[syzbot] [cgroups?] BUG: sleeping function called from invalid context in cgroup_rstat_flush

[syzbot] [cgroups?] BUG: sleeping function called from invalid context in cgroup_rstat_flush

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit:    7c16f0a4ed1c Merge tag 'i2c-for-6.10-rc5' of git://git.ker..
git tree:       upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=1511528e980000
kernel config:  https://syzkaller.appspot.com/x/.config?x=12f98862a3c0c799
dashboard link: https://syzkaller.appspot.com/bug?extid=b7f13b2d0cc156edf61a
compiler:       Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40

Unfortunately, I don't have any reproducer for this issue yet.

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/50560e9024e5/disk-7c16f0a4.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/080c27daee72/vmlinux-7c16f0a4.xz
kernel image: https://storage.googleapis.com/syzbot-assets/c528e0da4544/bzImage-7c16f0a4.xz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+b7f13b2d0cc156edf61a@syzkaller.appspotmail.com

BUG: sleeping function called from invalid context at kernel/cgroup/rstat.c:351
in_atomic(): 0, irqs_disabled(): 0, non_block: 0, pid: 17332, name: syz-executor.4
preempt_count: 0, expected: 0
RCU nest depth: 1, expected: 0
1 lock held by syz-executor.4/17332:
 #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: rcu_lock_acquire include/linux/rcupdate.h:329 [inline]
 #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: rcu_read_lock include/linux/rcupdate.h:781 [inline]
 #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: filemap_cachestat mm/filemap.c:4251 [inline]
 #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: __do_sys_cachestat mm/filemap.c:4407 [inline]
 #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: __se_sys_cachestat+0x3ee/0xbb0 mm/filemap.c:4372
CPU: 1 PID: 17332 Comm: syz-executor.4 Not tainted 6.10.0-rc4-syzkaller-00330-g7c16f0a4ed1c #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 06/07/2024
Call Trace:
 <TASK>
 __dump_stack lib/dump_stack.c:88 [inline]
 dump_stack_lvl+0x241/0x360 lib/dump_stack.c:114
 __might_resched+0x5d4/0x780 kernel/sched/core.c:10196
 cgroup_rstat_flush+0x1e/0x50 kernel/cgroup/rstat.c:351
 workingset_test_recent+0x48a/0xa90 mm/workingset.c:473
 filemap_cachestat mm/filemap.c:4314 [inline]
 __do_sys_cachestat mm/filemap.c:4407 [inline]
 __se_sys_cachestat+0x795/0xbb0 mm/filemap.c:4372
 do_syscall_x64 arch/x86/entry/common.c:52 [inline]
 do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
 entry_SYSCALL_64_after_hwframe+0x77/0x7f
RIP: 0033:0x7ff329e7d0a9
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 e1 20 00 00 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b0 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007ff32ac600c8 EFLAGS: 00000246 ORIG_RAX: 00000000000001c3
RAX: ffffffffffffffda RBX: 00007ff329fb4120 RCX: 00007ff329e7d0a9
RDX: 0000000020000080 RSI: 0000000020000040 RDI: 0000000000000005
RBP: 00007ff329eec074 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 000000000000006e R14: 00007ff329fb4120 R15: 00007ffd3e0ff4a8
 </TASK>


---
This report is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzkaller@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup

Re: [syzbot] [cgroups?] BUG: sleeping function called from invalid context in cgroup_rstat_flush

[Johannes Weiner]

On Thu, Jun 27, 2024 at 07:03:21AM -0700, syzbot wrote:
> Hello,
> 
> syzbot found the following issue on:
> 
> HEAD commit:    7c16f0a4ed1c Merge tag 'i2c-for-6.10-rc5' of git://git.ker..
> git tree:       upstream
> console output: https://syzkaller.appspot.com/x/log.txt?x=1511528e980000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=12f98862a3c0c799
> dashboard link: https://syzkaller.appspot.com/bug?extid=b7f13b2d0cc156edf61a
> compiler:       Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40
> 
> Unfortunately, I don't have any reproducer for this issue yet.
> 
> Downloadable assets:
> disk image: https://storage.googleapis.com/syzbot-assets/50560e9024e5/disk-7c16f0a4.raw.xz
> vmlinux: https://storage.googleapis.com/syzbot-assets/080c27daee72/vmlinux-7c16f0a4.xz
> kernel image: https://storage.googleapis.com/syzbot-assets/c528e0da4544/bzImage-7c16f0a4.xz
> 
> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> Reported-by: syzbot+b7f13b2d0cc156edf61a@syzkaller.appspotmail.com
> 
> BUG: sleeping function called from invalid context at kernel/cgroup/rstat.c:351
> in_atomic(): 0, irqs_disabled(): 0, non_block: 0, pid: 17332, name: syz-executor.4
> preempt_count: 0, expected: 0
> RCU nest depth: 1, expected: 0
> 1 lock held by syz-executor.4/17332:
>  #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: rcu_lock_acquire include/linux/rcupdate.h:329 [inline]
>  #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: rcu_read_lock include/linux/rcupdate.h:781 [inline]
>  #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: filemap_cachestat mm/filemap.c:4251 [inline]
>  #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: __do_sys_cachestat mm/filemap.c:4407 [inline]
>  #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: __se_sys_cachestat+0x3ee/0xbb0 mm/filemap.c:4372
> CPU: 1 PID: 17332 Comm: syz-executor.4 Not tainted 6.10.0-rc4-syzkaller-00330-g7c16f0a4ed1c #0
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 06/07/2024
> Call Trace:
>  <TASK>
>  __dump_stack lib/dump_stack.c:88 [inline]
>  dump_stack_lvl+0x241/0x360 lib/dump_stack.c:114
>  __might_resched+0x5d4/0x780 kernel/sched/core.c:10196
>  cgroup_rstat_flush+0x1e/0x50 kernel/cgroup/rstat.c:351
>  workingset_test_recent+0x48a/0xa90 mm/workingset.c:473
>  filemap_cachestat mm/filemap.c:4314 [inline]
>  __do_sys_cachestat mm/filemap.c:4407 [inline]
>  __se_sys_cachestat+0x795/0xbb0 mm/filemap.c:4372
>  do_syscall_x64 arch/x86/entry/common.c:52 [inline]
>  do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
>  entry_SYSCALL_64_after_hwframe+0x77/0x7f

Ok yeah, cachestat() holds the rcu read lock, so
workingset_test_recent() can't do a sleepable rstat flush.

I think the easiest fix would be to flush rstat from the root down
(NULL) in filemap_cachestat(), before the rcu section, and add a flag
to workingset_test_recent() to forego it. Nhat?

Re: [syzbot] [cgroups?] BUG: sleeping function called from invalid context in cgroup_rstat_flush

[Nhat Pham]

On Thu, Jun 27, 2024 at 9:31 AM Johannes Weiner <hannes@cmpxchg.org> wrote:
>
> On Thu, Jun 27, 2024 at 07:03:21AM -0700, syzbot wrote:
> > Hello,
> >
> > syzbot found the following issue on:
> >
> > HEAD commit:    7c16f0a4ed1c Merge tag 'i2c-for-6.10-rc5' of git://git.ker..
> > git tree:       upstream
> > console output: https://syzkaller.appspot.com/x/log.txt?x=1511528e980000
> > kernel config:  https://syzkaller.appspot.com/x/.config?x=12f98862a3c0c799
> > dashboard link: https://syzkaller.appspot.com/bug?extid=b7f13b2d0cc156edf61a
> > compiler:       Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40
> >
> > Unfortunately, I don't have any reproducer for this issue yet.
> >
> > Downloadable assets:
> > disk image: https://storage.googleapis.com/syzbot-assets/50560e9024e5/disk-7c16f0a4.raw.xz
> > vmlinux: https://storage.googleapis.com/syzbot-assets/080c27daee72/vmlinux-7c16f0a4.xz
> > kernel image: https://storage.googleapis.com/syzbot-assets/c528e0da4544/bzImage-7c16f0a4.xz
> >
> > IMPORTANT: if you fix the issue, please add the following tag to the commit:
> > Reported-by: syzbot+b7f13b2d0cc156edf61a@syzkaller.appspotmail.com
> >
> > BUG: sleeping function called from invalid context at kernel/cgroup/rstat.c:351
> > in_atomic(): 0, irqs_disabled(): 0, non_block: 0, pid: 17332, name: syz-executor.4
> > preempt_count: 0, expected: 0
> > RCU nest depth: 1, expected: 0
> > 1 lock held by syz-executor.4/17332:
> >  #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: rcu_lock_acquire include/linux/rcupdate.h:329 [inline]
> >  #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: rcu_read_lock include/linux/rcupdate.h:781 [inline]
> >  #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: filemap_cachestat mm/filemap.c:4251 [inline]
> >  #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: __do_sys_cachestat mm/filemap.c:4407 [inline]
> >  #0: ffffffff8e333fa0 (rcu_read_lock){....}-{1:2}, at: __se_sys_cachestat+0x3ee/0xbb0 mm/filemap.c:4372
> > CPU: 1 PID: 17332 Comm: syz-executor.4 Not tainted 6.10.0-rc4-syzkaller-00330-g7c16f0a4ed1c #0
> > Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 06/07/2024
> > Call Trace:
> >  <TASK>
> >  __dump_stack lib/dump_stack.c:88 [inline]
> >  dump_stack_lvl+0x241/0x360 lib/dump_stack.c:114
> >  __might_resched+0x5d4/0x780 kernel/sched/core.c:10196
> >  cgroup_rstat_flush+0x1e/0x50 kernel/cgroup/rstat.c:351
> >  workingset_test_recent+0x48a/0xa90 mm/workingset.c:473
> >  filemap_cachestat mm/filemap.c:4314 [inline]
> >  __do_sys_cachestat mm/filemap.c:4407 [inline]
> >  __se_sys_cachestat+0x795/0xbb0 mm/filemap.c:4372
> >  do_syscall_x64 arch/x86/entry/common.c:52 [inline]
> >  do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
> >  entry_SYSCALL_64_after_hwframe+0x77/0x7f
>
> Ok yeah, cachestat() holds the rcu read lock, so
> workingset_test_recent() can't do a sleepable rstat flush.
>
> I think the easiest fix would be to flush rstat from the root down
> (NULL) in filemap_cachestat(), before the rcu section, and add a flag
> to workingset_test_recent() to forego it. Nhat?

You're right. I think it's been broken since this commit:

b00684722262 mm: workingset: move the stats flush into workingset_test_recent()

which moves the stats flushing from the refault step (before rcu read
lock section) to inside workingset_test_recent(). I believe that's
6.8, 6.9, and 6.10 we need to fix?

The fix sounds reasonable to me :) Let me whip up something real quick.

[PATCH] cachestat: do not flush stats in recency check

[Nhat Pham]

syzbot detects that cachestat() is flushing stats, which can sleep, in
its RCU read section (see [1]). This is done in the
workingset_test_recent() step (which checks if the folio's eviction is
recent).

Move the stat flushing step to before the RCU read section of cachestat,
and skip stat flushing during the recency check.

[1]: https://lore.kernel.org/cgroups/000000000000f71227061bdf97e0@google.com/

Reported-by: syzbot+b7f13b2d0cc156edf61a@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/cgroups/000000000000f71227061bdf97e0@google.com/
Debugged-by: Johannes Weiner <hannes@cmpxchg.org>
Suggested-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Nhat Pham <nphamcs@gmail.com>
Fixes: b00684722262 ("mm: workingset: move the stats flush into workingset_test_recent()")
Cc: stable@vger.kernel.org # v6.8+
---
 include/linux/swap.h |  3 ++-
 mm/filemap.c         |  5 ++++-
 mm/workingset.c      | 14 +++++++++++---
 3 files changed, 17 insertions(+), 5 deletions(-)

diff --git a/include/linux/swap.h b/include/linux/swap.h
index bd450023b9a4..e685e93ba354 100644
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -354,7 +354,8 @@ static inline swp_entry_t page_swap_entry(struct page *page)
 }
 
 /* linux/mm/workingset.c */
-bool workingset_test_recent(void *shadow, bool file, bool *workingset);
+bool workingset_test_recent(void *shadow, bool file, bool *workingset,
+				bool flush);
 void workingset_age_nonresident(struct lruvec *lruvec, unsigned long nr_pages);
 void *workingset_eviction(struct folio *folio, struct mem_cgroup *target_memcg);
 void workingset_refault(struct folio *folio, void *shadow);
diff --git a/mm/filemap.c b/mm/filemap.c
index fedefb10d947..298485d4b992 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -4248,6 +4248,9 @@ static void filemap_cachestat(struct address_space *mapping,
 	XA_STATE(xas, &mapping->i_pages, first_index);
 	struct folio *folio;
 
+	/* Flush stats (and potentially sleep) outside the RCU read section. */
+	mem_cgroup_flush_stats_ratelimited(NULL);
+
 	rcu_read_lock();
 	xas_for_each(&xas, folio, last_index) {
 		int order;
@@ -4311,7 +4314,7 @@ static void filemap_cachestat(struct address_space *mapping,
 					goto resched;
 			}
 #endif
-			if (workingset_test_recent(shadow, true, &workingset))
+			if (workingset_test_recent(shadow, true, &workingset, false))
 				cs->nr_recently_evicted += nr_pages;
 
 			goto resched;
diff --git a/mm/workingset.c b/mm/workingset.c
index c22adb93622a..a2b28e356e68 100644
--- a/mm/workingset.c
+++ b/mm/workingset.c
@@ -412,10 +412,12 @@ void *workingset_eviction(struct folio *folio, struct mem_cgroup *target_memcg)
  * @file: whether the corresponding folio is from the file lru.
  * @workingset: where the workingset value unpacked from shadow should
  * be stored.
+ * @flush: whether to flush cgroup rstat.
  *
  * Return: true if the shadow is for a recently evicted folio; false otherwise.
  */
-bool workingset_test_recent(void *shadow, bool file, bool *workingset)
+bool workingset_test_recent(void *shadow, bool file, bool *workingset,
+				bool flush)
 {
 	struct mem_cgroup *eviction_memcg;
 	struct lruvec *eviction_lruvec;
@@ -467,10 +469,16 @@ bool workingset_test_recent(void *shadow, bool file, bool *workingset)
 
 	/*
 	 * Flush stats (and potentially sleep) outside the RCU read section.
+	 *
+	 * Note that workingset_test_recent() itself might be called in RCU read
+	 * section (for e.g, in cachestat) - these callers need to skip flushing
+	 * stats (via the flush argument).
+	 *
 	 * XXX: With per-memcg flushing and thresholding, is ratelimiting
 	 * still needed here?
 	 */
-	mem_cgroup_flush_stats_ratelimited(eviction_memcg);
+	if (flush)
+		mem_cgroup_flush_stats_ratelimited(eviction_memcg);
 
 	eviction_lruvec = mem_cgroup_lruvec(eviction_memcg, pgdat);
 	refault = atomic_long_read(&eviction_lruvec->nonresident_age);
@@ -558,7 +566,7 @@ void workingset_refault(struct folio *folio, void *shadow)
 
 	mod_lruvec_state(lruvec, WORKINGSET_REFAULT_BASE + file, nr);
 
-	if (!workingset_test_recent(shadow, file, &workingset))
+	if (!workingset_test_recent(shadow, file, &workingset, true))
 		return;
 
 	folio_set_active(folio);

base-commit: a5c6fededf806aba1ff9b0f01278f7d089da5725
-- 
2.43.0

Re: [PATCH] cachestat: do not flush stats in recency check

[Johannes Weiner]

On Thu, Jun 27, 2024 at 01:17:37PM -0700, Nhat Pham wrote:
> syzbot detects that cachestat() is flushing stats, which can sleep, in
> its RCU read section (see [1]). This is done in the
> workingset_test_recent() step (which checks if the folio's eviction is
> recent).
> 
> Move the stat flushing step to before the RCU read section of cachestat,
> and skip stat flushing during the recency check.
> 
> [1]: https://lore.kernel.org/cgroups/000000000000f71227061bdf97e0@google.com/
> 
> Reported-by: syzbot+b7f13b2d0cc156edf61a@syzkaller.appspotmail.com
> Closes: https://lore.kernel.org/cgroups/000000000000f71227061bdf97e0@google.com/
> Debugged-by: Johannes Weiner <hannes@cmpxchg.org>
> Suggested-by: Johannes Weiner <hannes@cmpxchg.org>
> Signed-off-by: Nhat Pham <nphamcs@gmail.com>
> Fixes: b00684722262 ("mm: workingset: move the stats flush into workingset_test_recent()")
> Cc: stable@vger.kernel.org # v6.8+

Acked-by: Johannes Weiner <hannes@cmpxchg.org>

Re: [PATCH] cachestat: do not flush stats in recency check

[Shakeel Butt]

On Thu, Jun 27, 2024 at 01:17:37PM GMT, Nhat Pham wrote:
> syzbot detects that cachestat() is flushing stats, which can sleep, in
> its RCU read section (see [1]). This is done in the
> workingset_test_recent() step (which checks if the folio's eviction is
> recent).
> 
> Move the stat flushing step to before the RCU read section of cachestat,
> and skip stat flushing during the recency check.
> 
> [1]: https://lore.kernel.org/cgroups/000000000000f71227061bdf97e0@google.com/
> 
> Reported-by: syzbot+b7f13b2d0cc156edf61a@syzkaller.appspotmail.com
> Closes: https://lore.kernel.org/cgroups/000000000000f71227061bdf97e0@google.com/
> Debugged-by: Johannes Weiner <hannes@cmpxchg.org>
> Suggested-by: Johannes Weiner <hannes@cmpxchg.org>
> Signed-off-by: Nhat Pham <nphamcs@gmail.com>
> Fixes: b00684722262 ("mm: workingset: move the stats flush into workingset_test_recent()")
> Cc: stable@vger.kernel.org # v6.8+

Acked-by: Shakeel Butt <shakeel.butt@linux.dev>