From: Willy Tarreau <w@1wt.eu>
Date: Tue, 10 Dec 2019 17:12:04 +0000 (+0100)
Subject: BUG/MEDIUM: proto_udp/threads: recv() and send() must not be exclusive.
X-Git-Tag: v2.2-dev1~194
X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=d7f76a0a50f4ac6b32d2317c675b3752133ef6d2;p=thirdparty%2Fhaproxy.git

BUG/MEDIUM: proto_udp/threads: recv() and send() must not be exclusive.

This is a complement to previous fix for bug #399. The exclusion between
the recv() and send() calls prevents send handlers from being called if
rx readiness is reported. The DNS code can trigger this situations with
threads where the fd_recv_ready() flag disappears between the test in
dgram_fd_handler() and the second test in dns_resolve_recv() while a
thread calls fd_cant_recv(), and this situation can sustain itself for
a while. With 8 threads and an error in the socket queue, placing a
printf on the return statement in dns_resolve_recv() scrolls very fast.

Simply removing the "else" in dgram_fd_handler() addresses the issue.

This fix must be backported as far as 1.6.
---

diff --git a/src/proto_udp.c b/src/proto_udp.c
index c4434d78ea..810c31416d 100644
--- a/src/proto_udp.c
+++ b/src/proto_udp.c
@@ -26,7 +26,7 @@ void dgram_fd_handler(int fd)
 
 	if (fd_recv_ready(fd))
 		dgram->data->recv(dgram);
-	else if (fd_send_ready(fd))
+	if (fd_send_ready(fd))
 		dgram->data->send(dgram);
 
 	return;