[PATCH] io_uring: remove MSG_NOSIGNAL from recvmsg

From: David Lamparter <equinox@diac24.net>
To: io-uring@vger.kernel.org
Cc: netdev@vger.kernel.org, David Lamparter <equinox@diac24.net>,
	Eric Dumazet <edumazet@google.com>, Jens Axboe <axboe@kernel.dk>
Subject: [PATCH] io_uring: remove MSG_NOSIGNAL from recvmsg
Date: Fri, 24 Feb 2023 16:01:24 +0100	[thread overview]
Message-ID: <20230224150123.128346-1-equinox@diac24.net> (raw)
In-Reply-To: <CANn89iJE6SpB2bfXEc=73km6B2xtBSWHj==WsYFnH089WPKtSA@mail.gmail.com>

MSG_NOSIGNAL is not applicable for the receiving side, SIGPIPE is
generated when trying to write to a "broken pipe".  AF_PACKET's
packet_recvmsg() does enforce this, giving back EINVAL when MSG_NOSIGNAL
is set - making it unuseable in io_uring's recvmsg.

Remove MSG_NOSIGNAL from io_recvmsg_prep().

Signed-off-by: David Lamparter <equinox@diac24.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jens Axboe <axboe@kernel.dk>
---

> Sure, or perhaps David wanted to take care of this.

Here you go.  But maybe give me a few hours to test/confirm...

diff --git a/io_uring/net.c b/io_uring/net.c
index cbd4b725f58c..b7f190ca528e 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -567,7 +567,7 @@ int io_recvmsg_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 	sr->flags = READ_ONCE(sqe->ioprio);
 	if (sr->flags & ~(RECVMSG_FLAGS))
 		return -EINVAL;
-	sr->msg_flags = READ_ONCE(sqe->msg_flags) | MSG_NOSIGNAL;
+	sr->msg_flags = READ_ONCE(sqe->msg_flags);
 	if (sr->msg_flags & MSG_DONTWAIT)
 		req->flags |= REQ_F_NOWAIT;
 	if (sr->msg_flags & MSG_ERRQUEUE)
-- 
2.39.2

