diff options
| author |   Pavel Begunkov <asml.silence@gmail.com> | 2020-03-07 01:15:22 +0300 |
|---|---|---|
| committer |   Jens Axboe <axboe@kernel.dk> | 2020-03-07 08:35:56 -0700 |
| commit | f0e20b8943509d81200cef5e30af2adfddba0f5c (patch) | |
| tree | 7dccd88cdb089208a69c95de6a3d11ed7737b057 /fs | |
| parent | io_uring: free fixed_file_data after RCU grace period (diff) | |
| download | linux-dev-f0e20b8943509d81200cef5e30af2adfddba0f5c.tar.xz linux-dev-f0e20b8943509d81200cef5e30af2adfddba0f5c.zip | |
io_uring: fix lockup with timeouts
There is a recipe to deadlock the kernel: submit a timeout sqe with a
linked_timeout (e.g. test_single_link_timeout_ception() from liburing),
and SIGKILL the process.
Then, io_kill_timeouts() takes @ctx->completion_lock, but the timeout
isn't flagged with REQ_F_COMP_LOCKED, and will try to double grab it
during io_put_free() to cancel the linked timeout. Probably, the same
can happen with another io_kill_timeout() call site, that is
io_commit_cqring().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'fs')
| -rw-r--r-- | fs/io_uring.c | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/fs/io_uring.c b/fs/io_uring.c index 68050b61ad0e..c06082bb039a 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -1000,6 +1000,7 @@ static void io_kill_timeout(struct io_kiocb *req) if (ret != -1) { atomic_inc(&req->ctx->cq_timeouts); list_del_init(&req->list); + req->flags |= REQ_F_COMP_LOCKED; io_cqring_fill_event(req, 0); io_put_req(req); } |