Re: [bug report] blktests block/022 failed with possible circular locking dependency detected at: __kernfs_remove+0x275/0x710 and del_gendisk+0xfd/0x190

[Ming Lei <ming.lei@redhat.com>]

On Wed, Mar 04, 2026 at 03:10:06PM +0800, Yi Zhang wrote:
> Hi
> 
> I found blktests block/022 failed on the latest
> linux-block/for-next[1], please help check it and let me know if you
> need any info/testing for it, thanks.
> 
> [1]
> commit b82c5e4f1bcb3fa78ba3c95a0d198f55553c0fc1
> Merge: 187d9a5a18b4 d47f7c173262
> Author: Jens Axboe <axboe@kernel.dk>
> Date:   Mon Mar 2 14:04:16 2026 -0700
> 
>     Merge branch 'for-7.1/block' into for-next
> 
> dmesg:
> [ 1329.227229] run blktests block/022 at 2026-03-03 13:51:07
> [ 1329.838331] null_blk: module loaded
> --snip--

...

> [ 1349.822314]  Possible unsafe locking scenario:
> 
> [ 1349.828232]        CPU0                    CPU1
> [ 1349.832766]        ----                    ----
> [ 1349.837296]   rlock(&set->update_nr_hwq_lock);
> [ 1349.841752]                                lock(kn->active#104);
> [ 1349.847767]                                lock(&set->update_nr_hwq_lock);
> [ 1349.854647]   lock(kn->active#104);
> [ 1349.858148]
>                 *** DEADLOCK ***

Looks one real deadlock risk, can you try the following patch?


diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
index f3b1968c80ce..55a1bbfef7d4 100644
--- a/block/blk-sysfs.c
+++ b/block/blk-sysfs.c
@@ -78,8 +78,14 @@ queue_requests_store(struct gendisk *disk, const char *page, size_t count)
 	/*
 	 * Serialize updating nr_requests with concurrent queue_requests_store()
 	 * and switching elevator.
+	 *
+	 * Use trylock to avoid circular lock dependency with kernfs active
+	 * reference during concurrent disk deletion:
+	 *   update_nr_hwq_lock -> kn->active (via del_gendisk -> kobject_del)
+	 *   kn->active -> update_nr_hwq_lock (via this sysfs write path)
 	 */
-	down_write(&set->update_nr_hwq_lock);
+	if (!down_write_trylock(&set->update_nr_hwq_lock))
+		return -EBUSY;
 
 	if (nr == q->nr_requests)
 		goto unlock;
diff --git a/block/elevator.c b/block/elevator.c
index ebe2a1fcf011..3bcd37c2aa34 100644
--- a/block/elevator.c
+++ b/block/elevator.c
@@ -807,7 +807,16 @@ ssize_t elv_iosched_store(struct gendisk *disk, const char *buf,
 	elv_iosched_load_module(ctx.name);
 	ctx.type = elevator_find_get(ctx.name);
 
-	down_read(&set->update_nr_hwq_lock);
+	/*
+	 * Use trylock to avoid circular lock dependency with kernfs active
+	 * reference during concurrent disk deletion:
+	 *   update_nr_hwq_lock -> kn->active (via del_gendisk -> kobject_del)
+	 *   kn->active -> update_nr_hwq_lock (via this sysfs write path)
+	 */
+	if (!down_read_trylock(&set->update_nr_hwq_lock)) {
+		ret = -EBUSY;
+		goto out;
+	}
 	if (!blk_queue_no_elv_switch(q)) {
 		ret = elevator_change(q, &ctx);
 		if (!ret)
@@ -817,6 +826,7 @@ ssize_t elv_iosched_store(struct gendisk *disk, const char *buf,
 	}
 	up_read(&set->update_nr_hwq_lock);
 
+out:
 	if (ctx.type)
 		elevator_put(ctx.type);
 	return ret;


Thanks,
Ming