650254 - Permanent fail on update xpcshell tests for Thunderbird | TEST-UNEXPECTED-FAIL | test_0110_general.js and co failing in head_update.js

Status: VERIFIED FIXED

(Toolkit :: Application Update, defect)

Description

[Mark Banner (:standard8)]

Attachment: Failure log

I believe that since bug 386760 / bug 642765 we've been experiencing permanent test failures on some of the update tests.

TEST-UNEXPECTED-FAIL | e:\buildbot\comm-central-win32-opt-unittest-xpcshell\build\xpcshell\tests\toolkit\mozapps\update\test\unit\test_0110_general.js | test failed (with xpcshell return code: 0), see following log:
TEST-UNEXPECTED-FAIL | e:/buildbot/comm-central-win32-opt-unittest-xpcshell/build/xpcshell/tests/toolkit/mozapps/update/test/unit/head_update.js | UPDATE TYPE complete
TEST-UNEXPECTED-FAIL | e:\buildbot\comm-central-win32-opt-unittest-xpcshell\build\xpcshell\tests\toolkit\mozapps\update\test\unit\test_0111_general.js | test failed (with xpcshell return code: 0), see following log:
TEST-UNEXPECTED-FAIL | e:/buildbot/comm-central-win32-opt-unittest-xpcshell/build/xpcshell/tests/toolkit/mozapps/update/test/unit/head_update.js | DoUpdate: changing update channel
TEST-UNEXPECTED-FAIL | e:\buildbot\comm-central-win32-opt-unittest-xpcshell\build\xpcshell\tests\toolkit\mozapps\update\test\unit\test_0112_general.js | test failed (with xpcshell return code: 0), see following log:
TEST-UNEXPECTED-FAIL | e:/buildbot/comm-central-win32-opt-unittest-xpcshell/build/xpcshell/tests/toolkit/mozapps/update/test/unit/head_update.js | DoUpdate: changing update channel
TEST-UNEXPECTED-FAIL | e:\buildbot\comm-central-win32-opt-unittest-xpcshell\build\xpcshell\tests\toolkit\mozapps\update\test\unit\test_0120_channelChange_complete.js | test failed (with xpcshell return code: 0), see following log:
TEST-UNEXPECTED-FAIL | e:/buildbot/comm-central-win32-opt-unittest-xpcshell/build/xpcshell/tests/toolkit/mozapps/update/test/unit/head_update.js | DoUpdate: changing update channel

Attaching a full log.

If I extra the first failure point, and do a manual diff between the log files, there there's some differences in the orders:

--- temp1.txt   2011-04-15 13:50:44.000000000 +0100
+++ temp2.txt   2011-04-15 13:51:32.000000000 +0100
@@ -82,8 +82,6 @@
 PREPARE REMOVEDIR a/b/8/81/
 PREPARE REMOVEDIR a/b/8/80/
 PREPARE REMOVEDIR a/b/8/80/
-PREPARE REMOVEFILE a/b/7/7text0
-PREPARE REMOVEFILE a/b/7/7text1
 PREPARE REMOVEFILE a/b/7/70/7xtest.exe
 PREPARE REMOVEFILE a/b/7/70/7xtext0
 PREPARE REMOVEFILE a/b/7/70/7xtext1
@@ -92,6 +90,8 @@
 PREPARE REMOVEFILE a/b/7/71/7xtext0
 PREPARE REMOVEFILE a/b/7/71/7xtext1
 PREPARE REMOVEDIR a/b/7/71/
+PREPARE REMOVEFILE a/b/7/7text0
+PREPARE REMOVEFILE a/b/7/7text1
 PREPARE REMOVEDIR a/b/7/
 PREPARE REMOVEDIR a/b/6/
 PREPARE REMOVEFILE a/b/5/5text1
@@ -189,8 +189,6 @@
 EXECUTE REMOVEDIR a/b/8/81/
 EXECUTE REMOVEDIR a/b/8/80/
 EXECUTE REMOVEDIR a/b/8/80/
-EXECUTE REMOVEFILE a/b/7/7text0
-EXECUTE REMOVEFILE a/b/7/7text1
 EXECUTE REMOVEFILE a/b/7/70/7xtest.exe
 EXECUTE REMOVEFILE a/b/7/70/7xtext0
 EXECUTE REMOVEFILE a/b/7/70/7xtext1
@@ -199,6 +197,8 @@
 EXECUTE REMOVEFILE a/b/7/71/7xtext0
 EXECUTE REMOVEFILE a/b/7/71/7xtext1
 EXECUTE REMOVEDIR a/b/7/71/
+EXECUTE REMOVEFILE a/b/7/7text0
+EXECUTE REMOVEFILE a/b/7/7text1
 EXECUTE REMOVEDIR a/b/7/
 EXECUTE REMOVEDIR a/b/6/
 EXECUTE REMOVEFILE a/b/5/5text1
@@ -320,8 +320,6 @@
 FINISH REMOVEDIR a/b/8/80/
 FINISH REMOVEDIR a/b/8/80/
 directory no longer exists; skipping
-FINISH REMOVEFILE a/b/7/7text0
-FINISH REMOVEFILE a/b/7/7text1
 FINISH REMOVEFILE a/b/7/70/7xtest.exe
 FINISH REMOVEFILE a/b/7/70/7xtext0
 FINISH REMOVEFILE a/b/7/70/7xtext1
@@ -330,6 +328,8 @@
 FINISH REMOVEFILE a/b/7/71/7xtext0
 FINISH REMOVEFILE a/b/7/71/7xtext1
 FINISH REMOVEDIR a/b/7/71/
+FINISH REMOVEFILE a/b/7/7text0
+FINISH REMOVEFILE a/b/7/7text1
 FINISH REMOVEDIR a/b/7/
 FINISH REMOVEDIR a/b/6/
 FINISH REMOVEFILE a/b/5/5text1
@@ -342,3 +342,4 @@
 FINISH REMOVEFILE a/b/3/3text1
 FINISH REMOVEFILE a/b/3/3text0
 succeeded

This doesn't seem to happen if I run it locally on Windows, nor does it happen on the SeaMonkey tinderboxes.

Comment 1

[Robert Strong (they/them - no direct email)]

Attachment: patch rev1

I went through the Thunderbird error logs and this should take care of it. I was a tad worried this part of the test would be fragile on different platforms and OS's but I think it is better to one off the parts that break vs. removing the check.

Comment 2

[Robert Strong (they/them - no direct email)]

Pushed to mozilla-central
http://hg.mozilla.org/mozilla-central/rev/8adda532dc63

Comment 3

[Robert Strong (they/them - no direct email)]

Pushed to mozilla-aurora
http://hg.mozilla.org/mozilla-aurora/rev/d2aaab228d11

Comment 4

[Robert Strong (they/them - no direct email)]

Thunderbird win xpcshell is now green

Comment 5

[Mark Banner (:standard8)]

Thanks Rob.