diff options
| author | Jens Axboe <axboe@kernel.dk> | 2020-09-14 18:30:38 +0300 |
|---|---|---|
| committer | Jens Axboe <axboe@kernel.dk> | 2020-09-21 16:50:56 +0300 |
| commit | f5cac8b156e8b7b67bb0fdfd19900855bf9569f3 (patch) | |
| tree | a54a7b2a1818cfb0b2996d6683a2d8620b17113a /fs | |
| parent | 8f3d749685e48c44dbe877ac9781079d85f914c8 (diff) | |
| download | linux-f5cac8b156e8b7b67bb0fdfd19900855bf9569f3.tar.xz | |
io_uring: don't use retry based buffered reads for non-async bdev
Some block devices, like dm, bubble back -EAGAIN through the completion
handler. We check for this in io_read(), but don't honor it for when
we have copied the iov. Return -EAGAIN for this case before retrying,
to force punt to io-wq.
Fixes: bcf5a06304d6 ("io_uring: support true async buffered reads, if file provides it")
Reported-by: Zorro Lang <zlang@redhat.com>
Tested-by: Zorro Lang <zlang@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'fs')
| -rw-r--r-- | fs/io_uring.c | 6 |
1 files changed, 5 insertions, 1 deletions
diff --git a/fs/io_uring.c b/fs/io_uring.c index b93355ccceeb..10d742391544 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -3130,6 +3130,7 @@ static int io_read(struct io_kiocb *req, bool force_nonblock, struct iov_iter __iter, *iter = &__iter; ssize_t io_size, ret, ret2; size_t iov_count; + bool no_async; if (req->io) iter = &req->io->rw.iter; @@ -3147,7 +3148,8 @@ static int io_read(struct io_kiocb *req, bool force_nonblock, kiocb->ki_flags &= ~IOCB_NOWAIT; /* If the file doesn't support async, just async punt */ - if (force_nonblock && !io_file_supports_async(req->file, READ)) + no_async = force_nonblock && !io_file_supports_async(req->file, READ); + if (no_async) goto copy_iov; ret = rw_verify_area(READ, req->file, io_kiocb_ppos(kiocb), iov_count); @@ -3191,6 +3193,8 @@ copy_iov: ret = ret2; goto out_free; } + if (no_async) + return -EAGAIN; /* it's copied and will be cleaned with ->io */ iovec = NULL; /* now use our persistent iterator, if we aren't already */ |