Merge lp:~clint-fewbar/ubuntu/natty/squid/no-expect-fork into lp:ubuntu/natty/squid

[Clint Byrum]

Fixes the bug by dropping expect fork

This does extend the already existing race between squid's started event and it listening for connections. There don't seem to be any blocking operations before the fork, so this race condition is already as long as its going to be, also I'm not aware of any job files that depend on squid's started event before connecting to it.

=== modified file 'debian/changelog'
--- debian/changelog	2010-09-27 23:31:35 +0000
+++ debian/changelog	2011-01-11 21:34:15 +0000
@@ -1,3 +1,10 @@
+squid (2.7.STABLE9-2.1ubuntu2) natty; urgency=low
+
+  * Run squid with -N and drop 'expect fork' to prevent HUP's from
+    causing upstart to lose track of squid. (LP: #573853)
+
+ -- Clint Byrum <[e-mail]>  Tue, 11 Jan 2011 10:49:37 -0800
+
 squid (2.7.STABLE9-2.1ubuntu1) natty; urgency=low
 
   * Merge from debian unstable.  Remaining changes:
 
=== modified file 'debian/squid.upstart'
--- debian/squid.upstart	2010-09-27 12:27:09 +0000
+++ debian/squid.upstart	2011-01-11 21:34:15 +0000
@@ -8,8 +8,6 @@
         and net-device-up IFACE!=lo)
 stop on runlevel[!2345]
 
-expect fork
-
 pre-start script
 	if [ -f /etc/default/squid ]; then
 		. /etc/default/squid
@@ -78,5 +76,5 @@
 		. /etc/default/squid
 	fi
 
-	exec /usr/sbin/squid $SQUID_ARGS
+	exec /usr/sbin/squid -N $SQUID_ARGS
 end script