summaryrefslogtreecommitdiff
path: root/source4/script/tests/selftest.pl
diff options
context:
space:
mode:
authorAndrew Bartlett <abartlet@samba.org>2007-04-11 01:32:38 +0000
committerGerald (Jerry) Carter <jerry@samba.org>2007-10-10 14:49:55 -0500
commitdbfb32ffe5925259994712c9b1f18c4b60b9cdee (patch)
tree71b75c10c0d2b91705140e07c98ddce749ff65ee /source4/script/tests/selftest.pl
parente338b4b47b6271ca46060a1fe8c61ab8afaecc8f (diff)
downloadsamba-dbfb32ffe5925259994712c9b1f18c4b60b9cdee.tar.gz
samba-dbfb32ffe5925259994712c9b1f18c4b60b9cdee.tar.bz2
samba-dbfb32ffe5925259994712c9b1f18c4b60b9cdee.zip
r22160: Only use test environment names that actually exist (get make
quicktest to pass). Ensure that when we fail on the build farm, this ends up in the total failures. I think having the 2 different functions for running the tests is just asking for this kind of failure. The build farm should be a presentation detail, and not alter the test behaviour. There is still a gremlin in there, we will fail horribly on the farm, but not on local tests... Andrew Bartlett (This used to be commit 519f3c0575748968ece2f556f370ade98c59ff05)
Diffstat (limited to 'source4/script/tests/selftest.pl')
-rwxr-xr-xsource4/script/tests/selftest.pl1
1 files changed, 1 insertions, 0 deletions
diff --git a/source4/script/tests/selftest.pl b/source4/script/tests/selftest.pl
index bb9510a00b..e6ca48bafb 100755
--- a/source4/script/tests/selftest.pl
+++ b/source4/script/tests/selftest.pl
@@ -223,6 +223,7 @@ sub run_test_buildfarm($$$$)
if ($ret == $expected_ret) {
print "TEST PASSED: $name\n";
} else {
+ push(@$suitesfailed, $name);
print "TEST FAILED: $name (status $ret)\n";
}
print "==========================================\n";