cc: disable pre-painting when using gpu rasterization.

cc: disable pre-painting when using gpu rasterization.

R=enne@chromium.org
BUG=333045
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=245479

[ernstm, 2014-01-14 22:48:48]

PTAL

[danakj, 2014-01-14 22:52:19]

https://codereview.chromium.org/137993013/diff/1/cc/trees/layer_tree_host_imp...
File cc/trees/layer_tree_host_impl.cc (right):

https://codereview.chromium.org/137993013/diff/1/cc/trees/layer_tree_host_imp...
cc/trees/layer_tree_host_impl.cc:1651: settings_.gpu_rasterization) {
Doesn't this affect all layers including ones that are not being gpu rasterized?

[enne (OOO), 2014-01-14 22:54:54]

This would maybe be better done at the layer level, since this is a per-layer
setting.  A better solution would be something more akin to having the live
tiles rect for the PictureLayerTiling be exactly the visible rect when some
option is set.

Unfortunately, doing just that would mean that you couldn't cache
out-of-viewport tiles drawn with Ganesh; so, maybe we need better separation of
which tiles get created and which get deleted without tying them so closely
together in order to implement such an approach.

[ernstm, 2014-01-16 23:55:00]

The current implementation enables GPU rasterization on all layers, if the
command line flag is set. It'll probably be a while until this changes, as our
current focus is getting good performance out of very simple test cases. This
patch basically allows us to get one known performance issue out of the
equation.

I completely agree that we will need a more sophisticated solution once we
actually use hybrid rasterization, but I think we should postpone the
implementation of that until we have proven that GPU rasterization outperforms
CPU rasterization for the simple cases.

Would you be okay with landing this patch in that context?

[enne (OOO), 2014-01-17 00:03:56]

Okay, lgtm if you create a bug to improve it and put a TODO with that bug number
in a comment in the code.

[commit-bot: I haz the power, 2014-01-17 01:11:17]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ernstm@chromium.org/137993013/180001

[commit-bot: I haz the power, 2014-01-17 02:14:56]

Retried try job too often on win_rel for step(s) net_unittests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[commit-bot: I haz the power, 2014-01-17 05:55:37]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ernstm@chromium.org/137993013/180001

[commit-bot: I haz the power, 2014-01-17 10:11:13]

Change committed as 245479