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/drivers/nvme/host/Kconfig b/drivers/nvme/host/Kconfig
index cf0ae71..494675a 100644
--- a/drivers/nvme/host/Kconfig
+++ b/drivers/nvme/host/Kconfig
@@ -1,12 +1,13 @@
# SPDX-License-Identifier: GPL-2.0-only
config NVME_CORE
tristate
+ select BLK_DEV_INTEGRITY_T10 if BLK_DEV_INTEGRITY
config BLK_DEV_NVME
tristate "NVM Express block device"
depends on PCI && BLOCK
select NVME_CORE
- ---help---
+ help
The NVM Express driver is for solid state drives directly
connected to the PCI or PCI Express bus. If you know you
don't have one of these, it is safe to answer N.
@@ -17,12 +18,20 @@
config NVME_MULTIPATH
bool "NVMe multipath support"
depends on NVME_CORE
- ---help---
+ help
This option enables support for multipath access to NVMe
subsystems. If this option is enabled only a single
/dev/nvmeXnY device will show up for each NVMe namespaces,
even if it is accessible through multiple controllers.
+config NVME_HWMON
+ bool "NVMe hardware monitoring"
+ depends on (NVME_CORE=y && HWMON=y) || (NVME_CORE=m && HWMON)
+ help
+ This provides support for NVMe hardware monitoring. If enabled,
+ a hardware monitoring device will be created for each NVMe drive
+ in the system.
+
config NVME_FABRICS
tristate
diff --git a/drivers/nvme/host/Makefile b/drivers/nvme/host/Makefile
index 8a4b671..d7f6a87 100644
--- a/drivers/nvme/host/Makefile
+++ b/drivers/nvme/host/Makefile
@@ -13,7 +13,9 @@
nvme-core-$(CONFIG_TRACING) += trace.o
nvme-core-$(CONFIG_NVME_MULTIPATH) += multipath.o
nvme-core-$(CONFIG_NVM) += lightnvm.o
+nvme-core-$(CONFIG_BLK_DEV_ZONED) += zns.o
nvme-core-$(CONFIG_FAULT_INJECTION_DEBUG_FS) += fault_inject.o
+nvme-core-$(CONFIG_NVME_HWMON) += hwmon.o
nvme-y += pci.o
diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index a5b5a23..71c85c9 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -13,13 +13,11 @@
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/backing-dev.h>
-#include <linux/list_sort.h>
#include <linux/slab.h>
#include <linux/types.h>
#include <linux/pr.h>
#include <linux/ptrace.h>
#include <linux/nvme_ioctl.h>
-#include <linux/t10-pi.h>
#include <linux/pm_qos.h>
#include <asm/unaligned.h>
@@ -90,26 +88,38 @@
static struct class *nvme_class;
static struct class *nvme_subsys_class;
-static int nvme_revalidate_disk(struct gendisk *disk);
static void nvme_put_subsystem(struct nvme_subsystem *subsys);
static void nvme_remove_invalid_namespaces(struct nvme_ctrl *ctrl,
unsigned nsid);
+static void nvme_update_bdev_size(struct gendisk *disk)
+{
+ struct block_device *bdev = bdget_disk(disk, 0);
+
+ if (bdev) {
+ bd_set_nr_sectors(bdev, get_capacity(disk));
+ bdput(bdev);
+ }
+}
+
+/*
+ * Prepare a queue for teardown.
+ *
+ * This must forcibly unquiesce queues to avoid blocking dispatch, and only set
+ * the capacity to 0 after that to avoid blocking dispatchers that may be
+ * holding bd_butex. This will end buffered writers dirtying pages that can't
+ * be synced.
+ */
static void nvme_set_queue_dying(struct nvme_ns *ns)
{
- /*
- * Revalidating a dead namespace sets capacity to 0. This will end
- * buffered writers dirtying pages that can't be synced.
- */
- if (!ns->disk || test_and_set_bit(NVME_NS_DEAD, &ns->flags))
+ if (test_and_set_bit(NVME_NS_DEAD, &ns->flags))
return;
+
blk_set_queue_dying(ns->queue);
- /* Forcibly unquiesce queues to avoid blocking dispatch */
blk_mq_unquiesce_queue(ns->queue);
- /*
- * Revalidate after unblocking dispatchers that may be holding bd_butex
- */
- revalidate_disk(ns->disk);
+
+ set_capacity(ns->disk, 0);
+ nvme_update_bdev_size(ns->disk);
}
static void nvme_queue_scan(struct nvme_ctrl *ctrl)
@@ -172,7 +182,6 @@
nvme_remove_namespaces(ctrl);
ctrl->ops->delete_ctrl(ctrl);
nvme_uninit_ctrl(ctrl);
- nvme_put_ctrl(ctrl);
}
static void nvme_delete_ctrl_work(struct work_struct *work)
@@ -193,26 +202,16 @@
}
EXPORT_SYMBOL_GPL(nvme_delete_ctrl);
-static int nvme_delete_ctrl_sync(struct nvme_ctrl *ctrl)
+static void nvme_delete_ctrl_sync(struct nvme_ctrl *ctrl)
{
- int ret = 0;
-
/*
* Keep a reference until nvme_do_delete_ctrl() complete,
* since ->delete_ctrl can free the controller.
*/
nvme_get_ctrl(ctrl);
- if (!nvme_change_ctrl_state(ctrl, NVME_CTRL_DELETING))
- ret = -EBUSY;
- if (!ret)
+ if (nvme_change_ctrl_state(ctrl, NVME_CTRL_DELETING))
nvme_do_delete_ctrl(ctrl);
nvme_put_ctrl(ctrl);
- return ret;
-}
-
-static inline bool nvme_ns_has_pi(struct nvme_ns *ns)
-{
- return ns->pi_type && ns->ms == sizeof(struct t10_pi_tuple);
}
static blk_status_t nvme_error_status(u16 status)
@@ -223,6 +222,8 @@
case NVME_SC_CAP_EXCEEDED:
return BLK_STS_NOSPC;
case NVME_SC_LBA_RANGE:
+ case NVME_SC_CMD_INTERRUPTED:
+ case NVME_SC_NS_NOT_READY:
return BLK_STS_TARGET;
case NVME_SC_BAD_ATTRIBUTES:
case NVME_SC_ONCS_NOT_SUPPORTED:
@@ -246,22 +247,15 @@
return BLK_STS_NEXUS;
case NVME_SC_HOST_PATH_ERROR:
return BLK_STS_TRANSPORT;
+ case NVME_SC_ZONE_TOO_MANY_ACTIVE:
+ return BLK_STS_ZONE_ACTIVE_RESOURCE;
+ case NVME_SC_ZONE_TOO_MANY_OPEN:
+ return BLK_STS_ZONE_OPEN_RESOURCE;
default:
return BLK_STS_IOERR;
}
}
-static inline bool nvme_req_needs_retry(struct request *req)
-{
- if (blk_noretry_request(req))
- return false;
- if (nvme_req(req)->status & NVME_SC_DNR)
- return false;
- if (nvme_req(req)->retries >= nvme_max_retries)
- return false;
- return true;
-}
-
static void nvme_retry_req(struct request *req)
{
struct nvme_ns *ns = req->q->queuedata;
@@ -278,27 +272,66 @@
blk_mq_delay_kick_requeue_list(req->q, delay);
}
-void nvme_complete_rq(struct request *req)
+enum nvme_disposition {
+ COMPLETE,
+ RETRY,
+ FAILOVER,
+};
+
+static inline enum nvme_disposition nvme_decide_disposition(struct request *req)
+{
+ if (likely(nvme_req(req)->status == 0))
+ return COMPLETE;
+
+ if (blk_noretry_request(req) ||
+ (nvme_req(req)->status & NVME_SC_DNR) ||
+ nvme_req(req)->retries >= nvme_max_retries)
+ return COMPLETE;
+
+ if (req->cmd_flags & REQ_NVME_MPATH) {
+ if (nvme_is_path_error(nvme_req(req)->status) ||
+ blk_queue_dying(req->q))
+ return FAILOVER;
+ } else {
+ if (blk_queue_dying(req->q))
+ return COMPLETE;
+ }
+
+ return RETRY;
+}
+
+static inline void nvme_end_req(struct request *req)
{
blk_status_t status = nvme_error_status(nvme_req(req)->status);
+ if (IS_ENABLED(CONFIG_BLK_DEV_ZONED) &&
+ req_op(req) == REQ_OP_ZONE_APPEND)
+ req->__sector = nvme_lba_to_sect(req->q->queuedata,
+ le64_to_cpu(nvme_req(req)->result.u64));
+
+ nvme_trace_bio_complete(req, status);
+ blk_mq_end_request(req, status);
+}
+
+void nvme_complete_rq(struct request *req)
+{
trace_nvme_complete_rq(req);
+ nvme_cleanup_cmd(req);
if (nvme_req(req)->ctrl->kas)
nvme_req(req)->ctrl->comp_seen = true;
- if (unlikely(status != BLK_STS_OK && nvme_req_needs_retry(req))) {
- if ((req->cmd_flags & REQ_NVME_MPATH) && nvme_failover_req(req))
- return;
-
- if (!blk_queue_dying(req->q)) {
- nvme_retry_req(req);
- return;
- }
+ switch (nvme_decide_disposition(req)) {
+ case COMPLETE:
+ nvme_end_req(req);
+ return;
+ case RETRY:
+ nvme_retry_req(req);
+ return;
+ case FAILOVER:
+ nvme_failover_req(req);
+ return;
}
-
- nvme_trace_bio_complete(req, status);
- blk_mq_end_request(req, status);
}
EXPORT_SYMBOL_GPL(nvme_complete_rq);
@@ -355,7 +388,7 @@
case NVME_CTRL_RESETTING:
case NVME_CTRL_CONNECTING:
changed = true;
- /* FALLTHRU */
+ fallthrough;
default:
break;
}
@@ -365,7 +398,7 @@
case NVME_CTRL_NEW:
case NVME_CTRL_LIVE:
changed = true;
- /* FALLTHRU */
+ fallthrough;
default:
break;
}
@@ -375,7 +408,7 @@
case NVME_CTRL_NEW:
case NVME_CTRL_RESETTING:
changed = true;
- /* FALLTHRU */
+ fallthrough;
default:
break;
}
@@ -386,7 +419,17 @@
case NVME_CTRL_RESETTING:
case NVME_CTRL_CONNECTING:
changed = true;
- /* FALLTHRU */
+ fallthrough;
+ default:
+ break;
+ }
+ break;
+ case NVME_CTRL_DELETING_NOIO:
+ switch (old_state) {
+ case NVME_CTRL_DELETING:
+ case NVME_CTRL_DEAD:
+ changed = true;
+ fallthrough;
default:
break;
}
@@ -395,7 +438,7 @@
switch (old_state) {
case NVME_CTRL_DELETING:
changed = true;
- /* FALLTHRU */
+ fallthrough;
default:
break;
}
@@ -428,6 +471,7 @@
case NVME_CTRL_CONNECTING:
return false;
case NVME_CTRL_DELETING:
+ case NVME_CTRL_DELETING_NOIO:
case NVME_CTRL_DEAD:
return true;
default:
@@ -479,10 +523,11 @@
kfree(ns);
}
-static void nvme_put_ns(struct nvme_ns *ns)
+void nvme_put_ns(struct nvme_ns *ns)
{
kref_put(&ns->kref, nvme_free_ns);
}
+EXPORT_SYMBOL_NS_GPL(nvme_put_ns, NVME_TARGET_PASSTHRU);
static inline void nvme_clear_nvme_request(struct request *req)
{
@@ -552,7 +597,7 @@
c.directive.opcode = nvme_admin_directive_recv;
c.directive.nsid = cpu_to_le32(nsid);
- c.directive.numd = cpu_to_le32((sizeof(*s) >> 2) - 1);
+ c.directive.numd = cpu_to_le32(nvme_bytes_to_numd(sizeof(*s)));
c.directive.doper = NVME_DIR_RCV_ST_OP_PARAM;
c.directive.dtype = NVME_DIR_STREAMS;
@@ -575,19 +620,22 @@
ret = nvme_get_stream_params(ctrl, &s, NVME_NSID_ALL);
if (ret)
- return ret;
+ goto out_disable_stream;
ctrl->nssa = le16_to_cpu(s.nssa);
if (ctrl->nssa < BLK_MAX_WRITE_HINTS - 1) {
dev_info(ctrl->device, "too few streams (%u) available\n",
ctrl->nssa);
- nvme_disable_streams(ctrl);
- return 0;
+ goto out_disable_stream;
}
- ctrl->nr_streams = min_t(unsigned, ctrl->nssa, BLK_MAX_WRITE_HINTS - 1);
+ ctrl->nr_streams = min_t(u16, ctrl->nssa, BLK_MAX_WRITE_HINTS - 1);
dev_info(ctrl->device, "Using %u streams\n", ctrl->nr_streams);
return 0;
+
+out_disable_stream:
+ nvme_disable_streams(ctrl);
+ return ret;
}
/*
@@ -615,6 +663,14 @@
req->q->write_hints[streamid] += blk_rq_bytes(req) >> 9;
}
+static void nvme_setup_passthrough(struct request *req,
+ struct nvme_command *cmd)
+{
+ memcpy(cmd, nvme_req(req)->cmd, sizeof(*cmd));
+ /* passthru commands should let the driver set the SGL flags */
+ cmd->common.flags &= ~NVME_CMD_SGL_ALL;
+}
+
static inline void nvme_setup_flush(struct nvme_ns *ns,
struct nvme_command *cmnd)
{
@@ -702,7 +758,8 @@
}
static inline blk_status_t nvme_setup_rw(struct nvme_ns *ns,
- struct request *req, struct nvme_command *cmnd)
+ struct request *req, struct nvme_command *cmnd,
+ enum nvme_opcode op)
{
struct nvme_ctrl *ctrl = ns->ctrl;
u16 control = 0;
@@ -716,7 +773,7 @@
if (req->cmd_flags & REQ_RAHEAD)
dsmgmt |= NVME_RW_DSM_FREQ_PREFETCH;
- cmnd->rw.opcode = (rq_data_dir(req) ? nvme_cmd_write : nvme_cmd_read);
+ cmnd->rw.opcode = op;
cmnd->rw.nsid = cpu_to_le32(ns->head->ns_id);
cmnd->rw.slba = cpu_to_le64(nvme_sect_to_lba(ns, blk_rq_pos(req)));
cmnd->rw.length = cpu_to_le16((blk_rq_bytes(req) >> ns->lba_shift) - 1);
@@ -745,6 +802,8 @@
case NVME_NS_DPS_PI_TYPE2:
control |= NVME_RW_PRINFO_PRCHK_GUARD |
NVME_RW_PRINFO_PRCHK_REF;
+ if (op == nvme_cmd_zone_append)
+ control |= NVME_RW_APPEND_PIREMAP;
cmnd->rw.reftag = cpu_to_le32(t10_pi_ref_tag(req));
break;
}
@@ -772,6 +831,7 @@
blk_status_t nvme_setup_cmd(struct nvme_ns *ns, struct request *req,
struct nvme_command *cmd)
{
+ struct nvme_ctrl *ctrl = nvme_req(req)->ctrl;
blk_status_t ret = BLK_STS_OK;
nvme_clear_nvme_request(req);
@@ -780,11 +840,24 @@
switch (req_op(req)) {
case REQ_OP_DRV_IN:
case REQ_OP_DRV_OUT:
- memcpy(cmd, nvme_req(req)->cmd, sizeof(*cmd));
+ nvme_setup_passthrough(req, cmd);
break;
case REQ_OP_FLUSH:
nvme_setup_flush(ns, cmd);
break;
+ case REQ_OP_ZONE_RESET_ALL:
+ case REQ_OP_ZONE_RESET:
+ ret = nvme_setup_zone_mgmt_send(ns, req, cmd, NVME_ZONE_RESET);
+ break;
+ case REQ_OP_ZONE_OPEN:
+ ret = nvme_setup_zone_mgmt_send(ns, req, cmd, NVME_ZONE_OPEN);
+ break;
+ case REQ_OP_ZONE_CLOSE:
+ ret = nvme_setup_zone_mgmt_send(ns, req, cmd, NVME_ZONE_CLOSE);
+ break;
+ case REQ_OP_ZONE_FINISH:
+ ret = nvme_setup_zone_mgmt_send(ns, req, cmd, NVME_ZONE_FINISH);
+ break;
case REQ_OP_WRITE_ZEROES:
ret = nvme_setup_write_zeroes(ns, req, cmd);
break;
@@ -792,15 +865,22 @@
ret = nvme_setup_discard(ns, req, cmd);
break;
case REQ_OP_READ:
+ ret = nvme_setup_rw(ns, req, cmd, nvme_cmd_read);
+ break;
case REQ_OP_WRITE:
- ret = nvme_setup_rw(ns, req, cmd);
+ ret = nvme_setup_rw(ns, req, cmd, nvme_cmd_write);
+ break;
+ case REQ_OP_ZONE_APPEND:
+ ret = nvme_setup_rw(ns, req, cmd, nvme_cmd_zone_append);
break;
default:
WARN_ON_ONCE(1);
return BLK_STS_IOERR;
}
- cmd->common.command_id = req->tag;
+ if (!(ctrl->quirks & NVME_QUIRK_SKIP_CID_GEN))
+ nvme_req(req)->genctr++;
+ cmd->common.command_id = nvme_cid(req);
trace_nvme_setup_cmd(req, cmd);
return ret;
}
@@ -913,6 +993,93 @@
return ERR_PTR(ret);
}
+static u32 nvme_known_admin_effects(u8 opcode)
+{
+ switch (opcode) {
+ case nvme_admin_format_nvm:
+ return NVME_CMD_EFFECTS_LBCC | NVME_CMD_EFFECTS_NCC |
+ NVME_CMD_EFFECTS_CSE_MASK;
+ case nvme_admin_sanitize_nvm:
+ return NVME_CMD_EFFECTS_LBCC | NVME_CMD_EFFECTS_CSE_MASK;
+ default:
+ break;
+ }
+ return 0;
+}
+
+u32 nvme_command_effects(struct nvme_ctrl *ctrl, struct nvme_ns *ns, u8 opcode)
+{
+ u32 effects = 0;
+
+ if (ns) {
+ if (ns->head->effects)
+ effects = le32_to_cpu(ns->head->effects->iocs[opcode]);
+ if (effects & ~(NVME_CMD_EFFECTS_CSUPP | NVME_CMD_EFFECTS_LBCC))
+ dev_warn(ctrl->device,
+ "IO command:%02x has unhandled effects:%08x\n",
+ opcode, effects);
+ return 0;
+ }
+
+ if (ctrl->effects)
+ effects = le32_to_cpu(ctrl->effects->acs[opcode]);
+ effects |= nvme_known_admin_effects(opcode);
+
+ return effects;
+}
+EXPORT_SYMBOL_NS_GPL(nvme_command_effects, NVME_TARGET_PASSTHRU);
+
+static u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
+ u8 opcode)
+{
+ u32 effects = nvme_command_effects(ctrl, ns, opcode);
+
+ /*
+ * For simplicity, IO to all namespaces is quiesced even if the command
+ * effects say only one namespace is affected.
+ */
+ if (effects & NVME_CMD_EFFECTS_CSE_MASK) {
+ mutex_lock(&ctrl->scan_lock);
+ mutex_lock(&ctrl->subsys->lock);
+ nvme_mpath_start_freeze(ctrl->subsys);
+ nvme_mpath_wait_freeze(ctrl->subsys);
+ nvme_start_freeze(ctrl);
+ nvme_wait_freeze(ctrl);
+ }
+ return effects;
+}
+
+static void nvme_passthru_end(struct nvme_ctrl *ctrl, u32 effects)
+{
+ if (effects & NVME_CMD_EFFECTS_CSE_MASK) {
+ nvme_unfreeze(ctrl);
+ nvme_mpath_unfreeze(ctrl->subsys);
+ mutex_unlock(&ctrl->subsys->lock);
+ nvme_remove_invalid_namespaces(ctrl, NVME_NSID_ALL);
+ mutex_unlock(&ctrl->scan_lock);
+ }
+ if (effects & NVME_CMD_EFFECTS_CCC)
+ nvme_init_identify(ctrl);
+ if (effects & (NVME_CMD_EFFECTS_NIC | NVME_CMD_EFFECTS_NCC)) {
+ nvme_queue_scan(ctrl);
+ flush_work(&ctrl->scan_work);
+ }
+}
+
+void nvme_execute_passthru_rq(struct request *rq)
+{
+ struct nvme_command *cmd = nvme_req(rq)->cmd;
+ struct nvme_ctrl *ctrl = nvme_req(rq)->ctrl;
+ struct nvme_ns *ns = rq->q->queuedata;
+ struct gendisk *disk = ns ? ns->disk : NULL;
+ u32 effects;
+
+ effects = nvme_passthru_start(ctrl, ns, cmd->common.opcode);
+ blk_execute_rq(rq->q, disk, rq, 0);
+ nvme_passthru_end(ctrl, effects);
+}
+EXPORT_SYMBOL_NS_GPL(nvme_execute_passthru_rq, NVME_TARGET_PASSTHRU);
+
static int nvme_submit_user_cmd(struct request_queue *q,
struct nvme_command *cmd, void __user *ubuffer,
unsigned bufflen, void __user *meta_buffer, unsigned meta_len,
@@ -951,7 +1118,7 @@
}
}
- blk_execute_rq(req->q, disk, req, 0);
+ nvme_execute_passthru_rq(req);
if (nvme_req(req)->flags & NVME_REQ_CANCELLED)
ret = -EINTR;
else
@@ -1085,15 +1252,67 @@
return error;
}
+static bool nvme_multi_css(struct nvme_ctrl *ctrl)
+{
+ return (ctrl->ctrl_config & NVME_CC_CSS_MASK) == NVME_CC_CSS_CSI;
+}
+
+static int nvme_process_ns_desc(struct nvme_ctrl *ctrl, struct nvme_ns_ids *ids,
+ struct nvme_ns_id_desc *cur, bool *csi_seen)
+{
+ const char *warn_str = "ctrl returned bogus length:";
+ void *data = cur;
+
+ switch (cur->nidt) {
+ case NVME_NIDT_EUI64:
+ if (cur->nidl != NVME_NIDT_EUI64_LEN) {
+ dev_warn(ctrl->device, "%s %d for NVME_NIDT_EUI64\n",
+ warn_str, cur->nidl);
+ return -1;
+ }
+ memcpy(ids->eui64, data + sizeof(*cur), NVME_NIDT_EUI64_LEN);
+ return NVME_NIDT_EUI64_LEN;
+ case NVME_NIDT_NGUID:
+ if (cur->nidl != NVME_NIDT_NGUID_LEN) {
+ dev_warn(ctrl->device, "%s %d for NVME_NIDT_NGUID\n",
+ warn_str, cur->nidl);
+ return -1;
+ }
+ memcpy(ids->nguid, data + sizeof(*cur), NVME_NIDT_NGUID_LEN);
+ return NVME_NIDT_NGUID_LEN;
+ case NVME_NIDT_UUID:
+ if (cur->nidl != NVME_NIDT_UUID_LEN) {
+ dev_warn(ctrl->device, "%s %d for NVME_NIDT_UUID\n",
+ warn_str, cur->nidl);
+ return -1;
+ }
+ uuid_copy(&ids->uuid, data + sizeof(*cur));
+ return NVME_NIDT_UUID_LEN;
+ case NVME_NIDT_CSI:
+ if (cur->nidl != NVME_NIDT_CSI_LEN) {
+ dev_warn(ctrl->device, "%s %d for NVME_NIDT_CSI\n",
+ warn_str, cur->nidl);
+ return -1;
+ }
+ memcpy(&ids->csi, data + sizeof(*cur), NVME_NIDT_CSI_LEN);
+ *csi_seen = true;
+ return NVME_NIDT_CSI_LEN;
+ default:
+ /* Skip unknown types */
+ return cur->nidl;
+ }
+}
+
static int nvme_identify_ns_descs(struct nvme_ctrl *ctrl, unsigned nsid,
struct nvme_ns_ids *ids)
{
struct nvme_command c = { };
- int status;
+ bool csi_seen = false;
+ int status, pos, len;
void *data;
- int pos;
- int len;
+ if (ctrl->vs < NVME_VS(1, 3, 0) && !nvme_multi_css(ctrl))
+ return 0;
if (ctrl->quirks & NVME_QUIRK_NO_NS_DESC_LIST)
return 0;
@@ -1119,63 +1338,26 @@
if (cur->nidl == 0)
break;
- switch (cur->nidt) {
- case NVME_NIDT_EUI64:
- if (cur->nidl != NVME_NIDT_EUI64_LEN) {
- dev_warn(ctrl->device,
- "ctrl returned bogus length: %d for NVME_NIDT_EUI64\n",
- cur->nidl);
- goto free_data;
- }
- len = NVME_NIDT_EUI64_LEN;
- memcpy(ids->eui64, data + pos + sizeof(*cur), len);
+ len = nvme_process_ns_desc(ctrl, ids, cur, &csi_seen);
+ if (len < 0)
break;
- case NVME_NIDT_NGUID:
- if (cur->nidl != NVME_NIDT_NGUID_LEN) {
- dev_warn(ctrl->device,
- "ctrl returned bogus length: %d for NVME_NIDT_NGUID\n",
- cur->nidl);
- goto free_data;
- }
- len = NVME_NIDT_NGUID_LEN;
- memcpy(ids->nguid, data + pos + sizeof(*cur), len);
- break;
- case NVME_NIDT_UUID:
- if (cur->nidl != NVME_NIDT_UUID_LEN) {
- dev_warn(ctrl->device,
- "ctrl returned bogus length: %d for NVME_NIDT_UUID\n",
- cur->nidl);
- goto free_data;
- }
- len = NVME_NIDT_UUID_LEN;
- uuid_copy(&ids->uuid, data + pos + sizeof(*cur));
- break;
- default:
- /* Skip unknown types */
- len = cur->nidl;
- break;
- }
len += sizeof(*cur);
}
+
+ if (nvme_multi_css(ctrl) && !csi_seen) {
+ dev_warn(ctrl->device, "Command set not reported for nsid:%d\n",
+ nsid);
+ status = -EINVAL;
+ }
+
free_data:
kfree(data);
return status;
}
-static int nvme_identify_ns_list(struct nvme_ctrl *dev, unsigned nsid, __le32 *ns_list)
-{
- struct nvme_command c = { };
-
- c.identify.opcode = nvme_admin_identify;
- c.identify.cns = NVME_ID_CNS_NS_ACTIVE_LIST;
- c.identify.nsid = cpu_to_le32(nsid);
- return nvme_submit_sync_cmd(dev->admin_q, &c, ns_list,
- NVME_IDENTIFY_DATA_SIZE);
-}
-
-static int nvme_identify_ns(struct nvme_ctrl *ctrl,
- unsigned nsid, struct nvme_id_ns **id)
+static int nvme_identify_ns(struct nvme_ctrl *ctrl, unsigned nsid,
+ struct nvme_ns_ids *ids, struct nvme_id_ns **id)
{
struct nvme_command c = { };
int error;
@@ -1192,9 +1374,24 @@
error = nvme_submit_sync_cmd(ctrl->admin_q, &c, *id, sizeof(**id));
if (error) {
dev_warn(ctrl->device, "Identify namespace failed (%d)\n", error);
- kfree(*id);
+ goto out_free_id;
}
+ error = NVME_SC_INVALID_NS | NVME_SC_DNR;
+ if ((*id)->ncap == 0) /* namespace not allocated or attached */
+ goto out_free_id;
+
+ if (ctrl->vs >= NVME_VS(1, 1, 0) &&
+ !memchr_inv(ids->eui64, 0, sizeof(ids->eui64)))
+ memcpy(ids->eui64, (*id)->eui64, sizeof(ids->eui64));
+ if (ctrl->vs >= NVME_VS(1, 2, 0) &&
+ !memchr_inv(ids->nguid, 0, sizeof(ids->nguid)))
+ memcpy(ids->nguid, (*id)->nguid, sizeof(ids->nguid));
+
+ return 0;
+
+out_free_id:
+ kfree(*id);
return error;
}
@@ -1334,7 +1531,7 @@
metadata = nvme_to_user_ptr(io.metadata);
}
- if (ns->ext) {
+ if (ns->features & NVME_NS_EXT_LBAS) {
length += meta_len;
meta_len = 0;
} else if (meta_len) {
@@ -1359,94 +1556,12 @@
metadata, meta_len, lower_32_bits(io.slba), NULL, 0);
}
-static u32 nvme_known_admin_effects(u8 opcode)
-{
- switch (opcode) {
- case nvme_admin_format_nvm:
- return NVME_CMD_EFFECTS_CSUPP | NVME_CMD_EFFECTS_LBCC |
- NVME_CMD_EFFECTS_CSE_MASK;
- case nvme_admin_sanitize_nvm:
- return NVME_CMD_EFFECTS_CSE_MASK;
- default:
- break;
- }
- return 0;
-}
-
-static u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
- u8 opcode)
-{
- u32 effects = 0;
-
- if (ns) {
- if (ctrl->effects)
- effects = le32_to_cpu(ctrl->effects->iocs[opcode]);
- if (effects & ~(NVME_CMD_EFFECTS_CSUPP | NVME_CMD_EFFECTS_LBCC))
- dev_warn(ctrl->device,
- "IO command:%02x has unhandled effects:%08x\n",
- opcode, effects);
- return 0;
- }
-
- if (ctrl->effects)
- effects = le32_to_cpu(ctrl->effects->acs[opcode]);
- effects |= nvme_known_admin_effects(opcode);
-
- /*
- * For simplicity, IO to all namespaces is quiesced even if the command
- * effects say only one namespace is affected.
- */
- if (effects & (NVME_CMD_EFFECTS_LBCC | NVME_CMD_EFFECTS_CSE_MASK)) {
- mutex_lock(&ctrl->scan_lock);
- mutex_lock(&ctrl->subsys->lock);
- nvme_mpath_start_freeze(ctrl->subsys);
- nvme_mpath_wait_freeze(ctrl->subsys);
- nvme_start_freeze(ctrl);
- nvme_wait_freeze(ctrl);
- }
- return effects;
-}
-
-static void nvme_update_formats(struct nvme_ctrl *ctrl)
-{
- struct nvme_ns *ns;
-
- down_read(&ctrl->namespaces_rwsem);
- list_for_each_entry(ns, &ctrl->namespaces, list)
- if (ns->disk && nvme_revalidate_disk(ns->disk))
- nvme_set_queue_dying(ns);
- up_read(&ctrl->namespaces_rwsem);
-}
-
-static void nvme_passthru_end(struct nvme_ctrl *ctrl, u32 effects)
-{
- /*
- * Revalidate LBA changes prior to unfreezing. This is necessary to
- * prevent memory corruption if a logical block size was changed by
- * this command.
- */
- if (effects & NVME_CMD_EFFECTS_LBCC)
- nvme_update_formats(ctrl);
- if (effects & (NVME_CMD_EFFECTS_LBCC | NVME_CMD_EFFECTS_CSE_MASK)) {
- nvme_unfreeze(ctrl);
- nvme_mpath_unfreeze(ctrl->subsys);
- mutex_unlock(&ctrl->subsys->lock);
- nvme_remove_invalid_namespaces(ctrl, NVME_NSID_ALL);
- mutex_unlock(&ctrl->scan_lock);
- }
- if (effects & NVME_CMD_EFFECTS_CCC)
- nvme_init_identify(ctrl);
- if (effects & (NVME_CMD_EFFECTS_NIC | NVME_CMD_EFFECTS_NCC))
- nvme_queue_scan(ctrl);
-}
-
static int nvme_user_cmd(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
struct nvme_passthru_cmd __user *ucmd)
{
struct nvme_passthru_cmd cmd;
struct nvme_command c;
unsigned timeout = 0;
- u32 effects;
u64 result;
int status;
@@ -1473,12 +1588,10 @@
if (cmd.timeout_ms)
timeout = msecs_to_jiffies(cmd.timeout_ms);
- effects = nvme_passthru_start(ctrl, ns, cmd.opcode);
status = nvme_submit_user_cmd(ns ? ns->queue : ctrl->admin_q, &c,
nvme_to_user_ptr(cmd.addr), cmd.data_len,
nvme_to_user_ptr(cmd.metadata), cmd.metadata_len,
0, &result, timeout);
- nvme_passthru_end(ctrl, effects);
if (status >= 0) {
if (put_user(result, &ucmd->result))
@@ -1494,7 +1607,6 @@
struct nvme_passthru_cmd64 cmd;
struct nvme_command c;
unsigned timeout = 0;
- u32 effects;
int status;
if (!capable(CAP_SYS_ADMIN))
@@ -1520,12 +1632,10 @@
if (cmd.timeout_ms)
timeout = msecs_to_jiffies(cmd.timeout_ms);
- effects = nvme_passthru_start(ctrl, ns, cmd.opcode);
status = nvme_submit_user_cmd(ns ? ns->queue : ctrl->admin_q, &c,
nvme_to_user_ptr(cmd.addr), cmd.data_len,
nvme_to_user_ptr(cmd.metadata), cmd.metadata_len,
0, &cmd.result, timeout);
- nvme_passthru_end(ctrl, effects);
if (status >= 0) {
if (put_user(cmd.result, &ucmd->result))
@@ -1539,7 +1649,7 @@
* Issue ioctl requests on the first available path. Note that unlike normal
* block layer requests we will not retry failed request on another controller.
*/
-static struct nvme_ns *nvme_get_ns_from_disk(struct gendisk *disk,
+struct nvme_ns *nvme_get_ns_from_disk(struct gendisk *disk,
struct nvme_ns_head **head, int *srcu_idx)
{
#ifdef CONFIG_NVME_MULTIPATH
@@ -1559,7 +1669,7 @@
return disk->private_data;
}
-static void nvme_put_ns_from_disk(struct nvme_ns_head *head, int idx)
+void nvme_put_ns_from_disk(struct nvme_ns_head *head, int idx)
{
if (head)
srcu_read_unlock(&head->srcu, idx);
@@ -1645,6 +1755,47 @@
return ret;
}
+#ifdef CONFIG_COMPAT
+struct nvme_user_io32 {
+ __u8 opcode;
+ __u8 flags;
+ __u16 control;
+ __u16 nblocks;
+ __u16 rsvd;
+ __u64 metadata;
+ __u64 addr;
+ __u64 slba;
+ __u32 dsmgmt;
+ __u32 reftag;
+ __u16 apptag;
+ __u16 appmask;
+} __attribute__((__packed__));
+
+#define NVME_IOCTL_SUBMIT_IO32 _IOW('N', 0x42, struct nvme_user_io32)
+
+static int nvme_compat_ioctl(struct block_device *bdev, fmode_t mode,
+ unsigned int cmd, unsigned long arg)
+{
+ /*
+ * Corresponds to the difference of NVME_IOCTL_SUBMIT_IO
+ * between 32 bit programs and 64 bit kernel.
+ * The cause is that the results of sizeof(struct nvme_user_io),
+ * which is used to define NVME_IOCTL_SUBMIT_IO,
+ * are not same between 32 bit compiler and 64 bit compiler.
+ * NVME_IOCTL_SUBMIT_IO32 is for 64 bit kernel handling
+ * NVME_IOCTL_SUBMIT_IO issued from 32 bit programs.
+ * Other IOCTL numbers are same between 32 bit and 64 bit.
+ * So there is nothing to do regarding to other IOCTL numbers.
+ */
+ if (cmd == NVME_IOCTL_SUBMIT_IO32)
+ return nvme_ioctl(bdev, mode, NVME_IOCTL_SUBMIT_IO, arg);
+
+ return nvme_ioctl(bdev, mode, cmd, arg);
+}
+#else
+#define nvme_compat_ioctl NULL
+#endif /* CONFIG_COMPAT */
+
static int nvme_open(struct block_device *bdev, fmode_t mode)
{
struct nvme_ns *ns = bdev->bd_disk->private_data;
@@ -1685,7 +1836,8 @@
}
#ifdef CONFIG_BLK_DEV_INTEGRITY
-static void nvme_init_integrity(struct gendisk *disk, u16 ms, u8 pi_type)
+static void nvme_init_integrity(struct gendisk *disk, u16 ms, u8 pi_type,
+ u32 max_integrity_segments)
{
struct blk_integrity integrity;
@@ -1708,10 +1860,11 @@
}
integrity.tuple_size = ms;
blk_integrity_register(disk, &integrity);
- blk_queue_max_integrity_segments(disk->queue, 1);
+ blk_queue_max_integrity_segments(disk->queue, max_integrity_segments);
}
#else
-static void nvme_init_integrity(struct gendisk *disk, u16 ms, u8 pi_type)
+static void nvme_init_integrity(struct gendisk *disk, u16 ms, u8 pi_type,
+ u32 max_integrity_segments)
{
}
#endif /* CONFIG_BLK_DEV_INTEGRITY */
@@ -1761,20 +1914,6 @@
blk_queue_max_write_zeroes_sectors(q, ctrl->max_hw_sectors);
}
-static int nvme_report_ns_ids(struct nvme_ctrl *ctrl, unsigned int nsid,
- struct nvme_id_ns *id, struct nvme_ns_ids *ids)
-{
- memset(ids, 0, sizeof(*ids));
-
- if (ctrl->vs >= NVME_VS(1, 1, 0))
- memcpy(ids->eui64, id->eui64, sizeof(id->eui64));
- if (ctrl->vs >= NVME_VS(1, 2, 0))
- memcpy(ids->nguid, id->nguid, sizeof(id->nguid));
- if (ctrl->vs >= NVME_VS(1, 3, 0))
- return nvme_identify_ns_descs(ctrl, nsid, ids);
- return 0;
-}
-
static bool nvme_ns_ids_valid(struct nvme_ns_ids *ids)
{
return !uuid_is_null(&ids->uuid) ||
@@ -1786,7 +1925,95 @@
{
return uuid_equal(&a->uuid, &b->uuid) &&
memcmp(&a->nguid, &b->nguid, sizeof(a->nguid)) == 0 &&
- memcmp(&a->eui64, &b->eui64, sizeof(a->eui64)) == 0;
+ memcmp(&a->eui64, &b->eui64, sizeof(a->eui64)) == 0 &&
+ a->csi == b->csi;
+}
+
+static int nvme_setup_streams_ns(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
+ u32 *phys_bs, u32 *io_opt)
+{
+ struct streams_directive_params s;
+ int ret;
+
+ if (!ctrl->nr_streams)
+ return 0;
+
+ ret = nvme_get_stream_params(ctrl, &s, ns->head->ns_id);
+ if (ret)
+ return ret;
+
+ ns->sws = le32_to_cpu(s.sws);
+ ns->sgs = le16_to_cpu(s.sgs);
+
+ if (ns->sws) {
+ *phys_bs = ns->sws * (1 << ns->lba_shift);
+ if (ns->sgs)
+ *io_opt = *phys_bs * ns->sgs;
+ }
+
+ return 0;
+}
+
+static int nvme_configure_metadata(struct nvme_ns *ns, struct nvme_id_ns *id)
+{
+ struct nvme_ctrl *ctrl = ns->ctrl;
+
+ /*
+ * The PI implementation requires the metadata size to be equal to the
+ * t10 pi tuple size.
+ */
+ ns->ms = le16_to_cpu(id->lbaf[id->flbas & NVME_NS_FLBAS_LBA_MASK].ms);
+ if (ns->ms == sizeof(struct t10_pi_tuple))
+ ns->pi_type = id->dps & NVME_NS_DPS_PI_MASK;
+ else
+ ns->pi_type = 0;
+
+ ns->features &= ~(NVME_NS_METADATA_SUPPORTED | NVME_NS_EXT_LBAS);
+ if (!ns->ms || !(ctrl->ops->flags & NVME_F_METADATA_SUPPORTED))
+ return 0;
+ if (ctrl->ops->flags & NVME_F_FABRICS) {
+ /*
+ * The NVMe over Fabrics specification only supports metadata as
+ * part of the extended data LBA. We rely on HCA/HBA support to
+ * remap the separate metadata buffer from the block layer.
+ */
+ if (WARN_ON_ONCE(!(id->flbas & NVME_NS_FLBAS_META_EXT)))
+ return -EINVAL;
+ if (ctrl->max_integrity_segments)
+ ns->features |=
+ (NVME_NS_METADATA_SUPPORTED | NVME_NS_EXT_LBAS);
+ } else {
+ /*
+ * For PCIe controllers, we can't easily remap the separate
+ * metadata buffer from the block layer and thus require a
+ * separate metadata buffer for block layer metadata/PI support.
+ * We allow extended LBAs for the passthrough interface, though.
+ */
+ if (id->flbas & NVME_NS_FLBAS_META_EXT)
+ ns->features |= NVME_NS_EXT_LBAS;
+ else
+ ns->features |= NVME_NS_METADATA_SUPPORTED;
+ }
+
+ return 0;
+}
+
+static void nvme_set_queue_limits(struct nvme_ctrl *ctrl,
+ struct request_queue *q)
+{
+ bool vwc = ctrl->vwc & NVME_CTRL_VWC_PRESENT;
+
+ if (ctrl->max_hw_sectors) {
+ u32 max_segments =
+ (ctrl->max_hw_sectors / (NVME_CTRL_PAGE_SIZE >> 9)) + 1;
+
+ max_segments = min_not_zero(max_segments, ctrl->max_segments);
+ blk_queue_max_hw_sectors(q, ctrl->max_hw_sectors);
+ blk_queue_max_segments(q, min_t(u32, max_segments, USHRT_MAX));
+ }
+ blk_queue_virt_boundary(q, NVME_CTRL_PAGE_SIZE - 1);
+ blk_queue_dma_alignment(q, 7);
+ blk_queue_write_cache(q, vwc, vwc);
}
static void nvme_update_disk_info(struct gendisk *disk,
@@ -1794,35 +2021,38 @@
{
sector_t capacity = nvme_lba_to_sect(ns, le64_to_cpu(id->nsze));
unsigned short bs = 1 << ns->lba_shift;
- u32 atomic_bs, phys_bs, io_opt;
+ u32 atomic_bs, phys_bs, io_opt = 0;
+ /*
+ * The block layer can't support LBA sizes larger than the page size
+ * yet, so catch this early and don't allow block I/O.
+ */
if (ns->lba_shift > PAGE_SHIFT) {
- /* unsupported block size, set capacity to 0 later */
+ capacity = 0;
bs = (1 << 9);
}
- blk_mq_freeze_queue(disk->queue);
+
blk_integrity_unregister(disk);
+ atomic_bs = phys_bs = bs;
+ nvme_setup_streams_ns(ns->ctrl, ns, &phys_bs, &io_opt);
if (id->nabo == 0) {
/*
* Bit 1 indicates whether NAWUPF is defined for this namespace
* and whether it should be used instead of AWUPF. If NAWUPF ==
* 0 then AWUPF must be used instead.
*/
- if (id->nsfeat & (1 << 1) && id->nawupf)
+ if (id->nsfeat & NVME_NS_FEAT_ATOMICS && id->nawupf)
atomic_bs = (1 + le16_to_cpu(id->nawupf)) * bs;
else
atomic_bs = (1 + ns->ctrl->subsys->awupf) * bs;
- } else {
- atomic_bs = bs;
}
- phys_bs = bs;
- io_opt = bs;
- if (id->nsfeat & (1 << 4)) {
+
+ if (id->nsfeat & NVME_NS_FEAT_IO_OPT) {
/* NPWG = Namespace Preferred Write Granularity */
- phys_bs *= 1 + le16_to_cpu(id->npwg);
+ phys_bs = bs * (1 + le16_to_cpu(id->npwg));
/* NOWS = Namespace Optimal Write Size */
- io_opt *= 1 + le16_to_cpu(id->nows);
+ io_opt = bs * (1 + le16_to_cpu(id->nows));
}
blk_queue_logical_block_size(disk->queue, bs);
@@ -1835,109 +2065,110 @@
blk_queue_io_min(disk->queue, phys_bs);
blk_queue_io_opt(disk->queue, io_opt);
- if (ns->ms && !ns->ext &&
- (ns->ctrl->ops->flags & NVME_F_METADATA_SUPPORTED))
- nvme_init_integrity(disk, ns->ms, ns->pi_type);
- if ((ns->ms && !nvme_ns_has_pi(ns) && !blk_get_integrity(disk)) ||
- ns->lba_shift > PAGE_SHIFT)
- capacity = 0;
+ /*
+ * Register a metadata profile for PI, or the plain non-integrity NVMe
+ * metadata masquerading as Type 0 if supported, otherwise reject block
+ * I/O to namespaces with metadata except when the namespace supports
+ * PI, as it can strip/insert in that case.
+ */
+ if (ns->ms) {
+ if (IS_ENABLED(CONFIG_BLK_DEV_INTEGRITY) &&
+ (ns->features & NVME_NS_METADATA_SUPPORTED))
+ nvme_init_integrity(disk, ns->ms, ns->pi_type,
+ ns->ctrl->max_integrity_segments);
+ else if (!nvme_ns_has_pi(ns))
+ capacity = 0;
+ }
- set_capacity(disk, capacity);
+ set_capacity_revalidate_and_notify(disk, capacity, false);
nvme_config_discard(disk, ns);
nvme_config_write_zeroes(disk->queue, ns->ctrl);
- if (id->nsattr & (1 << 0))
+ if (id->nsattr & NVME_NS_ATTR_RO)
set_disk_ro(disk, true);
- else
- set_disk_ro(disk, false);
-
- blk_mq_unfreeze_queue(disk->queue);
}
-static void __nvme_revalidate_disk(struct gendisk *disk, struct nvme_id_ns *id)
+static inline bool nvme_first_scan(struct gendisk *disk)
{
- struct nvme_ns *ns = disk->private_data;
+ /* nvme_alloc_ns() scans the disk prior to adding it */
+ return !(disk->flags & GENHD_FL_UP);
+}
+
+static void nvme_set_chunk_sectors(struct nvme_ns *ns, struct nvme_id_ns *id)
+{
+ struct nvme_ctrl *ctrl = ns->ctrl;
u32 iob;
- /*
- * If identify namespace failed, use default 512 byte block size so
- * block layer can use before failing read/write for 0 capacity.
- */
- ns->lba_shift = id->lbaf[id->flbas & NVME_NS_FLBAS_LBA_MASK].ds;
- if (ns->lba_shift == 0)
- ns->lba_shift = 9;
-
- if ((ns->ctrl->quirks & NVME_QUIRK_STRIPE_SIZE) &&
- is_power_of_2(ns->ctrl->max_hw_sectors))
- iob = ns->ctrl->max_hw_sectors;
+ if ((ctrl->quirks & NVME_QUIRK_STRIPE_SIZE) &&
+ is_power_of_2(ctrl->max_hw_sectors))
+ iob = ctrl->max_hw_sectors;
else
iob = nvme_lba_to_sect(ns, le16_to_cpu(id->noiob));
- ns->ms = le16_to_cpu(id->lbaf[id->flbas & NVME_NS_FLBAS_LBA_MASK].ms);
- ns->ext = ns->ms && (id->flbas & NVME_NS_FLBAS_META_EXT);
- /* the PI implementation requires metadata equal t10 pi tuple size */
- if (ns->ms == sizeof(struct t10_pi_tuple))
- ns->pi_type = id->dps & NVME_NS_DPS_PI_MASK;
- else
- ns->pi_type = 0;
+ if (!iob)
+ return;
- if (iob)
- blk_queue_chunk_sectors(ns->queue, rounddown_pow_of_two(iob));
- nvme_update_disk_info(disk, ns, id);
-#ifdef CONFIG_NVME_MULTIPATH
- if (ns->head->disk) {
- nvme_update_disk_info(ns->head->disk, ns, id);
- blk_queue_stack_limits(ns->head->disk->queue, ns->queue);
- nvme_mpath_update_disk_size(ns->head->disk);
+ if (!is_power_of_2(iob)) {
+ if (nvme_first_scan(ns->disk))
+ pr_warn("%s: ignoring unaligned IO boundary:%u\n",
+ ns->disk->disk_name, iob);
+ return;
}
-#endif
+
+ if (blk_queue_is_zoned(ns->disk->queue)) {
+ if (nvme_first_scan(ns->disk))
+ pr_warn("%s: ignoring zoned namespace IO boundary\n",
+ ns->disk->disk_name);
+ return;
+ }
+
+ blk_queue_chunk_sectors(ns->queue, iob);
}
-static int nvme_revalidate_disk(struct gendisk *disk)
+static int nvme_update_ns_info(struct nvme_ns *ns, struct nvme_id_ns *id)
{
- struct nvme_ns *ns = disk->private_data;
- struct nvme_ctrl *ctrl = ns->ctrl;
- struct nvme_id_ns *id;
- struct nvme_ns_ids ids;
- int ret = 0;
+ unsigned lbaf = id->flbas & NVME_NS_FLBAS_LBA_MASK;
+ int ret;
- if (test_bit(NVME_NS_DEAD, &ns->flags)) {
- set_capacity(disk, 0);
- return -ENODEV;
+ blk_mq_freeze_queue(ns->disk->queue);
+ ns->lba_shift = id->lbaf[lbaf].ds;
+ nvme_set_queue_limits(ns->ctrl, ns->queue);
+
+ if (ns->head->ids.csi == NVME_CSI_ZNS) {
+ ret = nvme_update_zone_info(ns, lbaf);
+ if (ret)
+ goto out_unfreeze;
}
- ret = nvme_identify_ns(ctrl, ns->head->ns_id, &id);
+ ret = nvme_configure_metadata(ns, id);
if (ret)
- goto out;
+ goto out_unfreeze;
+ nvme_set_chunk_sectors(ns, id);
+ nvme_update_disk_info(ns->disk, ns, id);
+ blk_mq_unfreeze_queue(ns->disk->queue);
- if (id->ncap == 0) {
- ret = -ENODEV;
- goto free_id;
+ if (blk_queue_is_zoned(ns->queue)) {
+ ret = nvme_revalidate_zones(ns);
+ if (ret && !nvme_first_scan(ns->disk))
+ return ret;
}
- __nvme_revalidate_disk(disk, id);
- ret = nvme_report_ns_ids(ctrl, ns->head->ns_id, id, &ids);
- if (ret)
- goto free_id;
-
- if (!nvme_ns_ids_equal(&ns->head->ids, &ids)) {
- dev_err(ctrl->device,
- "identifiers changed for nsid %d\n", ns->head->ns_id);
- ret = -ENODEV;
+#ifdef CONFIG_NVME_MULTIPATH
+ if (ns->head->disk) {
+ blk_mq_freeze_queue(ns->head->disk->queue);
+ nvme_update_disk_info(ns->head->disk, ns, id);
+ blk_stack_limits(&ns->head->disk->queue->limits,
+ &ns->queue->limits, 0);
+ blk_queue_update_readahead(ns->head->disk->queue);
+ nvme_update_bdev_size(ns->head->disk);
+ blk_mq_unfreeze_queue(ns->head->disk->queue);
}
+#endif
+ return 0;
-free_id:
- kfree(id);
-out:
- /*
- * Only fail the function if we got a fatal error back from the
- * device, otherwise ignore the error and just move on.
- */
- if (ret == -ENOMEM || (ret > 0 && !(ret & NVME_SC_DNR)))
- ret = 0;
- else if (ret > 0)
- ret = blk_status_to_errno(nvme_error_status(ret));
+out_unfreeze:
+ blk_mq_unfreeze_queue(ns->disk->queue);
return ret;
}
@@ -2066,11 +2297,11 @@
static const struct block_device_operations nvme_fops = {
.owner = THIS_MODULE,
.ioctl = nvme_ioctl,
- .compat_ioctl = nvme_ioctl,
+ .compat_ioctl = nvme_compat_ioctl,
.open = nvme_open,
.release = nvme_release,
.getgeo = nvme_getgeo,
- .revalidate_disk= nvme_revalidate_disk,
+ .report_zones = nvme_report_zones,
.pr_ops = &nvme_pr_ops,
};
@@ -2091,11 +2322,13 @@
const struct block_device_operations nvme_ns_head_ops = {
.owner = THIS_MODULE,
+ .submit_bio = nvme_ns_head_submit_bio,
.open = nvme_ns_head_open,
.release = nvme_ns_head_release,
.ioctl = nvme_ioctl,
- .compat_ioctl = nvme_ioctl,
+ .compat_ioctl = nvme_compat_ioctl,
.getgeo = nvme_getgeo,
+ .report_zones = nvme_report_zones,
.pr_ops = &nvme_pr_ops,
};
#endif /* CONFIG_NVME_MULTIPATH */
@@ -2113,13 +2346,13 @@
if ((csts & NVME_CSTS_RDY) == bit)
break;
- msleep(100);
+ usleep_range(1000, 2000);
if (fatal_signal_pending(current))
return -EINTR;
if (time_after(jiffies, timeout)) {
dev_err(ctrl->device,
- "Device not ready; aborting %s\n", enabled ?
- "initialisation" : "reset");
+ "Device not ready; aborting %s, CSTS=0x%x\n",
+ enabled ? "initialisation" : "reset", csts);
return -ENODEV;
}
}
@@ -2153,12 +2386,7 @@
int nvme_enable_ctrl(struct nvme_ctrl *ctrl)
{
- /*
- * Default to a 4K page size, with the intention to update this
- * path in the future to accomodate architectures with differing
- * kernel and IO page sizes.
- */
- unsigned dev_page_min, page_shift = 12;
+ unsigned dev_page_min;
int ret;
ret = ctrl->ops->reg_read64(ctrl, NVME_REG_CAP, &ctrl->cap);
@@ -2168,17 +2396,18 @@
}
dev_page_min = NVME_CAP_MPSMIN(ctrl->cap) + 12;
- if (page_shift < dev_page_min) {
+ if (NVME_CTRL_PAGE_SHIFT < dev_page_min) {
dev_err(ctrl->device,
"Minimum device page size %u too large for host (%u)\n",
- 1 << dev_page_min, 1 << page_shift);
+ 1 << dev_page_min, 1 << NVME_CTRL_PAGE_SHIFT);
return -ENODEV;
}
- ctrl->page_size = 1 << page_shift;
-
- ctrl->ctrl_config = NVME_CC_CSS_NVM;
- ctrl->ctrl_config |= (page_shift - 12) << NVME_CC_MPS_SHIFT;
+ if (NVME_CAP_CSS(ctrl->cap) & NVME_CAP_CSS_CSI)
+ ctrl->ctrl_config = NVME_CC_CSS_CSI;
+ else
+ ctrl->ctrl_config = NVME_CC_CSS_NVM;
+ ctrl->ctrl_config |= (NVME_CTRL_PAGE_SHIFT - 12) << NVME_CC_MPS_SHIFT;
ctrl->ctrl_config |= NVME_CC_AMS_RR | NVME_CC_SHN_NONE;
ctrl->ctrl_config |= NVME_CC_IOSQES | NVME_CC_IOCQES;
ctrl->ctrl_config |= NVME_CC_ENABLE;
@@ -2221,25 +2450,6 @@
}
EXPORT_SYMBOL_GPL(nvme_shutdown_ctrl);
-static void nvme_set_queue_limits(struct nvme_ctrl *ctrl,
- struct request_queue *q)
-{
- bool vwc = false;
-
- if (ctrl->max_hw_sectors) {
- u32 max_segments =
- (ctrl->max_hw_sectors / (ctrl->page_size >> 9)) + 1;
-
- max_segments = min_not_zero(max_segments, ctrl->max_segments);
- blk_queue_max_hw_sectors(q, ctrl->max_hw_sectors);
- blk_queue_max_segments(q, min_t(u32, max_segments, USHRT_MAX));
- }
- blk_queue_virt_boundary(q, ctrl->page_size - 1);
- if (ctrl->vwc & NVME_CTRL_VWC_PRESENT)
- vwc = true;
- blk_queue_write_cache(q, vwc, vwc);
-}
-
static int nvme_configure_timestamp(struct nvme_ctrl *ctrl)
{
__le64 ts;
@@ -2620,6 +2830,11 @@
NULL,
};
+static inline bool nvme_discovery_ctrl(struct nvme_ctrl *ctrl)
+{
+ return ctrl->opts && ctrl->opts->discovery_nqn;
+}
+
static bool nvme_validate_cntlid(struct nvme_subsystem *subsys,
struct nvme_ctrl *ctrl, struct nvme_id_ctrl *id)
{
@@ -2628,8 +2843,7 @@
lockdep_assert_held(&nvme_subsystems_lock);
list_for_each_entry(tmp, &subsys->ctrls, subsys_entry) {
- if (tmp->state == NVME_CTRL_DELETING ||
- tmp->state == NVME_CTRL_DEAD)
+ if (nvme_state_terminal(tmp))
continue;
if (tmp->cntlid == ctrl->cntlid) {
@@ -2639,8 +2853,8 @@
return false;
}
- if ((id->cmic & (1 << 1)) ||
- (ctrl->opts && ctrl->opts->discovery_nqn))
+ if ((id->cmic & NVME_CTRL_CMIC_MULTI_CTRL) ||
+ nvme_discovery_ctrl(ctrl))
continue;
dev_err(ctrl->device,
@@ -2726,11 +2940,11 @@
return ret;
}
-int nvme_get_log(struct nvme_ctrl *ctrl, u32 nsid, u8 log_page, u8 lsp,
+int nvme_get_log(struct nvme_ctrl *ctrl, u32 nsid, u8 log_page, u8 lsp, u8 csi,
void *log, size_t size, u64 offset)
{
struct nvme_command c = { };
- unsigned long dwlen = size / 4 - 1;
+ u32 dwlen = nvme_bytes_to_numd(size);
c.get_log_page.opcode = nvme_admin_get_log_page;
c.get_log_page.nsid = cpu_to_le32(nsid);
@@ -2740,27 +2954,35 @@
c.get_log_page.numdu = cpu_to_le16(dwlen >> 16);
c.get_log_page.lpol = cpu_to_le32(lower_32_bits(offset));
c.get_log_page.lpou = cpu_to_le32(upper_32_bits(offset));
+ c.get_log_page.csi = csi;
return nvme_submit_sync_cmd(ctrl->admin_q, &c, log, size);
}
-static int nvme_get_effects_log(struct nvme_ctrl *ctrl)
+static int nvme_get_effects_log(struct nvme_ctrl *ctrl, u8 csi,
+ struct nvme_effects_log **log)
{
+ struct nvme_effects_log *cel = xa_load(&ctrl->cels, csi);
int ret;
- if (!ctrl->effects)
- ctrl->effects = kzalloc(sizeof(*ctrl->effects), GFP_KERNEL);
+ if (cel)
+ goto out;
- if (!ctrl->effects)
- return 0;
+ cel = kzalloc(sizeof(*cel), GFP_KERNEL);
+ if (!cel)
+ return -ENOMEM;
- ret = nvme_get_log(ctrl, NVME_NSID_ALL, NVME_LOG_CMD_EFFECTS, 0,
- ctrl->effects, sizeof(*ctrl->effects), 0);
+ ret = nvme_get_log(ctrl, 0x00, NVME_LOG_CMD_EFFECTS, 0, csi,
+ cel, sizeof(*cel), 0);
if (ret) {
- kfree(ctrl->effects);
- ctrl->effects = NULL;
+ kfree(cel);
+ return ret;
}
- return ret;
+
+ xa_store(&ctrl->cels, csi, cel, GFP_KERNEL);
+out:
+ *log = cel;
+ return 0;
}
/*
@@ -2781,7 +3003,7 @@
return ret;
}
page_shift = NVME_CAP_MPSMIN(ctrl->cap) + 12;
- ctrl->sqsize = min_t(int, NVME_CAP_MQES(ctrl->cap), ctrl->sqsize);
+ ctrl->sqsize = min_t(u16, NVME_CAP_MQES(ctrl->cap), ctrl->sqsize);
if (ctrl->vs >= NVME_VS(1, 1, 0))
ctrl->subsystem = NVME_CAP_NSSRC(ctrl->cap);
@@ -2793,7 +3015,7 @@
}
if (id->lpa & NVME_CTRL_LPA_CMD_EFFECTS_LOG) {
- ret = nvme_get_effects_log(ctrl);
+ ret = nvme_get_effects_log(ctrl, NVME_CSI_NVM, &ctrl->effects);
if (ret < 0)
goto out_free;
}
@@ -2835,6 +3057,9 @@
ctrl->oncs = le16_to_cpu(id->oncs);
ctrl->mtfa = le16_to_cpu(id->mtfa);
ctrl->oaes = le32_to_cpu(id->oaes);
+ ctrl->wctemp = le16_to_cpu(id->wctemp);
+ ctrl->cctemp = le16_to_cpu(id->cctemp);
+
atomic_set(&ctrl->abort_limit, id->acl + 1);
ctrl->vwc = id->vwc;
if (id->mdts)
@@ -2852,7 +3077,7 @@
if (id->rtd3e) {
/* us -> s */
- u32 transition_time = le32_to_cpu(id->rtd3e) / 1000000;
+ u32 transition_time = le32_to_cpu(id->rtd3e) / USEC_PER_SEC;
ctrl->shutdown_timeout = clamp_t(unsigned int, transition_time,
shutdown_timeout, 60);
@@ -2890,11 +3115,15 @@
* admin connect
*/
if (ctrl->cntlid != le16_to_cpu(id->cntlid)) {
+ dev_err(ctrl->device,
+ "Mismatching cntlid: Connect %u vs Identify "
+ "%u, rejecting\n",
+ ctrl->cntlid, le16_to_cpu(id->cntlid));
ret = -EINVAL;
goto out_free;
}
- if (!ctrl->opts->discovery_nqn && !ctrl->kas) {
+ if (!nvme_discovery_ctrl(ctrl) && !ctrl->kas) {
dev_err(ctrl->device,
"keep-alive support is mandatory for fabrics\n");
ret = -EINVAL;
@@ -2934,6 +3163,12 @@
if (ret < 0)
return ret;
+ if (!ctrl->identified && !nvme_discovery_ctrl(ctrl)) {
+ ret = nvme_hwmon_init(ctrl);
+ if (ret < 0)
+ return ret;
+ }
+
ctrl->identified = true;
return 0;
@@ -3040,7 +3275,7 @@
.open = nvme_dev_open,
.release = nvme_dev_release,
.unlocked_ioctl = nvme_dev_ioctl,
- .compat_ioctl = nvme_dev_ioctl,
+ .compat_ioctl = compat_ptr_ioctl,
};
static ssize_t nvme_sysfs_reset(struct device *dev,
@@ -3237,10 +3472,6 @@
{
struct nvme_ctrl *ctrl = dev_get_drvdata(dev);
- /* Can't delete non-created controllers */
- if (!ctrl->created)
- return -EBUSY;
-
if (device_remove_file_self(dev, attr))
nvme_delete_ctrl_sync(ctrl);
return count;
@@ -3268,6 +3499,7 @@
[NVME_CTRL_RESETTING] = "resetting",
[NVME_CTRL_CONNECTING] = "connecting",
[NVME_CTRL_DELETING] = "deleting",
+ [NVME_CTRL_DELETING_NOIO]= "deleting (no IO)",
[NVME_CTRL_DEAD] = "dead",
};
@@ -3290,6 +3522,26 @@
}
static DEVICE_ATTR(subsysnqn, S_IRUGO, nvme_sysfs_show_subsysnqn, NULL);
+static ssize_t nvme_sysfs_show_hostnqn(struct device *dev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct nvme_ctrl *ctrl = dev_get_drvdata(dev);
+
+ return snprintf(buf, PAGE_SIZE, "%s\n", ctrl->opts->host->nqn);
+}
+static DEVICE_ATTR(hostnqn, S_IRUGO, nvme_sysfs_show_hostnqn, NULL);
+
+static ssize_t nvme_sysfs_show_hostid(struct device *dev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct nvme_ctrl *ctrl = dev_get_drvdata(dev);
+
+ return snprintf(buf, PAGE_SIZE, "%pU\n", &ctrl->opts->host->id);
+}
+static DEVICE_ATTR(hostid, S_IRUGO, nvme_sysfs_show_hostid, NULL);
+
static ssize_t nvme_sysfs_show_address(struct device *dev,
struct device_attribute *attr,
char *buf)
@@ -3300,6 +3552,66 @@
}
static DEVICE_ATTR(address, S_IRUGO, nvme_sysfs_show_address, NULL);
+static ssize_t nvme_ctrl_loss_tmo_show(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ struct nvme_ctrl *ctrl = dev_get_drvdata(dev);
+ struct nvmf_ctrl_options *opts = ctrl->opts;
+
+ if (ctrl->opts->max_reconnects == -1)
+ return sprintf(buf, "off\n");
+ return sprintf(buf, "%d\n",
+ opts->max_reconnects * opts->reconnect_delay);
+}
+
+static ssize_t nvme_ctrl_loss_tmo_store(struct device *dev,
+ struct device_attribute *attr, const char *buf, size_t count)
+{
+ struct nvme_ctrl *ctrl = dev_get_drvdata(dev);
+ struct nvmf_ctrl_options *opts = ctrl->opts;
+ int ctrl_loss_tmo, err;
+
+ err = kstrtoint(buf, 10, &ctrl_loss_tmo);
+ if (err)
+ return -EINVAL;
+
+ else if (ctrl_loss_tmo < 0)
+ opts->max_reconnects = -1;
+ else
+ opts->max_reconnects = DIV_ROUND_UP(ctrl_loss_tmo,
+ opts->reconnect_delay);
+ return count;
+}
+static DEVICE_ATTR(ctrl_loss_tmo, S_IRUGO | S_IWUSR,
+ nvme_ctrl_loss_tmo_show, nvme_ctrl_loss_tmo_store);
+
+static ssize_t nvme_ctrl_reconnect_delay_show(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ struct nvme_ctrl *ctrl = dev_get_drvdata(dev);
+
+ if (ctrl->opts->reconnect_delay == -1)
+ return sprintf(buf, "off\n");
+ return sprintf(buf, "%d\n", ctrl->opts->reconnect_delay);
+}
+
+static ssize_t nvme_ctrl_reconnect_delay_store(struct device *dev,
+ struct device_attribute *attr, const char *buf, size_t count)
+{
+ struct nvme_ctrl *ctrl = dev_get_drvdata(dev);
+ unsigned int v;
+ int err;
+
+ err = kstrtou32(buf, 10, &v);
+ if (err)
+ return err;
+
+ ctrl->opts->reconnect_delay = v;
+ return count;
+}
+static DEVICE_ATTR(reconnect_delay, S_IRUGO | S_IWUSR,
+ nvme_ctrl_reconnect_delay_show, nvme_ctrl_reconnect_delay_store);
+
static struct attribute *nvme_dev_attrs[] = {
&dev_attr_reset_controller.attr,
&dev_attr_rescan_controller.attr,
@@ -3315,6 +3627,10 @@
&dev_attr_numa_node.attr,
&dev_attr_queue_count.attr,
&dev_attr_sqsize.attr,
+ &dev_attr_hostnqn.attr,
+ &dev_attr_hostid.attr,
+ &dev_attr_ctrl_loss_tmo.attr,
+ &dev_attr_reconnect_delay.attr,
NULL
};
@@ -3328,6 +3644,14 @@
return 0;
if (a == &dev_attr_address.attr && !ctrl->ops->get_address)
return 0;
+ if (a == &dev_attr_hostnqn.attr && !ctrl->opts)
+ return 0;
+ if (a == &dev_attr_hostid.attr && !ctrl->opts)
+ return 0;
+ if (a == &dev_attr_ctrl_loss_tmo.attr && !ctrl->opts)
+ return 0;
+ if (a == &dev_attr_reconnect_delay.attr && !ctrl->opts)
+ return 0;
return a->mode;
}
@@ -3342,7 +3666,7 @@
NULL,
};
-static struct nvme_ns_head *__nvme_find_ns_head(struct nvme_subsystem *subsys,
+static struct nvme_ns_head *nvme_find_ns_head(struct nvme_subsystem *subsys,
unsigned nsid)
{
struct nvme_ns_head *h;
@@ -3374,7 +3698,7 @@
}
static struct nvme_ns_head *nvme_alloc_ns_head(struct nvme_ctrl *ctrl,
- unsigned nsid, struct nvme_id_ns *id)
+ unsigned nsid, struct nvme_ns_ids *ids)
{
struct nvme_ns_head *head;
size_t size = sizeof(*head);
@@ -3397,12 +3721,9 @@
goto out_ida_remove;
head->subsys = ctrl->subsys;
head->ns_id = nsid;
+ head->ids = *ids;
kref_init(&head->ref);
- ret = nvme_report_ns_ids(ctrl, nsid, id, &head->ids);
- if (ret)
- goto out_cleanup_srcu;
-
ret = __nvme_check_ids(ctrl->subsys, head);
if (ret) {
dev_err(ctrl->device,
@@ -3410,6 +3731,13 @@
goto out_cleanup_srcu;
}
+ if (head->ids.csi) {
+ ret = nvme_get_effects_log(ctrl, head->ids.csi, &head->effects);
+ if (ret)
+ goto out_cleanup_srcu;
+ } else
+ head->effects = ctrl->effects;
+
ret = nvme_mpath_alloc_disk(ctrl, head);
if (ret)
goto out_cleanup_srcu;
@@ -3432,58 +3760,49 @@
}
static int nvme_init_ns_head(struct nvme_ns *ns, unsigned nsid,
- struct nvme_id_ns *id)
+ struct nvme_ns_ids *ids, bool is_shared)
{
struct nvme_ctrl *ctrl = ns->ctrl;
- bool is_shared = id->nmic & (1 << 0);
struct nvme_ns_head *head = NULL;
int ret = 0;
mutex_lock(&ctrl->subsys->lock);
- if (is_shared)
- head = __nvme_find_ns_head(ctrl->subsys, nsid);
+ head = nvme_find_ns_head(ctrl->subsys, nsid);
if (!head) {
- head = nvme_alloc_ns_head(ctrl, nsid, id);
+ head = nvme_alloc_ns_head(ctrl, nsid, ids);
if (IS_ERR(head)) {
ret = PTR_ERR(head);
goto out_unlock;
}
+ head->shared = is_shared;
} else {
- struct nvme_ns_ids ids;
-
- ret = nvme_report_ns_ids(ctrl, nsid, id, &ids);
- if (ret)
- goto out_unlock;
-
- if (!nvme_ns_ids_equal(&head->ids, &ids)) {
+ ret = -EINVAL;
+ if (!is_shared || !head->shared) {
+ dev_err(ctrl->device,
+ "Duplicate unshared namespace %d\n", nsid);
+ goto out_put_ns_head;
+ }
+ if (!nvme_ns_ids_equal(&head->ids, ids)) {
dev_err(ctrl->device,
"IDs don't match for shared namespace %d\n",
nsid);
- ret = -EINVAL;
- nvme_put_ns_head(head);
- goto out_unlock;
+ goto out_put_ns_head;
}
}
list_add_tail(&ns->siblings, &head->list);
ns->head = head;
+ mutex_unlock(&ctrl->subsys->lock);
+ return 0;
+out_put_ns_head:
+ nvme_put_ns_head(head);
out_unlock:
mutex_unlock(&ctrl->subsys->lock);
- if (ret > 0)
- ret = blk_status_to_errno(nvme_error_status(ret));
return ret;
}
-static int ns_cmp(void *priv, struct list_head *a, struct list_head *b)
-{
- struct nvme_ns *nsa = container_of(a, struct nvme_ns, list);
- struct nvme_ns *nsb = container_of(b, struct nvme_ns, list);
-
- return nsa->head->ns_id - nsb->head->ns_id;
-}
-
-static struct nvme_ns *nvme_find_get_ns(struct nvme_ctrl *ctrl, unsigned nsid)
+struct nvme_ns *nvme_find_get_ns(struct nvme_ctrl *ctrl, unsigned nsid)
{
struct nvme_ns *ns, *ret = NULL;
@@ -3501,34 +3820,26 @@
up_read(&ctrl->namespaces_rwsem);
return ret;
}
+EXPORT_SYMBOL_NS_GPL(nvme_find_get_ns, NVME_TARGET_PASSTHRU);
-static int nvme_setup_streams_ns(struct nvme_ctrl *ctrl, struct nvme_ns *ns)
+/*
+ * Add the namespace to the controller list while keeping the list ordered.
+ */
+static void nvme_ns_add_to_ctrl_list(struct nvme_ns *ns)
{
- struct streams_directive_params s;
- int ret;
+ struct nvme_ns *tmp;
- if (!ctrl->nr_streams)
- return 0;
-
- ret = nvme_get_stream_params(ctrl, &s, ns->head->ns_id);
- if (ret)
- return ret;
-
- ns->sws = le32_to_cpu(s.sws);
- ns->sgs = le16_to_cpu(s.sgs);
-
- if (ns->sws) {
- unsigned int bs = 1 << ns->lba_shift;
-
- blk_queue_io_min(ns->queue, bs * ns->sws);
- if (ns->sgs)
- blk_queue_io_opt(ns->queue, bs * ns->sws * ns->sgs);
+ list_for_each_entry_reverse(tmp, &ns->ctrl->namespaces, list) {
+ if (tmp->head->ns_id < ns->head->ns_id) {
+ list_add(&ns->list, &tmp->list);
+ return;
+ }
}
-
- return 0;
+ list_add(&ns->list, &ns->ctrl->namespaces);
}
-static int nvme_alloc_ns(struct nvme_ctrl *ctrl, unsigned nsid)
+static void nvme_alloc_ns(struct nvme_ctrl *ctrl, unsigned nsid,
+ struct nvme_ns_ids *ids)
{
struct nvme_ns *ns;
struct gendisk *disk;
@@ -3536,19 +3847,19 @@
char disk_name[DISK_NAME_LEN];
int node = ctrl->numa_node, flags = GENHD_FL_EXT_DEVT, ret;
+ if (nvme_identify_ns(ctrl, nsid, ids, &id))
+ return;
+
ns = kzalloc_node(sizeof(*ns), GFP_KERNEL, node);
if (!ns)
- return -ENOMEM;
+ goto out_free_id;
ns->queue = blk_mq_init_queue(ctrl->tagset);
- if (IS_ERR(ns->queue)) {
- ret = PTR_ERR(ns->queue);
+ if (IS_ERR(ns->queue))
goto out_free_ns;
- }
if (ctrl->opts && ctrl->opts->data_digest)
- ns->queue->backing_dev_info->capabilities
- |= BDI_CAP_STABLE_WRITES;
+ blk_queue_flag_set(QUEUE_FLAG_STABLE_WRITES, ns->queue);
blk_queue_flag_set(QUEUE_FLAG_NONROT, ns->queue);
if (ctrl->ops->flags & NVME_F_PCI_P2PDMA)
@@ -3556,33 +3867,16 @@
ns->queue->queuedata = ns;
ns->ctrl = ctrl;
-
kref_init(&ns->kref);
- ns->lba_shift = 9; /* set to a default value for 512 until disk is validated */
- blk_queue_logical_block_size(ns->queue, 1 << ns->lba_shift);
- nvme_set_queue_limits(ctrl, ns->queue);
-
- ret = nvme_identify_ns(ctrl, nsid, &id);
+ ret = nvme_init_ns_head(ns, nsid, ids, id->nmic & NVME_NS_NMIC_SHARED);
if (ret)
goto out_free_queue;
-
- if (id->ncap == 0) {
- ret = -EINVAL;
- goto out_free_id;
- }
-
- ret = nvme_init_ns_head(ns, nsid, id);
- if (ret)
- goto out_free_id;
- nvme_setup_streams_ns(ctrl, ns);
nvme_set_disk_name(disk_name, ns, ctrl, &flags);
disk = alloc_disk_node(0, node);
- if (!disk) {
- ret = -ENOMEM;
+ if (!disk)
goto out_unlink_ns;
- }
disk->fops = &nvme_fops;
disk->private_data = ns;
@@ -3591,7 +3885,8 @@
memcpy(disk->disk_name, disk_name, DISK_NAME_LEN);
ns->disk = disk;
- __nvme_revalidate_disk(disk, id);
+ if (nvme_update_ns_info(ns, id))
+ goto out_put_disk;
if ((ctrl->quirks & NVME_QUIRK_LIGHTNVM) && id->vs[0] == 0x1) {
ret = nvme_nvm_register(ns, disk_name, node);
@@ -3602,9 +3897,8 @@
}
down_write(&ctrl->namespaces_rwsem);
- list_add_tail(&ns->list, &ctrl->namespaces);
+ nvme_ns_add_to_ctrl_list(ns);
up_write(&ctrl->namespaces_rwsem);
-
nvme_get_ctrl(ctrl);
device_add_disk(ctrl->device, ns->disk, nvme_ns_id_attr_groups);
@@ -3613,7 +3907,7 @@
nvme_fault_inject_init(&ns->fault_inject, ns->disk->disk_name);
kfree(id);
- return 0;
+ return;
out_put_disk:
/* prevent double queue cleanup */
ns->disk->queue = NULL;
@@ -3625,15 +3919,12 @@
list_del_init(&ns->head->entry);
mutex_unlock(&ctrl->subsys->lock);
nvme_put_ns_head(ns->head);
- out_free_id:
- kfree(id);
out_free_queue:
blk_cleanup_queue(ns->queue);
out_free_ns:
kfree(ns);
- if (ret > 0)
- ret = blk_status_to_errno(nvme_error_status(ret));
- return ret;
+ out_free_id:
+ kfree(id);
}
static void nvme_ns_remove(struct nvme_ns *ns)
@@ -3641,6 +3932,7 @@
if (test_and_set_bit(NVME_NS_REMOVING, &ns->flags))
return;
+ set_capacity(ns->disk, 0);
nvme_fault_inject_fini(&ns->fault_inject);
mutex_lock(&ns->ctrl->subsys->lock);
@@ -3653,7 +3945,7 @@
nvme_mpath_clear_current_path(ns);
synchronize_srcu(&ns->head->srcu); /* wait for concurrent submissions */
- if (ns->disk && ns->disk->flags & GENHD_FL_UP) {
+ if (ns->disk->flags & GENHD_FL_UP) {
del_gendisk(ns->disk);
blk_cleanup_queue(ns->queue);
if (blk_get_integrity(ns->disk))
@@ -3668,17 +3960,91 @@
nvme_put_ns(ns);
}
-static void nvme_validate_ns(struct nvme_ctrl *ctrl, unsigned nsid)
+static void nvme_ns_remove_by_nsid(struct nvme_ctrl *ctrl, u32 nsid)
{
+ struct nvme_ns *ns = nvme_find_get_ns(ctrl, nsid);
+
+ if (ns) {
+ nvme_ns_remove(ns);
+ nvme_put_ns(ns);
+ }
+}
+
+static void nvme_validate_ns(struct nvme_ns *ns, struct nvme_ns_ids *ids)
+{
+ struct nvme_id_ns *id;
+ int ret = NVME_SC_INVALID_NS | NVME_SC_DNR;
+
+ if (test_bit(NVME_NS_DEAD, &ns->flags))
+ goto out;
+
+ ret = nvme_identify_ns(ns->ctrl, ns->head->ns_id, ids, &id);
+ if (ret)
+ goto out;
+
+ ret = NVME_SC_INVALID_NS | NVME_SC_DNR;
+ if (!nvme_ns_ids_equal(&ns->head->ids, ids)) {
+ dev_err(ns->ctrl->device,
+ "identifiers changed for nsid %d\n", ns->head->ns_id);
+ goto out_free_id;
+ }
+
+ ret = nvme_update_ns_info(ns, id);
+
+out_free_id:
+ kfree(id);
+out:
+ /*
+ * Only remove the namespace if we got a fatal error back from the
+ * device, otherwise ignore the error and just move on.
+ *
+ * TODO: we should probably schedule a delayed retry here.
+ */
+ if (ret > 0 && (ret & NVME_SC_DNR))
+ nvme_ns_remove(ns);
+ else
+ revalidate_disk_size(ns->disk, true);
+}
+
+static void nvme_validate_or_alloc_ns(struct nvme_ctrl *ctrl, unsigned nsid)
+{
+ struct nvme_ns_ids ids = { };
struct nvme_ns *ns;
+ if (nvme_identify_ns_descs(ctrl, nsid, &ids))
+ return;
+
ns = nvme_find_get_ns(ctrl, nsid);
if (ns) {
- if (ns->disk && revalidate_disk(ns->disk))
- nvme_ns_remove(ns);
+ nvme_validate_ns(ns, &ids);
nvme_put_ns(ns);
- } else
- nvme_alloc_ns(ctrl, nsid);
+ return;
+ }
+
+ switch (ids.csi) {
+ case NVME_CSI_NVM:
+ nvme_alloc_ns(ctrl, nsid, &ids);
+ break;
+ case NVME_CSI_ZNS:
+ if (!IS_ENABLED(CONFIG_BLK_DEV_ZONED)) {
+ dev_warn(ctrl->device,
+ "nsid %u not supported without CONFIG_BLK_DEV_ZONED\n",
+ nsid);
+ break;
+ }
+ if (!nvme_multi_css(ctrl)) {
+ dev_warn(ctrl->device,
+ "command set not reported for nsid: %d\n",
+ nsid);
+ break;
+ }
+ nvme_alloc_ns(ctrl, nsid, &ids);
+ break;
+ default:
+ dev_warn(ctrl->device, "unknown csi %u for nsid %u\n",
+ ids.csi, nsid);
+ break;
+ }
}
static void nvme_remove_invalid_namespaces(struct nvme_ctrl *ctrl,
@@ -3699,39 +4065,41 @@
}
-static int nvme_scan_ns_list(struct nvme_ctrl *ctrl, unsigned nn)
+static int nvme_scan_ns_list(struct nvme_ctrl *ctrl)
{
- struct nvme_ns *ns;
+ const int nr_entries = NVME_IDENTIFY_DATA_SIZE / sizeof(__le32);
__le32 *ns_list;
- unsigned i, j, nsid, prev = 0;
- unsigned num_lists = DIV_ROUND_UP_ULL((u64)nn, 1024);
- int ret = 0;
+ u32 prev = 0;
+ int ret = 0, i;
+
+ if (nvme_ctrl_limited_cns(ctrl))
+ return -EOPNOTSUPP;
ns_list = kzalloc(NVME_IDENTIFY_DATA_SIZE, GFP_KERNEL);
if (!ns_list)
return -ENOMEM;
- for (i = 0; i < num_lists; i++) {
- ret = nvme_identify_ns_list(ctrl, prev, ns_list);
+ for (;;) {
+ struct nvme_command cmd = {
+ .identify.opcode = nvme_admin_identify,
+ .identify.cns = NVME_ID_CNS_NS_ACTIVE_LIST,
+ .identify.nsid = cpu_to_le32(prev),
+ };
+
+ ret = nvme_submit_sync_cmd(ctrl->admin_q, &cmd, ns_list,
+ NVME_IDENTIFY_DATA_SIZE);
if (ret)
goto free;
- for (j = 0; j < min(nn, 1024U); j++) {
- nsid = le32_to_cpu(ns_list[j]);
- if (!nsid)
+ for (i = 0; i < nr_entries; i++) {
+ u32 nsid = le32_to_cpu(ns_list[i]);
+
+ if (!nsid) /* end of the list? */
goto out;
-
- nvme_validate_ns(ctrl, nsid);
-
- while (++prev < nsid) {
- ns = nvme_find_get_ns(ctrl, prev);
- if (ns) {
- nvme_ns_remove(ns);
- nvme_put_ns(ns);
- }
- }
+ nvme_validate_or_alloc_ns(ctrl, nsid);
+ while (++prev < nsid)
+ nvme_ns_remove_by_nsid(ctrl, prev);
}
- nn -= j;
}
out:
nvme_remove_invalid_namespaces(ctrl, prev);
@@ -3740,12 +4108,18 @@
return ret;
}
-static void nvme_scan_ns_sequential(struct nvme_ctrl *ctrl, unsigned nn)
+static void nvme_scan_ns_sequential(struct nvme_ctrl *ctrl)
{
- unsigned i;
+ struct nvme_id_ctrl *id;
+ u32 nn, i;
+
+ if (nvme_identify_ctrl(ctrl, &id))
+ return;
+ nn = le32_to_cpu(id->nn);
+ kfree(id);
for (i = 1; i <= nn; i++)
- nvme_validate_ns(ctrl, i);
+ nvme_validate_or_alloc_ns(ctrl, i);
nvme_remove_invalid_namespaces(ctrl, nn);
}
@@ -3766,8 +4140,8 @@
* raced with us in reading the log page, which could cause us to miss
* updates.
*/
- error = nvme_get_log(ctrl, NVME_NSID_ALL, NVME_LOG_CHANGED_NS, 0, log,
- log_size, 0);
+ error = nvme_get_log(ctrl, NVME_NSID_ALL, NVME_LOG_CHANGED_NS, 0,
+ NVME_CSI_NVM, log, log_size, 0);
if (error)
dev_warn(ctrl->device,
"reading changed ns log failed: %d\n", error);
@@ -3779,8 +4153,6 @@
{
struct nvme_ctrl *ctrl =
container_of(work, struct nvme_ctrl, scan_work);
- struct nvme_id_ctrl *id;
- unsigned nn;
/* No tagset on a live ctrl means IO queues could not created */
if (ctrl->state != NVME_CTRL_LIVE || !ctrl->tagset)
@@ -3791,22 +4163,10 @@
nvme_clear_changed_ns_log(ctrl);
}
- if (nvme_identify_ctrl(ctrl, &id))
- return;
-
mutex_lock(&ctrl->scan_lock);
- nn = le32_to_cpu(id->nn);
- if (!nvme_ctrl_limited_cns(ctrl)) {
- if (!nvme_scan_ns_list(ctrl, nn))
- goto out_free_id;
- }
- nvme_scan_ns_sequential(ctrl, nn);
-out_free_id:
+ if (nvme_scan_ns_list(ctrl) != 0)
+ nvme_scan_ns_sequential(ctrl);
mutex_unlock(&ctrl->scan_lock);
- kfree(id);
- down_write(&ctrl->namespaces_rwsem);
- list_sort(NULL, &ctrl->namespaces, ns_cmp);
- up_write(&ctrl->namespaces_rwsem);
}
/*
@@ -3838,6 +4198,9 @@
if (ctrl->state == NVME_CTRL_DEAD)
nvme_kill_queues(ctrl);
+ /* this is a no-op when called from the controller reset handler */
+ nvme_change_ctrl_state(ctrl, NVME_CTRL_DELETING_NOIO);
+
down_write(&ctrl->namespaces_rwsem);
list_splice_init(&ctrl->namespaces, &ns_list);
up_write(&ctrl->namespaces_rwsem);
@@ -3896,7 +4259,14 @@
container_of(work, struct nvme_ctrl, async_event_work);
nvme_aen_uevent(ctrl);
- ctrl->ops->submit_async_event(ctrl);
+
+ /*
+ * The transport drivers must guarantee AER submission here is safe by
+ * flushing ctrl async_event_work after changing the controller state
+ * from LIVE and before freeing the admin queue.
+ */
+ if (ctrl->state == NVME_CTRL_LIVE)
+ ctrl->ops->submit_async_event(ctrl);
}
static bool nvme_ctrl_pp_status(struct nvme_ctrl *ctrl)
@@ -3921,8 +4291,8 @@
if (!log)
return;
- if (nvme_get_log(ctrl, NVME_NSID_ALL, NVME_LOG_FW_SLOT, 0, log,
- sizeof(*log), 0))
+ if (nvme_get_log(ctrl, NVME_NSID_ALL, NVME_LOG_FW_SLOT, 0, NVME_CSI_NVM,
+ log, sizeof(*log), 0))
dev_warn(ctrl->device, "Get FW SLOT INFO log error\n");
kfree(log);
}
@@ -4032,8 +4402,7 @@
void nvme_start_ctrl(struct nvme_ctrl *ctrl)
{
- if (ctrl->kato)
- nvme_start_keep_alive(ctrl);
+ nvme_start_keep_alive(ctrl);
nvme_enable_aen(ctrl);
@@ -4041,7 +4410,6 @@
nvme_queue_scan(ctrl);
nvme_start_queues(ctrl);
}
- ctrl->created = true;
}
EXPORT_SYMBOL_GPL(nvme_start_ctrl);
@@ -4050,9 +4418,23 @@
nvme_fault_inject_fini(&ctrl->fault_inject);
dev_pm_qos_hide_latency_tolerance(ctrl->device);
cdev_device_del(&ctrl->cdev, ctrl->device);
+ nvme_put_ctrl(ctrl);
}
EXPORT_SYMBOL_GPL(nvme_uninit_ctrl);
+static void nvme_free_cels(struct nvme_ctrl *ctrl)
+{
+ struct nvme_effects_log *cel;
+ unsigned long i;
+
+ xa_for_each (&ctrl->cels, i, cel) {
+ xa_erase(&ctrl->cels, i);
+ kfree(cel);
+ }
+
+ xa_destroy(&ctrl->cels);
+}
+
static void nvme_free_ctrl(struct device *dev)
{
struct nvme_ctrl *ctrl =
@@ -4062,7 +4444,7 @@
if (!subsys || ctrl->instance != subsys->instance)
ida_simple_remove(&nvme_instance_ida, ctrl->instance);
- kfree(ctrl->effects);
+ nvme_free_cels(ctrl);
nvme_mpath_uninit(ctrl);
__free_page(ctrl->discard_page);
@@ -4093,10 +4475,12 @@
spin_lock_init(&ctrl->lock);
mutex_init(&ctrl->scan_lock);
INIT_LIST_HEAD(&ctrl->namespaces);
+ xa_init(&ctrl->cels);
init_rwsem(&ctrl->namespaces_rwsem);
ctrl->dev = dev;
ctrl->ops = ops;
ctrl->quirks = quirks;
+ ctrl->numa_node = NUMA_NO_NODE;
INIT_WORK(&ctrl->scan_work, nvme_scan_work);
INIT_WORK(&ctrl->async_event_work, nvme_async_event_work);
INIT_WORK(&ctrl->fw_act_work, nvme_fw_act_work);
@@ -4276,6 +4660,14 @@
}
EXPORT_SYMBOL_GPL(nvme_sync_queues);
+struct nvme_ctrl *nvme_ctrl_from_file(struct file *file)
+{
+ if (file->f_op != &nvme_dev_fops)
+ return NULL;
+ return file->private_data;
+}
+EXPORT_SYMBOL_NS_GPL(nvme_ctrl_from_file, NVME_TARGET_PASSTHRU);
+
/*
* Check we didn't inadvertently grow the command structure sizes:
*/
@@ -4294,6 +4686,8 @@
BUILD_BUG_ON(sizeof(struct nvme_command) != 64);
BUILD_BUG_ON(sizeof(struct nvme_id_ctrl) != NVME_IDENTIFY_DATA_SIZE);
BUILD_BUG_ON(sizeof(struct nvme_id_ns) != NVME_IDENTIFY_DATA_SIZE);
+ BUILD_BUG_ON(sizeof(struct nvme_id_ns_zns) != NVME_IDENTIFY_DATA_SIZE);
+ BUILD_BUG_ON(sizeof(struct nvme_id_ctrl_zns) != NVME_IDENTIFY_DATA_SIZE);
BUILD_BUG_ON(sizeof(struct nvme_lba_range_type) != 64);
BUILD_BUG_ON(sizeof(struct nvme_smart_log) != 512);
BUILD_BUG_ON(sizeof(struct nvme_dbbuf) != 64);
@@ -4362,6 +4756,7 @@
destroy_workqueue(nvme_delete_wq);
destroy_workqueue(nvme_reset_wq);
destroy_workqueue(nvme_wq);
+ ida_destroy(&nvme_instance_ida);
}
MODULE_LICENSE("GPL");
diff --git a/drivers/nvme/host/fabrics.c b/drivers/nvme/host/fabrics.c
index d884187..7015fba 100644
--- a/drivers/nvme/host/fabrics.c
+++ b/drivers/nvme/host/fabrics.c
@@ -105,14 +105,14 @@
int len = 0;
if (ctrl->opts->mask & NVMF_OPT_TRADDR)
- len += snprintf(buf, size, "traddr=%s", ctrl->opts->traddr);
+ len += scnprintf(buf, size, "traddr=%s", ctrl->opts->traddr);
if (ctrl->opts->mask & NVMF_OPT_TRSVCID)
- len += snprintf(buf + len, size - len, "%strsvcid=%s",
+ len += scnprintf(buf + len, size - len, "%strsvcid=%s",
(len) ? "," : "", ctrl->opts->trsvcid);
if (ctrl->opts->mask & NVMF_OPT_HOST_TRADDR)
- len += snprintf(buf + len, size - len, "%shost_traddr=%s",
+ len += scnprintf(buf + len, size - len, "%shost_traddr=%s",
(len) ? "," : "", ctrl->opts->host_traddr);
- len += snprintf(buf + len, size - len, "\n");
+ len += scnprintf(buf + len, size - len, "\n");
return len;
}
@@ -552,7 +552,7 @@
blk_status_t nvmf_fail_nonready_command(struct nvme_ctrl *ctrl,
struct request *rq)
{
- if (ctrl->state != NVME_CTRL_DELETING &&
+ if (ctrl->state != NVME_CTRL_DELETING_NOIO &&
ctrl->state != NVME_CTRL_DEAD &&
!blk_noretry_request(rq) && !(rq->cmd_flags & REQ_NVME_MPATH))
return BLK_STS_RESOURCE;
diff --git a/drivers/nvme/host/fabrics.h b/drivers/nvme/host/fabrics.h
index a0ec40a..78467cb 100644
--- a/drivers/nvme/host/fabrics.h
+++ b/drivers/nvme/host/fabrics.h
@@ -153,6 +153,7 @@
struct nvmf_ctrl_options *opts)
{
if (ctrl->state == NVME_CTRL_DELETING ||
+ ctrl->state == NVME_CTRL_DELETING_NOIO ||
ctrl->state == NVME_CTRL_DEAD ||
strcmp(opts->subsysnqn, ctrl->opts->subsysnqn) ||
strcmp(opts->host->nqn, ctrl->opts->host->nqn) ||
@@ -182,7 +183,8 @@
static inline bool nvmf_check_ready(struct nvme_ctrl *ctrl, struct request *rq,
bool queue_live)
{
- if (likely(ctrl->state == NVME_CTRL_LIVE))
+ if (likely(ctrl->state == NVME_CTRL_LIVE ||
+ ctrl->state == NVME_CTRL_DELETING))
return true;
return __nvmf_check_ready(ctrl, rq, queue_live);
}
diff --git a/drivers/nvme/host/fc.c b/drivers/nvme/host/fc.c
index 0d2c22c..906cab3 100644
--- a/drivers/nvme/host/fc.c
+++ b/drivers/nvme/host/fc.c
@@ -14,6 +14,7 @@
#include "fabrics.h"
#include <linux/nvme-fc-driver.h>
#include <linux/nvme-fc.h>
+#include "fc.h"
#include <scsi/scsi_transport_fc.h>
/* *************************** Data Structures/Defines ****************** */
@@ -25,6 +26,10 @@
};
#define NVME_FC_DEFAULT_DEV_LOSS_TMO 60 /* seconds */
+#define NVME_FC_DEFAULT_RECONNECT_TMO 2 /* delay between reconnects
+ * when connected and a
+ * connection failure.
+ */
struct nvme_fc_queue {
struct nvme_fc_ctrl *ctrl;
@@ -61,6 +66,17 @@
bool req_queued;
};
+struct nvmefc_ls_rcv_op {
+ struct nvme_fc_rport *rport;
+ struct nvmefc_ls_rsp *lsrsp;
+ union nvmefc_ls_requests *rqstbuf;
+ union nvmefc_ls_responses *rspbuf;
+ u16 rqstdatalen;
+ bool handled;
+ dma_addr_t rspdma;
+ struct list_head lsrcv_list; /* rport->ls_rcv_list */
+} __aligned(sizeof(u64)); /* alignment for other things alloc'd with */
+
enum nvme_fcpop_state {
FCPOP_STATE_UNINIT = 0,
FCPOP_STATE_IDLE = 1,
@@ -95,8 +111,8 @@
struct nvme_fcp_op_w_sgl {
struct nvme_fc_fcp_op op;
- struct scatterlist sgl[SG_CHUNK_SIZE];
- uint8_t priv[0];
+ struct scatterlist sgl[NVME_INLINE_SG_CNT];
+ uint8_t priv[];
};
struct nvme_fc_lport {
@@ -117,6 +133,7 @@
struct list_head endp_list; /* for lport->endp_list */
struct list_head ctrl_list;
struct list_head ls_req_list;
+ struct list_head ls_rcv_list;
struct list_head disc_list;
struct device *dev; /* physical device for dma */
struct nvme_fc_lport *lport;
@@ -124,11 +141,13 @@
struct kref ref;
atomic_t act_ctrl_cnt;
unsigned long dev_loss_end;
+ struct work_struct lsrcv_work;
} __aligned(sizeof(u64)); /* alignment for other things alloc'd with */
-enum nvme_fcctrl_flags {
- FCCTRL_TERMIO = (1 << 0),
-};
+/* fc_ctrl flags values - specified as bit positions */
+#define ASSOC_ACTIVE 0
+#define ASSOC_FAILED 1
+#define FCCTRL_TERMIO 2
struct nvme_fc_ctrl {
spinlock_t lock;
@@ -139,20 +158,19 @@
u32 cnum;
bool ioq_live;
- bool assoc_active;
- atomic_t err_work_active;
u64 association_id;
+ struct nvmefc_ls_rcv_op *rcv_disconn;
struct list_head ctrl_list; /* rport->ctrl_list */
struct blk_mq_tag_set admin_tag_set;
struct blk_mq_tag_set tag_set;
+ struct work_struct ioerr_work;
struct delayed_work connect_work;
- struct work_struct err_work;
struct kref ref;
- u32 flags;
+ unsigned long flags;
u32 iocnt;
wait_queue_head_t ioabort_wait;
@@ -213,12 +231,16 @@
*/
static struct device *fc_udev_device;
+static void nvme_fc_complete_rq(struct request *rq);
/* *********************** FC-NVME Port Management ************************ */
static void __nvme_fc_delete_hw_queue(struct nvme_fc_ctrl *,
struct nvme_fc_queue *, unsigned int);
+static void nvme_fc_handle_ls_rqst_work(struct work_struct *work);
+
+
static void
nvme_fc_free_lport(struct kref *ref)
{
@@ -394,7 +416,10 @@
newrec->ops = template;
newrec->dev = dev;
ida_init(&newrec->endp_cnt);
- newrec->localport.private = &newrec[1];
+ if (template->local_priv_sz)
+ newrec->localport.private = &newrec[1];
+ else
+ newrec->localport.private = NULL;
newrec->localport.node_name = pinfo->node_name;
newrec->localport.port_name = pinfo->port_name;
newrec->localport.port_role = pinfo->port_role;
@@ -701,9 +726,13 @@
atomic_set(&newrec->act_ctrl_cnt, 0);
spin_lock_init(&newrec->lock);
newrec->remoteport.localport = &lport->localport;
+ INIT_LIST_HEAD(&newrec->ls_rcv_list);
newrec->dev = lport->dev;
newrec->lport = lport;
- newrec->remoteport.private = &newrec[1];
+ if (lport->ops->remote_priv_sz)
+ newrec->remoteport.private = &newrec[1];
+ else
+ newrec->remoteport.private = NULL;
newrec->remoteport.port_role = pinfo->port_role;
newrec->remoteport.node_name = pinfo->node_name;
newrec->remoteport.port_name = pinfo->port_name;
@@ -711,6 +740,7 @@
newrec->remoteport.port_state = FC_OBJSTATE_ONLINE;
newrec->remoteport.port_num = idx;
__nvme_fc_set_dev_loss_tmo(newrec, pinfo);
+ INIT_WORK(&newrec->lsrcv_work, nvme_fc_handle_ls_rqst_work);
spin_lock_irqsave(&nvme_fc_lock, flags);
list_add_tail(&newrec->endp_list, &lport->endp_list);
@@ -800,6 +830,7 @@
break;
case NVME_CTRL_DELETING:
+ case NVME_CTRL_DELETING_NOIO:
default:
/* no action to take - let it delete */
break;
@@ -1000,6 +1031,7 @@
static void nvme_fc_ctrl_put(struct nvme_fc_ctrl *);
static int nvme_fc_ctrl_get(struct nvme_fc_ctrl *);
+static void nvme_fc_error_recovery(struct nvme_fc_ctrl *ctrl, char *errmsg);
static void
__nvme_fc_finish_ls_req(struct nvmefc_ls_req_op *lsop)
@@ -1140,41 +1172,6 @@
return __nvme_fc_send_ls_req(rport, lsop, done);
}
-/* Validation Error indexes into the string table below */
-enum {
- VERR_NO_ERROR = 0,
- VERR_LSACC = 1,
- VERR_LSDESC_RQST = 2,
- VERR_LSDESC_RQST_LEN = 3,
- VERR_ASSOC_ID = 4,
- VERR_ASSOC_ID_LEN = 5,
- VERR_CONN_ID = 6,
- VERR_CONN_ID_LEN = 7,
- VERR_CR_ASSOC = 8,
- VERR_CR_ASSOC_ACC_LEN = 9,
- VERR_CR_CONN = 10,
- VERR_CR_CONN_ACC_LEN = 11,
- VERR_DISCONN = 12,
- VERR_DISCONN_ACC_LEN = 13,
-};
-
-static char *validation_errors[] = {
- "OK",
- "Not LS_ACC",
- "Not LSDESC_RQST",
- "Bad LSDESC_RQST Length",
- "Not Association ID",
- "Bad Association ID Length",
- "Not Connection ID",
- "Bad Connection ID Length",
- "Not CR_ASSOC Rqst",
- "Bad CR_ASSOC ACC Length",
- "Not CR_CONN Rqst",
- "Bad CR_CONN ACC Length",
- "Not Disconnect Rqst",
- "Bad Disconnect ACC Length",
-};
-
static int
nvme_fc_connect_admin_queue(struct nvme_fc_ctrl *ctrl,
struct nvme_fc_queue *queue, u16 qsize, u16 ersp_ratio)
@@ -1183,21 +1180,27 @@
struct nvmefc_ls_req *lsreq;
struct fcnvme_ls_cr_assoc_rqst *assoc_rqst;
struct fcnvme_ls_cr_assoc_acc *assoc_acc;
+ unsigned long flags;
int ret, fcret = 0;
lsop = kzalloc((sizeof(*lsop) +
- ctrl->lport->ops->lsrqst_priv_sz +
- sizeof(*assoc_rqst) + sizeof(*assoc_acc)), GFP_KERNEL);
+ sizeof(*assoc_rqst) + sizeof(*assoc_acc) +
+ ctrl->lport->ops->lsrqst_priv_sz), GFP_KERNEL);
if (!lsop) {
+ dev_info(ctrl->ctrl.device,
+ "NVME-FC{%d}: send Create Association failed: ENOMEM\n",
+ ctrl->cnum);
ret = -ENOMEM;
goto out_no_memory;
}
- lsreq = &lsop->ls_req;
- lsreq->private = (void *)&lsop[1];
- assoc_rqst = (struct fcnvme_ls_cr_assoc_rqst *)
- (lsreq->private + ctrl->lport->ops->lsrqst_priv_sz);
+ assoc_rqst = (struct fcnvme_ls_cr_assoc_rqst *)&lsop[1];
assoc_acc = (struct fcnvme_ls_cr_assoc_acc *)&assoc_rqst[1];
+ lsreq = &lsop->ls_req;
+ if (ctrl->lport->ops->lsrqst_priv_sz)
+ lsreq->private = &assoc_acc[1];
+ else
+ lsreq->private = NULL;
assoc_rqst->w0.ls_cmd = FCNVME_LS_CREATE_ASSOCIATION;
assoc_rqst->desc_list_len =
@@ -1224,7 +1227,7 @@
lsreq->rqstlen = sizeof(*assoc_rqst);
lsreq->rspaddr = assoc_acc;
lsreq->rsplen = sizeof(*assoc_acc);
- lsreq->timeout = NVME_FC_CONNECT_TIMEOUT_SEC;
+ lsreq->timeout = NVME_FC_LS_TIMEOUT_SEC;
ret = nvme_fc_send_ls_req(ctrl->rport, lsop);
if (ret)
@@ -1264,14 +1267,16 @@
if (fcret) {
ret = -EBADF;
dev_err(ctrl->dev,
- "q %d connect failed: %s\n",
+ "q %d Create Association LS failed: %s\n",
queue->qnum, validation_errors[fcret]);
} else {
+ spin_lock_irqsave(&ctrl->lock, flags);
ctrl->association_id =
be64_to_cpu(assoc_acc->associd.association_id);
queue->connection_id =
be64_to_cpu(assoc_acc->connectid.connection_id);
set_bit(NVME_FC_Q_CONNECTED, &queue->flags);
+ spin_unlock_irqrestore(&ctrl->lock, flags);
}
out_free_buffer:
@@ -1295,18 +1300,23 @@
int ret, fcret = 0;
lsop = kzalloc((sizeof(*lsop) +
- ctrl->lport->ops->lsrqst_priv_sz +
- sizeof(*conn_rqst) + sizeof(*conn_acc)), GFP_KERNEL);
+ sizeof(*conn_rqst) + sizeof(*conn_acc) +
+ ctrl->lport->ops->lsrqst_priv_sz), GFP_KERNEL);
if (!lsop) {
+ dev_info(ctrl->ctrl.device,
+ "NVME-FC{%d}: send Create Connection failed: ENOMEM\n",
+ ctrl->cnum);
ret = -ENOMEM;
goto out_no_memory;
}
- lsreq = &lsop->ls_req;
- lsreq->private = (void *)&lsop[1];
- conn_rqst = (struct fcnvme_ls_cr_conn_rqst *)
- (lsreq->private + ctrl->lport->ops->lsrqst_priv_sz);
+ conn_rqst = (struct fcnvme_ls_cr_conn_rqst *)&lsop[1];
conn_acc = (struct fcnvme_ls_cr_conn_acc *)&conn_rqst[1];
+ lsreq = &lsop->ls_req;
+ if (ctrl->lport->ops->lsrqst_priv_sz)
+ lsreq->private = (void *)&conn_acc[1];
+ else
+ lsreq->private = NULL;
conn_rqst->w0.ls_cmd = FCNVME_LS_CREATE_CONNECTION;
conn_rqst->desc_list_len = cpu_to_be32(
@@ -1332,7 +1342,7 @@
lsreq->rqstlen = sizeof(*conn_rqst);
lsreq->rspaddr = conn_acc;
lsreq->rsplen = sizeof(*conn_acc);
- lsreq->timeout = NVME_FC_CONNECT_TIMEOUT_SEC;
+ lsreq->timeout = NVME_FC_LS_TIMEOUT_SEC;
ret = nvme_fc_send_ls_req(ctrl->rport, lsop);
if (ret)
@@ -1363,7 +1373,7 @@
if (fcret) {
ret = -EBADF;
dev_err(ctrl->dev,
- "q %d connect failed: %s\n",
+ "q %d Create I/O Connection LS failed: %s\n",
queue->qnum, validation_errors[fcret]);
} else {
queue->connection_id =
@@ -1376,7 +1386,7 @@
out_no_memory:
if (ret)
dev_err(ctrl->dev,
- "queue %d connect command failed (%d).\n",
+ "queue %d connect I/O queue failed (%d).\n",
queue->qnum, ret);
return ret;
}
@@ -1413,67 +1423,393 @@
static void
nvme_fc_xmt_disconnect_assoc(struct nvme_fc_ctrl *ctrl)
{
- struct fcnvme_ls_disconnect_rqst *discon_rqst;
- struct fcnvme_ls_disconnect_acc *discon_acc;
+ struct fcnvme_ls_disconnect_assoc_rqst *discon_rqst;
+ struct fcnvme_ls_disconnect_assoc_acc *discon_acc;
struct nvmefc_ls_req_op *lsop;
struct nvmefc_ls_req *lsreq;
int ret;
lsop = kzalloc((sizeof(*lsop) +
- ctrl->lport->ops->lsrqst_priv_sz +
- sizeof(*discon_rqst) + sizeof(*discon_acc)),
- GFP_KERNEL);
- if (!lsop)
- /* couldn't sent it... too bad */
+ sizeof(*discon_rqst) + sizeof(*discon_acc) +
+ ctrl->lport->ops->lsrqst_priv_sz), GFP_KERNEL);
+ if (!lsop) {
+ dev_info(ctrl->ctrl.device,
+ "NVME-FC{%d}: send Disconnect Association "
+ "failed: ENOMEM\n",
+ ctrl->cnum);
return;
+ }
+ discon_rqst = (struct fcnvme_ls_disconnect_assoc_rqst *)&lsop[1];
+ discon_acc = (struct fcnvme_ls_disconnect_assoc_acc *)&discon_rqst[1];
lsreq = &lsop->ls_req;
+ if (ctrl->lport->ops->lsrqst_priv_sz)
+ lsreq->private = (void *)&discon_acc[1];
+ else
+ lsreq->private = NULL;
- lsreq->private = (void *)&lsop[1];
- discon_rqst = (struct fcnvme_ls_disconnect_rqst *)
- (lsreq->private + ctrl->lport->ops->lsrqst_priv_sz);
- discon_acc = (struct fcnvme_ls_disconnect_acc *)&discon_rqst[1];
-
- discon_rqst->w0.ls_cmd = FCNVME_LS_DISCONNECT;
- discon_rqst->desc_list_len = cpu_to_be32(
- sizeof(struct fcnvme_lsdesc_assoc_id) +
- sizeof(struct fcnvme_lsdesc_disconn_cmd));
-
- discon_rqst->associd.desc_tag = cpu_to_be32(FCNVME_LSDESC_ASSOC_ID);
- discon_rqst->associd.desc_len =
- fcnvme_lsdesc_len(
- sizeof(struct fcnvme_lsdesc_assoc_id));
-
- discon_rqst->associd.association_id = cpu_to_be64(ctrl->association_id);
-
- discon_rqst->discon_cmd.desc_tag = cpu_to_be32(
- FCNVME_LSDESC_DISCONN_CMD);
- discon_rqst->discon_cmd.desc_len =
- fcnvme_lsdesc_len(
- sizeof(struct fcnvme_lsdesc_disconn_cmd));
- discon_rqst->discon_cmd.scope = FCNVME_DISCONN_ASSOCIATION;
- discon_rqst->discon_cmd.id = cpu_to_be64(ctrl->association_id);
-
- lsreq->rqstaddr = discon_rqst;
- lsreq->rqstlen = sizeof(*discon_rqst);
- lsreq->rspaddr = discon_acc;
- lsreq->rsplen = sizeof(*discon_acc);
- lsreq->timeout = NVME_FC_CONNECT_TIMEOUT_SEC;
+ nvmefc_fmt_lsreq_discon_assoc(lsreq, discon_rqst, discon_acc,
+ ctrl->association_id);
ret = nvme_fc_send_ls_req_async(ctrl->rport, lsop,
nvme_fc_disconnect_assoc_done);
if (ret)
kfree(lsop);
-
- /* only meaningful part to terminating the association */
- ctrl->association_id = 0;
}
+static void
+nvme_fc_xmt_ls_rsp_done(struct nvmefc_ls_rsp *lsrsp)
+{
+ struct nvmefc_ls_rcv_op *lsop = lsrsp->nvme_fc_private;
+ struct nvme_fc_rport *rport = lsop->rport;
+ struct nvme_fc_lport *lport = rport->lport;
+ unsigned long flags;
+
+ spin_lock_irqsave(&rport->lock, flags);
+ list_del(&lsop->lsrcv_list);
+ spin_unlock_irqrestore(&rport->lock, flags);
+
+ fc_dma_sync_single_for_cpu(lport->dev, lsop->rspdma,
+ sizeof(*lsop->rspbuf), DMA_TO_DEVICE);
+ fc_dma_unmap_single(lport->dev, lsop->rspdma,
+ sizeof(*lsop->rspbuf), DMA_TO_DEVICE);
+
+ kfree(lsop);
+
+ nvme_fc_rport_put(rport);
+}
+
+static void
+nvme_fc_xmt_ls_rsp(struct nvmefc_ls_rcv_op *lsop)
+{
+ struct nvme_fc_rport *rport = lsop->rport;
+ struct nvme_fc_lport *lport = rport->lport;
+ struct fcnvme_ls_rqst_w0 *w0 = &lsop->rqstbuf->w0;
+ int ret;
+
+ fc_dma_sync_single_for_device(lport->dev, lsop->rspdma,
+ sizeof(*lsop->rspbuf), DMA_TO_DEVICE);
+
+ ret = lport->ops->xmt_ls_rsp(&lport->localport, &rport->remoteport,
+ lsop->lsrsp);
+ if (ret) {
+ dev_warn(lport->dev,
+ "LLDD rejected LS RSP xmt: LS %d status %d\n",
+ w0->ls_cmd, ret);
+ nvme_fc_xmt_ls_rsp_done(lsop->lsrsp);
+ return;
+ }
+}
+
+static struct nvme_fc_ctrl *
+nvme_fc_match_disconn_ls(struct nvme_fc_rport *rport,
+ struct nvmefc_ls_rcv_op *lsop)
+{
+ struct fcnvme_ls_disconnect_assoc_rqst *rqst =
+ &lsop->rqstbuf->rq_dis_assoc;
+ struct nvme_fc_ctrl *ctrl, *ret = NULL;
+ struct nvmefc_ls_rcv_op *oldls = NULL;
+ u64 association_id = be64_to_cpu(rqst->associd.association_id);
+ unsigned long flags;
+
+ spin_lock_irqsave(&rport->lock, flags);
+
+ list_for_each_entry(ctrl, &rport->ctrl_list, ctrl_list) {
+ if (!nvme_fc_ctrl_get(ctrl))
+ continue;
+ spin_lock(&ctrl->lock);
+ if (association_id == ctrl->association_id) {
+ oldls = ctrl->rcv_disconn;
+ ctrl->rcv_disconn = lsop;
+ ret = ctrl;
+ }
+ spin_unlock(&ctrl->lock);
+ if (ret)
+ /* leave the ctrl get reference */
+ break;
+ nvme_fc_ctrl_put(ctrl);
+ }
+
+ spin_unlock_irqrestore(&rport->lock, flags);
+
+ /* transmit a response for anything that was pending */
+ if (oldls) {
+ dev_info(rport->lport->dev,
+ "NVME-FC{%d}: Multiple Disconnect Association "
+ "LS's received\n", ctrl->cnum);
+ /* overwrite good response with bogus failure */
+ oldls->lsrsp->rsplen = nvme_fc_format_rjt(oldls->rspbuf,
+ sizeof(*oldls->rspbuf),
+ rqst->w0.ls_cmd,
+ FCNVME_RJT_RC_UNAB,
+ FCNVME_RJT_EXP_NONE, 0);
+ nvme_fc_xmt_ls_rsp(oldls);
+ }
+
+ return ret;
+}
+
+/*
+ * returns true to mean LS handled and ls_rsp can be sent
+ * returns false to defer ls_rsp xmt (will be done as part of
+ * association termination)
+ */
+static bool
+nvme_fc_ls_disconnect_assoc(struct nvmefc_ls_rcv_op *lsop)
+{
+ struct nvme_fc_rport *rport = lsop->rport;
+ struct fcnvme_ls_disconnect_assoc_rqst *rqst =
+ &lsop->rqstbuf->rq_dis_assoc;
+ struct fcnvme_ls_disconnect_assoc_acc *acc =
+ &lsop->rspbuf->rsp_dis_assoc;
+ struct nvme_fc_ctrl *ctrl = NULL;
+ int ret = 0;
+
+ memset(acc, 0, sizeof(*acc));
+
+ ret = nvmefc_vldt_lsreq_discon_assoc(lsop->rqstdatalen, rqst);
+ if (!ret) {
+ /* match an active association */
+ ctrl = nvme_fc_match_disconn_ls(rport, lsop);
+ if (!ctrl)
+ ret = VERR_NO_ASSOC;
+ }
+
+ if (ret) {
+ dev_info(rport->lport->dev,
+ "Disconnect LS failed: %s\n",
+ validation_errors[ret]);
+ lsop->lsrsp->rsplen = nvme_fc_format_rjt(acc,
+ sizeof(*acc), rqst->w0.ls_cmd,
+ (ret == VERR_NO_ASSOC) ?
+ FCNVME_RJT_RC_INV_ASSOC :
+ FCNVME_RJT_RC_LOGIC,
+ FCNVME_RJT_EXP_NONE, 0);
+ return true;
+ }
+
+ /* format an ACCept response */
+
+ lsop->lsrsp->rsplen = sizeof(*acc);
+
+ nvme_fc_format_rsp_hdr(acc, FCNVME_LS_ACC,
+ fcnvme_lsdesc_len(
+ sizeof(struct fcnvme_ls_disconnect_assoc_acc)),
+ FCNVME_LS_DISCONNECT_ASSOC);
+
+ /*
+ * the transmit of the response will occur after the exchanges
+ * for the association have been ABTS'd by
+ * nvme_fc_delete_association().
+ */
+
+ /* fail the association */
+ nvme_fc_error_recovery(ctrl, "Disconnect Association LS received");
+
+ /* release the reference taken by nvme_fc_match_disconn_ls() */
+ nvme_fc_ctrl_put(ctrl);
+
+ return false;
+}
+
+/*
+ * Actual Processing routine for received FC-NVME LS Requests from the LLD
+ * returns true if a response should be sent afterward, false if rsp will
+ * be sent asynchronously.
+ */
+static bool
+nvme_fc_handle_ls_rqst(struct nvmefc_ls_rcv_op *lsop)
+{
+ struct fcnvme_ls_rqst_w0 *w0 = &lsop->rqstbuf->w0;
+ bool ret = true;
+
+ lsop->lsrsp->nvme_fc_private = lsop;
+ lsop->lsrsp->rspbuf = lsop->rspbuf;
+ lsop->lsrsp->rspdma = lsop->rspdma;
+ lsop->lsrsp->done = nvme_fc_xmt_ls_rsp_done;
+ /* Be preventative. handlers will later set to valid length */
+ lsop->lsrsp->rsplen = 0;
+
+ /*
+ * handlers:
+ * parse request input, execute the request, and format the
+ * LS response
+ */
+ switch (w0->ls_cmd) {
+ case FCNVME_LS_DISCONNECT_ASSOC:
+ ret = nvme_fc_ls_disconnect_assoc(lsop);
+ break;
+ case FCNVME_LS_DISCONNECT_CONN:
+ lsop->lsrsp->rsplen = nvme_fc_format_rjt(lsop->rspbuf,
+ sizeof(*lsop->rspbuf), w0->ls_cmd,
+ FCNVME_RJT_RC_UNSUP, FCNVME_RJT_EXP_NONE, 0);
+ break;
+ case FCNVME_LS_CREATE_ASSOCIATION:
+ case FCNVME_LS_CREATE_CONNECTION:
+ lsop->lsrsp->rsplen = nvme_fc_format_rjt(lsop->rspbuf,
+ sizeof(*lsop->rspbuf), w0->ls_cmd,
+ FCNVME_RJT_RC_LOGIC, FCNVME_RJT_EXP_NONE, 0);
+ break;
+ default:
+ lsop->lsrsp->rsplen = nvme_fc_format_rjt(lsop->rspbuf,
+ sizeof(*lsop->rspbuf), w0->ls_cmd,
+ FCNVME_RJT_RC_INVAL, FCNVME_RJT_EXP_NONE, 0);
+ break;
+ }
+
+ return(ret);
+}
+
+static void
+nvme_fc_handle_ls_rqst_work(struct work_struct *work)
+{
+ struct nvme_fc_rport *rport =
+ container_of(work, struct nvme_fc_rport, lsrcv_work);
+ struct fcnvme_ls_rqst_w0 *w0;
+ struct nvmefc_ls_rcv_op *lsop;
+ unsigned long flags;
+ bool sendrsp;
+
+restart:
+ sendrsp = true;
+ spin_lock_irqsave(&rport->lock, flags);
+ list_for_each_entry(lsop, &rport->ls_rcv_list, lsrcv_list) {
+ if (lsop->handled)
+ continue;
+
+ lsop->handled = true;
+ if (rport->remoteport.port_state == FC_OBJSTATE_ONLINE) {
+ spin_unlock_irqrestore(&rport->lock, flags);
+ sendrsp = nvme_fc_handle_ls_rqst(lsop);
+ } else {
+ spin_unlock_irqrestore(&rport->lock, flags);
+ w0 = &lsop->rqstbuf->w0;
+ lsop->lsrsp->rsplen = nvme_fc_format_rjt(
+ lsop->rspbuf,
+ sizeof(*lsop->rspbuf),
+ w0->ls_cmd,
+ FCNVME_RJT_RC_UNAB,
+ FCNVME_RJT_EXP_NONE, 0);
+ }
+ if (sendrsp)
+ nvme_fc_xmt_ls_rsp(lsop);
+ goto restart;
+ }
+ spin_unlock_irqrestore(&rport->lock, flags);
+}
+
+/**
+ * nvme_fc_rcv_ls_req - transport entry point called by an LLDD
+ * upon the reception of a NVME LS request.
+ *
+ * The nvme-fc layer will copy payload to an internal structure for
+ * processing. As such, upon completion of the routine, the LLDD may
+ * immediately free/reuse the LS request buffer passed in the call.
+ *
+ * If this routine returns error, the LLDD should abort the exchange.
+ *
+ * @remoteport: pointer to the (registered) remote port that the LS
+ * was received from. The remoteport is associated with
+ * a specific localport.
+ * @lsrsp: pointer to a nvmefc_ls_rsp response structure to be
+ * used to reference the exchange corresponding to the LS
+ * when issuing an ls response.
+ * @lsreqbuf: pointer to the buffer containing the LS Request
+ * @lsreqbuf_len: length, in bytes, of the received LS request
+ */
+int
+nvme_fc_rcv_ls_req(struct nvme_fc_remote_port *portptr,
+ struct nvmefc_ls_rsp *lsrsp,
+ void *lsreqbuf, u32 lsreqbuf_len)
+{
+ struct nvme_fc_rport *rport = remoteport_to_rport(portptr);
+ struct nvme_fc_lport *lport = rport->lport;
+ struct fcnvme_ls_rqst_w0 *w0 = (struct fcnvme_ls_rqst_w0 *)lsreqbuf;
+ struct nvmefc_ls_rcv_op *lsop;
+ unsigned long flags;
+ int ret;
+
+ nvme_fc_rport_get(rport);
+
+ /* validate there's a routine to transmit a response */
+ if (!lport->ops->xmt_ls_rsp) {
+ dev_info(lport->dev,
+ "RCV %s LS failed: no LLDD xmt_ls_rsp\n",
+ (w0->ls_cmd <= NVME_FC_LAST_LS_CMD_VALUE) ?
+ nvmefc_ls_names[w0->ls_cmd] : "");
+ ret = -EINVAL;
+ goto out_put;
+ }
+
+ if (lsreqbuf_len > sizeof(union nvmefc_ls_requests)) {
+ dev_info(lport->dev,
+ "RCV %s LS failed: payload too large\n",
+ (w0->ls_cmd <= NVME_FC_LAST_LS_CMD_VALUE) ?
+ nvmefc_ls_names[w0->ls_cmd] : "");
+ ret = -E2BIG;
+ goto out_put;
+ }
+
+ lsop = kzalloc(sizeof(*lsop) +
+ sizeof(union nvmefc_ls_requests) +
+ sizeof(union nvmefc_ls_responses),
+ GFP_KERNEL);
+ if (!lsop) {
+ dev_info(lport->dev,
+ "RCV %s LS failed: No memory\n",
+ (w0->ls_cmd <= NVME_FC_LAST_LS_CMD_VALUE) ?
+ nvmefc_ls_names[w0->ls_cmd] : "");
+ ret = -ENOMEM;
+ goto out_put;
+ }
+ lsop->rqstbuf = (union nvmefc_ls_requests *)&lsop[1];
+ lsop->rspbuf = (union nvmefc_ls_responses *)&lsop->rqstbuf[1];
+
+ lsop->rspdma = fc_dma_map_single(lport->dev, lsop->rspbuf,
+ sizeof(*lsop->rspbuf),
+ DMA_TO_DEVICE);
+ if (fc_dma_mapping_error(lport->dev, lsop->rspdma)) {
+ dev_info(lport->dev,
+ "RCV %s LS failed: DMA mapping failure\n",
+ (w0->ls_cmd <= NVME_FC_LAST_LS_CMD_VALUE) ?
+ nvmefc_ls_names[w0->ls_cmd] : "");
+ ret = -EFAULT;
+ goto out_free;
+ }
+
+ lsop->rport = rport;
+ lsop->lsrsp = lsrsp;
+
+ memcpy(lsop->rqstbuf, lsreqbuf, lsreqbuf_len);
+ lsop->rqstdatalen = lsreqbuf_len;
+
+ spin_lock_irqsave(&rport->lock, flags);
+ if (rport->remoteport.port_state != FC_OBJSTATE_ONLINE) {
+ spin_unlock_irqrestore(&rport->lock, flags);
+ ret = -ENOTCONN;
+ goto out_unmap;
+ }
+ list_add_tail(&lsop->lsrcv_list, &rport->ls_rcv_list);
+ spin_unlock_irqrestore(&rport->lock, flags);
+
+ schedule_work(&rport->lsrcv_work);
+
+ return 0;
+
+out_unmap:
+ fc_dma_unmap_single(lport->dev, lsop->rspdma,
+ sizeof(*lsop->rspbuf), DMA_TO_DEVICE);
+out_free:
+ kfree(lsop);
+out_put:
+ nvme_fc_rport_put(rport);
+ return ret;
+}
+EXPORT_SYMBOL_GPL(nvme_fc_rcv_ls_req);
+
/* *********************** NVME Ctrl Routines **************************** */
-static void nvme_fc_error_recovery(struct nvme_fc_ctrl *ctrl, char *errmsg);
-
static void
__nvme_fc_exit_request(struct nvme_fc_ctrl *ctrl,
struct nvme_fc_fcp_op *op)
@@ -1505,8 +1841,10 @@
opstate = atomic_xchg(&op->state, FCPOP_STATE_ABORTED);
if (opstate != FCPOP_STATE_ACTIVE)
atomic_set(&op->state, opstate);
- else if (ctrl->flags & FCCTRL_TERMIO)
+ else if (test_bit(FCCTRL_TERMIO, &ctrl->flags)) {
+ op->flags |= FCOP_FLAGS_TERMIO;
ctrl->iocnt++;
+ }
spin_unlock_irqrestore(&ctrl->lock, flags);
if (opstate != FCPOP_STATE_ACTIVE)
@@ -1542,7 +1880,8 @@
if (opstate == FCPOP_STATE_ABORTED) {
spin_lock_irqsave(&ctrl->lock, flags);
- if (ctrl->flags & FCCTRL_TERMIO) {
+ if (test_bit(FCCTRL_TERMIO, &ctrl->flags) &&
+ op->flags & FCOP_FLAGS_TERMIO) {
if (!--ctrl->iocnt)
wake_up(&ctrl->ioabort_wait);
}
@@ -1551,6 +1890,15 @@
}
static void
+nvme_fc_ctrl_ioerr_work(struct work_struct *work)
+{
+ struct nvme_fc_ctrl *ctrl =
+ container_of(work, struct nvme_fc_ctrl, ioerr_work);
+
+ nvme_fc_error_recovery(ctrl, "transport detected io error");
+}
+
+static void
nvme_fc_fcpio_done(struct nvmefc_fcp_req *req)
{
struct nvme_fc_fcp_op *op = fcp_req_to_fcp_op(req);
@@ -1662,7 +2010,7 @@
(freq->rcv_rsplen / 4) ||
be32_to_cpu(op->rsp_iu.xfrd_len) !=
freq->transferred_length ||
- op->rsp_iu.status_code ||
+ op->rsp_iu.ersp_result ||
sqe->common.command_id != cqe->command_id)) {
status = cpu_to_le16(NVME_SC_HOST_PATH_ERROR << 1);
dev_info(ctrl->ctrl.device,
@@ -1672,7 +2020,7 @@
ctrl->cnum, be16_to_cpu(op->rsp_iu.iu_len),
be32_to_cpu(op->rsp_iu.xfrd_len),
freq->transferred_length,
- op->rsp_iu.status_code,
+ op->rsp_iu.ersp_result,
sqe->common.command_id,
cqe->command_id);
goto done;
@@ -1703,11 +2051,12 @@
}
__nvme_fc_fcpop_chk_teardowns(ctrl, op, opstate);
- nvme_end_request(rq, status, result);
+ if (!nvme_try_complete_req(rq, status, result))
+ nvme_fc_complete_rq(rq);
check_error:
- if (terminate_assoc)
- nvme_fc_error_recovery(ctrl, "transport detected io error");
+ if (terminate_assoc && ctrl->ctrl.state != NVME_CTRL_RESETTING)
+ queue_work(nvme_reset_wq, &ctrl->ioerr_work);
}
static int
@@ -1731,9 +2080,14 @@
op->rq = rq;
op->rqno = rqno;
- cmdiu->scsi_id = NVME_CMD_SCSI_ID;
+ cmdiu->format_id = NVME_CMD_FORMAT_ID;
cmdiu->fc_id = NVME_CMD_FC_ID;
cmdiu->iu_len = cpu_to_be16(sizeof(*cmdiu) / sizeof(u32));
+ if (queue->qnum)
+ cmdiu->rsv_cat = fccmnd_set_cat_css(0,
+ (NVME_CC_CSS_NVM >> NVME_CC_CSS_SHIFT));
+ else
+ cmdiu->rsv_cat = fccmnd_set_cat_admin(0);
op->fcp_req.cmddma = fc_dma_map_single(ctrl->lport->dev,
&op->cmd_iu, sizeof(op->cmd_iu), DMA_TO_DEVICE);
@@ -1771,7 +2125,7 @@
res = __nvme_fc_init_request(ctrl, queue, &op->op, rq, queue->rqcnt++);
if (res)
return res;
- op->op.fcp_req.first_sgl = &op->sgl[0];
+ op->op.fcp_req.first_sgl = op->sgl;
op->op.fcp_req.private = &op->priv[0];
nvme_req(rq)->ctrl = &ctrl->ctrl;
return res;
@@ -1783,15 +2137,17 @@
struct nvme_fc_fcp_op *aen_op;
struct nvme_fc_cmd_iu *cmdiu;
struct nvme_command *sqe;
- void *private;
+ void *private = NULL;
int i, ret;
aen_op = ctrl->aen_ops;
for (i = 0; i < NVME_NR_AEN_COMMANDS; i++, aen_op++) {
- private = kzalloc(ctrl->lport->ops->fcprqst_priv_sz,
+ if (ctrl->lport->ops->fcprqst_priv_sz) {
+ private = kzalloc(ctrl->lport->ops->fcprqst_priv_sz,
GFP_KERNEL);
- if (!private)
- return -ENOMEM;
+ if (!private)
+ return -ENOMEM;
+ }
cmdiu = &aen_op->cmd_iu;
sqe = &cmdiu->sqe;
@@ -1823,9 +2179,6 @@
cancel_work_sync(&ctrl->ctrl.async_event_work);
aen_op = ctrl->aen_ops;
for (i = 0; i < NVME_NR_AEN_COMMANDS; i++, aen_op++) {
- if (!aen_op->fcp_req.private)
- continue;
-
__nvme_fc_exit_request(ctrl, aen_op);
kfree(aen_op->fcp_req.private);
@@ -1977,7 +2330,7 @@
return 0;
delete_queues:
- for (; i >= 0; i--)
+ for (; i > 0; i--)
__nvme_fc_delete_hw_queue(ctrl, &ctrl->queues[i], i);
return ret;
}
@@ -2070,24 +2423,112 @@
nvme_fc_ctrl_put(ctrl);
}
+/*
+ * This routine is used by the transport when it needs to find active
+ * io on a queue that is to be terminated. The transport uses
+ * blk_mq_tagset_busy_itr() to find the busy requests, which then invoke
+ * this routine to kill them on a 1 by 1 basis.
+ *
+ * As FC allocates FC exchange for each io, the transport must contact
+ * the LLDD to terminate the exchange, thus releasing the FC exchange.
+ * After terminating the exchange the LLDD will call the transport's
+ * normal io done path for the request, but it will have an aborted
+ * status. The done path will return the io request back to the block
+ * layer with an error status.
+ */
+static bool
+nvme_fc_terminate_exchange(struct request *req, void *data, bool reserved)
+{
+ struct nvme_ctrl *nctrl = data;
+ struct nvme_fc_ctrl *ctrl = to_fc_ctrl(nctrl);
+ struct nvme_fc_fcp_op *op = blk_mq_rq_to_pdu(req);
+
+ op->nreq.flags |= NVME_REQ_CANCELLED;
+ __nvme_fc_abort_op(ctrl, op);
+ return true;
+}
+
+/*
+ * This routine runs through all outstanding commands on the association
+ * and aborts them. This routine is typically be called by the
+ * delete_association routine. It is also called due to an error during
+ * reconnect. In that scenario, it is most likely a command that initializes
+ * the controller, including fabric Connect commands on io queues, that
+ * may have timed out or failed thus the io must be killed for the connect
+ * thread to see the error.
+ */
+static void
+__nvme_fc_abort_outstanding_ios(struct nvme_fc_ctrl *ctrl, bool start_queues)
+{
+ int q;
+
+ /*
+ * if aborting io, the queues are no longer good, mark them
+ * all as not live.
+ */
+ if (ctrl->ctrl.queue_count > 1) {
+ for (q = 1; q < ctrl->ctrl.queue_count; q++)
+ clear_bit(NVME_FC_Q_LIVE, &ctrl->queues[q].flags);
+ }
+ clear_bit(NVME_FC_Q_LIVE, &ctrl->queues[0].flags);
+
+ /*
+ * If io queues are present, stop them and terminate all outstanding
+ * ios on them. As FC allocates FC exchange for each io, the
+ * transport must contact the LLDD to terminate the exchange,
+ * thus releasing the FC exchange. We use blk_mq_tagset_busy_itr()
+ * to tell us what io's are busy and invoke a transport routine
+ * to kill them with the LLDD. After terminating the exchange
+ * the LLDD will call the transport's normal io done path, but it
+ * will have an aborted status. The done path will return the
+ * io requests back to the block layer as part of normal completions
+ * (but with error status).
+ */
+ if (ctrl->ctrl.queue_count > 1) {
+ nvme_stop_queues(&ctrl->ctrl);
+ nvme_sync_io_queues(&ctrl->ctrl);
+ blk_mq_tagset_busy_iter(&ctrl->tag_set,
+ nvme_fc_terminate_exchange, &ctrl->ctrl);
+ blk_mq_tagset_wait_completed_request(&ctrl->tag_set);
+ if (start_queues)
+ nvme_start_queues(&ctrl->ctrl);
+ }
+
+ /*
+ * Other transports, which don't have link-level contexts bound
+ * to sqe's, would try to gracefully shutdown the controller by
+ * writing the registers for shutdown and polling (call
+ * nvme_shutdown_ctrl()). Given a bunch of i/o was potentially
+ * just aborted and we will wait on those contexts, and given
+ * there was no indication of how live the controlelr is on the
+ * link, don't send more io to create more contexts for the
+ * shutdown. Let the controller fail via keepalive failure if
+ * its still present.
+ */
+
+ /*
+ * clean up the admin queue. Same thing as above.
+ */
+ blk_mq_quiesce_queue(ctrl->ctrl.admin_q);
+ blk_sync_queue(ctrl->ctrl.admin_q);
+ blk_mq_tagset_busy_iter(&ctrl->admin_tag_set,
+ nvme_fc_terminate_exchange, &ctrl->ctrl);
+ blk_mq_tagset_wait_completed_request(&ctrl->admin_tag_set);
+}
+
static void
nvme_fc_error_recovery(struct nvme_fc_ctrl *ctrl, char *errmsg)
{
- int active;
-
/*
- * if an error (io timeout, etc) while (re)connecting,
- * it's an error on creating the new association.
- * Start the error recovery thread if it hasn't already
- * been started. It is expected there could be multiple
- * ios hitting this path before things are cleaned up.
+ * if an error (io timeout, etc) while (re)connecting, the remote
+ * port requested terminating of the association (disconnect_ls)
+ * or an error (timeout or abort) occurred on an io while creating
+ * the controller. Abort any ios on the association and let the
+ * create_association error path resolve things.
*/
if (ctrl->ctrl.state == NVME_CTRL_CONNECTING) {
- active = atomic_xchg(&ctrl->err_work_active, 1);
- if (!active && !queue_work(nvme_fc_wq, &ctrl->err_work)) {
- atomic_set(&ctrl->err_work_active, 0);
- WARN_ON(1);
- }
+ __nvme_fc_abort_outstanding_ios(ctrl, true);
+ set_bit(ASSOC_FAILED, &ctrl->flags);
return;
}
@@ -2096,7 +2537,7 @@
return;
dev_warn(ctrl->ctrl.device,
- "NVME-FC{%d}: transport association error detected: %s\n",
+ "NVME-FC{%d}: transport association event: %s\n",
ctrl->cnum, errmsg);
dev_warn(ctrl->ctrl.device,
"NVME-FC{%d}: resetting controller\n", ctrl->cnum);
@@ -2109,15 +2550,20 @@
{
struct nvme_fc_fcp_op *op = blk_mq_rq_to_pdu(rq);
struct nvme_fc_ctrl *ctrl = op->ctrl;
+ struct nvme_fc_cmd_iu *cmdiu = &op->cmd_iu;
+ struct nvme_command *sqe = &cmdiu->sqe;
/*
- * we can't individually ABTS an io without affecting the queue,
- * thus killing the queue, and thus the association.
- * So resolve by performing a controller reset, which will stop
- * the host/io stack, terminate the association on the link,
- * and recreate an association on the link.
+ * Attempt to abort the offending command. Command completion
+ * will detect the aborted io and will fail the connection.
*/
- nvme_fc_error_recovery(ctrl, "io timeout error");
+ dev_info(ctrl->ctrl.device,
+ "NVME-FC{%d.%d}: io timeout: opcode %d fctype %d w10/11: "
+ "x%08x/x%08x\n",
+ ctrl->cnum, op->queue->qnum, sqe->common.opcode,
+ sqe->connect.fctype, sqe->common.cdw10, sqe->common.cdw11);
+ if (__nvme_fc_abort_op(ctrl, op))
+ nvme_fc_error_recovery(ctrl, "io timeout abort failed");
/*
* the io abort has been initiated. Have the reset timer
@@ -2142,7 +2588,7 @@
freq->sg_table.sgl = freq->first_sgl;
ret = sg_alloc_table_chained(&freq->sg_table,
blk_rq_nr_phys_segments(rq), freq->sg_table.sgl,
- SG_CHUNK_SIZE);
+ NVME_INLINE_SG_CNT);
if (ret)
return -ENOMEM;
@@ -2151,7 +2597,7 @@
freq->sg_cnt = fc_dma_map_sg(ctrl->lport->dev, freq->sg_table.sgl,
op->nents, rq_dma_dir(rq));
if (unlikely(freq->sg_cnt <= 0)) {
- sg_free_table_chained(&freq->sg_table, SG_CHUNK_SIZE);
+ sg_free_table_chained(&freq->sg_table, NVME_INLINE_SG_CNT);
freq->sg_cnt = 0;
return -EFAULT;
}
@@ -2174,9 +2620,7 @@
fc_dma_unmap_sg(ctrl->lport->dev, freq->sg_table.sgl, op->nents,
rq_dma_dir(rq));
- nvme_cleanup_cmd(rq);
-
- sg_free_table_chained(&freq->sg_table, SG_CHUNK_SIZE);
+ sg_free_table_chained(&freq->sg_table, NVME_INLINE_SG_CNT);
freq->sg_cnt = 0;
}
@@ -2303,8 +2747,10 @@
opstate = atomic_xchg(&op->state, FCPOP_STATE_COMPLETE);
__nvme_fc_fcpop_chk_teardowns(ctrl, op, opstate);
- if (!(op->flags & FCOP_FLAGS_AEN))
+ if (!(op->flags & FCOP_FLAGS_AEN)) {
nvme_fc_unmap_data(ctrl, op->rq, op);
+ nvme_cleanup_cmd(op->rq);
+ }
nvme_fc_ctrl_put(ctrl);
@@ -2368,16 +2814,9 @@
{
struct nvme_fc_ctrl *ctrl = to_fc_ctrl(arg);
struct nvme_fc_fcp_op *aen_op;
- unsigned long flags;
- bool terminating = false;
blk_status_t ret;
- spin_lock_irqsave(&ctrl->lock, flags);
- if (ctrl->flags & FCCTRL_TERMIO)
- terminating = true;
- spin_unlock_irqrestore(&ctrl->lock, flags);
-
- if (terminating)
+ if (test_bit(FCCTRL_TERMIO, &ctrl->flags))
return;
aen_op = &ctrl->aen_ops[0];
@@ -2396,36 +2835,13 @@
struct nvme_fc_ctrl *ctrl = op->ctrl;
atomic_set(&op->state, FCPOP_STATE_IDLE);
+ op->flags &= ~FCOP_FLAGS_TERMIO;
nvme_fc_unmap_data(ctrl, rq, op);
nvme_complete_rq(rq);
nvme_fc_ctrl_put(ctrl);
}
-/*
- * This routine is used by the transport when it needs to find active
- * io on a queue that is to be terminated. The transport uses
- * blk_mq_tagset_busy_itr() to find the busy requests, which then invoke
- * this routine to kill them on a 1 by 1 basis.
- *
- * As FC allocates FC exchange for each io, the transport must contact
- * the LLDD to terminate the exchange, thus releasing the FC exchange.
- * After terminating the exchange the LLDD will call the transport's
- * normal io done path for the request, but it will have an aborted
- * status. The done path will return the io request back to the block
- * layer with an error status.
- */
-static bool
-nvme_fc_terminate_exchange(struct request *req, void *data, bool reserved)
-{
- struct nvme_ctrl *nctrl = data;
- struct nvme_fc_ctrl *ctrl = to_fc_ctrl(nctrl);
- struct nvme_fc_fcp_op *op = blk_mq_rq_to_pdu(req);
-
- __nvme_fc_abort_op(ctrl, op);
- return true;
-}
-
static const struct blk_mq_ops nvme_fc_mq_ops = {
.queue_rq = nvme_fc_queue_rq,
@@ -2538,6 +2954,15 @@
if (ctrl->ctrl.queue_count == 1)
return 0;
+ if (prior_ioq_cnt != nr_io_queues) {
+ dev_info(ctrl->ctrl.device,
+ "reconnect: revising io queue count from %d to %d\n",
+ prior_ioq_cnt, nr_io_queues);
+ nvme_wait_freeze(&ctrl->ctrl);
+ blk_mq_update_nr_hw_queues(&ctrl->tag_set, nr_io_queues);
+ nvme_unfreeze(&ctrl->ctrl);
+ }
+
ret = nvme_fc_create_hw_io_queues(ctrl, ctrl->ctrl.sqsize + 1);
if (ret)
goto out_free_io_queues;
@@ -2546,12 +2971,6 @@
if (ret)
goto out_delete_hw_queues;
- if (prior_ioq_cnt != nr_io_queues)
- dev_info(ctrl->ctrl.device,
- "reconnect: revising io queue count from %d to %d\n",
- prior_ioq_cnt, nr_io_queues);
- blk_mq_update_nr_hw_queues(&ctrl->tag_set, nr_io_queues);
-
return 0;
out_delete_hw_queues:
@@ -2586,10 +3005,9 @@
struct nvme_fc_rport *rport = ctrl->rport;
u32 cnt;
- if (ctrl->assoc_active)
+ if (test_and_set_bit(ASSOC_ACTIVE, &ctrl->flags))
return 1;
- ctrl->assoc_active = true;
cnt = atomic_inc_return(&rport->act_ctrl_cnt);
if (cnt == 1)
nvme_fc_rport_active_on_lport(rport);
@@ -2604,7 +3022,7 @@
struct nvme_fc_lport *lport = rport->lport;
u32 cnt;
- /* ctrl->assoc_active=false will be set independently */
+ /* clearing of ctrl->flags ASSOC_ACTIVE bit is in association delete */
cnt = atomic_dec_return(&rport->act_ctrl_cnt);
if (cnt == 0) {
@@ -2624,6 +3042,8 @@
nvme_fc_create_association(struct nvme_fc_ctrl *ctrl)
{
struct nvmf_ctrl_options *opts = ctrl->ctrl.opts;
+ struct nvmefc_ls_rcv_op *disls = NULL;
+ unsigned long flags;
int ret;
bool changed;
@@ -2641,6 +3061,8 @@
ctrl->cnum, ctrl->lport->localport.port_name,
ctrl->rport->remoteport.port_name, ctrl->ctrl.opts->subsysnqn);
+ clear_bit(ASSOC_FAILED, &ctrl->flags);
+
/*
* Create the admin queue
*/
@@ -2669,16 +3091,17 @@
*/
ret = nvme_enable_ctrl(&ctrl->ctrl);
- if (ret)
+ if (ret || test_bit(ASSOC_FAILED, &ctrl->flags))
goto out_disconnect_admin_queue;
- ctrl->ctrl.max_hw_sectors =
- (ctrl->lport->ops->max_sgl_segments - 1) << (PAGE_SHIFT - 9);
+ ctrl->ctrl.max_segments = ctrl->lport->ops->max_sgl_segments;
+ ctrl->ctrl.max_hw_sectors = ctrl->ctrl.max_segments <<
+ (ilog2(SZ_4K) - 9);
blk_mq_unquiesce_queue(ctrl->ctrl.admin_q);
ret = nvme_init_identify(&ctrl->ctrl);
- if (ret)
+ if (ret || test_bit(ASSOC_FAILED, &ctrl->flags))
goto out_disconnect_admin_queue;
/* sanity checks */
@@ -2696,7 +3119,7 @@
/* warn if maxcmd is lower than queue_size */
dev_warn(ctrl->ctrl.device,
"queue_size %zu > ctrl maxcmd %u, reducing "
- "to queue_size\n",
+ "to maxcmd\n",
opts->queue_size, ctrl->ctrl.maxcmd);
opts->queue_size = ctrl->ctrl.maxcmd;
}
@@ -2704,7 +3127,8 @@
if (opts->queue_size > ctrl->ctrl.sqsize + 1) {
/* warn if sqsize is lower than queue_size */
dev_warn(ctrl->ctrl.device,
- "queue_size %zu > ctrl sqsize %u, clamping down\n",
+ "queue_size %zu > ctrl sqsize %u, reducing "
+ "to sqsize\n",
opts->queue_size, ctrl->ctrl.sqsize + 1);
opts->queue_size = ctrl->ctrl.sqsize + 1;
}
@@ -2722,9 +3146,9 @@
ret = nvme_fc_create_io_queues(ctrl);
else
ret = nvme_fc_recreate_io_queues(ctrl);
- if (ret)
- goto out_term_aen_ops;
}
+ if (ret || test_bit(ASSOC_FAILED, &ctrl->flags))
+ goto out_term_aen_ops;
changed = nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_LIVE);
@@ -2740,16 +3164,24 @@
out_disconnect_admin_queue:
/* send a Disconnect(association) LS to fc-nvme target */
nvme_fc_xmt_disconnect_assoc(ctrl);
+ spin_lock_irqsave(&ctrl->lock, flags);
+ ctrl->association_id = 0;
+ disls = ctrl->rcv_disconn;
+ ctrl->rcv_disconn = NULL;
+ spin_unlock_irqrestore(&ctrl->lock, flags);
+ if (disls)
+ nvme_fc_xmt_ls_rsp(disls);
out_delete_hw_queue:
__nvme_fc_delete_hw_queue(ctrl, &ctrl->queues[0], 0);
out_free_queue:
nvme_fc_free_queue(&ctrl->queues[0]);
- ctrl->assoc_active = false;
+ clear_bit(ASSOC_ACTIVE, &ctrl->flags);
nvme_fc_ctlr_inactive_on_rport(ctrl);
return ret;
}
+
/*
* This routine stops operation of the controller on the host side.
* On the host os stack side: Admin and IO queues are stopped,
@@ -2759,57 +3191,18 @@
static void
nvme_fc_delete_association(struct nvme_fc_ctrl *ctrl)
{
+ struct nvmefc_ls_rcv_op *disls = NULL;
unsigned long flags;
- if (!ctrl->assoc_active)
+ if (!test_and_clear_bit(ASSOC_ACTIVE, &ctrl->flags))
return;
- ctrl->assoc_active = false;
spin_lock_irqsave(&ctrl->lock, flags);
- ctrl->flags |= FCCTRL_TERMIO;
+ set_bit(FCCTRL_TERMIO, &ctrl->flags);
ctrl->iocnt = 0;
spin_unlock_irqrestore(&ctrl->lock, flags);
- /*
- * If io queues are present, stop them and terminate all outstanding
- * ios on them. As FC allocates FC exchange for each io, the
- * transport must contact the LLDD to terminate the exchange,
- * thus releasing the FC exchange. We use blk_mq_tagset_busy_itr()
- * to tell us what io's are busy and invoke a transport routine
- * to kill them with the LLDD. After terminating the exchange
- * the LLDD will call the transport's normal io done path, but it
- * will have an aborted status. The done path will return the
- * io requests back to the block layer as part of normal completions
- * (but with error status).
- */
- if (ctrl->ctrl.queue_count > 1) {
- nvme_stop_queues(&ctrl->ctrl);
- blk_mq_tagset_busy_iter(&ctrl->tag_set,
- nvme_fc_terminate_exchange, &ctrl->ctrl);
- blk_mq_tagset_wait_completed_request(&ctrl->tag_set);
- }
-
- /*
- * Other transports, which don't have link-level contexts bound
- * to sqe's, would try to gracefully shutdown the controller by
- * writing the registers for shutdown and polling (call
- * nvme_shutdown_ctrl()). Given a bunch of i/o was potentially
- * just aborted and we will wait on those contexts, and given
- * there was no indication of how live the controlelr is on the
- * link, don't send more io to create more contexts for the
- * shutdown. Let the controller fail via keepalive failure if
- * its still present.
- */
-
- /*
- * clean up the admin queue. Same thing as above.
- * use blk_mq_tagset_busy_itr() and the transport routine to
- * terminate the exchanges.
- */
- blk_mq_quiesce_queue(ctrl->ctrl.admin_q);
- blk_mq_tagset_busy_iter(&ctrl->admin_tag_set,
- nvme_fc_terminate_exchange, &ctrl->ctrl);
- blk_mq_tagset_wait_completed_request(&ctrl->admin_tag_set);
+ __nvme_fc_abort_outstanding_ios(ctrl, false);
/* kill the aens as they are a separate path */
nvme_fc_abort_aen_ops(ctrl);
@@ -2817,7 +3210,7 @@
/* wait for all io that had to be aborted */
spin_lock_irq(&ctrl->lock);
wait_event_lock_irq(ctrl->ioabort_wait, ctrl->iocnt == 0, ctrl->lock);
- ctrl->flags &= ~FCCTRL_TERMIO;
+ clear_bit(FCCTRL_TERMIO, &ctrl->flags);
spin_unlock_irq(&ctrl->lock);
nvme_fc_term_aen_ops(ctrl);
@@ -2831,6 +3224,18 @@
if (ctrl->association_id)
nvme_fc_xmt_disconnect_assoc(ctrl);
+ spin_lock_irqsave(&ctrl->lock, flags);
+ ctrl->association_id = 0;
+ disls = ctrl->rcv_disconn;
+ ctrl->rcv_disconn = NULL;
+ spin_unlock_irqrestore(&ctrl->lock, flags);
+ if (disls)
+ /*
+ * if a Disconnect Request was waiting for a response, send
+ * now that all ABTS's have been issued (and are complete).
+ */
+ nvme_fc_xmt_ls_rsp(disls);
+
if (ctrl->ctrl.tagset) {
nvme_fc_delete_hw_io_queues(ctrl);
nvme_fc_free_io_queues(ctrl);
@@ -2853,7 +3258,7 @@
{
struct nvme_fc_ctrl *ctrl = to_fc_ctrl(nctrl);
- cancel_work_sync(&ctrl->err_work);
+ cancel_work_sync(&ctrl->ioerr_work);
cancel_delayed_work_sync(&ctrl->connect_work);
/*
* kill the association on the link side. this will block
@@ -2900,79 +3305,42 @@
dev_warn(ctrl->ctrl.device,
"NVME-FC{%d}: dev_loss_tmo (%d) expired "
"while waiting for remoteport connectivity.\n",
- ctrl->cnum, portptr->dev_loss_tmo);
+ ctrl->cnum, min_t(int, portptr->dev_loss_tmo,
+ (ctrl->ctrl.opts->max_reconnects *
+ ctrl->ctrl.opts->reconnect_delay)));
WARN_ON(nvme_delete_ctrl(&ctrl->ctrl));
}
}
static void
-__nvme_fc_terminate_io(struct nvme_fc_ctrl *ctrl)
-{
- /*
- * if state is connecting - the error occurred as part of a
- * reconnect attempt. The create_association error paths will
- * clean up any outstanding io.
- *
- * if it's a different state - ensure all pending io is
- * terminated. Given this can delay while waiting for the
- * aborted io to return, we recheck adapter state below
- * before changing state.
- */
- if (ctrl->ctrl.state != NVME_CTRL_CONNECTING) {
- nvme_stop_keep_alive(&ctrl->ctrl);
-
- /* will block will waiting for io to terminate */
- nvme_fc_delete_association(ctrl);
- }
-
- if (ctrl->ctrl.state != NVME_CTRL_CONNECTING &&
- !nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_CONNECTING))
- dev_err(ctrl->ctrl.device,
- "NVME-FC{%d}: error_recovery: Couldn't change state "
- "to CONNECTING\n", ctrl->cnum);
-}
-
-static void
nvme_fc_reset_ctrl_work(struct work_struct *work)
{
struct nvme_fc_ctrl *ctrl =
container_of(work, struct nvme_fc_ctrl, ctrl.reset_work);
- int ret;
-
- __nvme_fc_terminate_io(ctrl);
nvme_stop_ctrl(&ctrl->ctrl);
- if (ctrl->rport->remoteport.port_state == FC_OBJSTATE_ONLINE)
- ret = nvme_fc_create_association(ctrl);
- else
- ret = -ENOTCONN;
+ /* will block will waiting for io to terminate */
+ nvme_fc_delete_association(ctrl);
- if (ret)
- nvme_fc_reconnect_or_delete(ctrl, ret);
- else
- dev_info(ctrl->ctrl.device,
- "NVME-FC{%d}: controller reset complete\n",
- ctrl->cnum);
+ if (!nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_CONNECTING))
+ dev_err(ctrl->ctrl.device,
+ "NVME-FC{%d}: error_recovery: Couldn't change state "
+ "to CONNECTING\n", ctrl->cnum);
+
+ if (ctrl->rport->remoteport.port_state == FC_OBJSTATE_ONLINE) {
+ if (!queue_delayed_work(nvme_wq, &ctrl->connect_work, 0)) {
+ dev_err(ctrl->ctrl.device,
+ "NVME-FC{%d}: failed to schedule connect "
+ "after reset\n", ctrl->cnum);
+ } else {
+ flush_delayed_work(&ctrl->connect_work);
+ }
+ } else {
+ nvme_fc_reconnect_or_delete(ctrl, -ENOTCONN);
+ }
}
-static void
-nvme_fc_connect_err_work(struct work_struct *work)
-{
- struct nvme_fc_ctrl *ctrl =
- container_of(work, struct nvme_fc_ctrl, err_work);
-
- __nvme_fc_terminate_io(ctrl);
-
- atomic_set(&ctrl->err_work_active, 0);
-
- /*
- * Rescheduling the connection after recovering
- * from the io error is left to the reconnect work
- * item, which is what should have stalled waiting on
- * the io that had the error that scheduled this work.
- */
-}
static const struct nvme_ctrl_ops nvme_fc_ctrl_ops = {
.name = "fc",
@@ -3049,7 +3417,7 @@
{
struct nvme_fc_ctrl *ctrl;
unsigned long flags;
- int ret, idx;
+ int ret, idx, ctrl_loss_tmo;
if (!(rport->remoteport.port_role &
(FC_PORT_ROLE_NVME_DISCOVERY | FC_PORT_ROLE_NVME_TARGET))) {
@@ -3075,6 +3443,19 @@
goto out_free_ctrl;
}
+ /*
+ * if ctrl_loss_tmo is being enforced and the default reconnect delay
+ * is being used, change to a shorter reconnect delay for FC.
+ */
+ if (opts->max_reconnects != -1 &&
+ opts->reconnect_delay == NVMF_DEF_RECONNECT_DELAY &&
+ opts->reconnect_delay > NVME_FC_DEFAULT_RECONNECT_TMO) {
+ ctrl_loss_tmo = opts->max_reconnects * opts->reconnect_delay;
+ opts->reconnect_delay = NVME_FC_DEFAULT_RECONNECT_TMO;
+ opts->max_reconnects = DIV_ROUND_UP(ctrl_loss_tmo,
+ opts->reconnect_delay);
+ }
+
ctrl->ctrl.opts = opts;
ctrl->ctrl.nr_reconnects = 0;
if (lport->dev)
@@ -3087,8 +3468,6 @@
ctrl->dev = lport->dev;
ctrl->cnum = idx;
ctrl->ioq_live = false;
- ctrl->assoc_active = false;
- atomic_set(&ctrl->err_work_active, 0);
init_waitqueue_head(&ctrl->ioabort_wait);
get_device(ctrl->dev);
@@ -3096,7 +3475,7 @@
INIT_WORK(&ctrl->ctrl.reset_work, nvme_fc_reset_ctrl_work);
INIT_DELAYED_WORK(&ctrl->connect_work, nvme_fc_connect_ctrl_work);
- INIT_WORK(&ctrl->err_work, nvme_fc_connect_err_work);
+ INIT_WORK(&ctrl->ioerr_work, nvme_fc_ctrl_ioerr_work);
spin_lock_init(&ctrl->lock);
/* io queue count */
@@ -3188,15 +3567,14 @@
fail_ctrl:
nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_DELETING);
+ cancel_work_sync(&ctrl->ioerr_work);
cancel_work_sync(&ctrl->ctrl.reset_work);
- cancel_work_sync(&ctrl->err_work);
cancel_delayed_work_sync(&ctrl->connect_work);
ctrl->ctrl.opts = NULL;
/* initiate nvme ctrl ref counting teardown */
nvme_uninit_ctrl(&ctrl->ctrl);
- nvme_put_ctrl(&ctrl->ctrl);
/* Remove core ctrl ref. */
nvme_put_ctrl(&ctrl->ctrl);
diff --git a/drivers/nvme/host/fc.h b/drivers/nvme/host/fc.h
new file mode 100644
index 0000000..05ce566
--- /dev/null
+++ b/drivers/nvme/host/fc.h
@@ -0,0 +1,227 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (c) 2016, Avago Technologies
+ */
+
+#ifndef _NVME_FC_TRANSPORT_H
+#define _NVME_FC_TRANSPORT_H 1
+
+
+/*
+ * Common definitions between the nvme_fc (host) transport and
+ * nvmet_fc (target) transport implementation.
+ */
+
+/*
+ * ****************** FC-NVME LS HANDLING ******************
+ */
+
+union nvmefc_ls_requests {
+ struct fcnvme_ls_rqst_w0 w0;
+ struct fcnvme_ls_cr_assoc_rqst rq_cr_assoc;
+ struct fcnvme_ls_cr_conn_rqst rq_cr_conn;
+ struct fcnvme_ls_disconnect_assoc_rqst rq_dis_assoc;
+ struct fcnvme_ls_disconnect_conn_rqst rq_dis_conn;
+} __aligned(128); /* alignment for other things alloc'd with */
+
+union nvmefc_ls_responses {
+ struct fcnvme_ls_rjt rsp_rjt;
+ struct fcnvme_ls_cr_assoc_acc rsp_cr_assoc;
+ struct fcnvme_ls_cr_conn_acc rsp_cr_conn;
+ struct fcnvme_ls_disconnect_assoc_acc rsp_dis_assoc;
+ struct fcnvme_ls_disconnect_conn_acc rsp_dis_conn;
+} __aligned(128); /* alignment for other things alloc'd with */
+
+static inline void
+nvme_fc_format_rsp_hdr(void *buf, u8 ls_cmd, __be32 desc_len, u8 rqst_ls_cmd)
+{
+ struct fcnvme_ls_acc_hdr *acc = buf;
+
+ acc->w0.ls_cmd = ls_cmd;
+ acc->desc_list_len = desc_len;
+ acc->rqst.desc_tag = cpu_to_be32(FCNVME_LSDESC_RQST);
+ acc->rqst.desc_len =
+ fcnvme_lsdesc_len(sizeof(struct fcnvme_lsdesc_rqst));
+ acc->rqst.w0.ls_cmd = rqst_ls_cmd;
+}
+
+static inline int
+nvme_fc_format_rjt(void *buf, u16 buflen, u8 ls_cmd,
+ u8 reason, u8 explanation, u8 vendor)
+{
+ struct fcnvme_ls_rjt *rjt = buf;
+
+ nvme_fc_format_rsp_hdr(buf, FCNVME_LSDESC_RQST,
+ fcnvme_lsdesc_len(sizeof(struct fcnvme_ls_rjt)),
+ ls_cmd);
+ rjt->rjt.desc_tag = cpu_to_be32(FCNVME_LSDESC_RJT);
+ rjt->rjt.desc_len = fcnvme_lsdesc_len(sizeof(struct fcnvme_lsdesc_rjt));
+ rjt->rjt.reason_code = reason;
+ rjt->rjt.reason_explanation = explanation;
+ rjt->rjt.vendor = vendor;
+
+ return sizeof(struct fcnvme_ls_rjt);
+}
+
+/* Validation Error indexes into the string table below */
+enum {
+ VERR_NO_ERROR = 0,
+ VERR_CR_ASSOC_LEN = 1,
+ VERR_CR_ASSOC_RQST_LEN = 2,
+ VERR_CR_ASSOC_CMD = 3,
+ VERR_CR_ASSOC_CMD_LEN = 4,
+ VERR_ERSP_RATIO = 5,
+ VERR_ASSOC_ALLOC_FAIL = 6,
+ VERR_QUEUE_ALLOC_FAIL = 7,
+ VERR_CR_CONN_LEN = 8,
+ VERR_CR_CONN_RQST_LEN = 9,
+ VERR_ASSOC_ID = 10,
+ VERR_ASSOC_ID_LEN = 11,
+ VERR_NO_ASSOC = 12,
+ VERR_CONN_ID = 13,
+ VERR_CONN_ID_LEN = 14,
+ VERR_INVAL_CONN = 15,
+ VERR_CR_CONN_CMD = 16,
+ VERR_CR_CONN_CMD_LEN = 17,
+ VERR_DISCONN_LEN = 18,
+ VERR_DISCONN_RQST_LEN = 19,
+ VERR_DISCONN_CMD = 20,
+ VERR_DISCONN_CMD_LEN = 21,
+ VERR_DISCONN_SCOPE = 22,
+ VERR_RS_LEN = 23,
+ VERR_RS_RQST_LEN = 24,
+ VERR_RS_CMD = 25,
+ VERR_RS_CMD_LEN = 26,
+ VERR_RS_RCTL = 27,
+ VERR_RS_RO = 28,
+ VERR_LSACC = 29,
+ VERR_LSDESC_RQST = 30,
+ VERR_LSDESC_RQST_LEN = 31,
+ VERR_CR_ASSOC = 32,
+ VERR_CR_ASSOC_ACC_LEN = 33,
+ VERR_CR_CONN = 34,
+ VERR_CR_CONN_ACC_LEN = 35,
+ VERR_DISCONN = 36,
+ VERR_DISCONN_ACC_LEN = 37,
+};
+
+static char *validation_errors[] = {
+ "OK",
+ "Bad CR_ASSOC Length",
+ "Bad CR_ASSOC Rqst Length",
+ "Not CR_ASSOC Cmd",
+ "Bad CR_ASSOC Cmd Length",
+ "Bad Ersp Ratio",
+ "Association Allocation Failed",
+ "Queue Allocation Failed",
+ "Bad CR_CONN Length",
+ "Bad CR_CONN Rqst Length",
+ "Not Association ID",
+ "Bad Association ID Length",
+ "No Association",
+ "Not Connection ID",
+ "Bad Connection ID Length",
+ "Invalid Connection ID",
+ "Not CR_CONN Cmd",
+ "Bad CR_CONN Cmd Length",
+ "Bad DISCONN Length",
+ "Bad DISCONN Rqst Length",
+ "Not DISCONN Cmd",
+ "Bad DISCONN Cmd Length",
+ "Bad Disconnect Scope",
+ "Bad RS Length",
+ "Bad RS Rqst Length",
+ "Not RS Cmd",
+ "Bad RS Cmd Length",
+ "Bad RS R_CTL",
+ "Bad RS Relative Offset",
+ "Not LS_ACC",
+ "Not LSDESC_RQST",
+ "Bad LSDESC_RQST Length",
+ "Not CR_ASSOC Rqst",
+ "Bad CR_ASSOC ACC Length",
+ "Not CR_CONN Rqst",
+ "Bad CR_CONN ACC Length",
+ "Not Disconnect Rqst",
+ "Bad Disconnect ACC Length",
+};
+
+#define NVME_FC_LAST_LS_CMD_VALUE FCNVME_LS_DISCONNECT_CONN
+
+static char *nvmefc_ls_names[] = {
+ "Reserved (0)",
+ "RJT (1)",
+ "ACC (2)",
+ "Create Association",
+ "Create Connection",
+ "Disconnect Association",
+ "Disconnect Connection",
+};
+
+static inline void
+nvmefc_fmt_lsreq_discon_assoc(struct nvmefc_ls_req *lsreq,
+ struct fcnvme_ls_disconnect_assoc_rqst *discon_rqst,
+ struct fcnvme_ls_disconnect_assoc_acc *discon_acc,
+ u64 association_id)
+{
+ lsreq->rqstaddr = discon_rqst;
+ lsreq->rqstlen = sizeof(*discon_rqst);
+ lsreq->rspaddr = discon_acc;
+ lsreq->rsplen = sizeof(*discon_acc);
+ lsreq->timeout = NVME_FC_LS_TIMEOUT_SEC;
+
+ discon_rqst->w0.ls_cmd = FCNVME_LS_DISCONNECT_ASSOC;
+ discon_rqst->desc_list_len = cpu_to_be32(
+ sizeof(struct fcnvme_lsdesc_assoc_id) +
+ sizeof(struct fcnvme_lsdesc_disconn_cmd));
+
+ discon_rqst->associd.desc_tag = cpu_to_be32(FCNVME_LSDESC_ASSOC_ID);
+ discon_rqst->associd.desc_len =
+ fcnvme_lsdesc_len(
+ sizeof(struct fcnvme_lsdesc_assoc_id));
+
+ discon_rqst->associd.association_id = cpu_to_be64(association_id);
+
+ discon_rqst->discon_cmd.desc_tag = cpu_to_be32(
+ FCNVME_LSDESC_DISCONN_CMD);
+ discon_rqst->discon_cmd.desc_len =
+ fcnvme_lsdesc_len(
+ sizeof(struct fcnvme_lsdesc_disconn_cmd));
+}
+
+static inline int
+nvmefc_vldt_lsreq_discon_assoc(u32 rqstlen,
+ struct fcnvme_ls_disconnect_assoc_rqst *rqst)
+{
+ int ret = 0;
+
+ if (rqstlen < sizeof(struct fcnvme_ls_disconnect_assoc_rqst))
+ ret = VERR_DISCONN_LEN;
+ else if (rqst->desc_list_len !=
+ fcnvme_lsdesc_len(
+ sizeof(struct fcnvme_ls_disconnect_assoc_rqst)))
+ ret = VERR_DISCONN_RQST_LEN;
+ else if (rqst->associd.desc_tag != cpu_to_be32(FCNVME_LSDESC_ASSOC_ID))
+ ret = VERR_ASSOC_ID;
+ else if (rqst->associd.desc_len !=
+ fcnvme_lsdesc_len(
+ sizeof(struct fcnvme_lsdesc_assoc_id)))
+ ret = VERR_ASSOC_ID_LEN;
+ else if (rqst->discon_cmd.desc_tag !=
+ cpu_to_be32(FCNVME_LSDESC_DISCONN_CMD))
+ ret = VERR_DISCONN_CMD;
+ else if (rqst->discon_cmd.desc_len !=
+ fcnvme_lsdesc_len(
+ sizeof(struct fcnvme_lsdesc_disconn_cmd)))
+ ret = VERR_DISCONN_CMD_LEN;
+ /*
+ * As the standard changed on the LS, check if old format and scope
+ * something other than Association (e.g. 0).
+ */
+ else if (rqst->discon_cmd.rsvd8[0])
+ ret = VERR_DISCONN_SCOPE;
+
+ return ret;
+}
+
+#endif /* _NVME_FC_TRANSPORT_H */
diff --git a/drivers/nvme/host/hwmon.c b/drivers/nvme/host/hwmon.c
new file mode 100644
index 0000000..552dbc0
--- /dev/null
+++ b/drivers/nvme/host/hwmon.c
@@ -0,0 +1,255 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * NVM Express hardware monitoring support
+ * Copyright (c) 2019, Guenter Roeck
+ */
+
+#include <linux/hwmon.h>
+#include <linux/units.h>
+#include <asm/unaligned.h>
+
+#include "nvme.h"
+
+struct nvme_hwmon_data {
+ struct nvme_ctrl *ctrl;
+ struct nvme_smart_log log;
+ struct mutex read_lock;
+};
+
+static int nvme_get_temp_thresh(struct nvme_ctrl *ctrl, int sensor, bool under,
+ long *temp)
+{
+ unsigned int threshold = sensor << NVME_TEMP_THRESH_SELECT_SHIFT;
+ u32 status;
+ int ret;
+
+ if (under)
+ threshold |= NVME_TEMP_THRESH_TYPE_UNDER;
+
+ ret = nvme_get_features(ctrl, NVME_FEAT_TEMP_THRESH, threshold, NULL, 0,
+ &status);
+ if (ret > 0)
+ return -EIO;
+ if (ret < 0)
+ return ret;
+ *temp = kelvin_to_millicelsius(status & NVME_TEMP_THRESH_MASK);
+
+ return 0;
+}
+
+static int nvme_set_temp_thresh(struct nvme_ctrl *ctrl, int sensor, bool under,
+ long temp)
+{
+ unsigned int threshold = sensor << NVME_TEMP_THRESH_SELECT_SHIFT;
+ int ret;
+
+ temp = millicelsius_to_kelvin(temp);
+ threshold |= clamp_val(temp, 0, NVME_TEMP_THRESH_MASK);
+
+ if (under)
+ threshold |= NVME_TEMP_THRESH_TYPE_UNDER;
+
+ ret = nvme_set_features(ctrl, NVME_FEAT_TEMP_THRESH, threshold, NULL, 0,
+ NULL);
+ if (ret > 0)
+ return -EIO;
+
+ return ret;
+}
+
+static int nvme_hwmon_get_smart_log(struct nvme_hwmon_data *data)
+{
+ return nvme_get_log(data->ctrl, NVME_NSID_ALL, NVME_LOG_SMART, 0,
+ NVME_CSI_NVM, &data->log, sizeof(data->log), 0);
+}
+
+static int nvme_hwmon_read(struct device *dev, enum hwmon_sensor_types type,
+ u32 attr, int channel, long *val)
+{
+ struct nvme_hwmon_data *data = dev_get_drvdata(dev);
+ struct nvme_smart_log *log = &data->log;
+ int temp;
+ int err;
+
+ /*
+ * First handle attributes which don't require us to read
+ * the smart log.
+ */
+ switch (attr) {
+ case hwmon_temp_max:
+ return nvme_get_temp_thresh(data->ctrl, channel, false, val);
+ case hwmon_temp_min:
+ return nvme_get_temp_thresh(data->ctrl, channel, true, val);
+ case hwmon_temp_crit:
+ *val = kelvin_to_millicelsius(data->ctrl->cctemp);
+ return 0;
+ default:
+ break;
+ }
+
+ mutex_lock(&data->read_lock);
+ err = nvme_hwmon_get_smart_log(data);
+ if (err)
+ goto unlock;
+
+ switch (attr) {
+ case hwmon_temp_input:
+ if (!channel)
+ temp = get_unaligned_le16(log->temperature);
+ else
+ temp = le16_to_cpu(log->temp_sensor[channel - 1]);
+ *val = kelvin_to_millicelsius(temp);
+ break;
+ case hwmon_temp_alarm:
+ *val = !!(log->critical_warning & NVME_SMART_CRIT_TEMPERATURE);
+ break;
+ default:
+ err = -EOPNOTSUPP;
+ break;
+ }
+unlock:
+ mutex_unlock(&data->read_lock);
+ return err;
+}
+
+static int nvme_hwmon_write(struct device *dev, enum hwmon_sensor_types type,
+ u32 attr, int channel, long val)
+{
+ struct nvme_hwmon_data *data = dev_get_drvdata(dev);
+
+ switch (attr) {
+ case hwmon_temp_max:
+ return nvme_set_temp_thresh(data->ctrl, channel, false, val);
+ case hwmon_temp_min:
+ return nvme_set_temp_thresh(data->ctrl, channel, true, val);
+ default:
+ break;
+ }
+
+ return -EOPNOTSUPP;
+}
+
+static const char * const nvme_hwmon_sensor_names[] = {
+ "Composite",
+ "Sensor 1",
+ "Sensor 2",
+ "Sensor 3",
+ "Sensor 4",
+ "Sensor 5",
+ "Sensor 6",
+ "Sensor 7",
+ "Sensor 8",
+};
+
+static int nvme_hwmon_read_string(struct device *dev,
+ enum hwmon_sensor_types type, u32 attr,
+ int channel, const char **str)
+{
+ *str = nvme_hwmon_sensor_names[channel];
+ return 0;
+}
+
+static umode_t nvme_hwmon_is_visible(const void *_data,
+ enum hwmon_sensor_types type,
+ u32 attr, int channel)
+{
+ const struct nvme_hwmon_data *data = _data;
+
+ switch (attr) {
+ case hwmon_temp_crit:
+ if (!channel && data->ctrl->cctemp)
+ return 0444;
+ break;
+ case hwmon_temp_max:
+ case hwmon_temp_min:
+ if ((!channel && data->ctrl->wctemp) ||
+ (channel && data->log.temp_sensor[channel - 1])) {
+ if (data->ctrl->quirks &
+ NVME_QUIRK_NO_TEMP_THRESH_CHANGE)
+ return 0444;
+ return 0644;
+ }
+ break;
+ case hwmon_temp_alarm:
+ if (!channel)
+ return 0444;
+ break;
+ case hwmon_temp_input:
+ case hwmon_temp_label:
+ if (!channel || data->log.temp_sensor[channel - 1])
+ return 0444;
+ break;
+ default:
+ break;
+ }
+ return 0;
+}
+
+static const struct hwmon_channel_info *nvme_hwmon_info[] = {
+ HWMON_CHANNEL_INFO(chip, HWMON_C_REGISTER_TZ),
+ HWMON_CHANNEL_INFO(temp,
+ HWMON_T_INPUT | HWMON_T_MAX | HWMON_T_MIN |
+ HWMON_T_CRIT | HWMON_T_LABEL | HWMON_T_ALARM,
+ HWMON_T_INPUT | HWMON_T_MAX | HWMON_T_MIN |
+ HWMON_T_LABEL,
+ HWMON_T_INPUT | HWMON_T_MAX | HWMON_T_MIN |
+ HWMON_T_LABEL,
+ HWMON_T_INPUT | HWMON_T_MAX | HWMON_T_MIN |
+ HWMON_T_LABEL,
+ HWMON_T_INPUT | HWMON_T_MAX | HWMON_T_MIN |
+ HWMON_T_LABEL,
+ HWMON_T_INPUT | HWMON_T_MAX | HWMON_T_MIN |
+ HWMON_T_LABEL,
+ HWMON_T_INPUT | HWMON_T_MAX | HWMON_T_MIN |
+ HWMON_T_LABEL,
+ HWMON_T_INPUT | HWMON_T_MAX | HWMON_T_MIN |
+ HWMON_T_LABEL,
+ HWMON_T_INPUT | HWMON_T_MAX | HWMON_T_MIN |
+ HWMON_T_LABEL),
+ NULL
+};
+
+static const struct hwmon_ops nvme_hwmon_ops = {
+ .is_visible = nvme_hwmon_is_visible,
+ .read = nvme_hwmon_read,
+ .read_string = nvme_hwmon_read_string,
+ .write = nvme_hwmon_write,
+};
+
+static const struct hwmon_chip_info nvme_hwmon_chip_info = {
+ .ops = &nvme_hwmon_ops,
+ .info = nvme_hwmon_info,
+};
+
+int nvme_hwmon_init(struct nvme_ctrl *ctrl)
+{
+ struct device *dev = ctrl->dev;
+ struct nvme_hwmon_data *data;
+ struct device *hwmon;
+ int err;
+
+ data = devm_kzalloc(dev, sizeof(*data), GFP_KERNEL);
+ if (!data)
+ return 0;
+
+ data->ctrl = ctrl;
+ mutex_init(&data->read_lock);
+
+ err = nvme_hwmon_get_smart_log(data);
+ if (err) {
+ dev_warn(ctrl->device,
+ "Failed to read smart log (error %d)\n", err);
+ devm_kfree(dev, data);
+ return err;
+ }
+
+ hwmon = devm_hwmon_device_register_with_info(dev, "nvme", data,
+ &nvme_hwmon_chip_info,
+ NULL);
+ if (IS_ERR(hwmon)) {
+ dev_warn(dev, "Failed to instantiate hwmon device\n");
+ devm_kfree(dev, data);
+ }
+
+ return 0;
+}
diff --git a/drivers/nvme/host/lightnvm.c b/drivers/nvme/host/lightnvm.c
index ec46693..8e562d0 100644
--- a/drivers/nvme/host/lightnvm.c
+++ b/drivers/nvme/host/lightnvm.c
@@ -171,7 +171,7 @@
__le32 tdresv;
__le32 thresv;
__le32 rsvd2[8];
- __u8 blk[0];
+ __u8 blk[];
};
struct nvme_nvm_id20_addrf {
@@ -593,8 +593,8 @@
dev_meta_off = dev_meta;
ret = nvme_get_log(ctrl, ns->head->ns_id,
- NVME_NVM_LOG_REPORT_CHUNK, 0, dev_meta, len,
- offset);
+ NVME_NVM_LOG_REPORT_CHUNK, 0, NVME_CSI_NVM,
+ dev_meta, len, offset);
if (ret) {
dev_err(ctrl->device, "Get REPORT CHUNK log error\n");
break;
@@ -961,7 +961,10 @@
geo = &dev->geo;
geo->csecs = 1 << ns->lba_shift;
geo->sos = ns->ms;
- geo->ext = ns->ext;
+ if (ns->features & NVME_NS_EXT_LBAS)
+ geo->ext = true;
+ else
+ geo->ext = false;
geo->mdts = ns->ctrl->max_hw_sectors;
dev->q = q;
diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c
index 590b040..18a7564 100644
--- a/drivers/nvme/host/multipath.c
+++ b/drivers/nvme/host/multipath.c
@@ -65,51 +65,30 @@
}
}
-bool nvme_failover_req(struct request *req)
+void nvme_failover_req(struct request *req)
{
struct nvme_ns *ns = req->q->queuedata;
- u16 status = nvme_req(req)->status;
+ u16 status = nvme_req(req)->status & 0x7ff;
unsigned long flags;
- switch (status & 0x7ff) {
- case NVME_SC_ANA_TRANSITION:
- case NVME_SC_ANA_INACCESSIBLE:
- case NVME_SC_ANA_PERSISTENT_LOSS:
- /*
- * If we got back an ANA error we know the controller is alive,
- * but not ready to serve this namespaces. The spec suggests
- * we should update our general state here, but due to the fact
- * that the admin and I/O queues are not serialized that is
- * fundamentally racy. So instead just clear the current path,
- * mark the the path as pending and kick of a re-read of the ANA
- * log page ASAP.
- */
- nvme_mpath_clear_current_path(ns);
- if (ns->ctrl->ana_log_buf) {
- set_bit(NVME_NS_ANA_PENDING, &ns->flags);
- queue_work(nvme_wq, &ns->ctrl->ana_work);
- }
- break;
- case NVME_SC_HOST_PATH_ERROR:
- case NVME_SC_HOST_ABORTED_CMD:
- /*
- * Temporary transport disruption in talking to the controller.
- * Try to send on a new path.
- */
- nvme_mpath_clear_current_path(ns);
- break;
- default:
- /* This was a non-ANA error so follow the normal error path. */
- return false;
+ nvme_mpath_clear_current_path(ns);
+
+ /*
+ * If we got back an ANA error, we know the controller is alive but not
+ * ready to serve this namespace. Kick of a re-read of the ANA
+ * information page, and just try any other available path for now.
+ */
+ if (nvme_is_ana_error(status) && ns->ctrl->ana_log_buf) {
+ set_bit(NVME_NS_ANA_PENDING, &ns->flags);
+ queue_work(nvme_wq, &ns->ctrl->ana_work);
}
spin_lock_irqsave(&ns->head->requeue_lock, flags);
blk_steal_bios(&ns->head->requeue_list, req);
spin_unlock_irqrestore(&ns->head->requeue_lock, flags);
- blk_mq_end_request(req, 0);
+ blk_mq_end_request(req, 0);
kblockd_schedule_work(&ns->head->requeue_work);
- return true;
}
void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl)
@@ -156,20 +135,28 @@
{
struct nvme_ns *ns;
- mutex_lock(&ctrl->scan_lock);
down_read(&ctrl->namespaces_rwsem);
- list_for_each_entry(ns, &ctrl->namespaces, list)
- if (nvme_mpath_clear_current_path(ns))
- kblockd_schedule_work(&ns->head->requeue_work);
+ list_for_each_entry(ns, &ctrl->namespaces, list) {
+ nvme_mpath_clear_current_path(ns);
+ kblockd_schedule_work(&ns->head->requeue_work);
+ }
up_read(&ctrl->namespaces_rwsem);
- mutex_unlock(&ctrl->scan_lock);
}
static bool nvme_path_is_disabled(struct nvme_ns *ns)
{
- return ns->ctrl->state != NVME_CTRL_LIVE ||
- test_bit(NVME_NS_ANA_PENDING, &ns->flags) ||
- test_bit(NVME_NS_REMOVING, &ns->flags);
+ /*
+ * We don't treat NVME_CTRL_DELETING as a disabled path as I/O should
+ * still be able to complete assuming that the controller is connected.
+ * Otherwise it will fail immediately and return to the requeue list.
+ */
+ if (ns->ctrl->state != NVME_CTRL_LIVE &&
+ ns->ctrl->state != NVME_CTRL_DELETING)
+ return true;
+ if (test_bit(NVME_NS_ANA_PENDING, &ns->flags) ||
+ test_bit(NVME_NS_REMOVING, &ns->flags))
+ return true;
+ return false;
}
static struct nvme_ns *__nvme_find_path(struct nvme_ns_head *head, int node)
@@ -224,7 +211,7 @@
static struct nvme_ns *nvme_round_robin_path(struct nvme_ns_head *head,
int node, struct nvme_ns *old)
{
- struct nvme_ns *ns, *found, *fallback = NULL;
+ struct nvme_ns *ns, *found = NULL;
if (list_is_singular(&head->list)) {
if (nvme_path_is_disabled(old))
@@ -243,7 +230,7 @@
goto out;
}
if (ns->ana_state == NVME_ANA_NONOPTIMIZED)
- fallback = ns;
+ found = ns;
}
/*
@@ -254,12 +241,11 @@
*/
if (!nvme_path_is_disabled(old) &&
(old->ana_state == NVME_ANA_OPTIMIZED ||
- (!fallback && old->ana_state == NVME_ANA_NONOPTIMIZED)))
+ (!found && old->ana_state == NVME_ANA_NONOPTIMIZED)))
return old;
- if (!fallback)
+ if (!found)
return NULL;
- found = fallback;
out:
rcu_assign_pointer(head->current_path[node], found);
return found;
@@ -305,22 +291,20 @@
return false;
}
-static blk_qc_t nvme_ns_head_make_request(struct request_queue *q,
- struct bio *bio)
+blk_qc_t nvme_ns_head_submit_bio(struct bio *bio)
{
- struct nvme_ns_head *head = q->queuedata;
+ struct nvme_ns_head *head = bio->bi_disk->private_data;
struct device *dev = disk_to_dev(head->disk);
struct nvme_ns *ns;
blk_qc_t ret = BLK_QC_T_NONE;
int srcu_idx;
/*
- * The namespace might be going away and the bio might
- * be moved to a different queue via blk_steal_bios(),
- * so we need to use the bio_split pool from the original
- * queue to allocate the bvecs from.
+ * The namespace might be going away and the bio might be moved to a
+ * different queue via blk_steal_bios(), so we need to use the bio_split
+ * pool from the original queue to allocate the bvecs from.
*/
- blk_queue_split(q, &bio);
+ blk_queue_split(&bio);
srcu_idx = srcu_read_lock(&head->srcu);
ns = nvme_find_path(head);
@@ -330,7 +314,7 @@
trace_block_bio_remap(bio->bi_disk->queue, bio,
disk_devt(ns->head->disk),
bio->bi_iter.bi_sector);
- ret = generic_make_request(bio);
+ ret = submit_bio_noacct(bio);
} else if (nvme_available_path(head)) {
dev_warn_ratelimited(dev, "no usable path - requeuing I/O\n");
@@ -367,7 +351,7 @@
* path.
*/
bio->bi_disk = head->disk;
- generic_make_request(bio);
+ submit_bio_noacct(bio);
}
}
@@ -386,14 +370,12 @@
* We also do this for private namespaces as the namespace sharing data could
* change after a rescan.
*/
- if (!(ctrl->subsys->cmic & (1 << 1)) || !multipath)
+ if (!(ctrl->subsys->cmic & NVME_CTRL_CMIC_MULTI_CTRL) || !multipath)
return 0;
- q = blk_alloc_queue_node(GFP_KERNEL, ctrl->numa_node);
+ q = blk_alloc_queue(ctrl->numa_node);
if (!q)
goto out;
- q->queuedata = head;
- blk_queue_make_request(q, nvme_ns_head_make_request);
blk_queue_flag_set(QUEUE_FLAG_NONROT, q);
/* set to a default value for 512 until disk is validated */
blk_queue_logical_block_size(q, 512);
@@ -459,8 +441,14 @@
for (i = 0; i < le16_to_cpu(ctrl->ana_log_buf->ngrps); i++) {
struct nvme_ana_group_desc *desc = base + offset;
- u32 nr_nsids = le32_to_cpu(desc->nnsids);
- size_t nsid_buf_size = nr_nsids * sizeof(__le32);
+ u32 nr_nsids;
+ size_t nsid_buf_size;
+
+ if (WARN_ON_ONCE(offset > ctrl->ana_log_size - sizeof(*desc)))
+ return -EINVAL;
+
+ nr_nsids = le32_to_cpu(desc->nnsids);
+ nsid_buf_size = nr_nsids * sizeof(__le32);
if (WARN_ON_ONCE(desc->grpid == 0))
return -EINVAL;
@@ -480,8 +468,6 @@
return error;
offset += nsid_buf_size;
- if (WARN_ON_ONCE(offset > ctrl->ana_log_size - sizeof(*desc)))
- return -EINVAL;
}
return 0;
@@ -522,14 +508,17 @@
down_read(&ctrl->namespaces_rwsem);
list_for_each_entry(ns, &ctrl->namespaces, list) {
- unsigned nsid = le32_to_cpu(desc->nsids[n]);
-
+ unsigned nsid;
+again:
+ nsid = le32_to_cpu(desc->nsids[n]);
if (ns->head->ns_id < nsid)
continue;
if (ns->head->ns_id == nsid)
nvme_update_ns_ana_state(desc, ns);
if (++n == nr_nsids)
break;
+ if (ns->head->ns_id > nsid)
+ goto again;
}
up_read(&ctrl->namespaces_rwsem);
return 0;
@@ -541,7 +530,7 @@
int error;
mutex_lock(&ctrl->ana_lock);
- error = nvme_get_log(ctrl, NVME_NSID_ALL, NVME_LOG_ANA, 0,
+ error = nvme_get_log(ctrl, NVME_NSID_ALL, NVME_LOG_ANA, 0, NVME_CSI_NVM,
ctrl->ana_log_buf, ctrl->ana_log_size, 0);
if (error) {
dev_warn(ctrl->device, "Failed to get ANA log: %d\n", error);
@@ -577,6 +566,9 @@
{
struct nvme_ctrl *ctrl = container_of(work, struct nvme_ctrl, ana_work);
+ if (ctrl->state != NVME_CTRL_LIVE)
+ return;
+
nvme_read_ana_log(ctrl);
}
@@ -687,13 +679,13 @@
nvme_mpath_set_live(ns);
}
- if (bdi_cap_stable_pages_required(ns->queue->backing_dev_info)) {
- struct gendisk *disk = ns->head->disk;
-
- if (disk)
- disk->queue->backing_dev_info->capabilities |=
- BDI_CAP_STABLE_WRITES;
- }
+ if (blk_queue_stable_writes(ns->queue) && ns->head->disk)
+ blk_queue_flag_set(QUEUE_FLAG_STABLE_WRITES,
+ ns->head->disk->queue);
+#ifdef CONFIG_BLK_DEV_ZONED
+ if (blk_queue_is_zoned(ns->queue) && ns->head->disk)
+ ns->head->disk->queue->nr_zones = ns->queue->nr_zones;
+#endif
}
void nvme_mpath_remove_disk(struct nvme_ns_head *head)
@@ -732,7 +724,8 @@
int error = 0;
/* check if multipath is enabled and we have the capability */
- if (!multipath || !ctrl->subsys || !(ctrl->subsys->cmic & (1 << 3)))
+ if (!multipath || !ctrl->subsys ||
+ !(ctrl->subsys->cmic & NVME_CTRL_CMIC_ANA))
return 0;
ctrl->anacap = id->anacap;
diff --git a/drivers/nvme/host/nvme.h b/drivers/nvme/host/nvme.h
index 2df90d4..5dd1dd8 100644
--- a/drivers/nvme/host/nvme.h
+++ b/drivers/nvme/host/nvme.h
@@ -16,6 +16,7 @@
#include <linux/fault-inject.h>
#include <linux/rcupdate.h>
#include <linux/wait.h>
+#include <linux/t10-pi.h>
#include <trace/events/block.h>
@@ -28,6 +29,22 @@
#define NVME_DEFAULT_KATO 5
#define NVME_KATO_GRACE 10
+#ifdef CONFIG_ARCH_NO_SG_CHAIN
+#define NVME_INLINE_SG_CNT 0
+#define NVME_INLINE_METADATA_SG_CNT 0
+#else
+#define NVME_INLINE_SG_CNT 2
+#define NVME_INLINE_METADATA_SG_CNT 1
+#endif
+
+/*
+ * Default to a 4K page size, with the intention to update this
+ * path in the future to accommodate architectures with differing
+ * kernel and IO page sizes.
+ */
+#define NVME_CTRL_PAGE_SHIFT 12
+#define NVME_CTRL_PAGE_SIZE (1 << NVME_CTRL_PAGE_SHIFT)
+
extern struct workqueue_struct *nvme_wq;
extern struct workqueue_struct *nvme_reset_wq;
extern struct workqueue_struct *nvme_delete_wq;
@@ -117,11 +134,22 @@
NVME_QUIRK_SHARED_TAGS = (1 << 13),
/*
+ * Don't change the value of the temperature threshold feature
+ */
+ NVME_QUIRK_NO_TEMP_THRESH_CHANGE = (1 << 14),
+
+ /*
* The controller doesn't handle the Identify Namespace
* Identification Descriptor list subcommand despite claiming
* NVMe 1.3 compliance.
*/
NVME_QUIRK_NO_NS_DESC_LIST = (1 << 15),
+
+ /*
+ * The controller requires the command_id value be be limited, so skip
+ * encoding the generation sequence number.
+ */
+ NVME_QUIRK_SKIP_CID_GEN = (1 << 17),
};
/*
@@ -131,6 +159,7 @@
struct nvme_request {
struct nvme_command *cmd;
union nvme_result result;
+ u8 genctr;
u8 retries;
u8 flags;
u16 status;
@@ -154,7 +183,7 @@
static inline u16 nvme_req_qid(struct request *req)
{
- if (!req->rq_disk)
+ if (!req->q->queuedata)
return 0;
return blk_mq_unique_tag_to_hwq(blk_mq_unique_tag(req)) + 1;
}
@@ -166,12 +195,32 @@
*/
#define NVME_QUIRK_DELAY_AMOUNT 2300
+/*
+ * enum nvme_ctrl_state: Controller state
+ *
+ * @NVME_CTRL_NEW: New controller just allocated, initial state
+ * @NVME_CTRL_LIVE: Controller is connected and I/O capable
+ * @NVME_CTRL_RESETTING: Controller is resetting (or scheduled reset)
+ * @NVME_CTRL_CONNECTING: Controller is disconnected, now connecting the
+ * transport
+ * @NVME_CTRL_DELETING: Controller is deleting (or scheduled deletion)
+ * @NVME_CTRL_DELETING_NOIO: Controller is deleting and I/O is not
+ * disabled/failed immediately. This state comes
+ * after all async event processing took place and
+ * before ns removal and the controller deletion
+ * progress
+ * @NVME_CTRL_DEAD: Controller is non-present/unresponsive during
+ * shutdown or removal. In this case we forcibly
+ * kill all inflight I/O as they have no chance to
+ * complete
+ */
enum nvme_ctrl_state {
NVME_CTRL_NEW,
NVME_CTRL_LIVE,
NVME_CTRL_RESETTING,
NVME_CTRL_CONNECTING,
NVME_CTRL_DELETING,
+ NVME_CTRL_DELETING_NOIO,
NVME_CTRL_DEAD,
};
@@ -221,9 +270,12 @@
u32 queue_count;
u64 cap;
- u32 page_size;
u32 max_hw_sectors;
u32 max_segments;
+ u32 max_integrity_segments;
+#ifdef CONFIG_BLK_DEV_ZONED
+ u32 max_zone_append;
+#endif
u16 crdt[3];
u16 oncs;
u16 oacs;
@@ -238,6 +290,8 @@
u16 kas;
u8 npss;
u8 apsta;
+ u16 wctemp;
+ u16 cctemp;
u32 oaes;
u32 aen_result;
u32 ctratt;
@@ -247,13 +301,13 @@
unsigned long quirks;
struct nvme_id_power_state psd[32];
struct nvme_effects_log *effects;
+ struct xarray cels;
struct work_struct scan_work;
struct work_struct async_event_work;
struct delayed_work ka_work;
struct nvme_command ka_cmd;
struct work_struct fw_act_work;
unsigned long events;
- bool created;
#ifdef CONFIG_NVME_MULTIPATH
/* asymmetric namespace access: */
@@ -329,6 +383,7 @@
u8 eui64[8];
u8 nguid[16];
uuid_t uuid;
+ u8 csi;
};
/*
@@ -346,7 +401,9 @@
struct nvme_ns_ids ids;
struct list_head entry;
struct kref ref;
+ bool shared;
int instance;
+ struct nvme_effects_log *effects;
#ifdef CONFIG_NVME_MULTIPATH
struct gendisk *disk;
struct bio_list requeue_list;
@@ -359,6 +416,11 @@
#endif
};
+enum nvme_ns_features {
+ NVME_NS_EXT_LBAS = 1 << 0, /* support extended LBA format */
+ NVME_NS_METADATA_SUPPORTED = 1 << 1, /* support getting generated md */
+};
+
struct nvme_ns {
struct list_head list;
@@ -378,8 +440,11 @@
u16 ms;
u16 sgs;
u32 sws;
- bool ext;
u8 pi_type;
+#ifdef CONFIG_BLK_DEV_ZONED
+ u64 zsze;
+#endif
+ unsigned long features;
unsigned long flags;
#define NVME_NS_REMOVING 0
#define NVME_NS_DEAD 1
@@ -389,6 +454,12 @@
};
+/* NVMe ns supports metadata actions by the controller (generate/strip) */
+static inline bool nvme_ns_has_pi(struct nvme_ns *ns)
+{
+ return ns->pi_type && ns->ms == sizeof(struct t10_pi_tuple);
+}
+
struct nvme_ctrl_ops {
const char *name;
struct module *module;
@@ -405,6 +476,49 @@
int (*get_address)(struct nvme_ctrl *ctrl, char *buf, int size);
};
+/*
+ * nvme command_id is constructed as such:
+ * | xxxx | xxxxxxxxxxxx |
+ * gen request tag
+ */
+#define nvme_genctr_mask(gen) (gen & 0xf)
+#define nvme_cid_install_genctr(gen) (nvme_genctr_mask(gen) << 12)
+#define nvme_genctr_from_cid(cid) ((cid & 0xf000) >> 12)
+#define nvme_tag_from_cid(cid) (cid & 0xfff)
+
+static inline u16 nvme_cid(struct request *rq)
+{
+ return nvme_cid_install_genctr(nvme_req(rq)->genctr) | rq->tag;
+}
+
+static inline struct request *nvme_find_rq(struct blk_mq_tags *tags,
+ u16 command_id)
+{
+ u8 genctr = nvme_genctr_from_cid(command_id);
+ u16 tag = nvme_tag_from_cid(command_id);
+ struct request *rq;
+
+ rq = blk_mq_tag_to_rq(tags, tag);
+ if (unlikely(!rq)) {
+ pr_err("could not locate request for tag %#x\n",
+ tag);
+ return NULL;
+ }
+ if (unlikely(nvme_genctr_mask(nvme_req(rq)->genctr) != genctr)) {
+ dev_err(nvme_req(rq)->ctrl->device,
+ "request %#x genctr mismatch (got %#x expected %#x)\n",
+ tag, genctr, nvme_genctr_mask(nvme_req(rq)->genctr));
+ return NULL;
+ }
+ return rq;
+}
+
+static inline struct request *nvme_cid_to_rq(struct blk_mq_tags *tags,
+ u16 command_id)
+{
+ return blk_mq_tag_to_rq(tags, nvme_tag_from_cid(command_id));
+}
+
#ifdef CONFIG_FAULT_INJECTION_DEBUG_FS
void nvme_fault_inject_init(struct nvme_fault_inject *fault_inj,
const char *dev_name);
@@ -444,7 +558,39 @@
return lba << (ns->lba_shift - SECTOR_SHIFT);
}
-static inline void nvme_end_request(struct request *req, __le16 status,
+/*
+ * Convert byte length to nvme's 0-based num dwords
+ */
+static inline u32 nvme_bytes_to_numd(size_t len)
+{
+ return (len >> 2) - 1;
+}
+
+static inline bool nvme_is_ana_error(u16 status)
+{
+ switch (status & 0x7ff) {
+ case NVME_SC_ANA_TRANSITION:
+ case NVME_SC_ANA_INACCESSIBLE:
+ case NVME_SC_ANA_PERSISTENT_LOSS:
+ return true;
+ default:
+ return false;
+ }
+}
+
+static inline bool nvme_is_path_error(u16 status)
+{
+ /* check for a status code type of 'path related status' */
+ return (status & 0x700) == 0x300;
+}
+
+/*
+ * Fill in the status and result information from the CQE, and then figure out
+ * if blk-mq will need to use IPI magic to complete the request, and if yes do
+ * so. If not let the caller complete the request without an indirect function
+ * call.
+ */
+static inline bool nvme_try_complete_req(struct request *req, __le16 status,
union nvme_result result)
{
struct nvme_request *rq = nvme_req(req);
@@ -453,7 +599,9 @@
rq->result = result;
/* inject error when permitted by fault injection framework */
nvme_should_fail(req);
- blk_mq_complete_request(req);
+ if (unlikely(blk_should_fake_timeout(req->q)))
+ return true;
+ return blk_mq_complete_request_remote(req);
}
static inline void nvme_get_ctrl(struct nvme_ctrl *ctrl)
@@ -466,6 +614,12 @@
put_device(ctrl->device);
}
+static inline bool nvme_is_aen_req(u16 qid, __u16 command_id)
+{
+ return !qid &&
+ nvme_tag_from_cid(command_id) >= NVME_AQ_BLK_MQ_DEPTH;
+}
+
void nvme_complete_rq(struct request *req);
bool nvme_cancel_request(struct request *req, void *data, bool reserved);
void nvme_cancel_tagset(struct nvme_ctrl *ctrl);
@@ -481,7 +635,6 @@
void nvme_uninit_ctrl(struct nvme_ctrl *ctrl);
void nvme_start_ctrl(struct nvme_ctrl *ctrl);
void nvme_stop_ctrl(struct nvme_ctrl *ctrl);
-void nvme_put_ctrl(struct nvme_ctrl *ctrl);
int nvme_init_identify(struct nvme_ctrl *ctrl);
void nvme_remove_namespaces(struct nvme_ctrl *ctrl);
@@ -527,8 +680,11 @@
int nvme_try_sched_reset(struct nvme_ctrl *ctrl);
int nvme_delete_ctrl(struct nvme_ctrl *ctrl);
-int nvme_get_log(struct nvme_ctrl *ctrl, u32 nsid, u8 log_page, u8 lsp,
+int nvme_get_log(struct nvme_ctrl *ctrl, u32 nsid, u8 log_page, u8 lsp, u8 csi,
void *log, size_t size, u64 offset);
+struct nvme_ns *nvme_get_ns_from_disk(struct gendisk *disk,
+ struct nvme_ns_head **head, int *srcu_idx);
+void nvme_put_ns_from_disk(struct nvme_ns_head *head, int idx);
extern const struct attribute_group *nvme_ns_id_attr_groups[];
extern const struct block_device_operations nvme_ns_head_ops;
@@ -544,7 +700,7 @@
void nvme_mpath_start_freeze(struct nvme_subsystem *subsys);
void nvme_set_disk_name(char *disk_name, struct nvme_ns *ns,
struct nvme_ctrl *ctrl, int *flags);
-bool nvme_failover_req(struct request *req);
+void nvme_failover_req(struct request *req);
void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl);
int nvme_mpath_alloc_disk(struct nvme_ctrl *ctrl,struct nvme_ns_head *head);
void nvme_mpath_add_disk(struct nvme_ns *ns, struct nvme_id_ns *id);
@@ -556,6 +712,7 @@
bool nvme_mpath_clear_current_path(struct nvme_ns *ns);
void nvme_mpath_clear_ctrl_paths(struct nvme_ctrl *ctrl);
struct nvme_ns *nvme_find_path(struct nvme_ns_head *head);
+blk_qc_t nvme_ns_head_submit_bio(struct bio *bio);
static inline void nvme_mpath_check_last_path(struct nvme_ns *ns)
{
@@ -571,18 +728,7 @@
struct nvme_ns *ns = req->q->queuedata;
if (req->cmd_flags & REQ_NVME_MPATH)
- trace_block_bio_complete(ns->head->disk->queue,
- req->bio, status);
-}
-
-static inline void nvme_mpath_update_disk_size(struct gendisk *disk)
-{
- struct block_device *bdev = bdget_disk(disk, 0);
-
- if (bdev) {
- bd_set_size(bdev, get_capacity(disk) << SECTOR_SHIFT);
- bdput(bdev);
- }
+ trace_block_bio_complete(ns->head->disk->queue, req->bio);
}
extern struct device_attribute dev_attr_ana_grpid;
@@ -604,9 +750,8 @@
sprintf(disk_name, "nvme%dn%d", ctrl->instance, ns->head->instance);
}
-static inline bool nvme_failover_req(struct request *req)
+static inline void nvme_failover_req(struct request *req)
{
- return false;
}
static inline void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl)
{
@@ -663,11 +808,35 @@
static inline void nvme_mpath_start_freeze(struct nvme_subsystem *subsys)
{
}
-static inline void nvme_mpath_update_disk_size(struct gendisk *disk)
-{
-}
#endif /* CONFIG_NVME_MULTIPATH */
+int nvme_revalidate_zones(struct nvme_ns *ns);
+#ifdef CONFIG_BLK_DEV_ZONED
+int nvme_update_zone_info(struct nvme_ns *ns, unsigned lbaf);
+int nvme_report_zones(struct gendisk *disk, sector_t sector,
+ unsigned int nr_zones, report_zones_cb cb, void *data);
+
+blk_status_t nvme_setup_zone_mgmt_send(struct nvme_ns *ns, struct request *req,
+ struct nvme_command *cmnd,
+ enum nvme_zone_mgmt_action action);
+#else
+#define nvme_report_zones NULL
+
+static inline blk_status_t nvme_setup_zone_mgmt_send(struct nvme_ns *ns,
+ struct request *req, struct nvme_command *cmnd,
+ enum nvme_zone_mgmt_action action)
+{
+ return BLK_STS_NOTSUPP;
+}
+
+static inline int nvme_update_zone_info(struct nvme_ns *ns, unsigned lbaf)
+{
+ dev_warn(ns->ctrl->device,
+ "Please enable CONFIG_BLK_DEV_ZONED to support ZNS devices\n");
+ return -EPROTONOSUPPORT;
+}
+#endif
+
#ifdef CONFIG_NVM
int nvme_nvm_register(struct nvme_ns *ns, char *disk_name, int node);
void nvme_nvm_unregister(struct nvme_ns *ns);
@@ -693,4 +862,20 @@
return dev_to_disk(dev)->private_data;
}
+#ifdef CONFIG_NVME_HWMON
+int nvme_hwmon_init(struct nvme_ctrl *ctrl);
+#else
+static inline int nvme_hwmon_init(struct nvme_ctrl *ctrl)
+{
+ return 0;
+}
+#endif
+
+u32 nvme_command_effects(struct nvme_ctrl *ctrl, struct nvme_ns *ns,
+ u8 opcode);
+void nvme_execute_passthru_rq(struct request *rq);
+struct nvme_ctrl *nvme_ctrl_from_file(struct file *file);
+struct nvme_ns *nvme_find_get_ns(struct nvme_ctrl *ctrl, unsigned nsid);
+void nvme_put_ns(struct nvme_ns *ns);
+
#endif /* _NVME_H */
diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index af516c3..97afeb8 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -4,6 +4,7 @@
* Copyright (c) 2011-2014, Intel Corporation.
*/
+#include <linux/acpi.h>
#include <linux/aer.h>
#include <linux/async.h>
#include <linux/blkdev.h>
@@ -22,6 +23,7 @@
#include <linux/t10-pi.h>
#include <linux/types.h>
#include <linux/io-64-nonatomic-lo-hi.h>
+#include <linux/io-64-nonatomic-hi-lo.h>
#include <linux/sed-opal.h>
#include <linux/pci-p2pdma.h>
@@ -61,23 +63,43 @@
static int io_queue_depth_set(const char *val, const struct kernel_param *kp);
static const struct kernel_param_ops io_queue_depth_ops = {
.set = io_queue_depth_set,
- .get = param_get_int,
+ .get = param_get_uint,
};
-static int io_queue_depth = 1024;
+static unsigned int io_queue_depth = 1024;
module_param_cb(io_queue_depth, &io_queue_depth_ops, &io_queue_depth, 0644);
MODULE_PARM_DESC(io_queue_depth, "set io queue depth, should >= 2");
+static int io_queue_count_set(const char *val, const struct kernel_param *kp)
+{
+ unsigned int n;
+ int ret;
+
+ ret = kstrtouint(val, 10, &n);
+ if (ret != 0 || n > num_possible_cpus())
+ return -EINVAL;
+ return param_set_uint(val, kp);
+}
+
+static const struct kernel_param_ops io_queue_count_ops = {
+ .set = io_queue_count_set,
+ .get = param_get_uint,
+};
+
static unsigned int write_queues;
-module_param(write_queues, uint, 0644);
+module_param_cb(write_queues, &io_queue_count_ops, &write_queues, 0644);
MODULE_PARM_DESC(write_queues,
"Number of queues to use for writes. If not set, reads and writes "
"will share a queue set.");
static unsigned int poll_queues;
-module_param(poll_queues, uint, 0644);
+module_param_cb(poll_queues, &io_queue_count_ops, &poll_queues, 0644);
MODULE_PARM_DESC(poll_queues, "Number of queues to use for polled IO.");
+static bool noacpi;
+module_param(noacpi, bool, 0444);
+MODULE_PARM_DESC(noacpi, "disable acpi bios quirks");
+
struct nvme_dev;
struct nvme_queue;
@@ -99,7 +121,7 @@
unsigned max_qid;
unsigned io_queues[HCTX_MAX_TYPES];
unsigned int num_vecs;
- int q_depth;
+ u32 q_depth;
int io_sqes;
u32 db_stride;
void __iomem *bar;
@@ -135,13 +157,14 @@
static int io_queue_depth_set(const char *val, const struct kernel_param *kp)
{
- int n = 0, ret;
+ int ret;
+ u32 n;
- ret = kstrtoint(val, 10, &n);
+ ret = kstrtou32(val, 10, &n);
if (ret != 0 || n < 2)
return -EINVAL;
- return param_set_int(val, kp);
+ return param_set_uint(val, kp);
}
static inline unsigned int sq_idx(unsigned int qid, u32 stride)
@@ -169,16 +192,15 @@
void *sq_cmds;
/* only used for poll queues: */
spinlock_t cq_poll_lock ____cacheline_aligned_in_smp;
- volatile struct nvme_completion *cqes;
+ struct nvme_completion *cqes;
dma_addr_t sq_dma_addr;
dma_addr_t cq_dma_addr;
u32 __iomem *q_db;
- u16 q_depth;
+ u32 q_depth;
u16 cq_vector;
u16 sq_tail;
u16 last_sq_tail;
u16 cq_head;
- u16 last_cq_head;
u16 qid;
u8 cq_phase;
u8 sqes;
@@ -346,10 +368,10 @@
* as it only leads to a small amount of wasted memory for the lifetime of
* the I/O.
*/
-static int nvme_npages(unsigned size, struct nvme_dev *dev)
+static int nvme_pci_npages_prp(void)
{
- unsigned nprps = DIV_ROUND_UP(size + dev->ctrl.page_size,
- dev->ctrl.page_size);
+ unsigned nprps = DIV_ROUND_UP(NVME_MAX_KB_SZ + NVME_CTRL_PAGE_SIZE,
+ NVME_CTRL_PAGE_SIZE);
return DIV_ROUND_UP(8 * nprps, PAGE_SIZE - 8);
}
@@ -357,22 +379,18 @@
* Calculates the number of pages needed for the SGL segments. For example a 4k
* page can accommodate 256 SGL descriptors.
*/
-static int nvme_pci_npages_sgl(unsigned int num_seg)
+static int nvme_pci_npages_sgl(void)
{
- return DIV_ROUND_UP(num_seg * sizeof(struct nvme_sgl_desc), PAGE_SIZE);
+ return DIV_ROUND_UP(NVME_MAX_SEGS * sizeof(struct nvme_sgl_desc),
+ PAGE_SIZE);
}
-static unsigned int nvme_pci_iod_alloc_size(struct nvme_dev *dev,
- unsigned int size, unsigned int nseg, bool use_sgl)
+static size_t nvme_pci_iod_alloc_size(void)
{
- size_t alloc_size;
+ size_t npages = max(nvme_pci_npages_prp(), nvme_pci_npages_sgl());
- if (use_sgl)
- alloc_size = sizeof(__le64 *) * nvme_pci_npages_sgl(nseg);
- else
- alloc_size = sizeof(__le64 *) * nvme_npages(size, dev);
-
- return alloc_size + sizeof(struct scatterlist) * nseg;
+ return sizeof(__le64 *) * npages +
+ sizeof(struct scatterlist) * NVME_MAX_SEGS;
}
static int nvme_admin_init_hctx(struct blk_mq_hw_ctx *hctx, void *data,
@@ -514,9 +532,6 @@
int nseg = blk_rq_nr_phys_segments(req);
unsigned int avg_seg_size;
- if (nseg == 0)
- return false;
-
avg_seg_size = DIV_ROUND_UP(blk_rq_payload_bytes(req), nseg);
if (!(dev->ctrl.sgls & ((1 << 0) | (1 << 1))))
@@ -530,7 +545,7 @@
static void nvme_free_prps(struct nvme_dev *dev, struct request *req)
{
- const int last_prp = dev->ctrl.page_size / sizeof(__le64) - 1;
+ const int last_prp = NVME_CTRL_PAGE_SIZE / sizeof(__le64) - 1;
struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
dma_addr_t dma_addr = iod->first_dma;
int i;
@@ -619,34 +634,33 @@
struct scatterlist *sg = iod->sg;
int dma_len = sg_dma_len(sg);
u64 dma_addr = sg_dma_address(sg);
- u32 page_size = dev->ctrl.page_size;
- int offset = dma_addr & (page_size - 1);
+ int offset = dma_addr & (NVME_CTRL_PAGE_SIZE - 1);
__le64 *prp_list;
void **list = nvme_pci_iod_list(req);
dma_addr_t prp_dma;
int nprps, i;
- length -= (page_size - offset);
+ length -= (NVME_CTRL_PAGE_SIZE - offset);
if (length <= 0) {
iod->first_dma = 0;
goto done;
}
- dma_len -= (page_size - offset);
+ dma_len -= (NVME_CTRL_PAGE_SIZE - offset);
if (dma_len) {
- dma_addr += (page_size - offset);
+ dma_addr += (NVME_CTRL_PAGE_SIZE - offset);
} else {
sg = sg_next(sg);
dma_addr = sg_dma_address(sg);
dma_len = sg_dma_len(sg);
}
- if (length <= page_size) {
+ if (length <= NVME_CTRL_PAGE_SIZE) {
iod->first_dma = dma_addr;
goto done;
}
- nprps = DIV_ROUND_UP(length, page_size);
+ nprps = DIV_ROUND_UP(length, NVME_CTRL_PAGE_SIZE);
if (nprps <= (256 / 8)) {
pool = dev->prp_small_pool;
iod->npages = 0;
@@ -665,7 +679,7 @@
iod->first_dma = prp_dma;
i = 0;
for (;;) {
- if (i == page_size >> 3) {
+ if (i == NVME_CTRL_PAGE_SIZE >> 3) {
__le64 *old_prp_list = prp_list;
prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma);
if (!prp_list)
@@ -676,9 +690,9 @@
i = 1;
}
prp_list[i++] = cpu_to_le64(dma_addr);
- dma_len -= page_size;
- dma_addr += page_size;
- length -= page_size;
+ dma_len -= NVME_CTRL_PAGE_SIZE;
+ dma_addr += NVME_CTRL_PAGE_SIZE;
+ length -= NVME_CTRL_PAGE_SIZE;
if (length <= 0)
break;
if (dma_len > 0)
@@ -791,8 +805,8 @@
struct bio_vec *bv)
{
struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
- unsigned int offset = bv->bv_offset & (dev->ctrl.page_size - 1);
- unsigned int first_prp_len = dev->ctrl.page_size - offset;
+ unsigned int offset = bv->bv_offset & (NVME_CTRL_PAGE_SIZE - 1);
+ unsigned int first_prp_len = NVME_CTRL_PAGE_SIZE - offset;
iod->first_dma = dma_map_bvec(dev->dev, bv, rq_dma_dir(req), 0);
if (dma_mapping_error(dev->dev, iod->first_dma))
@@ -802,7 +816,7 @@
cmnd->dptr.prp1 = cpu_to_le64(iod->first_dma);
if (bv->bv_len > first_prp_len)
cmnd->dptr.prp2 = cpu_to_le64(iod->first_dma + first_prp_len);
- return 0;
+ return BLK_STS_OK;
}
static blk_status_t nvme_setup_sgl_simple(struct nvme_dev *dev,
@@ -820,7 +834,7 @@
cmnd->dptr.sgl.addr = cpu_to_le64(iod->first_dma);
cmnd->dptr.sgl.length = cpu_to_le32(iod->dma_len);
cmnd->dptr.sgl.type = NVME_SGL_FMT_DATA_DESC << 4;
- return 0;
+ return BLK_STS_OK;
}
static blk_status_t nvme_map_data(struct nvme_dev *dev, struct request *req,
@@ -834,7 +848,7 @@
struct bio_vec bv = req_bvec(req);
if (!is_pci_p2pdma_page(bv.bv_page)) {
- if (bv.bv_offset + bv.bv_len <= dev->ctrl.page_size * 2)
+ if (bv.bv_offset + bv.bv_len <= NVME_CTRL_PAGE_SIZE * 2)
return nvme_setup_prp_simple(dev, req,
&cmnd->rw, &bv);
@@ -889,7 +903,7 @@
if (dma_mapping_error(dev->dev, iod->meta_dma))
return BLK_STS_IOERR;
cmnd->rw.metadata = cpu_to_le64(iod->meta_dma);
- return 0;
+ return BLK_STS_OK;
}
/*
@@ -948,7 +962,6 @@
struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
struct nvme_dev *dev = iod->nvmeq->dev;
- nvme_cleanup_cmd(req);
if (blk_integrity_rq(req))
dma_unmap_page(dev->dev, iod->meta_dma,
rq_integrity_vec(req)->bv_len, rq_data_dir(req));
@@ -960,8 +973,9 @@
/* We read the CQE phase first to check if the rest of the entry is valid */
static inline bool nvme_cqe_pending(struct nvme_queue *nvmeq)
{
- return (le16_to_cpu(nvmeq->cqes[nvmeq->cq_head].status) & 1) ==
- nvmeq->cq_phase;
+ struct nvme_completion *hcqe = &nvmeq->cqes[nvmeq->cq_head];
+
+ return (le16_to_cpu(READ_ONCE(hcqe->status)) & 1) == nvmeq->cq_phase;
}
static inline void nvme_ring_cq_doorbell(struct nvme_queue *nvmeq)
@@ -982,7 +996,8 @@
static inline void nvme_handle_cqe(struct nvme_queue *nvmeq, u16 idx)
{
- volatile struct nvme_completion *cqe = &nvmeq->cqes[idx];
+ struct nvme_completion *cqe = &nvmeq->cqes[idx];
+ __u16 command_id = READ_ONCE(cqe->command_id);
struct request *req;
/*
@@ -991,58 +1006,53 @@
* aborts. We don't even bother to allocate a struct request
* for them but rather special case them here.
*/
- if (unlikely(nvmeq->qid == 0 &&
- cqe->command_id >= NVME_AQ_BLK_MQ_DEPTH)) {
+ if (unlikely(nvme_is_aen_req(nvmeq->qid, command_id))) {
nvme_complete_async_event(&nvmeq->dev->ctrl,
cqe->status, &cqe->result);
return;
}
- req = blk_mq_tag_to_rq(nvme_queue_tagset(nvmeq), cqe->command_id);
+ req = nvme_find_rq(nvme_queue_tagset(nvmeq), command_id);
if (unlikely(!req)) {
dev_warn(nvmeq->dev->ctrl.device,
"invalid id %d completed on queue %d\n",
- cqe->command_id, le16_to_cpu(cqe->sq_id));
+ command_id, le16_to_cpu(cqe->sq_id));
return;
}
trace_nvme_sq(req, cqe->sq_head, nvmeq->sq_tail);
- nvme_end_request(req, cqe->status, cqe->result);
-}
-
-static void nvme_complete_cqes(struct nvme_queue *nvmeq, u16 start, u16 end)
-{
- while (start != end) {
- nvme_handle_cqe(nvmeq, start);
- if (++start == nvmeq->q_depth)
- start = 0;
- }
+ if (!nvme_try_complete_req(req, cqe->status, cqe->result))
+ nvme_pci_complete_rq(req);
}
static inline void nvme_update_cq_head(struct nvme_queue *nvmeq)
{
- if (nvmeq->cq_head == nvmeq->q_depth - 1) {
+ u32 tmp = nvmeq->cq_head + 1;
+
+ if (tmp == nvmeq->q_depth) {
nvmeq->cq_head = 0;
- nvmeq->cq_phase = !nvmeq->cq_phase;
+ nvmeq->cq_phase ^= 1;
} else {
- nvmeq->cq_head++;
+ nvmeq->cq_head = tmp;
}
}
-static inline int nvme_process_cq(struct nvme_queue *nvmeq, u16 *start,
- u16 *end, unsigned int tag)
+static inline int nvme_process_cq(struct nvme_queue *nvmeq)
{
int found = 0;
- *start = nvmeq->cq_head;
while (nvme_cqe_pending(nvmeq)) {
- if (tag == -1U || nvmeq->cqes[nvmeq->cq_head].command_id == tag)
- found++;
+ found++;
+ /*
+ * load-load control dependency between phase and the rest of
+ * the cqe requires a full read memory barrier
+ */
+ dma_rmb();
+ nvme_handle_cqe(nvmeq, nvmeq->cq_head);
nvme_update_cq_head(nvmeq);
}
- *end = nvmeq->cq_head;
- if (*start != *end)
+ if (found)
nvme_ring_cq_doorbell(nvmeq);
return found;
}
@@ -1051,76 +1061,53 @@
{
struct nvme_queue *nvmeq = data;
irqreturn_t ret = IRQ_NONE;
- u16 start, end;
/*
* The rmb/wmb pair ensures we see all updates from a previous run of
* the irq handler, even if that was on another CPU.
*/
rmb();
- if (nvmeq->cq_head != nvmeq->last_cq_head)
+ if (nvme_process_cq(nvmeq))
ret = IRQ_HANDLED;
- nvme_process_cq(nvmeq, &start, &end, -1);
- nvmeq->last_cq_head = nvmeq->cq_head;
wmb();
- if (start != end) {
- nvme_complete_cqes(nvmeq, start, end);
- return IRQ_HANDLED;
- }
-
return ret;
}
static irqreturn_t nvme_irq_check(int irq, void *data)
{
struct nvme_queue *nvmeq = data;
+
if (nvme_cqe_pending(nvmeq))
return IRQ_WAKE_THREAD;
return IRQ_NONE;
}
/*
- * Poll for completions any queue, including those not dedicated to polling.
+ * Poll for completions for any interrupt driven queue
* Can be called from any context.
*/
-static int nvme_poll_irqdisable(struct nvme_queue *nvmeq, unsigned int tag)
+static void nvme_poll_irqdisable(struct nvme_queue *nvmeq)
{
struct pci_dev *pdev = to_pci_dev(nvmeq->dev->dev);
- u16 start, end;
- int found;
- /*
- * For a poll queue we need to protect against the polling thread
- * using the CQ lock. For normal interrupt driven threads we have
- * to disable the interrupt to avoid racing with it.
- */
- if (test_bit(NVMEQ_POLLED, &nvmeq->flags)) {
- spin_lock(&nvmeq->cq_poll_lock);
- found = nvme_process_cq(nvmeq, &start, &end, tag);
- spin_unlock(&nvmeq->cq_poll_lock);
- } else {
- disable_irq(pci_irq_vector(pdev, nvmeq->cq_vector));
- found = nvme_process_cq(nvmeq, &start, &end, tag);
- enable_irq(pci_irq_vector(pdev, nvmeq->cq_vector));
- }
+ WARN_ON_ONCE(test_bit(NVMEQ_POLLED, &nvmeq->flags));
- nvme_complete_cqes(nvmeq, start, end);
- return found;
+ disable_irq(pci_irq_vector(pdev, nvmeq->cq_vector));
+ nvme_process_cq(nvmeq);
+ enable_irq(pci_irq_vector(pdev, nvmeq->cq_vector));
}
static int nvme_poll(struct blk_mq_hw_ctx *hctx)
{
struct nvme_queue *nvmeq = hctx->driver_data;
- u16 start, end;
bool found;
if (!nvme_cqe_pending(nvmeq))
return 0;
spin_lock(&nvmeq->cq_poll_lock);
- found = nvme_process_cq(nvmeq, &start, &end, -1);
- nvme_complete_cqes(nvmeq, start, end);
+ found = nvme_process_cq(nvmeq);
spin_unlock(&nvmeq->cq_poll_lock);
return found;
@@ -1226,7 +1213,6 @@
static bool nvme_should_reset(struct nvme_dev *dev, u32 csts)
{
-
/* If true, indicates loss of adapter communication, possibly by a
* NVMe Subsystem reset.
*/
@@ -1297,7 +1283,12 @@
/*
* Did we miss an interrupt?
*/
- if (nvme_poll_irqdisable(nvmeq, req->tag)) {
+ if (test_bit(NVMEQ_POLLED, &nvmeq->flags))
+ nvme_poll(req->mq_hctx);
+ else
+ nvme_poll_irqdisable(nvmeq);
+
+ if (blk_mq_request_completed(req)) {
dev_warn(dev->ctrl.device,
"I/O %d QID %d timeout, completion polled\n",
req->tag, nvmeq->qid);
@@ -1313,7 +1304,7 @@
switch (dev->ctrl.state) {
case NVME_CTRL_CONNECTING:
nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_DELETING);
- /* fall through */
+ fallthrough;
case NVME_CTRL_DELETING:
dev_warn_ratelimited(dev->ctrl.device,
"I/O %d QID %d timeout, disable controller\n",
@@ -1328,9 +1319,9 @@
}
/*
- * Shutdown the controller immediately and schedule a reset if the
- * command was already aborted once before and still hasn't been
- * returned to the driver, or if this is the admin queue.
+ * Shutdown the controller immediately and schedule a reset if the
+ * command was already aborted once before and still hasn't been
+ * returned to the driver, or if this is the admin queue.
*/
if (!nvmeq->qid || iod->aborted) {
dev_warn(dev->ctrl.device,
@@ -1351,7 +1342,7 @@
memset(&cmd, 0, sizeof(cmd));
cmd.abort.opcode = nvme_admin_abort_cmd;
- cmd.abort.cid = req->tag;
+ cmd.abort.cid = nvme_cid(req);
cmd.abort.sqid = cpu_to_le16(nvmeq->qid);
dev_warn(nvmeq->dev->ctrl.device,
@@ -1440,23 +1431,23 @@
else
nvme_disable_ctrl(&dev->ctrl);
- nvme_poll_irqdisable(nvmeq, -1);
+ nvme_poll_irqdisable(nvmeq);
}
/*
* Called only on a device that has been disabled and after all other threads
- * that can check this device's completion queues have synced. This is the
- * last chance for the driver to see a natural completion before
- * nvme_cancel_request() terminates all incomplete requests.
+ * that can check this device's completion queues have synced, except
+ * nvme_poll(). This is the last chance for the driver to see a natural
+ * completion before nvme_cancel_request() terminates all incomplete requests.
*/
static void nvme_reap_pending_cqes(struct nvme_dev *dev)
{
- u16 start, end;
int i;
for (i = dev->ctrl.queue_count - 1; i > 0; i--) {
- nvme_process_cq(&dev->queues[i], &start, &end, -1);
- nvme_complete_cqes(&dev->queues[i], start, end);
+ spin_lock(&dev->queues[i].cq_poll_lock);
+ nvme_process_cq(&dev->queues[i]);
+ spin_unlock(&dev->queues[i].cq_poll_lock);
}
}
@@ -1465,11 +1456,12 @@
{
int q_depth = dev->q_depth;
unsigned q_size_aligned = roundup(q_depth * entry_size,
- dev->ctrl.page_size);
+ NVME_CTRL_PAGE_SIZE);
if (q_size_aligned * nr_io_queues > dev->cmb_size) {
u64 mem_per_q = div_u64(dev->cmb_size, nr_io_queues);
- mem_per_q = round_down(mem_per_q, dev->ctrl.page_size);
+
+ mem_per_q = round_down(mem_per_q, NVME_CTRL_PAGE_SIZE);
q_depth = div_u64(mem_per_q, entry_size);
/*
@@ -1598,7 +1590,7 @@
result = adapter_alloc_sq(dev, qid, nvmeq);
if (result < 0)
return result;
- else if (result)
+ if (result)
goto release_cq;
nvmeq->cq_vector = vector;
@@ -1662,7 +1654,7 @@
dev->admin_tagset.queue_depth = NVME_AQ_MQ_TAG_DEPTH;
dev->admin_tagset.timeout = ADMIN_TIMEOUT;
- dev->admin_tagset.numa_node = dev_to_node(dev->dev);
+ dev->admin_tagset.numa_node = dev->ctrl.numa_node;
dev->admin_tagset.cmd_size = sizeof(struct nvme_iod);
dev->admin_tagset.flags = BLK_MQ_F_NO_SCHED;
dev->admin_tagset.driver_data = dev;
@@ -1738,6 +1730,8 @@
if (result)
return result;
+ dev->ctrl.numa_node = dev_to_node(dev->dev);
+
nvmeq = &dev->queues[0];
aqa = nvmeq->q_depth - 1;
aqa |= aqa << 16;
@@ -1832,6 +1826,9 @@
if (dev->cmb_size)
return;
+ if (NVME_CAP_CMBS(dev->ctrl.cap))
+ writel(NVME_CMBMSC_CRE, dev->bar + NVME_REG_CMBMSC);
+
dev->cmbsz = readl(dev->bar + NVME_REG_CMBSZ);
if (!dev->cmbsz)
return;
@@ -1846,6 +1843,16 @@
return;
/*
+ * Tell the controller about the host side address mapping the CMB,
+ * and enable CMB decoding for the NVMe 1.4+ scheme:
+ */
+ if (NVME_CAP_CMBS(dev->ctrl.cap)) {
+ hi_lo_writeq(NVME_CMBMSC_CRE | NVME_CMBMSC_CMSE |
+ (pci_bus_address(pdev, bar) + offset),
+ dev->bar + NVME_REG_CMBMSC);
+ }
+
+ /*
* Controllers may support a CMB size larger than their BAR,
* for example, due to being behind a bridge. Reduce the CMB to
* the reported size of the BAR
@@ -1883,6 +1890,7 @@
static int nvme_set_host_mem(struct nvme_dev *dev, u32 bits)
{
+ u32 host_mem_size = dev->host_mem_size >> NVME_CTRL_PAGE_SHIFT;
u64 dma_addr = dev->host_mem_descs_dma;
struct nvme_command c;
int ret;
@@ -1891,8 +1899,7 @@
c.features.opcode = nvme_admin_set_features;
c.features.fid = cpu_to_le32(NVME_FEAT_HOST_MEM_BUF);
c.features.dword11 = cpu_to_le32(bits);
- c.features.dword12 = cpu_to_le32(dev->host_mem_size >>
- ilog2(dev->ctrl.page_size));
+ c.features.dword12 = cpu_to_le32(host_mem_size);
c.features.dword13 = cpu_to_le32(lower_32_bits(dma_addr));
c.features.dword14 = cpu_to_le32(upper_32_bits(dma_addr));
c.features.dword15 = cpu_to_le32(dev->nr_host_mem_descs);
@@ -1912,7 +1919,7 @@
for (i = 0; i < dev->nr_host_mem_descs; i++) {
struct nvme_host_mem_buf_desc *desc = &dev->host_mem_descs[i];
- size_t size = le32_to_cpu(desc->size) * dev->ctrl.page_size;
+ size_t size = le32_to_cpu(desc->size) * NVME_CTRL_PAGE_SIZE;
dma_free_attrs(dev->dev, size, dev->host_mem_desc_bufs[i],
le64_to_cpu(desc->addr),
@@ -1964,7 +1971,7 @@
break;
descs[i].addr = cpu_to_le64(dma_addr);
- descs[i].size = cpu_to_le32(len / dev->ctrl.page_size);
+ descs[i].size = cpu_to_le32(len / NVME_CTRL_PAGE_SIZE);
i++;
}
@@ -1980,7 +1987,7 @@
out_free_bufs:
while (--i >= 0) {
- size_t size = le32_to_cpu(descs[i].size) * dev->ctrl.page_size;
+ size_t size = le32_to_cpu(descs[i].size) * NVME_CTRL_PAGE_SIZE;
dma_free_attrs(dev->dev, size, bufs[i],
le64_to_cpu(descs[i].addr),
@@ -1998,12 +2005,12 @@
static int nvme_alloc_host_mem(struct nvme_dev *dev, u64 min, u64 preferred)
{
- u32 chunk_size;
+ u64 min_chunk = min_t(u64, preferred, PAGE_SIZE * MAX_ORDER_NR_PAGES);
+ u64 hmminds = max_t(u32, dev->ctrl.hmminds * 4096, PAGE_SIZE * 2);
+ u64 chunk_size;
/* start big and work our way down */
- for (chunk_size = min_t(u64, preferred, PAGE_SIZE * MAX_ORDER_NR_PAGES);
- chunk_size >= max_t(u32, dev->ctrl.hmminds * 4096, PAGE_SIZE * 2);
- chunk_size /= 2) {
+ for (chunk_size = min_chunk; chunk_size >= hmminds; chunk_size /= 2) {
if (!__nvme_alloc_host_mem(dev, preferred, chunk_size)) {
if (!min || dev->host_mem_size >= min)
return 0;
@@ -2069,7 +2076,7 @@
unsigned int nr_read_queues, nr_write_queues = dev->nr_write_queues;
/*
- * If there is no interupt available for queues, ensure that
+ * If there is no interrupt available for queues, ensure that
* the default queue is set to 1. The affinity set size is
* also set to one, but the irq core ignores it for this case.
*
@@ -2105,32 +2112,30 @@
.calc_sets = nvme_calc_irq_sets,
.priv = dev,
};
- unsigned int irq_queues, this_p_queues;
+ unsigned int irq_queues, poll_queues;
/*
- * Poll queues don't need interrupts, but we need at least one IO
- * queue left over for non-polled IO.
+ * Poll queues don't need interrupts, but we need at least one I/O queue
+ * left over for non-polled I/O.
*/
- this_p_queues = dev->nr_poll_queues;
- if (this_p_queues >= nr_io_queues) {
- this_p_queues = nr_io_queues - 1;
- irq_queues = 1;
- } else {
- irq_queues = nr_io_queues - this_p_queues + 1;
- }
- dev->io_queues[HCTX_TYPE_POLL] = this_p_queues;
+ poll_queues = min(dev->nr_poll_queues, nr_io_queues - 1);
+ dev->io_queues[HCTX_TYPE_POLL] = poll_queues;
- /* Initialize for the single interrupt case */
+ /*
+ * Initialize for the single interrupt case, will be updated in
+ * nvme_calc_irq_sets().
+ */
dev->io_queues[HCTX_TYPE_DEFAULT] = 1;
dev->io_queues[HCTX_TYPE_READ] = 0;
/*
- * Some Apple controllers require all queues to use the
- * first vector.
+ * We need interrupts for the admin queue and each non-polled I/O queue,
+ * but some Apple controllers require all queues to use the first
+ * vector.
*/
- if (dev->ctrl.quirks & NVME_QUIRK_SINGLE_VECTOR)
- irq_queues = 1;
-
+ irq_queues = 1;
+ if (!(dev->ctrl.quirks & NVME_QUIRK_SINGLE_VECTOR))
+ irq_queues += (nr_io_queues - poll_queues);
return pci_alloc_irq_vectors_affinity(pdev, 1, irq_queues,
PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd);
}
@@ -2326,9 +2331,9 @@
if (dev->io_queues[HCTX_TYPE_POLL])
dev->tagset.nr_maps++;
dev->tagset.timeout = NVME_IO_TIMEOUT;
- dev->tagset.numa_node = dev_to_node(dev->dev);
- dev->tagset.queue_depth =
- min_t(int, dev->q_depth, BLK_MQ_MAX_DEPTH) - 1;
+ dev->tagset.numa_node = dev->ctrl.numa_node;
+ dev->tagset.queue_depth = min_t(unsigned int, dev->q_depth,
+ BLK_MQ_MAX_DEPTH) - 1;
dev->tagset.cmd_size = sizeof(struct nvme_iod);
dev->tagset.flags = BLK_MQ_F_SHOULD_MERGE;
dev->tagset.driver_data = dev;
@@ -2387,7 +2392,7 @@
dev->ctrl.cap = lo_hi_readq(dev->bar + NVME_REG_CAP);
- dev->q_depth = min_t(int, NVME_CAP_MQES(dev->ctrl.cap) + 1,
+ dev->q_depth = min_t(u32, NVME_CAP_MQES(dev->ctrl.cap) + 1,
io_queue_depth);
dev->ctrl.sqsize = dev->q_depth - 1; /* 0's based queue depth */
dev->db_stride = 1 << NVME_CAP_STRIDE(dev->ctrl.cap);
@@ -2527,7 +2532,8 @@
static int nvme_setup_prp_pools(struct nvme_dev *dev)
{
dev->prp_page_pool = dma_pool_create("prp list page", dev->dev,
- PAGE_SIZE, PAGE_SIZE, 0);
+ NVME_CTRL_PAGE_SIZE,
+ NVME_CTRL_PAGE_SIZE, 0);
if (!dev->prp_page_pool)
return -ENOMEM;
@@ -2559,13 +2565,13 @@
struct nvme_dev *dev = to_nvme_dev(ctrl);
nvme_dbbuf_dma_free(dev);
- put_device(dev->dev);
nvme_free_tagset(dev);
if (dev->ctrl.admin_q)
blk_put_queue(dev->ctrl.admin_q);
- kfree(dev->queues);
free_opal_dev(dev->ctrl.opal_dev);
mempool_destroy(dev->iod_mempool);
+ put_device(dev->dev);
+ kfree(dev->queues);
kfree(dev);
}
@@ -2630,6 +2636,7 @@
* Don't limit the IOMMU merged segment size.
*/
dma_set_max_seg_size(dev->dev, 0xffffffff);
+ dma_set_min_align_mask(dev->dev, NVME_CTRL_PAGE_SIZE - 1);
mutex_unlock(&dev->shutdown_lock);
@@ -2644,6 +2651,12 @@
goto out;
}
+ /*
+ * We do not support an SGL for metadata (yet), so we are limited to a
+ * single integrity segment for the separate metadata pointer.
+ */
+ dev->ctrl.max_integrity_segments = 1;
+
result = nvme_init_identify(&dev->ctrl);
if (result)
goto out;
@@ -2747,7 +2760,7 @@
{
struct pci_dev *pdev = to_pci_dev(to_nvme_dev(ctrl)->dev);
- return snprintf(buf, size, "%s", dev_name(&pdev->dev));
+ return snprintf(buf, size, "%s\n", dev_name(&pdev->dev));
}
static const struct nvme_ctrl_ops nvme_pci_ctrl_ops = {
@@ -2822,6 +2835,32 @@
return 0;
}
+#ifdef CONFIG_ACPI
+static bool nvme_acpi_storage_d3(struct pci_dev *dev)
+{
+ struct acpi_device *adev = ACPI_COMPANION(&dev->dev);
+ u8 val;
+
+ /*
+ * Look for _DSD property specifying that the storage device on the port
+ * must use D3 to support deep platform power savings during
+ * suspend-to-idle.
+ */
+
+ if (!adev)
+ return false;
+ if (fwnode_property_read_u8(acpi_fwnode_handle(adev), "StorageD3Enable",
+ &val))
+ return false;
+ return val == 1;
+}
+#else
+static inline bool nvme_acpi_storage_d3(struct pci_dev *dev)
+{
+ return false;
+}
+#endif /* CONFIG_ACPI */
+
static void nvme_async_probe(void *data, async_cookie_t cookie)
{
struct nvme_dev *dev = data;
@@ -2871,12 +2910,21 @@
quirks |= check_vendor_combination_bug(pdev);
+ if (!noacpi && nvme_acpi_storage_d3(pdev)) {
+ /*
+ * Some systems use a bios work around to ask for D3 on
+ * platforms that support kernel managed suspend.
+ */
+ dev_info(&pdev->dev,
+ "platform quirk: setting simple suspend\n");
+ quirks |= NVME_QUIRK_SIMPLE_SUSPEND;
+ }
+
/*
* Double check that our mempool alloc size will cover the biggest
* command we support.
*/
- alloc_size = nvme_pci_iod_alloc_size(dev, NVME_MAX_KB_SZ,
- NVME_MAX_SEGS, true);
+ alloc_size = nvme_pci_iod_alloc_size();
WARN_ON_ONCE(alloc_size > PAGE_SIZE);
dev->iod_mempool = mempool_create_node(1, mempool_kmalloc,
@@ -2938,6 +2986,7 @@
static void nvme_shutdown(struct pci_dev *pdev)
{
struct nvme_dev *dev = pci_get_drvdata(pdev);
+
nvme_disable_prepare_reset(dev, true);
}
@@ -2966,10 +3015,9 @@
nvme_free_host_mem(dev);
nvme_dev_remove_admin(dev);
nvme_free_queues(dev, 0);
- nvme_uninit_ctrl(&dev->ctrl);
nvme_release_prp_pools(dev);
nvme_dev_unmap(dev);
- nvme_put_ctrl(&dev->ctrl);
+ nvme_uninit_ctrl(&dev->ctrl);
}
#ifdef CONFIG_PM_SLEEP
@@ -3055,7 +3103,7 @@
/*
* Clearing npss forces a controller reset on resume. The
- * correct value will be resdicovered then.
+ * correct value will be rediscovered then.
*/
ret = nvme_disable_prepare_reset(ndev, true);
ctrl->npss = 0;
@@ -3068,6 +3116,7 @@
static int nvme_simple_suspend(struct device *dev)
{
struct nvme_dev *ndev = pci_get_drvdata(to_pci_dev(dev));
+
return nvme_disable_prepare_reset(ndev, true);
}
@@ -3141,21 +3190,23 @@
};
static const struct pci_device_id nvme_id_table[] = {
- { PCI_VDEVICE(INTEL, 0x0953),
+ { PCI_VDEVICE(INTEL, 0x0953), /* Intel 750/P3500/P3600/P3700 */
.driver_data = NVME_QUIRK_STRIPE_SIZE |
NVME_QUIRK_DEALLOCATE_ZEROES, },
- { PCI_VDEVICE(INTEL, 0x0a53),
+ { PCI_VDEVICE(INTEL, 0x0a53), /* Intel P3520 */
.driver_data = NVME_QUIRK_STRIPE_SIZE |
NVME_QUIRK_DEALLOCATE_ZEROES, },
- { PCI_VDEVICE(INTEL, 0x0a54),
+ { PCI_VDEVICE(INTEL, 0x0a54), /* Intel P4500/P4600 */
.driver_data = NVME_QUIRK_STRIPE_SIZE |
- NVME_QUIRK_DEALLOCATE_ZEROES, },
- { PCI_VDEVICE(INTEL, 0x0a55),
+ NVME_QUIRK_DEALLOCATE_ZEROES |
+ NVME_QUIRK_IGNORE_DEV_SUBNQN, },
+ { PCI_VDEVICE(INTEL, 0x0a55), /* Dell Express Flash P4600 */
.driver_data = NVME_QUIRK_STRIPE_SIZE |
NVME_QUIRK_DEALLOCATE_ZEROES, },
{ PCI_VDEVICE(INTEL, 0xf1a5), /* Intel 600P/P3100 */
.driver_data = NVME_QUIRK_NO_DEEPEST_PS |
NVME_QUIRK_MEDIUM_PRIO_SQ |
+ NVME_QUIRK_NO_TEMP_THRESH_CHANGE |
NVME_QUIRK_DISABLE_WRITE_ZEROES, },
{ PCI_VDEVICE(INTEL, 0xf1a6), /* Intel 760p/Pro 7600p */
.driver_data = NVME_QUIRK_IGNORE_DEV_SUBNQN, },
@@ -3197,7 +3248,10 @@
NVME_QUIRK_IGNORE_DEV_SUBNQN, },
{ PCI_DEVICE(0x1c5c, 0x1504), /* SK Hynix PC400 */
.driver_data = NVME_QUIRK_DISABLE_WRITE_ZEROES, },
- { PCI_DEVICE_CLASS(PCI_CLASS_STORAGE_EXPRESS, 0xffffff) },
+ { PCI_DEVICE(0x15b7, 0x2001), /* Sandisk Skyhawk */
+ .driver_data = NVME_QUIRK_DISABLE_WRITE_ZEROES, },
+ { PCI_DEVICE(0x2646, 0x2262), /* KINGSTON SKC2000 NVMe SSD */
+ .driver_data = NVME_QUIRK_NO_DEEPEST_PS, },
{ PCI_DEVICE(0x2646, 0x2263), /* KINGSTON A2000 NVMe SSD */
.driver_data = NVME_QUIRK_NO_DEEPEST_PS, },
{ PCI_DEVICE(PCI_VENDOR_ID_APPLE, 0x2001),
@@ -3206,7 +3260,10 @@
{ PCI_DEVICE(PCI_VENDOR_ID_APPLE, 0x2005),
.driver_data = NVME_QUIRK_SINGLE_VECTOR |
NVME_QUIRK_128_BYTES_SQES |
- NVME_QUIRK_SHARED_TAGS },
+ NVME_QUIRK_SHARED_TAGS |
+ NVME_QUIRK_SKIP_CID_GEN },
+
+ { PCI_DEVICE_CLASS(PCI_CLASS_STORAGE_EXPRESS, 0xffffff) },
{ 0, }
};
MODULE_DEVICE_TABLE(pci, nvme_id_table);
@@ -3232,6 +3289,7 @@
BUILD_BUG_ON(sizeof(struct nvme_create_sq) != 64);
BUILD_BUG_ON(sizeof(struct nvme_delete_queue) != 64);
BUILD_BUG_ON(IRQ_AFFINITY_MAX_SETS < 2);
+
return pci_register_driver(&nvme_driver);
}
diff --git a/drivers/nvme/host/rdma.c b/drivers/nvme/host/rdma.c
index dcc3d23..8eacc9b 100644
--- a/drivers/nvme/host/rdma.c
+++ b/drivers/nvme/host/rdma.c
@@ -34,6 +34,11 @@
#define NVME_RDMA_MAX_INLINE_SEGMENTS 4
+#define NVME_RDMA_DATA_SGL_SIZE \
+ (sizeof(struct scatterlist) * NVME_INLINE_SG_CNT)
+#define NVME_RDMA_METADATA_SGL_SIZE \
+ (sizeof(struct scatterlist) * NVME_INLINE_METADATA_SG_CNT)
+
struct nvme_rdma_device {
struct ib_device *dev;
struct ib_pd *pd;
@@ -48,6 +53,11 @@
u64 dma;
};
+struct nvme_rdma_sgl {
+ int nents;
+ struct sg_table sg_table;
+};
+
struct nvme_rdma_queue;
struct nvme_rdma_request {
struct nvme_request req;
@@ -58,12 +68,12 @@
refcount_t ref;
struct ib_sge sge[1 + NVME_RDMA_MAX_INLINE_SEGMENTS];
u32 num_sge;
- int nents;
struct ib_reg_wr reg_wr;
struct ib_cqe reg_cqe;
struct nvme_rdma_queue *queue;
- struct sg_table sg_table;
- struct scatterlist first_sgl[];
+ struct nvme_rdma_sgl data_sgl;
+ struct nvme_rdma_sgl *metadata_sgl;
+ bool use_sig_mr;
};
enum nvme_rdma_queue_flags {
@@ -85,6 +95,9 @@
struct rdma_cm_id *cm_id;
int cm_error;
struct completion cm_done;
+ bool pi_support;
+ int cq_size;
+ struct mutex queue_lock;
};
struct nvme_rdma_ctrl {
@@ -138,18 +151,11 @@
static int nvme_rdma_cm_handler(struct rdma_cm_id *cm_id,
struct rdma_cm_event *event);
static void nvme_rdma_recv_done(struct ib_cq *cq, struct ib_wc *wc);
+static void nvme_rdma_complete_rq(struct request *rq);
static const struct blk_mq_ops nvme_rdma_mq_ops;
static const struct blk_mq_ops nvme_rdma_admin_mq_ops;
-/* XXX: really should move to a generic header sooner or later.. */
-static inline void put_unaligned_le24(u32 val, u8 *p)
-{
- *p++ = val;
- *p++ = val >> 8;
- *p++ = val >> 16;
-}
-
static inline int nvme_rdma_queue_idx(struct nvme_rdma_queue *queue)
{
return queue - queue->ctrl->queues;
@@ -269,6 +275,9 @@
init_attr.qp_type = IB_QPT_RC;
init_attr.send_cq = queue->ib_cq;
init_attr.recv_cq = queue->ib_cq;
+ if (queue->pi_support)
+ init_attr.create_flags |= IB_QP_CREATE_INTEGRITY_EN;
+ init_attr.qp_context = queue;
ret = rdma_create_qp(queue->cm_id, dev->pd, &init_attr);
@@ -298,6 +307,12 @@
if (!req->sqe.data)
return -ENOMEM;
+ /* metadata nvme_rdma_sgl struct is located after command's data SGL */
+ if (queue->pi_support)
+ req->metadata_sgl = (void *)nvme_req(rq) +
+ sizeof(struct nvme_rdma_request) +
+ NVME_RDMA_DATA_SGL_SIZE;
+
req->queue = queue;
return 0;
@@ -397,6 +412,14 @@
return NULL;
}
+static void nvme_rdma_free_cq(struct nvme_rdma_queue *queue)
+{
+ if (nvme_rdma_poll_queue(queue))
+ ib_free_cq(queue->ib_cq);
+ else
+ ib_cq_pool_put(queue->ib_cq, queue->cq_size);
+}
+
static void nvme_rdma_destroy_queue_ib(struct nvme_rdma_queue *queue)
{
struct nvme_rdma_device *dev;
@@ -408,6 +431,8 @@
dev = queue->device;
ibdev = dev->dev;
+ if (queue->pi_support)
+ ib_mr_pool_destroy(queue->qp, &queue->qp->sig_mrs);
ib_mr_pool_destroy(queue->qp, &queue->qp->rdma_mrs);
/*
@@ -416,7 +441,7 @@
* the destruction of the QP shouldn't use rdma_cm API.
*/
ib_destroy_qp(queue->qp);
- ib_free_cq(queue->ib_cq);
+ nvme_rdma_free_cq(queue);
nvme_rdma_free_ring(ibdev, queue->rsp_ring, queue->queue_size,
sizeof(struct nvme_completion), DMA_FROM_DEVICE);
@@ -424,10 +449,47 @@
nvme_rdma_dev_put(dev);
}
-static int nvme_rdma_get_max_fr_pages(struct ib_device *ibdev)
+static int nvme_rdma_get_max_fr_pages(struct ib_device *ibdev, bool pi_support)
{
- return min_t(u32, NVME_RDMA_MAX_SEGMENTS,
- ibdev->attrs.max_fast_reg_page_list_len - 1);
+ u32 max_page_list_len;
+
+ if (pi_support)
+ max_page_list_len = ibdev->attrs.max_pi_fast_reg_page_list_len;
+ else
+ max_page_list_len = ibdev->attrs.max_fast_reg_page_list_len;
+
+ return min_t(u32, NVME_RDMA_MAX_SEGMENTS, max_page_list_len - 1);
+}
+
+static int nvme_rdma_create_cq(struct ib_device *ibdev,
+ struct nvme_rdma_queue *queue)
+{
+ int ret, comp_vector, idx = nvme_rdma_queue_idx(queue);
+ enum ib_poll_context poll_ctx;
+
+ /*
+ * Spread I/O queues completion vectors according their queue index.
+ * Admin queues can always go on completion vector 0.
+ */
+ comp_vector = (idx == 0 ? idx : idx - 1) % ibdev->num_comp_vectors;
+
+ /* Polling queues need direct cq polling context */
+ if (nvme_rdma_poll_queue(queue)) {
+ poll_ctx = IB_POLL_DIRECT;
+ queue->ib_cq = ib_alloc_cq(ibdev, queue, queue->cq_size,
+ comp_vector, poll_ctx);
+ } else {
+ poll_ctx = IB_POLL_SOFTIRQ;
+ queue->ib_cq = ib_cq_pool_get(ibdev, queue->cq_size,
+ comp_vector, poll_ctx);
+ }
+
+ if (IS_ERR(queue->ib_cq)) {
+ ret = PTR_ERR(queue->ib_cq);
+ return ret;
+ }
+
+ return 0;
}
static int nvme_rdma_create_queue_ib(struct nvme_rdma_queue *queue)
@@ -435,8 +497,6 @@
struct ib_device *ibdev;
const int send_wr_factor = 3; /* MR, SEND, INV */
const int cq_factor = send_wr_factor + 1; /* + RECV */
- int comp_vector, idx = nvme_rdma_queue_idx(queue);
- enum ib_poll_context poll_ctx;
int ret, pages_per_mr;
queue->device = nvme_rdma_find_get_device(queue->cm_id);
@@ -447,26 +507,12 @@
}
ibdev = queue->device->dev;
- /*
- * Spread I/O queues completion vectors according their queue index.
- * Admin queues can always go on completion vector 0.
- */
- comp_vector = (idx == 0 ? idx : idx - 1) % ibdev->num_comp_vectors;
-
- /* Polling queues need direct cq polling context */
- if (nvme_rdma_poll_queue(queue))
- poll_ctx = IB_POLL_DIRECT;
- else
- poll_ctx = IB_POLL_SOFTIRQ;
-
/* +1 for ib_stop_cq */
- queue->ib_cq = ib_alloc_cq(ibdev, queue,
- cq_factor * queue->queue_size + 1,
- comp_vector, poll_ctx);
- if (IS_ERR(queue->ib_cq)) {
- ret = PTR_ERR(queue->ib_cq);
+ queue->cq_size = cq_factor * queue->queue_size + 1;
+
+ ret = nvme_rdma_create_cq(ibdev, queue);
+ if (ret)
goto out_put_dev;
- }
ret = nvme_rdma_create_qp(queue, send_wr_factor);
if (ret)
@@ -484,7 +530,7 @@
* misaligned we'll end up using two entries for a single data page,
* so one additional entry is required.
*/
- pages_per_mr = nvme_rdma_get_max_fr_pages(ibdev) + 1;
+ pages_per_mr = nvme_rdma_get_max_fr_pages(ibdev, queue->pi_support) + 1;
ret = ib_mr_pool_init(queue->qp, &queue->qp->rdma_mrs,
queue->queue_size,
IB_MR_TYPE_MEM_REG,
@@ -492,21 +538,35 @@
if (ret) {
dev_err(queue->ctrl->ctrl.device,
"failed to initialize MR pool sized %d for QID %d\n",
- queue->queue_size, idx);
+ queue->queue_size, nvme_rdma_queue_idx(queue));
goto out_destroy_ring;
}
+ if (queue->pi_support) {
+ ret = ib_mr_pool_init(queue->qp, &queue->qp->sig_mrs,
+ queue->queue_size, IB_MR_TYPE_INTEGRITY,
+ pages_per_mr, pages_per_mr);
+ if (ret) {
+ dev_err(queue->ctrl->ctrl.device,
+ "failed to initialize PI MR pool sized %d for QID %d\n",
+ queue->queue_size, nvme_rdma_queue_idx(queue));
+ goto out_destroy_mr_pool;
+ }
+ }
+
set_bit(NVME_RDMA_Q_TR_READY, &queue->flags);
return 0;
+out_destroy_mr_pool:
+ ib_mr_pool_destroy(queue->qp, &queue->qp->rdma_mrs);
out_destroy_ring:
nvme_rdma_free_ring(ibdev, queue->rsp_ring, queue->queue_size,
sizeof(struct nvme_completion), DMA_FROM_DEVICE);
out_destroy_qp:
rdma_destroy_qp(queue->cm_id);
out_destroy_ib_cq:
- ib_free_cq(queue->ib_cq);
+ nvme_rdma_free_cq(queue);
out_put_dev:
nvme_rdma_dev_put(queue->device);
return ret;
@@ -520,7 +580,12 @@
int ret;
queue = &ctrl->queues[idx];
+ mutex_init(&queue->queue_lock);
queue->ctrl = ctrl;
+ if (idx && ctrl->ctrl.max_integrity_segments)
+ queue->pi_support = true;
+ else
+ queue->pi_support = false;
init_completion(&queue->cm_done);
if (idx > 0)
@@ -535,7 +600,8 @@
if (IS_ERR(queue->cm_id)) {
dev_info(ctrl->ctrl.device,
"failed to create CM ID: %ld\n", PTR_ERR(queue->cm_id));
- return PTR_ERR(queue->cm_id);
+ ret = PTR_ERR(queue->cm_id);
+ goto out_destroy_mutex;
}
if (ctrl->ctrl.opts->mask & NVMF_OPT_HOST_TRADDR)
@@ -565,6 +631,8 @@
out_destroy_cm_id:
rdma_destroy_id(queue->cm_id);
nvme_rdma_destroy_queue_ib(queue);
+out_destroy_mutex:
+ mutex_destroy(&queue->queue_lock);
return ret;
}
@@ -576,9 +644,10 @@
static void nvme_rdma_stop_queue(struct nvme_rdma_queue *queue)
{
- if (!test_and_clear_bit(NVME_RDMA_Q_LIVE, &queue->flags))
- return;
- __nvme_rdma_stop_queue(queue);
+ mutex_lock(&queue->queue_lock);
+ if (test_and_clear_bit(NVME_RDMA_Q_LIVE, &queue->flags))
+ __nvme_rdma_stop_queue(queue);
+ mutex_unlock(&queue->queue_lock);
}
static void nvme_rdma_free_queue(struct nvme_rdma_queue *queue)
@@ -586,8 +655,9 @@
if (!test_and_clear_bit(NVME_RDMA_Q_ALLOCATED, &queue->flags))
return;
- nvme_rdma_destroy_queue_ib(queue);
rdma_destroy_id(queue->cm_id);
+ nvme_rdma_destroy_queue_ib(queue);
+ mutex_destroy(&queue->queue_lock);
}
static void nvme_rdma_free_io_queues(struct nvme_rdma_ctrl *ctrl)
@@ -734,7 +804,7 @@
set->reserved_tags = 2; /* connect + keep-alive */
set->numa_node = nctrl->numa_node;
set->cmd_size = sizeof(struct nvme_rdma_request) +
- SG_CHUNK_SIZE * sizeof(struct scatterlist);
+ NVME_RDMA_DATA_SGL_SIZE;
set->driver_data = ctrl;
set->nr_hw_queues = 1;
set->timeout = ADMIN_TIMEOUT;
@@ -748,7 +818,10 @@
set->numa_node = nctrl->numa_node;
set->flags = BLK_MQ_F_SHOULD_MERGE;
set->cmd_size = sizeof(struct nvme_rdma_request) +
- SG_CHUNK_SIZE * sizeof(struct scatterlist);
+ NVME_RDMA_DATA_SGL_SIZE;
+ if (nctrl->max_integrity_segments)
+ set->cmd_size += sizeof(struct nvme_rdma_sgl) +
+ NVME_RDMA_METADATA_SGL_SIZE;
set->driver_data = ctrl;
set->nr_hw_queues = nctrl->queue_count - 1;
set->timeout = NVME_IO_TIMEOUT;
@@ -782,6 +855,7 @@
static int nvme_rdma_configure_admin_queue(struct nvme_rdma_ctrl *ctrl,
bool new)
{
+ bool pi_capable = false;
int error;
error = nvme_rdma_alloc_queue(ctrl, 0, NVME_AQ_DEPTH);
@@ -789,9 +863,15 @@
return error;
ctrl->device = ctrl->queues[0].device;
- ctrl->ctrl.numa_node = dev_to_node(ctrl->device->dev->dma_device);
+ ctrl->ctrl.numa_node = ibdev_to_node(ctrl->device->dev);
- ctrl->max_fr_pages = nvme_rdma_get_max_fr_pages(ctrl->device->dev);
+ /* T10-PI support */
+ if (ctrl->device->dev->attrs.device_cap_flags &
+ IB_DEVICE_INTEGRITY_HANDOVER)
+ pi_capable = true;
+
+ ctrl->max_fr_pages = nvme_rdma_get_max_fr_pages(ctrl->device->dev,
+ pi_capable);
/*
* Bind the async event SQE DMA mapping to the admin queue lifetime.
@@ -833,6 +913,10 @@
ctrl->ctrl.max_segments = ctrl->max_fr_pages;
ctrl->ctrl.max_hw_sectors = ctrl->max_fr_pages << (ilog2(SZ_4K) - 9);
+ if (pi_capable)
+ ctrl->ctrl.max_integrity_segments = ctrl->max_fr_pages;
+ else
+ ctrl->ctrl.max_integrity_segments = 0;
blk_mq_unquiesce_queue(ctrl->ctrl.admin_q);
@@ -1019,11 +1103,13 @@
return ret;
if (ctrl->ctrl.icdoff) {
+ ret = -EOPNOTSUPP;
dev_err(ctrl->ctrl.device, "icdoff is not supported!\n");
goto destroy_admin;
}
if (!(ctrl->ctrl.sgls & (1 << 2))) {
+ ret = -EOPNOTSUPP;
dev_err(ctrl->ctrl.device,
"Mandatory keyed sgls are not supported!\n");
goto destroy_admin;
@@ -1053,8 +1139,14 @@
changed = nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_LIVE);
if (!changed) {
- /* state change failure is ok if we're in DELETING state */
- WARN_ON_ONCE(ctrl->ctrl.state != NVME_CTRL_DELETING);
+ /*
+ * state change failure is ok if we started ctrl delete,
+ * unless we're during creation of a new controller to
+ * avoid races with teardown flow.
+ */
+ WARN_ON_ONCE(ctrl->ctrl.state != NVME_CTRL_DELETING &&
+ ctrl->ctrl.state != NVME_CTRL_DELETING_NOIO);
+ WARN_ON_ONCE(new);
ret = -EINVAL;
goto destroy_io;
}
@@ -1108,14 +1200,16 @@
struct nvme_rdma_ctrl, err_work);
nvme_stop_keep_alive(&ctrl->ctrl);
+ flush_work(&ctrl->ctrl.async_event_work);
nvme_rdma_teardown_io_queues(ctrl, false);
nvme_start_queues(&ctrl->ctrl);
nvme_rdma_teardown_admin_queue(ctrl, false);
blk_mq_unquiesce_queue(ctrl->ctrl.admin_q);
if (!nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_CONNECTING)) {
- /* state change failure is ok if we're in DELETING state */
- WARN_ON_ONCE(ctrl->ctrl.state != NVME_CTRL_DELETING);
+ /* state change failure is ok if we started ctrl delete */
+ WARN_ON_ONCE(ctrl->ctrl.state != NVME_CTRL_DELETING &&
+ ctrl->ctrl.state != NVME_CTRL_DELETING_NOIO);
return;
}
@@ -1131,10 +1225,20 @@
queue_work(nvme_reset_wq, &ctrl->err_work);
}
+static void nvme_rdma_end_request(struct nvme_rdma_request *req)
+{
+ struct request *rq = blk_mq_rq_from_pdu(req);
+
+ if (!refcount_dec_and_test(&req->ref))
+ return;
+ if (!nvme_try_complete_req(rq, req->status, req->result))
+ nvme_rdma_complete_rq(rq);
+}
+
static void nvme_rdma_wr_error(struct ib_cq *cq, struct ib_wc *wc,
const char *op)
{
- struct nvme_rdma_queue *queue = cq->cq_context;
+ struct nvme_rdma_queue *queue = wc->qp->qp_context;
struct nvme_rdma_ctrl *ctrl = queue->ctrl;
if (ctrl->ctrl.state == NVME_CTRL_LIVE)
@@ -1155,16 +1259,11 @@
{
struct nvme_rdma_request *req =
container_of(wc->wr_cqe, struct nvme_rdma_request, reg_cqe);
- struct request *rq = blk_mq_rq_from_pdu(req);
- if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ if (unlikely(wc->status != IB_WC_SUCCESS))
nvme_rdma_wr_error(cq, wc, "LOCAL_INV");
- return;
- }
-
- if (refcount_dec_and_test(&req->ref))
- nvme_end_request(rq, req->status, req->result);
-
+ else
+ nvme_rdma_end_request(req);
}
static int nvme_rdma_inv_rkey(struct nvme_rdma_queue *queue,
@@ -1190,19 +1289,29 @@
struct nvme_rdma_request *req = blk_mq_rq_to_pdu(rq);
struct nvme_rdma_device *dev = queue->device;
struct ib_device *ibdev = dev->dev;
+ struct list_head *pool = &queue->qp->rdma_mrs;
if (!blk_rq_nr_phys_segments(rq))
return;
+ if (blk_integrity_rq(rq)) {
+ ib_dma_unmap_sg(ibdev, req->metadata_sgl->sg_table.sgl,
+ req->metadata_sgl->nents, rq_dma_dir(rq));
+ sg_free_table_chained(&req->metadata_sgl->sg_table,
+ NVME_INLINE_METADATA_SG_CNT);
+ }
+
+ if (req->use_sig_mr)
+ pool = &queue->qp->sig_mrs;
+
if (req->mr) {
- ib_mr_pool_put(queue->qp, &queue->qp->rdma_mrs, req->mr);
+ ib_mr_pool_put(queue->qp, pool, req->mr);
req->mr = NULL;
}
- ib_dma_unmap_sg(ibdev, req->sg_table.sgl, req->nents, rq_dma_dir(rq));
-
- nvme_cleanup_cmd(rq);
- sg_free_table_chained(&req->sg_table, SG_CHUNK_SIZE);
+ ib_dma_unmap_sg(ibdev, req->data_sgl.sg_table.sgl, req->data_sgl.nents,
+ rq_dma_dir(rq));
+ sg_free_table_chained(&req->data_sgl.sg_table, NVME_INLINE_SG_CNT);
}
static int nvme_rdma_set_sg_null(struct nvme_command *c)
@@ -1221,16 +1330,17 @@
int count)
{
struct nvme_sgl_desc *sg = &c->common.dptr.sgl;
- struct scatterlist *sgl = req->sg_table.sgl;
struct ib_sge *sge = &req->sge[1];
+ struct scatterlist *sgl;
u32 len = 0;
int i;
- for (i = 0; i < count; i++, sgl++, sge++) {
+ for_each_sg(req->data_sgl.sg_table.sgl, sgl, count, i) {
sge->addr = sg_dma_address(sgl);
sge->length = sg_dma_len(sgl);
sge->lkey = queue->device->pd->local_dma_lkey;
len += sge->length;
+ sge++;
}
sg->addr = cpu_to_le64(queue->ctrl->ctrl.icdoff);
@@ -1246,8 +1356,8 @@
{
struct nvme_keyed_sgl_desc *sg = &c->common.dptr.ksgl;
- sg->addr = cpu_to_le64(sg_dma_address(req->sg_table.sgl));
- put_unaligned_le24(sg_dma_len(req->sg_table.sgl), sg->length);
+ sg->addr = cpu_to_le64(sg_dma_address(req->data_sgl.sg_table.sgl));
+ put_unaligned_le24(sg_dma_len(req->data_sgl.sg_table.sgl), sg->length);
put_unaligned_le32(queue->device->pd->unsafe_global_rkey, sg->key);
sg->type = NVME_KEY_SGL_FMT_DATA_DESC << 4;
return 0;
@@ -1268,7 +1378,8 @@
* Align the MR to a 4K page size to match the ctrl page size and
* the block virtual boundary.
*/
- nr = ib_map_mr_sg(req->mr, req->sg_table.sgl, count, NULL, SZ_4K);
+ nr = ib_map_mr_sg(req->mr, req->data_sgl.sg_table.sgl, count, NULL,
+ SZ_4K);
if (unlikely(nr < count)) {
ib_mr_pool_put(queue->qp, &queue->qp->rdma_mrs, req->mr);
req->mr = NULL;
@@ -1299,12 +1410,125 @@
return 0;
}
+static void nvme_rdma_set_sig_domain(struct blk_integrity *bi,
+ struct nvme_command *cmd, struct ib_sig_domain *domain,
+ u16 control, u8 pi_type)
+{
+ domain->sig_type = IB_SIG_TYPE_T10_DIF;
+ domain->sig.dif.bg_type = IB_T10DIF_CRC;
+ domain->sig.dif.pi_interval = 1 << bi->interval_exp;
+ domain->sig.dif.ref_tag = le32_to_cpu(cmd->rw.reftag);
+ if (control & NVME_RW_PRINFO_PRCHK_REF)
+ domain->sig.dif.ref_remap = true;
+
+ domain->sig.dif.app_tag = le16_to_cpu(cmd->rw.apptag);
+ domain->sig.dif.apptag_check_mask = le16_to_cpu(cmd->rw.appmask);
+ domain->sig.dif.app_escape = true;
+ if (pi_type == NVME_NS_DPS_PI_TYPE3)
+ domain->sig.dif.ref_escape = true;
+}
+
+static void nvme_rdma_set_sig_attrs(struct blk_integrity *bi,
+ struct nvme_command *cmd, struct ib_sig_attrs *sig_attrs,
+ u8 pi_type)
+{
+ u16 control = le16_to_cpu(cmd->rw.control);
+
+ memset(sig_attrs, 0, sizeof(*sig_attrs));
+ if (control & NVME_RW_PRINFO_PRACT) {
+ /* for WRITE_INSERT/READ_STRIP no memory domain */
+ sig_attrs->mem.sig_type = IB_SIG_TYPE_NONE;
+ nvme_rdma_set_sig_domain(bi, cmd, &sig_attrs->wire, control,
+ pi_type);
+ /* Clear the PRACT bit since HCA will generate/verify the PI */
+ control &= ~NVME_RW_PRINFO_PRACT;
+ cmd->rw.control = cpu_to_le16(control);
+ } else {
+ /* for WRITE_PASS/READ_PASS both wire/memory domains exist */
+ nvme_rdma_set_sig_domain(bi, cmd, &sig_attrs->wire, control,
+ pi_type);
+ nvme_rdma_set_sig_domain(bi, cmd, &sig_attrs->mem, control,
+ pi_type);
+ }
+}
+
+static void nvme_rdma_set_prot_checks(struct nvme_command *cmd, u8 *mask)
+{
+ *mask = 0;
+ if (le16_to_cpu(cmd->rw.control) & NVME_RW_PRINFO_PRCHK_REF)
+ *mask |= IB_SIG_CHECK_REFTAG;
+ if (le16_to_cpu(cmd->rw.control) & NVME_RW_PRINFO_PRCHK_GUARD)
+ *mask |= IB_SIG_CHECK_GUARD;
+}
+
+static void nvme_rdma_sig_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+ if (unlikely(wc->status != IB_WC_SUCCESS))
+ nvme_rdma_wr_error(cq, wc, "SIG");
+}
+
+static int nvme_rdma_map_sg_pi(struct nvme_rdma_queue *queue,
+ struct nvme_rdma_request *req, struct nvme_command *c,
+ int count, int pi_count)
+{
+ struct nvme_rdma_sgl *sgl = &req->data_sgl;
+ struct ib_reg_wr *wr = &req->reg_wr;
+ struct request *rq = blk_mq_rq_from_pdu(req);
+ struct nvme_ns *ns = rq->q->queuedata;
+ struct bio *bio = rq->bio;
+ struct nvme_keyed_sgl_desc *sg = &c->common.dptr.ksgl;
+ int nr;
+
+ req->mr = ib_mr_pool_get(queue->qp, &queue->qp->sig_mrs);
+ if (WARN_ON_ONCE(!req->mr))
+ return -EAGAIN;
+
+ nr = ib_map_mr_sg_pi(req->mr, sgl->sg_table.sgl, count, NULL,
+ req->metadata_sgl->sg_table.sgl, pi_count, NULL,
+ SZ_4K);
+ if (unlikely(nr))
+ goto mr_put;
+
+ nvme_rdma_set_sig_attrs(blk_get_integrity(bio->bi_disk), c,
+ req->mr->sig_attrs, ns->pi_type);
+ nvme_rdma_set_prot_checks(c, &req->mr->sig_attrs->check_mask);
+
+ ib_update_fast_reg_key(req->mr, ib_inc_rkey(req->mr->rkey));
+
+ req->reg_cqe.done = nvme_rdma_sig_done;
+ memset(wr, 0, sizeof(*wr));
+ wr->wr.opcode = IB_WR_REG_MR_INTEGRITY;
+ wr->wr.wr_cqe = &req->reg_cqe;
+ wr->wr.num_sge = 0;
+ wr->wr.send_flags = 0;
+ wr->mr = req->mr;
+ wr->key = req->mr->rkey;
+ wr->access = IB_ACCESS_LOCAL_WRITE |
+ IB_ACCESS_REMOTE_READ |
+ IB_ACCESS_REMOTE_WRITE;
+
+ sg->addr = cpu_to_le64(req->mr->iova);
+ put_unaligned_le24(req->mr->length, sg->length);
+ put_unaligned_le32(req->mr->rkey, sg->key);
+ sg->type = NVME_KEY_SGL_FMT_DATA_DESC << 4;
+
+ return 0;
+
+mr_put:
+ ib_mr_pool_put(queue->qp, &queue->qp->sig_mrs, req->mr);
+ req->mr = NULL;
+ if (nr < 0)
+ return nr;
+ return -EINVAL;
+}
+
static int nvme_rdma_map_data(struct nvme_rdma_queue *queue,
struct request *rq, struct nvme_command *c)
{
struct nvme_rdma_request *req = blk_mq_rq_to_pdu(rq);
struct nvme_rdma_device *dev = queue->device;
struct ib_device *ibdev = dev->dev;
+ int pi_count = 0;
int count, ret;
req->num_sge = 1;
@@ -1315,22 +1539,52 @@
if (!blk_rq_nr_phys_segments(rq))
return nvme_rdma_set_sg_null(c);
- req->sg_table.sgl = req->first_sgl;
- ret = sg_alloc_table_chained(&req->sg_table,
- blk_rq_nr_phys_segments(rq), req->sg_table.sgl,
- SG_CHUNK_SIZE);
+ req->data_sgl.sg_table.sgl = (struct scatterlist *)(req + 1);
+ ret = sg_alloc_table_chained(&req->data_sgl.sg_table,
+ blk_rq_nr_phys_segments(rq), req->data_sgl.sg_table.sgl,
+ NVME_INLINE_SG_CNT);
if (ret)
return -ENOMEM;
- req->nents = blk_rq_map_sg(rq->q, rq, req->sg_table.sgl);
+ req->data_sgl.nents = blk_rq_map_sg(rq->q, rq,
+ req->data_sgl.sg_table.sgl);
- count = ib_dma_map_sg(ibdev, req->sg_table.sgl, req->nents,
- rq_dma_dir(rq));
+ count = ib_dma_map_sg(ibdev, req->data_sgl.sg_table.sgl,
+ req->data_sgl.nents, rq_dma_dir(rq));
if (unlikely(count <= 0)) {
ret = -EIO;
goto out_free_table;
}
+ if (blk_integrity_rq(rq)) {
+ req->metadata_sgl->sg_table.sgl =
+ (struct scatterlist *)(req->metadata_sgl + 1);
+ ret = sg_alloc_table_chained(&req->metadata_sgl->sg_table,
+ blk_rq_count_integrity_sg(rq->q, rq->bio),
+ req->metadata_sgl->sg_table.sgl,
+ NVME_INLINE_METADATA_SG_CNT);
+ if (unlikely(ret)) {
+ ret = -ENOMEM;
+ goto out_unmap_sg;
+ }
+
+ req->metadata_sgl->nents = blk_rq_map_integrity_sg(rq->q,
+ rq->bio, req->metadata_sgl->sg_table.sgl);
+ pi_count = ib_dma_map_sg(ibdev,
+ req->metadata_sgl->sg_table.sgl,
+ req->metadata_sgl->nents,
+ rq_dma_dir(rq));
+ if (unlikely(pi_count <= 0)) {
+ ret = -EIO;
+ goto out_free_pi_table;
+ }
+ }
+
+ if (req->use_sig_mr) {
+ ret = nvme_rdma_map_sg_pi(queue, req, c, count, pi_count);
+ goto out;
+ }
+
if (count <= dev->num_inline_segments) {
if (rq_data_dir(rq) == WRITE && nvme_rdma_queue_idx(queue) &&
queue->ctrl->use_inline_data &&
@@ -1349,14 +1603,23 @@
ret = nvme_rdma_map_sg_fr(queue, req, c, count);
out:
if (unlikely(ret))
- goto out_unmap_sg;
+ goto out_unmap_pi_sg;
return 0;
+out_unmap_pi_sg:
+ if (blk_integrity_rq(rq))
+ ib_dma_unmap_sg(ibdev, req->metadata_sgl->sg_table.sgl,
+ req->metadata_sgl->nents, rq_dma_dir(rq));
+out_free_pi_table:
+ if (blk_integrity_rq(rq))
+ sg_free_table_chained(&req->metadata_sgl->sg_table,
+ NVME_INLINE_METADATA_SG_CNT);
out_unmap_sg:
- ib_dma_unmap_sg(ibdev, req->sg_table.sgl, req->nents, rq_dma_dir(rq));
+ ib_dma_unmap_sg(ibdev, req->data_sgl.sg_table.sgl, req->data_sgl.nents,
+ rq_dma_dir(rq));
out_free_table:
- sg_free_table_chained(&req->sg_table, SG_CHUNK_SIZE);
+ sg_free_table_chained(&req->data_sgl.sg_table, NVME_INLINE_SG_CNT);
return ret;
}
@@ -1366,15 +1629,11 @@
container_of(wc->wr_cqe, struct nvme_rdma_qe, cqe);
struct nvme_rdma_request *req =
container_of(qe, struct nvme_rdma_request, sqe);
- struct request *rq = blk_mq_rq_from_pdu(req);
- if (unlikely(wc->status != IB_WC_SUCCESS)) {
+ if (unlikely(wc->status != IB_WC_SUCCESS))
nvme_rdma_wr_error(cq, wc, "SEND");
- return;
- }
-
- if (refcount_dec_and_test(&req->ref))
- nvme_end_request(rq, req->status, req->result);
+ else
+ nvme_rdma_end_request(req);
}
static int nvme_rdma_post_send(struct nvme_rdma_queue *queue,
@@ -1385,7 +1644,7 @@
int ret;
sge->addr = qe->dma;
- sge->length = sizeof(struct nvme_command),
+ sge->length = sizeof(struct nvme_command);
sge->lkey = queue->device->pd->local_dma_lkey;
wr.next = NULL;
@@ -1482,10 +1741,10 @@
struct request *rq;
struct nvme_rdma_request *req;
- rq = blk_mq_tag_to_rq(nvme_rdma_tagset(queue), cqe->command_id);
+ rq = nvme_find_rq(nvme_rdma_tagset(queue), cqe->command_id);
if (!rq) {
dev_err(queue->ctrl->ctrl.device,
- "tag 0x%x on QP %#x not found\n",
+ "got bad command_id %#x on QP %#x\n",
cqe->command_id, queue->qp->qp_num);
nvme_rdma_error_recovery(queue->ctrl);
return;
@@ -1496,10 +1755,11 @@
req->result = cqe->result;
if (wc->wc_flags & IB_WC_WITH_INVALIDATE) {
- if (unlikely(wc->ex.invalidate_rkey != req->mr->rkey)) {
+ if (unlikely(!req->mr ||
+ wc->ex.invalidate_rkey != req->mr->rkey)) {
dev_err(queue->ctrl->ctrl.device,
"Bogus remote invalidation for rkey %#x\n",
- req->mr->rkey);
+ req->mr ? req->mr->rkey : 0);
nvme_rdma_error_recovery(queue->ctrl);
}
} else if (req->mr) {
@@ -1516,15 +1776,14 @@
return;
}
- if (refcount_dec_and_test(&req->ref))
- nvme_end_request(rq, req->status, req->result);
+ nvme_rdma_end_request(req);
}
static void nvme_rdma_recv_done(struct ib_cq *cq, struct ib_wc *wc)
{
struct nvme_rdma_qe *qe =
container_of(wc->wr_cqe, struct nvme_rdma_qe, cqe);
- struct nvme_rdma_queue *queue = cq->cq_context;
+ struct nvme_rdma_queue *queue = wc->qp->qp_context;
struct ib_device *ibdev = queue->device->dev;
struct nvme_completion *cqe = qe->data;
const size_t len = sizeof(struct nvme_completion);
@@ -1549,8 +1808,8 @@
* aborts. We don't even bother to allocate a struct request
* for them but rather special case them here.
*/
- if (unlikely(nvme_rdma_queue_idx(queue) == 0 &&
- cqe->command_id >= NVME_AQ_BLK_MQ_DEPTH))
+ if (unlikely(nvme_is_aen_req(nvme_rdma_queue_idx(queue),
+ cqe->command_id)))
nvme_complete_async_event(&queue->ctrl->ctrl, cqe->status,
&cqe->result);
else
@@ -1567,14 +1826,10 @@
for (i = 0; i < queue->queue_size; i++) {
ret = nvme_rdma_post_recv(queue, &queue->rsp_ring[i]);
if (ret)
- goto out_destroy_queue_ib;
+ return ret;
}
return 0;
-
-out_destroy_queue_ib:
- nvme_rdma_destroy_queue_ib(queue);
- return ret;
}
static int nvme_rdma_conn_rejected(struct nvme_rdma_queue *queue,
@@ -1664,18 +1919,14 @@
priv.hsqsize = cpu_to_le16(queue->ctrl->ctrl.sqsize);
}
- ret = rdma_connect(queue->cm_id, ¶m);
+ ret = rdma_connect_locked(queue->cm_id, ¶m);
if (ret) {
dev_err(ctrl->ctrl.device,
- "rdma_connect failed (%d).\n", ret);
- goto out_destroy_queue_ib;
+ "rdma_connect_locked failed (%d).\n", ret);
+ return ret;
}
return 0;
-
-out_destroy_queue_ib:
- nvme_rdma_destroy_queue_ib(queue);
- return ret;
}
static int nvme_rdma_cm_handler(struct rdma_cm_id *cm_id,
@@ -1706,8 +1957,6 @@
case RDMA_CM_EVENT_ROUTE_ERROR:
case RDMA_CM_EVENT_CONNECT_ERROR:
case RDMA_CM_EVENT_UNREACHABLE:
- nvme_rdma_destroy_queue_ib(queue);
- /* fall through */
case RDMA_CM_EVENT_ADDR_ERROR:
dev_dbg(queue->ctrl->ctrl.device,
"CM error event %d\n", ev->event);
@@ -1823,11 +2072,19 @@
blk_mq_start_request(rq);
+ if (IS_ENABLED(CONFIG_BLK_DEV_INTEGRITY) &&
+ queue->pi_support &&
+ (c->common.opcode == nvme_cmd_write ||
+ c->common.opcode == nvme_cmd_read) &&
+ nvme_ns_has_pi(ns))
+ req->use_sig_mr = true;
+ else
+ req->use_sig_mr = false;
+
err = nvme_rdma_map_data(queue, rq, c);
if (unlikely(err < 0)) {
dev_err(queue->ctrl->ctrl.device,
"Failed to map data (%d)\n", err);
- nvme_cleanup_cmd(rq);
goto err;
}
@@ -1838,18 +2095,19 @@
err = nvme_rdma_post_send(queue, sqe, req->sge, req->num_sge,
req->mr ? &req->reg_wr.wr : NULL);
- if (unlikely(err)) {
- nvme_rdma_unmap_data(queue, rq);
- goto err;
- }
+ if (unlikely(err))
+ goto err_unmap;
return BLK_STS_OK;
+err_unmap:
+ nvme_rdma_unmap_data(queue, rq);
err:
if (err == -ENOMEM || err == -EAGAIN)
ret = BLK_STS_RESOURCE;
else
ret = BLK_STS_IOERR;
+ nvme_cleanup_cmd(rq);
unmap_qe:
ib_dma_unmap_single(dev, req->sqe.dma, sizeof(struct nvme_command),
DMA_TO_DEVICE);
@@ -1863,12 +2121,46 @@
return ib_process_cq_direct(queue->ib_cq, -1);
}
+static void nvme_rdma_check_pi_status(struct nvme_rdma_request *req)
+{
+ struct request *rq = blk_mq_rq_from_pdu(req);
+ struct ib_mr_status mr_status;
+ int ret;
+
+ ret = ib_check_mr_status(req->mr, IB_MR_CHECK_SIG_STATUS, &mr_status);
+ if (ret) {
+ pr_err("ib_check_mr_status failed, ret %d\n", ret);
+ nvme_req(rq)->status = NVME_SC_INVALID_PI;
+ return;
+ }
+
+ if (mr_status.fail_status & IB_MR_CHECK_SIG_STATUS) {
+ switch (mr_status.sig_err.err_type) {
+ case IB_SIG_BAD_GUARD:
+ nvme_req(rq)->status = NVME_SC_GUARD_CHECK;
+ break;
+ case IB_SIG_BAD_REFTAG:
+ nvme_req(rq)->status = NVME_SC_REFTAG_CHECK;
+ break;
+ case IB_SIG_BAD_APPTAG:
+ nvme_req(rq)->status = NVME_SC_APPTAG_CHECK;
+ break;
+ }
+ pr_err("PI error found type %d expected 0x%x vs actual 0x%x\n",
+ mr_status.sig_err.err_type, mr_status.sig_err.expected,
+ mr_status.sig_err.actual);
+ }
+}
+
static void nvme_rdma_complete_rq(struct request *rq)
{
struct nvme_rdma_request *req = blk_mq_rq_to_pdu(rq);
struct nvme_rdma_queue *queue = req->queue;
struct ib_device *ibdev = queue->device->dev;
+ if (req->use_sig_mr)
+ nvme_rdma_check_pi_status(req);
+
nvme_rdma_unmap_data(queue, rq);
ib_dma_unmap_single(ibdev, req->sqe.dma, sizeof(struct nvme_command),
DMA_TO_DEVICE);
@@ -1988,7 +2280,7 @@
static const struct nvme_ctrl_ops nvme_rdma_ctrl_ops = {
.name = "rdma",
.module = THIS_MODULE,
- .flags = NVME_F_FABRICS,
+ .flags = NVME_F_FABRICS | NVME_F_METADATA_SUPPORTED,
.reg_read32 = nvmf_reg_read32,
.reg_read64 = nvmf_reg_read64,
.reg_write32 = nvmf_reg_write32,
@@ -2113,7 +2405,6 @@
out_uninit_ctrl:
nvme_uninit_ctrl(&ctrl->ctrl);
nvme_put_ctrl(&ctrl->ctrl);
- nvme_put_ctrl(&ctrl->ctrl);
if (ret > 0)
ret = -EIO;
return ERR_PTR(ret);
diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
index 38bbbbb..6105894 100644
--- a/drivers/nvme/host/tcp.c
+++ b/drivers/nvme/host/tcp.c
@@ -20,6 +20,16 @@
struct nvme_tcp_queue;
+/* Define the socket priority to use for connections were it is desirable
+ * that the NIC consider performing optimized packet processing or filtering.
+ * A non-zero value being sufficient to indicate general consideration of any
+ * possible optimization. Making it a module param allows for alternative
+ * values that may be unique for some NIC implementations.
+ */
+static int so_priority;
+module_param(so_priority, int, 0644);
+MODULE_PARM_DESC(so_priority, "nvme tcp socket optimize priority");
+
enum nvme_tcp_send_state {
NVME_TCP_SEND_CMD_PDU = 0,
NVME_TCP_SEND_H2C_PDU,
@@ -36,6 +46,7 @@
u32 pdu_sent;
u16 ttag;
struct list_head entry;
+ struct llist_node lentry;
__le32 ddgst;
struct bio *curr_bio;
@@ -50,6 +61,7 @@
enum nvme_tcp_queue_flags {
NVME_TCP_Q_ALLOCATED = 0,
NVME_TCP_Q_LIVE = 1,
+ NVME_TCP_Q_POLLING = 2,
};
enum nvme_tcp_recv_state {
@@ -64,8 +76,11 @@
struct work_struct io_work;
int io_cpu;
- spinlock_t lock;
+ struct mutex queue_lock;
+ struct mutex send_mutex;
+ struct llist_head req_list;
struct list_head send_list;
+ bool more_requests;
/* recv state */
void *pdu;
@@ -119,8 +134,9 @@
static LIST_HEAD(nvme_tcp_ctrl_list);
static DEFINE_MUTEX(nvme_tcp_ctrl_mutex);
static struct workqueue_struct *nvme_tcp_wq;
-static struct blk_mq_ops nvme_tcp_mq_ops;
-static struct blk_mq_ops nvme_tcp_admin_mq_ops;
+static const struct blk_mq_ops nvme_tcp_mq_ops;
+static const struct blk_mq_ops nvme_tcp_admin_mq_ops;
+static int nvme_tcp_try_send(struct nvme_tcp_queue *queue);
static inline struct nvme_tcp_ctrl *to_tcp_ctrl(struct nvme_ctrl *ctrl)
{
@@ -247,15 +263,57 @@
}
}
-static inline void nvme_tcp_queue_request(struct nvme_tcp_request *req)
+static inline void nvme_tcp_send_all(struct nvme_tcp_queue *queue)
+{
+ int ret;
+
+ /* drain the send queue as much as we can... */
+ do {
+ ret = nvme_tcp_try_send(queue);
+ } while (ret > 0);
+}
+
+static inline bool nvme_tcp_queue_more(struct nvme_tcp_queue *queue)
+{
+ return !list_empty(&queue->send_list) ||
+ !llist_empty(&queue->req_list) || queue->more_requests;
+}
+
+static inline void nvme_tcp_queue_request(struct nvme_tcp_request *req,
+ bool sync, bool last)
{
struct nvme_tcp_queue *queue = req->queue;
+ bool empty;
- spin_lock(&queue->lock);
- list_add_tail(&req->entry, &queue->send_list);
- spin_unlock(&queue->lock);
+ empty = llist_add(&req->lentry, &queue->req_list) &&
+ list_empty(&queue->send_list) && !queue->request;
- queue_work_on(queue->io_cpu, nvme_tcp_wq, &queue->io_work);
+ /*
+ * if we're the first on the send_list and we can try to send
+ * directly, otherwise queue io_work. Also, only do that if we
+ * are on the same cpu, so we don't introduce contention.
+ */
+ if (queue->io_cpu == raw_smp_processor_id() &&
+ sync && empty && mutex_trylock(&queue->send_mutex)) {
+ queue->more_requests = !last;
+ nvme_tcp_send_all(queue);
+ queue->more_requests = false;
+ mutex_unlock(&queue->send_mutex);
+ }
+
+ if (last && nvme_tcp_queue_more(queue))
+ queue_work_on(queue->io_cpu, nvme_tcp_wq, &queue->io_work);
+}
+
+static void nvme_tcp_process_req_list(struct nvme_tcp_queue *queue)
+{
+ struct nvme_tcp_request *req;
+ struct llist_node *node;
+
+ for (node = llist_del_all(&queue->req_list); node; node = node->next) {
+ req = llist_entry(node, struct nvme_tcp_request, lentry);
+ list_add(&req->entry, &queue->send_list);
+ }
}
static inline struct nvme_tcp_request *
@@ -263,13 +321,17 @@
{
struct nvme_tcp_request *req;
- spin_lock(&queue->lock);
req = list_first_entry_or_null(&queue->send_list,
struct nvme_tcp_request, entry);
- if (req)
- list_del(&req->entry);
- spin_unlock(&queue->lock);
+ if (!req) {
+ nvme_tcp_process_req_list(queue);
+ req = list_first_entry_or_null(&queue->send_list,
+ struct nvme_tcp_request, entry);
+ if (unlikely(!req))
+ return NULL;
+ }
+ list_del(&req->entry);
return req;
}
@@ -429,16 +491,17 @@
{
struct request *rq;
- rq = blk_mq_tag_to_rq(nvme_tcp_tagset(queue), cqe->command_id);
+ rq = nvme_find_rq(nvme_tcp_tagset(queue), cqe->command_id);
if (!rq) {
dev_err(queue->ctrl->ctrl.device,
- "queue %d tag 0x%x not found\n",
- nvme_tcp_queue_id(queue), cqe->command_id);
+ "got bad cqe.command_id %#x on queue %d\n",
+ cqe->command_id, nvme_tcp_queue_id(queue));
nvme_tcp_error_recovery(&queue->ctrl->ctrl);
return -EINVAL;
}
- nvme_end_request(rq, cqe->status, cqe->result);
+ if (!nvme_try_complete_req(rq, cqe->status, cqe->result))
+ nvme_complete_rq(rq);
queue->nr_cqe++;
return 0;
@@ -449,11 +512,11 @@
{
struct request *rq;
- rq = blk_mq_tag_to_rq(nvme_tcp_tagset(queue), pdu->command_id);
+ rq = nvme_find_rq(nvme_tcp_tagset(queue), pdu->command_id);
if (!rq) {
dev_err(queue->ctrl->ctrl.device,
- "queue %d tag %#x not found\n",
- nvme_tcp_queue_id(queue), pdu->command_id);
+ "got bad c2hdata.command_id %#x on queue %d\n",
+ pdu->command_id, nvme_tcp_queue_id(queue));
return -ENOENT;
}
@@ -490,8 +553,8 @@
* aborts. We don't even bother to allocate a struct request
* for them but rather special case them here.
*/
- if (unlikely(nvme_tcp_queue_id(queue) == 0 &&
- cqe->command_id >= NVME_AQ_BLK_MQ_DEPTH))
+ if (unlikely(nvme_is_aen_req(nvme_tcp_queue_id(queue),
+ cqe->command_id)))
nvme_complete_async_event(&queue->ctrl->ctrl, cqe->status,
&cqe->result);
else
@@ -547,8 +610,8 @@
data->hdr.plen =
cpu_to_le32(data->hdr.hlen + hdgst + req->pdu_len + ddgst);
data->ttag = pdu->ttag;
- data->command_id = rq->tag;
- data->data_offset = cpu_to_le32(req->data_sent);
+ data->command_id = nvme_cid(rq);
+ data->data_offset = pdu->r2t_offset;
data->data_length = cpu_to_le32(req->pdu_len);
return 0;
}
@@ -560,11 +623,11 @@
struct request *rq;
int ret;
- rq = blk_mq_tag_to_rq(nvme_tcp_tagset(queue), pdu->command_id);
+ rq = nvme_find_rq(nvme_tcp_tagset(queue), pdu->command_id);
if (!rq) {
dev_err(queue->ctrl->ctrl.device,
- "queue %d tag %#x not found\n",
- nvme_tcp_queue_id(queue), pdu->command_id);
+ "got bad r2t.command_id %#x on queue %d\n",
+ pdu->command_id, nvme_tcp_queue_id(queue));
return -ENOENT;
}
req = blk_mq_rq_to_pdu(rq);
@@ -576,7 +639,7 @@
req->state = NVME_TCP_SEND_H2C_PDU;
req->offset = 0;
- nvme_tcp_queue_request(req);
+ nvme_tcp_queue_request(req, false, true);
return 0;
}
@@ -635,7 +698,8 @@
{
union nvme_result res = {};
- nvme_end_request(rq, cpu_to_le16(status << 1), res);
+ if (!nvme_try_complete_req(rq, cpu_to_le16(status << 1), res))
+ nvme_complete_rq(rq);
}
static int nvme_tcp_recv_data(struct nvme_tcp_queue *queue, struct sk_buff *skb,
@@ -643,7 +707,7 @@
{
struct nvme_tcp_data_pdu *pdu = (void *)queue->pdu;
struct request *rq =
- blk_mq_tag_to_rq(nvme_tcp_tagset(queue), pdu->command_id);
+ nvme_cid_to_rq(nvme_tcp_tagset(queue), pdu->command_id);
struct nvme_tcp_request *req = blk_mq_rq_to_pdu(rq);
while (true) {
@@ -736,8 +800,8 @@
}
if (pdu->hdr.flags & NVME_TCP_F_DATA_SUCCESS) {
- struct request *rq = blk_mq_tag_to_rq(nvme_tcp_tagset(queue),
- pdu->command_id);
+ struct request *rq = nvme_cid_to_rq(nvme_tcp_tagset(queue),
+ pdu->command_id);
nvme_tcp_end_request(rq, NVME_SC_SUCCESS);
queue->nr_cqe++;
@@ -786,7 +850,8 @@
read_lock_bh(&sk->sk_callback_lock);
queue = sk->sk_user_data;
- if (likely(queue && queue->rd_enabled))
+ if (likely(queue && queue->rd_enabled) &&
+ !test_bit(NVME_TCP_Q_POLLING, &queue->flags))
queue_work_on(queue->io_cpu, nvme_tcp_wq, &queue->io_work);
read_unlock_bh(&sk->sk_callback_lock);
}
@@ -819,7 +884,6 @@
case TCP_LAST_ACK:
case TCP_FIN_WAIT1:
case TCP_FIN_WAIT2:
- /* fallthrough */
nvme_tcp_error_recovery(&queue->ctrl->ctrl);
break;
default:
@@ -840,24 +904,34 @@
static void nvme_tcp_fail_request(struct nvme_tcp_request *req)
{
- nvme_tcp_end_request(blk_mq_rq_from_pdu(req), NVME_SC_HOST_PATH_ERROR);
+ if (nvme_tcp_async_req(req)) {
+ union nvme_result res = {};
+
+ nvme_complete_async_event(&req->queue->ctrl->ctrl,
+ cpu_to_le16(NVME_SC_HOST_PATH_ERROR), &res);
+ } else {
+ nvme_tcp_end_request(blk_mq_rq_from_pdu(req),
+ NVME_SC_HOST_PATH_ERROR);
+ }
}
static int nvme_tcp_try_send_data(struct nvme_tcp_request *req)
{
struct nvme_tcp_queue *queue = req->queue;
+ int req_data_len = req->data_len;
while (true) {
struct page *page = nvme_tcp_req_cur_page(req);
size_t offset = nvme_tcp_req_cur_offset(req);
size_t len = nvme_tcp_req_cur_length(req);
bool last = nvme_tcp_pdu_last_send(req, len);
+ int req_data_sent = req->data_sent;
int ret, flags = MSG_DONTWAIT;
- if (last && !queue->data_digest)
+ if (last && !queue->data_digest && !nvme_tcp_queue_more(queue))
flags |= MSG_EOR;
else
- flags |= MSG_MORE;
+ flags |= MSG_MORE | MSG_SENDPAGE_NOTLAST;
if (sendpage_ok(page)) {
ret = kernel_sendpage(queue->sock, page, offset, len,
@@ -869,12 +943,19 @@
if (ret <= 0)
return ret;
- nvme_tcp_advance_req(req, ret);
if (queue->data_digest)
nvme_tcp_ddgst_update(queue->snd_hash, page,
offset, ret);
- /* fully successful last write*/
+ /*
+ * update the request iterator except for the last payload send
+ * in the request where we don't want to modify it as we may
+ * compete with the RX path completing the request.
+ */
+ if (req_data_sent + ret < req_data_len)
+ nvme_tcp_advance_req(req, ret);
+
+ /* fully successful last send in current PDU */
if (last && ret == len) {
if (queue->data_digest) {
nvme_tcp_ddgst_final(queue->snd_hash,
@@ -895,11 +976,16 @@
struct nvme_tcp_queue *queue = req->queue;
struct nvme_tcp_cmd_pdu *pdu = req->pdu;
bool inline_data = nvme_tcp_has_inline_data(req);
- int flags = MSG_DONTWAIT | (inline_data ? MSG_MORE : MSG_EOR);
u8 hdgst = nvme_tcp_hdgst_len(queue);
int len = sizeof(*pdu) + hdgst - req->offset;
+ int flags = MSG_DONTWAIT;
int ret;
+ if (inline_data || nvme_tcp_queue_more(queue))
+ flags |= MSG_MORE | MSG_SENDPAGE_NOTLAST;
+ else
+ flags |= MSG_EOR;
+
if (queue->hdr_digest && !req->offset)
nvme_tcp_hdgst(queue->snd_hash, pdu, sizeof(*pdu));
@@ -938,7 +1024,7 @@
ret = kernel_sendpage(queue->sock, virt_to_page(pdu),
offset_in_page(pdu) + req->offset, len,
- MSG_DONTWAIT | MSG_MORE);
+ MSG_DONTWAIT | MSG_MORE | MSG_SENDPAGE_NOTLAST);
if (unlikely(ret <= 0))
return ret;
@@ -959,18 +1045,24 @@
static int nvme_tcp_try_send_ddgst(struct nvme_tcp_request *req)
{
struct nvme_tcp_queue *queue = req->queue;
+ size_t offset = req->offset;
int ret;
- struct msghdr msg = { .msg_flags = MSG_DONTWAIT | MSG_EOR };
+ struct msghdr msg = { .msg_flags = MSG_DONTWAIT };
struct kvec iov = {
- .iov_base = &req->ddgst + req->offset,
+ .iov_base = (u8 *)&req->ddgst + req->offset,
.iov_len = NVME_TCP_DIGEST_LENGTH - req->offset
};
+ if (nvme_tcp_queue_more(queue))
+ msg.msg_flags |= MSG_MORE;
+ else
+ msg.msg_flags |= MSG_EOR;
+
ret = kernel_sendmsg(queue->sock, &msg, &iov, 1, iov.iov_len);
if (unlikely(ret <= 0))
return ret;
- if (req->offset + ret == NVME_TCP_DIGEST_LENGTH) {
+ if (offset + ret == NVME_TCP_DIGEST_LENGTH) {
nvme_tcp_done_send_req(queue);
return 1;
}
@@ -1014,8 +1106,15 @@
if (req->state == NVME_TCP_SEND_DDGST)
ret = nvme_tcp_try_send_ddgst(req);
done:
- if (ret == -EAGAIN)
+ if (ret == -EAGAIN) {
ret = 0;
+ } else if (ret < 0) {
+ dev_err(queue->ctrl->ctrl.device,
+ "failed to send request %d\n", ret);
+ if (ret != -EPIPE && ret != -ECONNRESET)
+ nvme_tcp_fail_request(queue->request);
+ nvme_tcp_done_send_req(queue);
+ }
return ret;
}
@@ -1045,26 +1144,20 @@
bool pending = false;
int result;
- result = nvme_tcp_try_send(queue);
- if (result > 0) {
- pending = true;
- } else if (unlikely(result < 0)) {
- dev_err(queue->ctrl->ctrl.device,
- "failed to send request %d\n", result);
-
- /*
- * Fail the request unless peer closed the connection,
- * in which case error recovery flow will complete all.
- */
- if ((result != -EPIPE) && (result != -ECONNRESET))
- nvme_tcp_fail_request(queue->request);
- nvme_tcp_done_send_req(queue);
- return;
+ if (mutex_trylock(&queue->send_mutex)) {
+ result = nvme_tcp_try_send(queue);
+ mutex_unlock(&queue->send_mutex);
+ if (result > 0)
+ pending = true;
+ else if (unlikely(result < 0))
+ break;
}
result = nvme_tcp_try_recv(queue);
if (result > 0)
pending = true;
+ else if (unlikely(result < 0))
+ return;
if (!pending)
return;
@@ -1145,6 +1238,7 @@
sock_release(queue->sock);
kfree(queue->pdu);
+ mutex_destroy(&queue->queue_lock);
}
static int nvme_tcp_init_connection(struct nvme_tcp_queue *queue)
@@ -1245,17 +1339,72 @@
return ret;
}
+static bool nvme_tcp_admin_queue(struct nvme_tcp_queue *queue)
+{
+ return nvme_tcp_queue_id(queue) == 0;
+}
+
+static bool nvme_tcp_default_queue(struct nvme_tcp_queue *queue)
+{
+ struct nvme_tcp_ctrl *ctrl = queue->ctrl;
+ int qid = nvme_tcp_queue_id(queue);
+
+ return !nvme_tcp_admin_queue(queue) &&
+ qid < 1 + ctrl->io_queues[HCTX_TYPE_DEFAULT];
+}
+
+static bool nvme_tcp_read_queue(struct nvme_tcp_queue *queue)
+{
+ struct nvme_tcp_ctrl *ctrl = queue->ctrl;
+ int qid = nvme_tcp_queue_id(queue);
+
+ return !nvme_tcp_admin_queue(queue) &&
+ !nvme_tcp_default_queue(queue) &&
+ qid < 1 + ctrl->io_queues[HCTX_TYPE_DEFAULT] +
+ ctrl->io_queues[HCTX_TYPE_READ];
+}
+
+static bool nvme_tcp_poll_queue(struct nvme_tcp_queue *queue)
+{
+ struct nvme_tcp_ctrl *ctrl = queue->ctrl;
+ int qid = nvme_tcp_queue_id(queue);
+
+ return !nvme_tcp_admin_queue(queue) &&
+ !nvme_tcp_default_queue(queue) &&
+ !nvme_tcp_read_queue(queue) &&
+ qid < 1 + ctrl->io_queues[HCTX_TYPE_DEFAULT] +
+ ctrl->io_queues[HCTX_TYPE_READ] +
+ ctrl->io_queues[HCTX_TYPE_POLL];
+}
+
+static void nvme_tcp_set_queue_io_cpu(struct nvme_tcp_queue *queue)
+{
+ struct nvme_tcp_ctrl *ctrl = queue->ctrl;
+ int qid = nvme_tcp_queue_id(queue);
+ int n = 0;
+
+ if (nvme_tcp_default_queue(queue))
+ n = qid - 1;
+ else if (nvme_tcp_read_queue(queue))
+ n = qid - ctrl->io_queues[HCTX_TYPE_DEFAULT] - 1;
+ else if (nvme_tcp_poll_queue(queue))
+ n = qid - ctrl->io_queues[HCTX_TYPE_DEFAULT] -
+ ctrl->io_queues[HCTX_TYPE_READ] - 1;
+ queue->io_cpu = cpumask_next_wrap(n - 1, cpu_online_mask, -1, false);
+}
+
static int nvme_tcp_alloc_queue(struct nvme_ctrl *nctrl,
int qid, size_t queue_size)
{
struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(nctrl);
struct nvme_tcp_queue *queue = &ctrl->queues[qid];
- struct linger sol = { .l_onoff = 1, .l_linger = 0 };
- int ret, opt, rcv_pdu_size, n;
+ int ret, rcv_pdu_size;
+ mutex_init(&queue->queue_lock);
queue->ctrl = ctrl;
+ init_llist_head(&queue->req_list);
INIT_LIST_HEAD(&queue->send_list);
- spin_lock_init(&queue->lock);
+ mutex_init(&queue->send_mutex);
INIT_WORK(&queue->io_work, nvme_tcp_io_work);
queue->queue_size = queue_size;
@@ -1270,63 +1419,34 @@
if (ret) {
dev_err(nctrl->device,
"failed to create socket: %d\n", ret);
- return ret;
+ goto err_destroy_mutex;
}
/* Single syn retry */
- opt = 1;
- ret = kernel_setsockopt(queue->sock, IPPROTO_TCP, TCP_SYNCNT,
- (char *)&opt, sizeof(opt));
- if (ret) {
- dev_err(nctrl->device,
- "failed to set TCP_SYNCNT sock opt %d\n", ret);
- goto err_sock;
- }
+ tcp_sock_set_syncnt(queue->sock->sk, 1);
/* Set TCP no delay */
- opt = 1;
- ret = kernel_setsockopt(queue->sock, IPPROTO_TCP,
- TCP_NODELAY, (char *)&opt, sizeof(opt));
- if (ret) {
- dev_err(nctrl->device,
- "failed to set TCP_NODELAY sock opt %d\n", ret);
- goto err_sock;
- }
+ tcp_sock_set_nodelay(queue->sock->sk);
/*
* Cleanup whatever is sitting in the TCP transmit queue on socket
* close. This is done to prevent stale data from being sent should
* the network connection be restored before TCP times out.
*/
- ret = kernel_setsockopt(queue->sock, SOL_SOCKET, SO_LINGER,
- (char *)&sol, sizeof(sol));
- if (ret) {
- dev_err(nctrl->device,
- "failed to set SO_LINGER sock opt %d\n", ret);
- goto err_sock;
- }
+ sock_no_linger(queue->sock->sk);
+
+ if (so_priority > 0)
+ sock_set_priority(queue->sock->sk, so_priority);
/* Set socket type of service */
- if (nctrl->opts->tos >= 0) {
- opt = nctrl->opts->tos;
- ret = kernel_setsockopt(queue->sock, SOL_IP, IP_TOS,
- (char *)&opt, sizeof(opt));
- if (ret) {
- dev_err(nctrl->device,
- "failed to set IP_TOS sock opt %d\n", ret);
- goto err_sock;
- }
- }
+ if (nctrl->opts->tos >= 0)
+ ip_sock_set_tos(queue->sock->sk, nctrl->opts->tos);
/* Set 10 seconds timeout for icresp recvmsg */
queue->sock->sk->sk_rcvtimeo = 10 * HZ;
queue->sock->sk->sk_allocation = GFP_ATOMIC;
- if (!qid)
- n = 0;
- else
- n = (qid - 1) % num_online_cpus();
- queue->io_cpu = cpumask_next_wrap(n - 1, cpu_online_mask, -1, false);
+ nvme_tcp_set_queue_io_cpu(queue);
queue->request = NULL;
queue->data_remaining = 0;
queue->ddgst_remaining = 0;
@@ -1408,6 +1528,8 @@
err_sock:
sock_release(queue->sock);
queue->sock = NULL;
+err_destroy_mutex:
+ mutex_destroy(&queue->queue_lock);
return ret;
}
@@ -1435,9 +1557,10 @@
struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(nctrl);
struct nvme_tcp_queue *queue = &ctrl->queues[qid];
- if (!test_and_clear_bit(NVME_TCP_Q_LIVE, &queue->flags))
- return;
- __nvme_tcp_stop_queue(queue);
+ mutex_lock(&queue->queue_lock);
+ if (test_and_clear_bit(NVME_TCP_Q_LIVE, &queue->flags))
+ __nvme_tcp_stop_queue(queue);
+ mutex_unlock(&queue->queue_lock);
}
static int nvme_tcp_start_queue(struct nvme_ctrl *nctrl, int idx)
@@ -1474,7 +1597,8 @@
set->ops = &nvme_tcp_admin_mq_ops;
set->queue_depth = NVME_AQ_MQ_TAG_DEPTH;
set->reserved_tags = 2; /* connect + keep-alive */
- set->numa_node = NUMA_NO_NODE;
+ set->numa_node = nctrl->numa_node;
+ set->flags = BLK_MQ_F_BLOCKING;
set->cmd_size = sizeof(struct nvme_tcp_request);
set->driver_data = ctrl;
set->nr_hw_queues = 1;
@@ -1485,8 +1609,8 @@
set->ops = &nvme_tcp_mq_ops;
set->queue_depth = nctrl->sqsize + 1;
set->reserved_tags = 1; /* fabric connect */
- set->numa_node = NUMA_NO_NODE;
- set->flags = BLK_MQ_F_SHOULD_MERGE;
+ set->numa_node = nctrl->numa_node;
+ set->flags = BLK_MQ_F_SHOULD_MERGE | BLK_MQ_F_BLOCKING;
set->cmd_size = sizeof(struct nvme_tcp_request);
set->driver_data = ctrl;
set->nr_hw_queues = nctrl->queue_count - 1;
@@ -1890,8 +2014,14 @@
}
if (!nvme_change_ctrl_state(ctrl, NVME_CTRL_LIVE)) {
- /* state change failure is ok if we're in DELETING state */
- WARN_ON_ONCE(ctrl->state != NVME_CTRL_DELETING);
+ /*
+ * state change failure is ok if we started ctrl delete,
+ * unless we're during creation of a new controller to
+ * avoid races with teardown flow.
+ */
+ WARN_ON_ONCE(ctrl->state != NVME_CTRL_DELETING &&
+ ctrl->state != NVME_CTRL_DELETING_NOIO);
+ WARN_ON_ONCE(new);
ret = -EINVAL;
goto destroy_io;
}
@@ -1947,6 +2077,7 @@
struct nvme_ctrl *ctrl = &tcp_ctrl->ctrl;
nvme_stop_keep_alive(ctrl);
+ flush_work(&ctrl->async_event_work);
nvme_tcp_teardown_io_queues(ctrl, false);
/* unquiesce to fail fast pending requests */
nvme_start_queues(ctrl);
@@ -1954,8 +2085,9 @@
blk_mq_unquiesce_queue(ctrl->admin_q);
if (!nvme_change_ctrl_state(ctrl, NVME_CTRL_CONNECTING)) {
- /* state change failure is ok if we're in DELETING state */
- WARN_ON_ONCE(ctrl->state != NVME_CTRL_DELETING);
+ /* state change failure is ok if we started ctrl delete */
+ WARN_ON_ONCE(ctrl->state != NVME_CTRL_DELETING &&
+ ctrl->state != NVME_CTRL_DELETING_NOIO);
return;
}
@@ -1990,8 +2122,9 @@
nvme_tcp_teardown_ctrl(ctrl, false);
if (!nvme_change_ctrl_state(ctrl, NVME_CTRL_CONNECTING)) {
- /* state change failure is ok if we're in DELETING state */
- WARN_ON_ONCE(ctrl->state != NVME_CTRL_DELETING);
+ /* state change failure is ok if we started ctrl delete */
+ WARN_ON_ONCE(ctrl->state != NVME_CTRL_DELETING &&
+ ctrl->state != NVME_CTRL_DELETING_NOIO);
return;
}
@@ -2078,7 +2211,7 @@
ctrl->async_req.curr_bio = NULL;
ctrl->async_req.data_len = 0;
- nvme_tcp_queue_request(&ctrl->async_req);
+ nvme_tcp_queue_request(&ctrl->async_req, true, true);
}
static void nvme_tcp_complete_timed_out(struct request *rq)
@@ -2202,6 +2335,14 @@
return 0;
}
+static void nvme_tcp_commit_rqs(struct blk_mq_hw_ctx *hctx)
+{
+ struct nvme_tcp_queue *queue = hctx->driver_data;
+
+ if (!llist_empty(&queue->req_list))
+ queue_work_on(queue->io_cpu, nvme_tcp_wq, &queue->io_work);
+}
+
static blk_status_t nvme_tcp_queue_rq(struct blk_mq_hw_ctx *hctx,
const struct blk_mq_queue_data *bd)
{
@@ -2221,7 +2362,7 @@
blk_mq_start_request(rq);
- nvme_tcp_queue_request(req);
+ nvme_tcp_queue_request(req, true, bd->last);
return BLK_STS_OK;
}
@@ -2276,14 +2417,20 @@
struct nvme_tcp_queue *queue = hctx->driver_data;
struct sock *sk = queue->sock->sk;
+ if (!test_bit(NVME_TCP_Q_LIVE, &queue->flags))
+ return 0;
+
+ set_bit(NVME_TCP_Q_POLLING, &queue->flags);
if (sk_can_busy_loop(sk) && skb_queue_empty_lockless(&sk->sk_receive_queue))
sk_busy_loop(sk, true);
nvme_tcp_try_recv(queue);
+ clear_bit(NVME_TCP_Q_POLLING, &queue->flags);
return queue->nr_cqe;
}
-static struct blk_mq_ops nvme_tcp_mq_ops = {
+static const struct blk_mq_ops nvme_tcp_mq_ops = {
.queue_rq = nvme_tcp_queue_rq,
+ .commit_rqs = nvme_tcp_commit_rqs,
.complete = nvme_complete_rq,
.init_request = nvme_tcp_init_request,
.exit_request = nvme_tcp_exit_request,
@@ -2293,7 +2440,7 @@
.poll = nvme_tcp_poll,
};
-static struct blk_mq_ops nvme_tcp_admin_mq_ops = {
+static const struct blk_mq_ops nvme_tcp_admin_mq_ops = {
.queue_rq = nvme_tcp_queue_rq,
.complete = nvme_complete_rq,
.init_request = nvme_tcp_init_request,
@@ -2420,7 +2567,6 @@
out_uninit_ctrl:
nvme_uninit_ctrl(&ctrl->ctrl);
nvme_put_ctrl(&ctrl->ctrl);
- nvme_put_ctrl(&ctrl->ctrl);
if (ret > 0)
ret = -EIO;
return ERR_PTR(ret);
diff --git a/drivers/nvme/host/zns.c b/drivers/nvme/host/zns.c
new file mode 100644
index 0000000..67e87e9
--- /dev/null
+++ b/drivers/nvme/host/zns.c
@@ -0,0 +1,257 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2020 Western Digital Corporation or its affiliates.
+ */
+
+#include <linux/blkdev.h>
+#include <linux/vmalloc.h>
+#include "nvme.h"
+
+int nvme_revalidate_zones(struct nvme_ns *ns)
+{
+ struct request_queue *q = ns->queue;
+ int ret;
+
+ ret = blk_revalidate_disk_zones(ns->disk, NULL);
+ if (!ret)
+ blk_queue_max_zone_append_sectors(q, ns->ctrl->max_zone_append);
+ return ret;
+}
+
+static int nvme_set_max_append(struct nvme_ctrl *ctrl)
+{
+ struct nvme_command c = { };
+ struct nvme_id_ctrl_zns *id;
+ int status;
+
+ id = kzalloc(sizeof(*id), GFP_KERNEL);
+ if (!id)
+ return -ENOMEM;
+
+ c.identify.opcode = nvme_admin_identify;
+ c.identify.cns = NVME_ID_CNS_CS_CTRL;
+ c.identify.csi = NVME_CSI_ZNS;
+
+ status = nvme_submit_sync_cmd(ctrl->admin_q, &c, id, sizeof(*id));
+ if (status) {
+ kfree(id);
+ return status;
+ }
+
+ if (id->zasl)
+ ctrl->max_zone_append = 1 << (id->zasl + 3);
+ else
+ ctrl->max_zone_append = ctrl->max_hw_sectors;
+ kfree(id);
+ return 0;
+}
+
+int nvme_update_zone_info(struct nvme_ns *ns, unsigned lbaf)
+{
+ struct nvme_effects_log *log = ns->head->effects;
+ struct request_queue *q = ns->queue;
+ struct nvme_command c = { };
+ struct nvme_id_ns_zns *id;
+ int status;
+
+ /* Driver requires zone append support */
+ if (!(le32_to_cpu(log->iocs[nvme_cmd_zone_append]) &
+ NVME_CMD_EFFECTS_CSUPP)) {
+ dev_warn(ns->ctrl->device,
+ "append not supported for zoned namespace:%d\n",
+ ns->head->ns_id);
+ return -EINVAL;
+ }
+
+ /* Lazily query controller append limit for the first zoned namespace */
+ if (!ns->ctrl->max_zone_append) {
+ status = nvme_set_max_append(ns->ctrl);
+ if (status)
+ return status;
+ }
+
+ id = kzalloc(sizeof(*id), GFP_KERNEL);
+ if (!id)
+ return -ENOMEM;
+
+ c.identify.opcode = nvme_admin_identify;
+ c.identify.nsid = cpu_to_le32(ns->head->ns_id);
+ c.identify.cns = NVME_ID_CNS_CS_NS;
+ c.identify.csi = NVME_CSI_ZNS;
+
+ status = nvme_submit_sync_cmd(ns->ctrl->admin_q, &c, id, sizeof(*id));
+ if (status)
+ goto free_data;
+
+ /*
+ * We currently do not handle devices requiring any of the zoned
+ * operation characteristics.
+ */
+ if (id->zoc) {
+ dev_warn(ns->ctrl->device,
+ "zone operations:%x not supported for namespace:%u\n",
+ le16_to_cpu(id->zoc), ns->head->ns_id);
+ status = -EINVAL;
+ goto free_data;
+ }
+
+ ns->zsze = nvme_lba_to_sect(ns, le64_to_cpu(id->lbafe[lbaf].zsze));
+ if (!is_power_of_2(ns->zsze)) {
+ dev_warn(ns->ctrl->device,
+ "invalid zone size:%llu for namespace:%u\n",
+ ns->zsze, ns->head->ns_id);
+ status = -EINVAL;
+ goto free_data;
+ }
+
+ q->limits.zoned = BLK_ZONED_HM;
+ blk_queue_flag_set(QUEUE_FLAG_ZONE_RESETALL, q);
+ blk_queue_max_open_zones(q, le32_to_cpu(id->mor) + 1);
+ blk_queue_max_active_zones(q, le32_to_cpu(id->mar) + 1);
+free_data:
+ kfree(id);
+ return status;
+}
+
+static void *nvme_zns_alloc_report_buffer(struct nvme_ns *ns,
+ unsigned int nr_zones, size_t *buflen)
+{
+ struct request_queue *q = ns->disk->queue;
+ size_t bufsize;
+ void *buf;
+
+ const size_t min_bufsize = sizeof(struct nvme_zone_report) +
+ sizeof(struct nvme_zone_descriptor);
+
+ nr_zones = min_t(unsigned int, nr_zones,
+ get_capacity(ns->disk) >> ilog2(ns->zsze));
+
+ bufsize = sizeof(struct nvme_zone_report) +
+ nr_zones * sizeof(struct nvme_zone_descriptor);
+ bufsize = min_t(size_t, bufsize,
+ queue_max_hw_sectors(q) << SECTOR_SHIFT);
+ bufsize = min_t(size_t, bufsize, queue_max_segments(q) << PAGE_SHIFT);
+
+ while (bufsize >= min_bufsize) {
+ buf = __vmalloc(bufsize, GFP_KERNEL | __GFP_NORETRY);
+ if (buf) {
+ *buflen = bufsize;
+ return buf;
+ }
+ bufsize >>= 1;
+ }
+ return NULL;
+}
+
+static int nvme_zone_parse_entry(struct nvme_ns *ns,
+ struct nvme_zone_descriptor *entry,
+ unsigned int idx, report_zones_cb cb,
+ void *data)
+{
+ struct blk_zone zone = { };
+
+ if ((entry->zt & 0xf) != NVME_ZONE_TYPE_SEQWRITE_REQ) {
+ dev_err(ns->ctrl->device, "invalid zone type %#x\n",
+ entry->zt);
+ return -EINVAL;
+ }
+
+ zone.type = BLK_ZONE_TYPE_SEQWRITE_REQ;
+ zone.cond = entry->zs >> 4;
+ zone.len = ns->zsze;
+ zone.capacity = nvme_lba_to_sect(ns, le64_to_cpu(entry->zcap));
+ zone.start = nvme_lba_to_sect(ns, le64_to_cpu(entry->zslba));
+ zone.wp = nvme_lba_to_sect(ns, le64_to_cpu(entry->wp));
+
+ return cb(&zone, idx, data);
+}
+
+static int nvme_ns_report_zones(struct nvme_ns *ns, sector_t sector,
+ unsigned int nr_zones, report_zones_cb cb, void *data)
+{
+ struct nvme_zone_report *report;
+ struct nvme_command c = { };
+ int ret, zone_idx = 0;
+ unsigned int nz, i;
+ size_t buflen;
+
+ report = nvme_zns_alloc_report_buffer(ns, nr_zones, &buflen);
+ if (!report)
+ return -ENOMEM;
+
+ c.zmr.opcode = nvme_cmd_zone_mgmt_recv;
+ c.zmr.nsid = cpu_to_le32(ns->head->ns_id);
+ c.zmr.numd = cpu_to_le32(nvme_bytes_to_numd(buflen));
+ c.zmr.zra = NVME_ZRA_ZONE_REPORT;
+ c.zmr.zrasf = NVME_ZRASF_ZONE_REPORT_ALL;
+ c.zmr.pr = NVME_REPORT_ZONE_PARTIAL;
+
+ sector &= ~(ns->zsze - 1);
+ while (zone_idx < nr_zones && sector < get_capacity(ns->disk)) {
+ memset(report, 0, buflen);
+
+ c.zmr.slba = cpu_to_le64(nvme_sect_to_lba(ns, sector));
+ ret = nvme_submit_sync_cmd(ns->queue, &c, report, buflen);
+ if (ret) {
+ if (ret > 0)
+ ret = -EIO;
+ goto out_free;
+ }
+
+ nz = min((unsigned int)le64_to_cpu(report->nr_zones), nr_zones);
+ if (!nz)
+ break;
+
+ for (i = 0; i < nz && zone_idx < nr_zones; i++) {
+ ret = nvme_zone_parse_entry(ns, &report->entries[i],
+ zone_idx, cb, data);
+ if (ret)
+ goto out_free;
+ zone_idx++;
+ }
+
+ sector += ns->zsze * nz;
+ }
+
+ if (zone_idx > 0)
+ ret = zone_idx;
+ else
+ ret = -EINVAL;
+out_free:
+ kvfree(report);
+ return ret;
+}
+
+int nvme_report_zones(struct gendisk *disk, sector_t sector,
+ unsigned int nr_zones, report_zones_cb cb, void *data)
+{
+ struct nvme_ns_head *head = NULL;
+ struct nvme_ns *ns;
+ int srcu_idx, ret;
+
+ ns = nvme_get_ns_from_disk(disk, &head, &srcu_idx);
+ if (unlikely(!ns))
+ return -EWOULDBLOCK;
+
+ if (ns->head->ids.csi == NVME_CSI_ZNS)
+ ret = nvme_ns_report_zones(ns, sector, nr_zones, cb, data);
+ else
+ ret = -EINVAL;
+ nvme_put_ns_from_disk(head, srcu_idx);
+
+ return ret;
+}
+
+blk_status_t nvme_setup_zone_mgmt_send(struct nvme_ns *ns, struct request *req,
+ struct nvme_command *c, enum nvme_zone_mgmt_action action)
+{
+ c->zms.opcode = nvme_cmd_zone_mgmt_send;
+ c->zms.nsid = cpu_to_le32(ns->head->ns_id);
+ c->zms.slba = cpu_to_le64(nvme_sect_to_lba(ns, blk_rq_pos(req)));
+ c->zms.zsa = action;
+
+ if (req_op(req) == REQ_OP_ZONE_RESET_ALL)
+ c->zms.select_all = 1;
+
+ return BLK_STS_OK;
+}