Merge tag 'nvme-6.13-2024-11-21' of git://git.infradead.org/nvme into for-6.13/block

Pull NVMe updates from Keith:

"nvme updates for Linux 6.13

 - Use correct srcu list traversal (Breno)
 - Scatter-gather support for metadata (Keith)
 - Fabrics shutdown race condition fix (Nilay)
 - Persistent reservations updates (Guixin)"

* tag 'nvme-6.13-2024-11-21' of git://git.infradead.org/nvme:
  nvme: tuning pr code by using defined structs and macros
  nvme: introduce change ptpl and iekey definition
  nvme-fabrics: fix kernel crash while shutting down controller
  Revert "nvme: make keep-alive synchronous operation"
  nvme-pci: use sgls for all user requests if possible
  nvme: define the remaining used sgls constants
  nvme-pci: add support for sgl metadata
  nvme/multipath: Fix RCU list traversal to use SRCU primitive

Author: axboe

drivers/nvme/host/core.c

@@ -1294,9 +1294,10 @@ static void nvme_queue_keep_alive_work(struct nvme_ctrl *ctrl)
 	queue_delayed_work(nvme_wq, &ctrl->ka_work, delay);
 }
 
-static void nvme_keep_alive_finish(struct request *rq,
-		blk_status_t status, struct nvme_ctrl *ctrl)
+static enum rq_end_io_ret nvme_keep_alive_end_io(struct request *rq,
+						 blk_status_t status)
 {
+	struct nvme_ctrl *ctrl = rq->end_io_data;
 	unsigned long rtt = jiffies - (rq->deadline - rq->timeout);
 	unsigned long delay = nvme_keep_alive_work_period(ctrl);
 	enum nvme_ctrl_state state = nvme_ctrl_state(ctrl);
@@ -1313,17 +1314,20 @@ static void nvme_keep_alive_finish(struct request *rq,
 		delay = 0;
 	}
 
+	blk_mq_free_request(rq);
+
 	if (status) {
 		dev_err(ctrl->device,
 			"failed nvme_keep_alive_end_io error=%d\n",
 				status);
-		return;
+		return RQ_END_IO_NONE;
 	}
 
 	ctrl->ka_last_check_time = jiffies;
 	ctrl->comp_seen = false;
 	if (state == NVME_CTRL_LIVE || state == NVME_CTRL_CONNECTING)
 		queue_delayed_work(nvme_wq, &ctrl->ka_work, delay);
+	return RQ_END_IO_NONE;
 }
 
 static void nvme_keep_alive_work(struct work_struct *work)
@@ -1332,7 +1336,6 @@ static void nvme_keep_alive_work(struct work_struct *work)
 			struct nvme_ctrl, ka_work);
 	bool comp_seen = ctrl->comp_seen;
 	struct request *rq;
-	blk_status_t status;
 
 	ctrl->ka_last_check_time = jiffies;
 
@@ -1355,9 +1358,9 @@ static void nvme_keep_alive_work(struct work_struct *work)
 	nvme_init_request(rq, &ctrl->ka_cmd);
 
 	rq->timeout = ctrl->kato * HZ;
-	status = blk_execute_rq(rq, false);
-	nvme_keep_alive_finish(rq, status, ctrl);
-	blk_mq_free_request(rq);
+	rq->end_io = nvme_keep_alive_end_io;
+	rq->end_io_data = ctrl;
+	blk_execute_rq_nowait(rq, false);
 }
 
 static void nvme_start_keep_alive(struct nvme_ctrl *ctrl)
@@ -4571,6 +4574,11 @@ EXPORT_SYMBOL_GPL(nvme_alloc_admin_tag_set);
 
 void nvme_remove_admin_tag_set(struct nvme_ctrl *ctrl)
 {
+	/*
+	 * As we're about to destroy the queue and free tagset
+	 * we can not have keep-alive work running.
+	 */
+	nvme_stop_keep_alive(ctrl);
 	blk_mq_destroy_queue(ctrl->admin_q);
 	blk_put_queue(ctrl->admin_q);
 	if (ctrl->ops->flags & NVME_F_FABRICS) {

drivers/nvme/host/ioctl.c

@@ -120,12 +120,20 @@ static int nvme_map_user_request(struct request *req, u64 ubuffer,
 	struct nvme_ns *ns = q->queuedata;
 	struct block_device *bdev = ns ? ns->disk->part0 : NULL;
 	bool supports_metadata = bdev && blk_get_integrity(bdev->bd_disk);
+	struct nvme_ctrl *ctrl = nvme_req(req)->ctrl;
 	bool has_metadata = meta_buffer && meta_len;
 	struct bio *bio = NULL;
 	int ret;
 
-	if (has_metadata && !supports_metadata)
-		return -EINVAL;
+	if (!nvme_ctrl_sgl_supported(ctrl))
+		dev_warn_once(ctrl->device, "using unchecked data buffer\n");
+	if (has_metadata) {
+		if (!supports_metadata)
+			return -EINVAL;
+		if (!nvme_ctrl_meta_sgl_supported(ctrl))
+			dev_warn_once(ctrl->device,
+				      "using unchecked metadata buffer\n");
+	}
 
 	if (ioucmd && (ioucmd->flags & IORING_URING_CMD_FIXED)) {
 		struct iov_iter iter;

drivers/nvme/host/multipath.c

@@ -165,7 +165,8 @@ void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl)
 	int srcu_idx;
 
 	srcu_idx = srcu_read_lock(&ctrl->srcu);
-	list_for_each_entry_rcu(ns, &ctrl->namespaces, list) {
+	list_for_each_entry_srcu(ns, &ctrl->namespaces, list,
+				 srcu_read_lock_held(&ctrl->srcu)) {
 		if (!ns->head->disk)
 			continue;
 		kblockd_schedule_work(&ns->head->requeue_work);
@@ -209,7 +210,8 @@ void nvme_mpath_clear_ctrl_paths(struct nvme_ctrl *ctrl)
 	int srcu_idx;
 
 	srcu_idx = srcu_read_lock(&ctrl->srcu);
-	list_for_each_entry_rcu(ns, &ctrl->namespaces, list) {
+	list_for_each_entry_srcu(ns, &ctrl->namespaces, list,
+				 srcu_read_lock_held(&ctrl->srcu)) {
 		nvme_mpath_clear_current_path(ns);
 		kblockd_schedule_work(&ns->head->requeue_work);
 	}
@@ -224,7 +226,8 @@ void nvme_mpath_revalidate_paths(struct nvme_ns *ns)
 	int srcu_idx;
 
 	srcu_idx = srcu_read_lock(&head->srcu);
-	list_for_each_entry_rcu(ns, &head->list, siblings) {
+	list_for_each_entry_srcu(ns, &head->list, siblings,
+				 srcu_read_lock_held(&head->srcu)) {
 		if (capacity != get_capacity(ns->disk))
 			clear_bit(NVME_NS_READY, &ns->flags);
 	}
@@ -257,7 +260,8 @@ static struct nvme_ns *__nvme_find_path(struct nvme_ns_head *head, int node)
 	int found_distance = INT_MAX, fallback_distance = INT_MAX, distance;
 	struct nvme_ns *found = NULL, *fallback = NULL, *ns;
 
-	list_for_each_entry_rcu(ns, &head->list, siblings) {
+	list_for_each_entry_srcu(ns, &head->list, siblings,
+				 srcu_read_lock_held(&head->srcu)) {
 		if (nvme_path_is_disabled(ns))
 			continue;
 
@@ -356,7 +360,8 @@ static struct nvme_ns *nvme_queue_depth_path(struct nvme_ns_head *head)
 	unsigned int min_depth_opt = UINT_MAX, min_depth_nonopt = UINT_MAX;
 	unsigned int depth;
 
-	list_for_each_entry_rcu(ns, &head->list, siblings) {
+	list_for_each_entry_srcu(ns, &head->list, siblings,
+				 srcu_read_lock_held(&head->srcu)) {
 		if (nvme_path_is_disabled(ns))
 			continue;
 
@@ -424,7 +429,8 @@ static bool nvme_available_path(struct nvme_ns_head *head)
 	if (!test_bit(NVME_NSHEAD_DISK_LIVE, &head->flags))
 		return NULL;
 
-	list_for_each_entry_rcu(ns, &head->list, siblings) {
+	list_for_each_entry_srcu(ns, &head->list, siblings,
+				 srcu_read_lock_held(&head->srcu)) {
 		if (test_bit(NVME_CTRL_FAILFAST_EXPIRED, &ns->ctrl->flags))
 			continue;
 		switch (nvme_ctrl_state(ns->ctrl)) {
@@ -783,7 +789,8 @@ static int nvme_update_ana_state(struct nvme_ctrl *ctrl,
 		return 0;
 
 	srcu_idx = srcu_read_lock(&ctrl->srcu);
-	list_for_each_entry_rcu(ns, &ctrl->namespaces, list) {
+	list_for_each_entry_srcu(ns, &ctrl->namespaces, list,
+				 srcu_read_lock_held(&ctrl->srcu)) {
 		unsigned nsid;
 again:
 		nsid = le32_to_cpu(desc->nsids[n]);

drivers/nvme/host/nvme.h

@@ -1123,7 +1123,15 @@ static inline void nvme_start_request(struct request *rq)
 
 static inline bool nvme_ctrl_sgl_supported(struct nvme_ctrl *ctrl)
 {
-	return ctrl->sgls & ((1 << 0) | (1 << 1));
+	return ctrl->sgls & (NVME_CTRL_SGLS_BYTE_ALIGNED |
+			     NVME_CTRL_SGLS_DWORD_ALIGNED);
+}
+
+static inline bool nvme_ctrl_meta_sgl_supported(struct nvme_ctrl *ctrl)
+{
+	if (ctrl->ops->flags & NVME_F_FABRICS)
+		return true;
+	return ctrl->sgls & NVME_CTRL_SGLS_MSDS;
 }
 
 #ifdef CONFIG_NVME_HOST_AUTH