From: Jens Axboe <[email protected]>
To: [email protected]
Cc: [email protected], Jens Axboe <[email protected]>
Subject: [PATCH] io_uring: add IOSQE_IGNORE_NONBLOCK flag
Date: Fri, 18 Jun 2021 13:19:40 -0600 [thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>
If a file has O_NONBLOCK set, then io_uring will not arm poll and wait
for data/space if none is available. Instead, -EAGAIN is returned if
an IO attempt is made on the file descriptor.
For library use cases, the library may not be in full control of the
file descriptor, and hence cannot modify file flags through fcntl(2).
Or even if it can, it'd be inefficient and require 3 system calls to
check, set, and re-set.
IOSQE_IGNORE_NONBLOCK provides a way to tell io_uring to ignore O_NONBLOCK
and arm poll to wait for data/space instead, just like we would have if
O_NONBLOCK wasn't set on the file descriptor.
Suggested-by: Samuel Williams <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
---
fs/io_uring.c | 11 ++++++++---
include/uapi/linux/io_uring.h | 3 +++
2 files changed, 11 insertions(+), 3 deletions(-)
diff --git a/fs/io_uring.c b/fs/io_uring.c
index fc8637f591a6..2d42273bb50f 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -108,7 +108,7 @@
#define SQE_VALID_FLAGS (IOSQE_FIXED_FILE|IOSQE_IO_DRAIN|IOSQE_IO_LINK| \
IOSQE_IO_HARDLINK | IOSQE_ASYNC | \
- IOSQE_BUFFER_SELECT)
+ IOSQE_BUFFER_SELECT | IOSQE_IGNORE_NONBLOCK)
#define IO_REQ_CLEAN_FLAGS (REQ_F_BUFFER_SELECTED | REQ_F_NEED_CLEANUP | \
REQ_F_POLLED | REQ_F_INFLIGHT | REQ_F_CREDS)
@@ -706,6 +706,7 @@ enum {
REQ_F_HARDLINK_BIT = IOSQE_IO_HARDLINK_BIT,
REQ_F_FORCE_ASYNC_BIT = IOSQE_ASYNC_BIT,
REQ_F_BUFFER_SELECT_BIT = IOSQE_BUFFER_SELECT_BIT,
+ REQ_F_IGNORE_NONBLOCK_BIT = IOSQE_IGNORE_NONBLOCK_BIT,
/* first byte is taken by user flags, shift it to not overlap */
REQ_F_FAIL_BIT = 8,
@@ -743,6 +744,8 @@ enum {
REQ_F_FORCE_ASYNC = BIT(REQ_F_FORCE_ASYNC_BIT),
/* IOSQE_BUFFER_SELECT */
REQ_F_BUFFER_SELECT = BIT(REQ_F_BUFFER_SELECT_BIT),
+ /* IOSQE_IGNORE_NONBLOCK */
+ REQ_F_IGNORE_NONBLOCK = BIT(REQ_F_IGNORE_NONBLOCK_BIT),
/* fail rest of links */
REQ_F_FAIL = BIT(REQ_F_FAIL_BIT),
@@ -2686,7 +2689,8 @@ static int io_prep_rw(struct io_kiocb *req, const struct io_uring_sqe *sqe)
return ret;
/* don't allow async punt for O_NONBLOCK or RWF_NOWAIT */
- if ((kiocb->ki_flags & IOCB_NOWAIT) || (file->f_flags & O_NONBLOCK))
+ if ((kiocb->ki_flags & IOCB_NOWAIT) ||
+ ((file->f_flags & O_NONBLOCK) && !(req->flags & REQ_F_IGNORE_NONBLOCK)))
req->flags |= REQ_F_NOWAIT;
ioprio = READ_ONCE(sqe->ioprio);
@@ -4709,7 +4713,8 @@ static int io_accept(struct io_kiocb *req, unsigned int issue_flags)
unsigned int file_flags = force_nonblock ? O_NONBLOCK : 0;
int ret;
- if (req->file->f_flags & O_NONBLOCK)
+ if ((req->file->f_flags & O_NONBLOCK) &&
+ !(req->flags & REQ_F_IGNORE_NONBLOCK))
req->flags |= REQ_F_NOWAIT;
ret = __sys_accept4_file(req->file, file_flags, accept->addr,
diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index f1f9ac114b51..582eee6e898b 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -70,6 +70,7 @@ enum {
IOSQE_IO_HARDLINK_BIT,
IOSQE_ASYNC_BIT,
IOSQE_BUFFER_SELECT_BIT,
+ IOSQE_IGNORE_NONBLOCK_BIT,
};
/*
@@ -87,6 +88,8 @@ enum {
#define IOSQE_ASYNC (1U << IOSQE_ASYNC_BIT)
/* select buffer from sqe->buf_group */
#define IOSQE_BUFFER_SELECT (1U << IOSQE_BUFFER_SELECT_BIT)
+/* ignore if file has O_NONBLOCK set */
+#define IOSQE_IGNORE_NONBLOCK (1U << IOSQE_IGNORE_NONBLOCK_BIT)
/*
* io_uring_setup() flags
--
2.32.0
prev parent reply other threads:[~2021-06-18 19:19 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-06-18 19:19 [PATCH] Add option to ignore O_NONBLOCK for retry purposes Jens Axboe
2021-06-18 19:19 ` Jens Axboe [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
[email protected] \
[email protected] \
[email protected] \
[email protected] \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox