1529202 - TEST-UNEXPECTED-FAIL | security/manager/ssl/tests/unit/test_sss_sanitizeOnShutdown.js when run in Thunderbird

Status: RESOLVED FIXED

(Core :: Security: PSM, defect)

Description

[Jorg K (CEST = GMT+2)]

security/manager/ssl/tests/unit/test_sss_sanitizeOnShutdown.js when run in TB.

M-C last good: 57c4b2afc95bfb3c444c732802d99c8522
M-C first bad: bf3951daded0b4fefd0fc81ee22ce5d536
https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=57c4b2afc95bfb3c444c732802d99c8522&tochange=bf3951daded0b4fefd0fc81ee22ce5d536

The test is new in
https://hg.mozilla.org/mozilla-central/rev/2ee7c36ed3ac5cc94e55e4094f6e46bb1be689f5
from bug 1527372.

Maybe it should be disabled for Thunderbird.

Kai, would you like to join the "bustage fix team" ;-)

Comment 1

[Kai Engert [:KaiE:]]

[task 2019-02-20T06:05:42.344Z] 06:05:42 INFO - TEST-START | security/manager/ssl/tests/unit/test_sss_sanitizeOnShutdown.js
[task 2019-02-20T06:05:42.663Z] 06:05:42 WARNING - TEST-UNEXPECTED-FAIL | security/manager/ssl/tests/unit/test_sss_sanitizeOnShutdown.js | xpcshell return code: 0
[task 2019-02-20T06:05:42.663Z] 06:05:42 INFO - TEST-INFO took 316ms
[task 2019-02-20T06:05:42.664Z] 06:05:42 INFO - >>>>>>>
[task 2019-02-20T06:05:42.664Z] 06:05:42 INFO - PID 11089 | [11089, Unnamed thread ec30c100] WARNING: XPCOM objects created/destroyed from static ctor/dtor: file /builds/worker/workspace/build/src/xpcom/base/nsTraceRefcnt.cpp, line 194
[task 2019-02-20T06:05:42.665Z] 06:05:42 INFO - PID 11089 | [11089, Main Thread] WARNING: Couldn't get the user appdata directory. Crash events may not be produced.: file /builds/worker/workspace/build/src/toolkit/crashreporter/nsExceptionHandler.cpp, line 2528
[task 2019-02-20T06:05:42.665Z] 06:05:42 INFO - NS_ERROR_FILE_NOT_FOUND:
[task 2019-02-20T06:05:42.666Z] 06:05:42 INFO - @/builds/worker/workspace/build/tests/xpcshell/tests/security/manager/ssl/tests/unit/test_sss_sanitizeOnShutdown.js:14:1
[task 2019-02-20T06:05:42.667Z] 06:05:42 INFO - load_file@/builds/worker/workspace/build/tests/xpcshell/head.js:635:7
[task 2019-02-20T06:05:42.667Z] 06:05:42 INFO - _load_files@/builds/worker/workspace/build/tests/xpcshell/head.js:647:10
[task 2019-02-20T06:05:42.668Z] 06:05:42 INFO - _execute_test@/builds/worker/workspace/build/tests/xpcshell/head.js:501:3
[task 2019-02-20T06:05:42.668Z] 06:05:42 INFO - @-e:1:1

Comment 2

[Kai Engert [:KaiE:]]

Also, above comment 1, we had
INFO - TEST-FAIL | security/manager/ssl/tests/unit/test_sss_sanitizeOnShutdown.js | took 2586ms
INFO - security/manager/ssl/tests/unit/test_sss_sanitizeOnShutdown.js failed or timed out, will retry.

Why does it complain about test_sss_sanitizeOnShutdown.js:14:1 ? Line 14 ?
https://hg.mozilla.org/mozilla-central/diff/2ee7c36ed3ac/security/manager/ssl/tests/unit/test_sss_sanitizeOnShutdown.js

Is the file-not-found error related to this test? Which file cannot be found?

I see two Sanitizer.jsm files in the tree, is there a conflict in Thunderbird
./browser/modules/Sanitizer.jsm
./comm/suite/modules/Sanitizer.jsm

• // Simulate shutdown.
• Services.obs.notifyObservers(null, "profile-change-teardown");
• Services.obs.notifyObservers(null, "profile-before-change");
• equal(getStateFileContents(), "", "state file should be empty");

Are these notifications sync or async?
If they are async, is the check getStateFileContents() done too early?

Should the "equal" failure scenario print the result of getStateFileContents? Could that help us understand why the test is failing with Thunderbird (what content is still there)?

Comment 3

[Kai Engert [:KaiE:]]

from the checkin comment: "make sure that data was cleared at shutdown, if requested by the user"

I think this means the test is related to things like "delete cookies on exit"?
TB has this preference, too, so I conclude this test must pass in TB, or this feature probably isn't working correctly in TB.

Comment 4

[Dana Keeler (she/her) (use needinfo) [:keeler]]

Hmmm - I suppose if browser/modules/Sanitizer.jsm is a Firefox-specific thing, then it makes sense that it wouldn't be available in Thunderbird. Maybe the best option is to just disable this for TB?

Comment 5

[Jorg K (CEST = GMT+2)]

Attachment: Bug 1529202 - disable test test_sss_sanitizeOnShutdown.js for Thunderbird.

Comment 6

[Cosmin Sabou [:CosminS]]

This revision is blocked from landing in Phabricator. Please see the following info:

Lando says "This diff does not have the proper author information uploaded to Phabricator", but I see an author on Phabricator. What's wrong?

There are a few reasons this could happen:

The revision was created via the Phabricator Web UI or via an unsupported client.
    Use arcanist or moz-phab to submit the patch instead; see the Mozilla Phabricator User Guide for help.

The revision was created via arcanist or moz-phab, but the error still appears.
    This can happen if you have not set an author email in your .hgrc file (or git config). Set your author email in your .hgrc to your username, Firstname Lastname <[e-mail]>. Update your commit so it contains the new author data. Re-run arc diff so that the new commit+data is uploaded to Phabricator.

You are attempting to re-land a revision but Bug 1489126 happens
    This is an unfortunate bug that will be fixed in time, but for now you can get around this by reopening the revision on Phabricator and then resubmitting the revisions on your client. Follow the bug to see status updates.

Comment 7

[Kai Engert [:KaiE:]]

Jörg, let me know if you want me to help to land this manually into mozilla-inbound.

Comment 8

[Jorg K (CEST = GMT+2)]

Attachment: 1529202.diff

Sorry, I don't know what's wrong with Phab, it's certainly no enhancing my productivity :-(

Quoting from the FAQ:
The revision was created via the Phabricator Web UI

Yes, it was. So that Web UI is useless? Why is it there? Surely the sheriffs could have gotten the patch out of Phab and landed it.

Here it is as classic patch.

Comment 9

[Jorg K (CEST = GMT+2)]

Filed bug 1529506.

Comment 10

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/0b0638404dfc
disable test test_sss_sanitizeOnShutdown.js for Thunderbird. r=keeler

Comment 11

[Narcis Beleuzu [:NarcisB]]

https://hg.mozilla.org/mozilla-central/rev/0b0638404dfc

Comment 12

[Liz Henry (:lizzard) (relman/hg->git project)]

Can you request uplift to beta here? (for bug 1527372)

Comment 13

[Liz Henry (:lizzard) (relman/hg->git project)]

Or just go ahead and land on beta since this is a test-only change.

Comment 14

[Jorg K (CEST = GMT+2)]

Right, but bug 1527372 needs to go to beta first since we can't disable a test that doesn't exist yet. Hence I posted the request there.

Comment 15

[Liz Henry (:lizzard) (relman/hg->git project)]

Oh ok. thanks.

Comment 16

[Jorg K (CEST = GMT+2)]

It's a test-only change, in fact, only disabling a test for Thunderbird.

Comment 17

[Andreea Pavel [:apavel]]

https://hg.mozilla.org/releases/mozilla-beta/rev/eccb1f5d6a07