]> git.ipfire.org Git - thirdparty/linux.git/commitdiff
nilfs2: avoid having an active sc_timer before freeing sci
authorEdward Adam Davis <eadavis@qq.com>
Wed, 29 Oct 2025 22:51:52 +0000 (07:51 +0900)
committerAndrew Morton <akpm@linux-foundation.org>
Mon, 10 Nov 2025 05:19:46 +0000 (21:19 -0800)
Because kthread_stop did not stop sc_task properly and returned -EINTR,
the sc_timer was not properly closed, ultimately causing the problem [1]
reported by syzbot when freeing sci due to the sc_timer not being closed.

Because the thread sc_task main function nilfs_segctor_thread() returns 0
when it succeeds, when the return value of kthread_stop() is not 0 in
nilfs_segctor_destroy(), we believe that it has not properly closed
sc_timer.

We use timer_shutdown_sync() to sync wait for sc_timer to shutdown, and
set the value of sc_task to NULL under the protection of lock
sc_state_lock, so as to avoid the issue caused by sc_timer not being
properly shutdowned.

[1]
ODEBUG: free active (active state 0) object: 00000000dacb411a object type: timer_list hint: nilfs_construction_timeout
Call trace:
 nilfs_segctor_destroy fs/nilfs2/segment.c:2811 [inline]
 nilfs_detach_log_writer+0x668/0x8cc fs/nilfs2/segment.c:2877
 nilfs_put_super+0x4c/0x12c fs/nilfs2/super.c:509

Link: https://lkml.kernel.org/r/20251029225226.16044-1-konishi.ryusuke@gmail.com
Fixes: 3f66cc261ccb ("nilfs2: use kthread_create and kthread_stop for the log writer thread")
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@gmail.com>
Reported-by: syzbot+24d8b70f039151f65590@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=24d8b70f039151f65590
Tested-by: syzbot+24d8b70f039151f65590@syzkaller.appspotmail.com
Signed-off-by: Edward Adam Davis <eadavis@qq.com>
Cc: <stable@vger.kernel.org> [6.12+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
fs/nilfs2/segment.c

index f15ca6fc400d19ab39699cb948b7a7e52285c24c..deee16bc9d4eb20823188610d2715f42fdfce0ec 100644 (file)
@@ -2768,7 +2768,12 @@ static void nilfs_segctor_destroy(struct nilfs_sc_info *sci)
 
        if (sci->sc_task) {
                wake_up(&sci->sc_wait_daemon);
-               kthread_stop(sci->sc_task);
+               if (kthread_stop(sci->sc_task)) {
+                       spin_lock(&sci->sc_state_lock);
+                       sci->sc_task = NULL;
+                       timer_shutdown_sync(&sci->sc_timer);
+                       spin_unlock(&sci->sc_state_lock);
+               }
        }
 
        spin_lock(&sci->sc_state_lock);