Closed Bug 1694887 Opened 3 years ago Closed 3 years ago

Perma [tier 2] Win AArch64 text-stroke/webkit-text-stroke-property-001.html == text-stroke/webkit-text-stroke-property-001-ref.html | image comparison, max difference: 80, number of differing pixels: 2822

Categories

(Core :: Layout, defect, P5)

defect

Tracking

()

RESOLVED FIXED

People

(Reporter: intermittent-bug-filer, Unassigned)

Details

(Keywords: intermittent-failure)

Attachments

(1 file)

Filed by: malexandru [at] mozilla.com
Parsed log: https://treeherder.mozilla.org/logviewer?job_id=331148937&repo=mozilla-central
Full log: https://firefox-ci-tc.services.mozilla.com/api/queue/v1/task/R9aObJWdTe2OI-3aqwISAg/runs/0/artifacts/public/logs/live_backing.log
Reftest URL: https://hg.mozilla.org/mozilla-central/raw-file/tip/layout/tools/reftest/reftest-analyzer.xhtml#logurl=https://firefox-ci-tc.services.mozilla.com/api/queue/v1/task/R9aObJWdTe2OI-3aqwISAg/runs/0/artifacts/public/logs/live_backing.log&only_show_unexpected=1


[task 2021-02-25T06:35:25.630Z] 06:35:25     INFO - REFTEST TEST-START | layout/reftests/text-stroke/webkit-text-stroke-property-001.html == layout/reftests/text-stroke/webkit-text-stroke-property-001-ref.html
[task 2021-02-25T06:35:25.650Z] 06:35:25     INFO - REFTEST TEST-LOAD | file:///C:/tasks/task_1614203958/build/tests/reftest/tests/layout/reftests/text-stroke/webkit-text-stroke-property-001.html | 0 / 6 (0%)
[task 2021-02-25T06:35:25.878Z] 06:35:25     INFO - REFTEST INFO | drawWindow flags = DRAWWINDOW_DRAW_CARET | DRAWWINDOW_DRAW_VIEW | DRAWWINDOW_USE_WIDGET_LAYERS; window size = 816,1039; test browser size = 800,1000
[task 2021-02-25T06:35:25.900Z] 06:35:25     INFO - REFTEST TEST-LOAD | file:///C:/tasks/task_1614203958/build/tests/reftest/tests/layout/reftests/text-stroke/webkit-text-stroke-property-001-ref.html | 0 / 6 (0%)
[task 2021-02-25T06:35:25.946Z] 06:35:25     INFO - REFTEST INFO | REFTEST fuzzy test (0, 0) <= (80, 2822) <= (64, 52)
[task 2021-02-25T06:35:26.169Z] 06:35:26     INFO - REFTEST TEST-UNEXPECTED-FAIL | layout/reftests/text-stroke/webkit-text-stroke-property-001.html == layout/reftests/text-stroke/webkit-text-stroke-property-001-ref.html | image comparison, max difference: 80, number of differing pixels: 2822
[task 2021-02-25T06:35:26.171Z] 06:35:26     INFO - REFTEST   IMAGE 1 (TEST):
[task 2021-02-25T06:35:26.171Z] 06:35:26     INFO - REFTEST   IMAGE 2 (REFERENCE):
[task 2021-02-25T06:35:26.172Z] 06:35:26     INFO - REFTEST TEST-END | layout/reftests/text-stroke/webkit-text-stroke-property-001.html == layout/reftests/text-stroke/webkit-text-stroke-property-001-ref.html
[task 2021-02-25T06:35:26.173Z] 06:35:26     INFO - REFTEST TEST-START | layout/reftests/text-stroke/webkit-text-stroke-property-002.html == layout/reftests/text-stroke/webkit-text-stroke-property-002-ref.html
[task 2021-02-25T06:35:26.173Z] 06:35:26     INFO - REFTEST TEST-LOAD | file:///C:/tasks/task_1614203958/build/tests/reftest/tests/layout/reftests/text-stroke/webkit-text-stroke-property-002.html | 1 / 6 (16%)
[task 2021-02-25T06:35:26.177Z] 06:35:26     INFO - REFTEST TEST-LOAD | file:///C:/tasks/task_1614203958/build/tests/reftest/tests/layout/reftests/text-stroke/webkit-text-stroke-property-002-ref.html | 1 / 6 (16%)
[task 2021-02-25T06:35:26.212Z] 06:35:26     INFO - REFTEST INFO | REFTEST fuzzy test (0, 0) <= (125, 3725) <= (4, 24)
[task 2021-02-25T06:35:26.456Z] 06:35:26     INFO - REFTEST TEST-UNEXPECTED-FAIL | layout/reftests/text-stroke/webkit-text-stroke-property-002.html == layout/reftests/text-stroke/webkit-text-stroke-property-002-ref.html | image comparison, max difference: 125, number of differing pixels: 3725
[task 2021-02-25T06:35:26.458Z] 06:35:26     INFO - REFTEST   IMAGE 1 (TEST):
[task 2021-02-25T06:35:26.462Z] 06:35:26     INFO - REFTEST   IMAGE 2 (REFERENCE):
[task 2021-02-25T06:35:26.464Z] 06:35:26     INFO - REFTEST TEST-END | layout/reftests/text-stroke/webkit-text-stroke-property-002.html == layout/reftests/text-stroke/webkit-text-stroke-property-002-ref.html
[task 2021-02-25T06:35:26.465Z] 06:35:26     INFO - REFTEST TEST-START | layout/reftests/text-stroke/webkit-text-stroke-property-003.html == layout/reftests/text-stroke/webkit-text-stroke-property-003-ref.html
[task 2021-02-25T06:35:26.465Z] 06:35:26     INFO - REFTEST TEST-LOAD | file:///C:/tasks/task_1614203958/build/tests/reftest/tests/layout/reftests/text-stroke/webkit-text-stroke-property-003.html | 2 / 6 (33%)
[task 2021-02-25T06:35:26.468Z] 06:35:26     INFO - REFTEST TEST-LOAD | file:///C:/tasks/task_1614203958/build/tests/reftest/tests/layout/reftests/text-stroke/webkit-text-stroke-property-003-ref.html | 2 / 6 (33%)
[task 2021-02-25T06:35:26.485Z] 06:35:26     INFO - REFTEST INFO | REFTEST fuzzy test (0, 0) <= (74, 2596) <= (48, 28)
[task 2021-02-25T06:35:26.699Z] 06:35:26     INFO - REFTEST TEST-UNEXPECTED-FAIL | layout/reftests/text-stroke/webkit-text-stroke-property-003.html == layout/reftests/text-stroke/webkit-text-stroke-property-003-ref.html | image comparison, max difference: 74, number of differing pixels: 2596```

Should anything be done about these Windows AArch reftest failures? They are almost permanent (if the task doesn't fail with bug 1572900 instead).

Severity: S4 → --
Flags: needinfo?(dholbert)
Priority: P5 → --

Thanks for the heads-up. These look like minor antialiasing issues -- subtle painting differences between the way that we're drawing the outline in the testcase vs. the reference case. It's not an actual issue from a user's perspective, and not something we should worry about.

I'd suggest adding fuzzy annotations, specific to this platform & tuned to the levels of fuzziness that we're seeing in the failure logs here.

Is that something that a sheriff could take care of? (I'd be happy to r+.) If not, I could do it as well, though I'm focused on other things at the moment.

Flags: needinfo?(dholbert) → needinfo?(aryx.bugmail)
Assignee: nobody → apavel
Status: NEW → ASSIGNED
Pushed by apavel@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/a836a820c250
adjust reftest fuzzy annotations r=dholbert DONTBUILD
Assignee: apavel → nobody
Status: ASSIGNED → NEW
Keywords: leave-open
Flags: needinfo?(aryx.bugmail)
Status: NEW → RESOLVED
Closed: 3 years ago
Resolution: --- → FIXED
Assignee: nobody → apavel
Assignee: apavel → nobody
Status: RESOLVED → REOPENED
Keywords: leave-open
Resolution: FIXED → ---

Andreaa, why reopening? The test is fuzzy after all, so there doesn't seem to be much else actionable here? We could investigate the difference in our windows font code or something but given the test is also fuzzy in tier1 platforms it doesn't seem particularly valuable.

Severity: -- → N/A
Flags: needinfo?(apavel)
Priority: -- → P5

Yeah, let's close this - this feels like a situation where the feature is just hard to craft an exactly precise reference case without using the feature itself, and the test is doing as good a job as it can, which involves some fuzzy mismatch. So I think this is as "FIXED" as it's going to get.

Thanks again for adding these annotations, Andreea!

Status: REOPENED → RESOLVED
Closed: 3 years ago3 years ago
Flags: needinfo?(apavel)
Keywords: leave-open
Resolution: --- → FIXED

Hi, we usually leave-open all bugs for which we update annotations or disable the tests to keep track that they weren't actually fixed.
Thank you for clarifying why in this case updating the annotation is the actual fix. I'll ask in the future before reopening.

You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: