1807928 - browser/components/migration/tests/unit/test_IE_history.js fails against MSIX packages

Status: NEW

(Firefox :: Migration, defect, P3)

Description

[[e-mail] (:bhearsum)]

It's unclear to me what's happening here. Unless these tests rely on writing to the install directory (which seems unlikely), I don't see why they would fail on MSIX and not on win32 builds.

[task 2022-12-28T20:22:12.820Z] 20:22:12  WARNING -  TEST-UNEXPECTED-FAIL | browser/components/migration/tests/unit/test_IE_history.js | xpcshell return code: 0
[task 2022-12-28T20:22:12.821Z] 20:22:12     INFO -  TEST-INFO took 684ms
[task 2022-12-28T20:22:12.821Z] 20:22:12     INFO -  >>>>>>>
[task 2022-12-28T20:22:12.821Z] 20:22:12     INFO -  (xpcshell/head.js) | test MAIN run_test pending (1)
[task 2022-12-28T20:22:12.822Z] 20:22:12     INFO -  (xpcshell/head.js) | test run_next_test 0 pending (2)
[task 2022-12-28T20:22:12.822Z] 20:22:12     INFO -  (xpcshell/head.js) | test MAIN run_test finished (2)
[task 2022-12-28T20:22:12.822Z] 20:22:12     INFO -  running event loop
[task 2022-12-28T20:22:12.822Z] 20:22:12     INFO -  browser/components/migration/tests/unit/test_IE_history.js | Starting setup
[task 2022-12-28T20:22:12.822Z] 20:22:12     INFO -  (xpcshell/head.js) | test setup pending (2)
[task 2022-12-28T20:22:12.823Z] 20:22:12     INFO -  (xpcshell/head.js) | test run_next_test 0 finished (2)
[task 2022-12-28T20:22:12.823Z] 20:22:12     INFO -  TEST-PASS | browser/components/migration/tests/unit/test_IE_history.js | setup - [setup : 24] Should be done running - true == true
[task 2022-12-28T20:22:12.823Z] 20:22:12     INFO -  TEST-PASS | browser/components/migration/tests/unit/test_IE_history.js | setup - [setup : 25] Check exit code - 0 == 0
[task 2022-12-28T20:22:12.823Z] 20:22:12     INFO -  (xpcshell/head.js) | test run_next_test 1 pending (2)
[task 2022-12-28T20:22:12.824Z] 20:22:12     INFO -  (xpcshell/head.js) | test setup finished (2)
[task 2022-12-28T20:22:12.824Z] 20:22:12     INFO -  browser/components/migration/tests/unit/test_IE_history.js | Starting test_IE_history
[task 2022-12-28T20:22:12.824Z] 20:22:12     INFO -  (xpcshell/head.js) | test test_IE_history pending (2)
[task 2022-12-28T20:22:12.824Z] 20:22:12     INFO -  (xpcshell/head.js) | test run_next_test 1 finished (2)
[task 2022-12-28T20:22:12.824Z] 20:22:12     INFO -  TEST-PASS | browser/components/migration/tests/unit/test_IE_history.js | test_IE_history - [test_IE_history : 40] Source is available - true == true
[task 2022-12-28T20:22:12.825Z] 20:22:12     INFO -  TEST-PASS | browser/components/migration/tests/unit/test_IE_history.js | test_IE_history - [test_IE_history : 1] Resource supported by migrator - true == true
[task 2022-12-28T20:22:12.825Z] 20:22:12     INFO -  Unexpected exception TypeError: can't access property "url", entry is null at Z:/task_167225684093308/build/tests/xpcshell/tests/browser/components/migration/tests/unit/test_IE_history.js:46
[task 2022-12-28T20:22:12.825Z] 20:22:12     INFO -  test_IE_history@Z:/task_167225684093308/build/tests/xpcshell/tests/browser/components/migration/tests/unit/test_IE_history.js:46:5
[task 2022-12-28T20:22:12.825Z] 20:22:12     INFO -  _do_main@Z:\task_167225684093308\build\tests\xpcshell\head.js:238:6
[task 2022-12-28T20:22:12.826Z] 20:22:12     INFO -  _execute_test@Z:\task_167225684093308\build\tests\xpcshell\head.js:585:5
[task 2022-12-28T20:22:12.826Z] 20:22:12     INFO -  @-e:1:1
[task 2022-12-28T20:22:12.826Z] 20:22:12     INFO -  exiting test
[task 2022-12-28T20:22:12.826Z] 20:22:12     INFO -  <<<<<<<

Comment 1

[:Gijs (he/him)]

This test runs an external process called InsertIEHistory.exe, built from this file, and then tries to use ancient Win32 APIs to get internet history. Is it possible one or more of those things won't work with an msix build? It may need some logging in the production IE history reading code to work out why that isn't producing results, perhaps? It looks like failure just gets treated as success and "there are no entries".

Comment 2

[:Gijs (he/him)]

(holding off on setting priority because I guess it depends on whether this failure indicates the import won't ever work or if it's "just" the test that is broken)

Comment 3

[[e-mail] (:bhearsum)]

(In reply to :Gijs (he/him) from comment #1)

This test runs an external process called InsertIEHistory.exe, built from this file, and then tries to use ancient Win32 APIs to get internet history. Is it possible one or more of those things won't work with an msix build? It may need some logging in the production IE history reading code to work out why that isn't producing results, perhaps? It looks like failure just gets treated as success and "there are no entries".

It's plausible that we're unable to find it or launch it for some reason (we've had issues with some support processes like this due to the exe not sitting beside the DLLS we ship with Firefox -- but that doesn't appear it would be a problem here). There's really no reason why we shouldn't be able to launch this process - so it's probably a test specific issue. We can probably root it out a bit by adding some extra logging around finding and launching of the process. (I'll try to do this when I get a chance.)

Comment 4

[:Gijs (he/him)]

Lumping into the S3/P3 bucket based on comment #3. Gonna re-needinfo for the investigation outlined in comment 3 - feel free to clear again if you keep track of stuff like that some other way, Ben! :-)

Comment 5

[[e-mail] (:bhearsum)]

I don't think I'm going to be able to prioritize looking at this anytime soon. (It's also only IE history tests so....maybe not terribly important?)

Comment 6

[:Gijs (he/him)]

(In reply to [e-mail] (:bhearsum) from comment #5)

(It's also only IE history tests so....maybe not terribly important?)

Yep, I think so.