From 0bc2e80b9be51712b19e919db5abc97a418f8292 Mon Sep 17 00:00:00 2001 From: Yi Zhang Date: Wed, 10 Apr 2024 08:57:14 +0800 Subject: nvme: fix warn output about shared namespaces without CONFIG_NVME_MULTIPATH Move the stray '.' that is currently at the end of the line after newline '\n' to before newline character which is the right position. Fixes: ce8d78616a6b ("nvme: warn about shared namespaces without CONFIG_NVME_MULTIPATH") Signed-off-by: Yi Zhang Reviewed-by: Chaitanya Kulkarni Signed-off-by: Keith Busch --- drivers/nvme/host/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index 27281a9a8951..e4bec200ebeb 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -3681,7 +3681,7 @@ static int nvme_init_ns_head(struct nvme_ns *ns, struct nvme_ns_info *info) "Found shared namespace %d, but multipathing not supported.\n", info->nsid); dev_warn_once(ctrl->device, - "Support for shared namespaces without CONFIG_NVME_MULTIPATH is deprecated and will be removed in Linux 6.0\n."); + "Support for shared namespaces without CONFIG_NVME_MULTIPATH is deprecated and will be removed in Linux 6.0.\n"); } } -- cgit v1.2.3-58-ga151 From eaf4a9b19b9961f8ca294c39c5f8984a4cf42212 Mon Sep 17 00:00:00 2001 From: Uday Shankar Date: Tue, 30 Apr 2024 15:16:24 -0600 Subject: ublk: remove segment count and size limits ublk_drv currently creates block devices with the default max_segments and max_segment_size limits of BLK_MAX_SEGMENTS (128) and BLK_MAX_SEGMENT_SIZE (65536) respectively. These defaults can artificially constrain the I/O size seen by the ublk server - for example, suppose that the ublk server has configured itself to accept I/Os up to 1M and the application is also issuing 1M sized I/Os. If the I/O buffer used by the application is backed by 4K pages, the buffer could consist of up to 1M / 4K = 256 physically discontiguous segments (even if the buffer is virtually contiguous). As such, the I/O could exceed the default max_segments limit and get split. This can cause unnecessary performance issues if the ublk server is optimized to handle 1M I/Os. The block layer's segment count/size limits exist to model hardware constraints which don't exist in ublk_drv's case, so just remove those limits for the block devices created by ublk_drv. Signed-off-by: Uday Shankar Reviewed-by: Riley Thomasson Reviewed-by: Ming Lei Link: https://lore.kernel.org/r/20240430211623.2802036-1-ushankar@purestorage.com Signed-off-by: Jens Axboe --- drivers/block/ublk_drv.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/block/ublk_drv.c b/drivers/block/ublk_drv.c index bea3d5cf8a83..374e4efa8759 100644 --- a/drivers/block/ublk_drv.c +++ b/drivers/block/ublk_drv.c @@ -2177,7 +2177,8 @@ static int ublk_ctrl_start_dev(struct ublk_device *ub, struct io_uring_cmd *cmd) .max_hw_sectors = p->max_sectors, .chunk_sectors = p->chunk_sectors, .virt_boundary_mask = p->virt_boundary_mask, - + .max_segments = USHRT_MAX, + .max_segment_size = UINT_MAX, }; struct gendisk *disk; int ret = -EINVAL; -- cgit v1.2.3-58-ga151 From 863fe60ed27f2c85172654a63c5b827e72c8b2e6 Mon Sep 17 00:00:00 2001 From: Nilay Shroff Date: Tue, 16 Apr 2024 13:49:23 +0530 Subject: nvme: find numa distance only if controller has valid numa id On system where native nvme multipath is configured and iopolicy is set to numa but the nvme controller numa node id is undefined or -1 (NUMA_NO_NODE) then avoid calculating node distance for finding optimal io path. In such case we may access numa distance table with invalid index and that may potentially refer to incorrect memory. So this patch ensures that if the nvme controller numa node id is -1 then instead of calculating node distance for finding optimal io path, we set the numa node distance of such controller to default 10 (LOCAL_DISTANCE). Link: https://lore.kernel.org/all/20240413090614.678353-1-nilay@linux.ibm.com/ Signed-off-by: Nilay Shroff Reviewed-by: Christoph Hellwig Reviewed-by: Sagi Grimberg Reviewed-by: Chaitanya Kulkarni Signed-off-by: Keith Busch --- drivers/nvme/host/multipath.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c index 5397fb428b24..d16e976ae1a4 100644 --- a/drivers/nvme/host/multipath.c +++ b/drivers/nvme/host/multipath.c @@ -247,7 +247,8 @@ static struct nvme_ns *__nvme_find_path(struct nvme_ns_head *head, int node) if (nvme_path_is_disabled(ns)) continue; - if (READ_ONCE(head->subsys->iopolicy) == NVME_IOPOLICY_NUMA) + if (ns->ctrl->numa_node != NUMA_NO_NODE && + READ_ONCE(head->subsys->iopolicy) == NVME_IOPOLICY_NUMA) distance = node_distance(node, ns->ctrl->numa_node); else distance = LOCAL_DISTANCE; -- cgit v1.2.3-58-ga151 From 46b8f9f74f6d500871985e22eb19560b21f3bc81 Mon Sep 17 00:00:00 2001 From: Maurizio Lombardi Date: Wed, 10 Apr 2024 11:48:41 +0200 Subject: nvmet-auth: return the error code to the nvmet_auth_host_hash() callers If the nvmet_auth_host_hash() function fails, the error code should be returned to its callers. Signed-off-by: Maurizio Lombardi Reviewed-by: Sagi Grimberg Reviewed-by: Chaitanya Kulkarni Signed-off-by: Keith Busch --- drivers/nvme/target/auth.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/nvme/target/auth.c b/drivers/nvme/target/auth.c index 3ddbc3880cac..9e51c064b072 100644 --- a/drivers/nvme/target/auth.c +++ b/drivers/nvme/target/auth.c @@ -370,7 +370,7 @@ out_free_response: nvme_auth_free_key(transformed_key); out_free_tfm: crypto_free_shash(shash_tfm); - return 0; + return ret; } int nvmet_auth_ctrl_hash(struct nvmet_req *req, u8 *response, -- cgit v1.2.3-58-ga151 From 445f9119e70368ccc964575c2a6d3176966a9d65 Mon Sep 17 00:00:00 2001 From: Maurizio Lombardi Date: Wed, 10 Apr 2024 11:48:42 +0200 Subject: nvmet-auth: replace pr_debug() with pr_err() to report an error. In nvmet_auth_host_hash(), if a mismatch is detected in the hash length the kernel should print an error. Signed-off-by: Maurizio Lombardi Reviewed-by: Sagi Grimberg Reviewed-by: Chaitanya Kulkarni Signed-off-by: Keith Busch --- drivers/nvme/target/auth.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/nvme/target/auth.c b/drivers/nvme/target/auth.c index 9e51c064b072..fb518b00f71f 100644 --- a/drivers/nvme/target/auth.c +++ b/drivers/nvme/target/auth.c @@ -285,9 +285,9 @@ int nvmet_auth_host_hash(struct nvmet_req *req, u8 *response, } if (shash_len != crypto_shash_digestsize(shash_tfm)) { - pr_debug("%s: hash len mismatch (len %d digest %d)\n", - __func__, shash_len, - crypto_shash_digestsize(shash_tfm)); + pr_err("%s: hash len mismatch (len %d digest %d)\n", + __func__, shash_len, + crypto_shash_digestsize(shash_tfm)); ret = -EINVAL; goto out_free_tfm; } -- cgit v1.2.3-58-ga151 From 25bb3534ee21e39eb9301c4edd7182eb83cb0d07 Mon Sep 17 00:00:00 2001 From: Nilay Shroff Date: Thu, 25 Apr 2024 19:33:00 +0530 Subject: nvme: cancel pending I/O if nvme controller is in terminal state While I/O is running, if the pci bus error occurs then in-flight I/O can not complete. Worst, if at this time, user (logically) hot-unplug the nvme disk then the nvme_remove() code path can't forward progress until in-flight I/O is cancelled. So these sequence of events may potentially hang hot-unplug code path indefinitely. This patch helps cancel the pending/in-flight I/O from the nvme request timeout handler in case the nvme controller is in the terminal (DEAD/DELETING/DELETING_NOIO) state and that helps nvme_remove() code path forward progress and finish successfully. Link: https://lore.kernel.org/all/199be893-5dfa-41e5-b6f2-40ac90ebccc4@linux.ibm.com/ Signed-off-by: Nilay Shroff Reviewed-by: Sagi Grimberg Signed-off-by: Keith Busch --- drivers/nvme/host/core.c | 21 --------------------- drivers/nvme/host/nvme.h | 21 +++++++++++++++++++++ drivers/nvme/host/pci.c | 8 +++++++- 3 files changed, 28 insertions(+), 22 deletions(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index e4bec200ebeb..095f59e7aa93 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -628,27 +628,6 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl, } EXPORT_SYMBOL_GPL(nvme_change_ctrl_state); -/* - * Returns true for sink states that can't ever transition back to live. - */ -static bool nvme_state_terminal(struct nvme_ctrl *ctrl) -{ - switch (nvme_ctrl_state(ctrl)) { - case NVME_CTRL_NEW: - case NVME_CTRL_LIVE: - case NVME_CTRL_RESETTING: - case NVME_CTRL_CONNECTING: - return false; - case NVME_CTRL_DELETING: - case NVME_CTRL_DELETING_NOIO: - case NVME_CTRL_DEAD: - return true; - default: - WARN_ONCE(1, "Unhandled ctrl state:%d", ctrl->state); - return true; - } -} - /* * Waits for the controller state to be resetting, or returns false if it is * not possible to ever transition to that state. diff --git a/drivers/nvme/host/nvme.h b/drivers/nvme/host/nvme.h index d0ed64dc7380..b564c5f1450c 100644 --- a/drivers/nvme/host/nvme.h +++ b/drivers/nvme/host/nvme.h @@ -741,6 +741,27 @@ static inline bool nvme_is_aen_req(u16 qid, __u16 command_id) nvme_tag_from_cid(command_id) >= NVME_AQ_BLK_MQ_DEPTH; } +/* + * Returns true for sink states that can't ever transition back to live. + */ +static inline bool nvme_state_terminal(struct nvme_ctrl *ctrl) +{ + switch (nvme_ctrl_state(ctrl)) { + case NVME_CTRL_NEW: + case NVME_CTRL_LIVE: + case NVME_CTRL_RESETTING: + case NVME_CTRL_CONNECTING: + return false; + case NVME_CTRL_DELETING: + case NVME_CTRL_DELETING_NOIO: + case NVME_CTRL_DEAD: + return true; + default: + WARN_ONCE(1, "Unhandled ctrl state:%d", ctrl->state); + return true; + } +} + void nvme_complete_rq(struct request *req); void nvme_complete_batch_req(struct request *req); diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c index 8e0bb9692685..e393f6947ce4 100644 --- a/drivers/nvme/host/pci.c +++ b/drivers/nvme/host/pci.c @@ -1286,6 +1286,9 @@ static enum blk_eh_timer_return nvme_timeout(struct request *req) u32 csts = readl(dev->bar + NVME_REG_CSTS); u8 opcode; + if (nvme_state_terminal(&dev->ctrl)) + goto disable; + /* If PCI error recovery process is happening, we cannot reset or * the recovery mechanism will surely fail. */ @@ -1390,8 +1393,11 @@ static enum blk_eh_timer_return nvme_timeout(struct request *req) return BLK_EH_RESET_TIMER; disable: - if (!nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_RESETTING)) + if (!nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_RESETTING)) { + if (nvme_state_terminal(&dev->ctrl)) + nvme_dev_disable(dev, true); return BLK_EH_DONE; + } nvme_dev_disable(dev, false); if (nvme_try_sched_reset(&dev->ctrl)) -- cgit v1.2.3-58-ga151 From 6825bdde44340c5a9121f6d6fa25cc885bd9e821 Mon Sep 17 00:00:00 2001 From: Sagi Grimberg Date: Sun, 28 Apr 2024 11:49:49 +0300 Subject: nvmet-tcp: fix possible memory leak when tearing down a controller When we teardown the controller, we wait for pending I/Os to complete (sq->ref on all queues to drop to zero) and then we go over the commands, and free their command buffers in case they are still fetching data from the host (e.g. processing nvme writes) and have yet to take a reference on the sq. However, we may miss the case where commands have failed before executing and are queued for sending a response, but will never occur because the queue socket is already down. In this case we may miss deallocating command buffers. Solve this by freeing all commands buffers as nvmet_tcp_free_cmd_buffers is idempotent anyways. Reported-by: Yi Zhang Tested-by: Yi Zhang Signed-off-by: Sagi Grimberg Reviewed-by: Christoph Hellwig Signed-off-by: Keith Busch --- drivers/nvme/target/tcp.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) diff --git a/drivers/nvme/target/tcp.c b/drivers/nvme/target/tcp.c index a5422e2c979a..380f22ee3ebb 100644 --- a/drivers/nvme/target/tcp.c +++ b/drivers/nvme/target/tcp.c @@ -348,6 +348,7 @@ static int nvmet_tcp_check_ddgst(struct nvmet_tcp_queue *queue, void *pdu) return 0; } +/* If cmd buffers are NULL, no operation is performed */ static void nvmet_tcp_free_cmd_buffers(struct nvmet_tcp_cmd *cmd) { kfree(cmd->iov); @@ -1581,13 +1582,9 @@ static void nvmet_tcp_free_cmd_data_in_buffers(struct nvmet_tcp_queue *queue) struct nvmet_tcp_cmd *cmd = queue->cmds; int i; - for (i = 0; i < queue->nr_cmds; i++, cmd++) { - if (nvmet_tcp_need_data_in(cmd)) - nvmet_tcp_free_cmd_buffers(cmd); - } - - if (!queue->nr_cmds && nvmet_tcp_need_data_in(&queue->connect)) - nvmet_tcp_free_cmd_buffers(&queue->connect); + for (i = 0; i < queue->nr_cmds; i++, cmd++) + nvmet_tcp_free_cmd_buffers(cmd); + nvmet_tcp_free_cmd_buffers(&queue->connect); } static void nvmet_tcp_release_queue_work(struct work_struct *w) -- cgit v1.2.3-58-ga151 From 505363957fad35f7aed9a2b0d8dad73451a80fb5 Mon Sep 17 00:00:00 2001 From: Sagi Grimberg Date: Sun, 28 Apr 2024 12:25:40 +0300 Subject: nvmet: fix nvme status code when namespace is disabled If the user disabled a nvmet namespace, it is removed from the subsystem namespaces list. When nvmet processes a command directed to an nsid that was disabled, it cannot differentiate between a nsid that is disabled vs. a non-existent namespace, and resorts to return NVME_SC_INVALID_NS with the dnr bit set. This translates to a non-retryable status for the host, which translates to a user error. We should expect disabled namespaces to not cause an I/O error in a multipath environment. Address this by searching a configfs item for the namespace nvmet failed to find, and if we found one, conclude that the namespace is disabled (perhaps temporarily). Return NVME_SC_INTERNAL_PATH_ERROR in this case and keep DNR bit cleared. Reported-by: Jirong Feng Tested-by: Jirong Feng Signed-off-by: Sagi Grimberg Reviewed-by: Christoph Hellwig Signed-off-by: Keith Busch --- drivers/nvme/target/configfs.c | 13 +++++++++++++ drivers/nvme/target/core.c | 5 ++++- drivers/nvme/target/nvmet.h | 1 + 3 files changed, 18 insertions(+), 1 deletion(-) diff --git a/drivers/nvme/target/configfs.c b/drivers/nvme/target/configfs.c index a2325330bf22..89a001101a7f 100644 --- a/drivers/nvme/target/configfs.c +++ b/drivers/nvme/target/configfs.c @@ -754,6 +754,19 @@ static struct configfs_attribute *nvmet_ns_attrs[] = { NULL, }; +bool nvmet_subsys_nsid_exists(struct nvmet_subsys *subsys, u32 nsid) +{ + struct config_item *ns_item; + char name[4] = {}; + + if (sprintf(name, "%u", nsid) <= 0) + return false; + mutex_lock(&subsys->namespaces_group.cg_subsys->su_mutex); + ns_item = config_group_find_item(&subsys->namespaces_group, name); + mutex_unlock(&subsys->namespaces_group.cg_subsys->su_mutex); + return ns_item != NULL; +} + static void nvmet_ns_release(struct config_item *item) { struct nvmet_ns *ns = to_nvmet_ns(item); diff --git a/drivers/nvme/target/core.c b/drivers/nvme/target/core.c index 8860a3eb71ec..e06013c5dace 100644 --- a/drivers/nvme/target/core.c +++ b/drivers/nvme/target/core.c @@ -437,10 +437,13 @@ void nvmet_stop_keep_alive_timer(struct nvmet_ctrl *ctrl) u16 nvmet_req_find_ns(struct nvmet_req *req) { u32 nsid = le32_to_cpu(req->cmd->common.nsid); + struct nvmet_subsys *subsys = nvmet_req_subsys(req); - req->ns = xa_load(&nvmet_req_subsys(req)->namespaces, nsid); + req->ns = xa_load(&subsys->namespaces, nsid); if (unlikely(!req->ns)) { req->error_loc = offsetof(struct nvme_common_command, nsid); + if (nvmet_subsys_nsid_exists(subsys, nsid)) + return NVME_SC_INTERNAL_PATH_ERROR; return NVME_SC_INVALID_NS | NVME_SC_DNR; } diff --git a/drivers/nvme/target/nvmet.h b/drivers/nvme/target/nvmet.h index f460728e1df1..c1306de1f4dd 100644 --- a/drivers/nvme/target/nvmet.h +++ b/drivers/nvme/target/nvmet.h @@ -543,6 +543,7 @@ void nvmet_subsys_disc_changed(struct nvmet_subsys *subsys, struct nvmet_host *host); void nvmet_add_async_event(struct nvmet_ctrl *ctrl, u8 event_type, u8 event_info, u8 log_page); +bool nvmet_subsys_nsid_exists(struct nvmet_subsys *subsys, u32 nsid); #define NVMET_MIN_QUEUE_SIZE 16 #define NVMET_MAX_QUEUE_SIZE 1024 -- cgit v1.2.3-58-ga151 From 50abcc179e0c9ca667feb223b26ea406d5c4c556 Mon Sep 17 00:00:00 2001 From: Hannes Reinecke Date: Thu, 18 Apr 2024 12:39:45 +0200 Subject: nvme-tcp: strict pdu pacing to avoid send stalls on TLS TLS requires a strict pdu pacing via MSG_EOR to signal the end of a record and subsequent encryption. If we do not set MSG_EOR at the end of a sequence the record won't be closed, encryption doesn't start, and we end up with a send stall as the message will never be passed on to the TCP layer. So do not check for the queue status when TLS is enabled but rather make the MSG_MORE setting dependent on the current request only. Signed-off-by: Hannes Reinecke Reviewed-by: Sagi Grimberg Signed-off-by: Keith Busch --- drivers/nvme/host/tcp.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c index fdbcdcedcee9..28bc2f373cfa 100644 --- a/drivers/nvme/host/tcp.c +++ b/drivers/nvme/host/tcp.c @@ -360,12 +360,18 @@ static inline void nvme_tcp_send_all(struct nvme_tcp_queue *queue) } while (ret > 0); } -static inline bool nvme_tcp_queue_more(struct nvme_tcp_queue *queue) +static inline bool nvme_tcp_queue_has_pending(struct nvme_tcp_queue *queue) { return !list_empty(&queue->send_list) || !llist_empty(&queue->req_list); } +static inline bool nvme_tcp_queue_more(struct nvme_tcp_queue *queue) +{ + return !nvme_tcp_tls(&queue->ctrl->ctrl) && + nvme_tcp_queue_has_pending(queue); +} + static inline void nvme_tcp_queue_request(struct nvme_tcp_request *req, bool sync, bool last) { @@ -386,7 +392,7 @@ static inline void nvme_tcp_queue_request(struct nvme_tcp_request *req, mutex_unlock(&queue->send_mutex); } - if (last && nvme_tcp_queue_more(queue)) + if (last && nvme_tcp_queue_has_pending(queue)) queue_work_on(queue->io_cpu, nvme_tcp_wq, &queue->io_work); } -- cgit v1.2.3-58-ga151