* FAILED: patch "[PATCH] jbd2: check need_resched() when skipping busy checkpoint" failed to apply to 5.15-stable tree
@ 2026-09-03 13:59 gregkh
2026-09-08 16:06 ` [PATCH 5.15.y 1/3] jbd2: use shrink_type type instead of bool type for __jbd2_journal_clean_checkpoint_list() Sasha Levin
0 siblings, 1 reply; 4+ messages in thread
From: gregkh @ 2026-09-03 13:59 UTC (permalink / raw)
To: max.kellermann, jack, tytso, yi.zhang; +Cc: stable
The patch below does not apply to the 5.15-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable@vger.kernel.org>.
To reproduce the conflict and resubmit, you may use the following commands:
git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.15.y
git checkout FETCH_HEAD
git cherry-pick -x f213e12ff5c9590b1034ae8da0e6d09665c772d0
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<stable@vger.kernel.org>' --in-reply-to '2026090337-absolve-bonnet-5b19@gregkh' --subject-prefix 'PATCH 5.15.y' 'HEAD^..'
Possible dependencies:
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From f213e12ff5c9590b1034ae8da0e6d09665c772d0 Mon Sep 17 00:00:00 2001
From: Max Kellermann <max.kellermann@ionos.com>
Date: Mon, 13 Jul 2026 12:22:28 +0200
Subject: [PATCH] jbd2: check need_resched() when skipping busy checkpoint
buffers
journal_shrink_one_cp_list() skips busy checkpoint buffers when called
with JBD2_SHRINK_BUSY_SKIP. The continue statement on this path also
skips the need_resched() check at the end of the loop body.
Consequently, when a checkpoint list contains mostly busy buffers, the
shrinker can walk the entire list while holding journal->j_list_lock,
even when a reschedule has been requested. Large checkpoint lists under
memory pressure can therefore cause long lock hold times and leave other
CPUs spinning on j_list_lock, resulting in soft lockups or RCU stalls.
Route the busy-buffer path through the need_resched() check so that the
shrinker can release j_list_lock and reschedule promptly, restoring
parity with the clean-buffer path, which already checks need_resched().
This does not change which checkpoint buffers are eligible for removal.
Fixes: b98dba273a0e ("jbd2: remove journal_clean_one_cp_list()")
Cc: stable@vger.kernel.org
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://patch.msgid.link/20260713102229.1598812-2-max.kellermann@ionos.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c
index 1508e2f54462..5266017565ac 100644
--- a/fs/jbd2/checkpoint.c
+++ b/fs/jbd2/checkpoint.c
@@ -389,7 +389,7 @@ static unsigned long journal_shrink_one_cp_list(struct journal_head *jh,
ret = jbd2_journal_try_remove_checkpoint(jh);
if (ret < 0) {
if (type == JBD2_SHRINK_BUSY_SKIP)
- continue;
+ goto next;
break;
}
}
@@ -400,6 +400,7 @@ static unsigned long journal_shrink_one_cp_list(struct journal_head *jh,
break;
}
+next:
if (need_resched())
break;
} while (jh != last_jh);
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH 5.15.y 1/3] jbd2: use shrink_type type instead of bool type for __jbd2_journal_clean_checkpoint_list()
2026-09-03 13:59 FAILED: patch "[PATCH] jbd2: check need_resched() when skipping busy checkpoint" failed to apply to 5.15-stable tree gregkh
@ 2026-09-08 16:06 ` Sasha Levin
2026-09-08 16:06 ` [PATCH 5.15.y 2/3] jbd2: add prefix 'jbd2' for 'shrink_type' Sasha Levin
2026-09-08 16:06 ` [PATCH 5.15.y 3/3] jbd2: check need_resched() when skipping busy checkpoint buffers Sasha Levin
0 siblings, 2 replies; 4+ messages in thread
From: Sasha Levin @ 2026-09-08 16:06 UTC (permalink / raw)
To: stable; +Cc: Ye Bin, Jan Kara, Zhang Yi, Theodore Ts'o, Sasha Levin
From: Ye Bin <yebin10@huawei.com>
[ Upstream commit 078760d950016f5982751f5512e69f26ad8feb31 ]
"enum shrink_type" can clearly express the meaning of the parameter of
__jbd2_journal_clean_checkpoint_list(), and there is no need to use the
bool type.
Signed-off-by: Ye Bin <yebin10@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Zhang Yi <yi.zhang@huawei.com>
Link: https://lore.kernel.org/r/20240407065355.1528580-2-yebin10@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Stable-dep-of: f213e12ff5c9 ("jbd2: check need_resched() when skipping busy checkpoint buffers")
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
fs/jbd2/checkpoint.c | 16 +++++++++-------
fs/jbd2/commit.c | 2 +-
include/linux/jbd2.h | 4 +++-
3 files changed, 13 insertions(+), 9 deletions(-)
diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c
index 10ce0b6b4ffa2..9fcc68884e262 100644
--- a/fs/jbd2/checkpoint.c
+++ b/fs/jbd2/checkpoint.c
@@ -367,8 +367,6 @@ int jbd2_cleanup_journal_tail(journal_t *journal)
/* Checkpoint list management */
-enum shrink_type {SHRINK_DESTROY, SHRINK_BUSY_STOP, SHRINK_BUSY_SKIP};
-
/*
* journal_shrink_one_cp_list
*
@@ -505,21 +503,25 @@ unsigned long jbd2_journal_shrink_checkpoint_list(journal_t *journal,
* journal_clean_checkpoint_list
*
* Find all the written-back checkpoint buffers in the journal and release them.
- * If 'destroy' is set, release all buffers unconditionally.
+ * If 'type' is SHRINK_DESTROY, release all buffers unconditionally. If 'type'
+ * is SHRINK_BUSY_STOP, will stop release buffers if encounters a busy buffer.
+ * To avoid wasting CPU cycles scanning the buffer list in some cases, don't
+ * pass SHRINK_BUSY_SKIP 'type' for this function.
*
* Called with j_list_lock held.
*/
-void __jbd2_journal_clean_checkpoint_list(journal_t *journal, bool destroy)
+void __jbd2_journal_clean_checkpoint_list(journal_t *journal,
+ enum shrink_type type)
{
transaction_t *transaction, *last_transaction, *next_transaction;
- enum shrink_type type;
bool released;
+ WARN_ON_ONCE(type == SHRINK_BUSY_SKIP);
+
transaction = journal->j_checkpoint_transactions;
if (!transaction)
return;
- type = destroy ? SHRINK_DESTROY : SHRINK_BUSY_STOP;
last_transaction = transaction->t_cpprev;
next_transaction = transaction;
do {
@@ -560,7 +562,7 @@ void jbd2_journal_destroy_checkpoint(journal_t *journal)
spin_unlock(&journal->j_list_lock);
break;
}
- __jbd2_journal_clean_checkpoint_list(journal, true);
+ __jbd2_journal_clean_checkpoint_list(journal, SHRINK_DESTROY);
spin_unlock(&journal->j_list_lock);
cond_resched();
}
diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index 9b2225ab5fe5a..182d2c39ff0a4 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -536,7 +536,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
* frees some memory
*/
spin_lock(&journal->j_list_lock);
- __jbd2_journal_clean_checkpoint_list(journal, false);
+ __jbd2_journal_clean_checkpoint_list(journal, SHRINK_BUSY_STOP);
spin_unlock(&journal->j_list_lock);
jbd2_debug(3, "JBD2: commit phase 1\n");
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index 5377df8c85453..5583d735d4bf3 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -1435,7 +1435,9 @@ void jbd2_update_log_tail(journal_t *journal, tid_t tid, unsigned long block);
extern void jbd2_journal_commit_transaction(journal_t *);
/* Checkpoint list management */
-void __jbd2_journal_clean_checkpoint_list(journal_t *journal, bool destroy);
+enum shrink_type {SHRINK_DESTROY, SHRINK_BUSY_STOP, SHRINK_BUSY_SKIP};
+
+void __jbd2_journal_clean_checkpoint_list(journal_t *journal, enum shrink_type type);
unsigned long jbd2_journal_shrink_checkpoint_list(journal_t *journal, unsigned long *nr_to_scan);
int __jbd2_journal_remove_checkpoint(struct journal_head *);
int jbd2_journal_try_remove_checkpoint(struct journal_head *jh);
--
2.53.0
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH 5.15.y 2/3] jbd2: add prefix 'jbd2' for 'shrink_type'
2026-09-08 16:06 ` [PATCH 5.15.y 1/3] jbd2: use shrink_type type instead of bool type for __jbd2_journal_clean_checkpoint_list() Sasha Levin
@ 2026-09-08 16:06 ` Sasha Levin
2026-09-08 16:06 ` [PATCH 5.15.y 3/3] jbd2: check need_resched() when skipping busy checkpoint buffers Sasha Levin
1 sibling, 0 replies; 4+ messages in thread
From: Sasha Levin @ 2026-09-08 16:06 UTC (permalink / raw)
To: stable; +Cc: Ye Bin, Jan Kara, Zhang Yi, Theodore Ts'o, Sasha Levin
From: Ye Bin <yebin10@huawei.com>
[ Upstream commit 26770a717cac57041d9414725e3e01dd19b08dd2 ]
As 'shrink_type' is exported. The module prefix 'jbd2' is added to
distinguish from memory reclamation.
Signed-off-by: Ye Bin <yebin10@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Zhang Yi <yi.zhang@huawei.com>
Link: https://lore.kernel.org/r/20240407065355.1528580-3-yebin10@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Stable-dep-of: f213e12ff5c9 ("jbd2: check need_resched() when skipping busy checkpoint buffers")
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
fs/jbd2/checkpoint.c | 22 +++++++++++-----------
fs/jbd2/commit.c | 2 +-
include/linux/jbd2.h | 4 ++--
3 files changed, 14 insertions(+), 14 deletions(-)
diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c
index 9fcc68884e262..36c64cf44e005 100644
--- a/fs/jbd2/checkpoint.c
+++ b/fs/jbd2/checkpoint.c
@@ -378,7 +378,7 @@ int jbd2_cleanup_journal_tail(journal_t *journal)
* Called with j_list_lock held.
*/
static unsigned long journal_shrink_one_cp_list(struct journal_head *jh,
- enum shrink_type type,
+ enum jbd2_shrink_type type,
bool *released)
{
struct journal_head *last_jh;
@@ -395,12 +395,12 @@ static unsigned long journal_shrink_one_cp_list(struct journal_head *jh,
jh = next_jh;
next_jh = jh->b_cpnext;
- if (type == SHRINK_DESTROY) {
+ if (type == JBD2_SHRINK_DESTROY) {
ret = __jbd2_journal_remove_checkpoint(jh);
} else {
ret = jbd2_journal_try_remove_checkpoint(jh);
if (ret < 0) {
- if (type == SHRINK_BUSY_SKIP)
+ if (type == JBD2_SHRINK_BUSY_SKIP)
continue;
break;
}
@@ -470,7 +470,7 @@ unsigned long jbd2_journal_shrink_checkpoint_list(journal_t *journal,
tid = transaction->t_tid;
freed = journal_shrink_one_cp_list(transaction->t_checkpoint_list,
- SHRINK_BUSY_SKIP, &released);
+ JBD2_SHRINK_BUSY_SKIP, &released);
nr_freed += freed;
(*nr_to_scan) -= min(*nr_to_scan, freed);
if (*nr_to_scan == 0)
@@ -503,20 +503,20 @@ unsigned long jbd2_journal_shrink_checkpoint_list(journal_t *journal,
* journal_clean_checkpoint_list
*
* Find all the written-back checkpoint buffers in the journal and release them.
- * If 'type' is SHRINK_DESTROY, release all buffers unconditionally. If 'type'
- * is SHRINK_BUSY_STOP, will stop release buffers if encounters a busy buffer.
- * To avoid wasting CPU cycles scanning the buffer list in some cases, don't
- * pass SHRINK_BUSY_SKIP 'type' for this function.
+ * If 'type' is JBD2_SHRINK_DESTROY, release all buffers unconditionally. If
+ * 'type' is JBD2_SHRINK_BUSY_STOP, will stop release buffers if encounters a
+ * busy buffer. To avoid wasting CPU cycles scanning the buffer list in some
+ * cases, don't pass JBD2_SHRINK_BUSY_SKIP 'type' for this function.
*
* Called with j_list_lock held.
*/
void __jbd2_journal_clean_checkpoint_list(journal_t *journal,
- enum shrink_type type)
+ enum jbd2_shrink_type type)
{
transaction_t *transaction, *last_transaction, *next_transaction;
bool released;
- WARN_ON_ONCE(type == SHRINK_BUSY_SKIP);
+ WARN_ON_ONCE(type == JBD2_SHRINK_BUSY_SKIP);
transaction = journal->j_checkpoint_transactions;
if (!transaction)
@@ -562,7 +562,7 @@ void jbd2_journal_destroy_checkpoint(journal_t *journal)
spin_unlock(&journal->j_list_lock);
break;
}
- __jbd2_journal_clean_checkpoint_list(journal, SHRINK_DESTROY);
+ __jbd2_journal_clean_checkpoint_list(journal, JBD2_SHRINK_DESTROY);
spin_unlock(&journal->j_list_lock);
cond_resched();
}
diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index 182d2c39ff0a4..53908eb74c58f 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -536,7 +536,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
* frees some memory
*/
spin_lock(&journal->j_list_lock);
- __jbd2_journal_clean_checkpoint_list(journal, SHRINK_BUSY_STOP);
+ __jbd2_journal_clean_checkpoint_list(journal, JBD2_SHRINK_BUSY_STOP);
spin_unlock(&journal->j_list_lock);
jbd2_debug(3, "JBD2: commit phase 1\n");
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index 5583d735d4bf3..221e7dc87a296 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -1435,9 +1435,9 @@ void jbd2_update_log_tail(journal_t *journal, tid_t tid, unsigned long block);
extern void jbd2_journal_commit_transaction(journal_t *);
/* Checkpoint list management */
-enum shrink_type {SHRINK_DESTROY, SHRINK_BUSY_STOP, SHRINK_BUSY_SKIP};
+enum jbd2_shrink_type {JBD2_SHRINK_DESTROY, JBD2_SHRINK_BUSY_STOP, JBD2_SHRINK_BUSY_SKIP};
-void __jbd2_journal_clean_checkpoint_list(journal_t *journal, enum shrink_type type);
+void __jbd2_journal_clean_checkpoint_list(journal_t *journal, enum jbd2_shrink_type type);
unsigned long jbd2_journal_shrink_checkpoint_list(journal_t *journal, unsigned long *nr_to_scan);
int __jbd2_journal_remove_checkpoint(struct journal_head *);
int jbd2_journal_try_remove_checkpoint(struct journal_head *jh);
--
2.53.0
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH 5.15.y 3/3] jbd2: check need_resched() when skipping busy checkpoint buffers
2026-09-08 16:06 ` [PATCH 5.15.y 1/3] jbd2: use shrink_type type instead of bool type for __jbd2_journal_clean_checkpoint_list() Sasha Levin
2026-09-08 16:06 ` [PATCH 5.15.y 2/3] jbd2: add prefix 'jbd2' for 'shrink_type' Sasha Levin
@ 2026-09-08 16:06 ` Sasha Levin
1 sibling, 0 replies; 4+ messages in thread
From: Sasha Levin @ 2026-09-08 16:06 UTC (permalink / raw)
To: stable; +Cc: Max Kellermann, Zhang Yi, Jan Kara, Theodore Ts'o,
Sasha Levin
From: Max Kellermann <max.kellermann@ionos.com>
[ Upstream commit f213e12ff5c9590b1034ae8da0e6d09665c772d0 ]
journal_shrink_one_cp_list() skips busy checkpoint buffers when called
with JBD2_SHRINK_BUSY_SKIP. The continue statement on this path also
skips the need_resched() check at the end of the loop body.
Consequently, when a checkpoint list contains mostly busy buffers, the
shrinker can walk the entire list while holding journal->j_list_lock,
even when a reschedule has been requested. Large checkpoint lists under
memory pressure can therefore cause long lock hold times and leave other
CPUs spinning on j_list_lock, resulting in soft lockups or RCU stalls.
Route the busy-buffer path through the need_resched() check so that the
shrinker can release j_list_lock and reschedule promptly, restoring
parity with the clean-buffer path, which already checks need_resched().
This does not change which checkpoint buffers are eligible for removal.
Fixes: b98dba273a0e ("jbd2: remove journal_clean_one_cp_list()")
Cc: stable@vger.kernel.org
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Zhang Yi <yi.zhang@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://patch.msgid.link/20260713102229.1598812-2-max.kellermann@ionos.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
fs/jbd2/checkpoint.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c
index 36c64cf44e005..a888357a29964 100644
--- a/fs/jbd2/checkpoint.c
+++ b/fs/jbd2/checkpoint.c
@@ -401,7 +401,7 @@ static unsigned long journal_shrink_one_cp_list(struct journal_head *jh,
ret = jbd2_journal_try_remove_checkpoint(jh);
if (ret < 0) {
if (type == JBD2_SHRINK_BUSY_SKIP)
- continue;
+ goto next;
break;
}
}
@@ -412,6 +412,7 @@ static unsigned long journal_shrink_one_cp_list(struct journal_head *jh,
break;
}
+next:
if (need_resched())
break;
} while (jh != last_jh);
--
2.53.0
^ permalink raw reply related [flat|nested] 4+ messages in thread
end of thread, other threads:[~2026-09-08 16:06 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-09-03 13:59 FAILED: patch "[PATCH] jbd2: check need_resched() when skipping busy checkpoint" failed to apply to 5.15-stable tree gregkh
2026-09-08 16:06 ` [PATCH 5.15.y 1/3] jbd2: use shrink_type type instead of bool type for __jbd2_journal_clean_checkpoint_list() Sasha Levin
2026-09-08 16:06 ` [PATCH 5.15.y 2/3] jbd2: add prefix 'jbd2' for 'shrink_type' Sasha Levin
2026-09-08 16:06 ` [PATCH 5.15.y 3/3] jbd2: check need_resched() when skipping busy checkpoint buffers Sasha Levin
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.