[PATCH 3/4] nvme-multipath: check for NVME_NSHEAD_DISK_LIVE when selecting paths

[Hannes Reinecke]

During repetitive namespace remapping operations on the target the
namespace might have changed between the time the initial scan
was performed, and partition scan was invoked by device_add_disk()
in nvme_mpath_set_live(). We then end up with a stuck scanning process:

[<0>] folio_wait_bit_common+0x12a/0x310
[<0>] filemap_read_folio+0x97/0xd0
[<0>] do_read_cache_folio+0x108/0x390
[<0>] read_part_sector+0x31/0xa0
[<0>] read_lba+0xc5/0x160
[<0>] efi_partition+0xd9/0x8f0
[<0>] bdev_disk_changed+0x23d/0x6d0
[<0>] blkdev_get_whole+0x78/0xc0
[<0>] bdev_open+0x2c6/0x3b0
[<0>] bdev_file_open_by_dev+0xcb/0x120
[<0>] disk_scan_partitions+0x5d/0x100
[<0>] device_add_disk+0x402/0x420
[<0>] nvme_mpath_set_live+0x4f/0x1f0 [nvme_core]
[<0>] nvme_mpath_add_disk+0x107/0x120 [nvme_core]
[<0>] nvme_alloc_ns+0xac6/0xe60 [nvme_core]
[<0>] nvme_scan_ns+0x2dd/0x3e0 [nvme_core]
[<0>] nvme_scan_work+0x1a3/0x490 [nvme_core]

and another scanning process (triggered by AEN) trying to remove the namespaces:

[<0>] __synchronize_srcu+0x17c/0x1b0
[<0>] nvme_ns_remove+0x12e/0x210 [nvme_core]
[<0>] nvme_ns_remove_by_nsid+0x21/0x70 [nvme_core]
[<0>] nvme_scan_work+0x1b7/0x490 [nvme_core]
[<0>] process_scheduled_works+0x37d/0x6d0

none of which is able to make progress as the first process cannot submit I/O
(as all the namespace inforation is stale), and the second process is stalled
waiting on the srcu to be released by the first process.

This patch checks for the NVME_NSHEAD_DISK_LIVE bit when selecting a path,
causing I/O to be aborted and allowing the first process to make progress.

Signed-off-by: Hannes Reinecke <hare at kernel.org>
---
 drivers/nvme/host/multipath.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c
index 01dbbe866d61..f5f2e1667c64 100644
--- a/drivers/nvme/host/multipath.c
+++ b/drivers/nvme/host/multipath.c
@@ -407,6 +407,9 @@ static struct nvme_ns *nvme_numa_path(struct nvme_ns_head *head)
 
 inline struct nvme_ns *nvme_find_path(struct nvme_ns_head *head)
 {
+	if (!test_bit(NVME_NSHEAD_DISK_LIVE, &head->flags))
+		return NULL;
+
 	switch (READ_ONCE(head->subsys->iopolicy)) {
 	case NVME_IOPOLICY_QD:
 		return nvme_queue_depth_path(head);
@@ -421,6 +424,9 @@ static bool nvme_available_path(struct nvme_ns_head *head)
 {
 	struct nvme_ns *ns;
 
+	if (!test_bit(NVME_NSHEAD_DISK_LIVE, &head->flags))
+		return NULL;
+
 	list_for_each_entry_rcu(ns, &head->list, siblings) {
 		if (test_bit(NVME_CTRL_FAILFAST_EXPIRED, &ns->ctrl->flags))
 			continue;
-- 
2.35.3