[PATCH 6.0] io_uring/net: copy addr for zc on POLL_FIRST

[PATCH 6.0] io_uring/net: copy addr for zc on POLL_FIRST

[Pavel Begunkov]

Every time we return from an issue handler and expect the request to be
retried we should also setup it for async exec ourselves. Do that when
we return on IORING_RECVSEND_POLL_FIRST in io_sendzc(), otherwise it'll
re-read the address, which might be a surprise for the userspace.

Fixes: 092aeedb750a9 ("io_uring: allow to pass addr into sendzc")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 io_uring/net.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/io_uring/net.c b/io_uring/net.c
index 7047c1342541..e9efed40cf3d 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -1003,9 +1003,6 @@ int io_sendzc(struct io_kiocb *req, unsigned int issue_flags)
 	unsigned msg_flags, cflags;
 	int ret, min_ret = 0;
 
-	if (!(req->flags & REQ_F_POLLED) &&
-	    (zc->flags & IORING_RECVSEND_POLL_FIRST))
-		return -EAGAIN;
 	sock = sock_from_file(req->file);
 	if (unlikely(!sock))
 		return -ENOTSOCK;
@@ -1030,6 +1027,10 @@ int io_sendzc(struct io_kiocb *req, unsigned int issue_flags)
 		msg.msg_namelen = zc->addr_len;
 	}
 
+	if (!(req->flags & REQ_F_POLLED) &&
+	    (zc->flags & IORING_RECVSEND_POLL_FIRST))
+		return io_setup_async_addr(req, addr, issue_flags);
+
 	if (zc->flags & IORING_RECVSEND_FIXED_BUF) {
 		ret = io_import_fixed(WRITE, &msg.msg_iter, req->imu,
 					(u64)(uintptr_t)zc->buf, zc->len);
-- 
2.37.2

Re: [PATCH 6.0] io_uring/net: copy addr for zc on POLL_FIRST

[Jens Axboe]

On Thu, 8 Sep 2022 14:01:10 +0100, Pavel Begunkov wrote:
> Every time we return from an issue handler and expect the request to be
> retried we should also setup it for async exec ourselves. Do that when
> we return on IORING_RECVSEND_POLL_FIRST in io_sendzc(), otherwise it'll
> re-read the address, which might be a surprise for the userspace.
> 
> 

Applied, thanks!

[1/1] io_uring/net: copy addr for zc on POLL_FIRST
      commit: 3c8400532dd8305024ff6eea38707de20b1b9822

Best regards,
-- 
Jens Axboe