From: Jens Axboe <axboe@kernel.dk>
To: io-uring@vger.kernel.org, netdev@vger.kernel.org
Cc: kuba@kernel.org, Jens Axboe <axboe@kernel.dk>,
Norman Maurer <norman_maurer@apple.com>
Subject: [PATCH 4/4] io_uring/net: wire up IORING_CQE_F_SOCK_NONEMPTY for accept
Date: Thu, 9 May 2024 12:00:29 -0600 [thread overview]
Message-ID: <20240509180627.204155-5-axboe@kernel.dk> (raw)
In-Reply-To: <20240509180627.204155-1-axboe@kernel.dk>
If the given protocol supports passing back whether or not we had more
pending accept post this one, pass back this information to userspace.
This is done by setting IORING_CQE_F_SOCK_NONEMPTY in the CQE flags,
just like we do for recv/recvmsg if there's more data available post
a receive operation.
We can also use this information to be smarter about multishot retry,
as we don't need to do a pointless retry if we know for a fact that
there aren't any more connections to accept.
Suggested-by: Norman Maurer <norman_maurer@apple.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
io_uring/net.c | 20 ++++++++++++++++----
1 file changed, 16 insertions(+), 4 deletions(-)
diff --git a/io_uring/net.c b/io_uring/net.c
index d4d1fc93635c..0a48596429d9 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -1533,6 +1533,7 @@ int io_accept(struct io_kiocb *req, unsigned int issue_flags)
.flags = force_nonblock ? O_NONBLOCK : 0,
};
struct file *file;
+ unsigned cflags;
int ret, fd;
if (!(req->flags & REQ_F_POLLED) &&
@@ -1545,6 +1546,8 @@ int io_accept(struct io_kiocb *req, unsigned int issue_flags)
if (unlikely(fd < 0))
return fd;
}
+ arg.err = 0;
+ arg.is_empty = -1;
file = do_accept(req->file, &arg, accept->addr, accept->addr_len,
accept->flags);
if (IS_ERR(file)) {
@@ -1573,17 +1576,26 @@ int io_accept(struct io_kiocb *req, unsigned int issue_flags)
accept->file_slot);
}
+ cflags = 0;
+ if (!arg.is_empty)
+ cflags |= IORING_CQE_F_SOCK_NONEMPTY;
+
if (!(req->flags & REQ_F_APOLL_MULTISHOT)) {
- io_req_set_res(req, ret, 0);
+ io_req_set_res(req, ret, cflags);
return IOU_OK;
}
if (ret < 0)
return ret;
- if (io_req_post_cqe(req, ret, IORING_CQE_F_MORE))
- goto retry;
+ if (io_req_post_cqe(req, ret, cflags | IORING_CQE_F_MORE)) {
+ if (cflags & IORING_CQE_F_SOCK_NONEMPTY || arg.is_empty == -1)
+ goto retry;
+ if (issue_flags & IO_URING_F_MULTISHOT)
+ return IOU_ISSUE_SKIP_COMPLETE;
+ return -EAGAIN;
+ }
- io_req_set_res(req, ret, 0);
+ io_req_set_res(req, ret, cflags);
return IOU_STOP_MULTISHOT;
}
--
2.43.0
next prev parent reply other threads:[~2024-05-09 18:06 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-05-09 18:00 [PATCHSET RFC 0/4] Propagate back queue status on accept Jens Axboe
2024-05-09 18:00 ` [PATCH 1/4] net: change proto and proto_ops accept type Jens Axboe
2024-05-09 18:00 ` [PATCH 2/4] net: have do_accept() take a struct proto_accept_arg argument Jens Axboe
2024-05-09 18:00 ` [PATCH 3/4] net: pass back whether socket was empty post accept Jens Axboe
2024-05-09 18:00 ` Jens Axboe [this message]
2024-05-14 0:13 ` [PATCHSET RFC 0/4] Propagate back queue status on accept Jakub Kicinski
2024-05-14 0:24 ` Jens Axboe
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 \
--in-reply-to=20240509180627.204155-5-axboe@kernel.dk \
--to=axboe@kernel.dk \
--cc=io-uring@vger.kernel.org \
--cc=kuba@kernel.org \
--cc=netdev@vger.kernel.org \
--cc=norman_maurer@apple.com \
/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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).