Video controls cut and flickering when fading away

VERIFIED FIXED in Firefox 32

Status

()

defect
VERIFIED FIXED
5 years ago
5 years ago

People

(Reporter: FlorinMezei, Assigned: mattwoodrow)

Tracking

({regression})

31 Branch
mozilla33
x86_64
Linux
Points:
---
Dependency tree / graph

Firefox Tracking Flags

(firefox30 unaffected, firefox31- wontfix, firefox32+ verified, firefox33+ verified, firefox-esr31 wontfix)

Details

Attachments

(2 attachments, 1 obsolete attachment)

Reproducible on:
Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Firefox/31.0
20140619131915
Mozilla/5.0 (X11; Linux x86_64; rv:32.0) Gecko/20100101 Firefox/32.0
20140619004025
Mozilla/5.0 (X11; Linux x86_64; rv:33.0) Gecko/20100101 Firefox/33.0
20140619030203

Steps to reproduce:
1. Launch Firefox with a clean profile.
2. Go to, for example https://air.mozilla.org/webrtc-meet-up-francophone-paris/.
3. Play the video. 
4. During playback hover the video are to get the video controls, then move the cursor outside the video area.

Expected results:
When the cursor is moved outside the video area, the video controls should fade away in a 2 sec animation.

Actual results:
When the cursor is moved outside the video area, the video controls are cut, as part disappears instantly, while part continues to flicker for about 2 seconds and then disappears.

Notes: 
1. The issue appeared after fix for bug 994282.
2. On Windows and Mac there are no issues, as the controls fade away in a 2 sec second animation.
Matt, since you landed the patch for bug 994282, could you please have a look at this?
Flags: needinfo?(matt.woodrow)
Keywords: regression
Assignee: nobody → matt.woodrow
Attachment #8445596 - Flags: review?(roc)
Flags: needinfo?(matt.woodrow)
Comment on attachment 8445596 [details] [diff] [review]
Account for device offset for thebes layer masks

Review of attachment 8445596 [details] [diff] [review]:
-----------------------------------------------------------------

r+ on the other part.

::: widget/cocoa/nsChildView.mm
@@ +3723,5 @@
>    nsAutoRetainCocoaObject kungFuDeathGrip(self);
>    bool painted = false;
>    if (mGeckoChild->GetLayerManager()->GetBackendType() == LayersBackend::LAYERS_BASIC) {
>      nsBaseWidget::AutoLayerManagerSetup
> +      setupLayerManager(mGeckoChild, targetContext, BufferMode::BUFFERED);

Is this change supposed to be part of this patch?
(In reply to Robert O'Callahan (:roc) (Mozilla Corporation) from comment #4)
> Comment on attachment 8445596 [details] [diff] [review]
> Account for device offset for thebes layer masks
> 
> Review of attachment 8445596 [details] [diff] [review]:
> -----------------------------------------------------------------
> 
> r+ on the other part.
> 
> ::: widget/cocoa/nsChildView.mm
> @@ +3723,5 @@
> >    nsAutoRetainCocoaObject kungFuDeathGrip(self);
> >    bool painted = false;
> >    if (mGeckoChild->GetLayerManager()->GetBackendType() == LayersBackend::LAYERS_BASIC) {
> >      nsBaseWidget::AutoLayerManagerSetup
> > +      setupLayerManager(mGeckoChild, targetContext, BufferMode::BUFFERED);
> 
> Is this change supposed to be part of this patch?

Nope, accidental qref.
Attachment #8445596 - Attachment is obsolete: true
Attachment #8445596 - Flags: review?(roc)
Attachment #8447791 - Flags: review?(roc)
https://hg.mozilla.org/mozilla-central/rev/f79302ac0ede
Status: NEW → RESOLVED
Closed: 5 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla33
Verified as fixed on Ubuntu 13.04 x64 (User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:33.0) Gecko/20100101 Firefox/33.0) with latest Firefox 33 Nightly (BuildID: 20140702030201). The video controls now fade away in a 2 second animation as expected.
Status: RESOLVED → VERIFIED
Given QA has verified this fixed it would be good to uplift this, if it's low risk.
When will this be uplifted? We are getting really close to the release, and there is risk we wont be able to land it for Firefox 31. :/ Matt, what blocks you here from backporting? See bug 1035114 for another graphical fallout, which this patch would fix.
Flags: needinfo?(matt.woodrow)
Yeh, the patch seems low risk and have been in m-c for a week, I could take it in 31 in the RC.
Comment on attachment 8447791 [details] [diff] [review]
Account for device offset for thebes layer masks v2

Approval Request Comment
[Feature/regressing bug #]: Bug 994282
[User impact if declined]: Broken mask layers, reported issue with video controls
[Describe test coverage new/current, TBPL]: Tested manually and verified.
[Risks and why]: Very low risk, just extends an earlier change to cover all layer types.
[String/UUID change made/needed]: None.
Attachment #8447791 - Flags: approval-mozilla-beta?
Attachment #8447791 - Flags: approval-mozilla-aurora?
Flags: needinfo?(matt.woodrow)
Attachment #8447791 - Flags: approval-mozilla-beta?
Attachment #8447791 - Flags: approval-mozilla-beta+
Attachment #8447791 - Flags: approval-mozilla-aurora?
Attachment #8447791 - Flags: approval-mozilla-aurora+
Comment on attachment 8447791 [details] [diff] [review]
Account for device offset for thebes layer masks v2

Actually, I changed my mind for beta. I will however take for ESR 31.1.0
Attachment #8447791 - Flags: approval-mozilla-beta+
Keywords: verifyme
Works great with Mozilla/5.0 (X11; Linux x86_64; rv:32.0) Gecko/20100101 Firefox/32.0 ID:20140715004003 CSet: d32649a24965. Thanks for the backport!
Keywords: verifyme
Without active requests from the ESR community, this does not meet ESR landing criteria.
You need to log in before you can comment on or make changes to this bug.