cc: Remove all impl-side caps from RendererCapabilitiesImpl.

cc: Remove all impl-side caps from RendererCapabilitiesImpl.

Next up is destroy the struct itself, but need to see which main thread
ones can be removed, and generate them outside of the Renderer (and so
probably rename it all).

R=enne
BUG=606056
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel

Committed: https://crrev.com/1a86f6d6a4b658c88cb293430326614c8165ec00
Cr-Commit-Position: refs/heads/master@{#412626}

[danakj, 2016-08-17 02:04:43]

Description was changed from

==========
cc: Remove all impl-side caps from RendererCapabilitiesImpl.

Next up is destroy the struct it self, but need to see which main thread
ones can be removed, and generate them outside of the Renderer (and so
probably rename it all).

R=enne
BUG=606056
==========

to

==========
cc: Remove all impl-side caps from RendererCapabilitiesImpl.

Next up is destroy the struct it self, but need to see which main thread
ones can be removed, and generate them outside of the Renderer (and so
probably rename it all).

R=enne
BUG=606056
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel
==========

[danakj, 2016-08-17 02:06:43]

https://codereview.chromium.org/2253823002/diff/1/cc/output/direct_renderer.cc
File cc/output/direct_renderer.cc (right):

https://codereview.chromium.org/2253823002/diff/1/cc/output/direct_renderer.c...
cc/output/direct_renderer.cc:84: use_partial_swap_ =
settings_->partial_swap_enabled && CanPartialSwap();
This is a behaviour change. What it changed is that if
LayerTreeSettings::partial_swap_enabled = false, then we won't use partial swap
for software compositing (as demonstrated by a unit test change also to set the
setting to true). ui::Compositor is the only place that we have a Display that
does software compositing, and it sets it to true so that should be fine, and
making --ui-disable-partial-swap work for software too seems pretty reasonable?

[enne (OOO), 2016-08-17 17:48:43]

https://codereview.chromium.org/2253823002/diff/1/cc/output/direct_renderer.cc
File cc/output/direct_renderer.cc (right):

https://codereview.chromium.org/2253823002/diff/1/cc/output/direct_renderer.c...
cc/output/direct_renderer.cc:84: use_partial_swap_ =
settings_->partial_swap_enabled && CanPartialSwap();
On 2016/08/17 at 02:06:43, danakj wrote:
> This is a behaviour change. What it changed is that if
LayerTreeSettings::partial_swap_enabled = false, then we won't use partial swap
for software compositing (as demonstrated by a unit test change also to set the
setting to true). ui::Compositor is the only place that we have a Display that
does software compositing, and it sets it to true so that should be fine, and
making --ui-disable-partial-swap work for software too seems pretty reasonable?

Yeah, this sounds good to me.  We've had a bunch of partial swap issues over
time and being able to have a flag that makes it easy to know when that's the
case even in software seems great.

https://codereview.chromium.org/2253823002/diff/1/cc/output/direct_renderer.h
File cc/output/direct_renderer.h (right):

https://codereview.chromium.org/2253823002/diff/1/cc/output/direct_renderer.h...
cc/output/direct_renderer.h:43: struct Capabilities {
Where does this get used?

https://codereview.chromium.org/2253823002/diff/1/cc/output/gl_renderer_unitt...
File cc/output/gl_renderer_unittest.cc (right):

https://codereview.chromium.org/2253823002/diff/1/cc/output/gl_renderer_unitt...
cc/output/gl_renderer_unittest.cc:1076: renderer.Initialize();
Now that there's a required initialize, can there be dchecks at least in like
draw or swap that initialize has occurred?

https://codereview.chromium.org/2253823002/diff/1/cc/trees/layer_tree_host_im...
File cc/trees/layer_tree_host_impl.cc (right):

https://codereview.chromium.org/2253823002/diff/1/cc/trees/layer_tree_host_im...
cc/trees/layer_tree_host_impl.cc:1781: // TODO(danakj): Can we avoid having this
run when there's no output surface?
Maybe it should early out when there's no output surface in the meantime?

https://codereview.chromium.org/2253823002/diff/1/cc/trees/layer_tree_host_im...
File cc/trees/layer_tree_host_impl.h (right):

https://codereview.chromium.org/2253823002/diff/1/cc/trees/layer_tree_host_im...
cc/trees/layer_tree_host_impl.h:408: need_update_gpu_rasterization_status_ =
true;
Should this early out if it hasn't changed?

[danakj, 2016-08-17 18:15:00]

The CQ bit was checked by danakj@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-17 18:15:32]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[danakj, 2016-08-17 18:16:41]

PTAL

https://codereview.chromium.org/2253823002/diff/1/cc/output/direct_renderer.h
File cc/output/direct_renderer.h (right):

https://codereview.chromium.org/2253823002/diff/1/cc/output/direct_renderer.h...
cc/output/direct_renderer.h:43: struct Capabilities {
On 2016/08/17 17:48:42, enne wrote:
> Where does this get used?

I think I was going to do this but then I didn't so ya it doesn't.

https://codereview.chromium.org/2253823002/diff/1/cc/output/gl_renderer_unitt...
File cc/output/gl_renderer_unittest.cc (right):

https://codereview.chromium.org/2253823002/diff/1/cc/output/gl_renderer_unitt...
cc/output/gl_renderer_unittest.cc:1076: renderer.Initialize();
On 2016/08/17 17:48:42, enne wrote:
> Now that there's a required initialize, can there be dchecks at least in like
> draw or swap that initialize has occurred?

Ya I did that to find all the places that need it, I can leave it in. Since you
can't DrawAndSwap without SetVisible(true), I think dchecking init in SetVisible
should work out.

https://codereview.chromium.org/2253823002/diff/1/cc/trees/layer_tree_host_im...
File cc/trees/layer_tree_host_impl.cc (right):

https://codereview.chromium.org/2253823002/diff/1/cc/trees/layer_tree_host_im...
cc/trees/layer_tree_host_impl.cc:1781: // TODO(danakj): Can we avoid having this
run when there's no output surface?
On 2016/08/17 17:48:42, enne wrote:
> Maybe it should early out when there's no output surface in the meantime?

Good idea.

https://codereview.chromium.org/2253823002/diff/1/cc/trees/layer_tree_host_im...
File cc/trees/layer_tree_host_impl.h (right):

https://codereview.chromium.org/2253823002/diff/1/cc/trees/layer_tree_host_im...
cc/trees/layer_tree_host_impl.h:408: need_update_gpu_rasterization_status_ =
true;
On 2016/08/17 17:48:43, enne wrote:
> Should this early out if it hasn't changed?

Done.

[enne (OOO), 2016-08-17 18:30:33]

Description was changed from

==========
cc: Remove all impl-side caps from RendererCapabilitiesImpl.

Next up is destroy the struct it self, but need to see which main thread
ones can be removed, and generate them outside of the Renderer (and so
probably rename it all).

R=enne
BUG=606056
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel
==========

to

==========
cc: Remove all impl-side caps from RendererCapabilitiesImpl.

Next up is destroy the struct itself, but need to see which main thread
ones can be removed, and generate them outside of the Renderer (and so
probably rename it all).

R=enne
BUG=606056
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel
==========

[enne (OOO), 2016-08-17 18:31:59]

lgtm

[commit-bot: I haz the power, 2016-08-17 19:10:24]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-17 19:10:25]

Dry run: This issue passed the CQ dry run.

[danakj, 2016-08-17 20:04:04]

The CQ bit was checked by danakj@chromium.org

[commit-bot: I haz the power, 2016-08-17 20:04:40]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-17 20:11:20]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-08-17 20:18:18]

Description was changed from

==========
cc: Remove all impl-side caps from RendererCapabilitiesImpl.

Next up is destroy the struct itself, but need to see which main thread
ones can be removed, and generate them outside of the Renderer (and so
probably rename it all).

R=enne
BUG=606056
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel
==========

to

==========
cc: Remove all impl-side caps from RendererCapabilitiesImpl.

Next up is destroy the struct itself, but need to see which main thread
ones can be removed, and generate them outside of the Renderer (and so
probably rename it all).

R=enne
BUG=606056
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel

Committed: https://crrev.com/1a86f6d6a4b658c88cb293430326614c8165ec00
Cr-Commit-Position: refs/heads/master@{#412626}
==========

[commit-bot: I haz the power, 2016-08-17 20:18:20]

Patchset 2 (id:??) landed as
https://crrev.com/1a86f6d6a4b658c88cb293430326614c8165ec00
Cr-Commit-Position: refs/heads/master@{#412626}