diff mbox series

[net] mptcp: sendmsg: reset iter on error redux

Message ID 20200816211420.7337-1-fw@strlen.de
State New
Headers show
Series [net] mptcp: sendmsg: reset iter on error redux | expand

Commit Message

Florian Westphal Aug. 16, 2020, 9:14 p.m. UTC 
This fix wasn't correct: When this function is invoked from the
retransmission worker, the iterator contains garbage and resetting
it causes a crash.

As the work queue should not be performance critical also zero the
msghdr struct.

Fixes: 35759383133f64d "(mptcp: sendmsg: reset iter on error)"
Signed-off-by: Florian Westphal <fw@strlen.de>
---
 Brown paper bag patch.  I will see if having distinct functions
 for the mtcp_sendmsg and retransmit wq case is feasible/more
 appropriate.

 net/mptcp/protocol.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
diff mbox series

Patch

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index 2499757bf899..f6561d126110 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -740,7 +740,8 @@  static int mptcp_sendmsg_frag(struct sock *sk, struct sock *ssk,
 	ret = do_tcp_sendpages(ssk, page, offset, psize,
 			       msg->msg_flags | MSG_SENDPAGE_NOTLAST | MSG_DONTWAIT);
 	if (ret <= 0) {
-		iov_iter_revert(&msg->msg_iter, psize);
+		if (!retransmission)
+			iov_iter_revert(&msg->msg_iter, psize);
 		return ret;
 	}
 
@@ -1391,7 +1392,9 @@  static void mptcp_worker(struct work_struct *work)
 	struct mptcp_data_frag *dfrag;
 	u64 orig_write_seq;
 	size_t copied = 0;
-	struct msghdr msg;
+	struct msghdr msg = {
+		.msg_flags = MSG_DONTWAIT,
+	};
 	long timeo = 0;
 
 	lock_sock(sk);
@@ -1424,7 +1427,6 @@  static void mptcp_worker(struct work_struct *work)
 
 	lock_sock(ssk);
 
-	msg.msg_flags = MSG_DONTWAIT;
 	orig_len = dfrag->data_len;
 	orig_offset = dfrag->offset;
 	orig_write_seq = dfrag->data_seq;