Update Linux to v5.10.109
Sourced from [1]
[1] https://cdn.kernel.org/pub/linux/kernel/v5.x/linux-5.10.109.tar.xz
Change-Id: I19bca9fc6762d4e63bcf3e4cba88bbe560d9c76c
Signed-off-by: Olivier Deprez <olivier.deprez@arm.com>
diff --git a/arch/s390/pci/Makefile b/arch/s390/pci/Makefile
index 748626a..bf557a1 100644
--- a/arch/s390/pci/Makefile
+++ b/arch/s390/pci/Makefile
@@ -4,4 +4,6 @@
#
obj-$(CONFIG_PCI) += pci.o pci_irq.o pci_dma.o pci_clp.o pci_sysfs.o \
- pci_event.o pci_debug.o pci_insn.o pci_mmio.o
+ pci_event.o pci_debug.o pci_insn.o pci_mmio.o \
+ pci_bus.o
+obj-$(CONFIG_PCI_IOV) += pci_iov.o
diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
index b8ddacf..e14e4a3 100644
--- a/arch/s390/pci/pci.c
+++ b/arch/s390/pci/pci.c
@@ -27,6 +27,7 @@
#include <linux/seq_file.h>
#include <linux/jump_label.h>
#include <linux/pci.h>
+#include <linux/printk.h>
#include <asm/isc.h>
#include <asm/airq.h>
@@ -35,17 +36,22 @@
#include <asm/pci_clp.h>
#include <asm/pci_dma.h>
+#include "pci_bus.h"
+#include "pci_iov.h"
+
/* list of all detected zpci devices */
static LIST_HEAD(zpci_list);
static DEFINE_SPINLOCK(zpci_list_lock);
-static DECLARE_BITMAP(zpci_domain, ZPCI_NR_DEVICES);
+static DECLARE_BITMAP(zpci_domain, ZPCI_DOMAIN_BITMAP_SIZE);
static DEFINE_SPINLOCK(zpci_domain_lock);
#define ZPCI_IOMAP_ENTRIES \
- min(((unsigned long) ZPCI_NR_DEVICES * PCI_BAR_COUNT / 2), \
+ min(((unsigned long) ZPCI_NR_DEVICES * PCI_STD_NUM_BARS / 2), \
ZPCI_IOMAP_MAX_ENTRIES)
+unsigned int s390_pci_no_rid;
+
static DEFINE_SPINLOCK(zpci_iomap_lock);
static unsigned long *zpci_iomap_bitmap;
struct zpci_iomap_entry *zpci_iomap_start;
@@ -86,17 +92,12 @@
spin_unlock(&zpci_list_lock);
list_for_each_entry_safe(zdev, tmp, &remove, entry)
- zpci_remove_device(zdev);
-}
-
-static struct zpci_dev *get_zdev_by_bus(struct pci_bus *bus)
-{
- return (bus && bus->sysdata) ? (struct zpci_dev *) bus->sysdata : NULL;
+ zpci_device_reserved(zdev);
}
int pci_domain_nr(struct pci_bus *bus)
{
- return ((struct zpci_dev *) bus->sysdata)->domain;
+ return ((struct zpci_bus *) bus->sysdata)->domain_nr;
}
EXPORT_SYMBOL_GPL(pci_domain_nr);
@@ -226,34 +227,58 @@
zpci_memcpy_toio(to, from, count);
}
-void __iomem *ioremap(unsigned long ioaddr, unsigned long size)
+static void __iomem *__ioremap(phys_addr_t addr, size_t size, pgprot_t prot)
{
+ unsigned long offset, vaddr;
struct vm_struct *area;
- unsigned long offset;
+ phys_addr_t last_addr;
- if (!size)
+ last_addr = addr + size - 1;
+ if (!size || last_addr < addr)
return NULL;
if (!static_branch_unlikely(&have_mio))
- return (void __iomem *) ioaddr;
+ return (void __iomem *) addr;
- offset = ioaddr & ~PAGE_MASK;
- ioaddr &= PAGE_MASK;
+ offset = addr & ~PAGE_MASK;
+ addr &= PAGE_MASK;
size = PAGE_ALIGN(size + offset);
area = get_vm_area(size, VM_IOREMAP);
if (!area)
return NULL;
- if (ioremap_page_range((unsigned long) area->addr,
- (unsigned long) area->addr + size,
- ioaddr, PAGE_KERNEL)) {
- vunmap(area->addr);
+ vaddr = (unsigned long) area->addr;
+ if (ioremap_page_range(vaddr, vaddr + size, addr, prot)) {
+ free_vm_area(area);
return NULL;
}
return (void __iomem *) ((unsigned long) area->addr + offset);
}
+
+void __iomem *ioremap_prot(phys_addr_t addr, size_t size, unsigned long prot)
+{
+ return __ioremap(addr, size, __pgprot(prot));
+}
+EXPORT_SYMBOL(ioremap_prot);
+
+void __iomem *ioremap(phys_addr_t addr, size_t size)
+{
+ return __ioremap(addr, size, PAGE_KERNEL);
+}
EXPORT_SYMBOL(ioremap);
+void __iomem *ioremap_wc(phys_addr_t addr, size_t size)
+{
+ return __ioremap(addr, size, pgprot_writecombine(PAGE_KERNEL));
+}
+EXPORT_SYMBOL(ioremap_wc);
+
+void __iomem *ioremap_wt(phys_addr_t addr, size_t size)
+{
+ return __ioremap(addr, size, pgprot_writethrough(PAGE_KERNEL));
+}
+EXPORT_SYMBOL(ioremap_wt);
+
void iounmap(volatile void __iomem *addr)
{
if (static_branch_likely(&have_mio))
@@ -294,7 +319,7 @@
void __iomem *pci_iomap_range(struct pci_dev *pdev, int bar,
unsigned long offset, unsigned long max)
{
- if (!pci_resource_len(pdev, bar) || bar >= PCI_BAR_COUNT)
+ if (bar >= PCI_STD_NUM_BARS || !pci_resource_len(pdev, bar))
return NULL;
if (static_branch_likely(&have_mio))
@@ -324,7 +349,7 @@
void __iomem *pci_iomap_wc_range(struct pci_dev *pdev, int bar,
unsigned long offset, unsigned long max)
{
- if (!pci_resource_len(pdev, bar) || bar >= PCI_BAR_COUNT)
+ if (bar >= PCI_STD_NUM_BARS || !pci_resource_len(pdev, bar))
return NULL;
if (static_branch_likely(&have_mio))
@@ -371,29 +396,17 @@
static int pci_read(struct pci_bus *bus, unsigned int devfn, int where,
int size, u32 *val)
{
- struct zpci_dev *zdev = get_zdev_by_bus(bus);
- int ret;
+ struct zpci_dev *zdev = get_zdev_by_bus(bus, devfn);
- if (!zdev || devfn != ZPCI_DEVFN)
- ret = -ENODEV;
- else
- ret = zpci_cfg_load(zdev, where, val, size);
-
- return ret;
+ return (zdev) ? zpci_cfg_load(zdev, where, val, size) : -ENODEV;
}
static int pci_write(struct pci_bus *bus, unsigned int devfn, int where,
int size, u32 val)
{
- struct zpci_dev *zdev = get_zdev_by_bus(bus);
- int ret;
+ struct zpci_dev *zdev = get_zdev_by_bus(bus, devfn);
- if (!zdev || devfn != ZPCI_DEVFN)
- ret = -ENODEV;
- else
- ret = zpci_cfg_store(zdev, where, val, size);
-
- return ret;
+ return (zdev) ? zpci_cfg_store(zdev, where, val, size) : -ENODEV;
}
static struct pci_ops pci_root_ops = {
@@ -401,22 +414,13 @@
.write = pci_write,
};
-#ifdef CONFIG_PCI_IOV
-static struct resource iov_res = {
- .name = "PCI IOV res",
- .start = 0,
- .end = -1,
- .flags = IORESOURCE_MEM,
-};
-#endif
-
static void zpci_map_resources(struct pci_dev *pdev)
{
struct zpci_dev *zdev = to_zpci(pdev);
resource_size_t len;
int i;
- for (i = 0; i < PCI_BAR_COUNT; i++) {
+ for (i = 0; i < PCI_STD_NUM_BARS; i++) {
len = pci_resource_len(pdev, i);
if (!len)
continue;
@@ -430,16 +434,7 @@
pdev->resource[i].end = pdev->resource[i].start + len - 1;
}
-#ifdef CONFIG_PCI_IOV
- for (i = 0; i < PCI_SRIOV_NUM_BARS; i++) {
- int bar = i + PCI_IOV_RESOURCES;
-
- len = pci_resource_len(pdev, bar);
- if (!len)
- continue;
- pdev->resource[bar].parent = &iov_res;
- }
-#endif
+ zpci_iov_map_resources(pdev);
}
static void zpci_unmap_resources(struct pci_dev *pdev)
@@ -451,7 +446,7 @@
if (zpci_use_mio(zdev))
return;
- for (i = 0; i < PCI_BAR_COUNT; i++) {
+ for (i = 0; i < PCI_STD_NUM_BARS; i++) {
len = pci_resource_len(pdev, i);
if (!len)
continue;
@@ -504,17 +499,17 @@
return r;
}
-static int zpci_setup_bus_resources(struct zpci_dev *zdev,
- struct list_head *resources)
+int zpci_setup_bus_resources(struct zpci_dev *zdev,
+ struct list_head *resources)
{
unsigned long addr, size, flags;
struct resource *res;
int i, entry;
snprintf(zdev->res_name, sizeof(zdev->res_name),
- "PCI Bus %04x:%02x", zdev->domain, ZPCI_BUS_NR);
+ "PCI Bus %04x:%02x", zdev->uid, ZPCI_BUS_NR);
- for (i = 0; i < PCI_BAR_COUNT; i++) {
+ for (i = 0; i < PCI_STD_NUM_BARS; i++) {
if (!zdev->bars[i].size)
continue;
entry = zpci_alloc_iomap(zdev);
@@ -551,7 +546,7 @@
{
int i;
- for (i = 0; i < PCI_BAR_COUNT; i++) {
+ for (i = 0; i < PCI_STD_NUM_BARS; i++) {
if (!zdev->bars[i].size || !zdev->bars[i].res)
continue;
@@ -563,9 +558,12 @@
int pcibios_add_device(struct pci_dev *pdev)
{
+ struct zpci_dev *zdev = to_zpci(pdev);
struct resource *res;
int i;
+ /* The pdev has a reference to the zdev via its bus */
+ zpci_zdev_get(zdev);
if (pdev->is_physfn)
pdev->no_vf_scan = 1;
@@ -573,7 +571,7 @@
pdev->dev.dma_ops = &s390_pci_dma_ops;
zpci_map_resources(pdev);
- for (i = 0; i < PCI_BAR_COUNT; i++) {
+ for (i = 0; i < PCI_STD_NUM_BARS; i++) {
res = &pdev->resource[i];
if (res->parent || !res->flags)
continue;
@@ -585,7 +583,10 @@
void pcibios_release_device(struct pci_dev *pdev)
{
+ struct zpci_dev *zdev = to_zpci(pdev);
+
zpci_unmap_resources(pdev);
+ zpci_zdev_put(zdev);
}
int pcibios_enable_device(struct pci_dev *pdev, int mask)
@@ -606,136 +607,62 @@
zpci_debug_exit_device(zdev);
}
-#ifdef CONFIG_HIBERNATE_CALLBACKS
-static int zpci_restore(struct device *dev)
+static int __zpci_register_domain(int domain)
{
- struct pci_dev *pdev = to_pci_dev(dev);
- struct zpci_dev *zdev = to_zpci(pdev);
- int ret = 0;
-
- if (zdev->state != ZPCI_FN_STATE_ONLINE)
- goto out;
-
- ret = clp_enable_fh(zdev, ZPCI_NR_DMA_SPACES);
- if (ret)
- goto out;
-
- zpci_map_resources(pdev);
- zpci_register_ioat(zdev, 0, zdev->start_dma, zdev->end_dma,
- (u64) zdev->dma_table);
-
-out:
- return ret;
+ spin_lock(&zpci_domain_lock);
+ if (test_bit(domain, zpci_domain)) {
+ spin_unlock(&zpci_domain_lock);
+ pr_err("Domain %04x is already assigned\n", domain);
+ return -EEXIST;
+ }
+ set_bit(domain, zpci_domain);
+ spin_unlock(&zpci_domain_lock);
+ return domain;
}
-static int zpci_freeze(struct device *dev)
+static int __zpci_alloc_domain(void)
{
- struct pci_dev *pdev = to_pci_dev(dev);
- struct zpci_dev *zdev = to_zpci(pdev);
+ int domain;
- if (zdev->state != ZPCI_FN_STATE_ONLINE)
- return 0;
-
- zpci_unregister_ioat(zdev, 0);
- zpci_unmap_resources(pdev);
- return clp_disable_fh(zdev);
+ spin_lock(&zpci_domain_lock);
+ /*
+ * We can always auto allocate domains below ZPCI_NR_DEVICES.
+ * There is either a free domain or we have reached the maximum in
+ * which case we would have bailed earlier.
+ */
+ domain = find_first_zero_bit(zpci_domain, ZPCI_NR_DEVICES);
+ set_bit(domain, zpci_domain);
+ spin_unlock(&zpci_domain_lock);
+ return domain;
}
-struct dev_pm_ops pcibios_pm_ops = {
- .thaw_noirq = zpci_restore,
- .freeze_noirq = zpci_freeze,
- .restore_noirq = zpci_restore,
- .poweroff_noirq = zpci_freeze,
-};
-#endif /* CONFIG_HIBERNATE_CALLBACKS */
-
-static int zpci_alloc_domain(struct zpci_dev *zdev)
+int zpci_alloc_domain(int domain)
{
if (zpci_unique_uid) {
- zdev->domain = (u16) zdev->uid;
- if (zdev->domain >= ZPCI_NR_DEVICES)
- return 0;
-
- spin_lock(&zpci_domain_lock);
- if (test_bit(zdev->domain, zpci_domain)) {
- spin_unlock(&zpci_domain_lock);
- return -EEXIST;
- }
- set_bit(zdev->domain, zpci_domain);
- spin_unlock(&zpci_domain_lock);
- return 0;
+ if (domain)
+ return __zpci_register_domain(domain);
+ pr_warn("UID checking was active but no UID is provided: switching to automatic domain allocation\n");
+ update_uid_checking(false);
}
-
- spin_lock(&zpci_domain_lock);
- zdev->domain = find_first_zero_bit(zpci_domain, ZPCI_NR_DEVICES);
- if (zdev->domain == ZPCI_NR_DEVICES) {
- spin_unlock(&zpci_domain_lock);
- return -ENOSPC;
- }
- set_bit(zdev->domain, zpci_domain);
- spin_unlock(&zpci_domain_lock);
- return 0;
+ return __zpci_alloc_domain();
}
-static void zpci_free_domain(struct zpci_dev *zdev)
+void zpci_free_domain(int domain)
{
- if (zdev->domain >= ZPCI_NR_DEVICES)
- return;
-
spin_lock(&zpci_domain_lock);
- clear_bit(zdev->domain, zpci_domain);
+ clear_bit(domain, zpci_domain);
spin_unlock(&zpci_domain_lock);
}
-void pcibios_remove_bus(struct pci_bus *bus)
-{
- struct zpci_dev *zdev = get_zdev_by_bus(bus);
-
- zpci_exit_slot(zdev);
- zpci_cleanup_bus_resources(zdev);
- zpci_destroy_iommu(zdev);
- zpci_free_domain(zdev);
-
- spin_lock(&zpci_list_lock);
- list_del(&zdev->entry);
- spin_unlock(&zpci_list_lock);
-
- zpci_dbg(3, "rem fid:%x\n", zdev->fid);
- kfree(zdev);
-}
-
-static int zpci_scan_bus(struct zpci_dev *zdev)
-{
- LIST_HEAD(resources);
- int ret;
-
- ret = zpci_setup_bus_resources(zdev, &resources);
- if (ret)
- goto error;
-
- zdev->bus = pci_scan_root_bus(NULL, ZPCI_BUS_NR, &pci_root_ops,
- zdev, &resources);
- if (!zdev->bus) {
- ret = -EIO;
- goto error;
- }
- zdev->bus->max_bus_speed = zdev->max_bus_speed;
- pci_bus_add_devices(zdev->bus);
- return 0;
-
-error:
- zpci_cleanup_bus_resources(zdev);
- pci_free_resource_list(&resources);
- return ret;
-}
int zpci_enable_device(struct zpci_dev *zdev)
{
int rc;
- rc = clp_enable_fh(zdev, ZPCI_NR_DMA_SPACES);
- if (rc)
+ if (clp_enable_fh(zdev, ZPCI_NR_DMA_SPACES)) {
+ rc = -EIO;
goto out;
+ }
rc = zpci_dma_init_device(zdev);
if (rc)
@@ -754,58 +681,175 @@
int zpci_disable_device(struct zpci_dev *zdev)
{
zpci_dma_exit_device(zdev);
- return clp_disable_fh(zdev);
+ /*
+ * The zPCI function may already be disabled by the platform, this is
+ * detected in clp_disable_fh() which becomes a no-op.
+ */
+ return clp_disable_fh(zdev) ? -EIO : 0;
}
EXPORT_SYMBOL_GPL(zpci_disable_device);
-int zpci_create_device(struct zpci_dev *zdev)
+/* zpci_remove_device - Removes the given zdev from the PCI core
+ * @zdev: the zdev to be removed from the PCI core
+ * @set_error: if true the device's error state is set to permanent failure
+ *
+ * Sets a zPCI device to a configured but offline state; the zPCI
+ * device is still accessible through its hotplug slot and the zPCI
+ * API but is removed from the common code PCI bus, making it
+ * no longer available to drivers.
+ */
+void zpci_remove_device(struct zpci_dev *zdev, bool set_error)
{
+ struct zpci_bus *zbus = zdev->zbus;
+ struct pci_dev *pdev;
+
+ if (!zdev->zbus->bus)
+ return;
+
+ pdev = pci_get_slot(zbus->bus, zdev->devfn);
+ if (pdev) {
+ if (set_error)
+ pdev->error_state = pci_channel_io_perm_failure;
+ if (pdev->is_virtfn) {
+ zpci_iov_remove_virtfn(pdev, zdev->vfn);
+ /* balance pci_get_slot */
+ pci_dev_put(pdev);
+ return;
+ }
+ pci_stop_and_remove_bus_device_locked(pdev);
+ /* balance pci_get_slot */
+ pci_dev_put(pdev);
+ }
+}
+
+/**
+ * zpci_create_device() - Create a new zpci_dev and add it to the zbus
+ * @fid: Function ID of the device to be created
+ * @fh: Current Function Handle of the device to be created
+ * @state: Initial state after creation either Standby or Configured
+ *
+ * Creates a new zpci device and adds it to its, possibly newly created, zbus
+ * as well as zpci_list.
+ *
+ * Returns: 0 on success, an error value otherwise
+ */
+int zpci_create_device(u32 fid, u32 fh, enum zpci_state state)
+{
+ struct zpci_dev *zdev;
int rc;
- rc = zpci_alloc_domain(zdev);
+ zpci_dbg(3, "add fid:%x, fh:%x, c:%d\n", fid, fh, state);
+ zdev = kzalloc(sizeof(*zdev), GFP_KERNEL);
+ if (!zdev)
+ return -ENOMEM;
+
+ /* FID and Function Handle are the static/dynamic identifiers */
+ zdev->fid = fid;
+ zdev->fh = fh;
+
+ /* Query function properties and update zdev */
+ rc = clp_query_pci_fn(zdev);
if (rc)
- goto out;
+ goto error;
+ zdev->state = state;
+
+ kref_init(&zdev->kref);
+ mutex_init(&zdev->lock);
rc = zpci_init_iommu(zdev);
if (rc)
- goto out_free;
+ goto error;
- mutex_init(&zdev->lock);
if (zdev->state == ZPCI_FN_STATE_CONFIGURED) {
rc = zpci_enable_device(zdev);
if (rc)
- goto out_destroy_iommu;
+ goto error_destroy_iommu;
}
- rc = zpci_scan_bus(zdev);
+
+ rc = zpci_bus_device_register(zdev, &pci_root_ops);
if (rc)
- goto out_disable;
+ goto error_disable;
spin_lock(&zpci_list_lock);
list_add_tail(&zdev->entry, &zpci_list);
spin_unlock(&zpci_list_lock);
- zpci_init_slot(zdev);
-
return 0;
-out_disable:
+error_disable:
if (zdev->state == ZPCI_FN_STATE_ONLINE)
zpci_disable_device(zdev);
-out_destroy_iommu:
+error_destroy_iommu:
zpci_destroy_iommu(zdev);
-out_free:
- zpci_free_domain(zdev);
-out:
+error:
+ zpci_dbg(0, "add fid:%x, rc:%d\n", fid, rc);
+ kfree(zdev);
return rc;
}
-void zpci_remove_device(struct zpci_dev *zdev)
+bool zpci_is_device_configured(struct zpci_dev *zdev)
{
- if (!zdev->bus)
- return;
+ enum zpci_state state = zdev->state;
- pci_stop_root_bus(zdev->bus);
- pci_remove_root_bus(zdev->bus);
+ return state != ZPCI_FN_STATE_RESERVED &&
+ state != ZPCI_FN_STATE_STANDBY;
+}
+
+/**
+ * zpci_device_reserved() - Mark device as resverved
+ * @zdev: the zpci_dev that was reserved
+ *
+ * Handle the case that a given zPCI function was reserved by another system.
+ * After a call to this function the zpci_dev can not be found via
+ * get_zdev_by_fid() anymore but may still be accessible via existing
+ * references though it will not be functional anymore.
+ */
+void zpci_device_reserved(struct zpci_dev *zdev)
+{
+ if (zdev->has_hp_slot)
+ zpci_exit_slot(zdev);
+ /*
+ * Remove device from zpci_list as it is going away. This also
+ * makes sure we ignore subsequent zPCI events for this device.
+ */
+ spin_lock(&zpci_list_lock);
+ list_del(&zdev->entry);
+ spin_unlock(&zpci_list_lock);
+ zdev->state = ZPCI_FN_STATE_RESERVED;
+ zpci_dbg(3, "rsv fid:%x\n", zdev->fid);
+ zpci_zdev_put(zdev);
+}
+
+void zpci_release_device(struct kref *kref)
+{
+ struct zpci_dev *zdev = container_of(kref, struct zpci_dev, kref);
+
+ if (zdev->zbus->bus)
+ zpci_remove_device(zdev, false);
+
+ switch (zdev->state) {
+ case ZPCI_FN_STATE_ONLINE:
+ case ZPCI_FN_STATE_CONFIGURED:
+ zpci_disable_device(zdev);
+ fallthrough;
+ case ZPCI_FN_STATE_STANDBY:
+ if (zdev->has_hp_slot)
+ zpci_exit_slot(zdev);
+ spin_lock(&zpci_list_lock);
+ list_del(&zdev->entry);
+ spin_unlock(&zpci_list_lock);
+ zpci_dbg(3, "rsv fid:%x\n", zdev->fid);
+ fallthrough;
+ case ZPCI_FN_STATE_RESERVED:
+ zpci_cleanup_bus_resources(zdev);
+ zpci_bus_device_unregister(zdev);
+ zpci_destroy_iommu(zdev);
+ fallthrough;
+ default:
+ break;
+ }
+ zpci_dbg(3, "rem fid:%x\n", zdev->fid);
+ kfree(zdev);
}
int zpci_report_error(struct pci_dev *pdev,
@@ -837,6 +881,9 @@
if (!zpci_iomap_bitmap)
goto error_iomap_bitmap;
+ if (static_branch_likely(&have_mio))
+ clp_setup_writeback_mio();
+
return 0;
error_iomap_bitmap:
kfree(zpci_iomap_start);
@@ -871,6 +918,10 @@
s390_pci_force_floating = 1;
return NULL;
}
+ if (!strcmp(str, "norid")) {
+ s390_pci_no_rid = 1;
+ return NULL;
+ }
return str;
}
@@ -929,9 +980,3 @@
return rc;
}
subsys_initcall_sync(pci_base_init);
-
-void zpci_rescan(void)
-{
- if (zpci_is_enabled())
- clp_rescan_pci_devices_simple(NULL);
-}
diff --git a/arch/s390/pci/pci_bus.c b/arch/s390/pci/pci_bus.c
new file mode 100644
index 0000000..755b46f
--- /dev/null
+++ b/arch/s390/pci/pci_bus.c
@@ -0,0 +1,275 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright IBM Corp. 2020
+ *
+ * Author(s):
+ * Pierre Morel <pmorel@linux.ibm.com>
+ *
+ */
+
+#define KMSG_COMPONENT "zpci"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <linux/export.h>
+#include <linux/delay.h>
+#include <linux/seq_file.h>
+#include <linux/jump_label.h>
+#include <linux/pci.h>
+#include <linux/printk.h>
+
+#include <asm/pci_clp.h>
+#include <asm/pci_dma.h>
+
+#include "pci_bus.h"
+#include "pci_iov.h"
+
+static LIST_HEAD(zbus_list);
+static DEFINE_SPINLOCK(zbus_list_lock);
+static int zpci_nb_devices;
+
+/* zpci_bus_scan
+ * @zbus: the zbus holding the zdevices
+ * @ops: the pci operations
+ *
+ * The domain number must be set before pci_scan_root_bus is called.
+ * This function can be called once the domain is known, hence
+ * when the function_0 is dicovered.
+ */
+static int zpci_bus_scan(struct zpci_bus *zbus, int domain, struct pci_ops *ops)
+{
+ struct pci_bus *bus;
+ int rc;
+
+ rc = zpci_alloc_domain(domain);
+ if (rc < 0)
+ return rc;
+ zbus->domain_nr = rc;
+
+ bus = pci_scan_root_bus(NULL, ZPCI_BUS_NR, ops, zbus, &zbus->resources);
+ if (!bus) {
+ zpci_free_domain(zbus->domain_nr);
+ return -EFAULT;
+ }
+
+ zbus->bus = bus;
+ pci_bus_add_devices(bus);
+ return 0;
+}
+
+static void zpci_bus_release(struct kref *kref)
+{
+ struct zpci_bus *zbus = container_of(kref, struct zpci_bus, kref);
+
+ if (zbus->bus) {
+ pci_lock_rescan_remove();
+ pci_stop_root_bus(zbus->bus);
+
+ zpci_free_domain(zbus->domain_nr);
+ pci_free_resource_list(&zbus->resources);
+
+ pci_remove_root_bus(zbus->bus);
+ pci_unlock_rescan_remove();
+ }
+
+ spin_lock(&zbus_list_lock);
+ list_del(&zbus->bus_next);
+ spin_unlock(&zbus_list_lock);
+ kfree(zbus);
+}
+
+static void zpci_bus_put(struct zpci_bus *zbus)
+{
+ kref_put(&zbus->kref, zpci_bus_release);
+}
+
+static struct zpci_bus *zpci_bus_get(int pchid)
+{
+ struct zpci_bus *zbus;
+
+ spin_lock(&zbus_list_lock);
+ list_for_each_entry(zbus, &zbus_list, bus_next) {
+ if (pchid == zbus->pchid) {
+ kref_get(&zbus->kref);
+ goto out_unlock;
+ }
+ }
+ zbus = NULL;
+out_unlock:
+ spin_unlock(&zbus_list_lock);
+ return zbus;
+}
+
+static struct zpci_bus *zpci_bus_alloc(int pchid)
+{
+ struct zpci_bus *zbus;
+
+ zbus = kzalloc(sizeof(*zbus), GFP_KERNEL);
+ if (!zbus)
+ return NULL;
+
+ zbus->pchid = pchid;
+ INIT_LIST_HEAD(&zbus->bus_next);
+ spin_lock(&zbus_list_lock);
+ list_add_tail(&zbus->bus_next, &zbus_list);
+ spin_unlock(&zbus_list_lock);
+
+ kref_init(&zbus->kref);
+ INIT_LIST_HEAD(&zbus->resources);
+
+ zbus->bus_resource.start = 0;
+ zbus->bus_resource.end = ZPCI_BUS_NR;
+ zbus->bus_resource.flags = IORESOURCE_BUS;
+ pci_add_resource(&zbus->resources, &zbus->bus_resource);
+
+ return zbus;
+}
+
+void pcibios_bus_add_device(struct pci_dev *pdev)
+{
+ struct zpci_dev *zdev = to_zpci(pdev);
+
+ /*
+ * With pdev->no_vf_scan the common PCI probing code does not
+ * perform PF/VF linking.
+ */
+ if (zdev->vfn) {
+ zpci_iov_setup_virtfn(zdev->zbus, pdev, zdev->vfn);
+ pdev->no_command_memory = 1;
+ }
+}
+
+static int zpci_bus_add_device(struct zpci_bus *zbus, struct zpci_dev *zdev)
+{
+ struct pci_bus *bus;
+ struct resource_entry *window, *n;
+ struct resource *res;
+ struct pci_dev *pdev;
+ int rc;
+
+ bus = zbus->bus;
+ if (!bus)
+ return -EINVAL;
+
+ pdev = pci_get_slot(bus, zdev->devfn);
+ if (pdev) {
+ /* Device is already known. */
+ pci_dev_put(pdev);
+ return 0;
+ }
+
+ rc = zpci_init_slot(zdev);
+ if (rc)
+ return rc;
+ zdev->has_hp_slot = 1;
+
+ resource_list_for_each_entry_safe(window, n, &zbus->resources) {
+ res = window->res;
+ pci_bus_add_resource(bus, res, 0);
+ }
+
+ pdev = pci_scan_single_device(bus, zdev->devfn);
+ if (pdev)
+ pci_bus_add_device(pdev);
+
+ return 0;
+}
+
+static void zpci_bus_add_devices(struct zpci_bus *zbus)
+{
+ int i;
+
+ for (i = 1; i < ZPCI_FUNCTIONS_PER_BUS; i++)
+ if (zbus->function[i])
+ zpci_bus_add_device(zbus, zbus->function[i]);
+
+ pci_lock_rescan_remove();
+ pci_bus_add_devices(zbus->bus);
+ pci_unlock_rescan_remove();
+}
+
+int zpci_bus_device_register(struct zpci_dev *zdev, struct pci_ops *ops)
+{
+ struct zpci_bus *zbus = NULL;
+ int rc = -EBADF;
+
+ if (zpci_nb_devices == ZPCI_NR_DEVICES) {
+ pr_warn("Adding PCI function %08x failed because the configured limit of %d is reached\n",
+ zdev->fid, ZPCI_NR_DEVICES);
+ return -ENOSPC;
+ }
+ zpci_nb_devices++;
+
+ if (zdev->devfn >= ZPCI_FUNCTIONS_PER_BUS)
+ return -EINVAL;
+
+ if (!s390_pci_no_rid && zdev->rid_available)
+ zbus = zpci_bus_get(zdev->pchid);
+
+ if (!zbus) {
+ zbus = zpci_bus_alloc(zdev->pchid);
+ if (!zbus)
+ return -ENOMEM;
+ }
+
+ zdev->zbus = zbus;
+ if (zbus->function[zdev->devfn]) {
+ pr_err("devfn %04x is already assigned\n", zdev->devfn);
+ goto error; /* rc already set */
+ }
+ zbus->function[zdev->devfn] = zdev;
+
+ zpci_setup_bus_resources(zdev, &zbus->resources);
+
+ if (zbus->bus) {
+ if (!zbus->multifunction) {
+ WARN_ONCE(1, "zbus is not multifunction\n");
+ goto error_bus;
+ }
+ if (!zdev->rid_available) {
+ WARN_ONCE(1, "rid_available not set for multifunction\n");
+ goto error_bus;
+ }
+ rc = zpci_bus_add_device(zbus, zdev);
+ if (rc)
+ goto error_bus;
+ } else if (zdev->devfn == 0) {
+ if (zbus->multifunction && !zdev->rid_available) {
+ WARN_ONCE(1, "rid_available not set on function 0 for multifunction\n");
+ goto error_bus;
+ }
+ rc = zpci_bus_scan(zbus, (u16)zdev->uid, ops);
+ if (rc)
+ goto error_bus;
+ zpci_bus_add_devices(zbus);
+ rc = zpci_init_slot(zdev);
+ if (rc)
+ goto error_bus;
+ zdev->has_hp_slot = 1;
+ zbus->multifunction = zdev->rid_available;
+ zbus->max_bus_speed = zdev->max_bus_speed;
+ } else {
+ zbus->multifunction = 1;
+ }
+
+ return 0;
+
+error_bus:
+ zpci_nb_devices--;
+ zbus->function[zdev->devfn] = NULL;
+error:
+ pr_err("Adding PCI function %08x failed\n", zdev->fid);
+ zpci_bus_put(zbus);
+ return rc;
+}
+
+void zpci_bus_device_unregister(struct zpci_dev *zdev)
+{
+ struct zpci_bus *zbus = zdev->zbus;
+
+ zpci_nb_devices--;
+ zbus->function[zdev->devfn] = NULL;
+ zpci_bus_put(zbus);
+}
diff --git a/arch/s390/pci/pci_bus.h b/arch/s390/pci/pci_bus.h
new file mode 100644
index 0000000..55c9488
--- /dev/null
+++ b/arch/s390/pci/pci_bus.h
@@ -0,0 +1,36 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright IBM Corp. 2020
+ *
+ * Author(s):
+ * Pierre Morel <pmorel@linux.ibm.com>
+ *
+ */
+
+int zpci_bus_device_register(struct zpci_dev *zdev, struct pci_ops *ops);
+void zpci_bus_device_unregister(struct zpci_dev *zdev);
+
+void zpci_release_device(struct kref *kref);
+static inline void zpci_zdev_put(struct zpci_dev *zdev)
+{
+ kref_put(&zdev->kref, zpci_release_device);
+}
+
+static inline void zpci_zdev_get(struct zpci_dev *zdev)
+{
+ kref_get(&zdev->kref);
+}
+
+int zpci_alloc_domain(int domain);
+void zpci_free_domain(int domain);
+int zpci_setup_bus_resources(struct zpci_dev *zdev,
+ struct list_head *resources);
+
+static inline struct zpci_dev *get_zdev_by_bus(struct pci_bus *bus,
+ unsigned int devfn)
+{
+ struct zpci_bus *zbus = bus->sysdata;
+
+ return (devfn >= ZPCI_FUNCTIONS_PER_BUS) ? NULL : zbus->function[devfn];
+}
+
diff --git a/arch/s390/pci/pci_clp.c b/arch/s390/pci/pci_clp.c
index 20e093f..0a0e8b8 100644
--- a/arch/s390/pci/pci_clp.c
+++ b/arch/s390/pci/pci_clp.c
@@ -24,7 +24,7 @@
bool zpci_unique_uid;
-static void update_uid_checking(bool new)
+void update_uid_checking(bool new)
{
if (zpci_unique_uid != new)
zpci_dbg(1, "uid checking:%d\n", new);
@@ -102,6 +102,7 @@
zdev->msi_addr = response->msia;
zdev->max_msi = response->noi;
zdev->fmb_update = response->mui;
+ zdev->version = response->version;
switch (response->version) {
case 1:
@@ -145,7 +146,7 @@
{
int i;
- for (i = 0; i < PCI_BAR_COUNT; i++) {
+ for (i = 0; i < PCI_STD_NUM_BARS; i++) {
zdev->bars[i].val = le32_to_cpu(response->bar[i]);
zdev->bars[i].size = response->bar_size[i];
}
@@ -155,17 +156,23 @@
zdev->pfgid = response->pfgid;
zdev->pft = response->pft;
zdev->vfn = response->vfn;
+ zdev->port = response->port;
zdev->uid = response->uid;
zdev->fmb_length = sizeof(u32) * response->fmb_len;
+ zdev->rid_available = response->rid_avail;
+ zdev->is_physfn = response->is_physfn;
+ if (!s390_pci_no_rid && zdev->rid_available)
+ zdev->devfn = response->rid & ZPCI_RID_MASK_DEVFN;
memcpy(zdev->pfip, response->pfip, sizeof(zdev->pfip));
if (response->util_str_avail) {
memcpy(zdev->util_str, response->util_str,
sizeof(zdev->util_str));
+ zdev->util_str_avail = 1;
}
zdev->mio_capable = response->mio_addr_avail;
- for (i = 0; i < PCI_BAR_COUNT; i++) {
- if (!(response->mio.valid & (1 << (PCI_BAR_COUNT - i - 1))))
+ for (i = 0; i < PCI_STD_NUM_BARS; i++) {
+ if (!(response->mio.valid & (1 << (PCI_STD_NUM_BARS - i - 1))))
continue;
zdev->bars[i].mio_wb = (void __iomem *) response->mio.addr[i].wb;
@@ -174,7 +181,7 @@
return 0;
}
-static int clp_query_pci_fn(struct zpci_dev *zdev, u32 fh)
+int clp_query_pci_fn(struct zpci_dev *zdev)
{
struct clp_req_rsp_query_pci *rrb;
int rc;
@@ -187,7 +194,7 @@
rrb->request.hdr.len = sizeof(rrb->request);
rrb->request.hdr.cmd = CLP_QUERY_PCI_FN;
rrb->response.hdr.len = sizeof(rrb->response);
- rrb->request.fh = fh;
+ rrb->request.fh = zdev->fh;
rc = clp_req(rrb, CLP_LPS_PCI);
if (!rc && rrb->response.hdr.rsp == CLP_RC_OK) {
@@ -205,49 +212,20 @@
return rc;
}
-int clp_add_pci_device(u32 fid, u32 fh, int configured)
-{
- struct zpci_dev *zdev;
- int rc = -ENOMEM;
-
- zpci_dbg(3, "add fid:%x, fh:%x, c:%d\n", fid, fh, configured);
- zdev = kzalloc(sizeof(*zdev), GFP_KERNEL);
- if (!zdev)
- goto error;
-
- zdev->fh = fh;
- zdev->fid = fid;
-
- /* Query function properties and update zdev */
- rc = clp_query_pci_fn(zdev, fh);
- if (rc)
- goto error;
-
- if (configured)
- zdev->state = ZPCI_FN_STATE_CONFIGURED;
- else
- zdev->state = ZPCI_FN_STATE_STANDBY;
-
- rc = zpci_create_device(zdev);
- if (rc)
- goto error;
- return 0;
-
-error:
- zpci_dbg(0, "add fid:%x, rc:%d\n", fid, rc);
- kfree(zdev);
- return rc;
-}
-
-/*
- * Enable/Disable a given PCI function and update its function handle if
- * necessary
+static int clp_refresh_fh(u32 fid);
+/**
+ * clp_set_pci_fn() - Execute a command on a PCI function
+ * @zdev: Function that will be affected
+ * @nr_dma_as: DMA address space number
+ * @command: The command code to execute
+ *
+ * Returns: 0 on success, < 0 for Linux errors (e.g. -ENOMEM), and
+ * > 0 for non-success platform responses
*/
static int clp_set_pci_fn(struct zpci_dev *zdev, u8 nr_dma_as, u8 command)
{
struct clp_req_rsp_set_pci *rrb;
int rc, retries = 100;
- u32 fid = zdev->fid;
rrb = clp_alloc_block(GFP_KERNEL);
if (!rrb)
@@ -271,17 +249,50 @@
}
} while (rrb->response.hdr.rsp == CLP_RC_SETPCIFN_BUSY);
- if (rc || rrb->response.hdr.rsp != CLP_RC_OK) {
- zpci_err("Set PCI FN:\n");
- zpci_err_clp(rrb->response.hdr.rsp, rc);
- }
-
if (!rc && rrb->response.hdr.rsp == CLP_RC_OK) {
zdev->fh = rrb->response.fh;
- } else if (!rc && rrb->response.hdr.rsp == CLP_RC_SETPCIFN_ALRDY &&
- rrb->response.fh == 0) {
+ } else if (!rc && rrb->response.hdr.rsp == CLP_RC_SETPCIFN_ALRDY) {
/* Function is already in desired state - update handle */
- rc = clp_rescan_pci_devices_simple(&fid);
+ rc = clp_refresh_fh(zdev->fid);
+ } else {
+ zpci_err("Set PCI FN:\n");
+ zpci_err_clp(rrb->response.hdr.rsp, rc);
+ if (!rc)
+ rc = rrb->response.hdr.rsp;
+ }
+ clp_free_block(rrb);
+ return rc;
+}
+
+int clp_setup_writeback_mio(void)
+{
+ struct clp_req_rsp_slpc_pci *rrb;
+ u8 wb_bit_pos;
+ int rc;
+
+ rrb = clp_alloc_block(GFP_KERNEL);
+ if (!rrb)
+ return -ENOMEM;
+
+ memset(rrb, 0, sizeof(*rrb));
+ rrb->request.hdr.len = sizeof(rrb->request);
+ rrb->request.hdr.cmd = CLP_SLPC;
+ rrb->response.hdr.len = sizeof(rrb->response);
+
+ rc = clp_req(rrb, CLP_LPS_PCI);
+ if (!rc && rrb->response.hdr.rsp == CLP_RC_OK) {
+ if (rrb->response.vwb) {
+ wb_bit_pos = rrb->response.mio_wb;
+ set_bit_inv(wb_bit_pos, &mio_wb_bit_mask);
+ zpci_dbg(3, "wb bit: %d\n", wb_bit_pos);
+ } else {
+ zpci_dbg(3, "wb bit: n.a.\n");
+ }
+
+ } else {
+ zpci_err("SLPC PCI:\n");
+ zpci_err_clp(rrb->response.hdr.rsp, rc);
+ rc = -EIO;
}
clp_free_block(rrb);
return rc;
@@ -293,17 +304,13 @@
rc = clp_set_pci_fn(zdev, nr_dma_as, CLP_SET_ENABLE_PCI_FN);
zpci_dbg(3, "ena fid:%x, fh:%x, rc:%d\n", zdev->fid, zdev->fh, rc);
- if (rc)
- goto out;
-
- if (zpci_use_mio(zdev)) {
+ if (!rc && zpci_use_mio(zdev)) {
rc = clp_set_pci_fn(zdev, nr_dma_as, CLP_SET_ENABLE_MIO);
zpci_dbg(3, "ena mio fid:%x, fh:%x, rc:%d\n",
zdev->fid, zdev->fh, rc);
if (rc)
clp_disable_fh(zdev);
}
-out:
return rc;
}
@@ -366,25 +373,7 @@
zdev = get_zdev_by_fid(entry->fid);
if (!zdev)
- clp_add_pci_device(entry->fid, entry->fh, entry->config_state);
-}
-
-static void __clp_update(struct clp_fh_list_entry *entry, void *data)
-{
- struct zpci_dev *zdev;
- u32 *fid = data;
-
- if (!entry->vendor_id)
- return;
-
- if (fid && *fid != entry->fid)
- return;
-
- zdev = get_zdev_by_fid(entry->fid);
- if (!zdev)
- return;
-
- zdev->fh = entry->fh;
+ zpci_create_device(entry->fid, entry->fh, entry->config_state);
}
int clp_scan_pci_devices(void)
@@ -402,27 +391,25 @@
return rc;
}
-int clp_rescan_pci_devices(void)
+static void __clp_refresh_fh(struct clp_fh_list_entry *entry, void *data)
{
- struct clp_req_rsp_list_pci *rrb;
- int rc;
+ struct zpci_dev *zdev;
+ u32 fid = *((u32 *)data);
- zpci_remove_reserved_devices();
+ if (!entry->vendor_id || fid != entry->fid)
+ return;
- rrb = clp_alloc_block(GFP_KERNEL);
- if (!rrb)
- return -ENOMEM;
+ zdev = get_zdev_by_fid(fid);
+ if (!zdev)
+ return;
- rc = clp_list_pci(rrb, NULL, __clp_add);
-
- clp_free_block(rrb);
- return rc;
+ zdev->fh = entry->fh;
}
-/* Rescan PCI functions and refresh function handles. If fid is non-NULL only
- * refresh the handle of the function matching @fid
+/*
+ * Refresh the function handle of the function matching @fid
*/
-int clp_rescan_pci_devices_simple(u32 *fid)
+static int clp_refresh_fh(u32 fid)
{
struct clp_req_rsp_list_pci *rrb;
int rc;
@@ -431,7 +418,7 @@
if (!rrb)
return -ENOMEM;
- rc = clp_list_pci(rrb, fid, __clp_update);
+ rc = clp_list_pci(rrb, &fid, __clp_refresh_fh);
clp_free_block(rrb);
return rc;
@@ -490,7 +477,7 @@
}
}
-static int clp_pci_slpc(struct clp_req *req, struct clp_req_rsp_slpc *lpcb)
+static int clp_pci_slpc(struct clp_req *req, struct clp_req_rsp_slpc_pci *lpcb)
{
unsigned long limit = PAGE_SIZE - sizeof(lpcb->request);
diff --git a/arch/s390/pci/pci_dma.c b/arch/s390/pci/pci_dma.c
index 64b1399..ebc9a49 100644
--- a/arch/s390/pci/pci_dma.c
+++ b/arch/s390/pci/pci_dma.c
@@ -10,7 +10,7 @@
#include <linux/slab.h>
#include <linux/export.h>
#include <linux/iommu-helper.h>
-#include <linux/dma-mapping.h>
+#include <linux/dma-map-ops.h>
#include <linux/vmalloc.h>
#include <linux/pci.h>
#include <asm/pci_dma.h>
@@ -261,13 +261,11 @@
unsigned long start, int size)
{
struct zpci_dev *zdev = to_zpci(to_pci_dev(dev));
- unsigned long boundary_size;
- boundary_size = ALIGN(dma_get_seg_boundary(dev) + 1,
- PAGE_SIZE) >> PAGE_SHIFT;
return iommu_area_alloc(zdev->iommu_bitmap, zdev->iommu_pages,
start, size, zdev->start_dma >> PAGE_SHIFT,
- boundary_size, 0);
+ dma_get_seg_boundary_nr_pages(dev, PAGE_SHIFT),
+ 0);
}
static dma_addr_t dma_alloc_address(struct device *dev, int size)
@@ -670,6 +668,8 @@
.unmap_page = s390_dma_unmap_pages,
.mmap = dma_common_mmap,
.get_sgtable = dma_common_get_sgtable,
+ .alloc_pages = dma_common_alloc_pages,
+ .free_pages = dma_common_free_pages,
/* dma_supported is unconditionally true without a callback */
};
EXPORT_SYMBOL_GPL(s390_pci_dma_ops);
diff --git a/arch/s390/pci/pci_event.c b/arch/s390/pci/pci_event.c
index 8d6ee4a..b7cfde7 100644
--- a/arch/s390/pci/pci_event.c
+++ b/arch/s390/pci/pci_event.c
@@ -14,6 +14,8 @@
#include <asm/pci_debug.h>
#include <asm/sclp.h>
+#include "pci_bus.h"
+
/* Content Code Description for PCI Function Error */
struct zpci_ccdf_err {
u32 reserved1;
@@ -53,7 +55,7 @@
zpci_err_hex(ccdf, sizeof(*ccdf));
if (zdev)
- pdev = pci_get_slot(zdev->bus, ZPCI_DEVFN);
+ pdev = pci_get_slot(zdev->zbus->bus, zdev->devfn);
pr_err("%s: Event 0x%x reports an error for PCI function 0x%x\n",
pdev ? pci_name(pdev) : "n/a", ccdf->pec, ccdf->fid);
@@ -74,46 +76,52 @@
static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
{
struct zpci_dev *zdev = get_zdev_by_fid(ccdf->fid);
- struct pci_dev *pdev = NULL;
enum zpci_state state;
+ struct pci_dev *pdev;
int ret;
- if (zdev)
- pdev = pci_get_slot(zdev->bus, ZPCI_DEVFN);
-
- pr_info("%s: Event 0x%x reconfigured PCI function 0x%x\n",
- pdev ? pci_name(pdev) : "n/a", ccdf->pec, ccdf->fid);
zpci_err("avail CCDF:\n");
zpci_err_hex(ccdf, sizeof(*ccdf));
switch (ccdf->pec) {
case 0x0301: /* Reserved|Standby -> Configured */
if (!zdev) {
- ret = clp_add_pci_device(ccdf->fid, ccdf->fh, 0);
- if (ret)
- break;
- zdev = get_zdev_by_fid(ccdf->fid);
- }
- if (!zdev || zdev->state != ZPCI_FN_STATE_STANDBY)
+ zpci_create_device(ccdf->fid, ccdf->fh, ZPCI_FN_STATE_CONFIGURED);
break;
- zdev->state = ZPCI_FN_STATE_CONFIGURED;
+ }
+ /* the configuration request may be stale */
+ if (zdev->state != ZPCI_FN_STATE_STANDBY)
+ break;
zdev->fh = ccdf->fh;
+ zdev->state = ZPCI_FN_STATE_CONFIGURED;
ret = zpci_enable_device(zdev);
if (ret)
break;
+
+ /* the PCI function will be scanned once function 0 appears */
+ if (!zdev->zbus->bus)
+ break;
+
+ pdev = pci_scan_single_device(zdev->zbus->bus, zdev->devfn);
+ if (!pdev)
+ break;
+
+ pci_bus_add_device(pdev);
pci_lock_rescan_remove();
- pci_rescan_bus(zdev->bus);
+ pci_bus_add_devices(zdev->zbus->bus);
pci_unlock_rescan_remove();
break;
case 0x0302: /* Reserved -> Standby */
- if (!zdev)
- clp_add_pci_device(ccdf->fid, ccdf->fh, 0);
+ if (!zdev) {
+ zpci_create_device(ccdf->fid, ccdf->fh, ZPCI_FN_STATE_STANDBY);
+ break;
+ }
+ zdev->fh = ccdf->fh;
break;
case 0x0303: /* Deconfiguration requested */
if (!zdev)
break;
- if (pdev)
- pci_stop_and_remove_bus_device_locked(pdev);
+ zpci_remove_device(zdev, false);
ret = zpci_disable_device(zdev);
if (ret)
@@ -128,33 +136,31 @@
case 0x0304: /* Configured -> Standby|Reserved */
if (!zdev)
break;
- if (pdev) {
- /* Give the driver a hint that the function is
- * already unusable. */
- pdev->error_state = pci_channel_io_perm_failure;
- pci_stop_and_remove_bus_device_locked(pdev);
- }
+ /* Give the driver a hint that the function is
+ * already unusable.
+ */
+ zpci_remove_device(zdev, true);
zdev->fh = ccdf->fh;
zpci_disable_device(zdev);
zdev->state = ZPCI_FN_STATE_STANDBY;
if (!clp_get_state(ccdf->fid, &state) &&
state == ZPCI_FN_STATE_RESERVED) {
- zpci_remove_device(zdev);
+ zpci_device_reserved(zdev);
}
break;
case 0x0306: /* 0x308 or 0x302 for multiple devices */
- clp_rescan_pci_devices();
+ zpci_remove_reserved_devices();
+ clp_scan_pci_devices();
break;
case 0x0308: /* Standby -> Reserved */
if (!zdev)
break;
- zpci_remove_device(zdev);
+ zpci_device_reserved(zdev);
break;
default:
break;
}
- pci_dev_put(pdev);
}
void zpci_event_availability(void *data)
diff --git a/arch/s390/pci/pci_iov.c b/arch/s390/pci/pci_iov.c
new file mode 100644
index 0000000..ead062b
--- /dev/null
+++ b/arch/s390/pci/pci_iov.c
@@ -0,0 +1,99 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright IBM Corp. 2020
+ *
+ * Author(s):
+ * Niklas Schnelle <schnelle@linux.ibm.com>
+ *
+ */
+
+#define KMSG_COMPONENT "zpci"
+#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/pci.h>
+
+#include "pci_iov.h"
+
+static struct resource iov_res = {
+ .name = "PCI IOV res",
+ .start = 0,
+ .end = -1,
+ .flags = IORESOURCE_MEM,
+};
+
+void zpci_iov_map_resources(struct pci_dev *pdev)
+{
+ resource_size_t len;
+ int i;
+
+ for (i = 0; i < PCI_SRIOV_NUM_BARS; i++) {
+ int bar = i + PCI_IOV_RESOURCES;
+
+ len = pci_resource_len(pdev, bar);
+ if (!len)
+ continue;
+ pdev->resource[bar].parent = &iov_res;
+ }
+}
+
+void zpci_iov_remove_virtfn(struct pci_dev *pdev, int vfn)
+{
+ pci_lock_rescan_remove();
+ /* Linux' vfid's start at 0 vfn at 1 */
+ pci_iov_remove_virtfn(pdev->physfn, vfn - 1);
+ pci_unlock_rescan_remove();
+}
+
+static int zpci_iov_link_virtfn(struct pci_dev *pdev, struct pci_dev *virtfn, int vfid)
+{
+ int rc;
+
+ rc = pci_iov_sysfs_link(pdev, virtfn, vfid);
+ if (rc)
+ return rc;
+
+ virtfn->is_virtfn = 1;
+ virtfn->multifunction = 0;
+ virtfn->physfn = pci_dev_get(pdev);
+
+ return 0;
+}
+
+int zpci_iov_setup_virtfn(struct zpci_bus *zbus, struct pci_dev *virtfn, int vfn)
+{
+ int i, cand_devfn;
+ struct zpci_dev *zdev;
+ struct pci_dev *pdev;
+ int vfid = vfn - 1; /* Linux' vfid's start at 0 vfn at 1*/
+ int rc = 0;
+
+ if (!zbus->multifunction)
+ return 0;
+
+ /* If the parent PF for the given VF is also configured in the
+ * instance, it must be on the same zbus.
+ * We can then identify the parent PF by checking what
+ * devfn the VF would have if it belonged to that PF using the PF's
+ * stride and offset. Only if this candidate devfn matches the
+ * actual devfn will we link both functions.
+ */
+ for (i = 0; i < ZPCI_FUNCTIONS_PER_BUS; i++) {
+ zdev = zbus->function[i];
+ if (zdev && zdev->is_physfn) {
+ pdev = pci_get_slot(zbus->bus, zdev->devfn);
+ if (!pdev)
+ continue;
+ cand_devfn = pci_iov_virtfn_devfn(pdev, vfid);
+ if (cand_devfn == virtfn->devfn) {
+ rc = zpci_iov_link_virtfn(pdev, virtfn, vfid);
+ /* balance pci_get_slot() */
+ pci_dev_put(pdev);
+ break;
+ }
+ /* balance pci_get_slot() */
+ pci_dev_put(pdev);
+ }
+ }
+ return rc;
+}
diff --git a/arch/s390/pci/pci_iov.h b/arch/s390/pci/pci_iov.h
new file mode 100644
index 0000000..b2c8280
--- /dev/null
+++ b/arch/s390/pci/pci_iov.h
@@ -0,0 +1,30 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright IBM Corp. 2020
+ *
+ * Author(s):
+ * Niklas Schnelle <schnelle@linux.ibm.com>
+ *
+ */
+
+#ifndef __S390_PCI_IOV_H
+#define __S390_PCI_IOV_H
+
+#ifdef CONFIG_PCI_IOV
+void zpci_iov_remove_virtfn(struct pci_dev *pdev, int vfn);
+
+void zpci_iov_map_resources(struct pci_dev *pdev);
+
+int zpci_iov_setup_virtfn(struct zpci_bus *zbus, struct pci_dev *virtfn, int vfn);
+
+#else /* CONFIG_PCI_IOV */
+static inline void zpci_iov_remove_virtfn(struct pci_dev *pdev, int vfn) {}
+
+static inline void zpci_iov_map_resources(struct pci_dev *pdev) {}
+
+static inline int zpci_iov_setup_virtfn(struct zpci_bus *zbus, struct pci_dev *virtfn, int vfn)
+{
+ return 0;
+}
+#endif /* CONFIG_PCI_IOV */
+#endif /* __S390_PCI_IOV_h */
diff --git a/arch/s390/pci/pci_irq.c b/arch/s390/pci/pci_irq.c
index 743f257..75217fb 100644
--- a/arch/s390/pci/pci_irq.c
+++ b/arch/s390/pci/pci_irq.c
@@ -103,9 +103,10 @@
{
struct msi_desc *entry = irq_get_msi_desc(data->irq);
struct msi_msg msg = entry->msg;
+ int cpu_addr = smp_cpu_get_cpu_address(cpumask_first(dest));
msg.address_lo &= 0xff0000ff;
- msg.address_lo |= (cpumask_first(dest) << 8);
+ msg.address_lo |= (cpu_addr << 8);
pci_write_msi_msg(data->irq, &msg);
return IRQ_SET_MASK_OK;
@@ -238,6 +239,7 @@
unsigned long bit;
struct msi_desc *msi;
struct msi_msg msg;
+ int cpu_addr;
int rc, irq;
zdev->aisb = -1UL;
@@ -287,9 +289,15 @@
handle_percpu_irq);
msg.data = hwirq - bit;
if (irq_delivery == DIRECTED) {
+ if (msi->affinity)
+ cpu = cpumask_first(&msi->affinity->mask);
+ else
+ cpu = 0;
+ cpu_addr = smp_cpu_get_cpu_address(cpu);
+
msg.address_lo = zdev->msi_addr & 0xff0000ff;
- msg.address_lo |= msi->affinity ?
- (cpumask_first(&msi->affinity->mask) << 8) : 0;
+ msg.address_lo |= (cpu_addr << 8);
+
for_each_possible_cpu(cpu) {
airq_iv_set_data(zpci_ibv[cpu], hwirq, irq);
}
diff --git a/arch/s390/pci/pci_mmio.c b/arch/s390/pci/pci_mmio.c
index 020a2c5..37b1bbd 100644
--- a/arch/s390/pci/pci_mmio.c
+++ b/arch/s390/pci/pci_mmio.c
@@ -125,17 +125,17 @@
struct vm_area_struct *vma;
long ret;
- down_read(¤t->mm->mmap_sem);
+ mmap_read_lock(current->mm);
ret = -EINVAL;
vma = find_vma(current->mm, user_addr);
- if (!vma)
+ if (!vma || user_addr < vma->vm_start)
goto out;
ret = -EACCES;
if (!(vma->vm_flags & access))
goto out;
ret = follow_pfn(vma, user_addr, pfn);
out:
- up_read(¤t->mm->mmap_sem);
+ mmap_read_unlock(current->mm);
return ret;
}
@@ -155,10 +155,12 @@
return -EINVAL;
/*
- * Only support read access to MIO capable devices on a MIO enabled
- * system. Otherwise we would have to check for every address if it is
- * a special ZPCI_ADDR and we would have to do a get_pfn() which we
- * don't need for MIO capable devices.
+ * We only support write access to MIO capable devices if we are on
+ * a MIO enabled system. Otherwise we would have to check for every
+ * address if it is a special ZPCI_ADDR and would have to do
+ * a get_pfn() which we don't need for MIO capable devices. Currently
+ * ISM devices are the only devices without MIO support and there is no
+ * known need for accessing these from userspace.
*/
if (static_branch_likely(&have_mio)) {
ret = __memcpy_toio_inuser((void __iomem *) mmio_addr,
@@ -282,10 +284,12 @@
return -EINVAL;
/*
- * Only support write access to MIO capable devices on a MIO enabled
- * system. Otherwise we would have to check for every address if it is
- * a special ZPCI_ADDR and we would have to do a get_pfn() which we
- * don't need for MIO capable devices.
+ * We only support read access to MIO capable devices if we are on
+ * a MIO enabled system. Otherwise we would have to check for every
+ * address if it is a special ZPCI_ADDR and would have to do
+ * a get_pfn() which we don't need for MIO capable devices. Currently
+ * ISM devices are the only devices without MIO support and there is no
+ * known need for accessing these from userspace.
*/
if (static_branch_likely(&have_mio)) {
ret = __memcpy_fromio_inuser(
diff --git a/arch/s390/pci/pci_sysfs.c b/arch/s390/pci/pci_sysfs.c
index 215f174..5c028be 100644
--- a/arch/s390/pci/pci_sysfs.c
+++ b/arch/s390/pci/pci_sysfs.c
@@ -33,6 +33,7 @@
zpci_attr(pfgid, "0x%02x\n", pfgid);
zpci_attr(vfn, "0x%04x\n", vfn);
zpci_attr(pft, "0x%02x\n", pft);
+zpci_attr(port, "%d\n", port);
zpci_attr(uid, "0x%x\n", uid);
zpci_attr(segment0, "0x%02x\n", pfip[0]);
zpci_attr(segment1, "0x%02x\n", pfip[1]);
@@ -88,7 +89,7 @@
ret = zpci_enable_device(zdev);
if (ret)
goto out;
- pci_rescan_bus(zdev->bus);
+ pci_rescan_bus(zdev->zbus->bus);
}
out:
pci_unlock_rescan_remove();
@@ -142,6 +143,7 @@
&dev_attr_pchid.attr,
&dev_attr_pfgid.attr,
&dev_attr_pft.attr,
+ &dev_attr_port.attr,
&dev_attr_vfn.attr,
&dev_attr_uid.attr,
&dev_attr_recover.attr,