* [PATCH v3 1/3] dm thin: drop the grace period from thin_ctr()
2026-11-19 11:05 [PATCH v3 0/3] dm thin: stop waiting for grace periods when activating and removing thins Qiliang Yuan
@ 2026-11-19 11:05 ` Qiliang Yuan
2026-11-19 11:05 ` [PATCH v3 2/3] dm thin: keep a thin on active_thins until its last reference is dropped Qiliang Yuan
2026-11-19 11:05 ` [PATCH v3 3/3] dm thin: free thins with kfree_rcu() instead of waiting in thin_dtr() Qiliang Yuan
2 siblings, 0 replies; 6+ messages in thread
From: Qiliang Yuan @ 2026-11-19 11:05 UTC (permalink / raw)
To: Alasdair Kergon, Mike Snitzer, Mikulas Patocka,
Benjamin Marzinski
Cc: dm-devel, linux-kernel, Qiliang Yuan
thin_ctr() adds the new thin to pool->active_thins with
list_add_tail_rcu() and then calls synchronize_rcu(), so that a
wake_worker() call can't miss the bios of a thin that isn't visible yet.
The grace period doesn't make the entry any more visible. A worker that
walks the list after list_add_tail_rcu() finds the thin, and the thin
gets no bios before its table is resumed. A walk already in progress may
miss it, but the wake_worker() for the first bio queues the worker
again. Still, activating a thin device waits for a full grace period,
which takes about half of the activation.
Remove the synchronize_rcu() call.
Activating 100 thin devices of an idle pool with dmsetup create takes
per device (mean of five rounds, min-max of the rounds in parentheses):
before (ms) after (ms) change
15.5 (15.2-15.7) 7.7 (7.5-7.9) -50%
Signed-off-by: Qiliang Yuan <qiliang.yuan@linux.dev>
Reviewed-by: Benjamin Marzinski <bmarzins@redhat.com>
---
drivers/md/dm-thin.c | 7 -------
1 file changed, 7 deletions(-)
diff --git a/drivers/md/dm-thin.c b/drivers/md/dm-thin.c
index 0b65c1b09f47b..bdd250d1eb2e1 100644
--- a/drivers/md/dm-thin.c
+++ b/drivers/md/dm-thin.c
@@ -4292,13 +4292,6 @@ static int thin_ctr(struct dm_target *ti, unsigned int argc, char **argv)
init_completion(&tc->can_destroy);
list_add_tail_rcu(&tc->list, &tc->pool->active_thins);
spin_unlock_irq(&tc->pool->lock);
- /*
- * This synchronize_rcu() call is needed here otherwise we risk a
- * wake_worker() call finding no bios to process (because the newly
- * added tc isn't yet visible). So this reduces latency since we
- * aren't then dependent on the periodic commit to wake_worker().
- */
- synchronize_rcu();
dm_put(pool_md);
--
2.43.0
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH v3 2/3] dm thin: keep a thin on active_thins until its last reference is dropped
2026-11-19 11:05 [PATCH v3 0/3] dm thin: stop waiting for grace periods when activating and removing thins Qiliang Yuan
2026-11-19 11:05 ` [PATCH v3 1/3] dm thin: drop the grace period from thin_ctr() Qiliang Yuan
@ 2026-11-19 11:05 ` Qiliang Yuan
2026-11-19 19:20 ` Benjamin Marzinski
2026-11-19 11:05 ` [PATCH v3 3/3] dm thin: free thins with kfree_rcu() instead of waiting in thin_dtr() Qiliang Yuan
2 siblings, 1 reply; 6+ messages in thread
From: Qiliang Yuan @ 2026-11-19 11:05 UTC (permalink / raw)
To: Alasdair Kergon, Mike Snitzer, Mikulas Patocka,
Benjamin Marzinski
Cc: dm-devel, linux-kernel, Qiliang Yuan, stable
get_first_thin() and get_next_thin() let the pool worker process the
thins on pool->active_thins without holding rcu_read_lock() while it
does: they take a reference on a thin under the lock, and
get_next_thin() continues from that thin in a new read-side section.
thin_dtr() takes the thin off the list, waits for a grace period and
then for the references still held.
Once thin_dtr() has removed a thin that the worker holds, nothing
updates its list.next pointer any more. If the thin after it is removed
and freed before the worker calls get_next_thin(), the walk continues
from freed memory:
BUG: KASAN: slab-use-after-free in get_next_thin+0x8b/0x240 [dm_thin_pool]
Write of size 4 at addr ff11000116371cc8 by task kworker/u128:0/13
Workqueue: dm-thin do_worker [dm_thin_pool]
Call Trace:
get_next_thin+0x8b/0x240 [dm_thin_pool]
process_deferred_bios+0x34/0x650 [dm_thin_pool]
do_worker+0x2b7/0x4f0 [dm_thin_pool]
process_one_work+0x8ce/0x1850
worker_thread+0x639/0x1190
Allocated by task 2290:
thin_ctr+0x130/0x1260 [dm_thin_pool]
dm_table_add_target+0x65f/0xf60 [dm_mod]
table_load+0x1c2/0x480 [dm_mod]
Freed by task 2391:
kfree+0x15d/0x550
thin_dtr+0x21c/0x2c0 [dm_thin_pool]
dm_table_destroy+0x13c/0x570 [dm_mod]
__dm_destroy+0x2eb/0x5d0 [dm_mod]
dev_remove+0x2f1/0x3c0 [dm_mod]
Mark the thin as dying in thin_dtr() instead of removing it, and take it
off the list only after the last reference is dropped, so that removing
its neighbours keeps its list.next valid. Take references with
refcount_inc_not_zero() and skip dying thins, so that thin_dtr() only
waits for the references already taken, and have requeue_bios() and
error_retry_list_with_code() skip dying thins as they skipped the thins
already off the list.
With a test-only 3 s msleep() added to the worker after each thin it
processes, thins 1 and 2 of a pool with three thins are suspended and
then removed while the worker holds thin 1. Trials in which KASAN reports
a slab-use-after-free in get_next_thin():
before (trials) after (trials) change
KASAN report 3 of 3 0 of 5 -100%
Fixes: 17181fb7a0c3 ("dm thin: fix a race in thin_dtr")
Cc: stable@vger.kernel.org
Reported-by: Benjamin Marzinski <bmarzins@redhat.com>
Closes: https://lore.kernel.org/r/avz7qDWd2g-uubmL@redhat.com
Co-developed-by: Benjamin Marzinski <bmarzins@redhat.com>
Signed-off-by: Benjamin Marzinski <bmarzins@redhat.com>
Signed-off-by: Qiliang Yuan <qiliang.yuan@linux.dev>
---
drivers/md/dm-thin.c | 49 +++++++++++++++++++++++++++++++++----------------
1 file changed, 33 insertions(+), 16 deletions(-)
diff --git a/drivers/md/dm-thin.c b/drivers/md/dm-thin.c
index bdd250d1eb2e1..62ef96bf08cb3 100644
--- a/drivers/md/dm-thin.c
+++ b/drivers/md/dm-thin.c
@@ -355,11 +355,16 @@ struct thin_c {
struct rb_root sort_bio_list; /* sorted list of deferred bios */
/*
- * Ensures the thin is not destroyed until the worker has finished
- * iterating the active_thins list.
+ * Ensures the thin stays on the active_thins list until the workers
+ * have moved past it while iterating the list.
*/
refcount_t refcount;
struct completion can_destroy;
+ /*
+ * Keeps new active_thin walkers from grabbing the table while
+ * thin_dtr() is waiting for the existing holders to finish.
+ */
+ bool dying;
};
/*----------------------------------------------------------------*/
@@ -652,8 +657,11 @@ static void error_retry_list_with_code(struct pool *pool, blk_status_t error)
struct thin_c *tc;
rcu_read_lock();
- list_for_each_entry_rcu(tc, &pool->active_thins, list)
+ list_for_each_entry_rcu(tc, &pool->active_thins, list) {
+ if (READ_ONCE(tc->dying))
+ continue;
error_thin_bio_list(tc, &tc->retry_on_resume_list, error);
+ }
rcu_read_unlock();
}
@@ -2318,25 +2326,29 @@ static void process_thin_deferred_cells(struct thin_c *tc)
} while (!list_empty(&cells));
}
-static void thin_get(struct thin_c *tc);
+static bool thin_tryget(struct thin_c *tc);
static void thin_put(struct thin_c *tc);
/*
* We can't hold rcu_read_lock() around code that can block. So we
* find a thin with the rcu lock held; bump a refcount; then drop
- * the lock.
+ * the lock. Skip the thins marked as dying or whose refcount thin_dtr()
+ * already dropped to zero.
*/
static struct thin_c *get_first_thin(struct pool *pool)
{
- struct thin_c *tc = NULL;
+ struct thin_c *tc;
rcu_read_lock();
- tc = list_first_or_null_rcu(&pool->active_thins, struct thin_c, list);
- if (tc)
- thin_get(tc);
+ list_for_each_entry_rcu(tc, &pool->active_thins, list) {
+ if (thin_tryget(tc)) {
+ rcu_read_unlock();
+ return tc;
+ }
+ }
rcu_read_unlock();
- return tc;
+ return NULL;
}
static struct thin_c *get_next_thin(struct pool *pool, struct thin_c *tc)
@@ -2345,7 +2357,8 @@ static struct thin_c *get_next_thin(struct pool *pool, struct thin_c *tc)
rcu_read_lock();
list_for_each_entry_continue_rcu(tc, &pool->active_thins, list) {
- thin_get(tc);
+ if (!thin_tryget(tc))
+ continue;
thin_put(old_tc);
rcu_read_unlock();
return tc;
@@ -2817,6 +2830,8 @@ static void requeue_bios(struct pool *pool)
rcu_read_lock();
list_for_each_entry_rcu(tc, &pool->active_thins, list) {
+ if (READ_ONCE(tc->dying))
+ continue;
spin_lock_irq(&tc->lock);
bio_list_merge(&tc->deferred_bio_list, &tc->retry_on_resume_list);
bio_list_init(&tc->retry_on_resume_list);
@@ -4131,9 +4146,9 @@ static struct target_type pool_target = {
* Thin target methods
*--------------------------------------------------------------
*/
-static void thin_get(struct thin_c *tc)
+static bool thin_tryget(struct thin_c *tc)
{
- refcount_inc(&tc->refcount);
+ return !READ_ONCE(tc->dying) && refcount_inc_not_zero(&tc->refcount);
}
static void thin_put(struct thin_c *tc)
@@ -4146,14 +4161,16 @@ static void thin_dtr(struct dm_target *ti)
{
struct thin_c *tc = ti->private;
+ WRITE_ONCE(tc->dying, true);
+
+ thin_put(tc);
+ wait_for_completion(&tc->can_destroy);
+
spin_lock_irq(&tc->pool->lock);
list_del_rcu(&tc->list);
spin_unlock_irq(&tc->pool->lock);
synchronize_rcu();
- thin_put(tc);
- wait_for_completion(&tc->can_destroy);
-
mutex_lock(&dm_thin_pool_table.mutex);
__pool_dec(tc->pool);
--
2.43.0
^ permalink raw reply related [flat|nested] 6+ messages in thread* [PATCH v3 3/3] dm thin: free thins with kfree_rcu() instead of waiting in thin_dtr()
2026-11-19 11:05 [PATCH v3 0/3] dm thin: stop waiting for grace periods when activating and removing thins Qiliang Yuan
2026-11-19 11:05 ` [PATCH v3 1/3] dm thin: drop the grace period from thin_ctr() Qiliang Yuan
2026-11-19 11:05 ` [PATCH v3 2/3] dm thin: keep a thin on active_thins until its last reference is dropped Qiliang Yuan
@ 2026-11-19 11:05 ` Qiliang Yuan
2026-11-19 19:20 ` Benjamin Marzinski
2 siblings, 1 reply; 6+ messages in thread
From: Qiliang Yuan @ 2026-11-19 11:05 UTC (permalink / raw)
To: Alasdair Kergon, Mike Snitzer, Mikulas Patocka,
Benjamin Marzinski
Cc: dm-devel, linux-kernel, Qiliang Yuan
thin_dtr() takes the thin off pool->active_thins once its last
reference is dropped and then waits for a grace period, so that
requeue_bios() and error_retry_list_with_code(), which walk the list
under rcu_read_lock(), are done with the thin before it is freed.
Removing a thin device thus waits for a full grace period in thin_dtr(),
although those readers only need the memory of the thin to stay around.
Free the thin with kfree_rcu() instead of waiting for the grace period.
Removing 100 thin devices of an idle pool with dmsetup remove takes per
device (mean of five rounds, min-max of the rounds in parentheses). The
queue freeze in del_gendisk() waits for a grace period of its own, which
the one in thin_dtr() used to overlap, so only part of it is saved:
before (ms) after (ms) change
17.2 (17.0-17.4) 13.7 (13.2-14.4) -20%
Signed-off-by: Qiliang Yuan <qiliang.yuan@linux.dev>
---
drivers/md/dm-thin.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/md/dm-thin.c b/drivers/md/dm-thin.c
index 62ef96bf08cb3..e695fc356c4dc 100644
--- a/drivers/md/dm-thin.c
+++ b/drivers/md/dm-thin.c
@@ -360,6 +360,7 @@ struct thin_c {
*/
refcount_t refcount;
struct completion can_destroy;
+ struct rcu_head rcu;
/*
* Keeps new active_thin walkers from grabbing the table while
* thin_dtr() is waiting for the existing holders to finish.
@@ -4169,7 +4170,6 @@ static void thin_dtr(struct dm_target *ti)
spin_lock_irq(&tc->pool->lock);
list_del_rcu(&tc->list);
spin_unlock_irq(&tc->pool->lock);
- synchronize_rcu();
mutex_lock(&dm_thin_pool_table.mutex);
@@ -4178,7 +4178,8 @@ static void thin_dtr(struct dm_target *ti)
dm_put_device(ti, tc->pool_dev);
if (tc->origin_dev)
dm_put_device(ti, tc->origin_dev);
- kfree(tc);
+ /* Let RCU readers of active_thins finish with tc before freeing it */
+ kfree_rcu(tc, rcu);
mutex_unlock(&dm_thin_pool_table.mutex);
}
--
2.43.0
^ permalink raw reply related [flat|nested] 6+ messages in thread