From: Hao Xu <haoxu@xxxxxxxxxxxxxxxxx> [ Upstream commit e302f1046f4c209291b07ff7bc4d15ca26891f16 ] ctx->cq_extra should be protected by completion lock so that the req_need_defer() does the right check. Cc: stable@xxxxxxxxxxxxxxx Signed-off-by: Hao Xu <haoxu@xxxxxxxxxxxxxxxxx> Link: https://lore.kernel.org/r/20211125092103.224502-2-haoxu@xxxxxxxxxxxxxxxxx Signed-off-by: Jens Axboe <axboe@xxxxxxxxx> Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx> --- fs/io_uring.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/io_uring.c b/fs/io_uring.c index d7e49e87b49b9..156c54ebb62b7 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -6573,11 +6573,14 @@ static bool io_drain_req(struct io_kiocb *req) } /* Still need defer if there is pending req in defer list. */ + spin_lock(&ctx->completion_lock); if (likely(list_empty_careful(&ctx->defer_list) && !(req->flags & REQ_F_IO_DRAIN))) { + spin_unlock(&ctx->completion_lock); ctx->drain_active = false; return false; } + spin_unlock(&ctx->completion_lock); seq = io_get_sequence(req); /* Still a chance to pass the sequence check */ -- 2.34.1