Closed Bug 1530824 Opened 8 months ago Closed 8 months ago

Permafailing browser_startup_content.js | all processScripts whitelist entries should have been used - Got 1, expected 0 when Gecko 67 merges to beta on 2019-03-11

Categories

(Web Compatibility :: Interventions, defect)

defect
Not set

Tracking

(firefox-esr60 unaffected, firefox65 unaffected, firefox66 unaffected, firefox67+ verified)

VERIFIED FIXED
Tracking Status
firefox-esr60 --- unaffected
firefox65 --- unaffected
firefox66 --- unaffected
firefox67 + verified

People

(Reporter: nataliaCs, Unassigned)

References

Details

Attachments

(1 file)

[Tracking Requested - why for this release]:
Central as Beta: https://treeherder.mozilla.org/#/jobs?repo=try&revision=f6eef58baf63ab14ea1241924bb220f990a31be8&searchStr=browser-chrome&group_state=expanded&selectedJob=230585518

Failure log: https://treeherder.mozilla.org/logviewer.html#/jobs?job_id=230585518&repo=try&lineNumber=2291

[task 2019-02-26T18:16:00.394Z] 18:16:00 INFO - TEST-START | browser/base/content/test/performance/browser_startup_content.js
[task 2019-02-26T18:16:01.129Z] 18:16:01 INFO - TEST-PASS | browser/base/content/test/performance/browser_startup_content.js | should have no unexpected modules loaded on content process startup -
[task 2019-02-26T18:16:01.129Z] 18:16:01 INFO - TEST-PASS | browser/base/content/test/performance/browser_startup_content.js | all modules whitelist entries should have been used -
[task 2019-02-26T18:16:01.130Z] 18:16:01 INFO - TEST-PASS | browser/base/content/test/performance/browser_startup_content.js | should have no unexpected frameScripts loaded on content process startup -
[task 2019-02-26T18:16:01.131Z] 18:16:01 INFO - TEST-PASS | browser/base/content/test/performance/browser_startup_content.js | all frameScripts whitelist entries should have been used -
[task 2019-02-26T18:16:01.131Z] 18:16:01 INFO - TEST-PASS | browser/base/content/test/performance/browser_startup_content.js | should have no unexpected processScripts loaded on content process startup -
[task 2019-02-26T18:16:01.132Z] 18:16:01 INFO - TEST-INFO | started process screentopng
[task 2019-02-26T18:16:01.667Z] 18:16:01 INFO - TEST-INFO | screentopng: exit 0
[task 2019-02-26T18:16:01.669Z] 18:16:01 INFO - Buffered messages logged at 18:16:00
[task 2019-02-26T18:16:01.670Z] 18:16:01 INFO - Entering test bound
[task 2019-02-26T18:16:01.671Z] 18:16:01 INFO - Buffered messages finished
[task 2019-02-26T18:16:01.673Z] 18:16:01 INFO - TEST-UNEXPECTED-FAIL | browser/base/content/test/performance/browser_startup_content.js | all processScripts whitelist entries should have been used - Got 1, expected 0
[task 2019-02-26T18:16:01.674Z] 18:16:01 INFO - Stack trace:
[task 2019-02-26T18:16:01.675Z] 18:16:01 INFO - chrome://mochikit/content/browser-test.js:test_is:1315
[task 2019-02-26T18:16:01.676Z] 18:16:01 INFO - chrome://mochitests/content/browser/browser/base/content/test/performance/browser_startup_content.js:null:191
[task 2019-02-26T18:16:01.678Z] 18:16:01 INFO - Not taking screenshot here: see the one that was previously logged
[task 2019-02-26T18:16:01.679Z] 18:16:01 INFO - TEST-UNEXPECTED-FAIL | browser/base/content/test/performance/browser_startup_content.js | processScripts is whitelisted for content process startup but wasn't used: resource://webcompat/aboutPageProcessScript.js -
[task 2019-02-26T18:16:01.680Z] 18:16:01 INFO - Stack trace:
[task 2019-02-26T18:16:01.681Z] 18:16:01 INFO - chrome://mochikit/content/browser-test.js:test_ok:1304
[task 2019-02-26T18:16:01.682Z] 18:16:01 INFO - chrome://mochitests/content/browser/browser/base/content/test/performance/browser_startup_content.js:null:195
[task 2019-02-26T18:16:01.684Z] 18:16:01 INFO - Leaving test bound
[task 2019-02-26T18:16:01.688Z] 18:16:01 INFO - GECKO(1968) | MEMORY STAT | vsize 2878MB | residentFast 1406MB | heapAllocated 1072MB
[task 2019-02-26T18:16:01.689Z] 18:16:01 INFO - TEST-OK | browser/base/content/test/performance/browser_startup_content.js | took 798ms

Flags: needinfo?(twisniewski)

Interesting, this is supposed to be Nightly-only for now.

From the failure it probably is, but the test thinks it should be on (presumably because I used the wrong whitelist or something like that). I'll investigate (I just have to remember how to make a central-as-beta build...)

Flags: needinfo?(twisniewski)

only expect the about:compat process script to load on builds that actually start it

Alright, I've created a patch which should fix this, but I can't seem to test it properly on a local build, despite applying the same beta-as-central simulation patch that the try-run applied.

:nataliaCs, could we do another beta-sim try run with this patch applied to verify that it works before landing it? (it should only need to run the related mochitest, if that's easy to do)

Flags: needinfo?(ncsoregi)

(In reply to Thomas Wisniewski [:twisniewski] from comment #4)

Alright, I've created a patch which should fix this, but I can't seem to test it properly on a local build, despite applying the same beta-as-central simulation patch that the try-run applied.

:nataliaCs, could we do another beta-sim try run with this patch applied to verify that it works before landing it? (it should only need to run the related mochitest, if that's easy to do)

The next set of beta-sims will be done after the daily merges at around 10am UTC, and this patch will be included as well

Flags: needinfo?(ncsoregi)
Attachment #9046947 - Attachment description: Bug 1530824 - only expect the about:compat process script to load on builds that actually start it; r?kmag → Bug 1530824 - mark the about:compat process script as intermittently loaded in the browser_startup_content mochitest; r?kmag

Interesting... it failed on a different OS this time. In that case, the script must just be intermittently loaded by the time the test runs sometimes (due to how webextension experiments work). Given that, I've updated the patch with a fix which should do the trick.

Can we try again with the updated patch for the next beta-simulation run, Cristina?

Natalia, would you mind queueing up the latest patch for the next beta-sim run to see if it fixes it?

Flags: needinfo?(twisniewski) → needinfo?(ncsoregi)

(In reply to Thomas Wisniewski [:twisniewski] from comment #8)

Natalia, would you mind queueing up the latest patch for the next beta-sim run to see if it fixes it?

Yes, will import it in today's beta-sim run and will also update the bug with the results

Flags: needinfo?(ncsoregi)
Pushed by twisniewski@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/1a37e28f12b1
mark the about:compat process script as intermittently loaded in the browser_startup_content mochitest; r=kmag

Alright, then I've gone ahead and updated the comments in the patch as requested in the r+, and am landing the patch via Lando.

Status: NEW → RESOLVED
Closed: 8 months ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.