Merge lp:~charlesk/dee/fix-904995 into lp:dee

[Charles Kerr]

Fix a potential divide-by-zero error detected by Coverity.

Fix (or, rather, silence) a potential divide-by-zero error detected by Coverity.

This patch fell behind the couch cushions awhile back, but I was cleaning out my branches and found it... good thing it's not an important bug.

[lp:~charlesk/dee/fix-904995]

345. By Charles Kerr on 2014-03-13

sync with lp:dee

[PS Jenkins bot (ps-jenkins)]

PASSED: Continuous integration, rev:344
http://jenkins.qa.ubuntu.com/job/dee-ci/16/
Executed test runs:
    SUCCESS: http://jenkins.qa.ubuntu.com/job/dee-trusty-amd64-ci/4
    SUCCESS: http://jenkins.qa.ubuntu.com/job/dee-trusty-armhf-ci/4
        deb: http://jenkins.qa.ubuntu.com/job/dee-trusty-armhf-ci/4/artifact/work/output/*zip*/output.zip
    SUCCESS: http://jenkins.qa.ubuntu.com/job/dee-trusty-i386-ci/4

Click here to trigger a rebuild:
http://s-jenkins.ubuntu-ci:8080/job/dee-ci/16/rebuild

review: Approve (continuous-integration)

[PS Jenkins bot (ps-jenkins)]

PASSED: Continuous integration, rev:345
http://jenkins.qa.ubuntu.com/job/dee-ci/17/
Executed test runs:
    SUCCESS: http://jenkins.qa.ubuntu.com/job/dee-trusty-amd64-ci/5
    SUCCESS: http://jenkins.qa.ubuntu.com/job/dee-trusty-armhf-ci/5
        deb: http://jenkins.qa.ubuntu.com/job/dee-trusty-armhf-ci/5/artifact/work/output/*zip*/output.zip
    SUCCESS: http://jenkins.qa.ubuntu.com/job/dee-trusty-i386-ci/5

Click here to trigger a rebuild:
http://s-jenkins.ubuntu-ci:8080/job/dee-ci/17/rebuild

review: Approve (continuous-integration)

Unmerged revisions

345. By Charles Kerr on 2014-03-13

sync with lp:dee

344. By Charles Kerr on 2012-02-17

silence an improbable div-by-zero coverity warning (bug #904995)

=== modified file 'tests/test-benchmark.c'
--- tests/test-benchmark.c	2013-02-04 16:07:59 +0000
+++ tests/test-benchmark.c	2014-03-13 20:34:20 +0000
@@ -80,19 +80,28 @@
     if (bench->state == NULL)
       bench->benchmark_setup (bench);
   }
-  
-  /* Compute average runtime */
-  avg_runtime = total_runtime / bench->n_runs;
-  
-  /* Compute standard deviation */
-  std_dev = 0;
-  for (i = 0; i < bench->n_runs; i++)
-    {
-      gdouble delta = bench->runs[i].elapsed - avg_runtime;
-      std_dev += delta * delta;
-    }
-  std_dev = sqrt (std_dev / bench->n_runs);
-  coeff_of_var = (std_dev /avg_runtime) * 100;
+ 
+  if (bench->n_runs == 0)
+    {
+      avg_runtime = 0;
+      std_dev = 0;
+      coeff_of_var = 0;
+    }
+  else
+    {
+      /* Compute average runtime */
+      avg_runtime = total_runtime / bench->n_runs;
+  
+      /* Compute standard deviation */
+      std_dev = 0;
+      for (i = 0; i < bench->n_runs; i++)
+        {
+          gdouble delta = bench->runs[i].elapsed - avg_runtime;
+          std_dev += delta * delta;
+        }
+      std_dev = sqrt (std_dev / bench->n_runs);
+      coeff_of_var = (std_dev /avg_runtime) * 100;
+    }
   
   /* This is not a very precise way of judging the quality,
   *  but better than rigorous hand waving */