Closed Bug 1509242 Opened 6 years ago Closed 5 years ago

Categories

(Core :: CSS Parsing and Computation, defect, P5)

defect

Tracking

()

RESOLVED INCOMPLETE

People

(Reporter: intermittent-bug-filer, Unassigned)

References

(Regression)

Details

(Keywords: intermittent-failure, regression, Whiteboard: [comment 10])

Filed by: aciure [at] mozilla.com https://treeherder.mozilla.org/logviewer.html#?job_id=213318098&repo=autoland https://queue.taskcluster.net/v1/task/Y3Ge_svCR36Z-1yB04PBOg/runs/0/artifacts/public/logs/live_backing.log https://hg.mozilla.org/mozilla-central/raw-file/tip/layout/tools/reftest/reftest-analyzer.xhtml#logurl=https://queue.taskcluster.net/v1/task/Y3Ge_svCR36Z-1yB04PBOg/runs/0/artifacts/public/logs/live_backing.log&only_show_unexpected=1 20:25:45 INFO - TEST-START | /css/CSS2/colors/color-001.xht [...] 20:25:45 INFO - PID 707 | 1542860745578 Marionette INFO No differences allowed 20:25:45 INFO - PID 707 | 1542860745578 Marionette INFO Found 8735 pixels different, maximum difference per channel 255 20:25:45 INFO - TEST-UNEXPECTED-FAIL | /css/CSS2/colors/color-001.xht | Testing http://web-platform.test:8000/css/CSS2/colors/color-001.xht == http://web-platform.test:8000/css/CSS2/colors/color-001-ref.xht 20:25:45 INFO - TEST-INFO took 540ms
The screenshot shows a lot of large "X" characters that are expected to render as squares (but are instead rendering as "X"). Looks like the Ahem font failed to load. This is OS X where Ahem+WPT was having too many issues to be usable at all, until some recent change, IIRC -- this might be some lingering bit of whatever was broken before, perhaps?
See Also: → 1425698
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → INCOMPLETE
Status: RESOLVED → REOPENED
Resolution: INCOMPLETE → ---

I believe the bug-reopening here was for this recent failure:
https://treeherder.mozilla.org/logviewer.html#?job_id=252329030&repo=mozilla-beta
...which is a case where we failed every css/CSS2/colors/color-*.xht testcase it seems, due to Ahem failing to load on mac, as described in comment 1.

(I'm not sure if we failed "comprehensively" like this [every test] in comment 0 as well, though I'm betting we probably did. The log from comment 0 has expired now, so it's hard to know.)

James recent occurrences seems to be from bug 1505739, can you take a look?

Flags: needinfo?(james)
Regressed by: 1505739
Keywords: regression

(In reply to Andreea Pavel [:apavel] from comment #8)

James recent occurrences seems to be from bug 1505739, can you take a look?

I'm not James, but could you clarify what you mean? That bug's patch landed 8 months ago, which is not really "recent". :)

(It is possible that that bug may be where this started, since it's probably where these tests became enabled. But I'm curious if there's something else I'm misunderstanding RE recent failures and/or recent landings.)

Flags: needinfo?(apavel)

AFAICT there is a possible, apparently low-frequency, intermittent failure mode whereby the Ahem font isn't loading at all on macOS for some reason. There has been work to move away from using the system font and instead prefer to use webfonts, so I think this issue will go away on its own at some point in the near future.

Flags: needinfo?(james)

(In reply to Daniel Holbert [:dholbert] from comment #9)

(In reply to Andreea Pavel [:apavel] from comment #8)

James recent occurrences seems to be from bug 1505739, can you take a look?

I'm not James, but could you clarify what you mean? That bug's patch landed 8 months ago, which is not really "recent". :)

(It is possible that that bug may be where this started, since it's probably where these tests became enabled. But I'm curious if there's something else I'm misunderstanding RE recent failures and/or recent landings.)

i just looked for the last changes on that file in mozilla-files: https://hg.mozilla.org/mozilla-central/log?rev=css%2FCSS2%2Fcolors%2Fcolor-001.xht

Flags: needinfo?(apavel)
Whiteboard: [comment 10]
Status: REOPENED → RESOLVED
Closed: 6 years ago5 years ago
Resolution: --- → INCOMPLETE
Has Regression Range: --- → yes
You need to log in before you can comment on or make changes to this bug.