Resize output_rect of aggregated CompositorFrame to enable it to be drawn when resizing.

Resize output_rect of aggregated CompositorFrame to enable it to be drawn when resizing.

It's possible for the browser compositor to get at least one frame behind the resizes that come from Windows so the Surface that it creates is never the right size to be swapped to screen. In this case, resize the aggregated frame so that the it fills the backbuffer. This allows it to be drawn, even if there may be a resize gutter or clipping around it.

BUG=
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Committed: https://crrev.com/e95e17121740b79c447cb93d0e1d07bbfc407a14
Cr-Commit-Position: refs/heads/master@{#364897}

[jbauman, 2015-11-04 01:34:02]

Description was changed from

==========
Resize output_rect of aggregated CompositorFrame to enable it to be drawn when
resizing.

It's possible for the browser compositor to get at least one frame behind the
resizes that come from Windows, so the Surface that it creates are never the
right size to be swapped to screen. In this case, resize the aggregated frame so
that the it fills the backbuffer. This allows it to be drawn, even if there may
be a resize gutter or clipping around it.

BUG=
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
Resize output_rect of aggregated CompositorFrame to enable it to be drawn when
resizing.

It's possible for the browser compositor to get at least one frame behind the
resizes that come from Windows so the Surface that it creates is never the right
size to be swapped to screen. In this case, resize the aggregated frame so that
the it fills the backbuffer. This allows it to be drawn, even if there may be a
resize gutter or clipping around it.

BUG=
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[jbauman, 2015-11-04 01:34:29]

jbauman@chromium.org changed reviewers:
+ brianderson@chromium.org

[brianderson, 2015-11-04 01:40:25]

Nice way to avoid the stretching. What does the gutter look like with this
patch?

https://codereview.chromium.org/1426453006/diff/1/cc/surfaces/display.cc
File cc/surfaces/display.cc (right):

https://codereview.chromium.org/1426453006/diff/1/cc/surfaces/display.cc#newc...
cc/surfaces/display.cc:236: TRACE_EVENT_INSTANT0("cc", "Size mismatch.",
TRACE_EVENT_SCOPE_THREAD);
In what cases would we hit this trace after this patch?

[jbauman, 2015-11-04 01:53:45]

On 2015/11/04 01:40:25, brianderson wrote:
> Nice way to avoid the stretching. What does the gutter look like with this
> patch?

On Windows it's black. I don't know what it looks like on other systems.
> 
> https://codereview.chromium.org/1426453006/diff/1/cc/surfaces/display.cc
> File cc/surfaces/display.cc (right):
> 
>
https://codereview.chromium.org/1426453006/diff/1/cc/surfaces/display.cc#newc...
> cc/surfaces/display.cc:236: TRACE_EVENT_INSTANT0("cc", "Size mismatch.",
> TRACE_EVENT_SCOPE_THREAD);
> In what cases would we hit this trace after this patch?

This happens if current_surface_size_ is 0x0 (temporarily happens while the
win32 window is being resized) or the damage rect is not the full size of the
root render pass (because presentation fails if the first swap from a resizing
surface does not contain the entire surface, so we can't resize in this case).

[jbauman, 2015-11-10 01:44:15]

ping?

On 2015/11/04 01:53:45, jbauman wrote:
> On 2015/11/04 01:40:25, brianderson wrote:
> > Nice way to avoid the stretching. What does the gutter look like with this
> > patch?
> 
> On Windows it's black. I don't know what it looks like on other systems.
> > 
> > https://codereview.chromium.org/1426453006/diff/1/cc/surfaces/display.cc
> > File cc/surfaces/display.cc (right):
> > 
> >
>
https://codereview.chromium.org/1426453006/diff/1/cc/surfaces/display.cc#newc...
> > cc/surfaces/display.cc:236: TRACE_EVENT_INSTANT0("cc", "Size mismatch.",
> > TRACE_EVENT_SCOPE_THREAD);
> > In what cases would we hit this trace after this patch?
> 
> This happens if current_surface_size_ is 0x0 (temporarily happens while the
> win32 window is being resized) or the damage rect is not the full size of the
> root render pass (because presentation fails if the first swap from a resizing
> surface does not contain the entire surface, so we can't resize in this case).

[brianderson, 2015-11-13 03:52:39]

lgtm

[jbauman, 2015-12-12 01:42:57]

The CQ bit was checked by jbauman@chromium.org

[commit-bot: I haz the power, 2015-12-12 01:46:11]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1426453006/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1426453006/1

[commit-bot: I haz the power, 2015-12-12 04:14:00]

Description was changed from

==========
Resize output_rect of aggregated CompositorFrame to enable it to be drawn when
resizing.

It's possible for the browser compositor to get at least one frame behind the
resizes that come from Windows so the Surface that it creates is never the right
size to be swapped to screen. In this case, resize the aggregated frame so that
the it fills the backbuffer. This allows it to be drawn, even if there may be a
resize gutter or clipping around it.

BUG=
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
Resize output_rect of aggregated CompositorFrame to enable it to be drawn when
resizing.

It's possible for the browser compositor to get at least one frame behind the
resizes that come from Windows so the Surface that it creates is never the right
size to be swapped to screen. In this case, resize the aggregated frame so that
the it fills the backbuffer. This allows it to be drawn, even if there may be a
resize gutter or clipping around it.

BUG=
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

[commit-bot: I haz the power, 2015-12-12 04:14:01]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-12-12 04:15:19]

Description was changed from

==========
Resize output_rect of aggregated CompositorFrame to enable it to be drawn when
resizing.

It's possible for the browser compositor to get at least one frame behind the
resizes that come from Windows so the Surface that it creates is never the right
size to be swapped to screen. In this case, resize the aggregated frame so that
the it fills the backbuffer. This allows it to be drawn, even if there may be a
resize gutter or clipping around it.

BUG=
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
==========

to

==========
Resize output_rect of aggregated CompositorFrame to enable it to be drawn when
resizing.

It's possible for the browser compositor to get at least one frame behind the
resizes that come from Windows so the Surface that it creates is never the right
size to be swapped to screen. In this case, resize the aggregated frame so that
the it fills the backbuffer. This allows it to be drawn, even if there may be a
resize gutter or clipping around it.

BUG=
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Committed: https://crrev.com/e95e17121740b79c447cb93d0e1d07bbfc407a14
Cr-Commit-Position: refs/heads/master@{#364897}
==========

[commit-bot: I haz the power, 2015-12-12 04:15:20]

Patchset 1 (id:??) landed as
https://crrev.com/e95e17121740b79c447cb93d0e1d07bbfc407a14
Cr-Commit-Position: refs/heads/master@{#364897}