[PATCH v2] io_uring: fix short read retries for non-reg files

[PATCH v2] io_uring: fix short read retries for non-reg files

[Pavel Begunkov]

Sockets and other non-regular files may actually expect short reads to
happen, don't retry reads for them. Because non-reg files don't set
FMODE_BUF_RASYNC and so it won't do second/retry do_read, we can filter
out those cases after first do_read() attempt with ret>0.

Cc: stable@vger.kernel.org # 5.9+
Suggested-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---

v2: essentially same, but a bit cleaner check placement and
    extended commit message

 fs/io_uring.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 5f6f1e48954e..18920f9785d2 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -3552,7 +3552,7 @@ static int io_read(struct io_kiocb *req, bool force_nonblock,
 
 	/* read it all, or we did blocking attempt. no retry. */
 	if (!iov_iter_count(iter) || !force_nonblock ||
-	    (req->file->f_flags & O_NONBLOCK))
+	    (req->file->f_flags & O_NONBLOCK) || !(req->flags & REQ_F_ISREG))
 		goto done;
 
 	io_size -= ret;
-- 
2.24.0

Re: [PATCH v2] io_uring: fix short read retries for non-reg files

[Jens Axboe]

On 1/21/21 5:01 AM, Pavel Begunkov wrote:
> Sockets and other non-regular files may actually expect short reads to
> happen, don't retry reads for them. Because non-reg files don't set
> FMODE_BUF_RASYNC and so it won't do second/retry do_read, we can filter
> out those cases after first do_read() attempt with ret>0.

Applied, thanks.

-- 
Jens Axboe