[PATCH v5.10.y 1/1] io_uring: don't take uring_lock during iowq cancel

[Lee Jones <lee.jones@linaro.org>]

From: Pavel Begunkov <asml.silence@gmail.com>

[ Upstream commit 792bb6eb862333658bf1bd2260133f0507e2da8d ]

[   97.866748] a.out/2890 is trying to acquire lock:
[   97.867829] ffff8881046763e8 (&ctx->uring_lock){+.+.}-{3:3}, at:
io_wq_submit_work+0x155/0x240
[   97.869735]
[   97.869735] but task is already holding lock:
[   97.871033] ffff88810dfe0be8 (&ctx->uring_lock){+.+.}-{3:3}, at:
__x64_sys_io_uring_enter+0x3f0/0x5b0
[   97.873074]
[   97.873074] other info that might help us debug this:
[   97.874520]  Possible unsafe locking scenario:
[   97.874520]
[   97.875845]        CPU0
[   97.876440]        ----
[   97.877048]   lock(&ctx->uring_lock);
[   97.877961]   lock(&ctx->uring_lock);
[   97.878881]
[   97.878881]  *** DEADLOCK ***
[   97.878881]
[   97.880341]  May be due to missing lock nesting notation
[   97.880341]
[   97.881952] 1 lock held by a.out/2890:
[   97.882873]  #0: ffff88810dfe0be8 (&ctx->uring_lock){+.+.}-{3:3}, at:
__x64_sys_io_uring_enter+0x3f0/0x5b0
[   97.885108]
[   97.885108] stack backtrace:
[   97.890457] Call Trace:
[   97.891121]  dump_stack+0xac/0xe3
[   97.891972]  __lock_acquire+0xab6/0x13a0
[   97.892940]  lock_acquire+0x2c3/0x390
[   97.894894]  __mutex_lock+0xae/0x9f0
[   97.901101]  io_wq_submit_work+0x155/0x240
[   97.902112]  io_wq_cancel_cb+0x162/0x490
[   97.904126]  io_async_find_and_cancel+0x3b/0x140
[   97.905247]  io_issue_sqe+0x86d/0x13e0
[   97.909122]  __io_queue_sqe+0x10b/0x550
[   97.913971]  io_queue_sqe+0x235/0x470
[   97.914894]  io_submit_sqes+0xcce/0xf10
[   97.917872]  __x64_sys_io_uring_enter+0x3fb/0x5b0
[   97.921424]  do_syscall_64+0x2d/0x40
[   97.922329]  entry_SYSCALL_64_after_hwframe+0x44/0xa9

While holding uring_lock, e.g. from inline execution, async cancel
request may attempt cancellations through io_wq_submit_work, which may
try to grab a lock. Delay it to task_work, so we do it from a clean
context and don't have to worry about locking.

Cc: <stable@vger.kernel.org> # 5.5+
Fixes: c07e6719511e ("io_uring: hold uring_lock while completing failed polled io in io_wq_submit_work()")
Reported-by: Abaci <abaci@linux.alibaba.com>
Reported-by: Hao Xu <haoxu@linux.alibaba.com>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
[Lee: The first hunk solves a different (double free) issue in v5.10.
      Only the first hunk of the original patch is relevant to v5.10 AND
      the first hunk of the original patch is only relevant to v5.10]
Reported-by: syzbot+59d8a1f4e60c20c066cf@syzkaller.appspotmail.com
Signed-off-by: Lee Jones <lee.jones@linaro.org>
---
 fs/io_uring.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 26753d0cb4312..361f8ae96c36f 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -2075,7 +2075,9 @@ static void io_req_task_cancel(struct callback_head *cb)
 	struct io_kiocb *req = container_of(cb, struct io_kiocb, task_work);
 	struct io_ring_ctx *ctx = req->ctx;
 
+	mutex_lock(&ctx->uring_lock);
 	__io_req_task_cancel(req, -ECANCELED);
+	mutex_unlock(&ctx->uring_lock);
 	percpu_ref_put(&ctx->refs);
 }
 
-- 
2.33.0.1079.g6e70778dc9-goog