The abort_on_kill path in request_wait_answer() calls fuse_abort_conn()
and returns without waiting for FR_FINISHED.  If fuse_dev_do_write() is
concurrently processing the same request (FR_LOCKED set), the caller
frees req->args while it is still being accessed, causing a
use-after-free.

Fix this by jumping to the existing wait_event(FR_FINISHED) instead of
returning early.  The wait will not hang because fuse_abort_conn()
ensures all requests are ended.

Reported-by: syzbot+d6540a3fa1626e11360d@syzkaller.appspotmail.com
Fixes: 204aa22a686b ("fuse: abort on fatal signal during sync init")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-6
Signed-off-by: Rochan Avlur <rochan.avlur@gmail.com>
---
Tested by injecting a msleep(100) between spin_unlock(&fpq->lock) and
the req->args->page_replace access in fuse_dev_do_write() to widen the
race window, then running a reproducer that concurrently writes the INIT
reply while SIGKILL is delivered to the mount process.  KASAN reliably
reports the slab-use-after-free without the fix, and the report is absent
with the fix applied.

 fs/fuse/dev.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c
index c105aaf9ff5d..b946884d4d5b 100644
--- a/fs/fuse/dev.c
+++ b/fs/fuse/dev.c
@@ -572,7 +572,7 @@ static void request_wait_answer(struct fuse_req *req)
 
 		if (req->args->abort_on_kill) {
 			fuse_abort_conn(fc);
-			return;
+			goto wait_for_finish;
 		}
 
 		if (test_bit(FR_URING, &req->flags))
@@ -583,6 +583,7 @@ static void request_wait_answer(struct fuse_req *req)
 			return;
 	}
 
+wait_for_finish:
 	/*
 	 * Either request is already in userspace, or it was forced.
 	 * Wait it out.
-- 
2.45.2

On Thu, 13 Aug 2026 at 05:59, Rochan Avlur <rochan.avlur@gmail.com> wrote:
>
> The abort_on_kill path in request_wait_answer() calls fuse_abort_conn()
> and returns without waiting for FR_FINISHED.  If fuse_dev_do_write() is
> concurrently processing the same request (FR_LOCKED set), the caller
> frees req->args while it is still being accessed, causing a
> use-after-free.
>
> Fix this by jumping to the existing wait_event(FR_FINISHED) instead of
> returning early.  The wait will not hang because fuse_abort_conn()
> ensures all requests are ended.
>
> Reported-by: syzbot+d6540a3fa1626e11360d@syzkaller.appspotmail.com
> Fixes: 204aa22a686b ("fuse: abort on fatal signal during sync init")
> Cc: stable@vger.kernel.org
> Assisted-by: Claude:claude-opus-4-6
> Signed-off-by: Rochan Avlur <rochan.avlur@gmail.com>

Applied, thanks.

Miklos