Closed Bug 1441812 Opened 6 years ago Closed 6 years ago

FAIL | toolkit/components/places/tests/unifiedcomplete/test_PlacesSearchAutocompleteProvider.js and more in toolkit/components/search/tests/xpcshell/ and Mozmill failures

Categories

(Thunderbird :: Testing Infrastructure, defect)

Product:
Thunderbird
Component:
Testing Infrastructure
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

(Not tracked)

Status:
RESOLVED FIXED
Milestone:
Thunderbird 60.0

People

(Reporter: jorgk-bmo, Assigned: jorgk-bmo)

Details

(Whiteboard: [Thunderbird-testfailure: XZ all])

Attachments

(1 file)

1441812.patch
1.06 KB, patch
jorgk-bmo
: review+
Details | Diff | Splinter Review 
FAIL | toolkit/components/places/tests/unifiedcomplete/test_PlacesSearchAutocompleteProvider.js
xpcshell return code: 0

TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_addEngine_callback.js | xpcshell return code: 0 
TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_addEngine_callback.js | simple_callback_test - [simple_callback_test : 43] "Test search engine" != "Test search engine" 
TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_selectedEngine.js | xpcshell return code: 0 
TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_selectedEngine.js | test_defaultEngine - [test_defaultEngine : 13] "Test search engine" == "Bing" 
TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_addEngineWithDetails.js | xpcshell return code: 0 
TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_addEngineWithDetailsObject.js | xpcshell return code: 0 
TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_engineUpdate.js | xpcshell return code: 0 
TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_engineUpdate.js | test_engineUpdate - [test_engineUpdate : 38] Order not cleared by update - "undefined" == 1 
TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_defaultEngine.js | xpcshell return code: 0 
TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_defaultEngine.js | test_defaultEngine - [test_defaultEngine : 37] [object XPCWrappedNative_NoHelper] == null 
TEST-UNEXPECTED-TIMEOUT | toolkit/components/search/tests/xpcshell/test_nocache.js | Test timed out 
TEST-UNEXPECTED-TIMEOUT | toolkit/components/search/tests/xpcshell/test_geodefaults.js | Test timed out 
TEST-UNEXPECTED-TIMEOUT | toolkit/components/search/tests/xpcshell/test_searchReset.js | Test timed out 
TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_currentEngine_fallback.js | xpcshell return code: 0 
TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_currentEngine_fallback.js | run_test - [run_test : 5] false == true  

First seen: Wed, Feb 28, 2018, 12:27:35
M-C last good: 5297541590781af40ff09e067646f31159
M-C first bad: ee326c976eebdca48128054022c443d399
https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=5297541590781af40ff09e067646f31159&tochange=ee326c976eebdca48128054022c443d399

One failure in
https://taskcluster-artifacts.net/IJm8978_RMSLhq0cLBphdA/0/public/logs/live_backing.log
says:
WARNING -  TEST-UNEXPECTED-FAIL | toolkit/components/search/tests/xpcshell/test_selectedEngine.js | test_defaultEngine - [test_defaultEngine : 13] "Test search engine" == "Bing"

FRG, can you see the regression?
Flags: needinfo?(frgrahl)
This also causes Mozmill failures:
TEST-UNEXPECTED-FAIL | /Users/cltbld/tasks/task_1519819063/build/tests/mozmill/newmailaccount/test-newmailaccount.js | test-newmailaccount.js::test_get_an_account
TEST-UNEXPECTED-FAIL | /Users/cltbld/tasks/task_1519819063/build/tests/mozmill/newmailaccount/test-newmailaccount.js | test-newmailaccount.js::test_restored_ap_tab_works

Log
https://taskcluster-artifacts.net/TKWvWDW1RjmlfTuZRpKZ9g/0/public/logs/live_backing.log
says:
INFO -  SUMMARY-PASS | test-newmailaccount.js::setupModule
INFO -  SUMMARY-UNEXPECTED-FAIL | test-newmailaccount.js | test-newmailaccount.js::test_get_an_account
INFO -    EXCEPTION: a == b: '[xpconnect wrapped nsISearchEngine]' == '[xpconnect wrapped nsISearchEngine]'.
Summary: FAIL | toolkit/components/places/tests/unifiedcomplete/test_PlacesSearchAutocompleteProvider.js and more in toolkit/components/search/tests/xpcshell/ → FAIL | toolkit/components/places/tests/unifiedcomplete/test_PlacesSearchAutocompleteProvider.js and more in toolkit/components/search/tests/xpcshell/ and Mozmill failures
I installed a Daily and in the error console I see:
ASSERT: *** Search: Can't init an engine with no data!
 debug.js:55
NS_ASSERT resource://gre/modules/debug.js:55:3
ENSURE_WARN jar:file:///C:/Program%20Files/Mozilla%20Thunderbird%2060/omni.ja!/components/nsSearchService.js:262:3
SRCH_ENG_initFromData jar:file:///C:/Program%20Files/Mozilla%20Thunderbird%2060/omni.ja!/components/nsSearchService.js:1801:5
asyncInitFromURI jar:file:///C:/Program%20Files/Mozilla%20Thunderbird%2060/omni.ja!/components/nsSearchService.js:1416:5
next self-hosted:1211:9

Alice, could you check for us which M-C change caused this. I can't see it from the check-in messages.

STR: Start TB and look at the error console.
Flags: needinfo?(frgrahl) → needinfo?(alice0775)
Thanks, but I already had the M-C range in comment #0. I can't see what caused this from the check-in comments. It would need local bisecting.
I think that I can do local bisect if I can get about 24-36hrs.
I won't touch the bug and I don't think anyone else will work out what happened here by just looking at the regression range. Thanks in advance!
(omg!, local build fails, and also hg clone aborted...)
I ended up... due to "transaction abort!"
Could this come from bug 1439742?
Watching a local build I saw these errors:
 1:13.52 Makefile:58: Missing searchplugin: amazondotcom.xml
 1:13.52 Makefile:58: Missing searchplugin: wikipedia.xml
 1:13.52 Makefile:58: Missing searchplugin: bing.xml
 1:13.52 Makefile:58: Missing searchplugin: aol-web-search.xml
 1:13.52 Makefile:58: Missing searchplugin: yahoo.xml
 1:13.52 Makefile:58: Missing searchplugin: twitter.xml
I'm pretty sure they come from bug 1439742 and hopefully the patch in bug 1441791 might have already fixed the problem.

I'll try it later.
Bug 1441791 has now landed but I still see the "Missing searchplugin" errors in a local build. Here's the push:
https://treeherder.mozilla.org/#/jobs?repo=comm-central&revision=6c63128a12cad1d8340eb0eac611c71c9b03c672
So I suppose the test failures will still be there. Philipp, could you please take a look what's going on with those search plugins. This seems to be a build issue rather than a code issue since the M-C Xpcshell tests pass FF but not for TB.
Flags: needinfo?(philipp)
Whiteboard: [Thunderbird-testfailure: XZ all]
Note that bug 1439742 landed something to do with search plugins here:
https://hg.mozilla.org/mozilla-central/rev/d304b444c23d#l1.14

SM are restored this hunk in bug 1442149:
https://hg.mozilla.org/comm-central/rev/dada7b1b4e03#l1.12

I've looked, but TB's mail/locales/Makefile.in looks very different.
Attached patch 1441812.patch — — Splinter Review 
This does the trick. It's similar to https://hg.mozilla.org/comm-central/rev/dada7b1b4e03#l1.12 but doesn't bring back the MERGE_FILES since that's only used in suite/.

Bill Gianopoulos and FRG were trading patches via PM, so I grabbed one that does the job ;-)
Assignee: nobody → jorgk
Status: NEW → ASSIGNED
Flags: needinfo?(philipp)
Attachment #8955793 - Flags: review+
Pushed by mozilla@jorgk.com:
https://hg.mozilla.org/comm-central/rev/84c3d9dbc9ec
Port bug 1439742: Reinstate EN_US_OR_L10N_FILE (analysis by the SM team). r=jorgk
Status: ASSIGNED → RESOLVED
Closed: 6 years ago
Resolution: --- → FIXED
Target Milestone: --- → Thunderbird 60.0
This might work, but isn't the right fix. m-c now uses

SEARCHPLUGINS := $(foreach plugin,$(addsuffix .xml,$(SEARCHPLUGINS_FILENAMES)),$(or $(wildcard $(srcdir)/searchplugins/$(plugin)),$(warning Missing searchplugin: $(plugin))))

which drops EN_US_OR_L10N_FILE. The best course of action would be to find the bug where this was removed, and make a similar fix on our Makefile
From here?
https://hg.mozilla.org/mozilla-central/rev/872a8e47e894#l1.16
Feel free to reopen and give a better solution ;-)
More like using hg grep to find bug 1276740, which shows that we can just remove use of that macro. Can you look into that?
No time for this really right now. It's a very minor "perfection" issue compared with a Daily build I have to cancel since it's busted.
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: