Increase fixed waits in "pg_ctl start -w" from 5 seconds to 10.

In the 9.1 branch only, modify test_postmaster_connection() so that it
will wait up to 10 seconds, not 5, for the postmaster pid file to appear.
This is a much simpler and safer, if less complete, way of addressing
the buildfarm instability issues we hoped to solve with c869a7d5b.

Discussion: <20160618042812.5798.85609@wrigleys.postgresql.org>

src/bin/pg_ctl/pg_ctl.c

@@ -607,7 +607,7 @@ test_postmaster_connection(bool do_checkpoint)
 						 * timeout first.
 						 */
 						snprintf(connstr, sizeof(connstr),
-						"dbname=postgres port=%d host='%s' connect_timeout=5",
+						"dbname=postgres port=%d host='%s' connect_timeout=10",
 								 portnum, host_str);
 					}
 				}
@@ -624,14 +624,14 @@ test_postmaster_connection(bool do_checkpoint)
 
 		/*
 		 * The postmaster should create postmaster.pid very soon after being
-		 * started.  If it's not there after we've waited 5 or more seconds,
+		 * started.  If it's not there after we've waited 10 or more seconds,
 		 * assume startup failed and give up waiting.  (Note this covers both
 		 * cases where the pidfile was never created, and where it was created
 		 * and then removed during postmaster exit.)  Also, if there *is* a
 		 * file there but it appears stale, issue a suitable warning and give
 		 * up waiting.
 		 */
-		if (i >= 5)
+		if (i >= 10)
 		{
 			struct stat statbuf;