Bug 104195 – [WK2][EFL] Gardening. Rebase fast/block/basic/020.html

• Component: WebKit EFL
• Status: RESOLVED FIXED

[KyungTae Kim]

The layer for "visibility:hidden; position: absolute;" object should be shown in RenderLayer Tree when USE(ACCELERATED_COMPOSITING).

It's because the HiddenLayers are included in zOrderList when CompositingMode is on.

void RenderLayer::rebuildZOrderLists()
{
#if USE(ACCELERATED_COMPOSITING)
    bool includeHiddenLayers = compositor()->inCompositingMode();  // TRUE or FALSE by RenderLayerCompositor::m_compositing
#else
    bool includeHiddenLayers = false; // Always FALSE -> other ports' case
#endif
    for (RenderLayer* child = firstChild(); child; child = child->nextSibling())
        if (!m_reflection || reflectionLayer() != child)
            child->collectLayers(includeHiddenLayers, m_posZOrderList, m_negZOrderList);

[KyungTae Kim]

Created attachment 177890 [details]
Patch

[Kangil Han]

(In reply to comment #0)
> The layer for "visibility:hidden; position: absolute;" object should be shown in RenderLayer Tree when USE(ACCELERATED_COMPOSITING).
> 
> It's because the HiddenLayers are included in zOrderList when CompositingMode is on.
> 
> void RenderLayer::rebuildZOrderLists()
> {
> #if USE(ACCELERATED_COMPOSITING)
>     bool includeHiddenLayers = compositor()->inCompositingMode();  // TRUE or FALSE by RenderLayerCompositor::m_compositing
> #else
>     bool includeHiddenLayers = false; // Always FALSE -> other ports' case
> #endif
>     for (RenderLayer* child = firstChild(); child; child = child->nextSibling())
>         if (!m_reflection || reflectionLayer() != child)
>             child->collectLayers(includeHiddenLayers, m_posZOrderList, m_negZOrderList);

Love to know this! :)

[Ryuan Choi]

Comment on attachment 177890 [details]
Patch

I think that this is reasonable.

[WebKit Review Bot]

Comment on attachment 177890 [details]
Patch

Clearing flags on attachment: 177890

Committed r136786: <http://trac.webkit.org/changeset/136786>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

[KyungTae Kim]

(In reply to comment #2)
> (In reply to comment #0)
Just for reference)
That code was originated from the bug 38829,
fixing a bug that the first frame of the <video> was displayed even though the video has been set to hidden.