6.4-stable review patch. If anyone has any objections, please let me know.
------------------
From: Pavel Begunkov asml.silence@gmail.com
commit 45500dc4e01c167ee063f3dcc22f51ced5b2b1e9 upstream.
io-wq will retry iopoll even when it failed with -EAGAIN. If that races with task exit, which sets TIF_NOTIFY_SIGNAL for all its workers, such workers might potentially infinitely spin retrying iopoll again and again and each time failing on some allocation / waiting / etc. Don't keep spinning if io-wq is dying.
Fixes: 561fb04a6a225 ("io_uring: replace workqueue usage with io-wq") Cc: stable@vger.kernel.org Signed-off-by: Pavel Begunkov asml.silence@gmail.com Signed-off-by: Jens Axboe axboe@kernel.dk Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- io_uring/io-wq.c | 10 ++++++++++ io_uring/io-wq.h | 1 + io_uring/io_uring.c | 2 ++ 3 files changed, 13 insertions(+)
--- a/io_uring/io-wq.c +++ b/io_uring/io-wq.c @@ -174,6 +174,16 @@ static void io_worker_ref_put(struct io_ complete(&wq->worker_done); }
+bool io_wq_worker_stopped(void) +{ + struct io_worker *worker = current->worker_private; + + if (WARN_ON_ONCE(!io_wq_current_is_worker())) + return true; + + return test_bit(IO_WQ_BIT_EXIT, &worker->wq->state); +} + static void io_worker_cancel_cb(struct io_worker *worker) { struct io_wq_acct *acct = io_wq_get_acct(worker); --- a/io_uring/io-wq.h +++ b/io_uring/io-wq.h @@ -52,6 +52,7 @@ void io_wq_hash_work(struct io_wq_work *
int io_wq_cpu_affinity(struct io_uring_task *tctx, cpumask_var_t mask); int io_wq_max_workers(struct io_wq *wq, int *new_count); +bool io_wq_worker_stopped(void);
static inline bool io_wq_is_hashed(struct io_wq_work *work) { --- a/io_uring/io_uring.c +++ b/io_uring/io_uring.c @@ -2050,6 +2050,8 @@ fail: if (!needs_poll) { if (!(req->ctx->flags & IORING_SETUP_IOPOLL)) break; + if (io_wq_worker_stopped()) + break; cond_resched(); continue; }