summaryrefslogtreecommitdiff
path: root/drivers/nvme
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2022-05-24 00:04:14 +0300
committerLinus Torvalds <torvalds@linux-foundation.org>2022-05-24 00:04:14 +0300
commit5dc921868c507c1f0835932d3f255cf1b7415618 (patch)
tree3b458e2a75c36722fdb71f47d73e6f27fceae02a /drivers/nvme
parent115cd47132d71bd7e4aa1093e15d861a59e73a94 (diff)
parent537b9f2bf60f4bbd8ab89cea16aaab70f0c1560d (diff)
downloadlinux-5dc921868c507c1f0835932d3f255cf1b7415618.tar.xz
Merge tag 'for-5.19/drivers-2022-05-22' of git://git.kernel.dk/linux-block
Pull block driver updates from Jens Axboe: "Here are the driver updates queued up for 5.19. This contains: - NVMe pull requests via Christoph: - tighten the PCI presence check (Stefan Roese) - fix a potential NULL pointer dereference in an error path (Kyle Miller Smith) - fix interpretation of the DMRSL field (Tom Yan) - relax the data transfer alignment (Keith Busch) - verbose error logging improvements (Max Gurtovoy, Chaitanya Kulkarni) - misc cleanups (Chaitanya Kulkarni, Christoph) - set non-mdts limits in nvme_scan_work (Chaitanya Kulkarni) - add support for TP4084 - Time-to-Ready Enhancements (Christoph) - MD pull request via Song: - Improve annotation in raid5 code, by Logan Gunthorpe - Support MD_BROKEN flag in raid-1/5/10, by Mariusz Tkaczyk - Other small fixes/cleanups - null_blk series making the configfs side much saner (Damien) - Various minor drbd cleanups and fixes (Haowen, Uladzislau, Jiapeng, Arnd, Cai) - Avoid using the system workqueue (and hence flushing it) in rnbd (Jack) - Avoid using the system workqueue (and hence flushing it) in aoe (Tetsuo) - Series fixing discard_alignment issues in drivers (Christoph) - Small series fixing drivers poking at disk->part0 for openers information (Christoph) - Series fixing deadlocks in loop (Christoph, Tetsuo) - Remove loop.h and add SPDX headers (Christoph) - Various fixes and cleanups (Julia, Xie, Yu)" * tag 'for-5.19/drivers-2022-05-22' of git://git.kernel.dk/linux-block: (72 commits) mtip32xx: fix typo in comment nvme: set non-mdts limits in nvme_scan_work nvme: add support for TP4084 - Time-to-Ready Enhancements nvme: split the enum used for various register constants nbd: Fix hung on disconnect request if socket is closed before nvme-fabrics: add a request timeout helper nvme-pci: harden drive presence detect in nvme_dev_disable() nvme-pci: fix a NULL pointer dereference in nvme_alloc_admin_tags nvme: mark internal passthru request RQF_QUIET nvme: remove unneeded include from constants file nvme: add missing status values to verbose logging nvme: set dma alignment to dword nvme: fix interpretation of DMRSL loop: remove most the top-of-file boilerplate comment from the UAPI header loop: remove most the top-of-file boilerplate comment loop: add a SPDX header loop: remove loop.h block: null_blk: Improve device creation with configfs block: null_blk: Cleanup messages block: null_blk: Cleanup device creation and deletion ...
Diffstat (limited to 'drivers/nvme')
-rw-r--r--drivers/nvme/host/constants.c5
-rw-r--r--drivers/nvme/host/core.c105
-rw-r--r--drivers/nvme/host/fabrics.h8
-rw-r--r--drivers/nvme/host/nvme.h1
-rw-r--r--drivers/nvme/host/pci.c5
-rw-r--r--drivers/nvme/host/rdma.c5
-rw-r--r--drivers/nvme/host/tcp.c5
7 files changed, 110 insertions, 24 deletions
diff --git a/drivers/nvme/host/constants.c b/drivers/nvme/host/constants.c
index 7d49eb34b348..4910543f00ff 100644
--- a/drivers/nvme/host/constants.c
+++ b/drivers/nvme/host/constants.c
@@ -4,7 +4,6 @@
* Copyright (c) 2022, Oracle and/or its affiliates
*/
-#include <linux/blkdev.h>
#include "nvme.h"
#ifdef CONFIG_NVME_VERBOSE_ERRORS
@@ -92,6 +91,7 @@ static const char * const nvme_statuses[] = {
[NVME_SC_NS_WRITE_PROTECTED] = "Namespace is Write Protected",
[NVME_SC_CMD_INTERRUPTED] = "Command Interrupted",
[NVME_SC_TRANSIENT_TR_ERR] = "Transient Transport Error",
+ [NVME_SC_ADMIN_COMMAND_MEDIA_NOT_READY] = "Admin Command Media Not Ready",
[NVME_SC_INVALID_IO_CMD_SET] = "Invalid IO Command Set",
[NVME_SC_LBA_RANGE] = "LBA Out of Range",
[NVME_SC_CAP_EXCEEDED] = "Capacity Exceeded",
@@ -155,10 +155,13 @@ static const char * const nvme_statuses[] = {
[NVME_SC_COMPARE_FAILED] = "Compare Failure",
[NVME_SC_ACCESS_DENIED] = "Access Denied",
[NVME_SC_UNWRITTEN_BLOCK] = "Deallocated or Unwritten Logical Block",
+ [NVME_SC_INTERNAL_PATH_ERROR] = "Internal Pathing Error",
[NVME_SC_ANA_PERSISTENT_LOSS] = "Asymmetric Access Persistent Loss",
[NVME_SC_ANA_INACCESSIBLE] = "Asymmetric Access Inaccessible",
[NVME_SC_ANA_TRANSITION] = "Asymmetric Access Transition",
+ [NVME_SC_CTRL_PATH_ERROR] = "Controller Pathing Error",
[NVME_SC_HOST_PATH_ERROR] = "Host Pathing Error",
+ [NVME_SC_HOST_ABORTED_CMD] = "Host Aborted Command",
};
const unsigned char *nvme_get_error_status_str(u16 status)
diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index c951d7b2cf48..72f7c955c707 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -1207,6 +1207,7 @@ static void nvme_keep_alive_work(struct work_struct *work)
rq->timeout = ctrl->kato * HZ;
rq->end_io_data = ctrl;
+ rq->rq_flags |= RQF_QUIET;
blk_execute_rq_nowait(rq, false, nvme_keep_alive_end_io);
}
@@ -1426,6 +1427,32 @@ out_free_id:
return error;
}
+static int nvme_identify_ns_cs_indep(struct nvme_ctrl *ctrl, unsigned nsid,
+ struct nvme_id_ns_cs_indep **id)
+{
+ struct nvme_command c = {
+ .identify.opcode = nvme_admin_identify,
+ .identify.nsid = cpu_to_le32(nsid),
+ .identify.cns = NVME_ID_CNS_NS_CS_INDEP,
+ };
+ int ret;
+
+ *id = kmalloc(sizeof(**id), GFP_KERNEL);
+ if (!*id)
+ return -ENOMEM;
+
+ ret = nvme_submit_sync_cmd(ctrl->admin_q, &c, *id, sizeof(**id));
+ if (ret) {
+ dev_warn(ctrl->device,
+ "Identify namespace (CS independent) failed (%d)\n",
+ ret);
+ kfree(*id);
+ return ret;
+ }
+
+ return 0;
+}
+
static int nvme_features(struct nvme_ctrl *dev, u8 op, unsigned int fid,
unsigned int dword11, void *buffer, size_t buflen, u32 *result)
{
@@ -1628,13 +1655,15 @@ static void nvme_config_discard(struct gendisk *disk, struct nvme_ns *ns)
BUILD_BUG_ON(PAGE_SIZE / sizeof(struct nvme_dsm_range) <
NVME_DSM_MAX_RANGES);
- queue->limits.discard_alignment = 0;
queue->limits.discard_granularity = size;
/* If discard is already enabled, don't reset queue limits */
if (queue->limits.max_discard_sectors)
return;
+ if (ctrl->dmrsl && ctrl->dmrsl <= nvme_sect_to_lba(ns, UINT_MAX))
+ ctrl->max_discard_sectors = nvme_lba_to_sect(ns, ctrl->dmrsl);
+
blk_queue_max_discard_sectors(queue, ctrl->max_discard_sectors);
blk_queue_max_discard_segments(queue, ctrl->max_discard_segments);
@@ -1771,7 +1800,7 @@ static void nvme_set_queue_limits(struct nvme_ctrl *ctrl,
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_dma_alignment(q, 3);
blk_queue_write_cache(q, vwc, vwc);
}
@@ -2100,10 +2129,9 @@ static const struct block_device_operations nvme_bdev_ops = {
.pr_ops = &nvme_pr_ops,
};
-static int nvme_wait_ready(struct nvme_ctrl *ctrl, u64 cap, bool enabled)
+static int nvme_wait_ready(struct nvme_ctrl *ctrl, u32 timeout, bool enabled)
{
- unsigned long timeout =
- ((NVME_CAP_TIMEOUT(cap) + 1) * HZ / 2) + jiffies;
+ unsigned long timeout_jiffies = ((timeout + 1) * HZ / 2) + jiffies;
u32 csts, bit = enabled ? NVME_CSTS_RDY : 0;
int ret;
@@ -2116,7 +2144,7 @@ static int nvme_wait_ready(struct nvme_ctrl *ctrl, u64 cap, bool enabled)
usleep_range(1000, 2000);
if (fatal_signal_pending(current))
return -EINTR;
- if (time_after(jiffies, timeout)) {
+ if (time_after(jiffies, timeout_jiffies)) {
dev_err(ctrl->device,
"Device not ready; aborting %s, CSTS=0x%x\n",
enabled ? "initialisation" : "reset", csts);
@@ -2147,13 +2175,14 @@ int nvme_disable_ctrl(struct nvme_ctrl *ctrl)
if (ctrl->quirks & NVME_QUIRK_DELAY_BEFORE_CHK_RDY)
msleep(NVME_QUIRK_DELAY_AMOUNT);
- return nvme_wait_ready(ctrl, ctrl->cap, false);
+ return nvme_wait_ready(ctrl, NVME_CAP_TIMEOUT(ctrl->cap), false);
}
EXPORT_SYMBOL_GPL(nvme_disable_ctrl);
int nvme_enable_ctrl(struct nvme_ctrl *ctrl)
{
unsigned dev_page_min;
+ u32 timeout;
int ret;
ret = ctrl->ops->reg_read64(ctrl, NVME_REG_CAP, &ctrl->cap);
@@ -2174,6 +2203,27 @@ int nvme_enable_ctrl(struct nvme_ctrl *ctrl)
ctrl->ctrl_config = NVME_CC_CSS_CSI;
else
ctrl->ctrl_config = NVME_CC_CSS_NVM;
+
+ if (ctrl->cap & NVME_CAP_CRMS_CRWMS) {
+ u32 crto;
+
+ ret = ctrl->ops->reg_read32(ctrl, NVME_REG_CRTO, &crto);
+ if (ret) {
+ dev_err(ctrl->device, "Reading CRTO failed (%d)\n",
+ ret);
+ return ret;
+ }
+
+ if (ctrl->cap & NVME_CAP_CRMS_CRIMS) {
+ ctrl->ctrl_config |= NVME_CC_CRIME;
+ timeout = NVME_CRTO_CRIMT(crto);
+ } else {
+ timeout = NVME_CRTO_CRWMT(crto);
+ }
+ } else {
+ timeout = NVME_CAP_TIMEOUT(ctrl->cap);
+ }
+
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;
@@ -2182,7 +2232,7 @@ int nvme_enable_ctrl(struct nvme_ctrl *ctrl)
ret = ctrl->ops->reg_write32(ctrl, NVME_REG_CC, ctrl->ctrl_config);
if (ret)
return ret;
- return nvme_wait_ready(ctrl, ctrl->cap, true);
+ return nvme_wait_ready(ctrl, timeout, true);
}
EXPORT_SYMBOL_GPL(nvme_enable_ctrl);
@@ -2894,8 +2944,7 @@ static int nvme_init_non_mdts_limits(struct nvme_ctrl *ctrl)
if (id->dmrl)
ctrl->max_discard_segments = id->dmrl;
- if (id->dmrsl)
- ctrl->max_discard_sectors = le32_to_cpu(id->dmrsl);
+ ctrl->dmrsl = le32_to_cpu(id->dmrsl);
if (id->wzsl)
ctrl->max_zeroes_sectors = nvme_mps_to_sectors(ctrl, id->wzsl);
@@ -3080,10 +3129,6 @@ int nvme_init_ctrl_finish(struct nvme_ctrl *ctrl)
if (ret)
return ret;
- ret = nvme_init_non_mdts_limits(ctrl);
- if (ret < 0)
- return ret;
-
ret = nvme_configure_apst(ctrl);
if (ret < 0)
return ret;
@@ -4092,11 +4137,26 @@ out:
static void nvme_validate_or_alloc_ns(struct nvme_ctrl *ctrl, unsigned nsid)
{
struct nvme_ns_ids ids = { };
+ struct nvme_id_ns_cs_indep *id;
struct nvme_ns *ns;
+ bool ready = true;
if (nvme_identify_ns_descs(ctrl, nsid, &ids))
return;
+ /*
+ * Check if the namespace is ready. If not ignore it, we will get an
+ * AEN once it becomes ready and restart the scan.
+ */
+ if ((ctrl->cap & NVME_CAP_CRMS_CRIMS) &&
+ !nvme_identify_ns_cs_indep(ctrl, nsid, &id)) {
+ ready = id->nstat & NVME_NSTAT_NRDY;
+ kfree(id);
+ }
+
+ if (!ready)
+ return;
+
ns = nvme_find_get_ns(ctrl, nsid);
if (ns) {
nvme_validate_ns(ns, &ids);
@@ -4239,11 +4299,26 @@ static void nvme_scan_work(struct work_struct *work)
{
struct nvme_ctrl *ctrl =
container_of(work, struct nvme_ctrl, scan_work);
+ int ret;
/* No tagset on a live ctrl means IO queues could not created */
if (ctrl->state != NVME_CTRL_LIVE || !ctrl->tagset)
return;
+ /*
+ * Identify controller limits can change at controller reset due to
+ * new firmware download, even though it is not common we cannot ignore
+ * such scenario. Controller's non-mdts limits are reported in the unit
+ * of logical blocks that is dependent on the format of attached
+ * namespace. Hence re-read the limits at the time of ns allocation.
+ */
+ ret = nvme_init_non_mdts_limits(ctrl);
+ if (ret < 0) {
+ dev_warn(ctrl->device,
+ "reading non-mdts-limits failed: %d\n", ret);
+ return;
+ }
+
if (test_and_clear_bit(NVME_AER_NOTICE_NS_CHANGED, &ctrl->events)) {
dev_info(ctrl->device, "rescanning namespaces.\n");
nvme_clear_changed_ns_log(ctrl);
@@ -4841,6 +4916,8 @@ static inline void _nvme_check_size(void)
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_cs_indep) !=
+ NVME_IDENTIFY_DATA_SIZE);
BUILD_BUG_ON(sizeof(struct nvme_id_ns_zns) != NVME_IDENTIFY_DATA_SIZE);
BUILD_BUG_ON(sizeof(struct nvme_id_ns_nvm) != NVME_IDENTIFY_DATA_SIZE);
BUILD_BUG_ON(sizeof(struct nvme_id_ctrl_zns) != NVME_IDENTIFY_DATA_SIZE);
diff --git a/drivers/nvme/host/fabrics.h b/drivers/nvme/host/fabrics.h
index 1e3a09cad961..46d6e194ac2b 100644
--- a/drivers/nvme/host/fabrics.h
+++ b/drivers/nvme/host/fabrics.h
@@ -187,6 +187,14 @@ static inline char *nvmf_ctrl_subsysnqn(struct nvme_ctrl *ctrl)
return ctrl->subsys->subnqn;
}
+static inline void nvmf_complete_timed_out_request(struct request *rq)
+{
+ if (blk_mq_request_started(rq) && !blk_mq_request_completed(rq)) {
+ nvme_req(rq)->status = NVME_SC_HOST_ABORTED_CMD;
+ blk_mq_complete_request(rq);
+ }
+}
+
int nvmf_reg_read32(struct nvme_ctrl *ctrl, u32 off, u32 *val);
int nvmf_reg_read64(struct nvme_ctrl *ctrl, u32 off, u64 *val);
int nvmf_reg_write32(struct nvme_ctrl *ctrl, u32 off, u32 val);
diff --git a/drivers/nvme/host/nvme.h b/drivers/nvme/host/nvme.h
index 26d35c557588..9b72b6ecf33c 100644
--- a/drivers/nvme/host/nvme.h
+++ b/drivers/nvme/host/nvme.h
@@ -284,6 +284,7 @@ struct nvme_ctrl {
#endif
u16 crdt[3];
u16 oncs;
+ u32 dmrsl;
u16 oacs;
u16 sqsize;
u32 max_namespaces;
diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 3aacf1c0d5a5..5a98a7de0964 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1439,6 +1439,7 @@ static enum blk_eh_timer_return nvme_timeout(struct request *req, bool reserved)
nvme_init_request(abort_req, &cmd);
abort_req->end_io_data = NULL;
+ abort_req->rq_flags |= RQF_QUIET;
blk_execute_rq_nowait(abort_req, false, abort_endio);
/*
@@ -1775,6 +1776,7 @@ static int nvme_alloc_admin_tags(struct nvme_dev *dev)
dev->ctrl.admin_q = blk_mq_init_queue(&dev->admin_tagset);
if (IS_ERR(dev->ctrl.admin_q)) {
blk_mq_free_tag_set(&dev->admin_tagset);
+ dev->ctrl.admin_q = NULL;
return -ENOMEM;
}
if (!blk_get_queue(dev->ctrl.admin_q)) {
@@ -2486,6 +2488,7 @@ static int nvme_delete_queue(struct nvme_queue *nvmeq, u8 opcode)
req->end_io_data = nvmeq;
init_completion(&nvmeq->delete_done);
+ req->rq_flags |= RQF_QUIET;
blk_execute_rq_nowait(req, false, opcode == nvme_admin_delete_cq ?
nvme_del_cq_end : nvme_del_queue_end);
return 0;
@@ -2675,7 +2678,7 @@ static void nvme_dev_disable(struct nvme_dev *dev, bool shutdown)
struct pci_dev *pdev = to_pci_dev(dev->dev);
mutex_lock(&dev->shutdown_lock);
- if (pci_is_enabled(pdev)) {
+ if (pci_device_is_present(pdev) && pci_is_enabled(pdev)) {
u32 csts = readl(dev->bar + NVME_REG_CSTS);
if (dev->ctrl.state == NVME_CTRL_LIVE ||
diff --git a/drivers/nvme/host/rdma.c b/drivers/nvme/host/rdma.c
index d9f19d901313..b87c8ae41d9b 100644
--- a/drivers/nvme/host/rdma.c
+++ b/drivers/nvme/host/rdma.c
@@ -2010,10 +2010,7 @@ static void nvme_rdma_complete_timed_out(struct request *rq)
struct nvme_rdma_queue *queue = req->queue;
nvme_rdma_stop_queue(queue);
- if (blk_mq_request_started(rq) && !blk_mq_request_completed(rq)) {
- nvme_req(rq)->status = NVME_SC_HOST_ABORTED_CMD;
- blk_mq_complete_request(rq);
- }
+ nvmf_complete_timed_out_request(rq);
}
static enum blk_eh_timer_return
diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
index ad3a2bf2f1e9..bb67538d241b 100644
--- a/drivers/nvme/host/tcp.c
+++ b/drivers/nvme/host/tcp.c
@@ -2318,10 +2318,7 @@ static void nvme_tcp_complete_timed_out(struct request *rq)
struct nvme_ctrl *ctrl = &req->queue->ctrl->ctrl;
nvme_tcp_stop_queue(ctrl, nvme_tcp_queue_id(req->queue));
- if (blk_mq_request_started(rq) && !blk_mq_request_completed(rq)) {
- nvme_req(rq)->status = NVME_SC_HOST_ABORTED_CMD;
- blk_mq_complete_request(rq);
- }
+ nvmf_complete_timed_out_request(rq);
}
static enum blk_eh_timer_return