Closed Bug 1740413 Opened 4 years ago Closed 4 years ago

High frequency [tier2] bugs/1738700-1.html == bugs/1738700-1-ref.html | image comparison, max difference: 3, number of differing pixels: 1177

Categories

(Core :: Layout, defect, P5)

defect

Tracking

()

RESOLVED FIXED
96 Branch
Tracking Status
firefox-esr91 --- unaffected
firefox94 --- unaffected
firefox95 --- unaffected
firefox96 --- fixed

People

(Reporter: intermittent-bug-filer, Assigned: jfkthame)

References

(Regression)

Details

(Keywords: intermittent-failure, regression)

Attachments

(2 files)

Regressed by: 1727943
No longer regressed by: 1727943

Jonathan, this looks to be from Bug 1738700 https://hg.mozilla.org/integration/autoland/rev/3bd4e0246b821a1239b1c184b9e5bd8a99acbf25. Could you have a look over it? It's only android-hw-p2-8-0-android-aarch64-qr so far. Thank you.

Flags: needinfo?(jkew)
Regressed by: 1738700
Has Regression Range: --- → yes

Looks like some Android systems show more antialiasing fuzz (sigh), though I see it's not entirely consistent. With a max-difference of 3, though, it's not really a concern; we can just annotate it accordingly.

Flags: needinfo?(jkew)
Assignee: nobody → jfkthame
Status: NEW → ASSIGNED

Set release status flags based on info from the regressing bug 1738700

Pushed by jkew@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/7b18da1fbd9c Adjust fuzz annotation for Android antialiasing variability. r=lsalzman
Status: ASSIGNED → RESOLVED
Closed: 4 years ago
Resolution: --- → FIXED
Target Milestone: --- → 96 Branch

This is still happening: https://treeherder.mozilla.org/logviewer?job_id=357735273&repo=autoland&lineNumber=1930

Jonathan, could you take a look?

Status: RESOLVED → REOPENED
Flags: needinfo?(jfkthame)
Resolution: FIXED → ---
Target Milestone: 96 Branch → ---

That shows a difference on 1328 pixels instead of 1177, though still only a max-difference of 3 (insignificant). Not sure why it's so variable between different test runs, as they're all "passing" in the sense that the behavior being tested is working correctly; there's just some kind of difference between the precise antialiasing that we see. (Do all these tests always run on identical systems, or are there hardware- or driver-level differences that might affect GPU rendering, I wonder?)

Anyhow, the answer here is simply to adjust the fuzz number.

Flags: needinfo?(jfkthame)

Set release status flags based on info from the regressing bug 1738700

Pushed by jkew@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/54d201a9118a Adjust Android fuzz a bit more, for inconsistent antialiasing results. r=lsalzman
Status: REOPENED → RESOLVED
Closed: 4 years ago4 years ago
Resolution: --- → FIXED
Target Milestone: --- → 96 Branch
See Also: → 1740922
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: