[Webkit-unassigned] [Bug 56514] [chromium] Properly reset VideoLayerChromium textures after lost renderer context

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=56514





--- Comment #4 from Kenneth Russell <kbr at google.com>  2011-03-17 14:00:34 PST ---
(From update of attachment 86017)
View in context: https://bugs.webkit.org/attachment.cgi?id=86017&action=review

>>> Source/WebCore/platform/graphics/chromium/VideoLayerChromium.cpp:96
>>> +        resetFrameParameters();
>> 
>> LayerChromium::setLayerRenderer() will call cleanupResources() if (layerRenderer() && layerRenderer() != renderer), which will call releaseCurrentFrame() which contains a call to resetFrameParameters(). Doesn't that make this override redundant?
> 
> Good question! It's not redundant because m_currentFrame can be null, and in that case releaseCurrentFrame() will not call resetFrameParameters() because it exits out early.
> 
> And actually, saveCurrentFrame/releaseCurrentFrame are hooks to be used when we have hardware decoding implemented in Chrome. Right now we don't have hardware decode implemented, so m_currentFrame is _always_ null.

Wouldn't it be better to move the call to resetFrameParameters() to the top of releaseCurrentFrame() in that case? It shouldn't be necessary to override setLayerRenderer unless something more than simple resource reallocation needs to occur in response.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.