Fix seemingly random failures of 'volume-test' in particular on 32-bit
machines.  See <https://bugs.freedesktop.org/show_bug.cgi?id=72374> for
details.

From 27e47c72a25846e107b6e450c3a1480a2742382e Mon Sep 17 00:00:00 2001
From: Tanu Kaskinen <tanu.kaskinen@linux.intel.com>
Date: Sat, 14 Dec 2013 07:21:22 +0000
Subject: volume-test: Increase the allowed number of rouding errors

BugLink: https://bugs.freedesktop.org/show_bug.cgi?id=72374
---
diff --git a/src/tests/volume-test.c b/src/tests/volume-test.c
index a2daf3e..1ab0b5c 100644
--- a/src/tests/volume-test.c
+++ b/src/tests/volume-test.c
@@ -138,7 +138,13 @@ START_TEST (volume_test) {
     pa_log("max deviation: %lu n=%lu", (unsigned long) md, (unsigned long) mdn);
 
     fail_unless(md <= 1);
-    fail_unless(mdn <= 251);
+
+    /* mdn counts the times there were rounding errors during the test. The
+     * number of rounding errors seems to vary slightly depending on the
+     * hardware. The original limit was 251 errors, but it was increased to 253
+     * when the test was failing on Tanu's laptop.
+     * See https://bugs.freedesktop.org/show_bug.cgi?id=72374 */
+    fail_unless(mdn <= 253);
 }
 END_TEST