diff options
| author | Pavel Begunkov <asml.silence@gmail.com> | 2021-02-26 18:47:56 +0300 |
|---|---|---|
| committer | Jens Axboe <axboe@kernel.dk> | 2021-03-04 16:34:03 +0300 |
| commit | ba50a036f23c44608b1d903c34644a1acd5d21fa (patch) | |
| tree | e4d1f04645e03aefa8500270f262979da5bdd9d1 | |
| parent | e54945ae947fb881212a4b97d5599a01bba6ad06 (diff) | |
| download | linux-ba50a036f23c44608b1d903c34644a1acd5d21fa.tar.xz | |
io_uring: run fallback on cancellation
io_uring_try_cancel_requests() matches not only current's requests, but
also of other exiting tasks, so we need to actively cancel them and not
just wait, especially since the function can be called on flush during
do_exit() -> exit_files().
Even if it's not a problem for now, it's much nicer to know that the
function tries to cancel everything it can.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
| -rw-r--r-- | fs/io_uring.c | 7 |
1 files changed, 6 insertions, 1 deletions
diff --git a/fs/io_uring.c b/fs/io_uring.c index d55c9ab6314a..9d6696ff5748 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -8518,9 +8518,10 @@ static int io_remove_personalities(int id, void *p, void *data) return 0; } -static void io_run_ctx_fallback(struct io_ring_ctx *ctx) +static bool io_run_ctx_fallback(struct io_ring_ctx *ctx) { struct callback_head *work, *head, *next; + bool executed = false; do { do { @@ -8537,7 +8538,10 @@ static void io_run_ctx_fallback(struct io_ring_ctx *ctx) work = next; cond_resched(); } while (work); + executed = true; } while (1); + + return executed; } static void io_ring_exit_work(struct work_struct *work) @@ -8677,6 +8681,7 @@ static void io_uring_try_cancel_requests(struct io_ring_ctx *ctx, ret |= io_poll_remove_all(ctx, task, files); ret |= io_kill_timeouts(ctx, task, files); ret |= io_run_task_work(); + ret |= io_run_ctx_fallback(ctx); io_cqring_overflow_flush(ctx, true, task, files); if (!ret) break; |