device *spi, bool level)
+{
+	if (level)
+		cs_deassert(spi);
+	else
+		cs_assert(spi);
+}
+
+int pxa2xx_spi_flush(struct driver_data *drv_data)
+{
+	unsigned long limit = loops_per_jiffy << 1;
+
+	do {
+		while (pxa2xx_spi_read(drv_data, SSSR) & SSSR_RNE)
+			pxa2xx_spi_read(drv_data, SSDR);
+	} while ((pxa2xx_spi_read(drv_data, SSSR) & SSSR_BSY) && --limit);
+	write_SSSR_CS(drv_data, SSSR_ROR);
+
+	return limit;
+}
+
+static int null_writer(struct driver_data *drv_data)
+{
+	u8 n_bytes = drv_data->n_bytes;
+
+	if (pxa2xx_spi_txfifo_full(drv_data)
+		|| (drv_data->tx == drv_data->tx_end))
+		return 0;
+
+	pxa2xx_spi_write(drv_data, SSDR, 0);
+	drv_data->tx += n_bytes;
+
+	return 1;
+}
+
+static int null_reader(struct driver_data *drv_data)
+{
+	u8 n_bytes = drv_data->n_bytes;
+
+	while ((pxa2xx_spi_read(drv_data, SSSR) & SSSR_RNE)
+	       && (drv_data->rx < drv_data->rx_end)) {
+		pxa2xx_spi_read(drv_data, SSDR);
+		drv_data->rx += n_bytes;
+	}
+
+	return drv_data->rx == drv_data->rx_end;
+}
+
+static int u8_writer(struct driver_data *drv_data)
+{
+	if (pxa2xx_spi_txfifo_full(drv_data)
+		|| (drv_data->tx == drv_data->tx_end))
+		return 0;
+
+	pxa2xx_spi_write(drv_data, SSDR, *(u8 *)(drv_data->tx));
+	++drv_data->tx;
+
+	return 1;
+}
+
+static int u8_reader(struct driver_data *drv_data)
+{
+	while ((pxa2xx_spi_read(drv_data, SSSR) & SSSR_RNE)
+	       && (drv_data->rx < drv_data->rx_end)) {
+		*(u8 *)(drv_data->rx) = pxa2xx_spi_read(drv_data, SSDR);
+		++drv_data->rx;
+	}
+
+	return drv_data->rx == drv_data->rx_end;
+}
+
+static int u16_writer(struct driver_data *drv_data)
+{
+	if (pxa2xx_spi_txfifo_full(drv_data)
+		|| (drv_data->tx == drv_data->tx_end))
+		return 0;
+
+	pxa2xx_spi_write(drv_data, SSDR, *(u16 *)(drv_data->tx));
+	drv_data->tx += 2;
+
+	return 1;
+}
+
+static int u16_reader(struct driver_data *drv_data)
+{
+	while ((pxa2xx_spi_read(drv_data, SSSR) & SSSR_RNE)
+	       && (drv_data->rx < drv_data->rx_end)) {
+		*(u16 *)(drv_data->rx) = pxa2xx_spi_read(drv_data, SSDR);
+		drv_data->rx += 2;
+	}
+
+	return drv_data->rx == drv_data->rx_end;
+}
+
+static int u32_writer(struct driver_data *drv_data)
+{
+	if (pxa2xx_spi_txfifo_full(drv_data)
+		|| (drv_data->tx == drv_data->tx_end))
+		return 0;
+
+	pxa2xx_spi_write(drv_data, SSDR, *(u32 *)(drv_data->tx));
+	drv_data->tx += 4;
+
+	return 1;
+}
+
+static int u32_reader(struct driver_data *drv_data)
+{
+	while ((pxa2xx_spi_read(drv_data, SSSR) & SSSR_RNE)
+	       && (drv_data->rx < drv_data->rx_end)) {
+		*(u32 *)(drv_data->rx) = pxa2xx_spi_read(drv_data, SSDR);
+		drv_data->rx += 4;
+	}
+
+	return drv_data->rx == drv_data->rx_end;
+}
+
+static void reset_sccr1(struct driver_data *drv_data)
+{
+	struct chip_data *chip =
+		spi_get_ctldata(drv_data->master->cur_msg->spi);
+	u32 sccr1_reg;
+
+	sccr1_reg = pxa2xx_spi_read(drv_data, SSCR1) & ~drv_data->int_cr1;
+	switch (drv_data->ssp_type) {
+	case QUARK_X1000_SSP:
+		sccr1_reg &= ~QUARK_X1000_SSCR1_RFT;
+		break;
+	case CE4100_SSP:
+		sccr1_reg &= ~CE4100_SSCR1_RFT;
+		break;
+	default:
+		sccr1_reg &= ~SSCR1_RFT;
+		break;
+	}
+	sccr1_reg |= chip->threshold;
+	pxa2xx_spi_write(drv_data, SSCR1, sccr1_reg);
+}
+
+static void int_error_stop(struct driver_data *drv_data, const char* msg)
+{
+	/* Stop and reset SSP */
+	write_SSSR_CS(drv_data, drv_data->clear_sr);
+	reset_sccr1(drv_data);
+	if (!pxa25x_ssp_comp(drv_data))
+		pxa2xx_spi_write(drv_data, SSTO, 0);
+	pxa2xx_spi_flush(drv_data);
+	pxa2xx_spi_write(drv_data, SSCR0,
+			 pxa2xx_spi_read(drv_data, SSCR0) & ~SSCR0_SSE);
+
+	dev_err(&drv_data->pdev->dev, "%s\n", msg);
+
+	drv_data->master->cur_msg->status = -EIO;
+	spi_finalize_current_transfer(drv_data->master);
+}
+
+static void int_transfer_complete(struct driver_data *drv_data)
+{
+	/* Clear and disable interrupts */
+	write_SSSR_CS(drv_data, drv_data->clear_sr);
+	reset_sccr1(drv_data);
+	if (!pxa25x_ssp_comp(drv_data))
+		pxa2xx_spi_write(drv_data, SSTO, 0);
+
+	spi_finalize_current_transfer(drv_data->master);
+}
+
+static irqreturn_t interrupt_transfer(struct driver_data *drv_data)
+{
+	u32 irq_mask = (pxa2xx_spi_read(drv_data, SSCR1) & SSCR1_TIE) ?
+		       drv_data->mask_sr : drv_data->mask_sr & ~SSSR_TFS;
+
+	u32 irq_status = pxa2xx_spi_read(drv_data, SSSR) & irq_mask;
+
+	if (irq_status & SSSR_ROR) {
+		int_error_stop(drv_data, "interrupt_transfer: fifo overrun");
+		return IRQ_HANDLED;
+	}
+
+	if (irq_status & SSSR_TINT) {
+		pxa2xx_spi_write(drv_data, SSSR, SSSR_TINT);
+		if (drv_data->read(drv_data)) {
+			int_transfer_complete(drv_data);
+			return IRQ_HANDLED;
+		}
+	}
+
+	/* Drain rx fifo, Fill tx fifo and prevent overruns */
+	do {
+		if (drv_data->read(drv_data)) {
+			int_transfer_complete(drv_data);
+			return IRQ_HANDLED;
+		}
+	} while (drv_data->write(drv_data));
+
+	if (drv_data->read(drv_data)) {
+		int_transfer_complete(drv_data);
+		return IRQ_HANDLED;
+	}
+
+	if (drv_data->tx == drv_data->tx_end) {
+		u32 bytes_left;
+		u32 sccr1_reg;
+
+		sccr1_reg = pxa2xx_spi_read(drv_data, SSCR1);
+		sccr1_reg &= ~SSCR1_TIE;
+
+		/*
+		 * PXA25x_SSP has no timeout, set up rx threshould for the
+		 * remaining RX bytes.
+		 */
+		if (pxa25x_ssp_comp(drv_data)) {
+			u32 rx_thre;
+
+			pxa2xx_spi_clear_rx_thre(drv_data, &sccr1_reg);
+
+			bytes_left = drv_data->rx_end - drv_data->rx;
+			switch (drv_data->n_bytes) {
+			case 4:
+				bytes_left >>= 1;
+			case 2:
+				bytes_left >>= 1;
+			}
+
+			rx_thre = pxa2xx_spi_get_rx_default_thre(drv_data);
+			if (rx_thre > bytes_left)
+				rx_thre = bytes_left;
+
+			pxa2xx_spi_set_rx_thre(drv_data, &sccr1_reg, rx_thre);
+		}
+		pxa2xx_spi_write(drv_data, SSCR1, sccr1_reg);
+	}
+
+	/* We did something */
+	return IRQ_HANDLED;
+}
+
+static void handle_bad_msg(struct driver_data *drv_data)
+{
+	pxa2xx_spi_write(drv_data, SSCR0,
+			 pxa2xx_spi_read(drv_data, SSCR0) & ~SSCR0_SSE);
+	pxa2xx_spi_write(drv_data, SSCR1,
+			 pxa2xx_spi_read(drv_data, SSCR1) & ~drv_data->int_cr1);
+	if (!pxa25x_ssp_comp(drv_data))
+		pxa2xx_spi_write(drv_data, SSTO, 0);
+	write_SSSR_CS(drv_data, drv_data->clear_sr);
+
+	dev_err(&drv_data->pdev->dev,
+		"bad message state in interrupt handler\n");
+}
+
+static irqreturn_t ssp_int(int irq, void *dev_id)
+{
+	struct driver_data *drv_data = dev_id;
+	u32 sccr1_reg;
+	u32 mask = drv_data->mask_sr;
+	u32 status;
+
+	/*
+	 * The IRQ might be shared with other peripherals so we must first
+	 * check that are we RPM suspended or not. If we are we assume that
+	 * the IRQ was not for us (we shouldn't be RPM suspended when the
+	 * interrupt is enabled).
+	 */
+	if (pm_runtime_suspended(&drv_data->pdev->dev))
+		return IRQ_NONE;
+
+	/*
+	 * If the device is not yet in RPM suspended state and we get an
+	 * interrupt that is meant for another device, check if status bits
+	 * are all set to one. That means that the device is already
+	 * powered off.
+	 */
+	status = pxa2xx_spi_read(drv_data, SSSR);
+	if (status == ~0)
+		return IRQ_NONE;
+
+	sccr1_reg = pxa2xx_spi_read(drv_data, SSCR1);
+
+	/* Ignore possible writes if we don't need to write */
+	if (!(sccr1_reg & SSCR1_TIE))
+		mask &= ~SSSR_TFS;
+
+	/* Ignore RX timeout interrupt if it is disabled */
+	if (!(sccr1_reg & SSCR1_TINTE))
+		mask &= ~SSSR_TINT;
+
+	if (!(status & mask))
+		return IRQ_NONE;
+
+	pxa2xx_spi_write(drv_data, SSCR1, sccr1_reg & ~drv_data->int_cr1);
+	pxa2xx_spi_write(drv_data, SSCR1, sccr1_reg);
+
+	if (!drv_data->master->cur_msg) {
+		handle_bad_msg(drv_data);
+		/* Never fail */
+		return IRQ_HANDLED;
+	}
+
+	return drv_data->transfer_handler(drv_data);
+}
+
+/*
+ * The Quark SPI has an additional 24 bit register (DDS_CLK_RATE) to multiply
+ * input frequency by fractions of 2^24. It also has a divider by 5.
+ *
+ * There are formulas to get baud rate value for given input frequency and
+ * divider parameters, such as DDS_CLK_RATE and SCR:
+ *
+ * Fsys = 200MHz
+ *
+ * Fssp = Fsys * DDS_CLK_RATE / 2^24			(1)
+ * Baud rate = Fsclk = Fssp / (2 * (SCR + 1))		(2)
+ *
+ * DDS_CLK_RATE either 2^n or 2^n / 5.
+ * SCR is in range 0 .. 255
+ *
+ * Divisor = 5^i * 2^j * 2 * k
+ *       i = [0, 1]      i = 1 iff j = 0 or j > 3
+ *       j = [0, 23]     j = 0 iff i = 1
+ *       k = [1, 256]
+ * Special case: j = 0, i = 1: Divisor = 2 / 5
+ *
+ * Accordingly to the specification the recommended values for DDS_CLK_RATE
+ * are:
+ *	Case 1:		2^n, n = [0, 23]
+ *	Case 2:		2^24 * 2 / 5 (0x666666)
+ *	Case 3:		less than or equal to 2^24 / 5 / 16 (0x33333)
+ *
+ * In all cases the lowest possible value is better.
+ *
+ * The function calculates parameters for all cases and chooses the one closest
+ * to the asked baud rate.
+ */
+static unsigned int quark_x1000_get_clk_div(int rate, u32 *dds)
+{
+	unsigned long xtal = 200000000;
+	unsigned long fref = xtal / 2;		/* mandatory division by 2,
+						   see (2) */
+						/* case 3 */
+	unsigned long fref1 = fref / 2;		/* case 1 */
+	unsigned long fref2 = fref * 2 / 5;	/* case 2 */
+	unsigned long scale;
+	unsigned long q, q1, q2;
+	long r, r1, r2;
+	u32 mul;
+
+	/* Case 1 */
+
+	/* Set initial value for DDS_CLK_RATE */
+	mul = (1 << 24) >> 1;
+
+	/* Calculate initial quot */
+	q1 = DIV_ROUND_UP(fref1, rate);
+
+	/* Scale q1 if it's too big */
+	if (q1 > 256) {
+		/* Scale q1 to range [1, 512] */
+		scale = fls_long(q1 - 1);
+		if (scale > 9) {
+			q1 >>= scale - 9;
+			mul >>= scale - 9;
+		}
+
+		/* Round the result if we have a remainder */
+		q1 += q1 & 1;
+	}
+
+	/* Decrease DDS_CLK_RATE as much as we can without loss in precision */
+	scale = __ffs(q1);
+	q1 >>= scale;
+	mul >>= scale;
+
+	/* Get the remainder */
+	r1 = abs(fref1 / (1 << (24 - fls_long(mul))) / q1 - rate);
+
+	/* Case 2 */
+
+	q2 = DIV_ROUND_UP(fref2, rate);
+	r2 = abs(fref2 / q2 - rate);
+
+	/*
+	 * Choose the best between two: less remainder we have the better. We
+	 * can't go case 2 if q2 is greater than 256 since SCR register can
+	 * hold only values 0 .. 255.
+	 */
+	if (r2 >= r1 || q2 > 256) {
+		/* case 1 is better */
+		r = r1;
+		q = q1;
+	} else {
+		/* case 2 is better */
+		r = r2;
+		q = q2;
+		mul = (1 << 24) * 2 / 5;
+	}
+
+	/* Check case 3 only if the divisor is big enough */
+	if (fref / rate >= 80) {
+		u64 fssp;
+		u32 m;
+
+		/* Calculate initial quot */
+		q1 = DIV_ROUND_UP(fref, rate);
+		m = (1 << 24) / q1;
+
+		/* Get the remainder */
+		fssp = (u64)fref * m;
+		do_div(fssp, 1 << 24);
+		r1 = abs(fssp - rate);
+
+		/* Choose this one if it suits better */
+		if (r1 < r) {
+			/* case 3 is better */
+			q = 1;
+			mul = m;
+		}
+	}
+
+	*dds = mul;
+	return q - 1;
+}
+
+static unsigned int ssp_get_clk_div(struct driver_data *drv_data, int rate)
+{
+	unsigned long ssp_clk = drv_data->master->max_speed_hz;
+	const struct ssp_device *ssp = drv_data->ssp;
+
+	rate = min_t(int, ssp_clk, rate);
+
+	if (ssp->type == PXA25x_SSP || ssp->type == CE4100_SSP)
+		return (ssp_clk / (2 * rate) - 1) & 0xff;
+	else
+		return (ssp_clk / rate - 1) & 0xfff;
+}
+
+static unsigned int pxa2xx_ssp_get_clk_div(struct driver_data *drv_data,
+					   int rate)
+{
+	struct chip_data *chip =
+		spi_get_ctldata(drv_data->master->cur_msg->spi);
+	unsigned int clk_div;
+
+	switch (drv_data->ssp_type) {
+	case QUARK_X1000_SSP:
+		clk_div = quark_x1000_get_clk_div(rate, &chip->dds_rate);
+		break;
+	default:
+		clk_div = ssp_get_clk_div(drv_data, rate);
+		break;
+	}
+	return clk_div << 8;
+}
+
+static bool pxa2xx_spi_can_dma(struct spi_controller *master,
+			       struct spi_device *spi,
+			       struct spi_transfer *xfer)
+{
+	struct chip_data *chip = spi_get_ctldata(spi);
+
+	return chip->enable_dma &&
+	       xfer->len <= MAX_DMA_LEN &&
+	       xfer->len >= chip->dma_burst_size;
+}
+
+static int pxa2xx_spi_transfer_one(struct spi_controller *master,
+				   struct spi_device *spi,
+				   struct spi_transfer *transfer)
+{
+	struct driver_data *drv_data = spi_controller_get_devdata(master);
+	struct spi_message *message = master->cur_msg;
+	struct chip_data *chip = spi_get_ctldata(message->spi);
+	u32 dma_thresh = chip->dma_threshold;
+	u32 dma_burst = chip->dma_burst_size;
+	u32 change_mask = pxa2xx_spi_get_ssrc1_change_mask(drv_data);
+	u32 clk_div;
+	u8 bits;
+	u32 speed;
+	u32 cr0;
+	u32 cr1;
+	int err;
+	int dma_mapped;
+
+	/* Check if we can DMA this transfer */
+	if (transfer->len > MAX_DMA_LEN && chip->enable_dma) {
+
+		/* reject already-mapped transfers; PIO won't always work */
+		if (message->is_dma_mapped
+				|| transfer->rx_dma || transfer->tx_dma) {
+			dev_err(&drv_data->pdev->dev,
+				"Mapped transfer length of %u is greater than %d\n",
+				transfer->len, MAX_DMA_LEN);
+			return -EINVAL;
+		}
+
+		/* warn ... we force this to PIO mode */
+		dev_warn_ratelimited(&message->spi->dev,
+				     "DMA disabled for transfer length %ld greater than %d\n",
+				     (long)transfer->len, MAX_DMA_LEN);
+	}
+
+	/* Setup the transfer state based on the type of transfer */
+	if (pxa2xx_spi_flush(drv_data) == 0) {
+		dev_err(&drv_data->pdev->dev, "Flush failed\n");
+		return -EIO;
+	}
+	drv_data->n_bytes = chip->n_bytes;
+	drv_data->tx = (void *)transfer->tx_buf;
+	drv_data->tx_end = drv_data->tx + transfer->len;
+	drv_data->rx = transfer->rx_buf;
+	drv_data->rx_end = drv_data->rx + transfer->len;
+	drv_data->write = drv_data->tx ? chip->write : null_writer;
+	drv_data->read = drv_data->rx ? chip->read : null_reader;
+
+	/* Change speed and bit per word on a per transfer */
+	bits = transfer->bits_per_word;
+	speed = transfer->speed_hz;
+
+	clk_div = pxa2xx_ssp_get_clk_div(drv_data, speed);
+
+	if (bits <= 8) {
+		drv_data->n_bytes = 1;
+		drv_data->read = drv_data->read != null_reader ?
+					u8_reader : null_reader;
+		drv_data->write = drv_data->write != null_writer ?
+					u8_writer : null_writer;
+	} else if (bits <= 16) {
+		drv_data->n_bytes = 2;
+		drv_data->read = drv_data->read != null_reader ?
+					u16_reader : null_reader;
+		drv_data->write = drv_data->write != null_writer ?
+					u16_writer : null_writer;
+	} else if (bits <= 32) {
+		drv_data->n_bytes = 4;
+		drv_data->read = drv_data->read != null_reader ?
+					u32_reader : null_reader;
+		drv_data->write = drv_data->write != null_writer ?
+					u32_writer : null_writer;
+	}
+	/*
+	 * if bits/word is changed in dma mode, then must check the
+	 * thresholds and burst also
+	 */
+	if (chip->enable_dma) {
+		if (pxa2xx_spi_set_dma_burst_and_threshold(chip,
+						message->spi,
+						bits, &dma_burst,
+						&dma_thresh))
+			dev_warn_ratelimited(&message->spi->dev,
+					     "DMA burst size reduced to match bits_per_word\n");
+	}
+
+	dma_mapped = master->can_dma &&
+		     master->can_dma(master, message->spi, transfer) &&
+		     master->cur_msg_mapped;
+	if (dma_mapped) {
+
+		/* Ensure we have the correct interrupt handler */
+		drv_data->transfer_handler = pxa2xx_spi_dma_transfer;
+
+		err = pxa2xx_spi_dma_prepare(drv_data, transfer);
+		if (err)
+			return err;
+
+		/* Clear status and start DMA engine */
+		cr1 = chip->cr1 | dma_thresh | drv_data->dma_cr1;
+		pxa2xx_spi_write(drv_data, SSSR, drv_data->clear_sr);
+
+		pxa2xx_spi_dma_start(drv_data);
+	} else {
+		/* Ensure we have the correct interrupt handler	*/
+		drv_data->transfer_handler = interrupt_transfer;
+
+		/* Clear status  */
+		cr1 = chip->cr1 | chip->threshold | drv_data->int_cr1;
+		write_SSSR_CS(drv_data, drv_data->clear_sr);
+	}
+
+	/* NOTE:  PXA25x_SSP _could_ use external clocking ... */
+	cr0 = pxa2xx_configure_sscr0(drv_data, clk_div, bits);
+	if (!pxa25x_ssp_comp(drv_data))
+		dev_dbg(&message->spi->dev, "%u Hz actual, %s\n",
+			master->max_speed_hz
+				/ (1 + ((cr0 & SSCR0_SCR(0xfff)) >> 8)),
+			dma_mapped ? "DMA" : "PIO");
+	else
+		dev_dbg(&message->spi->dev, "%u Hz actual, %s\n",
+			master->max_speed_hz / 2
+				/ (1 + ((cr0 & SSCR0_SCR(0x0ff)) >> 8)),
+			dma_mapped ? "DMA" : "PIO");
+
+	if (is_lpss_ssp(drv_data)) {
+		if ((pxa2xx_spi_read(drv_data, SSIRF) & 0xff)
+		    != chip->lpss_rx_threshold)
+			pxa2xx_spi_write(drv_data, SSIRF,
+					 chip->lpss_rx_threshold);
+		if ((pxa2xx_spi_read(drv_data, SSITF) & 0xffff)
+		    != chip->lpss_tx_threshold)
+			pxa2xx_spi_write(drv_data, SSITF,
+					 chip->lpss_tx_threshold);
+	}
+
+	if (is_quark_x1000_ssp(drv_data) &&
+	    (pxa2xx_spi_read(drv_data, DDS_RATE) != chip->dds_rate))
+		pxa2xx_spi_write(drv_data, DDS_RATE, chip->dds_rate);
+
+	/* see if we need to reload the config registers */
+	if ((pxa2xx_spi_read(drv_data, SSCR0) != cr0)
+	    || (pxa2xx_spi_read(drv_data, SSCR1) & change_mask)
+	    != (cr1 & change_mask)) {
+		/* stop the SSP, and update the other bits */
+		pxa2xx_spi_write(drv_data, SSCR0, cr0 & ~SSCR0_SSE);
+		if (!pxa25x_ssp_comp(drv_data))
+			pxa2xx_spi_write(drv_data, SSTO, chip->timeout);
+		/* first set CR1 without interrupt and service enables */
+		pxa2xx_spi_write(drv_data, SSCR1, cr1 & change_mask);
+		/* restart the SSP */
+		pxa2xx_spi_write(drv_data, SSCR0, cr0);
+
+	} else {
+		if (!pxa25x_ssp_comp(drv_data))
+			pxa2xx_spi_write(drv_data, SSTO, chip->timeout);
+	}
+
+	/*
+	 * Release the data by enabling service requests and interrupts,
+	 * without changing any mode bits
+	 */
+	pxa2xx_spi_write(drv_data, SSCR1, cr1);
+
+	return 1;
+}
+
+static void pxa2xx_spi_handle_err(struct spi_controller *master,
+				 struct spi_message *msg)
+{
+	struct driver_data *drv_data = spi_controller_get_devdata(master);
+
+	/* Disable the SSP */
+	pxa2xx_spi_write(drv_data, SSCR0,
+			 pxa2xx_spi_read(drv_data, SSCR0) & ~SSCR0_SSE);
+	/* Clear and disable interrupts and service requests */
+	write_SSSR_CS(drv_data, drv_data->clear_sr);
+	pxa2xx_spi_write(drv_data, SSCR1,
+			 pxa2xx_spi_read(drv_data, SSCR1)
+			 & ~(drv_data->int_cr1 | drv_data->dma_cr1));
+	if (!pxa25x_ssp_comp(drv_data))
+		pxa2xx_spi_write(drv_data, SSTO, 0);
+
+	/*
+	 * Stop the DMA if running. Note DMA callback handler may have unset
+	 * the dma_running already, which is fine as stopping is not needed
+	 * then but we shouldn't rely this flag for anything else than
+	 * stopping. For instance to differentiate between PIO and DMA
+	 * transfers.
+	 */
+	if (atomic_read(&drv_data->dma_running))
+		pxa2xx_spi_dma_stop(drv_data);
+}
+
+static int pxa2xx_spi_unprepare_transfer(struct spi_controller *master)
+{
+	struct driver_data *drv_data = spi_controller_get_devdata(master);
+
+	/* Disable the SSP now */
+	pxa2xx_spi_write(drv_data, SSCR0,
+			 pxa2xx_spi_read(drv_data, SSCR0) & ~SSCR0_SSE);
+
+	return 0;
+}
+
+static int setup_cs(struct spi_device *spi, struct chip_data *chip,
+		    struct pxa2xx_spi_chip *chip_info)
+{
+	struct driver_data *drv_data =
+		spi_controller_get_devdata(spi->controller);
+	struct gpio_desc *gpiod;
+	int err = 0;
+
+	if (chip == NULL)
+		return 0;
+
+	if (drv_data->cs_gpiods) {
+		gpiod = drv_data->cs_gpiods[spi->chip_select];
+		if (gpiod) {
+			chip->gpiod_cs = gpiod;
+			chip->gpio_cs_inverted = spi->mode & SPI_CS_HIGH;
+			gpiod_set_value(gpiod, chip->gpio_cs_inverted);
+		}
+
+		return 0;
+	}
+
+	if (chip_info == NULL)
+		return 0;
+
+	/* NOTE: setup() can be called multiple times, possibly with
+	 * different chip_info, release previously requested GPIO
+	 */
+	if (chip->gpiod_cs) {
+		gpiod_put(chip->gpiod_cs);
+		chip->gpiod_cs = NULL;
+	}
+
+	/* If (*cs_control) is provided, ignore GPIO chip select */
+	if (chip_info->cs_control) {
+		chip->cs_control = chip_info->cs_control;
+		return 0;
+	}
+
+	if (gpio_is_valid(chip_info->gpio_cs)) {
+		err = gpio_request(chip_info->gpio_cs, "SPI_CS");
+		if (err) {
+			dev_err(&spi->dev, "failed to request chip select GPIO%d\n",
+				chip_info->gpio_cs);
+			return err;
+		}
+
+		gpiod = gpio_to_desc(chip_info->gpio_cs);
+		chip->gpiod_cs = gpiod;
+		chip->gpio_cs_inverted = spi->mode & SPI_CS_HIGH;
+
+		err = gpiod_direction_output(gpiod, !chip->gpio_cs_inverted);
+	}
+
+	return err;
+}
+
+static int setup(struct spi_device *spi)
+{
+	struct pxa2xx_spi_chip *chip_info;
+	struct chip_data *chip;
+	const struct lpss_config *config;
+	struct driver_data *drv_data =
+		spi_controller_get_devdata(spi->controller);
+	uint tx_thres, tx_hi_thres, rx_thres;
+
+	switch (drv_data->ssp_type) {
+	case QUARK_X1000_SSP:
+		tx_thres = TX_THRESH_QUARK_X1000_DFLT;
+		tx_hi_thres = 0;
+		rx_thres = RX_THRESH_QUARK_X1000_DFLT;
+		break;
+	case CE4100_SSP:
+		tx_thres = TX_THRESH_CE4100_DFLT;
+		tx_hi_thres = 0;
+		rx_thres = RX_THRESH_CE4100_DFLT;
+		break;
+	case LPSS_LPT_SSP:
+	case LPSS_BYT_SSP:
+	case LPSS_BSW_SSP:
+	case LPSS_SPT_SSP:
+	case LPSS_BXT_SSP:
+	case LPSS_CNL_SSP:
+		config = lpss_get_config(drv_data);
+		tx_thres = config->tx_threshold_lo;
+		tx_hi_thres = config->tx_threshold_hi;
+		rx_thres = config->rx_threshold;
+		break;
+	default:
+		tx_thres = TX_THRESH_DFLT;
+		tx_hi_thres = 0;
+		rx_thres = RX_THRESH_DFLT;
+		break;
+	}
+
+	/* Only alloc on first setup */
+	chip = spi_get_ctldata(spi);
+	if (!chip) {
+		chip = kzalloc(sizeof(struct chip_data), GFP_KERNEL);
+		if (!chip)
+			return -ENOMEM;
+
+		if (drv_data->ssp_type == CE4100_SSP) {
+			if (spi->chip_select > 4) {
+				dev_err(&spi->dev,
+					"failed setup: cs number must not be > 4.\n");
+				kfree(chip);
+				return -EINVAL;
+			}
+
+			chip->frm = spi->chip_select;
+		}
+		chip->enable_dma = drv_data->master_info->enable_dma;
+		chip->timeout = TIMOUT_DFLT;
+	}
+
+	/* protocol drivers may change the chip settings, so...
+	 * if chip_info exists, use it */
+	chip_info = spi->controller_data;
+
+	/* chip_info isn't always needed */
+	chip->cr1 = 0;
+	if (chip_info) {
+		if (chip_info->timeout)
+			chip->timeout = chip_info->timeout;
+		if (chip_info->tx_threshold)
+			tx_thres = chip_info->tx_threshold;
+		if (chip_info->tx_hi_threshold)
+			tx_hi_thres = chip_info->tx_hi_threshold;
+		if (chip_info->rx_threshold)
+			rx_thres = chip_info->rx_threshold;
+		chip->dma_threshold = 0;
+		if (chip_info->enable_loopback)
+			chip->cr1 = SSCR1_LBM;
+	}
+
+	chip->lpss_rx_threshold = SSIRF_RxThresh(rx_thres);
+	chip->lpss_tx_threshold = SSITF_TxLoThresh(tx_thres)
+				| SSITF_TxHiThresh(tx_hi_thres);
+
+	/* set dma burst and threshold outside of chip_info path so that if
+	 * chip_info goes away after setting chip->enable_dma, the
+	 * burst and threshold can still respond to changes in bits_per_word */
+	if (chip->enable_dma) {
+		/* set up legal burst and threshold for dma */
+		if (pxa2xx_spi_set_dma_burst_and_threshold(chip, spi,
+						spi->bits_per_word,
+						&chip->dma_burst_size,
+						&chip->dma_threshold)) {
+			dev_warn(&spi->dev,
+				 "in setup: DMA burst size reduced to match bits_per_word\n");
+		}
+	}
+
+	switch (drv_data->ssp_type) {
+	case QUARK_X1000_SSP:
+		chip->threshold = (QUARK_X1000_SSCR1_RxTresh(rx_thres)
+				   & QUARK_X1000_SSCR1_RFT)
+				   | (QUARK_X1000_SSCR1_TxTresh(tx_thres)
+				   & QUARK_X1000_SSCR1_TFT);
+		break;
+	case CE4100_SSP:
+		chip->threshold = (CE4100_SSCR1_RxTresh(rx_thres) & CE4100_SSCR1_RFT) |
+			(CE4100_SSCR1_TxTresh(tx_thres) & CE4100_SSCR1_TFT);
+		break;
+	default:
+		chip->threshold = (SSCR1_RxTresh(rx_thres) & SSCR1_RFT) |
+			(SSCR1_TxTresh(tx_thres) & SSCR1_TFT);
+		break;
+	}
+
+	chip->cr1 &= ~(SSCR1_SPO | SSCR1_SPH);
+	chip->cr1 |= (((spi->mode & SPI_CPHA) != 0) ? SSCR1_SPH : 0)
+			| (((spi->mode & SPI_CPOL) != 0) ? SSCR1_SPO : 0);
+
+	if (spi->mode & SPI_LOOP)
+		chip->cr1 |= SSCR1_LBM;
+
+	if (spi->bits_per_word <= 8) {
+		chip->n_bytes = 1;
+		chip->read = u8_reader;
+		chip->write = u8_writer;
+	} else if (spi->bits_per_word <= 16) {
+		chip->n_bytes = 2;
+		chip->read = u16_reader;
+		chip->write = u16_writer;
+	} else if (spi->bits_per_word <= 32) {
+		chip->n_bytes = 4;
+		chip->read = u32_reader;
+		chip->write = u32_writer;
+	}
+
+	spi_set_ctldata(spi, chip);
+
+	if (drv_data->ssp_type == CE4100_SSP)
+		return 0;
+
+	return setup_cs(spi, chip, chip_info);
+}
+
+static void cleanup(struct spi_device *spi)
+{
+	struct chip_data *chip = spi_get_ctldata(spi);
+	struct driver_data *drv_data =
+		spi_controller_get_devdata(spi->controller);
+
+	if (!chip)
+		return;
+
+	if (drv_data->ssp_type != CE4100_SSP && !drv_data->cs_gpiods &&
+	    chip->gpiod_cs)
+		gpiod_put(chip->gpiod_cs);
+
+	kfree(chip);
+}
+
+#ifdef CONFIG_PCI
+#ifdef CONFIG_ACPI
+
+static const struct acpi_device_id pxa2xx_spi_acpi_match[] = {
+	{ "INT33C0", LPSS_LPT_SSP },
+	{ "INT33C1", LPSS_LPT_SSP },
+	{ "INT3430", LPSS_LPT_SSP },
+	{ "INT3431", LPSS_LPT_SSP },
+	{ "80860F0E", LPSS_BYT_SSP },
+	{ "8086228E", LPSS_BSW_SSP },
+	{ },
+};
+MODULE_DEVICE_TABLE(acpi, pxa2xx_spi_acpi_match);
+
+static int pxa2xx_spi_get_port_id(struct acpi_device *adev)
+{
+	unsigned int devid;
+	int port_id = -1;
+
+	if (adev && adev->pnp.unique_id &&
+	    !kstrtouint(adev->pnp.unique_id, 0, &devid))
+		port_id = devid;
+	return port_id;
+}
+#else /* !CONFIG_ACPI */
+static int pxa2xx_spi_get_port_id(struct acpi_device *adev)
+{
+	return -1;
+}
+#endif
+
+/*
+ * PCI IDs of compound devices that integrate both host controller and private
+ * integrated DMA engine. Please note these are not used in module
+ * autoloading and probing in this module but matching the LPSS SSP type.
+ */
+static const struct pci_device_id pxa2xx_spi_pci_compound_match[] = {
+	/* SPT-LP */
+	{ PCI_VDEVICE(INTEL, 0x9d29), LPSS_SPT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x9d2a), LPSS_SPT_SSP },
+	/* SPT-H */
+	{ PCI_VDEVICE(INTEL, 0xa129), LPSS_SPT_SSP },
+	{ PCI_VDEVICE(INTEL, 0xa12a), LPSS_SPT_SSP },
+	/* KBL-H */
+	{ PCI_VDEVICE(INTEL, 0xa2a9), LPSS_SPT_SSP },
+	{ PCI_VDEVICE(INTEL, 0xa2aa), LPSS_SPT_SSP },
+	/* BXT A-Step */
+	{ PCI_VDEVICE(INTEL, 0x0ac2), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x0ac4), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x0ac6), LPSS_BXT_SSP },
+	/* BXT B-Step */
+	{ PCI_VDEVICE(INTEL, 0x1ac2), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x1ac4), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x1ac6), LPSS_BXT_SSP },
+	/* GLK */
+	{ PCI_VDEVICE(INTEL, 0x31c2), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x31c4), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x31c6), LPSS_BXT_SSP },
+	/* ICL-LP */
+	{ PCI_VDEVICE(INTEL, 0x34aa), LPSS_CNL_SSP },
+	{ PCI_VDEVICE(INTEL, 0x34ab), LPSS_CNL_SSP },
+	{ PCI_VDEVICE(INTEL, 0x34fb), LPSS_CNL_SSP },
+	/* APL */
+	{ PCI_VDEVICE(INTEL, 0x5ac2), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x5ac4), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x5ac6), LPSS_BXT_SSP },
+	/* CNL-LP */
+	{ PCI_VDEVICE(INTEL, 0x9daa), LPSS_CNL_SSP },
+	{ PCI_VDEVICE(INTEL, 0x9dab), LPSS_CNL_SSP },
+	{ PCI_VDEVICE(INTEL, 0x9dfb), LPSS_CNL_SSP },
+	/* CNL-H */
+	{ PCI_VDEVICE(INTEL, 0xa32a), LPSS_CNL_SSP },
+	{ PCI_VDEVICE(INTEL, 0xa32b), LPSS_CNL_SSP },
+	{ PCI_VDEVICE(INTEL, 0xa37b), LPSS_CNL_SSP },
+	{ },
+};
+
+static bool pxa2xx_spi_idma_filter(struct dma_chan *chan, void *param)
+{
+	struct device *dev = param;
+
+	if (dev != chan->device->dev->parent)
+		return false;
+
+	return true;
+}
+
+static struct pxa2xx_spi_master *
+pxa2xx_spi_init_pdata(struct platform_device *pdev)
+{
+	struct pxa2xx_spi_master *pdata;
+	struct acpi_device *adev;
+	struct ssp_device *ssp;
+	struct resource *res;
+	const struct acpi_device_id *adev_id = NULL;
+	const struct pci_device_id *pcidev_id = NULL;
+	int type;
+
+	adev = ACPI_COMPANION(&pdev->dev);
+
+	if (dev_is_pci(pdev->dev.parent))
+		pcidev_id = pci_match_id(pxa2xx_spi_pci_compound_match,
+					 to_pci_dev(pdev->dev.parent));
+	else if (adev)
+		adev_id = acpi_match_device(pdev->dev.driver->acpi_match_table,
+					    &pdev->dev);
+	else
+		return NULL;
+
+	if (adev_id)
+		type = (int)adev_id->driver_data;
+	else if (pcidev_id)
+		type = (int)pcidev_id->driver_data;
+	else
+		return NULL;
+
+	pdata = devm_kzalloc(&pdev->dev, sizeof(*pdata), GFP_KERNEL);
+	if (!pdata)
+		return NULL;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res)
+		return NULL;
+
+	ssp = &pdata->ssp;
+
+	ssp->phys_base = res->start;
+	ssp->mmio_base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(ssp->mmio_base))
+		return NULL;
+
+	if (pcidev_id) {
+		pdata->tx_param = pdev->dev.parent;
+		pdata->rx_param = pdev->dev.parent;
+		pdata->dma_filter = pxa2xx_spi_idma_filter;
+	}
+
+	ssp->clk = devm_clk_get(&pdev->dev, NULL);
+	ssp->irq = platform_get_irq(pdev, 0);
+	ssp->type = type;
+	ssp->pdev = pdev;
+	ssp->port_id = pxa2xx_spi_get_port_id(adev);
+
+	pdata->num_chipselect = 1;
+	pdata->enable_dma = true;
+
+	return pdata;
+}
+
+#else /* !CONFIG_PCI */
+static inline struct pxa2xx_spi_master *
+pxa2xx_spi_init_pdata(struct platform_device *pdev)
+{
+	return NULL;
+}
+#endif
+
+static int pxa2xx_spi_fw_translate_cs(struct spi_controller *master,
+				      unsigned int cs)
+{
+	struct driver_data *drv_data = spi_controller_get_devdata(master);
+
+	if (has_acpi_companion(&drv_data->pdev->dev)) {
+		switch (drv_data->ssp_type) {
+		/*
+		 * For Atoms the ACPI DeviceSelection used by the Windows
+		 * driver starts from 1 instead of 0 so translate it here
+		 * to match what Linux expects.
+		 */
+		case LPSS_BYT_SSP:
+		case LPSS_BSW_SSP:
+			return cs - 1;
+
+		default:
+			break;
+		}
+	}
+
+	return cs;
+}
+
+static int pxa2xx_spi_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct pxa2xx_spi_master *platform_info;
+	struct spi_controller *master;
+	struct driver_data *drv_data;
+	struct ssp_device *ssp;
+	const struct lpss_config *config;
+	int status, count;
+	u32 tmp;
+
+	platform_info = dev_get_platdata(dev);
+	if (!platform_info) {
+		platform_info = pxa2xx_spi_init_pdata(pdev);
+		if (!platform_info) {
+			dev_err(&pdev->dev, "missing platform data\n");
+			return -ENODEV;
+		}
+	}
+
+	ssp = pxa_ssp_request(pdev->id, pdev->name);
+	if (!ssp)
+		ssp = &platform_info->ssp;
+
+	if (!ssp->mmio_base) {
+		dev_err(&pdev->dev, "failed to get ssp\n");
+		return -ENODEV;
+	}
+
+	master = spi_alloc_master(dev, sizeof(struct driver_data));
+	if (!master) {
+		dev_err(&pdev->dev, "cannot alloc spi_master\n");
+		pxa_ssp_free(ssp);
+		return -ENOMEM;
+	}
+	drv_data = spi_controller_get_devdata(master);
+	drv_data->master = master;
+	drv_data->master_info = platform_info;
+	drv_data->pdev = pdev;
+	drv_data->ssp = ssp;
+
+	master->dev.of_node = pdev->dev.of_node;
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LOOP;
+
+	master->bus_num = ssp->port_id;
+	master->dma_alignment = DMA_ALIGNMENT;
+	master->cleanup = cleanup;
+	master->setup = setup;
+	master->set_cs = pxa2xx_spi_set_cs;
+	master->transfer_one = pxa2xx_spi_transfer_one;
+	master->handle_err = pxa2xx_spi_handle_err;
+	master->unprepare_transfer_hardware = pxa2xx_spi_unprepare_transfer;
+	master->fw_translate_cs = pxa2xx_spi_fw_translate_cs;
+	master->auto_runtime_pm = true;
+	master->flags = SPI_CONTROLLER_MUST_RX | SPI_CONTROLLER_MUST_TX;
+
+	drv_data->ssp_type = ssp->type;
+
+	drv_data->ioaddr = ssp->mmio_base;
+	drv_data->ssdr_physical = ssp->phys_base + SSDR;
+	if (pxa25x_ssp_comp(drv_data)) {
+		switch (drv_data->ssp_type) {
+		case QUARK_X1000_SSP:
+			master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 32);
+			break;
+		default:
+			master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 16);
+			break;
+		}
+
+		drv_data->int_cr1 = SSCR1_TIE | SSCR1_RIE;
+		drv_data->dma_cr1 = 0;
+		drv_data->clear_sr = SSSR_ROR;
+		drv_data->mask_sr = SSSR_RFS | SSSR_TFS | SSSR_ROR;
+	} else {
+		master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 32);
+		drv_data->int_cr1 = SSCR1_TIE | SSCR1_RIE | SSCR1_TINTE;
+		drv_data->dma_cr1 = DEFAULT_DMA_CR1;
+		drv_data->clear_sr = SSSR_ROR | SSSR_TINT;
+		drv_data->mask_sr = SSSR_TINT | SSSR_RFS | SSSR_TFS | SSSR_ROR;
+	}
+
+	status = request_irq(ssp->irq, ssp_int, IRQF_SHARED, dev_name(dev),
+			drv_data);
+	if (status < 0) {
+		dev_err(&pdev->dev, "cannot get IRQ %d\n", ssp->irq);
+		goto out_error_master_alloc;
+	}
+
+	/* Setup DMA if requested */
+	if (platform_info->enable_dma) {
+		status = pxa2xx_spi_dma_setup(drv_data);
+		if (status) {
+			dev_dbg(dev, "no DMA channels available, using PIO\n");
+			platform_info->enable_dma = false;
+		} else {
+			master->can_dma = pxa2xx_spi_can_dma;
+		}
+	}
+
+	/* Enable SOC clock */
+	status = clk_prepare_enable(ssp->clk);
+	if (status)
+		goto out_error_dma_irq_alloc;
+
+	master->max_speed_hz = clk_get_rate(ssp->clk);
+
+	/* Load default SSP configuration */
+	pxa2xx_spi_write(drv_data, SSCR0, 0);
+	switch (drv_data->ssp_type) {
+	case QUARK_X1000_SSP:
+		tmp = QUARK_X1000_SSCR1_RxTresh(RX_THRESH_QUARK_X1000_DFLT) |
+		      QUARK_X1000_SSCR1_TxTresh(TX_THRESH_QUARK_X1000_DFLT);
+		pxa2xx_spi_write(drv_data, SSCR1, tmp);
+
+		/* using the Motorola SPI protocol and use 8 bit frame */
+		tmp = QUARK_X1000_SSCR0_Motorola | QUARK_X1000_SSCR0_DataSize(8);
+		pxa2xx_spi_write(drv_data, SSCR0, tmp);
+		break;
+	case CE4100_SSP:
+		tmp = CE4100_SSCR1_RxTresh(RX_THRESH_CE4100_DFLT) |
+		      CE4100_SSCR1_TxTresh(TX_THRESH_CE4100_DFLT);
+		pxa2xx_spi_write(drv_data, SSCR1, tmp);
+		tmp = SSCR0_SCR(2) | SSCR0_Motorola | SSCR0_DataSize(8);
+		pxa2xx_spi_write(drv_data, SSCR0, tmp);
+		break;
+	default:
+		tmp = SSCR1_RxTresh(RX_THRESH_DFLT) |
+		      SSCR1_TxTresh(TX_THRESH_DFLT);
+		pxa2xx_spi_write(drv_data, SSCR1, tmp);
+		tmp = SSCR0_SCR(2) | SSCR0_Motorola | SSCR0_DataSize(8);
+		pxa2xx_spi_write(drv_data, SSCR0, tmp);
+		break;
+	}
+
+	if (!pxa25x_ssp_comp(drv_data))
+		pxa2xx_spi_write(drv_data, SSTO, 0);
+
+	if (!is_quark_x1000_ssp(drv_data))
+		pxa2xx_spi_write(drv_data, SSPSP, 0);
+
+	if (is_lpss_ssp(drv_data)) {
+		lpss_ssp_setup(drv_data);
+		config = lpss_get_config(drv_data);
+		if (config->reg_capabilities >= 0) {
+			tmp = __lpss_ssp_read_priv(drv_data,
+						   config->reg_capabilities);
+			tmp &= LPSS_CAPS_CS_EN_MASK;
+			tmp >>= LPSS_CAPS_CS_EN_SHIFT;
+			platform_info->num_chipselect = ffz(tmp);
+		} else if (config->cs_num) {
+			platform_info->num_chipselect = config->cs_num;
+		}
+	}
+	master->num_chipselect = platform_info->num_chipselect;
+
+	count = gpiod_count(&pdev->dev, "cs");
+	if (count > 0) {
+		int i;
+
+		master->num_chipselect = max_t(int, count,
+			master->num_chipselect);
+
+		drv_data->cs_gpiods = devm_kcalloc(&pdev->dev,
+			master->num_chipselect, sizeof(struct gpio_desc *),
+			GFP_KERNEL);
+		if (!drv_data->cs_gpiods) {
+			status = -ENOMEM;
+			goto out_error_clock_enabled;
+		}
+
+		for (i = 0; i < master->num_chipselect; i++) {
+			struct gpio_desc *gpiod;
+
+			gpiod = devm_gpiod_get_index(dev, "cs", i, GPIOD_ASIS);
+			if (IS_ERR(gpiod)) {
+				/* Means use native chip select */
+				if (PTR_ERR(gpiod) == -ENOENT)
+					continue;
+
+				status = (int)PTR_ERR(gpiod);
+				goto out_error_clock_enabled;
+			} else {
+				drv_data->cs_gpiods[i] = gpiod;
+			}
+		}
+	}
+
+	pm_runtime_set_autosuspend_delay(&pdev->dev, 50);
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+
+	/* Register with the SPI framework */
+	platform_set_drvdata(pdev, drv_data);
+	status = devm_spi_register_controller(&pdev->dev, master);
+	if (status != 0) {
+		dev_err(&pdev->dev, "problem registering spi master\n");
+		goto out_error_clock_enabled;
+	}
+
+	return status;
+
+out_error_clock_enabled:
+	pm_runtime_put_noidle(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+	clk_disable_unprepare(ssp->clk);
+
+out_error_dma_irq_alloc:
+	pxa2xx_spi_dma_release(drv_data);
+	free_irq(ssp->irq, drv_data);
+
+out_error_master_alloc:
+	spi_controller_put(master);
+	pxa_ssp_free(ssp);
+	return status;
+}
+
+static int pxa2xx_spi_remove(struct platform_device *pdev)
+{
+	struct driver_data *drv_data = platform_get_drvdata(pdev);
+	struct ssp_device *ssp;
+
+	if (!drv_data)
+		return 0;
+	ssp = drv_data->ssp;
+
+	pm_runtime_get_sync(&pdev->dev);
+
+	/* Disable the SSP at the peripheral and SOC level */
+	pxa2xx_spi_write(drv_data, SSCR0, 0);
+	clk_disable_unprepare(ssp->clk);
+
+	/* Release DMA */
+	if (drv_data->master_info->enable_dma)
+		pxa2xx_spi_dma_release(drv_data);
+
+	pm_runtime_put_noidle(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+
+	/* Release IRQ */
+	free_irq(ssp->irq, drv_data);
+
+	/* Release SSP */
+	pxa_ssp_free(ssp);
+
+	return 0;
+}
+
+static void pxa2xx_spi_shutdown(struct platform_device *pdev)
+{
+	int status = 0;
+
+	if ((status = pxa2xx_spi_remove(pdev)) != 0)
+		dev_err(&pdev->dev, "shutdown failed with %d\n", status);
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int pxa2xx_spi_suspend(struct device *dev)
+{
+	struct driver_data *drv_data = dev_get_drvdata(dev);
+	struct ssp_device *ssp = drv_data->ssp;
+	int status;
+
+	status = spi_controller_suspend(drv_data->master);
+	if (status != 0)
+		return status;
+	pxa2xx_spi_write(drv_data, SSCR0, 0);
+
+	if (!pm_runtime_suspended(dev))
+		clk_disable_unprepare(ssp->clk);
+
+	return 0;
+}
+
+static int pxa2xx_spi_resume(struct device *dev)
+{
+	struct driver_data *drv_data = dev_get_drvdata(dev);
+	struct ssp_device *ssp = drv_data->ssp;
+	int status;
+
+	/* Enable the SSP clock */
+	if (!pm_runtime_suspended(dev)) {
+		status = clk_prepare_enable(ssp->clk);
+		if (status)
+			return status;
+	}
+
+	/* Restore LPSS private register bits */
+	if (is_lpss_ssp(drv_data))
+		lpss_ssp_setup(drv_data);
+
+	/* Start the queue running */
+	status = spi_controller_resume(drv_data->master);
+	if (status != 0) {
+		dev_err(dev, "problem starting queue (%d)\n", status);
+		return status;
+	}
+
+	return 0;
+}
+#endif
+
+#ifdef CONFIG_PM
+static int pxa2xx_spi_runtime_suspend(struct device *dev)
+{
+	struct driver_data *drv_data = dev_get_drvdata(dev);
+
+	clk_disable_unprepare(drv_data->ssp->clk);
+	return 0;
+}
+
+static int pxa2xx_spi_runtime_resume(struct device *dev)
+{
+	struct driver_data *drv_data = dev_get_drvdata(dev);
+	int status;
+
+	status = clk_prepare_enable(drv_data->ssp->clk);
+	return status;
+}
+#endif
+
+static const struct dev_pm_ops pxa2xx_spi_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(pxa2xx_spi_suspend, pxa2xx_spi_resume)
+	SET_RUNTIME_PM_OPS(pxa2xx_spi_runtime_suspend,
+			   pxa2xx_spi_runtime_resume, NULL)
+};
+
+static struct platform_driver driver = {
+	.driver = {
+		.name	= "pxa2xx-spi",
+		.pm	= &pxa2xx_spi_pm_ops,
+		.acpi_match_table = ACPI_PTR(pxa2xx_spi_acpi_match),
+	},
+	.probe = pxa2xx_spi_probe,
+	.remove = pxa2xx_spi_remove,
+	.shutdown = pxa2xx_spi_shutdown,
+};
+
+static int __init pxa2xx_spi_init(void)
+{
+	return platform_driver_register(&driver);
+}
+subsys_initcall(pxa2xx_spi_init);
+
+static void __exit pxa2xx_spi_exit(void)
+{
+	platform_driver_unregister(&driver);
+}
+module_exit(pxa2xx_spi_exit);
diff --git a/drivers/spi/spi-pxa2xx.h b/drivers/spi/spi-pxa2xx.h
new file mode 100644
index 0000000..513c53a
--- /dev/null
+++ b/drivers/spi/spi-pxa2xx.h
@@ -0,0 +1,143 @@
+/*
+ * Copyright (C) 2005 Stephen Street / StreetFire Sound Labs
+ * Copyright (C) 2013, Intel Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef SPI_PXA2XX_H
+#define SPI_PXA2XX_H
+
+#include <linux/atomic.h>
+#include <linux/dmaengine.h>
+#include <linux/errno.h>
+#include <linux/io.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/pxa2xx_ssp.h>
+#include <linux/scatterlist.h>
+#include <linux/sizes.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/pxa2xx_spi.h>
+
+struct driver_data {
+	/* Driver model hookup */
+	struct platform_device *pdev;
+
+	/* SSP Info */
+	struct ssp_device *ssp;
+
+	/* SPI framework hookup */
+	enum pxa_ssp_type ssp_type;
+	struct spi_controller *master;
+
+	/* PXA hookup */
+	struct pxa2xx_spi_master *master_info;
+
+	/* SSP register addresses */
+	void __iomem *ioaddr;
+	phys_addr_t ssdr_physical;
+
+	/* SSP masks*/
+	u32 dma_cr1;
+	u32 int_cr1;
+	u32 clear_sr;
+	u32 mask_sr;
+
+	/* DMA engine support */
+	atomic_t dma_running;
+
+	/* Current transfer state info */
+	void *tx;
+	void *tx_end;
+	void *rx;
+	void *rx_end;
+	u8 n_bytes;
+	int (*write)(struct driver_data *drv_data);
+	int (*read)(struct driver_data *drv_data);
+	irqreturn_t (*transfer_handler)(struct driver_data *drv_data);
+	void (*cs_control)(u32 command);
+
+	void __iomem *lpss_base;
+
+	/* GPIOs for chip selects */
+	struct gpio_desc **cs_gpiods;
+};
+
+struct chip_data {
+	u32 cr1;
+	u32 dds_rate;
+	u32 timeout;
+	u8 n_bytes;
+	u32 dma_burst_size;
+	u32 threshold;
+	u32 dma_threshold;
+	u16 lpss_rx_threshold;
+	u16 lpss_tx_threshold;
+	u8 enable_dma;
+	union {
+		struct gpio_desc *gpiod_cs;
+		unsigned int frm;
+	};
+	int gpio_cs_inverted;
+	int (*write)(struct driver_data *drv_data);
+	int (*read)(struct driver_data *drv_data);
+	void (*cs_control)(u32 command);
+};
+
+static inline u32 pxa2xx_spi_read(const struct driver_data *drv_data,
+				  unsigned reg)
+{
+	return __raw_readl(drv_data->ioaddr + reg);
+}
+
+static  inline void pxa2xx_spi_write(const struct driver_data *drv_data,
+				     unsigned reg, u32 val)
+{
+	__raw_writel(val, drv_data->ioaddr + reg);
+}
+
+#define DMA_ALIGNMENT		8
+
+static inline int pxa25x_ssp_comp(struct driver_data *drv_data)
+{
+	switch (drv_data->ssp_type) {
+	case PXA25x_SSP:
+	case CE4100_SSP:
+	case QUARK_X1000_SSP:
+		return 1;
+	default:
+		return 0;
+	}
+}
+
+static inline void write_SSSR_CS(struct driver_data *drv_data, u32 val)
+{
+	if (drv_data->ssp_type == CE4100_SSP ||
+	    drv_data->ssp_type == QUARK_X1000_SSP)
+		val |= pxa2xx_spi_read(drv_data, SSSR) & SSSR_ALT_FRM_MASK;
+
+	pxa2xx_spi_write(drv_data, SSSR, val);
+}
+
+extern int pxa2xx_spi_flush(struct driver_data *drv_data);
+
+#define MAX_DMA_LEN		SZ_64K
+#define DEFAULT_DMA_CR1		(SSCR1_TSRE | SSCR1_RSRE | SSCR1_TRAIL)
+
+extern irqreturn_t pxa2xx_spi_dma_transfer(struct driver_data *drv_data);
+extern int pxa2xx_spi_dma_prepare(struct driver_data *drv_data,
+				  struct spi_transfer *xfer);
+extern void pxa2xx_spi_dma_start(struct driver_data *drv_data);
+extern void pxa2xx_spi_dma_stop(struct driver_data *drv_data);
+extern int pxa2xx_spi_dma_setup(struct driver_data *drv_data);
+extern void pxa2xx_spi_dma_release(struct driver_data *drv_data);
+extern int pxa2xx_spi_set_dma_burst_and_threshold(struct chip_data *chip,
+						  struct spi_device *spi,
+						  u8 bits_per_word,
+						  u32 *burst_code,
+						  u32 *threshold);
+
+#endif /* SPI_PXA2XX_H */
diff --git a/drivers/spi/spi-qup.c b/drivers/spi/spi-qup.c
new file mode 100644
index 0000000..974a8ce
--- /dev/null
+++ b/drivers/spi/spi-qup.c
@@ -0,0 +1,1282 @@
+/*
+ * Copyright (c) 2008-2014, The Linux foundation. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License rev 2 and
+ * only rev 2 as published by the free Software foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or fITNESS fOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/list.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/spi/spi.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+
+#define QUP_CONFIG			0x0000
+#define QUP_STATE			0x0004
+#define QUP_IO_M_MODES			0x0008
+#define QUP_SW_RESET			0x000c
+#define QUP_OPERATIONAL			0x0018
+#define QUP_ERROR_FLAGS			0x001c
+#define QUP_ERROR_FLAGS_EN		0x0020
+#define QUP_OPERATIONAL_MASK		0x0028
+#define QUP_HW_VERSION			0x0030
+#define QUP_MX_OUTPUT_CNT		0x0100
+#define QUP_OUTPUT_FIFO			0x0110
+#define QUP_MX_WRITE_CNT		0x0150
+#define QUP_MX_INPUT_CNT		0x0200
+#define QUP_MX_READ_CNT			0x0208
+#define QUP_INPUT_FIFO			0x0218
+
+#define SPI_CONFIG			0x0300
+#define SPI_IO_CONTROL			0x0304
+#define SPI_ERROR_FLAGS			0x0308
+#define SPI_ERROR_FLAGS_EN		0x030c
+
+/* QUP_CONFIG fields */
+#define QUP_CONFIG_SPI_MODE		(1 << 8)
+#define QUP_CONFIG_CLOCK_AUTO_GATE	BIT(13)
+#define QUP_CONFIG_NO_INPUT		BIT(7)
+#define QUP_CONFIG_NO_OUTPUT		BIT(6)
+#define QUP_CONFIG_N			0x001f
+
+/* QUP_STATE fields */
+#define QUP_STATE_VALID			BIT(2)
+#define QUP_STATE_RESET			0
+#define QUP_STATE_RUN			1
+#define QUP_STATE_PAUSE			3
+#define QUP_STATE_MASK			3
+#define QUP_STATE_CLEAR			2
+
+#define QUP_HW_VERSION_2_1_1		0x20010001
+
+/* QUP_IO_M_MODES fields */
+#define QUP_IO_M_PACK_EN		BIT(15)
+#define QUP_IO_M_UNPACK_EN		BIT(14)
+#define QUP_IO_M_INPUT_MODE_MASK_SHIFT	12
+#define QUP_IO_M_OUTPUT_MODE_MASK_SHIFT	10
+#define QUP_IO_M_INPUT_MODE_MASK	(3 << QUP_IO_M_INPUT_MODE_MASK_SHIFT)
+#define QUP_IO_M_OUTPUT_MODE_MASK	(3 << QUP_IO_M_OUTPUT_MODE_MASK_SHIFT)
+
+#define QUP_IO_M_OUTPUT_BLOCK_SIZE(x)	(((x) & (0x03 << 0)) >> 0)
+#define QUP_IO_M_OUTPUT_FIFO_SIZE(x)	(((x) & (0x07 << 2)) >> 2)
+#define QUP_IO_M_INPUT_BLOCK_SIZE(x)	(((x) & (0x03 << 5)) >> 5)
+#define QUP_IO_M_INPUT_FIFO_SIZE(x)	(((x) & (0x07 << 7)) >> 7)
+
+#define QUP_IO_M_MODE_FIFO		0
+#define QUP_IO_M_MODE_BLOCK		1
+#define QUP_IO_M_MODE_DMOV		2
+#define QUP_IO_M_MODE_BAM		3
+
+/* QUP_OPERATIONAL fields */
+#define QUP_OP_IN_BLOCK_READ_REQ	BIT(13)
+#define QUP_OP_OUT_BLOCK_WRITE_REQ	BIT(12)
+#define QUP_OP_MAX_INPUT_DONE_FLAG	BIT(11)
+#define QUP_OP_MAX_OUTPUT_DONE_FLAG	BIT(10)
+#define QUP_OP_IN_SERVICE_FLAG		BIT(9)
+#define QUP_OP_OUT_SERVICE_FLAG		BIT(8)
+#define QUP_OP_IN_FIFO_FULL		BIT(7)
+#define QUP_OP_OUT_FIFO_FULL		BIT(6)
+#define QUP_OP_IN_FIFO_NOT_EMPTY	BIT(5)
+#define QUP_OP_OUT_FIFO_NOT_EMPTY	BIT(4)
+
+/* QUP_ERROR_FLAGS and QUP_ERROR_FLAGS_EN fields */
+#define QUP_ERROR_OUTPUT_OVER_RUN	BIT(5)
+#define QUP_ERROR_INPUT_UNDER_RUN	BIT(4)
+#define QUP_ERROR_OUTPUT_UNDER_RUN	BIT(3)
+#define QUP_ERROR_INPUT_OVER_RUN	BIT(2)
+
+/* SPI_CONFIG fields */
+#define SPI_CONFIG_HS_MODE		BIT(10)
+#define SPI_CONFIG_INPUT_FIRST		BIT(9)
+#define SPI_CONFIG_LOOPBACK		BIT(8)
+
+/* SPI_IO_CONTROL fields */
+#define SPI_IO_C_FORCE_CS		BIT(11)
+#define SPI_IO_C_CLK_IDLE_HIGH		BIT(10)
+#define SPI_IO_C_MX_CS_MODE		BIT(8)
+#define SPI_IO_C_CS_N_POLARITY_0	BIT(4)
+#define SPI_IO_C_CS_SELECT(x)		(((x) & 3) << 2)
+#define SPI_IO_C_CS_SELECT_MASK		0x000c
+#define SPI_IO_C_TRISTATE_CS		BIT(1)
+#define SPI_IO_C_NO_TRI_STATE		BIT(0)
+
+/* SPI_ERROR_FLAGS and SPI_ERROR_FLAGS_EN fields */
+#define SPI_ERROR_CLK_OVER_RUN		BIT(1)
+#define SPI_ERROR_CLK_UNDER_RUN		BIT(0)
+
+#define SPI_NUM_CHIPSELECTS		4
+
+#define SPI_MAX_XFER			(SZ_64K - 64)
+
+/* high speed mode is when bus rate is greater then 26MHz */
+#define SPI_HS_MIN_RATE			26000000
+#define SPI_MAX_RATE			50000000
+
+#define SPI_DELAY_THRESHOLD		1
+#define SPI_DELAY_RETRY			10
+
+struct spi_qup {
+	void __iomem		*base;
+	struct device		*dev;
+	struct clk		*cclk;	/* core clock */
+	struct clk		*iclk;	/* interface clock */
+	int			irq;
+	spinlock_t		lock;
+
+	int			in_fifo_sz;
+	int			out_fifo_sz;
+	int			in_blk_sz;
+	int			out_blk_sz;
+
+	struct spi_transfer	*xfer;
+	struct completion	done;
+	int			error;
+	int			w_size;	/* bytes per SPI word */
+	int			n_words;
+	int			tx_bytes;
+	int			rx_bytes;
+	const u8		*tx_buf;
+	u8			*rx_buf;
+	int			qup_v1;
+
+	int			mode;
+	struct dma_slave_config	rx_conf;
+	struct dma_slave_config	tx_conf;
+};
+
+static int spi_qup_io_config(struct spi_device *spi, struct spi_transfer *xfer);
+
+static inline bool spi_qup_is_flag_set(struct spi_qup *controller, u32 flag)
+{
+	u32 opflag = readl_relaxed(controller->base + QUP_OPERATIONAL);
+
+	return (opflag & flag) != 0;
+}
+
+static inline bool spi_qup_is_dma_xfer(int mode)
+{
+	if (mode == QUP_IO_M_MODE_DMOV || mode == QUP_IO_M_MODE_BAM)
+		return true;
+
+	return false;
+}
+
+/* get's the transaction size length */
+static inline unsigned int spi_qup_len(struct spi_qup *controller)
+{
+	return controller->n_words * controller->w_size;
+}
+
+static inline bool spi_qup_is_valid_state(struct spi_qup *controller)
+{
+	u32 opstate = readl_relaxed(controller->base + QUP_STATE);
+
+	return opstate & QUP_STATE_VALID;
+}
+
+static int spi_qup_set_state(struct spi_qup *controller, u32 state)
+{
+	unsigned long loop;
+	u32 cur_state;
+
+	loop = 0;
+	while (!spi_qup_is_valid_state(controller)) {
+
+		usleep_range(SPI_DELAY_THRESHOLD, SPI_DELAY_THRESHOLD * 2);
+
+		if (++loop > SPI_DELAY_RETRY)
+			return -EIO;
+	}
+
+	if (loop)
+		dev_dbg(controller->dev, "invalid state for %ld,us %d\n",
+			loop, state);
+
+	cur_state = readl_relaxed(controller->base + QUP_STATE);
+	/*
+	 * Per spec: for PAUSE_STATE to RESET_STATE, two writes
+	 * of (b10) are required
+	 */
+	if (((cur_state & QUP_STATE_MASK) == QUP_STATE_PAUSE) &&
+	    (state == QUP_STATE_RESET)) {
+		writel_relaxed(QUP_STATE_CLEAR, controller->base + QUP_STATE);
+		writel_relaxed(QUP_STATE_CLEAR, controller->base + QUP_STATE);
+	} else {
+		cur_state &= ~QUP_STATE_MASK;
+		cur_state |= state;
+		writel_relaxed(cur_state, controller->base + QUP_STATE);
+	}
+
+	loop = 0;
+	while (!spi_qup_is_valid_state(controller)) {
+
+		usleep_range(SPI_DELAY_THRESHOLD, SPI_DELAY_THRESHOLD * 2);
+
+		if (++loop > SPI_DELAY_RETRY)
+			return -EIO;
+	}
+
+	return 0;
+}
+
+static void spi_qup_read_from_fifo(struct spi_qup *controller, u32 num_words)
+{
+	u8 *rx_buf = controller->rx_buf;
+	int i, shift, num_bytes;
+	u32 word;
+
+	for (; num_words; num_words--) {
+
+		word = readl_relaxed(controller->base + QUP_INPUT_FIFO);
+
+		num_bytes = min_t(int, spi_qup_len(controller) -
+				       controller->rx_bytes,
+				       controller->w_size);
+
+		if (!rx_buf) {
+			controller->rx_bytes += num_bytes;
+			continue;
+		}
+
+		for (i = 0; i < num_bytes; i++, controller->rx_bytes++) {
+			/*
+			 * The data format depends on bytes per SPI word:
+			 *  4 bytes: 0x12345678
+			 *  2 bytes: 0x00001234
+			 *  1 byte : 0x00000012
+			 */
+			shift = BITS_PER_BYTE;
+			shift *= (controller->w_size - i - 1);
+			rx_buf[controller->rx_bytes] = word >> shift;
+		}
+	}
+}
+
+static void spi_qup_read(struct spi_qup *controller, u32 *opflags)
+{
+	u32 remainder, words_per_block, num_words;
+	bool is_block_mode = controller->mode == QUP_IO_M_MODE_BLOCK;
+
+	remainder = DIV_ROUND_UP(spi_qup_len(controller) - controller->rx_bytes,
+				 controller->w_size);
+	words_per_block = controller->in_blk_sz >> 2;
+
+	do {
+		/* ACK by clearing service flag */
+		writel_relaxed(QUP_OP_IN_SERVICE_FLAG,
+			       controller->base + QUP_OPERATIONAL);
+
+		if (is_block_mode) {
+			num_words = (remainder > words_per_block) ?
+					words_per_block : remainder;
+		} else {
+			if (!spi_qup_is_flag_set(controller,
+						 QUP_OP_IN_FIFO_NOT_EMPTY))
+				break;
+
+			num_words = 1;
+		}
+
+		/* read up to the maximum transfer size available */
+		spi_qup_read_from_fifo(controller, num_words);
+
+		remainder -= num_words;
+
+		/* if block mode, check to see if next block is available */
+		if (is_block_mode && !spi_qup_is_flag_set(controller,
+					QUP_OP_IN_BLOCK_READ_REQ))
+			break;
+
+	} while (remainder);
+
+	/*
+	 * Due to extra stickiness of the QUP_OP_IN_SERVICE_FLAG during block
+	 * reads, it has to be cleared again at the very end.  However, be sure
+	 * to refresh opflags value because MAX_INPUT_DONE_FLAG may now be
+	 * present and this is used to determine if transaction is complete
+	 */
+	*opflags = readl_relaxed(controller->base + QUP_OPERATIONAL);
+	if (is_block_mode && *opflags & QUP_OP_MAX_INPUT_DONE_FLAG)
+		writel_relaxed(QUP_OP_IN_SERVICE_FLAG,
+			       controller->base + QUP_OPERATIONAL);
+
+}
+
+static void spi_qup_write_to_fifo(struct spi_qup *controller, u32 num_words)
+{
+	const u8 *tx_buf = controller->tx_buf;
+	int i, num_bytes;
+	u32 word, data;
+
+	for (; num_words; num_words--) {
+		word = 0;
+
+		num_bytes = min_t(int, spi_qup_len(controller) -
+				       controller->tx_bytes,
+				       controller->w_size);
+		if (tx_buf)
+			for (i = 0; i < num_bytes; i++) {
+				data = tx_buf[controller->tx_bytes + i];
+				word |= data << (BITS_PER_BYTE * (3 - i));
+			}
+
+		controller->tx_bytes += num_bytes;
+
+		writel_relaxed(word, controller->base + QUP_OUTPUT_FIFO);
+	}
+}
+
+static void spi_qup_dma_done(void *data)
+{
+	struct spi_qup *qup = data;
+
+	complete(&qup->done);
+}
+
+static void spi_qup_write(struct spi_qup *controller)
+{
+	bool is_block_mode = controller->mode == QUP_IO_M_MODE_BLOCK;
+	u32 remainder, words_per_block, num_words;
+
+	remainder = DIV_ROUND_UP(spi_qup_len(controller) - controller->tx_bytes,
+				 controller->w_size);
+	words_per_block = controller->out_blk_sz >> 2;
+
+	do {
+		/* ACK by clearing service flag */
+		writel_relaxed(QUP_OP_OUT_SERVICE_FLAG,
+			       controller->base + QUP_OPERATIONAL);
+
+		if (is_block_mode) {
+			num_words = (remainder > words_per_block) ?
+				words_per_block : remainder;
+		} else {
+			if (spi_qup_is_flag_set(controller,
+						QUP_OP_OUT_FIFO_FULL))
+				break;
+
+			num_words = 1;
+		}
+
+		spi_qup_write_to_fifo(controller, num_words);
+
+		remainder -= num_words;
+
+		/* if block mode, check to see if next block is available */
+		if (is_block_mode && !spi_qup_is_flag_set(controller,
+					QUP_OP_OUT_BLOCK_WRITE_REQ))
+			break;
+
+	} while (remainder);
+}
+
+static int spi_qup_prep_sg(struct spi_master *master, struct scatterlist *sgl,
+			   unsigned int nents, enum dma_transfer_direction dir,
+			   dma_async_tx_callback callback)
+{
+	struct spi_qup *qup = spi_master_get_devdata(master);
+	unsigned long flags = DMA_PREP_INTERRUPT | DMA_PREP_FENCE;
+	struct dma_async_tx_descriptor *desc;
+	struct dma_chan *chan;
+	dma_cookie_t cookie;
+
+	if (dir == DMA_MEM_TO_DEV)
+		chan = master->dma_tx;
+	else
+		chan = master->dma_rx;
+
+	desc = dmaengine_prep_slave_sg(chan, sgl, nents, dir, flags);
+	if (IS_ERR_OR_NULL(desc))
+		return desc ? PTR_ERR(desc) : -EINVAL;
+
+	desc->callback = callback;
+	desc->callback_param = qup;
+
+	cookie = dmaengine_submit(desc);
+
+	return dma_submit_error(cookie);
+}
+
+static void spi_qup_dma_terminate(struct spi_master *master,
+				  struct spi_transfer *xfer)
+{
+	if (xfer->tx_buf)
+		dmaengine_terminate_all(master->dma_tx);
+	if (xfer->rx_buf)
+		dmaengine_terminate_all(master->dma_rx);
+}
+
+static u32 spi_qup_sgl_get_nents_len(struct scatterlist *sgl, u32 max,
+				     u32 *nents)
+{
+	struct scatterlist *sg;
+	u32 total = 0;
+
+	for (sg = sgl; sg; sg = sg_next(sg)) {
+		unsigned int len = sg_dma_len(sg);
+
+		/* check for overflow as well as limit */
+		if (((total + len) < total) || ((total + len) > max))
+			break;
+
+		total += len;
+		(*nents)++;
+	}
+
+	return total;
+}
+
+static int spi_qup_do_dma(struct spi_device *spi, struct spi_transfer *xfer,
+			  unsigned long timeout)
+{
+	dma_async_tx_callback rx_done = NULL, tx_done = NULL;
+	struct spi_master *master = spi->master;
+	struct spi_qup *qup = spi_master_get_devdata(master);
+	struct scatterlist *tx_sgl, *rx_sgl;
+	int ret;
+
+	if (xfer->rx_buf)
+		rx_done = spi_qup_dma_done;
+	else if (xfer->tx_buf)
+		tx_done = spi_qup_dma_done;
+
+	rx_sgl = xfer->rx_sg.sgl;
+	tx_sgl = xfer->tx_sg.sgl;
+
+	do {
+		u32 rx_nents = 0, tx_nents = 0;
+
+		if (rx_sgl)
+			qup->n_words = spi_qup_sgl_get_nents_len(rx_sgl,
+					SPI_MAX_XFER, &rx_nents) / qup->w_size;
+		if (tx_sgl)
+			qup->n_words = spi_qup_sgl_get_nents_len(tx_sgl,
+					SPI_MAX_XFER, &tx_nents) / qup->w_size;
+		if (!qup->n_words)
+			return -EIO;
+
+		ret = spi_qup_io_config(spi, xfer);
+		if (ret)
+			return ret;
+
+		/* before issuing the descriptors, set the QUP to run */
+		ret = spi_qup_set_state(qup, QUP_STATE_RUN);
+		if (ret) {
+			dev_warn(qup->dev, "cannot set RUN state\n");
+			return ret;
+		}
+		if (rx_sgl) {
+			ret = spi_qup_prep_sg(master, rx_sgl, rx_nents,
+					      DMA_DEV_TO_MEM, rx_done);
+			if (ret)
+				return ret;
+			dma_async_issue_pending(master->dma_rx);
+		}
+
+		if (tx_sgl) {
+			ret = spi_qup_prep_sg(master, tx_sgl, tx_nents,
+					      DMA_MEM_TO_DEV, tx_done);
+			if (ret)
+				return ret;
+
+			dma_async_issue_pending(master->dma_tx);
+		}
+
+		if (!wait_for_completion_timeout(&qup->done, timeout))
+			return -ETIMEDOUT;
+
+		for (; rx_sgl && rx_nents--; rx_sgl = sg_next(rx_sgl))
+			;
+		for (; tx_sgl && tx_nents--; tx_sgl = sg_next(tx_sgl))
+			;
+
+	} while (rx_sgl || tx_sgl);
+
+	return 0;
+}
+
+static int spi_qup_do_pio(struct spi_device *spi, struct spi_transfer *xfer,
+			  unsigned long timeout)
+{
+	struct spi_master *master = spi->master;
+	struct spi_qup *qup = spi_master_get_devdata(master);
+	int ret, n_words, iterations, offset = 0;
+
+	n_words = qup->n_words;
+	iterations = n_words / SPI_MAX_XFER; /* round down */
+	qup->rx_buf = xfer->rx_buf;
+	qup->tx_buf = xfer->tx_buf;
+
+	do {
+		if (iterations)
+			qup->n_words = SPI_MAX_XFER;
+		else
+			qup->n_words = n_words % SPI_MAX_XFER;
+
+		if (qup->tx_buf && offset)
+			qup->tx_buf = xfer->tx_buf + offset * SPI_MAX_XFER;
+
+		if (qup->rx_buf && offset)
+			qup->rx_buf = xfer->rx_buf + offset * SPI_MAX_XFER;
+
+		/*
+		 * if the transaction is small enough, we need
+		 * to fallback to FIFO mode
+		 */
+		if (qup->n_words <= (qup->in_fifo_sz / sizeof(u32)))
+			qup->mode = QUP_IO_M_MODE_FIFO;
+
+		ret = spi_qup_io_config(spi, xfer);
+		if (ret)
+			return ret;
+
+		ret = spi_qup_set_state(qup, QUP_STATE_RUN);
+		if (ret) {
+			dev_warn(qup->dev, "cannot set RUN state\n");
+			return ret;
+		}
+
+		ret = spi_qup_set_state(qup, QUP_STATE_PAUSE);
+		if (ret) {
+			dev_warn(qup->dev, "cannot set PAUSE state\n");
+			return ret;
+		}
+
+		if (qup->mode == QUP_IO_M_MODE_FIFO)
+			spi_qup_write(qup);
+
+		ret = spi_qup_set_state(qup, QUP_STATE_RUN);
+		if (ret) {
+			dev_warn(qup->dev, "cannot set RUN state\n");
+			return ret;
+		}
+
+		if (!wait_for_completion_timeout(&qup->done, timeout))
+			return -ETIMEDOUT;
+
+		offset++;
+	} while (iterations--);
+
+	return 0;
+}
+
+static irqreturn_t spi_qup_qup_irq(int irq, void *dev_id)
+{
+	struct spi_qup *controller = dev_id;
+	u32 opflags, qup_err, spi_err;
+	int error = 0;
+
+	qup_err = readl_relaxed(controller->base + QUP_ERROR_FLAGS);
+	spi_err = readl_relaxed(controller->base + SPI_ERROR_FLAGS);
+	opflags = readl_relaxed(controller->base + QUP_OPERATIONAL);
+
+	writel_relaxed(qup_err, controller->base + QUP_ERROR_FLAGS);
+	writel_relaxed(spi_err, controller->base + SPI_ERROR_FLAGS);
+
+	if (qup_err) {
+		if (qup_err & QUP_ERROR_OUTPUT_OVER_RUN)
+			dev_warn(controller->dev, "OUTPUT_OVER_RUN\n");
+		if (qup_err & QUP_ERROR_INPUT_UNDER_RUN)
+			dev_warn(controller->dev, "INPUT_UNDER_RUN\n");
+		if (qup_err & QUP_ERROR_OUTPUT_UNDER_RUN)
+			dev_warn(controller->dev, "OUTPUT_UNDER_RUN\n");
+		if (qup_err & QUP_ERROR_INPUT_OVER_RUN)
+			dev_warn(controller->dev, "INPUT_OVER_RUN\n");
+
+		error = -EIO;
+	}
+
+	if (spi_err) {
+		if (spi_err & SPI_ERROR_CLK_OVER_RUN)
+			dev_warn(controller->dev, "CLK_OVER_RUN\n");
+		if (spi_err & SPI_ERROR_CLK_UNDER_RUN)
+			dev_warn(controller->dev, "CLK_UNDER_RUN\n");
+
+		error = -EIO;
+	}
+
+	if (spi_qup_is_dma_xfer(controller->mode)) {
+		writel_relaxed(opflags, controller->base + QUP_OPERATIONAL);
+	} else {
+		if (opflags & QUP_OP_IN_SERVICE_FLAG)
+			spi_qup_read(controller, &opflags);
+
+		if (opflags & QUP_OP_OUT_SERVICE_FLAG)
+			spi_qup_write(controller);
+	}
+
+	if ((opflags & QUP_OP_MAX_INPUT_DONE_FLAG) || error)
+		complete(&controller->done);
+
+	return IRQ_HANDLED;
+}
+
+/* set clock freq ... bits per word, determine mode */
+static int spi_qup_io_prep(struct spi_device *spi, struct spi_transfer *xfer)
+{
+	struct spi_qup *controller = spi_master_get_devdata(spi->master);
+	int ret;
+
+	if (spi->mode & SPI_LOOP && xfer->len > controller->in_fifo_sz) {
+		dev_err(controller->dev, "too big size for loopback %d > %d\n",
+			xfer->len, controller->in_fifo_sz);
+		return -EIO;
+	}
+
+	ret = clk_set_rate(controller->cclk, xfer->speed_hz);
+	if (ret) {
+		dev_err(controller->dev, "fail to set frequency %d",
+			xfer->speed_hz);
+		return -EIO;
+	}
+
+	controller->w_size = DIV_ROUND_UP(xfer->bits_per_word, 8);
+	controller->n_words = xfer->len / controller->w_size;
+
+	if (controller->n_words <= (controller->in_fifo_sz / sizeof(u32)))
+		controller->mode = QUP_IO_M_MODE_FIFO;
+	else if (spi->master->can_dma &&
+		 spi->master->can_dma(spi->master, spi, xfer) &&
+		 spi->master->cur_msg_mapped)
+		controller->mode = QUP_IO_M_MODE_BAM;
+	else
+		controller->mode = QUP_IO_M_MODE_BLOCK;
+
+	return 0;
+}
+
+/* prep qup for another spi transaction of specific type */
+static int spi_qup_io_config(struct spi_device *spi, struct spi_transfer *xfer)
+{
+	struct spi_qup *controller = spi_master_get_devdata(spi->master);
+	u32 config, iomode, control;
+	unsigned long flags;
+
+	spin_lock_irqsave(&controller->lock, flags);
+	controller->xfer     = xfer;
+	controller->error    = 0;
+	controller->rx_bytes = 0;
+	controller->tx_bytes = 0;
+	spin_unlock_irqrestore(&controller->lock, flags);
+
+
+	if (spi_qup_set_state(controller, QUP_STATE_RESET)) {
+		dev_err(controller->dev, "cannot set RESET state\n");
+		return -EIO;
+	}
+
+	switch (controller->mode) {
+	case QUP_IO_M_MODE_FIFO:
+		writel_relaxed(controller->n_words,
+			       controller->base + QUP_MX_READ_CNT);
+		writel_relaxed(controller->n_words,
+			       controller->base + QUP_MX_WRITE_CNT);
+		/* must be zero for FIFO */
+		writel_relaxed(0, controller->base + QUP_MX_INPUT_CNT);
+		writel_relaxed(0, controller->base + QUP_MX_OUTPUT_CNT);
+		break;
+	case QUP_IO_M_MODE_BAM:
+		writel_relaxed(controller->n_words,
+			       controller->base + QUP_MX_INPUT_CNT);
+		writel_relaxed(controller->n_words,
+			       controller->base + QUP_MX_OUTPUT_CNT);
+		/* must be zero for BLOCK and BAM */
+		writel_relaxed(0, controller->base + QUP_MX_READ_CNT);
+		writel_relaxed(0, controller->base + QUP_MX_WRITE_CNT);
+
+		if (!controller->qup_v1) {
+			void __iomem *input_cnt;
+
+			input_cnt = controller->base + QUP_MX_INPUT_CNT;
+			/*
+			 * for DMA transfers, both QUP_MX_INPUT_CNT and
+			 * QUP_MX_OUTPUT_CNT must be zero to all cases but one.
+			 * That case is a non-balanced transfer when there is
+			 * only a rx_buf.
+			 */
+			if (xfer->tx_buf)
+				writel_relaxed(0, input_cnt);
+			else
+				writel_relaxed(controller->n_words, input_cnt);
+
+			writel_relaxed(0, controller->base + QUP_MX_OUTPUT_CNT);
+		}
+		break;
+	case QUP_IO_M_MODE_BLOCK:
+		reinit_completion(&controller->done);
+		writel_relaxed(controller->n_words,
+			       controller->base + QUP_MX_INPUT_CNT);
+		writel_relaxed(controller->n_words,
+			       controller->base + QUP_MX_OUTPUT_CNT);
+		/* must be zero for BLOCK and BAM */
+		writel_relaxed(0, controller->base + QUP_MX_READ_CNT);
+		writel_relaxed(0, controller->base + QUP_MX_WRITE_CNT);
+		break;
+	default:
+		dev_err(controller->dev, "unknown mode = %d\n",
+				controller->mode);
+		return -EIO;
+	}
+
+	iomode = readl_relaxed(controller->base + QUP_IO_M_MODES);
+	/* Set input and output transfer mode */
+	iomode &= ~(QUP_IO_M_INPUT_MODE_MASK | QUP_IO_M_OUTPUT_MODE_MASK);
+
+	if (!spi_qup_is_dma_xfer(controller->mode))
+		iomode &= ~(QUP_IO_M_PACK_EN | QUP_IO_M_UNPACK_EN);
+	else
+		iomode |= QUP_IO_M_PACK_EN | QUP_IO_M_UNPACK_EN;
+
+	iomode |= (controller->mode << QUP_IO_M_OUTPUT_MODE_MASK_SHIFT);
+	iomode |= (controller->mode << QUP_IO_M_INPUT_MODE_MASK_SHIFT);
+
+	writel_relaxed(iomode, controller->base + QUP_IO_M_MODES);
+
+	control = readl_relaxed(controller->base + SPI_IO_CONTROL);
+
+	if (spi->mode & SPI_CPOL)
+		control |= SPI_IO_C_CLK_IDLE_HIGH;
+	else
+		control &= ~SPI_IO_C_CLK_IDLE_HIGH;
+
+	writel_relaxed(control, controller->base + SPI_IO_CONTROL);
+
+	config = readl_relaxed(controller->base + SPI_CONFIG);
+
+	if (spi->mode & SPI_LOOP)
+		config |= SPI_CONFIG_LOOPBACK;
+	else
+		config &= ~SPI_CONFIG_LOOPBACK;
+
+	if (spi->mode & SPI_CPHA)
+		config &= ~SPI_CONFIG_INPUT_FIRST;
+	else
+		config |= SPI_CONFIG_INPUT_FIRST;
+
+	/*
+	 * HS_MODE improves signal stability for spi-clk high rates,
+	 * but is invalid in loop back mode.
+	 */
+	if ((xfer->speed_hz >= SPI_HS_MIN_RATE) && !(spi->mode & SPI_LOOP))
+		config |= SPI_CONFIG_HS_MODE;
+	else
+		config &= ~SPI_CONFIG_HS_MODE;
+
+	writel_relaxed(config, controller->base + SPI_CONFIG);
+
+	config = readl_relaxed(controller->base + QUP_CONFIG);
+	config &= ~(QUP_CONFIG_NO_INPUT | QUP_CONFIG_NO_OUTPUT | QUP_CONFIG_N);
+	config |= xfer->bits_per_word - 1;
+	config |= QUP_CONFIG_SPI_MODE;
+
+	if (spi_qup_is_dma_xfer(controller->mode)) {
+		if (!xfer->tx_buf)
+			config |= QUP_CONFIG_NO_OUTPUT;
+		if (!xfer->rx_buf)
+			config |= QUP_CONFIG_NO_INPUT;
+	}
+
+	writel_relaxed(config, controller->base + QUP_CONFIG);
+
+	/* only write to OPERATIONAL_MASK when register is present */
+	if (!controller->qup_v1) {
+		u32 mask = 0;
+
+		/*
+		 * mask INPUT and OUTPUT service flags to prevent IRQs on FIFO
+		 * status change in BAM mode
+		 */
+
+		if (spi_qup_is_dma_xfer(controller->mode))
+			mask = QUP_OP_IN_SERVICE_FLAG | QUP_OP_OUT_SERVICE_FLAG;
+
+		writel_relaxed(mask, controller->base + QUP_OPERATIONAL_MASK);
+	}
+
+	return 0;
+}
+
+static int spi_qup_transfer_one(struct spi_master *master,
+			      struct spi_device *spi,
+			      struct spi_transfer *xfer)
+{
+	struct spi_qup *controller = spi_master_get_devdata(master);
+	unsigned long timeout, flags;
+	int ret = -EIO;
+
+	ret = spi_qup_io_prep(spi, xfer);
+	if (ret)
+		return ret;
+
+	timeout = DIV_ROUND_UP(xfer->speed_hz, MSEC_PER_SEC);
+	timeout = DIV_ROUND_UP(min_t(unsigned long, SPI_MAX_XFER,
+				     xfer->len) * 8, timeout);
+	timeout = 100 * msecs_to_jiffies(timeout);
+
+	reinit_completion(&controller->done);
+
+	spin_lock_irqsave(&controller->lock, flags);
+	controller->xfer     = xfer;
+	controller->error    = 0;
+	controller->rx_bytes = 0;
+	controller->tx_bytes = 0;
+	spin_unlock_irqrestore(&controller->lock, flags);
+
+	if (spi_qup_is_dma_xfer(controller->mode))
+		ret = spi_qup_do_dma(spi, xfer, timeout);
+	else
+		ret = spi_qup_do_pio(spi, xfer, timeout);
+
+	if (ret)
+		goto exit;
+
+exit:
+	spi_qup_set_state(controller, QUP_STATE_RESET);
+	spin_lock_irqsave(&controller->lock, flags);
+	if (!ret)
+		ret = controller->error;
+	spin_unlock_irqrestore(&controller->lock, flags);
+
+	if (ret && spi_qup_is_dma_xfer(controller->mode))
+		spi_qup_dma_terminate(master, xfer);
+
+	return ret;
+}
+
+static bool spi_qup_can_dma(struct spi_master *master, struct spi_device *spi,
+			    struct spi_transfer *xfer)
+{
+	struct spi_qup *qup = spi_master_get_devdata(master);
+	size_t dma_align = dma_get_cache_alignment();
+	int n_words;
+
+	if (xfer->rx_buf) {
+		if (!IS_ALIGNED((size_t)xfer->rx_buf, dma_align) ||
+		    IS_ERR_OR_NULL(master->dma_rx))
+			return false;
+		if (qup->qup_v1 && (xfer->len % qup->in_blk_sz))
+			return false;
+	}
+
+	if (xfer->tx_buf) {
+		if (!IS_ALIGNED((size_t)xfer->tx_buf, dma_align) ||
+		    IS_ERR_OR_NULL(master->dma_tx))
+			return false;
+		if (qup->qup_v1 && (xfer->len % qup->out_blk_sz))
+			return false;
+	}
+
+	n_words = xfer->len / DIV_ROUND_UP(xfer->bits_per_word, 8);
+	if (n_words <= (qup->in_fifo_sz / sizeof(u32)))
+		return false;
+
+	return true;
+}
+
+static void spi_qup_release_dma(struct spi_master *master)
+{
+	if (!IS_ERR_OR_NULL(master->dma_rx))
+		dma_release_channel(master->dma_rx);
+	if (!IS_ERR_OR_NULL(master->dma_tx))
+		dma_release_channel(master->dma_tx);
+}
+
+static int spi_qup_init_dma(struct spi_master *master, resource_size_t base)
+{
+	struct spi_qup *spi = spi_master_get_devdata(master);
+	struct dma_slave_config *rx_conf = &spi->rx_conf,
+				*tx_conf = &spi->tx_conf;
+	struct device *dev = spi->dev;
+	int ret;
+
+	/* allocate dma resources, if available */
+	master->dma_rx = dma_request_slave_channel_reason(dev, "rx");
+	if (IS_ERR(master->dma_rx))
+		return PTR_ERR(master->dma_rx);
+
+	master->dma_tx = dma_request_slave_channel_reason(dev, "tx");
+	if (IS_ERR(master->dma_tx)) {
+		ret = PTR_ERR(master->dma_tx);
+		goto err_tx;
+	}
+
+	/* set DMA parameters */
+	rx_conf->direction = DMA_DEV_TO_MEM;
+	rx_conf->device_fc = 1;
+	rx_conf->src_addr = base + QUP_INPUT_FIFO;
+	rx_conf->src_maxburst = spi->in_blk_sz;
+
+	tx_conf->direction = DMA_MEM_TO_DEV;
+	tx_conf->device_fc = 1;
+	tx_conf->dst_addr = base + QUP_OUTPUT_FIFO;
+	tx_conf->dst_maxburst = spi->out_blk_sz;
+
+	ret = dmaengine_slave_config(master->dma_rx, rx_conf);
+	if (ret) {
+		dev_err(dev, "failed to configure RX channel\n");
+		goto err;
+	}
+
+	ret = dmaengine_slave_config(master->dma_tx, tx_conf);
+	if (ret) {
+		dev_err(dev, "failed to configure TX channel\n");
+		goto err;
+	}
+
+	return 0;
+
+err:
+	dma_release_channel(master->dma_tx);
+err_tx:
+	dma_release_channel(master->dma_rx);
+	return ret;
+}
+
+static void spi_qup_set_cs(struct spi_device *spi, bool val)
+{
+	struct spi_qup *controller;
+	u32 spi_ioc;
+	u32 spi_ioc_orig;
+
+	controller = spi_master_get_devdata(spi->master);
+	spi_ioc = readl_relaxed(controller->base + SPI_IO_CONTROL);
+	spi_ioc_orig = spi_ioc;
+	if (!val)
+		spi_ioc |= SPI_IO_C_FORCE_CS;
+	else
+		spi_ioc &= ~SPI_IO_C_FORCE_CS;
+
+	if (spi_ioc != spi_ioc_orig)
+		writel_relaxed(spi_ioc, controller->base + SPI_IO_CONTROL);
+}
+
+static int spi_qup_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct clk *iclk, *cclk;
+	struct spi_qup *controller;
+	struct resource *res;
+	struct device *dev;
+	void __iomem *base;
+	u32 max_freq, iomode, num_cs;
+	int ret, irq, size;
+
+	dev = &pdev->dev;
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	base = devm_ioremap_resource(dev, res);
+	if (IS_ERR(base))
+		return PTR_ERR(base);
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0)
+		return irq;
+
+	cclk = devm_clk_get(dev, "core");
+	if (IS_ERR(cclk))
+		return PTR_ERR(cclk);
+
+	iclk = devm_clk_get(dev, "iface");
+	if (IS_ERR(iclk))
+		return PTR_ERR(iclk);
+
+	/* This is optional parameter */
+	if (of_property_read_u32(dev->of_node, "spi-max-frequency", &max_freq))
+		max_freq = SPI_MAX_RATE;
+
+	if (!max_freq || max_freq > SPI_MAX_RATE) {
+		dev_err(dev, "invalid clock frequency %d\n", max_freq);
+		return -ENXIO;
+	}
+
+	ret = clk_prepare_enable(cclk);
+	if (ret) {
+		dev_err(dev, "cannot enable core clock\n");
+		return ret;
+	}
+
+	ret = clk_prepare_enable(iclk);
+	if (ret) {
+		clk_disable_unprepare(cclk);
+		dev_err(dev, "cannot enable iface clock\n");
+		return ret;
+	}
+
+	master = spi_alloc_master(dev, sizeof(struct spi_qup));
+	if (!master) {
+		clk_disable_unprepare(cclk);
+		clk_disable_unprepare(iclk);
+		dev_err(dev, "cannot allocate master\n");
+		return -ENOMEM;
+	}
+
+	/* use num-cs unless not present or out of range */
+	if (of_property_read_u32(dev->of_node, "num-cs", &num_cs) ||
+	    num_cs > SPI_NUM_CHIPSELECTS)
+		master->num_chipselect = SPI_NUM_CHIPSELECTS;
+	else
+		master->num_chipselect = num_cs;
+
+	master->bus_num = pdev->id;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LOOP;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 32);
+	master->max_speed_hz = max_freq;
+	master->transfer_one = spi_qup_transfer_one;
+	master->dev.of_node = pdev->dev.of_node;
+	master->auto_runtime_pm = true;
+	master->dma_alignment = dma_get_cache_alignment();
+	master->max_dma_len = SPI_MAX_XFER;
+
+	platform_set_drvdata(pdev, master);
+
+	controller = spi_master_get_devdata(master);
+
+	controller->dev = dev;
+	controller->base = base;
+	controller->iclk = iclk;
+	controller->cclk = cclk;
+	controller->irq = irq;
+
+	ret = spi_qup_init_dma(master, res->start);
+	if (ret == -EPROBE_DEFER)
+		goto error;
+	else if (!ret)
+		master->can_dma = spi_qup_can_dma;
+
+	controller->qup_v1 = (uintptr_t)of_device_get_match_data(dev);
+
+	if (!controller->qup_v1)
+		master->set_cs = spi_qup_set_cs;
+
+	spin_lock_init(&controller->lock);
+	init_completion(&controller->done);
+
+	iomode = readl_relaxed(base + QUP_IO_M_MODES);
+
+	size = QUP_IO_M_OUTPUT_BLOCK_SIZE(iomode);
+	if (size)
+		controller->out_blk_sz = size * 16;
+	else
+		controller->out_blk_sz = 4;
+
+	size = QUP_IO_M_INPUT_BLOCK_SIZE(iomode);
+	if (size)
+		controller->in_blk_sz = size * 16;
+	else
+		controller->in_blk_sz = 4;
+
+	size = QUP_IO_M_OUTPUT_FIFO_SIZE(iomode);
+	controller->out_fifo_sz = controller->out_blk_sz * (2 << size);
+
+	size = QUP_IO_M_INPUT_FIFO_SIZE(iomode);
+	controller->in_fifo_sz = controller->in_blk_sz * (2 << size);
+
+	dev_info(dev, "IN:block:%d, fifo:%d, OUT:block:%d, fifo:%d\n",
+		 controller->in_blk_sz, controller->in_fifo_sz,
+		 controller->out_blk_sz, controller->out_fifo_sz);
+
+	writel_relaxed(1, base + QUP_SW_RESET);
+
+	ret = spi_qup_set_state(controller, QUP_STATE_RESET);
+	if (ret) {
+		dev_err(dev, "cannot set RESET state\n");
+		goto error_dma;
+	}
+
+	writel_relaxed(0, base + QUP_OPERATIONAL);
+	writel_relaxed(0, base + QUP_IO_M_MODES);
+
+	if (!controller->qup_v1)
+		writel_relaxed(0, base + QUP_OPERATIONAL_MASK);
+
+	writel_relaxed(SPI_ERROR_CLK_UNDER_RUN | SPI_ERROR_CLK_OVER_RUN,
+		       base + SPI_ERROR_FLAGS_EN);
+
+	/* if earlier version of the QUP, disable INPUT_OVERRUN */
+	if (controller->qup_v1)
+		writel_relaxed(QUP_ERROR_OUTPUT_OVER_RUN |
+			QUP_ERROR_INPUT_UNDER_RUN | QUP_ERROR_OUTPUT_UNDER_RUN,
+			base + QUP_ERROR_FLAGS_EN);
+
+	writel_relaxed(0, base + SPI_CONFIG);
+	writel_relaxed(SPI_IO_C_NO_TRI_STATE, base + SPI_IO_CONTROL);
+
+	ret = devm_request_irq(dev, irq, spi_qup_qup_irq,
+			       IRQF_TRIGGER_HIGH, pdev->name, controller);
+	if (ret)
+		goto error_dma;
+
+	pm_runtime_set_autosuspend_delay(dev, MSEC_PER_SEC);
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_set_active(dev);
+	pm_runtime_enable(dev);
+
+	ret = devm_spi_register_master(dev, master);
+	if (ret)
+		goto disable_pm;
+
+	return 0;
+
+disable_pm:
+	pm_runtime_disable(&pdev->dev);
+error_dma:
+	spi_qup_release_dma(master);
+error:
+	clk_disable_unprepare(cclk);
+	clk_disable_unprepare(iclk);
+	spi_master_put(master);
+	return ret;
+}
+
+#ifdef CONFIG_PM
+static int spi_qup_pm_suspend_runtime(struct device *device)
+{
+	struct spi_master *master = dev_get_drvdata(device);
+	struct spi_qup *controller = spi_master_get_devdata(master);
+	u32 config;
+
+	/* Enable clocks auto gaiting */
+	config = readl(controller->base + QUP_CONFIG);
+	config |= QUP_CONFIG_CLOCK_AUTO_GATE;
+	writel_relaxed(config, controller->base + QUP_CONFIG);
+
+	clk_disable_unprepare(controller->cclk);
+	clk_disable_unprepare(controller->iclk);
+
+	return 0;
+}
+
+static int spi_qup_pm_resume_runtime(struct device *device)
+{
+	struct spi_master *master = dev_get_drvdata(device);
+	struct spi_qup *controller = spi_master_get_devdata(master);
+	u32 config;
+	int ret;
+
+	ret = clk_prepare_enable(controller->iclk);
+	if (ret)
+		return ret;
+
+	ret = clk_prepare_enable(controller->cclk);
+	if (ret)
+		return ret;
+
+	/* Disable clocks auto gaiting */
+	config = readl_relaxed(controller->base + QUP_CONFIG);
+	config &= ~QUP_CONFIG_CLOCK_AUTO_GATE;
+	writel_relaxed(config, controller->base + QUP_CONFIG);
+	return 0;
+}
+#endif /* CONFIG_PM */
+
+#ifdef CONFIG_PM_SLEEP
+static int spi_qup_suspend(struct device *device)
+{
+	struct spi_master *master = dev_get_drvdata(device);
+	struct spi_qup *controller = spi_master_get_devdata(master);
+	int ret;
+
+	ret = spi_master_suspend(master);
+	if (ret)
+		return ret;
+
+	ret = spi_qup_set_state(controller, QUP_STATE_RESET);
+	if (ret)
+		return ret;
+
+	if (!pm_runtime_suspended(device)) {
+		clk_disable_unprepare(controller->cclk);
+		clk_disable_unprepare(controller->iclk);
+	}
+	return 0;
+}
+
+static int spi_qup_resume(struct device *device)
+{
+	struct spi_master *master = dev_get_drvdata(device);
+	struct spi_qup *controller = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(controller->iclk);
+	if (ret)
+		return ret;
+
+	ret = clk_prepare_enable(controller->cclk);
+	if (ret)
+		return ret;
+
+	ret = spi_qup_set_state(controller, QUP_STATE_RESET);
+	if (ret)
+		return ret;
+
+	return spi_master_resume(master);
+}
+#endif /* CONFIG_PM_SLEEP */
+
+static int spi_qup_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = dev_get_drvdata(&pdev->dev);
+	struct spi_qup *controller = spi_master_get_devdata(master);
+	int ret;
+
+	ret = pm_runtime_get_sync(&pdev->dev);
+	if (ret < 0)
+		return ret;
+
+	ret = spi_qup_set_state(controller, QUP_STATE_RESET);
+	if (ret)
+		return ret;
+
+	spi_qup_release_dma(master);
+
+	clk_disable_unprepare(controller->cclk);
+	clk_disable_unprepare(controller->iclk);
+
+	pm_runtime_put_noidle(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+
+	return 0;
+}
+
+static const struct of_device_id spi_qup_dt_match[] = {
+	{ .compatible = "qcom,spi-qup-v1.1.1", .data = (void *)1, },
+	{ .compatible = "qcom,spi-qup-v2.1.1", },
+	{ .compatible = "qcom,spi-qup-v2.2.1", },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, spi_qup_dt_match);
+
+static const struct dev_pm_ops spi_qup_dev_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(spi_qup_suspend, spi_qup_resume)
+	SET_RUNTIME_PM_OPS(spi_qup_pm_suspend_runtime,
+			   spi_qup_pm_resume_runtime,
+			   NULL)
+};
+
+static struct platform_driver spi_qup_driver = {
+	.driver = {
+		.name		= "spi_qup",
+		.pm		= &spi_qup_dev_pm_ops,
+		.of_match_table = spi_qup_dt_match,
+	},
+	.probe = spi_qup_probe,
+	.remove = spi_qup_remove,
+};
+module_platform_driver(spi_qup_driver);
+
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:spi_qup");
diff --git a/drivers/spi/spi-rb4xx.c b/drivers/spi/spi-rb4xx.c
new file mode 100644
index 0000000..3641d0e
--- /dev/null
+++ b/drivers/spi/spi-rb4xx.c
@@ -0,0 +1,210 @@
+/*
+ * SPI controller driver for the Mikrotik RB4xx boards
+ *
+ * Copyright (C) 2010 Gabor Juhos <juhosg@openwrt.org>
+ * Copyright (C) 2015 Bert Vermeulen <bert@biot.com>
+ *
+ * This file was based on the patches for Linux 2.6.27.39 published by
+ * MikroTik for their RouterBoard 4xx series devices.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/clk.h>
+#include <linux/spi/spi.h>
+
+#include <asm/mach-ath79/ar71xx_regs.h>
+
+struct rb4xx_spi {
+	void __iomem *base;
+	struct clk *clk;
+};
+
+static inline u32 rb4xx_read(struct rb4xx_spi *rbspi, u32 reg)
+{
+	return __raw_readl(rbspi->base + reg);
+}
+
+static inline void rb4xx_write(struct rb4xx_spi *rbspi, u32 reg, u32 value)
+{
+	__raw_writel(value, rbspi->base + reg);
+}
+
+static inline void do_spi_clk(struct rb4xx_spi *rbspi, u32 spi_ioc, int value)
+{
+	u32 regval;
+
+	regval = spi_ioc;
+	if (value & BIT(0))
+		regval |= AR71XX_SPI_IOC_DO;
+
+	rb4xx_write(rbspi, AR71XX_SPI_REG_IOC, regval);
+	rb4xx_write(rbspi, AR71XX_SPI_REG_IOC, regval | AR71XX_SPI_IOC_CLK);
+}
+
+static void do_spi_byte(struct rb4xx_spi *rbspi, u32 spi_ioc, u8 byte)
+{
+	int i;
+
+	for (i = 7; i >= 0; i--)
+		do_spi_clk(rbspi, spi_ioc, byte >> i);
+}
+
+/* The CS2 pin is used to clock in a second bit per clock cycle. */
+static inline void do_spi_clk_two(struct rb4xx_spi *rbspi, u32 spi_ioc,
+				   u8 value)
+{
+	u32 regval;
+
+	regval = spi_ioc;
+	if (value & BIT(1))
+		regval |= AR71XX_SPI_IOC_DO;
+	if (value & BIT(0))
+		regval |= AR71XX_SPI_IOC_CS2;
+
+	rb4xx_write(rbspi, AR71XX_SPI_REG_IOC, regval);
+	rb4xx_write(rbspi, AR71XX_SPI_REG_IOC, regval | AR71XX_SPI_IOC_CLK);
+}
+
+/* Two bits at a time, msb first */
+static void do_spi_byte_two(struct rb4xx_spi *rbspi, u32 spi_ioc, u8 byte)
+{
+	do_spi_clk_two(rbspi, spi_ioc, byte >> 6);
+	do_spi_clk_two(rbspi, spi_ioc, byte >> 4);
+	do_spi_clk_two(rbspi, spi_ioc, byte >> 2);
+	do_spi_clk_two(rbspi, spi_ioc, byte >> 0);
+}
+
+static void rb4xx_set_cs(struct spi_device *spi, bool enable)
+{
+	struct rb4xx_spi *rbspi = spi_master_get_devdata(spi->master);
+
+	/*
+	 * Setting CS is done along with bitbanging the actual values,
+	 * since it's all on the same hardware register. However the
+	 * CPLD needs CS deselected after every command.
+	 */
+	if (enable)
+		rb4xx_write(rbspi, AR71XX_SPI_REG_IOC,
+			    AR71XX_SPI_IOC_CS0 | AR71XX_SPI_IOC_CS1);
+}
+
+static int rb4xx_transfer_one(struct spi_master *master,
+			      struct spi_device *spi, struct spi_transfer *t)
+{
+	struct rb4xx_spi *rbspi = spi_master_get_devdata(master);
+	int i;
+	u32 spi_ioc;
+	u8 *rx_buf;
+	const u8 *tx_buf;
+
+	/*
+	 * Prime the SPI register with the SPI device selected. The m25p80 boot
+	 * flash and CPLD share the CS0 pin. This works because the CPLD's
+	 * command set was designed to almost not clash with that of the
+	 * boot flash.
+	 */
+	if (spi->chip_select == 2)
+		/* MMC */
+		spi_ioc = AR71XX_SPI_IOC_CS0;
+	else
+		/* Boot flash and CPLD */
+		spi_ioc = AR71XX_SPI_IOC_CS1;
+
+	tx_buf = t->tx_buf;
+	rx_buf = t->rx_buf;
+	for (i = 0; i < t->len; ++i) {
+		if (t->tx_nbits == SPI_NBITS_DUAL)
+			/* CPLD can use two-wire transfers */
+			do_spi_byte_two(rbspi, spi_ioc, tx_buf[i]);
+		else
+			do_spi_byte(rbspi, spi_ioc, tx_buf[i]);
+		if (!rx_buf)
+			continue;
+		rx_buf[i] = rb4xx_read(rbspi, AR71XX_SPI_REG_RDS);
+	}
+	spi_finalize_current_transfer(master);
+
+	return 0;
+}
+
+static int rb4xx_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct clk *ahb_clk;
+	struct rb4xx_spi *rbspi;
+	struct resource *r;
+	int err;
+	void __iomem *spi_base;
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	spi_base = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(spi_base))
+		return PTR_ERR(spi_base);
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*rbspi));
+	if (!master)
+		return -ENOMEM;
+
+	ahb_clk = devm_clk_get(&pdev->dev, "ahb");
+	if (IS_ERR(ahb_clk))
+		return PTR_ERR(ahb_clk);
+
+	master->bus_num = 0;
+	master->num_chipselect = 3;
+	master->mode_bits = SPI_TX_DUAL;
+	master->bits_per_word_mask = BIT(7);
+	master->flags = SPI_MASTER_MUST_TX;
+	master->transfer_one = rb4xx_transfer_one;
+	master->set_cs = rb4xx_set_cs;
+
+	err = devm_spi_register_master(&pdev->dev, master);
+	if (err) {
+		dev_err(&pdev->dev, "failed to register SPI master\n");
+		return err;
+	}
+
+	err = clk_prepare_enable(ahb_clk);
+	if (err)
+		return err;
+
+	rbspi = spi_master_get_devdata(master);
+	rbspi->base = spi_base;
+	rbspi->clk = ahb_clk;
+	platform_set_drvdata(pdev, rbspi);
+
+	/* Enable SPI */
+	rb4xx_write(rbspi, AR71XX_SPI_REG_FS, AR71XX_SPI_FS_GPIO);
+
+	return 0;
+}
+
+static int rb4xx_spi_remove(struct platform_device *pdev)
+{
+	struct rb4xx_spi *rbspi = platform_get_drvdata(pdev);
+
+	clk_disable_unprepare(rbspi->clk);
+
+	return 0;
+}
+
+static struct platform_driver rb4xx_spi_drv = {
+	.probe = rb4xx_spi_probe,
+	.remove = rb4xx_spi_remove,
+	.driver = {
+		.name = "rb4xx-spi",
+	},
+};
+
+module_platform_driver(rb4xx_spi_drv);
+
+MODULE_DESCRIPTION("Mikrotik RB4xx SPI controller driver");
+MODULE_AUTHOR("Gabor Juhos <juhosg@openwrt.org>");
+MODULE_AUTHOR("Bert Vermeulen <bert@biot.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-rockchip.c b/drivers/spi/spi-rockchip.c
new file mode 100644
index 0000000..fdcf307
--- /dev/null
+++ b/drivers/spi/spi-rockchip.c
@@ -0,0 +1,949 @@
+/*
+ * Copyright (c) 2014, Fuzhou Rockchip Electronics Co., Ltd
+ * Author: Addy Ke <addy.ke@rock-chips.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ */
+
+#include <linux/clk.h>
+#include <linux/dmaengine.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/pm_runtime.h>
+#include <linux/scatterlist.h>
+
+#define DRIVER_NAME "rockchip-spi"
+
+#define ROCKCHIP_SPI_CLR_BITS(reg, bits) \
+		writel_relaxed(readl_relaxed(reg) & ~(bits), reg)
+#define ROCKCHIP_SPI_SET_BITS(reg, bits) \
+		writel_relaxed(readl_relaxed(reg) | (bits), reg)
+
+/* SPI register offsets */
+#define ROCKCHIP_SPI_CTRLR0			0x0000
+#define ROCKCHIP_SPI_CTRLR1			0x0004
+#define ROCKCHIP_SPI_SSIENR			0x0008
+#define ROCKCHIP_SPI_SER			0x000c
+#define ROCKCHIP_SPI_BAUDR			0x0010
+#define ROCKCHIP_SPI_TXFTLR			0x0014
+#define ROCKCHIP_SPI_RXFTLR			0x0018
+#define ROCKCHIP_SPI_TXFLR			0x001c
+#define ROCKCHIP_SPI_RXFLR			0x0020
+#define ROCKCHIP_SPI_SR				0x0024
+#define ROCKCHIP_SPI_IPR			0x0028
+#define ROCKCHIP_SPI_IMR			0x002c
+#define ROCKCHIP_SPI_ISR			0x0030
+#define ROCKCHIP_SPI_RISR			0x0034
+#define ROCKCHIP_SPI_ICR			0x0038
+#define ROCKCHIP_SPI_DMACR			0x003c
+#define ROCKCHIP_SPI_DMATDLR		0x0040
+#define ROCKCHIP_SPI_DMARDLR		0x0044
+#define ROCKCHIP_SPI_TXDR			0x0400
+#define ROCKCHIP_SPI_RXDR			0x0800
+
+/* Bit fields in CTRLR0 */
+#define CR0_DFS_OFFSET				0
+
+#define CR0_CFS_OFFSET				2
+
+#define CR0_SCPH_OFFSET				6
+
+#define CR0_SCPOL_OFFSET			7
+
+#define CR0_CSM_OFFSET				8
+#define CR0_CSM_KEEP				0x0
+/* ss_n be high for half sclk_out cycles */
+#define CR0_CSM_HALF				0X1
+/* ss_n be high for one sclk_out cycle */
+#define CR0_CSM_ONE					0x2
+
+/* ss_n to sclk_out delay */
+#define CR0_SSD_OFFSET				10
+/*
+ * The period between ss_n active and
+ * sclk_out active is half sclk_out cycles
+ */
+#define CR0_SSD_HALF				0x0
+/*
+ * The period between ss_n active and
+ * sclk_out active is one sclk_out cycle
+ */
+#define CR0_SSD_ONE					0x1
+
+#define CR0_EM_OFFSET				11
+#define CR0_EM_LITTLE				0x0
+#define CR0_EM_BIG					0x1
+
+#define CR0_FBM_OFFSET				12
+#define CR0_FBM_MSB					0x0
+#define CR0_FBM_LSB					0x1
+
+#define CR0_BHT_OFFSET				13
+#define CR0_BHT_16BIT				0x0
+#define CR0_BHT_8BIT				0x1
+
+#define CR0_RSD_OFFSET				14
+
+#define CR0_FRF_OFFSET				16
+#define CR0_FRF_SPI					0x0
+#define CR0_FRF_SSP					0x1
+#define CR0_FRF_MICROWIRE			0x2
+
+#define CR0_XFM_OFFSET				18
+#define CR0_XFM_MASK				(0x03 << SPI_XFM_OFFSET)
+#define CR0_XFM_TR					0x0
+#define CR0_XFM_TO					0x1
+#define CR0_XFM_RO					0x2
+
+#define CR0_OPM_OFFSET				20
+#define CR0_OPM_MASTER				0x0
+#define CR0_OPM_SLAVE				0x1
+
+#define CR0_MTM_OFFSET				0x21
+
+/* Bit fields in SER, 2bit */
+#define SER_MASK					0x3
+
+/* Bit fields in SR, 5bit */
+#define SR_MASK						0x1f
+#define SR_BUSY						(1 << 0)
+#define SR_TF_FULL					(1 << 1)
+#define SR_TF_EMPTY					(1 << 2)
+#define SR_RF_EMPTY					(1 << 3)
+#define SR_RF_FULL					(1 << 4)
+
+/* Bit fields in ISR, IMR, ISR, RISR, 5bit */
+#define INT_MASK					0x1f
+#define INT_TF_EMPTY				(1 << 0)
+#define INT_TF_OVERFLOW				(1 << 1)
+#define INT_RF_UNDERFLOW			(1 << 2)
+#define INT_RF_OVERFLOW				(1 << 3)
+#define INT_RF_FULL					(1 << 4)
+
+/* Bit fields in ICR, 4bit */
+#define ICR_MASK					0x0f
+#define ICR_ALL						(1 << 0)
+#define ICR_RF_UNDERFLOW			(1 << 1)
+#define ICR_RF_OVERFLOW				(1 << 2)
+#define ICR_TF_OVERFLOW				(1 << 3)
+
+/* Bit fields in DMACR */
+#define RF_DMA_EN					(1 << 0)
+#define TF_DMA_EN					(1 << 1)
+
+#define RXBUSY						(1 << 0)
+#define TXBUSY						(1 << 1)
+
+/* sclk_out: spi master internal logic in rk3x can support 50Mhz */
+#define MAX_SCLK_OUT		50000000
+
+/*
+ * SPI_CTRLR1 is 16-bits, so we should support lengths of 0xffff + 1. However,
+ * the controller seems to hang when given 0x10000, so stick with this for now.
+ */
+#define ROCKCHIP_SPI_MAX_TRANLEN		0xffff
+
+#define ROCKCHIP_SPI_MAX_CS_NUM			2
+
+enum rockchip_ssi_type {
+	SSI_MOTO_SPI = 0,
+	SSI_TI_SSP,
+	SSI_NS_MICROWIRE,
+};
+
+struct rockchip_spi_dma_data {
+	struct dma_chan *ch;
+	enum dma_transfer_direction direction;
+	dma_addr_t addr;
+};
+
+struct rockchip_spi {
+	struct device *dev;
+	struct spi_master *master;
+
+	struct clk *spiclk;
+	struct clk *apb_pclk;
+
+	void __iomem *regs;
+	/*depth of the FIFO buffer */
+	u32 fifo_len;
+	/* max bus freq supported */
+	u32 max_freq;
+	/* supported slave numbers */
+	enum rockchip_ssi_type type;
+
+	u16 mode;
+	u8 tmode;
+	u8 bpw;
+	u8 n_bytes;
+	u32 rsd_nsecs;
+	unsigned len;
+	u32 speed;
+
+	const void *tx;
+	const void *tx_end;
+	void *rx;
+	void *rx_end;
+
+	u32 state;
+	/* protect state */
+	spinlock_t lock;
+
+	bool cs_asserted[ROCKCHIP_SPI_MAX_CS_NUM];
+
+	u32 use_dma;
+	struct sg_table tx_sg;
+	struct sg_table rx_sg;
+	struct rockchip_spi_dma_data dma_rx;
+	struct rockchip_spi_dma_data dma_tx;
+	struct dma_slave_caps dma_caps;
+};
+
+static inline void spi_enable_chip(struct rockchip_spi *rs, int enable)
+{
+	writel_relaxed((enable ? 1 : 0), rs->regs + ROCKCHIP_SPI_SSIENR);
+}
+
+static inline void spi_set_clk(struct rockchip_spi *rs, u16 div)
+{
+	writel_relaxed(div, rs->regs + ROCKCHIP_SPI_BAUDR);
+}
+
+static inline void flush_fifo(struct rockchip_spi *rs)
+{
+	while (readl_relaxed(rs->regs + ROCKCHIP_SPI_RXFLR))
+		readl_relaxed(rs->regs + ROCKCHIP_SPI_RXDR);
+}
+
+static inline void wait_for_idle(struct rockchip_spi *rs)
+{
+	unsigned long timeout = jiffies + msecs_to_jiffies(5);
+
+	do {
+		if (!(readl_relaxed(rs->regs + ROCKCHIP_SPI_SR) & SR_BUSY))
+			return;
+	} while (!time_after(jiffies, timeout));
+
+	dev_warn(rs->dev, "spi controller is in busy state!\n");
+}
+
+static u32 get_fifo_len(struct rockchip_spi *rs)
+{
+	u32 fifo;
+
+	for (fifo = 2; fifo < 32; fifo++) {
+		writel_relaxed(fifo, rs->regs + ROCKCHIP_SPI_TXFTLR);
+		if (fifo != readl_relaxed(rs->regs + ROCKCHIP_SPI_TXFTLR))
+			break;
+	}
+
+	writel_relaxed(0, rs->regs + ROCKCHIP_SPI_TXFTLR);
+
+	return (fifo == 31) ? 0 : fifo;
+}
+
+static inline u32 tx_max(struct rockchip_spi *rs)
+{
+	u32 tx_left, tx_room;
+
+	tx_left = (rs->tx_end - rs->tx) / rs->n_bytes;
+	tx_room = rs->fifo_len - readl_relaxed(rs->regs + ROCKCHIP_SPI_TXFLR);
+
+	return min(tx_left, tx_room);
+}
+
+static inline u32 rx_max(struct rockchip_spi *rs)
+{
+	u32 rx_left = (rs->rx_end - rs->rx) / rs->n_bytes;
+	u32 rx_room = (u32)readl_relaxed(rs->regs + ROCKCHIP_SPI_RXFLR);
+
+	return min(rx_left, rx_room);
+}
+
+static void rockchip_spi_set_cs(struct spi_device *spi, bool enable)
+{
+	struct spi_master *master = spi->master;
+	struct rockchip_spi *rs = spi_master_get_devdata(master);
+	bool cs_asserted = !enable;
+
+	/* Return immediately for no-op */
+	if (cs_asserted == rs->cs_asserted[spi->chip_select])
+		return;
+
+	if (cs_asserted) {
+		/* Keep things powered as long as CS is asserted */
+		pm_runtime_get_sync(rs->dev);
+
+		ROCKCHIP_SPI_SET_BITS(rs->regs + ROCKCHIP_SPI_SER,
+				      BIT(spi->chip_select));
+	} else {
+		ROCKCHIP_SPI_CLR_BITS(rs->regs + ROCKCHIP_SPI_SER,
+				      BIT(spi->chip_select));
+
+		/* Drop reference from when we first asserted CS */
+		pm_runtime_put(rs->dev);
+	}
+
+	rs->cs_asserted[spi->chip_select] = cs_asserted;
+}
+
+static int rockchip_spi_prepare_message(struct spi_master *master,
+					struct spi_message *msg)
+{
+	struct rockchip_spi *rs = spi_master_get_devdata(master);
+	struct spi_device *spi = msg->spi;
+
+	rs->mode = spi->mode;
+
+	return 0;
+}
+
+static void rockchip_spi_handle_err(struct spi_master *master,
+				    struct spi_message *msg)
+{
+	unsigned long flags;
+	struct rockchip_spi *rs = spi_master_get_devdata(master);
+
+	spin_lock_irqsave(&rs->lock, flags);
+
+	/*
+	 * For DMA mode, we need terminate DMA channel and flush
+	 * fifo for the next transfer if DMA thansfer timeout.
+	 * handle_err() was called by core if transfer failed.
+	 * Maybe it is reasonable for error handling here.
+	 */
+	if (rs->use_dma) {
+		if (rs->state & RXBUSY) {
+			dmaengine_terminate_async(rs->dma_rx.ch);
+			flush_fifo(rs);
+		}
+
+		if (rs->state & TXBUSY)
+			dmaengine_terminate_async(rs->dma_tx.ch);
+	}
+
+	spin_unlock_irqrestore(&rs->lock, flags);
+}
+
+static int rockchip_spi_unprepare_message(struct spi_master *master,
+					  struct spi_message *msg)
+{
+	struct rockchip_spi *rs = spi_master_get_devdata(master);
+
+	spi_enable_chip(rs, 0);
+
+	return 0;
+}
+
+static void rockchip_spi_pio_writer(struct rockchip_spi *rs)
+{
+	u32 max = tx_max(rs);
+	u32 txw = 0;
+
+	while (max--) {
+		if (rs->n_bytes == 1)
+			txw = *(u8 *)(rs->tx);
+		else
+			txw = *(u16 *)(rs->tx);
+
+		writel_relaxed(txw, rs->regs + ROCKCHIP_SPI_TXDR);
+		rs->tx += rs->n_bytes;
+	}
+}
+
+static void rockchip_spi_pio_reader(struct rockchip_spi *rs)
+{
+	u32 max = rx_max(rs);
+	u32 rxw;
+
+	while (max--) {
+		rxw = readl_relaxed(rs->regs + ROCKCHIP_SPI_RXDR);
+		if (rs->n_bytes == 1)
+			*(u8 *)(rs->rx) = (u8)rxw;
+		else
+			*(u16 *)(rs->rx) = (u16)rxw;
+		rs->rx += rs->n_bytes;
+	}
+}
+
+static int rockchip_spi_pio_transfer(struct rockchip_spi *rs)
+{
+	int remain = 0;
+
+	do {
+		if (rs->tx) {
+			remain = rs->tx_end - rs->tx;
+			rockchip_spi_pio_writer(rs);
+		}
+
+		if (rs->rx) {
+			remain = rs->rx_end - rs->rx;
+			rockchip_spi_pio_reader(rs);
+		}
+
+		cpu_relax();
+	} while (remain);
+
+	/* If tx, wait until the FIFO data completely. */
+	if (rs->tx)
+		wait_for_idle(rs);
+
+	spi_enable_chip(rs, 0);
+
+	return 0;
+}
+
+static void rockchip_spi_dma_rxcb(void *data)
+{
+	unsigned long flags;
+	struct rockchip_spi *rs = data;
+
+	spin_lock_irqsave(&rs->lock, flags);
+
+	rs->state &= ~RXBUSY;
+	if (!(rs->state & TXBUSY)) {
+		spi_enable_chip(rs, 0);
+		spi_finalize_current_transfer(rs->master);
+	}
+
+	spin_unlock_irqrestore(&rs->lock, flags);
+}
+
+static void rockchip_spi_dma_txcb(void *data)
+{
+	unsigned long flags;
+	struct rockchip_spi *rs = data;
+
+	/* Wait until the FIFO data completely. */
+	wait_for_idle(rs);
+
+	spin_lock_irqsave(&rs->lock, flags);
+
+	rs->state &= ~TXBUSY;
+	if (!(rs->state & RXBUSY)) {
+		spi_enable_chip(rs, 0);
+		spi_finalize_current_transfer(rs->master);
+	}
+
+	spin_unlock_irqrestore(&rs->lock, flags);
+}
+
+static int rockchip_spi_prepare_dma(struct rockchip_spi *rs)
+{
+	unsigned long flags;
+	struct dma_slave_config rxconf, txconf;
+	struct dma_async_tx_descriptor *rxdesc, *txdesc;
+
+	spin_lock_irqsave(&rs->lock, flags);
+	rs->state &= ~RXBUSY;
+	rs->state &= ~TXBUSY;
+	spin_unlock_irqrestore(&rs->lock, flags);
+
+	rxdesc = NULL;
+	if (rs->rx) {
+		rxconf.direction = rs->dma_rx.direction;
+		rxconf.src_addr = rs->dma_rx.addr;
+		rxconf.src_addr_width = rs->n_bytes;
+		if (rs->dma_caps.max_burst > 4)
+			rxconf.src_maxburst = 4;
+		else
+			rxconf.src_maxburst = 1;
+		dmaengine_slave_config(rs->dma_rx.ch, &rxconf);
+
+		rxdesc = dmaengine_prep_slave_sg(
+				rs->dma_rx.ch,
+				rs->rx_sg.sgl, rs->rx_sg.nents,
+				rs->dma_rx.direction, DMA_PREP_INTERRUPT);
+		if (!rxdesc)
+			return -EINVAL;
+
+		rxdesc->callback = rockchip_spi_dma_rxcb;
+		rxdesc->callback_param = rs;
+	}
+
+	txdesc = NULL;
+	if (rs->tx) {
+		txconf.direction = rs->dma_tx.direction;
+		txconf.dst_addr = rs->dma_tx.addr;
+		txconf.dst_addr_width = rs->n_bytes;
+		if (rs->dma_caps.max_burst > 4)
+			txconf.dst_maxburst = 4;
+		else
+			txconf.dst_maxburst = 1;
+		dmaengine_slave_config(rs->dma_tx.ch, &txconf);
+
+		txdesc = dmaengine_prep_slave_sg(
+				rs->dma_tx.ch,
+				rs->tx_sg.sgl, rs->tx_sg.nents,
+				rs->dma_tx.direction, DMA_PREP_INTERRUPT);
+		if (!txdesc) {
+			if (rxdesc)
+				dmaengine_terminate_sync(rs->dma_rx.ch);
+			return -EINVAL;
+		}
+
+		txdesc->callback = rockchip_spi_dma_txcb;
+		txdesc->callback_param = rs;
+	}
+
+	/* rx must be started before tx due to spi instinct */
+	if (rxdesc) {
+		spin_lock_irqsave(&rs->lock, flags);
+		rs->state |= RXBUSY;
+		spin_unlock_irqrestore(&rs->lock, flags);
+		dmaengine_submit(rxdesc);
+		dma_async_issue_pending(rs->dma_rx.ch);
+	}
+
+	if (txdesc) {
+		spin_lock_irqsave(&rs->lock, flags);
+		rs->state |= TXBUSY;
+		spin_unlock_irqrestore(&rs->lock, flags);
+		dmaengine_submit(txdesc);
+		dma_async_issue_pending(rs->dma_tx.ch);
+	}
+
+	return 0;
+}
+
+static void rockchip_spi_config(struct rockchip_spi *rs)
+{
+	u32 div = 0;
+	u32 dmacr = 0;
+	int rsd = 0;
+
+	u32 cr0 = (CR0_BHT_8BIT << CR0_BHT_OFFSET)
+		| (CR0_SSD_ONE << CR0_SSD_OFFSET)
+		| (CR0_EM_BIG << CR0_EM_OFFSET);
+
+	cr0 |= (rs->n_bytes << CR0_DFS_OFFSET);
+	cr0 |= ((rs->mode & 0x3) << CR0_SCPH_OFFSET);
+	cr0 |= (rs->tmode << CR0_XFM_OFFSET);
+	cr0 |= (rs->type << CR0_FRF_OFFSET);
+
+	if (rs->use_dma) {
+		if (rs->tx)
+			dmacr |= TF_DMA_EN;
+		if (rs->rx)
+			dmacr |= RF_DMA_EN;
+	}
+
+	if (WARN_ON(rs->speed > MAX_SCLK_OUT))
+		rs->speed = MAX_SCLK_OUT;
+
+	/* the minimum divisor is 2 */
+	if (rs->max_freq < 2 * rs->speed) {
+		clk_set_rate(rs->spiclk, 2 * rs->speed);
+		rs->max_freq = clk_get_rate(rs->spiclk);
+	}
+
+	/* div doesn't support odd number */
+	div = DIV_ROUND_UP(rs->max_freq, rs->speed);
+	div = (div + 1) & 0xfffe;
+
+	/* Rx sample delay is expressed in parent clock cycles (max 3) */
+	rsd = DIV_ROUND_CLOSEST(rs->rsd_nsecs * (rs->max_freq >> 8),
+				1000000000 >> 8);
+	if (!rsd && rs->rsd_nsecs) {
+		pr_warn_once("rockchip-spi: %u Hz are too slow to express %u ns delay\n",
+			     rs->max_freq, rs->rsd_nsecs);
+	} else if (rsd > 3) {
+		rsd = 3;
+		pr_warn_once("rockchip-spi: %u Hz are too fast to express %u ns delay, clamping at %u ns\n",
+			     rs->max_freq, rs->rsd_nsecs,
+			     rsd * 1000000000U / rs->max_freq);
+	}
+	cr0 |= rsd << CR0_RSD_OFFSET;
+
+	writel_relaxed(cr0, rs->regs + ROCKCHIP_SPI_CTRLR0);
+
+	if (rs->n_bytes == 1)
+		writel_relaxed(rs->len - 1, rs->regs + ROCKCHIP_SPI_CTRLR1);
+	else if (rs->n_bytes == 2)
+		writel_relaxed((rs->len / 2) - 1, rs->regs + ROCKCHIP_SPI_CTRLR1);
+	else
+		writel_relaxed((rs->len * 2) - 1, rs->regs + ROCKCHIP_SPI_CTRLR1);
+
+	writel_relaxed(rs->fifo_len / 2 - 1, rs->regs + ROCKCHIP_SPI_TXFTLR);
+	writel_relaxed(rs->fifo_len / 2 - 1, rs->regs + ROCKCHIP_SPI_RXFTLR);
+
+	writel_relaxed(0, rs->regs + ROCKCHIP_SPI_DMATDLR);
+	writel_relaxed(0, rs->regs + ROCKCHIP_SPI_DMARDLR);
+	writel_relaxed(dmacr, rs->regs + ROCKCHIP_SPI_DMACR);
+
+	spi_set_clk(rs, div);
+
+	dev_dbg(rs->dev, "cr0 0x%x, div %d\n", cr0, div);
+}
+
+static size_t rockchip_spi_max_transfer_size(struct spi_device *spi)
+{
+	return ROCKCHIP_SPI_MAX_TRANLEN;
+}
+
+static int rockchip_spi_transfer_one(
+		struct spi_master *master,
+		struct spi_device *spi,
+		struct spi_transfer *xfer)
+{
+	int ret = 0;
+	struct rockchip_spi *rs = spi_master_get_devdata(master);
+
+	WARN_ON(readl_relaxed(rs->regs + ROCKCHIP_SPI_SSIENR) &&
+		(readl_relaxed(rs->regs + ROCKCHIP_SPI_SR) & SR_BUSY));
+
+	if (!xfer->tx_buf && !xfer->rx_buf) {
+		dev_err(rs->dev, "No buffer for transfer\n");
+		return -EINVAL;
+	}
+
+	if (xfer->len > ROCKCHIP_SPI_MAX_TRANLEN) {
+		dev_err(rs->dev, "Transfer is too long (%d)\n", xfer->len);
+		return -EINVAL;
+	}
+
+	rs->speed = xfer->speed_hz;
+	rs->bpw = xfer->bits_per_word;
+	rs->n_bytes = rs->bpw >> 3;
+
+	rs->tx = xfer->tx_buf;
+	rs->tx_end = rs->tx + xfer->len;
+	rs->rx = xfer->rx_buf;
+	rs->rx_end = rs->rx + xfer->len;
+	rs->len = xfer->len;
+
+	rs->tx_sg = xfer->tx_sg;
+	rs->rx_sg = xfer->rx_sg;
+
+	if (rs->tx && rs->rx)
+		rs->tmode = CR0_XFM_TR;
+	else if (rs->tx)
+		rs->tmode = CR0_XFM_TO;
+	else if (rs->rx)
+		rs->tmode = CR0_XFM_RO;
+
+	/* we need prepare dma before spi was enabled */
+	if (master->can_dma && master->can_dma(master, spi, xfer))
+		rs->use_dma = 1;
+	else
+		rs->use_dma = 0;
+
+	rockchip_spi_config(rs);
+
+	if (rs->use_dma) {
+		if (rs->tmode == CR0_XFM_RO) {
+			/* rx: dma must be prepared first */
+			ret = rockchip_spi_prepare_dma(rs);
+			spi_enable_chip(rs, 1);
+		} else {
+			/* tx or tr: spi must be enabled first */
+			spi_enable_chip(rs, 1);
+			ret = rockchip_spi_prepare_dma(rs);
+		}
+		/* successful DMA prepare means the transfer is in progress */
+		ret = ret ? ret : 1;
+	} else {
+		spi_enable_chip(rs, 1);
+		ret = rockchip_spi_pio_transfer(rs);
+	}
+
+	return ret;
+}
+
+static bool rockchip_spi_can_dma(struct spi_master *master,
+				 struct spi_device *spi,
+				 struct spi_transfer *xfer)
+{
+	struct rockchip_spi *rs = spi_master_get_devdata(master);
+
+	return (xfer->len > rs->fifo_len);
+}
+
+static int rockchip_spi_probe(struct platform_device *pdev)
+{
+	int ret;
+	struct rockchip_spi *rs;
+	struct spi_master *master;
+	struct resource *mem;
+	u32 rsd_nsecs;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct rockchip_spi));
+	if (!master)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, master);
+
+	rs = spi_master_get_devdata(master);
+
+	/* Get basic io resource and map it */
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	rs->regs = devm_ioremap_resource(&pdev->dev, mem);
+	if (IS_ERR(rs->regs)) {
+		ret =  PTR_ERR(rs->regs);
+		goto err_put_master;
+	}
+
+	rs->apb_pclk = devm_clk_get(&pdev->dev, "apb_pclk");
+	if (IS_ERR(rs->apb_pclk)) {
+		dev_err(&pdev->dev, "Failed to get apb_pclk\n");
+		ret = PTR_ERR(rs->apb_pclk);
+		goto err_put_master;
+	}
+
+	rs->spiclk = devm_clk_get(&pdev->dev, "spiclk");
+	if (IS_ERR(rs->spiclk)) {
+		dev_err(&pdev->dev, "Failed to get spi_pclk\n");
+		ret = PTR_ERR(rs->spiclk);
+		goto err_put_master;
+	}
+
+	ret = clk_prepare_enable(rs->apb_pclk);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to enable apb_pclk\n");
+		goto err_put_master;
+	}
+
+	ret = clk_prepare_enable(rs->spiclk);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to enable spi_clk\n");
+		goto err_disable_apbclk;
+	}
+
+	spi_enable_chip(rs, 0);
+
+	rs->type = SSI_MOTO_SPI;
+	rs->master = master;
+	rs->dev = &pdev->dev;
+	rs->max_freq = clk_get_rate(rs->spiclk);
+
+	if (!of_property_read_u32(pdev->dev.of_node, "rx-sample-delay-ns",
+				  &rsd_nsecs))
+		rs->rsd_nsecs = rsd_nsecs;
+
+	rs->fifo_len = get_fifo_len(rs);
+	if (!rs->fifo_len) {
+		dev_err(&pdev->dev, "Failed to get fifo length\n");
+		ret = -EINVAL;
+		goto err_disable_spiclk;
+	}
+
+	spin_lock_init(&rs->lock);
+
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+
+	master->auto_runtime_pm = true;
+	master->bus_num = pdev->id;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_LOOP;
+	master->num_chipselect = ROCKCHIP_SPI_MAX_CS_NUM;
+	master->dev.of_node = pdev->dev.of_node;
+	master->bits_per_word_mask = SPI_BPW_MASK(16) | SPI_BPW_MASK(8);
+
+	master->set_cs = rockchip_spi_set_cs;
+	master->prepare_message = rockchip_spi_prepare_message;
+	master->unprepare_message = rockchip_spi_unprepare_message;
+	master->transfer_one = rockchip_spi_transfer_one;
+	master->max_transfer_size = rockchip_spi_max_transfer_size;
+	master->handle_err = rockchip_spi_handle_err;
+	master->flags = SPI_MASTER_GPIO_SS;
+
+	rs->dma_tx.ch = dma_request_chan(rs->dev, "tx");
+	if (IS_ERR(rs->dma_tx.ch)) {
+		/* Check tx to see if we need defer probing driver */
+		if (PTR_ERR(rs->dma_tx.ch) == -EPROBE_DEFER) {
+			ret = -EPROBE_DEFER;
+			goto err_disable_pm_runtime;
+		}
+		dev_warn(rs->dev, "Failed to request TX DMA channel\n");
+		rs->dma_tx.ch = NULL;
+	}
+
+	rs->dma_rx.ch = dma_request_chan(rs->dev, "rx");
+	if (IS_ERR(rs->dma_rx.ch)) {
+		if (PTR_ERR(rs->dma_rx.ch) == -EPROBE_DEFER) {
+			ret = -EPROBE_DEFER;
+			goto err_free_dma_tx;
+		}
+		dev_warn(rs->dev, "Failed to request RX DMA channel\n");
+		rs->dma_rx.ch = NULL;
+	}
+
+	if (rs->dma_tx.ch && rs->dma_rx.ch) {
+		dma_get_slave_caps(rs->dma_rx.ch, &(rs->dma_caps));
+		rs->dma_tx.addr = (dma_addr_t)(mem->start + ROCKCHIP_SPI_TXDR);
+		rs->dma_rx.addr = (dma_addr_t)(mem->start + ROCKCHIP_SPI_RXDR);
+		rs->dma_tx.direction = DMA_MEM_TO_DEV;
+		rs->dma_rx.direction = DMA_DEV_TO_MEM;
+
+		master->can_dma = rockchip_spi_can_dma;
+		master->dma_tx = rs->dma_tx.ch;
+		master->dma_rx = rs->dma_rx.ch;
+	}
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to register master\n");
+		goto err_free_dma_rx;
+	}
+
+	return 0;
+
+err_free_dma_rx:
+	if (rs->dma_rx.ch)
+		dma_release_channel(rs->dma_rx.ch);
+err_free_dma_tx:
+	if (rs->dma_tx.ch)
+		dma_release_channel(rs->dma_tx.ch);
+err_disable_pm_runtime:
+	pm_runtime_disable(&pdev->dev);
+err_disable_spiclk:
+	clk_disable_unprepare(rs->spiclk);
+err_disable_apbclk:
+	clk_disable_unprepare(rs->apb_pclk);
+err_put_master:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int rockchip_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = spi_master_get(platform_get_drvdata(pdev));
+	struct rockchip_spi *rs = spi_master_get_devdata(master);
+
+	pm_runtime_get_sync(&pdev->dev);
+
+	clk_disable_unprepare(rs->spiclk);
+	clk_disable_unprepare(rs->apb_pclk);
+
+	pm_runtime_put_noidle(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+	pm_runtime_set_suspended(&pdev->dev);
+
+	if (rs->dma_tx.ch)
+		dma_release_channel(rs->dma_tx.ch);
+	if (rs->dma_rx.ch)
+		dma_release_channel(rs->dma_rx.ch);
+
+	spi_master_put(master);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int rockchip_spi_suspend(struct device *dev)
+{
+	int ret;
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct rockchip_spi *rs = spi_master_get_devdata(master);
+
+	ret = spi_master_suspend(rs->master);
+	if (ret < 0)
+		return ret;
+
+	ret = pm_runtime_force_suspend(dev);
+	if (ret < 0)
+		return ret;
+
+	pinctrl_pm_select_sleep_state(dev);
+
+	return 0;
+}
+
+static int rockchip_spi_resume(struct device *dev)
+{
+	int ret;
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct rockchip_spi *rs = spi_master_get_devdata(master);
+
+	pinctrl_pm_select_default_state(dev);
+
+	ret = pm_runtime_force_resume(dev);
+	if (ret < 0)
+		return ret;
+
+	ret = spi_master_resume(rs->master);
+	if (ret < 0) {
+		clk_disable_unprepare(rs->spiclk);
+		clk_disable_unprepare(rs->apb_pclk);
+	}
+
+	return 0;
+}
+#endif /* CONFIG_PM_SLEEP */
+
+#ifdef CONFIG_PM
+static int rockchip_spi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct rockchip_spi *rs = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(rs->spiclk);
+	clk_disable_unprepare(rs->apb_pclk);
+
+	return 0;
+}
+
+static int rockchip_spi_runtime_resume(struct device *dev)
+{
+	int ret;
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct rockchip_spi *rs = spi_master_get_devdata(master);
+
+	ret = clk_prepare_enable(rs->apb_pclk);
+	if (ret < 0)
+		return ret;
+
+	ret = clk_prepare_enable(rs->spiclk);
+	if (ret < 0)
+		clk_disable_unprepare(rs->apb_pclk);
+
+	return 0;
+}
+#endif /* CONFIG_PM */
+
+static const struct dev_pm_ops rockchip_spi_pm = {
+	SET_SYSTEM_SLEEP_PM_OPS(rockchip_spi_suspend, rockchip_spi_resume)
+	SET_RUNTIME_PM_OPS(rockchip_spi_runtime_suspend,
+			   rockchip_spi_runtime_resume, NULL)
+};
+
+static const struct of_device_id rockchip_spi_dt_match[] = {
+	{ .compatible = "rockchip,rv1108-spi", },
+	{ .compatible = "rockchip,rk3036-spi", },
+	{ .compatible = "rockchip,rk3066-spi", },
+	{ .compatible = "rockchip,rk3188-spi", },
+	{ .compatible = "rockchip,rk3228-spi", },
+	{ .compatible = "rockchip,rk3288-spi", },
+	{ .compatible = "rockchip,rk3368-spi", },
+	{ .compatible = "rockchip,rk3399-spi", },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, rockchip_spi_dt_match);
+
+static struct platform_driver rockchip_spi_driver = {
+	.driver = {
+		.name	= DRIVER_NAME,
+		.pm = &rockchip_spi_pm,
+		.of_match_table = of_match_ptr(rockchip_spi_dt_match),
+	},
+	.probe = rockchip_spi_probe,
+	.remove = rockchip_spi_remove,
+};
+
+module_platform_driver(rockchip_spi_driver);
+
+MODULE_AUTHOR("Addy Ke <addy.ke@rock-chips.com>");
+MODULE_DESCRIPTION("ROCKCHIP SPI Controller Driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-rspi.c b/drivers/spi/spi-rspi.c
new file mode 100644
index 0000000..b37de1d
--- /dev/null
+++ b/drivers/spi/spi-rspi.c
@@ -0,0 +1,1393 @@
+/*
+ * SH RSPI driver
+ *
+ * Copyright (C) 2012, 2013  Renesas Solutions Corp.
+ * Copyright (C) 2014 Glider bvba
+ *
+ * Based on spi-sh.c:
+ * Copyright (C) 2011 Renesas Solutions Corp.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/io.h>
+#include <linux/clk.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/of_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/sh_dma.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/rspi.h>
+
+#define RSPI_SPCR		0x00	/* Control Register */
+#define RSPI_SSLP		0x01	/* Slave Select Polarity Register */
+#define RSPI_SPPCR		0x02	/* Pin Control Register */
+#define RSPI_SPSR		0x03	/* Status Register */
+#define RSPI_SPDR		0x04	/* Data Register */
+#define RSPI_SPSCR		0x08	/* Sequence Control Register */
+#define RSPI_SPSSR		0x09	/* Sequence Status Register */
+#define RSPI_SPBR		0x0a	/* Bit Rate Register */
+#define RSPI_SPDCR		0x0b	/* Data Control Register */
+#define RSPI_SPCKD		0x0c	/* Clock Delay Register */
+#define RSPI_SSLND		0x0d	/* Slave Select Negation Delay Register */
+#define RSPI_SPND		0x0e	/* Next-Access Delay Register */
+#define RSPI_SPCR2		0x0f	/* Control Register 2 (SH only) */
+#define RSPI_SPCMD0		0x10	/* Command Register 0 */
+#define RSPI_SPCMD1		0x12	/* Command Register 1 */
+#define RSPI_SPCMD2		0x14	/* Command Register 2 */
+#define RSPI_SPCMD3		0x16	/* Command Register 3 */
+#define RSPI_SPCMD4		0x18	/* Command Register 4 */
+#define RSPI_SPCMD5		0x1a	/* Command Register 5 */
+#define RSPI_SPCMD6		0x1c	/* Command Register 6 */
+#define RSPI_SPCMD7		0x1e	/* Command Register 7 */
+#define RSPI_SPCMD(i)		(RSPI_SPCMD0 + (i) * 2)
+#define RSPI_NUM_SPCMD		8
+#define RSPI_RZ_NUM_SPCMD	4
+#define QSPI_NUM_SPCMD		4
+
+/* RSPI on RZ only */
+#define RSPI_SPBFCR		0x20	/* Buffer Control Register */
+#define RSPI_SPBFDR		0x22	/* Buffer Data Count Setting Register */
+
+/* QSPI only */
+#define QSPI_SPBFCR		0x18	/* Buffer Control Register */
+#define QSPI_SPBDCR		0x1a	/* Buffer Data Count Register */
+#define QSPI_SPBMUL0		0x1c	/* Transfer Data Length Multiplier Setting Register 0 */
+#define QSPI_SPBMUL1		0x20	/* Transfer Data Length Multiplier Setting Register 1 */
+#define QSPI_SPBMUL2		0x24	/* Transfer Data Length Multiplier Setting Register 2 */
+#define QSPI_SPBMUL3		0x28	/* Transfer Data Length Multiplier Setting Register 3 */
+#define QSPI_SPBMUL(i)		(QSPI_SPBMUL0 + (i) * 4)
+
+/* SPCR - Control Register */
+#define SPCR_SPRIE		0x80	/* Receive Interrupt Enable */
+#define SPCR_SPE		0x40	/* Function Enable */
+#define SPCR_SPTIE		0x20	/* Transmit Interrupt Enable */
+#define SPCR_SPEIE		0x10	/* Error Interrupt Enable */
+#define SPCR_MSTR		0x08	/* Master/Slave Mode Select */
+#define SPCR_MODFEN		0x04	/* Mode Fault Error Detection Enable */
+/* RSPI on SH only */
+#define SPCR_TXMD		0x02	/* TX Only Mode (vs. Full Duplex) */
+#define SPCR_SPMS		0x01	/* 3-wire Mode (vs. 4-wire) */
+/* QSPI on R-Car Gen2 only */
+#define SPCR_WSWAP		0x02	/* Word Swap of read-data for DMAC */
+#define SPCR_BSWAP		0x01	/* Byte Swap of read-data for DMAC */
+
+/* SSLP - Slave Select Polarity Register */
+#define SSLP_SSL1P		0x02	/* SSL1 Signal Polarity Setting */
+#define SSLP_SSL0P		0x01	/* SSL0 Signal Polarity Setting */
+
+/* SPPCR - Pin Control Register */
+#define SPPCR_MOIFE		0x20	/* MOSI Idle Value Fixing Enable */
+#define SPPCR_MOIFV		0x10	/* MOSI Idle Fixed Value */
+#define SPPCR_SPOM		0x04
+#define SPPCR_SPLP2		0x02	/* Loopback Mode 2 (non-inverting) */
+#define SPPCR_SPLP		0x01	/* Loopback Mode (inverting) */
+
+#define SPPCR_IO3FV		0x04	/* Single-/Dual-SPI Mode IO3 Output Fixed Value */
+#define SPPCR_IO2FV		0x04	/* Single-/Dual-SPI Mode IO2 Output Fixed Value */
+
+/* SPSR - Status Register */
+#define SPSR_SPRF		0x80	/* Receive Buffer Full Flag */
+#define SPSR_TEND		0x40	/* Transmit End */
+#define SPSR_SPTEF		0x20	/* Transmit Buffer Empty Flag */
+#define SPSR_PERF		0x08	/* Parity Error Flag */
+#define SPSR_MODF		0x04	/* Mode Fault Error Flag */
+#define SPSR_IDLNF		0x02	/* RSPI Idle Flag */
+#define SPSR_OVRF		0x01	/* Overrun Error Flag (RSPI only) */
+
+/* SPSCR - Sequence Control Register */
+#define SPSCR_SPSLN_MASK	0x07	/* Sequence Length Specification */
+
+/* SPSSR - Sequence Status Register */
+#define SPSSR_SPECM_MASK	0x70	/* Command Error Mask */
+#define SPSSR_SPCP_MASK		0x07	/* Command Pointer Mask */
+
+/* SPDCR - Data Control Register */
+#define SPDCR_TXDMY		0x80	/* Dummy Data Transmission Enable */
+#define SPDCR_SPLW1		0x40	/* Access Width Specification (RZ) */
+#define SPDCR_SPLW0		0x20	/* Access Width Specification (RZ) */
+#define SPDCR_SPLLWORD		(SPDCR_SPLW1 | SPDCR_SPLW0)
+#define SPDCR_SPLWORD		SPDCR_SPLW1
+#define SPDCR_SPLBYTE		SPDCR_SPLW0
+#define SPDCR_SPLW		0x20	/* Access Width Specification (SH) */
+#define SPDCR_SPRDTD		0x10	/* Receive Transmit Data Select (SH) */
+#define SPDCR_SLSEL1		0x08
+#define SPDCR_SLSEL0		0x04
+#define SPDCR_SLSEL_MASK	0x0c	/* SSL1 Output Select (SH) */
+#define SPDCR_SPFC1		0x02
+#define SPDCR_SPFC0		0x01
+#define SPDCR_SPFC_MASK		0x03	/* Frame Count Setting (1-4) (SH) */
+
+/* SPCKD - Clock Delay Register */
+#define SPCKD_SCKDL_MASK	0x07	/* Clock Delay Setting (1-8) */
+
+/* SSLND - Slave Select Negation Delay Register */
+#define SSLND_SLNDL_MASK	0x07	/* SSL Negation Delay Setting (1-8) */
+
+/* SPND - Next-Access Delay Register */
+#define SPND_SPNDL_MASK		0x07	/* Next-Access Delay Setting (1-8) */
+
+/* SPCR2 - Control Register 2 */
+#define SPCR2_PTE		0x08	/* Parity Self-Test Enable */
+#define SPCR2_SPIE		0x04	/* Idle Interrupt Enable */
+#define SPCR2_SPOE		0x02	/* Odd Parity Enable (vs. Even) */
+#define SPCR2_SPPE		0x01	/* Parity Enable */
+
+/* SPCMDn - Command Registers */
+#define SPCMD_SCKDEN		0x8000	/* Clock Delay Setting Enable */
+#define SPCMD_SLNDEN		0x4000	/* SSL Negation Delay Setting Enable */
+#define SPCMD_SPNDEN		0x2000	/* Next-Access Delay Enable */
+#define SPCMD_LSBF		0x1000	/* LSB First */
+#define SPCMD_SPB_MASK		0x0f00	/* Data Length Setting */
+#define SPCMD_SPB_8_TO_16(bit)	(((bit - 1) << 8) & SPCMD_SPB_MASK)
+#define SPCMD_SPB_8BIT		0x0000	/* QSPI only */
+#define SPCMD_SPB_16BIT		0x0100
+#define SPCMD_SPB_20BIT		0x0000
+#define SPCMD_SPB_24BIT		0x0100
+#define SPCMD_SPB_32BIT		0x0200
+#define SPCMD_SSLKP		0x0080	/* SSL Signal Level Keeping */
+#define SPCMD_SPIMOD_MASK	0x0060	/* SPI Operating Mode (QSPI only) */
+#define SPCMD_SPIMOD1		0x0040
+#define SPCMD_SPIMOD0		0x0020
+#define SPCMD_SPIMOD_SINGLE	0
+#define SPCMD_SPIMOD_DUAL	SPCMD_SPIMOD0
+#define SPCMD_SPIMOD_QUAD	SPCMD_SPIMOD1
+#define SPCMD_SPRW		0x0010	/* SPI Read/Write Access (Dual/Quad) */
+#define SPCMD_SSLA_MASK		0x0030	/* SSL Assert Signal Setting (RSPI) */
+#define SPCMD_BRDV_MASK		0x000c	/* Bit Rate Division Setting */
+#define SPCMD_CPOL		0x0002	/* Clock Polarity Setting */
+#define SPCMD_CPHA		0x0001	/* Clock Phase Setting */
+
+/* SPBFCR - Buffer Control Register */
+#define SPBFCR_TXRST		0x80	/* Transmit Buffer Data Reset */
+#define SPBFCR_RXRST		0x40	/* Receive Buffer Data Reset */
+#define SPBFCR_TXTRG_MASK	0x30	/* Transmit Buffer Data Triggering Number */
+#define SPBFCR_RXTRG_MASK	0x07	/* Receive Buffer Data Triggering Number */
+/* QSPI on R-Car Gen2 */
+#define SPBFCR_TXTRG_1B		0x00	/* 31 bytes (1 byte available) */
+#define SPBFCR_TXTRG_32B	0x30	/* 0 byte (32 bytes available) */
+#define SPBFCR_RXTRG_1B		0x00	/* 1 byte (31 bytes available) */
+#define SPBFCR_RXTRG_32B	0x07	/* 32 bytes (0 byte available) */
+
+#define QSPI_BUFFER_SIZE        32u
+
+struct rspi_data {
+	void __iomem *addr;
+	u32 max_speed_hz;
+	struct spi_master *master;
+	wait_queue_head_t wait;
+	struct clk *clk;
+	u16 spcmd;
+	u8 spsr;
+	u8 sppcr;
+	int rx_irq, tx_irq;
+	const struct spi_ops *ops;
+
+	unsigned dma_callbacked:1;
+	unsigned byte_access:1;
+};
+
+static void rspi_write8(const struct rspi_data *rspi, u8 data, u16 offset)
+{
+	iowrite8(data, rspi->addr + offset);
+}
+
+static void rspi_write16(const struct rspi_data *rspi, u16 data, u16 offset)
+{
+	iowrite16(data, rspi->addr + offset);
+}
+
+static void rspi_write32(const struct rspi_data *rspi, u32 data, u16 offset)
+{
+	iowrite32(data, rspi->addr + offset);
+}
+
+static u8 rspi_read8(const struct rspi_data *rspi, u16 offset)
+{
+	return ioread8(rspi->addr + offset);
+}
+
+static u16 rspi_read16(const struct rspi_data *rspi, u16 offset)
+{
+	return ioread16(rspi->addr + offset);
+}
+
+static void rspi_write_data(const struct rspi_data *rspi, u16 data)
+{
+	if (rspi->byte_access)
+		rspi_write8(rspi, data, RSPI_SPDR);
+	else /* 16 bit */
+		rspi_write16(rspi, data, RSPI_SPDR);
+}
+
+static u16 rspi_read_data(const struct rspi_data *rspi)
+{
+	if (rspi->byte_access)
+		return rspi_read8(rspi, RSPI_SPDR);
+	else /* 16 bit */
+		return rspi_read16(rspi, RSPI_SPDR);
+}
+
+/* optional functions */
+struct spi_ops {
+	int (*set_config_register)(struct rspi_data *rspi, int access_size);
+	int (*transfer_one)(struct spi_master *master, struct spi_device *spi,
+			    struct spi_transfer *xfer);
+	u16 mode_bits;
+	u16 flags;
+	u16 fifo_size;
+};
+
+/*
+ * functions for RSPI on legacy SH
+ */
+static int rspi_set_config_register(struct rspi_data *rspi, int access_size)
+{
+	int spbr;
+
+	/* Sets output mode, MOSI signal, and (optionally) loopback */
+	rspi_write8(rspi, rspi->sppcr, RSPI_SPPCR);
+
+	/* Sets transfer bit rate */
+	spbr = DIV_ROUND_UP(clk_get_rate(rspi->clk),
+			    2 * rspi->max_speed_hz) - 1;
+	rspi_write8(rspi, clamp(spbr, 0, 255), RSPI_SPBR);
+
+	/* Disable dummy transmission, set 16-bit word access, 1 frame */
+	rspi_write8(rspi, 0, RSPI_SPDCR);
+	rspi->byte_access = 0;
+
+	/* Sets RSPCK, SSL, next-access delay value */
+	rspi_write8(rspi, 0x00, RSPI_SPCKD);
+	rspi_write8(rspi, 0x00, RSPI_SSLND);
+	rspi_write8(rspi, 0x00, RSPI_SPND);
+
+	/* Sets parity, interrupt mask */
+	rspi_write8(rspi, 0x00, RSPI_SPCR2);
+
+	/* Sets SPCMD */
+	rspi->spcmd |= SPCMD_SPB_8_TO_16(access_size);
+	rspi_write16(rspi, rspi->spcmd, RSPI_SPCMD0);
+
+	/* Sets RSPI mode */
+	rspi_write8(rspi, SPCR_MSTR, RSPI_SPCR);
+
+	return 0;
+}
+
+/*
+ * functions for RSPI on RZ
+ */
+static int rspi_rz_set_config_register(struct rspi_data *rspi, int access_size)
+{
+	int spbr;
+	int div = 0;
+	unsigned long clksrc;
+
+	/* Sets output mode, MOSI signal, and (optionally) loopback */
+	rspi_write8(rspi, rspi->sppcr, RSPI_SPPCR);
+
+	clksrc = clk_get_rate(rspi->clk);
+	while (div < 3) {
+		if (rspi->max_speed_hz >= clksrc/4) /* 4=(CLK/2)/2 */
+			break;
+		div++;
+		clksrc /= 2;
+	}
+
+	/* Sets transfer bit rate */
+	spbr = DIV_ROUND_UP(clksrc, 2 * rspi->max_speed_hz) - 1;
+	rspi_write8(rspi, clamp(spbr, 0, 255), RSPI_SPBR);
+	rspi->spcmd |= div << 2;
+
+	/* Disable dummy transmission, set byte access */
+	rspi_write8(rspi, SPDCR_SPLBYTE, RSPI_SPDCR);
+	rspi->byte_access = 1;
+
+	/* Sets RSPCK, SSL, next-access delay value */
+	rspi_write8(rspi, 0x00, RSPI_SPCKD);
+	rspi_write8(rspi, 0x00, RSPI_SSLND);
+	rspi_write8(rspi, 0x00, RSPI_SPND);
+
+	/* Sets SPCMD */
+	rspi->spcmd |= SPCMD_SPB_8_TO_16(access_size);
+	rspi_write16(rspi, rspi->spcmd, RSPI_SPCMD0);
+
+	/* Sets RSPI mode */
+	rspi_write8(rspi, SPCR_MSTR, RSPI_SPCR);
+
+	return 0;
+}
+
+/*
+ * functions for QSPI
+ */
+static int qspi_set_config_register(struct rspi_data *rspi, int access_size)
+{
+	int spbr;
+
+	/* Sets output mode, MOSI signal, and (optionally) loopback */
+	rspi_write8(rspi, rspi->sppcr, RSPI_SPPCR);
+
+	/* Sets transfer bit rate */
+	spbr = DIV_ROUND_UP(clk_get_rate(rspi->clk), 2 * rspi->max_speed_hz);
+	rspi_write8(rspi, clamp(spbr, 0, 255), RSPI_SPBR);
+
+	/* Disable dummy transmission, set byte access */
+	rspi_write8(rspi, 0, RSPI_SPDCR);
+	rspi->byte_access = 1;
+
+	/* Sets RSPCK, SSL, next-access delay value */
+	rspi_write8(rspi, 0x00, RSPI_SPCKD);
+	rspi_write8(rspi, 0x00, RSPI_SSLND);
+	rspi_write8(rspi, 0x00, RSPI_SPND);
+
+	/* Data Length Setting */
+	if (access_size == 8)
+		rspi->spcmd |= SPCMD_SPB_8BIT;
+	else if (access_size == 16)
+		rspi->spcmd |= SPCMD_SPB_16BIT;
+	else
+		rspi->spcmd |= SPCMD_SPB_32BIT;
+
+	rspi->spcmd |= SPCMD_SCKDEN | SPCMD_SLNDEN | SPCMD_SPNDEN;
+
+	/* Resets transfer data length */
+	rspi_write32(rspi, 0, QSPI_SPBMUL0);
+
+	/* Resets transmit and receive buffer */
+	rspi_write8(rspi, SPBFCR_TXRST | SPBFCR_RXRST, QSPI_SPBFCR);
+	/* Sets buffer to allow normal operation */
+	rspi_write8(rspi, 0x00, QSPI_SPBFCR);
+
+	/* Sets SPCMD */
+	rspi_write16(rspi, rspi->spcmd, RSPI_SPCMD0);
+
+	/* Sets RSPI mode */
+	rspi_write8(rspi, SPCR_MSTR, RSPI_SPCR);
+
+	return 0;
+}
+
+static void qspi_update(const struct rspi_data *rspi, u8 mask, u8 val, u8 reg)
+{
+	u8 data;
+
+	data = rspi_read8(rspi, reg);
+	data &= ~mask;
+	data |= (val & mask);
+	rspi_write8(rspi, data, reg);
+}
+
+static unsigned int qspi_set_send_trigger(struct rspi_data *rspi,
+					  unsigned int len)
+{
+	unsigned int n;
+
+	n = min(len, QSPI_BUFFER_SIZE);
+
+	if (len >= QSPI_BUFFER_SIZE) {
+		/* sets triggering number to 32 bytes */
+		qspi_update(rspi, SPBFCR_TXTRG_MASK,
+			     SPBFCR_TXTRG_32B, QSPI_SPBFCR);
+	} else {
+		/* sets triggering number to 1 byte */
+		qspi_update(rspi, SPBFCR_TXTRG_MASK,
+			     SPBFCR_TXTRG_1B, QSPI_SPBFCR);
+	}
+
+	return n;
+}
+
+static int qspi_set_receive_trigger(struct rspi_data *rspi, unsigned int len)
+{
+	unsigned int n;
+
+	n = min(len, QSPI_BUFFER_SIZE);
+
+	if (len >= QSPI_BUFFER_SIZE) {
+		/* sets triggering number to 32 bytes */
+		qspi_update(rspi, SPBFCR_RXTRG_MASK,
+			     SPBFCR_RXTRG_32B, QSPI_SPBFCR);
+	} else {
+		/* sets triggering number to 1 byte */
+		qspi_update(rspi, SPBFCR_RXTRG_MASK,
+			     SPBFCR_RXTRG_1B, QSPI_SPBFCR);
+	}
+	return n;
+}
+
+#define set_config_register(spi, n) spi->ops->set_config_register(spi, n)
+
+static void rspi_enable_irq(const struct rspi_data *rspi, u8 enable)
+{
+	rspi_write8(rspi, rspi_read8(rspi, RSPI_SPCR) | enable, RSPI_SPCR);
+}
+
+static void rspi_disable_irq(const struct rspi_data *rspi, u8 disable)
+{
+	rspi_write8(rspi, rspi_read8(rspi, RSPI_SPCR) & ~disable, RSPI_SPCR);
+}
+
+static int rspi_wait_for_interrupt(struct rspi_data *rspi, u8 wait_mask,
+				   u8 enable_bit)
+{
+	int ret;
+
+	rspi->spsr = rspi_read8(rspi, RSPI_SPSR);
+	if (rspi->spsr & wait_mask)
+		return 0;
+
+	rspi_enable_irq(rspi, enable_bit);
+	ret = wait_event_timeout(rspi->wait, rspi->spsr & wait_mask, HZ);
+	if (ret == 0 && !(rspi->spsr & wait_mask))
+		return -ETIMEDOUT;
+
+	return 0;
+}
+
+static inline int rspi_wait_for_tx_empty(struct rspi_data *rspi)
+{
+	return rspi_wait_for_interrupt(rspi, SPSR_SPTEF, SPCR_SPTIE);
+}
+
+static inline int rspi_wait_for_rx_full(struct rspi_data *rspi)
+{
+	return rspi_wait_for_interrupt(rspi, SPSR_SPRF, SPCR_SPRIE);
+}
+
+static int rspi_data_out(struct rspi_data *rspi, u8 data)
+{
+	int error = rspi_wait_for_tx_empty(rspi);
+	if (error < 0) {
+		dev_err(&rspi->master->dev, "transmit timeout\n");
+		return error;
+	}
+	rspi_write_data(rspi, data);
+	return 0;
+}
+
+static int rspi_data_in(struct rspi_data *rspi)
+{
+	int error;
+	u8 data;
+
+	error = rspi_wait_for_rx_full(rspi);
+	if (error < 0) {
+		dev_err(&rspi->master->dev, "receive timeout\n");
+		return error;
+	}
+	data = rspi_read_data(rspi);
+	return data;
+}
+
+static int rspi_pio_transfer(struct rspi_data *rspi, const u8 *tx, u8 *rx,
+			     unsigned int n)
+{
+	while (n-- > 0) {
+		if (tx) {
+			int ret = rspi_data_out(rspi, *tx++);
+			if (ret < 0)
+				return ret;
+		}
+		if (rx) {
+			int ret = rspi_data_in(rspi);
+			if (ret < 0)
+				return ret;
+			*rx++ = ret;
+		}
+	}
+
+	return 0;
+}
+
+static void rspi_dma_complete(void *arg)
+{
+	struct rspi_data *rspi = arg;
+
+	rspi->dma_callbacked = 1;
+	wake_up_interruptible(&rspi->wait);
+}
+
+static int rspi_dma_transfer(struct rspi_data *rspi, struct sg_table *tx,
+			     struct sg_table *rx)
+{
+	struct dma_async_tx_descriptor *desc_tx = NULL, *desc_rx = NULL;
+	u8 irq_mask = 0;
+	unsigned int other_irq = 0;
+	dma_cookie_t cookie;
+	int ret;
+
+	/* First prepare and submit the DMA request(s), as this may fail */
+	if (rx) {
+		desc_rx = dmaengine_prep_slave_sg(rspi->master->dma_rx,
+					rx->sgl, rx->nents, DMA_DEV_TO_MEM,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!desc_rx) {
+			ret = -EAGAIN;
+			goto no_dma_rx;
+		}
+
+		desc_rx->callback = rspi_dma_complete;
+		desc_rx->callback_param = rspi;
+		cookie = dmaengine_submit(desc_rx);
+		if (dma_submit_error(cookie)) {
+			ret = cookie;
+			goto no_dma_rx;
+		}
+
+		irq_mask |= SPCR_SPRIE;
+	}
+
+	if (tx) {
+		desc_tx = dmaengine_prep_slave_sg(rspi->master->dma_tx,
+					tx->sgl, tx->nents, DMA_MEM_TO_DEV,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!desc_tx) {
+			ret = -EAGAIN;
+			goto no_dma_tx;
+		}
+
+		if (rx) {
+			/* No callback */
+			desc_tx->callback = NULL;
+		} else {
+			desc_tx->callback = rspi_dma_complete;
+			desc_tx->callback_param = rspi;
+		}
+		cookie = dmaengine_submit(desc_tx);
+		if (dma_submit_error(cookie)) {
+			ret = cookie;
+			goto no_dma_tx;
+		}
+
+		irq_mask |= SPCR_SPTIE;
+	}
+
+	/*
+	 * DMAC needs SPxIE, but if SPxIE is set, the IRQ routine will be
+	 * called. So, this driver disables the IRQ while DMA transfer.
+	 */
+	if (tx)
+		disable_irq(other_irq = rspi->tx_irq);
+	if (rx && rspi->rx_irq != other_irq)
+		disable_irq(rspi->rx_irq);
+
+	rspi_enable_irq(rspi, irq_mask);
+	rspi->dma_callbacked = 0;
+
+	/* Now start DMA */
+	if (rx)
+		dma_async_issue_pending(rspi->master->dma_rx);
+	if (tx)
+		dma_async_issue_pending(rspi->master->dma_tx);
+
+	ret = wait_event_interruptible_timeout(rspi->wait,
+					       rspi->dma_callbacked, HZ);
+	if (ret > 0 && rspi->dma_callbacked) {
+		ret = 0;
+	} else {
+		if (!ret) {
+			dev_err(&rspi->master->dev, "DMA timeout\n");
+			ret = -ETIMEDOUT;
+		}
+		if (tx)
+			dmaengine_terminate_all(rspi->master->dma_tx);
+		if (rx)
+			dmaengine_terminate_all(rspi->master->dma_rx);
+	}
+
+	rspi_disable_irq(rspi, irq_mask);
+
+	if (tx)
+		enable_irq(rspi->tx_irq);
+	if (rx && rspi->rx_irq != other_irq)
+		enable_irq(rspi->rx_irq);
+
+	return ret;
+
+no_dma_tx:
+	if (rx)
+		dmaengine_terminate_all(rspi->master->dma_rx);
+no_dma_rx:
+	if (ret == -EAGAIN) {
+		pr_warn_once("%s %s: DMA not available, falling back to PIO\n",
+			     dev_driver_string(&rspi->master->dev),
+			     dev_name(&rspi->master->dev));
+	}
+	return ret;
+}
+
+static void rspi_receive_init(const struct rspi_data *rspi)
+{
+	u8 spsr;
+
+	spsr = rspi_read8(rspi, RSPI_SPSR);
+	if (spsr & SPSR_SPRF)
+		rspi_read_data(rspi);	/* dummy read */
+	if (spsr & SPSR_OVRF)
+		rspi_write8(rspi, rspi_read8(rspi, RSPI_SPSR) & ~SPSR_OVRF,
+			    RSPI_SPSR);
+}
+
+static void rspi_rz_receive_init(const struct rspi_data *rspi)
+{
+	rspi_receive_init(rspi);
+	rspi_write8(rspi, SPBFCR_TXRST | SPBFCR_RXRST, RSPI_SPBFCR);
+	rspi_write8(rspi, 0, RSPI_SPBFCR);
+}
+
+static void qspi_receive_init(const struct rspi_data *rspi)
+{
+	u8 spsr;
+
+	spsr = rspi_read8(rspi, RSPI_SPSR);
+	if (spsr & SPSR_SPRF)
+		rspi_read_data(rspi);   /* dummy read */
+	rspi_write8(rspi, SPBFCR_TXRST | SPBFCR_RXRST, QSPI_SPBFCR);
+	rspi_write8(rspi, 0, QSPI_SPBFCR);
+}
+
+static bool __rspi_can_dma(const struct rspi_data *rspi,
+			   const struct spi_transfer *xfer)
+{
+	return xfer->len > rspi->ops->fifo_size;
+}
+
+static bool rspi_can_dma(struct spi_master *master, struct spi_device *spi,
+			 struct spi_transfer *xfer)
+{
+	struct rspi_data *rspi = spi_master_get_devdata(master);
+
+	return __rspi_can_dma(rspi, xfer);
+}
+
+static int rspi_dma_check_then_transfer(struct rspi_data *rspi,
+					 struct spi_transfer *xfer)
+{
+	if (!rspi->master->can_dma || !__rspi_can_dma(rspi, xfer))
+		return -EAGAIN;
+
+	/* rx_buf can be NULL on RSPI on SH in TX-only Mode */
+	return rspi_dma_transfer(rspi, &xfer->tx_sg,
+				xfer->rx_buf ? &xfer->rx_sg : NULL);
+}
+
+static int rspi_common_transfer(struct rspi_data *rspi,
+				struct spi_transfer *xfer)
+{
+	int ret;
+
+	ret = rspi_dma_check_then_transfer(rspi, xfer);
+	if (ret != -EAGAIN)
+		return ret;
+
+	ret = rspi_pio_transfer(rspi, xfer->tx_buf, xfer->rx_buf, xfer->len);
+	if (ret < 0)
+		return ret;
+
+	/* Wait for the last transmission */
+	rspi_wait_for_tx_empty(rspi);
+
+	return 0;
+}
+
+static int rspi_transfer_one(struct spi_master *master, struct spi_device *spi,
+			     struct spi_transfer *xfer)
+{
+	struct rspi_data *rspi = spi_master_get_devdata(master);
+	u8 spcr;
+
+	spcr = rspi_read8(rspi, RSPI_SPCR);
+	if (xfer->rx_buf) {
+		rspi_receive_init(rspi);
+		spcr &= ~SPCR_TXMD;
+	} else {
+		spcr |= SPCR_TXMD;
+	}
+	rspi_write8(rspi, spcr, RSPI_SPCR);
+
+	return rspi_common_transfer(rspi, xfer);
+}
+
+static int rspi_rz_transfer_one(struct spi_master *master,
+				struct spi_device *spi,
+				struct spi_transfer *xfer)
+{
+	struct rspi_data *rspi = spi_master_get_devdata(master);
+
+	rspi_rz_receive_init(rspi);
+
+	return rspi_common_transfer(rspi, xfer);
+}
+
+static int qspi_trigger_transfer_out_in(struct rspi_data *rspi, const u8 *tx,
+					u8 *rx, unsigned int len)
+{
+	unsigned int i, n;
+	int ret;
+
+	while (len > 0) {
+		n = qspi_set_send_trigger(rspi, len);
+		qspi_set_receive_trigger(rspi, len);
+		if (n == QSPI_BUFFER_SIZE) {
+			ret = rspi_wait_for_tx_empty(rspi);
+			if (ret < 0) {
+				dev_err(&rspi->master->dev, "transmit timeout\n");
+				return ret;
+			}
+			for (i = 0; i < n; i++)
+				rspi_write_data(rspi, *tx++);
+
+			ret = rspi_wait_for_rx_full(rspi);
+			if (ret < 0) {
+				dev_err(&rspi->master->dev, "receive timeout\n");
+				return ret;
+			}
+			for (i = 0; i < n; i++)
+				*rx++ = rspi_read_data(rspi);
+		} else {
+			ret = rspi_pio_transfer(rspi, tx, rx, n);
+			if (ret < 0)
+				return ret;
+		}
+		len -= n;
+	}
+
+	return 0;
+}
+
+static int qspi_transfer_out_in(struct rspi_data *rspi,
+				struct spi_transfer *xfer)
+{
+	int ret;
+
+	qspi_receive_init(rspi);
+
+	ret = rspi_dma_check_then_transfer(rspi, xfer);
+	if (ret != -EAGAIN)
+		return ret;
+
+	return qspi_trigger_transfer_out_in(rspi, xfer->tx_buf,
+					    xfer->rx_buf, xfer->len);
+}
+
+static int qspi_transfer_out(struct rspi_data *rspi, struct spi_transfer *xfer)
+{
+	const u8 *tx = xfer->tx_buf;
+	unsigned int n = xfer->len;
+	unsigned int i, len;
+	int ret;
+
+	if (rspi->master->can_dma && __rspi_can_dma(rspi, xfer)) {
+		ret = rspi_dma_transfer(rspi, &xfer->tx_sg, NULL);
+		if (ret != -EAGAIN)
+			return ret;
+	}
+
+	while (n > 0) {
+		len = qspi_set_send_trigger(rspi, n);
+		if (len == QSPI_BUFFER_SIZE) {
+			ret = rspi_wait_for_tx_empty(rspi);
+			if (ret < 0) {
+				dev_err(&rspi->master->dev, "transmit timeout\n");
+				return ret;
+			}
+			for (i = 0; i < len; i++)
+				rspi_write_data(rspi, *tx++);
+		} else {
+			ret = rspi_pio_transfer(rspi, tx, NULL, len);
+			if (ret < 0)
+				return ret;
+		}
+		n -= len;
+	}
+
+	/* Wait for the last transmission */
+	rspi_wait_for_tx_empty(rspi);
+
+	return 0;
+}
+
+static int qspi_transfer_in(struct rspi_data *rspi, struct spi_transfer *xfer)
+{
+	u8 *rx = xfer->rx_buf;
+	unsigned int n = xfer->len;
+	unsigned int i, len;
+	int ret;
+
+	if (rspi->master->can_dma && __rspi_can_dma(rspi, xfer)) {
+		int ret = rspi_dma_transfer(rspi, NULL, &xfer->rx_sg);
+		if (ret != -EAGAIN)
+			return ret;
+	}
+
+	while (n > 0) {
+		len = qspi_set_receive_trigger(rspi, n);
+		if (len == QSPI_BUFFER_SIZE) {
+			ret = rspi_wait_for_rx_full(rspi);
+			if (ret < 0) {
+				dev_err(&rspi->master->dev, "receive timeout\n");
+				return ret;
+			}
+			for (i = 0; i < len; i++)
+				*rx++ = rspi_read_data(rspi);
+		} else {
+			ret = rspi_pio_transfer(rspi, NULL, rx, len);
+			if (ret < 0)
+				return ret;
+		}
+		n -= len;
+	}
+
+	return 0;
+}
+
+static int qspi_transfer_one(struct spi_master *master, struct spi_device *spi,
+			     struct spi_transfer *xfer)
+{
+	struct rspi_data *rspi = spi_master_get_devdata(master);
+
+	if (spi->mode & SPI_LOOP) {
+		return qspi_transfer_out_in(rspi, xfer);
+	} else if (xfer->tx_nbits > SPI_NBITS_SINGLE) {
+		/* Quad or Dual SPI Write */
+		return qspi_transfer_out(rspi, xfer);
+	} else if (xfer->rx_nbits > SPI_NBITS_SINGLE) {
+		/* Quad or Dual SPI Read */
+		return qspi_transfer_in(rspi, xfer);
+	} else {
+		/* Single SPI Transfer */
+		return qspi_transfer_out_in(rspi, xfer);
+	}
+}
+
+static int rspi_setup(struct spi_device *spi)
+{
+	struct rspi_data *rspi = spi_master_get_devdata(spi->master);
+
+	rspi->max_speed_hz = spi->max_speed_hz;
+
+	rspi->spcmd = SPCMD_SSLKP;
+	if (spi->mode & SPI_CPOL)
+		rspi->spcmd |= SPCMD_CPOL;
+	if (spi->mode & SPI_CPHA)
+		rspi->spcmd |= SPCMD_CPHA;
+
+	/* CMOS output mode and MOSI signal from previous transfer */
+	rspi->sppcr = 0;
+	if (spi->mode & SPI_LOOP)
+		rspi->sppcr |= SPPCR_SPLP;
+
+	set_config_register(rspi, 8);
+
+	return 0;
+}
+
+static u16 qspi_transfer_mode(const struct spi_transfer *xfer)
+{
+	if (xfer->tx_buf)
+		switch (xfer->tx_nbits) {
+		case SPI_NBITS_QUAD:
+			return SPCMD_SPIMOD_QUAD;
+		case SPI_NBITS_DUAL:
+			return SPCMD_SPIMOD_DUAL;
+		default:
+			return 0;
+		}
+	if (xfer->rx_buf)
+		switch (xfer->rx_nbits) {
+		case SPI_NBITS_QUAD:
+			return SPCMD_SPIMOD_QUAD | SPCMD_SPRW;
+		case SPI_NBITS_DUAL:
+			return SPCMD_SPIMOD_DUAL | SPCMD_SPRW;
+		default:
+			return 0;
+		}
+
+	return 0;
+}
+
+static int qspi_setup_sequencer(struct rspi_data *rspi,
+				const struct spi_message *msg)
+{
+	const struct spi_transfer *xfer;
+	unsigned int i = 0, len = 0;
+	u16 current_mode = 0xffff, mode;
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		mode = qspi_transfer_mode(xfer);
+		if (mode == current_mode) {
+			len += xfer->len;
+			continue;
+		}
+
+		/* Transfer mode change */
+		if (i) {
+			/* Set transfer data length of previous transfer */
+			rspi_write32(rspi, len, QSPI_SPBMUL(i - 1));
+		}
+
+		if (i >= QSPI_NUM_SPCMD) {
+			dev_err(&msg->spi->dev,
+				"Too many different transfer modes");
+			return -EINVAL;
+		}
+
+		/* Program transfer mode for this transfer */
+		rspi_write16(rspi, rspi->spcmd | mode, RSPI_SPCMD(i));
+		current_mode = mode;
+		len = xfer->len;
+		i++;
+	}
+	if (i) {
+		/* Set final transfer data length and sequence length */
+		rspi_write32(rspi, len, QSPI_SPBMUL(i - 1));
+		rspi_write8(rspi, i - 1, RSPI_SPSCR);
+	}
+
+	return 0;
+}
+
+static int rspi_prepare_message(struct spi_master *master,
+				struct spi_message *msg)
+{
+	struct rspi_data *rspi = spi_master_get_devdata(master);
+	int ret;
+
+	if (msg->spi->mode &
+	    (SPI_TX_DUAL | SPI_TX_QUAD | SPI_RX_DUAL | SPI_RX_QUAD)) {
+		/* Setup sequencer for messages with multiple transfer modes */
+		ret = qspi_setup_sequencer(rspi, msg);
+		if (ret < 0)
+			return ret;
+	}
+
+	/* Enable SPI function in master mode */
+	rspi_write8(rspi, rspi_read8(rspi, RSPI_SPCR) | SPCR_SPE, RSPI_SPCR);
+	return 0;
+}
+
+static int rspi_unprepare_message(struct spi_master *master,
+				  struct spi_message *msg)
+{
+	struct rspi_data *rspi = spi_master_get_devdata(master);
+
+	/* Disable SPI function */
+	rspi_write8(rspi, rspi_read8(rspi, RSPI_SPCR) & ~SPCR_SPE, RSPI_SPCR);
+
+	/* Reset sequencer for Single SPI Transfers */
+	rspi_write16(rspi, rspi->spcmd, RSPI_SPCMD0);
+	rspi_write8(rspi, 0, RSPI_SPSCR);
+	return 0;
+}
+
+static irqreturn_t rspi_irq_mux(int irq, void *_sr)
+{
+	struct rspi_data *rspi = _sr;
+	u8 spsr;
+	irqreturn_t ret = IRQ_NONE;
+	u8 disable_irq = 0;
+
+	rspi->spsr = spsr = rspi_read8(rspi, RSPI_SPSR);
+	if (spsr & SPSR_SPRF)
+		disable_irq |= SPCR_SPRIE;
+	if (spsr & SPSR_SPTEF)
+		disable_irq |= SPCR_SPTIE;
+
+	if (disable_irq) {
+		ret = IRQ_HANDLED;
+		rspi_disable_irq(rspi, disable_irq);
+		wake_up(&rspi->wait);
+	}
+
+	return ret;
+}
+
+static irqreturn_t rspi_irq_rx(int irq, void *_sr)
+{
+	struct rspi_data *rspi = _sr;
+	u8 spsr;
+
+	rspi->spsr = spsr = rspi_read8(rspi, RSPI_SPSR);
+	if (spsr & SPSR_SPRF) {
+		rspi_disable_irq(rspi, SPCR_SPRIE);
+		wake_up(&rspi->wait);
+		return IRQ_HANDLED;
+	}
+
+	return 0;
+}
+
+static irqreturn_t rspi_irq_tx(int irq, void *_sr)
+{
+	struct rspi_data *rspi = _sr;
+	u8 spsr;
+
+	rspi->spsr = spsr = rspi_read8(rspi, RSPI_SPSR);
+	if (spsr & SPSR_SPTEF) {
+		rspi_disable_irq(rspi, SPCR_SPTIE);
+		wake_up(&rspi->wait);
+		return IRQ_HANDLED;
+	}
+
+	return 0;
+}
+
+static struct dma_chan *rspi_request_dma_chan(struct device *dev,
+					      enum dma_transfer_direction dir,
+					      unsigned int id,
+					      dma_addr_t port_addr)
+{
+	dma_cap_mask_t mask;
+	struct dma_chan *chan;
+	struct dma_slave_config cfg;
+	int ret;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	chan = dma_request_slave_channel_compat(mask, shdma_chan_filter,
+				(void *)(unsigned long)id, dev,
+				dir == DMA_MEM_TO_DEV ? "tx" : "rx");
+	if (!chan) {
+		dev_warn(dev, "dma_request_slave_channel_compat failed\n");
+		return NULL;
+	}
+
+	memset(&cfg, 0, sizeof(cfg));
+	cfg.direction = dir;
+	if (dir == DMA_MEM_TO_DEV) {
+		cfg.dst_addr = port_addr;
+		cfg.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	} else {
+		cfg.src_addr = port_addr;
+		cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	}
+
+	ret = dmaengine_slave_config(chan, &cfg);
+	if (ret) {
+		dev_warn(dev, "dmaengine_slave_config failed %d\n", ret);
+		dma_release_channel(chan);
+		return NULL;
+	}
+
+	return chan;
+}
+
+static int rspi_request_dma(struct device *dev, struct spi_master *master,
+			    const struct resource *res)
+{
+	const struct rspi_plat_data *rspi_pd = dev_get_platdata(dev);
+	unsigned int dma_tx_id, dma_rx_id;
+
+	if (dev->of_node) {
+		/* In the OF case we will get the slave IDs from the DT */
+		dma_tx_id = 0;
+		dma_rx_id = 0;
+	} else if (rspi_pd && rspi_pd->dma_tx_id && rspi_pd->dma_rx_id) {
+		dma_tx_id = rspi_pd->dma_tx_id;
+		dma_rx_id = rspi_pd->dma_rx_id;
+	} else {
+		/* The driver assumes no error. */
+		return 0;
+	}
+
+	master->dma_tx = rspi_request_dma_chan(dev, DMA_MEM_TO_DEV, dma_tx_id,
+					       res->start + RSPI_SPDR);
+	if (!master->dma_tx)
+		return -ENODEV;
+
+	master->dma_rx = rspi_request_dma_chan(dev, DMA_DEV_TO_MEM, dma_rx_id,
+					       res->start + RSPI_SPDR);
+	if (!master->dma_rx) {
+		dma_release_channel(master->dma_tx);
+		master->dma_tx = NULL;
+		return -ENODEV;
+	}
+
+	master->can_dma = rspi_can_dma;
+	dev_info(dev, "DMA available");
+	return 0;
+}
+
+static void rspi_release_dma(struct spi_master *master)
+{
+	if (master->dma_tx)
+		dma_release_channel(master->dma_tx);
+	if (master->dma_rx)
+		dma_release_channel(master->dma_rx);
+}
+
+static int rspi_remove(struct platform_device *pdev)
+{
+	struct rspi_data *rspi = platform_get_drvdata(pdev);
+
+	rspi_release_dma(rspi->master);
+	pm_runtime_disable(&pdev->dev);
+
+	return 0;
+}
+
+static const struct spi_ops rspi_ops = {
+	.set_config_register =	rspi_set_config_register,
+	.transfer_one =		rspi_transfer_one,
+	.mode_bits =		SPI_CPHA | SPI_CPOL | SPI_LOOP,
+	.flags =		SPI_MASTER_MUST_TX,
+	.fifo_size =		8,
+};
+
+static const struct spi_ops rspi_rz_ops = {
+	.set_config_register =	rspi_rz_set_config_register,
+	.transfer_one =		rspi_rz_transfer_one,
+	.mode_bits =		SPI_CPHA | SPI_CPOL | SPI_LOOP,
+	.flags =		SPI_MASTER_MUST_RX | SPI_MASTER_MUST_TX,
+	.fifo_size =		8,	/* 8 for TX, 32 for RX */
+};
+
+static const struct spi_ops qspi_ops = {
+	.set_config_register =	qspi_set_config_register,
+	.transfer_one =		qspi_transfer_one,
+	.mode_bits =		SPI_CPHA | SPI_CPOL | SPI_LOOP |
+				SPI_TX_DUAL | SPI_TX_QUAD |
+				SPI_RX_DUAL | SPI_RX_QUAD,
+	.flags =		SPI_MASTER_MUST_RX | SPI_MASTER_MUST_TX,
+	.fifo_size =		32,
+};
+
+#ifdef CONFIG_OF
+static const struct of_device_id rspi_of_match[] = {
+	/* RSPI on legacy SH */
+	{ .compatible = "renesas,rspi", .data = &rspi_ops },
+	/* RSPI on RZ/A1H */
+	{ .compatible = "renesas,rspi-rz", .data = &rspi_rz_ops },
+	/* QSPI on R-Car Gen2 */
+	{ .compatible = "renesas,qspi", .data = &qspi_ops },
+	{ /* sentinel */ }
+};
+
+MODULE_DEVICE_TABLE(of, rspi_of_match);
+
+static int rspi_parse_dt(struct device *dev, struct spi_master *master)
+{
+	u32 num_cs;
+	int error;
+
+	/* Parse DT properties */
+	error = of_property_read_u32(dev->of_node, "num-cs", &num_cs);
+	if (error) {
+		dev_err(dev, "of_property_read_u32 num-cs failed %d\n", error);
+		return error;
+	}
+
+	master->num_chipselect = num_cs;
+	return 0;
+}
+#else
+#define rspi_of_match	NULL
+static inline int rspi_parse_dt(struct device *dev, struct spi_master *master)
+{
+	return -EINVAL;
+}
+#endif /* CONFIG_OF */
+
+static int rspi_request_irq(struct device *dev, unsigned int irq,
+			    irq_handler_t handler, const char *suffix,
+			    void *dev_id)
+{
+	const char *name = devm_kasprintf(dev, GFP_KERNEL, "%s:%s",
+					  dev_name(dev), suffix);
+	if (!name)
+		return -ENOMEM;
+
+	return devm_request_irq(dev, irq, handler, 0, name, dev_id);
+}
+
+static int rspi_probe(struct platform_device *pdev)
+{
+	struct resource *res;
+	struct spi_master *master;
+	struct rspi_data *rspi;
+	int ret;
+	const struct rspi_plat_data *rspi_pd;
+	const struct spi_ops *ops;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct rspi_data));
+	if (master == NULL)
+		return -ENOMEM;
+
+	ops = of_device_get_match_data(&pdev->dev);
+	if (ops) {
+		ret = rspi_parse_dt(&pdev->dev, master);
+		if (ret)
+			goto error1;
+	} else {
+		ops = (struct spi_ops *)pdev->id_entry->driver_data;
+		rspi_pd = dev_get_platdata(&pdev->dev);
+		if (rspi_pd && rspi_pd->num_chipselect)
+			master->num_chipselect = rspi_pd->num_chipselect;
+		else
+			master->num_chipselect = 2; /* default */
+	}
+
+	/* ops parameter check */
+	if (!ops->set_config_register) {
+		dev_err(&pdev->dev, "there is no set_config_register\n");
+		ret = -ENODEV;
+		goto error1;
+	}
+
+	rspi = spi_master_get_devdata(master);
+	platform_set_drvdata(pdev, rspi);
+	rspi->ops = ops;
+	rspi->master = master;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	rspi->addr = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(rspi->addr)) {
+		ret = PTR_ERR(rspi->addr);
+		goto error1;
+	}
+
+	rspi->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(rspi->clk)) {
+		dev_err(&pdev->dev, "cannot get clock\n");
+		ret = PTR_ERR(rspi->clk);
+		goto error1;
+	}
+
+	pm_runtime_enable(&pdev->dev);
+
+	init_waitqueue_head(&rspi->wait);
+
+	master->bus_num = pdev->id;
+	master->setup = rspi_setup;
+	master->auto_runtime_pm = true;
+	master->transfer_one = ops->transfer_one;
+	master->prepare_message = rspi_prepare_message;
+	master->unprepare_message = rspi_unprepare_message;
+	master->mode_bits = ops->mode_bits;
+	master->flags = ops->flags;
+	master->dev.of_node = pdev->dev.of_node;
+
+	ret = platform_get_irq_byname(pdev, "rx");
+	if (ret < 0) {
+		ret = platform_get_irq_byname(pdev, "mux");
+		if (ret < 0)
+			ret = platform_get_irq(pdev, 0);
+		if (ret >= 0)
+			rspi->rx_irq = rspi->tx_irq = ret;
+	} else {
+		rspi->rx_irq = ret;
+		ret = platform_get_irq_byname(pdev, "tx");
+		if (ret >= 0)
+			rspi->tx_irq = ret;
+	}
+	if (ret < 0) {
+		dev_err(&pdev->dev, "platform_get_irq error\n");
+		goto error2;
+	}
+
+	if (rspi->rx_irq == rspi->tx_irq) {
+		/* Single multiplexed interrupt */
+		ret = rspi_request_irq(&pdev->dev, rspi->rx_irq, rspi_irq_mux,
+				       "mux", rspi);
+	} else {
+		/* Multi-interrupt mode, only SPRI and SPTI are used */
+		ret = rspi_request_irq(&pdev->dev, rspi->rx_irq, rspi_irq_rx,
+				       "rx", rspi);
+		if (!ret)
+			ret = rspi_request_irq(&pdev->dev, rspi->tx_irq,
+					       rspi_irq_tx, "tx", rspi);
+	}
+	if (ret < 0) {
+		dev_err(&pdev->dev, "request_irq error\n");
+		goto error2;
+	}
+
+	ret = rspi_request_dma(&pdev->dev, master, res);
+	if (ret < 0)
+		dev_warn(&pdev->dev, "DMA not available, using PIO\n");
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "spi_register_master error.\n");
+		goto error3;
+	}
+
+	dev_info(&pdev->dev, "probed\n");
+
+	return 0;
+
+error3:
+	rspi_release_dma(master);
+error2:
+	pm_runtime_disable(&pdev->dev);
+error1:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static const struct platform_device_id spi_driver_ids[] = {
+	{ "rspi",	(kernel_ulong_t)&rspi_ops },
+	{ "rspi-rz",	(kernel_ulong_t)&rspi_rz_ops },
+	{ "qspi",	(kernel_ulong_t)&qspi_ops },
+	{},
+};
+
+MODULE_DEVICE_TABLE(platform, spi_driver_ids);
+
+#ifdef CONFIG_PM_SLEEP
+static int rspi_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct rspi_data *rspi = platform_get_drvdata(pdev);
+
+	return spi_master_suspend(rspi->master);
+}
+
+static int rspi_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct rspi_data *rspi = platform_get_drvdata(pdev);
+
+	return spi_master_resume(rspi->master);
+}
+
+static SIMPLE_DEV_PM_OPS(rspi_pm_ops, rspi_suspend, rspi_resume);
+#define DEV_PM_OPS	&rspi_pm_ops
+#else
+#define DEV_PM_OPS	NULL
+#endif /* CONFIG_PM_SLEEP */
+
+static struct platform_driver rspi_driver = {
+	.probe =	rspi_probe,
+	.remove =	rspi_remove,
+	.id_table =	spi_driver_ids,
+	.driver		= {
+		.name = "renesas_spi",
+		.pm = DEV_PM_OPS,
+		.of_match_table = of_match_ptr(rspi_of_match),
+	},
+};
+module_platform_driver(rspi_driver);
+
+MODULE_DESCRIPTION("Renesas RSPI bus driver");
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Yoshihiro Shimoda");
+MODULE_ALIAS("platform:rspi");
diff --git a/drivers/spi/spi-s3c24xx-fiq.S b/drivers/spi/spi-s3c24xx-fiq.S
new file mode 100644
index 0000000..059f2dc
--- /dev/null
+++ b/drivers/spi/spi-s3c24xx-fiq.S
@@ -0,0 +1,116 @@
+/* linux/drivers/spi/spi_s3c24xx_fiq.S
+ *
+ * Copyright 2009 Simtec Electronics
+ *	Ben Dooks <ben@simtec.co.uk>
+ *
+ * S3C24XX SPI - FIQ pseudo-DMA transfer code
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+*/
+
+#include <linux/linkage.h>
+#include <asm/assembler.h>
+
+#include <mach/map.h>
+#include <mach/regs-irq.h>
+#include <plat/regs-spi.h>
+
+#include "spi-s3c24xx-fiq.h"
+
+	.text
+
+	@ entry to these routines is as follows, with the register names
+	@ defined in fiq.h so that they can be shared with the C files which
+	@ setup the calling registers.
+	@
+	@ fiq_rirq	The base of the IRQ registers to find S3C2410_SRCPND
+	@ fiq_rtmp	Temporary register to hold tx/rx data
+	@ fiq_rspi	The base of the SPI register block
+	@ fiq_rtx	The tx buffer pointer
+	@ fiq_rrx	The rx buffer pointer
+	@ fiq_rcount	The number of bytes to move
+
+	@ each entry starts with a word entry of how long it is
+	@ and an offset to the irq acknowledgment word
+
+ENTRY(s3c24xx_spi_fiq_rx)
+s3c24xx_spi_fix_rx:
+	.word	fiq_rx_end - fiq_rx_start
+	.word	fiq_rx_irq_ack - fiq_rx_start
+fiq_rx_start:
+	ldr	fiq_rtmp, fiq_rx_irq_ack
+	str	fiq_rtmp, [ fiq_rirq, # S3C2410_SRCPND - S3C24XX_VA_IRQ ]
+
+	ldrb	fiq_rtmp, [ fiq_rspi, #  S3C2410_SPRDAT ]
+	strb	fiq_rtmp, [ fiq_rrx ], #1
+
+	mov	fiq_rtmp, #0xff
+	strb	fiq_rtmp, [ fiq_rspi, # S3C2410_SPTDAT ]
+
+	subs	fiq_rcount, fiq_rcount, #1
+	subnes	pc, lr, #4		@@ return, still have work to do
+
+	@@ set IRQ controller so that next op will trigger IRQ
+	mov	fiq_rtmp, #0
+	str	fiq_rtmp, [ fiq_rirq, # S3C2410_INTMOD  - S3C24XX_VA_IRQ ]
+	subs	pc, lr, #4
+
+fiq_rx_irq_ack:
+	.word	0
+fiq_rx_end:
+
+ENTRY(s3c24xx_spi_fiq_txrx)
+s3c24xx_spi_fiq_txrx:
+	.word	fiq_txrx_end - fiq_txrx_start
+	.word	fiq_txrx_irq_ack - fiq_txrx_start
+fiq_txrx_start:
+
+	ldrb	fiq_rtmp, [ fiq_rspi, #  S3C2410_SPRDAT ]
+	strb	fiq_rtmp, [ fiq_rrx ], #1
+
+	ldr	fiq_rtmp, fiq_txrx_irq_ack
+	str	fiq_rtmp, [ fiq_rirq, # S3C2410_SRCPND - S3C24XX_VA_IRQ ]
+
+	ldrb	fiq_rtmp, [ fiq_rtx ], #1
+	strb	fiq_rtmp, [ fiq_rspi, # S3C2410_SPTDAT ]
+
+	subs	fiq_rcount, fiq_rcount, #1
+	subnes	pc, lr, #4		@@ return, still have work to do
+
+	mov	fiq_rtmp, #0
+	str	fiq_rtmp, [ fiq_rirq, # S3C2410_INTMOD  - S3C24XX_VA_IRQ ]
+	subs	pc, lr, #4
+
+fiq_txrx_irq_ack:
+	.word	0
+
+fiq_txrx_end:
+
+ENTRY(s3c24xx_spi_fiq_tx)
+s3c24xx_spi_fix_tx:
+	.word	fiq_tx_end - fiq_tx_start
+	.word	fiq_tx_irq_ack - fiq_tx_start
+fiq_tx_start:
+	ldrb	fiq_rtmp, [ fiq_rspi, #  S3C2410_SPRDAT ]
+
+	ldr	fiq_rtmp, fiq_tx_irq_ack
+	str	fiq_rtmp, [ fiq_rirq, # S3C2410_SRCPND - S3C24XX_VA_IRQ ]
+
+	ldrb	fiq_rtmp, [ fiq_rtx ], #1
+	strb	fiq_rtmp, [ fiq_rspi, # S3C2410_SPTDAT ]
+
+	subs	fiq_rcount, fiq_rcount, #1
+	subnes	pc, lr, #4		@@ return, still have work to do
+
+	mov	fiq_rtmp, #0
+	str	fiq_rtmp, [ fiq_rirq, # S3C2410_INTMOD  - S3C24XX_VA_IRQ ]
+	subs	pc, lr, #4
+
+fiq_tx_irq_ack:
+	.word	0
+
+fiq_tx_end:
+
+	.end
diff --git a/drivers/spi/spi-s3c24xx-fiq.h b/drivers/spi/spi-s3c24xx-fiq.h
new file mode 100644
index 0000000..a5950bb
--- /dev/null
+++ b/drivers/spi/spi-s3c24xx-fiq.h
@@ -0,0 +1,26 @@
+/* linux/drivers/spi/spi_s3c24xx_fiq.h
+ *
+ * Copyright 2009 Simtec Electronics
+ *	Ben Dooks <ben@simtec.co.uk>
+ *
+ * S3C24XX SPI - FIQ pseudo-DMA transfer support
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+*/
+
+/* We have R8 through R13 to play with */
+
+#ifdef __ASSEMBLY__
+#define __REG_NR(x)     r##x
+#else
+#define __REG_NR(x)     (x)
+#endif
+
+#define fiq_rspi	__REG_NR(8)
+#define fiq_rtmp	__REG_NR(9)
+#define fiq_rrx		__REG_NR(10)
+#define fiq_rtx		__REG_NR(11)
+#define fiq_rcount	__REG_NR(12)
+#define fiq_rirq	__REG_NR(13)
diff --git a/drivers/spi/spi-s3c24xx.c b/drivers/spi/spi-s3c24xx.c
new file mode 100644
index 0000000..4e7d1bf
--- /dev/null
+++ b/drivers/spi/spi-s3c24xx.c
@@ -0,0 +1,672 @@
+/*
+ * Copyright (c) 2006 Ben Dooks
+ * Copyright 2006-2009 Simtec Electronics
+ *	Ben Dooks <ben@simtec.co.uk>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+*/
+
+#include <linux/spinlock.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/errno.h>
+#include <linux/err.h>
+#include <linux/clk.h>
+#include <linux/platform_device.h>
+#include <linux/gpio.h>
+#include <linux/io.h>
+#include <linux/slab.h>
+
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/spi/s3c24xx.h>
+#include <linux/module.h>
+
+#include <plat/regs-spi.h>
+
+#include <asm/fiq.h>
+
+#include "spi-s3c24xx-fiq.h"
+
+/**
+ * s3c24xx_spi_devstate - per device data
+ * @hz: Last frequency calculated for @sppre field.
+ * @mode: Last mode setting for the @spcon field.
+ * @spcon: Value to write to the SPCON register.
+ * @sppre: Value to write to the SPPRE register.
+ */
+struct s3c24xx_spi_devstate {
+	unsigned int	hz;
+	unsigned int	mode;
+	u8		spcon;
+	u8		sppre;
+};
+
+enum spi_fiq_mode {
+	FIQ_MODE_NONE	= 0,
+	FIQ_MODE_TX	= 1,
+	FIQ_MODE_RX	= 2,
+	FIQ_MODE_TXRX	= 3,
+};
+
+struct s3c24xx_spi {
+	/* bitbang has to be first */
+	struct spi_bitbang	 bitbang;
+	struct completion	 done;
+
+	void __iomem		*regs;
+	int			 irq;
+	int			 len;
+	int			 count;
+
+	struct fiq_handler	 fiq_handler;
+	enum spi_fiq_mode	 fiq_mode;
+	unsigned char		 fiq_inuse;
+	unsigned char		 fiq_claimed;
+
+	void			(*set_cs)(struct s3c2410_spi_info *spi,
+					  int cs, int pol);
+
+	/* data buffers */
+	const unsigned char	*tx;
+	unsigned char		*rx;
+
+	struct clk		*clk;
+	struct spi_master	*master;
+	struct spi_device	*curdev;
+	struct device		*dev;
+	struct s3c2410_spi_info *pdata;
+};
+
+#define SPCON_DEFAULT (S3C2410_SPCON_MSTR | S3C2410_SPCON_SMOD_INT)
+#define SPPIN_DEFAULT (S3C2410_SPPIN_KEEP)
+
+static inline struct s3c24xx_spi *to_hw(struct spi_device *sdev)
+{
+	return spi_master_get_devdata(sdev->master);
+}
+
+static void s3c24xx_spi_gpiocs(struct s3c2410_spi_info *spi, int cs, int pol)
+{
+	gpio_set_value(spi->pin_cs, pol);
+}
+
+static void s3c24xx_spi_chipsel(struct spi_device *spi, int value)
+{
+	struct s3c24xx_spi_devstate *cs = spi->controller_state;
+	struct s3c24xx_spi *hw = to_hw(spi);
+	unsigned int cspol = spi->mode & SPI_CS_HIGH ? 1 : 0;
+
+	/* change the chipselect state and the state of the spi engine clock */
+
+	switch (value) {
+	case BITBANG_CS_INACTIVE:
+		hw->set_cs(hw->pdata, spi->chip_select, cspol^1);
+		writeb(cs->spcon, hw->regs + S3C2410_SPCON);
+		break;
+
+	case BITBANG_CS_ACTIVE:
+		writeb(cs->spcon | S3C2410_SPCON_ENSCK,
+		       hw->regs + S3C2410_SPCON);
+		hw->set_cs(hw->pdata, spi->chip_select, cspol);
+		break;
+	}
+}
+
+static int s3c24xx_spi_update_state(struct spi_device *spi,
+				    struct spi_transfer *t)
+{
+	struct s3c24xx_spi *hw = to_hw(spi);
+	struct s3c24xx_spi_devstate *cs = spi->controller_state;
+	unsigned int hz;
+	unsigned int div;
+	unsigned long clk;
+
+	hz  = t ? t->speed_hz : spi->max_speed_hz;
+
+	if (!hz)
+		hz = spi->max_speed_hz;
+
+	if (spi->mode != cs->mode) {
+		u8 spcon = SPCON_DEFAULT | S3C2410_SPCON_ENSCK;
+
+		if (spi->mode & SPI_CPHA)
+			spcon |= S3C2410_SPCON_CPHA_FMTB;
+
+		if (spi->mode & SPI_CPOL)
+			spcon |= S3C2410_SPCON_CPOL_HIGH;
+
+		cs->mode = spi->mode;
+		cs->spcon = spcon;
+	}
+
+	if (cs->hz != hz) {
+		clk = clk_get_rate(hw->clk);
+		div = DIV_ROUND_UP(clk, hz * 2) - 1;
+
+		if (div > 255)
+			div = 255;
+
+		dev_dbg(&spi->dev, "pre-scaler=%d (wanted %d, got %ld)\n",
+			div, hz, clk / (2 * (div + 1)));
+
+		cs->hz = hz;
+		cs->sppre = div;
+	}
+
+	return 0;
+}
+
+static int s3c24xx_spi_setupxfer(struct spi_device *spi,
+				 struct spi_transfer *t)
+{
+	struct s3c24xx_spi_devstate *cs = spi->controller_state;
+	struct s3c24xx_spi *hw = to_hw(spi);
+	int ret;
+
+	ret = s3c24xx_spi_update_state(spi, t);
+	if (!ret)
+		writeb(cs->sppre, hw->regs + S3C2410_SPPRE);
+
+	return ret;
+}
+
+static int s3c24xx_spi_setup(struct spi_device *spi)
+{
+	struct s3c24xx_spi_devstate *cs = spi->controller_state;
+	struct s3c24xx_spi *hw = to_hw(spi);
+	int ret;
+
+	/* allocate settings on the first call */
+	if (!cs) {
+		cs = devm_kzalloc(&spi->dev,
+				  sizeof(struct s3c24xx_spi_devstate),
+				  GFP_KERNEL);
+		if (!cs)
+			return -ENOMEM;
+
+		cs->spcon = SPCON_DEFAULT;
+		cs->hz = -1;
+		spi->controller_state = cs;
+	}
+
+	/* initialise the state from the device */
+	ret = s3c24xx_spi_update_state(spi, NULL);
+	if (ret)
+		return ret;
+
+	mutex_lock(&hw->bitbang.lock);
+	if (!hw->bitbang.busy) {
+		hw->bitbang.chipselect(spi, BITBANG_CS_INACTIVE);
+		/* need to ndelay for 0.5 clocktick ? */
+	}
+	mutex_unlock(&hw->bitbang.lock);
+
+	return 0;
+}
+
+static inline unsigned int hw_txbyte(struct s3c24xx_spi *hw, int count)
+{
+	return hw->tx ? hw->tx[count] : 0;
+}
+
+#ifdef CONFIG_SPI_S3C24XX_FIQ
+/* Support for FIQ based pseudo-DMA to improve the transfer speed.
+ *
+ * This code uses the assembly helper in spi_s3c24xx_spi.S which is
+ * used by the FIQ core to move data between main memory and the peripheral
+ * block. Since this is code running on the processor, there is no problem
+ * with cache coherency of the buffers, so we can use any buffer we like.
+ */
+
+/**
+ * struct spi_fiq_code - FIQ code and header
+ * @length: The length of the code fragment, excluding this header.
+ * @ack_offset: The offset from @data to the word to place the IRQ ACK bit at.
+ * @data: The code itself to install as a FIQ handler.
+ */
+struct spi_fiq_code {
+	u32	length;
+	u32	ack_offset;
+	u8	data[0];
+};
+
+extern struct spi_fiq_code s3c24xx_spi_fiq_txrx;
+extern struct spi_fiq_code s3c24xx_spi_fiq_tx;
+extern struct spi_fiq_code s3c24xx_spi_fiq_rx;
+
+/**
+ * ack_bit - turn IRQ into IRQ acknowledgement bit
+ * @irq: The interrupt number
+ *
+ * Returns the bit to write to the interrupt acknowledge register.
+ */
+static inline u32 ack_bit(unsigned int irq)
+{
+	return 1 << (irq - IRQ_EINT0);
+}
+
+/**
+ * s3c24xx_spi_tryfiq - attempt to claim and setup FIQ for transfer
+ * @hw: The hardware state.
+ *
+ * Claim the FIQ handler (only one can be active at any one time) and
+ * then setup the correct transfer code for this transfer.
+ *
+ * This call updates all the necessary state information if successful,
+ * so the caller does not need to do anything more than start the transfer
+ * as normal, since the IRQ will have been re-routed to the FIQ handler.
+*/
+static void s3c24xx_spi_tryfiq(struct s3c24xx_spi *hw)
+{
+	struct pt_regs regs;
+	enum spi_fiq_mode mode;
+	struct spi_fiq_code *code;
+	int ret;
+
+	if (!hw->fiq_claimed) {
+		/* try and claim fiq if we haven't got it, and if not
+		 * then return and simply use another transfer method */
+
+		ret = claim_fiq(&hw->fiq_handler);
+		if (ret)
+			return;
+	}
+
+	if (hw->tx && !hw->rx)
+		mode = FIQ_MODE_TX;
+	else if (hw->rx && !hw->tx)
+		mode = FIQ_MODE_RX;
+	else
+		mode = FIQ_MODE_TXRX;
+
+	regs.uregs[fiq_rspi] = (long)hw->regs;
+	regs.uregs[fiq_rrx]  = (long)hw->rx;
+	regs.uregs[fiq_rtx]  = (long)hw->tx + 1;
+	regs.uregs[fiq_rcount] = hw->len - 1;
+	regs.uregs[fiq_rirq] = (long)S3C24XX_VA_IRQ;
+
+	set_fiq_regs(&regs);
+
+	if (hw->fiq_mode != mode) {
+		u32 *ack_ptr;
+
+		hw->fiq_mode = mode;
+
+		switch (mode) {
+		case FIQ_MODE_TX:
+			code = &s3c24xx_spi_fiq_tx;
+			break;
+		case FIQ_MODE_RX:
+			code = &s3c24xx_spi_fiq_rx;
+			break;
+		case FIQ_MODE_TXRX:
+			code = &s3c24xx_spi_fiq_txrx;
+			break;
+		default:
+			code = NULL;
+		}
+
+		BUG_ON(!code);
+
+		ack_ptr = (u32 *)&code->data[code->ack_offset];
+		*ack_ptr = ack_bit(hw->irq);
+
+		set_fiq_handler(&code->data, code->length);
+	}
+
+	s3c24xx_set_fiq(hw->irq, true);
+
+	hw->fiq_mode = mode;
+	hw->fiq_inuse = 1;
+}
+
+/**
+ * s3c24xx_spi_fiqop - FIQ core code callback
+ * @pw: Data registered with the handler
+ * @release: Whether this is a release or a return.
+ *
+ * Called by the FIQ code when another module wants to use the FIQ, so
+ * return whether we are currently using this or not and then update our
+ * internal state.
+ */
+static int s3c24xx_spi_fiqop(void *pw, int release)
+{
+	struct s3c24xx_spi *hw = pw;
+	int ret = 0;
+
+	if (release) {
+		if (hw->fiq_inuse)
+			ret = -EBUSY;
+
+		/* note, we do not need to unroute the FIQ, as the FIQ
+		 * vector code de-routes it to signal the end of transfer */
+
+		hw->fiq_mode = FIQ_MODE_NONE;
+		hw->fiq_claimed = 0;
+	} else {
+		hw->fiq_claimed = 1;
+	}
+
+	return ret;
+}
+
+/**
+ * s3c24xx_spi_initfiq - setup the information for the FIQ core
+ * @hw: The hardware state.
+ *
+ * Setup the fiq_handler block to pass to the FIQ core.
+ */
+static inline void s3c24xx_spi_initfiq(struct s3c24xx_spi *hw)
+{
+	hw->fiq_handler.dev_id = hw;
+	hw->fiq_handler.name = dev_name(hw->dev);
+	hw->fiq_handler.fiq_op = s3c24xx_spi_fiqop;
+}
+
+/**
+ * s3c24xx_spi_usefiq - return if we should be using FIQ.
+ * @hw: The hardware state.
+ *
+ * Return true if the platform data specifies whether this channel is
+ * allowed to use the FIQ.
+ */
+static inline bool s3c24xx_spi_usefiq(struct s3c24xx_spi *hw)
+{
+	return hw->pdata->use_fiq;
+}
+
+/**
+ * s3c24xx_spi_usingfiq - return if channel is using FIQ
+ * @spi: The hardware state.
+ *
+ * Return whether the channel is currently using the FIQ (separate from
+ * whether the FIQ is claimed).
+ */
+static inline bool s3c24xx_spi_usingfiq(struct s3c24xx_spi *spi)
+{
+	return spi->fiq_inuse;
+}
+#else
+
+static inline void s3c24xx_spi_initfiq(struct s3c24xx_spi *s) { }
+static inline void s3c24xx_spi_tryfiq(struct s3c24xx_spi *s) { }
+static inline bool s3c24xx_spi_usefiq(struct s3c24xx_spi *s) { return false; }
+static inline bool s3c24xx_spi_usingfiq(struct s3c24xx_spi *s) { return false; }
+
+#endif /* CONFIG_SPI_S3C24XX_FIQ */
+
+static int s3c24xx_spi_txrx(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct s3c24xx_spi *hw = to_hw(spi);
+
+	hw->tx = t->tx_buf;
+	hw->rx = t->rx_buf;
+	hw->len = t->len;
+	hw->count = 0;
+
+	init_completion(&hw->done);
+
+	hw->fiq_inuse = 0;
+	if (s3c24xx_spi_usefiq(hw) && t->len >= 3)
+		s3c24xx_spi_tryfiq(hw);
+
+	/* send the first byte */
+	writeb(hw_txbyte(hw, 0), hw->regs + S3C2410_SPTDAT);
+
+	wait_for_completion(&hw->done);
+	return hw->count;
+}
+
+static irqreturn_t s3c24xx_spi_irq(int irq, void *dev)
+{
+	struct s3c24xx_spi *hw = dev;
+	unsigned int spsta = readb(hw->regs + S3C2410_SPSTA);
+	unsigned int count = hw->count;
+
+	if (spsta & S3C2410_SPSTA_DCOL) {
+		dev_dbg(hw->dev, "data-collision\n");
+		complete(&hw->done);
+		goto irq_done;
+	}
+
+	if (!(spsta & S3C2410_SPSTA_READY)) {
+		dev_dbg(hw->dev, "spi not ready for tx?\n");
+		complete(&hw->done);
+		goto irq_done;
+	}
+
+	if (!s3c24xx_spi_usingfiq(hw)) {
+		hw->count++;
+
+		if (hw->rx)
+			hw->rx[count] = readb(hw->regs + S3C2410_SPRDAT);
+
+		count++;
+
+		if (count < hw->len)
+			writeb(hw_txbyte(hw, count), hw->regs + S3C2410_SPTDAT);
+		else
+			complete(&hw->done);
+	} else {
+		hw->count = hw->len;
+		hw->fiq_inuse = 0;
+
+		if (hw->rx)
+			hw->rx[hw->len-1] = readb(hw->regs + S3C2410_SPRDAT);
+
+		complete(&hw->done);
+	}
+
+ irq_done:
+	return IRQ_HANDLED;
+}
+
+static void s3c24xx_spi_initialsetup(struct s3c24xx_spi *hw)
+{
+	/* for the moment, permanently enable the clock */
+
+	clk_enable(hw->clk);
+
+	/* program defaults into the registers */
+
+	writeb(0xff, hw->regs + S3C2410_SPPRE);
+	writeb(SPPIN_DEFAULT, hw->regs + S3C2410_SPPIN);
+	writeb(SPCON_DEFAULT, hw->regs + S3C2410_SPCON);
+
+	if (hw->pdata) {
+		if (hw->set_cs == s3c24xx_spi_gpiocs)
+			gpio_direction_output(hw->pdata->pin_cs, 1);
+
+		if (hw->pdata->gpio_setup)
+			hw->pdata->gpio_setup(hw->pdata, 1);
+	}
+}
+
+static int s3c24xx_spi_probe(struct platform_device *pdev)
+{
+	struct s3c2410_spi_info *pdata;
+	struct s3c24xx_spi *hw;
+	struct spi_master *master;
+	struct resource *res;
+	int err = 0;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct s3c24xx_spi));
+	if (master == NULL) {
+		dev_err(&pdev->dev, "No memory for spi_master\n");
+		return -ENOMEM;
+	}
+
+	hw = spi_master_get_devdata(master);
+
+	hw->master = master;
+	hw->pdata = pdata = dev_get_platdata(&pdev->dev);
+	hw->dev = &pdev->dev;
+
+	if (pdata == NULL) {
+		dev_err(&pdev->dev, "No platform data supplied\n");
+		err = -ENOENT;
+		goto err_no_pdata;
+	}
+
+	platform_set_drvdata(pdev, hw);
+	init_completion(&hw->done);
+
+	/* initialise fiq handler */
+
+	s3c24xx_spi_initfiq(hw);
+
+	/* setup the master state. */
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+
+	master->num_chipselect = hw->pdata->num_cs;
+	master->bus_num = pdata->bus_num;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+
+	/* setup the state for the bitbang driver */
+
+	hw->bitbang.master         = hw->master;
+	hw->bitbang.setup_transfer = s3c24xx_spi_setupxfer;
+	hw->bitbang.chipselect     = s3c24xx_spi_chipsel;
+	hw->bitbang.txrx_bufs      = s3c24xx_spi_txrx;
+
+	hw->master->setup  = s3c24xx_spi_setup;
+
+	dev_dbg(hw->dev, "bitbang at %p\n", &hw->bitbang);
+
+	/* find and map our resources */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	hw->regs = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(hw->regs)) {
+		err = PTR_ERR(hw->regs);
+		goto err_no_pdata;
+	}
+
+	hw->irq = platform_get_irq(pdev, 0);
+	if (hw->irq < 0) {
+		dev_err(&pdev->dev, "No IRQ specified\n");
+		err = -ENOENT;
+		goto err_no_pdata;
+	}
+
+	err = devm_request_irq(&pdev->dev, hw->irq, s3c24xx_spi_irq, 0,
+				pdev->name, hw);
+	if (err) {
+		dev_err(&pdev->dev, "Cannot claim IRQ\n");
+		goto err_no_pdata;
+	}
+
+	hw->clk = devm_clk_get(&pdev->dev, "spi");
+	if (IS_ERR(hw->clk)) {
+		dev_err(&pdev->dev, "No clock for device\n");
+		err = PTR_ERR(hw->clk);
+		goto err_no_pdata;
+	}
+
+	/* setup any gpio we can */
+
+	if (!pdata->set_cs) {
+		if (pdata->pin_cs < 0) {
+			dev_err(&pdev->dev, "No chipselect pin\n");
+			err = -EINVAL;
+			goto err_register;
+		}
+
+		err = devm_gpio_request(&pdev->dev, pdata->pin_cs,
+					dev_name(&pdev->dev));
+		if (err) {
+			dev_err(&pdev->dev, "Failed to get gpio for cs\n");
+			goto err_register;
+		}
+
+		hw->set_cs = s3c24xx_spi_gpiocs;
+		gpio_direction_output(pdata->pin_cs, 1);
+	} else
+		hw->set_cs = pdata->set_cs;
+
+	s3c24xx_spi_initialsetup(hw);
+
+	/* register our spi controller */
+
+	err = spi_bitbang_start(&hw->bitbang);
+	if (err) {
+		dev_err(&pdev->dev, "Failed to register SPI master\n");
+		goto err_register;
+	}
+
+	return 0;
+
+ err_register:
+	clk_disable(hw->clk);
+
+ err_no_pdata:
+	spi_master_put(hw->master);
+	return err;
+}
+
+static int s3c24xx_spi_remove(struct platform_device *dev)
+{
+	struct s3c24xx_spi *hw = platform_get_drvdata(dev);
+
+	spi_bitbang_stop(&hw->bitbang);
+	clk_disable(hw->clk);
+	spi_master_put(hw->master);
+	return 0;
+}
+
+
+#ifdef CONFIG_PM
+
+static int s3c24xx_spi_suspend(struct device *dev)
+{
+	struct s3c24xx_spi *hw = dev_get_drvdata(dev);
+	int ret;
+
+	ret = spi_master_suspend(hw->master);
+	if (ret)
+		return ret;
+
+	if (hw->pdata && hw->pdata->gpio_setup)
+		hw->pdata->gpio_setup(hw->pdata, 0);
+
+	clk_disable(hw->clk);
+	return 0;
+}
+
+static int s3c24xx_spi_resume(struct device *dev)
+{
+	struct s3c24xx_spi *hw = dev_get_drvdata(dev);
+
+	s3c24xx_spi_initialsetup(hw);
+	return spi_master_resume(hw->master);
+}
+
+static const struct dev_pm_ops s3c24xx_spi_pmops = {
+	.suspend	= s3c24xx_spi_suspend,
+	.resume		= s3c24xx_spi_resume,
+};
+
+#define S3C24XX_SPI_PMOPS &s3c24xx_spi_pmops
+#else
+#define S3C24XX_SPI_PMOPS NULL
+#endif /* CONFIG_PM */
+
+MODULE_ALIAS("platform:s3c2410-spi");
+static struct platform_driver s3c24xx_spi_driver = {
+	.probe		= s3c24xx_spi_probe,
+	.remove		= s3c24xx_spi_remove,
+	.driver		= {
+		.name	= "s3c2410-spi",
+		.pm	= S3C24XX_SPI_PMOPS,
+	},
+};
+module_platform_driver(s3c24xx_spi_driver);
+
+MODULE_DESCRIPTION("S3C24XX SPI Driver");
+MODULE_AUTHOR("Ben Dooks, <ben@simtec.co.uk>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-s3c64xx.c b/drivers/spi/spi-s3c64xx.c
new file mode 100644
index 0000000..7b7151e
--- /dev/null
+++ b/drivers/spi/spi-s3c64xx.c
@@ -0,0 +1,1448 @@
+// SPDX-License-Identifier: GPL-2.0+
+//
+// Copyright (c) 2009 Samsung Electronics Co., Ltd.
+//      Jaswinder Singh <jassi.brar@samsung.com>
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/clk.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/spi/spi.h>
+#include <linux/gpio.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+
+#include <linux/platform_data/spi-s3c64xx.h>
+
+#define MAX_SPI_PORTS		6
+#define S3C64XX_SPI_QUIRK_POLL		(1 << 0)
+#define S3C64XX_SPI_QUIRK_CS_AUTO	(1 << 1)
+#define AUTOSUSPEND_TIMEOUT	2000
+
+/* Registers and bit-fields */
+
+#define S3C64XX_SPI_CH_CFG		0x00
+#define S3C64XX_SPI_CLK_CFG		0x04
+#define S3C64XX_SPI_MODE_CFG		0x08
+#define S3C64XX_SPI_SLAVE_SEL		0x0C
+#define S3C64XX_SPI_INT_EN		0x10
+#define S3C64XX_SPI_STATUS		0x14
+#define S3C64XX_SPI_TX_DATA		0x18
+#define S3C64XX_SPI_RX_DATA		0x1C
+#define S3C64XX_SPI_PACKET_CNT		0x20
+#define S3C64XX_SPI_PENDING_CLR		0x24
+#define S3C64XX_SPI_SWAP_CFG		0x28
+#define S3C64XX_SPI_FB_CLK		0x2C
+
+#define S3C64XX_SPI_CH_HS_EN		(1<<6)	/* High Speed Enable */
+#define S3C64XX_SPI_CH_SW_RST		(1<<5)
+#define S3C64XX_SPI_CH_SLAVE		(1<<4)
+#define S3C64XX_SPI_CPOL_L		(1<<3)
+#define S3C64XX_SPI_CPHA_B		(1<<2)
+#define S3C64XX_SPI_CH_RXCH_ON		(1<<1)
+#define S3C64XX_SPI_CH_TXCH_ON		(1<<0)
+
+#define S3C64XX_SPI_CLKSEL_SRCMSK	(3<<9)
+#define S3C64XX_SPI_CLKSEL_SRCSHFT	9
+#define S3C64XX_SPI_ENCLK_ENABLE	(1<<8)
+#define S3C64XX_SPI_PSR_MASK		0xff
+
+#define S3C64XX_SPI_MODE_CH_TSZ_BYTE		(0<<29)
+#define S3C64XX_SPI_MODE_CH_TSZ_HALFWORD	(1<<29)
+#define S3C64XX_SPI_MODE_CH_TSZ_WORD		(2<<29)
+#define S3C64XX_SPI_MODE_CH_TSZ_MASK		(3<<29)
+#define S3C64XX_SPI_MODE_BUS_TSZ_BYTE		(0<<17)
+#define S3C64XX_SPI_MODE_BUS_TSZ_HALFWORD	(1<<17)
+#define S3C64XX_SPI_MODE_BUS_TSZ_WORD		(2<<17)
+#define S3C64XX_SPI_MODE_BUS_TSZ_MASK		(3<<17)
+#define S3C64XX_SPI_MODE_RXDMA_ON		(1<<2)
+#define S3C64XX_SPI_MODE_TXDMA_ON		(1<<1)
+#define S3C64XX_SPI_MODE_4BURST			(1<<0)
+
+#define S3C64XX_SPI_SLAVE_AUTO			(1<<1)
+#define S3C64XX_SPI_SLAVE_SIG_INACT		(1<<0)
+#define S3C64XX_SPI_SLAVE_NSC_CNT_2		(2<<4)
+
+#define S3C64XX_SPI_INT_TRAILING_EN		(1<<6)
+#define S3C64XX_SPI_INT_RX_OVERRUN_EN		(1<<5)
+#define S3C64XX_SPI_INT_RX_UNDERRUN_EN		(1<<4)
+#define S3C64XX_SPI_INT_TX_OVERRUN_EN		(1<<3)
+#define S3C64XX_SPI_INT_TX_UNDERRUN_EN		(1<<2)
+#define S3C64XX_SPI_INT_RX_FIFORDY_EN		(1<<1)
+#define S3C64XX_SPI_INT_TX_FIFORDY_EN		(1<<0)
+
+#define S3C64XX_SPI_ST_RX_OVERRUN_ERR		(1<<5)
+#define S3C64XX_SPI_ST_RX_UNDERRUN_ERR		(1<<4)
+#define S3C64XX_SPI_ST_TX_OVERRUN_ERR		(1<<3)
+#define S3C64XX_SPI_ST_TX_UNDERRUN_ERR		(1<<2)
+#define S3C64XX_SPI_ST_RX_FIFORDY		(1<<1)
+#define S3C64XX_SPI_ST_TX_FIFORDY		(1<<0)
+
+#define S3C64XX_SPI_PACKET_CNT_EN		(1<<16)
+
+#define S3C64XX_SPI_PND_TX_UNDERRUN_CLR		(1<<4)
+#define S3C64XX_SPI_PND_TX_OVERRUN_CLR		(1<<3)
+#define S3C64XX_SPI_PND_RX_UNDERRUN_CLR		(1<<2)
+#define S3C64XX_SPI_PND_RX_OVERRUN_CLR		(1<<1)
+#define S3C64XX_SPI_PND_TRAILING_CLR		(1<<0)
+
+#define S3C64XX_SPI_SWAP_RX_HALF_WORD		(1<<7)
+#define S3C64XX_SPI_SWAP_RX_BYTE		(1<<6)
+#define S3C64XX_SPI_SWAP_RX_BIT			(1<<5)
+#define S3C64XX_SPI_SWAP_RX_EN			(1<<4)
+#define S3C64XX_SPI_SWAP_TX_HALF_WORD		(1<<3)
+#define S3C64XX_SPI_SWAP_TX_BYTE		(1<<2)
+#define S3C64XX_SPI_SWAP_TX_BIT			(1<<1)
+#define S3C64XX_SPI_SWAP_TX_EN			(1<<0)
+
+#define S3C64XX_SPI_FBCLK_MSK			(3<<0)
+
+#define FIFO_LVL_MASK(i) ((i)->port_conf->fifo_lvl_mask[i->port_id])
+#define S3C64XX_SPI_ST_TX_DONE(v, i) (((v) & \
+				(1 << (i)->port_conf->tx_st_done)) ? 1 : 0)
+#define TX_FIFO_LVL(v, i) (((v) >> 6) & FIFO_LVL_MASK(i))
+#define RX_FIFO_LVL(v, i) (((v) >> (i)->port_conf->rx_lvl_offset) & \
+					FIFO_LVL_MASK(i))
+
+#define S3C64XX_SPI_MAX_TRAILCNT	0x3ff
+#define S3C64XX_SPI_TRAILCNT_OFF	19
+
+#define S3C64XX_SPI_TRAILCNT		S3C64XX_SPI_MAX_TRAILCNT
+
+#define msecs_to_loops(t) (loops_per_jiffy / 1000 * HZ * t)
+#define is_polling(x)	(x->port_conf->quirks & S3C64XX_SPI_QUIRK_POLL)
+
+#define RXBUSY    (1<<2)
+#define TXBUSY    (1<<3)
+
+struct s3c64xx_spi_dma_data {
+	struct dma_chan *ch;
+	enum dma_transfer_direction direction;
+};
+
+/**
+ * struct s3c64xx_spi_info - SPI Controller hardware info
+ * @fifo_lvl_mask: Bit-mask for {TX|RX}_FIFO_LVL bits in SPI_STATUS register.
+ * @rx_lvl_offset: Bit offset of RX_FIFO_LVL bits in SPI_STATUS regiter.
+ * @tx_st_done: Bit offset of TX_DONE bit in SPI_STATUS regiter.
+ * @high_speed: True, if the controller supports HIGH_SPEED_EN bit.
+ * @clk_from_cmu: True, if the controller does not include a clock mux and
+ *	prescaler unit.
+ *
+ * The Samsung s3c64xx SPI controller are used on various Samsung SoC's but
+ * differ in some aspects such as the size of the fifo and spi bus clock
+ * setup. Such differences are specified to the driver using this structure
+ * which is provided as driver data to the driver.
+ */
+struct s3c64xx_spi_port_config {
+	int	fifo_lvl_mask[MAX_SPI_PORTS];
+	int	rx_lvl_offset;
+	int	tx_st_done;
+	int	quirks;
+	bool	high_speed;
+	bool	clk_from_cmu;
+	bool	clk_ioclk;
+};
+
+/**
+ * struct s3c64xx_spi_driver_data - Runtime info holder for SPI driver.
+ * @clk: Pointer to the spi clock.
+ * @src_clk: Pointer to the clock used to generate SPI signals.
+ * @ioclk: Pointer to the i/o clock between master and slave
+ * @master: Pointer to the SPI Protocol master.
+ * @cntrlr_info: Platform specific data for the controller this driver manages.
+ * @lock: Controller specific lock.
+ * @state: Set of FLAGS to indicate status.
+ * @rx_dmach: Controller's DMA channel for Rx.
+ * @tx_dmach: Controller's DMA channel for Tx.
+ * @sfr_start: BUS address of SPI controller regs.
+ * @regs: Pointer to ioremap'ed controller registers.
+ * @irq: interrupt
+ * @xfer_completion: To indicate completion of xfer task.
+ * @cur_mode: Stores the active configuration of the controller.
+ * @cur_bpw: Stores the active bits per word settings.
+ * @cur_speed: Stores the active xfer clock speed.
+ */
+struct s3c64xx_spi_driver_data {
+	void __iomem                    *regs;
+	struct clk                      *clk;
+	struct clk                      *src_clk;
+	struct clk                      *ioclk;
+	struct platform_device          *pdev;
+	struct spi_master               *master;
+	struct s3c64xx_spi_info  *cntrlr_info;
+	spinlock_t                      lock;
+	unsigned long                   sfr_start;
+	struct completion               xfer_completion;
+	unsigned                        state;
+	unsigned                        cur_mode, cur_bpw;
+	unsigned                        cur_speed;
+	struct s3c64xx_spi_dma_data	rx_dma;
+	struct s3c64xx_spi_dma_data	tx_dma;
+	struct s3c64xx_spi_port_config	*port_conf;
+	unsigned int			port_id;
+};
+
+static void s3c64xx_flush_fifo(struct s3c64xx_spi_driver_data *sdd)
+{
+	void __iomem *regs = sdd->regs;
+	unsigned long loops;
+	u32 val;
+
+	writel(0, regs + S3C64XX_SPI_PACKET_CNT);
+
+	val = readl(regs + S3C64XX_SPI_CH_CFG);
+	val &= ~(S3C64XX_SPI_CH_RXCH_ON | S3C64XX_SPI_CH_TXCH_ON);
+	writel(val, regs + S3C64XX_SPI_CH_CFG);
+
+	val = readl(regs + S3C64XX_SPI_CH_CFG);
+	val |= S3C64XX_SPI_CH_SW_RST;
+	val &= ~S3C64XX_SPI_CH_HS_EN;
+	writel(val, regs + S3C64XX_SPI_CH_CFG);
+
+	/* Flush TxFIFO*/
+	loops = msecs_to_loops(1);
+	do {
+		val = readl(regs + S3C64XX_SPI_STATUS);
+	} while (TX_FIFO_LVL(val, sdd) && loops--);
+
+	if (loops == 0)
+		dev_warn(&sdd->pdev->dev, "Timed out flushing TX FIFO\n");
+
+	/* Flush RxFIFO*/
+	loops = msecs_to_loops(1);
+	do {
+		val = readl(regs + S3C64XX_SPI_STATUS);
+		if (RX_FIFO_LVL(val, sdd))
+			readl(regs + S3C64XX_SPI_RX_DATA);
+		else
+			break;
+	} while (loops--);
+
+	if (loops == 0)
+		dev_warn(&sdd->pdev->dev, "Timed out flushing RX FIFO\n");
+
+	val = readl(regs + S3C64XX_SPI_CH_CFG);
+	val &= ~S3C64XX_SPI_CH_SW_RST;
+	writel(val, regs + S3C64XX_SPI_CH_CFG);
+
+	val = readl(regs + S3C64XX_SPI_MODE_CFG);
+	val &= ~(S3C64XX_SPI_MODE_TXDMA_ON | S3C64XX_SPI_MODE_RXDMA_ON);
+	writel(val, regs + S3C64XX_SPI_MODE_CFG);
+}
+
+static void s3c64xx_spi_dmacb(void *data)
+{
+	struct s3c64xx_spi_driver_data *sdd;
+	struct s3c64xx_spi_dma_data *dma = data;
+	unsigned long flags;
+
+	if (dma->direction == DMA_DEV_TO_MEM)
+		sdd = container_of(data,
+			struct s3c64xx_spi_driver_data, rx_dma);
+	else
+		sdd = container_of(data,
+			struct s3c64xx_spi_driver_data, tx_dma);
+
+	spin_lock_irqsave(&sdd->lock, flags);
+
+	if (dma->direction == DMA_DEV_TO_MEM) {
+		sdd->state &= ~RXBUSY;
+		if (!(sdd->state & TXBUSY))
+			complete(&sdd->xfer_completion);
+	} else {
+		sdd->state &= ~TXBUSY;
+		if (!(sdd->state & RXBUSY))
+			complete(&sdd->xfer_completion);
+	}
+
+	spin_unlock_irqrestore(&sdd->lock, flags);
+}
+
+static void prepare_dma(struct s3c64xx_spi_dma_data *dma,
+			struct sg_table *sgt)
+{
+	struct s3c64xx_spi_driver_data *sdd;
+	struct dma_slave_config config;
+	struct dma_async_tx_descriptor *desc;
+
+	memset(&config, 0, sizeof(config));
+
+	if (dma->direction == DMA_DEV_TO_MEM) {
+		sdd = container_of((void *)dma,
+			struct s3c64xx_spi_driver_data, rx_dma);
+		config.direction = dma->direction;
+		config.src_addr = sdd->sfr_start + S3C64XX_SPI_RX_DATA;
+		config.src_addr_width = sdd->cur_bpw / 8;
+		config.src_maxburst = 1;
+		dmaengine_slave_config(dma->ch, &config);
+	} else {
+		sdd = container_of((void *)dma,
+			struct s3c64xx_spi_driver_data, tx_dma);
+		config.direction = dma->direction;
+		config.dst_addr = sdd->sfr_start + S3C64XX_SPI_TX_DATA;
+		config.dst_addr_width = sdd->cur_bpw / 8;
+		config.dst_maxburst = 1;
+		dmaengine_slave_config(dma->ch, &config);
+	}
+
+	desc = dmaengine_prep_slave_sg(dma->ch, sgt->sgl, sgt->nents,
+				       dma->direction, DMA_PREP_INTERRUPT);
+
+	desc->callback = s3c64xx_spi_dmacb;
+	desc->callback_param = dma;
+
+	dmaengine_submit(desc);
+	dma_async_issue_pending(dma->ch);
+}
+
+static void s3c64xx_spi_set_cs(struct spi_device *spi, bool enable)
+{
+	struct s3c64xx_spi_driver_data *sdd =
+					spi_master_get_devdata(spi->master);
+
+	if (sdd->cntrlr_info->no_cs)
+		return;
+
+	if (enable) {
+		if (!(sdd->port_conf->quirks & S3C64XX_SPI_QUIRK_CS_AUTO)) {
+			writel(0, sdd->regs + S3C64XX_SPI_SLAVE_SEL);
+		} else {
+			u32 ssel = readl(sdd->regs + S3C64XX_SPI_SLAVE_SEL);
+
+			ssel |= (S3C64XX_SPI_SLAVE_AUTO |
+						S3C64XX_SPI_SLAVE_NSC_CNT_2);
+			writel(ssel, sdd->regs + S3C64XX_SPI_SLAVE_SEL);
+		}
+	} else {
+		if (!(sdd->port_conf->quirks & S3C64XX_SPI_QUIRK_CS_AUTO))
+			writel(S3C64XX_SPI_SLAVE_SIG_INACT,
+			       sdd->regs + S3C64XX_SPI_SLAVE_SEL);
+	}
+}
+
+static int s3c64xx_spi_prepare_transfer(struct spi_master *spi)
+{
+	struct s3c64xx_spi_driver_data *sdd = spi_master_get_devdata(spi);
+
+	if (is_polling(sdd))
+		return 0;
+
+	spi->dma_rx = sdd->rx_dma.ch;
+	spi->dma_tx = sdd->tx_dma.ch;
+
+	return 0;
+}
+
+static bool s3c64xx_spi_can_dma(struct spi_master *master,
+				struct spi_device *spi,
+				struct spi_transfer *xfer)
+{
+	struct s3c64xx_spi_driver_data *sdd = spi_master_get_devdata(master);
+
+	return xfer->len > (FIFO_LVL_MASK(sdd) >> 1) + 1;
+}
+
+static void s3c64xx_enable_datapath(struct s3c64xx_spi_driver_data *sdd,
+				    struct spi_transfer *xfer, int dma_mode)
+{
+	void __iomem *regs = sdd->regs;
+	u32 modecfg, chcfg;
+
+	modecfg = readl(regs + S3C64XX_SPI_MODE_CFG);
+	modecfg &= ~(S3C64XX_SPI_MODE_TXDMA_ON | S3C64XX_SPI_MODE_RXDMA_ON);
+
+	chcfg = readl(regs + S3C64XX_SPI_CH_CFG);
+	chcfg &= ~S3C64XX_SPI_CH_TXCH_ON;
+
+	if (dma_mode) {
+		chcfg &= ~S3C64XX_SPI_CH_RXCH_ON;
+	} else {
+		/* Always shift in data in FIFO, even if xfer is Tx only,
+		 * this helps setting PCKT_CNT value for generating clocks
+		 * as exactly needed.
+		 */
+		chcfg |= S3C64XX_SPI_CH_RXCH_ON;
+		writel(((xfer->len * 8 / sdd->cur_bpw) & 0xffff)
+					| S3C64XX_SPI_PACKET_CNT_EN,
+					regs + S3C64XX_SPI_PACKET_CNT);
+	}
+
+	if (xfer->tx_buf != NULL) {
+		sdd->state |= TXBUSY;
+		chcfg |= S3C64XX_SPI_CH_TXCH_ON;
+		if (dma_mode) {
+			modecfg |= S3C64XX_SPI_MODE_TXDMA_ON;
+			prepare_dma(&sdd->tx_dma, &xfer->tx_sg);
+		} else {
+			switch (sdd->cur_bpw) {
+			case 32:
+				iowrite32_rep(regs + S3C64XX_SPI_TX_DATA,
+					xfer->tx_buf, xfer->len / 4);
+				break;
+			case 16:
+				iowrite16_rep(regs + S3C64XX_SPI_TX_DATA,
+					xfer->tx_buf, xfer->len / 2);
+				break;
+			default:
+				iowrite8_rep(regs + S3C64XX_SPI_TX_DATA,
+					xfer->tx_buf, xfer->len);
+				break;
+			}
+		}
+	}
+
+	if (xfer->rx_buf != NULL) {
+		sdd->state |= RXBUSY;
+
+		if (sdd->port_conf->high_speed && sdd->cur_speed >= 30000000UL
+					&& !(sdd->cur_mode & SPI_CPHA))
+			chcfg |= S3C64XX_SPI_CH_HS_EN;
+
+		if (dma_mode) {
+			modecfg |= S3C64XX_SPI_MODE_RXDMA_ON;
+			chcfg |= S3C64XX_SPI_CH_RXCH_ON;
+			writel(((xfer->len * 8 / sdd->cur_bpw) & 0xffff)
+					| S3C64XX_SPI_PACKET_CNT_EN,
+					regs + S3C64XX_SPI_PACKET_CNT);
+			prepare_dma(&sdd->rx_dma, &xfer->rx_sg);
+		}
+	}
+
+	writel(modecfg, regs + S3C64XX_SPI_MODE_CFG);
+	writel(chcfg, regs + S3C64XX_SPI_CH_CFG);
+}
+
+static u32 s3c64xx_spi_wait_for_timeout(struct s3c64xx_spi_driver_data *sdd,
+					int timeout_ms)
+{
+	void __iomem *regs = sdd->regs;
+	unsigned long val = 1;
+	u32 status;
+
+	/* max fifo depth available */
+	u32 max_fifo = (FIFO_LVL_MASK(sdd) >> 1) + 1;
+
+	if (timeout_ms)
+		val = msecs_to_loops(timeout_ms);
+
+	do {
+		status = readl(regs + S3C64XX_SPI_STATUS);
+	} while (RX_FIFO_LVL(status, sdd) < max_fifo && --val);
+
+	/* return the actual received data length */
+	return RX_FIFO_LVL(status, sdd);
+}
+
+static int s3c64xx_wait_for_dma(struct s3c64xx_spi_driver_data *sdd,
+				struct spi_transfer *xfer)
+{
+	void __iomem *regs = sdd->regs;
+	unsigned long val;
+	u32 status;
+	int ms;
+
+	/* millisecs to xfer 'len' bytes @ 'cur_speed' */
+	ms = xfer->len * 8 * 1000 / sdd->cur_speed;
+	ms += 10; /* some tolerance */
+
+	val = msecs_to_jiffies(ms) + 10;
+	val = wait_for_completion_timeout(&sdd->xfer_completion, val);
+
+	/*
+	 * If the previous xfer was completed within timeout, then
+	 * proceed further else return -EIO.
+	 * DmaTx returns after simply writing data in the FIFO,
+	 * w/o waiting for real transmission on the bus to finish.
+	 * DmaRx returns only after Dma read data from FIFO which
+	 * needs bus transmission to finish, so we don't worry if
+	 * Xfer involved Rx(with or without Tx).
+	 */
+	if (val && !xfer->rx_buf) {
+		val = msecs_to_loops(10);
+		status = readl(regs + S3C64XX_SPI_STATUS);
+		while ((TX_FIFO_LVL(status, sdd)
+			|| !S3C64XX_SPI_ST_TX_DONE(status, sdd))
+		       && --val) {
+			cpu_relax();
+			status = readl(regs + S3C64XX_SPI_STATUS);
+		}
+
+	}
+
+	/* If timed out while checking rx/tx status return error */
+	if (!val)
+		return -EIO;
+
+	return 0;
+}
+
+static int s3c64xx_wait_for_pio(struct s3c64xx_spi_driver_data *sdd,
+				struct spi_transfer *xfer)
+{
+	void __iomem *regs = sdd->regs;
+	unsigned long val;
+	u32 status;
+	int loops;
+	u32 cpy_len;
+	u8 *buf;
+	int ms;
+
+	/* millisecs to xfer 'len' bytes @ 'cur_speed' */
+	ms = xfer->len * 8 * 1000 / sdd->cur_speed;
+	ms += 10; /* some tolerance */
+
+	val = msecs_to_loops(ms);
+	do {
+		status = readl(regs + S3C64XX_SPI_STATUS);
+	} while (RX_FIFO_LVL(status, sdd) < xfer->len && --val);
+
+	if (!val)
+		return -EIO;
+
+	/* If it was only Tx */
+	if (!xfer->rx_buf) {
+		sdd->state &= ~TXBUSY;
+		return 0;
+	}
+
+	/*
+	 * If the receive length is bigger than the controller fifo
+	 * size, calculate the loops and read the fifo as many times.
+	 * loops = length / max fifo size (calculated by using the
+	 * fifo mask).
+	 * For any size less than the fifo size the below code is
+	 * executed atleast once.
+	 */
+	loops = xfer->len / ((FIFO_LVL_MASK(sdd) >> 1) + 1);
+	buf = xfer->rx_buf;
+	do {
+		/* wait for data to be received in the fifo */
+		cpy_len = s3c64xx_spi_wait_for_timeout(sdd,
+						       (loops ? ms : 0));
+
+		switch (sdd->cur_bpw) {
+		case 32:
+			ioread32_rep(regs + S3C64XX_SPI_RX_DATA,
+				     buf, cpy_len / 4);
+			break;
+		case 16:
+			ioread16_rep(regs + S3C64XX_SPI_RX_DATA,
+				     buf, cpy_len / 2);
+			break;
+		default:
+			ioread8_rep(regs + S3C64XX_SPI_RX_DATA,
+				    buf, cpy_len);
+			break;
+		}
+
+		buf = buf + cpy_len;
+	} while (loops--);
+	sdd->state &= ~RXBUSY;
+
+	return 0;
+}
+
+static void s3c64xx_spi_config(struct s3c64xx_spi_driver_data *sdd)
+{
+	void __iomem *regs = sdd->regs;
+	u32 val;
+
+	/* Disable Clock */
+	if (!sdd->port_conf->clk_from_cmu) {
+		val = readl(regs + S3C64XX_SPI_CLK_CFG);
+		val &= ~S3C64XX_SPI_ENCLK_ENABLE;
+		writel(val, regs + S3C64XX_SPI_CLK_CFG);
+	}
+
+	/* Set Polarity and Phase */
+	val = readl(regs + S3C64XX_SPI_CH_CFG);
+	val &= ~(S3C64XX_SPI_CH_SLAVE |
+			S3C64XX_SPI_CPOL_L |
+			S3C64XX_SPI_CPHA_B);
+
+	if (sdd->cur_mode & SPI_CPOL)
+		val |= S3C64XX_SPI_CPOL_L;
+
+	if (sdd->cur_mode & SPI_CPHA)
+		val |= S3C64XX_SPI_CPHA_B;
+
+	writel(val, regs + S3C64XX_SPI_CH_CFG);
+
+	/* Set Channel & DMA Mode */
+	val = readl(regs + S3C64XX_SPI_MODE_CFG);
+	val &= ~(S3C64XX_SPI_MODE_BUS_TSZ_MASK
+			| S3C64XX_SPI_MODE_CH_TSZ_MASK);
+
+	switch (sdd->cur_bpw) {
+	case 32:
+		val |= S3C64XX_SPI_MODE_BUS_TSZ_WORD;
+		val |= S3C64XX_SPI_MODE_CH_TSZ_WORD;
+		break;
+	case 16:
+		val |= S3C64XX_SPI_MODE_BUS_TSZ_HALFWORD;
+		val |= S3C64XX_SPI_MODE_CH_TSZ_HALFWORD;
+		break;
+	default:
+		val |= S3C64XX_SPI_MODE_BUS_TSZ_BYTE;
+		val |= S3C64XX_SPI_MODE_CH_TSZ_BYTE;
+		break;
+	}
+
+	writel(val, regs + S3C64XX_SPI_MODE_CFG);
+
+	if (sdd->port_conf->clk_from_cmu) {
+		/* The src_clk clock is divided internally by 2 */
+		clk_set_rate(sdd->src_clk, sdd->cur_speed * 2);
+	} else {
+		/* Configure Clock */
+		val = readl(regs + S3C64XX_SPI_CLK_CFG);
+		val &= ~S3C64XX_SPI_PSR_MASK;
+		val |= ((clk_get_rate(sdd->src_clk) / sdd->cur_speed / 2 - 1)
+				& S3C64XX_SPI_PSR_MASK);
+		writel(val, regs + S3C64XX_SPI_CLK_CFG);
+
+		/* Enable Clock */
+		val = readl(regs + S3C64XX_SPI_CLK_CFG);
+		val |= S3C64XX_SPI_ENCLK_ENABLE;
+		writel(val, regs + S3C64XX_SPI_CLK_CFG);
+	}
+}
+
+#define XFER_DMAADDR_INVALID DMA_BIT_MASK(32)
+
+static int s3c64xx_spi_prepare_message(struct spi_master *master,
+				       struct spi_message *msg)
+{
+	struct s3c64xx_spi_driver_data *sdd = spi_master_get_devdata(master);
+	struct spi_device *spi = msg->spi;
+	struct s3c64xx_spi_csinfo *cs = spi->controller_data;
+
+	/* Configure feedback delay */
+	writel(cs->fb_delay & 0x3, sdd->regs + S3C64XX_SPI_FB_CLK);
+
+	return 0;
+}
+
+static int s3c64xx_spi_transfer_one(struct spi_master *master,
+				    struct spi_device *spi,
+				    struct spi_transfer *xfer)
+{
+	struct s3c64xx_spi_driver_data *sdd = spi_master_get_devdata(master);
+	const unsigned int fifo_len = (FIFO_LVL_MASK(sdd) >> 1) + 1;
+	const void *tx_buf = NULL;
+	void *rx_buf = NULL;
+	int target_len = 0, origin_len = 0;
+	int use_dma = 0;
+	int status;
+	u32 speed;
+	u8 bpw;
+	unsigned long flags;
+
+	reinit_completion(&sdd->xfer_completion);
+
+	/* Only BPW and Speed may change across transfers */
+	bpw = xfer->bits_per_word;
+	speed = xfer->speed_hz;
+
+	if (bpw != sdd->cur_bpw || speed != sdd->cur_speed) {
+		sdd->cur_bpw = bpw;
+		sdd->cur_speed = speed;
+		sdd->cur_mode = spi->mode;
+		s3c64xx_spi_config(sdd);
+	}
+
+	if (!is_polling(sdd) && (xfer->len > fifo_len) &&
+	    sdd->rx_dma.ch && sdd->tx_dma.ch) {
+		use_dma = 1;
+
+	} else if (is_polling(sdd) && xfer->len > fifo_len) {
+		tx_buf = xfer->tx_buf;
+		rx_buf = xfer->rx_buf;
+		origin_len = xfer->len;
+
+		target_len = xfer->len;
+		if (xfer->len > fifo_len)
+			xfer->len = fifo_len;
+	}
+
+	do {
+		spin_lock_irqsave(&sdd->lock, flags);
+
+		/* Pending only which is to be done */
+		sdd->state &= ~RXBUSY;
+		sdd->state &= ~TXBUSY;
+
+		s3c64xx_enable_datapath(sdd, xfer, use_dma);
+
+		/* Start the signals */
+		s3c64xx_spi_set_cs(spi, true);
+
+		spin_unlock_irqrestore(&sdd->lock, flags);
+
+		if (use_dma)
+			status = s3c64xx_wait_for_dma(sdd, xfer);
+		else
+			status = s3c64xx_wait_for_pio(sdd, xfer);
+
+		if (status) {
+			dev_err(&spi->dev,
+				"I/O Error: rx-%d tx-%d res:rx-%c tx-%c len-%d\n",
+				xfer->rx_buf ? 1 : 0, xfer->tx_buf ? 1 : 0,
+				(sdd->state & RXBUSY) ? 'f' : 'p',
+				(sdd->state & TXBUSY) ? 'f' : 'p',
+				xfer->len);
+
+			if (use_dma) {
+				if (xfer->tx_buf && (sdd->state & TXBUSY))
+					dmaengine_terminate_all(sdd->tx_dma.ch);
+				if (xfer->rx_buf && (sdd->state & RXBUSY))
+					dmaengine_terminate_all(sdd->rx_dma.ch);
+			}
+		} else {
+			s3c64xx_flush_fifo(sdd);
+		}
+		if (target_len > 0) {
+			target_len -= xfer->len;
+
+			if (xfer->tx_buf)
+				xfer->tx_buf += xfer->len;
+
+			if (xfer->rx_buf)
+				xfer->rx_buf += xfer->len;
+
+			if (target_len > fifo_len)
+				xfer->len = fifo_len;
+			else
+				xfer->len = target_len;
+		}
+	} while (target_len > 0);
+
+	if (origin_len) {
+		/* Restore original xfer buffers and length */
+		xfer->tx_buf = tx_buf;
+		xfer->rx_buf = rx_buf;
+		xfer->len = origin_len;
+	}
+
+	return status;
+}
+
+static struct s3c64xx_spi_csinfo *s3c64xx_get_slave_ctrldata(
+				struct spi_device *spi)
+{
+	struct s3c64xx_spi_csinfo *cs;
+	struct device_node *slave_np, *data_np = NULL;
+	u32 fb_delay = 0;
+
+	slave_np = spi->dev.of_node;
+	if (!slave_np) {
+		dev_err(&spi->dev, "device node not found\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	data_np = of_get_child_by_name(slave_np, "controller-data");
+	if (!data_np) {
+		dev_err(&spi->dev, "child node 'controller-data' not found\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	cs = kzalloc(sizeof(*cs), GFP_KERNEL);
+	if (!cs) {
+		of_node_put(data_np);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	of_property_read_u32(data_np, "samsung,spi-feedback-delay", &fb_delay);
+	cs->fb_delay = fb_delay;
+	of_node_put(data_np);
+	return cs;
+}
+
+/*
+ * Here we only check the validity of requested configuration
+ * and save the configuration in a local data-structure.
+ * The controller is actually configured only just before we
+ * get a message to transfer.
+ */
+static int s3c64xx_spi_setup(struct spi_device *spi)
+{
+	struct s3c64xx_spi_csinfo *cs = spi->controller_data;
+	struct s3c64xx_spi_driver_data *sdd;
+	int err;
+
+	sdd = spi_master_get_devdata(spi->master);
+	if (spi->dev.of_node) {
+		cs = s3c64xx_get_slave_ctrldata(spi);
+		spi->controller_data = cs;
+	} else if (cs) {
+		/* On non-DT platforms the SPI core will set spi->cs_gpio
+		 * to -ENOENT. The GPIO pin used to drive the chip select
+		 * is defined by using platform data so spi->cs_gpio value
+		 * has to be override to have the proper GPIO pin number.
+		 */
+		spi->cs_gpio = cs->line;
+	}
+
+	if (IS_ERR_OR_NULL(cs)) {
+		dev_err(&spi->dev, "No CS for SPI(%d)\n", spi->chip_select);
+		return -ENODEV;
+	}
+
+	if (!spi_get_ctldata(spi)) {
+		if (gpio_is_valid(spi->cs_gpio)) {
+			err = gpio_request_one(spi->cs_gpio, GPIOF_OUT_INIT_HIGH,
+					       dev_name(&spi->dev));
+			if (err) {
+				dev_err(&spi->dev,
+					"Failed to get /CS gpio [%d]: %d\n",
+					spi->cs_gpio, err);
+				goto err_gpio_req;
+			}
+		}
+
+		spi_set_ctldata(spi, cs);
+	}
+
+	pm_runtime_get_sync(&sdd->pdev->dev);
+
+	/* Check if we can provide the requested rate */
+	if (!sdd->port_conf->clk_from_cmu) {
+		u32 psr, speed;
+
+		/* Max possible */
+		speed = clk_get_rate(sdd->src_clk) / 2 / (0 + 1);
+
+		if (spi->max_speed_hz > speed)
+			spi->max_speed_hz = speed;
+
+		psr = clk_get_rate(sdd->src_clk) / 2 / spi->max_speed_hz - 1;
+		psr &= S3C64XX_SPI_PSR_MASK;
+		if (psr == S3C64XX_SPI_PSR_MASK)
+			psr--;
+
+		speed = clk_get_rate(sdd->src_clk) / 2 / (psr + 1);
+		if (spi->max_speed_hz < speed) {
+			if (psr+1 < S3C64XX_SPI_PSR_MASK) {
+				psr++;
+			} else {
+				err = -EINVAL;
+				goto setup_exit;
+			}
+		}
+
+		speed = clk_get_rate(sdd->src_clk) / 2 / (psr + 1);
+		if (spi->max_speed_hz >= speed) {
+			spi->max_speed_hz = speed;
+		} else {
+			dev_err(&spi->dev, "Can't set %dHz transfer speed\n",
+				spi->max_speed_hz);
+			err = -EINVAL;
+			goto setup_exit;
+		}
+	}
+
+	pm_runtime_mark_last_busy(&sdd->pdev->dev);
+	pm_runtime_put_autosuspend(&sdd->pdev->dev);
+	s3c64xx_spi_set_cs(spi, false);
+
+	return 0;
+
+setup_exit:
+	pm_runtime_mark_last_busy(&sdd->pdev->dev);
+	pm_runtime_put_autosuspend(&sdd->pdev->dev);
+	/* setup() returns with device de-selected */
+	s3c64xx_spi_set_cs(spi, false);
+
+	if (gpio_is_valid(spi->cs_gpio))
+		gpio_free(spi->cs_gpio);
+	spi_set_ctldata(spi, NULL);
+
+err_gpio_req:
+	if (spi->dev.of_node)
+		kfree(cs);
+
+	return err;
+}
+
+static void s3c64xx_spi_cleanup(struct spi_device *spi)
+{
+	struct s3c64xx_spi_csinfo *cs = spi_get_ctldata(spi);
+
+	if (gpio_is_valid(spi->cs_gpio)) {
+		gpio_free(spi->cs_gpio);
+		if (spi->dev.of_node)
+			kfree(cs);
+		else {
+			/* On non-DT platforms, the SPI core sets
+			 * spi->cs_gpio to -ENOENT and .setup()
+			 * overrides it with the GPIO pin value
+			 * passed using platform data.
+			 */
+			spi->cs_gpio = -ENOENT;
+		}
+	}
+
+	spi_set_ctldata(spi, NULL);
+}
+
+static irqreturn_t s3c64xx_spi_irq(int irq, void *data)
+{
+	struct s3c64xx_spi_driver_data *sdd = data;
+	struct spi_master *spi = sdd->master;
+	unsigned int val, clr = 0;
+
+	val = readl(sdd->regs + S3C64XX_SPI_STATUS);
+
+	if (val & S3C64XX_SPI_ST_RX_OVERRUN_ERR) {
+		clr = S3C64XX_SPI_PND_RX_OVERRUN_CLR;
+		dev_err(&spi->dev, "RX overrun\n");
+	}
+	if (val & S3C64XX_SPI_ST_RX_UNDERRUN_ERR) {
+		clr |= S3C64XX_SPI_PND_RX_UNDERRUN_CLR;
+		dev_err(&spi->dev, "RX underrun\n");
+	}
+	if (val & S3C64XX_SPI_ST_TX_OVERRUN_ERR) {
+		clr |= S3C64XX_SPI_PND_TX_OVERRUN_CLR;
+		dev_err(&spi->dev, "TX overrun\n");
+	}
+	if (val & S3C64XX_SPI_ST_TX_UNDERRUN_ERR) {
+		clr |= S3C64XX_SPI_PND_TX_UNDERRUN_CLR;
+		dev_err(&spi->dev, "TX underrun\n");
+	}
+
+	/* Clear the pending irq by setting and then clearing it */
+	writel(clr, sdd->regs + S3C64XX_SPI_PENDING_CLR);
+	writel(0, sdd->regs + S3C64XX_SPI_PENDING_CLR);
+
+	return IRQ_HANDLED;
+}
+
+static void s3c64xx_spi_hwinit(struct s3c64xx_spi_driver_data *sdd)
+{
+	struct s3c64xx_spi_info *sci = sdd->cntrlr_info;
+	void __iomem *regs = sdd->regs;
+	unsigned int val;
+
+	sdd->cur_speed = 0;
+
+	if (sci->no_cs)
+		writel(0, sdd->regs + S3C64XX_SPI_SLAVE_SEL);
+	else if (!(sdd->port_conf->quirks & S3C64XX_SPI_QUIRK_CS_AUTO))
+		writel(S3C64XX_SPI_SLAVE_SIG_INACT, sdd->regs + S3C64XX_SPI_SLAVE_SEL);
+
+	/* Disable Interrupts - we use Polling if not DMA mode */
+	writel(0, regs + S3C64XX_SPI_INT_EN);
+
+	if (!sdd->port_conf->clk_from_cmu)
+		writel(sci->src_clk_nr << S3C64XX_SPI_CLKSEL_SRCSHFT,
+				regs + S3C64XX_SPI_CLK_CFG);
+	writel(0, regs + S3C64XX_SPI_MODE_CFG);
+	writel(0, regs + S3C64XX_SPI_PACKET_CNT);
+
+	/* Clear any irq pending bits, should set and clear the bits */
+	val = S3C64XX_SPI_PND_RX_OVERRUN_CLR |
+		S3C64XX_SPI_PND_RX_UNDERRUN_CLR |
+		S3C64XX_SPI_PND_TX_OVERRUN_CLR |
+		S3C64XX_SPI_PND_TX_UNDERRUN_CLR;
+	writel(val, regs + S3C64XX_SPI_PENDING_CLR);
+	writel(0, regs + S3C64XX_SPI_PENDING_CLR);
+
+	writel(0, regs + S3C64XX_SPI_SWAP_CFG);
+
+	val = readl(regs + S3C64XX_SPI_MODE_CFG);
+	val &= ~S3C64XX_SPI_MODE_4BURST;
+	val &= ~(S3C64XX_SPI_MAX_TRAILCNT << S3C64XX_SPI_TRAILCNT_OFF);
+	val |= (S3C64XX_SPI_TRAILCNT << S3C64XX_SPI_TRAILCNT_OFF);
+	writel(val, regs + S3C64XX_SPI_MODE_CFG);
+
+	s3c64xx_flush_fifo(sdd);
+}
+
+#ifdef CONFIG_OF
+static struct s3c64xx_spi_info *s3c64xx_spi_parse_dt(struct device *dev)
+{
+	struct s3c64xx_spi_info *sci;
+	u32 temp;
+
+	sci = devm_kzalloc(dev, sizeof(*sci), GFP_KERNEL);
+	if (!sci)
+		return ERR_PTR(-ENOMEM);
+
+	if (of_property_read_u32(dev->of_node, "samsung,spi-src-clk", &temp)) {
+		dev_warn(dev, "spi bus clock parent not specified, using clock at index 0 as parent\n");
+		sci->src_clk_nr = 0;
+	} else {
+		sci->src_clk_nr = temp;
+	}
+
+	if (of_property_read_u32(dev->of_node, "num-cs", &temp)) {
+		dev_warn(dev, "number of chip select lines not specified, assuming 1 chip select line\n");
+		sci->num_cs = 1;
+	} else {
+		sci->num_cs = temp;
+	}
+
+	sci->no_cs = of_property_read_bool(dev->of_node, "no-cs-readback");
+
+	return sci;
+}
+#else
+static struct s3c64xx_spi_info *s3c64xx_spi_parse_dt(struct device *dev)
+{
+	return dev_get_platdata(dev);
+}
+#endif
+
+static const struct of_device_id s3c64xx_spi_dt_match[];
+
+static inline struct s3c64xx_spi_port_config *s3c64xx_spi_get_port_config(
+						struct platform_device *pdev)
+{
+#ifdef CONFIG_OF
+	if (pdev->dev.of_node) {
+		const struct of_device_id *match;
+		match = of_match_node(s3c64xx_spi_dt_match, pdev->dev.of_node);
+		return (struct s3c64xx_spi_port_config *)match->data;
+	}
+#endif
+	return (struct s3c64xx_spi_port_config *)
+			 platform_get_device_id(pdev)->driver_data;
+}
+
+static int s3c64xx_spi_probe(struct platform_device *pdev)
+{
+	struct resource	*mem_res;
+	struct s3c64xx_spi_driver_data *sdd;
+	struct s3c64xx_spi_info *sci = dev_get_platdata(&pdev->dev);
+	struct spi_master *master;
+	int ret, irq;
+	char clk_name[16];
+
+	if (!sci && pdev->dev.of_node) {
+		sci = s3c64xx_spi_parse_dt(&pdev->dev);
+		if (IS_ERR(sci))
+			return PTR_ERR(sci);
+	}
+
+	if (!sci) {
+		dev_err(&pdev->dev, "platform_data missing!\n");
+		return -ENODEV;
+	}
+
+	mem_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (mem_res == NULL) {
+		dev_err(&pdev->dev, "Unable to get SPI MEM resource\n");
+		return -ENXIO;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_warn(&pdev->dev, "Failed to get IRQ: %d\n", irq);
+		return irq;
+	}
+
+	master = spi_alloc_master(&pdev->dev,
+				sizeof(struct s3c64xx_spi_driver_data));
+	if (master == NULL) {
+		dev_err(&pdev->dev, "Unable to allocate SPI Master\n");
+		return -ENOMEM;
+	}
+
+	platform_set_drvdata(pdev, master);
+
+	sdd = spi_master_get_devdata(master);
+	sdd->port_conf = s3c64xx_spi_get_port_config(pdev);
+	sdd->master = master;
+	sdd->cntrlr_info = sci;
+	sdd->pdev = pdev;
+	sdd->sfr_start = mem_res->start;
+	if (pdev->dev.of_node) {
+		ret = of_alias_get_id(pdev->dev.of_node, "spi");
+		if (ret < 0) {
+			dev_err(&pdev->dev, "failed to get alias id, errno %d\n",
+				ret);
+			goto err_deref_master;
+		}
+		sdd->port_id = ret;
+	} else {
+		sdd->port_id = pdev->id;
+	}
+
+	sdd->cur_bpw = 8;
+
+	sdd->tx_dma.direction = DMA_MEM_TO_DEV;
+	sdd->rx_dma.direction = DMA_DEV_TO_MEM;
+
+	master->dev.of_node = pdev->dev.of_node;
+	master->bus_num = sdd->port_id;
+	master->setup = s3c64xx_spi_setup;
+	master->cleanup = s3c64xx_spi_cleanup;
+	master->prepare_transfer_hardware = s3c64xx_spi_prepare_transfer;
+	master->prepare_message = s3c64xx_spi_prepare_message;
+	master->transfer_one = s3c64xx_spi_transfer_one;
+	master->num_chipselect = sci->num_cs;
+	master->dma_alignment = 8;
+	master->bits_per_word_mask = SPI_BPW_MASK(32) | SPI_BPW_MASK(16) |
+					SPI_BPW_MASK(8);
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->auto_runtime_pm = true;
+	if (!is_polling(sdd))
+		master->can_dma = s3c64xx_spi_can_dma;
+
+	sdd->regs = devm_ioremap_resource(&pdev->dev, mem_res);
+	if (IS_ERR(sdd->regs)) {
+		ret = PTR_ERR(sdd->regs);
+		goto err_deref_master;
+	}
+
+	if (sci->cfg_gpio && sci->cfg_gpio()) {
+		dev_err(&pdev->dev, "Unable to config gpio\n");
+		ret = -EBUSY;
+		goto err_deref_master;
+	}
+
+	/* Setup clocks */
+	sdd->clk = devm_clk_get(&pdev->dev, "spi");
+	if (IS_ERR(sdd->clk)) {
+		dev_err(&pdev->dev, "Unable to acquire clock 'spi'\n");
+		ret = PTR_ERR(sdd->clk);
+		goto err_deref_master;
+	}
+
+	ret = clk_prepare_enable(sdd->clk);
+	if (ret) {
+		dev_err(&pdev->dev, "Couldn't enable clock 'spi'\n");
+		goto err_deref_master;
+	}
+
+	sprintf(clk_name, "spi_busclk%d", sci->src_clk_nr);
+	sdd->src_clk = devm_clk_get(&pdev->dev, clk_name);
+	if (IS_ERR(sdd->src_clk)) {
+		dev_err(&pdev->dev,
+			"Unable to acquire clock '%s'\n", clk_name);
+		ret = PTR_ERR(sdd->src_clk);
+		goto err_disable_clk;
+	}
+
+	ret = clk_prepare_enable(sdd->src_clk);
+	if (ret) {
+		dev_err(&pdev->dev, "Couldn't enable clock '%s'\n", clk_name);
+		goto err_disable_clk;
+	}
+
+	if (sdd->port_conf->clk_ioclk) {
+		sdd->ioclk = devm_clk_get(&pdev->dev, "spi_ioclk");
+		if (IS_ERR(sdd->ioclk)) {
+			dev_err(&pdev->dev, "Unable to acquire 'ioclk'\n");
+			ret = PTR_ERR(sdd->ioclk);
+			goto err_disable_src_clk;
+		}
+
+		ret = clk_prepare_enable(sdd->ioclk);
+		if (ret) {
+			dev_err(&pdev->dev, "Couldn't enable clock 'ioclk'\n");
+			goto err_disable_src_clk;
+		}
+	}
+
+	if (!is_polling(sdd)) {
+		/* Acquire DMA channels */
+		sdd->rx_dma.ch = dma_request_slave_channel_reason(&pdev->dev,
+								  "rx");
+		if (IS_ERR(sdd->rx_dma.ch)) {
+			dev_err(&pdev->dev, "Failed to get RX DMA channel\n");
+			ret = PTR_ERR(sdd->rx_dma.ch);
+			goto err_disable_io_clk;
+		}
+		sdd->tx_dma.ch = dma_request_slave_channel_reason(&pdev->dev,
+								  "tx");
+		if (IS_ERR(sdd->tx_dma.ch)) {
+			dev_err(&pdev->dev, "Failed to get TX DMA channel\n");
+			ret = PTR_ERR(sdd->tx_dma.ch);
+			goto err_release_rx_dma;
+		}
+	}
+
+	pm_runtime_set_autosuspend_delay(&pdev->dev, AUTOSUSPEND_TIMEOUT);
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+	pm_runtime_get_sync(&pdev->dev);
+
+	/* Setup Deufult Mode */
+	s3c64xx_spi_hwinit(sdd);
+
+	spin_lock_init(&sdd->lock);
+	init_completion(&sdd->xfer_completion);
+
+	ret = devm_request_irq(&pdev->dev, irq, s3c64xx_spi_irq, 0,
+				"spi-s3c64xx", sdd);
+	if (ret != 0) {
+		dev_err(&pdev->dev, "Failed to request IRQ %d: %d\n",
+			irq, ret);
+		goto err_pm_put;
+	}
+
+	writel(S3C64XX_SPI_INT_RX_OVERRUN_EN | S3C64XX_SPI_INT_RX_UNDERRUN_EN |
+	       S3C64XX_SPI_INT_TX_OVERRUN_EN | S3C64XX_SPI_INT_TX_UNDERRUN_EN,
+	       sdd->regs + S3C64XX_SPI_INT_EN);
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret != 0) {
+		dev_err(&pdev->dev, "cannot register SPI master: %d\n", ret);
+		goto err_pm_put;
+	}
+
+	dev_dbg(&pdev->dev, "Samsung SoC SPI Driver loaded for Bus SPI-%d with %d Slaves attached\n",
+					sdd->port_id, master->num_chipselect);
+	dev_dbg(&pdev->dev, "\tIOmem=[%pR]\tFIFO %dbytes\n",
+					mem_res, (FIFO_LVL_MASK(sdd) >> 1) + 1);
+
+	pm_runtime_mark_last_busy(&pdev->dev);
+	pm_runtime_put_autosuspend(&pdev->dev);
+
+	return 0;
+
+err_pm_put:
+	pm_runtime_put_noidle(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+	pm_runtime_set_suspended(&pdev->dev);
+
+	if (!is_polling(sdd))
+		dma_release_channel(sdd->tx_dma.ch);
+err_release_rx_dma:
+	if (!is_polling(sdd))
+		dma_release_channel(sdd->rx_dma.ch);
+err_disable_io_clk:
+	clk_disable_unprepare(sdd->ioclk);
+err_disable_src_clk:
+	clk_disable_unprepare(sdd->src_clk);
+err_disable_clk:
+	clk_disable_unprepare(sdd->clk);
+err_deref_master:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int s3c64xx_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct s3c64xx_spi_driver_data *sdd = spi_master_get_devdata(master);
+
+	pm_runtime_get_sync(&pdev->dev);
+
+	writel(0, sdd->regs + S3C64XX_SPI_INT_EN);
+
+	if (!is_polling(sdd)) {
+		dma_release_channel(sdd->rx_dma.ch);
+		dma_release_channel(sdd->tx_dma.ch);
+	}
+
+	clk_disable_unprepare(sdd->ioclk);
+
+	clk_disable_unprepare(sdd->src_clk);
+
+	clk_disable_unprepare(sdd->clk);
+
+	pm_runtime_put_noidle(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+	pm_runtime_set_suspended(&pdev->dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int s3c64xx_spi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct s3c64xx_spi_driver_data *sdd = spi_master_get_devdata(master);
+
+	int ret = spi_master_suspend(master);
+	if (ret)
+		return ret;
+
+	ret = pm_runtime_force_suspend(dev);
+	if (ret < 0)
+		return ret;
+
+	sdd->cur_speed = 0; /* Output Clock is stopped */
+
+	return 0;
+}
+
+static int s3c64xx_spi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct s3c64xx_spi_driver_data *sdd = spi_master_get_devdata(master);
+	struct s3c64xx_spi_info *sci = sdd->cntrlr_info;
+	int ret;
+
+	if (sci->cfg_gpio)
+		sci->cfg_gpio();
+
+	ret = pm_runtime_force_resume(dev);
+	if (ret < 0)
+		return ret;
+
+	return spi_master_resume(master);
+}
+#endif /* CONFIG_PM_SLEEP */
+
+#ifdef CONFIG_PM
+static int s3c64xx_spi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct s3c64xx_spi_driver_data *sdd = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(sdd->clk);
+	clk_disable_unprepare(sdd->src_clk);
+	clk_disable_unprepare(sdd->ioclk);
+
+	return 0;
+}
+
+static int s3c64xx_spi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct s3c64xx_spi_driver_data *sdd = spi_master_get_devdata(master);
+	int ret;
+
+	if (sdd->port_conf->clk_ioclk) {
+		ret = clk_prepare_enable(sdd->ioclk);
+		if (ret != 0)
+			return ret;
+	}
+
+	ret = clk_prepare_enable(sdd->src_clk);
+	if (ret != 0)
+		goto err_disable_ioclk;
+
+	ret = clk_prepare_enable(sdd->clk);
+	if (ret != 0)
+		goto err_disable_src_clk;
+
+	s3c64xx_spi_hwinit(sdd);
+
+	return 0;
+
+err_disable_src_clk:
+	clk_disable_unprepare(sdd->src_clk);
+err_disable_ioclk:
+	clk_disable_unprepare(sdd->ioclk);
+
+	return ret;
+}
+#endif /* CONFIG_PM */
+
+static const struct dev_pm_ops s3c64xx_spi_pm = {
+	SET_SYSTEM_SLEEP_PM_OPS(s3c64xx_spi_suspend, s3c64xx_spi_resume)
+	SET_RUNTIME_PM_OPS(s3c64xx_spi_runtime_suspend,
+			   s3c64xx_spi_runtime_resume, NULL)
+};
+
+static struct s3c64xx_spi_port_config s3c2443_spi_port_config = {
+	.fifo_lvl_mask	= { 0x7f },
+	.rx_lvl_offset	= 13,
+	.tx_st_done	= 21,
+	.high_speed	= true,
+};
+
+static struct s3c64xx_spi_port_config s3c6410_spi_port_config = {
+	.fifo_lvl_mask	= { 0x7f, 0x7F },
+	.rx_lvl_offset	= 13,
+	.tx_st_done	= 21,
+};
+
+static struct s3c64xx_spi_port_config s5pv210_spi_port_config = {
+	.fifo_lvl_mask	= { 0x1ff, 0x7F },
+	.rx_lvl_offset	= 15,
+	.tx_st_done	= 25,
+	.high_speed	= true,
+};
+
+static struct s3c64xx_spi_port_config exynos4_spi_port_config = {
+	.fifo_lvl_mask	= { 0x1ff, 0x7F, 0x7F },
+	.rx_lvl_offset	= 15,
+	.tx_st_done	= 25,
+	.high_speed	= true,
+	.clk_from_cmu	= true,
+};
+
+static struct s3c64xx_spi_port_config exynos7_spi_port_config = {
+	.fifo_lvl_mask	= { 0x1ff, 0x7F, 0x7F, 0x7F, 0x7F, 0x1ff},
+	.rx_lvl_offset	= 15,
+	.tx_st_done	= 25,
+	.high_speed	= true,
+	.clk_from_cmu	= true,
+	.quirks		= S3C64XX_SPI_QUIRK_CS_AUTO,
+};
+
+static struct s3c64xx_spi_port_config exynos5433_spi_port_config = {
+	.fifo_lvl_mask	= { 0x1ff, 0x7f, 0x7f, 0x7f, 0x7f, 0x1ff},
+	.rx_lvl_offset	= 15,
+	.tx_st_done	= 25,
+	.high_speed	= true,
+	.clk_from_cmu	= true,
+	.clk_ioclk	= true,
+	.quirks		= S3C64XX_SPI_QUIRK_CS_AUTO,
+};
+
+static const struct platform_device_id s3c64xx_spi_driver_ids[] = {
+	{
+		.name		= "s3c2443-spi",
+		.driver_data	= (kernel_ulong_t)&s3c2443_spi_port_config,
+	}, {
+		.name		= "s3c6410-spi",
+		.driver_data	= (kernel_ulong_t)&s3c6410_spi_port_config,
+	},
+	{ },
+};
+
+static const struct of_device_id s3c64xx_spi_dt_match[] = {
+	{ .compatible = "samsung,s3c2443-spi",
+			.data = (void *)&s3c2443_spi_port_config,
+	},
+	{ .compatible = "samsung,s3c6410-spi",
+			.data = (void *)&s3c6410_spi_port_config,
+	},
+	{ .compatible = "samsung,s5pv210-spi",
+			.data = (void *)&s5pv210_spi_port_config,
+	},
+	{ .compatible = "samsung,exynos4210-spi",
+			.data = (void *)&exynos4_spi_port_config,
+	},
+	{ .compatible = "samsung,exynos7-spi",
+			.data = (void *)&exynos7_spi_port_config,
+	},
+	{ .compatible = "samsung,exynos5433-spi",
+			.data = (void *)&exynos5433_spi_port_config,
+	},
+	{ },
+};
+MODULE_DEVICE_TABLE(of, s3c64xx_spi_dt_match);
+
+static struct platform_driver s3c64xx_spi_driver = {
+	.driver = {
+		.name	= "s3c64xx-spi",
+		.pm = &s3c64xx_spi_pm,
+		.of_match_table = of_match_ptr(s3c64xx_spi_dt_match),
+	},
+	.probe = s3c64xx_spi_probe,
+	.remove = s3c64xx_spi_remove,
+	.id_table = s3c64xx_spi_driver_ids,
+};
+MODULE_ALIAS("platform:s3c64xx-spi");
+
+module_platform_driver(s3c64xx_spi_driver);
+
+MODULE_AUTHOR("Jaswinder Singh <jassi.brar@samsung.com>");
+MODULE_DESCRIPTION("S3C64XX SPI Controller Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-sc18is602.c b/drivers/spi/spi-sc18is602.c
new file mode 100644
index 0000000..52cf0e9
--- /dev/null
+++ b/drivers/spi/spi-sc18is602.c
@@ -0,0 +1,360 @@
+/*
+ * NXP SC18IS602/603 SPI driver
+ *
+ * Copyright (C) Guenter Roeck <linux@roeck-us.net>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/err.h>
+#include <linux/module.h>
+#include <linux/spi/spi.h>
+#include <linux/i2c.h>
+#include <linux/delay.h>
+#include <linux/pm_runtime.h>
+#include <linux/of_device.h>
+#include <linux/of.h>
+#include <linux/platform_data/sc18is602.h>
+#include <linux/gpio/consumer.h>
+
+enum chips { sc18is602, sc18is602b, sc18is603 };
+
+#define SC18IS602_BUFSIZ		200
+#define SC18IS602_CLOCK			7372000
+
+#define SC18IS602_MODE_CPHA		BIT(2)
+#define SC18IS602_MODE_CPOL		BIT(3)
+#define SC18IS602_MODE_LSB_FIRST	BIT(5)
+#define SC18IS602_MODE_CLOCK_DIV_4	0x0
+#define SC18IS602_MODE_CLOCK_DIV_16	0x1
+#define SC18IS602_MODE_CLOCK_DIV_64	0x2
+#define SC18IS602_MODE_CLOCK_DIV_128	0x3
+
+struct sc18is602 {
+	struct spi_master	*master;
+	struct device		*dev;
+	u8			ctrl;
+	u32			freq;
+	u32			speed;
+
+	/* I2C data */
+	struct i2c_client	*client;
+	enum chips		id;
+	u8			buffer[SC18IS602_BUFSIZ + 1];
+	int			tlen;	/* Data queued for tx in buffer */
+	int			rindex;	/* Receive data index in buffer */
+
+	struct gpio_desc	*reset;
+};
+
+static int sc18is602_wait_ready(struct sc18is602 *hw, int len)
+{
+	int i, err;
+	int usecs = 1000000 * len / hw->speed + 1;
+	u8 dummy[1];
+
+	for (i = 0; i < 10; i++) {
+		err = i2c_master_recv(hw->client, dummy, 1);
+		if (err >= 0)
+			return 0;
+		usleep_range(usecs, usecs * 2);
+	}
+	return -ETIMEDOUT;
+}
+
+static int sc18is602_txrx(struct sc18is602 *hw, struct spi_message *msg,
+			  struct spi_transfer *t, bool do_transfer)
+{
+	unsigned int len = t->len;
+	int ret;
+
+	if (hw->tlen == 0) {
+		/* First byte (I2C command) is chip select */
+		hw->buffer[0] = 1 << msg->spi->chip_select;
+		hw->tlen = 1;
+		hw->rindex = 0;
+	}
+	/*
+	 * We can not immediately send data to the chip, since each I2C message
+	 * resembles a full SPI message (from CS active to CS inactive).
+	 * Enqueue messages up to the first read or until do_transfer is true.
+	 */
+	if (t->tx_buf) {
+		memcpy(&hw->buffer[hw->tlen], t->tx_buf, len);
+		hw->tlen += len;
+		if (t->rx_buf)
+			do_transfer = true;
+		else
+			hw->rindex = hw->tlen - 1;
+	} else if (t->rx_buf) {
+		/*
+		 * For receive-only transfers we still need to perform a dummy
+		 * write to receive data from the SPI chip.
+		 * Read data starts at the end of transmit data (minus 1 to
+		 * account for CS).
+		 */
+		hw->rindex = hw->tlen - 1;
+		memset(&hw->buffer[hw->tlen], 0, len);
+		hw->tlen += len;
+		do_transfer = true;
+	}
+
+	if (do_transfer && hw->tlen > 1) {
+		ret = sc18is602_wait_ready(hw, SC18IS602_BUFSIZ);
+		if (ret < 0)
+			return ret;
+		ret = i2c_master_send(hw->client, hw->buffer, hw->tlen);
+		if (ret < 0)
+			return ret;
+		if (ret != hw->tlen)
+			return -EIO;
+
+		if (t->rx_buf) {
+			int rlen = hw->rindex + len;
+
+			ret = sc18is602_wait_ready(hw, hw->tlen);
+			if (ret < 0)
+				return ret;
+			ret = i2c_master_recv(hw->client, hw->buffer, rlen);
+			if (ret < 0)
+				return ret;
+			if (ret != rlen)
+				return -EIO;
+			memcpy(t->rx_buf, &hw->buffer[hw->rindex], len);
+		}
+		hw->tlen = 0;
+	}
+	return len;
+}
+
+static int sc18is602_setup_transfer(struct sc18is602 *hw, u32 hz, u8 mode)
+{
+	u8 ctrl = 0;
+	int ret;
+
+	if (mode & SPI_CPHA)
+		ctrl |= SC18IS602_MODE_CPHA;
+	if (mode & SPI_CPOL)
+		ctrl |= SC18IS602_MODE_CPOL;
+	if (mode & SPI_LSB_FIRST)
+		ctrl |= SC18IS602_MODE_LSB_FIRST;
+
+	/* Find the closest clock speed */
+	if (hz >= hw->freq / 4) {
+		ctrl |= SC18IS602_MODE_CLOCK_DIV_4;
+		hw->speed = hw->freq / 4;
+	} else if (hz >= hw->freq / 16) {
+		ctrl |= SC18IS602_MODE_CLOCK_DIV_16;
+		hw->speed = hw->freq / 16;
+	} else if (hz >= hw->freq / 64) {
+		ctrl |= SC18IS602_MODE_CLOCK_DIV_64;
+		hw->speed = hw->freq / 64;
+	} else {
+		ctrl |= SC18IS602_MODE_CLOCK_DIV_128;
+		hw->speed = hw->freq / 128;
+	}
+
+	/*
+	 * Don't do anything if the control value did not change. The initial
+	 * value of 0xff for hw->ctrl ensures that the correct mode will be set
+	 * with the first call to this function.
+	 */
+	if (ctrl == hw->ctrl)
+		return 0;
+
+	ret = i2c_smbus_write_byte_data(hw->client, 0xf0, ctrl);
+	if (ret < 0)
+		return ret;
+
+	hw->ctrl = ctrl;
+
+	return 0;
+}
+
+static int sc18is602_check_transfer(struct spi_device *spi,
+				    struct spi_transfer *t, int tlen)
+{
+	if (t && t->len + tlen > SC18IS602_BUFSIZ)
+		return -EINVAL;
+
+	return 0;
+}
+
+static int sc18is602_transfer_one(struct spi_master *master,
+				  struct spi_message *m)
+{
+	struct sc18is602 *hw = spi_master_get_devdata(master);
+	struct spi_device *spi = m->spi;
+	struct spi_transfer *t;
+	int status = 0;
+
+	hw->tlen = 0;
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		bool do_transfer;
+
+		status = sc18is602_check_transfer(spi, t, hw->tlen);
+		if (status < 0)
+			break;
+
+		status = sc18is602_setup_transfer(hw, t->speed_hz, spi->mode);
+		if (status < 0)
+			break;
+
+		do_transfer = t->cs_change || list_is_last(&t->transfer_list,
+							   &m->transfers);
+
+		if (t->len) {
+			status = sc18is602_txrx(hw, m, t, do_transfer);
+			if (status < 0)
+				break;
+			m->actual_length += status;
+		}
+		status = 0;
+
+		if (t->delay_usecs)
+			udelay(t->delay_usecs);
+	}
+	m->status = status;
+	spi_finalize_current_message(master);
+
+	return status;
+}
+
+static int sc18is602_setup(struct spi_device *spi)
+{
+	struct sc18is602 *hw = spi_master_get_devdata(spi->master);
+
+	/* SC18IS602 does not support CS2 */
+	if (hw->id == sc18is602 && spi->chip_select == 2)
+		return -ENXIO;
+
+	return 0;
+}
+
+static int sc18is602_probe(struct i2c_client *client,
+			   const struct i2c_device_id *id)
+{
+	struct device *dev = &client->dev;
+	struct device_node *np = dev->of_node;
+	struct sc18is602_platform_data *pdata = dev_get_platdata(dev);
+	struct sc18is602 *hw;
+	struct spi_master *master;
+	int error;
+
+	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C |
+				     I2C_FUNC_SMBUS_WRITE_BYTE_DATA))
+		return -EINVAL;
+
+	master = spi_alloc_master(dev, sizeof(struct sc18is602));
+	if (!master)
+		return -ENOMEM;
+
+	hw = spi_master_get_devdata(master);
+	i2c_set_clientdata(client, hw);
+
+	/* assert reset and then release */
+	hw->reset = devm_gpiod_get_optional(dev, "reset", GPIOD_OUT_HIGH);
+	if (IS_ERR(hw->reset))
+		return PTR_ERR(hw->reset);
+	gpiod_set_value_cansleep(hw->reset, 0);
+
+	hw->master = master;
+	hw->client = client;
+	hw->dev = dev;
+	hw->ctrl = 0xff;
+
+	if (client->dev.of_node)
+		hw->id = (enum chips)of_device_get_match_data(&client->dev);
+	else
+		hw->id = id->driver_data;
+
+	switch (hw->id) {
+	case sc18is602:
+	case sc18is602b:
+		master->num_chipselect = 4;
+		hw->freq = SC18IS602_CLOCK;
+		break;
+	case sc18is603:
+		master->num_chipselect = 2;
+		if (pdata) {
+			hw->freq = pdata->clock_frequency;
+		} else {
+			const __be32 *val;
+			int len;
+
+			val = of_get_property(np, "clock-frequency", &len);
+			if (val && len >= sizeof(__be32))
+				hw->freq = be32_to_cpup(val);
+		}
+		if (!hw->freq)
+			hw->freq = SC18IS602_CLOCK;
+		break;
+	}
+	master->bus_num = np ? -1 : client->adapter->nr;
+	master->mode_bits = SPI_CPHA | SPI_CPOL | SPI_LSB_FIRST;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->setup = sc18is602_setup;
+	master->transfer_one_message = sc18is602_transfer_one;
+	master->dev.of_node = np;
+	master->min_speed_hz = hw->freq / 128;
+	master->max_speed_hz = hw->freq / 4;
+
+	error = devm_spi_register_master(dev, master);
+	if (error)
+		goto error_reg;
+
+	return 0;
+
+error_reg:
+	spi_master_put(master);
+	return error;
+}
+
+static const struct i2c_device_id sc18is602_id[] = {
+	{ "sc18is602", sc18is602 },
+	{ "sc18is602b", sc18is602b },
+	{ "sc18is603", sc18is603 },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, sc18is602_id);
+
+static const struct of_device_id sc18is602_of_match[] = {
+	{
+		.compatible = "nxp,sc18is602",
+		.data = (void *)sc18is602
+	},
+	{
+		.compatible = "nxp,sc18is602b",
+		.data = (void *)sc18is602b
+	},
+	{
+		.compatible = "nxp,sc18is603",
+		.data = (void *)sc18is603
+	},
+	{ },
+};
+MODULE_DEVICE_TABLE(of, sc18is602_of_match);
+
+static struct i2c_driver sc18is602_driver = {
+	.driver = {
+		.name = "sc18is602",
+		.of_match_table = of_match_ptr(sc18is602_of_match),
+	},
+	.probe = sc18is602_probe,
+	.id_table = sc18is602_id,
+};
+
+module_i2c_driver(sc18is602_driver);
+
+MODULE_DESCRIPTION("SC18IC602/603 SPI Master Driver");
+MODULE_AUTHOR("Guenter Roeck");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-sh-hspi.c b/drivers/spi/spi-sh-hspi.c
new file mode 100644
index 0000000..20e800e
--- /dev/null
+++ b/drivers/spi/spi-sh-hspi.c
@@ -0,0 +1,321 @@
+/*
+ * SuperH HSPI bus driver
+ *
+ * Copyright (C) 2011  Kuninori Morimoto
+ *
+ * Based on spi-sh.c:
+ * Based on pxa2xx_spi.c:
+ * Copyright (C) 2011 Renesas Solutions Corp.
+ * Copyright (C) 2005 Stephen Street / StreetFire Sound Labs
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/clk.h>
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/timer.h>
+#include <linux/delay.h>
+#include <linux/list.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/io.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/sh_hspi.h>
+
+#define SPCR	0x00
+#define SPSR	0x04
+#define SPSCR	0x08
+#define SPTBR	0x0C
+#define SPRBR	0x10
+#define SPCR2	0x14
+
+/* SPSR */
+#define RXFL	(1 << 2)
+
+struct hspi_priv {
+	void __iomem *addr;
+	struct spi_master *master;
+	struct device *dev;
+	struct clk *clk;
+};
+
+/*
+ *		basic function
+ */
+static void hspi_write(struct hspi_priv *hspi, int reg, u32 val)
+{
+	iowrite32(val, hspi->addr + reg);
+}
+
+static u32 hspi_read(struct hspi_priv *hspi, int reg)
+{
+	return ioread32(hspi->addr + reg);
+}
+
+static void hspi_bit_set(struct hspi_priv *hspi, int reg, u32 mask, u32 set)
+{
+	u32 val = hspi_read(hspi, reg);
+
+	val &= ~mask;
+	val |= set & mask;
+
+	hspi_write(hspi, reg, val);
+}
+
+/*
+ *		transfer function
+ */
+static int hspi_status_check_timeout(struct hspi_priv *hspi, u32 mask, u32 val)
+{
+	int t = 256;
+
+	while (t--) {
+		if ((mask & hspi_read(hspi, SPSR)) == val)
+			return 0;
+
+		udelay(10);
+	}
+
+	dev_err(hspi->dev, "timeout\n");
+	return -ETIMEDOUT;
+}
+
+/*
+ *		spi master function
+ */
+
+#define hspi_hw_cs_enable(hspi)		hspi_hw_cs_ctrl(hspi, 0)
+#define hspi_hw_cs_disable(hspi)	hspi_hw_cs_ctrl(hspi, 1)
+static void hspi_hw_cs_ctrl(struct hspi_priv *hspi, int hi)
+{
+	hspi_bit_set(hspi, SPSCR, (1 << 6), (hi) << 6);
+}
+
+static void hspi_hw_setup(struct hspi_priv *hspi,
+			  struct spi_message *msg,
+			  struct spi_transfer *t)
+{
+	struct spi_device *spi = msg->spi;
+	struct device *dev = hspi->dev;
+	u32 spcr, idiv_clk;
+	u32 rate, best_rate, min, tmp;
+
+	/*
+	 * find best IDIV/CLKCx settings
+	 */
+	min = ~0;
+	best_rate = 0;
+	spcr = 0;
+	for (idiv_clk = 0x00; idiv_clk <= 0x3F; idiv_clk++) {
+		rate = clk_get_rate(hspi->clk);
+
+		/* IDIV calculation */
+		if (idiv_clk & (1 << 5))
+			rate /= 128;
+		else
+			rate /= 16;
+
+		/* CLKCx calculation */
+		rate /= (((idiv_clk & 0x1F) + 1) * 2);
+
+		/* save best settings */
+		tmp = abs(t->speed_hz - rate);
+		if (tmp < min) {
+			min = tmp;
+			spcr = idiv_clk;
+			best_rate = rate;
+		}
+	}
+
+	if (spi->mode & SPI_CPHA)
+		spcr |= 1 << 7;
+	if (spi->mode & SPI_CPOL)
+		spcr |= 1 << 6;
+
+	dev_dbg(dev, "speed %d/%d\n", t->speed_hz, best_rate);
+
+	hspi_write(hspi, SPCR, spcr);
+	hspi_write(hspi, SPSR, 0x0);
+	hspi_write(hspi, SPSCR, 0x21);	/* master mode / CS control */
+}
+
+static int hspi_transfer_one_message(struct spi_master *master,
+				     struct spi_message *msg)
+{
+	struct hspi_priv *hspi = spi_master_get_devdata(master);
+	struct spi_transfer *t;
+	u32 tx;
+	u32 rx;
+	int ret, i;
+	unsigned int cs_change;
+	const int nsecs = 50;
+
+	dev_dbg(hspi->dev, "%s\n", __func__);
+
+	cs_change = 1;
+	ret = 0;
+	list_for_each_entry(t, &msg->transfers, transfer_list) {
+
+		if (cs_change) {
+			hspi_hw_setup(hspi, msg, t);
+			hspi_hw_cs_enable(hspi);
+			ndelay(nsecs);
+		}
+		cs_change = t->cs_change;
+
+		for (i = 0; i < t->len; i++) {
+
+			/* wait remains */
+			ret = hspi_status_check_timeout(hspi, 0x1, 0);
+			if (ret < 0)
+				break;
+
+			tx = 0;
+			if (t->tx_buf)
+				tx = (u32)((u8 *)t->tx_buf)[i];
+
+			hspi_write(hspi, SPTBR, tx);
+
+			/* wait receive */
+			ret = hspi_status_check_timeout(hspi, 0x4, 0x4);
+			if (ret < 0)
+				break;
+
+			rx = hspi_read(hspi, SPRBR);
+			if (t->rx_buf)
+				((u8 *)t->rx_buf)[i] = (u8)rx;
+
+		}
+
+		msg->actual_length += t->len;
+
+		if (t->delay_usecs)
+			udelay(t->delay_usecs);
+
+		if (cs_change) {
+			ndelay(nsecs);
+			hspi_hw_cs_disable(hspi);
+			ndelay(nsecs);
+		}
+	}
+
+	msg->status = ret;
+	if (!cs_change) {
+		ndelay(nsecs);
+		hspi_hw_cs_disable(hspi);
+	}
+	spi_finalize_current_message(master);
+
+	return ret;
+}
+
+static int hspi_probe(struct platform_device *pdev)
+{
+	struct resource *res;
+	struct spi_master *master;
+	struct hspi_priv *hspi;
+	struct clk *clk;
+	int ret;
+
+	/* get base addr */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res) {
+		dev_err(&pdev->dev, "invalid resource\n");
+		return -EINVAL;
+	}
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*hspi));
+	if (!master) {
+		dev_err(&pdev->dev, "spi_alloc_master error.\n");
+		return -ENOMEM;
+	}
+
+	clk = clk_get(&pdev->dev, NULL);
+	if (IS_ERR(clk)) {
+		dev_err(&pdev->dev, "couldn't get clock\n");
+		ret = -EINVAL;
+		goto error0;
+	}
+
+	hspi = spi_master_get_devdata(master);
+	platform_set_drvdata(pdev, hspi);
+
+	/* init hspi */
+	hspi->master	= master;
+	hspi->dev	= &pdev->dev;
+	hspi->clk	= clk;
+	hspi->addr	= devm_ioremap(hspi->dev,
+				       res->start, resource_size(res));
+	if (!hspi->addr) {
+		dev_err(&pdev->dev, "ioremap error.\n");
+		ret = -ENOMEM;
+		goto error1;
+	}
+
+	pm_runtime_enable(&pdev->dev);
+
+	master->bus_num		= pdev->id;
+	master->mode_bits	= SPI_CPOL | SPI_CPHA;
+	master->dev.of_node	= pdev->dev.of_node;
+	master->auto_runtime_pm = true;
+	master->transfer_one_message		= hspi_transfer_one_message;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "spi_register_master error.\n");
+		goto error2;
+	}
+
+	return 0;
+
+ error2:
+	pm_runtime_disable(&pdev->dev);
+ error1:
+	clk_put(clk);
+ error0:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int hspi_remove(struct platform_device *pdev)
+{
+	struct hspi_priv *hspi = platform_get_drvdata(pdev);
+
+	pm_runtime_disable(&pdev->dev);
+
+	clk_put(hspi->clk);
+
+	return 0;
+}
+
+static const struct of_device_id hspi_of_match[] = {
+	{ .compatible = "renesas,hspi", },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, hspi_of_match);
+
+static struct platform_driver hspi_driver = {
+	.probe = hspi_probe,
+	.remove = hspi_remove,
+	.driver = {
+		.name = "sh-hspi",
+		.of_match_table = hspi_of_match,
+	},
+};
+module_platform_driver(hspi_driver);
+
+MODULE_DESCRIPTION("SuperH HSPI bus driver");
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>");
+MODULE_ALIAS("platform:sh-hspi");
diff --git a/drivers/spi/spi-sh-msiof.c b/drivers/spi/spi-sh-msiof.c
new file mode 100644
index 0000000..101cd6a
--- /dev/null
+++ b/drivers/spi/spi-sh-msiof.c
@@ -0,0 +1,1469 @@
+/*
+ * SuperH MSIOF SPI Master Interface
+ *
+ * Copyright (c) 2009 Magnus Damm
+ * Copyright (C) 2014 Renesas Electronics Corporation
+ * Copyright (C) 2014-2017 Glider bvba
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/bitmap.h>
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/err.h>
+#include <linux/gpio.h>
+#include <linux/gpio/consumer.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/sh_dma.h>
+
+#include <linux/spi/sh_msiof.h>
+#include <linux/spi/spi.h>
+
+#include <asm/unaligned.h>
+
+struct sh_msiof_chipdata {
+	u16 tx_fifo_size;
+	u16 rx_fifo_size;
+	u16 master_flags;
+	u16 min_div_pow;
+};
+
+struct sh_msiof_spi_priv {
+	struct spi_master *master;
+	void __iomem *mapbase;
+	struct clk *clk;
+	struct platform_device *pdev;
+	struct sh_msiof_spi_info *info;
+	struct completion done;
+	struct completion done_txdma;
+	unsigned int tx_fifo_size;
+	unsigned int rx_fifo_size;
+	unsigned int min_div_pow;
+	void *tx_dma_page;
+	void *rx_dma_page;
+	dma_addr_t tx_dma_addr;
+	dma_addr_t rx_dma_addr;
+	unsigned short unused_ss;
+	bool native_cs_inited;
+	bool native_cs_high;
+	bool slave_aborted;
+};
+
+#define MAX_SS	3	/* Maximum number of native chip selects */
+
+#define TMDR1	0x00	/* Transmit Mode Register 1 */
+#define TMDR2	0x04	/* Transmit Mode Register 2 */
+#define TMDR3	0x08	/* Transmit Mode Register 3 */
+#define RMDR1	0x10	/* Receive Mode Register 1 */
+#define RMDR2	0x14	/* Receive Mode Register 2 */
+#define RMDR3	0x18	/* Receive Mode Register 3 */
+#define TSCR	0x20	/* Transmit Clock Select Register */
+#define RSCR	0x22	/* Receive Clock Select Register (SH, A1, APE6) */
+#define CTR	0x28	/* Control Register */
+#define FCTR	0x30	/* FIFO Control Register */
+#define STR	0x40	/* Status Register */
+#define IER	0x44	/* Interrupt Enable Register */
+#define TDR1	0x48	/* Transmit Control Data Register 1 (SH, A1) */
+#define TDR2	0x4c	/* Transmit Control Data Register 2 (SH, A1) */
+#define TFDR	0x50	/* Transmit FIFO Data Register */
+#define RDR1	0x58	/* Receive Control Data Register 1 (SH, A1) */
+#define RDR2	0x5c	/* Receive Control Data Register 2 (SH, A1) */
+#define RFDR	0x60	/* Receive FIFO Data Register */
+
+/* TMDR1 and RMDR1 */
+#define MDR1_TRMD	 0x80000000 /* Transfer Mode (1 = Master mode) */
+#define MDR1_SYNCMD_MASK 0x30000000 /* SYNC Mode */
+#define MDR1_SYNCMD_SPI	 0x20000000 /*   Level mode/SPI */
+#define MDR1_SYNCMD_LR	 0x30000000 /*   L/R mode */
+#define MDR1_SYNCAC_SHIFT	 25 /* Sync Polarity (1 = Active-low) */
+#define MDR1_BITLSB_SHIFT	 24 /* MSB/LSB First (1 = LSB first) */
+#define MDR1_DTDL_SHIFT		 20 /* Data Pin Bit Delay for MSIOF_SYNC */
+#define MDR1_SYNCDL_SHIFT	 16 /* Frame Sync Signal Timing Delay */
+#define MDR1_FLD_MASK	 0x0000000c /* Frame Sync Signal Interval (0-3) */
+#define MDR1_FLD_SHIFT		  2
+#define MDR1_XXSTP	 0x00000001 /* Transmission/Reception Stop on FIFO */
+/* TMDR1 */
+#define TMDR1_PCON	 0x40000000 /* Transfer Signal Connection */
+#define TMDR1_SYNCCH_MASK 0xc000000 /* Synchronization Signal Channel Select */
+#define TMDR1_SYNCCH_SHIFT	 26 /* 0=MSIOF_SYNC, 1=MSIOF_SS1, 2=MSIOF_SS2 */
+
+/* TMDR2 and RMDR2 */
+#define MDR2_BITLEN1(i)	(((i) - 1) << 24) /* Data Size (8-32 bits) */
+#define MDR2_WDLEN1(i)	(((i) - 1) << 16) /* Word Count (1-64/256 (SH, A1))) */
+#define MDR2_GRPMASK1	0x00000001 /* Group Output Mask 1 (SH, A1) */
+
+/* TSCR and RSCR */
+#define SCR_BRPS_MASK	    0x1f00 /* Prescaler Setting (1-32) */
+#define SCR_BRPS(i)	(((i) - 1) << 8)
+#define SCR_BRDV_MASK	    0x0007 /* Baud Rate Generator's Division Ratio */
+#define SCR_BRDV_DIV_2	    0x0000
+#define SCR_BRDV_DIV_4	    0x0001
+#define SCR_BRDV_DIV_8	    0x0002
+#define SCR_BRDV_DIV_16	    0x0003
+#define SCR_BRDV_DIV_32	    0x0004
+#define SCR_BRDV_DIV_1	    0x0007
+
+/* CTR */
+#define CTR_TSCKIZ_MASK	0xc0000000 /* Transmit Clock I/O Polarity Select */
+#define CTR_TSCKIZ_SCK	0x80000000 /*   Disable SCK when TX disabled */
+#define CTR_TSCKIZ_POL_SHIFT	30 /*   Transmit Clock Polarity */
+#define CTR_RSCKIZ_MASK	0x30000000 /* Receive Clock Polarity Select */
+#define CTR_RSCKIZ_SCK	0x20000000 /*   Must match CTR_TSCKIZ_SCK */
+#define CTR_RSCKIZ_POL_SHIFT	28 /*   Receive Clock Polarity */
+#define CTR_TEDG_SHIFT		27 /* Transmit Timing (1 = falling edge) */
+#define CTR_REDG_SHIFT		26 /* Receive Timing (1 = falling edge) */
+#define CTR_TXDIZ_MASK	0x00c00000 /* Pin Output When TX is Disabled */
+#define CTR_TXDIZ_LOW	0x00000000 /*   0 */
+#define CTR_TXDIZ_HIGH	0x00400000 /*   1 */
+#define CTR_TXDIZ_HIZ	0x00800000 /*   High-impedance */
+#define CTR_TSCKE	0x00008000 /* Transmit Serial Clock Output Enable */
+#define CTR_TFSE	0x00004000 /* Transmit Frame Sync Signal Output Enable */
+#define CTR_TXE		0x00000200 /* Transmit Enable */
+#define CTR_RXE		0x00000100 /* Receive Enable */
+
+/* FCTR */
+#define FCTR_TFWM_MASK	0xe0000000 /* Transmit FIFO Watermark */
+#define FCTR_TFWM_64	0x00000000 /*  Transfer Request when 64 empty stages */
+#define FCTR_TFWM_32	0x20000000 /*  Transfer Request when 32 empty stages */
+#define FCTR_TFWM_24	0x40000000 /*  Transfer Request when 24 empty stages */
+#define FCTR_TFWM_16	0x60000000 /*  Transfer Request when 16 empty stages */
+#define FCTR_TFWM_12	0x80000000 /*  Transfer Request when 12 empty stages */
+#define FCTR_TFWM_8	0xa0000000 /*  Transfer Request when 8 empty stages */
+#define FCTR_TFWM_4	0xc0000000 /*  Transfer Request when 4 empty stages */
+#define FCTR_TFWM_1	0xe0000000 /*  Transfer Request when 1 empty stage */
+#define FCTR_TFUA_MASK	0x07f00000 /* Transmit FIFO Usable Area */
+#define FCTR_TFUA_SHIFT		20
+#define FCTR_TFUA(i)	((i) << FCTR_TFUA_SHIFT)
+#define FCTR_RFWM_MASK	0x0000e000 /* Receive FIFO Watermark */
+#define FCTR_RFWM_1	0x00000000 /*  Transfer Request when 1 valid stages */
+#define FCTR_RFWM_4	0x00002000 /*  Transfer Request when 4 valid stages */
+#define FCTR_RFWM_8	0x00004000 /*  Transfer Request when 8 valid stages */
+#define FCTR_RFWM_16	0x00006000 /*  Transfer Request when 16 valid stages */
+#define FCTR_RFWM_32	0x00008000 /*  Transfer Request when 32 valid stages */
+#define FCTR_RFWM_64	0x0000a000 /*  Transfer Request when 64 valid stages */
+#define FCTR_RFWM_128	0x0000c000 /*  Transfer Request when 128 valid stages */
+#define FCTR_RFWM_256	0x0000e000 /*  Transfer Request when 256 valid stages */
+#define FCTR_RFUA_MASK	0x00001ff0 /* Receive FIFO Usable Area (0x40 = full) */
+#define FCTR_RFUA_SHIFT		 4
+#define FCTR_RFUA(i)	((i) << FCTR_RFUA_SHIFT)
+
+/* STR */
+#define STR_TFEMP	0x20000000 /* Transmit FIFO Empty */
+#define STR_TDREQ	0x10000000 /* Transmit Data Transfer Request */
+#define STR_TEOF	0x00800000 /* Frame Transmission End */
+#define STR_TFSERR	0x00200000 /* Transmit Frame Synchronization Error */
+#define STR_TFOVF	0x00100000 /* Transmit FIFO Overflow */
+#define STR_TFUDF	0x00080000 /* Transmit FIFO Underflow */
+#define STR_RFFUL	0x00002000 /* Receive FIFO Full */
+#define STR_RDREQ	0x00001000 /* Receive Data Transfer Request */
+#define STR_REOF	0x00000080 /* Frame Reception End */
+#define STR_RFSERR	0x00000020 /* Receive Frame Synchronization Error */
+#define STR_RFUDF	0x00000010 /* Receive FIFO Underflow */
+#define STR_RFOVF	0x00000008 /* Receive FIFO Overflow */
+
+/* IER */
+#define IER_TDMAE	0x80000000 /* Transmit Data DMA Transfer Req. Enable */
+#define IER_TFEMPE	0x20000000 /* Transmit FIFO Empty Enable */
+#define IER_TDREQE	0x10000000 /* Transmit Data Transfer Request Enable */
+#define IER_TEOFE	0x00800000 /* Frame Transmission End Enable */
+#define IER_TFSERRE	0x00200000 /* Transmit Frame Sync Error Enable */
+#define IER_TFOVFE	0x00100000 /* Transmit FIFO Overflow Enable */
+#define IER_TFUDFE	0x00080000 /* Transmit FIFO Underflow Enable */
+#define IER_RDMAE	0x00008000 /* Receive Data DMA Transfer Req. Enable */
+#define IER_RFFULE	0x00002000 /* Receive FIFO Full Enable */
+#define IER_RDREQE	0x00001000 /* Receive Data Transfer Request Enable */
+#define IER_REOFE	0x00000080 /* Frame Reception End Enable */
+#define IER_RFSERRE	0x00000020 /* Receive Frame Sync Error Enable */
+#define IER_RFUDFE	0x00000010 /* Receive FIFO Underflow Enable */
+#define IER_RFOVFE	0x00000008 /* Receive FIFO Overflow Enable */
+
+
+static u32 sh_msiof_read(struct sh_msiof_spi_priv *p, int reg_offs)
+{
+	switch (reg_offs) {
+	case TSCR:
+	case RSCR:
+		return ioread16(p->mapbase + reg_offs);
+	default:
+		return ioread32(p->mapbase + reg_offs);
+	}
+}
+
+static void sh_msiof_write(struct sh_msiof_spi_priv *p, int reg_offs,
+			   u32 value)
+{
+	switch (reg_offs) {
+	case TSCR:
+	case RSCR:
+		iowrite16(value, p->mapbase + reg_offs);
+		break;
+	default:
+		iowrite32(value, p->mapbase + reg_offs);
+		break;
+	}
+}
+
+static int sh_msiof_modify_ctr_wait(struct sh_msiof_spi_priv *p,
+				    u32 clr, u32 set)
+{
+	u32 mask = clr | set;
+	u32 data;
+	int k;
+
+	data = sh_msiof_read(p, CTR);
+	data &= ~clr;
+	data |= set;
+	sh_msiof_write(p, CTR, data);
+
+	for (k = 100; k > 0; k--) {
+		if ((sh_msiof_read(p, CTR) & mask) == set)
+			break;
+
+		udelay(10);
+	}
+
+	return k > 0 ? 0 : -ETIMEDOUT;
+}
+
+static irqreturn_t sh_msiof_spi_irq(int irq, void *data)
+{
+	struct sh_msiof_spi_priv *p = data;
+
+	/* just disable the interrupt and wake up */
+	sh_msiof_write(p, IER, 0);
+	complete(&p->done);
+
+	return IRQ_HANDLED;
+}
+
+static const u32 sh_msiof_spi_div_array[] = {
+	SCR_BRDV_DIV_1, SCR_BRDV_DIV_2,	 SCR_BRDV_DIV_4,
+	SCR_BRDV_DIV_8,	SCR_BRDV_DIV_16, SCR_BRDV_DIV_32,
+};
+
+static void sh_msiof_spi_set_clk_regs(struct sh_msiof_spi_priv *p,
+				      unsigned long parent_rate, u32 spi_hz)
+{
+	unsigned long div;
+	u32 brps, scr;
+	unsigned int div_pow = p->min_div_pow;
+
+	if (!spi_hz || !parent_rate) {
+		WARN(1, "Invalid clock rate parameters %lu and %u\n",
+		     parent_rate, spi_hz);
+		return;
+	}
+
+	div = DIV_ROUND_UP(parent_rate, spi_hz);
+	if (div <= 1024) {
+		/* SCR_BRDV_DIV_1 is valid only if BRPS is x 1/1 or x 1/2 */
+		if (!div_pow && div <= 32 && div > 2)
+			div_pow = 1;
+
+		if (div_pow)
+			brps = (div + 1) >> div_pow;
+		else
+			brps = div;
+
+		for (; brps > 32; div_pow++)
+			brps = (brps + 1) >> 1;
+	} else {
+		/* Set transfer rate composite divisor to 2^5 * 32 = 1024 */
+		dev_err(&p->pdev->dev,
+			"Requested SPI transfer rate %d is too low\n", spi_hz);
+		div_pow = 5;
+		brps = 32;
+	}
+
+	scr = sh_msiof_spi_div_array[div_pow] | SCR_BRPS(brps);
+	sh_msiof_write(p, TSCR, scr);
+	if (!(p->master->flags & SPI_MASTER_MUST_TX))
+		sh_msiof_write(p, RSCR, scr);
+}
+
+static u32 sh_msiof_get_delay_bit(u32 dtdl_or_syncdl)
+{
+	/*
+	 * DTDL/SYNCDL bit	: p->info->dtdl or p->info->syncdl
+	 * b'000		: 0
+	 * b'001		: 100
+	 * b'010		: 200
+	 * b'011 (SYNCDL only)	: 300
+	 * b'101		: 50
+	 * b'110		: 150
+	 */
+	if (dtdl_or_syncdl % 100)
+		return dtdl_or_syncdl / 100 + 5;
+	else
+		return dtdl_or_syncdl / 100;
+}
+
+static u32 sh_msiof_spi_get_dtdl_and_syncdl(struct sh_msiof_spi_priv *p)
+{
+	u32 val;
+
+	if (!p->info)
+		return 0;
+
+	/* check if DTDL and SYNCDL is allowed value */
+	if (p->info->dtdl > 200 || p->info->syncdl > 300) {
+		dev_warn(&p->pdev->dev, "DTDL or SYNCDL is too large\n");
+		return 0;
+	}
+
+	/* check if the sum of DTDL and SYNCDL becomes an integer value  */
+	if ((p->info->dtdl + p->info->syncdl) % 100) {
+		dev_warn(&p->pdev->dev, "the sum of DTDL/SYNCDL is not good\n");
+		return 0;
+	}
+
+	val = sh_msiof_get_delay_bit(p->info->dtdl) << MDR1_DTDL_SHIFT;
+	val |= sh_msiof_get_delay_bit(p->info->syncdl) << MDR1_SYNCDL_SHIFT;
+
+	return val;
+}
+
+static void sh_msiof_spi_set_pin_regs(struct sh_msiof_spi_priv *p, u32 ss,
+				      u32 cpol, u32 cpha,
+				      u32 tx_hi_z, u32 lsb_first, u32 cs_high)
+{
+	u32 tmp;
+	int edge;
+
+	/*
+	 * CPOL CPHA     TSCKIZ RSCKIZ TEDG REDG
+	 *    0    0         10     10    1    1
+	 *    0    1         10     10    0    0
+	 *    1    0         11     11    0    0
+	 *    1    1         11     11    1    1
+	 */
+	tmp = MDR1_SYNCMD_SPI | 1 << MDR1_FLD_SHIFT | MDR1_XXSTP;
+	tmp |= !cs_high << MDR1_SYNCAC_SHIFT;
+	tmp |= lsb_first << MDR1_BITLSB_SHIFT;
+	tmp |= sh_msiof_spi_get_dtdl_and_syncdl(p);
+	if (spi_controller_is_slave(p->master)) {
+		sh_msiof_write(p, TMDR1, tmp | TMDR1_PCON);
+	} else {
+		sh_msiof_write(p, TMDR1,
+			       tmp | MDR1_TRMD | TMDR1_PCON |
+			       (ss < MAX_SS ? ss : 0) << TMDR1_SYNCCH_SHIFT);
+	}
+	if (p->master->flags & SPI_MASTER_MUST_TX) {
+		/* These bits are reserved if RX needs TX */
+		tmp &= ~0x0000ffff;
+	}
+	sh_msiof_write(p, RMDR1, tmp);
+
+	tmp = 0;
+	tmp |= CTR_TSCKIZ_SCK | cpol << CTR_TSCKIZ_POL_SHIFT;
+	tmp |= CTR_RSCKIZ_SCK | cpol << CTR_RSCKIZ_POL_SHIFT;
+
+	edge = cpol ^ !cpha;
+
+	tmp |= edge << CTR_TEDG_SHIFT;
+	tmp |= edge << CTR_REDG_SHIFT;
+	tmp |= tx_hi_z ? CTR_TXDIZ_HIZ : CTR_TXDIZ_LOW;
+	sh_msiof_write(p, CTR, tmp);
+}
+
+static void sh_msiof_spi_set_mode_regs(struct sh_msiof_spi_priv *p,
+				       const void *tx_buf, void *rx_buf,
+				       u32 bits, u32 words)
+{
+	u32 dr2 = MDR2_BITLEN1(bits) | MDR2_WDLEN1(words);
+
+	if (tx_buf || (p->master->flags & SPI_MASTER_MUST_TX))
+		sh_msiof_write(p, TMDR2, dr2);
+	else
+		sh_msiof_write(p, TMDR2, dr2 | MDR2_GRPMASK1);
+
+	if (rx_buf)
+		sh_msiof_write(p, RMDR2, dr2);
+}
+
+static void sh_msiof_reset_str(struct sh_msiof_spi_priv *p)
+{
+	sh_msiof_write(p, STR,
+		       sh_msiof_read(p, STR) & ~(STR_TDREQ | STR_RDREQ));
+}
+
+static void sh_msiof_spi_write_fifo_8(struct sh_msiof_spi_priv *p,
+				      const void *tx_buf, int words, int fs)
+{
+	const u8 *buf_8 = tx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		sh_msiof_write(p, TFDR, buf_8[k] << fs);
+}
+
+static void sh_msiof_spi_write_fifo_16(struct sh_msiof_spi_priv *p,
+				       const void *tx_buf, int words, int fs)
+{
+	const u16 *buf_16 = tx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		sh_msiof_write(p, TFDR, buf_16[k] << fs);
+}
+
+static void sh_msiof_spi_write_fifo_16u(struct sh_msiof_spi_priv *p,
+					const void *tx_buf, int words, int fs)
+{
+	const u16 *buf_16 = tx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		sh_msiof_write(p, TFDR, get_unaligned(&buf_16[k]) << fs);
+}
+
+static void sh_msiof_spi_write_fifo_32(struct sh_msiof_spi_priv *p,
+				       const void *tx_buf, int words, int fs)
+{
+	const u32 *buf_32 = tx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		sh_msiof_write(p, TFDR, buf_32[k] << fs);
+}
+
+static void sh_msiof_spi_write_fifo_32u(struct sh_msiof_spi_priv *p,
+					const void *tx_buf, int words, int fs)
+{
+	const u32 *buf_32 = tx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		sh_msiof_write(p, TFDR, get_unaligned(&buf_32[k]) << fs);
+}
+
+static void sh_msiof_spi_write_fifo_s32(struct sh_msiof_spi_priv *p,
+					const void *tx_buf, int words, int fs)
+{
+	const u32 *buf_32 = tx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		sh_msiof_write(p, TFDR, swab32(buf_32[k] << fs));
+}
+
+static void sh_msiof_spi_write_fifo_s32u(struct sh_msiof_spi_priv *p,
+					 const void *tx_buf, int words, int fs)
+{
+	const u32 *buf_32 = tx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		sh_msiof_write(p, TFDR, swab32(get_unaligned(&buf_32[k]) << fs));
+}
+
+static void sh_msiof_spi_read_fifo_8(struct sh_msiof_spi_priv *p,
+				     void *rx_buf, int words, int fs)
+{
+	u8 *buf_8 = rx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		buf_8[k] = sh_msiof_read(p, RFDR) >> fs;
+}
+
+static void sh_msiof_spi_read_fifo_16(struct sh_msiof_spi_priv *p,
+				      void *rx_buf, int words, int fs)
+{
+	u16 *buf_16 = rx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		buf_16[k] = sh_msiof_read(p, RFDR) >> fs;
+}
+
+static void sh_msiof_spi_read_fifo_16u(struct sh_msiof_spi_priv *p,
+				       void *rx_buf, int words, int fs)
+{
+	u16 *buf_16 = rx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		put_unaligned(sh_msiof_read(p, RFDR) >> fs, &buf_16[k]);
+}
+
+static void sh_msiof_spi_read_fifo_32(struct sh_msiof_spi_priv *p,
+				      void *rx_buf, int words, int fs)
+{
+	u32 *buf_32 = rx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		buf_32[k] = sh_msiof_read(p, RFDR) >> fs;
+}
+
+static void sh_msiof_spi_read_fifo_32u(struct sh_msiof_spi_priv *p,
+				       void *rx_buf, int words, int fs)
+{
+	u32 *buf_32 = rx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		put_unaligned(sh_msiof_read(p, RFDR) >> fs, &buf_32[k]);
+}
+
+static void sh_msiof_spi_read_fifo_s32(struct sh_msiof_spi_priv *p,
+				       void *rx_buf, int words, int fs)
+{
+	u32 *buf_32 = rx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		buf_32[k] = swab32(sh_msiof_read(p, RFDR) >> fs);
+}
+
+static void sh_msiof_spi_read_fifo_s32u(struct sh_msiof_spi_priv *p,
+				       void *rx_buf, int words, int fs)
+{
+	u32 *buf_32 = rx_buf;
+	int k;
+
+	for (k = 0; k < words; k++)
+		put_unaligned(swab32(sh_msiof_read(p, RFDR) >> fs), &buf_32[k]);
+}
+
+static int sh_msiof_spi_setup(struct spi_device *spi)
+{
+	struct device_node	*np = spi->master->dev.of_node;
+	struct sh_msiof_spi_priv *p = spi_master_get_devdata(spi->master);
+	u32 clr, set, tmp;
+
+	if (!np) {
+		/*
+		 * Use spi->controller_data for CS (same strategy as spi_gpio),
+		 * if any. otherwise let HW control CS
+		 */
+		spi->cs_gpio = (uintptr_t)spi->controller_data;
+	}
+
+	if (gpio_is_valid(spi->cs_gpio)) {
+		gpio_direction_output(spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
+		return 0;
+	}
+
+	if (spi_controller_is_slave(p->master))
+		return 0;
+
+	if (p->native_cs_inited &&
+	    (p->native_cs_high == !!(spi->mode & SPI_CS_HIGH)))
+		return 0;
+
+	/* Configure native chip select mode/polarity early */
+	clr = MDR1_SYNCMD_MASK;
+	set = MDR1_SYNCMD_SPI;
+	if (spi->mode & SPI_CS_HIGH)
+		clr |= BIT(MDR1_SYNCAC_SHIFT);
+	else
+		set |= BIT(MDR1_SYNCAC_SHIFT);
+	pm_runtime_get_sync(&p->pdev->dev);
+	tmp = sh_msiof_read(p, TMDR1) & ~clr;
+	sh_msiof_write(p, TMDR1, tmp | set | MDR1_TRMD | TMDR1_PCON);
+	tmp = sh_msiof_read(p, RMDR1) & ~clr;
+	sh_msiof_write(p, RMDR1, tmp | set);
+	pm_runtime_put(&p->pdev->dev);
+	p->native_cs_high = spi->mode & SPI_CS_HIGH;
+	p->native_cs_inited = true;
+	return 0;
+}
+
+static int sh_msiof_prepare_message(struct spi_master *master,
+				    struct spi_message *msg)
+{
+	struct sh_msiof_spi_priv *p = spi_master_get_devdata(master);
+	const struct spi_device *spi = msg->spi;
+	u32 ss, cs_high;
+
+	/* Configure pins before asserting CS */
+	if (gpio_is_valid(spi->cs_gpio)) {
+		ss = p->unused_ss;
+		cs_high = p->native_cs_high;
+	} else {
+		ss = spi->chip_select;
+		cs_high = !!(spi->mode & SPI_CS_HIGH);
+	}
+	sh_msiof_spi_set_pin_regs(p, ss, !!(spi->mode & SPI_CPOL),
+				  !!(spi->mode & SPI_CPHA),
+				  !!(spi->mode & SPI_3WIRE),
+				  !!(spi->mode & SPI_LSB_FIRST), cs_high);
+	return 0;
+}
+
+static int sh_msiof_spi_start(struct sh_msiof_spi_priv *p, void *rx_buf)
+{
+	bool slave = spi_controller_is_slave(p->master);
+	int ret = 0;
+
+	/* setup clock and rx/tx signals */
+	if (!slave)
+		ret = sh_msiof_modify_ctr_wait(p, 0, CTR_TSCKE);
+	if (rx_buf && !ret)
+		ret = sh_msiof_modify_ctr_wait(p, 0, CTR_RXE);
+	if (!ret)
+		ret = sh_msiof_modify_ctr_wait(p, 0, CTR_TXE);
+
+	/* start by setting frame bit */
+	if (!ret && !slave)
+		ret = sh_msiof_modify_ctr_wait(p, 0, CTR_TFSE);
+
+	return ret;
+}
+
+static int sh_msiof_spi_stop(struct sh_msiof_spi_priv *p, void *rx_buf)
+{
+	bool slave = spi_controller_is_slave(p->master);
+	int ret = 0;
+
+	/* shut down frame, rx/tx and clock signals */
+	if (!slave)
+		ret = sh_msiof_modify_ctr_wait(p, CTR_TFSE, 0);
+	if (!ret)
+		ret = sh_msiof_modify_ctr_wait(p, CTR_TXE, 0);
+	if (rx_buf && !ret)
+		ret = sh_msiof_modify_ctr_wait(p, CTR_RXE, 0);
+	if (!ret && !slave)
+		ret = sh_msiof_modify_ctr_wait(p, CTR_TSCKE, 0);
+
+	return ret;
+}
+
+static int sh_msiof_slave_abort(struct spi_master *master)
+{
+	struct sh_msiof_spi_priv *p = spi_master_get_devdata(master);
+
+	p->slave_aborted = true;
+	complete(&p->done);
+	complete(&p->done_txdma);
+	return 0;
+}
+
+static int sh_msiof_wait_for_completion(struct sh_msiof_spi_priv *p,
+					struct completion *x)
+{
+	if (spi_controller_is_slave(p->master)) {
+		if (wait_for_completion_interruptible(x) ||
+		    p->slave_aborted) {
+			dev_dbg(&p->pdev->dev, "interrupted\n");
+			return -EINTR;
+		}
+	} else {
+		if (!wait_for_completion_timeout(x, HZ)) {
+			dev_err(&p->pdev->dev, "timeout\n");
+			return -ETIMEDOUT;
+		}
+	}
+
+	return 0;
+}
+
+static int sh_msiof_spi_txrx_once(struct sh_msiof_spi_priv *p,
+				  void (*tx_fifo)(struct sh_msiof_spi_priv *,
+						  const void *, int, int),
+				  void (*rx_fifo)(struct sh_msiof_spi_priv *,
+						  void *, int, int),
+				  const void *tx_buf, void *rx_buf,
+				  int words, int bits)
+{
+	int fifo_shift;
+	int ret;
+
+	/* limit maximum word transfer to rx/tx fifo size */
+	if (tx_buf)
+		words = min_t(int, words, p->tx_fifo_size);
+	if (rx_buf)
+		words = min_t(int, words, p->rx_fifo_size);
+
+	/* the fifo contents need shifting */
+	fifo_shift = 32 - bits;
+
+	/* default FIFO watermarks for PIO */
+	sh_msiof_write(p, FCTR, 0);
+
+	/* setup msiof transfer mode registers */
+	sh_msiof_spi_set_mode_regs(p, tx_buf, rx_buf, bits, words);
+	sh_msiof_write(p, IER, IER_TEOFE | IER_REOFE);
+
+	/* write tx fifo */
+	if (tx_buf)
+		tx_fifo(p, tx_buf, words, fifo_shift);
+
+	reinit_completion(&p->done);
+	p->slave_aborted = false;
+
+	ret = sh_msiof_spi_start(p, rx_buf);
+	if (ret) {
+		dev_err(&p->pdev->dev, "failed to start hardware\n");
+		goto stop_ier;
+	}
+
+	/* wait for tx fifo to be emptied / rx fifo to be filled */
+	ret = sh_msiof_wait_for_completion(p, &p->done);
+	if (ret)
+		goto stop_reset;
+
+	/* read rx fifo */
+	if (rx_buf)
+		rx_fifo(p, rx_buf, words, fifo_shift);
+
+	/* clear status bits */
+	sh_msiof_reset_str(p);
+
+	ret = sh_msiof_spi_stop(p, rx_buf);
+	if (ret) {
+		dev_err(&p->pdev->dev, "failed to shut down hardware\n");
+		return ret;
+	}
+
+	return words;
+
+stop_reset:
+	sh_msiof_reset_str(p);
+	sh_msiof_spi_stop(p, rx_buf);
+stop_ier:
+	sh_msiof_write(p, IER, 0);
+	return ret;
+}
+
+static void sh_msiof_dma_complete(void *arg)
+{
+	complete(arg);
+}
+
+static int sh_msiof_dma_once(struct sh_msiof_spi_priv *p, const void *tx,
+			     void *rx, unsigned int len)
+{
+	u32 ier_bits = 0;
+	struct dma_async_tx_descriptor *desc_tx = NULL, *desc_rx = NULL;
+	dma_cookie_t cookie;
+	int ret;
+
+	/* First prepare and submit the DMA request(s), as this may fail */
+	if (rx) {
+		ier_bits |= IER_RDREQE | IER_RDMAE;
+		desc_rx = dmaengine_prep_slave_single(p->master->dma_rx,
+					p->rx_dma_addr, len, DMA_DEV_TO_MEM,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!desc_rx)
+			return -EAGAIN;
+
+		desc_rx->callback = sh_msiof_dma_complete;
+		desc_rx->callback_param = &p->done;
+		cookie = dmaengine_submit(desc_rx);
+		if (dma_submit_error(cookie))
+			return cookie;
+	}
+
+	if (tx) {
+		ier_bits |= IER_TDREQE | IER_TDMAE;
+		dma_sync_single_for_device(p->master->dma_tx->device->dev,
+					   p->tx_dma_addr, len, DMA_TO_DEVICE);
+		desc_tx = dmaengine_prep_slave_single(p->master->dma_tx,
+					p->tx_dma_addr, len, DMA_MEM_TO_DEV,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!desc_tx) {
+			ret = -EAGAIN;
+			goto no_dma_tx;
+		}
+
+		desc_tx->callback = sh_msiof_dma_complete;
+		desc_tx->callback_param = &p->done_txdma;
+		cookie = dmaengine_submit(desc_tx);
+		if (dma_submit_error(cookie)) {
+			ret = cookie;
+			goto no_dma_tx;
+		}
+	}
+
+	/* 1 stage FIFO watermarks for DMA */
+	sh_msiof_write(p, FCTR, FCTR_TFWM_1 | FCTR_RFWM_1);
+
+	/* setup msiof transfer mode registers (32-bit words) */
+	sh_msiof_spi_set_mode_regs(p, tx, rx, 32, len / 4);
+
+	sh_msiof_write(p, IER, ier_bits);
+
+	reinit_completion(&p->done);
+	if (tx)
+		reinit_completion(&p->done_txdma);
+	p->slave_aborted = false;
+
+	/* Now start DMA */
+	if (rx)
+		dma_async_issue_pending(p->master->dma_rx);
+	if (tx)
+		dma_async_issue_pending(p->master->dma_tx);
+
+	ret = sh_msiof_spi_start(p, rx);
+	if (ret) {
+		dev_err(&p->pdev->dev, "failed to start hardware\n");
+		goto stop_dma;
+	}
+
+	if (tx) {
+		/* wait for tx DMA completion */
+		ret = sh_msiof_wait_for_completion(p, &p->done_txdma);
+		if (ret)
+			goto stop_reset;
+	}
+
+	if (rx) {
+		/* wait for rx DMA completion */
+		ret = sh_msiof_wait_for_completion(p, &p->done);
+		if (ret)
+			goto stop_reset;
+
+		sh_msiof_write(p, IER, 0);
+	} else {
+		/* wait for tx fifo to be emptied */
+		sh_msiof_write(p, IER, IER_TEOFE);
+		ret = sh_msiof_wait_for_completion(p, &p->done);
+		if (ret)
+			goto stop_reset;
+	}
+
+	/* clear status bits */
+	sh_msiof_reset_str(p);
+
+	ret = sh_msiof_spi_stop(p, rx);
+	if (ret) {
+		dev_err(&p->pdev->dev, "failed to shut down hardware\n");
+		return ret;
+	}
+
+	if (rx)
+		dma_sync_single_for_cpu(p->master->dma_rx->device->dev,
+					p->rx_dma_addr, len,
+					DMA_FROM_DEVICE);
+
+	return 0;
+
+stop_reset:
+	sh_msiof_reset_str(p);
+	sh_msiof_spi_stop(p, rx);
+stop_dma:
+	if (tx)
+		dmaengine_terminate_all(p->master->dma_tx);
+no_dma_tx:
+	if (rx)
+		dmaengine_terminate_all(p->master->dma_rx);
+	sh_msiof_write(p, IER, 0);
+	return ret;
+}
+
+static void copy_bswap32(u32 *dst, const u32 *src, unsigned int words)
+{
+	/* src or dst can be unaligned, but not both */
+	if ((unsigned long)src & 3) {
+		while (words--) {
+			*dst++ = swab32(get_unaligned(src));
+			src++;
+		}
+	} else if ((unsigned long)dst & 3) {
+		while (words--) {
+			put_unaligned(swab32(*src++), dst);
+			dst++;
+		}
+	} else {
+		while (words--)
+			*dst++ = swab32(*src++);
+	}
+}
+
+static void copy_wswap32(u32 *dst, const u32 *src, unsigned int words)
+{
+	/* src or dst can be unaligned, but not both */
+	if ((unsigned long)src & 3) {
+		while (words--) {
+			*dst++ = swahw32(get_unaligned(src));
+			src++;
+		}
+	} else if ((unsigned long)dst & 3) {
+		while (words--) {
+			put_unaligned(swahw32(*src++), dst);
+			dst++;
+		}
+	} else {
+		while (words--)
+			*dst++ = swahw32(*src++);
+	}
+}
+
+static void copy_plain32(u32 *dst, const u32 *src, unsigned int words)
+{
+	memcpy(dst, src, words * 4);
+}
+
+static int sh_msiof_transfer_one(struct spi_master *master,
+				 struct spi_device *spi,
+				 struct spi_transfer *t)
+{
+	struct sh_msiof_spi_priv *p = spi_master_get_devdata(master);
+	void (*copy32)(u32 *, const u32 *, unsigned int);
+	void (*tx_fifo)(struct sh_msiof_spi_priv *, const void *, int, int);
+	void (*rx_fifo)(struct sh_msiof_spi_priv *, void *, int, int);
+	const void *tx_buf = t->tx_buf;
+	void *rx_buf = t->rx_buf;
+	unsigned int len = t->len;
+	unsigned int bits = t->bits_per_word;
+	unsigned int bytes_per_word;
+	unsigned int words;
+	int n;
+	bool swab;
+	int ret;
+
+	/* setup clocks (clock already enabled in chipselect()) */
+	if (!spi_controller_is_slave(p->master))
+		sh_msiof_spi_set_clk_regs(p, clk_get_rate(p->clk), t->speed_hz);
+
+	while (master->dma_tx && len > 15) {
+		/*
+		 *  DMA supports 32-bit words only, hence pack 8-bit and 16-bit
+		 *  words, with byte resp. word swapping.
+		 */
+		unsigned int l = 0;
+
+		if (tx_buf)
+			l = min(len, p->tx_fifo_size * 4);
+		if (rx_buf)
+			l = min(len, p->rx_fifo_size * 4);
+
+		if (bits <= 8) {
+			if (l & 3)
+				break;
+			copy32 = copy_bswap32;
+		} else if (bits <= 16) {
+			if (l & 3)
+				break;
+			copy32 = copy_wswap32;
+		} else {
+			copy32 = copy_plain32;
+		}
+
+		if (tx_buf)
+			copy32(p->tx_dma_page, tx_buf, l / 4);
+
+		ret = sh_msiof_dma_once(p, tx_buf, rx_buf, l);
+		if (ret == -EAGAIN) {
+			dev_warn_once(&p->pdev->dev,
+				"DMA not available, falling back to PIO\n");
+			break;
+		}
+		if (ret)
+			return ret;
+
+		if (rx_buf) {
+			copy32(rx_buf, p->rx_dma_page, l / 4);
+			rx_buf += l;
+		}
+		if (tx_buf)
+			tx_buf += l;
+
+		len -= l;
+		if (!len)
+			return 0;
+	}
+
+	if (bits <= 8 && len > 15 && !(len & 3)) {
+		bits = 32;
+		swab = true;
+	} else {
+		swab = false;
+	}
+
+	/* setup bytes per word and fifo read/write functions */
+	if (bits <= 8) {
+		bytes_per_word = 1;
+		tx_fifo = sh_msiof_spi_write_fifo_8;
+		rx_fifo = sh_msiof_spi_read_fifo_8;
+	} else if (bits <= 16) {
+		bytes_per_word = 2;
+		if ((unsigned long)tx_buf & 0x01)
+			tx_fifo = sh_msiof_spi_write_fifo_16u;
+		else
+			tx_fifo = sh_msiof_spi_write_fifo_16;
+
+		if ((unsigned long)rx_buf & 0x01)
+			rx_fifo = sh_msiof_spi_read_fifo_16u;
+		else
+			rx_fifo = sh_msiof_spi_read_fifo_16;
+	} else if (swab) {
+		bytes_per_word = 4;
+		if ((unsigned long)tx_buf & 0x03)
+			tx_fifo = sh_msiof_spi_write_fifo_s32u;
+		else
+			tx_fifo = sh_msiof_spi_write_fifo_s32;
+
+		if ((unsigned long)rx_buf & 0x03)
+			rx_fifo = sh_msiof_spi_read_fifo_s32u;
+		else
+			rx_fifo = sh_msiof_spi_read_fifo_s32;
+	} else {
+		bytes_per_word = 4;
+		if ((unsigned long)tx_buf & 0x03)
+			tx_fifo = sh_msiof_spi_write_fifo_32u;
+		else
+			tx_fifo = sh_msiof_spi_write_fifo_32;
+
+		if ((unsigned long)rx_buf & 0x03)
+			rx_fifo = sh_msiof_spi_read_fifo_32u;
+		else
+			rx_fifo = sh_msiof_spi_read_fifo_32;
+	}
+
+	/* transfer in fifo sized chunks */
+	words = len / bytes_per_word;
+
+	while (words > 0) {
+		n = sh_msiof_spi_txrx_once(p, tx_fifo, rx_fifo, tx_buf, rx_buf,
+					   words, bits);
+		if (n < 0)
+			return n;
+
+		if (tx_buf)
+			tx_buf += n * bytes_per_word;
+		if (rx_buf)
+			rx_buf += n * bytes_per_word;
+		words -= n;
+	}
+
+	return 0;
+}
+
+static const struct sh_msiof_chipdata sh_data = {
+	.tx_fifo_size = 64,
+	.rx_fifo_size = 64,
+	.master_flags = 0,
+	.min_div_pow = 0,
+};
+
+static const struct sh_msiof_chipdata rcar_gen2_data = {
+	.tx_fifo_size = 64,
+	.rx_fifo_size = 64,
+	.master_flags = SPI_MASTER_MUST_TX,
+	.min_div_pow = 0,
+};
+
+static const struct sh_msiof_chipdata rcar_gen3_data = {
+	.tx_fifo_size = 64,
+	.rx_fifo_size = 64,
+	.master_flags = SPI_MASTER_MUST_TX,
+	.min_div_pow = 1,
+};
+
+static const struct of_device_id sh_msiof_match[] = {
+	{ .compatible = "renesas,sh-mobile-msiof", .data = &sh_data },
+	{ .compatible = "renesas,msiof-r8a7743",   .data = &rcar_gen2_data },
+	{ .compatible = "renesas,msiof-r8a7745",   .data = &rcar_gen2_data },
+	{ .compatible = "renesas,msiof-r8a7790",   .data = &rcar_gen2_data },
+	{ .compatible = "renesas,msiof-r8a7791",   .data = &rcar_gen2_data },
+	{ .compatible = "renesas,msiof-r8a7792",   .data = &rcar_gen2_data },
+	{ .compatible = "renesas,msiof-r8a7793",   .data = &rcar_gen2_data },
+	{ .compatible = "renesas,msiof-r8a7794",   .data = &rcar_gen2_data },
+	{ .compatible = "renesas,rcar-gen2-msiof", .data = &rcar_gen2_data },
+	{ .compatible = "renesas,msiof-r8a7796",   .data = &rcar_gen3_data },
+	{ .compatible = "renesas,rcar-gen3-msiof", .data = &rcar_gen3_data },
+	{ .compatible = "renesas,sh-msiof",        .data = &sh_data }, /* Deprecated */
+	{},
+};
+MODULE_DEVICE_TABLE(of, sh_msiof_match);
+
+#ifdef CONFIG_OF
+static struct sh_msiof_spi_info *sh_msiof_spi_parse_dt(struct device *dev)
+{
+	struct sh_msiof_spi_info *info;
+	struct device_node *np = dev->of_node;
+	u32 num_cs = 1;
+
+	info = devm_kzalloc(dev, sizeof(struct sh_msiof_spi_info), GFP_KERNEL);
+	if (!info)
+		return NULL;
+
+	info->mode = of_property_read_bool(np, "spi-slave") ? MSIOF_SPI_SLAVE
+							    : MSIOF_SPI_MASTER;
+
+	/* Parse the MSIOF properties */
+	if (info->mode == MSIOF_SPI_MASTER)
+		of_property_read_u32(np, "num-cs", &num_cs);
+	of_property_read_u32(np, "renesas,tx-fifo-size",
+					&info->tx_fifo_override);
+	of_property_read_u32(np, "renesas,rx-fifo-size",
+					&info->rx_fifo_override);
+	of_property_read_u32(np, "renesas,dtdl", &info->dtdl);
+	of_property_read_u32(np, "renesas,syncdl", &info->syncdl);
+
+	info->num_chipselect = num_cs;
+
+	return info;
+}
+#else
+static struct sh_msiof_spi_info *sh_msiof_spi_parse_dt(struct device *dev)
+{
+	return NULL;
+}
+#endif
+
+static int sh_msiof_get_cs_gpios(struct sh_msiof_spi_priv *p)
+{
+	struct device *dev = &p->pdev->dev;
+	unsigned int used_ss_mask = 0;
+	unsigned int cs_gpios = 0;
+	unsigned int num_cs, i;
+	int ret;
+
+	ret = gpiod_count(dev, "cs");
+	if (ret <= 0)
+		return 0;
+
+	num_cs = max_t(unsigned int, ret, p->master->num_chipselect);
+	for (i = 0; i < num_cs; i++) {
+		struct gpio_desc *gpiod;
+
+		gpiod = devm_gpiod_get_index(dev, "cs", i, GPIOD_ASIS);
+		if (!IS_ERR(gpiod)) {
+			cs_gpios++;
+			continue;
+		}
+
+		if (PTR_ERR(gpiod) != -ENOENT)
+			return PTR_ERR(gpiod);
+
+		if (i >= MAX_SS) {
+			dev_err(dev, "Invalid native chip select %d\n", i);
+			return -EINVAL;
+		}
+		used_ss_mask |= BIT(i);
+	}
+	p->unused_ss = ffz(used_ss_mask);
+	if (cs_gpios && p->unused_ss >= MAX_SS) {
+		dev_err(dev, "No unused native chip select available\n");
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static struct dma_chan *sh_msiof_request_dma_chan(struct device *dev,
+	enum dma_transfer_direction dir, unsigned int id, dma_addr_t port_addr)
+{
+	dma_cap_mask_t mask;
+	struct dma_chan *chan;
+	struct dma_slave_config cfg;
+	int ret;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	chan = dma_request_slave_channel_compat(mask, shdma_chan_filter,
+				(void *)(unsigned long)id, dev,
+				dir == DMA_MEM_TO_DEV ? "tx" : "rx");
+	if (!chan) {
+		dev_warn(dev, "dma_request_slave_channel_compat failed\n");
+		return NULL;
+	}
+
+	memset(&cfg, 0, sizeof(cfg));
+	cfg.direction = dir;
+	if (dir == DMA_MEM_TO_DEV) {
+		cfg.dst_addr = port_addr;
+		cfg.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+	} else {
+		cfg.src_addr = port_addr;
+		cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+	}
+
+	ret = dmaengine_slave_config(chan, &cfg);
+	if (ret) {
+		dev_warn(dev, "dmaengine_slave_config failed %d\n", ret);
+		dma_release_channel(chan);
+		return NULL;
+	}
+
+	return chan;
+}
+
+static int sh_msiof_request_dma(struct sh_msiof_spi_priv *p)
+{
+	struct platform_device *pdev = p->pdev;
+	struct device *dev = &pdev->dev;
+	const struct sh_msiof_spi_info *info = dev_get_platdata(dev);
+	unsigned int dma_tx_id, dma_rx_id;
+	const struct resource *res;
+	struct spi_master *master;
+	struct device *tx_dev, *rx_dev;
+
+	if (dev->of_node) {
+		/* In the OF case we will get the slave IDs from the DT */
+		dma_tx_id = 0;
+		dma_rx_id = 0;
+	} else if (info && info->dma_tx_id && info->dma_rx_id) {
+		dma_tx_id = info->dma_tx_id;
+		dma_rx_id = info->dma_rx_id;
+	} else {
+		/* The driver assumes no error */
+		return 0;
+	}
+
+	/* The DMA engine uses the second register set, if present */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	if (!res)
+		res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+
+	master = p->master;
+	master->dma_tx = sh_msiof_request_dma_chan(dev, DMA_MEM_TO_DEV,
+						   dma_tx_id,
+						   res->start + TFDR);
+	if (!master->dma_tx)
+		return -ENODEV;
+
+	master->dma_rx = sh_msiof_request_dma_chan(dev, DMA_DEV_TO_MEM,
+						   dma_rx_id,
+						   res->start + RFDR);
+	if (!master->dma_rx)
+		goto free_tx_chan;
+
+	p->tx_dma_page = (void *)__get_free_page(GFP_KERNEL | GFP_DMA);
+	if (!p->tx_dma_page)
+		goto free_rx_chan;
+
+	p->rx_dma_page = (void *)__get_free_page(GFP_KERNEL | GFP_DMA);
+	if (!p->rx_dma_page)
+		goto free_tx_page;
+
+	tx_dev = master->dma_tx->device->dev;
+	p->tx_dma_addr = dma_map_single(tx_dev, p->tx_dma_page, PAGE_SIZE,
+					DMA_TO_DEVICE);
+	if (dma_mapping_error(tx_dev, p->tx_dma_addr))
+		goto free_rx_page;
+
+	rx_dev = master->dma_rx->device->dev;
+	p->rx_dma_addr = dma_map_single(rx_dev, p->rx_dma_page, PAGE_SIZE,
+					DMA_FROM_DEVICE);
+	if (dma_mapping_error(rx_dev, p->rx_dma_addr))
+		goto unmap_tx_page;
+
+	dev_info(dev, "DMA available");
+	return 0;
+
+unmap_tx_page:
+	dma_unmap_single(tx_dev, p->tx_dma_addr, PAGE_SIZE, DMA_TO_DEVICE);
+free_rx_page:
+	free_page((unsigned long)p->rx_dma_page);
+free_tx_page:
+	free_page((unsigned long)p->tx_dma_page);
+free_rx_chan:
+	dma_release_channel(master->dma_rx);
+free_tx_chan:
+	dma_release_channel(master->dma_tx);
+	master->dma_tx = NULL;
+	return -ENODEV;
+}
+
+static void sh_msiof_release_dma(struct sh_msiof_spi_priv *p)
+{
+	struct spi_master *master = p->master;
+
+	if (!master->dma_tx)
+		return;
+
+	dma_unmap_single(master->dma_rx->device->dev, p->rx_dma_addr,
+			 PAGE_SIZE, DMA_FROM_DEVICE);
+	dma_unmap_single(master->dma_tx->device->dev, p->tx_dma_addr,
+			 PAGE_SIZE, DMA_TO_DEVICE);
+	free_page((unsigned long)p->rx_dma_page);
+	free_page((unsigned long)p->tx_dma_page);
+	dma_release_channel(master->dma_rx);
+	dma_release_channel(master->dma_tx);
+}
+
+static int sh_msiof_spi_probe(struct platform_device *pdev)
+{
+	struct resource	*r;
+	struct spi_master *master;
+	const struct sh_msiof_chipdata *chipdata;
+	struct sh_msiof_spi_info *info;
+	struct sh_msiof_spi_priv *p;
+	int i;
+	int ret;
+
+	chipdata = of_device_get_match_data(&pdev->dev);
+	if (chipdata) {
+		info = sh_msiof_spi_parse_dt(&pdev->dev);
+	} else {
+		chipdata = (const void *)pdev->id_entry->driver_data;
+		info = dev_get_platdata(&pdev->dev);
+	}
+
+	if (!info) {
+		dev_err(&pdev->dev, "failed to obtain device info\n");
+		return -ENXIO;
+	}
+
+	if (info->mode == MSIOF_SPI_SLAVE)
+		master = spi_alloc_slave(&pdev->dev,
+					 sizeof(struct sh_msiof_spi_priv));
+	else
+		master = spi_alloc_master(&pdev->dev,
+					  sizeof(struct sh_msiof_spi_priv));
+	if (master == NULL)
+		return -ENOMEM;
+
+	p = spi_master_get_devdata(master);
+
+	platform_set_drvdata(pdev, p);
+	p->master = master;
+	p->info = info;
+	p->min_div_pow = chipdata->min_div_pow;
+
+	init_completion(&p->done);
+	init_completion(&p->done_txdma);
+
+	p->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(p->clk)) {
+		dev_err(&pdev->dev, "cannot get clock\n");
+		ret = PTR_ERR(p->clk);
+		goto err1;
+	}
+
+	i = platform_get_irq(pdev, 0);
+	if (i < 0) {
+		dev_err(&pdev->dev, "cannot get platform IRQ\n");
+		ret = -ENOENT;
+		goto err1;
+	}
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	p->mapbase = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(p->mapbase)) {
+		ret = PTR_ERR(p->mapbase);
+		goto err1;
+	}
+
+	ret = devm_request_irq(&pdev->dev, i, sh_msiof_spi_irq, 0,
+			       dev_name(&pdev->dev), p);
+	if (ret) {
+		dev_err(&pdev->dev, "unable to request irq\n");
+		goto err1;
+	}
+
+	p->pdev = pdev;
+	pm_runtime_enable(&pdev->dev);
+
+	/* Platform data may override FIFO sizes */
+	p->tx_fifo_size = chipdata->tx_fifo_size;
+	p->rx_fifo_size = chipdata->rx_fifo_size;
+	if (p->info->tx_fifo_override)
+		p->tx_fifo_size = p->info->tx_fifo_override;
+	if (p->info->rx_fifo_override)
+		p->rx_fifo_size = p->info->rx_fifo_override;
+
+	/* Setup GPIO chip selects */
+	master->num_chipselect = p->info->num_chipselect;
+	ret = sh_msiof_get_cs_gpios(p);
+	if (ret)
+		goto err1;
+
+	/* init master code */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->mode_bits |= SPI_LSB_FIRST | SPI_3WIRE;
+	master->flags = chipdata->master_flags;
+	master->bus_num = pdev->id;
+	master->dev.of_node = pdev->dev.of_node;
+	master->setup = sh_msiof_spi_setup;
+	master->prepare_message = sh_msiof_prepare_message;
+	master->slave_abort = sh_msiof_slave_abort;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(8, 32);
+	master->auto_runtime_pm = true;
+	master->transfer_one = sh_msiof_transfer_one;
+
+	ret = sh_msiof_request_dma(p);
+	if (ret < 0)
+		dev_warn(&pdev->dev, "DMA not available, using PIO\n");
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "spi_register_master error.\n");
+		goto err2;
+	}
+
+	return 0;
+
+ err2:
+	sh_msiof_release_dma(p);
+	pm_runtime_disable(&pdev->dev);
+ err1:
+	spi_master_put(master);
+	return ret;
+}
+
+static int sh_msiof_spi_remove(struct platform_device *pdev)
+{
+	struct sh_msiof_spi_priv *p = platform_get_drvdata(pdev);
+
+	sh_msiof_release_dma(p);
+	pm_runtime_disable(&pdev->dev);
+	return 0;
+}
+
+static const struct platform_device_id spi_driver_ids[] = {
+	{ "spi_sh_msiof",	(kernel_ulong_t)&sh_data },
+	{},
+};
+MODULE_DEVICE_TABLE(platform, spi_driver_ids);
+
+#ifdef CONFIG_PM_SLEEP
+static int sh_msiof_spi_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct sh_msiof_spi_priv *p = platform_get_drvdata(pdev);
+
+	return spi_master_suspend(p->master);
+}
+
+static int sh_msiof_spi_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct sh_msiof_spi_priv *p = platform_get_drvdata(pdev);
+
+	return spi_master_resume(p->master);
+}
+
+static SIMPLE_DEV_PM_OPS(sh_msiof_spi_pm_ops, sh_msiof_spi_suspend,
+			 sh_msiof_spi_resume);
+#define DEV_PM_OPS	&sh_msiof_spi_pm_ops
+#else
+#define DEV_PM_OPS	NULL
+#endif /* CONFIG_PM_SLEEP */
+
+static struct platform_driver sh_msiof_spi_drv = {
+	.probe		= sh_msiof_spi_probe,
+	.remove		= sh_msiof_spi_remove,
+	.id_table	= spi_driver_ids,
+	.driver		= {
+		.name		= "spi_sh_msiof",
+		.pm		= DEV_PM_OPS,
+		.of_match_table = of_match_ptr(sh_msiof_match),
+	},
+};
+module_platform_driver(sh_msiof_spi_drv);
+
+MODULE_DESCRIPTION("SuperH MSIOF SPI Master Interface Driver");
+MODULE_AUTHOR("Magnus Damm");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:spi_sh_msiof");
diff --git a/drivers/spi/spi-sh-sci.c b/drivers/spi/spi-sh-sci.c
new file mode 100644
index 0000000..393701c
--- /dev/null
+++ b/drivers/spi/spi-sh-sci.c
@@ -0,0 +1,201 @@
+/*
+ * SH SCI SPI interface
+ *
+ * Copyright (c) 2008 Magnus Damm
+ *
+ * Based on S3C24XX GPIO based SPI driver, which is:
+ *   Copyright (c) 2006 Ben Dooks
+ *   Copyright (c) 2006 Simtec Electronics
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/delay.h>
+#include <linux/spinlock.h>
+#include <linux/platform_device.h>
+
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/module.h>
+
+#include <asm/spi.h>
+#include <asm/io.h>
+
+struct sh_sci_spi {
+	struct spi_bitbang bitbang;
+
+	void __iomem *membase;
+	unsigned char val;
+	struct sh_spi_info *info;
+	struct platform_device *dev;
+};
+
+#define SCSPTR(sp)	(sp->membase + 0x1c)
+#define PIN_SCK		(1 << 2)
+#define PIN_TXD		(1 << 0)
+#define PIN_RXD		PIN_TXD
+#define PIN_INIT	((1 << 1) | (1 << 3) | PIN_SCK | PIN_TXD)
+
+static inline void setbits(struct sh_sci_spi *sp, int bits, int on)
+{
+	/*
+	 * We are the only user of SCSPTR so no locking is required.
+	 * Reading bit 2 and 0 in SCSPTR gives pin state as input.
+	 * Writing the same bits sets the output value.
+	 * This makes regular read-modify-write difficult so we
+	 * use sp->val to keep track of the latest register value.
+	 */
+
+	if (on)
+		sp->val |= bits;
+	else
+		sp->val &= ~bits;
+
+	iowrite8(sp->val, SCSPTR(sp));
+}
+
+static inline void setsck(struct spi_device *dev, int on)
+{
+	setbits(spi_master_get_devdata(dev->master), PIN_SCK, on);
+}
+
+static inline void setmosi(struct spi_device *dev, int on)
+{
+	setbits(spi_master_get_devdata(dev->master), PIN_TXD, on);
+}
+
+static inline u32 getmiso(struct spi_device *dev)
+{
+	struct sh_sci_spi *sp = spi_master_get_devdata(dev->master);
+
+	return (ioread8(SCSPTR(sp)) & PIN_RXD) ? 1 : 0;
+}
+
+#define spidelay(x) ndelay(x)
+
+#include "spi-bitbang-txrx.h"
+
+static u32 sh_sci_spi_txrx_mode0(struct spi_device *spi,
+				 unsigned nsecs, u32 word, u8 bits,
+				 unsigned flags)
+{
+	return bitbang_txrx_be_cpha0(spi, nsecs, 0, flags, word, bits);
+}
+
+static u32 sh_sci_spi_txrx_mode1(struct spi_device *spi,
+				 unsigned nsecs, u32 word, u8 bits,
+				 unsigned flags)
+{
+	return bitbang_txrx_be_cpha1(spi, nsecs, 0, flags, word, bits);
+}
+
+static u32 sh_sci_spi_txrx_mode2(struct spi_device *spi,
+				 unsigned nsecs, u32 word, u8 bits,
+				 unsigned flags)
+{
+	return bitbang_txrx_be_cpha0(spi, nsecs, 1, flags, word, bits);
+}
+
+static u32 sh_sci_spi_txrx_mode3(struct spi_device *spi,
+				 unsigned nsecs, u32 word, u8 bits,
+				 unsigned flags)
+{
+	return bitbang_txrx_be_cpha1(spi, nsecs, 1, flags, word, bits);
+}
+
+static void sh_sci_spi_chipselect(struct spi_device *dev, int value)
+{
+	struct sh_sci_spi *sp = spi_master_get_devdata(dev->master);
+
+	if (sp->info->chip_select)
+		(sp->info->chip_select)(sp->info, dev->chip_select, value);
+}
+
+static int sh_sci_spi_probe(struct platform_device *dev)
+{
+	struct resource	*r;
+	struct spi_master *master;
+	struct sh_sci_spi *sp;
+	int ret;
+
+	master = spi_alloc_master(&dev->dev, sizeof(struct sh_sci_spi));
+	if (master == NULL) {
+		dev_err(&dev->dev, "failed to allocate spi master\n");
+		ret = -ENOMEM;
+		goto err0;
+	}
+
+	sp = spi_master_get_devdata(master);
+
+	platform_set_drvdata(dev, sp);
+	sp->info = dev_get_platdata(&dev->dev);
+	if (!sp->info) {
+		dev_err(&dev->dev, "platform data is missing\n");
+		ret = -ENOENT;
+		goto err1;
+	}
+
+	/* setup spi bitbang adaptor */
+	sp->bitbang.master = master;
+	sp->bitbang.master->bus_num = sp->info->bus_num;
+	sp->bitbang.master->num_chipselect = sp->info->num_chipselect;
+	sp->bitbang.chipselect = sh_sci_spi_chipselect;
+
+	sp->bitbang.txrx_word[SPI_MODE_0] = sh_sci_spi_txrx_mode0;
+	sp->bitbang.txrx_word[SPI_MODE_1] = sh_sci_spi_txrx_mode1;
+	sp->bitbang.txrx_word[SPI_MODE_2] = sh_sci_spi_txrx_mode2;
+	sp->bitbang.txrx_word[SPI_MODE_3] = sh_sci_spi_txrx_mode3;
+
+	r = platform_get_resource(dev, IORESOURCE_MEM, 0);
+	if (r == NULL) {
+		ret = -ENOENT;
+		goto err1;
+	}
+	sp->membase = ioremap(r->start, resource_size(r));
+	if (!sp->membase) {
+		ret = -ENXIO;
+		goto err1;
+	}
+	sp->val = ioread8(SCSPTR(sp));
+	setbits(sp, PIN_INIT, 1);
+
+	ret = spi_bitbang_start(&sp->bitbang);
+	if (!ret)
+		return 0;
+
+	setbits(sp, PIN_INIT, 0);
+	iounmap(sp->membase);
+ err1:
+	spi_master_put(sp->bitbang.master);
+ err0:
+	return ret;
+}
+
+static int sh_sci_spi_remove(struct platform_device *dev)
+{
+	struct sh_sci_spi *sp = platform_get_drvdata(dev);
+
+	spi_bitbang_stop(&sp->bitbang);
+	setbits(sp, PIN_INIT, 0);
+	iounmap(sp->membase);
+	spi_master_put(sp->bitbang.master);
+	return 0;
+}
+
+static struct platform_driver sh_sci_spi_drv = {
+	.probe		= sh_sci_spi_probe,
+	.remove		= sh_sci_spi_remove,
+	.driver		= {
+		.name	= "spi_sh_sci",
+	},
+};
+module_platform_driver(sh_sci_spi_drv);
+
+MODULE_DESCRIPTION("SH SCI SPI Driver");
+MODULE_AUTHOR("Magnus Damm <damm@opensource.se>");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:spi_sh_sci");
diff --git a/drivers/spi/spi-sh.c b/drivers/spi/spi-sh.c
new file mode 100644
index 0000000..50e0ea9
--- /dev/null
+++ b/drivers/spi/spi-sh.c
@@ -0,0 +1,527 @@
+/*
+ * SH SPI bus driver
+ *
+ * Copyright (C) 2011  Renesas Solutions Corp.
+ *
+ * Based on pxa2xx_spi.c:
+ * Copyright (C) 2005 Stephen Street / StreetFire Sound Labs
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/errno.h>
+#include <linux/timer.h>
+#include <linux/delay.h>
+#include <linux/list.h>
+#include <linux/workqueue.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/io.h>
+#include <linux/spi/spi.h>
+
+#define SPI_SH_TBR		0x00
+#define SPI_SH_RBR		0x00
+#define SPI_SH_CR1		0x08
+#define SPI_SH_CR2		0x10
+#define SPI_SH_CR3		0x18
+#define SPI_SH_CR4		0x20
+#define SPI_SH_CR5		0x28
+
+/* CR1 */
+#define SPI_SH_TBE		0x80
+#define SPI_SH_TBF		0x40
+#define SPI_SH_RBE		0x20
+#define SPI_SH_RBF		0x10
+#define SPI_SH_PFONRD		0x08
+#define SPI_SH_SSDB		0x04
+#define SPI_SH_SSD		0x02
+#define SPI_SH_SSA		0x01
+
+/* CR2 */
+#define SPI_SH_RSTF		0x80
+#define SPI_SH_LOOPBK		0x40
+#define SPI_SH_CPOL		0x20
+#define SPI_SH_CPHA		0x10
+#define SPI_SH_L1M0		0x08
+
+/* CR3 */
+#define SPI_SH_MAX_BYTE		0xFF
+
+/* CR4 */
+#define SPI_SH_TBEI		0x80
+#define SPI_SH_TBFI		0x40
+#define SPI_SH_RBEI		0x20
+#define SPI_SH_RBFI		0x10
+#define SPI_SH_WPABRT		0x04
+#define SPI_SH_SSS		0x01
+
+/* CR8 */
+#define SPI_SH_P1L0		0x80
+#define SPI_SH_PP1L0		0x40
+#define SPI_SH_MUXI		0x20
+#define SPI_SH_MUXIRQ		0x10
+
+#define SPI_SH_FIFO_SIZE	32
+#define SPI_SH_SEND_TIMEOUT	(3 * HZ)
+#define SPI_SH_RECEIVE_TIMEOUT	(HZ >> 3)
+
+#undef DEBUG
+
+struct spi_sh_data {
+	void __iomem *addr;
+	int irq;
+	struct spi_master *master;
+	struct list_head queue;
+	struct work_struct ws;
+	unsigned long cr1;
+	wait_queue_head_t wait;
+	spinlock_t lock;
+	int width;
+};
+
+static void spi_sh_write(struct spi_sh_data *ss, unsigned long data,
+			     unsigned long offset)
+{
+	if (ss->width == 8)
+		iowrite8(data, ss->addr + (offset >> 2));
+	else if (ss->width == 32)
+		iowrite32(data, ss->addr + offset);
+}
+
+static unsigned long spi_sh_read(struct spi_sh_data *ss, unsigned long offset)
+{
+	if (ss->width == 8)
+		return ioread8(ss->addr + (offset >> 2));
+	else if (ss->width == 32)
+		return ioread32(ss->addr + offset);
+	else
+		return 0;
+}
+
+static void spi_sh_set_bit(struct spi_sh_data *ss, unsigned long val,
+				unsigned long offset)
+{
+	unsigned long tmp;
+
+	tmp = spi_sh_read(ss, offset);
+	tmp |= val;
+	spi_sh_write(ss, tmp, offset);
+}
+
+static void spi_sh_clear_bit(struct spi_sh_data *ss, unsigned long val,
+				unsigned long offset)
+{
+	unsigned long tmp;
+
+	tmp = spi_sh_read(ss, offset);
+	tmp &= ~val;
+	spi_sh_write(ss, tmp, offset);
+}
+
+static void clear_fifo(struct spi_sh_data *ss)
+{
+	spi_sh_set_bit(ss, SPI_SH_RSTF, SPI_SH_CR2);
+	spi_sh_clear_bit(ss, SPI_SH_RSTF, SPI_SH_CR2);
+}
+
+static int spi_sh_wait_receive_buffer(struct spi_sh_data *ss)
+{
+	int timeout = 100000;
+
+	while (spi_sh_read(ss, SPI_SH_CR1) & SPI_SH_RBE) {
+		udelay(10);
+		if (timeout-- < 0)
+			return -ETIMEDOUT;
+	}
+	return 0;
+}
+
+static int spi_sh_wait_write_buffer_empty(struct spi_sh_data *ss)
+{
+	int timeout = 100000;
+
+	while (!(spi_sh_read(ss, SPI_SH_CR1) & SPI_SH_TBE)) {
+		udelay(10);
+		if (timeout-- < 0)
+			return -ETIMEDOUT;
+	}
+	return 0;
+}
+
+static int spi_sh_send(struct spi_sh_data *ss, struct spi_message *mesg,
+			struct spi_transfer *t)
+{
+	int i, retval = 0;
+	int remain = t->len;
+	int cur_len;
+	unsigned char *data;
+	long ret;
+
+	if (t->len)
+		spi_sh_set_bit(ss, SPI_SH_SSA, SPI_SH_CR1);
+
+	data = (unsigned char *)t->tx_buf;
+	while (remain > 0) {
+		cur_len = min(SPI_SH_FIFO_SIZE, remain);
+		for (i = 0; i < cur_len &&
+				!(spi_sh_read(ss, SPI_SH_CR4) &
+							SPI_SH_WPABRT) &&
+				!(spi_sh_read(ss, SPI_SH_CR1) & SPI_SH_TBF);
+				i++)
+			spi_sh_write(ss, (unsigned long)data[i], SPI_SH_TBR);
+
+		if (spi_sh_read(ss, SPI_SH_CR4) & SPI_SH_WPABRT) {
+			/* Abort SPI operation */
+			spi_sh_set_bit(ss, SPI_SH_WPABRT, SPI_SH_CR4);
+			retval = -EIO;
+			break;
+		}
+
+		cur_len = i;
+
+		remain -= cur_len;
+		data += cur_len;
+
+		if (remain > 0) {
+			ss->cr1 &= ~SPI_SH_TBE;
+			spi_sh_set_bit(ss, SPI_SH_TBE, SPI_SH_CR4);
+			ret = wait_event_interruptible_timeout(ss->wait,
+						 ss->cr1 & SPI_SH_TBE,
+						 SPI_SH_SEND_TIMEOUT);
+			if (ret == 0 && !(ss->cr1 & SPI_SH_TBE)) {
+				printk(KERN_ERR "%s: timeout\n", __func__);
+				return -ETIMEDOUT;
+			}
+		}
+	}
+
+	if (list_is_last(&t->transfer_list, &mesg->transfers)) {
+		spi_sh_clear_bit(ss, SPI_SH_SSD | SPI_SH_SSDB, SPI_SH_CR1);
+		spi_sh_set_bit(ss, SPI_SH_SSA, SPI_SH_CR1);
+
+		ss->cr1 &= ~SPI_SH_TBE;
+		spi_sh_set_bit(ss, SPI_SH_TBE, SPI_SH_CR4);
+		ret = wait_event_interruptible_timeout(ss->wait,
+					 ss->cr1 & SPI_SH_TBE,
+					 SPI_SH_SEND_TIMEOUT);
+		if (ret == 0 && (ss->cr1 & SPI_SH_TBE)) {
+			printk(KERN_ERR "%s: timeout\n", __func__);
+			return -ETIMEDOUT;
+		}
+	}
+
+	return retval;
+}
+
+static int spi_sh_receive(struct spi_sh_data *ss, struct spi_message *mesg,
+			  struct spi_transfer *t)
+{
+	int i;
+	int remain = t->len;
+	int cur_len;
+	unsigned char *data;
+	long ret;
+
+	if (t->len > SPI_SH_MAX_BYTE)
+		spi_sh_write(ss, SPI_SH_MAX_BYTE, SPI_SH_CR3);
+	else
+		spi_sh_write(ss, t->len, SPI_SH_CR3);
+
+	spi_sh_clear_bit(ss, SPI_SH_SSD | SPI_SH_SSDB, SPI_SH_CR1);
+	spi_sh_set_bit(ss, SPI_SH_SSA, SPI_SH_CR1);
+
+	spi_sh_wait_write_buffer_empty(ss);
+
+	data = (unsigned char *)t->rx_buf;
+	while (remain > 0) {
+		if (remain >= SPI_SH_FIFO_SIZE) {
+			ss->cr1 &= ~SPI_SH_RBF;
+			spi_sh_set_bit(ss, SPI_SH_RBF, SPI_SH_CR4);
+			ret = wait_event_interruptible_timeout(ss->wait,
+						 ss->cr1 & SPI_SH_RBF,
+						 SPI_SH_RECEIVE_TIMEOUT);
+			if (ret == 0 &&
+			    spi_sh_read(ss, SPI_SH_CR1) & SPI_SH_RBE) {
+				printk(KERN_ERR "%s: timeout\n", __func__);
+				return -ETIMEDOUT;
+			}
+		}
+
+		cur_len = min(SPI_SH_FIFO_SIZE, remain);
+		for (i = 0; i < cur_len; i++) {
+			if (spi_sh_wait_receive_buffer(ss))
+				break;
+			data[i] = (unsigned char)spi_sh_read(ss, SPI_SH_RBR);
+		}
+
+		remain -= cur_len;
+		data += cur_len;
+	}
+
+	/* deassert CS when SPI is receiving. */
+	if (t->len > SPI_SH_MAX_BYTE) {
+		clear_fifo(ss);
+		spi_sh_write(ss, 1, SPI_SH_CR3);
+	} else {
+		spi_sh_write(ss, 0, SPI_SH_CR3);
+	}
+
+	return 0;
+}
+
+static void spi_sh_work(struct work_struct *work)
+{
+	struct spi_sh_data *ss = container_of(work, struct spi_sh_data, ws);
+	struct spi_message *mesg;
+	struct spi_transfer *t;
+	unsigned long flags;
+	int ret;
+
+	pr_debug("%s: enter\n", __func__);
+
+	spin_lock_irqsave(&ss->lock, flags);
+	while (!list_empty(&ss->queue)) {
+		mesg = list_entry(ss->queue.next, struct spi_message, queue);
+		list_del_init(&mesg->queue);
+
+		spin_unlock_irqrestore(&ss->lock, flags);
+		list_for_each_entry(t, &mesg->transfers, transfer_list) {
+			pr_debug("tx_buf = %p, rx_buf = %p\n",
+					t->tx_buf, t->rx_buf);
+			pr_debug("len = %d, delay_usecs = %d\n",
+					t->len, t->delay_usecs);
+
+			if (t->tx_buf) {
+				ret = spi_sh_send(ss, mesg, t);
+				if (ret < 0)
+					goto error;
+			}
+			if (t->rx_buf) {
+				ret = spi_sh_receive(ss, mesg, t);
+				if (ret < 0)
+					goto error;
+			}
+			mesg->actual_length += t->len;
+		}
+		spin_lock_irqsave(&ss->lock, flags);
+
+		mesg->status = 0;
+		if (mesg->complete)
+			mesg->complete(mesg->context);
+	}
+
+	clear_fifo(ss);
+	spi_sh_set_bit(ss, SPI_SH_SSD, SPI_SH_CR1);
+	udelay(100);
+
+	spi_sh_clear_bit(ss, SPI_SH_SSA | SPI_SH_SSDB | SPI_SH_SSD,
+			 SPI_SH_CR1);
+
+	clear_fifo(ss);
+
+	spin_unlock_irqrestore(&ss->lock, flags);
+
+	return;
+
+ error:
+	mesg->status = ret;
+	if (mesg->complete)
+		mesg->complete(mesg->context);
+
+	spi_sh_clear_bit(ss, SPI_SH_SSA | SPI_SH_SSDB | SPI_SH_SSD,
+			 SPI_SH_CR1);
+	clear_fifo(ss);
+
+}
+
+static int spi_sh_setup(struct spi_device *spi)
+{
+	struct spi_sh_data *ss = spi_master_get_devdata(spi->master);
+
+	pr_debug("%s: enter\n", __func__);
+
+	spi_sh_write(ss, 0xfe, SPI_SH_CR1);	/* SPI sycle stop */
+	spi_sh_write(ss, 0x00, SPI_SH_CR1);	/* CR1 init */
+	spi_sh_write(ss, 0x00, SPI_SH_CR3);	/* CR3 init */
+
+	clear_fifo(ss);
+
+	/* 1/8 clock */
+	spi_sh_write(ss, spi_sh_read(ss, SPI_SH_CR2) | 0x07, SPI_SH_CR2);
+	udelay(10);
+
+	return 0;
+}
+
+static int spi_sh_transfer(struct spi_device *spi, struct spi_message *mesg)
+{
+	struct spi_sh_data *ss = spi_master_get_devdata(spi->master);
+	unsigned long flags;
+
+	pr_debug("%s: enter\n", __func__);
+	pr_debug("\tmode = %02x\n", spi->mode);
+
+	spin_lock_irqsave(&ss->lock, flags);
+
+	mesg->actual_length = 0;
+	mesg->status = -EINPROGRESS;
+
+	spi_sh_clear_bit(ss, SPI_SH_SSA, SPI_SH_CR1);
+
+	list_add_tail(&mesg->queue, &ss->queue);
+	schedule_work(&ss->ws);
+
+	spin_unlock_irqrestore(&ss->lock, flags);
+
+	return 0;
+}
+
+static void spi_sh_cleanup(struct spi_device *spi)
+{
+	struct spi_sh_data *ss = spi_master_get_devdata(spi->master);
+
+	pr_debug("%s: enter\n", __func__);
+
+	spi_sh_clear_bit(ss, SPI_SH_SSA | SPI_SH_SSDB | SPI_SH_SSD,
+			 SPI_SH_CR1);
+}
+
+static irqreturn_t spi_sh_irq(int irq, void *_ss)
+{
+	struct spi_sh_data *ss = (struct spi_sh_data *)_ss;
+	unsigned long cr1;
+
+	cr1 = spi_sh_read(ss, SPI_SH_CR1);
+	if (cr1 & SPI_SH_TBE)
+		ss->cr1 |= SPI_SH_TBE;
+	if (cr1 & SPI_SH_TBF)
+		ss->cr1 |= SPI_SH_TBF;
+	if (cr1 & SPI_SH_RBE)
+		ss->cr1 |= SPI_SH_RBE;
+	if (cr1 & SPI_SH_RBF)
+		ss->cr1 |= SPI_SH_RBF;
+
+	if (ss->cr1) {
+		spi_sh_clear_bit(ss, ss->cr1, SPI_SH_CR4);
+		wake_up(&ss->wait);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int spi_sh_remove(struct platform_device *pdev)
+{
+	struct spi_sh_data *ss = platform_get_drvdata(pdev);
+
+	spi_unregister_master(ss->master);
+	flush_work(&ss->ws);
+	free_irq(ss->irq, ss);
+
+	return 0;
+}
+
+static int spi_sh_probe(struct platform_device *pdev)
+{
+	struct resource *res;
+	struct spi_master *master;
+	struct spi_sh_data *ss;
+	int ret, irq;
+
+	/* get base addr */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (unlikely(res == NULL)) {
+		dev_err(&pdev->dev, "invalid resource\n");
+		return -EINVAL;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "platform_get_irq error: %d\n", irq);
+		return irq;
+	}
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct spi_sh_data));
+	if (master == NULL) {
+		dev_err(&pdev->dev, "spi_alloc_master error.\n");
+		return -ENOMEM;
+	}
+
+	ss = spi_master_get_devdata(master);
+	platform_set_drvdata(pdev, ss);
+
+	switch (res->flags & IORESOURCE_MEM_TYPE_MASK) {
+	case IORESOURCE_MEM_8BIT:
+		ss->width = 8;
+		break;
+	case IORESOURCE_MEM_32BIT:
+		ss->width = 32;
+		break;
+	default:
+		dev_err(&pdev->dev, "No support width\n");
+		ret = -ENODEV;
+		goto error1;
+	}
+	ss->irq = irq;
+	ss->master = master;
+	ss->addr = devm_ioremap(&pdev->dev, res->start, resource_size(res));
+	if (ss->addr == NULL) {
+		dev_err(&pdev->dev, "ioremap error.\n");
+		ret = -ENOMEM;
+		goto error1;
+	}
+	INIT_LIST_HEAD(&ss->queue);
+	spin_lock_init(&ss->lock);
+	INIT_WORK(&ss->ws, spi_sh_work);
+	init_waitqueue_head(&ss->wait);
+
+	ret = request_irq(irq, spi_sh_irq, 0, "spi_sh", ss);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "request_irq error\n");
+		goto error1;
+	}
+
+	master->num_chipselect = 2;
+	master->bus_num = pdev->id;
+	master->setup = spi_sh_setup;
+	master->transfer = spi_sh_transfer;
+	master->cleanup = spi_sh_cleanup;
+
+	ret = spi_register_master(master);
+	if (ret < 0) {
+		printk(KERN_ERR "spi_register_master error.\n");
+		goto error3;
+	}
+
+	return 0;
+
+ error3:
+	free_irq(irq, ss);
+ error1:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static struct platform_driver spi_sh_driver = {
+	.probe = spi_sh_probe,
+	.remove = spi_sh_remove,
+	.driver = {
+		.name = "sh_spi",
+	},
+};
+module_platform_driver(spi_sh_driver);
+
+MODULE_DESCRIPTION("SH SPI bus driver");
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Yoshihiro Shimoda");
+MODULE_ALIAS("platform:sh_spi");
diff --git a/drivers/spi/spi-sirf.c b/drivers/spi/spi-sirf.c
new file mode 100644
index 0000000..f009d76
--- /dev/null
+++ b/drivers/spi/spi-sirf.c
@@ -0,0 +1,1239 @@
+/*
+ * SPI bus driver for CSR SiRFprimaII
+ *
+ * Copyright (c) 2011 Cambridge Silicon Radio Limited, a CSR plc group company.
+ *
+ * Licensed under GPLv2 or later.
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/of.h>
+#include <linux/bitops.h>
+#include <linux/err.h>
+#include <linux/platform_device.h>
+#include <linux/of_gpio.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-direction.h>
+#include <linux/dma-mapping.h>
+#include <linux/reset.h>
+
+#define DRIVER_NAME "sirfsoc_spi"
+/* SPI CTRL register defines */
+#define SIRFSOC_SPI_SLV_MODE		BIT(16)
+#define SIRFSOC_SPI_CMD_MODE		BIT(17)
+#define SIRFSOC_SPI_CS_IO_OUT		BIT(18)
+#define SIRFSOC_SPI_CS_IO_MODE		BIT(19)
+#define SIRFSOC_SPI_CLK_IDLE_STAT	BIT(20)
+#define SIRFSOC_SPI_CS_IDLE_STAT	BIT(21)
+#define SIRFSOC_SPI_TRAN_MSB		BIT(22)
+#define SIRFSOC_SPI_DRV_POS_EDGE	BIT(23)
+#define SIRFSOC_SPI_CS_HOLD_TIME	BIT(24)
+#define SIRFSOC_SPI_CLK_SAMPLE_MODE	BIT(25)
+#define SIRFSOC_SPI_TRAN_DAT_FORMAT_8	(0 << 26)
+#define SIRFSOC_SPI_TRAN_DAT_FORMAT_12	(1 << 26)
+#define SIRFSOC_SPI_TRAN_DAT_FORMAT_16	(2 << 26)
+#define SIRFSOC_SPI_TRAN_DAT_FORMAT_32	(3 << 26)
+#define SIRFSOC_SPI_CMD_BYTE_NUM(x)	((x & 3) << 28)
+#define SIRFSOC_SPI_ENA_AUTO_CLR	BIT(30)
+#define SIRFSOC_SPI_MUL_DAT_MODE	BIT(31)
+
+/* Interrupt Enable */
+#define SIRFSOC_SPI_RX_DONE_INT_EN	BIT(0)
+#define SIRFSOC_SPI_TX_DONE_INT_EN	BIT(1)
+#define SIRFSOC_SPI_RX_OFLOW_INT_EN	BIT(2)
+#define SIRFSOC_SPI_TX_UFLOW_INT_EN	BIT(3)
+#define SIRFSOC_SPI_RX_IO_DMA_INT_EN	BIT(4)
+#define SIRFSOC_SPI_TX_IO_DMA_INT_EN	BIT(5)
+#define SIRFSOC_SPI_RXFIFO_FULL_INT_EN	BIT(6)
+#define SIRFSOC_SPI_TXFIFO_EMPTY_INT_EN	BIT(7)
+#define SIRFSOC_SPI_RXFIFO_THD_INT_EN	BIT(8)
+#define SIRFSOC_SPI_TXFIFO_THD_INT_EN	BIT(9)
+#define SIRFSOC_SPI_FRM_END_INT_EN	BIT(10)
+
+/* Interrupt status */
+#define SIRFSOC_SPI_RX_DONE		BIT(0)
+#define SIRFSOC_SPI_TX_DONE		BIT(1)
+#define SIRFSOC_SPI_RX_OFLOW		BIT(2)
+#define SIRFSOC_SPI_TX_UFLOW		BIT(3)
+#define SIRFSOC_SPI_RX_IO_DMA		BIT(4)
+#define SIRFSOC_SPI_RX_FIFO_FULL	BIT(6)
+#define SIRFSOC_SPI_TXFIFO_EMPTY	BIT(7)
+#define SIRFSOC_SPI_RXFIFO_THD_REACH	BIT(8)
+#define SIRFSOC_SPI_TXFIFO_THD_REACH	BIT(9)
+#define SIRFSOC_SPI_FRM_END		BIT(10)
+
+/* TX RX enable */
+#define SIRFSOC_SPI_RX_EN		BIT(0)
+#define SIRFSOC_SPI_TX_EN		BIT(1)
+#define SIRFSOC_SPI_CMD_TX_EN		BIT(2)
+
+#define SIRFSOC_SPI_IO_MODE_SEL		BIT(0)
+#define SIRFSOC_SPI_RX_DMA_FLUSH	BIT(2)
+
+/* FIFO OPs */
+#define SIRFSOC_SPI_FIFO_RESET		BIT(0)
+#define SIRFSOC_SPI_FIFO_START		BIT(1)
+
+/* FIFO CTRL */
+#define SIRFSOC_SPI_FIFO_WIDTH_BYTE	(0 << 0)
+#define SIRFSOC_SPI_FIFO_WIDTH_WORD	(1 << 0)
+#define SIRFSOC_SPI_FIFO_WIDTH_DWORD	(2 << 0)
+/* USP related */
+#define SIRFSOC_USP_SYNC_MODE		BIT(0)
+#define SIRFSOC_USP_SLV_MODE		BIT(1)
+#define SIRFSOC_USP_LSB			BIT(4)
+#define SIRFSOC_USP_EN			BIT(5)
+#define SIRFSOC_USP_RXD_FALLING_EDGE	BIT(6)
+#define SIRFSOC_USP_TXD_FALLING_EDGE	BIT(7)
+#define SIRFSOC_USP_CS_HIGH_VALID	BIT(9)
+#define SIRFSOC_USP_SCLK_IDLE_STAT	BIT(11)
+#define SIRFSOC_USP_TFS_IO_MODE		BIT(14)
+#define SIRFSOC_USP_TFS_IO_INPUT	BIT(19)
+
+#define SIRFSOC_USP_RXD_DELAY_LEN_MASK	0xFF
+#define SIRFSOC_USP_TXD_DELAY_LEN_MASK	0xFF
+#define SIRFSOC_USP_RXD_DELAY_OFFSET	0
+#define SIRFSOC_USP_TXD_DELAY_OFFSET	8
+#define SIRFSOC_USP_RXD_DELAY_LEN	1
+#define SIRFSOC_USP_TXD_DELAY_LEN	1
+#define SIRFSOC_USP_CLK_DIVISOR_OFFSET	21
+#define SIRFSOC_USP_CLK_DIVISOR_MASK	0x3FF
+#define SIRFSOC_USP_CLK_10_11_MASK	0x3
+#define SIRFSOC_USP_CLK_10_11_OFFSET	30
+#define SIRFSOC_USP_CLK_12_15_MASK	0xF
+#define SIRFSOC_USP_CLK_12_15_OFFSET	24
+
+#define SIRFSOC_USP_TX_DATA_OFFSET	0
+#define SIRFSOC_USP_TX_SYNC_OFFSET	8
+#define SIRFSOC_USP_TX_FRAME_OFFSET	16
+#define SIRFSOC_USP_TX_SHIFTER_OFFSET	24
+
+#define SIRFSOC_USP_TX_DATA_MASK	0xFF
+#define SIRFSOC_USP_TX_SYNC_MASK	0xFF
+#define SIRFSOC_USP_TX_FRAME_MASK	0xFF
+#define SIRFSOC_USP_TX_SHIFTER_MASK	0x1F
+
+#define SIRFSOC_USP_RX_DATA_OFFSET	0
+#define SIRFSOC_USP_RX_FRAME_OFFSET	8
+#define SIRFSOC_USP_RX_SHIFTER_OFFSET	16
+
+#define SIRFSOC_USP_RX_DATA_MASK	0xFF
+#define SIRFSOC_USP_RX_FRAME_MASK	0xFF
+#define SIRFSOC_USP_RX_SHIFTER_MASK	0x1F
+#define SIRFSOC_USP_CS_HIGH_VALUE	BIT(1)
+
+#define SIRFSOC_SPI_FIFO_SC_OFFSET	0
+#define SIRFSOC_SPI_FIFO_LC_OFFSET	10
+#define SIRFSOC_SPI_FIFO_HC_OFFSET	20
+
+#define SIRFSOC_SPI_FIFO_FULL_MASK(s)	(1 << ((s)->fifo_full_offset))
+#define SIRFSOC_SPI_FIFO_EMPTY_MASK(s)	(1 << ((s)->fifo_full_offset + 1))
+#define SIRFSOC_SPI_FIFO_THD_MASK(s)	((s)->fifo_size - 1)
+#define SIRFSOC_SPI_FIFO_THD_OFFSET	2
+#define SIRFSOC_SPI_FIFO_LEVEL_CHK_MASK(s, val)	\
+	((val) & (s)->fifo_level_chk_mask)
+
+enum sirf_spi_type {
+	SIRF_REAL_SPI,
+	SIRF_USP_SPI_P2,
+	SIRF_USP_SPI_A7,
+};
+
+/*
+ * only if the rx/tx buffer and transfer size are 4-bytes aligned, we use dma
+ * due to the limitation of dma controller
+ */
+
+#define ALIGNED(x) (!((u32)x & 0x3))
+#define IS_DMA_VALID(x) (x && ALIGNED(x->tx_buf) && ALIGNED(x->rx_buf) && \
+	ALIGNED(x->len) && (x->len < 2 * PAGE_SIZE))
+
+#define SIRFSOC_MAX_CMD_BYTES	4
+#define SIRFSOC_SPI_DEFAULT_FRQ 1000000
+
+struct sirf_spi_register {
+	/*SPI and USP-SPI common*/
+	u32 tx_rx_en;
+	u32 int_en;
+	u32 int_st;
+	u32 tx_dma_io_ctrl;
+	u32 tx_dma_io_len;
+	u32 txfifo_ctrl;
+	u32 txfifo_level_chk;
+	u32 txfifo_op;
+	u32 txfifo_st;
+	u32 txfifo_data;
+	u32 rx_dma_io_ctrl;
+	u32 rx_dma_io_len;
+	u32 rxfifo_ctrl;
+	u32 rxfifo_level_chk;
+	u32 rxfifo_op;
+	u32 rxfifo_st;
+	u32 rxfifo_data;
+	/*SPI self*/
+	u32 spi_ctrl;
+	u32 spi_cmd;
+	u32 spi_dummy_delay_ctrl;
+	/*USP-SPI self*/
+	u32 usp_mode1;
+	u32 usp_mode2;
+	u32 usp_tx_frame_ctrl;
+	u32 usp_rx_frame_ctrl;
+	u32 usp_pin_io_data;
+	u32 usp_risc_dsp_mode;
+	u32 usp_async_param_reg;
+	u32 usp_irda_x_mode_div;
+	u32 usp_sm_cfg;
+	u32 usp_int_en_clr;
+};
+
+static const struct sirf_spi_register real_spi_register = {
+	.tx_rx_en		= 0x8,
+	.int_en		= 0xc,
+	.int_st		= 0x10,
+	.tx_dma_io_ctrl	= 0x100,
+	.tx_dma_io_len	= 0x104,
+	.txfifo_ctrl	= 0x108,
+	.txfifo_level_chk	= 0x10c,
+	.txfifo_op		= 0x110,
+	.txfifo_st		= 0x114,
+	.txfifo_data	= 0x118,
+	.rx_dma_io_ctrl	= 0x120,
+	.rx_dma_io_len	= 0x124,
+	.rxfifo_ctrl	= 0x128,
+	.rxfifo_level_chk	= 0x12c,
+	.rxfifo_op		= 0x130,
+	.rxfifo_st		= 0x134,
+	.rxfifo_data	= 0x138,
+	.spi_ctrl		= 0x0,
+	.spi_cmd		= 0x4,
+	.spi_dummy_delay_ctrl	= 0x144,
+};
+
+static const struct sirf_spi_register usp_spi_register = {
+	.tx_rx_en		= 0x10,
+	.int_en		= 0x14,
+	.int_st		= 0x18,
+	.tx_dma_io_ctrl	= 0x100,
+	.tx_dma_io_len	= 0x104,
+	.txfifo_ctrl	= 0x108,
+	.txfifo_level_chk	= 0x10c,
+	.txfifo_op		= 0x110,
+	.txfifo_st		= 0x114,
+	.txfifo_data	= 0x118,
+	.rx_dma_io_ctrl	= 0x120,
+	.rx_dma_io_len	= 0x124,
+	.rxfifo_ctrl	= 0x128,
+	.rxfifo_level_chk	= 0x12c,
+	.rxfifo_op		= 0x130,
+	.rxfifo_st		= 0x134,
+	.rxfifo_data	= 0x138,
+	.usp_mode1		= 0x0,
+	.usp_mode2		= 0x4,
+	.usp_tx_frame_ctrl	= 0x8,
+	.usp_rx_frame_ctrl	= 0xc,
+	.usp_pin_io_data	= 0x1c,
+	.usp_risc_dsp_mode	= 0x20,
+	.usp_async_param_reg	= 0x24,
+	.usp_irda_x_mode_div	= 0x28,
+	.usp_sm_cfg		= 0x2c,
+	.usp_int_en_clr		= 0x140,
+};
+
+struct sirfsoc_spi {
+	struct spi_bitbang bitbang;
+	struct completion rx_done;
+	struct completion tx_done;
+
+	void __iomem *base;
+	u32 ctrl_freq;  /* SPI controller clock speed */
+	struct clk *clk;
+
+	/* rx & tx bufs from the spi_transfer */
+	const void *tx;
+	void *rx;
+
+	/* place received word into rx buffer */
+	void (*rx_word) (struct sirfsoc_spi *);
+	/* get word from tx buffer for sending */
+	void (*tx_word) (struct sirfsoc_spi *);
+
+	/* number of words left to be tranmitted/received */
+	unsigned int left_tx_word;
+	unsigned int left_rx_word;
+
+	/* rx & tx DMA channels */
+	struct dma_chan *rx_chan;
+	struct dma_chan *tx_chan;
+	dma_addr_t src_start;
+	dma_addr_t dst_start;
+	int word_width; /* in bytes */
+
+	/*
+	 * if tx size is not more than 4 and rx size is NULL, use
+	 * command model
+	 */
+	bool	tx_by_cmd;
+	bool	hw_cs;
+	enum sirf_spi_type type;
+	const struct sirf_spi_register *regs;
+	unsigned int fifo_size;
+	/* fifo empty offset is (fifo full offset + 1)*/
+	unsigned int fifo_full_offset;
+	/* fifo_level_chk_mask is (fifo_size/4 - 1) */
+	unsigned int fifo_level_chk_mask;
+	unsigned int dat_max_frm_len;
+};
+
+struct sirf_spi_comp_data {
+	const struct sirf_spi_register *regs;
+	enum sirf_spi_type type;
+	unsigned int dat_max_frm_len;
+	unsigned int fifo_size;
+	void (*hwinit)(struct sirfsoc_spi *sspi);
+};
+
+static void sirfsoc_usp_hwinit(struct sirfsoc_spi *sspi)
+{
+	/* reset USP and let USP can operate */
+	writel(readl(sspi->base + sspi->regs->usp_mode1) &
+		~SIRFSOC_USP_EN, sspi->base + sspi->regs->usp_mode1);
+	writel(readl(sspi->base + sspi->regs->usp_mode1) |
+		SIRFSOC_USP_EN, sspi->base + sspi->regs->usp_mode1);
+}
+
+static void spi_sirfsoc_rx_word_u8(struct sirfsoc_spi *sspi)
+{
+	u32 data;
+	u8 *rx = sspi->rx;
+
+	data = readl(sspi->base + sspi->regs->rxfifo_data);
+
+	if (rx) {
+		*rx++ = (u8) data;
+		sspi->rx = rx;
+	}
+
+	sspi->left_rx_word--;
+}
+
+static void spi_sirfsoc_tx_word_u8(struct sirfsoc_spi *sspi)
+{
+	u32 data = 0;
+	const u8 *tx = sspi->tx;
+
+	if (tx) {
+		data = *tx++;
+		sspi->tx = tx;
+	}
+	writel(data, sspi->base + sspi->regs->txfifo_data);
+	sspi->left_tx_word--;
+}
+
+static void spi_sirfsoc_rx_word_u16(struct sirfsoc_spi *sspi)
+{
+	u32 data;
+	u16 *rx = sspi->rx;
+
+	data = readl(sspi->base + sspi->regs->rxfifo_data);
+
+	if (rx) {
+		*rx++ = (u16) data;
+		sspi->rx = rx;
+	}
+
+	sspi->left_rx_word--;
+}
+
+static void spi_sirfsoc_tx_word_u16(struct sirfsoc_spi *sspi)
+{
+	u32 data = 0;
+	const u16 *tx = sspi->tx;
+
+	if (tx) {
+		data = *tx++;
+		sspi->tx = tx;
+	}
+
+	writel(data, sspi->base + sspi->regs->txfifo_data);
+	sspi->left_tx_word--;
+}
+
+static void spi_sirfsoc_rx_word_u32(struct sirfsoc_spi *sspi)
+{
+	u32 data;
+	u32 *rx = sspi->rx;
+
+	data = readl(sspi->base + sspi->regs->rxfifo_data);
+
+	if (rx) {
+		*rx++ = (u32) data;
+		sspi->rx = rx;
+	}
+
+	sspi->left_rx_word--;
+
+}
+
+static void spi_sirfsoc_tx_word_u32(struct sirfsoc_spi *sspi)
+{
+	u32 data = 0;
+	const u32 *tx = sspi->tx;
+
+	if (tx) {
+		data = *tx++;
+		sspi->tx = tx;
+	}
+
+	writel(data, sspi->base + sspi->regs->txfifo_data);
+	sspi->left_tx_word--;
+}
+
+static irqreturn_t spi_sirfsoc_irq(int irq, void *dev_id)
+{
+	struct sirfsoc_spi *sspi = dev_id;
+	u32 spi_stat;
+
+	spi_stat = readl(sspi->base + sspi->regs->int_st);
+	if (sspi->tx_by_cmd && sspi->type == SIRF_REAL_SPI
+		&& (spi_stat & SIRFSOC_SPI_FRM_END)) {
+		complete(&sspi->tx_done);
+		writel(0x0, sspi->base + sspi->regs->int_en);
+		writel(readl(sspi->base + sspi->regs->int_st),
+				sspi->base + sspi->regs->int_st);
+		return IRQ_HANDLED;
+	}
+	/* Error Conditions */
+	if (spi_stat & SIRFSOC_SPI_RX_OFLOW ||
+			spi_stat & SIRFSOC_SPI_TX_UFLOW) {
+		complete(&sspi->tx_done);
+		complete(&sspi->rx_done);
+		switch (sspi->type) {
+		case SIRF_REAL_SPI:
+		case SIRF_USP_SPI_P2:
+			writel(0x0, sspi->base + sspi->regs->int_en);
+			break;
+		case SIRF_USP_SPI_A7:
+			writel(~0UL, sspi->base + sspi->regs->usp_int_en_clr);
+			break;
+		}
+		writel(readl(sspi->base + sspi->regs->int_st),
+				sspi->base + sspi->regs->int_st);
+		return IRQ_HANDLED;
+	}
+	if (spi_stat & SIRFSOC_SPI_TXFIFO_EMPTY)
+		complete(&sspi->tx_done);
+	while (!(readl(sspi->base + sspi->regs->int_st) &
+		SIRFSOC_SPI_RX_IO_DMA))
+		cpu_relax();
+	complete(&sspi->rx_done);
+	switch (sspi->type) {
+	case SIRF_REAL_SPI:
+	case SIRF_USP_SPI_P2:
+		writel(0x0, sspi->base + sspi->regs->int_en);
+		break;
+	case SIRF_USP_SPI_A7:
+		writel(~0UL, sspi->base + sspi->regs->usp_int_en_clr);
+		break;
+	}
+	writel(readl(sspi->base + sspi->regs->int_st),
+			sspi->base + sspi->regs->int_st);
+
+	return IRQ_HANDLED;
+}
+
+static void spi_sirfsoc_dma_fini_callback(void *data)
+{
+	struct completion *dma_complete = data;
+
+	complete(dma_complete);
+}
+
+static void spi_sirfsoc_cmd_transfer(struct spi_device *spi,
+	struct spi_transfer *t)
+{
+	struct sirfsoc_spi *sspi;
+	int timeout = t->len * 10;
+	u32 cmd;
+
+	sspi = spi_master_get_devdata(spi->master);
+	writel(SIRFSOC_SPI_FIFO_RESET, sspi->base + sspi->regs->txfifo_op);
+	writel(SIRFSOC_SPI_FIFO_START, sspi->base + sspi->regs->txfifo_op);
+	memcpy(&cmd, sspi->tx, t->len);
+	if (sspi->word_width == 1 && !(spi->mode & SPI_LSB_FIRST))
+		cmd = cpu_to_be32(cmd) >>
+			((SIRFSOC_MAX_CMD_BYTES - t->len) * 8);
+	if (sspi->word_width == 2 && t->len == 4 &&
+			(!(spi->mode & SPI_LSB_FIRST)))
+		cmd = ((cmd & 0xffff) << 16) | (cmd >> 16);
+	writel(cmd, sspi->base + sspi->regs->spi_cmd);
+	writel(SIRFSOC_SPI_FRM_END_INT_EN,
+		sspi->base + sspi->regs->int_en);
+	writel(SIRFSOC_SPI_CMD_TX_EN,
+		sspi->base + sspi->regs->tx_rx_en);
+	if (wait_for_completion_timeout(&sspi->tx_done, timeout) == 0) {
+		dev_err(&spi->dev, "cmd transfer timeout\n");
+		return;
+	}
+	sspi->left_rx_word -= t->len;
+}
+
+static void spi_sirfsoc_dma_transfer(struct spi_device *spi,
+	struct spi_transfer *t)
+{
+	struct sirfsoc_spi *sspi;
+	struct dma_async_tx_descriptor *rx_desc, *tx_desc;
+	int timeout = t->len * 10;
+
+	sspi = spi_master_get_devdata(spi->master);
+	writel(SIRFSOC_SPI_FIFO_RESET, sspi->base + sspi->regs->rxfifo_op);
+	writel(SIRFSOC_SPI_FIFO_RESET, sspi->base + sspi->regs->txfifo_op);
+	switch (sspi->type) {
+	case SIRF_REAL_SPI:
+		writel(SIRFSOC_SPI_FIFO_START,
+			sspi->base + sspi->regs->rxfifo_op);
+		writel(SIRFSOC_SPI_FIFO_START,
+			sspi->base + sspi->regs->txfifo_op);
+		writel(0, sspi->base + sspi->regs->int_en);
+		break;
+	case SIRF_USP_SPI_P2:
+		writel(0x0, sspi->base + sspi->regs->rxfifo_op);
+		writel(0x0, sspi->base + sspi->regs->txfifo_op);
+		writel(0, sspi->base + sspi->regs->int_en);
+		break;
+	case SIRF_USP_SPI_A7:
+		writel(0x0, sspi->base + sspi->regs->rxfifo_op);
+		writel(0x0, sspi->base + sspi->regs->txfifo_op);
+		writel(~0UL, sspi->base + sspi->regs->usp_int_en_clr);
+		break;
+	}
+	writel(readl(sspi->base + sspi->regs->int_st),
+		sspi->base + sspi->regs->int_st);
+	if (sspi->left_tx_word < sspi->dat_max_frm_len) {
+		switch (sspi->type) {
+		case SIRF_REAL_SPI:
+			writel(readl(sspi->base + sspi->regs->spi_ctrl) |
+				SIRFSOC_SPI_ENA_AUTO_CLR |
+				SIRFSOC_SPI_MUL_DAT_MODE,
+				sspi->base + sspi->regs->spi_ctrl);
+			writel(sspi->left_tx_word - 1,
+				sspi->base + sspi->regs->tx_dma_io_len);
+			writel(sspi->left_tx_word - 1,
+				sspi->base + sspi->regs->rx_dma_io_len);
+			break;
+		case SIRF_USP_SPI_P2:
+		case SIRF_USP_SPI_A7:
+			/*USP simulate SPI, tx/rx_dma_io_len indicates bytes*/
+			writel(sspi->left_tx_word * sspi->word_width,
+				sspi->base + sspi->regs->tx_dma_io_len);
+			writel(sspi->left_tx_word * sspi->word_width,
+				sspi->base + sspi->regs->rx_dma_io_len);
+			break;
+		}
+	} else {
+		if (sspi->type == SIRF_REAL_SPI)
+			writel(readl(sspi->base + sspi->regs->spi_ctrl),
+				sspi->base + sspi->regs->spi_ctrl);
+		writel(0, sspi->base + sspi->regs->tx_dma_io_len);
+		writel(0, sspi->base + sspi->regs->rx_dma_io_len);
+	}
+	sspi->dst_start = dma_map_single(&spi->dev, sspi->rx, t->len,
+					(t->tx_buf != t->rx_buf) ?
+					DMA_FROM_DEVICE : DMA_BIDIRECTIONAL);
+	rx_desc = dmaengine_prep_slave_single(sspi->rx_chan,
+		sspi->dst_start, t->len, DMA_DEV_TO_MEM,
+		DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	rx_desc->callback = spi_sirfsoc_dma_fini_callback;
+	rx_desc->callback_param = &sspi->rx_done;
+
+	sspi->src_start = dma_map_single(&spi->dev, (void *)sspi->tx, t->len,
+					(t->tx_buf != t->rx_buf) ?
+					DMA_TO_DEVICE : DMA_BIDIRECTIONAL);
+	tx_desc = dmaengine_prep_slave_single(sspi->tx_chan,
+		sspi->src_start, t->len, DMA_MEM_TO_DEV,
+		DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	tx_desc->callback = spi_sirfsoc_dma_fini_callback;
+	tx_desc->callback_param = &sspi->tx_done;
+
+	dmaengine_submit(tx_desc);
+	dmaengine_submit(rx_desc);
+	dma_async_issue_pending(sspi->tx_chan);
+	dma_async_issue_pending(sspi->rx_chan);
+	writel(SIRFSOC_SPI_RX_EN | SIRFSOC_SPI_TX_EN,
+			sspi->base + sspi->regs->tx_rx_en);
+	if (sspi->type == SIRF_USP_SPI_P2 ||
+		sspi->type == SIRF_USP_SPI_A7) {
+		writel(SIRFSOC_SPI_FIFO_START,
+			sspi->base + sspi->regs->rxfifo_op);
+		writel(SIRFSOC_SPI_FIFO_START,
+			sspi->base + sspi->regs->txfifo_op);
+	}
+	if (wait_for_completion_timeout(&sspi->rx_done, timeout) == 0) {
+		dev_err(&spi->dev, "transfer timeout\n");
+		dmaengine_terminate_all(sspi->rx_chan);
+	} else
+		sspi->left_rx_word = 0;
+	/*
+	 * we only wait tx-done event if transferring by DMA. for PIO,
+	 * we get rx data by writing tx data, so if rx is done, tx has
+	 * done earlier
+	 */
+	if (wait_for_completion_timeout(&sspi->tx_done, timeout) == 0) {
+		dev_err(&spi->dev, "transfer timeout\n");
+		if (sspi->type == SIRF_USP_SPI_P2 ||
+			sspi->type == SIRF_USP_SPI_A7)
+			writel(0, sspi->base + sspi->regs->tx_rx_en);
+		dmaengine_terminate_all(sspi->tx_chan);
+	}
+	dma_unmap_single(&spi->dev, sspi->src_start, t->len, DMA_TO_DEVICE);
+	dma_unmap_single(&spi->dev, sspi->dst_start, t->len, DMA_FROM_DEVICE);
+	/* TX, RX FIFO stop */
+	writel(0, sspi->base + sspi->regs->rxfifo_op);
+	writel(0, sspi->base + sspi->regs->txfifo_op);
+	if (sspi->left_tx_word >= sspi->dat_max_frm_len)
+		writel(0, sspi->base + sspi->regs->tx_rx_en);
+	if (sspi->type == SIRF_USP_SPI_P2 ||
+		sspi->type == SIRF_USP_SPI_A7)
+		writel(0, sspi->base + sspi->regs->tx_rx_en);
+}
+
+static void spi_sirfsoc_pio_transfer(struct spi_device *spi,
+		struct spi_transfer *t)
+{
+	struct sirfsoc_spi *sspi;
+	int timeout = t->len * 10;
+	unsigned int data_units;
+
+	sspi = spi_master_get_devdata(spi->master);
+	do {
+		writel(SIRFSOC_SPI_FIFO_RESET,
+			sspi->base + sspi->regs->rxfifo_op);
+		writel(SIRFSOC_SPI_FIFO_RESET,
+			sspi->base + sspi->regs->txfifo_op);
+		switch (sspi->type) {
+		case SIRF_USP_SPI_P2:
+			writel(0x0, sspi->base + sspi->regs->rxfifo_op);
+			writel(0x0, sspi->base + sspi->regs->txfifo_op);
+			writel(0, sspi->base + sspi->regs->int_en);
+			writel(readl(sspi->base + sspi->regs->int_st),
+				sspi->base + sspi->regs->int_st);
+			writel(min((sspi->left_tx_word * sspi->word_width),
+				sspi->fifo_size),
+				sspi->base + sspi->regs->tx_dma_io_len);
+			writel(min((sspi->left_rx_word * sspi->word_width),
+				sspi->fifo_size),
+				sspi->base + sspi->regs->rx_dma_io_len);
+			break;
+		case SIRF_USP_SPI_A7:
+			writel(0x0, sspi->base + sspi->regs->rxfifo_op);
+			writel(0x0, sspi->base + sspi->regs->txfifo_op);
+			writel(~0UL, sspi->base + sspi->regs->usp_int_en_clr);
+			writel(readl(sspi->base + sspi->regs->int_st),
+				sspi->base + sspi->regs->int_st);
+			writel(min((sspi->left_tx_word * sspi->word_width),
+				sspi->fifo_size),
+				sspi->base + sspi->regs->tx_dma_io_len);
+			writel(min((sspi->left_rx_word * sspi->word_width),
+				sspi->fifo_size),
+				sspi->base + sspi->regs->rx_dma_io_len);
+			break;
+		case SIRF_REAL_SPI:
+			writel(SIRFSOC_SPI_FIFO_START,
+				sspi->base + sspi->regs->rxfifo_op);
+			writel(SIRFSOC_SPI_FIFO_START,
+				sspi->base + sspi->regs->txfifo_op);
+			writel(0, sspi->base + sspi->regs->int_en);
+			writel(readl(sspi->base + sspi->regs->int_st),
+				sspi->base + sspi->regs->int_st);
+			writel(readl(sspi->base + sspi->regs->spi_ctrl) |
+				SIRFSOC_SPI_MUL_DAT_MODE |
+				SIRFSOC_SPI_ENA_AUTO_CLR,
+				sspi->base + sspi->regs->spi_ctrl);
+			data_units = sspi->fifo_size / sspi->word_width;
+			writel(min(sspi->left_tx_word, data_units) - 1,
+				sspi->base + sspi->regs->tx_dma_io_len);
+			writel(min(sspi->left_rx_word, data_units) - 1,
+				sspi->base + sspi->regs->rx_dma_io_len);
+			break;
+		}
+		while (!((readl(sspi->base + sspi->regs->txfifo_st)
+			& SIRFSOC_SPI_FIFO_FULL_MASK(sspi))) &&
+			sspi->left_tx_word)
+			sspi->tx_word(sspi);
+		writel(SIRFSOC_SPI_TXFIFO_EMPTY_INT_EN |
+			SIRFSOC_SPI_TX_UFLOW_INT_EN |
+			SIRFSOC_SPI_RX_OFLOW_INT_EN |
+			SIRFSOC_SPI_RX_IO_DMA_INT_EN,
+			sspi->base + sspi->regs->int_en);
+		writel(SIRFSOC_SPI_RX_EN | SIRFSOC_SPI_TX_EN,
+			sspi->base + sspi->regs->tx_rx_en);
+		if (sspi->type == SIRF_USP_SPI_P2 ||
+			sspi->type == SIRF_USP_SPI_A7) {
+			writel(SIRFSOC_SPI_FIFO_START,
+				sspi->base + sspi->regs->rxfifo_op);
+			writel(SIRFSOC_SPI_FIFO_START,
+				sspi->base + sspi->regs->txfifo_op);
+		}
+		if (!wait_for_completion_timeout(&sspi->tx_done, timeout) ||
+			!wait_for_completion_timeout(&sspi->rx_done, timeout)) {
+			dev_err(&spi->dev, "transfer timeout\n");
+			if (sspi->type == SIRF_USP_SPI_P2 ||
+				sspi->type == SIRF_USP_SPI_A7)
+				writel(0, sspi->base + sspi->regs->tx_rx_en);
+			break;
+		}
+		while (!((readl(sspi->base + sspi->regs->rxfifo_st)
+			& SIRFSOC_SPI_FIFO_EMPTY_MASK(sspi))) &&
+			sspi->left_rx_word)
+			sspi->rx_word(sspi);
+		if (sspi->type == SIRF_USP_SPI_P2 ||
+			sspi->type == SIRF_USP_SPI_A7)
+			writel(0, sspi->base + sspi->regs->tx_rx_en);
+		writel(0, sspi->base + sspi->regs->rxfifo_op);
+		writel(0, sspi->base + sspi->regs->txfifo_op);
+	} while (sspi->left_tx_word != 0 || sspi->left_rx_word != 0);
+}
+
+static int spi_sirfsoc_transfer(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct sirfsoc_spi *sspi;
+
+	sspi = spi_master_get_devdata(spi->master);
+	sspi->tx = t->tx_buf;
+	sspi->rx = t->rx_buf;
+	sspi->left_tx_word = sspi->left_rx_word = t->len / sspi->word_width;
+	reinit_completion(&sspi->rx_done);
+	reinit_completion(&sspi->tx_done);
+	/*
+	 * in the transfer, if transfer data using command register with rx_buf
+	 * null, just fill command data into command register and wait for its
+	 * completion.
+	 */
+	if (sspi->type == SIRF_REAL_SPI && sspi->tx_by_cmd)
+		spi_sirfsoc_cmd_transfer(spi, t);
+	else if (IS_DMA_VALID(t))
+		spi_sirfsoc_dma_transfer(spi, t);
+	else
+		spi_sirfsoc_pio_transfer(spi, t);
+
+	return t->len - sspi->left_rx_word * sspi->word_width;
+}
+
+static void spi_sirfsoc_chipselect(struct spi_device *spi, int value)
+{
+	struct sirfsoc_spi *sspi = spi_master_get_devdata(spi->master);
+
+	if (sspi->hw_cs) {
+		u32 regval;
+
+		switch (sspi->type) {
+		case SIRF_REAL_SPI:
+			regval = readl(sspi->base + sspi->regs->spi_ctrl);
+			switch (value) {
+			case BITBANG_CS_ACTIVE:
+				if (spi->mode & SPI_CS_HIGH)
+					regval |= SIRFSOC_SPI_CS_IO_OUT;
+				else
+					regval &= ~SIRFSOC_SPI_CS_IO_OUT;
+				break;
+			case BITBANG_CS_INACTIVE:
+				if (spi->mode & SPI_CS_HIGH)
+					regval &= ~SIRFSOC_SPI_CS_IO_OUT;
+				else
+					regval |= SIRFSOC_SPI_CS_IO_OUT;
+				break;
+			}
+			writel(regval, sspi->base + sspi->regs->spi_ctrl);
+			break;
+		case SIRF_USP_SPI_P2:
+		case SIRF_USP_SPI_A7:
+			regval = readl(sspi->base +
+					sspi->regs->usp_pin_io_data);
+			switch (value) {
+			case BITBANG_CS_ACTIVE:
+				if (spi->mode & SPI_CS_HIGH)
+					regval |= SIRFSOC_USP_CS_HIGH_VALUE;
+				else
+					regval &= ~(SIRFSOC_USP_CS_HIGH_VALUE);
+				break;
+			case BITBANG_CS_INACTIVE:
+				if (spi->mode & SPI_CS_HIGH)
+					regval &= ~(SIRFSOC_USP_CS_HIGH_VALUE);
+				else
+					regval |= SIRFSOC_USP_CS_HIGH_VALUE;
+				break;
+			}
+			writel(regval,
+				sspi->base + sspi->regs->usp_pin_io_data);
+			break;
+		}
+	} else {
+		switch (value) {
+		case BITBANG_CS_ACTIVE:
+			gpio_direction_output(spi->cs_gpio,
+					spi->mode & SPI_CS_HIGH ? 1 : 0);
+			break;
+		case BITBANG_CS_INACTIVE:
+			gpio_direction_output(spi->cs_gpio,
+					spi->mode & SPI_CS_HIGH ? 0 : 1);
+			break;
+		}
+	}
+}
+
+static int spi_sirfsoc_config_mode(struct spi_device *spi)
+{
+	struct sirfsoc_spi *sspi;
+	u32 regval, usp_mode1;
+
+	sspi = spi_master_get_devdata(spi->master);
+	regval = readl(sspi->base + sspi->regs->spi_ctrl);
+	usp_mode1 = readl(sspi->base + sspi->regs->usp_mode1);
+	if (!(spi->mode & SPI_CS_HIGH)) {
+		regval |= SIRFSOC_SPI_CS_IDLE_STAT;
+		usp_mode1 &= ~SIRFSOC_USP_CS_HIGH_VALID;
+	} else {
+		regval &= ~SIRFSOC_SPI_CS_IDLE_STAT;
+		usp_mode1 |= SIRFSOC_USP_CS_HIGH_VALID;
+	}
+	if (!(spi->mode & SPI_LSB_FIRST)) {
+		regval |= SIRFSOC_SPI_TRAN_MSB;
+		usp_mode1 &= ~SIRFSOC_USP_LSB;
+	} else {
+		regval &= ~SIRFSOC_SPI_TRAN_MSB;
+		usp_mode1 |= SIRFSOC_USP_LSB;
+	}
+	if (spi->mode & SPI_CPOL) {
+		regval |= SIRFSOC_SPI_CLK_IDLE_STAT;
+		usp_mode1 |= SIRFSOC_USP_SCLK_IDLE_STAT;
+	} else {
+		regval &= ~SIRFSOC_SPI_CLK_IDLE_STAT;
+		usp_mode1 &= ~SIRFSOC_USP_SCLK_IDLE_STAT;
+	}
+	/*
+	 * Data should be driven at least 1/2 cycle before the fetch edge
+	 * to make sure that data gets stable at the fetch edge.
+	 */
+	if (((spi->mode & SPI_CPOL) && (spi->mode & SPI_CPHA)) ||
+	    (!(spi->mode & SPI_CPOL) && !(spi->mode & SPI_CPHA))) {
+		regval &= ~SIRFSOC_SPI_DRV_POS_EDGE;
+		usp_mode1 |= (SIRFSOC_USP_TXD_FALLING_EDGE |
+				SIRFSOC_USP_RXD_FALLING_EDGE);
+	} else {
+		regval |= SIRFSOC_SPI_DRV_POS_EDGE;
+		usp_mode1 &= ~(SIRFSOC_USP_RXD_FALLING_EDGE |
+				SIRFSOC_USP_TXD_FALLING_EDGE);
+	}
+	writel((SIRFSOC_SPI_FIFO_LEVEL_CHK_MASK(sspi, sspi->fifo_size - 2) <<
+		SIRFSOC_SPI_FIFO_SC_OFFSET) |
+		(SIRFSOC_SPI_FIFO_LEVEL_CHK_MASK(sspi, sspi->fifo_size / 2) <<
+		SIRFSOC_SPI_FIFO_LC_OFFSET) |
+		(SIRFSOC_SPI_FIFO_LEVEL_CHK_MASK(sspi, 2) <<
+		SIRFSOC_SPI_FIFO_HC_OFFSET),
+		sspi->base + sspi->regs->txfifo_level_chk);
+	writel((SIRFSOC_SPI_FIFO_LEVEL_CHK_MASK(sspi, 2) <<
+		SIRFSOC_SPI_FIFO_SC_OFFSET) |
+		(SIRFSOC_SPI_FIFO_LEVEL_CHK_MASK(sspi, sspi->fifo_size / 2) <<
+		SIRFSOC_SPI_FIFO_LC_OFFSET) |
+		(SIRFSOC_SPI_FIFO_LEVEL_CHK_MASK(sspi, sspi->fifo_size - 2) <<
+		SIRFSOC_SPI_FIFO_HC_OFFSET),
+		sspi->base + sspi->regs->rxfifo_level_chk);
+	/*
+	 * it should never set to hardware cs mode because in hardware cs mode,
+	 * cs signal can't controlled by driver.
+	 */
+	switch (sspi->type) {
+	case SIRF_REAL_SPI:
+		regval |= SIRFSOC_SPI_CS_IO_MODE;
+		writel(regval, sspi->base + sspi->regs->spi_ctrl);
+		break;
+	case SIRF_USP_SPI_P2:
+	case SIRF_USP_SPI_A7:
+		usp_mode1 |= SIRFSOC_USP_SYNC_MODE;
+		usp_mode1 |= SIRFSOC_USP_TFS_IO_MODE;
+		usp_mode1 &= ~SIRFSOC_USP_TFS_IO_INPUT;
+		writel(usp_mode1, sspi->base + sspi->regs->usp_mode1);
+		break;
+	}
+
+	return 0;
+}
+
+static int
+spi_sirfsoc_setup_transfer(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct sirfsoc_spi *sspi;
+	u8 bits_per_word = 0;
+	int hz = 0;
+	u32 regval, txfifo_ctrl, rxfifo_ctrl, tx_frm_ctl, rx_frm_ctl, usp_mode2;
+
+	sspi = spi_master_get_devdata(spi->master);
+
+	bits_per_word = (t) ? t->bits_per_word : spi->bits_per_word;
+	hz = t && t->speed_hz ? t->speed_hz : spi->max_speed_hz;
+
+	usp_mode2 = regval = (sspi->ctrl_freq / (2 * hz)) - 1;
+	if (regval > 0xFFFF || regval < 0) {
+		dev_err(&spi->dev, "Speed %d not supported\n", hz);
+		return -EINVAL;
+	}
+	switch (bits_per_word) {
+	case 8:
+		regval |= SIRFSOC_SPI_TRAN_DAT_FORMAT_8;
+		sspi->rx_word = spi_sirfsoc_rx_word_u8;
+		sspi->tx_word = spi_sirfsoc_tx_word_u8;
+		break;
+	case 12:
+	case 16:
+		regval |= (bits_per_word ==  12) ?
+			SIRFSOC_SPI_TRAN_DAT_FORMAT_12 :
+			SIRFSOC_SPI_TRAN_DAT_FORMAT_16;
+		sspi->rx_word = spi_sirfsoc_rx_word_u16;
+		sspi->tx_word = spi_sirfsoc_tx_word_u16;
+		break;
+	case 32:
+		regval |= SIRFSOC_SPI_TRAN_DAT_FORMAT_32;
+		sspi->rx_word = spi_sirfsoc_rx_word_u32;
+		sspi->tx_word = spi_sirfsoc_tx_word_u32;
+		break;
+	default:
+		dev_err(&spi->dev, "bpw %d not supported\n", bits_per_word);
+		return -EINVAL;
+	}
+	sspi->word_width = DIV_ROUND_UP(bits_per_word, 8);
+	txfifo_ctrl = (((sspi->fifo_size / 2) &
+			SIRFSOC_SPI_FIFO_THD_MASK(sspi))
+			<< SIRFSOC_SPI_FIFO_THD_OFFSET) |
+			(sspi->word_width >> 1);
+	rxfifo_ctrl = (((sspi->fifo_size / 2) &
+			SIRFSOC_SPI_FIFO_THD_MASK(sspi))
+			<< SIRFSOC_SPI_FIFO_THD_OFFSET) |
+			(sspi->word_width >> 1);
+	writel(txfifo_ctrl, sspi->base + sspi->regs->txfifo_ctrl);
+	writel(rxfifo_ctrl, sspi->base + sspi->regs->rxfifo_ctrl);
+	if (sspi->type == SIRF_USP_SPI_P2 ||
+		sspi->type == SIRF_USP_SPI_A7) {
+		tx_frm_ctl = 0;
+		tx_frm_ctl |= ((bits_per_word - 1) & SIRFSOC_USP_TX_DATA_MASK)
+				<< SIRFSOC_USP_TX_DATA_OFFSET;
+		tx_frm_ctl |= ((bits_per_word + 1 + SIRFSOC_USP_TXD_DELAY_LEN
+				- 1) & SIRFSOC_USP_TX_SYNC_MASK) <<
+				SIRFSOC_USP_TX_SYNC_OFFSET;
+		tx_frm_ctl |= ((bits_per_word + 1 + SIRFSOC_USP_TXD_DELAY_LEN
+				+ 2 - 1) & SIRFSOC_USP_TX_FRAME_MASK) <<
+				SIRFSOC_USP_TX_FRAME_OFFSET;
+		tx_frm_ctl |= ((bits_per_word - 1) &
+				SIRFSOC_USP_TX_SHIFTER_MASK) <<
+				SIRFSOC_USP_TX_SHIFTER_OFFSET;
+		rx_frm_ctl = 0;
+		rx_frm_ctl |= ((bits_per_word - 1) & SIRFSOC_USP_RX_DATA_MASK)
+				<< SIRFSOC_USP_RX_DATA_OFFSET;
+		rx_frm_ctl |= ((bits_per_word + 1 + SIRFSOC_USP_RXD_DELAY_LEN
+				+ 2 - 1) & SIRFSOC_USP_RX_FRAME_MASK) <<
+				SIRFSOC_USP_RX_FRAME_OFFSET;
+		rx_frm_ctl |= ((bits_per_word - 1)
+				& SIRFSOC_USP_RX_SHIFTER_MASK) <<
+				SIRFSOC_USP_RX_SHIFTER_OFFSET;
+		writel(tx_frm_ctl | (((usp_mode2 >> 10) &
+			SIRFSOC_USP_CLK_10_11_MASK) <<
+			SIRFSOC_USP_CLK_10_11_OFFSET),
+			sspi->base + sspi->regs->usp_tx_frame_ctrl);
+		writel(rx_frm_ctl | (((usp_mode2 >> 12) &
+			SIRFSOC_USP_CLK_12_15_MASK) <<
+			SIRFSOC_USP_CLK_12_15_OFFSET),
+			sspi->base + sspi->regs->usp_rx_frame_ctrl);
+		writel(readl(sspi->base + sspi->regs->usp_mode2) |
+			((usp_mode2 & SIRFSOC_USP_CLK_DIVISOR_MASK) <<
+			SIRFSOC_USP_CLK_DIVISOR_OFFSET) |
+			(SIRFSOC_USP_RXD_DELAY_LEN <<
+			 SIRFSOC_USP_RXD_DELAY_OFFSET) |
+			(SIRFSOC_USP_TXD_DELAY_LEN <<
+			 SIRFSOC_USP_TXD_DELAY_OFFSET),
+			sspi->base + sspi->regs->usp_mode2);
+	}
+	if (sspi->type == SIRF_REAL_SPI)
+		writel(regval, sspi->base + sspi->regs->spi_ctrl);
+	spi_sirfsoc_config_mode(spi);
+	if (sspi->type == SIRF_REAL_SPI) {
+		if (t && t->tx_buf && !t->rx_buf &&
+			(t->len <= SIRFSOC_MAX_CMD_BYTES)) {
+			sspi->tx_by_cmd = true;
+			writel(readl(sspi->base + sspi->regs->spi_ctrl) |
+				(SIRFSOC_SPI_CMD_BYTE_NUM((t->len - 1)) |
+				SIRFSOC_SPI_CMD_MODE),
+				sspi->base + sspi->regs->spi_ctrl);
+		} else {
+			sspi->tx_by_cmd = false;
+			writel(readl(sspi->base + sspi->regs->spi_ctrl) &
+				~SIRFSOC_SPI_CMD_MODE,
+				sspi->base + sspi->regs->spi_ctrl);
+		}
+	}
+	if (IS_DMA_VALID(t)) {
+		/* Enable DMA mode for RX, TX */
+		writel(0, sspi->base + sspi->regs->tx_dma_io_ctrl);
+		writel(SIRFSOC_SPI_RX_DMA_FLUSH,
+			sspi->base + sspi->regs->rx_dma_io_ctrl);
+	} else {
+		/* Enable IO mode for RX, TX */
+		writel(SIRFSOC_SPI_IO_MODE_SEL,
+			sspi->base + sspi->regs->tx_dma_io_ctrl);
+		writel(SIRFSOC_SPI_IO_MODE_SEL,
+			sspi->base + sspi->regs->rx_dma_io_ctrl);
+	}
+	return 0;
+}
+
+static int spi_sirfsoc_setup(struct spi_device *spi)
+{
+	struct sirfsoc_spi *sspi;
+	int ret = 0;
+
+	sspi = spi_master_get_devdata(spi->master);
+	if (spi->cs_gpio == -ENOENT)
+		sspi->hw_cs = true;
+	else {
+		sspi->hw_cs = false;
+		if (!spi_get_ctldata(spi)) {
+			void *cs = kmalloc(sizeof(int), GFP_KERNEL);
+			if (!cs) {
+				ret = -ENOMEM;
+				goto exit;
+			}
+			ret = gpio_is_valid(spi->cs_gpio);
+			if (!ret) {
+				dev_err(&spi->dev, "no valid gpio\n");
+				ret = -ENOENT;
+				goto exit;
+			}
+			ret = gpio_request(spi->cs_gpio, DRIVER_NAME);
+			if (ret) {
+				dev_err(&spi->dev, "failed to request gpio\n");
+				goto exit;
+			}
+			spi_set_ctldata(spi, cs);
+		}
+	}
+	spi_sirfsoc_config_mode(spi);
+	spi_sirfsoc_chipselect(spi, BITBANG_CS_INACTIVE);
+exit:
+	return ret;
+}
+
+static void spi_sirfsoc_cleanup(struct spi_device *spi)
+{
+	if (spi_get_ctldata(spi)) {
+		gpio_free(spi->cs_gpio);
+		kfree(spi_get_ctldata(spi));
+	}
+}
+
+static const struct sirf_spi_comp_data sirf_real_spi = {
+	.regs = &real_spi_register,
+	.type = SIRF_REAL_SPI,
+	.dat_max_frm_len = 64 * 1024,
+	.fifo_size = 256,
+};
+
+static const struct sirf_spi_comp_data sirf_usp_spi_p2 = {
+	.regs = &usp_spi_register,
+	.type = SIRF_USP_SPI_P2,
+	.dat_max_frm_len = 1024 * 1024,
+	.fifo_size = 128,
+	.hwinit = sirfsoc_usp_hwinit,
+};
+
+static const struct sirf_spi_comp_data sirf_usp_spi_a7 = {
+	.regs = &usp_spi_register,
+	.type = SIRF_USP_SPI_A7,
+	.dat_max_frm_len = 1024 * 1024,
+	.fifo_size = 512,
+	.hwinit = sirfsoc_usp_hwinit,
+};
+
+static const struct of_device_id spi_sirfsoc_of_match[] = {
+	{ .compatible = "sirf,prima2-spi", .data = &sirf_real_spi},
+	{ .compatible = "sirf,prima2-usp-spi", .data = &sirf_usp_spi_p2},
+	{ .compatible = "sirf,atlas7-usp-spi", .data = &sirf_usp_spi_a7},
+	{}
+};
+MODULE_DEVICE_TABLE(of, spi_sirfsoc_of_match);
+
+static int spi_sirfsoc_probe(struct platform_device *pdev)
+{
+	struct sirfsoc_spi *sspi;
+	struct spi_master *master;
+	struct resource *mem_res;
+	const struct sirf_spi_comp_data *spi_comp_data;
+	int irq;
+	int ret;
+	const struct of_device_id *match;
+
+	ret = device_reset(&pdev->dev);
+	if (ret) {
+		dev_err(&pdev->dev, "SPI reset failed!\n");
+		return ret;
+	}
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*sspi));
+	if (!master) {
+		dev_err(&pdev->dev, "Unable to allocate SPI master\n");
+		return -ENOMEM;
+	}
+	match = of_match_node(spi_sirfsoc_of_match, pdev->dev.of_node);
+	platform_set_drvdata(pdev, master);
+	sspi = spi_master_get_devdata(master);
+	sspi->fifo_full_offset = ilog2(sspi->fifo_size);
+	spi_comp_data = match->data;
+	sspi->regs = spi_comp_data->regs;
+	sspi->type = spi_comp_data->type;
+	sspi->fifo_level_chk_mask = (sspi->fifo_size / 4) - 1;
+	sspi->dat_max_frm_len = spi_comp_data->dat_max_frm_len;
+	sspi->fifo_size = spi_comp_data->fifo_size;
+	mem_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	sspi->base = devm_ioremap_resource(&pdev->dev, mem_res);
+	if (IS_ERR(sspi->base)) {
+		ret = PTR_ERR(sspi->base);
+		goto free_master;
+	}
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		ret = -ENXIO;
+		goto free_master;
+	}
+	ret = devm_request_irq(&pdev->dev, irq, spi_sirfsoc_irq, 0,
+				DRIVER_NAME, sspi);
+	if (ret)
+		goto free_master;
+
+	sspi->bitbang.master = master;
+	sspi->bitbang.chipselect = spi_sirfsoc_chipselect;
+	sspi->bitbang.setup_transfer = spi_sirfsoc_setup_transfer;
+	sspi->bitbang.txrx_bufs = spi_sirfsoc_transfer;
+	sspi->bitbang.master->setup = spi_sirfsoc_setup;
+	sspi->bitbang.master->cleanup = spi_sirfsoc_cleanup;
+	master->bus_num = pdev->id;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_LSB_FIRST | SPI_CS_HIGH;
+	master->bits_per_word_mask = SPI_BPW_MASK(8) | SPI_BPW_MASK(12) |
+					SPI_BPW_MASK(16) | SPI_BPW_MASK(32);
+	master->max_speed_hz = SIRFSOC_SPI_DEFAULT_FRQ;
+	master->flags = SPI_MASTER_MUST_RX | SPI_MASTER_MUST_TX;
+	sspi->bitbang.master->dev.of_node = pdev->dev.of_node;
+
+	/* request DMA channels */
+	sspi->rx_chan = dma_request_slave_channel(&pdev->dev, "rx");
+	if (!sspi->rx_chan) {
+		dev_err(&pdev->dev, "can not allocate rx dma channel\n");
+		ret = -ENODEV;
+		goto free_master;
+	}
+	sspi->tx_chan = dma_request_slave_channel(&pdev->dev, "tx");
+	if (!sspi->tx_chan) {
+		dev_err(&pdev->dev, "can not allocate tx dma channel\n");
+		ret = -ENODEV;
+		goto free_rx_dma;
+	}
+
+	sspi->clk = clk_get(&pdev->dev, NULL);
+	if (IS_ERR(sspi->clk)) {
+		ret = PTR_ERR(sspi->clk);
+		goto free_tx_dma;
+	}
+	clk_prepare_enable(sspi->clk);
+	if (spi_comp_data->hwinit)
+		spi_comp_data->hwinit(sspi);
+	sspi->ctrl_freq = clk_get_rate(sspi->clk);
+
+	init_completion(&sspi->rx_done);
+	init_completion(&sspi->tx_done);
+
+	ret = spi_bitbang_start(&sspi->bitbang);
+	if (ret)
+		goto free_clk;
+	dev_info(&pdev->dev, "registered, bus number = %d\n", master->bus_num);
+
+	return 0;
+free_clk:
+	clk_disable_unprepare(sspi->clk);
+	clk_put(sspi->clk);
+free_tx_dma:
+	dma_release_channel(sspi->tx_chan);
+free_rx_dma:
+	dma_release_channel(sspi->rx_chan);
+free_master:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int  spi_sirfsoc_remove(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct sirfsoc_spi *sspi;
+
+	master = platform_get_drvdata(pdev);
+	sspi = spi_master_get_devdata(master);
+	spi_bitbang_stop(&sspi->bitbang);
+	clk_disable_unprepare(sspi->clk);
+	clk_put(sspi->clk);
+	dma_release_channel(sspi->rx_chan);
+	dma_release_channel(sspi->tx_chan);
+	spi_master_put(master);
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int spi_sirfsoc_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct sirfsoc_spi *sspi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = spi_master_suspend(master);
+	if (ret)
+		return ret;
+
+	clk_disable(sspi->clk);
+	return 0;
+}
+
+static int spi_sirfsoc_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct sirfsoc_spi *sspi = spi_master_get_devdata(master);
+
+	clk_enable(sspi->clk);
+	writel(SIRFSOC_SPI_FIFO_RESET, sspi->base + sspi->regs->txfifo_op);
+	writel(SIRFSOC_SPI_FIFO_RESET, sspi->base + sspi->regs->rxfifo_op);
+	writel(SIRFSOC_SPI_FIFO_START, sspi->base + sspi->regs->txfifo_op);
+	writel(SIRFSOC_SPI_FIFO_START, sspi->base + sspi->regs->rxfifo_op);
+	return 0;
+}
+#endif
+
+static SIMPLE_DEV_PM_OPS(spi_sirfsoc_pm_ops, spi_sirfsoc_suspend,
+			 spi_sirfsoc_resume);
+
+static struct platform_driver spi_sirfsoc_driver = {
+	.driver = {
+		.name = DRIVER_NAME,
+		.pm     = &spi_sirfsoc_pm_ops,
+		.of_match_table = spi_sirfsoc_of_match,
+	},
+	.probe = spi_sirfsoc_probe,
+	.remove = spi_sirfsoc_remove,
+};
+module_platform_driver(spi_sirfsoc_driver);
+MODULE_DESCRIPTION("SiRF SoC SPI master driver");
+MODULE_AUTHOR("Zhiwu Song <Zhiwu.Song@csr.com>");
+MODULE_AUTHOR("Barry Song <Baohua.Song@csr.com>");
+MODULE_AUTHOR("Qipan Li <Qipan.Li@csr.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-slave-system-control.c b/drivers/spi/spi-slave-system-control.c
new file mode 100644
index 0000000..c0257e9
--- /dev/null
+++ b/drivers/spi/spi-slave-system-control.c
@@ -0,0 +1,154 @@
+/*
+ * SPI slave handler controlling system state
+ *
+ * This SPI slave handler allows remote control of system reboot, power off,
+ * halt, and suspend.
+ *
+ * Copyright (C) 2016-2017 Glider bvba
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Usage (assuming /dev/spidev2.0 corresponds to the SPI master on the remote
+ * system):
+ *
+ *   # reboot='\x7c\x50'
+ *   # poweroff='\x71\x3f'
+ *   # halt='\x38\x76'
+ *   # suspend='\x1b\x1b'
+ *   # spidev_test -D /dev/spidev2.0 -p $suspend # or $reboot, $poweroff, $halt
+ */
+
+#include <linux/completion.h>
+#include <linux/module.h>
+#include <linux/reboot.h>
+#include <linux/suspend.h>
+#include <linux/spi/spi.h>
+
+/*
+ * The numbers are chosen to display something human-readable on two 7-segment
+ * displays connected to two 74HC595 shift registers
+ */
+#define CMD_REBOOT	0x7c50	/* rb */
+#define CMD_POWEROFF	0x713f	/* OF */
+#define CMD_HALT	0x3876	/* HL */
+#define CMD_SUSPEND	0x1b1b	/* ZZ */
+
+struct spi_slave_system_control_priv {
+	struct spi_device *spi;
+	struct completion finished;
+	struct spi_transfer xfer;
+	struct spi_message msg;
+	__be16 cmd;
+};
+
+static
+int spi_slave_system_control_submit(struct spi_slave_system_control_priv *priv);
+
+static void spi_slave_system_control_complete(void *arg)
+{
+	struct spi_slave_system_control_priv *priv = arg;
+	u16 cmd;
+	int ret;
+
+	if (priv->msg.status)
+		goto terminate;
+
+	cmd = be16_to_cpu(priv->cmd);
+	switch (cmd) {
+	case CMD_REBOOT:
+		dev_info(&priv->spi->dev, "Rebooting system...\n");
+		kernel_restart(NULL);
+
+	case CMD_POWEROFF:
+		dev_info(&priv->spi->dev, "Powering off system...\n");
+		kernel_power_off();
+		break;
+
+	case CMD_HALT:
+		dev_info(&priv->spi->dev, "Halting system...\n");
+		kernel_halt();
+		break;
+
+	case CMD_SUSPEND:
+		dev_info(&priv->spi->dev, "Suspending system...\n");
+		pm_suspend(PM_SUSPEND_MEM);
+		break;
+
+	default:
+		dev_warn(&priv->spi->dev, "Unknown command 0x%x\n", cmd);
+		break;
+	}
+
+	ret = spi_slave_system_control_submit(priv);
+	if (ret)
+		goto terminate;
+
+	return;
+
+terminate:
+	dev_info(&priv->spi->dev, "Terminating\n");
+	complete(&priv->finished);
+}
+
+static
+int spi_slave_system_control_submit(struct spi_slave_system_control_priv *priv)
+{
+	int ret;
+
+	spi_message_init_with_transfers(&priv->msg, &priv->xfer, 1);
+
+	priv->msg.complete = spi_slave_system_control_complete;
+	priv->msg.context = priv;
+
+	ret = spi_async(priv->spi, &priv->msg);
+	if (ret)
+		dev_err(&priv->spi->dev, "spi_async() failed %d\n", ret);
+
+	return ret;
+}
+
+static int spi_slave_system_control_probe(struct spi_device *spi)
+{
+	struct spi_slave_system_control_priv *priv;
+	int ret;
+
+	priv = devm_kzalloc(&spi->dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	priv->spi = spi;
+	init_completion(&priv->finished);
+	priv->xfer.rx_buf = &priv->cmd;
+	priv->xfer.len = sizeof(priv->cmd);
+
+	ret = spi_slave_system_control_submit(priv);
+	if (ret)
+		return ret;
+
+	spi_set_drvdata(spi, priv);
+	return 0;
+}
+
+static int spi_slave_system_control_remove(struct spi_device *spi)
+{
+	struct spi_slave_system_control_priv *priv = spi_get_drvdata(spi);
+
+	spi_slave_abort(spi);
+	wait_for_completion(&priv->finished);
+	return 0;
+}
+
+static struct spi_driver spi_slave_system_control_driver = {
+	.driver = {
+		.name	= "spi-slave-system-control",
+	},
+	.probe		= spi_slave_system_control_probe,
+	.remove		= spi_slave_system_control_remove,
+};
+module_spi_driver(spi_slave_system_control_driver);
+
+MODULE_AUTHOR("Geert Uytterhoeven <geert+renesas@glider.be>");
+MODULE_DESCRIPTION("SPI slave handler controlling system state");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-slave-time.c b/drivers/spi/spi-slave-time.c
new file mode 100644
index 0000000..f2e07a3
--- /dev/null
+++ b/drivers/spi/spi-slave-time.c
@@ -0,0 +1,129 @@
+/*
+ * SPI slave handler reporting uptime at reception of previous SPI message
+ *
+ * This SPI slave handler sends the time of reception of the last SPI message
+ * as two 32-bit unsigned integers in binary format and in network byte order,
+ * representing the number of seconds and fractional seconds (in microseconds)
+ * since boot up.
+ *
+ * Copyright (C) 2016-2017 Glider bvba
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Usage (assuming /dev/spidev2.0 corresponds to the SPI master on the remote
+ * system):
+ *
+ *   # spidev_test -D /dev/spidev2.0 -p dummy-8B
+ *   spi mode: 0x0
+ *   bits per word: 8
+ *   max speed: 500000 Hz (500 KHz)
+ *   RX | 00 00 04 6D 00 09 5B BB ...
+ *		^^^^^    ^^^^^^^^
+ *		seconds  microseconds
+ */
+
+#include <linux/completion.h>
+#include <linux/module.h>
+#include <linux/sched/clock.h>
+#include <linux/spi/spi.h>
+
+
+struct spi_slave_time_priv {
+	struct spi_device *spi;
+	struct completion finished;
+	struct spi_transfer xfer;
+	struct spi_message msg;
+	__be32 buf[2];
+};
+
+static int spi_slave_time_submit(struct spi_slave_time_priv *priv);
+
+static void spi_slave_time_complete(void *arg)
+{
+	struct spi_slave_time_priv *priv = arg;
+	int ret;
+
+	ret = priv->msg.status;
+	if (ret)
+		goto terminate;
+
+	ret = spi_slave_time_submit(priv);
+	if (ret)
+		goto terminate;
+
+	return;
+
+terminate:
+	dev_info(&priv->spi->dev, "Terminating\n");
+	complete(&priv->finished);
+}
+
+static int spi_slave_time_submit(struct spi_slave_time_priv *priv)
+{
+	u32 rem_us;
+	int ret;
+	u64 ts;
+
+	ts = local_clock();
+	rem_us = do_div(ts, 1000000000) / 1000;
+
+	priv->buf[0] = cpu_to_be32(ts);
+	priv->buf[1] = cpu_to_be32(rem_us);
+
+	spi_message_init_with_transfers(&priv->msg, &priv->xfer, 1);
+
+	priv->msg.complete = spi_slave_time_complete;
+	priv->msg.context = priv;
+
+	ret = spi_async(priv->spi, &priv->msg);
+	if (ret)
+		dev_err(&priv->spi->dev, "spi_async() failed %d\n", ret);
+
+	return ret;
+}
+
+static int spi_slave_time_probe(struct spi_device *spi)
+{
+	struct spi_slave_time_priv *priv;
+	int ret;
+
+	priv = devm_kzalloc(&spi->dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	priv->spi = spi;
+	init_completion(&priv->finished);
+	priv->xfer.tx_buf = priv->buf;
+	priv->xfer.len = sizeof(priv->buf);
+
+	ret = spi_slave_time_submit(priv);
+	if (ret)
+		return ret;
+
+	spi_set_drvdata(spi, priv);
+	return 0;
+}
+
+static int spi_slave_time_remove(struct spi_device *spi)
+{
+	struct spi_slave_time_priv *priv = spi_get_drvdata(spi);
+
+	spi_slave_abort(spi);
+	wait_for_completion(&priv->finished);
+	return 0;
+}
+
+static struct spi_driver spi_slave_time_driver = {
+	.driver = {
+		.name	= "spi-slave-time",
+	},
+	.probe		= spi_slave_time_probe,
+	.remove		= spi_slave_time_remove,
+};
+module_spi_driver(spi_slave_time_driver);
+
+MODULE_AUTHOR("Geert Uytterhoeven <geert+renesas@glider.be>");
+MODULE_DESCRIPTION("SPI slave reporting uptime at previous SPI message");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-sprd-adi.c b/drivers/spi/spi-sprd-adi.c
new file mode 100644
index 0000000..df5960b
--- /dev/null
+++ b/drivers/spi/spi-sprd-adi.c
@@ -0,0 +1,533 @@
+/*
+ * Copyright (C) 2017 Spreadtrum Communications Inc.
+ *
+ * SPDX-License-Identifier: GPL-2.0
+ */
+
+#include <linux/delay.h>
+#include <linux/hwspinlock.h>
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/reboot.h>
+#include <linux/spi/spi.h>
+#include <linux/sizes.h>
+
+/* Registers definitions for ADI controller */
+#define REG_ADI_CTRL0			0x4
+#define REG_ADI_CHN_PRIL		0x8
+#define REG_ADI_CHN_PRIH		0xc
+#define REG_ADI_INT_EN			0x10
+#define REG_ADI_INT_RAW			0x14
+#define REG_ADI_INT_MASK		0x18
+#define REG_ADI_INT_CLR			0x1c
+#define REG_ADI_GSSI_CFG0		0x20
+#define REG_ADI_GSSI_CFG1		0x24
+#define REG_ADI_RD_CMD			0x28
+#define REG_ADI_RD_DATA			0x2c
+#define REG_ADI_ARM_FIFO_STS		0x30
+#define REG_ADI_STS			0x34
+#define REG_ADI_EVT_FIFO_STS		0x38
+#define REG_ADI_ARM_CMD_STS		0x3c
+#define REG_ADI_CHN_EN			0x40
+#define REG_ADI_CHN_ADDR(id)		(0x44 + (id - 2) * 4)
+#define REG_ADI_CHN_EN1			0x20c
+
+/* Bits definitions for register REG_ADI_GSSI_CFG0 */
+#define BIT_CLK_ALL_ON			BIT(30)
+
+/* Bits definitions for register REG_ADI_RD_DATA */
+#define BIT_RD_CMD_BUSY			BIT(31)
+#define RD_ADDR_SHIFT			16
+#define RD_VALUE_MASK			GENMASK(15, 0)
+#define RD_ADDR_MASK			GENMASK(30, 16)
+
+/* Bits definitions for register REG_ADI_ARM_FIFO_STS */
+#define BIT_FIFO_FULL			BIT(11)
+#define BIT_FIFO_EMPTY			BIT(10)
+
+/*
+ * ADI slave devices include RTC, ADC, regulator, charger, thermal and so on.
+ * The slave devices address offset is always 0x8000 and size is 4K.
+ */
+#define ADI_SLAVE_ADDR_SIZE		SZ_4K
+#define ADI_SLAVE_OFFSET		0x8000
+
+/* Timeout (ms) for the trylock of hardware spinlocks */
+#define ADI_HWSPINLOCK_TIMEOUT		5000
+/*
+ * ADI controller has 50 channels including 2 software channels
+ * and 48 hardware channels.
+ */
+#define ADI_HW_CHNS			50
+
+#define ADI_FIFO_DRAIN_TIMEOUT		1000
+#define ADI_READ_TIMEOUT		2000
+#define REG_ADDR_LOW_MASK		GENMASK(11, 0)
+
+/* Registers definitions for PMIC watchdog controller */
+#define REG_WDG_LOAD_LOW		0x80
+#define REG_WDG_LOAD_HIGH		0x84
+#define REG_WDG_CTRL			0x88
+#define REG_WDG_LOCK			0xa0
+
+/* Bits definitions for register REG_WDG_CTRL */
+#define BIT_WDG_RUN			BIT(1)
+#define BIT_WDG_RST			BIT(3)
+
+/* Registers definitions for PMIC */
+#define PMIC_RST_STATUS			0xee8
+#define PMIC_MODULE_EN			0xc08
+#define PMIC_CLK_EN			0xc18
+#define BIT_WDG_EN			BIT(2)
+
+/* Definition of PMIC reset status register */
+#define HWRST_STATUS_RECOVERY		0x20
+#define HWRST_STATUS_NORMAL		0x40
+#define HWRST_STATUS_ALARM		0x50
+#define HWRST_STATUS_SLEEP		0x60
+#define HWRST_STATUS_FASTBOOT		0x30
+#define HWRST_STATUS_SPECIAL		0x70
+#define HWRST_STATUS_PANIC		0x80
+#define HWRST_STATUS_CFTREBOOT		0x90
+#define HWRST_STATUS_AUTODLOADER	0xa0
+#define HWRST_STATUS_IQMODE		0xb0
+#define HWRST_STATUS_SPRDISK		0xc0
+
+/* Use default timeout 50 ms that converts to watchdog values */
+#define WDG_LOAD_VAL			((50 * 1000) / 32768)
+#define WDG_LOAD_MASK			GENMASK(15, 0)
+#define WDG_UNLOCK_KEY			0xe551
+
+struct sprd_adi {
+	struct spi_controller	*ctlr;
+	struct device		*dev;
+	void __iomem		*base;
+	struct hwspinlock	*hwlock;
+	unsigned long		slave_vbase;
+	unsigned long		slave_pbase;
+	struct notifier_block	restart_handler;
+};
+
+static int sprd_adi_check_paddr(struct sprd_adi *sadi, u32 paddr)
+{
+	if (paddr < sadi->slave_pbase || paddr >
+	    (sadi->slave_pbase + ADI_SLAVE_ADDR_SIZE)) {
+		dev_err(sadi->dev,
+			"slave physical address is incorrect, addr = 0x%x\n",
+			paddr);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static unsigned long sprd_adi_to_vaddr(struct sprd_adi *sadi, u32 paddr)
+{
+	return (paddr - sadi->slave_pbase + sadi->slave_vbase);
+}
+
+static int sprd_adi_drain_fifo(struct sprd_adi *sadi)
+{
+	u32 timeout = ADI_FIFO_DRAIN_TIMEOUT;
+	u32 sts;
+
+	do {
+		sts = readl_relaxed(sadi->base + REG_ADI_ARM_FIFO_STS);
+		if (sts & BIT_FIFO_EMPTY)
+			break;
+
+		cpu_relax();
+	} while (--timeout);
+
+	if (timeout == 0) {
+		dev_err(sadi->dev, "drain write fifo timeout\n");
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+static int sprd_adi_fifo_is_full(struct sprd_adi *sadi)
+{
+	return readl_relaxed(sadi->base + REG_ADI_ARM_FIFO_STS) & BIT_FIFO_FULL;
+}
+
+static int sprd_adi_read(struct sprd_adi *sadi, u32 reg_paddr, u32 *read_val)
+{
+	int read_timeout = ADI_READ_TIMEOUT;
+	unsigned long flags;
+	u32 val, rd_addr;
+	int ret;
+
+	ret = hwspin_lock_timeout_irqsave(sadi->hwlock,
+					  ADI_HWSPINLOCK_TIMEOUT,
+					  &flags);
+	if (ret) {
+		dev_err(sadi->dev, "get the hw lock failed\n");
+		return ret;
+	}
+
+	/*
+	 * Set the physical register address need to read into RD_CMD register,
+	 * then ADI controller will start to transfer automatically.
+	 */
+	writel_relaxed(reg_paddr, sadi->base + REG_ADI_RD_CMD);
+
+	/*
+	 * Wait read operation complete, the BIT_RD_CMD_BUSY will be set
+	 * simultaneously when writing read command to register, and the
+	 * BIT_RD_CMD_BUSY will be cleared after the read operation is
+	 * completed.
+	 */
+	do {
+		val = readl_relaxed(sadi->base + REG_ADI_RD_DATA);
+		if (!(val & BIT_RD_CMD_BUSY))
+			break;
+
+		cpu_relax();
+	} while (--read_timeout);
+
+	if (read_timeout == 0) {
+		dev_err(sadi->dev, "ADI read timeout\n");
+		ret = -EBUSY;
+		goto out;
+	}
+
+	/*
+	 * The return value includes data and read register address, from bit 0
+	 * to bit 15 are data, and from bit 16 to bit 30 are read register
+	 * address. Then we can check the returned register address to validate
+	 * data.
+	 */
+	rd_addr = (val & RD_ADDR_MASK ) >> RD_ADDR_SHIFT;
+
+	if (rd_addr != (reg_paddr & REG_ADDR_LOW_MASK)) {
+		dev_err(sadi->dev, "read error, reg addr = 0x%x, val = 0x%x\n",
+			reg_paddr, val);
+		ret = -EIO;
+		goto out;
+	}
+
+	*read_val = val & RD_VALUE_MASK;
+
+out:
+	hwspin_unlock_irqrestore(sadi->hwlock, &flags);
+	return ret;
+}
+
+static int sprd_adi_write(struct sprd_adi *sadi, u32 reg_paddr, u32 val)
+{
+	unsigned long reg = sprd_adi_to_vaddr(sadi, reg_paddr);
+	u32 timeout = ADI_FIFO_DRAIN_TIMEOUT;
+	unsigned long flags;
+	int ret;
+
+	ret = hwspin_lock_timeout_irqsave(sadi->hwlock,
+					  ADI_HWSPINLOCK_TIMEOUT,
+					  &flags);
+	if (ret) {
+		dev_err(sadi->dev, "get the hw lock failed\n");
+		return ret;
+	}
+
+	ret = sprd_adi_drain_fifo(sadi);
+	if (ret < 0)
+		goto out;
+
+	/*
+	 * we should wait for write fifo is empty before writing data to PMIC
+	 * registers.
+	 */
+	do {
+		if (!sprd_adi_fifo_is_full(sadi)) {
+			writel_relaxed(val, (void __iomem *)reg);
+			break;
+		}
+
+		cpu_relax();
+	} while (--timeout);
+
+	if (timeout == 0) {
+		dev_err(sadi->dev, "write fifo is full\n");
+		ret = -EBUSY;
+	}
+
+out:
+	hwspin_unlock_irqrestore(sadi->hwlock, &flags);
+	return ret;
+}
+
+static int sprd_adi_transfer_one(struct spi_controller *ctlr,
+				 struct spi_device *spi_dev,
+				 struct spi_transfer *t)
+{
+	struct sprd_adi *sadi = spi_controller_get_devdata(ctlr);
+	u32 phy_reg, val;
+	int ret;
+
+	if (t->rx_buf) {
+		phy_reg = *(u32 *)t->rx_buf + sadi->slave_pbase;
+
+		ret = sprd_adi_check_paddr(sadi, phy_reg);
+		if (ret)
+			return ret;
+
+		ret = sprd_adi_read(sadi, phy_reg, &val);
+		if (ret)
+			return ret;
+
+		*(u32 *)t->rx_buf = val;
+	} else if (t->tx_buf) {
+		u32 *p = (u32 *)t->tx_buf;
+
+		/*
+		 * Get the physical register address need to write and convert
+		 * the physical address to virtual address. Since we need
+		 * virtual register address to write.
+		 */
+		phy_reg = *p++ + sadi->slave_pbase;
+		ret = sprd_adi_check_paddr(sadi, phy_reg);
+		if (ret)
+			return ret;
+
+		val = *p;
+		ret = sprd_adi_write(sadi, phy_reg, val);
+		if (ret)
+			return ret;
+	} else {
+		dev_err(sadi->dev, "no buffer for transfer\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int sprd_adi_restart_handler(struct notifier_block *this,
+				    unsigned long mode, void *cmd)
+{
+	struct sprd_adi *sadi = container_of(this, struct sprd_adi,
+					     restart_handler);
+	u32 val, reboot_mode = 0;
+
+	if (!cmd)
+		reboot_mode = HWRST_STATUS_NORMAL;
+	else if (!strncmp(cmd, "recovery", 8))
+		reboot_mode = HWRST_STATUS_RECOVERY;
+	else if (!strncmp(cmd, "alarm", 5))
+		reboot_mode = HWRST_STATUS_ALARM;
+	else if (!strncmp(cmd, "fastsleep", 9))
+		reboot_mode = HWRST_STATUS_SLEEP;
+	else if (!strncmp(cmd, "bootloader", 10))
+		reboot_mode = HWRST_STATUS_FASTBOOT;
+	else if (!strncmp(cmd, "panic", 5))
+		reboot_mode = HWRST_STATUS_PANIC;
+	else if (!strncmp(cmd, "special", 7))
+		reboot_mode = HWRST_STATUS_SPECIAL;
+	else if (!strncmp(cmd, "cftreboot", 9))
+		reboot_mode = HWRST_STATUS_CFTREBOOT;
+	else if (!strncmp(cmd, "autodloader", 11))
+		reboot_mode = HWRST_STATUS_AUTODLOADER;
+	else if (!strncmp(cmd, "iqmode", 6))
+		reboot_mode = HWRST_STATUS_IQMODE;
+	else if (!strncmp(cmd, "sprdisk", 7))
+		reboot_mode = HWRST_STATUS_SPRDISK;
+	else
+		reboot_mode = HWRST_STATUS_NORMAL;
+
+	/* Record the reboot mode */
+	sprd_adi_read(sadi, sadi->slave_pbase + PMIC_RST_STATUS, &val);
+	val |= reboot_mode;
+	sprd_adi_write(sadi, sadi->slave_pbase + PMIC_RST_STATUS, val);
+
+	/* Enable the interface clock of the watchdog */
+	sprd_adi_read(sadi, sadi->slave_pbase + PMIC_MODULE_EN, &val);
+	val |= BIT_WDG_EN;
+	sprd_adi_write(sadi, sadi->slave_pbase + PMIC_MODULE_EN, val);
+
+	/* Enable the work clock of the watchdog */
+	sprd_adi_read(sadi, sadi->slave_pbase + PMIC_CLK_EN, &val);
+	val |= BIT_WDG_EN;
+	sprd_adi_write(sadi, sadi->slave_pbase + PMIC_CLK_EN, val);
+
+	/* Unlock the watchdog */
+	sprd_adi_write(sadi, sadi->slave_pbase + REG_WDG_LOCK, WDG_UNLOCK_KEY);
+
+	/* Load the watchdog timeout value, 50ms is always enough. */
+	sprd_adi_write(sadi, sadi->slave_pbase + REG_WDG_LOAD_LOW,
+		       WDG_LOAD_VAL & WDG_LOAD_MASK);
+	sprd_adi_write(sadi, sadi->slave_pbase + REG_WDG_LOAD_HIGH, 0);
+
+	/* Start the watchdog to reset system */
+	sprd_adi_read(sadi, sadi->slave_pbase + REG_WDG_CTRL, &val);
+	val |= BIT_WDG_RUN | BIT_WDG_RST;
+	sprd_adi_write(sadi, sadi->slave_pbase + REG_WDG_CTRL, val);
+
+	mdelay(1000);
+
+	dev_emerg(sadi->dev, "Unable to restart system\n");
+	return NOTIFY_DONE;
+}
+
+static void sprd_adi_hw_init(struct sprd_adi *sadi)
+{
+	struct device_node *np = sadi->dev->of_node;
+	int i, size, chn_cnt;
+	const __be32 *list;
+	u32 tmp;
+
+	/* Address bits select default 12 bits */
+	writel_relaxed(0, sadi->base + REG_ADI_CTRL0);
+
+	/* Set all channels as default priority */
+	writel_relaxed(0, sadi->base + REG_ADI_CHN_PRIL);
+	writel_relaxed(0, sadi->base + REG_ADI_CHN_PRIH);
+
+	/* Set clock auto gate mode */
+	tmp = readl_relaxed(sadi->base + REG_ADI_GSSI_CFG0);
+	tmp &= ~BIT_CLK_ALL_ON;
+	writel_relaxed(tmp, sadi->base + REG_ADI_GSSI_CFG0);
+
+	/* Set hardware channels setting */
+	list = of_get_property(np, "sprd,hw-channels", &size);
+	if (!list || !size) {
+		dev_info(sadi->dev, "no hw channels setting in node\n");
+		return;
+	}
+
+	chn_cnt = size / 8;
+	for (i = 0; i < chn_cnt; i++) {
+		u32 value;
+		u32 chn_id = be32_to_cpu(*list++);
+		u32 chn_config = be32_to_cpu(*list++);
+
+		/* Channel 0 and 1 are software channels */
+		if (chn_id < 2)
+			continue;
+
+		writel_relaxed(chn_config, sadi->base +
+			       REG_ADI_CHN_ADDR(chn_id));
+
+		if (chn_id < 32) {
+			value = readl_relaxed(sadi->base + REG_ADI_CHN_EN);
+			value |= BIT(chn_id);
+			writel_relaxed(value, sadi->base + REG_ADI_CHN_EN);
+		} else if (chn_id < ADI_HW_CHNS) {
+			value = readl_relaxed(sadi->base + REG_ADI_CHN_EN1);
+			value |= BIT(chn_id - 32);
+			writel_relaxed(value, sadi->base + REG_ADI_CHN_EN1);
+		}
+	}
+}
+
+static int sprd_adi_probe(struct platform_device *pdev)
+{
+	struct device_node *np = pdev->dev.of_node;
+	struct spi_controller *ctlr;
+	struct sprd_adi *sadi;
+	struct resource *res;
+	u32 num_chipselect;
+	int ret;
+
+	if (!np) {
+		dev_err(&pdev->dev, "can not find the adi bus node\n");
+		return -ENODEV;
+	}
+
+	pdev->id = of_alias_get_id(np, "spi");
+	num_chipselect = of_get_child_count(np);
+
+	ctlr = spi_alloc_master(&pdev->dev, sizeof(struct sprd_adi));
+	if (!ctlr)
+		return -ENOMEM;
+
+	dev_set_drvdata(&pdev->dev, ctlr);
+	sadi = spi_controller_get_devdata(ctlr);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	sadi->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(sadi->base)) {
+		ret = PTR_ERR(sadi->base);
+		goto put_ctlr;
+	}
+
+	sadi->slave_vbase = (unsigned long)sadi->base + ADI_SLAVE_OFFSET;
+	sadi->slave_pbase = res->start + ADI_SLAVE_OFFSET;
+	sadi->ctlr = ctlr;
+	sadi->dev = &pdev->dev;
+	ret = of_hwspin_lock_get_id_byname(np, "adi");
+	if (ret < 0) {
+		dev_err(&pdev->dev, "can not get the hardware spinlock\n");
+		goto put_ctlr;
+	}
+
+	sadi->hwlock = devm_hwspin_lock_request_specific(&pdev->dev, ret);
+	if (!sadi->hwlock) {
+		ret = -ENXIO;
+		goto put_ctlr;
+	}
+
+	sprd_adi_hw_init(sadi);
+
+	ctlr->dev.of_node = pdev->dev.of_node;
+	ctlr->bus_num = pdev->id;
+	ctlr->num_chipselect = num_chipselect;
+	ctlr->flags = SPI_MASTER_HALF_DUPLEX;
+	ctlr->bits_per_word_mask = 0;
+	ctlr->transfer_one = sprd_adi_transfer_one;
+
+	ret = devm_spi_register_controller(&pdev->dev, ctlr);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to register SPI controller\n");
+		goto put_ctlr;
+	}
+
+	sadi->restart_handler.notifier_call = sprd_adi_restart_handler;
+	sadi->restart_handler.priority = 128;
+	ret = register_restart_handler(&sadi->restart_handler);
+	if (ret) {
+		dev_err(&pdev->dev, "can not register restart handler\n");
+		goto put_ctlr;
+	}
+
+	return 0;
+
+put_ctlr:
+	spi_controller_put(ctlr);
+	return ret;
+}
+
+static int sprd_adi_remove(struct platform_device *pdev)
+{
+	struct spi_controller *ctlr = dev_get_drvdata(&pdev->dev);
+	struct sprd_adi *sadi = spi_controller_get_devdata(ctlr);
+
+	unregister_restart_handler(&sadi->restart_handler);
+	return 0;
+}
+
+static const struct of_device_id sprd_adi_of_match[] = {
+	{
+		.compatible = "sprd,sc9860-adi",
+	},
+	{ },
+};
+MODULE_DEVICE_TABLE(of, sprd_adi_of_match);
+
+static struct platform_driver sprd_adi_driver = {
+	.driver = {
+		.name = "sprd-adi",
+		.of_match_table = sprd_adi_of_match,
+	},
+	.probe = sprd_adi_probe,
+	.remove = sprd_adi_remove,
+};
+module_platform_driver(sprd_adi_driver);
+
+MODULE_DESCRIPTION("Spreadtrum ADI Controller Driver");
+MODULE_AUTHOR("Baolin Wang <Baolin.Wang@spreadtrum.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-st-ssc4.c b/drivers/spi/spi-st-ssc4.c
new file mode 100644
index 0000000..a4e43fc
--- /dev/null
+++ b/drivers/spi/spi-st-ssc4.c
@@ -0,0 +1,483 @@
+/*
+ *  Copyright (c) 2008-2014 STMicroelectronics Limited
+ *
+ *  Author: Angus Clark <Angus.Clark@st.com>
+ *          Patrice Chotard <patrice.chotard@st.com>
+ *          Lee Jones <lee.jones@linaro.org>
+ *
+ *  SPI master mode controller driver, used in STMicroelectronics devices.
+ *
+ *  May be copied or modified under the terms of the GNU General Public
+ *  License Version 2.0 only.  See linux/COPYING for more information.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/platform_device.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+#include <linux/of_irq.h>
+#include <linux/pm_runtime.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+
+/* SSC registers */
+#define SSC_BRG				0x000
+#define SSC_TBUF			0x004
+#define SSC_RBUF			0x008
+#define SSC_CTL				0x00C
+#define SSC_IEN				0x010
+#define SSC_I2C				0x018
+
+/* SSC Control */
+#define SSC_CTL_DATA_WIDTH_9		0x8
+#define SSC_CTL_DATA_WIDTH_MSK		0xf
+#define SSC_CTL_BM			0xf
+#define SSC_CTL_HB			BIT(4)
+#define SSC_CTL_PH			BIT(5)
+#define SSC_CTL_PO			BIT(6)
+#define SSC_CTL_SR			BIT(7)
+#define SSC_CTL_MS			BIT(8)
+#define SSC_CTL_EN			BIT(9)
+#define SSC_CTL_LPB			BIT(10)
+#define SSC_CTL_EN_TX_FIFO		BIT(11)
+#define SSC_CTL_EN_RX_FIFO		BIT(12)
+#define SSC_CTL_EN_CLST_RX		BIT(13)
+
+/* SSC Interrupt Enable */
+#define SSC_IEN_TEEN			BIT(2)
+
+#define FIFO_SIZE			8
+
+struct spi_st {
+	/* SSC SPI Controller */
+	void __iomem		*base;
+	struct clk		*clk;
+	struct device		*dev;
+
+	/* SSC SPI current transaction */
+	const u8		*tx_ptr;
+	u8			*rx_ptr;
+	u16			bytes_per_word;
+	unsigned int		words_remaining;
+	unsigned int		baud;
+	struct completion	done;
+};
+
+/* Load the TX FIFO */
+static void ssc_write_tx_fifo(struct spi_st *spi_st)
+{
+	unsigned int count, i;
+	uint32_t word = 0;
+
+	if (spi_st->words_remaining > FIFO_SIZE)
+		count = FIFO_SIZE;
+	else
+		count = spi_st->words_remaining;
+
+	for (i = 0; i < count; i++) {
+		if (spi_st->tx_ptr) {
+			if (spi_st->bytes_per_word == 1) {
+				word = *spi_st->tx_ptr++;
+			} else {
+				word = *spi_st->tx_ptr++;
+				word = *spi_st->tx_ptr++ | (word << 8);
+			}
+		}
+		writel_relaxed(word, spi_st->base + SSC_TBUF);
+	}
+}
+
+/* Read the RX FIFO */
+static void ssc_read_rx_fifo(struct spi_st *spi_st)
+{
+	unsigned int count, i;
+	uint32_t word = 0;
+
+	if (spi_st->words_remaining > FIFO_SIZE)
+		count = FIFO_SIZE;
+	else
+		count = spi_st->words_remaining;
+
+	for (i = 0; i < count; i++) {
+		word = readl_relaxed(spi_st->base + SSC_RBUF);
+
+		if (spi_st->rx_ptr) {
+			if (spi_st->bytes_per_word == 1) {
+				*spi_st->rx_ptr++ = (uint8_t)word;
+			} else {
+				*spi_st->rx_ptr++ = (word >> 8);
+				*spi_st->rx_ptr++ = word & 0xff;
+			}
+		}
+	}
+	spi_st->words_remaining -= count;
+}
+
+static int spi_st_transfer_one(struct spi_master *master,
+			       struct spi_device *spi, struct spi_transfer *t)
+{
+	struct spi_st *spi_st = spi_master_get_devdata(master);
+	uint32_t ctl = 0;
+
+	/* Setup transfer */
+	spi_st->tx_ptr = t->tx_buf;
+	spi_st->rx_ptr = t->rx_buf;
+
+	if (spi->bits_per_word > 8) {
+		/*
+		 * Anything greater than 8 bits-per-word requires 2
+		 * bytes-per-word in the RX/TX buffers
+		 */
+		spi_st->bytes_per_word = 2;
+		spi_st->words_remaining = t->len / 2;
+
+	} else if (spi->bits_per_word == 8 && !(t->len & 0x1)) {
+		/*
+		 * If transfer is even-length, and 8 bits-per-word, then
+		 * implement as half-length 16 bits-per-word transfer
+		 */
+		spi_st->bytes_per_word = 2;
+		spi_st->words_remaining = t->len / 2;
+
+		/* Set SSC_CTL to 16 bits-per-word */
+		ctl = readl_relaxed(spi_st->base + SSC_CTL);
+		writel_relaxed((ctl | 0xf), spi_st->base + SSC_CTL);
+
+		readl_relaxed(spi_st->base + SSC_RBUF);
+
+	} else {
+		spi_st->bytes_per_word = 1;
+		spi_st->words_remaining = t->len;
+	}
+
+	reinit_completion(&spi_st->done);
+
+	/* Start transfer by writing to the TX FIFO */
+	ssc_write_tx_fifo(spi_st);
+	writel_relaxed(SSC_IEN_TEEN, spi_st->base + SSC_IEN);
+
+	/* Wait for transfer to complete */
+	wait_for_completion(&spi_st->done);
+
+	/* Restore SSC_CTL if necessary */
+	if (ctl)
+		writel_relaxed(ctl, spi_st->base + SSC_CTL);
+
+	spi_finalize_current_transfer(spi->master);
+
+	return t->len;
+}
+
+static void spi_st_cleanup(struct spi_device *spi)
+{
+	gpio_free(spi->cs_gpio);
+}
+
+/* the spi->mode bits understood by this driver: */
+#define MODEBITS  (SPI_CPOL | SPI_CPHA | SPI_LSB_FIRST | SPI_LOOP | SPI_CS_HIGH)
+static int spi_st_setup(struct spi_device *spi)
+{
+	struct spi_st *spi_st = spi_master_get_devdata(spi->master);
+	u32 spi_st_clk, sscbrg, var;
+	u32 hz = spi->max_speed_hz;
+	int cs = spi->cs_gpio;
+	int ret;
+
+	if (!hz)  {
+		dev_err(&spi->dev, "max_speed_hz unspecified\n");
+		return -EINVAL;
+	}
+
+	if (!gpio_is_valid(cs)) {
+		dev_err(&spi->dev, "%d is not a valid gpio\n", cs);
+		return -EINVAL;
+	}
+
+	ret = gpio_request(cs, dev_name(&spi->dev));
+	if (ret) {
+		dev_err(&spi->dev, "could not request gpio:%d\n", cs);
+		return ret;
+	}
+
+	ret = gpio_direction_output(cs, spi->mode & SPI_CS_HIGH);
+	if (ret)
+		goto out_free_gpio;
+
+	spi_st_clk = clk_get_rate(spi_st->clk);
+
+	/* Set SSC_BRF */
+	sscbrg = spi_st_clk / (2 * hz);
+	if (sscbrg < 0x07 || sscbrg > BIT(16)) {
+		dev_err(&spi->dev,
+			"baudrate %d outside valid range %d\n", sscbrg, hz);
+		ret = -EINVAL;
+		goto out_free_gpio;
+	}
+
+	spi_st->baud = spi_st_clk / (2 * sscbrg);
+	if (sscbrg == BIT(16)) /* 16-bit counter wraps */
+		sscbrg = 0x0;
+
+	writel_relaxed(sscbrg, spi_st->base + SSC_BRG);
+
+	dev_dbg(&spi->dev,
+		"setting baudrate:target= %u hz, actual= %u hz, sscbrg= %u\n",
+		hz, spi_st->baud, sscbrg);
+
+	/* Set SSC_CTL and enable SSC */
+	var = readl_relaxed(spi_st->base + SSC_CTL);
+	var |= SSC_CTL_MS;
+
+	if (spi->mode & SPI_CPOL)
+		var |= SSC_CTL_PO;
+	else
+		var &= ~SSC_CTL_PO;
+
+	if (spi->mode & SPI_CPHA)
+		var |= SSC_CTL_PH;
+	else
+		var &= ~SSC_CTL_PH;
+
+	if ((spi->mode & SPI_LSB_FIRST) == 0)
+		var |= SSC_CTL_HB;
+	else
+		var &= ~SSC_CTL_HB;
+
+	if (spi->mode & SPI_LOOP)
+		var |= SSC_CTL_LPB;
+	else
+		var &= ~SSC_CTL_LPB;
+
+	var &= ~SSC_CTL_DATA_WIDTH_MSK;
+	var |= (spi->bits_per_word - 1);
+
+	var |= SSC_CTL_EN_TX_FIFO | SSC_CTL_EN_RX_FIFO;
+	var |= SSC_CTL_EN;
+
+	writel_relaxed(var, spi_st->base + SSC_CTL);
+
+	/* Clear the status register */
+	readl_relaxed(spi_st->base + SSC_RBUF);
+
+	return 0;
+
+out_free_gpio:
+	gpio_free(cs);
+	return ret;
+}
+
+/* Interrupt fired when TX shift register becomes empty */
+static irqreturn_t spi_st_irq(int irq, void *dev_id)
+{
+	struct spi_st *spi_st = (struct spi_st *)dev_id;
+
+	/* Read RX FIFO */
+	ssc_read_rx_fifo(spi_st);
+
+	/* Fill TX FIFO */
+	if (spi_st->words_remaining) {
+		ssc_write_tx_fifo(spi_st);
+	} else {
+		/* TX/RX complete */
+		writel_relaxed(0x0, spi_st->base + SSC_IEN);
+		/*
+		 * read SSC_IEN to ensure that this bit is set
+		 * before re-enabling interrupt
+		 */
+		readl(spi_st->base + SSC_IEN);
+		complete(&spi_st->done);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int spi_st_probe(struct platform_device *pdev)
+{
+	struct device_node *np = pdev->dev.of_node;
+	struct spi_master *master;
+	struct resource *res;
+	struct spi_st *spi_st;
+	int irq, ret = 0;
+	u32 var;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*spi_st));
+	if (!master)
+		return -ENOMEM;
+
+	master->dev.of_node		= np;
+	master->mode_bits		= MODEBITS;
+	master->setup			= spi_st_setup;
+	master->cleanup			= spi_st_cleanup;
+	master->transfer_one		= spi_st_transfer_one;
+	master->bits_per_word_mask	= SPI_BPW_MASK(8) | SPI_BPW_MASK(16);
+	master->auto_runtime_pm		= true;
+	master->bus_num			= pdev->id;
+	spi_st				= spi_master_get_devdata(master);
+
+	spi_st->clk = devm_clk_get(&pdev->dev, "ssc");
+	if (IS_ERR(spi_st->clk)) {
+		dev_err(&pdev->dev, "Unable to request clock\n");
+		ret = PTR_ERR(spi_st->clk);
+		goto put_master;
+	}
+
+	ret = clk_prepare_enable(spi_st->clk);
+	if (ret)
+		goto put_master;
+
+	init_completion(&spi_st->done);
+
+	/* Get resources */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	spi_st->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(spi_st->base)) {
+		ret = PTR_ERR(spi_st->base);
+		goto clk_disable;
+	}
+
+	/* Disable I2C and Reset SSC */
+	writel_relaxed(0x0, spi_st->base + SSC_I2C);
+	var = readw_relaxed(spi_st->base + SSC_CTL);
+	var |= SSC_CTL_SR;
+	writel_relaxed(var, spi_st->base + SSC_CTL);
+
+	udelay(1);
+	var = readl_relaxed(spi_st->base + SSC_CTL);
+	var &= ~SSC_CTL_SR;
+	writel_relaxed(var, spi_st->base + SSC_CTL);
+
+	/* Set SSC into slave mode before reconfiguring PIO pins */
+	var = readl_relaxed(spi_st->base + SSC_CTL);
+	var &= ~SSC_CTL_MS;
+	writel_relaxed(var, spi_st->base + SSC_CTL);
+
+	irq = irq_of_parse_and_map(np, 0);
+	if (!irq) {
+		dev_err(&pdev->dev, "IRQ missing or invalid\n");
+		ret = -EINVAL;
+		goto clk_disable;
+	}
+
+	ret = devm_request_irq(&pdev->dev, irq, spi_st_irq, 0,
+			       pdev->name, spi_st);
+	if (ret) {
+		dev_err(&pdev->dev, "Failed to request irq %d\n", irq);
+		goto clk_disable;
+	}
+
+	/* by default the device is on */
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+
+	platform_set_drvdata(pdev, master);
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret) {
+		dev_err(&pdev->dev, "Failed to register master\n");
+		goto clk_disable;
+	}
+
+	return 0;
+
+clk_disable:
+	clk_disable_unprepare(spi_st->clk);
+put_master:
+	spi_master_put(master);
+	return ret;
+}
+
+static int spi_st_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct spi_st *spi_st = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(spi_st->clk);
+
+	pinctrl_pm_select_sleep_state(&pdev->dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int spi_st_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct spi_st *spi_st = spi_master_get_devdata(master);
+
+	writel_relaxed(0, spi_st->base + SSC_IEN);
+	pinctrl_pm_select_sleep_state(dev);
+
+	clk_disable_unprepare(spi_st->clk);
+
+	return 0;
+}
+
+static int spi_st_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct spi_st *spi_st = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(spi_st->clk);
+	pinctrl_pm_select_default_state(dev);
+
+	return ret;
+}
+#endif
+
+#ifdef CONFIG_PM_SLEEP
+static int spi_st_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	int ret;
+
+	ret = spi_master_suspend(master);
+	if (ret)
+		return ret;
+
+	return pm_runtime_force_suspend(dev);
+}
+
+static int spi_st_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	int ret;
+
+	ret = spi_master_resume(master);
+	if (ret)
+		return ret;
+
+	return pm_runtime_force_resume(dev);
+}
+#endif
+
+static const struct dev_pm_ops spi_st_pm = {
+	SET_SYSTEM_SLEEP_PM_OPS(spi_st_suspend, spi_st_resume)
+	SET_RUNTIME_PM_OPS(spi_st_runtime_suspend, spi_st_runtime_resume, NULL)
+};
+
+static const struct of_device_id stm_spi_match[] = {
+	{ .compatible = "st,comms-ssc4-spi", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, stm_spi_match);
+
+static struct platform_driver spi_st_driver = {
+	.driver = {
+		.name = "spi-st",
+		.pm = &spi_st_pm,
+		.of_match_table = of_match_ptr(stm_spi_match),
+	},
+	.probe = spi_st_probe,
+	.remove = spi_st_remove,
+};
+module_platform_driver(spi_st_driver);
+
+MODULE_AUTHOR("Patrice Chotard <patrice.chotard@st.com>");
+MODULE_DESCRIPTION("STM SSC SPI driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-stm32.c b/drivers/spi/spi-stm32.c
new file mode 100644
index 0000000..ad1e55d
--- /dev/null
+++ b/drivers/spi/spi-stm32.c
@@ -0,0 +1,1322 @@
+/*
+ * STMicroelectronics STM32 SPI Controller driver (master mode only)
+ *
+ * Copyright (C) 2017, STMicroelectronics - All Rights Reserved
+ * Author(s): Amelie Delaunay <amelie.delaunay@st.com> for STMicroelectronics.
+ *
+ * License terms: GPL V2.0.
+ *
+ * spi_stm32 driver is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * spi_stm32 driver is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
+ * details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * spi_stm32 driver. If not, see <http://www.gnu.org/licenses/>.
+ */
+#include <linux/debugfs.h>
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/dmaengine.h>
+#include <linux/gpio.h>
+#include <linux/interrupt.h>
+#include <linux/iopoll.h>
+#include <linux/module.h>
+#include <linux/of_platform.h>
+#include <linux/pm_runtime.h>
+#include <linux/reset.h>
+#include <linux/spi/spi.h>
+
+#define DRIVER_NAME "spi_stm32"
+
+/* STM32 SPI registers */
+#define STM32_SPI_CR1		0x00
+#define STM32_SPI_CR2		0x04
+#define STM32_SPI_CFG1		0x08
+#define STM32_SPI_CFG2		0x0C
+#define STM32_SPI_IER		0x10
+#define STM32_SPI_SR		0x14
+#define STM32_SPI_IFCR		0x18
+#define STM32_SPI_TXDR		0x20
+#define STM32_SPI_RXDR		0x30
+#define STM32_SPI_I2SCFGR	0x50
+
+/* STM32_SPI_CR1 bit fields */
+#define SPI_CR1_SPE		BIT(0)
+#define SPI_CR1_MASRX		BIT(8)
+#define SPI_CR1_CSTART		BIT(9)
+#define SPI_CR1_CSUSP		BIT(10)
+#define SPI_CR1_HDDIR		BIT(11)
+#define SPI_CR1_SSI		BIT(12)
+
+/* STM32_SPI_CR2 bit fields */
+#define SPI_CR2_TSIZE_SHIFT	0
+#define SPI_CR2_TSIZE		GENMASK(15, 0)
+
+/* STM32_SPI_CFG1 bit fields */
+#define SPI_CFG1_DSIZE_SHIFT	0
+#define SPI_CFG1_DSIZE		GENMASK(4, 0)
+#define SPI_CFG1_FTHLV_SHIFT	5
+#define SPI_CFG1_FTHLV		GENMASK(8, 5)
+#define SPI_CFG1_RXDMAEN	BIT(14)
+#define SPI_CFG1_TXDMAEN	BIT(15)
+#define SPI_CFG1_MBR_SHIFT	28
+#define SPI_CFG1_MBR		GENMASK(30, 28)
+#define SPI_CFG1_MBR_MIN	0
+#define SPI_CFG1_MBR_MAX	(GENMASK(30, 28) >> 28)
+
+/* STM32_SPI_CFG2 bit fields */
+#define SPI_CFG2_MIDI_SHIFT	4
+#define SPI_CFG2_MIDI		GENMASK(7, 4)
+#define SPI_CFG2_COMM_SHIFT	17
+#define SPI_CFG2_COMM		GENMASK(18, 17)
+#define SPI_CFG2_SP_SHIFT	19
+#define SPI_CFG2_SP		GENMASK(21, 19)
+#define SPI_CFG2_MASTER		BIT(22)
+#define SPI_CFG2_LSBFRST	BIT(23)
+#define SPI_CFG2_CPHA		BIT(24)
+#define SPI_CFG2_CPOL		BIT(25)
+#define SPI_CFG2_SSM		BIT(26)
+#define SPI_CFG2_AFCNTR		BIT(31)
+
+/* STM32_SPI_IER bit fields */
+#define SPI_IER_RXPIE		BIT(0)
+#define SPI_IER_TXPIE		BIT(1)
+#define SPI_IER_DXPIE		BIT(2)
+#define SPI_IER_EOTIE		BIT(3)
+#define SPI_IER_TXTFIE		BIT(4)
+#define SPI_IER_OVRIE		BIT(6)
+#define SPI_IER_MODFIE		BIT(9)
+#define SPI_IER_ALL		GENMASK(10, 0)
+
+/* STM32_SPI_SR bit fields */
+#define SPI_SR_RXP		BIT(0)
+#define SPI_SR_TXP		BIT(1)
+#define SPI_SR_EOT		BIT(3)
+#define SPI_SR_OVR		BIT(6)
+#define SPI_SR_MODF		BIT(9)
+#define SPI_SR_SUSP		BIT(11)
+#define SPI_SR_RXPLVL_SHIFT	13
+#define SPI_SR_RXPLVL		GENMASK(14, 13)
+#define SPI_SR_RXWNE		BIT(15)
+
+/* STM32_SPI_IFCR bit fields */
+#define SPI_IFCR_ALL		GENMASK(11, 3)
+
+/* STM32_SPI_I2SCFGR bit fields */
+#define SPI_I2SCFGR_I2SMOD	BIT(0)
+
+/* SPI Master Baud Rate min/max divisor */
+#define SPI_MBR_DIV_MIN		(2 << SPI_CFG1_MBR_MIN)
+#define SPI_MBR_DIV_MAX		(2 << SPI_CFG1_MBR_MAX)
+
+/* SPI Communication mode */
+#define SPI_FULL_DUPLEX		0
+#define SPI_SIMPLEX_TX		1
+#define SPI_SIMPLEX_RX		2
+#define SPI_HALF_DUPLEX		3
+
+#define SPI_1HZ_NS		1000000000
+
+/**
+ * struct stm32_spi - private data of the SPI controller
+ * @dev: driver model representation of the controller
+ * @master: controller master interface
+ * @base: virtual memory area
+ * @clk: hw kernel clock feeding the SPI clock generator
+ * @clk_rate: rate of the hw kernel clock feeding the SPI clock generator
+ * @rst: SPI controller reset line
+ * @lock: prevent I/O concurrent access
+ * @irq: SPI controller interrupt line
+ * @fifo_size: size of the embedded fifo in bytes
+ * @cur_midi: master inter-data idleness in ns
+ * @cur_speed: speed configured in Hz
+ * @cur_bpw: number of bits in a single SPI data frame
+ * @cur_fthlv: fifo threshold level (data frames in a single data packet)
+ * @cur_comm: SPI communication mode
+ * @cur_xferlen: current transfer length in bytes
+ * @cur_usedma: boolean to know if dma is used in current transfer
+ * @tx_buf: data to be written, or NULL
+ * @rx_buf: data to be read, or NULL
+ * @tx_len: number of data to be written in bytes
+ * @rx_len: number of data to be read in bytes
+ * @dma_tx: dma channel for TX transfer
+ * @dma_rx: dma channel for RX transfer
+ * @phys_addr: SPI registers physical base address
+ */
+struct stm32_spi {
+	struct device *dev;
+	struct spi_master *master;
+	void __iomem *base;
+	struct clk *clk;
+	u32 clk_rate;
+	struct reset_control *rst;
+	spinlock_t lock; /* prevent I/O concurrent access */
+	int irq;
+	unsigned int fifo_size;
+
+	unsigned int cur_midi;
+	unsigned int cur_speed;
+	unsigned int cur_bpw;
+	unsigned int cur_fthlv;
+	unsigned int cur_comm;
+	unsigned int cur_xferlen;
+	bool cur_usedma;
+
+	const void *tx_buf;
+	void *rx_buf;
+	int tx_len;
+	int rx_len;
+	struct dma_chan *dma_tx;
+	struct dma_chan *dma_rx;
+	dma_addr_t phys_addr;
+};
+
+static inline void stm32_spi_set_bits(struct stm32_spi *spi,
+				      u32 offset, u32 bits)
+{
+	writel_relaxed(readl_relaxed(spi->base + offset) | bits,
+		       spi->base + offset);
+}
+
+static inline void stm32_spi_clr_bits(struct stm32_spi *spi,
+				      u32 offset, u32 bits)
+{
+	writel_relaxed(readl_relaxed(spi->base + offset) & ~bits,
+		       spi->base + offset);
+}
+
+/**
+ * stm32_spi_get_fifo_size - Return fifo size
+ * @spi: pointer to the spi controller data structure
+ */
+static int stm32_spi_get_fifo_size(struct stm32_spi *spi)
+{
+	unsigned long flags;
+	u32 count = 0;
+
+	spin_lock_irqsave(&spi->lock, flags);
+
+	stm32_spi_set_bits(spi, STM32_SPI_CR1, SPI_CR1_SPE);
+
+	while (readl_relaxed(spi->base + STM32_SPI_SR) & SPI_SR_TXP)
+		writeb_relaxed(++count, spi->base + STM32_SPI_TXDR);
+
+	stm32_spi_clr_bits(spi, STM32_SPI_CR1, SPI_CR1_SPE);
+
+	spin_unlock_irqrestore(&spi->lock, flags);
+
+	dev_dbg(spi->dev, "%d x 8-bit fifo size\n", count);
+
+	return count;
+}
+
+/**
+ * stm32_spi_get_bpw_mask - Return bits per word mask
+ * @spi: pointer to the spi controller data structure
+ */
+static int stm32_spi_get_bpw_mask(struct stm32_spi *spi)
+{
+	unsigned long flags;
+	u32 cfg1, max_bpw;
+
+	spin_lock_irqsave(&spi->lock, flags);
+
+	/*
+	 * The most significant bit at DSIZE bit field is reserved when the
+	 * maximum data size of periperal instances is limited to 16-bit
+	 */
+	stm32_spi_set_bits(spi, STM32_SPI_CFG1, SPI_CFG1_DSIZE);
+
+	cfg1 = readl_relaxed(spi->base + STM32_SPI_CFG1);
+	max_bpw = (cfg1 & SPI_CFG1_DSIZE) >> SPI_CFG1_DSIZE_SHIFT;
+	max_bpw += 1;
+
+	spin_unlock_irqrestore(&spi->lock, flags);
+
+	dev_dbg(spi->dev, "%d-bit maximum data frame\n", max_bpw);
+
+	return SPI_BPW_RANGE_MASK(4, max_bpw);
+}
+
+/**
+ * stm32_spi_prepare_mbr - Determine SPI_CFG1.MBR value
+ * @spi: pointer to the spi controller data structure
+ * @speed_hz: requested speed
+ *
+ * Return SPI_CFG1.MBR value in case of success or -EINVAL
+ */
+static int stm32_spi_prepare_mbr(struct stm32_spi *spi, u32 speed_hz)
+{
+	u32 div, mbrdiv;
+
+	div = DIV_ROUND_UP(spi->clk_rate, speed_hz);
+
+	/*
+	 * SPI framework set xfer->speed_hz to master->max_speed_hz if
+	 * xfer->speed_hz is greater than master->max_speed_hz, and it returns
+	 * an error when xfer->speed_hz is lower than master->min_speed_hz, so
+	 * no need to check it there.
+	 * However, we need to ensure the following calculations.
+	 */
+	if (div < SPI_MBR_DIV_MIN ||
+	    div > SPI_MBR_DIV_MAX)
+		return -EINVAL;
+
+	/* Determine the first power of 2 greater than or equal to div */
+	if (div & (div - 1))
+		mbrdiv = fls(div);
+	else
+		mbrdiv = fls(div) - 1;
+
+	spi->cur_speed = spi->clk_rate / (1 << mbrdiv);
+
+	return mbrdiv - 1;
+}
+
+/**
+ * stm32_spi_prepare_fthlv - Determine FIFO threshold level
+ * @spi: pointer to the spi controller data structure
+ */
+static u32 stm32_spi_prepare_fthlv(struct stm32_spi *spi)
+{
+	u32 fthlv, half_fifo;
+
+	/* data packet should not exceed 1/2 of fifo space */
+	half_fifo = (spi->fifo_size / 2);
+
+	if (spi->cur_bpw <= 8)
+		fthlv = half_fifo;
+	else if (spi->cur_bpw <= 16)
+		fthlv = half_fifo / 2;
+	else
+		fthlv = half_fifo / 4;
+
+	/* align packet size with data registers access */
+	if (spi->cur_bpw > 8)
+		fthlv -= (fthlv % 2); /* multiple of 2 */
+	else
+		fthlv -= (fthlv % 4); /* multiple of 4 */
+
+	return fthlv;
+}
+
+/**
+ * stm32_spi_write_txfifo - Write bytes in Transmit Data Register
+ * @spi: pointer to the spi controller data structure
+ *
+ * Read from tx_buf depends on remaining bytes to avoid to read beyond
+ * tx_buf end.
+ */
+static void stm32_spi_write_txfifo(struct stm32_spi *spi)
+{
+	while ((spi->tx_len > 0) &&
+	       (readl_relaxed(spi->base + STM32_SPI_SR) & SPI_SR_TXP)) {
+		u32 offs = spi->cur_xferlen - spi->tx_len;
+
+		if (spi->tx_len >= sizeof(u32)) {
+			const u32 *tx_buf32 = (const u32 *)(spi->tx_buf + offs);
+
+			writel_relaxed(*tx_buf32, spi->base + STM32_SPI_TXDR);
+			spi->tx_len -= sizeof(u32);
+		} else if (spi->tx_len >= sizeof(u16)) {
+			const u16 *tx_buf16 = (const u16 *)(spi->tx_buf + offs);
+
+			writew_relaxed(*tx_buf16, spi->base + STM32_SPI_TXDR);
+			spi->tx_len -= sizeof(u16);
+		} else {
+			const u8 *tx_buf8 = (const u8 *)(spi->tx_buf + offs);
+
+			writeb_relaxed(*tx_buf8, spi->base + STM32_SPI_TXDR);
+			spi->tx_len -= sizeof(u8);
+		}
+	}
+
+	dev_dbg(spi->dev, "%s: %d bytes left\n", __func__, spi->tx_len);
+}
+
+/**
+ * stm32_spi_read_rxfifo - Read bytes in Receive Data Register
+ * @spi: pointer to the spi controller data structure
+ *
+ * Write in rx_buf depends on remaining bytes to avoid to write beyond
+ * rx_buf end.
+ */
+static void stm32_spi_read_rxfifo(struct stm32_spi *spi, bool flush)
+{
+	u32 sr = readl_relaxed(spi->base + STM32_SPI_SR);
+	u32 rxplvl = (sr & SPI_SR_RXPLVL) >> SPI_SR_RXPLVL_SHIFT;
+
+	while ((spi->rx_len > 0) &&
+	       ((sr & SPI_SR_RXP) ||
+		(flush && ((sr & SPI_SR_RXWNE) || (rxplvl > 0))))) {
+		u32 offs = spi->cur_xferlen - spi->rx_len;
+
+		if ((spi->rx_len >= sizeof(u32)) ||
+		    (flush && (sr & SPI_SR_RXWNE))) {
+			u32 *rx_buf32 = (u32 *)(spi->rx_buf + offs);
+
+			*rx_buf32 = readl_relaxed(spi->base + STM32_SPI_RXDR);
+			spi->rx_len -= sizeof(u32);
+		} else if ((spi->rx_len >= sizeof(u16)) ||
+			   (flush && (rxplvl >= 2 || spi->cur_bpw > 8))) {
+			u16 *rx_buf16 = (u16 *)(spi->rx_buf + offs);
+
+			*rx_buf16 = readw_relaxed(spi->base + STM32_SPI_RXDR);
+			spi->rx_len -= sizeof(u16);
+		} else {
+			u8 *rx_buf8 = (u8 *)(spi->rx_buf + offs);
+
+			*rx_buf8 = readb_relaxed(spi->base + STM32_SPI_RXDR);
+			spi->rx_len -= sizeof(u8);
+		}
+
+		sr = readl_relaxed(spi->base + STM32_SPI_SR);
+		rxplvl = (sr & SPI_SR_RXPLVL) >> SPI_SR_RXPLVL_SHIFT;
+	}
+
+	dev_dbg(spi->dev, "%s%s: %d bytes left\n", __func__,
+		flush ? "(flush)" : "", spi->rx_len);
+}
+
+/**
+ * stm32_spi_enable - Enable SPI controller
+ * @spi: pointer to the spi controller data structure
+ *
+ * SPI data transfer is enabled but spi_ker_ck is idle.
+ * SPI_CFG1 and SPI_CFG2 are now write protected.
+ */
+static void stm32_spi_enable(struct stm32_spi *spi)
+{
+	dev_dbg(spi->dev, "enable controller\n");
+
+	stm32_spi_set_bits(spi, STM32_SPI_CR1, SPI_CR1_SPE);
+}
+
+/**
+ * stm32_spi_disable - Disable SPI controller
+ * @spi: pointer to the spi controller data structure
+ *
+ * RX-Fifo is flushed when SPI controller is disabled. To prevent any data
+ * loss, use stm32_spi_read_rxfifo(flush) to read the remaining bytes in
+ * RX-Fifo.
+ */
+static void stm32_spi_disable(struct stm32_spi *spi)
+{
+	unsigned long flags;
+	u32 cr1, sr;
+
+	dev_dbg(spi->dev, "disable controller\n");
+
+	spin_lock_irqsave(&spi->lock, flags);
+
+	cr1 = readl_relaxed(spi->base + STM32_SPI_CR1);
+
+	if (!(cr1 & SPI_CR1_SPE)) {
+		spin_unlock_irqrestore(&spi->lock, flags);
+		return;
+	}
+
+	/* Wait on EOT or suspend the flow */
+	if (readl_relaxed_poll_timeout_atomic(spi->base + STM32_SPI_SR,
+					      sr, !(sr & SPI_SR_EOT),
+					      10, 100000) < 0) {
+		if (cr1 & SPI_CR1_CSTART) {
+			writel_relaxed(cr1 | SPI_CR1_CSUSP,
+				       spi->base + STM32_SPI_CR1);
+			if (readl_relaxed_poll_timeout_atomic(
+						spi->base + STM32_SPI_SR,
+						sr, !(sr & SPI_SR_SUSP),
+						10, 100000) < 0)
+				dev_warn(spi->dev,
+					 "Suspend request timeout\n");
+		}
+	}
+
+	if (!spi->cur_usedma && spi->rx_buf && (spi->rx_len > 0))
+		stm32_spi_read_rxfifo(spi, true);
+
+	if (spi->cur_usedma && spi->tx_buf)
+		dmaengine_terminate_all(spi->dma_tx);
+	if (spi->cur_usedma && spi->rx_buf)
+		dmaengine_terminate_all(spi->dma_rx);
+
+	stm32_spi_clr_bits(spi, STM32_SPI_CR1, SPI_CR1_SPE);
+
+	stm32_spi_clr_bits(spi, STM32_SPI_CFG1, SPI_CFG1_TXDMAEN |
+						SPI_CFG1_RXDMAEN);
+
+	/* Disable interrupts and clear status flags */
+	writel_relaxed(0, spi->base + STM32_SPI_IER);
+	writel_relaxed(SPI_IFCR_ALL, spi->base + STM32_SPI_IFCR);
+
+	spin_unlock_irqrestore(&spi->lock, flags);
+}
+
+/**
+ * stm32_spi_can_dma - Determine if the transfer is eligible for DMA use
+ *
+ * If the current transfer size is greater than fifo size, use DMA.
+ */
+static bool stm32_spi_can_dma(struct spi_master *master,
+			      struct spi_device *spi_dev,
+			      struct spi_transfer *transfer)
+{
+	struct stm32_spi *spi = spi_master_get_devdata(master);
+
+	dev_dbg(spi->dev, "%s: %s\n", __func__,
+		(transfer->len > spi->fifo_size) ? "true" : "false");
+
+	return (transfer->len > spi->fifo_size);
+}
+
+/**
+ * stm32_spi_irq - Interrupt handler for SPI controller events
+ * @irq: interrupt line
+ * @dev_id: SPI controller master interface
+ */
+static irqreturn_t stm32_spi_irq(int irq, void *dev_id)
+{
+	struct spi_master *master = dev_id;
+	struct stm32_spi *spi = spi_master_get_devdata(master);
+	u32 sr, ier, mask;
+	unsigned long flags;
+	bool end = false;
+
+	spin_lock_irqsave(&spi->lock, flags);
+
+	sr = readl_relaxed(spi->base + STM32_SPI_SR);
+	ier = readl_relaxed(spi->base + STM32_SPI_IER);
+
+	mask = ier;
+	/* EOTIE is triggered on EOT, SUSP and TXC events. */
+	mask |= SPI_SR_SUSP;
+	/*
+	 * When TXTF is set, DXPIE and TXPIE are cleared. So in case of
+	 * Full-Duplex, need to poll RXP event to know if there are remaining
+	 * data, before disabling SPI.
+	 */
+	if (spi->rx_buf && !spi->cur_usedma)
+		mask |= SPI_SR_RXP;
+
+	if (!(sr & mask)) {
+		dev_dbg(spi->dev, "spurious IT (sr=0x%08x, ier=0x%08x)\n",
+			sr, ier);
+		spin_unlock_irqrestore(&spi->lock, flags);
+		return IRQ_NONE;
+	}
+
+	if (sr & SPI_SR_SUSP) {
+		dev_warn(spi->dev, "Communication suspended\n");
+		if (!spi->cur_usedma && (spi->rx_buf && (spi->rx_len > 0)))
+			stm32_spi_read_rxfifo(spi, false);
+		/*
+		 * If communication is suspended while using DMA, it means
+		 * that something went wrong, so stop the current transfer
+		 */
+		if (spi->cur_usedma)
+			end = true;
+	}
+
+	if (sr & SPI_SR_MODF) {
+		dev_warn(spi->dev, "Mode fault: transfer aborted\n");
+		end = true;
+	}
+
+	if (sr & SPI_SR_OVR) {
+		dev_warn(spi->dev, "Overrun: received value discarded\n");
+		if (!spi->cur_usedma && (spi->rx_buf && (spi->rx_len > 0)))
+			stm32_spi_read_rxfifo(spi, false);
+		/*
+		 * If overrun is detected while using DMA, it means that
+		 * something went wrong, so stop the current transfer
+		 */
+		if (spi->cur_usedma)
+			end = true;
+	}
+
+	if (sr & SPI_SR_EOT) {
+		if (!spi->cur_usedma && (spi->rx_buf && (spi->rx_len > 0)))
+			stm32_spi_read_rxfifo(spi, true);
+		end = true;
+	}
+
+	if (sr & SPI_SR_TXP)
+		if (!spi->cur_usedma && (spi->tx_buf && (spi->tx_len > 0)))
+			stm32_spi_write_txfifo(spi);
+
+	if (sr & SPI_SR_RXP)
+		if (!spi->cur_usedma && (spi->rx_buf && (spi->rx_len > 0)))
+			stm32_spi_read_rxfifo(spi, false);
+
+	writel_relaxed(mask, spi->base + STM32_SPI_IFCR);
+
+	spin_unlock_irqrestore(&spi->lock, flags);
+
+	if (end) {
+		spi_finalize_current_transfer(master);
+		stm32_spi_disable(spi);
+	}
+
+	return IRQ_HANDLED;
+}
+
+/**
+ * stm32_spi_setup - setup device chip select
+ */
+static int stm32_spi_setup(struct spi_device *spi_dev)
+{
+	int ret = 0;
+
+	if (!gpio_is_valid(spi_dev->cs_gpio)) {
+		dev_err(&spi_dev->dev, "%d is not a valid gpio\n",
+			spi_dev->cs_gpio);
+		return -EINVAL;
+	}
+
+	dev_dbg(&spi_dev->dev, "%s: set gpio%d output %s\n", __func__,
+		spi_dev->cs_gpio,
+		(spi_dev->mode & SPI_CS_HIGH) ? "low" : "high");
+
+	ret = gpio_direction_output(spi_dev->cs_gpio,
+				    !(spi_dev->mode & SPI_CS_HIGH));
+
+	return ret;
+}
+
+/**
+ * stm32_spi_prepare_msg - set up the controller to transfer a single message
+ */
+static int stm32_spi_prepare_msg(struct spi_master *master,
+				 struct spi_message *msg)
+{
+	struct stm32_spi *spi = spi_master_get_devdata(master);
+	struct spi_device *spi_dev = msg->spi;
+	struct device_node *np = spi_dev->dev.of_node;
+	unsigned long flags;
+	u32 cfg2_clrb = 0, cfg2_setb = 0;
+
+	/* SPI slave device may need time between data frames */
+	spi->cur_midi = 0;
+	if (np && !of_property_read_u32(np, "st,spi-midi-ns", &spi->cur_midi))
+		dev_dbg(spi->dev, "%dns inter-data idleness\n", spi->cur_midi);
+
+	if (spi_dev->mode & SPI_CPOL)
+		cfg2_setb |= SPI_CFG2_CPOL;
+	else
+		cfg2_clrb |= SPI_CFG2_CPOL;
+
+	if (spi_dev->mode & SPI_CPHA)
+		cfg2_setb |= SPI_CFG2_CPHA;
+	else
+		cfg2_clrb |= SPI_CFG2_CPHA;
+
+	if (spi_dev->mode & SPI_LSB_FIRST)
+		cfg2_setb |= SPI_CFG2_LSBFRST;
+	else
+		cfg2_clrb |= SPI_CFG2_LSBFRST;
+
+	dev_dbg(spi->dev, "cpol=%d cpha=%d lsb_first=%d cs_high=%d\n",
+		spi_dev->mode & SPI_CPOL,
+		spi_dev->mode & SPI_CPHA,
+		spi_dev->mode & SPI_LSB_FIRST,
+		spi_dev->mode & SPI_CS_HIGH);
+
+	spin_lock_irqsave(&spi->lock, flags);
+
+	if (cfg2_clrb || cfg2_setb)
+		writel_relaxed(
+			(readl_relaxed(spi->base + STM32_SPI_CFG2) &
+				~cfg2_clrb) | cfg2_setb,
+			       spi->base + STM32_SPI_CFG2);
+
+	spin_unlock_irqrestore(&spi->lock, flags);
+
+	return 0;
+}
+
+/**
+ * stm32_spi_dma_cb - dma callback
+ *
+ * DMA callback is called when the transfer is complete or when an error
+ * occurs. If the transfer is complete, EOT flag is raised.
+ */
+static void stm32_spi_dma_cb(void *data)
+{
+	struct stm32_spi *spi = data;
+	unsigned long flags;
+	u32 sr;
+
+	spin_lock_irqsave(&spi->lock, flags);
+
+	sr = readl_relaxed(spi->base + STM32_SPI_SR);
+
+	spin_unlock_irqrestore(&spi->lock, flags);
+
+	if (!(sr & SPI_SR_EOT))
+		dev_warn(spi->dev, "DMA error (sr=0x%08x)\n", sr);
+
+	/* Now wait for EOT, or SUSP or OVR in case of error */
+}
+
+/**
+ * stm32_spi_dma_config - configure dma slave channel depending on current
+ *			  transfer bits_per_word.
+ */
+static void stm32_spi_dma_config(struct stm32_spi *spi,
+				 struct dma_slave_config *dma_conf,
+				 enum dma_transfer_direction dir)
+{
+	enum dma_slave_buswidth buswidth;
+	u32 maxburst;
+
+	if (spi->cur_bpw <= 8)
+		buswidth = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	else if (spi->cur_bpw <= 16)
+		buswidth = DMA_SLAVE_BUSWIDTH_2_BYTES;
+	else
+		buswidth = DMA_SLAVE_BUSWIDTH_4_BYTES;
+
+	/* Valid for DMA Half or Full Fifo threshold */
+	if (spi->cur_fthlv == 2)
+		maxburst = 1;
+	else
+		maxburst = spi->cur_fthlv;
+
+	memset(dma_conf, 0, sizeof(struct dma_slave_config));
+	dma_conf->direction = dir;
+	if (dma_conf->direction == DMA_DEV_TO_MEM) { /* RX */
+		dma_conf->src_addr = spi->phys_addr + STM32_SPI_RXDR;
+		dma_conf->src_addr_width = buswidth;
+		dma_conf->src_maxburst = maxburst;
+
+		dev_dbg(spi->dev, "Rx DMA config buswidth=%d, maxburst=%d\n",
+			buswidth, maxburst);
+	} else if (dma_conf->direction == DMA_MEM_TO_DEV) { /* TX */
+		dma_conf->dst_addr = spi->phys_addr + STM32_SPI_TXDR;
+		dma_conf->dst_addr_width = buswidth;
+		dma_conf->dst_maxburst = maxburst;
+
+		dev_dbg(spi->dev, "Tx DMA config buswidth=%d, maxburst=%d\n",
+			buswidth, maxburst);
+	}
+}
+
+/**
+ * stm32_spi_transfer_one_irq - transfer a single spi_transfer using
+ *				interrupts
+ *
+ * It must returns 0 if the transfer is finished or 1 if the transfer is still
+ * in progress.
+ */
+static int stm32_spi_transfer_one_irq(struct stm32_spi *spi)
+{
+	unsigned long flags;
+	u32 ier = 0;
+
+	/* Enable the interrupts relative to the current communication mode */
+	if (spi->tx_buf && spi->rx_buf)	/* Full Duplex */
+		ier |= SPI_IER_DXPIE;
+	else if (spi->tx_buf)		/* Half-Duplex TX dir or Simplex TX */
+		ier |= SPI_IER_TXPIE;
+	else if (spi->rx_buf)		/* Half-Duplex RX dir or Simplex RX */
+		ier |= SPI_IER_RXPIE;
+
+	/* Enable the interrupts relative to the end of transfer */
+	ier |= SPI_IER_EOTIE | SPI_IER_TXTFIE |	SPI_IER_OVRIE |	SPI_IER_MODFIE;
+
+	spin_lock_irqsave(&spi->lock, flags);
+
+	stm32_spi_enable(spi);
+
+	/* Be sure to have data in fifo before starting data transfer */
+	if (spi->tx_buf)
+		stm32_spi_write_txfifo(spi);
+
+	stm32_spi_set_bits(spi, STM32_SPI_CR1, SPI_CR1_CSTART);
+
+	writel_relaxed(ier, spi->base + STM32_SPI_IER);
+
+	spin_unlock_irqrestore(&spi->lock, flags);
+
+	return 1;
+}
+
+/**
+ * stm32_spi_transfer_one_dma - transfer a single spi_transfer using DMA
+ *
+ * It must returns 0 if the transfer is finished or 1 if the transfer is still
+ * in progress.
+ */
+static int stm32_spi_transfer_one_dma(struct stm32_spi *spi,
+				      struct spi_transfer *xfer)
+{
+	struct dma_slave_config tx_dma_conf, rx_dma_conf;
+	struct dma_async_tx_descriptor *tx_dma_desc, *rx_dma_desc;
+	unsigned long flags;
+	u32 ier = 0;
+
+	spin_lock_irqsave(&spi->lock, flags);
+
+	rx_dma_desc = NULL;
+	if (spi->rx_buf) {
+		stm32_spi_dma_config(spi, &rx_dma_conf, DMA_DEV_TO_MEM);
+		dmaengine_slave_config(spi->dma_rx, &rx_dma_conf);
+
+		/* Enable Rx DMA request */
+		stm32_spi_set_bits(spi, STM32_SPI_CFG1, SPI_CFG1_RXDMAEN);
+
+		rx_dma_desc = dmaengine_prep_slave_sg(
+					spi->dma_rx, xfer->rx_sg.sgl,
+					xfer->rx_sg.nents,
+					rx_dma_conf.direction,
+					DMA_PREP_INTERRUPT);
+	}
+
+	tx_dma_desc = NULL;
+	if (spi->tx_buf) {
+		stm32_spi_dma_config(spi, &tx_dma_conf, DMA_MEM_TO_DEV);
+		dmaengine_slave_config(spi->dma_tx, &tx_dma_conf);
+
+		tx_dma_desc = dmaengine_prep_slave_sg(
+					spi->dma_tx, xfer->tx_sg.sgl,
+					xfer->tx_sg.nents,
+					tx_dma_conf.direction,
+					DMA_PREP_INTERRUPT);
+	}
+
+	if ((spi->tx_buf && !tx_dma_desc) ||
+	    (spi->rx_buf && !rx_dma_desc))
+		goto dma_desc_error;
+
+	if (rx_dma_desc) {
+		rx_dma_desc->callback = stm32_spi_dma_cb;
+		rx_dma_desc->callback_param = spi;
+
+		if (dma_submit_error(dmaengine_submit(rx_dma_desc))) {
+			dev_err(spi->dev, "Rx DMA submit failed\n");
+			goto dma_desc_error;
+		}
+		/* Enable Rx DMA channel */
+		dma_async_issue_pending(spi->dma_rx);
+	}
+
+	if (tx_dma_desc) {
+		if (spi->cur_comm == SPI_SIMPLEX_TX) {
+			tx_dma_desc->callback = stm32_spi_dma_cb;
+			tx_dma_desc->callback_param = spi;
+		}
+
+		if (dma_submit_error(dmaengine_submit(tx_dma_desc))) {
+			dev_err(spi->dev, "Tx DMA submit failed\n");
+			goto dma_submit_error;
+		}
+		/* Enable Tx DMA channel */
+		dma_async_issue_pending(spi->dma_tx);
+
+		/* Enable Tx DMA request */
+		stm32_spi_set_bits(spi, STM32_SPI_CFG1, SPI_CFG1_TXDMAEN);
+	}
+
+	/* Enable the interrupts relative to the end of transfer */
+	ier |= SPI_IER_EOTIE | SPI_IER_TXTFIE |	SPI_IER_OVRIE |	SPI_IER_MODFIE;
+	writel_relaxed(ier, spi->base + STM32_SPI_IER);
+
+	stm32_spi_enable(spi);
+
+	stm32_spi_set_bits(spi, STM32_SPI_CR1, SPI_CR1_CSTART);
+
+	spin_unlock_irqrestore(&spi->lock, flags);
+
+	return 1;
+
+dma_submit_error:
+	if (spi->rx_buf)
+		dmaengine_terminate_all(spi->dma_rx);
+
+dma_desc_error:
+	stm32_spi_clr_bits(spi, STM32_SPI_CFG1, SPI_CFG1_RXDMAEN);
+
+	spin_unlock_irqrestore(&spi->lock, flags);
+
+	dev_info(spi->dev, "DMA issue: fall back to irq transfer\n");
+
+	return stm32_spi_transfer_one_irq(spi);
+}
+
+/**
+ * stm32_spi_transfer_one_setup - common setup to transfer a single
+ *				  spi_transfer either using DMA or
+ *				  interrupts.
+ */
+static int stm32_spi_transfer_one_setup(struct stm32_spi *spi,
+					struct spi_device *spi_dev,
+					struct spi_transfer *transfer)
+{
+	unsigned long flags;
+	u32 cfg1_clrb = 0, cfg1_setb = 0, cfg2_clrb = 0, cfg2_setb = 0;
+	u32 mode, nb_words;
+	int ret = 0;
+
+	spin_lock_irqsave(&spi->lock, flags);
+
+	if (spi->cur_bpw != transfer->bits_per_word) {
+		u32 bpw, fthlv;
+
+		spi->cur_bpw = transfer->bits_per_word;
+		bpw = spi->cur_bpw - 1;
+
+		cfg1_clrb |= SPI_CFG1_DSIZE;
+		cfg1_setb |= (bpw << SPI_CFG1_DSIZE_SHIFT) & SPI_CFG1_DSIZE;
+
+		spi->cur_fthlv = stm32_spi_prepare_fthlv(spi);
+		fthlv = spi->cur_fthlv - 1;
+
+		cfg1_clrb |= SPI_CFG1_FTHLV;
+		cfg1_setb |= (fthlv << SPI_CFG1_FTHLV_SHIFT) & SPI_CFG1_FTHLV;
+	}
+
+	if (spi->cur_speed != transfer->speed_hz) {
+		int mbr;
+
+		/* Update spi->cur_speed with real clock speed */
+		mbr = stm32_spi_prepare_mbr(spi, transfer->speed_hz);
+		if (mbr < 0) {
+			ret = mbr;
+			goto out;
+		}
+
+		transfer->speed_hz = spi->cur_speed;
+
+		cfg1_clrb |= SPI_CFG1_MBR;
+		cfg1_setb |= ((u32)mbr << SPI_CFG1_MBR_SHIFT) & SPI_CFG1_MBR;
+	}
+
+	if (cfg1_clrb || cfg1_setb)
+		writel_relaxed((readl_relaxed(spi->base + STM32_SPI_CFG1) &
+				~cfg1_clrb) | cfg1_setb,
+			       spi->base + STM32_SPI_CFG1);
+
+	mode = SPI_FULL_DUPLEX;
+	if (spi_dev->mode & SPI_3WIRE) { /* MISO/MOSI signals shared */
+		/*
+		 * SPI_3WIRE and xfer->tx_buf != NULL and xfer->rx_buf != NULL
+		 * is forbidden und unvalidated by SPI subsystem so depending
+		 * on the valid buffer, we can determine the direction of the
+		 * transfer.
+		 */
+		mode = SPI_HALF_DUPLEX;
+		if (!transfer->tx_buf)
+			stm32_spi_clr_bits(spi, STM32_SPI_CR1, SPI_CR1_HDDIR);
+		else if (!transfer->rx_buf)
+			stm32_spi_set_bits(spi, STM32_SPI_CR1, SPI_CR1_HDDIR);
+	} else {
+		if (!transfer->tx_buf)
+			mode = SPI_SIMPLEX_RX;
+		else if (!transfer->rx_buf)
+			mode = SPI_SIMPLEX_TX;
+	}
+	if (spi->cur_comm != mode) {
+		spi->cur_comm = mode;
+
+		cfg2_clrb |= SPI_CFG2_COMM;
+		cfg2_setb |= (mode << SPI_CFG2_COMM_SHIFT) & SPI_CFG2_COMM;
+	}
+
+	cfg2_clrb |= SPI_CFG2_MIDI;
+	if ((transfer->len > 1) && (spi->cur_midi > 0)) {
+		u32 sck_period_ns = DIV_ROUND_UP(SPI_1HZ_NS, spi->cur_speed);
+		u32 midi = min((u32)DIV_ROUND_UP(spi->cur_midi, sck_period_ns),
+			       (u32)SPI_CFG2_MIDI >> SPI_CFG2_MIDI_SHIFT);
+
+		dev_dbg(spi->dev, "period=%dns, midi=%d(=%dns)\n",
+			sck_period_ns, midi, midi * sck_period_ns);
+
+		cfg2_setb |= (midi << SPI_CFG2_MIDI_SHIFT) & SPI_CFG2_MIDI;
+	}
+
+	if (cfg2_clrb || cfg2_setb)
+		writel_relaxed((readl_relaxed(spi->base + STM32_SPI_CFG2) &
+				~cfg2_clrb) | cfg2_setb,
+			       spi->base + STM32_SPI_CFG2);
+
+	if (spi->cur_bpw <= 8)
+		nb_words = transfer->len;
+	else if (spi->cur_bpw <= 16)
+		nb_words = DIV_ROUND_UP(transfer->len * 8, 16);
+	else
+		nb_words = DIV_ROUND_UP(transfer->len * 8, 32);
+	nb_words <<= SPI_CR2_TSIZE_SHIFT;
+
+	if (nb_words <= SPI_CR2_TSIZE) {
+		writel_relaxed(nb_words, spi->base + STM32_SPI_CR2);
+	} else {
+		ret = -EMSGSIZE;
+		goto out;
+	}
+
+	spi->cur_xferlen = transfer->len;
+
+	dev_dbg(spi->dev, "transfer communication mode set to %d\n",
+		spi->cur_comm);
+	dev_dbg(spi->dev,
+		"data frame of %d-bit, data packet of %d data frames\n",
+		spi->cur_bpw, spi->cur_fthlv);
+	dev_dbg(spi->dev, "speed set to %dHz\n", spi->cur_speed);
+	dev_dbg(spi->dev, "transfer of %d bytes (%d data frames)\n",
+		spi->cur_xferlen, nb_words);
+	dev_dbg(spi->dev, "dma %s\n",
+		(spi->cur_usedma) ? "enabled" : "disabled");
+
+out:
+	spin_unlock_irqrestore(&spi->lock, flags);
+
+	return ret;
+}
+
+/**
+ * stm32_spi_transfer_one - transfer a single spi_transfer
+ *
+ * It must return 0 if the transfer is finished or 1 if the transfer is still
+ * in progress.
+ */
+static int stm32_spi_transfer_one(struct spi_master *master,
+				  struct spi_device *spi_dev,
+				  struct spi_transfer *transfer)
+{
+	struct stm32_spi *spi = spi_master_get_devdata(master);
+	int ret;
+
+	spi->tx_buf = transfer->tx_buf;
+	spi->rx_buf = transfer->rx_buf;
+	spi->tx_len = spi->tx_buf ? transfer->len : 0;
+	spi->rx_len = spi->rx_buf ? transfer->len : 0;
+
+	spi->cur_usedma = (master->can_dma &&
+			   stm32_spi_can_dma(master, spi_dev, transfer));
+
+	ret = stm32_spi_transfer_one_setup(spi, spi_dev, transfer);
+	if (ret) {
+		dev_err(spi->dev, "SPI transfer setup failed\n");
+		return ret;
+	}
+
+	if (spi->cur_usedma)
+		return stm32_spi_transfer_one_dma(spi, transfer);
+	else
+		return stm32_spi_transfer_one_irq(spi);
+}
+
+/**
+ * stm32_spi_unprepare_msg - relax the hardware
+ *
+ * Normally, if TSIZE has been configured, we should relax the hardware at the
+ * reception of the EOT interrupt. But in case of error, EOT will not be
+ * raised. So the subsystem unprepare_message call allows us to properly
+ * complete the transfer from an hardware point of view.
+ */
+static int stm32_spi_unprepare_msg(struct spi_master *master,
+				   struct spi_message *msg)
+{
+	struct stm32_spi *spi = spi_master_get_devdata(master);
+
+	stm32_spi_disable(spi);
+
+	return 0;
+}
+
+/**
+ * stm32_spi_config - Configure SPI controller as SPI master
+ */
+static int stm32_spi_config(struct stm32_spi *spi)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&spi->lock, flags);
+
+	/* Ensure I2SMOD bit is kept cleared */
+	stm32_spi_clr_bits(spi, STM32_SPI_I2SCFGR, SPI_I2SCFGR_I2SMOD);
+
+	/*
+	 * - SS input value high
+	 * - transmitter half duplex direction
+	 * - automatic communication suspend when RX-Fifo is full
+	 */
+	stm32_spi_set_bits(spi, STM32_SPI_CR1, SPI_CR1_SSI |
+					       SPI_CR1_HDDIR |
+					       SPI_CR1_MASRX);
+
+	/*
+	 * - Set the master mode (default Motorola mode)
+	 * - Consider 1 master/n slaves configuration and
+	 *   SS input value is determined by the SSI bit
+	 * - keep control of all associated GPIOs
+	 */
+	stm32_spi_set_bits(spi, STM32_SPI_CFG2, SPI_CFG2_MASTER |
+						SPI_CFG2_SSM |
+						SPI_CFG2_AFCNTR);
+
+	spin_unlock_irqrestore(&spi->lock, flags);
+
+	return 0;
+}
+
+static const struct of_device_id stm32_spi_of_match[] = {
+	{ .compatible = "st,stm32h7-spi", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, stm32_spi_of_match);
+
+static int stm32_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct stm32_spi *spi;
+	struct resource *res;
+	int i, ret;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct stm32_spi));
+	if (!master) {
+		dev_err(&pdev->dev, "spi master allocation failed\n");
+		return -ENOMEM;
+	}
+	platform_set_drvdata(pdev, master);
+
+	spi = spi_master_get_devdata(master);
+	spi->dev = &pdev->dev;
+	spi->master = master;
+	spin_lock_init(&spi->lock);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	spi->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(spi->base)) {
+		ret = PTR_ERR(spi->base);
+		goto err_master_put;
+	}
+	spi->phys_addr = (dma_addr_t)res->start;
+
+	spi->irq = platform_get_irq(pdev, 0);
+	if (spi->irq <= 0) {
+		dev_err(&pdev->dev, "no irq: %d\n", spi->irq);
+		ret = -ENOENT;
+		goto err_master_put;
+	}
+	ret = devm_request_threaded_irq(&pdev->dev, spi->irq, NULL,
+					stm32_spi_irq, IRQF_ONESHOT,
+					pdev->name, master);
+	if (ret) {
+		dev_err(&pdev->dev, "irq%d request failed: %d\n", spi->irq,
+			ret);
+		goto err_master_put;
+	}
+
+	spi->clk = devm_clk_get(&pdev->dev, 0);
+	if (IS_ERR(spi->clk)) {
+		ret = PTR_ERR(spi->clk);
+		dev_err(&pdev->dev, "clk get failed: %d\n", ret);
+		goto err_master_put;
+	}
+
+	ret = clk_prepare_enable(spi->clk);
+	if (ret) {
+		dev_err(&pdev->dev, "clk enable failed: %d\n", ret);
+		goto err_master_put;
+	}
+	spi->clk_rate = clk_get_rate(spi->clk);
+	if (!spi->clk_rate) {
+		dev_err(&pdev->dev, "clk rate = 0\n");
+		ret = -EINVAL;
+		goto err_clk_disable;
+	}
+
+	spi->rst = devm_reset_control_get_exclusive(&pdev->dev, NULL);
+	if (!IS_ERR(spi->rst)) {
+		reset_control_assert(spi->rst);
+		udelay(2);
+		reset_control_deassert(spi->rst);
+	}
+
+	spi->fifo_size = stm32_spi_get_fifo_size(spi);
+
+	ret = stm32_spi_config(spi);
+	if (ret) {
+		dev_err(&pdev->dev, "controller configuration failed: %d\n",
+			ret);
+		goto err_clk_disable;
+	}
+
+	master->dev.of_node = pdev->dev.of_node;
+	master->auto_runtime_pm = true;
+	master->bus_num = pdev->id;
+	master->mode_bits = SPI_MODE_3 | SPI_CS_HIGH | SPI_LSB_FIRST |
+			    SPI_3WIRE | SPI_LOOP;
+	master->bits_per_word_mask = stm32_spi_get_bpw_mask(spi);
+	master->max_speed_hz = spi->clk_rate / SPI_MBR_DIV_MIN;
+	master->min_speed_hz = spi->clk_rate / SPI_MBR_DIV_MAX;
+	master->setup = stm32_spi_setup;
+	master->prepare_message = stm32_spi_prepare_msg;
+	master->transfer_one = stm32_spi_transfer_one;
+	master->unprepare_message = stm32_spi_unprepare_msg;
+
+	spi->dma_tx = dma_request_slave_channel(spi->dev, "tx");
+	if (!spi->dma_tx)
+		dev_warn(&pdev->dev, "failed to request tx dma channel\n");
+	else
+		master->dma_tx = spi->dma_tx;
+
+	spi->dma_rx = dma_request_slave_channel(spi->dev, "rx");
+	if (!spi->dma_rx)
+		dev_warn(&pdev->dev, "failed to request rx dma channel\n");
+	else
+		master->dma_rx = spi->dma_rx;
+
+	if (spi->dma_tx || spi->dma_rx)
+		master->can_dma = stm32_spi_can_dma;
+
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret) {
+		dev_err(&pdev->dev, "spi master registration failed: %d\n",
+			ret);
+		goto err_dma_release;
+	}
+
+	if (!master->cs_gpios) {
+		dev_err(&pdev->dev, "no CS gpios available\n");
+		ret = -EINVAL;
+		goto err_dma_release;
+	}
+
+	for (i = 0; i < master->num_chipselect; i++) {
+		if (!gpio_is_valid(master->cs_gpios[i])) {
+			dev_err(&pdev->dev, "%i is not a valid gpio\n",
+				master->cs_gpios[i]);
+			ret = -EINVAL;
+			goto err_dma_release;
+		}
+
+		ret = devm_gpio_request(&pdev->dev, master->cs_gpios[i],
+					DRIVER_NAME);
+		if (ret) {
+			dev_err(&pdev->dev, "can't get CS gpio %i\n",
+				master->cs_gpios[i]);
+			goto err_dma_release;
+		}
+	}
+
+	dev_info(&pdev->dev, "driver initialized\n");
+
+	return 0;
+
+err_dma_release:
+	if (spi->dma_tx)
+		dma_release_channel(spi->dma_tx);
+	if (spi->dma_rx)
+		dma_release_channel(spi->dma_rx);
+
+	pm_runtime_disable(&pdev->dev);
+err_clk_disable:
+	clk_disable_unprepare(spi->clk);
+err_master_put:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int stm32_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct stm32_spi *spi = spi_master_get_devdata(master);
+
+	stm32_spi_disable(spi);
+
+	if (master->dma_tx)
+		dma_release_channel(master->dma_tx);
+	if (master->dma_rx)
+		dma_release_channel(master->dma_rx);
+
+	clk_disable_unprepare(spi->clk);
+
+	pm_runtime_disable(&pdev->dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+static int stm32_spi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct stm32_spi *spi = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(spi->clk);
+
+	return 0;
+}
+
+static int stm32_spi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct stm32_spi *spi = spi_master_get_devdata(master);
+
+	return clk_prepare_enable(spi->clk);
+}
+#endif
+
+#ifdef CONFIG_PM_SLEEP
+static int stm32_spi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	int ret;
+
+	ret = spi_master_suspend(master);
+	if (ret)
+		return ret;
+
+	return pm_runtime_force_suspend(dev);
+}
+
+static int stm32_spi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct stm32_spi *spi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = pm_runtime_force_resume(dev);
+	if (ret)
+		return ret;
+
+	ret = spi_master_resume(master);
+	if (ret)
+		clk_disable_unprepare(spi->clk);
+
+	return ret;
+}
+#endif
+
+static const struct dev_pm_ops stm32_spi_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(stm32_spi_suspend, stm32_spi_resume)
+	SET_RUNTIME_PM_OPS(stm32_spi_runtime_suspend,
+			   stm32_spi_runtime_resume, NULL)
+};
+
+static struct platform_driver stm32_spi_driver = {
+	.probe = stm32_spi_probe,
+	.remove = stm32_spi_remove,
+	.driver = {
+		.name = DRIVER_NAME,
+		.pm = &stm32_spi_pm_ops,
+		.of_match_table = stm32_spi_of_match,
+	},
+};
+
+module_platform_driver(stm32_spi_driver);
+
+MODULE_ALIAS("platform:" DRIVER_NAME);
+MODULE_DESCRIPTION("STMicroelectronics STM32 SPI Controller driver");
+MODULE_AUTHOR("Amelie Delaunay <amelie.delaunay@st.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-sun4i.c b/drivers/spi/spi-sun4i.c
new file mode 100644
index 0000000..4141003
--- /dev/null
+++ b/drivers/spi/spi-sun4i.c
@@ -0,0 +1,558 @@
+/*
+ * Copyright (C) 2012 - 2014 Allwinner Tech
+ * Pan Nan <pannan@allwinnertech.com>
+ *
+ * Copyright (C) 2014 Maxime Ripard
+ * Maxime Ripard <maxime.ripard@free-electrons.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+
+#include <linux/spi/spi.h>
+
+#define SUN4I_FIFO_DEPTH		64
+
+#define SUN4I_RXDATA_REG		0x00
+
+#define SUN4I_TXDATA_REG		0x04
+
+#define SUN4I_CTL_REG			0x08
+#define SUN4I_CTL_ENABLE			BIT(0)
+#define SUN4I_CTL_MASTER			BIT(1)
+#define SUN4I_CTL_CPHA				BIT(2)
+#define SUN4I_CTL_CPOL				BIT(3)
+#define SUN4I_CTL_CS_ACTIVE_LOW			BIT(4)
+#define SUN4I_CTL_LMTF				BIT(6)
+#define SUN4I_CTL_TF_RST			BIT(8)
+#define SUN4I_CTL_RF_RST			BIT(9)
+#define SUN4I_CTL_XCH				BIT(10)
+#define SUN4I_CTL_CS_MASK			0x3000
+#define SUN4I_CTL_CS(cs)			(((cs) << 12) & SUN4I_CTL_CS_MASK)
+#define SUN4I_CTL_DHB				BIT(15)
+#define SUN4I_CTL_CS_MANUAL			BIT(16)
+#define SUN4I_CTL_CS_LEVEL			BIT(17)
+#define SUN4I_CTL_TP				BIT(18)
+
+#define SUN4I_INT_CTL_REG		0x0c
+#define SUN4I_INT_CTL_RF_F34			BIT(4)
+#define SUN4I_INT_CTL_TF_E34			BIT(12)
+#define SUN4I_INT_CTL_TC			BIT(16)
+
+#define SUN4I_INT_STA_REG		0x10
+
+#define SUN4I_DMA_CTL_REG		0x14
+
+#define SUN4I_WAIT_REG			0x18
+
+#define SUN4I_CLK_CTL_REG		0x1c
+#define SUN4I_CLK_CTL_CDR2_MASK			0xff
+#define SUN4I_CLK_CTL_CDR2(div)			((div) & SUN4I_CLK_CTL_CDR2_MASK)
+#define SUN4I_CLK_CTL_CDR1_MASK			0xf
+#define SUN4I_CLK_CTL_CDR1(div)			(((div) & SUN4I_CLK_CTL_CDR1_MASK) << 8)
+#define SUN4I_CLK_CTL_DRS			BIT(12)
+
+#define SUN4I_MAX_XFER_SIZE			0xffffff
+
+#define SUN4I_BURST_CNT_REG		0x20
+#define SUN4I_BURST_CNT(cnt)			((cnt) & SUN4I_MAX_XFER_SIZE)
+
+#define SUN4I_XMIT_CNT_REG		0x24
+#define SUN4I_XMIT_CNT(cnt)			((cnt) & SUN4I_MAX_XFER_SIZE)
+
+
+#define SUN4I_FIFO_STA_REG		0x28
+#define SUN4I_FIFO_STA_RF_CNT_MASK		0x7f
+#define SUN4I_FIFO_STA_RF_CNT_BITS		0
+#define SUN4I_FIFO_STA_TF_CNT_MASK		0x7f
+#define SUN4I_FIFO_STA_TF_CNT_BITS		16
+
+struct sun4i_spi {
+	struct spi_master	*master;
+	void __iomem		*base_addr;
+	struct clk		*hclk;
+	struct clk		*mclk;
+
+	struct completion	done;
+
+	const u8		*tx_buf;
+	u8			*rx_buf;
+	int			len;
+};
+
+static inline u32 sun4i_spi_read(struct sun4i_spi *sspi, u32 reg)
+{
+	return readl(sspi->base_addr + reg);
+}
+
+static inline void sun4i_spi_write(struct sun4i_spi *sspi, u32 reg, u32 value)
+{
+	writel(value, sspi->base_addr + reg);
+}
+
+static inline u32 sun4i_spi_get_tx_fifo_count(struct sun4i_spi *sspi)
+{
+	u32 reg = sun4i_spi_read(sspi, SUN4I_FIFO_STA_REG);
+
+	reg >>= SUN4I_FIFO_STA_TF_CNT_BITS;
+
+	return reg & SUN4I_FIFO_STA_TF_CNT_MASK;
+}
+
+static inline void sun4i_spi_enable_interrupt(struct sun4i_spi *sspi, u32 mask)
+{
+	u32 reg = sun4i_spi_read(sspi, SUN4I_INT_CTL_REG);
+
+	reg |= mask;
+	sun4i_spi_write(sspi, SUN4I_INT_CTL_REG, reg);
+}
+
+static inline void sun4i_spi_disable_interrupt(struct sun4i_spi *sspi, u32 mask)
+{
+	u32 reg = sun4i_spi_read(sspi, SUN4I_INT_CTL_REG);
+
+	reg &= ~mask;
+	sun4i_spi_write(sspi, SUN4I_INT_CTL_REG, reg);
+}
+
+static inline void sun4i_spi_drain_fifo(struct sun4i_spi *sspi, int len)
+{
+	u32 reg, cnt;
+	u8 byte;
+
+	/* See how much data is available */
+	reg = sun4i_spi_read(sspi, SUN4I_FIFO_STA_REG);
+	reg &= SUN4I_FIFO_STA_RF_CNT_MASK;
+	cnt = reg >> SUN4I_FIFO_STA_RF_CNT_BITS;
+
+	if (len > cnt)
+		len = cnt;
+
+	while (len--) {
+		byte = readb(sspi->base_addr + SUN4I_RXDATA_REG);
+		if (sspi->rx_buf)
+			*sspi->rx_buf++ = byte;
+	}
+}
+
+static inline void sun4i_spi_fill_fifo(struct sun4i_spi *sspi, int len)
+{
+	u32 cnt;
+	u8 byte;
+
+	/* See how much data we can fit */
+	cnt = SUN4I_FIFO_DEPTH - sun4i_spi_get_tx_fifo_count(sspi);
+
+	len = min3(len, (int)cnt, sspi->len);
+
+	while (len--) {
+		byte = sspi->tx_buf ? *sspi->tx_buf++ : 0;
+		writeb(byte, sspi->base_addr + SUN4I_TXDATA_REG);
+		sspi->len--;
+	}
+}
+
+static void sun4i_spi_set_cs(struct spi_device *spi, bool enable)
+{
+	struct sun4i_spi *sspi = spi_master_get_devdata(spi->master);
+	u32 reg;
+
+	reg = sun4i_spi_read(sspi, SUN4I_CTL_REG);
+
+	reg &= ~SUN4I_CTL_CS_MASK;
+	reg |= SUN4I_CTL_CS(spi->chip_select);
+
+	/* We want to control the chip select manually */
+	reg |= SUN4I_CTL_CS_MANUAL;
+
+	if (enable)
+		reg |= SUN4I_CTL_CS_LEVEL;
+	else
+		reg &= ~SUN4I_CTL_CS_LEVEL;
+
+	/*
+	 * Even though this looks irrelevant since we are supposed to
+	 * be controlling the chip select manually, this bit also
+	 * controls the levels of the chip select for inactive
+	 * devices.
+	 *
+	 * If we don't set it, the chip select level will go low by
+	 * default when the device is idle, which is not really
+	 * expected in the common case where the chip select is active
+	 * low.
+	 */
+	if (spi->mode & SPI_CS_HIGH)
+		reg &= ~SUN4I_CTL_CS_ACTIVE_LOW;
+	else
+		reg |= SUN4I_CTL_CS_ACTIVE_LOW;
+
+	sun4i_spi_write(sspi, SUN4I_CTL_REG, reg);
+}
+
+static size_t sun4i_spi_max_transfer_size(struct spi_device *spi)
+{
+	return SUN4I_FIFO_DEPTH - 1;
+}
+
+static int sun4i_spi_transfer_one(struct spi_master *master,
+				  struct spi_device *spi,
+				  struct spi_transfer *tfr)
+{
+	struct sun4i_spi *sspi = spi_master_get_devdata(master);
+	unsigned int mclk_rate, div, timeout;
+	unsigned int start, end, tx_time;
+	unsigned int tx_len = 0;
+	int ret = 0;
+	u32 reg;
+
+	/* We don't support transfer larger than the FIFO */
+	if (tfr->len > SUN4I_MAX_XFER_SIZE)
+		return -EMSGSIZE;
+
+	if (tfr->tx_buf && tfr->len >= SUN4I_MAX_XFER_SIZE)
+		return -EMSGSIZE;
+
+	reinit_completion(&sspi->done);
+	sspi->tx_buf = tfr->tx_buf;
+	sspi->rx_buf = tfr->rx_buf;
+	sspi->len = tfr->len;
+
+	/* Clear pending interrupts */
+	sun4i_spi_write(sspi, SUN4I_INT_STA_REG, ~0);
+
+
+	reg = sun4i_spi_read(sspi, SUN4I_CTL_REG);
+
+	/* Reset FIFOs */
+	sun4i_spi_write(sspi, SUN4I_CTL_REG,
+			reg | SUN4I_CTL_RF_RST | SUN4I_CTL_TF_RST);
+
+	/*
+	 * Setup the transfer control register: Chip Select,
+	 * polarities, etc.
+	 */
+	if (spi->mode & SPI_CPOL)
+		reg |= SUN4I_CTL_CPOL;
+	else
+		reg &= ~SUN4I_CTL_CPOL;
+
+	if (spi->mode & SPI_CPHA)
+		reg |= SUN4I_CTL_CPHA;
+	else
+		reg &= ~SUN4I_CTL_CPHA;
+
+	if (spi->mode & SPI_LSB_FIRST)
+		reg |= SUN4I_CTL_LMTF;
+	else
+		reg &= ~SUN4I_CTL_LMTF;
+
+
+	/*
+	 * If it's a TX only transfer, we don't want to fill the RX
+	 * FIFO with bogus data
+	 */
+	if (sspi->rx_buf)
+		reg &= ~SUN4I_CTL_DHB;
+	else
+		reg |= SUN4I_CTL_DHB;
+
+	sun4i_spi_write(sspi, SUN4I_CTL_REG, reg);
+
+	/* Ensure that we have a parent clock fast enough */
+	mclk_rate = clk_get_rate(sspi->mclk);
+	if (mclk_rate < (2 * tfr->speed_hz)) {
+		clk_set_rate(sspi->mclk, 2 * tfr->speed_hz);
+		mclk_rate = clk_get_rate(sspi->mclk);
+	}
+
+	/*
+	 * Setup clock divider.
+	 *
+	 * We have two choices there. Either we can use the clock
+	 * divide rate 1, which is calculated thanks to this formula:
+	 * SPI_CLK = MOD_CLK / (2 ^ (cdr + 1))
+	 * Or we can use CDR2, which is calculated with the formula:
+	 * SPI_CLK = MOD_CLK / (2 * (cdr + 1))
+	 * Wether we use the former or the latter is set through the
+	 * DRS bit.
+	 *
+	 * First try CDR2, and if we can't reach the expected
+	 * frequency, fall back to CDR1.
+	 */
+	div = mclk_rate / (2 * tfr->speed_hz);
+	if (div <= (SUN4I_CLK_CTL_CDR2_MASK + 1)) {
+		if (div > 0)
+			div--;
+
+		reg = SUN4I_CLK_CTL_CDR2(div) | SUN4I_CLK_CTL_DRS;
+	} else {
+		div = ilog2(mclk_rate) - ilog2(tfr->speed_hz);
+		reg = SUN4I_CLK_CTL_CDR1(div);
+	}
+
+	sun4i_spi_write(sspi, SUN4I_CLK_CTL_REG, reg);
+
+	/* Setup the transfer now... */
+	if (sspi->tx_buf)
+		tx_len = tfr->len;
+
+	/* Setup the counters */
+	sun4i_spi_write(sspi, SUN4I_BURST_CNT_REG, SUN4I_BURST_CNT(tfr->len));
+	sun4i_spi_write(sspi, SUN4I_XMIT_CNT_REG, SUN4I_XMIT_CNT(tx_len));
+
+	/*
+	 * Fill the TX FIFO
+	 * Filling the FIFO fully causes timeout for some reason
+	 * at least on spi2 on A10s
+	 */
+	sun4i_spi_fill_fifo(sspi, SUN4I_FIFO_DEPTH - 1);
+
+	/* Enable the interrupts */
+	sun4i_spi_enable_interrupt(sspi, SUN4I_INT_CTL_TC |
+					 SUN4I_INT_CTL_RF_F34);
+	/* Only enable Tx FIFO interrupt if we really need it */
+	if (tx_len > SUN4I_FIFO_DEPTH)
+		sun4i_spi_enable_interrupt(sspi, SUN4I_INT_CTL_TF_E34);
+
+	/* Start the transfer */
+	reg = sun4i_spi_read(sspi, SUN4I_CTL_REG);
+	sun4i_spi_write(sspi, SUN4I_CTL_REG, reg | SUN4I_CTL_XCH);
+
+	tx_time = max(tfr->len * 8 * 2 / (tfr->speed_hz / 1000), 100U);
+	start = jiffies;
+	timeout = wait_for_completion_timeout(&sspi->done,
+					      msecs_to_jiffies(tx_time));
+	end = jiffies;
+	if (!timeout) {
+		dev_warn(&master->dev,
+			 "%s: timeout transferring %u bytes@%iHz for %i(%i)ms",
+			 dev_name(&spi->dev), tfr->len, tfr->speed_hz,
+			 jiffies_to_msecs(end - start), tx_time);
+		ret = -ETIMEDOUT;
+		goto out;
+	}
+
+
+out:
+	sun4i_spi_write(sspi, SUN4I_INT_CTL_REG, 0);
+
+	return ret;
+}
+
+static irqreturn_t sun4i_spi_handler(int irq, void *dev_id)
+{
+	struct sun4i_spi *sspi = dev_id;
+	u32 status = sun4i_spi_read(sspi, SUN4I_INT_STA_REG);
+
+	/* Transfer complete */
+	if (status & SUN4I_INT_CTL_TC) {
+		sun4i_spi_write(sspi, SUN4I_INT_STA_REG, SUN4I_INT_CTL_TC);
+		sun4i_spi_drain_fifo(sspi, SUN4I_FIFO_DEPTH);
+		complete(&sspi->done);
+		return IRQ_HANDLED;
+	}
+
+	/* Receive FIFO 3/4 full */
+	if (status & SUN4I_INT_CTL_RF_F34) {
+		sun4i_spi_drain_fifo(sspi, SUN4I_FIFO_DEPTH);
+		/* Only clear the interrupt _after_ draining the FIFO */
+		sun4i_spi_write(sspi, SUN4I_INT_STA_REG, SUN4I_INT_CTL_RF_F34);
+		return IRQ_HANDLED;
+	}
+
+	/* Transmit FIFO 3/4 empty */
+	if (status & SUN4I_INT_CTL_TF_E34) {
+		sun4i_spi_fill_fifo(sspi, SUN4I_FIFO_DEPTH);
+
+		if (!sspi->len)
+			/* nothing left to transmit */
+			sun4i_spi_disable_interrupt(sspi, SUN4I_INT_CTL_TF_E34);
+
+		/* Only clear the interrupt _after_ re-seeding the FIFO */
+		sun4i_spi_write(sspi, SUN4I_INT_STA_REG, SUN4I_INT_CTL_TF_E34);
+
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_NONE;
+}
+
+static int sun4i_spi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct sun4i_spi *sspi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(sspi->hclk);
+	if (ret) {
+		dev_err(dev, "Couldn't enable AHB clock\n");
+		goto out;
+	}
+
+	ret = clk_prepare_enable(sspi->mclk);
+	if (ret) {
+		dev_err(dev, "Couldn't enable module clock\n");
+		goto err;
+	}
+
+	sun4i_spi_write(sspi, SUN4I_CTL_REG,
+			SUN4I_CTL_ENABLE | SUN4I_CTL_MASTER | SUN4I_CTL_TP);
+
+	return 0;
+
+err:
+	clk_disable_unprepare(sspi->hclk);
+out:
+	return ret;
+}
+
+static int sun4i_spi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct sun4i_spi *sspi = spi_master_get_devdata(master);
+
+	clk_disable_unprepare(sspi->mclk);
+	clk_disable_unprepare(sspi->hclk);
+
+	return 0;
+}
+
+static int sun4i_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct sun4i_spi *sspi;
+	struct resource	*res;
+	int ret = 0, irq;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct sun4i_spi));
+	if (!master) {
+		dev_err(&pdev->dev, "Unable to allocate SPI Master\n");
+		return -ENOMEM;
+	}
+
+	platform_set_drvdata(pdev, master);
+	sspi = spi_master_get_devdata(master);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	sspi->base_addr = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(sspi->base_addr)) {
+		ret = PTR_ERR(sspi->base_addr);
+		goto err_free_master;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "No spi IRQ specified\n");
+		ret = -ENXIO;
+		goto err_free_master;
+	}
+
+	ret = devm_request_irq(&pdev->dev, irq, sun4i_spi_handler,
+			       0, "sun4i-spi", sspi);
+	if (ret) {
+		dev_err(&pdev->dev, "Cannot request IRQ\n");
+		goto err_free_master;
+	}
+
+	sspi->master = master;
+	master->max_speed_hz = 100 * 1000 * 1000;
+	master->min_speed_hz = 3 * 1000;
+	master->set_cs = sun4i_spi_set_cs;
+	master->transfer_one = sun4i_spi_transfer_one;
+	master->num_chipselect = 4;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LSB_FIRST;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->dev.of_node = pdev->dev.of_node;
+	master->auto_runtime_pm = true;
+	master->max_transfer_size = sun4i_spi_max_transfer_size;
+
+	sspi->hclk = devm_clk_get(&pdev->dev, "ahb");
+	if (IS_ERR(sspi->hclk)) {
+		dev_err(&pdev->dev, "Unable to acquire AHB clock\n");
+		ret = PTR_ERR(sspi->hclk);
+		goto err_free_master;
+	}
+
+	sspi->mclk = devm_clk_get(&pdev->dev, "mod");
+	if (IS_ERR(sspi->mclk)) {
+		dev_err(&pdev->dev, "Unable to acquire module clock\n");
+		ret = PTR_ERR(sspi->mclk);
+		goto err_free_master;
+	}
+
+	init_completion(&sspi->done);
+
+	/*
+	 * This wake-up/shutdown pattern is to be able to have the
+	 * device woken up, even if runtime_pm is disabled
+	 */
+	ret = sun4i_spi_runtime_resume(&pdev->dev);
+	if (ret) {
+		dev_err(&pdev->dev, "Couldn't resume the device\n");
+		goto err_free_master;
+	}
+
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+	pm_runtime_idle(&pdev->dev);
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret) {
+		dev_err(&pdev->dev, "cannot register SPI master\n");
+		goto err_pm_disable;
+	}
+
+	return 0;
+
+err_pm_disable:
+	pm_runtime_disable(&pdev->dev);
+	sun4i_spi_runtime_suspend(&pdev->dev);
+err_free_master:
+	spi_master_put(master);
+	return ret;
+}
+
+static int sun4i_spi_remove(struct platform_device *pdev)
+{
+	pm_runtime_force_suspend(&pdev->dev);
+
+	return 0;
+}
+
+static const struct of_device_id sun4i_spi_match[] = {
+	{ .compatible = "allwinner,sun4i-a10-spi", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, sun4i_spi_match);
+
+static const struct dev_pm_ops sun4i_spi_pm_ops = {
+	.runtime_resume		= sun4i_spi_runtime_resume,
+	.runtime_suspend	= sun4i_spi_runtime_suspend,
+};
+
+static struct platform_driver sun4i_spi_driver = {
+	.probe	= sun4i_spi_probe,
+	.remove	= sun4i_spi_remove,
+	.driver	= {
+		.name		= "sun4i-spi",
+		.of_match_table	= sun4i_spi_match,
+		.pm		= &sun4i_spi_pm_ops,
+	},
+};
+module_platform_driver(sun4i_spi_driver);
+
+MODULE_AUTHOR("Pan Nan <pannan@allwinnertech.com>");
+MODULE_AUTHOR("Maxime Ripard <maxime.ripard@free-electrons.com>");
+MODULE_DESCRIPTION("Allwinner A1X/A20 SPI controller driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-sun6i.c b/drivers/spi/spi-sun6i.c
new file mode 100644
index 0000000..8533f4e
--- /dev/null
+++ b/drivers/spi/spi-sun6i.c
@@ -0,0 +1,575 @@
+/*
+ * Copyright (C) 2012 - 2014 Allwinner Tech
+ * Pan Nan <pannan@allwinnertech.com>
+ *
+ * Copyright (C) 2014 Maxime Ripard
+ * Maxime Ripard <maxime.ripard@free-electrons.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/reset.h>
+
+#include <linux/spi/spi.h>
+
+#define SUN6I_FIFO_DEPTH		128
+#define SUN8I_FIFO_DEPTH		64
+
+#define SUN6I_GBL_CTL_REG		0x04
+#define SUN6I_GBL_CTL_BUS_ENABLE		BIT(0)
+#define SUN6I_GBL_CTL_MASTER			BIT(1)
+#define SUN6I_GBL_CTL_TP			BIT(7)
+#define SUN6I_GBL_CTL_RST			BIT(31)
+
+#define SUN6I_TFR_CTL_REG		0x08
+#define SUN6I_TFR_CTL_CPHA			BIT(0)
+#define SUN6I_TFR_CTL_CPOL			BIT(1)
+#define SUN6I_TFR_CTL_SPOL			BIT(2)
+#define SUN6I_TFR_CTL_CS_MASK			0x30
+#define SUN6I_TFR_CTL_CS(cs)			(((cs) << 4) & SUN6I_TFR_CTL_CS_MASK)
+#define SUN6I_TFR_CTL_CS_MANUAL			BIT(6)
+#define SUN6I_TFR_CTL_CS_LEVEL			BIT(7)
+#define SUN6I_TFR_CTL_DHB			BIT(8)
+#define SUN6I_TFR_CTL_FBS			BIT(12)
+#define SUN6I_TFR_CTL_XCH			BIT(31)
+
+#define SUN6I_INT_CTL_REG		0x10
+#define SUN6I_INT_CTL_RF_RDY			BIT(0)
+#define SUN6I_INT_CTL_TF_ERQ			BIT(4)
+#define SUN6I_INT_CTL_RF_OVF			BIT(8)
+#define SUN6I_INT_CTL_TC			BIT(12)
+
+#define SUN6I_INT_STA_REG		0x14
+
+#define SUN6I_FIFO_CTL_REG		0x18
+#define SUN6I_FIFO_CTL_RF_RDY_TRIG_LEVEL_MASK	0xff
+#define SUN6I_FIFO_CTL_RF_RDY_TRIG_LEVEL_BITS	0
+#define SUN6I_FIFO_CTL_RF_RST			BIT(15)
+#define SUN6I_FIFO_CTL_TF_ERQ_TRIG_LEVEL_MASK	0xff
+#define SUN6I_FIFO_CTL_TF_ERQ_TRIG_LEVEL_BITS	16
+#define SUN6I_FIFO_CTL_TF_RST			BIT(31)
+
+#define SUN6I_FIFO_STA_REG		0x1c
+#define SUN6I_FIFO_STA_RF_CNT_MASK		0x7f
+#define SUN6I_FIFO_STA_RF_CNT_BITS		0
+#define SUN6I_FIFO_STA_TF_CNT_MASK		0x7f
+#define SUN6I_FIFO_STA_TF_CNT_BITS		16
+
+#define SUN6I_CLK_CTL_REG		0x24
+#define SUN6I_CLK_CTL_CDR2_MASK			0xff
+#define SUN6I_CLK_CTL_CDR2(div)			(((div) & SUN6I_CLK_CTL_CDR2_MASK) << 0)
+#define SUN6I_CLK_CTL_CDR1_MASK			0xf
+#define SUN6I_CLK_CTL_CDR1(div)			(((div) & SUN6I_CLK_CTL_CDR1_MASK) << 8)
+#define SUN6I_CLK_CTL_DRS			BIT(12)
+
+#define SUN6I_MAX_XFER_SIZE		0xffffff
+
+#define SUN6I_BURST_CNT_REG		0x30
+#define SUN6I_BURST_CNT(cnt)			((cnt) & SUN6I_MAX_XFER_SIZE)
+
+#define SUN6I_XMIT_CNT_REG		0x34
+#define SUN6I_XMIT_CNT(cnt)			((cnt) & SUN6I_MAX_XFER_SIZE)
+
+#define SUN6I_BURST_CTL_CNT_REG		0x38
+#define SUN6I_BURST_CTL_CNT_STC(cnt)		((cnt) & SUN6I_MAX_XFER_SIZE)
+
+#define SUN6I_TXDATA_REG		0x200
+#define SUN6I_RXDATA_REG		0x300
+
+struct sun6i_spi {
+	struct spi_master	*master;
+	void __iomem		*base_addr;
+	struct clk		*hclk;
+	struct clk		*mclk;
+	struct reset_control	*rstc;
+
+	struct completion	done;
+
+	const u8		*tx_buf;
+	u8			*rx_buf;
+	int			len;
+	unsigned long		fifo_depth;
+};
+
+static inline u32 sun6i_spi_read(struct sun6i_spi *sspi, u32 reg)
+{
+	return readl(sspi->base_addr + reg);
+}
+
+static inline void sun6i_spi_write(struct sun6i_spi *sspi, u32 reg, u32 value)
+{
+	writel(value, sspi->base_addr + reg);
+}
+
+static inline u32 sun6i_spi_get_tx_fifo_count(struct sun6i_spi *sspi)
+{
+	u32 reg = sun6i_spi_read(sspi, SUN6I_FIFO_STA_REG);
+
+	reg >>= SUN6I_FIFO_STA_TF_CNT_BITS;
+
+	return reg & SUN6I_FIFO_STA_TF_CNT_MASK;
+}
+
+static inline void sun6i_spi_enable_interrupt(struct sun6i_spi *sspi, u32 mask)
+{
+	u32 reg = sun6i_spi_read(sspi, SUN6I_INT_CTL_REG);
+
+	reg |= mask;
+	sun6i_spi_write(sspi, SUN6I_INT_CTL_REG, reg);
+}
+
+static inline void sun6i_spi_disable_interrupt(struct sun6i_spi *sspi, u32 mask)
+{
+	u32 reg = sun6i_spi_read(sspi, SUN6I_INT_CTL_REG);
+
+	reg &= ~mask;
+	sun6i_spi_write(sspi, SUN6I_INT_CTL_REG, reg);
+}
+
+static inline void sun6i_spi_drain_fifo(struct sun6i_spi *sspi, int len)
+{
+	u32 reg, cnt;
+	u8 byte;
+
+	/* See how much data is available */
+	reg = sun6i_spi_read(sspi, SUN6I_FIFO_STA_REG);
+	reg &= SUN6I_FIFO_STA_RF_CNT_MASK;
+	cnt = reg >> SUN6I_FIFO_STA_RF_CNT_BITS;
+
+	if (len > cnt)
+		len = cnt;
+
+	while (len--) {
+		byte = readb(sspi->base_addr + SUN6I_RXDATA_REG);
+		if (sspi->rx_buf)
+			*sspi->rx_buf++ = byte;
+	}
+}
+
+static inline void sun6i_spi_fill_fifo(struct sun6i_spi *sspi, int len)
+{
+	u32 cnt;
+	u8 byte;
+
+	/* See how much data we can fit */
+	cnt = sspi->fifo_depth - sun6i_spi_get_tx_fifo_count(sspi);
+
+	len = min3(len, (int)cnt, sspi->len);
+
+	while (len--) {
+		byte = sspi->tx_buf ? *sspi->tx_buf++ : 0;
+		writeb(byte, sspi->base_addr + SUN6I_TXDATA_REG);
+		sspi->len--;
+	}
+}
+
+static void sun6i_spi_set_cs(struct spi_device *spi, bool enable)
+{
+	struct sun6i_spi *sspi = spi_master_get_devdata(spi->master);
+	u32 reg;
+
+	reg = sun6i_spi_read(sspi, SUN6I_TFR_CTL_REG);
+	reg &= ~SUN6I_TFR_CTL_CS_MASK;
+	reg |= SUN6I_TFR_CTL_CS(spi->chip_select);
+
+	if (enable)
+		reg |= SUN6I_TFR_CTL_CS_LEVEL;
+	else
+		reg &= ~SUN6I_TFR_CTL_CS_LEVEL;
+
+	sun6i_spi_write(sspi, SUN6I_TFR_CTL_REG, reg);
+}
+
+static size_t sun6i_spi_max_transfer_size(struct spi_device *spi)
+{
+	return SUN6I_MAX_XFER_SIZE - 1;
+}
+
+static int sun6i_spi_transfer_one(struct spi_master *master,
+				  struct spi_device *spi,
+				  struct spi_transfer *tfr)
+{
+	struct sun6i_spi *sspi = spi_master_get_devdata(master);
+	unsigned int mclk_rate, div, timeout;
+	unsigned int start, end, tx_time;
+	unsigned int trig_level;
+	unsigned int tx_len = 0;
+	int ret = 0;
+	u32 reg;
+
+	if (tfr->len > SUN6I_MAX_XFER_SIZE)
+		return -EINVAL;
+
+	reinit_completion(&sspi->done);
+	sspi->tx_buf = tfr->tx_buf;
+	sspi->rx_buf = tfr->rx_buf;
+	sspi->len = tfr->len;
+
+	/* Clear pending interrupts */
+	sun6i_spi_write(sspi, SUN6I_INT_STA_REG, ~0);
+
+	/* Reset FIFO */
+	sun6i_spi_write(sspi, SUN6I_FIFO_CTL_REG,
+			SUN6I_FIFO_CTL_RF_RST | SUN6I_FIFO_CTL_TF_RST);
+
+	/*
+	 * Setup FIFO interrupt trigger level
+	 * Here we choose 3/4 of the full fifo depth, as it's the hardcoded
+	 * value used in old generation of Allwinner SPI controller.
+	 * (See spi-sun4i.c)
+	 */
+	trig_level = sspi->fifo_depth / 4 * 3;
+	sun6i_spi_write(sspi, SUN6I_FIFO_CTL_REG,
+			(trig_level << SUN6I_FIFO_CTL_RF_RDY_TRIG_LEVEL_BITS) |
+			(trig_level << SUN6I_FIFO_CTL_TF_ERQ_TRIG_LEVEL_BITS));
+
+	/*
+	 * Setup the transfer control register: Chip Select,
+	 * polarities, etc.
+	 */
+	reg = sun6i_spi_read(sspi, SUN6I_TFR_CTL_REG);
+
+	if (spi->mode & SPI_CPOL)
+		reg |= SUN6I_TFR_CTL_CPOL;
+	else
+		reg &= ~SUN6I_TFR_CTL_CPOL;
+
+	if (spi->mode & SPI_CPHA)
+		reg |= SUN6I_TFR_CTL_CPHA;
+	else
+		reg &= ~SUN6I_TFR_CTL_CPHA;
+
+	if (spi->mode & SPI_LSB_FIRST)
+		reg |= SUN6I_TFR_CTL_FBS;
+	else
+		reg &= ~SUN6I_TFR_CTL_FBS;
+
+	/*
+	 * If it's a TX only transfer, we don't want to fill the RX
+	 * FIFO with bogus data
+	 */
+	if (sspi->rx_buf)
+		reg &= ~SUN6I_TFR_CTL_DHB;
+	else
+		reg |= SUN6I_TFR_CTL_DHB;
+
+	/* We want to control the chip select manually */
+	reg |= SUN6I_TFR_CTL_CS_MANUAL;
+
+	sun6i_spi_write(sspi, SUN6I_TFR_CTL_REG, reg);
+
+	/* Ensure that we have a parent clock fast enough */
+	mclk_rate = clk_get_rate(sspi->mclk);
+	if (mclk_rate < (2 * tfr->speed_hz)) {
+		clk_set_rate(sspi->mclk, 2 * tfr->speed_hz);
+		mclk_rate = clk_get_rate(sspi->mclk);
+	}
+
+	/*
+	 * Setup clock divider.
+	 *
+	 * We have two choices there. Either we can use the clock
+	 * divide rate 1, which is calculated thanks to this formula:
+	 * SPI_CLK = MOD_CLK / (2 ^ cdr)
+	 * Or we can use CDR2, which is calculated with the formula:
+	 * SPI_CLK = MOD_CLK / (2 * (cdr + 1))
+	 * Wether we use the former or the latter is set through the
+	 * DRS bit.
+	 *
+	 * First try CDR2, and if we can't reach the expected
+	 * frequency, fall back to CDR1.
+	 */
+	div = mclk_rate / (2 * tfr->speed_hz);
+	if (div <= (SUN6I_CLK_CTL_CDR2_MASK + 1)) {
+		if (div > 0)
+			div--;
+
+		reg = SUN6I_CLK_CTL_CDR2(div) | SUN6I_CLK_CTL_DRS;
+	} else {
+		div = ilog2(mclk_rate) - ilog2(tfr->speed_hz);
+		reg = SUN6I_CLK_CTL_CDR1(div);
+	}
+
+	sun6i_spi_write(sspi, SUN6I_CLK_CTL_REG, reg);
+
+	/* Setup the transfer now... */
+	if (sspi->tx_buf)
+		tx_len = tfr->len;
+
+	/* Setup the counters */
+	sun6i_spi_write(sspi, SUN6I_BURST_CNT_REG, SUN6I_BURST_CNT(tfr->len));
+	sun6i_spi_write(sspi, SUN6I_XMIT_CNT_REG, SUN6I_XMIT_CNT(tx_len));
+	sun6i_spi_write(sspi, SUN6I_BURST_CTL_CNT_REG,
+			SUN6I_BURST_CTL_CNT_STC(tx_len));
+
+	/* Fill the TX FIFO */
+	sun6i_spi_fill_fifo(sspi, sspi->fifo_depth);
+
+	/* Enable the interrupts */
+	sun6i_spi_write(sspi, SUN6I_INT_CTL_REG, SUN6I_INT_CTL_TC);
+	sun6i_spi_enable_interrupt(sspi, SUN6I_INT_CTL_TC |
+					 SUN6I_INT_CTL_RF_RDY);
+	if (tx_len > sspi->fifo_depth)
+		sun6i_spi_enable_interrupt(sspi, SUN6I_INT_CTL_TF_ERQ);
+
+	/* Start the transfer */
+	reg = sun6i_spi_read(sspi, SUN6I_TFR_CTL_REG);
+	sun6i_spi_write(sspi, SUN6I_TFR_CTL_REG, reg | SUN6I_TFR_CTL_XCH);
+
+	tx_time = max(tfr->len * 8 * 2 / (tfr->speed_hz / 1000), 100U);
+	start = jiffies;
+	timeout = wait_for_completion_timeout(&sspi->done,
+					      msecs_to_jiffies(tx_time));
+	end = jiffies;
+	if (!timeout) {
+		dev_warn(&master->dev,
+			 "%s: timeout transferring %u bytes@%iHz for %i(%i)ms",
+			 dev_name(&spi->dev), tfr->len, tfr->speed_hz,
+			 jiffies_to_msecs(end - start), tx_time);
+		ret = -ETIMEDOUT;
+		goto out;
+	}
+
+out:
+	sun6i_spi_write(sspi, SUN6I_INT_CTL_REG, 0);
+
+	return ret;
+}
+
+static irqreturn_t sun6i_spi_handler(int irq, void *dev_id)
+{
+	struct sun6i_spi *sspi = dev_id;
+	u32 status = sun6i_spi_read(sspi, SUN6I_INT_STA_REG);
+
+	/* Transfer complete */
+	if (status & SUN6I_INT_CTL_TC) {
+		sun6i_spi_write(sspi, SUN6I_INT_STA_REG, SUN6I_INT_CTL_TC);
+		sun6i_spi_drain_fifo(sspi, sspi->fifo_depth);
+		complete(&sspi->done);
+		return IRQ_HANDLED;
+	}
+
+	/* Receive FIFO 3/4 full */
+	if (status & SUN6I_INT_CTL_RF_RDY) {
+		sun6i_spi_drain_fifo(sspi, SUN6I_FIFO_DEPTH);
+		/* Only clear the interrupt _after_ draining the FIFO */
+		sun6i_spi_write(sspi, SUN6I_INT_STA_REG, SUN6I_INT_CTL_RF_RDY);
+		return IRQ_HANDLED;
+	}
+
+	/* Transmit FIFO 3/4 empty */
+	if (status & SUN6I_INT_CTL_TF_ERQ) {
+		sun6i_spi_fill_fifo(sspi, SUN6I_FIFO_DEPTH);
+
+		if (!sspi->len)
+			/* nothing left to transmit */
+			sun6i_spi_disable_interrupt(sspi, SUN6I_INT_CTL_TF_ERQ);
+
+		/* Only clear the interrupt _after_ re-seeding the FIFO */
+		sun6i_spi_write(sspi, SUN6I_INT_STA_REG, SUN6I_INT_CTL_TF_ERQ);
+
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_NONE;
+}
+
+static int sun6i_spi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct sun6i_spi *sspi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(sspi->hclk);
+	if (ret) {
+		dev_err(dev, "Couldn't enable AHB clock\n");
+		goto out;
+	}
+
+	ret = clk_prepare_enable(sspi->mclk);
+	if (ret) {
+		dev_err(dev, "Couldn't enable module clock\n");
+		goto err;
+	}
+
+	ret = reset_control_deassert(sspi->rstc);
+	if (ret) {
+		dev_err(dev, "Couldn't deassert the device from reset\n");
+		goto err2;
+	}
+
+	sun6i_spi_write(sspi, SUN6I_GBL_CTL_REG,
+			SUN6I_GBL_CTL_BUS_ENABLE | SUN6I_GBL_CTL_MASTER | SUN6I_GBL_CTL_TP);
+
+	return 0;
+
+err2:
+	clk_disable_unprepare(sspi->mclk);
+err:
+	clk_disable_unprepare(sspi->hclk);
+out:
+	return ret;
+}
+
+static int sun6i_spi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct sun6i_spi *sspi = spi_master_get_devdata(master);
+
+	reset_control_assert(sspi->rstc);
+	clk_disable_unprepare(sspi->mclk);
+	clk_disable_unprepare(sspi->hclk);
+
+	return 0;
+}
+
+static int sun6i_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct sun6i_spi *sspi;
+	struct resource	*res;
+	int ret = 0, irq;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct sun6i_spi));
+	if (!master) {
+		dev_err(&pdev->dev, "Unable to allocate SPI Master\n");
+		return -ENOMEM;
+	}
+
+	platform_set_drvdata(pdev, master);
+	sspi = spi_master_get_devdata(master);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	sspi->base_addr = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(sspi->base_addr)) {
+		ret = PTR_ERR(sspi->base_addr);
+		goto err_free_master;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "No spi IRQ specified\n");
+		ret = -ENXIO;
+		goto err_free_master;
+	}
+
+	ret = devm_request_irq(&pdev->dev, irq, sun6i_spi_handler,
+			       0, "sun6i-spi", sspi);
+	if (ret) {
+		dev_err(&pdev->dev, "Cannot request IRQ\n");
+		goto err_free_master;
+	}
+
+	sspi->master = master;
+	sspi->fifo_depth = (unsigned long)of_device_get_match_data(&pdev->dev);
+
+	master->max_speed_hz = 100 * 1000 * 1000;
+	master->min_speed_hz = 3 * 1000;
+	master->set_cs = sun6i_spi_set_cs;
+	master->transfer_one = sun6i_spi_transfer_one;
+	master->num_chipselect = 4;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LSB_FIRST;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->dev.of_node = pdev->dev.of_node;
+	master->auto_runtime_pm = true;
+	master->max_transfer_size = sun6i_spi_max_transfer_size;
+
+	sspi->hclk = devm_clk_get(&pdev->dev, "ahb");
+	if (IS_ERR(sspi->hclk)) {
+		dev_err(&pdev->dev, "Unable to acquire AHB clock\n");
+		ret = PTR_ERR(sspi->hclk);
+		goto err_free_master;
+	}
+
+	sspi->mclk = devm_clk_get(&pdev->dev, "mod");
+	if (IS_ERR(sspi->mclk)) {
+		dev_err(&pdev->dev, "Unable to acquire module clock\n");
+		ret = PTR_ERR(sspi->mclk);
+		goto err_free_master;
+	}
+
+	init_completion(&sspi->done);
+
+	sspi->rstc = devm_reset_control_get_exclusive(&pdev->dev, NULL);
+	if (IS_ERR(sspi->rstc)) {
+		dev_err(&pdev->dev, "Couldn't get reset controller\n");
+		ret = PTR_ERR(sspi->rstc);
+		goto err_free_master;
+	}
+
+	/*
+	 * This wake-up/shutdown pattern is to be able to have the
+	 * device woken up, even if runtime_pm is disabled
+	 */
+	ret = sun6i_spi_runtime_resume(&pdev->dev);
+	if (ret) {
+		dev_err(&pdev->dev, "Couldn't resume the device\n");
+		goto err_free_master;
+	}
+
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+	pm_runtime_idle(&pdev->dev);
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret) {
+		dev_err(&pdev->dev, "cannot register SPI master\n");
+		goto err_pm_disable;
+	}
+
+	return 0;
+
+err_pm_disable:
+	pm_runtime_disable(&pdev->dev);
+	sun6i_spi_runtime_suspend(&pdev->dev);
+err_free_master:
+	spi_master_put(master);
+	return ret;
+}
+
+static int sun6i_spi_remove(struct platform_device *pdev)
+{
+	pm_runtime_force_suspend(&pdev->dev);
+
+	return 0;
+}
+
+static const struct of_device_id sun6i_spi_match[] = {
+	{ .compatible = "allwinner,sun6i-a31-spi", .data = (void *)SUN6I_FIFO_DEPTH },
+	{ .compatible = "allwinner,sun8i-h3-spi",  .data = (void *)SUN8I_FIFO_DEPTH },
+	{}
+};
+MODULE_DEVICE_TABLE(of, sun6i_spi_match);
+
+static const struct dev_pm_ops sun6i_spi_pm_ops = {
+	.runtime_resume		= sun6i_spi_runtime_resume,
+	.runtime_suspend	= sun6i_spi_runtime_suspend,
+};
+
+static struct platform_driver sun6i_spi_driver = {
+	.probe	= sun6i_spi_probe,
+	.remove	= sun6i_spi_remove,
+	.driver	= {
+		.name		= "sun6i-spi",
+		.of_match_table	= sun6i_spi_match,
+		.pm		= &sun6i_spi_pm_ops,
+	},
+};
+module_platform_driver(sun6i_spi_driver);
+
+MODULE_AUTHOR("Pan Nan <pannan@allwinnertech.com>");
+MODULE_AUTHOR("Maxime Ripard <maxime.ripard@free-electrons.com>");
+MODULE_DESCRIPTION("Allwinner A31 SPI controller driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-tegra114.c b/drivers/spi/spi-tegra114.c
new file mode 100644
index 0000000..a76aced
--- /dev/null
+++ b/drivers/spi/spi-tegra114.c
@@ -0,0 +1,1237 @@
+/*
+ * SPI driver for NVIDIA's Tegra114 SPI Controller.
+ *
+ * Copyright (c) 2013, NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmapool.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/kthread.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/reset.h>
+#include <linux/spi/spi.h>
+
+#define SPI_COMMAND1				0x000
+#define SPI_BIT_LENGTH(x)			(((x) & 0x1f) << 0)
+#define SPI_PACKED				(1 << 5)
+#define SPI_TX_EN				(1 << 11)
+#define SPI_RX_EN				(1 << 12)
+#define SPI_BOTH_EN_BYTE			(1 << 13)
+#define SPI_BOTH_EN_BIT				(1 << 14)
+#define SPI_LSBYTE_FE				(1 << 15)
+#define SPI_LSBIT_FE				(1 << 16)
+#define SPI_BIDIROE				(1 << 17)
+#define SPI_IDLE_SDA_DRIVE_LOW			(0 << 18)
+#define SPI_IDLE_SDA_DRIVE_HIGH			(1 << 18)
+#define SPI_IDLE_SDA_PULL_LOW			(2 << 18)
+#define SPI_IDLE_SDA_PULL_HIGH			(3 << 18)
+#define SPI_IDLE_SDA_MASK			(3 << 18)
+#define SPI_CS_SW_VAL				(1 << 20)
+#define SPI_CS_SW_HW				(1 << 21)
+/* SPI_CS_POL_INACTIVE bits are default high */
+						/* n from 0 to 3 */
+#define SPI_CS_POL_INACTIVE(n)			(1 << (22 + (n)))
+#define SPI_CS_POL_INACTIVE_MASK		(0xF << 22)
+
+#define SPI_CS_SEL_0				(0 << 26)
+#define SPI_CS_SEL_1				(1 << 26)
+#define SPI_CS_SEL_2				(2 << 26)
+#define SPI_CS_SEL_3				(3 << 26)
+#define SPI_CS_SEL_MASK				(3 << 26)
+#define SPI_CS_SEL(x)				(((x) & 0x3) << 26)
+#define SPI_CONTROL_MODE_0			(0 << 28)
+#define SPI_CONTROL_MODE_1			(1 << 28)
+#define SPI_CONTROL_MODE_2			(2 << 28)
+#define SPI_CONTROL_MODE_3			(3 << 28)
+#define SPI_CONTROL_MODE_MASK			(3 << 28)
+#define SPI_MODE_SEL(x)				(((x) & 0x3) << 28)
+#define SPI_M_S					(1 << 30)
+#define SPI_PIO					(1 << 31)
+
+#define SPI_COMMAND2				0x004
+#define SPI_TX_TAP_DELAY(x)			(((x) & 0x3F) << 6)
+#define SPI_RX_TAP_DELAY(x)			(((x) & 0x3F) << 0)
+
+#define SPI_CS_TIMING1				0x008
+#define SPI_SETUP_HOLD(setup, hold)		(((setup) << 4) | (hold))
+#define SPI_CS_SETUP_HOLD(reg, cs, val)			\
+		((((val) & 0xFFu) << ((cs) * 8)) |	\
+		((reg) & ~(0xFFu << ((cs) * 8))))
+
+#define SPI_CS_TIMING2				0x00C
+#define CYCLES_BETWEEN_PACKETS_0(x)		(((x) & 0x1F) << 0)
+#define CS_ACTIVE_BETWEEN_PACKETS_0		(1 << 5)
+#define CYCLES_BETWEEN_PACKETS_1(x)		(((x) & 0x1F) << 8)
+#define CS_ACTIVE_BETWEEN_PACKETS_1		(1 << 13)
+#define CYCLES_BETWEEN_PACKETS_2(x)		(((x) & 0x1F) << 16)
+#define CS_ACTIVE_BETWEEN_PACKETS_2		(1 << 21)
+#define CYCLES_BETWEEN_PACKETS_3(x)		(((x) & 0x1F) << 24)
+#define CS_ACTIVE_BETWEEN_PACKETS_3		(1 << 29)
+#define SPI_SET_CS_ACTIVE_BETWEEN_PACKETS(reg, cs, val)		\
+		(reg = (((val) & 0x1) << ((cs) * 8 + 5)) |	\
+			((reg) & ~(1 << ((cs) * 8 + 5))))
+#define SPI_SET_CYCLES_BETWEEN_PACKETS(reg, cs, val)		\
+		(reg = (((val) & 0xF) << ((cs) * 8)) |		\
+			((reg) & ~(0xF << ((cs) * 8))))
+
+#define SPI_TRANS_STATUS			0x010
+#define SPI_BLK_CNT(val)			(((val) >> 0) & 0xFFFF)
+#define SPI_SLV_IDLE_COUNT(val)			(((val) >> 16) & 0xFF)
+#define SPI_RDY					(1 << 30)
+
+#define SPI_FIFO_STATUS				0x014
+#define SPI_RX_FIFO_EMPTY			(1 << 0)
+#define SPI_RX_FIFO_FULL			(1 << 1)
+#define SPI_TX_FIFO_EMPTY			(1 << 2)
+#define SPI_TX_FIFO_FULL			(1 << 3)
+#define SPI_RX_FIFO_UNF				(1 << 4)
+#define SPI_RX_FIFO_OVF				(1 << 5)
+#define SPI_TX_FIFO_UNF				(1 << 6)
+#define SPI_TX_FIFO_OVF				(1 << 7)
+#define SPI_ERR					(1 << 8)
+#define SPI_TX_FIFO_FLUSH			(1 << 14)
+#define SPI_RX_FIFO_FLUSH			(1 << 15)
+#define SPI_TX_FIFO_EMPTY_COUNT(val)		(((val) >> 16) & 0x7F)
+#define SPI_RX_FIFO_FULL_COUNT(val)		(((val) >> 23) & 0x7F)
+#define SPI_FRAME_END				(1 << 30)
+#define SPI_CS_INACTIVE				(1 << 31)
+
+#define SPI_FIFO_ERROR				(SPI_RX_FIFO_UNF | \
+			SPI_RX_FIFO_OVF | SPI_TX_FIFO_UNF | SPI_TX_FIFO_OVF)
+#define SPI_FIFO_EMPTY			(SPI_RX_FIFO_EMPTY | SPI_TX_FIFO_EMPTY)
+
+#define SPI_TX_DATA				0x018
+#define SPI_RX_DATA				0x01C
+
+#define SPI_DMA_CTL				0x020
+#define SPI_TX_TRIG_1				(0 << 15)
+#define SPI_TX_TRIG_4				(1 << 15)
+#define SPI_TX_TRIG_8				(2 << 15)
+#define SPI_TX_TRIG_16				(3 << 15)
+#define SPI_TX_TRIG_MASK			(3 << 15)
+#define SPI_RX_TRIG_1				(0 << 19)
+#define SPI_RX_TRIG_4				(1 << 19)
+#define SPI_RX_TRIG_8				(2 << 19)
+#define SPI_RX_TRIG_16				(3 << 19)
+#define SPI_RX_TRIG_MASK			(3 << 19)
+#define SPI_IE_TX				(1 << 28)
+#define SPI_IE_RX				(1 << 29)
+#define SPI_CONT				(1 << 30)
+#define SPI_DMA					(1 << 31)
+#define SPI_DMA_EN				SPI_DMA
+
+#define SPI_DMA_BLK				0x024
+#define SPI_DMA_BLK_SET(x)			(((x) & 0xFFFF) << 0)
+
+#define SPI_TX_FIFO				0x108
+#define SPI_RX_FIFO				0x188
+#define MAX_CHIP_SELECT				4
+#define SPI_FIFO_DEPTH				64
+#define DATA_DIR_TX				(1 << 0)
+#define DATA_DIR_RX				(1 << 1)
+
+#define SPI_DMA_TIMEOUT				(msecs_to_jiffies(1000))
+#define DEFAULT_SPI_DMA_BUF_LEN			(16*1024)
+#define TX_FIFO_EMPTY_COUNT_MAX			SPI_TX_FIFO_EMPTY_COUNT(0x40)
+#define RX_FIFO_FULL_COUNT_ZERO			SPI_RX_FIFO_FULL_COUNT(0)
+#define MAX_HOLD_CYCLES				16
+#define SPI_DEFAULT_SPEED			25000000
+
+struct tegra_spi_data {
+	struct device				*dev;
+	struct spi_master			*master;
+	spinlock_t				lock;
+
+	struct clk				*clk;
+	struct reset_control			*rst;
+	void __iomem				*base;
+	phys_addr_t				phys;
+	unsigned				irq;
+	u32					cur_speed;
+
+	struct spi_device			*cur_spi;
+	struct spi_device			*cs_control;
+	unsigned				cur_pos;
+	unsigned				words_per_32bit;
+	unsigned				bytes_per_word;
+	unsigned				curr_dma_words;
+	unsigned				cur_direction;
+
+	unsigned				cur_rx_pos;
+	unsigned				cur_tx_pos;
+
+	unsigned				dma_buf_size;
+	unsigned				max_buf_size;
+	bool					is_curr_dma_xfer;
+
+	struct completion			rx_dma_complete;
+	struct completion			tx_dma_complete;
+
+	u32					tx_status;
+	u32					rx_status;
+	u32					status_reg;
+	bool					is_packed;
+
+	u32					command1_reg;
+	u32					dma_control_reg;
+	u32					def_command1_reg;
+
+	struct completion			xfer_completion;
+	struct spi_transfer			*curr_xfer;
+	struct dma_chan				*rx_dma_chan;
+	u32					*rx_dma_buf;
+	dma_addr_t				rx_dma_phys;
+	struct dma_async_tx_descriptor		*rx_dma_desc;
+
+	struct dma_chan				*tx_dma_chan;
+	u32					*tx_dma_buf;
+	dma_addr_t				tx_dma_phys;
+	struct dma_async_tx_descriptor		*tx_dma_desc;
+};
+
+static int tegra_spi_runtime_suspend(struct device *dev);
+static int tegra_spi_runtime_resume(struct device *dev);
+
+static inline u32 tegra_spi_readl(struct tegra_spi_data *tspi,
+		unsigned long reg)
+{
+	return readl(tspi->base + reg);
+}
+
+static inline void tegra_spi_writel(struct tegra_spi_data *tspi,
+		u32 val, unsigned long reg)
+{
+	writel(val, tspi->base + reg);
+
+	/* Read back register to make sure that register writes completed */
+	if (reg != SPI_TX_FIFO)
+		readl(tspi->base + SPI_COMMAND1);
+}
+
+static void tegra_spi_clear_status(struct tegra_spi_data *tspi)
+{
+	u32 val;
+
+	/* Write 1 to clear status register */
+	val = tegra_spi_readl(tspi, SPI_TRANS_STATUS);
+	tegra_spi_writel(tspi, val, SPI_TRANS_STATUS);
+
+	/* Clear fifo status error if any */
+	val = tegra_spi_readl(tspi, SPI_FIFO_STATUS);
+	if (val & SPI_ERR)
+		tegra_spi_writel(tspi, SPI_ERR | SPI_FIFO_ERROR,
+				SPI_FIFO_STATUS);
+}
+
+static unsigned tegra_spi_calculate_curr_xfer_param(
+	struct spi_device *spi, struct tegra_spi_data *tspi,
+	struct spi_transfer *t)
+{
+	unsigned remain_len = t->len - tspi->cur_pos;
+	unsigned max_word;
+	unsigned bits_per_word = t->bits_per_word;
+	unsigned max_len;
+	unsigned total_fifo_words;
+
+	tspi->bytes_per_word = DIV_ROUND_UP(bits_per_word, 8);
+
+	if (bits_per_word == 8 || bits_per_word == 16) {
+		tspi->is_packed = 1;
+		tspi->words_per_32bit = 32/bits_per_word;
+	} else {
+		tspi->is_packed = 0;
+		tspi->words_per_32bit = 1;
+	}
+
+	if (tspi->is_packed) {
+		max_len = min(remain_len, tspi->max_buf_size);
+		tspi->curr_dma_words = max_len/tspi->bytes_per_word;
+		total_fifo_words = (max_len + 3) / 4;
+	} else {
+		max_word = (remain_len - 1) / tspi->bytes_per_word + 1;
+		max_word = min(max_word, tspi->max_buf_size/4);
+		tspi->curr_dma_words = max_word;
+		total_fifo_words = max_word;
+	}
+	return total_fifo_words;
+}
+
+static unsigned tegra_spi_fill_tx_fifo_from_client_txbuf(
+	struct tegra_spi_data *tspi, struct spi_transfer *t)
+{
+	unsigned nbytes;
+	unsigned tx_empty_count;
+	u32 fifo_status;
+	unsigned max_n_32bit;
+	unsigned i, count;
+	unsigned int written_words;
+	unsigned fifo_words_left;
+	u8 *tx_buf = (u8 *)t->tx_buf + tspi->cur_tx_pos;
+
+	fifo_status = tegra_spi_readl(tspi, SPI_FIFO_STATUS);
+	tx_empty_count = SPI_TX_FIFO_EMPTY_COUNT(fifo_status);
+
+	if (tspi->is_packed) {
+		fifo_words_left = tx_empty_count * tspi->words_per_32bit;
+		written_words = min(fifo_words_left, tspi->curr_dma_words);
+		nbytes = written_words * tspi->bytes_per_word;
+		max_n_32bit = DIV_ROUND_UP(nbytes, 4);
+		for (count = 0; count < max_n_32bit; count++) {
+			u32 x = 0;
+
+			for (i = 0; (i < 4) && nbytes; i++, nbytes--)
+				x |= (u32)(*tx_buf++) << (i * 8);
+			tegra_spi_writel(tspi, x, SPI_TX_FIFO);
+		}
+	} else {
+		max_n_32bit = min(tspi->curr_dma_words,  tx_empty_count);
+		written_words = max_n_32bit;
+		nbytes = written_words * tspi->bytes_per_word;
+		for (count = 0; count < max_n_32bit; count++) {
+			u32 x = 0;
+
+			for (i = 0; nbytes && (i < tspi->bytes_per_word);
+							i++, nbytes--)
+				x |= (u32)(*tx_buf++) << (i * 8);
+			tegra_spi_writel(tspi, x, SPI_TX_FIFO);
+		}
+	}
+	tspi->cur_tx_pos += written_words * tspi->bytes_per_word;
+	return written_words;
+}
+
+static unsigned int tegra_spi_read_rx_fifo_to_client_rxbuf(
+		struct tegra_spi_data *tspi, struct spi_transfer *t)
+{
+	unsigned rx_full_count;
+	u32 fifo_status;
+	unsigned i, count;
+	unsigned int read_words = 0;
+	unsigned len;
+	u8 *rx_buf = (u8 *)t->rx_buf + tspi->cur_rx_pos;
+
+	fifo_status = tegra_spi_readl(tspi, SPI_FIFO_STATUS);
+	rx_full_count = SPI_RX_FIFO_FULL_COUNT(fifo_status);
+	if (tspi->is_packed) {
+		len = tspi->curr_dma_words * tspi->bytes_per_word;
+		for (count = 0; count < rx_full_count; count++) {
+			u32 x = tegra_spi_readl(tspi, SPI_RX_FIFO);
+
+			for (i = 0; len && (i < 4); i++, len--)
+				*rx_buf++ = (x >> i*8) & 0xFF;
+		}
+		tspi->cur_rx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
+		read_words += tspi->curr_dma_words;
+	} else {
+		u32 rx_mask = ((u32)1 << t->bits_per_word) - 1;
+
+		for (count = 0; count < rx_full_count; count++) {
+			u32 x = tegra_spi_readl(tspi, SPI_RX_FIFO) & rx_mask;
+
+			for (i = 0; (i < tspi->bytes_per_word); i++)
+				*rx_buf++ = (x >> (i*8)) & 0xFF;
+		}
+		tspi->cur_rx_pos += rx_full_count * tspi->bytes_per_word;
+		read_words += rx_full_count;
+	}
+	return read_words;
+}
+
+static void tegra_spi_copy_client_txbuf_to_spi_txbuf(
+		struct tegra_spi_data *tspi, struct spi_transfer *t)
+{
+	/* Make the dma buffer to read by cpu */
+	dma_sync_single_for_cpu(tspi->dev, tspi->tx_dma_phys,
+				tspi->dma_buf_size, DMA_TO_DEVICE);
+
+	if (tspi->is_packed) {
+		unsigned len = tspi->curr_dma_words * tspi->bytes_per_word;
+
+		memcpy(tspi->tx_dma_buf, t->tx_buf + tspi->cur_pos, len);
+	} else {
+		unsigned int i;
+		unsigned int count;
+		u8 *tx_buf = (u8 *)t->tx_buf + tspi->cur_tx_pos;
+		unsigned consume = tspi->curr_dma_words * tspi->bytes_per_word;
+
+		for (count = 0; count < tspi->curr_dma_words; count++) {
+			u32 x = 0;
+
+			for (i = 0; consume && (i < tspi->bytes_per_word);
+							i++, consume--)
+				x |= (u32)(*tx_buf++) << (i * 8);
+			tspi->tx_dma_buf[count] = x;
+		}
+	}
+	tspi->cur_tx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
+
+	/* Make the dma buffer to read by dma */
+	dma_sync_single_for_device(tspi->dev, tspi->tx_dma_phys,
+				tspi->dma_buf_size, DMA_TO_DEVICE);
+}
+
+static void tegra_spi_copy_spi_rxbuf_to_client_rxbuf(
+		struct tegra_spi_data *tspi, struct spi_transfer *t)
+{
+	/* Make the dma buffer to read by cpu */
+	dma_sync_single_for_cpu(tspi->dev, tspi->rx_dma_phys,
+		tspi->dma_buf_size, DMA_FROM_DEVICE);
+
+	if (tspi->is_packed) {
+		unsigned len = tspi->curr_dma_words * tspi->bytes_per_word;
+
+		memcpy(t->rx_buf + tspi->cur_rx_pos, tspi->rx_dma_buf, len);
+	} else {
+		unsigned int i;
+		unsigned int count;
+		unsigned char *rx_buf = t->rx_buf + tspi->cur_rx_pos;
+		u32 rx_mask = ((u32)1 << t->bits_per_word) - 1;
+
+		for (count = 0; count < tspi->curr_dma_words; count++) {
+			u32 x = tspi->rx_dma_buf[count] & rx_mask;
+
+			for (i = 0; (i < tspi->bytes_per_word); i++)
+				*rx_buf++ = (x >> (i*8)) & 0xFF;
+		}
+	}
+	tspi->cur_rx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
+
+	/* Make the dma buffer to read by dma */
+	dma_sync_single_for_device(tspi->dev, tspi->rx_dma_phys,
+		tspi->dma_buf_size, DMA_FROM_DEVICE);
+}
+
+static void tegra_spi_dma_complete(void *args)
+{
+	struct completion *dma_complete = args;
+
+	complete(dma_complete);
+}
+
+static int tegra_spi_start_tx_dma(struct tegra_spi_data *tspi, int len)
+{
+	reinit_completion(&tspi->tx_dma_complete);
+	tspi->tx_dma_desc = dmaengine_prep_slave_single(tspi->tx_dma_chan,
+				tspi->tx_dma_phys, len, DMA_MEM_TO_DEV,
+				DMA_PREP_INTERRUPT |  DMA_CTRL_ACK);
+	if (!tspi->tx_dma_desc) {
+		dev_err(tspi->dev, "Not able to get desc for Tx\n");
+		return -EIO;
+	}
+
+	tspi->tx_dma_desc->callback = tegra_spi_dma_complete;
+	tspi->tx_dma_desc->callback_param = &tspi->tx_dma_complete;
+
+	dmaengine_submit(tspi->tx_dma_desc);
+	dma_async_issue_pending(tspi->tx_dma_chan);
+	return 0;
+}
+
+static int tegra_spi_start_rx_dma(struct tegra_spi_data *tspi, int len)
+{
+	reinit_completion(&tspi->rx_dma_complete);
+	tspi->rx_dma_desc = dmaengine_prep_slave_single(tspi->rx_dma_chan,
+				tspi->rx_dma_phys, len, DMA_DEV_TO_MEM,
+				DMA_PREP_INTERRUPT |  DMA_CTRL_ACK);
+	if (!tspi->rx_dma_desc) {
+		dev_err(tspi->dev, "Not able to get desc for Rx\n");
+		return -EIO;
+	}
+
+	tspi->rx_dma_desc->callback = tegra_spi_dma_complete;
+	tspi->rx_dma_desc->callback_param = &tspi->rx_dma_complete;
+
+	dmaengine_submit(tspi->rx_dma_desc);
+	dma_async_issue_pending(tspi->rx_dma_chan);
+	return 0;
+}
+
+static int tegra_spi_start_dma_based_transfer(
+		struct tegra_spi_data *tspi, struct spi_transfer *t)
+{
+	u32 val;
+	unsigned int len;
+	int ret = 0;
+	u32 status;
+
+	/* Make sure that Rx and Tx fifo are empty */
+	status = tegra_spi_readl(tspi, SPI_FIFO_STATUS);
+	if ((status & SPI_FIFO_EMPTY) != SPI_FIFO_EMPTY) {
+		dev_err(tspi->dev, "Rx/Tx fifo are not empty status 0x%08x\n",
+			(unsigned)status);
+		return -EIO;
+	}
+
+	val = SPI_DMA_BLK_SET(tspi->curr_dma_words - 1);
+	tegra_spi_writel(tspi, val, SPI_DMA_BLK);
+
+	if (tspi->is_packed)
+		len = DIV_ROUND_UP(tspi->curr_dma_words * tspi->bytes_per_word,
+					4) * 4;
+	else
+		len = tspi->curr_dma_words * 4;
+
+	/* Set attention level based on length of transfer */
+	if (len & 0xF)
+		val |= SPI_TX_TRIG_1 | SPI_RX_TRIG_1;
+	else if (((len) >> 4) & 0x1)
+		val |= SPI_TX_TRIG_4 | SPI_RX_TRIG_4;
+	else
+		val |= SPI_TX_TRIG_8 | SPI_RX_TRIG_8;
+
+	if (tspi->cur_direction & DATA_DIR_TX)
+		val |= SPI_IE_TX;
+
+	if (tspi->cur_direction & DATA_DIR_RX)
+		val |= SPI_IE_RX;
+
+	tegra_spi_writel(tspi, val, SPI_DMA_CTL);
+	tspi->dma_control_reg = val;
+
+	if (tspi->cur_direction & DATA_DIR_TX) {
+		tegra_spi_copy_client_txbuf_to_spi_txbuf(tspi, t);
+		ret = tegra_spi_start_tx_dma(tspi, len);
+		if (ret < 0) {
+			dev_err(tspi->dev,
+				"Starting tx dma failed, err %d\n", ret);
+			return ret;
+		}
+	}
+
+	if (tspi->cur_direction & DATA_DIR_RX) {
+		/* Make the dma buffer to read by dma */
+		dma_sync_single_for_device(tspi->dev, tspi->rx_dma_phys,
+				tspi->dma_buf_size, DMA_FROM_DEVICE);
+
+		ret = tegra_spi_start_rx_dma(tspi, len);
+		if (ret < 0) {
+			dev_err(tspi->dev,
+				"Starting rx dma failed, err %d\n", ret);
+			if (tspi->cur_direction & DATA_DIR_TX)
+				dmaengine_terminate_all(tspi->tx_dma_chan);
+			return ret;
+		}
+	}
+	tspi->is_curr_dma_xfer = true;
+	tspi->dma_control_reg = val;
+
+	val |= SPI_DMA_EN;
+	tegra_spi_writel(tspi, val, SPI_DMA_CTL);
+	return ret;
+}
+
+static int tegra_spi_start_cpu_based_transfer(
+		struct tegra_spi_data *tspi, struct spi_transfer *t)
+{
+	u32 val;
+	unsigned cur_words;
+
+	if (tspi->cur_direction & DATA_DIR_TX)
+		cur_words = tegra_spi_fill_tx_fifo_from_client_txbuf(tspi, t);
+	else
+		cur_words = tspi->curr_dma_words;
+
+	val = SPI_DMA_BLK_SET(cur_words - 1);
+	tegra_spi_writel(tspi, val, SPI_DMA_BLK);
+
+	val = 0;
+	if (tspi->cur_direction & DATA_DIR_TX)
+		val |= SPI_IE_TX;
+
+	if (tspi->cur_direction & DATA_DIR_RX)
+		val |= SPI_IE_RX;
+
+	tegra_spi_writel(tspi, val, SPI_DMA_CTL);
+	tspi->dma_control_reg = val;
+
+	tspi->is_curr_dma_xfer = false;
+
+	val |= SPI_DMA_EN;
+	tegra_spi_writel(tspi, val, SPI_DMA_CTL);
+	return 0;
+}
+
+static int tegra_spi_init_dma_param(struct tegra_spi_data *tspi,
+			bool dma_to_memory)
+{
+	struct dma_chan *dma_chan;
+	u32 *dma_buf;
+	dma_addr_t dma_phys;
+	int ret;
+	struct dma_slave_config dma_sconfig;
+
+	dma_chan = dma_request_slave_channel_reason(tspi->dev,
+					dma_to_memory ? "rx" : "tx");
+	if (IS_ERR(dma_chan)) {
+		ret = PTR_ERR(dma_chan);
+		if (ret != -EPROBE_DEFER)
+			dev_err(tspi->dev,
+				"Dma channel is not available: %d\n", ret);
+		return ret;
+	}
+
+	dma_buf = dma_alloc_coherent(tspi->dev, tspi->dma_buf_size,
+				&dma_phys, GFP_KERNEL);
+	if (!dma_buf) {
+		dev_err(tspi->dev, " Not able to allocate the dma buffer\n");
+		dma_release_channel(dma_chan);
+		return -ENOMEM;
+	}
+
+	if (dma_to_memory) {
+		dma_sconfig.src_addr = tspi->phys + SPI_RX_FIFO;
+		dma_sconfig.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		dma_sconfig.src_maxburst = 0;
+	} else {
+		dma_sconfig.dst_addr = tspi->phys + SPI_TX_FIFO;
+		dma_sconfig.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		dma_sconfig.dst_maxburst = 0;
+	}
+
+	ret = dmaengine_slave_config(dma_chan, &dma_sconfig);
+	if (ret)
+		goto scrub;
+	if (dma_to_memory) {
+		tspi->rx_dma_chan = dma_chan;
+		tspi->rx_dma_buf = dma_buf;
+		tspi->rx_dma_phys = dma_phys;
+	} else {
+		tspi->tx_dma_chan = dma_chan;
+		tspi->tx_dma_buf = dma_buf;
+		tspi->tx_dma_phys = dma_phys;
+	}
+	return 0;
+
+scrub:
+	dma_free_coherent(tspi->dev, tspi->dma_buf_size, dma_buf, dma_phys);
+	dma_release_channel(dma_chan);
+	return ret;
+}
+
+static void tegra_spi_deinit_dma_param(struct tegra_spi_data *tspi,
+	bool dma_to_memory)
+{
+	u32 *dma_buf;
+	dma_addr_t dma_phys;
+	struct dma_chan *dma_chan;
+
+	if (dma_to_memory) {
+		dma_buf = tspi->rx_dma_buf;
+		dma_chan = tspi->rx_dma_chan;
+		dma_phys = tspi->rx_dma_phys;
+		tspi->rx_dma_chan = NULL;
+		tspi->rx_dma_buf = NULL;
+	} else {
+		dma_buf = tspi->tx_dma_buf;
+		dma_chan = tspi->tx_dma_chan;
+		dma_phys = tspi->tx_dma_phys;
+		tspi->tx_dma_buf = NULL;
+		tspi->tx_dma_chan = NULL;
+	}
+	if (!dma_chan)
+		return;
+
+	dma_free_coherent(tspi->dev, tspi->dma_buf_size, dma_buf, dma_phys);
+	dma_release_channel(dma_chan);
+}
+
+static u32 tegra_spi_setup_transfer_one(struct spi_device *spi,
+		struct spi_transfer *t, bool is_first_of_msg)
+{
+	struct tegra_spi_data *tspi = spi_master_get_devdata(spi->master);
+	u32 speed = t->speed_hz;
+	u8 bits_per_word = t->bits_per_word;
+	u32 command1;
+	int req_mode;
+
+	if (speed != tspi->cur_speed) {
+		clk_set_rate(tspi->clk, speed);
+		tspi->cur_speed = speed;
+	}
+
+	tspi->cur_spi = spi;
+	tspi->cur_pos = 0;
+	tspi->cur_rx_pos = 0;
+	tspi->cur_tx_pos = 0;
+	tspi->curr_xfer = t;
+
+	if (is_first_of_msg) {
+		tegra_spi_clear_status(tspi);
+
+		command1 = tspi->def_command1_reg;
+		command1 |= SPI_BIT_LENGTH(bits_per_word - 1);
+
+		command1 &= ~SPI_CONTROL_MODE_MASK;
+		req_mode = spi->mode & 0x3;
+		if (req_mode == SPI_MODE_0)
+			command1 |= SPI_CONTROL_MODE_0;
+		else if (req_mode == SPI_MODE_1)
+			command1 |= SPI_CONTROL_MODE_1;
+		else if (req_mode == SPI_MODE_2)
+			command1 |= SPI_CONTROL_MODE_2;
+		else if (req_mode == SPI_MODE_3)
+			command1 |= SPI_CONTROL_MODE_3;
+
+		if (tspi->cs_control) {
+			if (tspi->cs_control != spi)
+				tegra_spi_writel(tspi, command1, SPI_COMMAND1);
+			tspi->cs_control = NULL;
+		} else
+			tegra_spi_writel(tspi, command1, SPI_COMMAND1);
+
+		command1 |= SPI_CS_SW_HW;
+		if (spi->mode & SPI_CS_HIGH)
+			command1 |= SPI_CS_SW_VAL;
+		else
+			command1 &= ~SPI_CS_SW_VAL;
+
+		tegra_spi_writel(tspi, 0, SPI_COMMAND2);
+	} else {
+		command1 = tspi->command1_reg;
+		command1 &= ~SPI_BIT_LENGTH(~0);
+		command1 |= SPI_BIT_LENGTH(bits_per_word - 1);
+	}
+
+	return command1;
+}
+
+static int tegra_spi_start_transfer_one(struct spi_device *spi,
+		struct spi_transfer *t, u32 command1)
+{
+	struct tegra_spi_data *tspi = spi_master_get_devdata(spi->master);
+	unsigned total_fifo_words;
+	int ret;
+
+	total_fifo_words = tegra_spi_calculate_curr_xfer_param(spi, tspi, t);
+
+	if (tspi->is_packed)
+		command1 |= SPI_PACKED;
+
+	command1 &= ~(SPI_CS_SEL_MASK | SPI_TX_EN | SPI_RX_EN);
+	tspi->cur_direction = 0;
+	if (t->rx_buf) {
+		command1 |= SPI_RX_EN;
+		tspi->cur_direction |= DATA_DIR_RX;
+	}
+	if (t->tx_buf) {
+		command1 |= SPI_TX_EN;
+		tspi->cur_direction |= DATA_DIR_TX;
+	}
+	command1 |= SPI_CS_SEL(spi->chip_select);
+	tegra_spi_writel(tspi, command1, SPI_COMMAND1);
+	tspi->command1_reg = command1;
+
+	dev_dbg(tspi->dev, "The def 0x%x and written 0x%x\n",
+		tspi->def_command1_reg, (unsigned)command1);
+
+	if (total_fifo_words > SPI_FIFO_DEPTH)
+		ret = tegra_spi_start_dma_based_transfer(tspi, t);
+	else
+		ret = tegra_spi_start_cpu_based_transfer(tspi, t);
+	return ret;
+}
+
+static int tegra_spi_setup(struct spi_device *spi)
+{
+	struct tegra_spi_data *tspi = spi_master_get_devdata(spi->master);
+	u32 val;
+	unsigned long flags;
+	int ret;
+
+	dev_dbg(&spi->dev, "setup %d bpw, %scpol, %scpha, %dHz\n",
+		spi->bits_per_word,
+		spi->mode & SPI_CPOL ? "" : "~",
+		spi->mode & SPI_CPHA ? "" : "~",
+		spi->max_speed_hz);
+
+	ret = pm_runtime_get_sync(tspi->dev);
+	if (ret < 0) {
+		dev_err(tspi->dev, "pm runtime failed, e = %d\n", ret);
+		return ret;
+	}
+
+	spin_lock_irqsave(&tspi->lock, flags);
+	val = tspi->def_command1_reg;
+	if (spi->mode & SPI_CS_HIGH)
+		val &= ~SPI_CS_POL_INACTIVE(spi->chip_select);
+	else
+		val |= SPI_CS_POL_INACTIVE(spi->chip_select);
+	tspi->def_command1_reg = val;
+	tegra_spi_writel(tspi, tspi->def_command1_reg, SPI_COMMAND1);
+	spin_unlock_irqrestore(&tspi->lock, flags);
+
+	pm_runtime_put(tspi->dev);
+	return 0;
+}
+
+static void tegra_spi_transfer_delay(int delay)
+{
+	if (!delay)
+		return;
+
+	if (delay >= 1000)
+		mdelay(delay / 1000);
+
+	udelay(delay % 1000);
+}
+
+static int tegra_spi_transfer_one_message(struct spi_master *master,
+			struct spi_message *msg)
+{
+	bool is_first_msg = true;
+	struct tegra_spi_data *tspi = spi_master_get_devdata(master);
+	struct spi_transfer *xfer;
+	struct spi_device *spi = msg->spi;
+	int ret;
+	bool skip = false;
+
+	msg->status = 0;
+	msg->actual_length = 0;
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		u32 cmd1;
+
+		reinit_completion(&tspi->xfer_completion);
+
+		cmd1 = tegra_spi_setup_transfer_one(spi, xfer, is_first_msg);
+
+		if (!xfer->len) {
+			ret = 0;
+			skip = true;
+			goto complete_xfer;
+		}
+
+		ret = tegra_spi_start_transfer_one(spi, xfer, cmd1);
+		if (ret < 0) {
+			dev_err(tspi->dev,
+				"spi can not start transfer, err %d\n", ret);
+			goto complete_xfer;
+		}
+
+		is_first_msg = false;
+		ret = wait_for_completion_timeout(&tspi->xfer_completion,
+						SPI_DMA_TIMEOUT);
+		if (WARN_ON(ret == 0)) {
+			dev_err(tspi->dev,
+				"spi transfer timeout, err %d\n", ret);
+			ret = -EIO;
+			goto complete_xfer;
+		}
+
+		if (tspi->tx_status ||  tspi->rx_status) {
+			dev_err(tspi->dev, "Error in Transfer\n");
+			ret = -EIO;
+			goto complete_xfer;
+		}
+		msg->actual_length += xfer->len;
+
+complete_xfer:
+		if (ret < 0 || skip) {
+			tegra_spi_writel(tspi, tspi->def_command1_reg,
+					SPI_COMMAND1);
+			tegra_spi_transfer_delay(xfer->delay_usecs);
+			goto exit;
+		} else if (list_is_last(&xfer->transfer_list,
+					&msg->transfers)) {
+			if (xfer->cs_change)
+				tspi->cs_control = spi;
+			else {
+				tegra_spi_writel(tspi, tspi->def_command1_reg,
+						SPI_COMMAND1);
+				tegra_spi_transfer_delay(xfer->delay_usecs);
+			}
+		} else if (xfer->cs_change) {
+			tegra_spi_writel(tspi, tspi->def_command1_reg,
+					SPI_COMMAND1);
+			tegra_spi_transfer_delay(xfer->delay_usecs);
+		}
+
+	}
+	ret = 0;
+exit:
+	msg->status = ret;
+	spi_finalize_current_message(master);
+	return ret;
+}
+
+static irqreturn_t handle_cpu_based_xfer(struct tegra_spi_data *tspi)
+{
+	struct spi_transfer *t = tspi->curr_xfer;
+	unsigned long flags;
+
+	spin_lock_irqsave(&tspi->lock, flags);
+	if (tspi->tx_status ||  tspi->rx_status) {
+		dev_err(tspi->dev, "CpuXfer ERROR bit set 0x%x\n",
+			tspi->status_reg);
+		dev_err(tspi->dev, "CpuXfer 0x%08x:0x%08x\n",
+			tspi->command1_reg, tspi->dma_control_reg);
+		reset_control_assert(tspi->rst);
+		udelay(2);
+		reset_control_deassert(tspi->rst);
+		complete(&tspi->xfer_completion);
+		goto exit;
+	}
+
+	if (tspi->cur_direction & DATA_DIR_RX)
+		tegra_spi_read_rx_fifo_to_client_rxbuf(tspi, t);
+
+	if (tspi->cur_direction & DATA_DIR_TX)
+		tspi->cur_pos = tspi->cur_tx_pos;
+	else
+		tspi->cur_pos = tspi->cur_rx_pos;
+
+	if (tspi->cur_pos == t->len) {
+		complete(&tspi->xfer_completion);
+		goto exit;
+	}
+
+	tegra_spi_calculate_curr_xfer_param(tspi->cur_spi, tspi, t);
+	tegra_spi_start_cpu_based_transfer(tspi, t);
+exit:
+	spin_unlock_irqrestore(&tspi->lock, flags);
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t handle_dma_based_xfer(struct tegra_spi_data *tspi)
+{
+	struct spi_transfer *t = tspi->curr_xfer;
+	long wait_status;
+	int err = 0;
+	unsigned total_fifo_words;
+	unsigned long flags;
+
+	/* Abort dmas if any error */
+	if (tspi->cur_direction & DATA_DIR_TX) {
+		if (tspi->tx_status) {
+			dmaengine_terminate_all(tspi->tx_dma_chan);
+			err += 1;
+		} else {
+			wait_status = wait_for_completion_interruptible_timeout(
+				&tspi->tx_dma_complete, SPI_DMA_TIMEOUT);
+			if (wait_status <= 0) {
+				dmaengine_terminate_all(tspi->tx_dma_chan);
+				dev_err(tspi->dev, "TxDma Xfer failed\n");
+				err += 1;
+			}
+		}
+	}
+
+	if (tspi->cur_direction & DATA_DIR_RX) {
+		if (tspi->rx_status) {
+			dmaengine_terminate_all(tspi->rx_dma_chan);
+			err += 2;
+		} else {
+			wait_status = wait_for_completion_interruptible_timeout(
+				&tspi->rx_dma_complete, SPI_DMA_TIMEOUT);
+			if (wait_status <= 0) {
+				dmaengine_terminate_all(tspi->rx_dma_chan);
+				dev_err(tspi->dev, "RxDma Xfer failed\n");
+				err += 2;
+			}
+		}
+	}
+
+	spin_lock_irqsave(&tspi->lock, flags);
+	if (err) {
+		dev_err(tspi->dev, "DmaXfer: ERROR bit set 0x%x\n",
+			tspi->status_reg);
+		dev_err(tspi->dev, "DmaXfer 0x%08x:0x%08x\n",
+			tspi->command1_reg, tspi->dma_control_reg);
+		reset_control_assert(tspi->rst);
+		udelay(2);
+		reset_control_deassert(tspi->rst);
+		complete(&tspi->xfer_completion);
+		spin_unlock_irqrestore(&tspi->lock, flags);
+		return IRQ_HANDLED;
+	}
+
+	if (tspi->cur_direction & DATA_DIR_RX)
+		tegra_spi_copy_spi_rxbuf_to_client_rxbuf(tspi, t);
+
+	if (tspi->cur_direction & DATA_DIR_TX)
+		tspi->cur_pos = tspi->cur_tx_pos;
+	else
+		tspi->cur_pos = tspi->cur_rx_pos;
+
+	if (tspi->cur_pos == t->len) {
+		complete(&tspi->xfer_completion);
+		goto exit;
+	}
+
+	/* Continue transfer in current message */
+	total_fifo_words = tegra_spi_calculate_curr_xfer_param(tspi->cur_spi,
+							tspi, t);
+	if (total_fifo_words > SPI_FIFO_DEPTH)
+		err = tegra_spi_start_dma_based_transfer(tspi, t);
+	else
+		err = tegra_spi_start_cpu_based_transfer(tspi, t);
+
+exit:
+	spin_unlock_irqrestore(&tspi->lock, flags);
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t tegra_spi_isr_thread(int irq, void *context_data)
+{
+	struct tegra_spi_data *tspi = context_data;
+
+	if (!tspi->is_curr_dma_xfer)
+		return handle_cpu_based_xfer(tspi);
+	return handle_dma_based_xfer(tspi);
+}
+
+static irqreturn_t tegra_spi_isr(int irq, void *context_data)
+{
+	struct tegra_spi_data *tspi = context_data;
+
+	tspi->status_reg = tegra_spi_readl(tspi, SPI_FIFO_STATUS);
+	if (tspi->cur_direction & DATA_DIR_TX)
+		tspi->tx_status = tspi->status_reg &
+					(SPI_TX_FIFO_UNF | SPI_TX_FIFO_OVF);
+
+	if (tspi->cur_direction & DATA_DIR_RX)
+		tspi->rx_status = tspi->status_reg &
+					(SPI_RX_FIFO_OVF | SPI_RX_FIFO_UNF);
+	tegra_spi_clear_status(tspi);
+
+	return IRQ_WAKE_THREAD;
+}
+
+static const struct of_device_id tegra_spi_of_match[] = {
+	{ .compatible = "nvidia,tegra114-spi", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, tegra_spi_of_match);
+
+static int tegra_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master	*master;
+	struct tegra_spi_data	*tspi;
+	struct resource		*r;
+	int ret, spi_irq;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*tspi));
+	if (!master) {
+		dev_err(&pdev->dev, "master allocation failed\n");
+		return -ENOMEM;
+	}
+	platform_set_drvdata(pdev, master);
+	tspi = spi_master_get_devdata(master);
+
+	if (of_property_read_u32(pdev->dev.of_node, "spi-max-frequency",
+				 &master->max_speed_hz))
+		master->max_speed_hz = 25000000; /* 25MHz */
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->setup = tegra_spi_setup;
+	master->transfer_one_message = tegra_spi_transfer_one_message;
+	master->num_chipselect = MAX_CHIP_SELECT;
+	master->auto_runtime_pm = true;
+
+	tspi->master = master;
+	tspi->dev = &pdev->dev;
+	spin_lock_init(&tspi->lock);
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	tspi->base = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(tspi->base)) {
+		ret = PTR_ERR(tspi->base);
+		goto exit_free_master;
+	}
+	tspi->phys = r->start;
+
+	spi_irq = platform_get_irq(pdev, 0);
+	tspi->irq = spi_irq;
+	ret = request_threaded_irq(tspi->irq, tegra_spi_isr,
+			tegra_spi_isr_thread, IRQF_ONESHOT,
+			dev_name(&pdev->dev), tspi);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to register ISR for IRQ %d\n",
+					tspi->irq);
+		goto exit_free_master;
+	}
+
+	tspi->clk = devm_clk_get(&pdev->dev, "spi");
+	if (IS_ERR(tspi->clk)) {
+		dev_err(&pdev->dev, "can not get clock\n");
+		ret = PTR_ERR(tspi->clk);
+		goto exit_free_irq;
+	}
+
+	tspi->rst = devm_reset_control_get_exclusive(&pdev->dev, "spi");
+	if (IS_ERR(tspi->rst)) {
+		dev_err(&pdev->dev, "can not get reset\n");
+		ret = PTR_ERR(tspi->rst);
+		goto exit_free_irq;
+	}
+
+	tspi->max_buf_size = SPI_FIFO_DEPTH << 2;
+	tspi->dma_buf_size = DEFAULT_SPI_DMA_BUF_LEN;
+
+	ret = tegra_spi_init_dma_param(tspi, true);
+	if (ret < 0)
+		goto exit_free_irq;
+	ret = tegra_spi_init_dma_param(tspi, false);
+	if (ret < 0)
+		goto exit_rx_dma_free;
+	tspi->max_buf_size = tspi->dma_buf_size;
+	init_completion(&tspi->tx_dma_complete);
+	init_completion(&tspi->rx_dma_complete);
+
+	init_completion(&tspi->xfer_completion);
+
+	pm_runtime_enable(&pdev->dev);
+	if (!pm_runtime_enabled(&pdev->dev)) {
+		ret = tegra_spi_runtime_resume(&pdev->dev);
+		if (ret)
+			goto exit_pm_disable;
+	}
+
+	ret = pm_runtime_get_sync(&pdev->dev);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "pm runtime get failed, e = %d\n", ret);
+		goto exit_pm_disable;
+	}
+	tspi->def_command1_reg  = SPI_M_S;
+	tegra_spi_writel(tspi, tspi->def_command1_reg, SPI_COMMAND1);
+	pm_runtime_put(&pdev->dev);
+
+	master->dev.of_node = pdev->dev.of_node;
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "can not register to master err %d\n", ret);
+		goto exit_pm_disable;
+	}
+	return ret;
+
+exit_pm_disable:
+	pm_runtime_disable(&pdev->dev);
+	if (!pm_runtime_status_suspended(&pdev->dev))
+		tegra_spi_runtime_suspend(&pdev->dev);
+	tegra_spi_deinit_dma_param(tspi, false);
+exit_rx_dma_free:
+	tegra_spi_deinit_dma_param(tspi, true);
+exit_free_irq:
+	free_irq(spi_irq, tspi);
+exit_free_master:
+	spi_master_put(master);
+	return ret;
+}
+
+static int tegra_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct tegra_spi_data	*tspi = spi_master_get_devdata(master);
+
+	free_irq(tspi->irq, tspi);
+
+	if (tspi->tx_dma_chan)
+		tegra_spi_deinit_dma_param(tspi, false);
+
+	if (tspi->rx_dma_chan)
+		tegra_spi_deinit_dma_param(tspi, true);
+
+	pm_runtime_disable(&pdev->dev);
+	if (!pm_runtime_status_suspended(&pdev->dev))
+		tegra_spi_runtime_suspend(&pdev->dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int tegra_spi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+
+	return spi_master_suspend(master);
+}
+
+static int tegra_spi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct tegra_spi_data *tspi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = pm_runtime_get_sync(dev);
+	if (ret < 0) {
+		dev_err(dev, "pm runtime failed, e = %d\n", ret);
+		return ret;
+	}
+	tegra_spi_writel(tspi, tspi->command1_reg, SPI_COMMAND1);
+	pm_runtime_put(dev);
+
+	return spi_master_resume(master);
+}
+#endif
+
+static int tegra_spi_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct tegra_spi_data *tspi = spi_master_get_devdata(master);
+
+	/* Flush all write which are in PPSB queue by reading back */
+	tegra_spi_readl(tspi, SPI_COMMAND1);
+
+	clk_disable_unprepare(tspi->clk);
+	return 0;
+}
+
+static int tegra_spi_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct tegra_spi_data *tspi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(tspi->clk);
+	if (ret < 0) {
+		dev_err(tspi->dev, "clk_prepare failed: %d\n", ret);
+		return ret;
+	}
+	return 0;
+}
+
+static const struct dev_pm_ops tegra_spi_pm_ops = {
+	SET_RUNTIME_PM_OPS(tegra_spi_runtime_suspend,
+		tegra_spi_runtime_resume, NULL)
+	SET_SYSTEM_SLEEP_PM_OPS(tegra_spi_suspend, tegra_spi_resume)
+};
+static struct platform_driver tegra_spi_driver = {
+	.driver = {
+		.name		= "spi-tegra114",
+		.pm		= &tegra_spi_pm_ops,
+		.of_match_table	= tegra_spi_of_match,
+	},
+	.probe =	tegra_spi_probe,
+	.remove =	tegra_spi_remove,
+};
+module_platform_driver(tegra_spi_driver);
+
+MODULE_ALIAS("platform:spi-tegra114");
+MODULE_DESCRIPTION("NVIDIA Tegra114 SPI Controller Driver");
+MODULE_AUTHOR("Laxman Dewangan <ldewangan@nvidia.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-tegra20-sflash.c b/drivers/spi/spi-tegra20-sflash.c
new file mode 100644
index 0000000..22893a7
--- /dev/null
+++ b/drivers/spi/spi-tegra20-sflash.c
@@ -0,0 +1,622 @@
+/*
+ * SPI driver for Nvidia's Tegra20 Serial Flash Controller.
+ *
+ * Copyright (c) 2012, NVIDIA CORPORATION.  All rights reserved.
+ *
+ * Author: Laxman Dewangan <ldewangan@nvidia.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/kthread.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/reset.h>
+#include <linux/spi/spi.h>
+
+#define SPI_COMMAND				0x000
+#define SPI_GO					BIT(30)
+#define SPI_M_S					BIT(28)
+#define SPI_ACTIVE_SCLK_MASK			(0x3 << 26)
+#define SPI_ACTIVE_SCLK_DRIVE_LOW		(0 << 26)
+#define SPI_ACTIVE_SCLK_DRIVE_HIGH		(1 << 26)
+#define SPI_ACTIVE_SCLK_PULL_LOW		(2 << 26)
+#define SPI_ACTIVE_SCLK_PULL_HIGH		(3 << 26)
+
+#define SPI_CK_SDA_FALLING			(1 << 21)
+#define SPI_CK_SDA_RISING			(0 << 21)
+#define SPI_CK_SDA_MASK				(1 << 21)
+#define SPI_ACTIVE_SDA				(0x3 << 18)
+#define SPI_ACTIVE_SDA_DRIVE_LOW		(0 << 18)
+#define SPI_ACTIVE_SDA_DRIVE_HIGH		(1 << 18)
+#define SPI_ACTIVE_SDA_PULL_LOW			(2 << 18)
+#define SPI_ACTIVE_SDA_PULL_HIGH		(3 << 18)
+
+#define SPI_CS_POL_INVERT			BIT(16)
+#define SPI_TX_EN				BIT(15)
+#define SPI_RX_EN				BIT(14)
+#define SPI_CS_VAL_HIGH				BIT(13)
+#define SPI_CS_VAL_LOW				0x0
+#define SPI_CS_SW				BIT(12)
+#define SPI_CS_HW				0x0
+#define SPI_CS_DELAY_MASK			(7 << 9)
+#define SPI_CS3_EN				BIT(8)
+#define SPI_CS2_EN				BIT(7)
+#define SPI_CS1_EN				BIT(6)
+#define SPI_CS0_EN				BIT(5)
+
+#define SPI_CS_MASK			(SPI_CS3_EN | SPI_CS2_EN |	\
+					SPI_CS1_EN | SPI_CS0_EN)
+#define SPI_BIT_LENGTH(x)		(((x) & 0x1f) << 0)
+
+#define SPI_MODES			(SPI_ACTIVE_SCLK_MASK | SPI_CK_SDA_MASK)
+
+#define SPI_STATUS			0x004
+#define SPI_BSY				BIT(31)
+#define SPI_RDY				BIT(30)
+#define SPI_TXF_FLUSH			BIT(29)
+#define SPI_RXF_FLUSH			BIT(28)
+#define SPI_RX_UNF			BIT(27)
+#define SPI_TX_OVF			BIT(26)
+#define SPI_RXF_EMPTY			BIT(25)
+#define SPI_RXF_FULL			BIT(24)
+#define SPI_TXF_EMPTY			BIT(23)
+#define SPI_TXF_FULL			BIT(22)
+#define SPI_BLK_CNT(count)		(((count) & 0xffff) + 1)
+
+#define SPI_FIFO_ERROR			(SPI_RX_UNF | SPI_TX_OVF)
+#define SPI_FIFO_EMPTY			(SPI_TX_EMPTY | SPI_RX_EMPTY)
+
+#define SPI_RX_CMP			0x8
+#define SPI_DMA_CTL			0x0C
+#define SPI_DMA_EN			BIT(31)
+#define SPI_IE_RXC			BIT(27)
+#define SPI_IE_TXC			BIT(26)
+#define SPI_PACKED			BIT(20)
+#define SPI_RX_TRIG_MASK		(0x3 << 18)
+#define SPI_RX_TRIG_1W			(0x0 << 18)
+#define SPI_RX_TRIG_4W			(0x1 << 18)
+#define SPI_TX_TRIG_MASK		(0x3 << 16)
+#define SPI_TX_TRIG_1W			(0x0 << 16)
+#define SPI_TX_TRIG_4W			(0x1 << 16)
+#define SPI_DMA_BLK_COUNT(count)	(((count) - 1) & 0xFFFF)
+
+#define SPI_TX_FIFO			0x10
+#define SPI_RX_FIFO			0x20
+
+#define DATA_DIR_TX			(1 << 0)
+#define DATA_DIR_RX			(1 << 1)
+
+#define MAX_CHIP_SELECT			4
+#define SPI_FIFO_DEPTH			4
+#define SPI_DMA_TIMEOUT               (msecs_to_jiffies(1000))
+
+struct tegra_sflash_data {
+	struct device				*dev;
+	struct spi_master			*master;
+	spinlock_t				lock;
+
+	struct clk				*clk;
+	struct reset_control			*rst;
+	void __iomem				*base;
+	unsigned				irq;
+	u32					cur_speed;
+
+	struct spi_device			*cur_spi;
+	unsigned				cur_pos;
+	unsigned				cur_len;
+	unsigned				bytes_per_word;
+	unsigned				cur_direction;
+	unsigned				curr_xfer_words;
+
+	unsigned				cur_rx_pos;
+	unsigned				cur_tx_pos;
+
+	u32					tx_status;
+	u32					rx_status;
+	u32					status_reg;
+
+	u32					def_command_reg;
+	u32					command_reg;
+	u32					dma_control_reg;
+
+	struct completion			xfer_completion;
+	struct spi_transfer			*curr_xfer;
+};
+
+static int tegra_sflash_runtime_suspend(struct device *dev);
+static int tegra_sflash_runtime_resume(struct device *dev);
+
+static inline u32 tegra_sflash_readl(struct tegra_sflash_data *tsd,
+		unsigned long reg)
+{
+	return readl(tsd->base + reg);
+}
+
+static inline void tegra_sflash_writel(struct tegra_sflash_data *tsd,
+		u32 val, unsigned long reg)
+{
+	writel(val, tsd->base + reg);
+}
+
+static void tegra_sflash_clear_status(struct tegra_sflash_data *tsd)
+{
+	/* Write 1 to clear status register */
+	tegra_sflash_writel(tsd, SPI_RDY | SPI_FIFO_ERROR, SPI_STATUS);
+}
+
+static unsigned tegra_sflash_calculate_curr_xfer_param(
+	struct spi_device *spi, struct tegra_sflash_data *tsd,
+	struct spi_transfer *t)
+{
+	unsigned remain_len = t->len - tsd->cur_pos;
+	unsigned max_word;
+
+	tsd->bytes_per_word = DIV_ROUND_UP(t->bits_per_word, 8);
+	max_word = remain_len / tsd->bytes_per_word;
+	if (max_word > SPI_FIFO_DEPTH)
+		max_word = SPI_FIFO_DEPTH;
+	tsd->curr_xfer_words = max_word;
+	return max_word;
+}
+
+static unsigned tegra_sflash_fill_tx_fifo_from_client_txbuf(
+	struct tegra_sflash_data *tsd, struct spi_transfer *t)
+{
+	unsigned nbytes;
+	u32 status;
+	unsigned max_n_32bit = tsd->curr_xfer_words;
+	u8 *tx_buf = (u8 *)t->tx_buf + tsd->cur_tx_pos;
+
+	if (max_n_32bit > SPI_FIFO_DEPTH)
+		max_n_32bit = SPI_FIFO_DEPTH;
+	nbytes = max_n_32bit * tsd->bytes_per_word;
+
+	status = tegra_sflash_readl(tsd, SPI_STATUS);
+	while (!(status & SPI_TXF_FULL)) {
+		int i;
+		u32 x = 0;
+
+		for (i = 0; nbytes && (i < tsd->bytes_per_word);
+							i++, nbytes--)
+			x |= (u32)(*tx_buf++) << (i * 8);
+		tegra_sflash_writel(tsd, x, SPI_TX_FIFO);
+		if (!nbytes)
+			break;
+
+		status = tegra_sflash_readl(tsd, SPI_STATUS);
+	}
+	tsd->cur_tx_pos += max_n_32bit * tsd->bytes_per_word;
+	return max_n_32bit;
+}
+
+static int tegra_sflash_read_rx_fifo_to_client_rxbuf(
+		struct tegra_sflash_data *tsd, struct spi_transfer *t)
+{
+	u32 status;
+	unsigned int read_words = 0;
+	u8 *rx_buf = (u8 *)t->rx_buf + tsd->cur_rx_pos;
+
+	status = tegra_sflash_readl(tsd, SPI_STATUS);
+	while (!(status & SPI_RXF_EMPTY)) {
+		int i;
+		u32 x = tegra_sflash_readl(tsd, SPI_RX_FIFO);
+
+		for (i = 0; (i < tsd->bytes_per_word); i++)
+			*rx_buf++ = (x >> (i*8)) & 0xFF;
+		read_words++;
+		status = tegra_sflash_readl(tsd, SPI_STATUS);
+	}
+	tsd->cur_rx_pos += read_words * tsd->bytes_per_word;
+	return 0;
+}
+
+static int tegra_sflash_start_cpu_based_transfer(
+		struct tegra_sflash_data *tsd, struct spi_transfer *t)
+{
+	u32 val = 0;
+	unsigned cur_words;
+
+	if (tsd->cur_direction & DATA_DIR_TX)
+		val |= SPI_IE_TXC;
+
+	if (tsd->cur_direction & DATA_DIR_RX)
+		val |= SPI_IE_RXC;
+
+	tegra_sflash_writel(tsd, val, SPI_DMA_CTL);
+	tsd->dma_control_reg = val;
+
+	if (tsd->cur_direction & DATA_DIR_TX)
+		cur_words = tegra_sflash_fill_tx_fifo_from_client_txbuf(tsd, t);
+	else
+		cur_words = tsd->curr_xfer_words;
+	val |= SPI_DMA_BLK_COUNT(cur_words);
+	tegra_sflash_writel(tsd, val, SPI_DMA_CTL);
+	tsd->dma_control_reg = val;
+	val |= SPI_DMA_EN;
+	tegra_sflash_writel(tsd, val, SPI_DMA_CTL);
+	return 0;
+}
+
+static int tegra_sflash_start_transfer_one(struct spi_device *spi,
+		struct spi_transfer *t, bool is_first_of_msg,
+		bool is_single_xfer)
+{
+	struct tegra_sflash_data *tsd = spi_master_get_devdata(spi->master);
+	u32 speed;
+	u32 command;
+
+	speed = t->speed_hz;
+	if (speed != tsd->cur_speed) {
+		clk_set_rate(tsd->clk, speed);
+		tsd->cur_speed = speed;
+	}
+
+	tsd->cur_spi = spi;
+	tsd->cur_pos = 0;
+	tsd->cur_rx_pos = 0;
+	tsd->cur_tx_pos = 0;
+	tsd->curr_xfer = t;
+	tegra_sflash_calculate_curr_xfer_param(spi, tsd, t);
+	if (is_first_of_msg) {
+		command = tsd->def_command_reg;
+		command |= SPI_BIT_LENGTH(t->bits_per_word - 1);
+		command |= SPI_CS_VAL_HIGH;
+
+		command &= ~SPI_MODES;
+		if (spi->mode & SPI_CPHA)
+			command |= SPI_CK_SDA_FALLING;
+
+		if (spi->mode & SPI_CPOL)
+			command |= SPI_ACTIVE_SCLK_DRIVE_HIGH;
+		else
+			command |= SPI_ACTIVE_SCLK_DRIVE_LOW;
+		command |= SPI_CS0_EN << spi->chip_select;
+	} else {
+		command = tsd->command_reg;
+		command &= ~SPI_BIT_LENGTH(~0);
+		command |= SPI_BIT_LENGTH(t->bits_per_word - 1);
+		command &= ~(SPI_RX_EN | SPI_TX_EN);
+	}
+
+	tsd->cur_direction = 0;
+	if (t->rx_buf) {
+		command |= SPI_RX_EN;
+		tsd->cur_direction |= DATA_DIR_RX;
+	}
+	if (t->tx_buf) {
+		command |= SPI_TX_EN;
+		tsd->cur_direction |= DATA_DIR_TX;
+	}
+	tegra_sflash_writel(tsd, command, SPI_COMMAND);
+	tsd->command_reg = command;
+
+	return tegra_sflash_start_cpu_based_transfer(tsd, t);
+}
+
+static int tegra_sflash_transfer_one_message(struct spi_master *master,
+			struct spi_message *msg)
+{
+	bool is_first_msg = true;
+	int single_xfer;
+	struct tegra_sflash_data *tsd = spi_master_get_devdata(master);
+	struct spi_transfer *xfer;
+	struct spi_device *spi = msg->spi;
+	int ret;
+
+	msg->status = 0;
+	msg->actual_length = 0;
+	single_xfer = list_is_singular(&msg->transfers);
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		reinit_completion(&tsd->xfer_completion);
+		ret = tegra_sflash_start_transfer_one(spi, xfer,
+					is_first_msg, single_xfer);
+		if (ret < 0) {
+			dev_err(tsd->dev,
+				"spi can not start transfer, err %d\n", ret);
+			goto exit;
+		}
+		is_first_msg = false;
+		ret = wait_for_completion_timeout(&tsd->xfer_completion,
+						SPI_DMA_TIMEOUT);
+		if (WARN_ON(ret == 0)) {
+			dev_err(tsd->dev,
+				"spi transfer timeout, err %d\n", ret);
+			ret = -EIO;
+			goto exit;
+		}
+
+		if (tsd->tx_status ||  tsd->rx_status) {
+			dev_err(tsd->dev, "Error in Transfer\n");
+			ret = -EIO;
+			goto exit;
+		}
+		msg->actual_length += xfer->len;
+		if (xfer->cs_change && xfer->delay_usecs) {
+			tegra_sflash_writel(tsd, tsd->def_command_reg,
+					SPI_COMMAND);
+			udelay(xfer->delay_usecs);
+		}
+	}
+	ret = 0;
+exit:
+	tegra_sflash_writel(tsd, tsd->def_command_reg, SPI_COMMAND);
+	msg->status = ret;
+	spi_finalize_current_message(master);
+	return ret;
+}
+
+static irqreturn_t handle_cpu_based_xfer(struct tegra_sflash_data *tsd)
+{
+	struct spi_transfer *t = tsd->curr_xfer;
+	unsigned long flags;
+
+	spin_lock_irqsave(&tsd->lock, flags);
+	if (tsd->tx_status || tsd->rx_status || (tsd->status_reg & SPI_BSY)) {
+		dev_err(tsd->dev,
+			"CpuXfer ERROR bit set 0x%x\n", tsd->status_reg);
+		dev_err(tsd->dev,
+			"CpuXfer 0x%08x:0x%08x\n", tsd->command_reg,
+				tsd->dma_control_reg);
+		reset_control_assert(tsd->rst);
+		udelay(2);
+		reset_control_deassert(tsd->rst);
+		complete(&tsd->xfer_completion);
+		goto exit;
+	}
+
+	if (tsd->cur_direction & DATA_DIR_RX)
+		tegra_sflash_read_rx_fifo_to_client_rxbuf(tsd, t);
+
+	if (tsd->cur_direction & DATA_DIR_TX)
+		tsd->cur_pos = tsd->cur_tx_pos;
+	else
+		tsd->cur_pos = tsd->cur_rx_pos;
+
+	if (tsd->cur_pos == t->len) {
+		complete(&tsd->xfer_completion);
+		goto exit;
+	}
+
+	tegra_sflash_calculate_curr_xfer_param(tsd->cur_spi, tsd, t);
+	tegra_sflash_start_cpu_based_transfer(tsd, t);
+exit:
+	spin_unlock_irqrestore(&tsd->lock, flags);
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t tegra_sflash_isr(int irq, void *context_data)
+{
+	struct tegra_sflash_data *tsd = context_data;
+
+	tsd->status_reg = tegra_sflash_readl(tsd, SPI_STATUS);
+	if (tsd->cur_direction & DATA_DIR_TX)
+		tsd->tx_status = tsd->status_reg & SPI_TX_OVF;
+
+	if (tsd->cur_direction & DATA_DIR_RX)
+		tsd->rx_status = tsd->status_reg & SPI_RX_UNF;
+	tegra_sflash_clear_status(tsd);
+
+	return handle_cpu_based_xfer(tsd);
+}
+
+static const struct of_device_id tegra_sflash_of_match[] = {
+	{ .compatible = "nvidia,tegra20-sflash", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, tegra_sflash_of_match);
+
+static int tegra_sflash_probe(struct platform_device *pdev)
+{
+	struct spi_master	*master;
+	struct tegra_sflash_data	*tsd;
+	struct resource		*r;
+	int ret;
+	const struct of_device_id *match;
+
+	match = of_match_device(tegra_sflash_of_match, &pdev->dev);
+	if (!match) {
+		dev_err(&pdev->dev, "Error: No device match found\n");
+		return -ENODEV;
+	}
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*tsd));
+	if (!master) {
+		dev_err(&pdev->dev, "master allocation failed\n");
+		return -ENOMEM;
+	}
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA;
+	master->transfer_one_message = tegra_sflash_transfer_one_message;
+	master->auto_runtime_pm = true;
+	master->num_chipselect = MAX_CHIP_SELECT;
+
+	platform_set_drvdata(pdev, master);
+	tsd = spi_master_get_devdata(master);
+	tsd->master = master;
+	tsd->dev = &pdev->dev;
+	spin_lock_init(&tsd->lock);
+
+	if (of_property_read_u32(tsd->dev->of_node, "spi-max-frequency",
+				 &master->max_speed_hz))
+		master->max_speed_hz = 25000000; /* 25MHz */
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	tsd->base = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(tsd->base)) {
+		ret = PTR_ERR(tsd->base);
+		goto exit_free_master;
+	}
+
+	tsd->irq = platform_get_irq(pdev, 0);
+	ret = request_irq(tsd->irq, tegra_sflash_isr, 0,
+			dev_name(&pdev->dev), tsd);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to register ISR for IRQ %d\n",
+					tsd->irq);
+		goto exit_free_master;
+	}
+
+	tsd->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(tsd->clk)) {
+		dev_err(&pdev->dev, "can not get clock\n");
+		ret = PTR_ERR(tsd->clk);
+		goto exit_free_irq;
+	}
+
+	tsd->rst = devm_reset_control_get_exclusive(&pdev->dev, "spi");
+	if (IS_ERR(tsd->rst)) {
+		dev_err(&pdev->dev, "can not get reset\n");
+		ret = PTR_ERR(tsd->rst);
+		goto exit_free_irq;
+	}
+
+	init_completion(&tsd->xfer_completion);
+	pm_runtime_enable(&pdev->dev);
+	if (!pm_runtime_enabled(&pdev->dev)) {
+		ret = tegra_sflash_runtime_resume(&pdev->dev);
+		if (ret)
+			goto exit_pm_disable;
+	}
+
+	ret = pm_runtime_get_sync(&pdev->dev);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "pm runtime get failed, e = %d\n", ret);
+		goto exit_pm_disable;
+	}
+
+	/* Reset controller */
+	reset_control_assert(tsd->rst);
+	udelay(2);
+	reset_control_deassert(tsd->rst);
+
+	tsd->def_command_reg  = SPI_M_S | SPI_CS_SW;
+	tegra_sflash_writel(tsd, tsd->def_command_reg, SPI_COMMAND);
+	pm_runtime_put(&pdev->dev);
+
+	master->dev.of_node = pdev->dev.of_node;
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "can not register to master err %d\n", ret);
+		goto exit_pm_disable;
+	}
+	return ret;
+
+exit_pm_disable:
+	pm_runtime_disable(&pdev->dev);
+	if (!pm_runtime_status_suspended(&pdev->dev))
+		tegra_sflash_runtime_suspend(&pdev->dev);
+exit_free_irq:
+	free_irq(tsd->irq, tsd);
+exit_free_master:
+	spi_master_put(master);
+	return ret;
+}
+
+static int tegra_sflash_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct tegra_sflash_data	*tsd = spi_master_get_devdata(master);
+
+	free_irq(tsd->irq, tsd);
+
+	pm_runtime_disable(&pdev->dev);
+	if (!pm_runtime_status_suspended(&pdev->dev))
+		tegra_sflash_runtime_suspend(&pdev->dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int tegra_sflash_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+
+	return spi_master_suspend(master);
+}
+
+static int tegra_sflash_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct tegra_sflash_data *tsd = spi_master_get_devdata(master);
+	int ret;
+
+	ret = pm_runtime_get_sync(dev);
+	if (ret < 0) {
+		dev_err(dev, "pm runtime failed, e = %d\n", ret);
+		return ret;
+	}
+	tegra_sflash_writel(tsd, tsd->command_reg, SPI_COMMAND);
+	pm_runtime_put(dev);
+
+	return spi_master_resume(master);
+}
+#endif
+
+static int tegra_sflash_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct tegra_sflash_data *tsd = spi_master_get_devdata(master);
+
+	/* Flush all write which are in PPSB queue by reading back */
+	tegra_sflash_readl(tsd, SPI_COMMAND);
+
+	clk_disable_unprepare(tsd->clk);
+	return 0;
+}
+
+static int tegra_sflash_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct tegra_sflash_data *tsd = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(tsd->clk);
+	if (ret < 0) {
+		dev_err(tsd->dev, "clk_prepare failed: %d\n", ret);
+		return ret;
+	}
+	return 0;
+}
+
+static const struct dev_pm_ops slink_pm_ops = {
+	SET_RUNTIME_PM_OPS(tegra_sflash_runtime_suspend,
+		tegra_sflash_runtime_resume, NULL)
+	SET_SYSTEM_SLEEP_PM_OPS(tegra_sflash_suspend, tegra_sflash_resume)
+};
+static struct platform_driver tegra_sflash_driver = {
+	.driver = {
+		.name		= "spi-tegra-sflash",
+		.pm		= &slink_pm_ops,
+		.of_match_table	= tegra_sflash_of_match,
+	},
+	.probe =	tegra_sflash_probe,
+	.remove =	tegra_sflash_remove,
+};
+module_platform_driver(tegra_sflash_driver);
+
+MODULE_ALIAS("platform:spi-tegra-sflash");
+MODULE_DESCRIPTION("NVIDIA Tegra20 Serial Flash Controller Driver");
+MODULE_AUTHOR("Laxman Dewangan <ldewangan@nvidia.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-tegra20-slink.c b/drivers/spi/spi-tegra20-slink.c
new file mode 100644
index 0000000..1427f34
--- /dev/null
+++ b/drivers/spi/spi-tegra20-slink.c
@@ -0,0 +1,1253 @@
+/*
+ * SPI driver for Nvidia's Tegra20/Tegra30 SLINK Controller.
+ *
+ * Copyright (c) 2012, NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmapool.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/kthread.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/reset.h>
+#include <linux/spi/spi.h>
+
+#define SLINK_COMMAND			0x000
+#define SLINK_BIT_LENGTH(x)		(((x) & 0x1f) << 0)
+#define SLINK_WORD_SIZE(x)		(((x) & 0x1f) << 5)
+#define SLINK_BOTH_EN			(1 << 10)
+#define SLINK_CS_SW			(1 << 11)
+#define SLINK_CS_VALUE			(1 << 12)
+#define SLINK_CS_POLARITY		(1 << 13)
+#define SLINK_IDLE_SDA_DRIVE_LOW	(0 << 16)
+#define SLINK_IDLE_SDA_DRIVE_HIGH	(1 << 16)
+#define SLINK_IDLE_SDA_PULL_LOW		(2 << 16)
+#define SLINK_IDLE_SDA_PULL_HIGH	(3 << 16)
+#define SLINK_IDLE_SDA_MASK		(3 << 16)
+#define SLINK_CS_POLARITY1		(1 << 20)
+#define SLINK_CK_SDA			(1 << 21)
+#define SLINK_CS_POLARITY2		(1 << 22)
+#define SLINK_CS_POLARITY3		(1 << 23)
+#define SLINK_IDLE_SCLK_DRIVE_LOW	(0 << 24)
+#define SLINK_IDLE_SCLK_DRIVE_HIGH	(1 << 24)
+#define SLINK_IDLE_SCLK_PULL_LOW	(2 << 24)
+#define SLINK_IDLE_SCLK_PULL_HIGH	(3 << 24)
+#define SLINK_IDLE_SCLK_MASK		(3 << 24)
+#define SLINK_M_S			(1 << 28)
+#define SLINK_WAIT			(1 << 29)
+#define SLINK_GO			(1 << 30)
+#define SLINK_ENB			(1 << 31)
+
+#define SLINK_MODES			(SLINK_IDLE_SCLK_MASK | SLINK_CK_SDA)
+
+#define SLINK_COMMAND2			0x004
+#define SLINK_LSBFE			(1 << 0)
+#define SLINK_SSOE			(1 << 1)
+#define SLINK_SPIE			(1 << 4)
+#define SLINK_BIDIROE			(1 << 6)
+#define SLINK_MODFEN			(1 << 7)
+#define SLINK_INT_SIZE(x)		(((x) & 0x1f) << 8)
+#define SLINK_CS_ACTIVE_BETWEEN		(1 << 17)
+#define SLINK_SS_EN_CS(x)		(((x) & 0x3) << 18)
+#define SLINK_SS_SETUP(x)		(((x) & 0x3) << 20)
+#define SLINK_FIFO_REFILLS_0		(0 << 22)
+#define SLINK_FIFO_REFILLS_1		(1 << 22)
+#define SLINK_FIFO_REFILLS_2		(2 << 22)
+#define SLINK_FIFO_REFILLS_3		(3 << 22)
+#define SLINK_FIFO_REFILLS_MASK		(3 << 22)
+#define SLINK_WAIT_PACK_INT(x)		(((x) & 0x7) << 26)
+#define SLINK_SPC0			(1 << 29)
+#define SLINK_TXEN			(1 << 30)
+#define SLINK_RXEN			(1 << 31)
+
+#define SLINK_STATUS			0x008
+#define SLINK_COUNT(val)		(((val) >> 0) & 0x1f)
+#define SLINK_WORD(val)			(((val) >> 5) & 0x1f)
+#define SLINK_BLK_CNT(val)		(((val) >> 0) & 0xffff)
+#define SLINK_MODF			(1 << 16)
+#define SLINK_RX_UNF			(1 << 18)
+#define SLINK_TX_OVF			(1 << 19)
+#define SLINK_TX_FULL			(1 << 20)
+#define SLINK_TX_EMPTY			(1 << 21)
+#define SLINK_RX_FULL			(1 << 22)
+#define SLINK_RX_EMPTY			(1 << 23)
+#define SLINK_TX_UNF			(1 << 24)
+#define SLINK_RX_OVF			(1 << 25)
+#define SLINK_TX_FLUSH			(1 << 26)
+#define SLINK_RX_FLUSH			(1 << 27)
+#define SLINK_SCLK			(1 << 28)
+#define SLINK_ERR			(1 << 29)
+#define SLINK_RDY			(1 << 30)
+#define SLINK_BSY			(1 << 31)
+#define SLINK_FIFO_ERROR		(SLINK_TX_OVF | SLINK_RX_UNF |	\
+					SLINK_TX_UNF | SLINK_RX_OVF)
+
+#define SLINK_FIFO_EMPTY		(SLINK_TX_EMPTY | SLINK_RX_EMPTY)
+
+#define SLINK_MAS_DATA			0x010
+#define SLINK_SLAVE_DATA		0x014
+
+#define SLINK_DMA_CTL			0x018
+#define SLINK_DMA_BLOCK_SIZE(x)		(((x) & 0xffff) << 0)
+#define SLINK_TX_TRIG_1			(0 << 16)
+#define SLINK_TX_TRIG_4			(1 << 16)
+#define SLINK_TX_TRIG_8			(2 << 16)
+#define SLINK_TX_TRIG_16		(3 << 16)
+#define SLINK_TX_TRIG_MASK		(3 << 16)
+#define SLINK_RX_TRIG_1			(0 << 18)
+#define SLINK_RX_TRIG_4			(1 << 18)
+#define SLINK_RX_TRIG_8			(2 << 18)
+#define SLINK_RX_TRIG_16		(3 << 18)
+#define SLINK_RX_TRIG_MASK		(3 << 18)
+#define SLINK_PACKED			(1 << 20)
+#define SLINK_PACK_SIZE_4		(0 << 21)
+#define SLINK_PACK_SIZE_8		(1 << 21)
+#define SLINK_PACK_SIZE_16		(2 << 21)
+#define SLINK_PACK_SIZE_32		(3 << 21)
+#define SLINK_PACK_SIZE_MASK		(3 << 21)
+#define SLINK_IE_TXC			(1 << 26)
+#define SLINK_IE_RXC			(1 << 27)
+#define SLINK_DMA_EN			(1 << 31)
+
+#define SLINK_STATUS2			0x01c
+#define SLINK_TX_FIFO_EMPTY_COUNT(val)	(((val) & 0x3f) >> 0)
+#define SLINK_RX_FIFO_FULL_COUNT(val)	(((val) & 0x3f0000) >> 16)
+#define SLINK_SS_HOLD_TIME(val)		(((val) & 0xF) << 6)
+
+#define SLINK_TX_FIFO			0x100
+#define SLINK_RX_FIFO			0x180
+
+#define DATA_DIR_TX			(1 << 0)
+#define DATA_DIR_RX			(1 << 1)
+
+#define SLINK_DMA_TIMEOUT		(msecs_to_jiffies(1000))
+
+#define DEFAULT_SPI_DMA_BUF_LEN		(16*1024)
+#define TX_FIFO_EMPTY_COUNT_MAX		SLINK_TX_FIFO_EMPTY_COUNT(0x20)
+#define RX_FIFO_FULL_COUNT_ZERO		SLINK_RX_FIFO_FULL_COUNT(0)
+
+#define SLINK_STATUS2_RESET \
+	(TX_FIFO_EMPTY_COUNT_MAX | RX_FIFO_FULL_COUNT_ZERO << 16)
+
+#define MAX_CHIP_SELECT			4
+#define SLINK_FIFO_DEPTH		32
+
+struct tegra_slink_chip_data {
+	bool cs_hold_time;
+};
+
+struct tegra_slink_data {
+	struct device				*dev;
+	struct spi_master			*master;
+	const struct tegra_slink_chip_data	*chip_data;
+	spinlock_t				lock;
+
+	struct clk				*clk;
+	struct reset_control			*rst;
+	void __iomem				*base;
+	phys_addr_t				phys;
+	unsigned				irq;
+	u32					cur_speed;
+
+	struct spi_device			*cur_spi;
+	unsigned				cur_pos;
+	unsigned				cur_len;
+	unsigned				words_per_32bit;
+	unsigned				bytes_per_word;
+	unsigned				curr_dma_words;
+	unsigned				cur_direction;
+
+	unsigned				cur_rx_pos;
+	unsigned				cur_tx_pos;
+
+	unsigned				dma_buf_size;
+	unsigned				max_buf_size;
+	bool					is_curr_dma_xfer;
+
+	struct completion			rx_dma_complete;
+	struct completion			tx_dma_complete;
+
+	u32					tx_status;
+	u32					rx_status;
+	u32					status_reg;
+	bool					is_packed;
+	u32					packed_size;
+
+	u32					command_reg;
+	u32					command2_reg;
+	u32					dma_control_reg;
+	u32					def_command_reg;
+	u32					def_command2_reg;
+
+	struct completion			xfer_completion;
+	struct spi_transfer			*curr_xfer;
+	struct dma_chan				*rx_dma_chan;
+	u32					*rx_dma_buf;
+	dma_addr_t				rx_dma_phys;
+	struct dma_async_tx_descriptor		*rx_dma_desc;
+
+	struct dma_chan				*tx_dma_chan;
+	u32					*tx_dma_buf;
+	dma_addr_t				tx_dma_phys;
+	struct dma_async_tx_descriptor		*tx_dma_desc;
+};
+
+static int tegra_slink_runtime_suspend(struct device *dev);
+static int tegra_slink_runtime_resume(struct device *dev);
+
+static inline u32 tegra_slink_readl(struct tegra_slink_data *tspi,
+		unsigned long reg)
+{
+	return readl(tspi->base + reg);
+}
+
+static inline void tegra_slink_writel(struct tegra_slink_data *tspi,
+		u32 val, unsigned long reg)
+{
+	writel(val, tspi->base + reg);
+
+	/* Read back register to make sure that register writes completed */
+	if (reg != SLINK_TX_FIFO)
+		readl(tspi->base + SLINK_MAS_DATA);
+}
+
+static void tegra_slink_clear_status(struct tegra_slink_data *tspi)
+{
+	u32 val_write;
+
+	tegra_slink_readl(tspi, SLINK_STATUS);
+
+	/* Write 1 to clear status register */
+	val_write = SLINK_RDY | SLINK_FIFO_ERROR;
+	tegra_slink_writel(tspi, val_write, SLINK_STATUS);
+}
+
+static u32 tegra_slink_get_packed_size(struct tegra_slink_data *tspi,
+				  struct spi_transfer *t)
+{
+	switch (tspi->bytes_per_word) {
+	case 0:
+		return SLINK_PACK_SIZE_4;
+	case 1:
+		return SLINK_PACK_SIZE_8;
+	case 2:
+		return SLINK_PACK_SIZE_16;
+	case 4:
+		return SLINK_PACK_SIZE_32;
+	default:
+		return 0;
+	}
+}
+
+static unsigned tegra_slink_calculate_curr_xfer_param(
+	struct spi_device *spi, struct tegra_slink_data *tspi,
+	struct spi_transfer *t)
+{
+	unsigned remain_len = t->len - tspi->cur_pos;
+	unsigned max_word;
+	unsigned bits_per_word;
+	unsigned max_len;
+	unsigned total_fifo_words;
+
+	bits_per_word = t->bits_per_word;
+	tspi->bytes_per_word = DIV_ROUND_UP(bits_per_word, 8);
+
+	if (bits_per_word == 8 || bits_per_word == 16) {
+		tspi->is_packed = true;
+		tspi->words_per_32bit = 32/bits_per_word;
+	} else {
+		tspi->is_packed = false;
+		tspi->words_per_32bit = 1;
+	}
+	tspi->packed_size = tegra_slink_get_packed_size(tspi, t);
+
+	if (tspi->is_packed) {
+		max_len = min(remain_len, tspi->max_buf_size);
+		tspi->curr_dma_words = max_len/tspi->bytes_per_word;
+		total_fifo_words = max_len/4;
+	} else {
+		max_word = (remain_len - 1) / tspi->bytes_per_word + 1;
+		max_word = min(max_word, tspi->max_buf_size/4);
+		tspi->curr_dma_words = max_word;
+		total_fifo_words = max_word;
+	}
+	return total_fifo_words;
+}
+
+static unsigned tegra_slink_fill_tx_fifo_from_client_txbuf(
+	struct tegra_slink_data *tspi, struct spi_transfer *t)
+{
+	unsigned nbytes;
+	unsigned tx_empty_count;
+	u32 fifo_status;
+	unsigned max_n_32bit;
+	unsigned i, count;
+	unsigned int written_words;
+	unsigned fifo_words_left;
+	u8 *tx_buf = (u8 *)t->tx_buf + tspi->cur_tx_pos;
+
+	fifo_status = tegra_slink_readl(tspi, SLINK_STATUS2);
+	tx_empty_count = SLINK_TX_FIFO_EMPTY_COUNT(fifo_status);
+
+	if (tspi->is_packed) {
+		fifo_words_left = tx_empty_count * tspi->words_per_32bit;
+		written_words = min(fifo_words_left, tspi->curr_dma_words);
+		nbytes = written_words * tspi->bytes_per_word;
+		max_n_32bit = DIV_ROUND_UP(nbytes, 4);
+		for (count = 0; count < max_n_32bit; count++) {
+			u32 x = 0;
+			for (i = 0; (i < 4) && nbytes; i++, nbytes--)
+				x |= (u32)(*tx_buf++) << (i * 8);
+			tegra_slink_writel(tspi, x, SLINK_TX_FIFO);
+		}
+	} else {
+		max_n_32bit = min(tspi->curr_dma_words,  tx_empty_count);
+		written_words = max_n_32bit;
+		nbytes = written_words * tspi->bytes_per_word;
+		for (count = 0; count < max_n_32bit; count++) {
+			u32 x = 0;
+			for (i = 0; nbytes && (i < tspi->bytes_per_word);
+							i++, nbytes--)
+				x |= (u32)(*tx_buf++) << (i * 8);
+			tegra_slink_writel(tspi, x, SLINK_TX_FIFO);
+		}
+	}
+	tspi->cur_tx_pos += written_words * tspi->bytes_per_word;
+	return written_words;
+}
+
+static unsigned int tegra_slink_read_rx_fifo_to_client_rxbuf(
+		struct tegra_slink_data *tspi, struct spi_transfer *t)
+{
+	unsigned rx_full_count;
+	u32 fifo_status;
+	unsigned i, count;
+	unsigned int read_words = 0;
+	unsigned len;
+	u8 *rx_buf = (u8 *)t->rx_buf + tspi->cur_rx_pos;
+
+	fifo_status = tegra_slink_readl(tspi, SLINK_STATUS2);
+	rx_full_count = SLINK_RX_FIFO_FULL_COUNT(fifo_status);
+	if (tspi->is_packed) {
+		len = tspi->curr_dma_words * tspi->bytes_per_word;
+		for (count = 0; count < rx_full_count; count++) {
+			u32 x = tegra_slink_readl(tspi, SLINK_RX_FIFO);
+			for (i = 0; len && (i < 4); i++, len--)
+				*rx_buf++ = (x >> i*8) & 0xFF;
+		}
+		tspi->cur_rx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
+		read_words += tspi->curr_dma_words;
+	} else {
+		for (count = 0; count < rx_full_count; count++) {
+			u32 x = tegra_slink_readl(tspi, SLINK_RX_FIFO);
+			for (i = 0; (i < tspi->bytes_per_word); i++)
+				*rx_buf++ = (x >> (i*8)) & 0xFF;
+		}
+		tspi->cur_rx_pos += rx_full_count * tspi->bytes_per_word;
+		read_words += rx_full_count;
+	}
+	return read_words;
+}
+
+static void tegra_slink_copy_client_txbuf_to_spi_txbuf(
+		struct tegra_slink_data *tspi, struct spi_transfer *t)
+{
+	/* Make the dma buffer to read by cpu */
+	dma_sync_single_for_cpu(tspi->dev, tspi->tx_dma_phys,
+				tspi->dma_buf_size, DMA_TO_DEVICE);
+
+	if (tspi->is_packed) {
+		unsigned len = tspi->curr_dma_words * tspi->bytes_per_word;
+		memcpy(tspi->tx_dma_buf, t->tx_buf + tspi->cur_pos, len);
+	} else {
+		unsigned int i;
+		unsigned int count;
+		u8 *tx_buf = (u8 *)t->tx_buf + tspi->cur_tx_pos;
+		unsigned consume = tspi->curr_dma_words * tspi->bytes_per_word;
+
+		for (count = 0; count < tspi->curr_dma_words; count++) {
+			u32 x = 0;
+			for (i = 0; consume && (i < tspi->bytes_per_word);
+							i++, consume--)
+				x |= (u32)(*tx_buf++) << (i * 8);
+			tspi->tx_dma_buf[count] = x;
+		}
+	}
+	tspi->cur_tx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
+
+	/* Make the dma buffer to read by dma */
+	dma_sync_single_for_device(tspi->dev, tspi->tx_dma_phys,
+				tspi->dma_buf_size, DMA_TO_DEVICE);
+}
+
+static void tegra_slink_copy_spi_rxbuf_to_client_rxbuf(
+		struct tegra_slink_data *tspi, struct spi_transfer *t)
+{
+	unsigned len;
+
+	/* Make the dma buffer to read by cpu */
+	dma_sync_single_for_cpu(tspi->dev, tspi->rx_dma_phys,
+		tspi->dma_buf_size, DMA_FROM_DEVICE);
+
+	if (tspi->is_packed) {
+		len = tspi->curr_dma_words * tspi->bytes_per_word;
+		memcpy(t->rx_buf + tspi->cur_rx_pos, tspi->rx_dma_buf, len);
+	} else {
+		unsigned int i;
+		unsigned int count;
+		unsigned char *rx_buf = t->rx_buf + tspi->cur_rx_pos;
+		u32 rx_mask = ((u32)1 << t->bits_per_word) - 1;
+
+		for (count = 0; count < tspi->curr_dma_words; count++) {
+			u32 x = tspi->rx_dma_buf[count] & rx_mask;
+			for (i = 0; (i < tspi->bytes_per_word); i++)
+				*rx_buf++ = (x >> (i*8)) & 0xFF;
+		}
+	}
+	tspi->cur_rx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
+
+	/* Make the dma buffer to read by dma */
+	dma_sync_single_for_device(tspi->dev, tspi->rx_dma_phys,
+		tspi->dma_buf_size, DMA_FROM_DEVICE);
+}
+
+static void tegra_slink_dma_complete(void *args)
+{
+	struct completion *dma_complete = args;
+
+	complete(dma_complete);
+}
+
+static int tegra_slink_start_tx_dma(struct tegra_slink_data *tspi, int len)
+{
+	reinit_completion(&tspi->tx_dma_complete);
+	tspi->tx_dma_desc = dmaengine_prep_slave_single(tspi->tx_dma_chan,
+				tspi->tx_dma_phys, len, DMA_MEM_TO_DEV,
+				DMA_PREP_INTERRUPT |  DMA_CTRL_ACK);
+	if (!tspi->tx_dma_desc) {
+		dev_err(tspi->dev, "Not able to get desc for Tx\n");
+		return -EIO;
+	}
+
+	tspi->tx_dma_desc->callback = tegra_slink_dma_complete;
+	tspi->tx_dma_desc->callback_param = &tspi->tx_dma_complete;
+
+	dmaengine_submit(tspi->tx_dma_desc);
+	dma_async_issue_pending(tspi->tx_dma_chan);
+	return 0;
+}
+
+static int tegra_slink_start_rx_dma(struct tegra_slink_data *tspi, int len)
+{
+	reinit_completion(&tspi->rx_dma_complete);
+	tspi->rx_dma_desc = dmaengine_prep_slave_single(tspi->rx_dma_chan,
+				tspi->rx_dma_phys, len, DMA_DEV_TO_MEM,
+				DMA_PREP_INTERRUPT |  DMA_CTRL_ACK);
+	if (!tspi->rx_dma_desc) {
+		dev_err(tspi->dev, "Not able to get desc for Rx\n");
+		return -EIO;
+	}
+
+	tspi->rx_dma_desc->callback = tegra_slink_dma_complete;
+	tspi->rx_dma_desc->callback_param = &tspi->rx_dma_complete;
+
+	dmaengine_submit(tspi->rx_dma_desc);
+	dma_async_issue_pending(tspi->rx_dma_chan);
+	return 0;
+}
+
+static int tegra_slink_start_dma_based_transfer(
+		struct tegra_slink_data *tspi, struct spi_transfer *t)
+{
+	u32 val;
+	unsigned int len;
+	int ret = 0;
+	u32 status;
+
+	/* Make sure that Rx and Tx fifo are empty */
+	status = tegra_slink_readl(tspi, SLINK_STATUS);
+	if ((status & SLINK_FIFO_EMPTY) != SLINK_FIFO_EMPTY) {
+		dev_err(tspi->dev, "Rx/Tx fifo are not empty status 0x%08x\n",
+			(unsigned)status);
+		return -EIO;
+	}
+
+	val = SLINK_DMA_BLOCK_SIZE(tspi->curr_dma_words - 1);
+	val |= tspi->packed_size;
+	if (tspi->is_packed)
+		len = DIV_ROUND_UP(tspi->curr_dma_words * tspi->bytes_per_word,
+					4) * 4;
+	else
+		len = tspi->curr_dma_words * 4;
+
+	/* Set attention level based on length of transfer */
+	if (len & 0xF)
+		val |= SLINK_TX_TRIG_1 | SLINK_RX_TRIG_1;
+	else if (((len) >> 4) & 0x1)
+		val |= SLINK_TX_TRIG_4 | SLINK_RX_TRIG_4;
+	else
+		val |= SLINK_TX_TRIG_8 | SLINK_RX_TRIG_8;
+
+	if (tspi->cur_direction & DATA_DIR_TX)
+		val |= SLINK_IE_TXC;
+
+	if (tspi->cur_direction & DATA_DIR_RX)
+		val |= SLINK_IE_RXC;
+
+	tegra_slink_writel(tspi, val, SLINK_DMA_CTL);
+	tspi->dma_control_reg = val;
+
+	if (tspi->cur_direction & DATA_DIR_TX) {
+		tegra_slink_copy_client_txbuf_to_spi_txbuf(tspi, t);
+		wmb();
+		ret = tegra_slink_start_tx_dma(tspi, len);
+		if (ret < 0) {
+			dev_err(tspi->dev,
+				"Starting tx dma failed, err %d\n", ret);
+			return ret;
+		}
+
+		/* Wait for tx fifo to be fill before starting slink */
+		status = tegra_slink_readl(tspi, SLINK_STATUS);
+		while (!(status & SLINK_TX_FULL))
+			status = tegra_slink_readl(tspi, SLINK_STATUS);
+	}
+
+	if (tspi->cur_direction & DATA_DIR_RX) {
+		/* Make the dma buffer to read by dma */
+		dma_sync_single_for_device(tspi->dev, tspi->rx_dma_phys,
+				tspi->dma_buf_size, DMA_FROM_DEVICE);
+
+		ret = tegra_slink_start_rx_dma(tspi, len);
+		if (ret < 0) {
+			dev_err(tspi->dev,
+				"Starting rx dma failed, err %d\n", ret);
+			if (tspi->cur_direction & DATA_DIR_TX)
+				dmaengine_terminate_all(tspi->tx_dma_chan);
+			return ret;
+		}
+	}
+	tspi->is_curr_dma_xfer = true;
+	if (tspi->is_packed) {
+		val |= SLINK_PACKED;
+		tegra_slink_writel(tspi, val, SLINK_DMA_CTL);
+		/* HW need small delay after settign Packed mode */
+		udelay(1);
+	}
+	tspi->dma_control_reg = val;
+
+	val |= SLINK_DMA_EN;
+	tegra_slink_writel(tspi, val, SLINK_DMA_CTL);
+	return ret;
+}
+
+static int tegra_slink_start_cpu_based_transfer(
+		struct tegra_slink_data *tspi, struct spi_transfer *t)
+{
+	u32 val;
+	unsigned cur_words;
+
+	val = tspi->packed_size;
+	if (tspi->cur_direction & DATA_DIR_TX)
+		val |= SLINK_IE_TXC;
+
+	if (tspi->cur_direction & DATA_DIR_RX)
+		val |= SLINK_IE_RXC;
+
+	tegra_slink_writel(tspi, val, SLINK_DMA_CTL);
+	tspi->dma_control_reg = val;
+
+	if (tspi->cur_direction & DATA_DIR_TX)
+		cur_words = tegra_slink_fill_tx_fifo_from_client_txbuf(tspi, t);
+	else
+		cur_words = tspi->curr_dma_words;
+	val |= SLINK_DMA_BLOCK_SIZE(cur_words - 1);
+	tegra_slink_writel(tspi, val, SLINK_DMA_CTL);
+	tspi->dma_control_reg = val;
+
+	tspi->is_curr_dma_xfer = false;
+	if (tspi->is_packed) {
+		val |= SLINK_PACKED;
+		tegra_slink_writel(tspi, val, SLINK_DMA_CTL);
+		udelay(1);
+		wmb();
+	}
+	tspi->dma_control_reg = val;
+	val |= SLINK_DMA_EN;
+	tegra_slink_writel(tspi, val, SLINK_DMA_CTL);
+	return 0;
+}
+
+static int tegra_slink_init_dma_param(struct tegra_slink_data *tspi,
+			bool dma_to_memory)
+{
+	struct dma_chan *dma_chan;
+	u32 *dma_buf;
+	dma_addr_t dma_phys;
+	int ret;
+	struct dma_slave_config dma_sconfig;
+
+	dma_chan = dma_request_slave_channel_reason(tspi->dev,
+						dma_to_memory ? "rx" : "tx");
+	if (IS_ERR(dma_chan)) {
+		ret = PTR_ERR(dma_chan);
+		if (ret != -EPROBE_DEFER)
+			dev_err(tspi->dev,
+				"Dma channel is not available: %d\n", ret);
+		return ret;
+	}
+
+	dma_buf = dma_alloc_coherent(tspi->dev, tspi->dma_buf_size,
+				&dma_phys, GFP_KERNEL);
+	if (!dma_buf) {
+		dev_err(tspi->dev, " Not able to allocate the dma buffer\n");
+		dma_release_channel(dma_chan);
+		return -ENOMEM;
+	}
+
+	if (dma_to_memory) {
+		dma_sconfig.src_addr = tspi->phys + SLINK_RX_FIFO;
+		dma_sconfig.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		dma_sconfig.src_maxburst = 0;
+	} else {
+		dma_sconfig.dst_addr = tspi->phys + SLINK_TX_FIFO;
+		dma_sconfig.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		dma_sconfig.dst_maxburst = 0;
+	}
+
+	ret = dmaengine_slave_config(dma_chan, &dma_sconfig);
+	if (ret)
+		goto scrub;
+	if (dma_to_memory) {
+		tspi->rx_dma_chan = dma_chan;
+		tspi->rx_dma_buf = dma_buf;
+		tspi->rx_dma_phys = dma_phys;
+	} else {
+		tspi->tx_dma_chan = dma_chan;
+		tspi->tx_dma_buf = dma_buf;
+		tspi->tx_dma_phys = dma_phys;
+	}
+	return 0;
+
+scrub:
+	dma_free_coherent(tspi->dev, tspi->dma_buf_size, dma_buf, dma_phys);
+	dma_release_channel(dma_chan);
+	return ret;
+}
+
+static void tegra_slink_deinit_dma_param(struct tegra_slink_data *tspi,
+	bool dma_to_memory)
+{
+	u32 *dma_buf;
+	dma_addr_t dma_phys;
+	struct dma_chan *dma_chan;
+
+	if (dma_to_memory) {
+		dma_buf = tspi->rx_dma_buf;
+		dma_chan = tspi->rx_dma_chan;
+		dma_phys = tspi->rx_dma_phys;
+		tspi->rx_dma_chan = NULL;
+		tspi->rx_dma_buf = NULL;
+	} else {
+		dma_buf = tspi->tx_dma_buf;
+		dma_chan = tspi->tx_dma_chan;
+		dma_phys = tspi->tx_dma_phys;
+		tspi->tx_dma_buf = NULL;
+		tspi->tx_dma_chan = NULL;
+	}
+	if (!dma_chan)
+		return;
+
+	dma_free_coherent(tspi->dev, tspi->dma_buf_size, dma_buf, dma_phys);
+	dma_release_channel(dma_chan);
+}
+
+static int tegra_slink_start_transfer_one(struct spi_device *spi,
+		struct spi_transfer *t)
+{
+	struct tegra_slink_data *tspi = spi_master_get_devdata(spi->master);
+	u32 speed;
+	u8 bits_per_word;
+	unsigned total_fifo_words;
+	int ret;
+	u32 command;
+	u32 command2;
+
+	bits_per_word = t->bits_per_word;
+	speed = t->speed_hz;
+	if (speed != tspi->cur_speed) {
+		clk_set_rate(tspi->clk, speed * 4);
+		tspi->cur_speed = speed;
+	}
+
+	tspi->cur_spi = spi;
+	tspi->cur_pos = 0;
+	tspi->cur_rx_pos = 0;
+	tspi->cur_tx_pos = 0;
+	tspi->curr_xfer = t;
+	total_fifo_words = tegra_slink_calculate_curr_xfer_param(spi, tspi, t);
+
+	command = tspi->command_reg;
+	command &= ~SLINK_BIT_LENGTH(~0);
+	command |= SLINK_BIT_LENGTH(bits_per_word - 1);
+
+	command2 = tspi->command2_reg;
+	command2 &= ~(SLINK_RXEN | SLINK_TXEN);
+
+	tegra_slink_writel(tspi, command, SLINK_COMMAND);
+	tspi->command_reg = command;
+
+	tspi->cur_direction = 0;
+	if (t->rx_buf) {
+		command2 |= SLINK_RXEN;
+		tspi->cur_direction |= DATA_DIR_RX;
+	}
+	if (t->tx_buf) {
+		command2 |= SLINK_TXEN;
+		tspi->cur_direction |= DATA_DIR_TX;
+	}
+	tegra_slink_writel(tspi, command2, SLINK_COMMAND2);
+	tspi->command2_reg = command2;
+
+	if (total_fifo_words > SLINK_FIFO_DEPTH)
+		ret = tegra_slink_start_dma_based_transfer(tspi, t);
+	else
+		ret = tegra_slink_start_cpu_based_transfer(tspi, t);
+	return ret;
+}
+
+static int tegra_slink_setup(struct spi_device *spi)
+{
+	static const u32 cs_pol_bit[MAX_CHIP_SELECT] = {
+			SLINK_CS_POLARITY,
+			SLINK_CS_POLARITY1,
+			SLINK_CS_POLARITY2,
+			SLINK_CS_POLARITY3,
+	};
+
+	struct tegra_slink_data *tspi = spi_master_get_devdata(spi->master);
+	u32 val;
+	unsigned long flags;
+	int ret;
+
+	dev_dbg(&spi->dev, "setup %d bpw, %scpol, %scpha, %dHz\n",
+		spi->bits_per_word,
+		spi->mode & SPI_CPOL ? "" : "~",
+		spi->mode & SPI_CPHA ? "" : "~",
+		spi->max_speed_hz);
+
+	ret = pm_runtime_get_sync(tspi->dev);
+	if (ret < 0) {
+		dev_err(tspi->dev, "pm runtime failed, e = %d\n", ret);
+		return ret;
+	}
+
+	spin_lock_irqsave(&tspi->lock, flags);
+	val = tspi->def_command_reg;
+	if (spi->mode & SPI_CS_HIGH)
+		val |= cs_pol_bit[spi->chip_select];
+	else
+		val &= ~cs_pol_bit[spi->chip_select];
+	tspi->def_command_reg = val;
+	tegra_slink_writel(tspi, tspi->def_command_reg, SLINK_COMMAND);
+	spin_unlock_irqrestore(&tspi->lock, flags);
+
+	pm_runtime_put(tspi->dev);
+	return 0;
+}
+
+static int tegra_slink_prepare_message(struct spi_master *master,
+				       struct spi_message *msg)
+{
+	struct tegra_slink_data *tspi = spi_master_get_devdata(master);
+	struct spi_device *spi = msg->spi;
+
+	tegra_slink_clear_status(tspi);
+
+	tspi->command_reg = tspi->def_command_reg;
+	tspi->command_reg |= SLINK_CS_SW | SLINK_CS_VALUE;
+
+	tspi->command2_reg = tspi->def_command2_reg;
+	tspi->command2_reg |= SLINK_SS_EN_CS(spi->chip_select);
+
+	tspi->command_reg &= ~SLINK_MODES;
+	if (spi->mode & SPI_CPHA)
+		tspi->command_reg |= SLINK_CK_SDA;
+
+	if (spi->mode & SPI_CPOL)
+		tspi->command_reg |= SLINK_IDLE_SCLK_DRIVE_HIGH;
+	else
+		tspi->command_reg |= SLINK_IDLE_SCLK_DRIVE_LOW;
+
+	return 0;
+}
+
+static int tegra_slink_transfer_one(struct spi_master *master,
+				    struct spi_device *spi,
+				    struct spi_transfer *xfer)
+{
+	struct tegra_slink_data *tspi = spi_master_get_devdata(master);
+	int ret;
+
+	reinit_completion(&tspi->xfer_completion);
+	ret = tegra_slink_start_transfer_one(spi, xfer);
+	if (ret < 0) {
+		dev_err(tspi->dev,
+			"spi can not start transfer, err %d\n", ret);
+		return ret;
+	}
+
+	ret = wait_for_completion_timeout(&tspi->xfer_completion,
+					  SLINK_DMA_TIMEOUT);
+	if (WARN_ON(ret == 0)) {
+		dev_err(tspi->dev,
+			"spi transfer timeout, err %d\n", ret);
+		return -EIO;
+	}
+
+	if (tspi->tx_status)
+		return tspi->tx_status;
+	if (tspi->rx_status)
+		return tspi->rx_status;
+
+	return 0;
+}
+
+static int tegra_slink_unprepare_message(struct spi_master *master,
+					 struct spi_message *msg)
+{
+	struct tegra_slink_data *tspi = spi_master_get_devdata(master);
+
+	tegra_slink_writel(tspi, tspi->def_command_reg, SLINK_COMMAND);
+	tegra_slink_writel(tspi, tspi->def_command2_reg, SLINK_COMMAND2);
+
+	return 0;
+}
+
+static irqreturn_t handle_cpu_based_xfer(struct tegra_slink_data *tspi)
+{
+	struct spi_transfer *t = tspi->curr_xfer;
+	unsigned long flags;
+
+	spin_lock_irqsave(&tspi->lock, flags);
+	if (tspi->tx_status ||  tspi->rx_status ||
+				(tspi->status_reg & SLINK_BSY)) {
+		dev_err(tspi->dev,
+			"CpuXfer ERROR bit set 0x%x\n", tspi->status_reg);
+		dev_err(tspi->dev,
+			"CpuXfer 0x%08x:0x%08x:0x%08x\n", tspi->command_reg,
+				tspi->command2_reg, tspi->dma_control_reg);
+		reset_control_assert(tspi->rst);
+		udelay(2);
+		reset_control_deassert(tspi->rst);
+		complete(&tspi->xfer_completion);
+		goto exit;
+	}
+
+	if (tspi->cur_direction & DATA_DIR_RX)
+		tegra_slink_read_rx_fifo_to_client_rxbuf(tspi, t);
+
+	if (tspi->cur_direction & DATA_DIR_TX)
+		tspi->cur_pos = tspi->cur_tx_pos;
+	else
+		tspi->cur_pos = tspi->cur_rx_pos;
+
+	if (tspi->cur_pos == t->len) {
+		complete(&tspi->xfer_completion);
+		goto exit;
+	}
+
+	tegra_slink_calculate_curr_xfer_param(tspi->cur_spi, tspi, t);
+	tegra_slink_start_cpu_based_transfer(tspi, t);
+exit:
+	spin_unlock_irqrestore(&tspi->lock, flags);
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t handle_dma_based_xfer(struct tegra_slink_data *tspi)
+{
+	struct spi_transfer *t = tspi->curr_xfer;
+	long wait_status;
+	int err = 0;
+	unsigned total_fifo_words;
+	unsigned long flags;
+
+	/* Abort dmas if any error */
+	if (tspi->cur_direction & DATA_DIR_TX) {
+		if (tspi->tx_status) {
+			dmaengine_terminate_all(tspi->tx_dma_chan);
+			err += 1;
+		} else {
+			wait_status = wait_for_completion_interruptible_timeout(
+				&tspi->tx_dma_complete, SLINK_DMA_TIMEOUT);
+			if (wait_status <= 0) {
+				dmaengine_terminate_all(tspi->tx_dma_chan);
+				dev_err(tspi->dev, "TxDma Xfer failed\n");
+				err += 1;
+			}
+		}
+	}
+
+	if (tspi->cur_direction & DATA_DIR_RX) {
+		if (tspi->rx_status) {
+			dmaengine_terminate_all(tspi->rx_dma_chan);
+			err += 2;
+		} else {
+			wait_status = wait_for_completion_interruptible_timeout(
+				&tspi->rx_dma_complete, SLINK_DMA_TIMEOUT);
+			if (wait_status <= 0) {
+				dmaengine_terminate_all(tspi->rx_dma_chan);
+				dev_err(tspi->dev, "RxDma Xfer failed\n");
+				err += 2;
+			}
+		}
+	}
+
+	spin_lock_irqsave(&tspi->lock, flags);
+	if (err) {
+		dev_err(tspi->dev,
+			"DmaXfer: ERROR bit set 0x%x\n", tspi->status_reg);
+		dev_err(tspi->dev,
+			"DmaXfer 0x%08x:0x%08x:0x%08x\n", tspi->command_reg,
+				tspi->command2_reg, tspi->dma_control_reg);
+		reset_control_assert(tspi->rst);
+		udelay(2);
+		reset_control_assert(tspi->rst);
+		complete(&tspi->xfer_completion);
+		spin_unlock_irqrestore(&tspi->lock, flags);
+		return IRQ_HANDLED;
+	}
+
+	if (tspi->cur_direction & DATA_DIR_RX)
+		tegra_slink_copy_spi_rxbuf_to_client_rxbuf(tspi, t);
+
+	if (tspi->cur_direction & DATA_DIR_TX)
+		tspi->cur_pos = tspi->cur_tx_pos;
+	else
+		tspi->cur_pos = tspi->cur_rx_pos;
+
+	if (tspi->cur_pos == t->len) {
+		complete(&tspi->xfer_completion);
+		goto exit;
+	}
+
+	/* Continue transfer in current message */
+	total_fifo_words = tegra_slink_calculate_curr_xfer_param(tspi->cur_spi,
+							tspi, t);
+	if (total_fifo_words > SLINK_FIFO_DEPTH)
+		err = tegra_slink_start_dma_based_transfer(tspi, t);
+	else
+		err = tegra_slink_start_cpu_based_transfer(tspi, t);
+
+exit:
+	spin_unlock_irqrestore(&tspi->lock, flags);
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t tegra_slink_isr_thread(int irq, void *context_data)
+{
+	struct tegra_slink_data *tspi = context_data;
+
+	if (!tspi->is_curr_dma_xfer)
+		return handle_cpu_based_xfer(tspi);
+	return handle_dma_based_xfer(tspi);
+}
+
+static irqreturn_t tegra_slink_isr(int irq, void *context_data)
+{
+	struct tegra_slink_data *tspi = context_data;
+
+	tspi->status_reg = tegra_slink_readl(tspi, SLINK_STATUS);
+	if (tspi->cur_direction & DATA_DIR_TX)
+		tspi->tx_status = tspi->status_reg &
+					(SLINK_TX_OVF | SLINK_TX_UNF);
+
+	if (tspi->cur_direction & DATA_DIR_RX)
+		tspi->rx_status = tspi->status_reg &
+					(SLINK_RX_OVF | SLINK_RX_UNF);
+	tegra_slink_clear_status(tspi);
+
+	return IRQ_WAKE_THREAD;
+}
+
+static const struct tegra_slink_chip_data tegra30_spi_cdata = {
+	.cs_hold_time = true,
+};
+
+static const struct tegra_slink_chip_data tegra20_spi_cdata = {
+	.cs_hold_time = false,
+};
+
+static const struct of_device_id tegra_slink_of_match[] = {
+	{ .compatible = "nvidia,tegra30-slink", .data = &tegra30_spi_cdata, },
+	{ .compatible = "nvidia,tegra20-slink", .data = &tegra20_spi_cdata, },
+	{}
+};
+MODULE_DEVICE_TABLE(of, tegra_slink_of_match);
+
+static int tegra_slink_probe(struct platform_device *pdev)
+{
+	struct spi_master	*master;
+	struct tegra_slink_data	*tspi;
+	struct resource		*r;
+	int ret, spi_irq;
+	const struct tegra_slink_chip_data *cdata = NULL;
+	const struct of_device_id *match;
+
+	match = of_match_device(tegra_slink_of_match, &pdev->dev);
+	if (!match) {
+		dev_err(&pdev->dev, "Error: No device match found\n");
+		return -ENODEV;
+	}
+	cdata = match->data;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*tspi));
+	if (!master) {
+		dev_err(&pdev->dev, "master allocation failed\n");
+		return -ENOMEM;
+	}
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->setup = tegra_slink_setup;
+	master->prepare_message = tegra_slink_prepare_message;
+	master->transfer_one = tegra_slink_transfer_one;
+	master->unprepare_message = tegra_slink_unprepare_message;
+	master->auto_runtime_pm = true;
+	master->num_chipselect = MAX_CHIP_SELECT;
+
+	platform_set_drvdata(pdev, master);
+	tspi = spi_master_get_devdata(master);
+	tspi->master = master;
+	tspi->dev = &pdev->dev;
+	tspi->chip_data = cdata;
+	spin_lock_init(&tspi->lock);
+
+	if (of_property_read_u32(tspi->dev->of_node, "spi-max-frequency",
+				 &master->max_speed_hz))
+		master->max_speed_hz = 25000000; /* 25MHz */
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!r) {
+		dev_err(&pdev->dev, "No IO memory resource\n");
+		ret = -ENODEV;
+		goto exit_free_master;
+	}
+	tspi->phys = r->start;
+	tspi->base = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(tspi->base)) {
+		ret = PTR_ERR(tspi->base);
+		goto exit_free_master;
+	}
+
+	/* disabled clock may cause interrupt storm upon request */
+	tspi->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(tspi->clk)) {
+		ret = PTR_ERR(tspi->clk);
+		dev_err(&pdev->dev, "Can not get clock %d\n", ret);
+		goto exit_free_master;
+	}
+	ret = clk_prepare(tspi->clk);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Clock prepare failed %d\n", ret);
+		goto exit_free_master;
+	}
+	ret = clk_enable(tspi->clk);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Clock enable failed %d\n", ret);
+		goto exit_free_master;
+	}
+
+	spi_irq = platform_get_irq(pdev, 0);
+	tspi->irq = spi_irq;
+	ret = request_threaded_irq(tspi->irq, tegra_slink_isr,
+			tegra_slink_isr_thread, IRQF_ONESHOT,
+			dev_name(&pdev->dev), tspi);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to register ISR for IRQ %d\n",
+					tspi->irq);
+		goto exit_clk_disable;
+	}
+
+	tspi->rst = devm_reset_control_get_exclusive(&pdev->dev, "spi");
+	if (IS_ERR(tspi->rst)) {
+		dev_err(&pdev->dev, "can not get reset\n");
+		ret = PTR_ERR(tspi->rst);
+		goto exit_free_irq;
+	}
+
+	tspi->max_buf_size = SLINK_FIFO_DEPTH << 2;
+	tspi->dma_buf_size = DEFAULT_SPI_DMA_BUF_LEN;
+
+	ret = tegra_slink_init_dma_param(tspi, true);
+	if (ret < 0)
+		goto exit_free_irq;
+	ret = tegra_slink_init_dma_param(tspi, false);
+	if (ret < 0)
+		goto exit_rx_dma_free;
+	tspi->max_buf_size = tspi->dma_buf_size;
+	init_completion(&tspi->tx_dma_complete);
+	init_completion(&tspi->rx_dma_complete);
+
+	init_completion(&tspi->xfer_completion);
+
+	pm_runtime_enable(&pdev->dev);
+	if (!pm_runtime_enabled(&pdev->dev)) {
+		ret = tegra_slink_runtime_resume(&pdev->dev);
+		if (ret)
+			goto exit_pm_disable;
+	}
+
+	ret = pm_runtime_get_sync(&pdev->dev);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "pm runtime get failed, e = %d\n", ret);
+		goto exit_pm_disable;
+	}
+	tspi->def_command_reg  = SLINK_M_S;
+	tspi->def_command2_reg = SLINK_CS_ACTIVE_BETWEEN;
+	tegra_slink_writel(tspi, tspi->def_command_reg, SLINK_COMMAND);
+	tegra_slink_writel(tspi, tspi->def_command2_reg, SLINK_COMMAND2);
+	pm_runtime_put(&pdev->dev);
+
+	master->dev.of_node = pdev->dev.of_node;
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "can not register to master err %d\n", ret);
+		goto exit_pm_disable;
+	}
+	return ret;
+
+exit_pm_disable:
+	pm_runtime_disable(&pdev->dev);
+	if (!pm_runtime_status_suspended(&pdev->dev))
+		tegra_slink_runtime_suspend(&pdev->dev);
+	tegra_slink_deinit_dma_param(tspi, false);
+exit_rx_dma_free:
+	tegra_slink_deinit_dma_param(tspi, true);
+exit_free_irq:
+	free_irq(spi_irq, tspi);
+exit_clk_disable:
+	clk_disable(tspi->clk);
+exit_free_master:
+	spi_master_put(master);
+	return ret;
+}
+
+static int tegra_slink_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct tegra_slink_data	*tspi = spi_master_get_devdata(master);
+
+	free_irq(tspi->irq, tspi);
+
+	clk_disable(tspi->clk);
+
+	if (tspi->tx_dma_chan)
+		tegra_slink_deinit_dma_param(tspi, false);
+
+	if (tspi->rx_dma_chan)
+		tegra_slink_deinit_dma_param(tspi, true);
+
+	pm_runtime_disable(&pdev->dev);
+	if (!pm_runtime_status_suspended(&pdev->dev))
+		tegra_slink_runtime_suspend(&pdev->dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int tegra_slink_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+
+	return spi_master_suspend(master);
+}
+
+static int tegra_slink_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct tegra_slink_data *tspi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = pm_runtime_get_sync(dev);
+	if (ret < 0) {
+		dev_err(dev, "pm runtime failed, e = %d\n", ret);
+		return ret;
+	}
+	tegra_slink_writel(tspi, tspi->command_reg, SLINK_COMMAND);
+	tegra_slink_writel(tspi, tspi->command2_reg, SLINK_COMMAND2);
+	pm_runtime_put(dev);
+
+	return spi_master_resume(master);
+}
+#endif
+
+static int tegra_slink_runtime_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct tegra_slink_data *tspi = spi_master_get_devdata(master);
+
+	/* Flush all write which are in PPSB queue by reading back */
+	tegra_slink_readl(tspi, SLINK_MAS_DATA);
+
+	clk_disable_unprepare(tspi->clk);
+	return 0;
+}
+
+static int tegra_slink_runtime_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct tegra_slink_data *tspi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_prepare_enable(tspi->clk);
+	if (ret < 0) {
+		dev_err(tspi->dev, "clk_prepare failed: %d\n", ret);
+		return ret;
+	}
+	return 0;
+}
+
+static const struct dev_pm_ops slink_pm_ops = {
+	SET_RUNTIME_PM_OPS(tegra_slink_runtime_suspend,
+		tegra_slink_runtime_resume, NULL)
+	SET_SYSTEM_SLEEP_PM_OPS(tegra_slink_suspend, tegra_slink_resume)
+};
+static struct platform_driver tegra_slink_driver = {
+	.driver = {
+		.name		= "spi-tegra-slink",
+		.pm		= &slink_pm_ops,
+		.of_match_table	= tegra_slink_of_match,
+	},
+	.probe =	tegra_slink_probe,
+	.remove =	tegra_slink_remove,
+};
+module_platform_driver(tegra_slink_driver);
+
+MODULE_ALIAS("platform:spi-tegra-slink");
+MODULE_DESCRIPTION("NVIDIA Tegra20/Tegra30 SLINK Controller Driver");
+MODULE_AUTHOR("Laxman Dewangan <ldewangan@nvidia.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-test.h b/drivers/spi/spi-test.h
new file mode 100644
index 0000000..6ed7b89
--- /dev/null
+++ b/drivers/spi/spi-test.h
@@ -0,0 +1,137 @@
+/*
+ *  linux/drivers/spi/spi-test.h
+ *
+ *  (c) Martin Sperl <kernel@martin.sperl.org>
+ *
+ *  spi_test definitions
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ */
+
+#include <linux/spi/spi.h>
+
+#define SPI_TEST_MAX_TRANSFERS 4
+#define SPI_TEST_MAX_SIZE (32 * PAGE_SIZE)
+#define SPI_TEST_MAX_ITERATE 32
+
+/* the "dummy" start addresses used in spi_test
+ * these addresses get translated at a later stage
+ */
+#define RX_START	BIT(30)
+#define TX_START	BIT(31)
+#define RX(off)		((void *)(RX_START + off))
+#define TX(off)		((void *)(TX_START + off))
+
+/* some special defines for offsets */
+#define SPI_TEST_MAX_SIZE_HALF	BIT(29)
+
+/* detection pattern for unfinished reads...
+ * - 0x00 or 0xff could be valid levels for tx_buf = NULL,
+ * so we do not use either of them
+ */
+#define SPI_TEST_PATTERN_UNWRITTEN 0xAA
+#define SPI_TEST_PATTERN_DO_NOT_WRITE 0x55
+#define SPI_TEST_CHECK_DO_NOT_WRITE 64
+
+/**
+ * struct spi_test - describes a specific (set of) tests to execute
+ *
+ * @description:      description of the test
+ *
+ * @msg:              a template @spi_message usedfor the default settings
+ * @transfers:        array of @spi_transfers that are part of the
+ *                    resulting spi_message.
+ * @transfer_count:   number of transfers
+ *
+ * @run_test:         run a specific spi_test - this allows to override
+ *                    the default implementation of @spi_test_run_transfer
+ *                    either to add some custom filters for a specific test
+ *                    or to effectively run some very custom tests...
+ * @execute_msg:      run the spi_message for real - this allows to override
+ *                    @spi_test_execute_msg to apply final modifications
+ *                    on the spi_message
+ * @expected_return:  the expected return code - in some cases we want to
+ *                    test also for error conditions
+ *
+ * @iterate_len:      list of length to iterate on
+ * @iterate_tx_align: change the alignment of @spi_transfer.tx_buf
+ *                    for all values in the below range if set.
+ *                    the ranges are:
+ *                    [0 : @spi_master.dma_alignment[ if set
+ *                    [0 : iterate_tx_align[ if unset
+ * @iterate_rx_align: change the alignment of @spi_transfer.rx_buf
+ *                    see @iterate_tx_align for details
+ * @iterate_transfer_mask: the bitmask of transfers to which the iterations
+ *                         apply - if 0, then it applies to all transfer
+ *
+ * @fill_option:      define the way how tx_buf is filled
+ * @fill_pattern:     fill pattern to apply to the tx_buf
+ *                    (used in some of the @fill_options)
+ * @elapsed_time:     elapsed time in nanoseconds
+ */
+
+struct spi_test {
+	char description[64];
+	struct spi_message msg;
+	struct spi_transfer transfers[SPI_TEST_MAX_TRANSFERS];
+	unsigned int transfer_count;
+	int (*run_test)(struct spi_device *spi, struct spi_test *test,
+			void *tx, void *rx);
+	int (*execute_msg)(struct spi_device *spi, struct spi_test *test,
+			   void *tx, void *rx);
+	int expected_return;
+	/* iterate over all values, terminated by a -1 */
+	int iterate_len[SPI_TEST_MAX_ITERATE];
+	int iterate_tx_align;
+	int iterate_rx_align;
+	u32 iterate_transfer_mask;
+	/* the tx-fill operation */
+	u32 fill_option;
+#define FILL_MEMSET_8	0	/* just memset with 8 bit */
+#define FILL_MEMSET_16	1	/* just memset with 16 bit */
+#define FILL_MEMSET_24	2	/* just memset with 24 bit */
+#define FILL_MEMSET_32	3	/* just memset with 32 bit */
+#define FILL_COUNT_8	4	/* fill with a 8 byte counter */
+#define FILL_COUNT_16	5	/* fill with a 16 bit counter */
+#define FILL_COUNT_24	6	/* fill with a 24 bit counter */
+#define FILL_COUNT_32	7	/* fill with a 32 bit counter */
+#define FILL_TRANSFER_BYTE_8  8	/* fill with the transfer byte - 8 bit */
+#define FILL_TRANSFER_BYTE_16 9	/* fill with the transfer byte - 16 bit */
+#define FILL_TRANSFER_BYTE_24 10 /* fill with the transfer byte - 24 bit */
+#define FILL_TRANSFER_BYTE_32 11 /* fill with the transfer byte - 32 bit */
+#define FILL_TRANSFER_NUM     16 /* fill with the transfer number */
+	u32 fill_pattern;
+	unsigned long long elapsed_time;
+};
+
+/* default implementation for @spi_test.run_test */
+int spi_test_run_test(struct spi_device *spi,
+		      const struct spi_test *test,
+		      void *tx, void *rx);
+
+/* default implementation for @spi_test.execute_msg */
+int spi_test_execute_msg(struct spi_device *spi,
+			 struct spi_test *test,
+			 void *tx, void *rx);
+
+/* function to execute a set of tests */
+int spi_test_run_tests(struct spi_device *spi,
+		       struct spi_test *tests);
+
+#define ITERATE_LEN_LIST 0, 1, 2, 3, 7, 11, 16, 31, 32, 64, 97, 128, 251, 256, \
+		1021, 1024, 1031, 4093, PAGE_SIZE, 4099, 65536, 65537
+/* some of the default @spi_transfer.len to test, terminated by a -1 */
+#define ITERATE_LEN ITERATE_LEN_LIST, -1
+#define ITERATE_MAX_LEN ITERATE_LEN_LIST, (SPI_TEST_MAX_SIZE - 1), \
+		SPI_TEST_MAX_SIZE, -1
+
+/* the default alignment to test */
+#define ITERATE_ALIGN sizeof(int)
diff --git a/drivers/spi/spi-ti-qspi.c b/drivers/spi/spi-ti-qspi.c
new file mode 100644
index 0000000..5f19016
--- /dev/null
+++ b/drivers/spi/spi-ti-qspi.c
@@ -0,0 +1,862 @@
+/*
+ * TI QSPI driver
+ *
+ * Copyright (C) 2013 Texas Instruments Incorporated - http://www.ti.com
+ * Author: Sourav Poddar <sourav.poddar@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GPLv2.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR /PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/omap-dma.h>
+#include <linux/platform_device.h>
+#include <linux/err.h>
+#include <linux/clk.h>
+#include <linux/io.h>
+#include <linux/slab.h>
+#include <linux/pm_runtime.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/mfd/syscon.h>
+#include <linux/regmap.h>
+#include <linux/sizes.h>
+
+#include <linux/spi/spi.h>
+#include <linux/spi/spi-mem.h>
+
+struct ti_qspi_regs {
+	u32 clkctrl;
+};
+
+struct ti_qspi {
+	struct completion	transfer_complete;
+
+	/* list synchronization */
+	struct mutex            list_lock;
+
+	struct spi_master	*master;
+	void __iomem            *base;
+	void __iomem            *mmap_base;
+	size_t			mmap_size;
+	struct regmap		*ctrl_base;
+	unsigned int		ctrl_reg;
+	struct clk		*fclk;
+	struct device           *dev;
+
+	struct ti_qspi_regs     ctx_reg;
+
+	dma_addr_t		mmap_phys_base;
+	dma_addr_t		rx_bb_dma_addr;
+	void			*rx_bb_addr;
+	struct dma_chan		*rx_chan;
+
+	u32 spi_max_frequency;
+	u32 cmd;
+	u32 dc;
+
+	bool mmap_enabled;
+};
+
+#define QSPI_PID			(0x0)
+#define QSPI_SYSCONFIG			(0x10)
+#define QSPI_SPI_CLOCK_CNTRL_REG	(0x40)
+#define QSPI_SPI_DC_REG			(0x44)
+#define QSPI_SPI_CMD_REG		(0x48)
+#define QSPI_SPI_STATUS_REG		(0x4c)
+#define QSPI_SPI_DATA_REG		(0x50)
+#define QSPI_SPI_SETUP_REG(n)		((0x54 + 4 * n))
+#define QSPI_SPI_SWITCH_REG		(0x64)
+#define QSPI_SPI_DATA_REG_1		(0x68)
+#define QSPI_SPI_DATA_REG_2		(0x6c)
+#define QSPI_SPI_DATA_REG_3		(0x70)
+
+#define QSPI_COMPLETION_TIMEOUT		msecs_to_jiffies(2000)
+
+#define QSPI_FCLK			192000000
+
+/* Clock Control */
+#define QSPI_CLK_EN			(1 << 31)
+#define QSPI_CLK_DIV_MAX		0xffff
+
+/* Command */
+#define QSPI_EN_CS(n)			(n << 28)
+#define QSPI_WLEN(n)			((n - 1) << 19)
+#define QSPI_3_PIN			(1 << 18)
+#define QSPI_RD_SNGL			(1 << 16)
+#define QSPI_WR_SNGL			(2 << 16)
+#define QSPI_RD_DUAL			(3 << 16)
+#define QSPI_RD_QUAD			(7 << 16)
+#define QSPI_INVAL			(4 << 16)
+#define QSPI_FLEN(n)			((n - 1) << 0)
+#define QSPI_WLEN_MAX_BITS		128
+#define QSPI_WLEN_MAX_BYTES		16
+#define QSPI_WLEN_MASK			QSPI_WLEN(QSPI_WLEN_MAX_BITS)
+
+/* STATUS REGISTER */
+#define BUSY				0x01
+#define WC				0x02
+
+/* Device Control */
+#define QSPI_DD(m, n)			(m << (3 + n * 8))
+#define QSPI_CKPHA(n)			(1 << (2 + n * 8))
+#define QSPI_CSPOL(n)			(1 << (1 + n * 8))
+#define QSPI_CKPOL(n)			(1 << (n * 8))
+
+#define	QSPI_FRAME			4096
+
+#define QSPI_AUTOSUSPEND_TIMEOUT         2000
+
+#define MEM_CS_EN(n)			((n + 1) << 8)
+#define MEM_CS_MASK			(7 << 8)
+
+#define MM_SWITCH			0x1
+
+#define QSPI_SETUP_RD_NORMAL		(0x0 << 12)
+#define QSPI_SETUP_RD_DUAL		(0x1 << 12)
+#define QSPI_SETUP_RD_QUAD		(0x3 << 12)
+#define QSPI_SETUP_ADDR_SHIFT		8
+#define QSPI_SETUP_DUMMY_SHIFT		10
+
+#define QSPI_DMA_BUFFER_SIZE            SZ_64K
+
+static inline unsigned long ti_qspi_read(struct ti_qspi *qspi,
+		unsigned long reg)
+{
+	return readl(qspi->base + reg);
+}
+
+static inline void ti_qspi_write(struct ti_qspi *qspi,
+		unsigned long val, unsigned long reg)
+{
+	writel(val, qspi->base + reg);
+}
+
+static int ti_qspi_setup(struct spi_device *spi)
+{
+	struct ti_qspi	*qspi = spi_master_get_devdata(spi->master);
+	struct ti_qspi_regs *ctx_reg = &qspi->ctx_reg;
+	int clk_div = 0, ret;
+	u32 clk_ctrl_reg, clk_rate, clk_mask;
+
+	if (spi->master->busy) {
+		dev_dbg(qspi->dev, "master busy doing other transfers\n");
+		return -EBUSY;
+	}
+
+	if (!qspi->spi_max_frequency) {
+		dev_err(qspi->dev, "spi max frequency not defined\n");
+		return -EINVAL;
+	}
+
+	clk_rate = clk_get_rate(qspi->fclk);
+
+	clk_div = DIV_ROUND_UP(clk_rate, qspi->spi_max_frequency) - 1;
+
+	if (clk_div < 0) {
+		dev_dbg(qspi->dev, "clock divider < 0, using /1 divider\n");
+		return -EINVAL;
+	}
+
+	if (clk_div > QSPI_CLK_DIV_MAX) {
+		dev_dbg(qspi->dev, "clock divider >%d , using /%d divider\n",
+				QSPI_CLK_DIV_MAX, QSPI_CLK_DIV_MAX + 1);
+		return -EINVAL;
+	}
+
+	dev_dbg(qspi->dev, "hz: %d, clock divider %d\n",
+			qspi->spi_max_frequency, clk_div);
+
+	ret = pm_runtime_get_sync(qspi->dev);
+	if (ret < 0) {
+		dev_err(qspi->dev, "pm_runtime_get_sync() failed\n");
+		return ret;
+	}
+
+	clk_ctrl_reg = ti_qspi_read(qspi, QSPI_SPI_CLOCK_CNTRL_REG);
+
+	clk_ctrl_reg &= ~QSPI_CLK_EN;
+
+	/* disable SCLK */
+	ti_qspi_write(qspi, clk_ctrl_reg, QSPI_SPI_CLOCK_CNTRL_REG);
+
+	/* enable SCLK */
+	clk_mask = QSPI_CLK_EN | clk_div;
+	ti_qspi_write(qspi, clk_mask, QSPI_SPI_CLOCK_CNTRL_REG);
+	ctx_reg->clkctrl = clk_mask;
+
+	pm_runtime_mark_last_busy(qspi->dev);
+	ret = pm_runtime_put_autosuspend(qspi->dev);
+	if (ret < 0) {
+		dev_err(qspi->dev, "pm_runtime_put_autosuspend() failed\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static void ti_qspi_restore_ctx(struct ti_qspi *qspi)
+{
+	struct ti_qspi_regs *ctx_reg = &qspi->ctx_reg;
+
+	ti_qspi_write(qspi, ctx_reg->clkctrl, QSPI_SPI_CLOCK_CNTRL_REG);
+}
+
+static inline u32 qspi_is_busy(struct ti_qspi *qspi)
+{
+	u32 stat;
+	unsigned long timeout = jiffies + QSPI_COMPLETION_TIMEOUT;
+
+	stat = ti_qspi_read(qspi, QSPI_SPI_STATUS_REG);
+	while ((stat & BUSY) && time_after(timeout, jiffies)) {
+		cpu_relax();
+		stat = ti_qspi_read(qspi, QSPI_SPI_STATUS_REG);
+	}
+
+	WARN(stat & BUSY, "qspi busy\n");
+	return stat & BUSY;
+}
+
+static inline int ti_qspi_poll_wc(struct ti_qspi *qspi)
+{
+	u32 stat;
+	unsigned long timeout = jiffies + QSPI_COMPLETION_TIMEOUT;
+
+	do {
+		stat = ti_qspi_read(qspi, QSPI_SPI_STATUS_REG);
+		if (stat & WC)
+			return 0;
+		cpu_relax();
+	} while (time_after(timeout, jiffies));
+
+	stat = ti_qspi_read(qspi, QSPI_SPI_STATUS_REG);
+	if (stat & WC)
+		return 0;
+	return  -ETIMEDOUT;
+}
+
+static int qspi_write_msg(struct ti_qspi *qspi, struct spi_transfer *t,
+			  int count)
+{
+	int wlen, xfer_len;
+	unsigned int cmd;
+	const u8 *txbuf;
+	u32 data;
+
+	txbuf = t->tx_buf;
+	cmd = qspi->cmd | QSPI_WR_SNGL;
+	wlen = t->bits_per_word >> 3;	/* in bytes */
+	xfer_len = wlen;
+
+	while (count) {
+		if (qspi_is_busy(qspi))
+			return -EBUSY;
+
+		switch (wlen) {
+		case 1:
+			dev_dbg(qspi->dev, "tx cmd %08x dc %08x data %02x\n",
+					cmd, qspi->dc, *txbuf);
+			if (count >= QSPI_WLEN_MAX_BYTES) {
+				u32 *txp = (u32 *)txbuf;
+
+				data = cpu_to_be32(*txp++);
+				writel(data, qspi->base +
+				       QSPI_SPI_DATA_REG_3);
+				data = cpu_to_be32(*txp++);
+				writel(data, qspi->base +
+				       QSPI_SPI_DATA_REG_2);
+				data = cpu_to_be32(*txp++);
+				writel(data, qspi->base +
+				       QSPI_SPI_DATA_REG_1);
+				data = cpu_to_be32(*txp++);
+				writel(data, qspi->base +
+				       QSPI_SPI_DATA_REG);
+				xfer_len = QSPI_WLEN_MAX_BYTES;
+				cmd |= QSPI_WLEN(QSPI_WLEN_MAX_BITS);
+			} else {
+				writeb(*txbuf, qspi->base + QSPI_SPI_DATA_REG);
+				cmd = qspi->cmd | QSPI_WR_SNGL;
+				xfer_len = wlen;
+				cmd |= QSPI_WLEN(wlen);
+			}
+			break;
+		case 2:
+			dev_dbg(qspi->dev, "tx cmd %08x dc %08x data %04x\n",
+					cmd, qspi->dc, *txbuf);
+			writew(*((u16 *)txbuf), qspi->base + QSPI_SPI_DATA_REG);
+			break;
+		case 4:
+			dev_dbg(qspi->dev, "tx cmd %08x dc %08x data %08x\n",
+					cmd, qspi->dc, *txbuf);
+			writel(*((u32 *)txbuf), qspi->base + QSPI_SPI_DATA_REG);
+			break;
+		}
+
+		ti_qspi_write(qspi, cmd, QSPI_SPI_CMD_REG);
+		if (ti_qspi_poll_wc(qspi)) {
+			dev_err(qspi->dev, "write timed out\n");
+			return -ETIMEDOUT;
+		}
+		txbuf += xfer_len;
+		count -= xfer_len;
+	}
+
+	return 0;
+}
+
+static int qspi_read_msg(struct ti_qspi *qspi, struct spi_transfer *t,
+			 int count)
+{
+	int wlen;
+	unsigned int cmd;
+	u8 *rxbuf;
+
+	rxbuf = t->rx_buf;
+	cmd = qspi->cmd;
+	switch (t->rx_nbits) {
+	case SPI_NBITS_DUAL:
+		cmd |= QSPI_RD_DUAL;
+		break;
+	case SPI_NBITS_QUAD:
+		cmd |= QSPI_RD_QUAD;
+		break;
+	default:
+		cmd |= QSPI_RD_SNGL;
+		break;
+	}
+	wlen = t->bits_per_word >> 3;	/* in bytes */
+
+	while (count) {
+		dev_dbg(qspi->dev, "rx cmd %08x dc %08x\n", cmd, qspi->dc);
+		if (qspi_is_busy(qspi))
+			return -EBUSY;
+
+		ti_qspi_write(qspi, cmd, QSPI_SPI_CMD_REG);
+		if (ti_qspi_poll_wc(qspi)) {
+			dev_err(qspi->dev, "read timed out\n");
+			return -ETIMEDOUT;
+		}
+		switch (wlen) {
+		case 1:
+			*rxbuf = readb(qspi->base + QSPI_SPI_DATA_REG);
+			break;
+		case 2:
+			*((u16 *)rxbuf) = readw(qspi->base + QSPI_SPI_DATA_REG);
+			break;
+		case 4:
+			*((u32 *)rxbuf) = readl(qspi->base + QSPI_SPI_DATA_REG);
+			break;
+		}
+		rxbuf += wlen;
+		count -= wlen;
+	}
+
+	return 0;
+}
+
+static int qspi_transfer_msg(struct ti_qspi *qspi, struct spi_transfer *t,
+			     int count)
+{
+	int ret;
+
+	if (t->tx_buf) {
+		ret = qspi_write_msg(qspi, t, count);
+		if (ret) {
+			dev_dbg(qspi->dev, "Error while writing\n");
+			return ret;
+		}
+	}
+
+	if (t->rx_buf) {
+		ret = qspi_read_msg(qspi, t, count);
+		if (ret) {
+			dev_dbg(qspi->dev, "Error while reading\n");
+			return ret;
+		}
+	}
+
+	return 0;
+}
+
+static void ti_qspi_dma_callback(void *param)
+{
+	struct ti_qspi *qspi = param;
+
+	complete(&qspi->transfer_complete);
+}
+
+static int ti_qspi_dma_xfer(struct ti_qspi *qspi, dma_addr_t dma_dst,
+			    dma_addr_t dma_src, size_t len)
+{
+	struct dma_chan *chan = qspi->rx_chan;
+	dma_cookie_t cookie;
+	enum dma_ctrl_flags flags = DMA_CTRL_ACK | DMA_PREP_INTERRUPT;
+	struct dma_async_tx_descriptor *tx;
+	int ret;
+
+	tx = dmaengine_prep_dma_memcpy(chan, dma_dst, dma_src, len, flags);
+	if (!tx) {
+		dev_err(qspi->dev, "device_prep_dma_memcpy error\n");
+		return -EIO;
+	}
+
+	tx->callback = ti_qspi_dma_callback;
+	tx->callback_param = qspi;
+	cookie = tx->tx_submit(tx);
+	reinit_completion(&qspi->transfer_complete);
+
+	ret = dma_submit_error(cookie);
+	if (ret) {
+		dev_err(qspi->dev, "dma_submit_error %d\n", cookie);
+		return -EIO;
+	}
+
+	dma_async_issue_pending(chan);
+	ret = wait_for_completion_timeout(&qspi->transfer_complete,
+					  msecs_to_jiffies(len));
+	if (ret <= 0) {
+		dmaengine_terminate_sync(chan);
+		dev_err(qspi->dev, "DMA wait_for_completion_timeout\n");
+		return -ETIMEDOUT;
+	}
+
+	return 0;
+}
+
+static int ti_qspi_dma_bounce_buffer(struct ti_qspi *qspi, loff_t offs,
+				     void *to, size_t readsize)
+{
+	dma_addr_t dma_src = qspi->mmap_phys_base + offs;
+	int ret = 0;
+
+	/*
+	 * Use bounce buffer as FS like jffs2, ubifs may pass
+	 * buffers that does not belong to kernel lowmem region.
+	 */
+	while (readsize != 0) {
+		size_t xfer_len = min_t(size_t, QSPI_DMA_BUFFER_SIZE,
+					readsize);
+
+		ret = ti_qspi_dma_xfer(qspi, qspi->rx_bb_dma_addr,
+				       dma_src, xfer_len);
+		if (ret != 0)
+			return ret;
+		memcpy(to, qspi->rx_bb_addr, xfer_len);
+		readsize -= xfer_len;
+		dma_src += xfer_len;
+		to += xfer_len;
+	}
+
+	return ret;
+}
+
+static int ti_qspi_dma_xfer_sg(struct ti_qspi *qspi, struct sg_table rx_sg,
+			       loff_t from)
+{
+	struct scatterlist *sg;
+	dma_addr_t dma_src = qspi->mmap_phys_base + from;
+	dma_addr_t dma_dst;
+	int i, len, ret;
+
+	for_each_sg(rx_sg.sgl, sg, rx_sg.nents, i) {
+		dma_dst = sg_dma_address(sg);
+		len = sg_dma_len(sg);
+		ret = ti_qspi_dma_xfer(qspi, dma_dst, dma_src, len);
+		if (ret)
+			return ret;
+		dma_src += len;
+	}
+
+	return 0;
+}
+
+static void ti_qspi_enable_memory_map(struct spi_device *spi)
+{
+	struct ti_qspi  *qspi = spi_master_get_devdata(spi->master);
+
+	ti_qspi_write(qspi, MM_SWITCH, QSPI_SPI_SWITCH_REG);
+	if (qspi->ctrl_base) {
+		regmap_update_bits(qspi->ctrl_base, qspi->ctrl_reg,
+				   MEM_CS_EN(spi->chip_select),
+				   MEM_CS_MASK);
+	}
+	qspi->mmap_enabled = true;
+}
+
+static void ti_qspi_disable_memory_map(struct spi_device *spi)
+{
+	struct ti_qspi  *qspi = spi_master_get_devdata(spi->master);
+
+	ti_qspi_write(qspi, 0, QSPI_SPI_SWITCH_REG);
+	if (qspi->ctrl_base)
+		regmap_update_bits(qspi->ctrl_base, qspi->ctrl_reg,
+				   0, MEM_CS_MASK);
+	qspi->mmap_enabled = false;
+}
+
+static void ti_qspi_setup_mmap_read(struct spi_device *spi, u8 opcode,
+				    u8 data_nbits, u8 addr_width,
+				    u8 dummy_bytes)
+{
+	struct ti_qspi  *qspi = spi_master_get_devdata(spi->master);
+	u32 memval = opcode;
+
+	switch (data_nbits) {
+	case SPI_NBITS_QUAD:
+		memval |= QSPI_SETUP_RD_QUAD;
+		break;
+	case SPI_NBITS_DUAL:
+		memval |= QSPI_SETUP_RD_DUAL;
+		break;
+	default:
+		memval |= QSPI_SETUP_RD_NORMAL;
+		break;
+	}
+	memval |= ((addr_width - 1) << QSPI_SETUP_ADDR_SHIFT |
+		   dummy_bytes << QSPI_SETUP_DUMMY_SHIFT);
+	ti_qspi_write(qspi, memval,
+		      QSPI_SPI_SETUP_REG(spi->chip_select));
+}
+
+static int ti_qspi_exec_mem_op(struct spi_mem *mem,
+			       const struct spi_mem_op *op)
+{
+	struct ti_qspi *qspi = spi_master_get_devdata(mem->spi->master);
+	u32 from = 0;
+	int ret = 0;
+
+	/* Only optimize read path. */
+	if (!op->data.nbytes || op->data.dir != SPI_MEM_DATA_IN ||
+	    !op->addr.nbytes || op->addr.nbytes > 4)
+		return -ENOTSUPP;
+
+	/* Address exceeds MMIO window size, fall back to regular mode. */
+	from = op->addr.val;
+	if (from + op->data.nbytes > qspi->mmap_size)
+		return -ENOTSUPP;
+
+	mutex_lock(&qspi->list_lock);
+
+	if (!qspi->mmap_enabled)
+		ti_qspi_enable_memory_map(mem->spi);
+	ti_qspi_setup_mmap_read(mem->spi, op->cmd.opcode, op->data.buswidth,
+				op->addr.nbytes, op->dummy.nbytes);
+
+	if (qspi->rx_chan) {
+		struct sg_table sgt;
+
+		if (virt_addr_valid(op->data.buf.in) &&
+		    !spi_controller_dma_map_mem_op_data(mem->spi->master, op,
+							&sgt)) {
+			ret = ti_qspi_dma_xfer_sg(qspi, sgt, from);
+			spi_controller_dma_unmap_mem_op_data(mem->spi->master,
+							     op, &sgt);
+		} else {
+			ret = ti_qspi_dma_bounce_buffer(qspi, from,
+							op->data.buf.in,
+							op->data.nbytes);
+		}
+	} else {
+		memcpy_fromio(op->data.buf.in, qspi->mmap_base + from,
+			      op->data.nbytes);
+	}
+
+	mutex_unlock(&qspi->list_lock);
+
+	return ret;
+}
+
+static const struct spi_controller_mem_ops ti_qspi_mem_ops = {
+	.exec_op = ti_qspi_exec_mem_op,
+};
+
+static int ti_qspi_start_transfer_one(struct spi_master *master,
+		struct spi_message *m)
+{
+	struct ti_qspi *qspi = spi_master_get_devdata(master);
+	struct spi_device *spi = m->spi;
+	struct spi_transfer *t;
+	int status = 0, ret;
+	unsigned int frame_len_words, transfer_len_words;
+	int wlen;
+
+	/* setup device control reg */
+	qspi->dc = 0;
+
+	if (spi->mode & SPI_CPHA)
+		qspi->dc |= QSPI_CKPHA(spi->chip_select);
+	if (spi->mode & SPI_CPOL)
+		qspi->dc |= QSPI_CKPOL(spi->chip_select);
+	if (spi->mode & SPI_CS_HIGH)
+		qspi->dc |= QSPI_CSPOL(spi->chip_select);
+
+	frame_len_words = 0;
+	list_for_each_entry(t, &m->transfers, transfer_list)
+		frame_len_words += t->len / (t->bits_per_word >> 3);
+	frame_len_words = min_t(unsigned int, frame_len_words, QSPI_FRAME);
+
+	/* setup command reg */
+	qspi->cmd = 0;
+	qspi->cmd |= QSPI_EN_CS(spi->chip_select);
+	qspi->cmd |= QSPI_FLEN(frame_len_words);
+
+	ti_qspi_write(qspi, qspi->dc, QSPI_SPI_DC_REG);
+
+	mutex_lock(&qspi->list_lock);
+
+	if (qspi->mmap_enabled)
+		ti_qspi_disable_memory_map(spi);
+
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		qspi->cmd = ((qspi->cmd & ~QSPI_WLEN_MASK) |
+			     QSPI_WLEN(t->bits_per_word));
+
+		wlen = t->bits_per_word >> 3;
+		transfer_len_words = min(t->len / wlen, frame_len_words);
+
+		ret = qspi_transfer_msg(qspi, t, transfer_len_words * wlen);
+		if (ret) {
+			dev_dbg(qspi->dev, "transfer message failed\n");
+			mutex_unlock(&qspi->list_lock);
+			return -EINVAL;
+		}
+
+		m->actual_length += transfer_len_words * wlen;
+		frame_len_words -= transfer_len_words;
+		if (frame_len_words == 0)
+			break;
+	}
+
+	mutex_unlock(&qspi->list_lock);
+
+	ti_qspi_write(qspi, qspi->cmd | QSPI_INVAL, QSPI_SPI_CMD_REG);
+	m->status = status;
+	spi_finalize_current_message(master);
+
+	return status;
+}
+
+static int ti_qspi_runtime_resume(struct device *dev)
+{
+	struct ti_qspi      *qspi;
+
+	qspi = dev_get_drvdata(dev);
+	ti_qspi_restore_ctx(qspi);
+
+	return 0;
+}
+
+static const struct of_device_id ti_qspi_match[] = {
+	{.compatible = "ti,dra7xxx-qspi" },
+	{.compatible = "ti,am4372-qspi" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, ti_qspi_match);
+
+static int ti_qspi_probe(struct platform_device *pdev)
+{
+	struct  ti_qspi *qspi;
+	struct spi_master *master;
+	struct resource         *r, *res_mmap;
+	struct device_node *np = pdev->dev.of_node;
+	u32 max_freq;
+	int ret = 0, num_cs, irq;
+	dma_cap_mask_t mask;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*qspi));
+	if (!master)
+		return -ENOMEM;
+
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_RX_DUAL | SPI_RX_QUAD;
+
+	master->flags = SPI_MASTER_HALF_DUPLEX;
+	master->setup = ti_qspi_setup;
+	master->auto_runtime_pm = true;
+	master->transfer_one_message = ti_qspi_start_transfer_one;
+	master->dev.of_node = pdev->dev.of_node;
+	master->bits_per_word_mask = SPI_BPW_MASK(32) | SPI_BPW_MASK(16) |
+				     SPI_BPW_MASK(8);
+	master->mem_ops = &ti_qspi_mem_ops;
+
+	if (!of_property_read_u32(np, "num-cs", &num_cs))
+		master->num_chipselect = num_cs;
+
+	qspi = spi_master_get_devdata(master);
+	qspi->master = master;
+	qspi->dev = &pdev->dev;
+	platform_set_drvdata(pdev, qspi);
+
+	r = platform_get_resource_byname(pdev, IORESOURCE_MEM, "qspi_base");
+	if (r == NULL) {
+		r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+		if (r == NULL) {
+			dev_err(&pdev->dev, "missing platform data\n");
+			ret = -ENODEV;
+			goto free_master;
+		}
+	}
+
+	res_mmap = platform_get_resource_byname(pdev,
+			IORESOURCE_MEM, "qspi_mmap");
+	if (res_mmap == NULL) {
+		res_mmap = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+		if (res_mmap == NULL) {
+			dev_err(&pdev->dev,
+				"memory mapped resource not required\n");
+		}
+	}
+
+	if (res_mmap)
+		qspi->mmap_size = resource_size(res_mmap);
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "no irq resource?\n");
+		ret = irq;
+		goto free_master;
+	}
+
+	mutex_init(&qspi->list_lock);
+
+	qspi->base = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(qspi->base)) {
+		ret = PTR_ERR(qspi->base);
+		goto free_master;
+	}
+
+
+	if (of_property_read_bool(np, "syscon-chipselects")) {
+		qspi->ctrl_base =
+		syscon_regmap_lookup_by_phandle(np,
+						"syscon-chipselects");
+		if (IS_ERR(qspi->ctrl_base)) {
+			ret = PTR_ERR(qspi->ctrl_base);
+			goto free_master;
+		}
+		ret = of_property_read_u32_index(np,
+						 "syscon-chipselects",
+						 1, &qspi->ctrl_reg);
+		if (ret) {
+			dev_err(&pdev->dev,
+				"couldn't get ctrl_mod reg index\n");
+			goto free_master;
+		}
+	}
+
+	qspi->fclk = devm_clk_get(&pdev->dev, "fck");
+	if (IS_ERR(qspi->fclk)) {
+		ret = PTR_ERR(qspi->fclk);
+		dev_err(&pdev->dev, "could not get clk: %d\n", ret);
+	}
+
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_runtime_set_autosuspend_delay(&pdev->dev, QSPI_AUTOSUSPEND_TIMEOUT);
+	pm_runtime_enable(&pdev->dev);
+
+	if (!of_property_read_u32(np, "spi-max-frequency", &max_freq))
+		qspi->spi_max_frequency = max_freq;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_MEMCPY, mask);
+
+	qspi->rx_chan = dma_request_chan_by_mask(&mask);
+	if (IS_ERR(qspi->rx_chan)) {
+		dev_err(qspi->dev,
+			"No Rx DMA available, trying mmap mode\n");
+		qspi->rx_chan = NULL;
+		ret = 0;
+		goto no_dma;
+	}
+	qspi->rx_bb_addr = dma_alloc_coherent(qspi->dev,
+					      QSPI_DMA_BUFFER_SIZE,
+					      &qspi->rx_bb_dma_addr,
+					      GFP_KERNEL | GFP_DMA);
+	if (!qspi->rx_bb_addr) {
+		dev_err(qspi->dev,
+			"dma_alloc_coherent failed, using PIO mode\n");
+		dma_release_channel(qspi->rx_chan);
+		goto no_dma;
+	}
+	master->dma_rx = qspi->rx_chan;
+	init_completion(&qspi->transfer_complete);
+	if (res_mmap)
+		qspi->mmap_phys_base = (dma_addr_t)res_mmap->start;
+
+no_dma:
+	if (!qspi->rx_chan && res_mmap) {
+		qspi->mmap_base = devm_ioremap_resource(&pdev->dev, res_mmap);
+		if (IS_ERR(qspi->mmap_base)) {
+			dev_info(&pdev->dev,
+				 "mmap failed with error %ld using PIO mode\n",
+				 PTR_ERR(qspi->mmap_base));
+			qspi->mmap_base = NULL;
+			master->mem_ops = NULL;
+		}
+	}
+	qspi->mmap_enabled = false;
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (!ret)
+		return 0;
+
+	pm_runtime_disable(&pdev->dev);
+free_master:
+	spi_master_put(master);
+	return ret;
+}
+
+static int ti_qspi_remove(struct platform_device *pdev)
+{
+	struct ti_qspi *qspi = platform_get_drvdata(pdev);
+	int rc;
+
+	rc = spi_master_suspend(qspi->master);
+	if (rc)
+		return rc;
+
+	pm_runtime_put_sync(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+
+	if (qspi->rx_bb_addr)
+		dma_free_coherent(qspi->dev, QSPI_DMA_BUFFER_SIZE,
+				  qspi->rx_bb_addr,
+				  qspi->rx_bb_dma_addr);
+	if (qspi->rx_chan)
+		dma_release_channel(qspi->rx_chan);
+
+	return 0;
+}
+
+static const struct dev_pm_ops ti_qspi_pm_ops = {
+	.runtime_resume = ti_qspi_runtime_resume,
+};
+
+static struct platform_driver ti_qspi_driver = {
+	.probe	= ti_qspi_probe,
+	.remove = ti_qspi_remove,
+	.driver = {
+		.name	= "ti-qspi",
+		.pm =   &ti_qspi_pm_ops,
+		.of_match_table = ti_qspi_match,
+	}
+};
+
+module_platform_driver(ti_qspi_driver);
+
+MODULE_AUTHOR("Sourav Poddar <sourav.poddar@ti.com>");
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("TI QSPI controller driver");
+MODULE_ALIAS("platform:ti-qspi");
diff --git a/drivers/spi/spi-tle62x0.c b/drivers/spi/spi-tle62x0.c
new file mode 100644
index 0000000..c6ae775
--- /dev/null
+++ b/drivers/spi/spi-tle62x0.c
@@ -0,0 +1,320 @@
+/*
+ * Support Infineon TLE62x0 driver chips
+ *
+ * Copyright (c) 2007 Simtec Electronics
+ *	Ben Dooks, <ben@simtec.co.uk>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+
+#include <linux/spi/spi.h>
+#include <linux/spi/tle62x0.h>
+
+
+#define CMD_READ	0x00
+#define CMD_SET		0xff
+
+#define DIAG_NORMAL	0x03
+#define DIAG_OVERLOAD	0x02
+#define DIAG_OPEN	0x01
+#define DIAG_SHORTGND	0x00
+
+struct tle62x0_state {
+	struct spi_device	*us;
+	struct mutex		lock;
+	unsigned int		nr_gpio;
+	unsigned int		gpio_state;
+
+	unsigned char		tx_buff[4];
+	unsigned char		rx_buff[4];
+};
+
+static int to_gpio_num(struct device_attribute *attr);
+
+static inline int tle62x0_write(struct tle62x0_state *st)
+{
+	unsigned char *buff = st->tx_buff;
+	unsigned int gpio_state = st->gpio_state;
+
+	buff[0] = CMD_SET;
+
+	if (st->nr_gpio == 16) {
+		buff[1] = gpio_state >> 8;
+		buff[2] = gpio_state;
+	} else {
+		buff[1] = gpio_state;
+	}
+
+	dev_dbg(&st->us->dev, "buff %3ph\n", buff);
+
+	return spi_write(st->us, buff, (st->nr_gpio == 16) ? 3 : 2);
+}
+
+static inline int tle62x0_read(struct tle62x0_state *st)
+{
+	unsigned char *txbuff = st->tx_buff;
+	struct spi_transfer xfer = {
+		.tx_buf		= txbuff,
+		.rx_buf		= st->rx_buff,
+		.len		= (st->nr_gpio * 2) / 8,
+	};
+	struct spi_message msg;
+
+	txbuff[0] = CMD_READ;
+	txbuff[1] = 0x00;
+	txbuff[2] = 0x00;
+	txbuff[3] = 0x00;
+
+	spi_message_init(&msg);
+	spi_message_add_tail(&xfer, &msg);
+
+	return spi_sync(st->us, &msg);
+}
+
+static unsigned char *decode_fault(unsigned int fault_code)
+{
+	fault_code &= 3;
+
+	switch (fault_code) {
+	case DIAG_NORMAL:
+		return "N";
+	case DIAG_OVERLOAD:
+		return "V";
+	case DIAG_OPEN:
+		return "O";
+	case DIAG_SHORTGND:
+		return "G";
+	}
+
+	return "?";
+}
+
+static ssize_t tle62x0_status_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct tle62x0_state *st = dev_get_drvdata(dev);
+	char *bp = buf;
+	unsigned char *buff = st->rx_buff;
+	unsigned long fault = 0;
+	int ptr;
+	int ret;
+
+	mutex_lock(&st->lock);
+	ret = tle62x0_read(st);
+	dev_dbg(dev, "tle62x0_read() returned %d\n", ret);
+	if (ret < 0) {
+		mutex_unlock(&st->lock);
+		return ret;
+	}
+
+	for (ptr = 0; ptr < (st->nr_gpio * 2)/8; ptr += 1) {
+		fault <<= 8;
+		fault  |= ((unsigned long)buff[ptr]);
+
+		dev_dbg(dev, "byte %d is %02x\n", ptr, buff[ptr]);
+	}
+
+	for (ptr = 0; ptr < st->nr_gpio; ptr++) {
+		bp += sprintf(bp, "%s ", decode_fault(fault >> (ptr * 2)));
+	}
+
+	*bp++ = '\n';
+
+	mutex_unlock(&st->lock);
+	return bp - buf;
+}
+
+static DEVICE_ATTR(status_show, S_IRUGO, tle62x0_status_show, NULL);
+
+static ssize_t tle62x0_gpio_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct tle62x0_state *st = dev_get_drvdata(dev);
+	int gpio_num = to_gpio_num(attr);
+	int value;
+
+	mutex_lock(&st->lock);
+	value = (st->gpio_state >> gpio_num) & 1;
+	mutex_unlock(&st->lock);
+
+	return snprintf(buf, PAGE_SIZE, "%d", value);
+}
+
+static ssize_t tle62x0_gpio_store(struct device *dev,
+		struct device_attribute *attr,
+		const char *buf, size_t len)
+{
+	struct tle62x0_state *st = dev_get_drvdata(dev);
+	int gpio_num = to_gpio_num(attr);
+	unsigned long val;
+	char *endp;
+
+	val = simple_strtoul(buf, &endp, 0);
+	if (buf == endp)
+		return -EINVAL;
+
+	dev_dbg(dev, "setting gpio %d to %ld\n", gpio_num, val);
+
+	mutex_lock(&st->lock);
+
+	if (val)
+		st->gpio_state |= 1 << gpio_num;
+	else
+		st->gpio_state &= ~(1 << gpio_num);
+
+	tle62x0_write(st);
+	mutex_unlock(&st->lock);
+
+	return len;
+}
+
+static DEVICE_ATTR(gpio1, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio2, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio3, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio4, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio5, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio6, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio7, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio8, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio9, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio10, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio11, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio12, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio13, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio14, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio15, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+static DEVICE_ATTR(gpio16, S_IWUSR|S_IRUGO,
+		tle62x0_gpio_show, tle62x0_gpio_store);
+
+static struct device_attribute *gpio_attrs[] = {
+	[0]		= &dev_attr_gpio1,
+	[1]		= &dev_attr_gpio2,
+	[2]		= &dev_attr_gpio3,
+	[3]		= &dev_attr_gpio4,
+	[4]		= &dev_attr_gpio5,
+	[5]		= &dev_attr_gpio6,
+	[6]		= &dev_attr_gpio7,
+	[7]		= &dev_attr_gpio8,
+	[8]		= &dev_attr_gpio9,
+	[9]		= &dev_attr_gpio10,
+	[10]		= &dev_attr_gpio11,
+	[11]		= &dev_attr_gpio12,
+	[12]		= &dev_attr_gpio13,
+	[13]		= &dev_attr_gpio14,
+	[14]		= &dev_attr_gpio15,
+	[15]		= &dev_attr_gpio16
+};
+
+static int to_gpio_num(struct device_attribute *attr)
+{
+	int ptr;
+
+	for (ptr = 0; ptr < ARRAY_SIZE(gpio_attrs); ptr++) {
+		if (gpio_attrs[ptr] == attr)
+			return ptr;
+	}
+
+	return -1;
+}
+
+static int tle62x0_probe(struct spi_device *spi)
+{
+	struct tle62x0_state *st;
+	struct tle62x0_pdata *pdata;
+	int ptr;
+	int ret;
+
+	pdata = dev_get_platdata(&spi->dev);
+	if (pdata == NULL) {
+		dev_err(&spi->dev, "no device data specified\n");
+		return -EINVAL;
+	}
+
+	st = kzalloc(sizeof(struct tle62x0_state), GFP_KERNEL);
+	if (st == NULL)
+		return -ENOMEM;
+
+	st->us = spi;
+	st->nr_gpio = pdata->gpio_count;
+	st->gpio_state = pdata->init_state;
+
+	mutex_init(&st->lock);
+
+	ret = device_create_file(&spi->dev, &dev_attr_status_show);
+	if (ret) {
+		dev_err(&spi->dev, "cannot create status attribute\n");
+		goto err_status;
+	}
+
+	for (ptr = 0; ptr < pdata->gpio_count; ptr++) {
+		ret = device_create_file(&spi->dev, gpio_attrs[ptr]);
+		if (ret) {
+			dev_err(&spi->dev, "cannot create gpio attribute\n");
+			goto err_gpios;
+		}
+	}
+
+	/* tle62x0_write(st); */
+	spi_set_drvdata(spi, st);
+	return 0;
+
+ err_gpios:
+	while (--ptr >= 0)
+		device_remove_file(&spi->dev, gpio_attrs[ptr]);
+
+	device_remove_file(&spi->dev, &dev_attr_status_show);
+
+ err_status:
+	kfree(st);
+	return ret;
+}
+
+static int tle62x0_remove(struct spi_device *spi)
+{
+	struct tle62x0_state *st = spi_get_drvdata(spi);
+	int ptr;
+
+	for (ptr = 0; ptr < st->nr_gpio; ptr++)
+		device_remove_file(&spi->dev, gpio_attrs[ptr]);
+
+	device_remove_file(&spi->dev, &dev_attr_status_show);
+	kfree(st);
+	return 0;
+}
+
+static struct spi_driver tle62x0_driver = {
+	.driver = {
+		.name	= "tle62x0",
+	},
+	.probe		= tle62x0_probe,
+	.remove		= tle62x0_remove,
+};
+
+module_spi_driver(tle62x0_driver);
+
+MODULE_AUTHOR("Ben Dooks <ben@simtec.co.uk>");
+MODULE_DESCRIPTION("TLE62x0 SPI driver");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("spi:tle62x0");
diff --git a/drivers/spi/spi-topcliff-pch.c b/drivers/spi/spi-topcliff-pch.c
new file mode 100644
index 0000000..97d1375
--- /dev/null
+++ b/drivers/spi/spi-topcliff-pch.c
@@ -0,0 +1,1722 @@
+/*
+ * SPI bus driver for the Topcliff PCH used by Intel SoCs
+ *
+ * Copyright (C) 2011 LAPIS Semiconductor Co., Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/delay.h>
+#include <linux/pci.h>
+#include <linux/wait.h>
+#include <linux/spi/spi.h>
+#include <linux/interrupt.h>
+#include <linux/sched.h>
+#include <linux/spi/spidev.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/platform_device.h>
+
+#include <linux/dmaengine.h>
+#include <linux/pch_dma.h>
+
+/* Register offsets */
+#define PCH_SPCR		0x00	/* SPI control register */
+#define PCH_SPBRR		0x04	/* SPI baud rate register */
+#define PCH_SPSR		0x08	/* SPI status register */
+#define PCH_SPDWR		0x0C	/* SPI write data register */
+#define PCH_SPDRR		0x10	/* SPI read data register */
+#define PCH_SSNXCR		0x18	/* SSN Expand Control Register */
+#define PCH_SRST		0x1C	/* SPI reset register */
+#define PCH_ADDRESS_SIZE	0x20
+
+#define PCH_SPSR_TFD		0x000007C0
+#define PCH_SPSR_RFD		0x0000F800
+
+#define PCH_READABLE(x)		(((x) & PCH_SPSR_RFD)>>11)
+#define PCH_WRITABLE(x)		(((x) & PCH_SPSR_TFD)>>6)
+
+#define PCH_RX_THOLD		7
+#define PCH_RX_THOLD_MAX	15
+
+#define PCH_TX_THOLD		2
+
+#define PCH_MAX_BAUDRATE	5000000
+#define PCH_MAX_FIFO_DEPTH	16
+
+#define STATUS_RUNNING		1
+#define STATUS_EXITING		2
+#define PCH_SLEEP_TIME		10
+
+#define SSN_LOW			0x02U
+#define SSN_HIGH		0x03U
+#define SSN_NO_CONTROL		0x00U
+#define PCH_MAX_CS		0xFF
+#define PCI_DEVICE_ID_GE_SPI	0x8816
+
+#define SPCR_SPE_BIT		(1 << 0)
+#define SPCR_MSTR_BIT		(1 << 1)
+#define SPCR_LSBF_BIT		(1 << 4)
+#define SPCR_CPHA_BIT		(1 << 5)
+#define SPCR_CPOL_BIT		(1 << 6)
+#define SPCR_TFIE_BIT		(1 << 8)
+#define SPCR_RFIE_BIT		(1 << 9)
+#define SPCR_FIE_BIT		(1 << 10)
+#define SPCR_ORIE_BIT		(1 << 11)
+#define SPCR_MDFIE_BIT		(1 << 12)
+#define SPCR_FICLR_BIT		(1 << 24)
+#define SPSR_TFI_BIT		(1 << 0)
+#define SPSR_RFI_BIT		(1 << 1)
+#define SPSR_FI_BIT		(1 << 2)
+#define SPSR_ORF_BIT		(1 << 3)
+#define SPBRR_SIZE_BIT		(1 << 10)
+
+#define PCH_ALL			(SPCR_TFIE_BIT|SPCR_RFIE_BIT|SPCR_FIE_BIT|\
+				SPCR_ORIE_BIT|SPCR_MDFIE_BIT)
+
+#define SPCR_RFIC_FIELD		20
+#define SPCR_TFIC_FIELD		16
+
+#define MASK_SPBRR_SPBR_BITS	((1 << 10) - 1)
+#define MASK_RFIC_SPCR_BITS	(0xf << SPCR_RFIC_FIELD)
+#define MASK_TFIC_SPCR_BITS	(0xf << SPCR_TFIC_FIELD)
+
+#define PCH_CLOCK_HZ		50000000
+#define PCH_MAX_SPBR		1023
+
+/* Definition for ML7213/ML7223/ML7831 by LAPIS Semiconductor */
+#define PCI_VENDOR_ID_ROHM		0x10DB
+#define PCI_DEVICE_ID_ML7213_SPI	0x802c
+#define PCI_DEVICE_ID_ML7223_SPI	0x800F
+#define PCI_DEVICE_ID_ML7831_SPI	0x8816
+
+/*
+ * Set the number of SPI instance max
+ * Intel EG20T PCH :		1ch
+ * LAPIS Semiconductor ML7213 IOH :	2ch
+ * LAPIS Semiconductor ML7223 IOH :	1ch
+ * LAPIS Semiconductor ML7831 IOH :	1ch
+*/
+#define PCH_SPI_MAX_DEV			2
+
+#define PCH_BUF_SIZE		4096
+#define PCH_DMA_TRANS_SIZE	12
+
+static int use_dma = 1;
+
+struct pch_spi_dma_ctrl {
+	struct dma_async_tx_descriptor	*desc_tx;
+	struct dma_async_tx_descriptor	*desc_rx;
+	struct pch_dma_slave		param_tx;
+	struct pch_dma_slave		param_rx;
+	struct dma_chan		*chan_tx;
+	struct dma_chan		*chan_rx;
+	struct scatterlist		*sg_tx_p;
+	struct scatterlist		*sg_rx_p;
+	struct scatterlist		sg_tx;
+	struct scatterlist		sg_rx;
+	int				nent;
+	void				*tx_buf_virt;
+	void				*rx_buf_virt;
+	dma_addr_t			tx_buf_dma;
+	dma_addr_t			rx_buf_dma;
+};
+/**
+ * struct pch_spi_data - Holds the SPI channel specific details
+ * @io_remap_addr:		The remapped PCI base address
+ * @master:			Pointer to the SPI master structure
+ * @work:			Reference to work queue handler
+ * @wait:			Wait queue for waking up upon receiving an
+ *				interrupt.
+ * @transfer_complete:		Status of SPI Transfer
+ * @bcurrent_msg_processing:	Status flag for message processing
+ * @lock:			Lock for protecting this structure
+ * @queue:			SPI Message queue
+ * @status:			Status of the SPI driver
+ * @bpw_len:			Length of data to be transferred in bits per
+ *				word
+ * @transfer_active:		Flag showing active transfer
+ * @tx_index:			Transmit data count; for bookkeeping during
+ *				transfer
+ * @rx_index:			Receive data count; for bookkeeping during
+ *				transfer
+ * @tx_buff:			Buffer for data to be transmitted
+ * @rx_index:			Buffer for Received data
+ * @n_curnt_chip:		The chip number that this SPI driver currently
+ *				operates on
+ * @current_chip:		Reference to the current chip that this SPI
+ *				driver currently operates on
+ * @current_msg:		The current message that this SPI driver is
+ *				handling
+ * @cur_trans:			The current transfer that this SPI driver is
+ *				handling
+ * @board_dat:			Reference to the SPI device data structure
+ * @plat_dev:			platform_device structure
+ * @ch:				SPI channel number
+ * @irq_reg_sts:		Status of IRQ registration
+ */
+struct pch_spi_data {
+	void __iomem *io_remap_addr;
+	unsigned long io_base_addr;
+	struct spi_master *master;
+	struct work_struct work;
+	wait_queue_head_t wait;
+	u8 transfer_complete;
+	u8 bcurrent_msg_processing;
+	spinlock_t lock;
+	struct list_head queue;
+	u8 status;
+	u32 bpw_len;
+	u8 transfer_active;
+	u32 tx_index;
+	u32 rx_index;
+	u16 *pkt_tx_buff;
+	u16 *pkt_rx_buff;
+	u8 n_curnt_chip;
+	struct spi_device *current_chip;
+	struct spi_message *current_msg;
+	struct spi_transfer *cur_trans;
+	struct pch_spi_board_data *board_dat;
+	struct platform_device	*plat_dev;
+	int ch;
+	struct pch_spi_dma_ctrl dma;
+	int use_dma;
+	u8 irq_reg_sts;
+	int save_total_len;
+};
+
+/**
+ * struct pch_spi_board_data - Holds the SPI device specific details
+ * @pdev:		Pointer to the PCI device
+ * @suspend_sts:	Status of suspend
+ * @num:		The number of SPI device instance
+ */
+struct pch_spi_board_data {
+	struct pci_dev *pdev;
+	u8 suspend_sts;
+	int num;
+};
+
+struct pch_pd_dev_save {
+	int num;
+	struct platform_device *pd_save[PCH_SPI_MAX_DEV];
+	struct pch_spi_board_data *board_dat;
+};
+
+static const struct pci_device_id pch_spi_pcidev_id[] = {
+	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_GE_SPI),    1, },
+	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7213_SPI), 2, },
+	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_SPI), 1, },
+	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7831_SPI), 1, },
+	{ }
+};
+
+/**
+ * pch_spi_writereg() - Performs  register writes
+ * @master:	Pointer to struct spi_master.
+ * @idx:	Register offset.
+ * @val:	Value to be written to register.
+ */
+static inline void pch_spi_writereg(struct spi_master *master, int idx, u32 val)
+{
+	struct pch_spi_data *data = spi_master_get_devdata(master);
+	iowrite32(val, (data->io_remap_addr + idx));
+}
+
+/**
+ * pch_spi_readreg() - Performs register reads
+ * @master:	Pointer to struct spi_master.
+ * @idx:	Register offset.
+ */
+static inline u32 pch_spi_readreg(struct spi_master *master, int idx)
+{
+	struct pch_spi_data *data = spi_master_get_devdata(master);
+	return ioread32(data->io_remap_addr + idx);
+}
+
+static inline void pch_spi_setclr_reg(struct spi_master *master, int idx,
+				      u32 set, u32 clr)
+{
+	u32 tmp = pch_spi_readreg(master, idx);
+	tmp = (tmp & ~clr) | set;
+	pch_spi_writereg(master, idx, tmp);
+}
+
+static void pch_spi_set_master_mode(struct spi_master *master)
+{
+	pch_spi_setclr_reg(master, PCH_SPCR, SPCR_MSTR_BIT, 0);
+}
+
+/**
+ * pch_spi_clear_fifo() - Clears the Transmit and Receive FIFOs
+ * @master:	Pointer to struct spi_master.
+ */
+static void pch_spi_clear_fifo(struct spi_master *master)
+{
+	pch_spi_setclr_reg(master, PCH_SPCR, SPCR_FICLR_BIT, 0);
+	pch_spi_setclr_reg(master, PCH_SPCR, 0, SPCR_FICLR_BIT);
+}
+
+static void pch_spi_handler_sub(struct pch_spi_data *data, u32 reg_spsr_val,
+				void __iomem *io_remap_addr)
+{
+	u32 n_read, tx_index, rx_index, bpw_len;
+	u16 *pkt_rx_buffer, *pkt_tx_buff;
+	int read_cnt;
+	u32 reg_spcr_val;
+	void __iomem *spsr;
+	void __iomem *spdrr;
+	void __iomem *spdwr;
+
+	spsr = io_remap_addr + PCH_SPSR;
+	iowrite32(reg_spsr_val, spsr);
+
+	if (data->transfer_active) {
+		rx_index = data->rx_index;
+		tx_index = data->tx_index;
+		bpw_len = data->bpw_len;
+		pkt_rx_buffer = data->pkt_rx_buff;
+		pkt_tx_buff = data->pkt_tx_buff;
+
+		spdrr = io_remap_addr + PCH_SPDRR;
+		spdwr = io_remap_addr + PCH_SPDWR;
+
+		n_read = PCH_READABLE(reg_spsr_val);
+
+		for (read_cnt = 0; (read_cnt < n_read); read_cnt++) {
+			pkt_rx_buffer[rx_index++] = ioread32(spdrr);
+			if (tx_index < bpw_len)
+				iowrite32(pkt_tx_buff[tx_index++], spdwr);
+		}
+
+		/* disable RFI if not needed */
+		if ((bpw_len - rx_index) <= PCH_MAX_FIFO_DEPTH) {
+			reg_spcr_val = ioread32(io_remap_addr + PCH_SPCR);
+			reg_spcr_val &= ~SPCR_RFIE_BIT; /* disable RFI */
+
+			/* reset rx threshold */
+			reg_spcr_val &= ~MASK_RFIC_SPCR_BITS;
+			reg_spcr_val |= (PCH_RX_THOLD_MAX << SPCR_RFIC_FIELD);
+
+			iowrite32(reg_spcr_val, (io_remap_addr + PCH_SPCR));
+		}
+
+		/* update counts */
+		data->tx_index = tx_index;
+		data->rx_index = rx_index;
+
+		/* if transfer complete interrupt */
+		if (reg_spsr_val & SPSR_FI_BIT) {
+			if ((tx_index == bpw_len) && (rx_index == tx_index)) {
+				/* disable interrupts */
+				pch_spi_setclr_reg(data->master, PCH_SPCR, 0,
+						   PCH_ALL);
+
+				/* transfer is completed;
+				   inform pch_spi_process_messages */
+				data->transfer_complete = true;
+				data->transfer_active = false;
+				wake_up(&data->wait);
+			} else {
+				dev_vdbg(&data->master->dev,
+					"%s : Transfer is not completed",
+					__func__);
+			}
+		}
+	}
+}
+
+/**
+ * pch_spi_handler() - Interrupt handler
+ * @irq:	The interrupt number.
+ * @dev_id:	Pointer to struct pch_spi_board_data.
+ */
+static irqreturn_t pch_spi_handler(int irq, void *dev_id)
+{
+	u32 reg_spsr_val;
+	void __iomem *spsr;
+	void __iomem *io_remap_addr;
+	irqreturn_t ret = IRQ_NONE;
+	struct pch_spi_data *data = dev_id;
+	struct pch_spi_board_data *board_dat = data->board_dat;
+
+	if (board_dat->suspend_sts) {
+		dev_dbg(&board_dat->pdev->dev,
+			"%s returning due to suspend\n", __func__);
+		return IRQ_NONE;
+	}
+
+	io_remap_addr = data->io_remap_addr;
+	spsr = io_remap_addr + PCH_SPSR;
+
+	reg_spsr_val = ioread32(spsr);
+
+	if (reg_spsr_val & SPSR_ORF_BIT) {
+		dev_err(&board_dat->pdev->dev, "%s Over run error\n", __func__);
+		if (data->current_msg->complete) {
+			data->transfer_complete = true;
+			data->current_msg->status = -EIO;
+			data->current_msg->complete(data->current_msg->context);
+			data->bcurrent_msg_processing = false;
+			data->current_msg = NULL;
+			data->cur_trans = NULL;
+		}
+	}
+
+	if (data->use_dma)
+		return IRQ_NONE;
+
+	/* Check if the interrupt is for SPI device */
+	if (reg_spsr_val & (SPSR_FI_BIT | SPSR_RFI_BIT)) {
+		pch_spi_handler_sub(data, reg_spsr_val, io_remap_addr);
+		ret = IRQ_HANDLED;
+	}
+
+	dev_dbg(&board_dat->pdev->dev, "%s EXIT return value=%d\n",
+		__func__, ret);
+
+	return ret;
+}
+
+/**
+ * pch_spi_set_baud_rate() - Sets SPBR field in SPBRR
+ * @master:	Pointer to struct spi_master.
+ * @speed_hz:	Baud rate.
+ */
+static void pch_spi_set_baud_rate(struct spi_master *master, u32 speed_hz)
+{
+	u32 n_spbr = PCH_CLOCK_HZ / (speed_hz * 2);
+
+	/* if baud rate is less than we can support limit it */
+	if (n_spbr > PCH_MAX_SPBR)
+		n_spbr = PCH_MAX_SPBR;
+
+	pch_spi_setclr_reg(master, PCH_SPBRR, n_spbr, MASK_SPBRR_SPBR_BITS);
+}
+
+/**
+ * pch_spi_set_bits_per_word() - Sets SIZE field in SPBRR
+ * @master:		Pointer to struct spi_master.
+ * @bits_per_word:	Bits per word for SPI transfer.
+ */
+static void pch_spi_set_bits_per_word(struct spi_master *master,
+				      u8 bits_per_word)
+{
+	if (bits_per_word == 8)
+		pch_spi_setclr_reg(master, PCH_SPBRR, 0, SPBRR_SIZE_BIT);
+	else
+		pch_spi_setclr_reg(master, PCH_SPBRR, SPBRR_SIZE_BIT, 0);
+}
+
+/**
+ * pch_spi_setup_transfer() - Configures the PCH SPI hardware for transfer
+ * @spi:	Pointer to struct spi_device.
+ */
+static void pch_spi_setup_transfer(struct spi_device *spi)
+{
+	u32 flags = 0;
+
+	dev_dbg(&spi->dev, "%s SPBRR content =%x setting baud rate=%d\n",
+		__func__, pch_spi_readreg(spi->master, PCH_SPBRR),
+		spi->max_speed_hz);
+	pch_spi_set_baud_rate(spi->master, spi->max_speed_hz);
+
+	/* set bits per word */
+	pch_spi_set_bits_per_word(spi->master, spi->bits_per_word);
+
+	if (!(spi->mode & SPI_LSB_FIRST))
+		flags |= SPCR_LSBF_BIT;
+	if (spi->mode & SPI_CPOL)
+		flags |= SPCR_CPOL_BIT;
+	if (spi->mode & SPI_CPHA)
+		flags |= SPCR_CPHA_BIT;
+	pch_spi_setclr_reg(spi->master, PCH_SPCR, flags,
+			   (SPCR_LSBF_BIT | SPCR_CPOL_BIT | SPCR_CPHA_BIT));
+
+	/* Clear the FIFO by toggling  FICLR to 1 and back to 0 */
+	pch_spi_clear_fifo(spi->master);
+}
+
+/**
+ * pch_spi_reset() - Clears SPI registers
+ * @master:	Pointer to struct spi_master.
+ */
+static void pch_spi_reset(struct spi_master *master)
+{
+	/* write 1 to reset SPI */
+	pch_spi_writereg(master, PCH_SRST, 0x1);
+
+	/* clear reset */
+	pch_spi_writereg(master, PCH_SRST, 0x0);
+}
+
+static int pch_spi_transfer(struct spi_device *pspi, struct spi_message *pmsg)
+{
+
+	struct spi_transfer *transfer;
+	struct pch_spi_data *data = spi_master_get_devdata(pspi->master);
+	int retval;
+	unsigned long flags;
+
+	spin_lock_irqsave(&data->lock, flags);
+	/* validate Tx/Rx buffers and Transfer length */
+	list_for_each_entry(transfer, &pmsg->transfers, transfer_list) {
+		if (!transfer->tx_buf && !transfer->rx_buf) {
+			dev_err(&pspi->dev,
+				"%s Tx and Rx buffer NULL\n", __func__);
+			retval = -EINVAL;
+			goto err_return_spinlock;
+		}
+
+		if (!transfer->len) {
+			dev_err(&pspi->dev, "%s Transfer length invalid\n",
+				__func__);
+			retval = -EINVAL;
+			goto err_return_spinlock;
+		}
+
+		dev_dbg(&pspi->dev,
+			"%s Tx/Rx buffer valid. Transfer length valid\n",
+			__func__);
+	}
+	spin_unlock_irqrestore(&data->lock, flags);
+
+	/* We won't process any messages if we have been asked to terminate */
+	if (data->status == STATUS_EXITING) {
+		dev_err(&pspi->dev, "%s status = STATUS_EXITING.\n", __func__);
+		retval = -ESHUTDOWN;
+		goto err_out;
+	}
+
+	/* If suspended ,return -EINVAL */
+	if (data->board_dat->suspend_sts) {
+		dev_err(&pspi->dev, "%s suspend; returning EINVAL\n", __func__);
+		retval = -EINVAL;
+		goto err_out;
+	}
+
+	/* set status of message */
+	pmsg->actual_length = 0;
+	dev_dbg(&pspi->dev, "%s - pmsg->status =%d\n", __func__, pmsg->status);
+
+	pmsg->status = -EINPROGRESS;
+	spin_lock_irqsave(&data->lock, flags);
+	/* add message to queue */
+	list_add_tail(&pmsg->queue, &data->queue);
+	spin_unlock_irqrestore(&data->lock, flags);
+
+	dev_dbg(&pspi->dev, "%s - Invoked list_add_tail\n", __func__);
+
+	schedule_work(&data->work);
+	dev_dbg(&pspi->dev, "%s - Invoked queue work\n", __func__);
+
+	retval = 0;
+
+err_out:
+	dev_dbg(&pspi->dev, "%s RETURN=%d\n", __func__, retval);
+	return retval;
+err_return_spinlock:
+	dev_dbg(&pspi->dev, "%s RETURN=%d\n", __func__, retval);
+	spin_unlock_irqrestore(&data->lock, flags);
+	return retval;
+}
+
+static inline void pch_spi_select_chip(struct pch_spi_data *data,
+				       struct spi_device *pspi)
+{
+	if (data->current_chip != NULL) {
+		if (pspi->chip_select != data->n_curnt_chip) {
+			dev_dbg(&pspi->dev, "%s : different slave\n", __func__);
+			data->current_chip = NULL;
+		}
+	}
+
+	data->current_chip = pspi;
+
+	data->n_curnt_chip = data->current_chip->chip_select;
+
+	dev_dbg(&pspi->dev, "%s :Invoking pch_spi_setup_transfer\n", __func__);
+	pch_spi_setup_transfer(pspi);
+}
+
+static void pch_spi_set_tx(struct pch_spi_data *data, int *bpw)
+{
+	int size;
+	u32 n_writes;
+	int j;
+	struct spi_message *pmsg, *tmp;
+	const u8 *tx_buf;
+	const u16 *tx_sbuf;
+
+	/* set baud rate if needed */
+	if (data->cur_trans->speed_hz) {
+		dev_dbg(&data->master->dev, "%s:setting baud rate\n", __func__);
+		pch_spi_set_baud_rate(data->master, data->cur_trans->speed_hz);
+	}
+
+	/* set bits per word if needed */
+	if (data->cur_trans->bits_per_word &&
+	    (data->current_msg->spi->bits_per_word != data->cur_trans->bits_per_word)) {
+		dev_dbg(&data->master->dev, "%s:set bits per word\n", __func__);
+		pch_spi_set_bits_per_word(data->master,
+					  data->cur_trans->bits_per_word);
+		*bpw = data->cur_trans->bits_per_word;
+	} else {
+		*bpw = data->current_msg->spi->bits_per_word;
+	}
+
+	/* reset Tx/Rx index */
+	data->tx_index = 0;
+	data->rx_index = 0;
+
+	data->bpw_len = data->cur_trans->len / (*bpw / 8);
+
+	/* find alloc size */
+	size = data->cur_trans->len * sizeof(*data->pkt_tx_buff);
+
+	/* allocate memory for pkt_tx_buff & pkt_rx_buffer */
+	data->pkt_tx_buff = kzalloc(size, GFP_KERNEL);
+	if (data->pkt_tx_buff != NULL) {
+		data->pkt_rx_buff = kzalloc(size, GFP_KERNEL);
+		if (!data->pkt_rx_buff)
+			kfree(data->pkt_tx_buff);
+	}
+
+	if (!data->pkt_rx_buff) {
+		/* flush queue and set status of all transfers to -ENOMEM */
+		list_for_each_entry_safe(pmsg, tmp, data->queue.next, queue) {
+			pmsg->status = -ENOMEM;
+
+			if (pmsg->complete)
+				pmsg->complete(pmsg->context);
+
+			/* delete from queue */
+			list_del_init(&pmsg->queue);
+		}
+		return;
+	}
+
+	/* copy Tx Data */
+	if (data->cur_trans->tx_buf != NULL) {
+		if (*bpw == 8) {
+			tx_buf = data->cur_trans->tx_buf;
+			for (j = 0; j < data->bpw_len; j++)
+				data->pkt_tx_buff[j] = *tx_buf++;
+		} else {
+			tx_sbuf = data->cur_trans->tx_buf;
+			for (j = 0; j < data->bpw_len; j++)
+				data->pkt_tx_buff[j] = *tx_sbuf++;
+		}
+	}
+
+	/* if len greater than PCH_MAX_FIFO_DEPTH, write 16,else len bytes */
+	n_writes = data->bpw_len;
+	if (n_writes > PCH_MAX_FIFO_DEPTH)
+		n_writes = PCH_MAX_FIFO_DEPTH;
+
+	dev_dbg(&data->master->dev,
+		"\n%s:Pulling down SSN low - writing 0x2 to SSNXCR\n",
+		__func__);
+	pch_spi_writereg(data->master, PCH_SSNXCR, SSN_LOW);
+
+	for (j = 0; j < n_writes; j++)
+		pch_spi_writereg(data->master, PCH_SPDWR, data->pkt_tx_buff[j]);
+
+	/* update tx_index */
+	data->tx_index = j;
+
+	/* reset transfer complete flag */
+	data->transfer_complete = false;
+	data->transfer_active = true;
+}
+
+static void pch_spi_nomore_transfer(struct pch_spi_data *data)
+{
+	struct spi_message *pmsg, *tmp;
+	dev_dbg(&data->master->dev, "%s called\n", __func__);
+	/* Invoke complete callback
+	 * [To the spi core..indicating end of transfer] */
+	data->current_msg->status = 0;
+
+	if (data->current_msg->complete) {
+		dev_dbg(&data->master->dev,
+			"%s:Invoking callback of SPI core\n", __func__);
+		data->current_msg->complete(data->current_msg->context);
+	}
+
+	/* update status in global variable */
+	data->bcurrent_msg_processing = false;
+
+	dev_dbg(&data->master->dev,
+		"%s:data->bcurrent_msg_processing = false\n", __func__);
+
+	data->current_msg = NULL;
+	data->cur_trans = NULL;
+
+	/* check if we have items in list and not suspending
+	 * return 1 if list empty */
+	if ((list_empty(&data->queue) == 0) &&
+	    (!data->board_dat->suspend_sts) &&
+	    (data->status != STATUS_EXITING)) {
+		/* We have some more work to do (either there is more tranint
+		 * bpw;sfer requests in the current message or there are
+		 *more messages)
+		 */
+		dev_dbg(&data->master->dev, "%s:Invoke queue_work\n", __func__);
+		schedule_work(&data->work);
+	} else if (data->board_dat->suspend_sts ||
+		   data->status == STATUS_EXITING) {
+		dev_dbg(&data->master->dev,
+			"%s suspend/remove initiated, flushing queue\n",
+			__func__);
+		list_for_each_entry_safe(pmsg, tmp, data->queue.next, queue) {
+			pmsg->status = -EIO;
+
+			if (pmsg->complete)
+				pmsg->complete(pmsg->context);
+
+			/* delete from queue */
+			list_del_init(&pmsg->queue);
+		}
+	}
+}
+
+static void pch_spi_set_ir(struct pch_spi_data *data)
+{
+	/* enable interrupts, set threshold, enable SPI */
+	if ((data->bpw_len) > PCH_MAX_FIFO_DEPTH)
+		/* set receive threshold to PCH_RX_THOLD */
+		pch_spi_setclr_reg(data->master, PCH_SPCR,
+				   PCH_RX_THOLD << SPCR_RFIC_FIELD |
+				   SPCR_FIE_BIT | SPCR_RFIE_BIT |
+				   SPCR_ORIE_BIT | SPCR_SPE_BIT,
+				   MASK_RFIC_SPCR_BITS | PCH_ALL);
+	else
+		/* set receive threshold to maximum */
+		pch_spi_setclr_reg(data->master, PCH_SPCR,
+				   PCH_RX_THOLD_MAX << SPCR_RFIC_FIELD |
+				   SPCR_FIE_BIT | SPCR_ORIE_BIT |
+				   SPCR_SPE_BIT,
+				   MASK_RFIC_SPCR_BITS | PCH_ALL);
+
+	/* Wait until the transfer completes; go to sleep after
+				 initiating the transfer. */
+	dev_dbg(&data->master->dev,
+		"%s:waiting for transfer to get over\n", __func__);
+
+	wait_event_interruptible(data->wait, data->transfer_complete);
+
+	/* clear all interrupts */
+	pch_spi_writereg(data->master, PCH_SPSR,
+			 pch_spi_readreg(data->master, PCH_SPSR));
+	/* Disable interrupts and SPI transfer */
+	pch_spi_setclr_reg(data->master, PCH_SPCR, 0, PCH_ALL | SPCR_SPE_BIT);
+	/* clear FIFO */
+	pch_spi_clear_fifo(data->master);
+}
+
+static void pch_spi_copy_rx_data(struct pch_spi_data *data, int bpw)
+{
+	int j;
+	u8 *rx_buf;
+	u16 *rx_sbuf;
+
+	/* copy Rx Data */
+	if (!data->cur_trans->rx_buf)
+		return;
+
+	if (bpw == 8) {
+		rx_buf = data->cur_trans->rx_buf;
+		for (j = 0; j < data->bpw_len; j++)
+			*rx_buf++ = data->pkt_rx_buff[j] & 0xFF;
+	} else {
+		rx_sbuf = data->cur_trans->rx_buf;
+		for (j = 0; j < data->bpw_len; j++)
+			*rx_sbuf++ = data->pkt_rx_buff[j];
+	}
+}
+
+static void pch_spi_copy_rx_data_for_dma(struct pch_spi_data *data, int bpw)
+{
+	int j;
+	u8 *rx_buf;
+	u16 *rx_sbuf;
+	const u8 *rx_dma_buf;
+	const u16 *rx_dma_sbuf;
+
+	/* copy Rx Data */
+	if (!data->cur_trans->rx_buf)
+		return;
+
+	if (bpw == 8) {
+		rx_buf = data->cur_trans->rx_buf;
+		rx_dma_buf = data->dma.rx_buf_virt;
+		for (j = 0; j < data->bpw_len; j++)
+			*rx_buf++ = *rx_dma_buf++ & 0xFF;
+		data->cur_trans->rx_buf = rx_buf;
+	} else {
+		rx_sbuf = data->cur_trans->rx_buf;
+		rx_dma_sbuf = data->dma.rx_buf_virt;
+		for (j = 0; j < data->bpw_len; j++)
+			*rx_sbuf++ = *rx_dma_sbuf++;
+		data->cur_trans->rx_buf = rx_sbuf;
+	}
+}
+
+static int pch_spi_start_transfer(struct pch_spi_data *data)
+{
+	struct pch_spi_dma_ctrl *dma;
+	unsigned long flags;
+	int rtn;
+
+	dma = &data->dma;
+
+	spin_lock_irqsave(&data->lock, flags);
+
+	/* disable interrupts, SPI set enable */
+	pch_spi_setclr_reg(data->master, PCH_SPCR, SPCR_SPE_BIT, PCH_ALL);
+
+	spin_unlock_irqrestore(&data->lock, flags);
+
+	/* Wait until the transfer completes; go to sleep after
+				 initiating the transfer. */
+	dev_dbg(&data->master->dev,
+		"%s:waiting for transfer to get over\n", __func__);
+	rtn = wait_event_interruptible_timeout(data->wait,
+					       data->transfer_complete,
+					       msecs_to_jiffies(2 * HZ));
+	if (!rtn)
+		dev_err(&data->master->dev,
+			"%s wait-event timeout\n", __func__);
+
+	dma_sync_sg_for_cpu(&data->master->dev, dma->sg_rx_p, dma->nent,
+			    DMA_FROM_DEVICE);
+
+	dma_sync_sg_for_cpu(&data->master->dev, dma->sg_tx_p, dma->nent,
+			    DMA_FROM_DEVICE);
+	memset(data->dma.tx_buf_virt, 0, PAGE_SIZE);
+
+	async_tx_ack(dma->desc_rx);
+	async_tx_ack(dma->desc_tx);
+	kfree(dma->sg_tx_p);
+	kfree(dma->sg_rx_p);
+
+	spin_lock_irqsave(&data->lock, flags);
+
+	/* clear fifo threshold, disable interrupts, disable SPI transfer */
+	pch_spi_setclr_reg(data->master, PCH_SPCR, 0,
+			   MASK_RFIC_SPCR_BITS | MASK_TFIC_SPCR_BITS | PCH_ALL |
+			   SPCR_SPE_BIT);
+	/* clear all interrupts */
+	pch_spi_writereg(data->master, PCH_SPSR,
+			 pch_spi_readreg(data->master, PCH_SPSR));
+	/* clear FIFO */
+	pch_spi_clear_fifo(data->master);
+
+	spin_unlock_irqrestore(&data->lock, flags);
+
+	return rtn;
+}
+
+static void pch_dma_rx_complete(void *arg)
+{
+	struct pch_spi_data *data = arg;
+
+	/* transfer is completed;inform pch_spi_process_messages_dma */
+	data->transfer_complete = true;
+	wake_up_interruptible(&data->wait);
+}
+
+static bool pch_spi_filter(struct dma_chan *chan, void *slave)
+{
+	struct pch_dma_slave *param = slave;
+
+	if ((chan->chan_id == param->chan_id) &&
+	    (param->dma_dev == chan->device->dev)) {
+		chan->private = param;
+		return true;
+	} else {
+		return false;
+	}
+}
+
+static void pch_spi_request_dma(struct pch_spi_data *data, int bpw)
+{
+	dma_cap_mask_t mask;
+	struct dma_chan *chan;
+	struct pci_dev *dma_dev;
+	struct pch_dma_slave *param;
+	struct pch_spi_dma_ctrl *dma;
+	unsigned int width;
+
+	if (bpw == 8)
+		width = PCH_DMA_WIDTH_1_BYTE;
+	else
+		width = PCH_DMA_WIDTH_2_BYTES;
+
+	dma = &data->dma;
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	/* Get DMA's dev information */
+	dma_dev = pci_get_slot(data->board_dat->pdev->bus,
+			PCI_DEVFN(PCI_SLOT(data->board_dat->pdev->devfn), 0));
+
+	/* Set Tx DMA */
+	param = &dma->param_tx;
+	param->dma_dev = &dma_dev->dev;
+	param->chan_id = data->ch * 2; /* Tx = 0, 2 */;
+	param->tx_reg = data->io_base_addr + PCH_SPDWR;
+	param->width = width;
+	chan = dma_request_channel(mask, pch_spi_filter, param);
+	if (!chan) {
+		dev_err(&data->master->dev,
+			"ERROR: dma_request_channel FAILS(Tx)\n");
+		data->use_dma = 0;
+		return;
+	}
+	dma->chan_tx = chan;
+
+	/* Set Rx DMA */
+	param = &dma->param_rx;
+	param->dma_dev = &dma_dev->dev;
+	param->chan_id = data->ch * 2 + 1; /* Rx = Tx + 1 */;
+	param->rx_reg = data->io_base_addr + PCH_SPDRR;
+	param->width = width;
+	chan = dma_request_channel(mask, pch_spi_filter, param);
+	if (!chan) {
+		dev_err(&data->master->dev,
+			"ERROR: dma_request_channel FAILS(Rx)\n");
+		dma_release_channel(dma->chan_tx);
+		dma->chan_tx = NULL;
+		data->use_dma = 0;
+		return;
+	}
+	dma->chan_rx = chan;
+}
+
+static void pch_spi_release_dma(struct pch_spi_data *data)
+{
+	struct pch_spi_dma_ctrl *dma;
+
+	dma = &data->dma;
+	if (dma->chan_tx) {
+		dma_release_channel(dma->chan_tx);
+		dma->chan_tx = NULL;
+	}
+	if (dma->chan_rx) {
+		dma_release_channel(dma->chan_rx);
+		dma->chan_rx = NULL;
+	}
+}
+
+static void pch_spi_handle_dma(struct pch_spi_data *data, int *bpw)
+{
+	const u8 *tx_buf;
+	const u16 *tx_sbuf;
+	u8 *tx_dma_buf;
+	u16 *tx_dma_sbuf;
+	struct scatterlist *sg;
+	struct dma_async_tx_descriptor *desc_tx;
+	struct dma_async_tx_descriptor *desc_rx;
+	int num;
+	int i;
+	int size;
+	int rem;
+	int head;
+	unsigned long flags;
+	struct pch_spi_dma_ctrl *dma;
+
+	dma = &data->dma;
+
+	/* set baud rate if needed */
+	if (data->cur_trans->speed_hz) {
+		dev_dbg(&data->master->dev, "%s:setting baud rate\n", __func__);
+		spin_lock_irqsave(&data->lock, flags);
+		pch_spi_set_baud_rate(data->master, data->cur_trans->speed_hz);
+		spin_unlock_irqrestore(&data->lock, flags);
+	}
+
+	/* set bits per word if needed */
+	if (data->cur_trans->bits_per_word &&
+	    (data->current_msg->spi->bits_per_word !=
+	     data->cur_trans->bits_per_word)) {
+		dev_dbg(&data->master->dev, "%s:set bits per word\n", __func__);
+		spin_lock_irqsave(&data->lock, flags);
+		pch_spi_set_bits_per_word(data->master,
+					  data->cur_trans->bits_per_word);
+		spin_unlock_irqrestore(&data->lock, flags);
+		*bpw = data->cur_trans->bits_per_word;
+	} else {
+		*bpw = data->current_msg->spi->bits_per_word;
+	}
+	data->bpw_len = data->cur_trans->len / (*bpw / 8);
+
+	if (data->bpw_len > PCH_BUF_SIZE) {
+		data->bpw_len = PCH_BUF_SIZE;
+		data->cur_trans->len -= PCH_BUF_SIZE;
+	}
+
+	/* copy Tx Data */
+	if (data->cur_trans->tx_buf != NULL) {
+		if (*bpw == 8) {
+			tx_buf = data->cur_trans->tx_buf;
+			tx_dma_buf = dma->tx_buf_virt;
+			for (i = 0; i < data->bpw_len; i++)
+				*tx_dma_buf++ = *tx_buf++;
+		} else {
+			tx_sbuf = data->cur_trans->tx_buf;
+			tx_dma_sbuf = dma->tx_buf_virt;
+			for (i = 0; i < data->bpw_len; i++)
+				*tx_dma_sbuf++ = *tx_sbuf++;
+		}
+	}
+
+	/* Calculate Rx parameter for DMA transmitting */
+	if (data->bpw_len > PCH_DMA_TRANS_SIZE) {
+		if (data->bpw_len % PCH_DMA_TRANS_SIZE) {
+			num = data->bpw_len / PCH_DMA_TRANS_SIZE + 1;
+			rem = data->bpw_len % PCH_DMA_TRANS_SIZE;
+		} else {
+			num = data->bpw_len / PCH_DMA_TRANS_SIZE;
+			rem = PCH_DMA_TRANS_SIZE;
+		}
+		size = PCH_DMA_TRANS_SIZE;
+	} else {
+		num = 1;
+		size = data->bpw_len;
+		rem = data->bpw_len;
+	}
+	dev_dbg(&data->master->dev, "%s num=%d size=%d rem=%d\n",
+		__func__, num, size, rem);
+	spin_lock_irqsave(&data->lock, flags);
+
+	/* set receive fifo threshold and transmit fifo threshold */
+	pch_spi_setclr_reg(data->master, PCH_SPCR,
+			   ((size - 1) << SPCR_RFIC_FIELD) |
+			   (PCH_TX_THOLD << SPCR_TFIC_FIELD),
+			   MASK_RFIC_SPCR_BITS | MASK_TFIC_SPCR_BITS);
+
+	spin_unlock_irqrestore(&data->lock, flags);
+
+	/* RX */
+	dma->sg_rx_p = kcalloc(num, sizeof(*dma->sg_rx_p), GFP_ATOMIC);
+	sg_init_table(dma->sg_rx_p, num); /* Initialize SG table */
+	/* offset, length setting */
+	sg = dma->sg_rx_p;
+	for (i = 0; i < num; i++, sg++) {
+		if (i == (num - 2)) {
+			sg->offset = size * i;
+			sg->offset = sg->offset * (*bpw / 8);
+			sg_set_page(sg, virt_to_page(dma->rx_buf_virt), rem,
+				    sg->offset);
+			sg_dma_len(sg) = rem;
+		} else if (i == (num - 1)) {
+			sg->offset = size * (i - 1) + rem;
+			sg->offset = sg->offset * (*bpw / 8);
+			sg_set_page(sg, virt_to_page(dma->rx_buf_virt), size,
+				    sg->offset);
+			sg_dma_len(sg) = size;
+		} else {
+			sg->offset = size * i;
+			sg->offset = sg->offset * (*bpw / 8);
+			sg_set_page(sg, virt_to_page(dma->rx_buf_virt), size,
+				    sg->offset);
+			sg_dma_len(sg) = size;
+		}
+		sg_dma_address(sg) = dma->rx_buf_dma + sg->offset;
+	}
+	sg = dma->sg_rx_p;
+	desc_rx = dmaengine_prep_slave_sg(dma->chan_rx, sg,
+					num, DMA_DEV_TO_MEM,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!desc_rx) {
+		dev_err(&data->master->dev,
+			"%s:dmaengine_prep_slave_sg Failed\n", __func__);
+		return;
+	}
+	dma_sync_sg_for_device(&data->master->dev, sg, num, DMA_FROM_DEVICE);
+	desc_rx->callback = pch_dma_rx_complete;
+	desc_rx->callback_param = data;
+	dma->nent = num;
+	dma->desc_rx = desc_rx;
+
+	/* Calculate Tx parameter for DMA transmitting */
+	if (data->bpw_len > PCH_MAX_FIFO_DEPTH) {
+		head = PCH_MAX_FIFO_DEPTH - PCH_DMA_TRANS_SIZE;
+		if (data->bpw_len % PCH_DMA_TRANS_SIZE > 4) {
+			num = data->bpw_len / PCH_DMA_TRANS_SIZE + 1;
+			rem = data->bpw_len % PCH_DMA_TRANS_SIZE - head;
+		} else {
+			num = data->bpw_len / PCH_DMA_TRANS_SIZE;
+			rem = data->bpw_len % PCH_DMA_TRANS_SIZE +
+			      PCH_DMA_TRANS_SIZE - head;
+		}
+		size = PCH_DMA_TRANS_SIZE;
+	} else {
+		num = 1;
+		size = data->bpw_len;
+		rem = data->bpw_len;
+		head = 0;
+	}
+
+	dma->sg_tx_p = kcalloc(num, sizeof(*dma->sg_tx_p), GFP_ATOMIC);
+	sg_init_table(dma->sg_tx_p, num); /* Initialize SG table */
+	/* offset, length setting */
+	sg = dma->sg_tx_p;
+	for (i = 0; i < num; i++, sg++) {
+		if (i == 0) {
+			sg->offset = 0;
+			sg_set_page(sg, virt_to_page(dma->tx_buf_virt), size + head,
+				    sg->offset);
+			sg_dma_len(sg) = size + head;
+		} else if (i == (num - 1)) {
+			sg->offset = head + size * i;
+			sg->offset = sg->offset * (*bpw / 8);
+			sg_set_page(sg, virt_to_page(dma->tx_buf_virt), rem,
+				    sg->offset);
+			sg_dma_len(sg) = rem;
+		} else {
+			sg->offset = head + size * i;
+			sg->offset = sg->offset * (*bpw / 8);
+			sg_set_page(sg, virt_to_page(dma->tx_buf_virt), size,
+				    sg->offset);
+			sg_dma_len(sg) = size;
+		}
+		sg_dma_address(sg) = dma->tx_buf_dma + sg->offset;
+	}
+	sg = dma->sg_tx_p;
+	desc_tx = dmaengine_prep_slave_sg(dma->chan_tx,
+					sg, num, DMA_MEM_TO_DEV,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!desc_tx) {
+		dev_err(&data->master->dev,
+			"%s:dmaengine_prep_slave_sg Failed\n", __func__);
+		return;
+	}
+	dma_sync_sg_for_device(&data->master->dev, sg, num, DMA_TO_DEVICE);
+	desc_tx->callback = NULL;
+	desc_tx->callback_param = data;
+	dma->nent = num;
+	dma->desc_tx = desc_tx;
+
+	dev_dbg(&data->master->dev, "%s:Pulling down SSN low - writing 0x2 to SSNXCR\n", __func__);
+
+	spin_lock_irqsave(&data->lock, flags);
+	pch_spi_writereg(data->master, PCH_SSNXCR, SSN_LOW);
+	desc_rx->tx_submit(desc_rx);
+	desc_tx->tx_submit(desc_tx);
+	spin_unlock_irqrestore(&data->lock, flags);
+
+	/* reset transfer complete flag */
+	data->transfer_complete = false;
+}
+
+static void pch_spi_process_messages(struct work_struct *pwork)
+{
+	struct spi_message *pmsg, *tmp;
+	struct pch_spi_data *data;
+	int bpw;
+
+	data = container_of(pwork, struct pch_spi_data, work);
+	dev_dbg(&data->master->dev, "%s data initialized\n", __func__);
+
+	spin_lock(&data->lock);
+	/* check if suspend has been initiated;if yes flush queue */
+	if (data->board_dat->suspend_sts || (data->status == STATUS_EXITING)) {
+		dev_dbg(&data->master->dev,
+			"%s suspend/remove initiated, flushing queue\n", __func__);
+		list_for_each_entry_safe(pmsg, tmp, data->queue.next, queue) {
+			pmsg->status = -EIO;
+
+			if (pmsg->complete) {
+				spin_unlock(&data->lock);
+				pmsg->complete(pmsg->context);
+				spin_lock(&data->lock);
+			}
+
+			/* delete from queue */
+			list_del_init(&pmsg->queue);
+		}
+
+		spin_unlock(&data->lock);
+		return;
+	}
+
+	data->bcurrent_msg_processing = true;
+	dev_dbg(&data->master->dev,
+		"%s Set data->bcurrent_msg_processing= true\n", __func__);
+
+	/* Get the message from the queue and delete it from there. */
+	data->current_msg = list_entry(data->queue.next, struct spi_message,
+					queue);
+
+	list_del_init(&data->current_msg->queue);
+
+	data->current_msg->status = 0;
+
+	pch_spi_select_chip(data, data->current_msg->spi);
+
+	spin_unlock(&data->lock);
+
+	if (data->use_dma)
+		pch_spi_request_dma(data,
+				    data->current_msg->spi->bits_per_word);
+	pch_spi_writereg(data->master, PCH_SSNXCR, SSN_NO_CONTROL);
+	do {
+		int cnt;
+		/* If we are already processing a message get the next
+		transfer structure from the message otherwise retrieve
+		the 1st transfer request from the message. */
+		spin_lock(&data->lock);
+		if (data->cur_trans == NULL) {
+			data->cur_trans =
+				list_entry(data->current_msg->transfers.next,
+					   struct spi_transfer, transfer_list);
+			dev_dbg(&data->master->dev,
+				"%s :Getting 1st transfer message\n",
+				__func__);
+		} else {
+			data->cur_trans =
+				list_entry(data->cur_trans->transfer_list.next,
+					   struct spi_transfer, transfer_list);
+			dev_dbg(&data->master->dev,
+				"%s :Getting next transfer message\n",
+				__func__);
+		}
+		spin_unlock(&data->lock);
+
+		if (!data->cur_trans->len)
+			goto out;
+		cnt = (data->cur_trans->len - 1) / PCH_BUF_SIZE + 1;
+		data->save_total_len = data->cur_trans->len;
+		if (data->use_dma) {
+			int i;
+			char *save_rx_buf = data->cur_trans->rx_buf;
+			for (i = 0; i < cnt; i ++) {
+				pch_spi_handle_dma(data, &bpw);
+				if (!pch_spi_start_transfer(data)) {
+					data->transfer_complete = true;
+					data->current_msg->status = -EIO;
+					data->current_msg->complete
+						   (data->current_msg->context);
+					data->bcurrent_msg_processing = false;
+					data->current_msg = NULL;
+					data->cur_trans = NULL;
+					goto out;
+				}
+				pch_spi_copy_rx_data_for_dma(data, bpw);
+			}
+			data->cur_trans->rx_buf = save_rx_buf;
+		} else {
+			pch_spi_set_tx(data, &bpw);
+			pch_spi_set_ir(data);
+			pch_spi_copy_rx_data(data, bpw);
+			kfree(data->pkt_rx_buff);
+			data->pkt_rx_buff = NULL;
+			kfree(data->pkt_tx_buff);
+			data->pkt_tx_buff = NULL;
+		}
+		/* increment message count */
+		data->cur_trans->len = data->save_total_len;
+		data->current_msg->actual_length += data->cur_trans->len;
+
+		dev_dbg(&data->master->dev,
+			"%s:data->current_msg->actual_length=%d\n",
+			__func__, data->current_msg->actual_length);
+
+		/* check for delay */
+		if (data->cur_trans->delay_usecs) {
+			dev_dbg(&data->master->dev, "%s:delay in usec=%d\n",
+				__func__, data->cur_trans->delay_usecs);
+			udelay(data->cur_trans->delay_usecs);
+		}
+
+		spin_lock(&data->lock);
+
+		/* No more transfer in this message. */
+		if ((data->cur_trans->transfer_list.next) ==
+		    &(data->current_msg->transfers)) {
+			pch_spi_nomore_transfer(data);
+		}
+
+		spin_unlock(&data->lock);
+
+	} while (data->cur_trans != NULL);
+
+out:
+	pch_spi_writereg(data->master, PCH_SSNXCR, SSN_HIGH);
+	if (data->use_dma)
+		pch_spi_release_dma(data);
+}
+
+static void pch_spi_free_resources(struct pch_spi_board_data *board_dat,
+				   struct pch_spi_data *data)
+{
+	dev_dbg(&board_dat->pdev->dev, "%s ENTRY\n", __func__);
+
+	flush_work(&data->work);
+}
+
+static int pch_spi_get_resources(struct pch_spi_board_data *board_dat,
+				 struct pch_spi_data *data)
+{
+	dev_dbg(&board_dat->pdev->dev, "%s ENTRY\n", __func__);
+
+	/* reset PCH SPI h/w */
+	pch_spi_reset(data->master);
+	dev_dbg(&board_dat->pdev->dev,
+		"%s pch_spi_reset invoked successfully\n", __func__);
+
+	dev_dbg(&board_dat->pdev->dev, "%s data->irq_reg_sts=true\n", __func__);
+
+	return 0;
+}
+
+static void pch_free_dma_buf(struct pch_spi_board_data *board_dat,
+			     struct pch_spi_data *data)
+{
+	struct pch_spi_dma_ctrl *dma;
+
+	dma = &data->dma;
+	if (dma->tx_buf_dma)
+		dma_free_coherent(&board_dat->pdev->dev, PCH_BUF_SIZE,
+				  dma->tx_buf_virt, dma->tx_buf_dma);
+	if (dma->rx_buf_dma)
+		dma_free_coherent(&board_dat->pdev->dev, PCH_BUF_SIZE,
+				  dma->rx_buf_virt, dma->rx_buf_dma);
+}
+
+static void pch_alloc_dma_buf(struct pch_spi_board_data *board_dat,
+			      struct pch_spi_data *data)
+{
+	struct pch_spi_dma_ctrl *dma;
+
+	dma = &data->dma;
+	/* Get Consistent memory for Tx DMA */
+	dma->tx_buf_virt = dma_alloc_coherent(&board_dat->pdev->dev,
+				PCH_BUF_SIZE, &dma->tx_buf_dma, GFP_KERNEL);
+	/* Get Consistent memory for Rx DMA */
+	dma->rx_buf_virt = dma_alloc_coherent(&board_dat->pdev->dev,
+				PCH_BUF_SIZE, &dma->rx_buf_dma, GFP_KERNEL);
+}
+
+static int pch_spi_pd_probe(struct platform_device *plat_dev)
+{
+	int ret;
+	struct spi_master *master;
+	struct pch_spi_board_data *board_dat = dev_get_platdata(&plat_dev->dev);
+	struct pch_spi_data *data;
+
+	dev_dbg(&plat_dev->dev, "%s:debug\n", __func__);
+
+	master = spi_alloc_master(&board_dat->pdev->dev,
+				  sizeof(struct pch_spi_data));
+	if (!master) {
+		dev_err(&plat_dev->dev, "spi_alloc_master[%d] failed.\n",
+			plat_dev->id);
+		return -ENOMEM;
+	}
+
+	data = spi_master_get_devdata(master);
+	data->master = master;
+
+	platform_set_drvdata(plat_dev, data);
+
+	/* baseaddress + address offset) */
+	data->io_base_addr = pci_resource_start(board_dat->pdev, 1) +
+					 PCH_ADDRESS_SIZE * plat_dev->id;
+	data->io_remap_addr = pci_iomap(board_dat->pdev, 1, 0);
+	if (!data->io_remap_addr) {
+		dev_err(&plat_dev->dev, "%s pci_iomap failed\n", __func__);
+		ret = -ENOMEM;
+		goto err_pci_iomap;
+	}
+	data->io_remap_addr += PCH_ADDRESS_SIZE * plat_dev->id;
+
+	dev_dbg(&plat_dev->dev, "[ch%d] remap_addr=%p\n",
+		plat_dev->id, data->io_remap_addr);
+
+	/* initialize members of SPI master */
+	master->num_chipselect = PCH_MAX_CS;
+	master->transfer = pch_spi_transfer;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_LSB_FIRST;
+	master->bits_per_word_mask = SPI_BPW_MASK(8) | SPI_BPW_MASK(16);
+	master->max_speed_hz = PCH_MAX_BAUDRATE;
+
+	data->board_dat = board_dat;
+	data->plat_dev = plat_dev;
+	data->n_curnt_chip = 255;
+	data->status = STATUS_RUNNING;
+	data->ch = plat_dev->id;
+	data->use_dma = use_dma;
+
+	INIT_LIST_HEAD(&data->queue);
+	spin_lock_init(&data->lock);
+	INIT_WORK(&data->work, pch_spi_process_messages);
+	init_waitqueue_head(&data->wait);
+
+	ret = pch_spi_get_resources(board_dat, data);
+	if (ret) {
+		dev_err(&plat_dev->dev, "%s fail(retval=%d)\n", __func__, ret);
+		goto err_spi_get_resources;
+	}
+
+	ret = request_irq(board_dat->pdev->irq, pch_spi_handler,
+			  IRQF_SHARED, KBUILD_MODNAME, data);
+	if (ret) {
+		dev_err(&plat_dev->dev,
+			"%s request_irq failed\n", __func__);
+		goto err_request_irq;
+	}
+	data->irq_reg_sts = true;
+
+	pch_spi_set_master_mode(master);
+
+	if (use_dma) {
+		dev_info(&plat_dev->dev, "Use DMA for data transfers\n");
+		pch_alloc_dma_buf(board_dat, data);
+	}
+
+	ret = spi_register_master(master);
+	if (ret != 0) {
+		dev_err(&plat_dev->dev,
+			"%s spi_register_master FAILED\n", __func__);
+		goto err_spi_register_master;
+	}
+
+	return 0;
+
+err_spi_register_master:
+	pch_free_dma_buf(board_dat, data);
+	free_irq(board_dat->pdev->irq, data);
+err_request_irq:
+	pch_spi_free_resources(board_dat, data);
+err_spi_get_resources:
+	pci_iounmap(board_dat->pdev, data->io_remap_addr);
+err_pci_iomap:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int pch_spi_pd_remove(struct platform_device *plat_dev)
+{
+	struct pch_spi_board_data *board_dat = dev_get_platdata(&plat_dev->dev);
+	struct pch_spi_data *data = platform_get_drvdata(plat_dev);
+	int count;
+	unsigned long flags;
+
+	dev_dbg(&plat_dev->dev, "%s:[ch%d] irq=%d\n",
+		__func__, plat_dev->id, board_dat->pdev->irq);
+
+	if (use_dma)
+		pch_free_dma_buf(board_dat, data);
+
+	/* check for any pending messages; no action is taken if the queue
+	 * is still full; but at least we tried.  Unload anyway */
+	count = 500;
+	spin_lock_irqsave(&data->lock, flags);
+	data->status = STATUS_EXITING;
+	while ((list_empty(&data->queue) == 0) && --count) {
+		dev_dbg(&board_dat->pdev->dev, "%s :queue not empty\n",
+			__func__);
+		spin_unlock_irqrestore(&data->lock, flags);
+		msleep(PCH_SLEEP_TIME);
+		spin_lock_irqsave(&data->lock, flags);
+	}
+	spin_unlock_irqrestore(&data->lock, flags);
+
+	pch_spi_free_resources(board_dat, data);
+	/* disable interrupts & free IRQ */
+	if (data->irq_reg_sts) {
+		/* disable interrupts */
+		pch_spi_setclr_reg(data->master, PCH_SPCR, 0, PCH_ALL);
+		data->irq_reg_sts = false;
+		free_irq(board_dat->pdev->irq, data);
+	}
+
+	pci_iounmap(board_dat->pdev, data->io_remap_addr);
+	spi_unregister_master(data->master);
+
+	return 0;
+}
+#ifdef CONFIG_PM
+static int pch_spi_pd_suspend(struct platform_device *pd_dev,
+			      pm_message_t state)
+{
+	u8 count;
+	struct pch_spi_board_data *board_dat = dev_get_platdata(&pd_dev->dev);
+	struct pch_spi_data *data = platform_get_drvdata(pd_dev);
+
+	dev_dbg(&pd_dev->dev, "%s ENTRY\n", __func__);
+
+	if (!board_dat) {
+		dev_err(&pd_dev->dev,
+			"%s pci_get_drvdata returned NULL\n", __func__);
+		return -EFAULT;
+	}
+
+	/* check if the current message is processed:
+	   Only after thats done the transfer will be suspended */
+	count = 255;
+	while ((--count) > 0) {
+		if (!(data->bcurrent_msg_processing))
+			break;
+		msleep(PCH_SLEEP_TIME);
+	}
+
+	/* Free IRQ */
+	if (data->irq_reg_sts) {
+		/* disable all interrupts */
+		pch_spi_setclr_reg(data->master, PCH_SPCR, 0, PCH_ALL);
+		pch_spi_reset(data->master);
+		free_irq(board_dat->pdev->irq, data);
+
+		data->irq_reg_sts = false;
+		dev_dbg(&pd_dev->dev,
+			"%s free_irq invoked successfully.\n", __func__);
+	}
+
+	return 0;
+}
+
+static int pch_spi_pd_resume(struct platform_device *pd_dev)
+{
+	struct pch_spi_board_data *board_dat = dev_get_platdata(&pd_dev->dev);
+	struct pch_spi_data *data = platform_get_drvdata(pd_dev);
+	int retval;
+
+	if (!board_dat) {
+		dev_err(&pd_dev->dev,
+			"%s pci_get_drvdata returned NULL\n", __func__);
+		return -EFAULT;
+	}
+
+	if (!data->irq_reg_sts) {
+		/* register IRQ */
+		retval = request_irq(board_dat->pdev->irq, pch_spi_handler,
+				     IRQF_SHARED, KBUILD_MODNAME, data);
+		if (retval < 0) {
+			dev_err(&pd_dev->dev,
+				"%s request_irq failed\n", __func__);
+			return retval;
+		}
+
+		/* reset PCH SPI h/w */
+		pch_spi_reset(data->master);
+		pch_spi_set_master_mode(data->master);
+		data->irq_reg_sts = true;
+	}
+	return 0;
+}
+#else
+#define pch_spi_pd_suspend NULL
+#define pch_spi_pd_resume NULL
+#endif
+
+static struct platform_driver pch_spi_pd_driver = {
+	.driver = {
+		.name = "pch-spi",
+	},
+	.probe = pch_spi_pd_probe,
+	.remove = pch_spi_pd_remove,
+	.suspend = pch_spi_pd_suspend,
+	.resume = pch_spi_pd_resume
+};
+
+static int pch_spi_probe(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+	struct pch_spi_board_data *board_dat;
+	struct platform_device *pd_dev = NULL;
+	int retval;
+	int i;
+	struct pch_pd_dev_save *pd_dev_save;
+
+	pd_dev_save = kzalloc(sizeof(*pd_dev_save), GFP_KERNEL);
+	if (!pd_dev_save)
+		return -ENOMEM;
+
+	board_dat = kzalloc(sizeof(*board_dat), GFP_KERNEL);
+	if (!board_dat) {
+		retval = -ENOMEM;
+		goto err_no_mem;
+	}
+
+	retval = pci_request_regions(pdev, KBUILD_MODNAME);
+	if (retval) {
+		dev_err(&pdev->dev, "%s request_region failed\n", __func__);
+		goto pci_request_regions;
+	}
+
+	board_dat->pdev = pdev;
+	board_dat->num = id->driver_data;
+	pd_dev_save->num = id->driver_data;
+	pd_dev_save->board_dat = board_dat;
+
+	retval = pci_enable_device(pdev);
+	if (retval) {
+		dev_err(&pdev->dev, "%s pci_enable_device failed\n", __func__);
+		goto pci_enable_device;
+	}
+
+	for (i = 0; i < board_dat->num; i++) {
+		pd_dev = platform_device_alloc("pch-spi", i);
+		if (!pd_dev) {
+			dev_err(&pdev->dev, "platform_device_alloc failed\n");
+			retval = -ENOMEM;
+			goto err_platform_device;
+		}
+		pd_dev_save->pd_save[i] = pd_dev;
+		pd_dev->dev.parent = &pdev->dev;
+
+		retval = platform_device_add_data(pd_dev, board_dat,
+						  sizeof(*board_dat));
+		if (retval) {
+			dev_err(&pdev->dev,
+				"platform_device_add_data failed\n");
+			platform_device_put(pd_dev);
+			goto err_platform_device;
+		}
+
+		retval = platform_device_add(pd_dev);
+		if (retval) {
+			dev_err(&pdev->dev, "platform_device_add failed\n");
+			platform_device_put(pd_dev);
+			goto err_platform_device;
+		}
+	}
+
+	pci_set_drvdata(pdev, pd_dev_save);
+
+	return 0;
+
+err_platform_device:
+	while (--i >= 0)
+		platform_device_unregister(pd_dev_save->pd_save[i]);
+	pci_disable_device(pdev);
+pci_enable_device:
+	pci_release_regions(pdev);
+pci_request_regions:
+	kfree(board_dat);
+err_no_mem:
+	kfree(pd_dev_save);
+
+	return retval;
+}
+
+static void pch_spi_remove(struct pci_dev *pdev)
+{
+	int i;
+	struct pch_pd_dev_save *pd_dev_save = pci_get_drvdata(pdev);
+
+	dev_dbg(&pdev->dev, "%s ENTRY:pdev=%p\n", __func__, pdev);
+
+	for (i = 0; i < pd_dev_save->num; i++)
+		platform_device_unregister(pd_dev_save->pd_save[i]);
+
+	pci_disable_device(pdev);
+	pci_release_regions(pdev);
+	kfree(pd_dev_save->board_dat);
+	kfree(pd_dev_save);
+}
+
+#ifdef CONFIG_PM
+static int pch_spi_suspend(struct pci_dev *pdev, pm_message_t state)
+{
+	int retval;
+	struct pch_pd_dev_save *pd_dev_save = pci_get_drvdata(pdev);
+
+	dev_dbg(&pdev->dev, "%s ENTRY\n", __func__);
+
+	pd_dev_save->board_dat->suspend_sts = true;
+
+	/* save config space */
+	retval = pci_save_state(pdev);
+	if (retval == 0) {
+		pci_enable_wake(pdev, PCI_D3hot, 0);
+		pci_disable_device(pdev);
+		pci_set_power_state(pdev, PCI_D3hot);
+	} else {
+		dev_err(&pdev->dev, "%s pci_save_state failed\n", __func__);
+	}
+
+	return retval;
+}
+
+static int pch_spi_resume(struct pci_dev *pdev)
+{
+	int retval;
+	struct pch_pd_dev_save *pd_dev_save = pci_get_drvdata(pdev);
+	dev_dbg(&pdev->dev, "%s ENTRY\n", __func__);
+
+	pci_set_power_state(pdev, PCI_D0);
+	pci_restore_state(pdev);
+
+	retval = pci_enable_device(pdev);
+	if (retval < 0) {
+		dev_err(&pdev->dev,
+			"%s pci_enable_device failed\n", __func__);
+	} else {
+		pci_enable_wake(pdev, PCI_D3hot, 0);
+
+		/* set suspend status to false */
+		pd_dev_save->board_dat->suspend_sts = false;
+	}
+
+	return retval;
+}
+#else
+#define pch_spi_suspend NULL
+#define pch_spi_resume NULL
+
+#endif
+
+static struct pci_driver pch_spi_pcidev_driver = {
+	.name = "pch_spi",
+	.id_table = pch_spi_pcidev_id,
+	.probe = pch_spi_probe,
+	.remove = pch_spi_remove,
+	.suspend = pch_spi_suspend,
+	.resume = pch_spi_resume,
+};
+
+static int __init pch_spi_init(void)
+{
+	int ret;
+	ret = platform_driver_register(&pch_spi_pd_driver);
+	if (ret)
+		return ret;
+
+	ret = pci_register_driver(&pch_spi_pcidev_driver);
+	if (ret) {
+		platform_driver_unregister(&pch_spi_pd_driver);
+		return ret;
+	}
+
+	return 0;
+}
+module_init(pch_spi_init);
+
+static void __exit pch_spi_exit(void)
+{
+	pci_unregister_driver(&pch_spi_pcidev_driver);
+	platform_driver_unregister(&pch_spi_pd_driver);
+}
+module_exit(pch_spi_exit);
+
+module_param(use_dma, int, 0644);
+MODULE_PARM_DESC(use_dma,
+		 "to use DMA for data transfers pass 1 else 0; default 1");
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Intel EG20T PCH/LAPIS Semiconductor ML7xxx IOH SPI Driver");
+MODULE_DEVICE_TABLE(pci, pch_spi_pcidev_id);
+
diff --git a/drivers/spi/spi-txx9.c b/drivers/spi/spi-txx9.c
new file mode 100644
index 0000000..51759d3
--- /dev/null
+++ b/drivers/spi/spi-txx9.c
@@ -0,0 +1,437 @@
+/*
+ * TXx9 SPI controller driver.
+ *
+ * Based on linux/arch/mips/tx4938/toshiba_rbtx4938/spi_txx9.c
+ * Copyright (C) 2000-2001 Toshiba Corporation
+ *
+ * 2003-2005 (c) MontaVista Software, Inc. This file is licensed under the
+ * terms of the GNU General Public License version 2. This program is
+ * licensed "as is" without any warranty of any kind, whether express
+ * or implied.
+ *
+ * Support for TX4938 in 2.6 - Manish Lachwani (mlachwani@mvista.com)
+ *
+ * Convert to generic SPI framework - Atsushi Nemoto (anemo@mba.ocn.ne.jp)
+ */
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/sched.h>
+#include <linux/spinlock.h>
+#include <linux/workqueue.h>
+#include <linux/spi/spi.h>
+#include <linux/err.h>
+#include <linux/clk.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/gpio.h>
+
+
+#define SPI_FIFO_SIZE 4
+#define SPI_MAX_DIVIDER 0xff	/* Max. value for SPCR1.SER */
+#define SPI_MIN_DIVIDER 1	/* Min. value for SPCR1.SER */
+
+#define TXx9_SPMCR		0x00
+#define TXx9_SPCR0		0x04
+#define TXx9_SPCR1		0x08
+#define TXx9_SPFS		0x0c
+#define TXx9_SPSR		0x14
+#define TXx9_SPDR		0x18
+
+/* SPMCR : SPI Master Control */
+#define TXx9_SPMCR_OPMODE	0xc0
+#define TXx9_SPMCR_CONFIG	0x40
+#define TXx9_SPMCR_ACTIVE	0x80
+#define TXx9_SPMCR_SPSTP	0x02
+#define TXx9_SPMCR_BCLR		0x01
+
+/* SPCR0 : SPI Control 0 */
+#define TXx9_SPCR0_TXIFL_MASK	0xc000
+#define TXx9_SPCR0_RXIFL_MASK	0x3000
+#define TXx9_SPCR0_SIDIE	0x0800
+#define TXx9_SPCR0_SOEIE	0x0400
+#define TXx9_SPCR0_RBSIE	0x0200
+#define TXx9_SPCR0_TBSIE	0x0100
+#define TXx9_SPCR0_IFSPSE	0x0010
+#define TXx9_SPCR0_SBOS		0x0004
+#define TXx9_SPCR0_SPHA		0x0002
+#define TXx9_SPCR0_SPOL		0x0001
+
+/* SPSR : SPI Status */
+#define TXx9_SPSR_TBSI		0x8000
+#define TXx9_SPSR_RBSI		0x4000
+#define TXx9_SPSR_TBS_MASK	0x3800
+#define TXx9_SPSR_RBS_MASK	0x0700
+#define TXx9_SPSR_SPOE		0x0080
+#define TXx9_SPSR_IFSD		0x0008
+#define TXx9_SPSR_SIDLE		0x0004
+#define TXx9_SPSR_STRDY		0x0002
+#define TXx9_SPSR_SRRDY		0x0001
+
+
+struct txx9spi {
+	struct work_struct work;
+	spinlock_t lock;	/* protect 'queue' */
+	struct list_head queue;
+	wait_queue_head_t waitq;
+	void __iomem *membase;
+	int baseclk;
+	struct clk *clk;
+	int last_chipselect;
+	int last_chipselect_val;
+};
+
+static u32 txx9spi_rd(struct txx9spi *c, int reg)
+{
+	return __raw_readl(c->membase + reg);
+}
+static void txx9spi_wr(struct txx9spi *c, u32 val, int reg)
+{
+	__raw_writel(val, c->membase + reg);
+}
+
+static void txx9spi_cs_func(struct spi_device *spi, struct txx9spi *c,
+		int on, unsigned int cs_delay)
+{
+	int val = (spi->mode & SPI_CS_HIGH) ? on : !on;
+
+	if (on) {
+		/* deselect the chip with cs_change hint in last transfer */
+		if (c->last_chipselect >= 0)
+			gpio_set_value(c->last_chipselect,
+					!c->last_chipselect_val);
+		c->last_chipselect = spi->chip_select;
+		c->last_chipselect_val = val;
+	} else {
+		c->last_chipselect = -1;
+		ndelay(cs_delay);	/* CS Hold Time */
+	}
+	gpio_set_value(spi->chip_select, val);
+	ndelay(cs_delay);	/* CS Setup Time / CS Recovery Time */
+}
+
+static int txx9spi_setup(struct spi_device *spi)
+{
+	struct txx9spi *c = spi_master_get_devdata(spi->master);
+
+	if (!spi->max_speed_hz)
+		return -EINVAL;
+
+	if (gpio_direction_output(spi->chip_select,
+			!(spi->mode & SPI_CS_HIGH))) {
+		dev_err(&spi->dev, "Cannot setup GPIO for chipselect.\n");
+		return -EINVAL;
+	}
+
+	/* deselect chip */
+	spin_lock(&c->lock);
+	txx9spi_cs_func(spi, c, 0, (NSEC_PER_SEC / 2) / spi->max_speed_hz);
+	spin_unlock(&c->lock);
+
+	return 0;
+}
+
+static irqreturn_t txx9spi_interrupt(int irq, void *dev_id)
+{
+	struct txx9spi *c = dev_id;
+
+	/* disable rx intr */
+	txx9spi_wr(c, txx9spi_rd(c, TXx9_SPCR0) & ~TXx9_SPCR0_RBSIE,
+			TXx9_SPCR0);
+	wake_up(&c->waitq);
+	return IRQ_HANDLED;
+}
+
+static void txx9spi_work_one(struct txx9spi *c, struct spi_message *m)
+{
+	struct spi_device *spi = m->spi;
+	struct spi_transfer *t;
+	unsigned int cs_delay;
+	unsigned int cs_change = 1;
+	int status = 0;
+	u32 mcr;
+	u32 prev_speed_hz = 0;
+	u8 prev_bits_per_word = 0;
+
+	/* CS setup/hold/recovery time in nsec */
+	cs_delay = 100 + (NSEC_PER_SEC / 2) / spi->max_speed_hz;
+
+	mcr = txx9spi_rd(c, TXx9_SPMCR);
+	if (unlikely((mcr & TXx9_SPMCR_OPMODE) == TXx9_SPMCR_ACTIVE)) {
+		dev_err(&spi->dev, "Bad mode.\n");
+		status = -EIO;
+		goto exit;
+	}
+	mcr &= ~(TXx9_SPMCR_OPMODE | TXx9_SPMCR_SPSTP | TXx9_SPMCR_BCLR);
+
+	/* enter config mode */
+	txx9spi_wr(c, mcr | TXx9_SPMCR_CONFIG | TXx9_SPMCR_BCLR, TXx9_SPMCR);
+	txx9spi_wr(c, TXx9_SPCR0_SBOS
+			| ((spi->mode & SPI_CPOL) ? TXx9_SPCR0_SPOL : 0)
+			| ((spi->mode & SPI_CPHA) ? TXx9_SPCR0_SPHA : 0)
+			| 0x08,
+			TXx9_SPCR0);
+
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		const void *txbuf = t->tx_buf;
+		void *rxbuf = t->rx_buf;
+		u32 data;
+		unsigned int len = t->len;
+		unsigned int wsize;
+		u32 speed_hz = t->speed_hz;
+		u8 bits_per_word = t->bits_per_word;
+
+		wsize = bits_per_word >> 3; /* in bytes */
+
+		if (prev_speed_hz != speed_hz
+				|| prev_bits_per_word != bits_per_word) {
+			int n = DIV_ROUND_UP(c->baseclk, speed_hz) - 1;
+
+			n = clamp(n, SPI_MIN_DIVIDER, SPI_MAX_DIVIDER);
+			/* enter config mode */
+			txx9spi_wr(c, mcr | TXx9_SPMCR_CONFIG | TXx9_SPMCR_BCLR,
+					TXx9_SPMCR);
+			txx9spi_wr(c, (n << 8) | bits_per_word, TXx9_SPCR1);
+			/* enter active mode */
+			txx9spi_wr(c, mcr | TXx9_SPMCR_ACTIVE, TXx9_SPMCR);
+
+			prev_speed_hz = speed_hz;
+			prev_bits_per_word = bits_per_word;
+		}
+
+		if (cs_change)
+			txx9spi_cs_func(spi, c, 1, cs_delay);
+		cs_change = t->cs_change;
+		while (len) {
+			unsigned int count = SPI_FIFO_SIZE;
+			int i;
+			u32 cr0;
+
+			if (len < count * wsize)
+				count = len / wsize;
+			/* now tx must be idle... */
+			while (!(txx9spi_rd(c, TXx9_SPSR) & TXx9_SPSR_SIDLE))
+				cpu_relax();
+			cr0 = txx9spi_rd(c, TXx9_SPCR0);
+			cr0 &= ~TXx9_SPCR0_RXIFL_MASK;
+			cr0 |= (count - 1) << 12;
+			/* enable rx intr */
+			cr0 |= TXx9_SPCR0_RBSIE;
+			txx9spi_wr(c, cr0, TXx9_SPCR0);
+			/* send */
+			for (i = 0; i < count; i++) {
+				if (txbuf) {
+					data = (wsize == 1)
+						? *(const u8 *)txbuf
+						: *(const u16 *)txbuf;
+					txx9spi_wr(c, data, TXx9_SPDR);
+					txbuf += wsize;
+				} else
+					txx9spi_wr(c, 0, TXx9_SPDR);
+			}
+			/* wait all rx data */
+			wait_event(c->waitq,
+				txx9spi_rd(c, TXx9_SPSR) & TXx9_SPSR_RBSI);
+			/* receive */
+			for (i = 0; i < count; i++) {
+				data = txx9spi_rd(c, TXx9_SPDR);
+				if (rxbuf) {
+					if (wsize == 1)
+						*(u8 *)rxbuf = data;
+					else
+						*(u16 *)rxbuf = data;
+					rxbuf += wsize;
+				}
+			}
+			len -= count * wsize;
+		}
+		m->actual_length += t->len;
+		if (t->delay_usecs)
+			udelay(t->delay_usecs);
+
+		if (!cs_change)
+			continue;
+		if (t->transfer_list.next == &m->transfers)
+			break;
+		/* sometimes a short mid-message deselect of the chip
+		 * may be needed to terminate a mode or command
+		 */
+		txx9spi_cs_func(spi, c, 0, cs_delay);
+	}
+
+exit:
+	m->status = status;
+	if (m->complete)
+		m->complete(m->context);
+
+	/* normally deactivate chipselect ... unless no error and
+	 * cs_change has hinted that the next message will probably
+	 * be for this chip too.
+	 */
+	if (!(status == 0 && cs_change))
+		txx9spi_cs_func(spi, c, 0, cs_delay);
+
+	/* enter config mode */
+	txx9spi_wr(c, mcr | TXx9_SPMCR_CONFIG | TXx9_SPMCR_BCLR, TXx9_SPMCR);
+}
+
+static void txx9spi_work(struct work_struct *work)
+{
+	struct txx9spi *c = container_of(work, struct txx9spi, work);
+	unsigned long flags;
+
+	spin_lock_irqsave(&c->lock, flags);
+	while (!list_empty(&c->queue)) {
+		struct spi_message *m;
+
+		m = container_of(c->queue.next, struct spi_message, queue);
+		list_del_init(&m->queue);
+		spin_unlock_irqrestore(&c->lock, flags);
+
+		txx9spi_work_one(c, m);
+
+		spin_lock_irqsave(&c->lock, flags);
+	}
+	spin_unlock_irqrestore(&c->lock, flags);
+}
+
+static int txx9spi_transfer(struct spi_device *spi, struct spi_message *m)
+{
+	struct spi_master *master = spi->master;
+	struct txx9spi *c = spi_master_get_devdata(master);
+	struct spi_transfer *t;
+	unsigned long flags;
+
+	m->actual_length = 0;
+
+	/* check each transfer's parameters */
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if (!t->tx_buf && !t->rx_buf && t->len)
+			return -EINVAL;
+	}
+
+	spin_lock_irqsave(&c->lock, flags);
+	list_add_tail(&m->queue, &c->queue);
+	schedule_work(&c->work);
+	spin_unlock_irqrestore(&c->lock, flags);
+
+	return 0;
+}
+
+static int txx9spi_probe(struct platform_device *dev)
+{
+	struct spi_master *master;
+	struct txx9spi *c;
+	struct resource *res;
+	int ret = -ENODEV;
+	u32 mcr;
+	int irq;
+
+	master = spi_alloc_master(&dev->dev, sizeof(*c));
+	if (!master)
+		return ret;
+	c = spi_master_get_devdata(master);
+	platform_set_drvdata(dev, master);
+
+	INIT_WORK(&c->work, txx9spi_work);
+	spin_lock_init(&c->lock);
+	INIT_LIST_HEAD(&c->queue);
+	init_waitqueue_head(&c->waitq);
+
+	c->clk = devm_clk_get(&dev->dev, "spi-baseclk");
+	if (IS_ERR(c->clk)) {
+		ret = PTR_ERR(c->clk);
+		c->clk = NULL;
+		goto exit;
+	}
+	ret = clk_prepare_enable(c->clk);
+	if (ret) {
+		c->clk = NULL;
+		goto exit;
+	}
+	c->baseclk = clk_get_rate(c->clk);
+	master->min_speed_hz = DIV_ROUND_UP(c->baseclk, SPI_MAX_DIVIDER + 1);
+	master->max_speed_hz = c->baseclk / (SPI_MIN_DIVIDER + 1);
+
+	res = platform_get_resource(dev, IORESOURCE_MEM, 0);
+	c->membase = devm_ioremap_resource(&dev->dev, res);
+	if (IS_ERR(c->membase))
+		goto exit_busy;
+
+	/* enter config mode */
+	mcr = txx9spi_rd(c, TXx9_SPMCR);
+	mcr &= ~(TXx9_SPMCR_OPMODE | TXx9_SPMCR_SPSTP | TXx9_SPMCR_BCLR);
+	txx9spi_wr(c, mcr | TXx9_SPMCR_CONFIG | TXx9_SPMCR_BCLR, TXx9_SPMCR);
+
+	irq = platform_get_irq(dev, 0);
+	if (irq < 0)
+		goto exit_busy;
+	ret = devm_request_irq(&dev->dev, irq, txx9spi_interrupt, 0,
+			       "spi_txx9", c);
+	if (ret)
+		goto exit;
+
+	c->last_chipselect = -1;
+
+	dev_info(&dev->dev, "at %#llx, irq %d, %dMHz\n",
+		 (unsigned long long)res->start, irq,
+		 (c->baseclk + 500000) / 1000000);
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CS_HIGH | SPI_CPOL | SPI_CPHA;
+
+	master->bus_num = dev->id;
+	master->setup = txx9spi_setup;
+	master->transfer = txx9spi_transfer;
+	master->num_chipselect = (u16)UINT_MAX; /* any GPIO numbers */
+	master->bits_per_word_mask = SPI_BPW_MASK(8) | SPI_BPW_MASK(16);
+
+	ret = devm_spi_register_master(&dev->dev, master);
+	if (ret)
+		goto exit;
+	return 0;
+exit_busy:
+	ret = -EBUSY;
+exit:
+	clk_disable_unprepare(c->clk);
+	spi_master_put(master);
+	return ret;
+}
+
+static int txx9spi_remove(struct platform_device *dev)
+{
+	struct spi_master *master = platform_get_drvdata(dev);
+	struct txx9spi *c = spi_master_get_devdata(master);
+
+	flush_work(&c->work);
+	clk_disable_unprepare(c->clk);
+	return 0;
+}
+
+/* work with hotplug and coldplug */
+MODULE_ALIAS("platform:spi_txx9");
+
+static struct platform_driver txx9spi_driver = {
+	.probe = txx9spi_probe,
+	.remove = txx9spi_remove,
+	.driver = {
+		.name = "spi_txx9",
+	},
+};
+
+static int __init txx9spi_init(void)
+{
+	return platform_driver_register(&txx9spi_driver);
+}
+subsys_initcall(txx9spi_init);
+
+static void __exit txx9spi_exit(void)
+{
+	platform_driver_unregister(&txx9spi_driver);
+}
+module_exit(txx9spi_exit);
+
+MODULE_DESCRIPTION("TXx9 SPI Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-uniphier.c b/drivers/spi/spi-uniphier.c
new file mode 100644
index 0000000..5a6137f
--- /dev/null
+++ b/drivers/spi/spi-uniphier.c
@@ -0,0 +1,523 @@
+// SPDX-License-Identifier: GPL-2.0
+// spi-uniphier.c - Socionext UniPhier SPI controller driver
+// Copyright 2012      Panasonic Corporation
+// Copyright 2016-2018 Socionext Inc.
+
+#include <linux/kernel.h>
+#include <linux/bitfield.h>
+#include <linux/bitops.h>
+#include <linux/clk.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+
+#include <asm/unaligned.h>
+
+#define SSI_TIMEOUT_MS		2000
+#define SSI_MAX_CLK_DIVIDER	254
+#define SSI_MIN_CLK_DIVIDER	4
+
+struct uniphier_spi_priv {
+	void __iomem *base;
+	struct clk *clk;
+	struct spi_master *master;
+	struct completion xfer_done;
+
+	int error;
+	unsigned int tx_bytes;
+	unsigned int rx_bytes;
+	const u8 *tx_buf;
+	u8 *rx_buf;
+
+	bool is_save_param;
+	u8 bits_per_word;
+	u16 mode;
+	u32 speed_hz;
+};
+
+#define SSI_CTL			0x00
+#define   SSI_CTL_EN		BIT(0)
+
+#define SSI_CKS			0x04
+#define   SSI_CKS_CKRAT_MASK	GENMASK(7, 0)
+#define   SSI_CKS_CKPHS		BIT(14)
+#define   SSI_CKS_CKINIT	BIT(13)
+#define   SSI_CKS_CKDLY		BIT(12)
+
+#define SSI_TXWDS		0x08
+#define   SSI_TXWDS_WDLEN_MASK	GENMASK(13, 8)
+#define   SSI_TXWDS_TDTF_MASK	GENMASK(7, 6)
+#define   SSI_TXWDS_DTLEN_MASK	GENMASK(5, 0)
+
+#define SSI_RXWDS		0x0c
+#define   SSI_RXWDS_DTLEN_MASK	GENMASK(5, 0)
+
+#define SSI_FPS			0x10
+#define   SSI_FPS_FSPOL		BIT(15)
+#define   SSI_FPS_FSTRT		BIT(14)
+
+#define SSI_SR			0x14
+#define   SSI_SR_RNE		BIT(0)
+
+#define SSI_IE			0x18
+#define   SSI_IE_RCIE		BIT(3)
+#define   SSI_IE_RORIE		BIT(0)
+
+#define SSI_IS			0x1c
+#define   SSI_IS_RXRS		BIT(9)
+#define   SSI_IS_RCID		BIT(3)
+#define   SSI_IS_RORID		BIT(0)
+
+#define SSI_IC			0x1c
+#define   SSI_IC_TCIC		BIT(4)
+#define   SSI_IC_RCIC		BIT(3)
+#define   SSI_IC_RORIC		BIT(0)
+
+#define SSI_FC			0x20
+#define   SSI_FC_TXFFL		BIT(12)
+#define   SSI_FC_TXFTH_MASK	GENMASK(11, 8)
+#define   SSI_FC_RXFFL		BIT(4)
+#define   SSI_FC_RXFTH_MASK	GENMASK(3, 0)
+
+#define SSI_TXDR		0x24
+#define SSI_RXDR		0x24
+
+#define SSI_FIFO_DEPTH		8U
+
+static inline unsigned int bytes_per_word(unsigned int bits)
+{
+	return bits <= 8 ? 1 : (bits <= 16 ? 2 : 4);
+}
+
+static inline void uniphier_spi_irq_enable(struct spi_device *spi, u32 mask)
+{
+	struct uniphier_spi_priv *priv = spi_master_get_devdata(spi->master);
+	u32 val;
+
+	val = readl(priv->base + SSI_IE);
+	val |= mask;
+	writel(val, priv->base + SSI_IE);
+}
+
+static inline void uniphier_spi_irq_disable(struct spi_device *spi, u32 mask)
+{
+	struct uniphier_spi_priv *priv = spi_master_get_devdata(spi->master);
+	u32 val;
+
+	val = readl(priv->base + SSI_IE);
+	val &= ~mask;
+	writel(val, priv->base + SSI_IE);
+}
+
+static void uniphier_spi_set_mode(struct spi_device *spi)
+{
+	struct uniphier_spi_priv *priv = spi_master_get_devdata(spi->master);
+	u32 val1, val2;
+
+	/*
+	 * clock setting
+	 * CKPHS    capture timing. 0:rising edge, 1:falling edge
+	 * CKINIT   clock initial level. 0:low, 1:high
+	 * CKDLY    clock delay. 0:no delay, 1:delay depending on FSTRT
+	 *          (FSTRT=0: 1 clock, FSTRT=1: 0.5 clock)
+	 *
+	 * frame setting
+	 * FSPOL    frame signal porarity. 0: low, 1: high
+	 * FSTRT    start frame timing
+	 *          0: rising edge of clock, 1: falling edge of clock
+	 */
+	switch (spi->mode & (SPI_CPOL | SPI_CPHA)) {
+	case SPI_MODE_0:
+		/* CKPHS=1, CKINIT=0, CKDLY=1, FSTRT=0 */
+		val1 = SSI_CKS_CKPHS | SSI_CKS_CKDLY;
+		val2 = 0;
+		break;
+	case SPI_MODE_1:
+		/* CKPHS=0, CKINIT=0, CKDLY=0, FSTRT=1 */
+		val1 = 0;
+		val2 = SSI_FPS_FSTRT;
+		break;
+	case SPI_MODE_2:
+		/* CKPHS=0, CKINIT=1, CKDLY=1, FSTRT=1 */
+		val1 = SSI_CKS_CKINIT | SSI_CKS_CKDLY;
+		val2 = SSI_FPS_FSTRT;
+		break;
+	case SPI_MODE_3:
+		/* CKPHS=1, CKINIT=1, CKDLY=0, FSTRT=0 */
+		val1 = SSI_CKS_CKPHS | SSI_CKS_CKINIT;
+		val2 = 0;
+		break;
+	}
+
+	if (!(spi->mode & SPI_CS_HIGH))
+		val2 |= SSI_FPS_FSPOL;
+
+	writel(val1, priv->base + SSI_CKS);
+	writel(val2, priv->base + SSI_FPS);
+
+	val1 = 0;
+	if (spi->mode & SPI_LSB_FIRST)
+		val1 |= FIELD_PREP(SSI_TXWDS_TDTF_MASK, 1);
+	writel(val1, priv->base + SSI_TXWDS);
+	writel(val1, priv->base + SSI_RXWDS);
+}
+
+static void uniphier_spi_set_transfer_size(struct spi_device *spi, int size)
+{
+	struct uniphier_spi_priv *priv = spi_master_get_devdata(spi->master);
+	u32 val;
+
+	val = readl(priv->base + SSI_TXWDS);
+	val &= ~(SSI_TXWDS_WDLEN_MASK | SSI_TXWDS_DTLEN_MASK);
+	val |= FIELD_PREP(SSI_TXWDS_WDLEN_MASK, size);
+	val |= FIELD_PREP(SSI_TXWDS_DTLEN_MASK, size);
+	writel(val, priv->base + SSI_TXWDS);
+
+	val = readl(priv->base + SSI_RXWDS);
+	val &= ~SSI_RXWDS_DTLEN_MASK;
+	val |= FIELD_PREP(SSI_RXWDS_DTLEN_MASK, size);
+	writel(val, priv->base + SSI_RXWDS);
+}
+
+static void uniphier_spi_set_baudrate(struct spi_device *spi,
+				      unsigned int speed)
+{
+	struct uniphier_spi_priv *priv = spi_master_get_devdata(spi->master);
+	u32 val, ckdiv;
+
+	/*
+	 * the supported rates are even numbers from 4 to 254. (4,6,8...254)
+	 * round up as we look for equal or less speed
+	 */
+	ckdiv = DIV_ROUND_UP(clk_get_rate(priv->clk), speed);
+	ckdiv = round_up(ckdiv, 2);
+
+	val = readl(priv->base + SSI_CKS);
+	val &= ~SSI_CKS_CKRAT_MASK;
+	val |= ckdiv & SSI_CKS_CKRAT_MASK;
+	writel(val, priv->base + SSI_CKS);
+}
+
+static void uniphier_spi_setup_transfer(struct spi_device *spi,
+				       struct spi_transfer *t)
+{
+	struct uniphier_spi_priv *priv = spi_master_get_devdata(spi->master);
+	u32 val;
+
+	priv->error = 0;
+	priv->tx_buf = t->tx_buf;
+	priv->rx_buf = t->rx_buf;
+	priv->tx_bytes = priv->rx_bytes = t->len;
+
+	if (!priv->is_save_param || priv->mode != spi->mode) {
+		uniphier_spi_set_mode(spi);
+		priv->mode = spi->mode;
+	}
+
+	if (!priv->is_save_param || priv->bits_per_word != t->bits_per_word) {
+		uniphier_spi_set_transfer_size(spi, t->bits_per_word);
+		priv->bits_per_word = t->bits_per_word;
+	}
+
+	if (!priv->is_save_param || priv->speed_hz != t->speed_hz) {
+		uniphier_spi_set_baudrate(spi, t->speed_hz);
+		priv->speed_hz = t->speed_hz;
+	}
+
+	if (!priv->is_save_param)
+		priv->is_save_param = true;
+
+	/* reset FIFOs */
+	val = SSI_FC_TXFFL | SSI_FC_RXFFL;
+	writel(val, priv->base + SSI_FC);
+}
+
+static void uniphier_spi_send(struct uniphier_spi_priv *priv)
+{
+	int wsize;
+	u32 val = 0;
+
+	wsize = min(bytes_per_word(priv->bits_per_word), priv->tx_bytes);
+	priv->tx_bytes -= wsize;
+
+	if (priv->tx_buf) {
+		switch (wsize) {
+		case 1:
+			val = *priv->tx_buf;
+			break;
+		case 2:
+			val = get_unaligned_le16(priv->tx_buf);
+			break;
+		case 4:
+			val = get_unaligned_le32(priv->tx_buf);
+			break;
+		}
+
+		priv->tx_buf += wsize;
+	}
+
+	writel(val, priv->base + SSI_TXDR);
+}
+
+static void uniphier_spi_recv(struct uniphier_spi_priv *priv)
+{
+	int rsize;
+	u32 val;
+
+	rsize = min(bytes_per_word(priv->bits_per_word), priv->rx_bytes);
+	priv->rx_bytes -= rsize;
+
+	val = readl(priv->base + SSI_RXDR);
+
+	if (priv->rx_buf) {
+		switch (rsize) {
+		case 1:
+			*priv->rx_buf = val;
+			break;
+		case 2:
+			put_unaligned_le16(val, priv->rx_buf);
+			break;
+		case 4:
+			put_unaligned_le32(val, priv->rx_buf);
+			break;
+		}
+
+		priv->rx_buf += rsize;
+	}
+}
+
+static void uniphier_spi_fill_tx_fifo(struct uniphier_spi_priv *priv)
+{
+	unsigned int tx_count;
+	u32 val;
+
+	tx_count = DIV_ROUND_UP(priv->tx_bytes,
+				bytes_per_word(priv->bits_per_word));
+	tx_count = min(tx_count, SSI_FIFO_DEPTH);
+
+	/* set fifo threshold */
+	val = readl(priv->base + SSI_FC);
+	val &= ~(SSI_FC_TXFTH_MASK | SSI_FC_RXFTH_MASK);
+	val |= FIELD_PREP(SSI_FC_TXFTH_MASK, tx_count);
+	val |= FIELD_PREP(SSI_FC_RXFTH_MASK, tx_count);
+	writel(val, priv->base + SSI_FC);
+
+	while (tx_count--)
+		uniphier_spi_send(priv);
+}
+
+static void uniphier_spi_set_cs(struct spi_device *spi, bool enable)
+{
+	struct uniphier_spi_priv *priv = spi_master_get_devdata(spi->master);
+	u32 val;
+
+	val = readl(priv->base + SSI_FPS);
+
+	if (enable)
+		val |= SSI_FPS_FSPOL;
+	else
+		val &= ~SSI_FPS_FSPOL;
+
+	writel(val, priv->base + SSI_FPS);
+}
+
+static int uniphier_spi_transfer_one(struct spi_master *master,
+				     struct spi_device *spi,
+				     struct spi_transfer *t)
+{
+	struct uniphier_spi_priv *priv = spi_master_get_devdata(master);
+	int status;
+
+	uniphier_spi_setup_transfer(spi, t);
+
+	reinit_completion(&priv->xfer_done);
+
+	uniphier_spi_fill_tx_fifo(priv);
+
+	uniphier_spi_irq_enable(spi, SSI_IE_RCIE | SSI_IE_RORIE);
+
+	status = wait_for_completion_timeout(&priv->xfer_done,
+					     msecs_to_jiffies(SSI_TIMEOUT_MS));
+
+	uniphier_spi_irq_disable(spi, SSI_IE_RCIE | SSI_IE_RORIE);
+
+	if (status < 0)
+		return status;
+
+	return priv->error;
+}
+
+static int uniphier_spi_prepare_transfer_hardware(struct spi_master *master)
+{
+	struct uniphier_spi_priv *priv = spi_master_get_devdata(master);
+
+	writel(SSI_CTL_EN, priv->base + SSI_CTL);
+
+	return 0;
+}
+
+static int uniphier_spi_unprepare_transfer_hardware(struct spi_master *master)
+{
+	struct uniphier_spi_priv *priv = spi_master_get_devdata(master);
+
+	writel(0, priv->base + SSI_CTL);
+
+	return 0;
+}
+
+static irqreturn_t uniphier_spi_handler(int irq, void *dev_id)
+{
+	struct uniphier_spi_priv *priv = dev_id;
+	u32 val, stat;
+
+	stat = readl(priv->base + SSI_IS);
+	val = SSI_IC_TCIC | SSI_IC_RCIC | SSI_IC_RORIC;
+	writel(val, priv->base + SSI_IC);
+
+	/* rx fifo overrun */
+	if (stat & SSI_IS_RORID) {
+		priv->error = -EIO;
+		goto done;
+	}
+
+	/* rx complete */
+	if ((stat & SSI_IS_RCID) && (stat & SSI_IS_RXRS)) {
+		while ((readl(priv->base + SSI_SR) & SSI_SR_RNE) &&
+				(priv->rx_bytes - priv->tx_bytes) > 0)
+			uniphier_spi_recv(priv);
+
+		if ((readl(priv->base + SSI_SR) & SSI_SR_RNE) ||
+				(priv->rx_bytes != priv->tx_bytes)) {
+			priv->error = -EIO;
+			goto done;
+		} else if (priv->rx_bytes == 0)
+			goto done;
+
+		/* next tx transfer */
+		uniphier_spi_fill_tx_fifo(priv);
+
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_NONE;
+
+done:
+	complete(&priv->xfer_done);
+	return IRQ_HANDLED;
+}
+
+static int uniphier_spi_probe(struct platform_device *pdev)
+{
+	struct uniphier_spi_priv *priv;
+	struct spi_master *master;
+	struct resource *res;
+	unsigned long clk_rate;
+	int irq;
+	int ret;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*priv));
+	if (!master)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, master);
+
+	priv = spi_master_get_devdata(master);
+	priv->master = master;
+	priv->is_save_param = false;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	priv->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(priv->base)) {
+		ret = PTR_ERR(priv->base);
+		goto out_master_put;
+	}
+
+	priv->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(priv->clk)) {
+		dev_err(&pdev->dev, "failed to get clock\n");
+		ret = PTR_ERR(priv->clk);
+		goto out_master_put;
+	}
+
+	ret = clk_prepare_enable(priv->clk);
+	if (ret)
+		goto out_master_put;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "failed to get IRQ\n");
+		ret = irq;
+		goto out_disable_clk;
+	}
+
+	ret = devm_request_irq(&pdev->dev, irq, uniphier_spi_handler,
+			       0, "uniphier-spi", priv);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to request IRQ\n");
+		goto out_disable_clk;
+	}
+
+	init_completion(&priv->xfer_done);
+
+	clk_rate = clk_get_rate(priv->clk);
+
+	master->max_speed_hz = DIV_ROUND_UP(clk_rate, SSI_MIN_CLK_DIVIDER);
+	master->min_speed_hz = DIV_ROUND_UP(clk_rate, SSI_MAX_CLK_DIVIDER);
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LSB_FIRST;
+	master->dev.of_node = pdev->dev.of_node;
+	master->bus_num = pdev->id;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(1, 32);
+
+	master->set_cs = uniphier_spi_set_cs;
+	master->transfer_one = uniphier_spi_transfer_one;
+	master->prepare_transfer_hardware
+				= uniphier_spi_prepare_transfer_hardware;
+	master->unprepare_transfer_hardware
+				= uniphier_spi_unprepare_transfer_hardware;
+	master->num_chipselect = 1;
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret)
+		goto out_disable_clk;
+
+	return 0;
+
+out_disable_clk:
+	clk_disable_unprepare(priv->clk);
+
+out_master_put:
+	spi_master_put(master);
+	return ret;
+}
+
+static int uniphier_spi_remove(struct platform_device *pdev)
+{
+	struct uniphier_spi_priv *priv = platform_get_drvdata(pdev);
+
+	clk_disable_unprepare(priv->clk);
+
+	return 0;
+}
+
+static const struct of_device_id uniphier_spi_match[] = {
+	{ .compatible = "socionext,uniphier-scssi" },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, uniphier_spi_match);
+
+static struct platform_driver uniphier_spi_driver = {
+	.probe = uniphier_spi_probe,
+	.remove = uniphier_spi_remove,
+	.driver = {
+		.name = "uniphier-spi",
+		.of_match_table = uniphier_spi_match,
+	},
+};
+module_platform_driver(uniphier_spi_driver);
+
+MODULE_AUTHOR("Kunihiko Hayashi <hayashi.kunihiko@socionext.com>");
+MODULE_AUTHOR("Keiji Hayashibara <hayashibara.keiji@socionext.com>");
+MODULE_DESCRIPTION("Socionext UniPhier SPI controller driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-xcomm.c b/drivers/spi/spi-xcomm.c
new file mode 100644
index 0000000..3c28e24
--- /dev/null
+++ b/drivers/spi/spi-xcomm.c
@@ -0,0 +1,253 @@
+/*
+ * Analog Devices AD-FMCOMMS1-EBZ board I2C-SPI bridge driver
+ *
+ * Copyright 2012 Analog Devices Inc.
+ * Author: Lars-Peter Clausen <lars@metafoo.de>
+ *
+ * Licensed under the GPL-2 or later.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/delay.h>
+#include <linux/i2c.h>
+#include <linux/spi/spi.h>
+#include <asm/unaligned.h>
+
+#define SPI_XCOMM_SETTINGS_LEN_OFFSET		10
+#define SPI_XCOMM_SETTINGS_3WIRE		BIT(6)
+#define SPI_XCOMM_SETTINGS_CS_HIGH		BIT(5)
+#define SPI_XCOMM_SETTINGS_SAMPLE_END		BIT(4)
+#define SPI_XCOMM_SETTINGS_CPHA			BIT(3)
+#define SPI_XCOMM_SETTINGS_CPOL			BIT(2)
+#define SPI_XCOMM_SETTINGS_CLOCK_DIV_MASK	0x3
+#define SPI_XCOMM_SETTINGS_CLOCK_DIV_64		0x2
+#define SPI_XCOMM_SETTINGS_CLOCK_DIV_16		0x1
+#define SPI_XCOMM_SETTINGS_CLOCK_DIV_4		0x0
+
+#define SPI_XCOMM_CMD_UPDATE_CONFIG	0x03
+#define SPI_XCOMM_CMD_WRITE		0x04
+
+#define SPI_XCOMM_CLOCK 48000000
+
+struct spi_xcomm {
+	struct i2c_client *i2c;
+
+	uint16_t settings;
+	uint16_t chipselect;
+
+	unsigned int current_speed;
+
+	uint8_t buf[63];
+};
+
+static int spi_xcomm_sync_config(struct spi_xcomm *spi_xcomm, unsigned int len)
+{
+	uint16_t settings;
+	uint8_t *buf = spi_xcomm->buf;
+
+	settings = spi_xcomm->settings;
+	settings |= len << SPI_XCOMM_SETTINGS_LEN_OFFSET;
+
+	buf[0] = SPI_XCOMM_CMD_UPDATE_CONFIG;
+	put_unaligned_be16(settings, &buf[1]);
+	put_unaligned_be16(spi_xcomm->chipselect, &buf[3]);
+
+	return i2c_master_send(spi_xcomm->i2c, buf, 5);
+}
+
+static void spi_xcomm_chipselect(struct spi_xcomm *spi_xcomm,
+	struct spi_device *spi, int is_active)
+{
+	unsigned long cs = spi->chip_select;
+	uint16_t chipselect = spi_xcomm->chipselect;
+
+	if (is_active)
+		chipselect |= BIT(cs);
+	else
+		chipselect &= ~BIT(cs);
+
+	spi_xcomm->chipselect = chipselect;
+}
+
+static int spi_xcomm_setup_transfer(struct spi_xcomm *spi_xcomm,
+	struct spi_device *spi, struct spi_transfer *t, unsigned int *settings)
+{
+	if (t->len > 62)
+		return -EINVAL;
+
+	if (t->speed_hz != spi_xcomm->current_speed) {
+		unsigned int divider;
+
+		divider = DIV_ROUND_UP(SPI_XCOMM_CLOCK, t->speed_hz);
+		if (divider >= 64)
+			*settings |= SPI_XCOMM_SETTINGS_CLOCK_DIV_64;
+		else if (divider >= 16)
+			*settings |= SPI_XCOMM_SETTINGS_CLOCK_DIV_16;
+		else
+			*settings |= SPI_XCOMM_SETTINGS_CLOCK_DIV_4;
+
+		spi_xcomm->current_speed = t->speed_hz;
+	}
+
+	if (spi->mode & SPI_CPOL)
+		*settings |= SPI_XCOMM_SETTINGS_CPOL;
+	else
+		*settings &= ~SPI_XCOMM_SETTINGS_CPOL;
+
+	if (spi->mode & SPI_CPHA)
+		*settings &= ~SPI_XCOMM_SETTINGS_CPHA;
+	else
+		*settings |= SPI_XCOMM_SETTINGS_CPHA;
+
+	if (spi->mode & SPI_3WIRE)
+		*settings |= SPI_XCOMM_SETTINGS_3WIRE;
+	else
+		*settings &= ~SPI_XCOMM_SETTINGS_3WIRE;
+
+	return 0;
+}
+
+static int spi_xcomm_txrx_bufs(struct spi_xcomm *spi_xcomm,
+	struct spi_device *spi, struct spi_transfer *t)
+{
+	int ret;
+
+	if (t->tx_buf) {
+		spi_xcomm->buf[0] = SPI_XCOMM_CMD_WRITE;
+		memcpy(spi_xcomm->buf + 1, t->tx_buf, t->len);
+
+		ret = i2c_master_send(spi_xcomm->i2c, spi_xcomm->buf, t->len + 1);
+		if (ret < 0)
+			return ret;
+		else if (ret != t->len + 1)
+			return -EIO;
+	} else if (t->rx_buf) {
+		ret = i2c_master_recv(spi_xcomm->i2c, t->rx_buf, t->len);
+		if (ret < 0)
+			return ret;
+		else if (ret != t->len)
+			return -EIO;
+	}
+
+	return t->len;
+}
+
+static int spi_xcomm_transfer_one(struct spi_master *master,
+	struct spi_message *msg)
+{
+	struct spi_xcomm *spi_xcomm = spi_master_get_devdata(master);
+	unsigned int settings = spi_xcomm->settings;
+	struct spi_device *spi = msg->spi;
+	unsigned cs_change = 0;
+	struct spi_transfer *t;
+	bool is_first = true;
+	int status = 0;
+	bool is_last;
+
+	spi_xcomm_chipselect(spi_xcomm, spi, true);
+
+	list_for_each_entry(t, &msg->transfers, transfer_list) {
+
+		if (!t->tx_buf && !t->rx_buf && t->len) {
+			status = -EINVAL;
+			break;
+		}
+
+		status = spi_xcomm_setup_transfer(spi_xcomm, spi, t, &settings);
+		if (status < 0)
+			break;
+
+		is_last = list_is_last(&t->transfer_list, &msg->transfers);
+		cs_change = t->cs_change;
+
+		if (cs_change ^ is_last)
+			settings |= BIT(5);
+		else
+			settings &= ~BIT(5);
+
+		if (t->rx_buf) {
+			spi_xcomm->settings = settings;
+			status = spi_xcomm_sync_config(spi_xcomm, t->len);
+			if (status < 0)
+				break;
+		} else if (settings != spi_xcomm->settings || is_first) {
+			spi_xcomm->settings = settings;
+			status = spi_xcomm_sync_config(spi_xcomm, 0);
+			if (status < 0)
+				break;
+		}
+
+		if (t->len) {
+			status = spi_xcomm_txrx_bufs(spi_xcomm, spi, t);
+
+			if (status < 0)
+				break;
+
+			if (status > 0)
+				msg->actual_length += status;
+		}
+		status = 0;
+
+		if (t->delay_usecs)
+			udelay(t->delay_usecs);
+
+		is_first = false;
+	}
+
+	if (status != 0 || !cs_change)
+		spi_xcomm_chipselect(spi_xcomm, spi, false);
+
+	msg->status = status;
+	spi_finalize_current_message(master);
+
+	return status;
+}
+
+static int spi_xcomm_probe(struct i2c_client *i2c,
+	const struct i2c_device_id *id)
+{
+	struct spi_xcomm *spi_xcomm;
+	struct spi_master *master;
+	int ret;
+
+	master = spi_alloc_master(&i2c->dev, sizeof(*spi_xcomm));
+	if (!master)
+		return -ENOMEM;
+
+	spi_xcomm = spi_master_get_devdata(master);
+	spi_xcomm->i2c = i2c;
+
+	master->num_chipselect = 16;
+	master->mode_bits = SPI_CPHA | SPI_CPOL | SPI_3WIRE;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->flags = SPI_MASTER_HALF_DUPLEX;
+	master->transfer_one_message = spi_xcomm_transfer_one;
+	master->dev.of_node = i2c->dev.of_node;
+	i2c_set_clientdata(i2c, master);
+
+	ret = devm_spi_register_master(&i2c->dev, master);
+	if (ret < 0)
+		spi_master_put(master);
+
+	return ret;
+}
+
+static const struct i2c_device_id spi_xcomm_ids[] = {
+	{ "spi-xcomm" },
+	{ },
+};
+MODULE_DEVICE_TABLE(i2c, spi_xcomm_ids);
+
+static struct i2c_driver spi_xcomm_driver = {
+	.driver = {
+		.name	= "spi-xcomm",
+	},
+	.id_table	= spi_xcomm_ids,
+	.probe		= spi_xcomm_probe,
+};
+module_i2c_driver(spi_xcomm_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Lars-Peter Clausen <lars@metafoo.de>");
+MODULE_DESCRIPTION("Analog Devices AD-FMCOMMS1-EBZ board I2C-SPI bridge driver");
diff --git a/drivers/spi/spi-xilinx.c b/drivers/spi/spi-xilinx.c
new file mode 100644
index 0000000..63fedc4
--- /dev/null
+++ b/drivers/spi/spi-xilinx.c
@@ -0,0 +1,541 @@
+/*
+ * Xilinx SPI controller driver (master mode only)
+ *
+ * Author: MontaVista Software, Inc.
+ *	source@mvista.com
+ *
+ * Copyright (c) 2010 Secret Lab Technologies, Ltd.
+ * Copyright (c) 2009 Intel Corporation
+ * 2002-2007 (c) MontaVista Software, Inc.
+
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/spi/xilinx_spi.h>
+#include <linux/io.h>
+
+#define XILINX_SPI_MAX_CS	32
+
+#define XILINX_SPI_NAME "xilinx_spi"
+
+/* Register definitions as per "OPB Serial Peripheral Interface (SPI) (v1.00e)
+ * Product Specification", DS464
+ */
+#define XSPI_CR_OFFSET		0x60	/* Control Register */
+
+#define XSPI_CR_LOOP		0x01
+#define XSPI_CR_ENABLE		0x02
+#define XSPI_CR_MASTER_MODE	0x04
+#define XSPI_CR_CPOL		0x08
+#define XSPI_CR_CPHA		0x10
+#define XSPI_CR_MODE_MASK	(XSPI_CR_CPHA | XSPI_CR_CPOL | \
+				 XSPI_CR_LSB_FIRST | XSPI_CR_LOOP)
+#define XSPI_CR_TXFIFO_RESET	0x20
+#define XSPI_CR_RXFIFO_RESET	0x40
+#define XSPI_CR_MANUAL_SSELECT	0x80
+#define XSPI_CR_TRANS_INHIBIT	0x100
+#define XSPI_CR_LSB_FIRST	0x200
+
+#define XSPI_SR_OFFSET		0x64	/* Status Register */
+
+#define XSPI_SR_RX_EMPTY_MASK	0x01	/* Receive FIFO is empty */
+#define XSPI_SR_RX_FULL_MASK	0x02	/* Receive FIFO is full */
+#define XSPI_SR_TX_EMPTY_MASK	0x04	/* Transmit FIFO is empty */
+#define XSPI_SR_TX_FULL_MASK	0x08	/* Transmit FIFO is full */
+#define XSPI_SR_MODE_FAULT_MASK	0x10	/* Mode fault error */
+
+#define XSPI_TXD_OFFSET		0x68	/* Data Transmit Register */
+#define XSPI_RXD_OFFSET		0x6c	/* Data Receive Register */
+
+#define XSPI_SSR_OFFSET		0x70	/* 32-bit Slave Select Register */
+
+/* Register definitions as per "OPB IPIF (v3.01c) Product Specification", DS414
+ * IPIF registers are 32 bit
+ */
+#define XIPIF_V123B_DGIER_OFFSET	0x1c	/* IPIF global int enable reg */
+#define XIPIF_V123B_GINTR_ENABLE	0x80000000
+
+#define XIPIF_V123B_IISR_OFFSET		0x20	/* IPIF interrupt status reg */
+#define XIPIF_V123B_IIER_OFFSET		0x28	/* IPIF interrupt enable reg */
+
+#define XSPI_INTR_MODE_FAULT		0x01	/* Mode fault error */
+#define XSPI_INTR_SLAVE_MODE_FAULT	0x02	/* Selected as slave while
+						 * disabled */
+#define XSPI_INTR_TX_EMPTY		0x04	/* TxFIFO is empty */
+#define XSPI_INTR_TX_UNDERRUN		0x08	/* TxFIFO was underrun */
+#define XSPI_INTR_RX_FULL		0x10	/* RxFIFO is full */
+#define XSPI_INTR_RX_OVERRUN		0x20	/* RxFIFO was overrun */
+#define XSPI_INTR_TX_HALF_EMPTY		0x40	/* TxFIFO is half empty */
+
+#define XIPIF_V123B_RESETR_OFFSET	0x40	/* IPIF reset register */
+#define XIPIF_V123B_RESET_MASK		0x0a	/* the value to write */
+
+struct xilinx_spi {
+	/* bitbang has to be first */
+	struct spi_bitbang bitbang;
+	struct completion done;
+	void __iomem	*regs;	/* virt. address of the control registers */
+
+	int		irq;
+
+	u8 *rx_ptr;		/* pointer in the Tx buffer */
+	const u8 *tx_ptr;	/* pointer in the Rx buffer */
+	u8 bytes_per_word;
+	int buffer_size;	/* buffer size in words */
+	u32 cs_inactive;	/* Level of the CS pins when inactive*/
+	unsigned int (*read_fn)(void __iomem *);
+	void (*write_fn)(u32, void __iomem *);
+};
+
+static void xspi_write32(u32 val, void __iomem *addr)
+{
+	iowrite32(val, addr);
+}
+
+static unsigned int xspi_read32(void __iomem *addr)
+{
+	return ioread32(addr);
+}
+
+static void xspi_write32_be(u32 val, void __iomem *addr)
+{
+	iowrite32be(val, addr);
+}
+
+static unsigned int xspi_read32_be(void __iomem *addr)
+{
+	return ioread32be(addr);
+}
+
+static void xilinx_spi_tx(struct xilinx_spi *xspi)
+{
+	u32 data = 0;
+
+	if (!xspi->tx_ptr) {
+		xspi->write_fn(0, xspi->regs + XSPI_TXD_OFFSET);
+		return;
+	}
+
+	switch (xspi->bytes_per_word) {
+	case 1:
+		data = *(u8 *)(xspi->tx_ptr);
+		break;
+	case 2:
+		data = *(u16 *)(xspi->tx_ptr);
+		break;
+	case 4:
+		data = *(u32 *)(xspi->tx_ptr);
+		break;
+	}
+
+	xspi->write_fn(data, xspi->regs + XSPI_TXD_OFFSET);
+	xspi->tx_ptr += xspi->bytes_per_word;
+}
+
+static void xilinx_spi_rx(struct xilinx_spi *xspi)
+{
+	u32 data = xspi->read_fn(xspi->regs + XSPI_RXD_OFFSET);
+
+	if (!xspi->rx_ptr)
+		return;
+
+	switch (xspi->bytes_per_word) {
+	case 1:
+		*(u8 *)(xspi->rx_ptr) = data;
+		break;
+	case 2:
+		*(u16 *)(xspi->rx_ptr) = data;
+		break;
+	case 4:
+		*(u32 *)(xspi->rx_ptr) = data;
+		break;
+	}
+
+	xspi->rx_ptr += xspi->bytes_per_word;
+}
+
+static void xspi_init_hw(struct xilinx_spi *xspi)
+{
+	void __iomem *regs_base = xspi->regs;
+
+	/* Reset the SPI device */
+	xspi->write_fn(XIPIF_V123B_RESET_MASK,
+		regs_base + XIPIF_V123B_RESETR_OFFSET);
+	/* Enable the transmit empty interrupt, which we use to determine
+	 * progress on the transmission.
+	 */
+	xspi->write_fn(XSPI_INTR_TX_EMPTY,
+			regs_base + XIPIF_V123B_IIER_OFFSET);
+	/* Disable the global IPIF interrupt */
+	xspi->write_fn(0, regs_base + XIPIF_V123B_DGIER_OFFSET);
+	/* Deselect the slave on the SPI bus */
+	xspi->write_fn(0xffff, regs_base + XSPI_SSR_OFFSET);
+	/* Disable the transmitter, enable Manual Slave Select Assertion,
+	 * put SPI controller into master mode, and enable it */
+	xspi->write_fn(XSPI_CR_MANUAL_SSELECT |	XSPI_CR_MASTER_MODE |
+		XSPI_CR_ENABLE | XSPI_CR_TXFIFO_RESET |	XSPI_CR_RXFIFO_RESET,
+		regs_base + XSPI_CR_OFFSET);
+}
+
+static void xilinx_spi_chipselect(struct spi_device *spi, int is_on)
+{
+	struct xilinx_spi *xspi = spi_master_get_devdata(spi->master);
+	u16 cr;
+	u32 cs;
+
+	if (is_on == BITBANG_CS_INACTIVE) {
+		/* Deselect the slave on the SPI bus */
+		xspi->write_fn(xspi->cs_inactive, xspi->regs + XSPI_SSR_OFFSET);
+		return;
+	}
+
+	/* Set the SPI clock phase and polarity */
+	cr = xspi->read_fn(xspi->regs + XSPI_CR_OFFSET)	& ~XSPI_CR_MODE_MASK;
+	if (spi->mode & SPI_CPHA)
+		cr |= XSPI_CR_CPHA;
+	if (spi->mode & SPI_CPOL)
+		cr |= XSPI_CR_CPOL;
+	if (spi->mode & SPI_LSB_FIRST)
+		cr |= XSPI_CR_LSB_FIRST;
+	if (spi->mode & SPI_LOOP)
+		cr |= XSPI_CR_LOOP;
+	xspi->write_fn(cr, xspi->regs + XSPI_CR_OFFSET);
+
+	/* We do not check spi->max_speed_hz here as the SPI clock
+	 * frequency is not software programmable (the IP block design
+	 * parameter)
+	 */
+
+	cs = xspi->cs_inactive;
+	cs ^= BIT(spi->chip_select);
+
+	/* Activate the chip select */
+	xspi->write_fn(cs, xspi->regs + XSPI_SSR_OFFSET);
+}
+
+/* spi_bitbang requires custom setup_transfer() to be defined if there is a
+ * custom txrx_bufs().
+ */
+static int xilinx_spi_setup_transfer(struct spi_device *spi,
+		struct spi_transfer *t)
+{
+	struct xilinx_spi *xspi = spi_master_get_devdata(spi->master);
+
+	if (spi->mode & SPI_CS_HIGH)
+		xspi->cs_inactive &= ~BIT(spi->chip_select);
+	else
+		xspi->cs_inactive |= BIT(spi->chip_select);
+
+	return 0;
+}
+
+static int xilinx_spi_txrx_bufs(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct xilinx_spi *xspi = spi_master_get_devdata(spi->master);
+	int remaining_words;	/* the number of words left to transfer */
+	bool use_irq = false;
+	u16 cr = 0;
+
+	/* We get here with transmitter inhibited */
+
+	xspi->tx_ptr = t->tx_buf;
+	xspi->rx_ptr = t->rx_buf;
+	remaining_words = t->len / xspi->bytes_per_word;
+
+	if (xspi->irq >= 0 &&  remaining_words > xspi->buffer_size) {
+		u32 isr;
+		use_irq = true;
+		/* Inhibit irq to avoid spurious irqs on tx_empty*/
+		cr = xspi->read_fn(xspi->regs + XSPI_CR_OFFSET);
+		xspi->write_fn(cr | XSPI_CR_TRANS_INHIBIT,
+			       xspi->regs + XSPI_CR_OFFSET);
+		/* ACK old irqs (if any) */
+		isr = xspi->read_fn(xspi->regs + XIPIF_V123B_IISR_OFFSET);
+		if (isr)
+			xspi->write_fn(isr,
+				       xspi->regs + XIPIF_V123B_IISR_OFFSET);
+		/* Enable the global IPIF interrupt */
+		xspi->write_fn(XIPIF_V123B_GINTR_ENABLE,
+				xspi->regs + XIPIF_V123B_DGIER_OFFSET);
+		reinit_completion(&xspi->done);
+	}
+
+	while (remaining_words) {
+		int n_words, tx_words, rx_words;
+		u32 sr;
+		int stalled;
+
+		n_words = min(remaining_words, xspi->buffer_size);
+
+		tx_words = n_words;
+		while (tx_words--)
+			xilinx_spi_tx(xspi);
+
+		/* Start the transfer by not inhibiting the transmitter any
+		 * longer
+		 */
+
+		if (use_irq) {
+			xspi->write_fn(cr, xspi->regs + XSPI_CR_OFFSET);
+			wait_for_completion(&xspi->done);
+			/* A transmit has just completed. Process received data
+			 * and check for more data to transmit. Always inhibit
+			 * the transmitter while the Isr refills the transmit
+			 * register/FIFO, or make sure it is stopped if we're
+			 * done.
+			 */
+			xspi->write_fn(cr | XSPI_CR_TRANS_INHIBIT,
+				       xspi->regs + XSPI_CR_OFFSET);
+			sr = XSPI_SR_TX_EMPTY_MASK;
+		} else
+			sr = xspi->read_fn(xspi->regs + XSPI_SR_OFFSET);
+
+		/* Read out all the data from the Rx FIFO */
+		rx_words = n_words;
+		stalled = 10;
+		while (rx_words) {
+			if (rx_words == n_words && !(stalled--) &&
+			    !(sr & XSPI_SR_TX_EMPTY_MASK) &&
+			    (sr & XSPI_SR_RX_EMPTY_MASK)) {
+				dev_err(&spi->dev,
+					"Detected stall. Check C_SPI_MODE and C_SPI_MEMORY\n");
+				xspi_init_hw(xspi);
+				return -EIO;
+			}
+
+			if ((sr & XSPI_SR_TX_EMPTY_MASK) && (rx_words > 1)) {
+				xilinx_spi_rx(xspi);
+				rx_words--;
+				continue;
+			}
+
+			sr = xspi->read_fn(xspi->regs + XSPI_SR_OFFSET);
+			if (!(sr & XSPI_SR_RX_EMPTY_MASK)) {
+				xilinx_spi_rx(xspi);
+				rx_words--;
+			}
+		}
+
+		remaining_words -= n_words;
+	}
+
+	if (use_irq) {
+		xspi->write_fn(0, xspi->regs + XIPIF_V123B_DGIER_OFFSET);
+		xspi->write_fn(cr, xspi->regs + XSPI_CR_OFFSET);
+	}
+
+	return t->len;
+}
+
+
+/* This driver supports single master mode only. Hence Tx FIFO Empty
+ * is the only interrupt we care about.
+ * Receive FIFO Overrun, Transmit FIFO Underrun, Mode Fault, and Slave Mode
+ * Fault are not to happen.
+ */
+static irqreturn_t xilinx_spi_irq(int irq, void *dev_id)
+{
+	struct xilinx_spi *xspi = dev_id;
+	u32 ipif_isr;
+
+	/* Get the IPIF interrupts, and clear them immediately */
+	ipif_isr = xspi->read_fn(xspi->regs + XIPIF_V123B_IISR_OFFSET);
+	xspi->write_fn(ipif_isr, xspi->regs + XIPIF_V123B_IISR_OFFSET);
+
+	if (ipif_isr & XSPI_INTR_TX_EMPTY) {	/* Transmission completed */
+		complete(&xspi->done);
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_NONE;
+}
+
+static int xilinx_spi_find_buffer_size(struct xilinx_spi *xspi)
+{
+	u8 sr;
+	int n_words = 0;
+
+	/*
+	 * Before the buffer_size detection we reset the core
+	 * to make sure we start with a clean state.
+	 */
+	xspi->write_fn(XIPIF_V123B_RESET_MASK,
+		xspi->regs + XIPIF_V123B_RESETR_OFFSET);
+
+	/* Fill the Tx FIFO with as many words as possible */
+	do {
+		xspi->write_fn(0, xspi->regs + XSPI_TXD_OFFSET);
+		sr = xspi->read_fn(xspi->regs + XSPI_SR_OFFSET);
+		n_words++;
+	} while (!(sr & XSPI_SR_TX_FULL_MASK));
+
+	return n_words;
+}
+
+static const struct of_device_id xilinx_spi_of_match[] = {
+	{ .compatible = "xlnx,axi-quad-spi-1.00.a", },
+	{ .compatible = "xlnx,xps-spi-2.00.a", },
+	{ .compatible = "xlnx,xps-spi-2.00.b", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, xilinx_spi_of_match);
+
+static int xilinx_spi_probe(struct platform_device *pdev)
+{
+	struct xilinx_spi *xspi;
+	struct xspi_platform_data *pdata;
+	struct resource *res;
+	int ret, num_cs = 0, bits_per_word = 8;
+	struct spi_master *master;
+	u32 tmp;
+	u8 i;
+
+	pdata = dev_get_platdata(&pdev->dev);
+	if (pdata) {
+		num_cs = pdata->num_chipselect;
+		bits_per_word = pdata->bits_per_word;
+	} else {
+		of_property_read_u32(pdev->dev.of_node, "xlnx,num-ss-bits",
+					  &num_cs);
+	}
+
+	if (!num_cs) {
+		dev_err(&pdev->dev,
+			"Missing slave select configuration data\n");
+		return -EINVAL;
+	}
+
+	if (num_cs > XILINX_SPI_MAX_CS) {
+		dev_err(&pdev->dev, "Invalid number of spi slaves\n");
+		return -EINVAL;
+	}
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct xilinx_spi));
+	if (!master)
+		return -ENODEV;
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_LSB_FIRST | SPI_LOOP |
+			    SPI_CS_HIGH;
+
+	xspi = spi_master_get_devdata(master);
+	xspi->cs_inactive = 0xffffffff;
+	xspi->bitbang.master = master;
+	xspi->bitbang.chipselect = xilinx_spi_chipselect;
+	xspi->bitbang.setup_transfer = xilinx_spi_setup_transfer;
+	xspi->bitbang.txrx_bufs = xilinx_spi_txrx_bufs;
+	init_completion(&xspi->done);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	xspi->regs = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(xspi->regs)) {
+		ret = PTR_ERR(xspi->regs);
+		goto put_master;
+	}
+
+	master->bus_num = pdev->id;
+	master->num_chipselect = num_cs;
+	master->dev.of_node = pdev->dev.of_node;
+
+	/*
+	 * Detect endianess on the IP via loop bit in CR. Detection
+	 * must be done before reset is sent because incorrect reset
+	 * value generates error interrupt.
+	 * Setup little endian helper functions first and try to use them
+	 * and check if bit was correctly setup or not.
+	 */
+	xspi->read_fn = xspi_read32;
+	xspi->write_fn = xspi_write32;
+
+	xspi->write_fn(XSPI_CR_LOOP, xspi->regs + XSPI_CR_OFFSET);
+	tmp = xspi->read_fn(xspi->regs + XSPI_CR_OFFSET);
+	tmp &= XSPI_CR_LOOP;
+	if (tmp != XSPI_CR_LOOP) {
+		xspi->read_fn = xspi_read32_be;
+		xspi->write_fn = xspi_write32_be;
+	}
+
+	master->bits_per_word_mask = SPI_BPW_MASK(bits_per_word);
+	xspi->bytes_per_word = bits_per_word / 8;
+	xspi->buffer_size = xilinx_spi_find_buffer_size(xspi);
+
+	xspi->irq = platform_get_irq(pdev, 0);
+	if (xspi->irq < 0 && xspi->irq != -ENXIO) {
+		ret = xspi->irq;
+		goto put_master;
+	} else if (xspi->irq >= 0) {
+		/* Register for SPI Interrupt */
+		ret = devm_request_irq(&pdev->dev, xspi->irq, xilinx_spi_irq, 0,
+				dev_name(&pdev->dev), xspi);
+		if (ret)
+			goto put_master;
+	}
+
+	/* SPI controller initializations */
+	xspi_init_hw(xspi);
+
+	ret = spi_bitbang_start(&xspi->bitbang);
+	if (ret) {
+		dev_err(&pdev->dev, "spi_bitbang_start FAILED\n");
+		goto put_master;
+	}
+
+	dev_info(&pdev->dev, "at 0x%08llX mapped to 0x%p, irq=%d\n",
+		(unsigned long long)res->start, xspi->regs, xspi->irq);
+
+	if (pdata) {
+		for (i = 0; i < pdata->num_devices; i++)
+			spi_new_device(master, pdata->devices + i);
+	}
+
+	platform_set_drvdata(pdev, master);
+	return 0;
+
+put_master:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int xilinx_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct xilinx_spi *xspi = spi_master_get_devdata(master);
+	void __iomem *regs_base = xspi->regs;
+
+	spi_bitbang_stop(&xspi->bitbang);
+
+	/* Disable all the interrupts just in case */
+	xspi->write_fn(0, regs_base + XIPIF_V123B_IIER_OFFSET);
+	/* Disable the global IPIF interrupt */
+	xspi->write_fn(0, regs_base + XIPIF_V123B_DGIER_OFFSET);
+
+	spi_master_put(xspi->bitbang.master);
+
+	return 0;
+}
+
+/* work with hotplug and coldplug */
+MODULE_ALIAS("platform:" XILINX_SPI_NAME);
+
+static struct platform_driver xilinx_spi_driver = {
+	.probe = xilinx_spi_probe,
+	.remove = xilinx_spi_remove,
+	.driver = {
+		.name = XILINX_SPI_NAME,
+		.of_match_table = xilinx_spi_of_match,
+	},
+};
+module_platform_driver(xilinx_spi_driver);
+
+MODULE_AUTHOR("MontaVista Software, Inc. <source@mvista.com>");
+MODULE_DESCRIPTION("Xilinx SPI driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-xlp.c b/drivers/spi/spi-xlp.c
new file mode 100644
index 0000000..74a01b0
--- /dev/null
+++ b/drivers/spi/spi-xlp.c
@@ -0,0 +1,469 @@
+/*
+ * Copyright (C) 2003-2015 Broadcom Corporation
+ * All Rights Reserved
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 (GPL v2)
+ * as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+#include <linux/acpi.h>
+#include <linux/clk.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/of.h>
+#include <linux/interrupt.h>
+
+/* SPI Configuration Register */
+#define XLP_SPI_CONFIG			0x00
+#define XLP_SPI_CPHA			BIT(0)
+#define XLP_SPI_CPOL			BIT(1)
+#define XLP_SPI_CS_POL			BIT(2)
+#define XLP_SPI_TXMISO_EN		BIT(3)
+#define XLP_SPI_TXMOSI_EN		BIT(4)
+#define XLP_SPI_RXMISO_EN		BIT(5)
+#define XLP_SPI_CS_LSBFE		BIT(10)
+#define XLP_SPI_RXCAP_EN		BIT(11)
+
+/* SPI Frequency Divider Register */
+#define XLP_SPI_FDIV			0x04
+
+/* SPI Command Register */
+#define XLP_SPI_CMD			0x08
+#define XLP_SPI_CMD_IDLE_MASK		0x0
+#define XLP_SPI_CMD_TX_MASK		0x1
+#define XLP_SPI_CMD_RX_MASK		0x2
+#define XLP_SPI_CMD_TXRX_MASK		0x3
+#define XLP_SPI_CMD_CONT		BIT(4)
+#define XLP_SPI_XFR_BITCNT_SHIFT	16
+
+/* SPI Status Register */
+#define XLP_SPI_STATUS			0x0c
+#define XLP_SPI_XFR_PENDING		BIT(0)
+#define XLP_SPI_XFR_DONE		BIT(1)
+#define XLP_SPI_TX_INT			BIT(2)
+#define XLP_SPI_RX_INT			BIT(3)
+#define XLP_SPI_TX_UF			BIT(4)
+#define XLP_SPI_RX_OF			BIT(5)
+#define XLP_SPI_STAT_MASK		0x3f
+
+/* SPI Interrupt Enable Register */
+#define XLP_SPI_INTR_EN			0x10
+#define XLP_SPI_INTR_DONE		BIT(0)
+#define XLP_SPI_INTR_TXTH		BIT(1)
+#define XLP_SPI_INTR_RXTH		BIT(2)
+#define XLP_SPI_INTR_TXUF		BIT(3)
+#define XLP_SPI_INTR_RXOF		BIT(4)
+
+/* SPI FIFO Threshold Register */
+#define XLP_SPI_FIFO_THRESH		0x14
+
+/* SPI FIFO Word Count Register */
+#define XLP_SPI_FIFO_WCNT		0x18
+#define XLP_SPI_RXFIFO_WCNT_MASK	0xf
+#define XLP_SPI_TXFIFO_WCNT_MASK	0xf0
+#define XLP_SPI_TXFIFO_WCNT_SHIFT	4
+
+/* SPI Transmit Data FIFO Register */
+#define XLP_SPI_TXDATA_FIFO		0x1c
+
+/* SPI Receive Data FIFO Register */
+#define XLP_SPI_RXDATA_FIFO		0x20
+
+/* SPI System Control Register */
+#define XLP_SPI_SYSCTRL			0x100
+#define XLP_SPI_SYS_RESET		BIT(0)
+#define XLP_SPI_SYS_CLKDIS		BIT(1)
+#define XLP_SPI_SYS_PMEN		BIT(8)
+
+#define SPI_CS_OFFSET			0x40
+#define XLP_SPI_TXRXTH			0x80
+#define XLP_SPI_FIFO_SIZE		8
+#define XLP_SPI_MAX_CS			4
+#define XLP_SPI_DEFAULT_FREQ		133333333
+#define XLP_SPI_FDIV_MIN		4
+#define XLP_SPI_FDIV_MAX		65535
+/*
+ * SPI can transfer only 28 bytes properly at a time. So split the
+ * transfer into 28 bytes size.
+ */
+#define XLP_SPI_XFER_SIZE		28
+
+struct xlp_spi_priv {
+	struct device		dev;		/* device structure */
+	void __iomem		*base;		/* spi registers base address */
+	const u8		*tx_buf;	/* tx data buffer */
+	u8			*rx_buf;	/* rx data buffer */
+	int			tx_len;		/* tx xfer length */
+	int			rx_len;		/* rx xfer length */
+	int			txerrors;	/* TXFIFO underflow count */
+	int			rxerrors;	/* RXFIFO overflow count */
+	int			cs;		/* slave device chip select */
+	u32			spi_clk;	/* spi clock frequency */
+	bool			cmd_cont;	/* cs active */
+	struct completion	done;		/* completion notification */
+};
+
+static inline u32 xlp_spi_reg_read(struct xlp_spi_priv *priv,
+				int cs, int regoff)
+{
+	return readl(priv->base + regoff + cs * SPI_CS_OFFSET);
+}
+
+static inline void xlp_spi_reg_write(struct xlp_spi_priv *priv, int cs,
+				int regoff, u32 val)
+{
+	writel(val, priv->base + regoff + cs * SPI_CS_OFFSET);
+}
+
+static inline void xlp_spi_sysctl_write(struct xlp_spi_priv *priv,
+				int regoff, u32 val)
+{
+	writel(val, priv->base + regoff);
+}
+
+/*
+ * Setup global SPI_SYSCTRL register for all SPI channels.
+ */
+static void xlp_spi_sysctl_setup(struct xlp_spi_priv *xspi)
+{
+	int cs;
+
+	for (cs = 0; cs < XLP_SPI_MAX_CS; cs++)
+		xlp_spi_sysctl_write(xspi, XLP_SPI_SYSCTRL,
+				XLP_SPI_SYS_RESET << cs);
+	xlp_spi_sysctl_write(xspi, XLP_SPI_SYSCTRL, XLP_SPI_SYS_PMEN);
+}
+
+static int xlp_spi_setup(struct spi_device *spi)
+{
+	struct xlp_spi_priv *xspi;
+	u32 fdiv, cfg;
+	int cs;
+
+	xspi = spi_master_get_devdata(spi->master);
+	cs = spi->chip_select;
+	/*
+	 * The value of fdiv must be between 4 and 65535.
+	 */
+	fdiv = DIV_ROUND_UP(xspi->spi_clk, spi->max_speed_hz);
+	if (fdiv > XLP_SPI_FDIV_MAX)
+		fdiv = XLP_SPI_FDIV_MAX;
+	else if (fdiv < XLP_SPI_FDIV_MIN)
+		fdiv = XLP_SPI_FDIV_MIN;
+
+	xlp_spi_reg_write(xspi, cs, XLP_SPI_FDIV, fdiv);
+	xlp_spi_reg_write(xspi, cs, XLP_SPI_FIFO_THRESH, XLP_SPI_TXRXTH);
+	cfg = xlp_spi_reg_read(xspi, cs, XLP_SPI_CONFIG);
+	if (spi->mode & SPI_CPHA)
+		cfg |= XLP_SPI_CPHA;
+	else
+		cfg &= ~XLP_SPI_CPHA;
+	if (spi->mode & SPI_CPOL)
+		cfg |= XLP_SPI_CPOL;
+	else
+		cfg &= ~XLP_SPI_CPOL;
+	if (!(spi->mode & SPI_CS_HIGH))
+		cfg |= XLP_SPI_CS_POL;
+	else
+		cfg &= ~XLP_SPI_CS_POL;
+	if (spi->mode & SPI_LSB_FIRST)
+		cfg |= XLP_SPI_CS_LSBFE;
+	else
+		cfg &= ~XLP_SPI_CS_LSBFE;
+
+	cfg |= XLP_SPI_TXMOSI_EN | XLP_SPI_RXMISO_EN;
+	if (fdiv == 4)
+		cfg |= XLP_SPI_RXCAP_EN;
+	xlp_spi_reg_write(xspi, cs, XLP_SPI_CONFIG, cfg);
+
+	return 0;
+}
+
+static void xlp_spi_read_rxfifo(struct xlp_spi_priv *xspi)
+{
+	u32 rx_data, rxfifo_cnt;
+	int i, j, nbytes;
+
+	rxfifo_cnt = xlp_spi_reg_read(xspi, xspi->cs, XLP_SPI_FIFO_WCNT);
+	rxfifo_cnt &= XLP_SPI_RXFIFO_WCNT_MASK;
+	while (rxfifo_cnt) {
+		rx_data = xlp_spi_reg_read(xspi, xspi->cs, XLP_SPI_RXDATA_FIFO);
+		j = 0;
+		nbytes = min(xspi->rx_len, 4);
+		for (i = nbytes - 1; i >= 0; i--, j++)
+			xspi->rx_buf[i] = (rx_data >> (j * 8)) & 0xff;
+
+		xspi->rx_len -= nbytes;
+		xspi->rx_buf += nbytes;
+		rxfifo_cnt--;
+	}
+}
+
+static void xlp_spi_fill_txfifo(struct xlp_spi_priv *xspi)
+{
+	u32 tx_data, txfifo_cnt;
+	int i, j, nbytes;
+
+	txfifo_cnt = xlp_spi_reg_read(xspi, xspi->cs, XLP_SPI_FIFO_WCNT);
+	txfifo_cnt &= XLP_SPI_TXFIFO_WCNT_MASK;
+	txfifo_cnt >>= XLP_SPI_TXFIFO_WCNT_SHIFT;
+	while (xspi->tx_len && (txfifo_cnt < XLP_SPI_FIFO_SIZE)) {
+		j = 0;
+		tx_data = 0;
+		nbytes = min(xspi->tx_len, 4);
+		for (i = nbytes - 1; i >= 0; i--, j++)
+			tx_data |= xspi->tx_buf[i] << (j * 8);
+
+		xlp_spi_reg_write(xspi, xspi->cs, XLP_SPI_TXDATA_FIFO, tx_data);
+		xspi->tx_len -= nbytes;
+		xspi->tx_buf += nbytes;
+		txfifo_cnt++;
+	}
+}
+
+static irqreturn_t xlp_spi_interrupt(int irq, void *dev_id)
+{
+	struct xlp_spi_priv *xspi = dev_id;
+	u32 stat;
+
+	stat = xlp_spi_reg_read(xspi, xspi->cs, XLP_SPI_STATUS) &
+		XLP_SPI_STAT_MASK;
+	if (!stat)
+		return IRQ_NONE;
+
+	if (stat & XLP_SPI_TX_INT) {
+		if (xspi->tx_len)
+			xlp_spi_fill_txfifo(xspi);
+		if (stat & XLP_SPI_TX_UF)
+			xspi->txerrors++;
+	}
+
+	if (stat & XLP_SPI_RX_INT) {
+		if (xspi->rx_len)
+			xlp_spi_read_rxfifo(xspi);
+		if (stat & XLP_SPI_RX_OF)
+			xspi->rxerrors++;
+	}
+
+	/* write status back to clear interrupts */
+	xlp_spi_reg_write(xspi, xspi->cs, XLP_SPI_STATUS, stat);
+	if (stat & XLP_SPI_XFR_DONE)
+		complete(&xspi->done);
+
+	return IRQ_HANDLED;
+}
+
+static void xlp_spi_send_cmd(struct xlp_spi_priv *xspi, int xfer_len,
+			int cmd_cont)
+{
+	u32 cmd = 0;
+
+	if (xspi->tx_buf)
+		cmd |= XLP_SPI_CMD_TX_MASK;
+	if (xspi->rx_buf)
+		cmd |= XLP_SPI_CMD_RX_MASK;
+	if (cmd_cont)
+		cmd |= XLP_SPI_CMD_CONT;
+	cmd |= ((xfer_len * 8 - 1) << XLP_SPI_XFR_BITCNT_SHIFT);
+	xlp_spi_reg_write(xspi, xspi->cs, XLP_SPI_CMD, cmd);
+}
+
+static int xlp_spi_xfer_block(struct  xlp_spi_priv *xs,
+		const unsigned char *tx_buf,
+		unsigned char *rx_buf, int xfer_len, int cmd_cont)
+{
+	int timeout;
+	u32 intr_mask = 0;
+
+	xs->tx_buf = tx_buf;
+	xs->rx_buf = rx_buf;
+	xs->tx_len = (xs->tx_buf == NULL) ? 0 : xfer_len;
+	xs->rx_len = (xs->rx_buf == NULL) ? 0 : xfer_len;
+	xs->txerrors = xs->rxerrors = 0;
+
+	/* fill TXDATA_FIFO, then send the CMD */
+	if (xs->tx_len)
+		xlp_spi_fill_txfifo(xs);
+
+	xlp_spi_send_cmd(xs, xfer_len, cmd_cont);
+
+	/*
+	 * We are getting some spurious tx interrupts, so avoid enabling
+	 * tx interrupts when only rx is in process.
+	 * Enable all the interrupts in tx case.
+	 */
+	if (xs->tx_len)
+		intr_mask |= XLP_SPI_INTR_TXTH | XLP_SPI_INTR_TXUF |
+				XLP_SPI_INTR_RXTH | XLP_SPI_INTR_RXOF;
+	else
+		intr_mask |= XLP_SPI_INTR_RXTH | XLP_SPI_INTR_RXOF;
+
+	intr_mask |= XLP_SPI_INTR_DONE;
+	xlp_spi_reg_write(xs, xs->cs, XLP_SPI_INTR_EN, intr_mask);
+
+	timeout = wait_for_completion_timeout(&xs->done,
+				msecs_to_jiffies(1000));
+	/* Disable interrupts */
+	xlp_spi_reg_write(xs, xs->cs, XLP_SPI_INTR_EN, 0x0);
+	if (!timeout) {
+		dev_err(&xs->dev, "xfer timedout!\n");
+		goto out;
+	}
+	if (xs->txerrors || xs->rxerrors)
+		dev_err(&xs->dev, "Over/Underflow rx %d tx %d xfer %d!\n",
+				xs->rxerrors, xs->txerrors, xfer_len);
+
+	return xfer_len;
+out:
+	return -ETIMEDOUT;
+}
+
+static int xlp_spi_txrx_bufs(struct xlp_spi_priv *xs, struct spi_transfer *t)
+{
+	int bytesleft, sz;
+	unsigned char *rx_buf;
+	const unsigned char *tx_buf;
+
+	tx_buf = t->tx_buf;
+	rx_buf = t->rx_buf;
+	bytesleft = t->len;
+	while (bytesleft) {
+		if (bytesleft > XLP_SPI_XFER_SIZE)
+			sz = xlp_spi_xfer_block(xs, tx_buf, rx_buf,
+					XLP_SPI_XFER_SIZE, 1);
+		else
+			sz = xlp_spi_xfer_block(xs, tx_buf, rx_buf,
+					bytesleft, xs->cmd_cont);
+		if (sz < 0)
+			return sz;
+		bytesleft -= sz;
+		if (tx_buf)
+			tx_buf += sz;
+		if (rx_buf)
+			rx_buf += sz;
+	}
+	return bytesleft;
+}
+
+static int xlp_spi_transfer_one(struct spi_master *master,
+					struct spi_device *spi,
+					struct spi_transfer *t)
+{
+	struct xlp_spi_priv *xspi = spi_master_get_devdata(master);
+	int ret = 0;
+
+	xspi->cs = spi->chip_select;
+	xspi->dev = spi->dev;
+
+	if (spi_transfer_is_last(master, t))
+		xspi->cmd_cont = 0;
+	else
+		xspi->cmd_cont = 1;
+
+	if (xlp_spi_txrx_bufs(xspi, t))
+		ret = -EIO;
+
+	spi_finalize_current_transfer(master);
+	return ret;
+}
+
+static int xlp_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct xlp_spi_priv *xspi;
+	struct resource *res;
+	struct clk *clk;
+	int irq, err;
+
+	xspi = devm_kzalloc(&pdev->dev, sizeof(*xspi), GFP_KERNEL);
+	if (!xspi)
+		return -ENOMEM;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	xspi->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(xspi->base))
+		return PTR_ERR(xspi->base);
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "no IRQ resource found: %d\n", irq);
+		return irq;
+	}
+	err = devm_request_irq(&pdev->dev, irq, xlp_spi_interrupt, 0,
+			pdev->name, xspi);
+	if (err) {
+		dev_err(&pdev->dev, "unable to request irq %d\n", irq);
+		return err;
+	}
+
+	clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(clk)) {
+		dev_err(&pdev->dev, "could not get spi clock\n");
+		return PTR_ERR(clk);
+	}
+
+	xspi->spi_clk = clk_get_rate(clk);
+
+	master = spi_alloc_master(&pdev->dev, 0);
+	if (!master) {
+		dev_err(&pdev->dev, "could not alloc master\n");
+		return -ENOMEM;
+	}
+
+	master->bus_num = 0;
+	master->num_chipselect = XLP_SPI_MAX_CS;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->setup = xlp_spi_setup;
+	master->transfer_one = xlp_spi_transfer_one;
+	master->dev.of_node = pdev->dev.of_node;
+
+	init_completion(&xspi->done);
+	spi_master_set_devdata(master, xspi);
+	xlp_spi_sysctl_setup(xspi);
+
+	/* register spi controller */
+	err = devm_spi_register_master(&pdev->dev, master);
+	if (err) {
+		dev_err(&pdev->dev, "spi register master failed!\n");
+		spi_master_put(master);
+		return err;
+	}
+
+	return 0;
+}
+
+#ifdef CONFIG_ACPI
+static const struct acpi_device_id xlp_spi_acpi_match[] = {
+	{ "BRCM900D", 0 },
+	{ "CAV900D",  0 },
+	{ },
+};
+MODULE_DEVICE_TABLE(acpi, xlp_spi_acpi_match);
+#endif
+
+static const struct of_device_id xlp_spi_dt_id[] = {
+	{ .compatible = "netlogic,xlp832-spi" },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, xlp_spi_dt_id);
+
+static struct platform_driver xlp_spi_driver = {
+	.probe	= xlp_spi_probe,
+	.driver = {
+		.name	= "xlp-spi",
+		.of_match_table = xlp_spi_dt_id,
+		.acpi_match_table = ACPI_PTR(xlp_spi_acpi_match),
+	},
+};
+module_platform_driver(xlp_spi_driver);
+
+MODULE_AUTHOR("Kamlakant Patel <kamlakant.patel@broadcom.com>");
+MODULE_DESCRIPTION("Netlogic XLP SPI controller driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-xtensa-xtfpga.c b/drivers/spi/spi-xtensa-xtfpga.c
new file mode 100644
index 0000000..8ce04f8
--- /dev/null
+++ b/drivers/spi/spi-xtensa-xtfpga.c
@@ -0,0 +1,170 @@
+/*
+ * Xtensa xtfpga SPI controller driver
+ *
+ * Copyright (c) 2014 Cadence Design Systems Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/delay.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+
+#define XTFPGA_SPI_NAME "xtfpga_spi"
+
+#define XTFPGA_SPI_START	0x0
+#define XTFPGA_SPI_BUSY		0x4
+#define XTFPGA_SPI_DATA		0x8
+
+#define BUSY_WAIT_US		100
+
+struct xtfpga_spi {
+	struct spi_bitbang bitbang;
+	void __iomem *regs;
+	u32 data;
+	unsigned data_sz;
+};
+
+static inline void xtfpga_spi_write32(const struct xtfpga_spi *spi,
+				      unsigned addr, u32 val)
+{
+	__raw_writel(val, spi->regs + addr);
+}
+
+static inline unsigned int xtfpga_spi_read32(const struct xtfpga_spi *spi,
+					     unsigned addr)
+{
+	return __raw_readl(spi->regs + addr);
+}
+
+static inline void xtfpga_spi_wait_busy(struct xtfpga_spi *xspi)
+{
+	unsigned i;
+
+	for (i = 0; xtfpga_spi_read32(xspi, XTFPGA_SPI_BUSY) &&
+	     i < BUSY_WAIT_US; ++i)
+		udelay(1);
+	WARN_ON_ONCE(i == BUSY_WAIT_US);
+}
+
+static u32 xtfpga_spi_txrx_word(struct spi_device *spi, unsigned nsecs,
+				u32 v, u8 bits, unsigned flags)
+{
+	struct xtfpga_spi *xspi = spi_master_get_devdata(spi->master);
+
+	xspi->data = (xspi->data << bits) | (v & GENMASK(bits - 1, 0));
+	xspi->data_sz += bits;
+	if (xspi->data_sz >= 16) {
+		xtfpga_spi_write32(xspi, XTFPGA_SPI_DATA,
+				   xspi->data >> (xspi->data_sz - 16));
+		xspi->data_sz -= 16;
+		xtfpga_spi_write32(xspi, XTFPGA_SPI_START, 1);
+		xtfpga_spi_wait_busy(xspi);
+		xtfpga_spi_write32(xspi, XTFPGA_SPI_START, 0);
+	}
+
+	return 0;
+}
+
+static void xtfpga_spi_chipselect(struct spi_device *spi, int is_on)
+{
+	struct xtfpga_spi *xspi = spi_master_get_devdata(spi->master);
+
+	WARN_ON(xspi->data_sz != 0);
+	xspi->data_sz = 0;
+}
+
+static int xtfpga_spi_probe(struct platform_device *pdev)
+{
+	struct xtfpga_spi *xspi;
+	struct resource *mem;
+	int ret;
+	struct spi_master *master;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(struct xtfpga_spi));
+	if (!master)
+		return -ENOMEM;
+
+	master->flags = SPI_MASTER_NO_RX;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(1, 16);
+	master->bus_num = pdev->dev.id;
+	master->dev.of_node = pdev->dev.of_node;
+
+	xspi = spi_master_get_devdata(master);
+	xspi->bitbang.master = master;
+	xspi->bitbang.chipselect = xtfpga_spi_chipselect;
+	xspi->bitbang.txrx_word[SPI_MODE_0] = xtfpga_spi_txrx_word;
+
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!mem) {
+		dev_err(&pdev->dev, "No memory resource\n");
+		ret = -ENODEV;
+		goto err;
+	}
+	xspi->regs = devm_ioremap_resource(&pdev->dev, mem);
+	if (IS_ERR(xspi->regs)) {
+		ret = PTR_ERR(xspi->regs);
+		goto err;
+	}
+
+	xtfpga_spi_write32(xspi, XTFPGA_SPI_START, 0);
+	usleep_range(1000, 2000);
+	if (xtfpga_spi_read32(xspi, XTFPGA_SPI_BUSY)) {
+		dev_err(&pdev->dev, "Device stuck in busy state\n");
+		ret = -EBUSY;
+		goto err;
+	}
+
+	ret = spi_bitbang_start(&xspi->bitbang);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "spi_bitbang_start failed\n");
+		goto err;
+	}
+
+	platform_set_drvdata(pdev, master);
+	return 0;
+err:
+	spi_master_put(master);
+	return ret;
+}
+
+static int xtfpga_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct xtfpga_spi *xspi = spi_master_get_devdata(master);
+
+	spi_bitbang_stop(&xspi->bitbang);
+	spi_master_put(master);
+
+	return 0;
+}
+
+MODULE_ALIAS("platform:" XTFPGA_SPI_NAME);
+
+#ifdef CONFIG_OF
+static const struct of_device_id xtfpga_spi_of_match[] = {
+	{ .compatible = "cdns,xtfpga-spi", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, xtfpga_spi_of_match);
+#endif
+
+static struct platform_driver xtfpga_spi_driver = {
+	.probe = xtfpga_spi_probe,
+	.remove = xtfpga_spi_remove,
+	.driver = {
+		.name = XTFPGA_SPI_NAME,
+		.of_match_table = of_match_ptr(xtfpga_spi_of_match),
+	},
+};
+module_platform_driver(xtfpga_spi_driver);
+
+MODULE_AUTHOR("Max Filippov <jcmvbkbc@gmail.com>");
+MODULE_DESCRIPTION("xtensa xtfpga SPI driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-zynqmp-gqspi.c b/drivers/spi/spi-zynqmp-gqspi.c
new file mode 100644
index 0000000..cc4d310
--- /dev/null
+++ b/drivers/spi/spi-zynqmp-gqspi.c
@@ -0,0 +1,1172 @@
+/*
+ * Xilinx Zynq UltraScale+ MPSoC Quad-SPI (QSPI) controller driver
+ * (master mode only)
+ *
+ * Copyright (C) 2009 - 2015 Xilinx, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published
+ * by the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of_irq.h>
+#include <linux/of_address.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/spi/spi.h>
+#include <linux/spinlock.h>
+#include <linux/workqueue.h>
+
+/* Generic QSPI register offsets */
+#define GQSPI_CONFIG_OFST		0x00000100
+#define GQSPI_ISR_OFST			0x00000104
+#define GQSPI_IDR_OFST			0x0000010C
+#define GQSPI_IER_OFST			0x00000108
+#define GQSPI_IMASK_OFST		0x00000110
+#define GQSPI_EN_OFST			0x00000114
+#define GQSPI_TXD_OFST			0x0000011C
+#define GQSPI_RXD_OFST			0x00000120
+#define GQSPI_TX_THRESHOLD_OFST		0x00000128
+#define GQSPI_RX_THRESHOLD_OFST		0x0000012C
+#define GQSPI_LPBK_DLY_ADJ_OFST		0x00000138
+#define GQSPI_GEN_FIFO_OFST		0x00000140
+#define GQSPI_SEL_OFST			0x00000144
+#define GQSPI_GF_THRESHOLD_OFST		0x00000150
+#define GQSPI_FIFO_CTRL_OFST		0x0000014C
+#define GQSPI_QSPIDMA_DST_CTRL_OFST	0x0000080C
+#define GQSPI_QSPIDMA_DST_SIZE_OFST	0x00000804
+#define GQSPI_QSPIDMA_DST_STS_OFST	0x00000808
+#define GQSPI_QSPIDMA_DST_I_STS_OFST	0x00000814
+#define GQSPI_QSPIDMA_DST_I_EN_OFST	0x00000818
+#define GQSPI_QSPIDMA_DST_I_DIS_OFST	0x0000081C
+#define GQSPI_QSPIDMA_DST_I_MASK_OFST	0x00000820
+#define GQSPI_QSPIDMA_DST_ADDR_OFST	0x00000800
+#define GQSPI_QSPIDMA_DST_ADDR_MSB_OFST 0x00000828
+
+/* GQSPI register bit masks */
+#define GQSPI_SEL_MASK				0x00000001
+#define GQSPI_EN_MASK				0x00000001
+#define GQSPI_LPBK_DLY_ADJ_USE_LPBK_MASK	0x00000020
+#define GQSPI_ISR_WR_TO_CLR_MASK		0x00000002
+#define GQSPI_IDR_ALL_MASK			0x00000FBE
+#define GQSPI_CFG_MODE_EN_MASK			0xC0000000
+#define GQSPI_CFG_GEN_FIFO_START_MODE_MASK	0x20000000
+#define GQSPI_CFG_ENDIAN_MASK			0x04000000
+#define GQSPI_CFG_EN_POLL_TO_MASK		0x00100000
+#define GQSPI_CFG_WP_HOLD_MASK			0x00080000
+#define GQSPI_CFG_BAUD_RATE_DIV_MASK		0x00000038
+#define GQSPI_CFG_CLK_PHA_MASK			0x00000004
+#define GQSPI_CFG_CLK_POL_MASK			0x00000002
+#define GQSPI_CFG_START_GEN_FIFO_MASK		0x10000000
+#define GQSPI_GENFIFO_IMM_DATA_MASK		0x000000FF
+#define GQSPI_GENFIFO_DATA_XFER			0x00000100
+#define GQSPI_GENFIFO_EXP			0x00000200
+#define GQSPI_GENFIFO_MODE_SPI			0x00000400
+#define GQSPI_GENFIFO_MODE_DUALSPI		0x00000800
+#define GQSPI_GENFIFO_MODE_QUADSPI		0x00000C00
+#define GQSPI_GENFIFO_MODE_MASK			0x00000C00
+#define GQSPI_GENFIFO_CS_LOWER			0x00001000
+#define GQSPI_GENFIFO_CS_UPPER			0x00002000
+#define GQSPI_GENFIFO_BUS_LOWER			0x00004000
+#define GQSPI_GENFIFO_BUS_UPPER			0x00008000
+#define GQSPI_GENFIFO_BUS_BOTH			0x0000C000
+#define GQSPI_GENFIFO_BUS_MASK			0x0000C000
+#define GQSPI_GENFIFO_TX			0x00010000
+#define GQSPI_GENFIFO_RX			0x00020000
+#define GQSPI_GENFIFO_STRIPE			0x00040000
+#define GQSPI_GENFIFO_POLL			0x00080000
+#define GQSPI_GENFIFO_EXP_START			0x00000100
+#define GQSPI_FIFO_CTRL_RST_RX_FIFO_MASK	0x00000004
+#define GQSPI_FIFO_CTRL_RST_TX_FIFO_MASK	0x00000002
+#define GQSPI_FIFO_CTRL_RST_GEN_FIFO_MASK	0x00000001
+#define GQSPI_ISR_RXEMPTY_MASK			0x00000800
+#define GQSPI_ISR_GENFIFOFULL_MASK		0x00000400
+#define GQSPI_ISR_GENFIFONOT_FULL_MASK		0x00000200
+#define GQSPI_ISR_TXEMPTY_MASK			0x00000100
+#define GQSPI_ISR_GENFIFOEMPTY_MASK		0x00000080
+#define GQSPI_ISR_RXFULL_MASK			0x00000020
+#define GQSPI_ISR_RXNEMPTY_MASK			0x00000010
+#define GQSPI_ISR_TXFULL_MASK			0x00000008
+#define GQSPI_ISR_TXNOT_FULL_MASK		0x00000004
+#define GQSPI_ISR_POLL_TIME_EXPIRE_MASK		0x00000002
+#define GQSPI_IER_TXNOT_FULL_MASK		0x00000004
+#define GQSPI_IER_RXEMPTY_MASK			0x00000800
+#define GQSPI_IER_POLL_TIME_EXPIRE_MASK		0x00000002
+#define GQSPI_IER_RXNEMPTY_MASK			0x00000010
+#define GQSPI_IER_GENFIFOEMPTY_MASK		0x00000080
+#define GQSPI_IER_TXEMPTY_MASK			0x00000100
+#define GQSPI_QSPIDMA_DST_INTR_ALL_MASK		0x000000FE
+#define GQSPI_QSPIDMA_DST_STS_WTC		0x0000E000
+#define GQSPI_CFG_MODE_EN_DMA_MASK		0x80000000
+#define GQSPI_ISR_IDR_MASK			0x00000994
+#define GQSPI_QSPIDMA_DST_I_EN_DONE_MASK	0x00000002
+#define GQSPI_QSPIDMA_DST_I_STS_DONE_MASK	0x00000002
+#define GQSPI_IRQ_MASK				0x00000980
+
+#define GQSPI_CFG_BAUD_RATE_DIV_SHIFT		3
+#define GQSPI_GENFIFO_CS_SETUP			0x4
+#define GQSPI_GENFIFO_CS_HOLD			0x3
+#define GQSPI_TXD_DEPTH				64
+#define GQSPI_RX_FIFO_THRESHOLD			32
+#define GQSPI_RX_FIFO_FILL	(GQSPI_RX_FIFO_THRESHOLD * 4)
+#define GQSPI_TX_FIFO_THRESHOLD_RESET_VAL	32
+#define GQSPI_TX_FIFO_FILL	(GQSPI_TXD_DEPTH -\
+				GQSPI_TX_FIFO_THRESHOLD_RESET_VAL)
+#define GQSPI_GEN_FIFO_THRESHOLD_RESET_VAL	0X10
+#define GQSPI_QSPIDMA_DST_CTRL_RESET_VAL	0x803FFA00
+#define GQSPI_SELECT_FLASH_CS_LOWER		0x1
+#define GQSPI_SELECT_FLASH_CS_UPPER		0x2
+#define GQSPI_SELECT_FLASH_CS_BOTH		0x3
+#define GQSPI_SELECT_FLASH_BUS_LOWER		0x1
+#define GQSPI_SELECT_FLASH_BUS_UPPER		0x2
+#define GQSPI_SELECT_FLASH_BUS_BOTH		0x3
+#define GQSPI_BAUD_DIV_MAX	7	/* Baud rate divisor maximum */
+#define GQSPI_BAUD_DIV_SHIFT	2	/* Baud rate divisor shift */
+#define GQSPI_SELECT_MODE_SPI		0x1
+#define GQSPI_SELECT_MODE_DUALSPI	0x2
+#define GQSPI_SELECT_MODE_QUADSPI	0x4
+#define GQSPI_DMA_UNALIGN		0x3
+#define GQSPI_DEFAULT_NUM_CS	1	/* Default number of chip selects */
+
+#define SPI_AUTOSUSPEND_TIMEOUT		3000
+enum mode_type {GQSPI_MODE_IO, GQSPI_MODE_DMA};
+
+/**
+ * struct zynqmp_qspi - Defines qspi driver instance
+ * @regs:		Virtual address of the QSPI controller registers
+ * @refclk:		Pointer to the peripheral clock
+ * @pclk:		Pointer to the APB clock
+ * @irq:		IRQ number
+ * @dev:		Pointer to struct device
+ * @txbuf:		Pointer to the TX buffer
+ * @rxbuf:		Pointer to the RX buffer
+ * @bytes_to_transfer:	Number of bytes left to transfer
+ * @bytes_to_receive:	Number of bytes left to receive
+ * @genfifocs:		Used for chip select
+ * @genfifobus:		Used to select the upper or lower bus
+ * @dma_rx_bytes:	Remaining bytes to receive by DMA mode
+ * @dma_addr:		DMA address after mapping the kernel buffer
+ * @genfifoentry:	Used for storing the genfifoentry instruction.
+ * @mode:		Defines the mode in which QSPI is operating
+ */
+struct zynqmp_qspi {
+	void __iomem *regs;
+	struct clk *refclk;
+	struct clk *pclk;
+	int irq;
+	struct device *dev;
+	const void *txbuf;
+	void *rxbuf;
+	int bytes_to_transfer;
+	int bytes_to_receive;
+	u32 genfifocs;
+	u32 genfifobus;
+	u32 dma_rx_bytes;
+	dma_addr_t dma_addr;
+	u32 genfifoentry;
+	enum mode_type mode;
+};
+
+/**
+ * zynqmp_gqspi_read:	For GQSPI controller read operation
+ * @xqspi:	Pointer to the zynqmp_qspi structure
+ * @offset:	Offset from where to read
+ */
+static u32 zynqmp_gqspi_read(struct zynqmp_qspi *xqspi, u32 offset)
+{
+	return readl_relaxed(xqspi->regs + offset);
+}
+
+/**
+ * zynqmp_gqspi_write:	For GQSPI controller write operation
+ * @xqspi:	Pointer to the zynqmp_qspi structure
+ * @offset:	Offset where to write
+ * @val:	Value to be written
+ */
+static inline void zynqmp_gqspi_write(struct zynqmp_qspi *xqspi, u32 offset,
+				      u32 val)
+{
+	writel_relaxed(val, (xqspi->regs + offset));
+}
+
+/**
+ * zynqmp_gqspi_selectslave:	For selection of slave device
+ * @instanceptr:	Pointer to the zynqmp_qspi structure
+ * @flashcs:	For chip select
+ * @flashbus:	To check which bus is selected- upper or lower
+ */
+static void zynqmp_gqspi_selectslave(struct zynqmp_qspi *instanceptr,
+				     u8 slavecs, u8 slavebus)
+{
+	/*
+	 * Bus and CS lines selected here will be updated in the instance and
+	 * used for subsequent GENFIFO entries during transfer.
+	 */
+
+	/* Choose slave select line */
+	switch (slavecs) {
+	case GQSPI_SELECT_FLASH_CS_BOTH:
+		instanceptr->genfifocs = GQSPI_GENFIFO_CS_LOWER |
+			GQSPI_GENFIFO_CS_UPPER;
+		break;
+	case GQSPI_SELECT_FLASH_CS_UPPER:
+		instanceptr->genfifocs = GQSPI_GENFIFO_CS_UPPER;
+		break;
+	case GQSPI_SELECT_FLASH_CS_LOWER:
+		instanceptr->genfifocs = GQSPI_GENFIFO_CS_LOWER;
+		break;
+	default:
+		dev_warn(instanceptr->dev, "Invalid slave select\n");
+	}
+
+	/* Choose the bus */
+	switch (slavebus) {
+	case GQSPI_SELECT_FLASH_BUS_BOTH:
+		instanceptr->genfifobus = GQSPI_GENFIFO_BUS_LOWER |
+			GQSPI_GENFIFO_BUS_UPPER;
+		break;
+	case GQSPI_SELECT_FLASH_BUS_UPPER:
+		instanceptr->genfifobus = GQSPI_GENFIFO_BUS_UPPER;
+		break;
+	case GQSPI_SELECT_FLASH_BUS_LOWER:
+		instanceptr->genfifobus = GQSPI_GENFIFO_BUS_LOWER;
+		break;
+	default:
+		dev_warn(instanceptr->dev, "Invalid slave bus\n");
+	}
+}
+
+/**
+ * zynqmp_qspi_init_hw:	Initialize the hardware
+ * @xqspi:	Pointer to the zynqmp_qspi structure
+ *
+ * The default settings of the QSPI controller's configurable parameters on
+ * reset are
+ *	- Master mode
+ *	- TX threshold set to 1
+ *	- RX threshold set to 1
+ *	- Flash memory interface mode enabled
+ * This function performs the following actions
+ *	- Disable and clear all the interrupts
+ *	- Enable manual slave select
+ *	- Enable manual start
+ *	- Deselect all the chip select lines
+ *	- Set the little endian mode of TX FIFO and
+ *	- Enable the QSPI controller
+ */
+static void zynqmp_qspi_init_hw(struct zynqmp_qspi *xqspi)
+{
+	u32 config_reg;
+
+	/* Select the GQSPI mode */
+	zynqmp_gqspi_write(xqspi, GQSPI_SEL_OFST, GQSPI_SEL_MASK);
+	/* Clear and disable interrupts */
+	zynqmp_gqspi_write(xqspi, GQSPI_ISR_OFST,
+			   zynqmp_gqspi_read(xqspi, GQSPI_ISR_OFST) |
+			   GQSPI_ISR_WR_TO_CLR_MASK);
+	/* Clear the DMA STS */
+	zynqmp_gqspi_write(xqspi, GQSPI_QSPIDMA_DST_I_STS_OFST,
+			   zynqmp_gqspi_read(xqspi,
+					     GQSPI_QSPIDMA_DST_I_STS_OFST));
+	zynqmp_gqspi_write(xqspi, GQSPI_QSPIDMA_DST_STS_OFST,
+			   zynqmp_gqspi_read(xqspi,
+					     GQSPI_QSPIDMA_DST_STS_OFST) |
+					     GQSPI_QSPIDMA_DST_STS_WTC);
+	zynqmp_gqspi_write(xqspi, GQSPI_IDR_OFST, GQSPI_IDR_ALL_MASK);
+	zynqmp_gqspi_write(xqspi,
+			   GQSPI_QSPIDMA_DST_I_DIS_OFST,
+			   GQSPI_QSPIDMA_DST_INTR_ALL_MASK);
+	/* Disable the GQSPI */
+	zynqmp_gqspi_write(xqspi, GQSPI_EN_OFST, 0x0);
+	config_reg = zynqmp_gqspi_read(xqspi, GQSPI_CONFIG_OFST);
+	config_reg &= ~GQSPI_CFG_MODE_EN_MASK;
+	/* Manual start */
+	config_reg |= GQSPI_CFG_GEN_FIFO_START_MODE_MASK;
+	/* Little endian by default */
+	config_reg &= ~GQSPI_CFG_ENDIAN_MASK;
+	/* Disable poll time out */
+	config_reg &= ~GQSPI_CFG_EN_POLL_TO_MASK;
+	/* Set hold bit */
+	config_reg |= GQSPI_CFG_WP_HOLD_MASK;
+	/* Clear pre-scalar by default */
+	config_reg &= ~GQSPI_CFG_BAUD_RATE_DIV_MASK;
+	/* CPHA 0 */
+	config_reg &= ~GQSPI_CFG_CLK_PHA_MASK;
+	/* CPOL 0 */
+	config_reg &= ~GQSPI_CFG_CLK_POL_MASK;
+	zynqmp_gqspi_write(xqspi, GQSPI_CONFIG_OFST, config_reg);
+
+	/* Clear the TX and RX FIFO */
+	zynqmp_gqspi_write(xqspi, GQSPI_FIFO_CTRL_OFST,
+			   GQSPI_FIFO_CTRL_RST_RX_FIFO_MASK |
+			   GQSPI_FIFO_CTRL_RST_TX_FIFO_MASK |
+			   GQSPI_FIFO_CTRL_RST_GEN_FIFO_MASK);
+	/* Set by default to allow for high frequencies */
+	zynqmp_gqspi_write(xqspi, GQSPI_LPBK_DLY_ADJ_OFST,
+			   zynqmp_gqspi_read(xqspi, GQSPI_LPBK_DLY_ADJ_OFST) |
+			   GQSPI_LPBK_DLY_ADJ_USE_LPBK_MASK);
+	/* Reset thresholds */
+	zynqmp_gqspi_write(xqspi, GQSPI_TX_THRESHOLD_OFST,
+			   GQSPI_TX_FIFO_THRESHOLD_RESET_VAL);
+	zynqmp_gqspi_write(xqspi, GQSPI_RX_THRESHOLD_OFST,
+			   GQSPI_RX_FIFO_THRESHOLD);
+	zynqmp_gqspi_write(xqspi, GQSPI_GF_THRESHOLD_OFST,
+			   GQSPI_GEN_FIFO_THRESHOLD_RESET_VAL);
+	zynqmp_gqspi_selectslave(xqspi,
+				 GQSPI_SELECT_FLASH_CS_LOWER,
+				 GQSPI_SELECT_FLASH_BUS_LOWER);
+	/* Initialize DMA */
+	zynqmp_gqspi_write(xqspi,
+			GQSPI_QSPIDMA_DST_CTRL_OFST,
+			GQSPI_QSPIDMA_DST_CTRL_RESET_VAL);
+
+	/* Enable the GQSPI */
+	zynqmp_gqspi_write(xqspi, GQSPI_EN_OFST, GQSPI_EN_MASK);
+}
+
+/**
+ * zynqmp_qspi_copy_read_data:	Copy data to RX buffer
+ * @xqspi:	Pointer to the zynqmp_qspi structure
+ * @data:	The variable where data is stored
+ * @size:	Number of bytes to be copied from data to RX buffer
+ */
+static void zynqmp_qspi_copy_read_data(struct zynqmp_qspi *xqspi,
+				       ulong data, u8 size)
+{
+	memcpy(xqspi->rxbuf, &data, size);
+	xqspi->rxbuf += size;
+	xqspi->bytes_to_receive -= size;
+}
+
+/**
+ * zynqmp_prepare_transfer_hardware:	Prepares hardware for transfer.
+ * @master:	Pointer to the spi_master structure which provides
+ *		information about the controller.
+ *
+ * This function enables SPI master controller.
+ *
+ * Return:	0 on success; error value otherwise
+ */
+static int zynqmp_prepare_transfer_hardware(struct spi_master *master)
+{
+	struct zynqmp_qspi *xqspi = spi_master_get_devdata(master);
+
+	zynqmp_gqspi_write(xqspi, GQSPI_EN_OFST, GQSPI_EN_MASK);
+	return 0;
+}
+
+/**
+ * zynqmp_unprepare_transfer_hardware:	Relaxes hardware after transfer
+ * @master:	Pointer to the spi_master structure which provides
+ *		information about the controller.
+ *
+ * This function disables the SPI master controller.
+ *
+ * Return:	Always 0
+ */
+static int zynqmp_unprepare_transfer_hardware(struct spi_master *master)
+{
+	struct zynqmp_qspi *xqspi = spi_master_get_devdata(master);
+
+	zynqmp_gqspi_write(xqspi, GQSPI_EN_OFST, 0x0);
+	return 0;
+}
+
+/**
+ * zynqmp_qspi_chipselect:	Select or deselect the chip select line
+ * @qspi:	Pointer to the spi_device structure
+ * @is_high:	Select(0) or deselect (1) the chip select line
+ */
+static void zynqmp_qspi_chipselect(struct spi_device *qspi, bool is_high)
+{
+	struct zynqmp_qspi *xqspi = spi_master_get_devdata(qspi->master);
+	ulong timeout;
+	u32 genfifoentry = 0x0, statusreg;
+
+	genfifoentry |= GQSPI_GENFIFO_MODE_SPI;
+	genfifoentry |= xqspi->genfifobus;
+
+	if (!is_high) {
+		genfifoentry |= xqspi->genfifocs;
+		genfifoentry |= GQSPI_GENFIFO_CS_SETUP;
+	} else {
+		genfifoentry |= GQSPI_GENFIFO_CS_HOLD;
+	}
+
+	zynqmp_gqspi_write(xqspi, GQSPI_GEN_FIFO_OFST, genfifoentry);
+
+	/* Dummy generic FIFO entry */
+	zynqmp_gqspi_write(xqspi, GQSPI_GEN_FIFO_OFST, 0x0);
+
+	/* Manually start the generic FIFO command */
+	zynqmp_gqspi_write(xqspi, GQSPI_CONFIG_OFST,
+			zynqmp_gqspi_read(xqspi, GQSPI_CONFIG_OFST) |
+			GQSPI_CFG_START_GEN_FIFO_MASK);
+
+	timeout = jiffies + msecs_to_jiffies(1000);
+
+	/* Wait until the generic FIFO command is empty */
+	do {
+		statusreg = zynqmp_gqspi_read(xqspi, GQSPI_ISR_OFST);
+
+		if ((statusreg & GQSPI_ISR_GENFIFOEMPTY_MASK) &&
+			(statusreg & GQSPI_ISR_TXEMPTY_MASK))
+			break;
+		else
+			cpu_relax();
+	} while (!time_after_eq(jiffies, timeout));
+
+	if (time_after_eq(jiffies, timeout))
+		dev_err(xqspi->dev, "Chip select timed out\n");
+}
+
+/**
+ * zynqmp_qspi_setup_transfer:	Configure QSPI controller for specified
+ *				transfer
+ * @qspi:	Pointer to the spi_device structure
+ * @transfer:	Pointer to the spi_transfer structure which provides
+ *		information about next transfer setup parameters
+ *
+ * Sets the operational mode of QSPI controller for the next QSPI transfer and
+ * sets the requested clock frequency.
+ *
+ * Return:	Always 0
+ *
+ * Note:
+ *	If the requested frequency is not an exact match with what can be
+ *	obtained using the pre-scalar value, the driver sets the clock
+ *	frequency which is lower than the requested frequency (maximum lower)
+ *	for the transfer.
+ *
+ *	If the requested frequency is higher or lower than that is supported
+ *	by the QSPI controller the driver will set the highest or lowest
+ *	frequency supported by controller.
+ */
+static int zynqmp_qspi_setup_transfer(struct spi_device *qspi,
+				      struct spi_transfer *transfer)
+{
+	struct zynqmp_qspi *xqspi = spi_master_get_devdata(qspi->master);
+	ulong clk_rate;
+	u32 config_reg, req_hz, baud_rate_val = 0;
+
+	if (transfer)
+		req_hz = transfer->speed_hz;
+	else
+		req_hz = qspi->max_speed_hz;
+
+	/* Set the clock frequency */
+	/* If req_hz == 0, default to lowest speed */
+	clk_rate = clk_get_rate(xqspi->refclk);
+
+	while ((baud_rate_val < GQSPI_BAUD_DIV_MAX) &&
+	       (clk_rate /
+		(GQSPI_BAUD_DIV_SHIFT << baud_rate_val)) > req_hz)
+		baud_rate_val++;
+
+	config_reg = zynqmp_gqspi_read(xqspi, GQSPI_CONFIG_OFST);
+
+	/* Set the QSPI clock phase and clock polarity */
+	config_reg &= (~GQSPI_CFG_CLK_PHA_MASK) & (~GQSPI_CFG_CLK_POL_MASK);
+
+	if (qspi->mode & SPI_CPHA)
+		config_reg |= GQSPI_CFG_CLK_PHA_MASK;
+	if (qspi->mode & SPI_CPOL)
+		config_reg |= GQSPI_CFG_CLK_POL_MASK;
+
+	config_reg &= ~GQSPI_CFG_BAUD_RATE_DIV_MASK;
+	config_reg |= (baud_rate_val << GQSPI_CFG_BAUD_RATE_DIV_SHIFT);
+	zynqmp_gqspi_write(xqspi, GQSPI_CONFIG_OFST, config_reg);
+	return 0;
+}
+
+/**
+ * zynqmp_qspi_setup:	Configure the QSPI controller
+ * @qspi:	Pointer to the spi_device structure
+ *
+ * Sets the operational mode of QSPI controller for the next QSPI transfer,
+ * baud rate and divisor value to setup the requested qspi clock.
+ *
+ * Return:	0 on success; error value otherwise.
+ */
+static int zynqmp_qspi_setup(struct spi_device *qspi)
+{
+	if (qspi->master->busy)
+		return -EBUSY;
+	return 0;
+}
+
+/**
+ * zynqmp_qspi_filltxfifo:	Fills the TX FIFO as long as there is room in
+ *				the FIFO or the bytes required to be
+ *				transmitted.
+ * @xqspi:	Pointer to the zynqmp_qspi structure
+ * @size:	Number of bytes to be copied from TX buffer to TX FIFO
+ */
+static void zynqmp_qspi_filltxfifo(struct zynqmp_qspi *xqspi, int size)
+{
+	u32 count = 0, intermediate;
+
+	while ((xqspi->bytes_to_transfer > 0) && (count < size)) {
+		memcpy(&intermediate, xqspi->txbuf, 4);
+		zynqmp_gqspi_write(xqspi, GQSPI_TXD_OFST, intermediate);
+
+		if (xqspi->bytes_to_transfer >= 4) {
+			xqspi->txbuf += 4;
+			xqspi->bytes_to_transfer -= 4;
+		} else {
+			xqspi->txbuf += xqspi->bytes_to_transfer;
+			xqspi->bytes_to_transfer = 0;
+		}
+		count++;
+	}
+}
+
+/**
+ * zynqmp_qspi_readrxfifo:	Fills the RX FIFO as long as there is room in
+ *				the FIFO.
+ * @xqspi:	Pointer to the zynqmp_qspi structure
+ * @size:	Number of bytes to be copied from RX buffer to RX FIFO
+ */
+static void zynqmp_qspi_readrxfifo(struct zynqmp_qspi *xqspi, u32 size)
+{
+	ulong data;
+	int count = 0;
+
+	while ((count < size) && (xqspi->bytes_to_receive > 0)) {
+		if (xqspi->bytes_to_receive >= 4) {
+			(*(u32 *) xqspi->rxbuf) =
+			zynqmp_gqspi_read(xqspi, GQSPI_RXD_OFST);
+			xqspi->rxbuf += 4;
+			xqspi->bytes_to_receive -= 4;
+			count += 4;
+		} else {
+			data = zynqmp_gqspi_read(xqspi, GQSPI_RXD_OFST);
+			count += xqspi->bytes_to_receive;
+			zynqmp_qspi_copy_read_data(xqspi, data,
+						   xqspi->bytes_to_receive);
+			xqspi->bytes_to_receive = 0;
+		}
+	}
+}
+
+/**
+ * zynqmp_process_dma_irq:	Handler for DMA done interrupt of QSPI
+ *				controller
+ * @xqspi:	zynqmp_qspi instance pointer
+ *
+ * This function handles DMA interrupt only.
+ */
+static void zynqmp_process_dma_irq(struct zynqmp_qspi *xqspi)
+{
+	u32 config_reg, genfifoentry;
+
+	dma_unmap_single(xqspi->dev, xqspi->dma_addr,
+				xqspi->dma_rx_bytes, DMA_FROM_DEVICE);
+	xqspi->rxbuf += xqspi->dma_rx_bytes;
+	xqspi->bytes_to_receive -= xqspi->dma_rx_bytes;
+	xqspi->dma_rx_bytes = 0;
+
+	/* Disabling the DMA interrupts */
+	zynqmp_gqspi_write(xqspi, GQSPI_QSPIDMA_DST_I_DIS_OFST,
+					GQSPI_QSPIDMA_DST_I_EN_DONE_MASK);
+
+	if (xqspi->bytes_to_receive > 0) {
+		/* Switch to IO mode,for remaining bytes to receive */
+		config_reg = zynqmp_gqspi_read(xqspi, GQSPI_CONFIG_OFST);
+		config_reg &= ~GQSPI_CFG_MODE_EN_MASK;
+		zynqmp_gqspi_write(xqspi, GQSPI_CONFIG_OFST, config_reg);
+
+		/* Initiate the transfer of remaining bytes */
+		genfifoentry = xqspi->genfifoentry;
+		genfifoentry |= xqspi->bytes_to_receive;
+		zynqmp_gqspi_write(xqspi, GQSPI_GEN_FIFO_OFST, genfifoentry);
+
+		/* Dummy generic FIFO entry */
+		zynqmp_gqspi_write(xqspi, GQSPI_GEN_FIFO_OFST, 0x0);
+
+		/* Manual start */
+		zynqmp_gqspi_write(xqspi, GQSPI_CONFIG_OFST,
+			(zynqmp_gqspi_read(xqspi, GQSPI_CONFIG_OFST) |
+			GQSPI_CFG_START_GEN_FIFO_MASK));
+
+		/* Enable the RX interrupts for IO mode */
+		zynqmp_gqspi_write(xqspi, GQSPI_IER_OFST,
+				GQSPI_IER_GENFIFOEMPTY_MASK |
+				GQSPI_IER_RXNEMPTY_MASK |
+				GQSPI_IER_RXEMPTY_MASK);
+	}
+}
+
+/**
+ * zynqmp_qspi_irq:	Interrupt service routine of the QSPI controller
+ * @irq:	IRQ number
+ * @dev_id:	Pointer to the xqspi structure
+ *
+ * This function handles TX empty only.
+ * On TX empty interrupt this function reads the received data from RX FIFO
+ * and fills the TX FIFO if there is any data remaining to be transferred.
+ *
+ * Return:	IRQ_HANDLED when interrupt is handled
+ *		IRQ_NONE otherwise.
+ */
+static irqreturn_t zynqmp_qspi_irq(int irq, void *dev_id)
+{
+	struct spi_master *master = dev_id;
+	struct zynqmp_qspi *xqspi = spi_master_get_devdata(master);
+	int ret = IRQ_NONE;
+	u32 status, mask, dma_status = 0;
+
+	status = zynqmp_gqspi_read(xqspi, GQSPI_ISR_OFST);
+	zynqmp_gqspi_write(xqspi, GQSPI_ISR_OFST, status);
+	mask = (status & ~(zynqmp_gqspi_read(xqspi, GQSPI_IMASK_OFST)));
+
+	/* Read and clear DMA status */
+	if (xqspi->mode == GQSPI_MODE_DMA) {
+		dma_status =
+			zynqmp_gqspi_read(xqspi, GQSPI_QSPIDMA_DST_I_STS_OFST);
+		zynqmp_gqspi_write(xqspi, GQSPI_QSPIDMA_DST_I_STS_OFST,
+								dma_status);
+	}
+
+	if (mask & GQSPI_ISR_TXNOT_FULL_MASK) {
+		zynqmp_qspi_filltxfifo(xqspi, GQSPI_TX_FIFO_FILL);
+		ret = IRQ_HANDLED;
+	}
+
+	if (dma_status & GQSPI_QSPIDMA_DST_I_STS_DONE_MASK) {
+		zynqmp_process_dma_irq(xqspi);
+		ret = IRQ_HANDLED;
+	} else if (!(mask & GQSPI_IER_RXEMPTY_MASK) &&
+			(mask & GQSPI_IER_GENFIFOEMPTY_MASK)) {
+		zynqmp_qspi_readrxfifo(xqspi, GQSPI_RX_FIFO_FILL);
+		ret = IRQ_HANDLED;
+	}
+
+	if ((xqspi->bytes_to_receive == 0) && (xqspi->bytes_to_transfer == 0)
+			&& ((status & GQSPI_IRQ_MASK) == GQSPI_IRQ_MASK)) {
+		zynqmp_gqspi_write(xqspi, GQSPI_IDR_OFST, GQSPI_ISR_IDR_MASK);
+		spi_finalize_current_transfer(master);
+		ret = IRQ_HANDLED;
+	}
+	return ret;
+}
+
+/**
+ * zynqmp_qspi_selectspimode:	Selects SPI mode - x1 or x2 or x4.
+ * @xqspi:	xqspi is a pointer to the GQSPI instance
+ * @spimode:	spimode - SPI or DUAL or QUAD.
+ * Return:	Mask to set desired SPI mode in GENFIFO entry.
+ */
+static inline u32 zynqmp_qspi_selectspimode(struct zynqmp_qspi *xqspi,
+						u8 spimode)
+{
+	u32 mask = 0;
+
+	switch (spimode) {
+	case GQSPI_SELECT_MODE_DUALSPI:
+		mask = GQSPI_GENFIFO_MODE_DUALSPI;
+		break;
+	case GQSPI_SELECT_MODE_QUADSPI:
+		mask = GQSPI_GENFIFO_MODE_QUADSPI;
+		break;
+	case GQSPI_SELECT_MODE_SPI:
+		mask = GQSPI_GENFIFO_MODE_SPI;
+		break;
+	default:
+		dev_warn(xqspi->dev, "Invalid SPI mode\n");
+	}
+
+	return mask;
+}
+
+/**
+ * zynq_qspi_setuprxdma:	This function sets up the RX DMA operation
+ * @xqspi:	xqspi is a pointer to the GQSPI instance.
+ */
+static void zynq_qspi_setuprxdma(struct zynqmp_qspi *xqspi)
+{
+	u32 rx_bytes, rx_rem, config_reg;
+	dma_addr_t addr;
+	u64 dma_align =  (u64)(uintptr_t)xqspi->rxbuf;
+
+	if ((xqspi->bytes_to_receive < 8) ||
+		((dma_align & GQSPI_DMA_UNALIGN) != 0x0)) {
+		/* Setting to IO mode */
+		config_reg = zynqmp_gqspi_read(xqspi, GQSPI_CONFIG_OFST);
+		config_reg &= ~GQSPI_CFG_MODE_EN_MASK;
+		zynqmp_gqspi_write(xqspi, GQSPI_CONFIG_OFST, config_reg);
+		xqspi->mode = GQSPI_MODE_IO;
+		xqspi->dma_rx_bytes = 0;
+		return;
+	}
+
+	rx_rem = xqspi->bytes_to_receive % 4;
+	rx_bytes = (xqspi->bytes_to_receive - rx_rem);
+
+	addr = dma_map_single(xqspi->dev, (void *)xqspi->rxbuf,
+						rx_bytes, DMA_FROM_DEVICE);
+	if (dma_mapping_error(xqspi->dev, addr))
+		dev_err(xqspi->dev, "ERR:rxdma:memory not mapped\n");
+
+	xqspi->dma_rx_bytes = rx_bytes;
+	xqspi->dma_addr = addr;
+	zynqmp_gqspi_write(xqspi, GQSPI_QSPIDMA_DST_ADDR_OFST,
+				(u32)(addr & 0xffffffff));
+	addr = ((addr >> 16) >> 16);
+	zynqmp_gqspi_write(xqspi, GQSPI_QSPIDMA_DST_ADDR_MSB_OFST,
+				((u32)addr) & 0xfff);
+
+	/* Enabling the DMA mode */
+	config_reg = zynqmp_gqspi_read(xqspi, GQSPI_CONFIG_OFST);
+	config_reg &= ~GQSPI_CFG_MODE_EN_MASK;
+	config_reg |= GQSPI_CFG_MODE_EN_DMA_MASK;
+	zynqmp_gqspi_write(xqspi, GQSPI_CONFIG_OFST, config_reg);
+
+	/* Switch to DMA mode */
+	xqspi->mode = GQSPI_MODE_DMA;
+
+	/* Write the number of bytes to transfer */
+	zynqmp_gqspi_write(xqspi, GQSPI_QSPIDMA_DST_SIZE_OFST, rx_bytes);
+}
+
+/**
+ * zynqmp_qspi_txrxsetup:	This function checks the TX/RX buffers in
+ *				the transfer and sets up the GENFIFO entries,
+ *				TX FIFO as required.
+ * @xqspi:	xqspi is a pointer to the GQSPI instance.
+ * @transfer:	It is a pointer to the structure containing transfer data.
+ * @genfifoentry:	genfifoentry is pointer to the variable in which
+ *			GENFIFO	mask is returned to calling function
+ */
+static void zynqmp_qspi_txrxsetup(struct zynqmp_qspi *xqspi,
+				  struct spi_transfer *transfer,
+				  u32 *genfifoentry)
+{
+	u32 config_reg;
+
+	/* Transmit */
+	if ((xqspi->txbuf != NULL) && (xqspi->rxbuf == NULL)) {
+		/* Setup data to be TXed */
+		*genfifoentry &= ~GQSPI_GENFIFO_RX;
+		*genfifoentry |= GQSPI_GENFIFO_DATA_XFER;
+		*genfifoentry |= GQSPI_GENFIFO_TX;
+		*genfifoentry |=
+			zynqmp_qspi_selectspimode(xqspi, transfer->tx_nbits);
+		xqspi->bytes_to_transfer = transfer->len;
+		if (xqspi->mode == GQSPI_MODE_DMA) {
+			config_reg = zynqmp_gqspi_read(xqspi,
+							GQSPI_CONFIG_OFST);
+			config_reg &= ~GQSPI_CFG_MODE_EN_MASK;
+			zynqmp_gqspi_write(xqspi, GQSPI_CONFIG_OFST,
+								config_reg);
+			xqspi->mode = GQSPI_MODE_IO;
+		}
+		zynqmp_qspi_filltxfifo(xqspi, GQSPI_TXD_DEPTH);
+		/* Discard RX data */
+		xqspi->bytes_to_receive = 0;
+	} else if ((xqspi->txbuf == NULL) && (xqspi->rxbuf != NULL)) {
+		/* Receive */
+
+		/* TX auto fill */
+		*genfifoentry &= ~GQSPI_GENFIFO_TX;
+		/* Setup RX */
+		*genfifoentry |= GQSPI_GENFIFO_DATA_XFER;
+		*genfifoentry |= GQSPI_GENFIFO_RX;
+		*genfifoentry |=
+			zynqmp_qspi_selectspimode(xqspi, transfer->rx_nbits);
+		xqspi->bytes_to_transfer = 0;
+		xqspi->bytes_to_receive = transfer->len;
+		zynq_qspi_setuprxdma(xqspi);
+	}
+}
+
+/**
+ * zynqmp_qspi_start_transfer:	Initiates the QSPI transfer
+ * @master:	Pointer to the spi_master structure which provides
+ *		information about the controller.
+ * @qspi:	Pointer to the spi_device structure
+ * @transfer:	Pointer to the spi_transfer structure which provide information
+ *		about next transfer parameters
+ *
+ * This function fills the TX FIFO, starts the QSPI transfer, and waits for the
+ * transfer to be completed.
+ *
+ * Return:	Number of bytes transferred in the last transfer
+ */
+static int zynqmp_qspi_start_transfer(struct spi_master *master,
+				      struct spi_device *qspi,
+				      struct spi_transfer *transfer)
+{
+	struct zynqmp_qspi *xqspi = spi_master_get_devdata(master);
+	u32 genfifoentry = 0x0, transfer_len;
+
+	xqspi->txbuf = transfer->tx_buf;
+	xqspi->rxbuf = transfer->rx_buf;
+
+	zynqmp_qspi_setup_transfer(qspi, transfer);
+
+	genfifoentry |= xqspi->genfifocs;
+	genfifoentry |= xqspi->genfifobus;
+
+	zynqmp_qspi_txrxsetup(xqspi, transfer, &genfifoentry);
+
+	if (xqspi->mode == GQSPI_MODE_DMA)
+		transfer_len = xqspi->dma_rx_bytes;
+	else
+		transfer_len = transfer->len;
+
+	xqspi->genfifoentry = genfifoentry;
+	if ((transfer_len) < GQSPI_GENFIFO_IMM_DATA_MASK) {
+		genfifoentry &= ~GQSPI_GENFIFO_IMM_DATA_MASK;
+		genfifoentry |= transfer_len;
+		zynqmp_gqspi_write(xqspi, GQSPI_GEN_FIFO_OFST, genfifoentry);
+	} else {
+		int tempcount = transfer_len;
+		u32 exponent = 8;	/* 2^8 = 256 */
+		u8 imm_data = tempcount & 0xFF;
+
+		tempcount &= ~(tempcount & 0xFF);
+		/* Immediate entry */
+		if (tempcount != 0) {
+			/* Exponent entries */
+			genfifoentry |= GQSPI_GENFIFO_EXP;
+			while (tempcount != 0) {
+				if (tempcount & GQSPI_GENFIFO_EXP_START) {
+					genfifoentry &=
+					    ~GQSPI_GENFIFO_IMM_DATA_MASK;
+					genfifoentry |= exponent;
+					zynqmp_gqspi_write(xqspi,
+							   GQSPI_GEN_FIFO_OFST,
+							   genfifoentry);
+				}
+				tempcount = tempcount >> 1;
+				exponent++;
+			}
+		}
+		if (imm_data != 0) {
+			genfifoentry &= ~GQSPI_GENFIFO_EXP;
+			genfifoentry &= ~GQSPI_GENFIFO_IMM_DATA_MASK;
+			genfifoentry |= (u8) (imm_data & 0xFF);
+			zynqmp_gqspi_write(xqspi,
+					   GQSPI_GEN_FIFO_OFST, genfifoentry);
+		}
+	}
+
+	if ((xqspi->mode == GQSPI_MODE_IO) &&
+			(xqspi->rxbuf != NULL)) {
+		/* Dummy generic FIFO entry */
+		zynqmp_gqspi_write(xqspi, GQSPI_GEN_FIFO_OFST, 0x0);
+	}
+
+	/* Since we are using manual mode */
+	zynqmp_gqspi_write(xqspi, GQSPI_CONFIG_OFST,
+			   zynqmp_gqspi_read(xqspi, GQSPI_CONFIG_OFST) |
+			   GQSPI_CFG_START_GEN_FIFO_MASK);
+
+	if (xqspi->txbuf != NULL)
+		/* Enable interrupts for TX */
+		zynqmp_gqspi_write(xqspi, GQSPI_IER_OFST,
+				   GQSPI_IER_TXEMPTY_MASK |
+					GQSPI_IER_GENFIFOEMPTY_MASK |
+					GQSPI_IER_TXNOT_FULL_MASK);
+
+	if (xqspi->rxbuf != NULL) {
+		/* Enable interrupts for RX */
+		if (xqspi->mode == GQSPI_MODE_DMA) {
+			/* Enable DMA interrupts */
+			zynqmp_gqspi_write(xqspi,
+					GQSPI_QSPIDMA_DST_I_EN_OFST,
+					GQSPI_QSPIDMA_DST_I_EN_DONE_MASK);
+		} else {
+			zynqmp_gqspi_write(xqspi, GQSPI_IER_OFST,
+					GQSPI_IER_GENFIFOEMPTY_MASK |
+					GQSPI_IER_RXNEMPTY_MASK |
+					GQSPI_IER_RXEMPTY_MASK);
+		}
+	}
+
+	return transfer->len;
+}
+
+/**
+ * zynqmp_qspi_suspend:	Suspend method for the QSPI driver
+ * @_dev:	Address of the platform_device structure
+ *
+ * This function stops the QSPI driver queue and disables the QSPI controller
+ *
+ * Return:	Always 0
+ */
+static int __maybe_unused zynqmp_qspi_suspend(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+
+	spi_master_suspend(master);
+
+	zynqmp_unprepare_transfer_hardware(master);
+
+	return 0;
+}
+
+/**
+ * zynqmp_qspi_resume:	Resume method for the QSPI driver
+ * @dev:	Address of the platform_device structure
+ *
+ * The function starts the QSPI driver queue and initializes the QSPI
+ * controller
+ *
+ * Return:	0 on success; error value otherwise
+ */
+static int __maybe_unused zynqmp_qspi_resume(struct device *dev)
+{
+	struct spi_master *master = dev_get_drvdata(dev);
+	struct zynqmp_qspi *xqspi = spi_master_get_devdata(master);
+	int ret = 0;
+
+	ret = clk_enable(xqspi->pclk);
+	if (ret) {
+		dev_err(dev, "Cannot enable APB clock.\n");
+		return ret;
+	}
+
+	ret = clk_enable(xqspi->refclk);
+	if (ret) {
+		dev_err(dev, "Cannot enable device clock.\n");
+		clk_disable(xqspi->pclk);
+		return ret;
+	}
+
+	spi_master_resume(master);
+
+	clk_disable(xqspi->refclk);
+	clk_disable(xqspi->pclk);
+	return 0;
+}
+
+/**
+ * zynqmp_runtime_suspend - Runtime suspend method for the SPI driver
+ * @dev:	Address of the platform_device structure
+ *
+ * This function disables the clocks
+ *
+ * Return:	Always 0
+ */
+static int __maybe_unused zynqmp_runtime_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct zynqmp_qspi *xqspi = spi_master_get_devdata(master);
+
+	clk_disable(xqspi->refclk);
+	clk_disable(xqspi->pclk);
+
+	return 0;
+}
+
+/**
+ * zynqmp_runtime_resume - Runtime resume method for the SPI driver
+ * @dev:	Address of the platform_device structure
+ *
+ * This function enables the clocks
+ *
+ * Return:	0 on success and error value on error
+ */
+static int __maybe_unused zynqmp_runtime_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct zynqmp_qspi *xqspi = spi_master_get_devdata(master);
+	int ret;
+
+	ret = clk_enable(xqspi->pclk);
+	if (ret) {
+		dev_err(dev, "Cannot enable APB clock.\n");
+		return ret;
+	}
+
+	ret = clk_enable(xqspi->refclk);
+	if (ret) {
+		dev_err(dev, "Cannot enable device clock.\n");
+		clk_disable(xqspi->pclk);
+		return ret;
+	}
+
+	return 0;
+}
+
+static const struct dev_pm_ops zynqmp_qspi_dev_pm_ops = {
+	SET_RUNTIME_PM_OPS(zynqmp_runtime_suspend,
+			   zynqmp_runtime_resume, NULL)
+	SET_SYSTEM_SLEEP_PM_OPS(zynqmp_qspi_suspend, zynqmp_qspi_resume)
+};
+
+/**
+ * zynqmp_qspi_probe:	Probe method for the QSPI driver
+ * @pdev:	Pointer to the platform_device structure
+ *
+ * This function initializes the driver data structures and the hardware.
+ *
+ * Return:	0 on success; error value otherwise
+ */
+static int zynqmp_qspi_probe(struct platform_device *pdev)
+{
+	int ret = 0;
+	struct spi_master *master;
+	struct zynqmp_qspi *xqspi;
+	struct resource *res;
+	struct device *dev = &pdev->dev;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*xqspi));
+	if (!master)
+		return -ENOMEM;
+
+	xqspi = spi_master_get_devdata(master);
+	master->dev.of_node = pdev->dev.of_node;
+	platform_set_drvdata(pdev, master);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	xqspi->regs = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(xqspi->regs)) {
+		ret = PTR_ERR(xqspi->regs);
+		goto remove_master;
+	}
+
+	xqspi->dev = dev;
+	xqspi->pclk = devm_clk_get(&pdev->dev, "pclk");
+	if (IS_ERR(xqspi->pclk)) {
+		dev_err(dev, "pclk clock not found.\n");
+		ret = PTR_ERR(xqspi->pclk);
+		goto remove_master;
+	}
+
+	ret = clk_prepare_enable(xqspi->pclk);
+	if (ret) {
+		dev_err(dev, "Unable to enable APB clock.\n");
+		goto remove_master;
+	}
+
+	xqspi->refclk = devm_clk_get(&pdev->dev, "ref_clk");
+	if (IS_ERR(xqspi->refclk)) {
+		dev_err(dev, "ref_clk clock not found.\n");
+		ret = PTR_ERR(xqspi->refclk);
+		goto clk_dis_pclk;
+	}
+
+	ret = clk_prepare_enable(xqspi->refclk);
+	if (ret) {
+		dev_err(dev, "Unable to enable device clock.\n");
+		goto clk_dis_pclk;
+	}
+
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_runtime_set_autosuspend_delay(&pdev->dev, SPI_AUTOSUSPEND_TIMEOUT);
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+	/* QSPI controller initializations */
+	zynqmp_qspi_init_hw(xqspi);
+
+	pm_runtime_mark_last_busy(&pdev->dev);
+	pm_runtime_put_autosuspend(&pdev->dev);
+	xqspi->irq = platform_get_irq(pdev, 0);
+	if (xqspi->irq <= 0) {
+		ret = -ENXIO;
+		dev_err(dev, "irq resource not found\n");
+		goto clk_dis_all;
+	}
+	ret = devm_request_irq(&pdev->dev, xqspi->irq, zynqmp_qspi_irq,
+			       0, pdev->name, master);
+	if (ret != 0) {
+		ret = -ENXIO;
+		dev_err(dev, "request_irq failed\n");
+		goto clk_dis_all;
+	}
+
+	master->num_chipselect = GQSPI_DEFAULT_NUM_CS;
+
+	master->setup = zynqmp_qspi_setup;
+	master->set_cs = zynqmp_qspi_chipselect;
+	master->transfer_one = zynqmp_qspi_start_transfer;
+	master->prepare_transfer_hardware = zynqmp_prepare_transfer_hardware;
+	master->unprepare_transfer_hardware =
+					zynqmp_unprepare_transfer_hardware;
+	master->max_speed_hz = clk_get_rate(xqspi->refclk) / 2;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_RX_DUAL | SPI_RX_QUAD |
+			    SPI_TX_DUAL | SPI_TX_QUAD;
+
+	if (master->dev.parent == NULL)
+		master->dev.parent = &master->dev;
+
+	ret = spi_register_master(master);
+	if (ret)
+		goto clk_dis_all;
+
+	return 0;
+
+clk_dis_all:
+	pm_runtime_set_suspended(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+	clk_disable_unprepare(xqspi->refclk);
+clk_dis_pclk:
+	clk_disable_unprepare(xqspi->pclk);
+remove_master:
+	spi_master_put(master);
+
+	return ret;
+}
+
+/**
+ * zynqmp_qspi_remove:	Remove method for the QSPI driver
+ * @pdev:	Pointer to the platform_device structure
+ *
+ * This function is called if a device is physically removed from the system or
+ * if the driver module is being unloaded. It frees all resources allocated to
+ * the device.
+ *
+ * Return:	0 Always
+ */
+static int zynqmp_qspi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct zynqmp_qspi *xqspi = spi_master_get_devdata(master);
+
+	zynqmp_gqspi_write(xqspi, GQSPI_EN_OFST, 0x0);
+	clk_disable_unprepare(xqspi->refclk);
+	clk_disable_unprepare(xqspi->pclk);
+	pm_runtime_set_suspended(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+
+	spi_unregister_master(master);
+
+	return 0;
+}
+
+static const struct of_device_id zynqmp_qspi_of_match[] = {
+	{ .compatible = "xlnx,zynqmp-qspi-1.0", },
+	{ /* End of table */ }
+};
+
+MODULE_DEVICE_TABLE(of, zynqmp_qspi_of_match);
+
+static struct platform_driver zynqmp_qspi_driver = {
+	.probe = zynqmp_qspi_probe,
+	.remove = zynqmp_qspi_remove,
+	.driver = {
+		.name = "zynqmp-qspi",
+		.of_match_table = zynqmp_qspi_of_match,
+		.pm = &zynqmp_qspi_dev_pm_ops,
+	},
+};
+
+module_platform_driver(zynqmp_qspi_driver);
+
+MODULE_AUTHOR("Xilinx, Inc.");
+MODULE_DESCRIPTION("Xilinx Zynqmp QSPI driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
new file mode 100644
index 0000000..9da0bc5
--- /dev/null
+++ b/drivers/spi/spi.c
@@ -0,0 +1,3543 @@
+/*
+ * SPI init/core code
+ *
+ * Copyright (C) 2005 David Brownell
+ * Copyright (C) 2008 Secret Lab Technologies Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/device.h>
+#include <linux/init.h>
+#include <linux/cache.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/mutex.h>
+#include <linux/of_device.h>
+#include <linux/of_irq.h>
+#include <linux/clk/clk-conf.h>
+#include <linux/slab.h>
+#include <linux/mod_devicetable.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi-mem.h>
+#include <linux/of_gpio.h>
+#include <linux/pm_runtime.h>
+#include <linux/pm_domain.h>
+#include <linux/property.h>
+#include <linux/export.h>
+#include <linux/sched/rt.h>
+#include <uapi/linux/sched/types.h>
+#include <linux/delay.h>
+#include <linux/kthread.h>
+#include <linux/ioport.h>
+#include <linux/acpi.h>
+#include <linux/highmem.h>
+#include <linux/idr.h>
+#include <linux/platform_data/x86/apple.h>
+
+#define CREATE_TRACE_POINTS
+#include <trace/events/spi.h>
+
+#include "internals.h"
+
+static DEFINE_IDR(spi_master_idr);
+
+static void spidev_release(struct device *dev)
+{
+	struct spi_device	*spi = to_spi_device(dev);
+
+	/* spi controllers may cleanup for released devices */
+	if (spi->controller->cleanup)
+		spi->controller->cleanup(spi);
+
+	spi_controller_put(spi->controller);
+	kfree(spi);
+}
+
+static ssize_t
+modalias_show(struct device *dev, struct device_attribute *a, char *buf)
+{
+	const struct spi_device	*spi = to_spi_device(dev);
+	int len;
+
+	len = acpi_device_modalias(dev, buf, PAGE_SIZE - 1);
+	if (len != -ENODEV)
+		return len;
+
+	return sprintf(buf, "%s%s\n", SPI_MODULE_PREFIX, spi->modalias);
+}
+static DEVICE_ATTR_RO(modalias);
+
+#define SPI_STATISTICS_ATTRS(field, file)				\
+static ssize_t spi_controller_##field##_show(struct device *dev,	\
+					     struct device_attribute *attr, \
+					     char *buf)			\
+{									\
+	struct spi_controller *ctlr = container_of(dev,			\
+					 struct spi_controller, dev);	\
+	return spi_statistics_##field##_show(&ctlr->statistics, buf);	\
+}									\
+static struct device_attribute dev_attr_spi_controller_##field = {	\
+	.attr = { .name = file, .mode = 0444 },				\
+	.show = spi_controller_##field##_show,				\
+};									\
+static ssize_t spi_device_##field##_show(struct device *dev,		\
+					 struct device_attribute *attr,	\
+					char *buf)			\
+{									\
+	struct spi_device *spi = to_spi_device(dev);			\
+	return spi_statistics_##field##_show(&spi->statistics, buf);	\
+}									\
+static struct device_attribute dev_attr_spi_device_##field = {		\
+	.attr = { .name = file, .mode = 0444 },				\
+	.show = spi_device_##field##_show,				\
+}
+
+#define SPI_STATISTICS_SHOW_NAME(name, file, field, format_string)	\
+static ssize_t spi_statistics_##name##_show(struct spi_statistics *stat, \
+					    char *buf)			\
+{									\
+	unsigned long flags;						\
+	ssize_t len;							\
+	spin_lock_irqsave(&stat->lock, flags);				\
+	len = sprintf(buf, format_string, stat->field);			\
+	spin_unlock_irqrestore(&stat->lock, flags);			\
+	return len;							\
+}									\
+SPI_STATISTICS_ATTRS(name, file)
+
+#define SPI_STATISTICS_SHOW(field, format_string)			\
+	SPI_STATISTICS_SHOW_NAME(field, __stringify(field),		\
+				 field, format_string)
+
+SPI_STATISTICS_SHOW(messages, "%lu");
+SPI_STATISTICS_SHOW(transfers, "%lu");
+SPI_STATISTICS_SHOW(errors, "%lu");
+SPI_STATISTICS_SHOW(timedout, "%lu");
+
+SPI_STATISTICS_SHOW(spi_sync, "%lu");
+SPI_STATISTICS_SHOW(spi_sync_immediate, "%lu");
+SPI_STATISTICS_SHOW(spi_async, "%lu");
+
+SPI_STATISTICS_SHOW(bytes, "%llu");
+SPI_STATISTICS_SHOW(bytes_rx, "%llu");
+SPI_STATISTICS_SHOW(bytes_tx, "%llu");
+
+#define SPI_STATISTICS_TRANSFER_BYTES_HISTO(index, number)		\
+	SPI_STATISTICS_SHOW_NAME(transfer_bytes_histo##index,		\
+				 "transfer_bytes_histo_" number,	\
+				 transfer_bytes_histo[index],  "%lu")
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(0,  "0-1");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(1,  "2-3");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(2,  "4-7");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(3,  "8-15");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(4,  "16-31");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(5,  "32-63");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(6,  "64-127");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(7,  "128-255");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(8,  "256-511");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(9,  "512-1023");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(10, "1024-2047");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(11, "2048-4095");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(12, "4096-8191");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(13, "8192-16383");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(14, "16384-32767");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(15, "32768-65535");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(16, "65536+");
+
+SPI_STATISTICS_SHOW(transfers_split_maxsize, "%lu");
+
+static struct attribute *spi_dev_attrs[] = {
+	&dev_attr_modalias.attr,
+	NULL,
+};
+
+static const struct attribute_group spi_dev_group = {
+	.attrs  = spi_dev_attrs,
+};
+
+static struct attribute *spi_device_statistics_attrs[] = {
+	&dev_attr_spi_device_messages.attr,
+	&dev_attr_spi_device_transfers.attr,
+	&dev_attr_spi_device_errors.attr,
+	&dev_attr_spi_device_timedout.attr,
+	&dev_attr_spi_device_spi_sync.attr,
+	&dev_attr_spi_device_spi_sync_immediate.attr,
+	&dev_attr_spi_device_spi_async.attr,
+	&dev_attr_spi_device_bytes.attr,
+	&dev_attr_spi_device_bytes_rx.attr,
+	&dev_attr_spi_device_bytes_tx.attr,
+	&dev_attr_spi_device_transfer_bytes_histo0.attr,
+	&dev_attr_spi_device_transfer_bytes_histo1.attr,
+	&dev_attr_spi_device_transfer_bytes_histo2.attr,
+	&dev_attr_spi_device_transfer_bytes_histo3.attr,
+	&dev_attr_spi_device_transfer_bytes_histo4.attr,
+	&dev_attr_spi_device_transfer_bytes_histo5.attr,
+	&dev_attr_spi_device_transfer_bytes_histo6.attr,
+	&dev_attr_spi_device_transfer_bytes_histo7.attr,
+	&dev_attr_spi_device_transfer_bytes_histo8.attr,
+	&dev_attr_spi_device_transfer_bytes_histo9.attr,
+	&dev_attr_spi_device_transfer_bytes_histo10.attr,
+	&dev_attr_spi_device_transfer_bytes_histo11.attr,
+	&dev_attr_spi_device_transfer_bytes_histo12.attr,
+	&dev_attr_spi_device_transfer_bytes_histo13.attr,
+	&dev_attr_spi_device_transfer_bytes_histo14.attr,
+	&dev_attr_spi_device_transfer_bytes_histo15.attr,
+	&dev_attr_spi_device_transfer_bytes_histo16.attr,
+	&dev_attr_spi_device_transfers_split_maxsize.attr,
+	NULL,
+};
+
+static const struct attribute_group spi_device_statistics_group = {
+	.name  = "statistics",
+	.attrs  = spi_device_statistics_attrs,
+};
+
+static const struct attribute_group *spi_dev_groups[] = {
+	&spi_dev_group,
+	&spi_device_statistics_group,
+	NULL,
+};
+
+static struct attribute *spi_controller_statistics_attrs[] = {
+	&dev_attr_spi_controller_messages.attr,
+	&dev_attr_spi_controller_transfers.attr,
+	&dev_attr_spi_controller_errors.attr,
+	&dev_attr_spi_controller_timedout.attr,
+	&dev_attr_spi_controller_spi_sync.attr,
+	&dev_attr_spi_controller_spi_sync_immediate.attr,
+	&dev_attr_spi_controller_spi_async.attr,
+	&dev_attr_spi_controller_bytes.attr,
+	&dev_attr_spi_controller_bytes_rx.attr,
+	&dev_attr_spi_controller_bytes_tx.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo0.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo1.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo2.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo3.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo4.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo5.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo6.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo7.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo8.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo9.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo10.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo11.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo12.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo13.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo14.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo15.attr,
+	&dev_attr_spi_controller_transfer_bytes_histo16.attr,
+	&dev_attr_spi_controller_transfers_split_maxsize.attr,
+	NULL,
+};
+
+static const struct attribute_group spi_controller_statistics_group = {
+	.name  = "statistics",
+	.attrs  = spi_controller_statistics_attrs,
+};
+
+static const struct attribute_group *spi_master_groups[] = {
+	&spi_controller_statistics_group,
+	NULL,
+};
+
+void spi_statistics_add_transfer_stats(struct spi_statistics *stats,
+				       struct spi_transfer *xfer,
+				       struct spi_controller *ctlr)
+{
+	unsigned long flags;
+	int l2len = min(fls(xfer->len), SPI_STATISTICS_HISTO_SIZE) - 1;
+
+	if (l2len < 0)
+		l2len = 0;
+
+	spin_lock_irqsave(&stats->lock, flags);
+
+	stats->transfers++;
+	stats->transfer_bytes_histo[l2len]++;
+
+	stats->bytes += xfer->len;
+	if ((xfer->tx_buf) &&
+	    (xfer->tx_buf != ctlr->dummy_tx))
+		stats->bytes_tx += xfer->len;
+	if ((xfer->rx_buf) &&
+	    (xfer->rx_buf != ctlr->dummy_rx))
+		stats->bytes_rx += xfer->len;
+
+	spin_unlock_irqrestore(&stats->lock, flags);
+}
+EXPORT_SYMBOL_GPL(spi_statistics_add_transfer_stats);
+
+/* modalias support makes "modprobe $MODALIAS" new-style hotplug work,
+ * and the sysfs version makes coldplug work too.
+ */
+
+static const struct spi_device_id *spi_match_id(const struct spi_device_id *id,
+						const struct spi_device *sdev)
+{
+	while (id->name[0]) {
+		if (!strcmp(sdev->modalias, id->name))
+			return id;
+		id++;
+	}
+	return NULL;
+}
+
+const struct spi_device_id *spi_get_device_id(const struct spi_device *sdev)
+{
+	const struct spi_driver *sdrv = to_spi_driver(sdev->dev.driver);
+
+	return spi_match_id(sdrv->id_table, sdev);
+}
+EXPORT_SYMBOL_GPL(spi_get_device_id);
+
+static int spi_match_device(struct device *dev, struct device_driver *drv)
+{
+	const struct spi_device	*spi = to_spi_device(dev);
+	const struct spi_driver	*sdrv = to_spi_driver(drv);
+
+	/* Attempt an OF style match */
+	if (of_driver_match_device(dev, drv))
+		return 1;
+
+	/* Then try ACPI */
+	if (acpi_driver_match_device(dev, drv))
+		return 1;
+
+	if (sdrv->id_table)
+		return !!spi_match_id(sdrv->id_table, spi);
+
+	return strcmp(spi->modalias, drv->name) == 0;
+}
+
+static int spi_uevent(struct device *dev, struct kobj_uevent_env *env)
+{
+	const struct spi_device		*spi = to_spi_device(dev);
+	int rc;
+
+	rc = acpi_device_uevent_modalias(dev, env);
+	if (rc != -ENODEV)
+		return rc;
+
+	return add_uevent_var(env, "MODALIAS=%s%s", SPI_MODULE_PREFIX, spi->modalias);
+}
+
+struct bus_type spi_bus_type = {
+	.name		= "spi",
+	.dev_groups	= spi_dev_groups,
+	.match		= spi_match_device,
+	.uevent		= spi_uevent,
+};
+EXPORT_SYMBOL_GPL(spi_bus_type);
+
+
+static int spi_drv_probe(struct device *dev)
+{
+	const struct spi_driver		*sdrv = to_spi_driver(dev->driver);
+	struct spi_device		*spi = to_spi_device(dev);
+	int ret;
+
+	ret = of_clk_set_defaults(dev->of_node, false);
+	if (ret)
+		return ret;
+
+	if (dev->of_node) {
+		spi->irq = of_irq_get(dev->of_node, 0);
+		if (spi->irq == -EPROBE_DEFER)
+			return -EPROBE_DEFER;
+		if (spi->irq < 0)
+			spi->irq = 0;
+	}
+
+	ret = dev_pm_domain_attach(dev, true);
+	if (ret)
+		return ret;
+
+	ret = sdrv->probe(spi);
+	if (ret)
+		dev_pm_domain_detach(dev, true);
+
+	return ret;
+}
+
+static int spi_drv_remove(struct device *dev)
+{
+	const struct spi_driver		*sdrv = to_spi_driver(dev->driver);
+	int ret;
+
+	ret = sdrv->remove(to_spi_device(dev));
+	dev_pm_domain_detach(dev, true);
+
+	return ret;
+}
+
+static void spi_drv_shutdown(struct device *dev)
+{
+	const struct spi_driver		*sdrv = to_spi_driver(dev->driver);
+
+	sdrv->shutdown(to_spi_device(dev));
+}
+
+/**
+ * __spi_register_driver - register a SPI driver
+ * @owner: owner module of the driver to register
+ * @sdrv: the driver to register
+ * Context: can sleep
+ *
+ * Return: zero on success, else a negative error code.
+ */
+int __spi_register_driver(struct module *owner, struct spi_driver *sdrv)
+{
+	sdrv->driver.owner = owner;
+	sdrv->driver.bus = &spi_bus_type;
+	if (sdrv->probe)
+		sdrv->driver.probe = spi_drv_probe;
+	if (sdrv->remove)
+		sdrv->driver.remove = spi_drv_remove;
+	if (sdrv->shutdown)
+		sdrv->driver.shutdown = spi_drv_shutdown;
+	return driver_register(&sdrv->driver);
+}
+EXPORT_SYMBOL_GPL(__spi_register_driver);
+
+/*-------------------------------------------------------------------------*/
+
+/* SPI devices should normally not be created by SPI device drivers; that
+ * would make them board-specific.  Similarly with SPI controller drivers.
+ * Device registration normally goes into like arch/.../mach.../board-YYY.c
+ * with other readonly (flashable) information about mainboard devices.
+ */
+
+struct boardinfo {
+	struct list_head	list;
+	struct spi_board_info	board_info;
+};
+
+static LIST_HEAD(board_list);
+static LIST_HEAD(spi_controller_list);
+
+/*
+ * Used to protect add/del opertion for board_info list and
+ * spi_controller list, and their matching process
+ * also used to protect object of type struct idr
+ */
+static DEFINE_MUTEX(board_lock);
+
+/**
+ * spi_alloc_device - Allocate a new SPI device
+ * @ctlr: Controller to which device is connected
+ * Context: can sleep
+ *
+ * Allows a driver to allocate and initialize a spi_device without
+ * registering it immediately.  This allows a driver to directly
+ * fill the spi_device with device parameters before calling
+ * spi_add_device() on it.
+ *
+ * Caller is responsible to call spi_add_device() on the returned
+ * spi_device structure to add it to the SPI controller.  If the caller
+ * needs to discard the spi_device without adding it, then it should
+ * call spi_dev_put() on it.
+ *
+ * Return: a pointer to the new device, or NULL.
+ */
+struct spi_device *spi_alloc_device(struct spi_controller *ctlr)
+{
+	struct spi_device	*spi;
+
+	if (!spi_controller_get(ctlr))
+		return NULL;
+
+	spi = kzalloc(sizeof(*spi), GFP_KERNEL);
+	if (!spi) {
+		spi_controller_put(ctlr);
+		return NULL;
+	}
+
+	spi->master = spi->controller = ctlr;
+	spi->dev.parent = &ctlr->dev;
+	spi->dev.bus = &spi_bus_type;
+	spi->dev.release = spidev_release;
+	spi->cs_gpio = -ENOENT;
+
+	spin_lock_init(&spi->statistics.lock);
+
+	device_initialize(&spi->dev);
+	return spi;
+}
+EXPORT_SYMBOL_GPL(spi_alloc_device);
+
+static void spi_dev_set_name(struct spi_device *spi)
+{
+	struct acpi_device *adev = ACPI_COMPANION(&spi->dev);
+
+	if (adev) {
+		dev_set_name(&spi->dev, "spi-%s", acpi_dev_name(adev));
+		return;
+	}
+
+	dev_set_name(&spi->dev, "%s.%u", dev_name(&spi->controller->dev),
+		     spi->chip_select);
+}
+
+static int spi_dev_check(struct device *dev, void *data)
+{
+	struct spi_device *spi = to_spi_device(dev);
+	struct spi_device *new_spi = data;
+
+	if (spi->controller == new_spi->controller &&
+	    spi->chip_select == new_spi->chip_select)
+		return -EBUSY;
+	return 0;
+}
+
+/**
+ * spi_add_device - Add spi_device allocated with spi_alloc_device
+ * @spi: spi_device to register
+ *
+ * Companion function to spi_alloc_device.  Devices allocated with
+ * spi_alloc_device can be added onto the spi bus with this function.
+ *
+ * Return: 0 on success; negative errno on failure
+ */
+int spi_add_device(struct spi_device *spi)
+{
+	static DEFINE_MUTEX(spi_add_lock);
+	struct spi_controller *ctlr = spi->controller;
+	struct device *dev = ctlr->dev.parent;
+	int status;
+
+	/* Chipselects are numbered 0..max; validate. */
+	if (spi->chip_select >= ctlr->num_chipselect) {
+		dev_err(dev, "cs%d >= max %d\n", spi->chip_select,
+			ctlr->num_chipselect);
+		return -EINVAL;
+	}
+
+	/* Set the bus ID string */
+	spi_dev_set_name(spi);
+
+	/* We need to make sure there's no other device with this
+	 * chipselect **BEFORE** we call setup(), else we'll trash
+	 * its configuration.  Lock against concurrent add() calls.
+	 */
+	mutex_lock(&spi_add_lock);
+
+	status = bus_for_each_dev(&spi_bus_type, NULL, spi, spi_dev_check);
+	if (status) {
+		dev_err(dev, "chipselect %d already in use\n",
+				spi->chip_select);
+		goto done;
+	}
+
+	if (ctlr->cs_gpios)
+		spi->cs_gpio = ctlr->cs_gpios[spi->chip_select];
+
+	/* Drivers may modify this initial i/o setup, but will
+	 * normally rely on the device being setup.  Devices
+	 * using SPI_CS_HIGH can't coexist well otherwise...
+	 */
+	status = spi_setup(spi);
+	if (status < 0) {
+		dev_err(dev, "can't setup %s, status %d\n",
+				dev_name(&spi->dev), status);
+		goto done;
+	}
+
+	/* Device may be bound to an active driver when this returns */
+	status = device_add(&spi->dev);
+	if (status < 0)
+		dev_err(dev, "can't add %s, status %d\n",
+				dev_name(&spi->dev), status);
+	else
+		dev_dbg(dev, "registered child %s\n", dev_name(&spi->dev));
+
+done:
+	mutex_unlock(&spi_add_lock);
+	return status;
+}
+EXPORT_SYMBOL_GPL(spi_add_device);
+
+/**
+ * spi_new_device - instantiate one new SPI device
+ * @ctlr: Controller to which device is connected
+ * @chip: Describes the SPI device
+ * Context: can sleep
+ *
+ * On typical mainboards, this is purely internal; and it's not needed
+ * after board init creates the hard-wired devices.  Some development
+ * platforms may not be able to use spi_register_board_info though, and
+ * this is exported so that for example a USB or parport based adapter
+ * driver could add devices (which it would learn about out-of-band).
+ *
+ * Return: the new device, or NULL.
+ */
+struct spi_device *spi_new_device(struct spi_controller *ctlr,
+				  struct spi_board_info *chip)
+{
+	struct spi_device	*proxy;
+	int			status;
+
+	/* NOTE:  caller did any chip->bus_num checks necessary.
+	 *
+	 * Also, unless we change the return value convention to use
+	 * error-or-pointer (not NULL-or-pointer), troubleshootability
+	 * suggests syslogged diagnostics are best here (ugh).
+	 */
+
+	proxy = spi_alloc_device(ctlr);
+	if (!proxy)
+		return NULL;
+
+	WARN_ON(strlen(chip->modalias) >= sizeof(proxy->modalias));
+
+	proxy->chip_select = chip->chip_select;
+	proxy->max_speed_hz = chip->max_speed_hz;
+	proxy->mode = chip->mode;
+	proxy->irq = chip->irq;
+	strlcpy(proxy->modalias, chip->modalias, sizeof(proxy->modalias));
+	proxy->dev.platform_data = (void *) chip->platform_data;
+	proxy->controller_data = chip->controller_data;
+	proxy->controller_state = NULL;
+
+	if (chip->properties) {
+		status = device_add_properties(&proxy->dev, chip->properties);
+		if (status) {
+			dev_err(&ctlr->dev,
+				"failed to add properties to '%s': %d\n",
+				chip->modalias, status);
+			goto err_dev_put;
+		}
+	}
+
+	status = spi_add_device(proxy);
+	if (status < 0)
+		goto err_remove_props;
+
+	return proxy;
+
+err_remove_props:
+	if (chip->properties)
+		device_remove_properties(&proxy->dev);
+err_dev_put:
+	spi_dev_put(proxy);
+	return NULL;
+}
+EXPORT_SYMBOL_GPL(spi_new_device);
+
+/**
+ * spi_unregister_device - unregister a single SPI device
+ * @spi: spi_device to unregister
+ *
+ * Start making the passed SPI device vanish. Normally this would be handled
+ * by spi_unregister_controller().
+ */
+void spi_unregister_device(struct spi_device *spi)
+{
+	if (!spi)
+		return;
+
+	if (spi->dev.of_node) {
+		of_node_clear_flag(spi->dev.of_node, OF_POPULATED);
+		of_node_put(spi->dev.of_node);
+	}
+	if (ACPI_COMPANION(&spi->dev))
+		acpi_device_clear_enumerated(ACPI_COMPANION(&spi->dev));
+	device_unregister(&spi->dev);
+}
+EXPORT_SYMBOL_GPL(spi_unregister_device);
+
+static void spi_match_controller_to_boardinfo(struct spi_controller *ctlr,
+					      struct spi_board_info *bi)
+{
+	struct spi_device *dev;
+
+	if (ctlr->bus_num != bi->bus_num)
+		return;
+
+	dev = spi_new_device(ctlr, bi);
+	if (!dev)
+		dev_err(ctlr->dev.parent, "can't create new device for %s\n",
+			bi->modalias);
+}
+
+/**
+ * spi_register_board_info - register SPI devices for a given board
+ * @info: array of chip descriptors
+ * @n: how many descriptors are provided
+ * Context: can sleep
+ *
+ * Board-specific early init code calls this (probably during arch_initcall)
+ * with segments of the SPI device table.  Any device nodes are created later,
+ * after the relevant parent SPI controller (bus_num) is defined.  We keep
+ * this table of devices forever, so that reloading a controller driver will
+ * not make Linux forget about these hard-wired devices.
+ *
+ * Other code can also call this, e.g. a particular add-on board might provide
+ * SPI devices through its expansion connector, so code initializing that board
+ * would naturally declare its SPI devices.
+ *
+ * The board info passed can safely be __initdata ... but be careful of
+ * any embedded pointers (platform_data, etc), they're copied as-is.
+ * Device properties are deep-copied though.
+ *
+ * Return: zero on success, else a negative error code.
+ */
+int spi_register_board_info(struct spi_board_info const *info, unsigned n)
+{
+	struct boardinfo *bi;
+	int i;
+
+	if (!n)
+		return 0;
+
+	bi = kcalloc(n, sizeof(*bi), GFP_KERNEL);
+	if (!bi)
+		return -ENOMEM;
+
+	for (i = 0; i < n; i++, bi++, info++) {
+		struct spi_controller *ctlr;
+
+		memcpy(&bi->board_info, info, sizeof(*info));
+		if (info->properties) {
+			bi->board_info.properties =
+					property_entries_dup(info->properties);
+			if (IS_ERR(bi->board_info.properties))
+				return PTR_ERR(bi->board_info.properties);
+		}
+
+		mutex_lock(&board_lock);
+		list_add_tail(&bi->list, &board_list);
+		list_for_each_entry(ctlr, &spi_controller_list, list)
+			spi_match_controller_to_boardinfo(ctlr,
+							  &bi->board_info);
+		mutex_unlock(&board_lock);
+	}
+
+	return 0;
+}
+
+/*-------------------------------------------------------------------------*/
+
+static void spi_set_cs(struct spi_device *spi, bool enable)
+{
+	if (spi->mode & SPI_CS_HIGH)
+		enable = !enable;
+
+	if (gpio_is_valid(spi->cs_gpio)) {
+		gpio_set_value(spi->cs_gpio, !enable);
+		/* Some SPI masters need both GPIO CS & slave_select */
+		if ((spi->controller->flags & SPI_MASTER_GPIO_SS) &&
+		    spi->controller->set_cs)
+			spi->controller->set_cs(spi, !enable);
+	} else if (spi->controller->set_cs) {
+		spi->controller->set_cs(spi, !enable);
+	}
+}
+
+#ifdef CONFIG_HAS_DMA
+int spi_map_buf(struct spi_controller *ctlr, struct device *dev,
+		struct sg_table *sgt, void *buf, size_t len,
+		enum dma_data_direction dir)
+{
+	const bool vmalloced_buf = is_vmalloc_addr(buf);
+	unsigned int max_seg_size = dma_get_max_seg_size(dev);
+#ifdef CONFIG_HIGHMEM
+	const bool kmap_buf = ((unsigned long)buf >= PKMAP_BASE &&
+				(unsigned long)buf < (PKMAP_BASE +
+					(LAST_PKMAP * PAGE_SIZE)));
+#else
+	const bool kmap_buf = false;
+#endif
+	int desc_len;
+	int sgs;
+	struct page *vm_page;
+	struct scatterlist *sg;
+	void *sg_buf;
+	size_t min;
+	int i, ret;
+
+	if (vmalloced_buf || kmap_buf) {
+		desc_len = min_t(int, max_seg_size, PAGE_SIZE);
+		sgs = DIV_ROUND_UP(len + offset_in_page(buf), desc_len);
+	} else if (virt_addr_valid(buf)) {
+		desc_len = min_t(int, max_seg_size, ctlr->max_dma_len);
+		sgs = DIV_ROUND_UP(len, desc_len);
+	} else {
+		return -EINVAL;
+	}
+
+	ret = sg_alloc_table(sgt, sgs, GFP_KERNEL);
+	if (ret != 0)
+		return ret;
+
+	sg = &sgt->sgl[0];
+	for (i = 0; i < sgs; i++) {
+
+		if (vmalloced_buf || kmap_buf) {
+			/*
+			 * Next scatterlist entry size is the minimum between
+			 * the desc_len and the remaining buffer length that
+			 * fits in a page.
+			 */
+			min = min_t(size_t, desc_len,
+				    min_t(size_t, len,
+					  PAGE_SIZE - offset_in_page(buf)));
+			if (vmalloced_buf)
+				vm_page = vmalloc_to_page(buf);
+			else
+				vm_page = kmap_to_page(buf);
+			if (!vm_page) {
+				sg_free_table(sgt);
+				return -ENOMEM;
+			}
+			sg_set_page(sg, vm_page,
+				    min, offset_in_page(buf));
+		} else {
+			min = min_t(size_t, len, desc_len);
+			sg_buf = buf;
+			sg_set_buf(sg, sg_buf, min);
+		}
+
+		buf += min;
+		len -= min;
+		sg = sg_next(sg);
+	}
+
+	ret = dma_map_sg(dev, sgt->sgl, sgt->nents, dir);
+	if (!ret)
+		ret = -ENOMEM;
+	if (ret < 0) {
+		sg_free_table(sgt);
+		return ret;
+	}
+
+	sgt->nents = ret;
+
+	return 0;
+}
+
+void spi_unmap_buf(struct spi_controller *ctlr, struct device *dev,
+		   struct sg_table *sgt, enum dma_data_direction dir)
+{
+	if (sgt->orig_nents) {
+		dma_unmap_sg(dev, sgt->sgl, sgt->orig_nents, dir);
+		sg_free_table(sgt);
+	}
+}
+
+static int __spi_map_msg(struct spi_controller *ctlr, struct spi_message *msg)
+{
+	struct device *tx_dev, *rx_dev;
+	struct spi_transfer *xfer;
+	int ret;
+
+	if (!ctlr->can_dma)
+		return 0;
+
+	if (ctlr->dma_tx)
+		tx_dev = ctlr->dma_tx->device->dev;
+	else
+		tx_dev = ctlr->dev.parent;
+
+	if (ctlr->dma_rx)
+		rx_dev = ctlr->dma_rx->device->dev;
+	else
+		rx_dev = ctlr->dev.parent;
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		if (!ctlr->can_dma(ctlr, msg->spi, xfer))
+			continue;
+
+		if (xfer->tx_buf != NULL) {
+			ret = spi_map_buf(ctlr, tx_dev, &xfer->tx_sg,
+					  (void *)xfer->tx_buf, xfer->len,
+					  DMA_TO_DEVICE);
+			if (ret != 0)
+				return ret;
+		}
+
+		if (xfer->rx_buf != NULL) {
+			ret = spi_map_buf(ctlr, rx_dev, &xfer->rx_sg,
+					  xfer->rx_buf, xfer->len,
+					  DMA_FROM_DEVICE);
+			if (ret != 0) {
+				spi_unmap_buf(ctlr, tx_dev, &xfer->tx_sg,
+					      DMA_TO_DEVICE);
+				return ret;
+			}
+		}
+	}
+
+	ctlr->cur_msg_mapped = true;
+
+	return 0;
+}
+
+static int __spi_unmap_msg(struct spi_controller *ctlr, struct spi_message *msg)
+{
+	struct spi_transfer *xfer;
+	struct device *tx_dev, *rx_dev;
+
+	if (!ctlr->cur_msg_mapped || !ctlr->can_dma)
+		return 0;
+
+	if (ctlr->dma_tx)
+		tx_dev = ctlr->dma_tx->device->dev;
+	else
+		tx_dev = ctlr->dev.parent;
+
+	if (ctlr->dma_rx)
+		rx_dev = ctlr->dma_rx->device->dev;
+	else
+		rx_dev = ctlr->dev.parent;
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		if (!ctlr->can_dma(ctlr, msg->spi, xfer))
+			continue;
+
+		spi_unmap_buf(ctlr, rx_dev, &xfer->rx_sg, DMA_FROM_DEVICE);
+		spi_unmap_buf(ctlr, tx_dev, &xfer->tx_sg, DMA_TO_DEVICE);
+	}
+
+	return 0;
+}
+#else /* !CONFIG_HAS_DMA */
+static inline int __spi_map_msg(struct spi_controller *ctlr,
+				struct spi_message *msg)
+{
+	return 0;
+}
+
+static inline int __spi_unmap_msg(struct spi_controller *ctlr,
+				  struct spi_message *msg)
+{
+	return 0;
+}
+#endif /* !CONFIG_HAS_DMA */
+
+static inline int spi_unmap_msg(struct spi_controller *ctlr,
+				struct spi_message *msg)
+{
+	struct spi_transfer *xfer;
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		/*
+		 * Restore the original value of tx_buf or rx_buf if they are
+		 * NULL.
+		 */
+		if (xfer->tx_buf == ctlr->dummy_tx)
+			xfer->tx_buf = NULL;
+		if (xfer->rx_buf == ctlr->dummy_rx)
+			xfer->rx_buf = NULL;
+	}
+
+	return __spi_unmap_msg(ctlr, msg);
+}
+
+static int spi_map_msg(struct spi_controller *ctlr, struct spi_message *msg)
+{
+	struct spi_transfer *xfer;
+	void *tmp;
+	unsigned int max_tx, max_rx;
+
+	if (ctlr->flags & (SPI_CONTROLLER_MUST_RX | SPI_CONTROLLER_MUST_TX)) {
+		max_tx = 0;
+		max_rx = 0;
+
+		list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+			if ((ctlr->flags & SPI_CONTROLLER_MUST_TX) &&
+			    !xfer->tx_buf)
+				max_tx = max(xfer->len, max_tx);
+			if ((ctlr->flags & SPI_CONTROLLER_MUST_RX) &&
+			    !xfer->rx_buf)
+				max_rx = max(xfer->len, max_rx);
+		}
+
+		if (max_tx) {
+			tmp = krealloc(ctlr->dummy_tx, max_tx,
+				       GFP_KERNEL | GFP_DMA);
+			if (!tmp)
+				return -ENOMEM;
+			ctlr->dummy_tx = tmp;
+			memset(tmp, 0, max_tx);
+		}
+
+		if (max_rx) {
+			tmp = krealloc(ctlr->dummy_rx, max_rx,
+				       GFP_KERNEL | GFP_DMA);
+			if (!tmp)
+				return -ENOMEM;
+			ctlr->dummy_rx = tmp;
+		}
+
+		if (max_tx || max_rx) {
+			list_for_each_entry(xfer, &msg->transfers,
+					    transfer_list) {
+				if (!xfer->tx_buf)
+					xfer->tx_buf = ctlr->dummy_tx;
+				if (!xfer->rx_buf)
+					xfer->rx_buf = ctlr->dummy_rx;
+			}
+		}
+	}
+
+	return __spi_map_msg(ctlr, msg);
+}
+
+/*
+ * spi_transfer_one_message - Default implementation of transfer_one_message()
+ *
+ * This is a standard implementation of transfer_one_message() for
+ * drivers which implement a transfer_one() operation.  It provides
+ * standard handling of delays and chip select management.
+ */
+static int spi_transfer_one_message(struct spi_controller *ctlr,
+				    struct spi_message *msg)
+{
+	struct spi_transfer *xfer;
+	bool keep_cs = false;
+	int ret = 0;
+	unsigned long long ms = 1;
+	struct spi_statistics *statm = &ctlr->statistics;
+	struct spi_statistics *stats = &msg->spi->statistics;
+
+	spi_set_cs(msg->spi, true);
+
+	SPI_STATISTICS_INCREMENT_FIELD(statm, messages);
+	SPI_STATISTICS_INCREMENT_FIELD(stats, messages);
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		trace_spi_transfer_start(msg, xfer);
+
+		spi_statistics_add_transfer_stats(statm, xfer, ctlr);
+		spi_statistics_add_transfer_stats(stats, xfer, ctlr);
+
+		if (xfer->tx_buf || xfer->rx_buf) {
+			reinit_completion(&ctlr->xfer_completion);
+
+			ret = ctlr->transfer_one(ctlr, msg->spi, xfer);
+			if (ret < 0) {
+				SPI_STATISTICS_INCREMENT_FIELD(statm,
+							       errors);
+				SPI_STATISTICS_INCREMENT_FIELD(stats,
+							       errors);
+				dev_err(&msg->spi->dev,
+					"SPI transfer failed: %d\n", ret);
+				goto out;
+			}
+
+			if (ret > 0) {
+				ret = 0;
+				ms = 8LL * 1000LL * xfer->len;
+				do_div(ms, xfer->speed_hz);
+				ms += ms + 200; /* some tolerance */
+
+				if (ms > UINT_MAX)
+					ms = UINT_MAX;
+
+				ms = wait_for_completion_timeout(&ctlr->xfer_completion,
+								 msecs_to_jiffies(ms));
+			}
+
+			if (ms == 0) {
+				SPI_STATISTICS_INCREMENT_FIELD(statm,
+							       timedout);
+				SPI_STATISTICS_INCREMENT_FIELD(stats,
+							       timedout);
+				dev_err(&msg->spi->dev,
+					"SPI transfer timed out\n");
+				msg->status = -ETIMEDOUT;
+			}
+		} else {
+			if (xfer->len)
+				dev_err(&msg->spi->dev,
+					"Bufferless transfer has length %u\n",
+					xfer->len);
+		}
+
+		trace_spi_transfer_stop(msg, xfer);
+
+		if (msg->status != -EINPROGRESS)
+			goto out;
+
+		if (xfer->delay_usecs) {
+			u16 us = xfer->delay_usecs;
+
+			if (us <= 10)
+				udelay(us);
+			else
+				usleep_range(us, us + DIV_ROUND_UP(us, 10));
+		}
+
+		if (xfer->cs_change) {
+			if (list_is_last(&xfer->transfer_list,
+					 &msg->transfers)) {
+				keep_cs = true;
+			} else {
+				spi_set_cs(msg->spi, false);
+				udelay(10);
+				spi_set_cs(msg->spi, true);
+			}
+		}
+
+		msg->actual_length += xfer->len;
+	}
+
+out:
+	if (ret != 0 || !keep_cs)
+		spi_set_cs(msg->spi, false);
+
+	if (msg->status == -EINPROGRESS)
+		msg->status = ret;
+
+	if (msg->status && ctlr->handle_err)
+		ctlr->handle_err(ctlr, msg);
+
+	spi_res_release(ctlr, msg);
+
+	spi_finalize_current_message(ctlr);
+
+	return ret;
+}
+
+/**
+ * spi_finalize_current_transfer - report completion of a transfer
+ * @ctlr: the controller reporting completion
+ *
+ * Called by SPI drivers using the core transfer_one_message()
+ * implementation to notify it that the current interrupt driven
+ * transfer has finished and the next one may be scheduled.
+ */
+void spi_finalize_current_transfer(struct spi_controller *ctlr)
+{
+	complete(&ctlr->xfer_completion);
+}
+EXPORT_SYMBOL_GPL(spi_finalize_current_transfer);
+
+/**
+ * __spi_pump_messages - function which processes spi message queue
+ * @ctlr: controller to process queue for
+ * @in_kthread: true if we are in the context of the message pump thread
+ *
+ * This function checks if there is any spi message in the queue that
+ * needs processing and if so call out to the driver to initialize hardware
+ * and transfer each message.
+ *
+ * Note that it is called both from the kthread itself and also from
+ * inside spi_sync(); the queue extraction handling at the top of the
+ * function should deal with this safely.
+ */
+static void __spi_pump_messages(struct spi_controller *ctlr, bool in_kthread)
+{
+	unsigned long flags;
+	bool was_busy = false;
+	int ret;
+
+	/* Lock queue */
+	spin_lock_irqsave(&ctlr->queue_lock, flags);
+
+	/* Make sure we are not already running a message */
+	if (ctlr->cur_msg) {
+		spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+		return;
+	}
+
+	/* If another context is idling the device then defer */
+	if (ctlr->idling) {
+		kthread_queue_work(&ctlr->kworker, &ctlr->pump_messages);
+		spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+		return;
+	}
+
+	/* Check if the queue is idle */
+	if (list_empty(&ctlr->queue) || !ctlr->running) {
+		if (!ctlr->busy) {
+			spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+			return;
+		}
+
+		/* Only do teardown in the thread */
+		if (!in_kthread) {
+			kthread_queue_work(&ctlr->kworker,
+					   &ctlr->pump_messages);
+			spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+			return;
+		}
+
+		ctlr->busy = false;
+		ctlr->idling = true;
+		spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+
+		kfree(ctlr->dummy_rx);
+		ctlr->dummy_rx = NULL;
+		kfree(ctlr->dummy_tx);
+		ctlr->dummy_tx = NULL;
+		if (ctlr->unprepare_transfer_hardware &&
+		    ctlr->unprepare_transfer_hardware(ctlr))
+			dev_err(&ctlr->dev,
+				"failed to unprepare transfer hardware\n");
+		if (ctlr->auto_runtime_pm) {
+			pm_runtime_mark_last_busy(ctlr->dev.parent);
+			pm_runtime_put_autosuspend(ctlr->dev.parent);
+		}
+		trace_spi_controller_idle(ctlr);
+
+		spin_lock_irqsave(&ctlr->queue_lock, flags);
+		ctlr->idling = false;
+		spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+		return;
+	}
+
+	/* Extract head of queue */
+	ctlr->cur_msg =
+		list_first_entry(&ctlr->queue, struct spi_message, queue);
+
+	list_del_init(&ctlr->cur_msg->queue);
+	if (ctlr->busy)
+		was_busy = true;
+	else
+		ctlr->busy = true;
+	spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+
+	mutex_lock(&ctlr->io_mutex);
+
+	if (!was_busy && ctlr->auto_runtime_pm) {
+		ret = pm_runtime_get_sync(ctlr->dev.parent);
+		if (ret < 0) {
+			pm_runtime_put_noidle(ctlr->dev.parent);
+			dev_err(&ctlr->dev, "Failed to power device: %d\n",
+				ret);
+			mutex_unlock(&ctlr->io_mutex);
+			return;
+		}
+	}
+
+	if (!was_busy)
+		trace_spi_controller_busy(ctlr);
+
+	if (!was_busy && ctlr->prepare_transfer_hardware) {
+		ret = ctlr->prepare_transfer_hardware(ctlr);
+		if (ret) {
+			dev_err(&ctlr->dev,
+				"failed to prepare transfer hardware\n");
+
+			if (ctlr->auto_runtime_pm)
+				pm_runtime_put(ctlr->dev.parent);
+			mutex_unlock(&ctlr->io_mutex);
+			return;
+		}
+	}
+
+	trace_spi_message_start(ctlr->cur_msg);
+
+	if (ctlr->prepare_message) {
+		ret = ctlr->prepare_message(ctlr, ctlr->cur_msg);
+		if (ret) {
+			dev_err(&ctlr->dev, "failed to prepare message: %d\n",
+				ret);
+			ctlr->cur_msg->status = ret;
+			spi_finalize_current_message(ctlr);
+			goto out;
+		}
+		ctlr->cur_msg_prepared = true;
+	}
+
+	ret = spi_map_msg(ctlr, ctlr->cur_msg);
+	if (ret) {
+		ctlr->cur_msg->status = ret;
+		spi_finalize_current_message(ctlr);
+		goto out;
+	}
+
+	ret = ctlr->transfer_one_message(ctlr, ctlr->cur_msg);
+	if (ret) {
+		dev_err(&ctlr->dev,
+			"failed to transfer one message from queue\n");
+		goto out;
+	}
+
+out:
+	mutex_unlock(&ctlr->io_mutex);
+
+	/* Prod the scheduler in case transfer_one() was busy waiting */
+	if (!ret)
+		cond_resched();
+}
+
+/**
+ * spi_pump_messages - kthread work function which processes spi message queue
+ * @work: pointer to kthread work struct contained in the controller struct
+ */
+static void spi_pump_messages(struct kthread_work *work)
+{
+	struct spi_controller *ctlr =
+		container_of(work, struct spi_controller, pump_messages);
+
+	__spi_pump_messages(ctlr, true);
+}
+
+static int spi_init_queue(struct spi_controller *ctlr)
+{
+	struct sched_param param = { .sched_priority = MAX_RT_PRIO - 1 };
+
+	ctlr->running = false;
+	ctlr->busy = false;
+
+	kthread_init_worker(&ctlr->kworker);
+	ctlr->kworker_task = kthread_run(kthread_worker_fn, &ctlr->kworker,
+					 "%s", dev_name(&ctlr->dev));
+	if (IS_ERR(ctlr->kworker_task)) {
+		dev_err(&ctlr->dev, "failed to create message pump task\n");
+		return PTR_ERR(ctlr->kworker_task);
+	}
+	kthread_init_work(&ctlr->pump_messages, spi_pump_messages);
+
+	/*
+	 * Controller config will indicate if this controller should run the
+	 * message pump with high (realtime) priority to reduce the transfer
+	 * latency on the bus by minimising the delay between a transfer
+	 * request and the scheduling of the message pump thread. Without this
+	 * setting the message pump thread will remain at default priority.
+	 */
+	if (ctlr->rt) {
+		dev_info(&ctlr->dev,
+			"will run message pump with realtime priority\n");
+		sched_setscheduler(ctlr->kworker_task, SCHED_FIFO, &param);
+	}
+
+	return 0;
+}
+
+/**
+ * spi_get_next_queued_message() - called by driver to check for queued
+ * messages
+ * @ctlr: the controller to check for queued messages
+ *
+ * If there are more messages in the queue, the next message is returned from
+ * this call.
+ *
+ * Return: the next message in the queue, else NULL if the queue is empty.
+ */
+struct spi_message *spi_get_next_queued_message(struct spi_controller *ctlr)
+{
+	struct spi_message *next;
+	unsigned long flags;
+
+	/* get a pointer to the next message, if any */
+	spin_lock_irqsave(&ctlr->queue_lock, flags);
+	next = list_first_entry_or_null(&ctlr->queue, struct spi_message,
+					queue);
+	spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+
+	return next;
+}
+EXPORT_SYMBOL_GPL(spi_get_next_queued_message);
+
+/**
+ * spi_finalize_current_message() - the current message is complete
+ * @ctlr: the controller to return the message to
+ *
+ * Called by the driver to notify the core that the message in the front of the
+ * queue is complete and can be removed from the queue.
+ */
+void spi_finalize_current_message(struct spi_controller *ctlr)
+{
+	struct spi_message *mesg;
+	unsigned long flags;
+	int ret;
+
+	spin_lock_irqsave(&ctlr->queue_lock, flags);
+	mesg = ctlr->cur_msg;
+	spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+
+	spi_unmap_msg(ctlr, mesg);
+
+	if (ctlr->cur_msg_prepared && ctlr->unprepare_message) {
+		ret = ctlr->unprepare_message(ctlr, mesg);
+		if (ret) {
+			dev_err(&ctlr->dev, "failed to unprepare message: %d\n",
+				ret);
+		}
+	}
+
+	spin_lock_irqsave(&ctlr->queue_lock, flags);
+	ctlr->cur_msg = NULL;
+	ctlr->cur_msg_prepared = false;
+	kthread_queue_work(&ctlr->kworker, &ctlr->pump_messages);
+	spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+
+	trace_spi_message_done(mesg);
+
+	mesg->state = NULL;
+	if (mesg->complete)
+		mesg->complete(mesg->context);
+}
+EXPORT_SYMBOL_GPL(spi_finalize_current_message);
+
+static int spi_start_queue(struct spi_controller *ctlr)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&ctlr->queue_lock, flags);
+
+	if (ctlr->running || ctlr->busy) {
+		spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+		return -EBUSY;
+	}
+
+	ctlr->running = true;
+	ctlr->cur_msg = NULL;
+	spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+
+	kthread_queue_work(&ctlr->kworker, &ctlr->pump_messages);
+
+	return 0;
+}
+
+static int spi_stop_queue(struct spi_controller *ctlr)
+{
+	unsigned long flags;
+	unsigned limit = 500;
+	int ret = 0;
+
+	spin_lock_irqsave(&ctlr->queue_lock, flags);
+
+	/*
+	 * This is a bit lame, but is optimized for the common execution path.
+	 * A wait_queue on the ctlr->busy could be used, but then the common
+	 * execution path (pump_messages) would be required to call wake_up or
+	 * friends on every SPI message. Do this instead.
+	 */
+	while ((!list_empty(&ctlr->queue) || ctlr->busy) && limit--) {
+		spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+		usleep_range(10000, 11000);
+		spin_lock_irqsave(&ctlr->queue_lock, flags);
+	}
+
+	if (!list_empty(&ctlr->queue) || ctlr->busy)
+		ret = -EBUSY;
+	else
+		ctlr->running = false;
+
+	spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+
+	if (ret) {
+		dev_warn(&ctlr->dev, "could not stop message queue\n");
+		return ret;
+	}
+	return ret;
+}
+
+static int spi_destroy_queue(struct spi_controller *ctlr)
+{
+	int ret;
+
+	ret = spi_stop_queue(ctlr);
+
+	/*
+	 * kthread_flush_worker will block until all work is done.
+	 * If the reason that stop_queue timed out is that the work will never
+	 * finish, then it does no good to call flush/stop thread, so
+	 * return anyway.
+	 */
+	if (ret) {
+		dev_err(&ctlr->dev, "problem destroying queue\n");
+		return ret;
+	}
+
+	kthread_flush_worker(&ctlr->kworker);
+	kthread_stop(ctlr->kworker_task);
+
+	return 0;
+}
+
+static int __spi_queued_transfer(struct spi_device *spi,
+				 struct spi_message *msg,
+				 bool need_pump)
+{
+	struct spi_controller *ctlr = spi->controller;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ctlr->queue_lock, flags);
+
+	if (!ctlr->running) {
+		spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+		return -ESHUTDOWN;
+	}
+	msg->actual_length = 0;
+	msg->status = -EINPROGRESS;
+
+	list_add_tail(&msg->queue, &ctlr->queue);
+	if (!ctlr->busy && need_pump)
+		kthread_queue_work(&ctlr->kworker, &ctlr->pump_messages);
+
+	spin_unlock_irqrestore(&ctlr->queue_lock, flags);
+	return 0;
+}
+
+/**
+ * spi_queued_transfer - transfer function for queued transfers
+ * @spi: spi device which is requesting transfer
+ * @msg: spi message which is to handled is queued to driver queue
+ *
+ * Return: zero on success, else a negative error code.
+ */
+static int spi_queued_transfer(struct spi_device *spi, struct spi_message *msg)
+{
+	return __spi_queued_transfer(spi, msg, true);
+}
+
+static int spi_controller_initialize_queue(struct spi_controller *ctlr)
+{
+	int ret;
+
+	ctlr->transfer = spi_queued_transfer;
+	if (!ctlr->transfer_one_message)
+		ctlr->transfer_one_message = spi_transfer_one_message;
+
+	/* Initialize and start queue */
+	ret = spi_init_queue(ctlr);
+	if (ret) {
+		dev_err(&ctlr->dev, "problem initializing queue\n");
+		goto err_init_queue;
+	}
+	ctlr->queued = true;
+	ret = spi_start_queue(ctlr);
+	if (ret) {
+		dev_err(&ctlr->dev, "problem starting queue\n");
+		goto err_start_queue;
+	}
+
+	return 0;
+
+err_start_queue:
+	spi_destroy_queue(ctlr);
+err_init_queue:
+	return ret;
+}
+
+/**
+ * spi_flush_queue - Send all pending messages in the queue from the callers'
+ *		     context
+ * @ctlr: controller to process queue for
+ *
+ * This should be used when one wants to ensure all pending messages have been
+ * sent before doing something. Is used by the spi-mem code to make sure SPI
+ * memory operations do not preempt regular SPI transfers that have been queued
+ * before the spi-mem operation.
+ */
+void spi_flush_queue(struct spi_controller *ctlr)
+{
+	if (ctlr->transfer == spi_queued_transfer)
+		__spi_pump_messages(ctlr, false);
+}
+
+/*-------------------------------------------------------------------------*/
+
+#if defined(CONFIG_OF)
+static int of_spi_parse_dt(struct spi_controller *ctlr, struct spi_device *spi,
+			   struct device_node *nc)
+{
+	u32 value;
+	int rc;
+
+	/* Mode (clock phase/polarity/etc.) */
+	if (of_property_read_bool(nc, "spi-cpha"))
+		spi->mode |= SPI_CPHA;
+	if (of_property_read_bool(nc, "spi-cpol"))
+		spi->mode |= SPI_CPOL;
+	if (of_property_read_bool(nc, "spi-cs-high"))
+		spi->mode |= SPI_CS_HIGH;
+	if (of_property_read_bool(nc, "spi-3wire"))
+		spi->mode |= SPI_3WIRE;
+	if (of_property_read_bool(nc, "spi-lsb-first"))
+		spi->mode |= SPI_LSB_FIRST;
+
+	/* Device DUAL/QUAD mode */
+	if (!of_property_read_u32(nc, "spi-tx-bus-width", &value)) {
+		switch (value) {
+		case 1:
+			break;
+		case 2:
+			spi->mode |= SPI_TX_DUAL;
+			break;
+		case 4:
+			spi->mode |= SPI_TX_QUAD;
+			break;
+		default:
+			dev_warn(&ctlr->dev,
+				"spi-tx-bus-width %d not supported\n",
+				value);
+			break;
+		}
+	}
+
+	if (!of_property_read_u32(nc, "spi-rx-bus-width", &value)) {
+		switch (value) {
+		case 1:
+			break;
+		case 2:
+			spi->mode |= SPI_RX_DUAL;
+			break;
+		case 4:
+			spi->mode |= SPI_RX_QUAD;
+			break;
+		default:
+			dev_warn(&ctlr->dev,
+				"spi-rx-bus-width %d not supported\n",
+				value);
+			break;
+		}
+	}
+
+	if (spi_controller_is_slave(ctlr)) {
+		if (strcmp(nc->name, "slave")) {
+			dev_err(&ctlr->dev, "%pOF is not called 'slave'\n",
+				nc);
+			return -EINVAL;
+		}
+		return 0;
+	}
+
+	/* Device address */
+	rc = of_property_read_u32(nc, "reg", &value);
+	if (rc) {
+		dev_err(&ctlr->dev, "%pOF has no valid 'reg' property (%d)\n",
+			nc, rc);
+		return rc;
+	}
+	spi->chip_select = value;
+
+	/* Device speed */
+	rc = of_property_read_u32(nc, "spi-max-frequency", &value);
+	if (rc) {
+		dev_err(&ctlr->dev,
+			"%pOF has no valid 'spi-max-frequency' property (%d)\n", nc, rc);
+		return rc;
+	}
+	spi->max_speed_hz = value;
+
+	return 0;
+}
+
+static struct spi_device *
+of_register_spi_device(struct spi_controller *ctlr, struct device_node *nc)
+{
+	struct spi_device *spi;
+	int rc;
+
+	/* Alloc an spi_device */
+	spi = spi_alloc_device(ctlr);
+	if (!spi) {
+		dev_err(&ctlr->dev, "spi_device alloc error for %pOF\n", nc);
+		rc = -ENOMEM;
+		goto err_out;
+	}
+
+	/* Select device driver */
+	rc = of_modalias_node(nc, spi->modalias,
+				sizeof(spi->modalias));
+	if (rc < 0) {
+		dev_err(&ctlr->dev, "cannot find modalias for %pOF\n", nc);
+		goto err_out;
+	}
+
+	rc = of_spi_parse_dt(ctlr, spi, nc);
+	if (rc)
+		goto err_out;
+
+	/* Store a pointer to the node in the device structure */
+	of_node_get(nc);
+	spi->dev.of_node = nc;
+
+	/* Register the new device */
+	rc = spi_add_device(spi);
+	if (rc) {
+		dev_err(&ctlr->dev, "spi_device register error %pOF\n", nc);
+		goto err_of_node_put;
+	}
+
+	return spi;
+
+err_of_node_put:
+	of_node_put(nc);
+err_out:
+	spi_dev_put(spi);
+	return ERR_PTR(rc);
+}
+
+/**
+ * of_register_spi_devices() - Register child devices onto the SPI bus
+ * @ctlr:	Pointer to spi_controller device
+ *
+ * Registers an spi_device for each child node of controller node which
+ * represents a valid SPI slave.
+ */
+static void of_register_spi_devices(struct spi_controller *ctlr)
+{
+	struct spi_device *spi;
+	struct device_node *nc;
+
+	if (!ctlr->dev.of_node)
+		return;
+
+	for_each_available_child_of_node(ctlr->dev.of_node, nc) {
+		if (of_node_test_and_set_flag(nc, OF_POPULATED))
+			continue;
+		spi = of_register_spi_device(ctlr, nc);
+		if (IS_ERR(spi)) {
+			dev_warn(&ctlr->dev,
+				 "Failed to create SPI device for %pOF\n", nc);
+			of_node_clear_flag(nc, OF_POPULATED);
+		}
+	}
+}
+#else
+static void of_register_spi_devices(struct spi_controller *ctlr) { }
+#endif
+
+#ifdef CONFIG_ACPI
+static void acpi_spi_parse_apple_properties(struct spi_device *spi)
+{
+	struct acpi_device *dev = ACPI_COMPANION(&spi->dev);
+	const union acpi_object *obj;
+
+	if (!x86_apple_machine)
+		return;
+
+	if (!acpi_dev_get_property(dev, "spiSclkPeriod", ACPI_TYPE_BUFFER, &obj)
+	    && obj->buffer.length >= 4)
+		spi->max_speed_hz  = NSEC_PER_SEC / *(u32 *)obj->buffer.pointer;
+
+	if (!acpi_dev_get_property(dev, "spiWordSize", ACPI_TYPE_BUFFER, &obj)
+	    && obj->buffer.length == 8)
+		spi->bits_per_word = *(u64 *)obj->buffer.pointer;
+
+	if (!acpi_dev_get_property(dev, "spiBitOrder", ACPI_TYPE_BUFFER, &obj)
+	    && obj->buffer.length == 8 && !*(u64 *)obj->buffer.pointer)
+		spi->mode |= SPI_LSB_FIRST;
+
+	if (!acpi_dev_get_property(dev, "spiSPO", ACPI_TYPE_BUFFER, &obj)
+	    && obj->buffer.length == 8 &&  *(u64 *)obj->buffer.pointer)
+		spi->mode |= SPI_CPOL;
+
+	if (!acpi_dev_get_property(dev, "spiSPH", ACPI_TYPE_BUFFER, &obj)
+	    && obj->buffer.length == 8 &&  *(u64 *)obj->buffer.pointer)
+		spi->mode |= SPI_CPHA;
+}
+
+static int acpi_spi_add_resource(struct acpi_resource *ares, void *data)
+{
+	struct spi_device *spi = data;
+	struct spi_controller *ctlr = spi->controller;
+
+	if (ares->type == ACPI_RESOURCE_TYPE_SERIAL_BUS) {
+		struct acpi_resource_spi_serialbus *sb;
+
+		sb = &ares->data.spi_serial_bus;
+		if (sb->type == ACPI_RESOURCE_SERIAL_TYPE_SPI) {
+			/*
+			 * ACPI DeviceSelection numbering is handled by the
+			 * host controller driver in Windows and can vary
+			 * from driver to driver. In Linux we always expect
+			 * 0 .. max - 1 so we need to ask the driver to
+			 * translate between the two schemes.
+			 */
+			if (ctlr->fw_translate_cs) {
+				int cs = ctlr->fw_translate_cs(ctlr,
+						sb->device_selection);
+				if (cs < 0)
+					return cs;
+				spi->chip_select = cs;
+			} else {
+				spi->chip_select = sb->device_selection;
+			}
+
+			spi->max_speed_hz = sb->connection_speed;
+
+			if (sb->clock_phase == ACPI_SPI_SECOND_PHASE)
+				spi->mode |= SPI_CPHA;
+			if (sb->clock_polarity == ACPI_SPI_START_HIGH)
+				spi->mode |= SPI_CPOL;
+			if (sb->device_polarity == ACPI_SPI_ACTIVE_HIGH)
+				spi->mode |= SPI_CS_HIGH;
+		}
+	} else if (spi->irq < 0) {
+		struct resource r;
+
+		if (acpi_dev_resource_interrupt(ares, 0, &r))
+			spi->irq = r.start;
+	}
+
+	/* Always tell the ACPI core to skip this resource */
+	return 1;
+}
+
+static acpi_status acpi_register_spi_device(struct spi_controller *ctlr,
+					    struct acpi_device *adev)
+{
+	struct list_head resource_list;
+	struct spi_device *spi;
+	int ret;
+
+	if (acpi_bus_get_status(adev) || !adev->status.present ||
+	    acpi_device_enumerated(adev))
+		return AE_OK;
+
+	spi = spi_alloc_device(ctlr);
+	if (!spi) {
+		dev_err(&ctlr->dev, "failed to allocate SPI device for %s\n",
+			dev_name(&adev->dev));
+		return AE_NO_MEMORY;
+	}
+
+	ACPI_COMPANION_SET(&spi->dev, adev);
+	spi->irq = -1;
+
+	INIT_LIST_HEAD(&resource_list);
+	ret = acpi_dev_get_resources(adev, &resource_list,
+				     acpi_spi_add_resource, spi);
+	acpi_dev_free_resource_list(&resource_list);
+
+	acpi_spi_parse_apple_properties(spi);
+
+	if (ret < 0 || !spi->max_speed_hz) {
+		spi_dev_put(spi);
+		return AE_OK;
+	}
+
+	acpi_set_modalias(adev, acpi_device_hid(adev), spi->modalias,
+			  sizeof(spi->modalias));
+
+	if (spi->irq < 0)
+		spi->irq = acpi_dev_gpio_irq_get(adev, 0);
+
+	acpi_device_set_enumerated(adev);
+
+	adev->power.flags.ignore_parent = true;
+	if (spi_add_device(spi)) {
+		adev->power.flags.ignore_parent = false;
+		dev_err(&ctlr->dev, "failed to add SPI device %s from ACPI\n",
+			dev_name(&adev->dev));
+		spi_dev_put(spi);
+	}
+
+	return AE_OK;
+}
+
+static acpi_status acpi_spi_add_device(acpi_handle handle, u32 level,
+				       void *data, void **return_value)
+{
+	struct spi_controller *ctlr = data;
+	struct acpi_device *adev;
+
+	if (acpi_bus_get_device(handle, &adev))
+		return AE_OK;
+
+	return acpi_register_spi_device(ctlr, adev);
+}
+
+static void acpi_register_spi_devices(struct spi_controller *ctlr)
+{
+	acpi_status status;
+	acpi_handle handle;
+
+	handle = ACPI_HANDLE(ctlr->dev.parent);
+	if (!handle)
+		return;
+
+	status = acpi_walk_namespace(ACPI_TYPE_DEVICE, handle, 1,
+				     acpi_spi_add_device, NULL, ctlr, NULL);
+	if (ACPI_FAILURE(status))
+		dev_warn(&ctlr->dev, "failed to enumerate SPI slaves\n");
+}
+#else
+static inline void acpi_register_spi_devices(struct spi_controller *ctlr) {}
+#endif /* CONFIG_ACPI */
+
+static void spi_controller_release(struct device *dev)
+{
+	struct spi_controller *ctlr;
+
+	ctlr = container_of(dev, struct spi_controller, dev);
+	kfree(ctlr);
+}
+
+static struct class spi_master_class = {
+	.name		= "spi_master",
+	.owner		= THIS_MODULE,
+	.dev_release	= spi_controller_release,
+	.dev_groups	= spi_master_groups,
+};
+
+#ifdef CONFIG_SPI_SLAVE
+/**
+ * spi_slave_abort - abort the ongoing transfer request on an SPI slave
+ *		     controller
+ * @spi: device used for the current transfer
+ */
+int spi_slave_abort(struct spi_device *spi)
+{
+	struct spi_controller *ctlr = spi->controller;
+
+	if (spi_controller_is_slave(ctlr) && ctlr->slave_abort)
+		return ctlr->slave_abort(ctlr);
+
+	return -ENOTSUPP;
+}
+EXPORT_SYMBOL_GPL(spi_slave_abort);
+
+static int match_true(struct device *dev, void *data)
+{
+	return 1;
+}
+
+static ssize_t spi_slave_show(struct device *dev,
+			      struct device_attribute *attr, char *buf)
+{
+	struct spi_controller *ctlr = container_of(dev, struct spi_controller,
+						   dev);
+	struct device *child;
+
+	child = device_find_child(&ctlr->dev, NULL, match_true);
+	return sprintf(buf, "%s\n",
+		       child ? to_spi_device(child)->modalias : NULL);
+}
+
+static ssize_t spi_slave_store(struct device *dev,
+			       struct device_attribute *attr, const char *buf,
+			       size_t count)
+{
+	struct spi_controller *ctlr = container_of(dev, struct spi_controller,
+						   dev);
+	struct spi_device *spi;
+	struct device *child;
+	char name[32];
+	int rc;
+
+	rc = sscanf(buf, "%31s", name);
+	if (rc != 1 || !name[0])
+		return -EINVAL;
+
+	child = device_find_child(&ctlr->dev, NULL, match_true);
+	if (child) {
+		/* Remove registered slave */
+		device_unregister(child);
+		put_device(child);
+	}
+
+	if (strcmp(name, "(null)")) {
+		/* Register new slave */
+		spi = spi_alloc_device(ctlr);
+		if (!spi)
+			return -ENOMEM;
+
+		strlcpy(spi->modalias, name, sizeof(spi->modalias));
+
+		rc = spi_add_device(spi);
+		if (rc) {
+			spi_dev_put(spi);
+			return rc;
+		}
+	}
+
+	return count;
+}
+
+static DEVICE_ATTR(slave, 0644, spi_slave_show, spi_slave_store);
+
+static struct attribute *spi_slave_attrs[] = {
+	&dev_attr_slave.attr,
+	NULL,
+};
+
+static const struct attribute_group spi_slave_group = {
+	.attrs = spi_slave_attrs,
+};
+
+static const struct attribute_group *spi_slave_groups[] = {
+	&spi_controller_statistics_group,
+	&spi_slave_group,
+	NULL,
+};
+
+static struct class spi_slave_class = {
+	.name		= "spi_slave",
+	.owner		= THIS_MODULE,
+	.dev_release	= spi_controller_release,
+	.dev_groups	= spi_slave_groups,
+};
+#else
+extern struct class spi_slave_class;	/* dummy */
+#endif
+
+/**
+ * __spi_alloc_controller - allocate an SPI master or slave controller
+ * @dev: the controller, possibly using the platform_bus
+ * @size: how much zeroed driver-private data to allocate; the pointer to this
+ *	memory is in the driver_data field of the returned device,
+ *	accessible with spi_controller_get_devdata().
+ * @slave: flag indicating whether to allocate an SPI master (false) or SPI
+ *	slave (true) controller
+ * Context: can sleep
+ *
+ * This call is used only by SPI controller drivers, which are the
+ * only ones directly touching chip registers.  It's how they allocate
+ * an spi_controller structure, prior to calling spi_register_controller().
+ *
+ * This must be called from context that can sleep.
+ *
+ * The caller is responsible for assigning the bus number and initializing the
+ * controller's methods before calling spi_register_controller(); and (after
+ * errors adding the device) calling spi_controller_put() to prevent a memory
+ * leak.
+ *
+ * Return: the SPI controller structure on success, else NULL.
+ */
+struct spi_controller *__spi_alloc_controller(struct device *dev,
+					      unsigned int size, bool slave)
+{
+	struct spi_controller	*ctlr;
+
+	if (!dev)
+		return NULL;
+
+	ctlr = kzalloc(size + sizeof(*ctlr), GFP_KERNEL);
+	if (!ctlr)
+		return NULL;
+
+	device_initialize(&ctlr->dev);
+	ctlr->bus_num = -1;
+	ctlr->num_chipselect = 1;
+	ctlr->slave = slave;
+	if (IS_ENABLED(CONFIG_SPI_SLAVE) && slave)
+		ctlr->dev.class = &spi_slave_class;
+	else
+		ctlr->dev.class = &spi_master_class;
+	ctlr->dev.parent = dev;
+	pm_suspend_ignore_children(&ctlr->dev, true);
+	spi_controller_set_devdata(ctlr, &ctlr[1]);
+
+	return ctlr;
+}
+EXPORT_SYMBOL_GPL(__spi_alloc_controller);
+
+#ifdef CONFIG_OF
+static int of_spi_register_master(struct spi_controller *ctlr)
+{
+	int nb, i, *cs;
+	struct device_node *np = ctlr->dev.of_node;
+
+	if (!np)
+		return 0;
+
+	nb = of_gpio_named_count(np, "cs-gpios");
+	ctlr->num_chipselect = max_t(int, nb, ctlr->num_chipselect);
+
+	/* Return error only for an incorrectly formed cs-gpios property */
+	if (nb == 0 || nb == -ENOENT)
+		return 0;
+	else if (nb < 0)
+		return nb;
+
+	cs = devm_kcalloc(&ctlr->dev, ctlr->num_chipselect, sizeof(int),
+			  GFP_KERNEL);
+	ctlr->cs_gpios = cs;
+
+	if (!ctlr->cs_gpios)
+		return -ENOMEM;
+
+	for (i = 0; i < ctlr->num_chipselect; i++)
+		cs[i] = -ENOENT;
+
+	for (i = 0; i < nb; i++)
+		cs[i] = of_get_named_gpio(np, "cs-gpios", i);
+
+	return 0;
+}
+#else
+static int of_spi_register_master(struct spi_controller *ctlr)
+{
+	return 0;
+}
+#endif
+
+static int spi_controller_check_ops(struct spi_controller *ctlr)
+{
+	/*
+	 * The controller may implement only the high-level SPI-memory like
+	 * operations if it does not support regular SPI transfers, and this is
+	 * valid use case.
+	 * If ->mem_ops is NULL, we request that at least one of the
+	 * ->transfer_xxx() method be implemented.
+	 */
+	if (ctlr->mem_ops) {
+		if (!ctlr->mem_ops->exec_op)
+			return -EINVAL;
+	} else if (!ctlr->transfer && !ctlr->transfer_one &&
+		   !ctlr->transfer_one_message) {
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+/**
+ * spi_register_controller - register SPI master or slave controller
+ * @ctlr: initialized master, originally from spi_alloc_master() or
+ *	spi_alloc_slave()
+ * Context: can sleep
+ *
+ * SPI controllers connect to their drivers using some non-SPI bus,
+ * such as the platform bus.  The final stage of probe() in that code
+ * includes calling spi_register_controller() to hook up to this SPI bus glue.
+ *
+ * SPI controllers use board specific (often SOC specific) bus numbers,
+ * and board-specific addressing for SPI devices combines those numbers
+ * with chip select numbers.  Since SPI does not directly support dynamic
+ * device identification, boards need configuration tables telling which
+ * chip is at which address.
+ *
+ * This must be called from context that can sleep.  It returns zero on
+ * success, else a negative error code (dropping the controller's refcount).
+ * After a successful return, the caller is responsible for calling
+ * spi_unregister_controller().
+ *
+ * Return: zero on success, else a negative error code.
+ */
+int spi_register_controller(struct spi_controller *ctlr)
+{
+	struct device		*dev = ctlr->dev.parent;
+	struct boardinfo	*bi;
+	int			status = -ENODEV;
+	int			id, first_dynamic;
+
+	if (!dev)
+		return -ENODEV;
+
+	/*
+	 * Make sure all necessary hooks are implemented before registering
+	 * the SPI controller.
+	 */
+	status = spi_controller_check_ops(ctlr);
+	if (status)
+		return status;
+
+	if (!spi_controller_is_slave(ctlr)) {
+		status = of_spi_register_master(ctlr);
+		if (status)
+			return status;
+	}
+
+	/* even if it's just one always-selected device, there must
+	 * be at least one chipselect
+	 */
+	if (ctlr->num_chipselect == 0)
+		return -EINVAL;
+	if (ctlr->bus_num >= 0) {
+		/* devices with a fixed bus num must check-in with the num */
+		mutex_lock(&board_lock);
+		id = idr_alloc(&spi_master_idr, ctlr, ctlr->bus_num,
+			ctlr->bus_num + 1, GFP_KERNEL);
+		mutex_unlock(&board_lock);
+		if (WARN(id < 0, "couldn't get idr"))
+			return id == -ENOSPC ? -EBUSY : id;
+		ctlr->bus_num = id;
+	} else if (ctlr->dev.of_node) {
+		/* allocate dynamic bus number using Linux idr */
+		id = of_alias_get_id(ctlr->dev.of_node, "spi");
+		if (id >= 0) {
+			ctlr->bus_num = id;
+			mutex_lock(&board_lock);
+			id = idr_alloc(&spi_master_idr, ctlr, ctlr->bus_num,
+				       ctlr->bus_num + 1, GFP_KERNEL);
+			mutex_unlock(&board_lock);
+			if (WARN(id < 0, "couldn't get idr"))
+				return id == -ENOSPC ? -EBUSY : id;
+		}
+	}
+	if (ctlr->bus_num < 0) {
+		first_dynamic = of_alias_get_highest_id("spi");
+		if (first_dynamic < 0)
+			first_dynamic = 0;
+		else
+			first_dynamic++;
+
+		mutex_lock(&board_lock);
+		id = idr_alloc(&spi_master_idr, ctlr, first_dynamic,
+			       0, GFP_KERNEL);
+		mutex_unlock(&board_lock);
+		if (WARN(id < 0, "couldn't get idr"))
+			return id;
+		ctlr->bus_num = id;
+	}
+	INIT_LIST_HEAD(&ctlr->queue);
+	spin_lock_init(&ctlr->queue_lock);
+	spin_lock_init(&ctlr->bus_lock_spinlock);
+	mutex_init(&ctlr->bus_lock_mutex);
+	mutex_init(&ctlr->io_mutex);
+	ctlr->bus_lock_flag = 0;
+	init_completion(&ctlr->xfer_completion);
+	if (!ctlr->max_dma_len)
+		ctlr->max_dma_len = INT_MAX;
+
+	/* register the device, then userspace will see it.
+	 * registration fails if the bus ID is in use.
+	 */
+	dev_set_name(&ctlr->dev, "spi%u", ctlr->bus_num);
+	status = device_add(&ctlr->dev);
+	if (status < 0) {
+		/* free bus id */
+		mutex_lock(&board_lock);
+		idr_remove(&spi_master_idr, ctlr->bus_num);
+		mutex_unlock(&board_lock);
+		goto done;
+	}
+	dev_dbg(dev, "registered %s %s\n",
+			spi_controller_is_slave(ctlr) ? "slave" : "master",
+			dev_name(&ctlr->dev));
+
+	/*
+	 * If we're using a queued driver, start the queue. Note that we don't
+	 * need the queueing logic if the driver is only supporting high-level
+	 * memory operations.
+	 */
+	if (ctlr->transfer) {
+		dev_info(dev, "controller is unqueued, this is deprecated\n");
+	} else if (ctlr->transfer_one || ctlr->transfer_one_message) {
+		status = spi_controller_initialize_queue(ctlr);
+		if (status) {
+			device_del(&ctlr->dev);
+			/* free bus id */
+			mutex_lock(&board_lock);
+			idr_remove(&spi_master_idr, ctlr->bus_num);
+			mutex_unlock(&board_lock);
+			goto done;
+		}
+	}
+	/* add statistics */
+	spin_lock_init(&ctlr->statistics.lock);
+
+	mutex_lock(&board_lock);
+	list_add_tail(&ctlr->list, &spi_controller_list);
+	list_for_each_entry(bi, &board_list, list)
+		spi_match_controller_to_boardinfo(ctlr, &bi->board_info);
+	mutex_unlock(&board_lock);
+
+	/* Register devices from the device tree and ACPI */
+	of_register_spi_devices(ctlr);
+	acpi_register_spi_devices(ctlr);
+done:
+	return status;
+}
+EXPORT_SYMBOL_GPL(spi_register_controller);
+
+static void devm_spi_unregister(struct device *dev, void *res)
+{
+	spi_unregister_controller(*(struct spi_controller **)res);
+}
+
+/**
+ * devm_spi_register_controller - register managed SPI master or slave
+ *	controller
+ * @dev:    device managing SPI controller
+ * @ctlr: initialized controller, originally from spi_alloc_master() or
+ *	spi_alloc_slave()
+ * Context: can sleep
+ *
+ * Register a SPI device as with spi_register_controller() which will
+ * automatically be unregistered and freed.
+ *
+ * Return: zero on success, else a negative error code.
+ */
+int devm_spi_register_controller(struct device *dev,
+				 struct spi_controller *ctlr)
+{
+	struct spi_controller **ptr;
+	int ret;
+
+	ptr = devres_alloc(devm_spi_unregister, sizeof(*ptr), GFP_KERNEL);
+	if (!ptr)
+		return -ENOMEM;
+
+	ret = spi_register_controller(ctlr);
+	if (!ret) {
+		*ptr = ctlr;
+		devres_add(dev, ptr);
+	} else {
+		devres_free(ptr);
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(devm_spi_register_controller);
+
+static int __unregister(struct device *dev, void *null)
+{
+	spi_unregister_device(to_spi_device(dev));
+	return 0;
+}
+
+/**
+ * spi_unregister_controller - unregister SPI master or slave controller
+ * @ctlr: the controller being unregistered
+ * Context: can sleep
+ *
+ * This call is used only by SPI controller drivers, which are the
+ * only ones directly touching chip registers.
+ *
+ * This must be called from context that can sleep.
+ *
+ * Note that this function also drops a reference to the controller.
+ */
+void spi_unregister_controller(struct spi_controller *ctlr)
+{
+	struct spi_controller *found;
+	int id = ctlr->bus_num;
+	int dummy;
+
+	/* First make sure that this controller was ever added */
+	mutex_lock(&board_lock);
+	found = idr_find(&spi_master_idr, id);
+	mutex_unlock(&board_lock);
+	if (ctlr->queued) {
+		if (spi_destroy_queue(ctlr))
+			dev_err(&ctlr->dev, "queue remove failed\n");
+	}
+	mutex_lock(&board_lock);
+	list_del(&ctlr->list);
+	mutex_unlock(&board_lock);
+
+	dummy = device_for_each_child(&ctlr->dev, NULL, __unregister);
+	device_unregister(&ctlr->dev);
+	/* free bus id */
+	mutex_lock(&board_lock);
+	if (found == ctlr)
+		idr_remove(&spi_master_idr, id);
+	mutex_unlock(&board_lock);
+}
+EXPORT_SYMBOL_GPL(spi_unregister_controller);
+
+int spi_controller_suspend(struct spi_controller *ctlr)
+{
+	int ret;
+
+	/* Basically no-ops for non-queued controllers */
+	if (!ctlr->queued)
+		return 0;
+
+	ret = spi_stop_queue(ctlr);
+	if (ret)
+		dev_err(&ctlr->dev, "queue stop failed\n");
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(spi_controller_suspend);
+
+int spi_controller_resume(struct spi_controller *ctlr)
+{
+	int ret;
+
+	if (!ctlr->queued)
+		return 0;
+
+	ret = spi_start_queue(ctlr);
+	if (ret)
+		dev_err(&ctlr->dev, "queue restart failed\n");
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(spi_controller_resume);
+
+static int __spi_controller_match(struct device *dev, const void *data)
+{
+	struct spi_controller *ctlr;
+	const u16 *bus_num = data;
+
+	ctlr = container_of(dev, struct spi_controller, dev);
+	return ctlr->bus_num == *bus_num;
+}
+
+/**
+ * spi_busnum_to_master - look up master associated with bus_num
+ * @bus_num: the master's bus number
+ * Context: can sleep
+ *
+ * This call may be used with devices that are registered after
+ * arch init time.  It returns a refcounted pointer to the relevant
+ * spi_controller (which the caller must release), or NULL if there is
+ * no such master registered.
+ *
+ * Return: the SPI master structure on success, else NULL.
+ */
+struct spi_controller *spi_busnum_to_master(u16 bus_num)
+{
+	struct device		*dev;
+	struct spi_controller	*ctlr = NULL;
+
+	dev = class_find_device(&spi_master_class, NULL, &bus_num,
+				__spi_controller_match);
+	if (dev)
+		ctlr = container_of(dev, struct spi_controller, dev);
+	/* reference got in class_find_device */
+	return ctlr;
+}
+EXPORT_SYMBOL_GPL(spi_busnum_to_master);
+
+/*-------------------------------------------------------------------------*/
+
+/* Core methods for SPI resource management */
+
+/**
+ * spi_res_alloc - allocate a spi resource that is life-cycle managed
+ *                 during the processing of a spi_message while using
+ *                 spi_transfer_one
+ * @spi:     the spi device for which we allocate memory
+ * @release: the release code to execute for this resource
+ * @size:    size to alloc and return
+ * @gfp:     GFP allocation flags
+ *
+ * Return: the pointer to the allocated data
+ *
+ * This may get enhanced in the future to allocate from a memory pool
+ * of the @spi_device or @spi_controller to avoid repeated allocations.
+ */
+void *spi_res_alloc(struct spi_device *spi,
+		    spi_res_release_t release,
+		    size_t size, gfp_t gfp)
+{
+	struct spi_res *sres;
+
+	sres = kzalloc(sizeof(*sres) + size, gfp);
+	if (!sres)
+		return NULL;
+
+	INIT_LIST_HEAD(&sres->entry);
+	sres->release = release;
+
+	return sres->data;
+}
+EXPORT_SYMBOL_GPL(spi_res_alloc);
+
+/**
+ * spi_res_free - free an spi resource
+ * @res: pointer to the custom data of a resource
+ *
+ */
+void spi_res_free(void *res)
+{
+	struct spi_res *sres = container_of(res, struct spi_res, data);
+
+	if (!res)
+		return;
+
+	WARN_ON(!list_empty(&sres->entry));
+	kfree(sres);
+}
+EXPORT_SYMBOL_GPL(spi_res_free);
+
+/**
+ * spi_res_add - add a spi_res to the spi_message
+ * @message: the spi message
+ * @res:     the spi_resource
+ */
+void spi_res_add(struct spi_message *message, void *res)
+{
+	struct spi_res *sres = container_of(res, struct spi_res, data);
+
+	WARN_ON(!list_empty(&sres->entry));
+	list_add_tail(&sres->entry, &message->resources);
+}
+EXPORT_SYMBOL_GPL(spi_res_add);
+
+/**
+ * spi_res_release - release all spi resources for this message
+ * @ctlr:  the @spi_controller
+ * @message: the @spi_message
+ */
+void spi_res_release(struct spi_controller *ctlr, struct spi_message *message)
+{
+	struct spi_res *res;
+
+	while (!list_empty(&message->resources)) {
+		res = list_last_entry(&message->resources,
+				      struct spi_res, entry);
+
+		if (res->release)
+			res->release(ctlr, message, res->data);
+
+		list_del(&res->entry);
+
+		kfree(res);
+	}
+}
+EXPORT_SYMBOL_GPL(spi_res_release);
+
+/*-------------------------------------------------------------------------*/
+
+/* Core methods for spi_message alterations */
+
+static void __spi_replace_transfers_release(struct spi_controller *ctlr,
+					    struct spi_message *msg,
+					    void *res)
+{
+	struct spi_replaced_transfers *rxfer = res;
+	size_t i;
+
+	/* call extra callback if requested */
+	if (rxfer->release)
+		rxfer->release(ctlr, msg, res);
+
+	/* insert replaced transfers back into the message */
+	list_splice(&rxfer->replaced_transfers, rxfer->replaced_after);
+
+	/* remove the formerly inserted entries */
+	for (i = 0; i < rxfer->inserted; i++)
+		list_del(&rxfer->inserted_transfers[i].transfer_list);
+}
+
+/**
+ * spi_replace_transfers - replace transfers with several transfers
+ *                         and register change with spi_message.resources
+ * @msg:           the spi_message we work upon
+ * @xfer_first:    the first spi_transfer we want to replace
+ * @remove:        number of transfers to remove
+ * @insert:        the number of transfers we want to insert instead
+ * @release:       extra release code necessary in some circumstances
+ * @extradatasize: extra data to allocate (with alignment guarantees
+ *                 of struct @spi_transfer)
+ * @gfp:           gfp flags
+ *
+ * Returns: pointer to @spi_replaced_transfers,
+ *          PTR_ERR(...) in case of errors.
+ */
+struct spi_replaced_transfers *spi_replace_transfers(
+	struct spi_message *msg,
+	struct spi_transfer *xfer_first,
+	size_t remove,
+	size_t insert,
+	spi_replaced_release_t release,
+	size_t extradatasize,
+	gfp_t gfp)
+{
+	struct spi_replaced_transfers *rxfer;
+	struct spi_transfer *xfer;
+	size_t i;
+
+	/* allocate the structure using spi_res */
+	rxfer = spi_res_alloc(msg->spi, __spi_replace_transfers_release,
+			      insert * sizeof(struct spi_transfer)
+			      + sizeof(struct spi_replaced_transfers)
+			      + extradatasize,
+			      gfp);
+	if (!rxfer)
+		return ERR_PTR(-ENOMEM);
+
+	/* the release code to invoke before running the generic release */
+	rxfer->release = release;
+
+	/* assign extradata */
+	if (extradatasize)
+		rxfer->extradata =
+			&rxfer->inserted_transfers[insert];
+
+	/* init the replaced_transfers list */
+	INIT_LIST_HEAD(&rxfer->replaced_transfers);
+
+	/* assign the list_entry after which we should reinsert
+	 * the @replaced_transfers - it may be spi_message.messages!
+	 */
+	rxfer->replaced_after = xfer_first->transfer_list.prev;
+
+	/* remove the requested number of transfers */
+	for (i = 0; i < remove; i++) {
+		/* if the entry after replaced_after it is msg->transfers
+		 * then we have been requested to remove more transfers
+		 * than are in the list
+		 */
+		if (rxfer->replaced_after->next == &msg->transfers) {
+			dev_err(&msg->spi->dev,
+				"requested to remove more spi_transfers than are available\n");
+			/* insert replaced transfers back into the message */
+			list_splice(&rxfer->replaced_transfers,
+				    rxfer->replaced_after);
+
+			/* free the spi_replace_transfer structure */
+			spi_res_free(rxfer);
+
+			/* and return with an error */
+			return ERR_PTR(-EINVAL);
+		}
+
+		/* remove the entry after replaced_after from list of
+		 * transfers and add it to list of replaced_transfers
+		 */
+		list_move_tail(rxfer->replaced_after->next,
+			       &rxfer->replaced_transfers);
+	}
+
+	/* create copy of the given xfer with identical settings
+	 * based on the first transfer to get removed
+	 */
+	for (i = 0; i < insert; i++) {
+		/* we need to run in reverse order */
+		xfer = &rxfer->inserted_transfers[insert - 1 - i];
+
+		/* copy all spi_transfer data */
+		memcpy(xfer, xfer_first, sizeof(*xfer));
+
+		/* add to list */
+		list_add(&xfer->transfer_list, rxfer->replaced_after);
+
+		/* clear cs_change and delay_usecs for all but the last */
+		if (i) {
+			xfer->cs_change = false;
+			xfer->delay_usecs = 0;
+		}
+	}
+
+	/* set up inserted */
+	rxfer->inserted = insert;
+
+	/* and register it with spi_res/spi_message */
+	spi_res_add(msg, rxfer);
+
+	return rxfer;
+}
+EXPORT_SYMBOL_GPL(spi_replace_transfers);
+
+static int __spi_split_transfer_maxsize(struct spi_controller *ctlr,
+					struct spi_message *msg,
+					struct spi_transfer **xferp,
+					size_t maxsize,
+					gfp_t gfp)
+{
+	struct spi_transfer *xfer = *xferp, *xfers;
+	struct spi_replaced_transfers *srt;
+	size_t offset;
+	size_t count, i;
+
+	/* warn once about this fact that we are splitting a transfer */
+	dev_warn_once(&msg->spi->dev,
+		      "spi_transfer of length %i exceed max length of %zu - needed to split transfers\n",
+		      xfer->len, maxsize);
+
+	/* calculate how many we have to replace */
+	count = DIV_ROUND_UP(xfer->len, maxsize);
+
+	/* create replacement */
+	srt = spi_replace_transfers(msg, xfer, 1, count, NULL, 0, gfp);
+	if (IS_ERR(srt))
+		return PTR_ERR(srt);
+	xfers = srt->inserted_transfers;
+
+	/* now handle each of those newly inserted spi_transfers
+	 * note that the replacements spi_transfers all are preset
+	 * to the same values as *xferp, so tx_buf, rx_buf and len
+	 * are all identical (as well as most others)
+	 * so we just have to fix up len and the pointers.
+	 *
+	 * this also includes support for the depreciated
+	 * spi_message.is_dma_mapped interface
+	 */
+
+	/* the first transfer just needs the length modified, so we
+	 * run it outside the loop
+	 */
+	xfers[0].len = min_t(size_t, maxsize, xfer[0].len);
+
+	/* all the others need rx_buf/tx_buf also set */
+	for (i = 1, offset = maxsize; i < count; offset += maxsize, i++) {
+		/* update rx_buf, tx_buf and dma */
+		if (xfers[i].rx_buf)
+			xfers[i].rx_buf += offset;
+		if (xfers[i].rx_dma)
+			xfers[i].rx_dma += offset;
+		if (xfers[i].tx_buf)
+			xfers[i].tx_buf += offset;
+		if (xfers[i].tx_dma)
+			xfers[i].tx_dma += offset;
+
+		/* update length */
+		xfers[i].len = min(maxsize, xfers[i].len - offset);
+	}
+
+	/* we set up xferp to the last entry we have inserted,
+	 * so that we skip those already split transfers
+	 */
+	*xferp = &xfers[count - 1];
+
+	/* increment statistics counters */
+	SPI_STATISTICS_INCREMENT_FIELD(&ctlr->statistics,
+				       transfers_split_maxsize);
+	SPI_STATISTICS_INCREMENT_FIELD(&msg->spi->statistics,
+				       transfers_split_maxsize);
+
+	return 0;
+}
+
+/**
+ * spi_split_tranfers_maxsize - split spi transfers into multiple transfers
+ *                              when an individual transfer exceeds a
+ *                              certain size
+ * @ctlr:    the @spi_controller for this transfer
+ * @msg:   the @spi_message to transform
+ * @maxsize:  the maximum when to apply this
+ * @gfp: GFP allocation flags
+ *
+ * Return: status of transformation
+ */
+int spi_split_transfers_maxsize(struct spi_controller *ctlr,
+				struct spi_message *msg,
+				size_t maxsize,
+				gfp_t gfp)
+{
+	struct spi_transfer *xfer;
+	int ret;
+
+	/* iterate over the transfer_list,
+	 * but note that xfer is advanced to the last transfer inserted
+	 * to avoid checking sizes again unnecessarily (also xfer does
+	 * potentiall belong to a different list by the time the
+	 * replacement has happened
+	 */
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		if (xfer->len > maxsize) {
+			ret = __spi_split_transfer_maxsize(ctlr, msg, &xfer,
+							   maxsize, gfp);
+			if (ret)
+				return ret;
+		}
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(spi_split_transfers_maxsize);
+
+/*-------------------------------------------------------------------------*/
+
+/* Core methods for SPI controller protocol drivers.  Some of the
+ * other core methods are currently defined as inline functions.
+ */
+
+static int __spi_validate_bits_per_word(struct spi_controller *ctlr,
+					u8 bits_per_word)
+{
+	if (ctlr->bits_per_word_mask) {
+		/* Only 32 bits fit in the mask */
+		if (bits_per_word > 32)
+			return -EINVAL;
+		if (!(ctlr->bits_per_word_mask & SPI_BPW_MASK(bits_per_word)))
+			return -EINVAL;
+	}
+
+	return 0;
+}
+
+/**
+ * spi_setup - setup SPI mode and clock rate
+ * @spi: the device whose settings are being modified
+ * Context: can sleep, and no requests are queued to the device
+ *
+ * SPI protocol drivers may need to update the transfer mode if the
+ * device doesn't work with its default.  They may likewise need
+ * to update clock rates or word sizes from initial values.  This function
+ * changes those settings, and must be called from a context that can sleep.
+ * Except for SPI_CS_HIGH, which takes effect immediately, the changes take
+ * effect the next time the device is selected and data is transferred to
+ * or from it.  When this function returns, the spi device is deselected.
+ *
+ * Note that this call will fail if the protocol driver specifies an option
+ * that the underlying controller or its driver does not support.  For
+ * example, not all hardware supports wire transfers using nine bit words,
+ * LSB-first wire encoding, or active-high chipselects.
+ *
+ * Return: zero on success, else a negative error code.
+ */
+int spi_setup(struct spi_device *spi)
+{
+	unsigned	bad_bits, ugly_bits;
+	int		status;
+
+	/* check mode to prevent that DUAL and QUAD set at the same time
+	 */
+	if (((spi->mode & SPI_TX_DUAL) && (spi->mode & SPI_TX_QUAD)) ||
+		((spi->mode & SPI_RX_DUAL) && (spi->mode & SPI_RX_QUAD))) {
+		dev_err(&spi->dev,
+		"setup: can not select dual and quad at the same time\n");
+		return -EINVAL;
+	}
+	/* if it is SPI_3WIRE mode, DUAL and QUAD should be forbidden
+	 */
+	if ((spi->mode & SPI_3WIRE) && (spi->mode &
+		(SPI_TX_DUAL | SPI_TX_QUAD | SPI_RX_DUAL | SPI_RX_QUAD)))
+		return -EINVAL;
+	/* help drivers fail *cleanly* when they need options
+	 * that aren't supported with their current controller
+	 */
+	bad_bits = spi->mode & ~spi->controller->mode_bits;
+	ugly_bits = bad_bits &
+		    (SPI_TX_DUAL | SPI_TX_QUAD | SPI_RX_DUAL | SPI_RX_QUAD);
+	if (ugly_bits) {
+		dev_warn(&spi->dev,
+			 "setup: ignoring unsupported mode bits %x\n",
+			 ugly_bits);
+		spi->mode &= ~ugly_bits;
+		bad_bits &= ~ugly_bits;
+	}
+	if (bad_bits) {
+		dev_err(&spi->dev, "setup: unsupported mode bits %x\n",
+			bad_bits);
+		return -EINVAL;
+	}
+
+	if (!spi->bits_per_word)
+		spi->bits_per_word = 8;
+
+	status = __spi_validate_bits_per_word(spi->controller,
+					      spi->bits_per_word);
+	if (status)
+		return status;
+
+	if (!spi->max_speed_hz)
+		spi->max_speed_hz = spi->controller->max_speed_hz;
+
+	if (spi->controller->setup)
+		status = spi->controller->setup(spi);
+
+	spi_set_cs(spi, false);
+
+	dev_dbg(&spi->dev, "setup mode %d, %s%s%s%s%u bits/w, %u Hz max --> %d\n",
+			(int) (spi->mode & (SPI_CPOL | SPI_CPHA)),
+			(spi->mode & SPI_CS_HIGH) ? "cs_high, " : "",
+			(spi->mode & SPI_LSB_FIRST) ? "lsb, " : "",
+			(spi->mode & SPI_3WIRE) ? "3wire, " : "",
+			(spi->mode & SPI_LOOP) ? "loopback, " : "",
+			spi->bits_per_word, spi->max_speed_hz,
+			status);
+
+	return status;
+}
+EXPORT_SYMBOL_GPL(spi_setup);
+
+static int __spi_validate(struct spi_device *spi, struct spi_message *message)
+{
+	struct spi_controller *ctlr = spi->controller;
+	struct spi_transfer *xfer;
+	int w_size;
+
+	if (list_empty(&message->transfers))
+		return -EINVAL;
+
+	/* Half-duplex links include original MicroWire, and ones with
+	 * only one data pin like SPI_3WIRE (switches direction) or where
+	 * either MOSI or MISO is missing.  They can also be caused by
+	 * software limitations.
+	 */
+	if ((ctlr->flags & SPI_CONTROLLER_HALF_DUPLEX) ||
+	    (spi->mode & SPI_3WIRE)) {
+		unsigned flags = ctlr->flags;
+
+		list_for_each_entry(xfer, &message->transfers, transfer_list) {
+			if (xfer->rx_buf && xfer->tx_buf)
+				return -EINVAL;
+			if ((flags & SPI_CONTROLLER_NO_TX) && xfer->tx_buf)
+				return -EINVAL;
+			if ((flags & SPI_CONTROLLER_NO_RX) && xfer->rx_buf)
+				return -EINVAL;
+		}
+	}
+
+	/**
+	 * Set transfer bits_per_word and max speed as spi device default if
+	 * it is not set for this transfer.
+	 * Set transfer tx_nbits and rx_nbits as single transfer default
+	 * (SPI_NBITS_SINGLE) if it is not set for this transfer.
+	 */
+	message->frame_length = 0;
+	list_for_each_entry(xfer, &message->transfers, transfer_list) {
+		message->frame_length += xfer->len;
+		if (!xfer->bits_per_word)
+			xfer->bits_per_word = spi->bits_per_word;
+
+		if (!xfer->speed_hz)
+			xfer->speed_hz = spi->max_speed_hz;
+		if (!xfer->speed_hz)
+			xfer->speed_hz = ctlr->max_speed_hz;
+
+		if (ctlr->max_speed_hz && xfer->speed_hz > ctlr->max_speed_hz)
+			xfer->speed_hz = ctlr->max_speed_hz;
+
+		if (__spi_validate_bits_per_word(ctlr, xfer->bits_per_word))
+			return -EINVAL;
+
+		/*
+		 * SPI transfer length should be multiple of SPI word size
+		 * where SPI word size should be power-of-two multiple
+		 */
+		if (xfer->bits_per_word <= 8)
+			w_size = 1;
+		else if (xfer->bits_per_word <= 16)
+			w_size = 2;
+		else
+			w_size = 4;
+
+		/* No partial transfers accepted */
+		if (xfer->len % w_size)
+			return -EINVAL;
+
+		if (xfer->speed_hz && ctlr->min_speed_hz &&
+		    xfer->speed_hz < ctlr->min_speed_hz)
+			return -EINVAL;
+
+		if (xfer->tx_buf && !xfer->tx_nbits)
+			xfer->tx_nbits = SPI_NBITS_SINGLE;
+		if (xfer->rx_buf && !xfer->rx_nbits)
+			xfer->rx_nbits = SPI_NBITS_SINGLE;
+		/* check transfer tx/rx_nbits:
+		 * 1. check the value matches one of single, dual and quad
+		 * 2. check tx/rx_nbits match the mode in spi_device
+		 */
+		if (xfer->tx_buf) {
+			if (xfer->tx_nbits != SPI_NBITS_SINGLE &&
+				xfer->tx_nbits != SPI_NBITS_DUAL &&
+				xfer->tx_nbits != SPI_NBITS_QUAD)
+				return -EINVAL;
+			if ((xfer->tx_nbits == SPI_NBITS_DUAL) &&
+				!(spi->mode & (SPI_TX_DUAL | SPI_TX_QUAD)))
+				return -EINVAL;
+			if ((xfer->tx_nbits == SPI_NBITS_QUAD) &&
+				!(spi->mode & SPI_TX_QUAD))
+				return -EINVAL;
+		}
+		/* check transfer rx_nbits */
+		if (xfer->rx_buf) {
+			if (xfer->rx_nbits != SPI_NBITS_SINGLE &&
+				xfer->rx_nbits != SPI_NBITS_DUAL &&
+				xfer->rx_nbits != SPI_NBITS_QUAD)
+				return -EINVAL;
+			if ((xfer->rx_nbits == SPI_NBITS_DUAL) &&
+				!(spi->mode & (SPI_RX_DUAL | SPI_RX_QUAD)))
+				return -EINVAL;
+			if ((xfer->rx_nbits == SPI_NBITS_QUAD) &&
+				!(spi->mode & SPI_RX_QUAD))
+				return -EINVAL;
+		}
+	}
+
+	message->status = -EINPROGRESS;
+
+	return 0;
+}
+
+static int __spi_async(struct spi_device *spi, struct spi_message *message)
+{
+	struct spi_controller *ctlr = spi->controller;
+
+	/*
+	 * Some controllers do not support doing regular SPI transfers. Return
+	 * ENOTSUPP when this is the case.
+	 */
+	if (!ctlr->transfer)
+		return -ENOTSUPP;
+
+	message->spi = spi;
+
+	SPI_STATISTICS_INCREMENT_FIELD(&ctlr->statistics, spi_async);
+	SPI_STATISTICS_INCREMENT_FIELD(&spi->statistics, spi_async);
+
+	trace_spi_message_submit(message);
+
+	return ctlr->transfer(spi, message);
+}
+
+/**
+ * spi_async - asynchronous SPI transfer
+ * @spi: device with which data will be exchanged
+ * @message: describes the data transfers, including completion callback
+ * Context: any (irqs may be blocked, etc)
+ *
+ * This call may be used in_irq and other contexts which can't sleep,
+ * as well as from task contexts which can sleep.
+ *
+ * The completion callback is invoked in a context which can't sleep.
+ * Before that invocation, the value of message->status is undefined.
+ * When the callback is issued, message->status holds either zero (to
+ * indicate complete success) or a negative error code.  After that
+ * callback returns, the driver which issued the transfer request may
+ * deallocate the associated memory; it's no longer in use by any SPI
+ * core or controller driver code.
+ *
+ * Note that although all messages to a spi_device are handled in
+ * FIFO order, messages may go to different devices in other orders.
+ * Some device might be higher priority, or have various "hard" access
+ * time requirements, for example.
+ *
+ * On detection of any fault during the transfer, processing of
+ * the entire message is aborted, and the device is deselected.
+ * Until returning from the associated message completion callback,
+ * no other spi_message queued to that device will be processed.
+ * (This rule applies equally to all the synchronous transfer calls,
+ * which are wrappers around this core asynchronous primitive.)
+ *
+ * Return: zero on success, else a negative error code.
+ */
+int spi_async(struct spi_device *spi, struct spi_message *message)
+{
+	struct spi_controller *ctlr = spi->controller;
+	int ret;
+	unsigned long flags;
+
+	ret = __spi_validate(spi, message);
+	if (ret != 0)
+		return ret;
+
+	spin_lock_irqsave(&ctlr->bus_lock_spinlock, flags);
+
+	if (ctlr->bus_lock_flag)
+		ret = -EBUSY;
+	else
+		ret = __spi_async(spi, message);
+
+	spin_unlock_irqrestore(&ctlr->bus_lock_spinlock, flags);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(spi_async);
+
+/**
+ * spi_async_locked - version of spi_async with exclusive bus usage
+ * @spi: device with which data will be exchanged
+ * @message: describes the data transfers, including completion callback
+ * Context: any (irqs may be blocked, etc)
+ *
+ * This call may be used in_irq and other contexts which can't sleep,
+ * as well as from task contexts which can sleep.
+ *
+ * The completion callback is invoked in a context which can't sleep.
+ * Before that invocation, the value of message->status is undefined.
+ * When the callback is issued, message->status holds either zero (to
+ * indicate complete success) or a negative error code.  After that
+ * callback returns, the driver which issued the transfer request may
+ * deallocate the associated memory; it's no longer in use by any SPI
+ * core or controller driver code.
+ *
+ * Note that although all messages to a spi_device are handled in
+ * FIFO order, messages may go to different devices in other orders.
+ * Some device might be higher priority, or have various "hard" access
+ * time requirements, for example.
+ *
+ * On detection of any fault during the transfer, processing of
+ * the entire message is aborted, and the device is deselected.
+ * Until returning from the associated message completion callback,
+ * no other spi_message queued to that device will be processed.
+ * (This rule applies equally to all the synchronous transfer calls,
+ * which are wrappers around this core asynchronous primitive.)
+ *
+ * Return: zero on success, else a negative error code.
+ */
+int spi_async_locked(struct spi_device *spi, struct spi_message *message)
+{
+	struct spi_controller *ctlr = spi->controller;
+	int ret;
+	unsigned long flags;
+
+	ret = __spi_validate(spi, message);
+	if (ret != 0)
+		return ret;
+
+	spin_lock_irqsave(&ctlr->bus_lock_spinlock, flags);
+
+	ret = __spi_async(spi, message);
+
+	spin_unlock_irqrestore(&ctlr->bus_lock_spinlock, flags);
+
+	return ret;
+
+}
+EXPORT_SYMBOL_GPL(spi_async_locked);
+
+/*-------------------------------------------------------------------------*/
+
+/* Utility methods for SPI protocol drivers, layered on
+ * top of the core.  Some other utility methods are defined as
+ * inline functions.
+ */
+
+static void spi_complete(void *arg)
+{
+	complete(arg);
+}
+
+static int __spi_sync(struct spi_device *spi, struct spi_message *message)
+{
+	DECLARE_COMPLETION_ONSTACK(done);
+	int status;
+	struct spi_controller *ctlr = spi->controller;
+	unsigned long flags;
+
+	status = __spi_validate(spi, message);
+	if (status != 0)
+		return status;
+
+	message->complete = spi_complete;
+	message->context = &done;
+	message->spi = spi;
+
+	SPI_STATISTICS_INCREMENT_FIELD(&ctlr->statistics, spi_sync);
+	SPI_STATISTICS_INCREMENT_FIELD(&spi->statistics, spi_sync);
+
+	/* If we're not using the legacy transfer method then we will
+	 * try to transfer in the calling context so special case.
+	 * This code would be less tricky if we could remove the
+	 * support for driver implemented message queues.
+	 */
+	if (ctlr->transfer == spi_queued_transfer) {
+		spin_lock_irqsave(&ctlr->bus_lock_spinlock, flags);
+
+		trace_spi_message_submit(message);
+
+		status = __spi_queued_transfer(spi, message, false);
+
+		spin_unlock_irqrestore(&ctlr->bus_lock_spinlock, flags);
+	} else {
+		status = spi_async_locked(spi, message);
+	}
+
+	if (status == 0) {
+		/* Push out the messages in the calling context if we
+		 * can.
+		 */
+		if (ctlr->transfer == spi_queued_transfer) {
+			SPI_STATISTICS_INCREMENT_FIELD(&ctlr->statistics,
+						       spi_sync_immediate);
+			SPI_STATISTICS_INCREMENT_FIELD(&spi->statistics,
+						       spi_sync_immediate);
+			__spi_pump_messages(ctlr, false);
+		}
+
+		wait_for_completion(&done);
+		status = message->status;
+	}
+	message->context = NULL;
+	return status;
+}
+
+/**
+ * spi_sync - blocking/synchronous SPI data transfers
+ * @spi: device with which data will be exchanged
+ * @message: describes the data transfers
+ * Context: can sleep
+ *
+ * This call may only be used from a context that may sleep.  The sleep
+ * is non-interruptible, and has no timeout.  Low-overhead controller
+ * drivers may DMA directly into and out of the message buffers.
+ *
+ * Note that the SPI device's chip select is active during the message,
+ * and then is normally disabled between messages.  Drivers for some
+ * frequently-used devices may want to minimize costs of selecting a chip,
+ * by leaving it selected in anticipation that the next message will go
+ * to the same chip.  (That may increase power usage.)
+ *
+ * Also, the caller is guaranteeing that the memory associated with the
+ * message will not be freed before this call returns.
+ *
+ * Return: zero on success, else a negative error code.
+ */
+int spi_sync(struct spi_device *spi, struct spi_message *message)
+{
+	int ret;
+
+	mutex_lock(&spi->controller->bus_lock_mutex);
+	ret = __spi_sync(spi, message);
+	mutex_unlock(&spi->controller->bus_lock_mutex);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(spi_sync);
+
+/**
+ * spi_sync_locked - version of spi_sync with exclusive bus usage
+ * @spi: device with which data will be exchanged
+ * @message: describes the data transfers
+ * Context: can sleep
+ *
+ * This call may only be used from a context that may sleep.  The sleep
+ * is non-interruptible, and has no timeout.  Low-overhead controller
+ * drivers may DMA directly into and out of the message buffers.
+ *
+ * This call should be used by drivers that require exclusive access to the
+ * SPI bus. It has to be preceded by a spi_bus_lock call. The SPI bus must
+ * be released by a spi_bus_unlock call when the exclusive access is over.
+ *
+ * Return: zero on success, else a negative error code.
+ */
+int spi_sync_locked(struct spi_device *spi, struct spi_message *message)
+{
+	return __spi_sync(spi, message);
+}
+EXPORT_SYMBOL_GPL(spi_sync_locked);
+
+/**
+ * spi_bus_lock - obtain a lock for exclusive SPI bus usage
+ * @ctlr: SPI bus master that should be locked for exclusive bus access
+ * Context: can sleep
+ *
+ * This call may only be used from a context that may sleep.  The sleep
+ * is non-interruptible, and has no timeout.
+ *
+ * This call should be used by drivers that require exclusive access to the
+ * SPI bus. The SPI bus must be released by a spi_bus_unlock call when the
+ * exclusive access is over. Data transfer must be done by spi_sync_locked
+ * and spi_async_locked calls when the SPI bus lock is held.
+ *
+ * Return: always zero.
+ */
+int spi_bus_lock(struct spi_controller *ctlr)
+{
+	unsigned long flags;
+
+	mutex_lock(&ctlr->bus_lock_mutex);
+
+	spin_lock_irqsave(&ctlr->bus_lock_spinlock, flags);
+	ctlr->bus_lock_flag = 1;
+	spin_unlock_irqrestore(&ctlr->bus_lock_spinlock, flags);
+
+	/* mutex remains locked until spi_bus_unlock is called */
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(spi_bus_lock);
+
+/**
+ * spi_bus_unlock - release the lock for exclusive SPI bus usage
+ * @ctlr: SPI bus master that was locked for exclusive bus access
+ * Context: can sleep
+ *
+ * This call may only be used from a context that may sleep.  The sleep
+ * is non-interruptible, and has no timeout.
+ *
+ * This call releases an SPI bus lock previously obtained by an spi_bus_lock
+ * call.
+ *
+ * Return: always zero.
+ */
+int spi_bus_unlock(struct spi_controller *ctlr)
+{
+	ctlr->bus_lock_flag = 0;
+
+	mutex_unlock(&ctlr->bus_lock_mutex);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(spi_bus_unlock);
+
+/* portable code must never pass more than 32 bytes */
+#define	SPI_BUFSIZ	max(32, SMP_CACHE_BYTES)
+
+static u8	*buf;
+
+/**
+ * spi_write_then_read - SPI synchronous write followed by read
+ * @spi: device with which data will be exchanged
+ * @txbuf: data to be written (need not be dma-safe)
+ * @n_tx: size of txbuf, in bytes
+ * @rxbuf: buffer into which data will be read (need not be dma-safe)
+ * @n_rx: size of rxbuf, in bytes
+ * Context: can sleep
+ *
+ * This performs a half duplex MicroWire style transaction with the
+ * device, sending txbuf and then reading rxbuf.  The return value
+ * is zero for success, else a negative errno status code.
+ * This call may only be used from a context that may sleep.
+ *
+ * Parameters to this routine are always copied using a small buffer;
+ * portable code should never use this for more than 32 bytes.
+ * Performance-sensitive or bulk transfer code should instead use
+ * spi_{async,sync}() calls with dma-safe buffers.
+ *
+ * Return: zero on success, else a negative error code.
+ */
+int spi_write_then_read(struct spi_device *spi,
+		const void *txbuf, unsigned n_tx,
+		void *rxbuf, unsigned n_rx)
+{
+	static DEFINE_MUTEX(lock);
+
+	int			status;
+	struct spi_message	message;
+	struct spi_transfer	x[2];
+	u8			*local_buf;
+
+	/* Use preallocated DMA-safe buffer if we can.  We can't avoid
+	 * copying here, (as a pure convenience thing), but we can
+	 * keep heap costs out of the hot path unless someone else is
+	 * using the pre-allocated buffer or the transfer is too large.
+	 */
+	if ((n_tx + n_rx) > SPI_BUFSIZ || !mutex_trylock(&lock)) {
+		local_buf = kmalloc(max((unsigned)SPI_BUFSIZ, n_tx + n_rx),
+				    GFP_KERNEL | GFP_DMA);
+		if (!local_buf)
+			return -ENOMEM;
+	} else {
+		local_buf = buf;
+	}
+
+	spi_message_init(&message);
+	memset(x, 0, sizeof(x));
+	if (n_tx) {
+		x[0].len = n_tx;
+		spi_message_add_tail(&x[0], &message);
+	}
+	if (n_rx) {
+		x[1].len = n_rx;
+		spi_message_add_tail(&x[1], &message);
+	}
+
+	memcpy(local_buf, txbuf, n_tx);
+	x[0].tx_buf = local_buf;
+	x[1].rx_buf = local_buf + n_tx;
+
+	/* do the i/o */
+	status = spi_sync(spi, &message);
+	if (status == 0)
+		memcpy(rxbuf, x[1].rx_buf, n_rx);
+
+	if (x[0].tx_buf == buf)
+		mutex_unlock(&lock);
+	else
+		kfree(local_buf);
+
+	return status;
+}
+EXPORT_SYMBOL_GPL(spi_write_then_read);
+
+/*-------------------------------------------------------------------------*/
+
+#if IS_ENABLED(CONFIG_OF_DYNAMIC)
+static int __spi_of_device_match(struct device *dev, void *data)
+{
+	return dev->of_node == data;
+}
+
+/* must call put_device() when done with returned spi_device device */
+static struct spi_device *of_find_spi_device_by_node(struct device_node *node)
+{
+	struct device *dev = bus_find_device(&spi_bus_type, NULL, node,
+						__spi_of_device_match);
+	return dev ? to_spi_device(dev) : NULL;
+}
+
+static int __spi_of_controller_match(struct device *dev, const void *data)
+{
+	return dev->of_node == data;
+}
+
+/* the spi controllers are not using spi_bus, so we find it with another way */
+static struct spi_controller *of_find_spi_controller_by_node(struct device_node *node)
+{
+	struct device *dev;
+
+	dev = class_find_device(&spi_master_class, NULL, node,
+				__spi_of_controller_match);
+	if (!dev && IS_ENABLED(CONFIG_SPI_SLAVE))
+		dev = class_find_device(&spi_slave_class, NULL, node,
+					__spi_of_controller_match);
+	if (!dev)
+		return NULL;
+
+	/* reference got in class_find_device */
+	return container_of(dev, struct spi_controller, dev);
+}
+
+static int of_spi_notify(struct notifier_block *nb, unsigned long action,
+			 void *arg)
+{
+	struct of_reconfig_data *rd = arg;
+	struct spi_controller *ctlr;
+	struct spi_device *spi;
+
+	switch (of_reconfig_get_state_change(action, arg)) {
+	case OF_RECONFIG_CHANGE_ADD:
+		ctlr = of_find_spi_controller_by_node(rd->dn->parent);
+		if (ctlr == NULL)
+			return NOTIFY_OK;	/* not for us */
+
+		if (of_node_test_and_set_flag(rd->dn, OF_POPULATED)) {
+			put_device(&ctlr->dev);
+			return NOTIFY_OK;
+		}
+
+		spi = of_register_spi_device(ctlr, rd->dn);
+		put_device(&ctlr->dev);
+
+		if (IS_ERR(spi)) {
+			pr_err("%s: failed to create for '%pOF'\n",
+					__func__, rd->dn);
+			of_node_clear_flag(rd->dn, OF_POPULATED);
+			return notifier_from_errno(PTR_ERR(spi));
+		}
+		break;
+
+	case OF_RECONFIG_CHANGE_REMOVE:
+		/* already depopulated? */
+		if (!of_node_check_flag(rd->dn, OF_POPULATED))
+			return NOTIFY_OK;
+
+		/* find our device by node */
+		spi = of_find_spi_device_by_node(rd->dn);
+		if (spi == NULL)
+			return NOTIFY_OK;	/* no? not meant for us */
+
+		/* unregister takes one ref away */
+		spi_unregister_device(spi);
+
+		/* and put the reference of the find */
+		put_device(&spi->dev);
+		break;
+	}
+
+	return NOTIFY_OK;
+}
+
+static struct notifier_block spi_of_notifier = {
+	.notifier_call = of_spi_notify,
+};
+#else /* IS_ENABLED(CONFIG_OF_DYNAMIC) */
+extern struct notifier_block spi_of_notifier;
+#endif /* IS_ENABLED(CONFIG_OF_DYNAMIC) */
+
+#if IS_ENABLED(CONFIG_ACPI)
+static int spi_acpi_controller_match(struct device *dev, const void *data)
+{
+	return ACPI_COMPANION(dev->parent) == data;
+}
+
+static int spi_acpi_device_match(struct device *dev, void *data)
+{
+	return ACPI_COMPANION(dev) == data;
+}
+
+static struct spi_controller *acpi_spi_find_controller_by_adev(struct acpi_device *adev)
+{
+	struct device *dev;
+
+	dev = class_find_device(&spi_master_class, NULL, adev,
+				spi_acpi_controller_match);
+	if (!dev && IS_ENABLED(CONFIG_SPI_SLAVE))
+		dev = class_find_device(&spi_slave_class, NULL, adev,
+					spi_acpi_controller_match);
+	if (!dev)
+		return NULL;
+
+	return container_of(dev, struct spi_controller, dev);
+}
+
+static struct spi_device *acpi_spi_find_device_by_adev(struct acpi_device *adev)
+{
+	struct device *dev;
+
+	dev = bus_find_device(&spi_bus_type, NULL, adev, spi_acpi_device_match);
+
+	return dev ? to_spi_device(dev) : NULL;
+}
+
+static int acpi_spi_notify(struct notifier_block *nb, unsigned long value,
+			   void *arg)
+{
+	struct acpi_device *adev = arg;
+	struct spi_controller *ctlr;
+	struct spi_device *spi;
+
+	switch (value) {
+	case ACPI_RECONFIG_DEVICE_ADD:
+		ctlr = acpi_spi_find_controller_by_adev(adev->parent);
+		if (!ctlr)
+			break;
+
+		acpi_register_spi_device(ctlr, adev);
+		put_device(&ctlr->dev);
+		break;
+	case ACPI_RECONFIG_DEVICE_REMOVE:
+		if (!acpi_device_enumerated(adev))
+			break;
+
+		spi = acpi_spi_find_device_by_adev(adev);
+		if (!spi)
+			break;
+
+		spi_unregister_device(spi);
+		put_device(&spi->dev);
+		break;
+	}
+
+	return NOTIFY_OK;
+}
+
+static struct notifier_block spi_acpi_notifier = {
+	.notifier_call = acpi_spi_notify,
+};
+#else
+extern struct notifier_block spi_acpi_notifier;
+#endif
+
+static int __init spi_init(void)
+{
+	int	status;
+
+	buf = kmalloc(SPI_BUFSIZ, GFP_KERNEL);
+	if (!buf) {
+		status = -ENOMEM;
+		goto err0;
+	}
+
+	status = bus_register(&spi_bus_type);
+	if (status < 0)
+		goto err1;
+
+	status = class_register(&spi_master_class);
+	if (status < 0)
+		goto err2;
+
+	if (IS_ENABLED(CONFIG_SPI_SLAVE)) {
+		status = class_register(&spi_slave_class);
+		if (status < 0)
+			goto err3;
+	}
+
+	if (IS_ENABLED(CONFIG_OF_DYNAMIC))
+		WARN_ON(of_reconfig_notifier_register(&spi_of_notifier));
+	if (IS_ENABLED(CONFIG_ACPI))
+		WARN_ON(acpi_reconfig_notifier_register(&spi_acpi_notifier));
+
+	return 0;
+
+err3:
+	class_unregister(&spi_master_class);
+err2:
+	bus_unregister(&spi_bus_type);
+err1:
+	kfree(buf);
+	buf = NULL;
+err0:
+	return status;
+}
+
+/* board_info is normally registered in arch_initcall(),
+ * but even essential drivers wait till later
+ *
+ * REVISIT only boardinfo really needs static linking. the rest (device and
+ * driver registration) _could_ be dynamically linked (modular) ... costs
+ * include needing to have boardinfo data structures be much more public.
+ */
+postcore_initcall(spi_init);
+
diff --git a/drivers/spi/spidev.c b/drivers/spi/spidev.c
new file mode 100644
index 0000000..cda1071
--- /dev/null
+++ b/drivers/spi/spidev.c
@@ -0,0 +1,858 @@
+/*
+ * Simple synchronous userspace interface to SPI devices
+ *
+ * Copyright (C) 2006 SWAPP
+ *	Andrea Paterniani <a.paterniani@swapp-eng.it>
+ * Copyright (C) 2007 David Brownell (simplification, cleanup)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/ioctl.h>
+#include <linux/fs.h>
+#include <linux/device.h>
+#include <linux/err.h>
+#include <linux/list.h>
+#include <linux/errno.h>
+#include <linux/mutex.h>
+#include <linux/slab.h>
+#include <linux/compat.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/acpi.h>
+
+#include <linux/spi/spi.h>
+#include <linux/spi/spidev.h>
+
+#include <linux/uaccess.h>
+
+
+/*
+ * This supports access to SPI devices using normal userspace I/O calls.
+ * Note that while traditional UNIX/POSIX I/O semantics are half duplex,
+ * and often mask message boundaries, full SPI support requires full duplex
+ * transfers.  There are several kinds of internal message boundaries to
+ * handle chipselect management and other protocol options.
+ *
+ * SPI has a character major number assigned.  We allocate minor numbers
+ * dynamically using a bitmask.  You must use hotplug tools, such as udev
+ * (or mdev with busybox) to create and destroy the /dev/spidevB.C device
+ * nodes, since there is no fixed association of minor numbers with any
+ * particular SPI bus or device.
+ */
+#define SPIDEV_MAJOR			153	/* assigned */
+#define N_SPI_MINORS			32	/* ... up to 256 */
+
+static DECLARE_BITMAP(minors, N_SPI_MINORS);
+
+
+/* Bit masks for spi_device.mode management.  Note that incorrect
+ * settings for some settings can cause *lots* of trouble for other
+ * devices on a shared bus:
+ *
+ *  - CS_HIGH ... this device will be active when it shouldn't be
+ *  - 3WIRE ... when active, it won't behave as it should
+ *  - NO_CS ... there will be no explicit message boundaries; this
+ *	is completely incompatible with the shared bus model
+ *  - READY ... transfers may proceed when they shouldn't.
+ *
+ * REVISIT should changing those flags be privileged?
+ */
+#define SPI_MODE_MASK		(SPI_CPHA | SPI_CPOL | SPI_CS_HIGH \
+				| SPI_LSB_FIRST | SPI_3WIRE | SPI_LOOP \
+				| SPI_NO_CS | SPI_READY | SPI_TX_DUAL \
+				| SPI_TX_QUAD | SPI_RX_DUAL | SPI_RX_QUAD)
+
+struct spidev_data {
+	dev_t			devt;
+	spinlock_t		spi_lock;
+	struct spi_device	*spi;
+	struct list_head	device_entry;
+
+	/* TX/RX buffers are NULL unless this device is open (users > 0) */
+	struct mutex		buf_lock;
+	unsigned		users;
+	u8			*tx_buffer;
+	u8			*rx_buffer;
+	u32			speed_hz;
+};
+
+static LIST_HEAD(device_list);
+static DEFINE_MUTEX(device_list_lock);
+
+static unsigned bufsiz = 4096;
+module_param(bufsiz, uint, S_IRUGO);
+MODULE_PARM_DESC(bufsiz, "data bytes in biggest supported SPI message");
+
+/*-------------------------------------------------------------------------*/
+
+static ssize_t
+spidev_sync(struct spidev_data *spidev, struct spi_message *message)
+{
+	int status;
+	struct spi_device *spi;
+
+	spin_lock_irq(&spidev->spi_lock);
+	spi = spidev->spi;
+	spin_unlock_irq(&spidev->spi_lock);
+
+	if (spi == NULL)
+		status = -ESHUTDOWN;
+	else
+		status = spi_sync(spi, message);
+
+	if (status == 0)
+		status = message->actual_length;
+
+	return status;
+}
+
+static inline ssize_t
+spidev_sync_write(struct spidev_data *spidev, size_t len)
+{
+	struct spi_transfer	t = {
+			.tx_buf		= spidev->tx_buffer,
+			.len		= len,
+			.speed_hz	= spidev->speed_hz,
+		};
+	struct spi_message	m;
+
+	spi_message_init(&m);
+	spi_message_add_tail(&t, &m);
+	return spidev_sync(spidev, &m);
+}
+
+static inline ssize_t
+spidev_sync_read(struct spidev_data *spidev, size_t len)
+{
+	struct spi_transfer	t = {
+			.rx_buf		= spidev->rx_buffer,
+			.len		= len,
+			.speed_hz	= spidev->speed_hz,
+		};
+	struct spi_message	m;
+
+	spi_message_init(&m);
+	spi_message_add_tail(&t, &m);
+	return spidev_sync(spidev, &m);
+}
+
+/*-------------------------------------------------------------------------*/
+
+/* Read-only message with current device setup */
+static ssize_t
+spidev_read(struct file *filp, char __user *buf, size_t count, loff_t *f_pos)
+{
+	struct spidev_data	*spidev;
+	ssize_t			status = 0;
+
+	/* chipselect only toggles at start or end of operation */
+	if (count > bufsiz)
+		return -EMSGSIZE;
+
+	spidev = filp->private_data;
+
+	mutex_lock(&spidev->buf_lock);
+	status = spidev_sync_read(spidev, count);
+	if (status > 0) {
+		unsigned long	missing;
+
+		missing = copy_to_user(buf, spidev->rx_buffer, status);
+		if (missing == status)
+			status = -EFAULT;
+		else
+			status = status - missing;
+	}
+	mutex_unlock(&spidev->buf_lock);
+
+	return status;
+}
+
+/* Write-only message with current device setup */
+static ssize_t
+spidev_write(struct file *filp, const char __user *buf,
+		size_t count, loff_t *f_pos)
+{
+	struct spidev_data	*spidev;
+	ssize_t			status = 0;
+	unsigned long		missing;
+
+	/* chipselect only toggles at start or end of operation */
+	if (count > bufsiz)
+		return -EMSGSIZE;
+
+	spidev = filp->private_data;
+
+	mutex_lock(&spidev->buf_lock);
+	missing = copy_from_user(spidev->tx_buffer, buf, count);
+	if (missing == 0)
+		status = spidev_sync_write(spidev, count);
+	else
+		status = -EFAULT;
+	mutex_unlock(&spidev->buf_lock);
+
+	return status;
+}
+
+static int spidev_message(struct spidev_data *spidev,
+		struct spi_ioc_transfer *u_xfers, unsigned n_xfers)
+{
+	struct spi_message	msg;
+	struct spi_transfer	*k_xfers;
+	struct spi_transfer	*k_tmp;
+	struct spi_ioc_transfer *u_tmp;
+	unsigned		n, total, tx_total, rx_total;
+	u8			*tx_buf, *rx_buf;
+	int			status = -EFAULT;
+
+	spi_message_init(&msg);
+	k_xfers = kcalloc(n_xfers, sizeof(*k_tmp), GFP_KERNEL);
+	if (k_xfers == NULL)
+		return -ENOMEM;
+
+	/* Construct spi_message, copying any tx data to bounce buffer.
+	 * We walk the array of user-provided transfers, using each one
+	 * to initialize a kernel version of the same transfer.
+	 */
+	tx_buf = spidev->tx_buffer;
+	rx_buf = spidev->rx_buffer;
+	total = 0;
+	tx_total = 0;
+	rx_total = 0;
+	for (n = n_xfers, k_tmp = k_xfers, u_tmp = u_xfers;
+			n;
+			n--, k_tmp++, u_tmp++) {
+		k_tmp->len = u_tmp->len;
+
+		total += k_tmp->len;
+		/* Since the function returns the total length of transfers
+		 * on success, restrict the total to positive int values to
+		 * avoid the return value looking like an error.  Also check
+		 * each transfer length to avoid arithmetic overflow.
+		 */
+		if (total > INT_MAX || k_tmp->len > INT_MAX) {
+			status = -EMSGSIZE;
+			goto done;
+		}
+
+		if (u_tmp->rx_buf) {
+			/* this transfer needs space in RX bounce buffer */
+			rx_total += k_tmp->len;
+			if (rx_total > bufsiz) {
+				status = -EMSGSIZE;
+				goto done;
+			}
+			k_tmp->rx_buf = rx_buf;
+			rx_buf += k_tmp->len;
+		}
+		if (u_tmp->tx_buf) {
+			/* this transfer needs space in TX bounce buffer */
+			tx_total += k_tmp->len;
+			if (tx_total > bufsiz) {
+				status = -EMSGSIZE;
+				goto done;
+			}
+			k_tmp->tx_buf = tx_buf;
+			if (copy_from_user(tx_buf, (const u8 __user *)
+						(uintptr_t) u_tmp->tx_buf,
+					u_tmp->len))
+				goto done;
+			tx_buf += k_tmp->len;
+		}
+
+		k_tmp->cs_change = !!u_tmp->cs_change;
+		k_tmp->tx_nbits = u_tmp->tx_nbits;
+		k_tmp->rx_nbits = u_tmp->rx_nbits;
+		k_tmp->bits_per_word = u_tmp->bits_per_word;
+		k_tmp->delay_usecs = u_tmp->delay_usecs;
+		k_tmp->speed_hz = u_tmp->speed_hz;
+		if (!k_tmp->speed_hz)
+			k_tmp->speed_hz = spidev->speed_hz;
+#ifdef VERBOSE
+		dev_dbg(&spidev->spi->dev,
+			"  xfer len %u %s%s%s%dbits %u usec %uHz\n",
+			u_tmp->len,
+			u_tmp->rx_buf ? "rx " : "",
+			u_tmp->tx_buf ? "tx " : "",
+			u_tmp->cs_change ? "cs " : "",
+			u_tmp->bits_per_word ? : spidev->spi->bits_per_word,
+			u_tmp->delay_usecs,
+			u_tmp->speed_hz ? : spidev->spi->max_speed_hz);
+#endif
+		spi_message_add_tail(k_tmp, &msg);
+	}
+
+	status = spidev_sync(spidev, &msg);
+	if (status < 0)
+		goto done;
+
+	/* copy any rx data out of bounce buffer */
+	rx_buf = spidev->rx_buffer;
+	for (n = n_xfers, u_tmp = u_xfers; n; n--, u_tmp++) {
+		if (u_tmp->rx_buf) {
+			if (copy_to_user((u8 __user *)
+					(uintptr_t) u_tmp->rx_buf, rx_buf,
+					u_tmp->len)) {
+				status = -EFAULT;
+				goto done;
+			}
+			rx_buf += u_tmp->len;
+		}
+	}
+	status = total;
+
+done:
+	kfree(k_xfers);
+	return status;
+}
+
+static struct spi_ioc_transfer *
+spidev_get_ioc_message(unsigned int cmd, struct spi_ioc_transfer __user *u_ioc,
+		unsigned *n_ioc)
+{
+	u32	tmp;
+
+	/* Check type, command number and direction */
+	if (_IOC_TYPE(cmd) != SPI_IOC_MAGIC
+			|| _IOC_NR(cmd) != _IOC_NR(SPI_IOC_MESSAGE(0))
+			|| _IOC_DIR(cmd) != _IOC_WRITE)
+		return ERR_PTR(-ENOTTY);
+
+	tmp = _IOC_SIZE(cmd);
+	if ((tmp % sizeof(struct spi_ioc_transfer)) != 0)
+		return ERR_PTR(-EINVAL);
+	*n_ioc = tmp / sizeof(struct spi_ioc_transfer);
+	if (*n_ioc == 0)
+		return NULL;
+
+	/* copy into scratch area */
+	return memdup_user(u_ioc, tmp);
+}
+
+static long
+spidev_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
+{
+	int			retval = 0;
+	struct spidev_data	*spidev;
+	struct spi_device	*spi;
+	u32			tmp;
+	unsigned		n_ioc;
+	struct spi_ioc_transfer	*ioc;
+
+	/* Check type and command number */
+	if (_IOC_TYPE(cmd) != SPI_IOC_MAGIC)
+		return -ENOTTY;
+
+	/* guard against device removal before, or while,
+	 * we issue this ioctl.
+	 */
+	spidev = filp->private_data;
+	spin_lock_irq(&spidev->spi_lock);
+	spi = spi_dev_get(spidev->spi);
+	spin_unlock_irq(&spidev->spi_lock);
+
+	if (spi == NULL)
+		return -ESHUTDOWN;
+
+	/* use the buffer lock here for triple duty:
+	 *  - prevent I/O (from us) so calling spi_setup() is safe;
+	 *  - prevent concurrent SPI_IOC_WR_* from morphing
+	 *    data fields while SPI_IOC_RD_* reads them;
+	 *  - SPI_IOC_MESSAGE needs the buffer locked "normally".
+	 */
+	mutex_lock(&spidev->buf_lock);
+
+	switch (cmd) {
+	/* read requests */
+	case SPI_IOC_RD_MODE:
+		retval = put_user(spi->mode & SPI_MODE_MASK,
+					(__u8 __user *)arg);
+		break;
+	case SPI_IOC_RD_MODE32:
+		retval = put_user(spi->mode & SPI_MODE_MASK,
+					(__u32 __user *)arg);
+		break;
+	case SPI_IOC_RD_LSB_FIRST:
+		retval = put_user((spi->mode & SPI_LSB_FIRST) ?  1 : 0,
+					(__u8 __user *)arg);
+		break;
+	case SPI_IOC_RD_BITS_PER_WORD:
+		retval = put_user(spi->bits_per_word, (__u8 __user *)arg);
+		break;
+	case SPI_IOC_RD_MAX_SPEED_HZ:
+		retval = put_user(spidev->speed_hz, (__u32 __user *)arg);
+		break;
+
+	/* write requests */
+	case SPI_IOC_WR_MODE:
+	case SPI_IOC_WR_MODE32:
+		if (cmd == SPI_IOC_WR_MODE)
+			retval = get_user(tmp, (u8 __user *)arg);
+		else
+			retval = get_user(tmp, (u32 __user *)arg);
+		if (retval == 0) {
+			u32	save = spi->mode;
+
+			if (tmp & ~SPI_MODE_MASK) {
+				retval = -EINVAL;
+				break;
+			}
+
+			tmp |= spi->mode & ~SPI_MODE_MASK;
+			spi->mode = (u16)tmp;
+			retval = spi_setup(spi);
+			if (retval < 0)
+				spi->mode = save;
+			else
+				dev_dbg(&spi->dev, "spi mode %x\n", tmp);
+		}
+		break;
+	case SPI_IOC_WR_LSB_FIRST:
+		retval = get_user(tmp, (__u8 __user *)arg);
+		if (retval == 0) {
+			u32	save = spi->mode;
+
+			if (tmp)
+				spi->mode |= SPI_LSB_FIRST;
+			else
+				spi->mode &= ~SPI_LSB_FIRST;
+			retval = spi_setup(spi);
+			if (retval < 0)
+				spi->mode = save;
+			else
+				dev_dbg(&spi->dev, "%csb first\n",
+						tmp ? 'l' : 'm');
+		}
+		break;
+	case SPI_IOC_WR_BITS_PER_WORD:
+		retval = get_user(tmp, (__u8 __user *)arg);
+		if (retval == 0) {
+			u8	save = spi->bits_per_word;
+
+			spi->bits_per_word = tmp;
+			retval = spi_setup(spi);
+			if (retval < 0)
+				spi->bits_per_word = save;
+			else
+				dev_dbg(&spi->dev, "%d bits per word\n", tmp);
+		}
+		break;
+	case SPI_IOC_WR_MAX_SPEED_HZ:
+		retval = get_user(tmp, (__u32 __user *)arg);
+		if (retval == 0) {
+			u32	save = spi->max_speed_hz;
+
+			spi->max_speed_hz = tmp;
+			retval = spi_setup(spi);
+			if (retval >= 0)
+				spidev->speed_hz = tmp;
+			else
+				dev_dbg(&spi->dev, "%d Hz (max)\n", tmp);
+			spi->max_speed_hz = save;
+		}
+		break;
+
+	default:
+		/* segmented and/or full-duplex I/O request */
+		/* Check message and copy into scratch area */
+		ioc = spidev_get_ioc_message(cmd,
+				(struct spi_ioc_transfer __user *)arg, &n_ioc);
+		if (IS_ERR(ioc)) {
+			retval = PTR_ERR(ioc);
+			break;
+		}
+		if (!ioc)
+			break;	/* n_ioc is also 0 */
+
+		/* translate to spi_message, execute */
+		retval = spidev_message(spidev, ioc, n_ioc);
+		kfree(ioc);
+		break;
+	}
+
+	mutex_unlock(&spidev->buf_lock);
+	spi_dev_put(spi);
+	return retval;
+}
+
+#ifdef CONFIG_COMPAT
+static long
+spidev_compat_ioc_message(struct file *filp, unsigned int cmd,
+		unsigned long arg)
+{
+	struct spi_ioc_transfer __user	*u_ioc;
+	int				retval = 0;
+	struct spidev_data		*spidev;
+	struct spi_device		*spi;
+	unsigned			n_ioc, n;
+	struct spi_ioc_transfer		*ioc;
+
+	u_ioc = (struct spi_ioc_transfer __user *) compat_ptr(arg);
+
+	/* guard against device removal before, or while,
+	 * we issue this ioctl.
+	 */
+	spidev = filp->private_data;
+	spin_lock_irq(&spidev->spi_lock);
+	spi = spi_dev_get(spidev->spi);
+	spin_unlock_irq(&spidev->spi_lock);
+
+	if (spi == NULL)
+		return -ESHUTDOWN;
+
+	/* SPI_IOC_MESSAGE needs the buffer locked "normally" */
+	mutex_lock(&spidev->buf_lock);
+
+	/* Check message and copy into scratch area */
+	ioc = spidev_get_ioc_message(cmd, u_ioc, &n_ioc);
+	if (IS_ERR(ioc)) {
+		retval = PTR_ERR(ioc);
+		goto done;
+	}
+	if (!ioc)
+		goto done;	/* n_ioc is also 0 */
+
+	/* Convert buffer pointers */
+	for (n = 0; n < n_ioc; n++) {
+		ioc[n].rx_buf = (uintptr_t) compat_ptr(ioc[n].rx_buf);
+		ioc[n].tx_buf = (uintptr_t) compat_ptr(ioc[n].tx_buf);
+	}
+
+	/* translate to spi_message, execute */
+	retval = spidev_message(spidev, ioc, n_ioc);
+	kfree(ioc);
+
+done:
+	mutex_unlock(&spidev->buf_lock);
+	spi_dev_put(spi);
+	return retval;
+}
+
+static long
+spidev_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
+{
+	if (_IOC_TYPE(cmd) == SPI_IOC_MAGIC
+			&& _IOC_NR(cmd) == _IOC_NR(SPI_IOC_MESSAGE(0))
+			&& _IOC_DIR(cmd) == _IOC_WRITE)
+		return spidev_compat_ioc_message(filp, cmd, arg);
+
+	return spidev_ioctl(filp, cmd, (unsigned long)compat_ptr(arg));
+}
+#else
+#define spidev_compat_ioctl NULL
+#endif /* CONFIG_COMPAT */
+
+static int spidev_open(struct inode *inode, struct file *filp)
+{
+	struct spidev_data	*spidev;
+	int			status = -ENXIO;
+
+	mutex_lock(&device_list_lock);
+
+	list_for_each_entry(spidev, &device_list, device_entry) {
+		if (spidev->devt == inode->i_rdev) {
+			status = 0;
+			break;
+		}
+	}
+
+	if (status) {
+		pr_debug("spidev: nothing for minor %d\n", iminor(inode));
+		goto err_find_dev;
+	}
+
+	if (!spidev->tx_buffer) {
+		spidev->tx_buffer = kmalloc(bufsiz, GFP_KERNEL);
+		if (!spidev->tx_buffer) {
+			dev_dbg(&spidev->spi->dev, "open/ENOMEM\n");
+			status = -ENOMEM;
+			goto err_find_dev;
+		}
+	}
+
+	if (!spidev->rx_buffer) {
+		spidev->rx_buffer = kmalloc(bufsiz, GFP_KERNEL);
+		if (!spidev->rx_buffer) {
+			dev_dbg(&spidev->spi->dev, "open/ENOMEM\n");
+			status = -ENOMEM;
+			goto err_alloc_rx_buf;
+		}
+	}
+
+	spidev->users++;
+	filp->private_data = spidev;
+	nonseekable_open(inode, filp);
+
+	mutex_unlock(&device_list_lock);
+	return 0;
+
+err_alloc_rx_buf:
+	kfree(spidev->tx_buffer);
+	spidev->tx_buffer = NULL;
+err_find_dev:
+	mutex_unlock(&device_list_lock);
+	return status;
+}
+
+static int spidev_release(struct inode *inode, struct file *filp)
+{
+	struct spidev_data	*spidev;
+
+	mutex_lock(&device_list_lock);
+	spidev = filp->private_data;
+	filp->private_data = NULL;
+
+	/* last close? */
+	spidev->users--;
+	if (!spidev->users) {
+		int		dofree;
+
+		kfree(spidev->tx_buffer);
+		spidev->tx_buffer = NULL;
+
+		kfree(spidev->rx_buffer);
+		spidev->rx_buffer = NULL;
+
+		spin_lock_irq(&spidev->spi_lock);
+		if (spidev->spi)
+			spidev->speed_hz = spidev->spi->max_speed_hz;
+
+		/* ... after we unbound from the underlying device? */
+		dofree = (spidev->spi == NULL);
+		spin_unlock_irq(&spidev->spi_lock);
+
+		if (dofree)
+			kfree(spidev);
+	}
+	mutex_unlock(&device_list_lock);
+
+	return 0;
+}
+
+static const struct file_operations spidev_fops = {
+	.owner =	THIS_MODULE,
+	/* REVISIT switch to aio primitives, so that userspace
+	 * gets more complete API coverage.  It'll simplify things
+	 * too, except for the locking.
+	 */
+	.write =	spidev_write,
+	.read =		spidev_read,
+	.unlocked_ioctl = spidev_ioctl,
+	.compat_ioctl = spidev_compat_ioctl,
+	.open =		spidev_open,
+	.release =	spidev_release,
+	.llseek =	no_llseek,
+};
+
+/*-------------------------------------------------------------------------*/
+
+/* The main reason to have this class is to make mdev/udev create the
+ * /dev/spidevB.C character device nodes exposing our userspace API.
+ * It also simplifies memory management.
+ */
+
+static struct class *spidev_class;
+
+#ifdef CONFIG_OF
+static const struct of_device_id spidev_dt_ids[] = {
+	{ .compatible = "rohm,dh2228fv" },
+	{ .compatible = "lineartechnology,ltc2488" },
+	{ .compatible = "ge,achc" },
+	{ .compatible = "semtech,sx1301" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, spidev_dt_ids);
+#endif
+
+#ifdef CONFIG_ACPI
+
+/* Dummy SPI devices not to be used in production systems */
+#define SPIDEV_ACPI_DUMMY	1
+
+static const struct acpi_device_id spidev_acpi_ids[] = {
+	/*
+	 * The ACPI SPT000* devices are only meant for development and
+	 * testing. Systems used in production should have a proper ACPI
+	 * description of the connected peripheral and they should also use
+	 * a proper driver instead of poking directly to the SPI bus.
+	 */
+	{ "SPT0001", SPIDEV_ACPI_DUMMY },
+	{ "SPT0002", SPIDEV_ACPI_DUMMY },
+	{ "SPT0003", SPIDEV_ACPI_DUMMY },
+	{},
+};
+MODULE_DEVICE_TABLE(acpi, spidev_acpi_ids);
+
+static void spidev_probe_acpi(struct spi_device *spi)
+{
+	const struct acpi_device_id *id;
+
+	if (!has_acpi_companion(&spi->dev))
+		return;
+
+	id = acpi_match_device(spidev_acpi_ids, &spi->dev);
+	if (WARN_ON(!id))
+		return;
+
+	if (id->driver_data == SPIDEV_ACPI_DUMMY)
+		dev_warn(&spi->dev, "do not use this driver in production systems!\n");
+}
+#else
+static inline void spidev_probe_acpi(struct spi_device *spi) {}
+#endif
+
+/*-------------------------------------------------------------------------*/
+
+static int spidev_probe(struct spi_device *spi)
+{
+	struct spidev_data	*spidev;
+	int			status;
+	unsigned long		minor;
+
+	/*
+	 * spidev should never be referenced in DT without a specific
+	 * compatible string, it is a Linux implementation thing
+	 * rather than a description of the hardware.
+	 */
+	if (spi->dev.of_node && !of_match_device(spidev_dt_ids, &spi->dev)) {
+		dev_err(&spi->dev, "buggy DT: spidev listed directly in DT\n");
+		WARN_ON(spi->dev.of_node &&
+			!of_match_device(spidev_dt_ids, &spi->dev));
+	}
+
+	spidev_probe_acpi(spi);
+
+	/* Allocate driver data */
+	spidev = kzalloc(sizeof(*spidev), GFP_KERNEL);
+	if (!spidev)
+		return -ENOMEM;
+
+	/* Initialize the driver data */
+	spidev->spi = spi;
+	spin_lock_init(&spidev->spi_lock);
+	mutex_init(&spidev->buf_lock);
+
+	INIT_LIST_HEAD(&spidev->device_entry);
+
+	/* If we can allocate a minor number, hook up this device.
+	 * Reusing minors is fine so long as udev or mdev is working.
+	 */
+	mutex_lock(&device_list_lock);
+	minor = find_first_zero_bit(minors, N_SPI_MINORS);
+	if (minor < N_SPI_MINORS) {
+		struct device *dev;
+
+		spidev->devt = MKDEV(SPIDEV_MAJOR, minor);
+		dev = device_create(spidev_class, &spi->dev, spidev->devt,
+				    spidev, "spidev%d.%d",
+				    spi->master->bus_num, spi->chip_select);
+		status = PTR_ERR_OR_ZERO(dev);
+	} else {
+		dev_dbg(&spi->dev, "no minor number available!\n");
+		status = -ENODEV;
+	}
+	if (status == 0) {
+		set_bit(minor, minors);
+		list_add(&spidev->device_entry, &device_list);
+	}
+	mutex_unlock(&device_list_lock);
+
+	spidev->speed_hz = spi->max_speed_hz;
+
+	if (status == 0)
+		spi_set_drvdata(spi, spidev);
+	else
+		kfree(spidev);
+
+	return status;
+}
+
+static int spidev_remove(struct spi_device *spi)
+{
+	struct spidev_data	*spidev = spi_get_drvdata(spi);
+
+	/* make sure ops on existing fds can abort cleanly */
+	spin_lock_irq(&spidev->spi_lock);
+	spidev->spi = NULL;
+	spin_unlock_irq(&spidev->spi_lock);
+
+	/* prevent new opens */
+	mutex_lock(&device_list_lock);
+	list_del(&spidev->device_entry);
+	device_destroy(spidev_class, spidev->devt);
+	clear_bit(MINOR(spidev->devt), minors);
+	if (spidev->users == 0)
+		kfree(spidev);
+	mutex_unlock(&device_list_lock);
+
+	return 0;
+}
+
+static struct spi_driver spidev_spi_driver = {
+	.driver = {
+		.name =		"spidev",
+		.of_match_table = of_match_ptr(spidev_dt_ids),
+		.acpi_match_table = ACPI_PTR(spidev_acpi_ids),
+	},
+	.probe =	spidev_probe,
+	.remove =	spidev_remove,
+
+	/* NOTE:  suspend/resume methods are not necessary here.
+	 * We don't do anything except pass the requests to/from
+	 * the underlying controller.  The refrigerator handles
+	 * most issues; the controller driver handles the rest.
+	 */
+};
+
+/*-------------------------------------------------------------------------*/
+
+static int __init spidev_init(void)
+{
+	int status;
+
+	/* Claim our 256 reserved device numbers.  Then register a class
+	 * that will key udev/mdev to add/remove /dev nodes.  Last, register
+	 * the driver which manages those device numbers.
+	 */
+	BUILD_BUG_ON(N_SPI_MINORS > 256);
+	status = register_chrdev(SPIDEV_MAJOR, "spi", &spidev_fops);
+	if (status < 0)
+		return status;
+
+	spidev_class = class_create(THIS_MODULE, "spidev");
+	if (IS_ERR(spidev_class)) {
+		unregister_chrdev(SPIDEV_MAJOR, spidev_spi_driver.driver.name);
+		return PTR_ERR(spidev_class);
+	}
+
+	status = spi_register_driver(&spidev_spi_driver);
+	if (status < 0) {
+		class_destroy(spidev_class);
+		unregister_chrdev(SPIDEV_MAJOR, spidev_spi_driver.driver.name);
+	}
+	return status;
+}
+module_init(spidev_init);
+
+static void __exit spidev_exit(void)
+{
+	spi_unregister_driver(&spidev_spi_driver);
+	class_destroy(spidev_class);
+	unregister_chrdev(SPIDEV_MAJOR, spidev_spi_driver.driver.name);
+}
+module_exit(spidev_exit);
+
+MODULE_AUTHOR("Andrea Paterniani, <a.paterniani@swapp-eng.it>");
+MODULE_DESCRIPTION("User mode SPI device interface");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("spi:spidev");