Closed Bug 1210882 Opened 9 years ago Closed 6 years ago

z-order is wrong on CSS 3D animation

Categories

(Core :: Layout, defect)

41 Branch
All
Windows 7
defect
Not set
normal

Tracking

()

RESOLVED WORKSFORME
Tracking Status
firefox41 --- wontfix
firefox42 + wontfix
firefox43 + wontfix
firefox44 - wontfix
firefox45 - wontfix
firefox-esr38 --- unaffected

People

(Reporter: alice0775, Unassigned, NeedInfo)

References

Details

(Keywords: regression)

Attachments

(3 files)

Attached file index.html (deleted) —
[Tracking Requested - why for this release]: Regression since Firefox41

I found a bug when I text Bug 1210784.

Steps To Reproduce:
1. Open attached

Actual Results:
z-order is wrong on CSS 3D animation.
See attached screenshot.


Pushlog:
https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=ef53c6c25fb1&tochange=eed5d2d610e2

Regressed by:
eed5d2d610e2	L. David Baron — Bug 980770 - Enable off-main-thread animations on all platforms other than Linux (and other X11 based platforms) with off-main-thread compositing, for nightly/aurora. r=birtles
Attached image screenshot (deleted) —
Flags: needinfo?(dbaron)
Setting layers.offmainthreadcomposition.async-animations = false fixes the z-order problem.
I haven't looked closely yet, but it might be that the right fix for this is bug 1097464.
Looks like we will be shipping 42 with this regression.
David, do you think it is worth tracking it in 43? Thanks
[Tracking Requested - why for this release]:

Is this still affecting 43 and 44?  If it's looking ok in 44, maybe bug 1097464 did fix it. 

Either way, I think we will be wontfixing this for 43.  If anyone disagrees or has a patch that looks upliftable, please let me know.
Flags: needinfo?(ehumphries)
Flags: needinfo?(andrei.vaida)
Unable to reproduce this in 45.0a1 (2015-11-11) MacOS 10.11.1 (15B42)
Flags: needinfo?(ehumphries)
(In reply to Liz Henry (:lizzard) (needinfo? me) from comment #6)
> [Tracking Requested - why for this release]:
> 
> Is this still affecting 43 and 44?  If it's looking ok in 44, maybe bug
> 1097464 did fix it. 
> 
> Either way, I think we will be wontfixing this for 43.  If anyone disagrees
> or has a patch that looks upliftable, please let me know.

I think OMTA should be simply disabled for 43.
(In reply to Alice0775 White from comment #8)
> I think OMTA should be simply disabled for 43.

I disagree, at least on the basis of this bug (given that I believe we still haven't shipped correct z-ordering for 3-D transforms in any release).
I just re-read this and now see what Alice0774 was describing and it is *still* happening in 45 nightly.
I'm still seeing this on Windows 7 64-bit using:
- latest Aurora 44.0a2, build ID 20151111004055
- Firefox 43 beta 2, build ID 20151109145326.
Flags: needinfo?(andrei.vaida)
Thinker, I am trying to follow up on 44+ tracked bugs and wondering if you might be able to help with this one. I know you worked on bug 1210882 that DBaron pointed to, so any help would be awesome.
Flags: needinfo?(tlee)
Sure! I would look into it.
Flags: needinfo?(tlee)
Stop tracking this one as we shipped already several versions with it.
Please resubmit when we are ready to take that in a branch. Thanks
Flags: needinfo?(dbaron)

I can reproduce the issue on 45.0esr windows10.
However I cannot reproduce the issue on Firefox60esr, 64.02, 65.0 and Nightly66.0a1(no WebRender) windows10.

Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → WORKSFORME

It would be nice to have the test case in comment 0 as a reftest. Also for future references, it would be nice to know which change fixed this issue.

I'd really like a reftest for this. Hopefully I can take time for that some day.

Flags: needinfo?(hikezoe)
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: