[PATCH for-next] io_uring/net: always set kmsg->msg.msg_control_user before issue

[PATCH for-next] io_uring/net: always set kmsg->msg.msg_control_user before issue

[Jens Axboe]

We currently set this separately for async/sync entry, but let's just
move it to a generic pre-issue spot and eliminate the difference
between the two.

Signed-off-by: Jens Axboe <axboe@kernel.dk>

---

This is for the 6.10 branch, was looking into dealing with the
msg_control_user fix for 6.9 SENDMSG_ZC, and I think this is a good
pre cleanup that just makes it easier to deal with while merging the
sync/async paths.

diff --git a/io_uring/net.c b/io_uring/net.c
index d095d3e48fcd..b08c0ae5951a 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -391,7 +391,6 @@ int io_sendmsg(struct io_kiocb *req, unsigned int issue_flags)
 
 	if (req_has_async_data(req)) {
 		kmsg = req->async_data;
-		kmsg->msg.msg_control_user = sr->msg_control;
 	} else {
 		kmsg = io_msg_alloc_async(req, issue_flags);
 		if (unlikely(!kmsg))
@@ -411,6 +410,8 @@ int io_sendmsg(struct io_kiocb *req, unsigned int issue_flags)
 	if (flags & MSG_WAITALL)
 		min_ret = iov_iter_count(&kmsg->msg.msg_iter);
 
+	kmsg->msg.msg_control_user = sr->msg_control;
+
 	ret = __sys_sendmsg_sock(sock, &kmsg->msg, flags);
 
 	if (ret < min_ret) {
@@ -1271,7 +1272,6 @@ int io_sendmsg_zc(struct io_kiocb *req, unsigned int issue_flags)
 
 	if (req_has_async_data(req)) {
 		kmsg = req->async_data;
-		kmsg->msg.msg_control_user = sr->msg_control;
 	} else {
 		kmsg = io_msg_alloc_async(req, issue_flags);
 		if (unlikely(!kmsg))
@@ -1291,6 +1291,7 @@ int io_sendmsg_zc(struct io_kiocb *req, unsigned int issue_flags)
 	if (flags & MSG_WAITALL)
 		min_ret = iov_iter_count(&kmsg->msg.msg_iter);
 
+	kmsg->msg.msg_control_user = sr->msg_control;
 	kmsg->msg.msg_ubuf = &io_notif_to_data(sr->notif)->uarg;
 	kmsg->msg.sg_from_iter = io_sg_from_iter_iovec;
 	ret = __sys_sendmsg_sock(sock, &kmsg->msg, flags);

-- 
Jens Axboe