Closed Bug 1388955 Opened 7 years ago Closed 7 years ago

Intermittent Windows 10 gfx/layers/apz/test/mochitest/test_group_touchevents.html | helper_iframe_pan.html | check that the iframe scrolled - got 49, expected 50

Categories

(Core :: Panning and Zooming, defect, P5)

57 Branch
x86_64
Windows 10
defect

Tracking

()

RESOLVED FIXED
mozilla57
Tracking Status
firefox57 --- fixed

People

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

Details

(Keywords: intermittent-failure, Whiteboard: [gfx-noted][stockwell fixed:other])

Attachments

(1 file)

Filed by: philringnalda [at] gmail.com

https://treeherder.mozilla.org/logviewer.html#?job_id=122093912&repo=mozilla-inbound

https://queue.taskcluster.net/v1/task/cMSOqtOSRRSLA4kanzl-3w/runs/0/artifacts/public/logs/live_backing.log

Not some new bustage, we just started running mochitests on Win10 today (other than a few misstars of try runs as a different bug).
OS: Unspecified → Windows 10
Hardware: Unspecified → x86_64
Whiteboard: [gfx-noted]
Version: unspecified → 57 Branch
Seems frequent, I'll investigate.
Assignee: nobody → bugmail
Reproduced in https://treeherder.mozilla.org/#/jobs?repo=try&revision=cf12857a2378d83ad216a34ac661005bc1c05a2d with logging. Seems like occasionally Windows eats a touchmove right at the start of the drag sequence.

Good case:

16:32:50     INFO -  GECKO(2572) | 00000245B8E93000 got a touch-start ((27,117)) in state 0
16:32:50     INFO -  GECKO(2572) | 00000245B8E93000 got a touch-move ((27,116)) in state 2
16:32:50     INFO -  GECKO(2572) | 00000245B8E93000 got a touch-move ((27,115)) in state 3
16:32:50     INFO -  GECKO(2572) | 00000245B8E93000 got a touch-move ((27,114)) in state 3
...
16:32:50     INFO -  GECKO(2572) | 00000245B8E93000 got a touch-move ((27,68)) in state 3
16:32:50     INFO -  GECKO(2572) | 00000245B8E93000 got a touch-move ((27,67)) in state 3
16:32:50     INFO -  GECKO(2572) | 00000245B8E93000 got a touch-move ((27,66)) in state 3
16:32:50     INFO -  GECKO(2572) | 00000245B8E93000 got a touch-end in state 3

Bad case:

16:32:51     INFO -  GECKO(13620) | 000001777F7FF000 got a touch-start ((27,117)) in state 0
16:32:51     INFO -  GECKO(13620) | 000001777F7FF000 got a touch-move ((27,115)) in state 2
16:32:51     INFO -  GECKO(13620) | 000001777F7FF000 got a touch-move ((27,114)) in state 3
...
16:32:51     INFO -  GECKO(13620) | 000001777F7FF000 got a touch-move ((27,67)) in state 3
16:32:51     INFO -  GECKO(13620) | 000001777F7FF000 got a touch-move ((27,66)) in state 3
16:32:51     INFO -  GECKO(13620) | 000001777F7FF000 got a touch-end in state 3

So in the "bad case" APZ only gets 49 pixels of scroll. On the one hand this can probably be fixed by setting apz.test.fails_with_native_injection=true on this test, but I'd like to avoid doing that everywhere since it's better to use the native injection where we can as it's more representative of production behaviour. Another alternative is just making the test a little fuzzy so it allows a value of 49 (or anything in the range 45-50) as success.
Whiteboard: [gfx-noted] → [gfx-noted][stockwell needswork]
Comment on attachment 8897124 [details]
Bug 1388955 - Allow helper_iframe_pan to scroll a little less than the expected 50 pixels on Windows.

https://reviewboard.mozilla.org/r/168412/#review173666
Attachment #8897124 - Flags: review?(botond) → review+
Pushed by kgupta@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/86ba3b37f10c
Allow helper_iframe_pan to scroll a little less than the expected 50 pixels on Windows. r=botond
https://hg.mozilla.org/mozilla-central/rev/86ba3b37f10c
Status: NEW → RESOLVED
Closed: 7 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla57
Whiteboard: [gfx-noted][stockwell needswork] → [gfx-noted][stockwell fixed:other]
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: