cc: Fix overlay damage tracking bug

cc: Fix overlay damage tracking bug

Track the region that is covered by overlays and ensure that that is
included in the damage rect of future frames. This fixes bugs where
something that was previously an overlay becomes a non-overlay due to
reasons that didn't involve doing damage to the frame in that quad's
region (e.g, creating a slightly overlapping overlay quad).

Move the removal of overlays from the damage rect to being done in the
OverlayProcessor instead of being done in each individual strategy.

BUG=551861
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Committed: https://crrev.com/cb27ead70f32946c71ad1d6672437dbd914b5b6d
Cr-Commit-Position: refs/heads/master@{#360513}

[ccameron, 2015-11-17 00:05:39]

ccameron@chromium.org changed reviewers:
+ dcastagna@chromium.org

[ccameron, 2015-11-17 00:05:39]

This fixes the overlapping video bug that we were looking at.

[ccameron, 2015-11-17 21:12:53]

ccameron@chromium.org changed reviewers:
+ watk@chromium.org

[watk, 2015-11-17 23:22:59]

On 2015/11/17 21:12:53, ccameron wrote:

lgtm

[kalyank, 2015-11-17 23:29:19]

kalyan.kondapally@intel.com changed reviewers:
+ kalyan.kondapally@intel.com

[kalyank, 2015-11-17 23:29:20]

https://codereview.chromium.org/1452943002/diff/20001/cc/output/overlay_proce...
File cc/output/overlay_processor.cc (right):

https://codereview.chromium.org/1452943002/diff/20001/cc/output/overlay_proce...
cc/output/overlay_processor.cc:84: if (overlay.plane_z_order > 0) {
We should also check if the quad being composited using Overlay is transparent,
in that case we cannot subtract the damage rect.

[ccameron, 2015-11-18 20:16:30]

On 2015/11/17 23:29:20, kalyank wrote:
>
https://codereview.chromium.org/1452943002/diff/20001/cc/output/overlay_proce...
> File cc/output/overlay_processor.cc (right):
> 
>
https://codereview.chromium.org/1452943002/diff/20001/cc/output/overlay_proce...
> cc/output/overlay_processor.cc:84: if (overlay.plane_z_order > 0) {
> We should also check if the quad being composited using Overlay is
transparent,
> in that case we cannot subtract the damage rect.

We currently assume that all overlays are opaque (we remove them from the damage
rect -- if they had visible content behind them, then we would see stale data
behind them). I agree that we should be more explicit about this requirement.

[kalyank, 2015-11-18 21:03:29]

On 2015/11/18 20:16:30, ccameron wrote:
> On 2015/11/17 23:29:20, kalyank wrote:
> >
>
https://codereview.chromium.org/1452943002/diff/20001/cc/output/overlay_proce...
> > File cc/output/overlay_processor.cc (right):
> > 
> >
>
https://codereview.chromium.org/1452943002/diff/20001/cc/output/overlay_proce...
> > cc/output/overlay_processor.cc:84: if (overlay.plane_z_order > 0) {
> > We should also check if the quad being composited using Overlay is
> transparent,
> > in that case we cannot subtract the damage rect.
> 
> We currently assume that all overlays are opaque (we remove them from the
damage
> rect -- if they had visible content behind them, then we would see stale data
> behind them). I agree that we should be more explicit about this requirement.

Ya, I forgot that we don't actually promote quads which need blending/opacity
to Overlays (Checks in OverlayCandidate).

[kalyank, 2015-11-18 21:18:33]

https://codereview.chromium.org/1452943002/diff/40001/cc/output/overlay_proce...
File cc/output/overlay_processor.cc (right):

https://codereview.chromium.org/1452943002/diff/40001/cc/output/overlay_proce...
cc/output/overlay_processor.cc:69: if (ProcessForCALayers(resource_provider,
render_passes, candidates,
What happens on platforms which don't use CALayers ? I didn't find any early opt
out in ProcessForCALayers.

[ccameron, 2015-11-18 21:22:12]

https://codereview.chromium.org/1452943002/diff/40001/cc/output/overlay_proce...
File cc/output/overlay_processor.cc (right):

https://codereview.chromium.org/1452943002/diff/40001/cc/output/overlay_proce...
cc/output/overlay_processor.cc:69: if (ProcessForCALayers(resource_provider,
render_passes, candidates,
On 2015/11/18 21:18:33, kalyank wrote:
> What happens on platforms which don't use CALayers ? I didn't find any early
opt
> out in ProcessForCALayers.

The overlay_validator->AllowCALayerOverlays() call will return false on
platforms that don't support CALayers.

[ccameron, 2015-11-19 02:39:49]

The CQ bit was checked by ccameron@chromium.org

[ccameron, 2015-11-19 02:39:50]

The patchset sent to the CQ was uploaded after l-g-t-m from watk@chromium.org
Link to the patchset: https://codereview.chromium.org/1452943002/#ps60001
(title: "Add buffer queue update")

[commit-bot: I haz the power, 2015-11-19 02:40:11]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1452943002/60001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1452943002/60001

[commit-bot: I haz the power, 2015-11-19 03:35:00]

Committed patchset #4 (id:60001)

[commit-bot: I haz the power, 2015-11-19 03:36:01]

Patchset 4 (id:??) landed as
https://crrev.com/cb27ead70f32946c71ad1d6672437dbd914b5b6d
Cr-Commit-Position: refs/heads/master@{#360513}