Remove usage of kReturnToCache_FlagBit in C2D layer bridge.

Remove usage of kReturnToCache_FlagBit in C2D layer bridge.

BUG=skia:2889
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=184191

[bsalomon, 2014-10-16 19:25:09]

bsalomon@google.com changed reviewers:
+ junov@chromium.org

[bsalomon, 2014-10-16 19:25:10]

This flag no longer does anything. The return to cache behavior is inferred.
After this CL I can delete the definition in Skia.

[bsalomon, 2014-10-16 20:23:57]

bsalomon@google.com changed reviewers:
+ senorblanco@chromium.org

[Stephen White, 2014-10-16 20:28:53]

LGTM, but will leave for Justin.

[Justin Novosad, 2014-10-17 14:45:35]

On 2014/10/16 20:28:53, Stephen White wrote:
> LGTM, but will leave for Justin.

FWIW, logic behind this is that if we know a canvas is going out of use
permanently, the likelihood of requiring another texture of the exact same size
is low, so no need to return it to cache.  This strategy helped improve scratch
pool performance in cases where the window is being resized.

So what happens now?

[bsalomon, 2014-10-22 14:30:13]

On 2014/10/17 14:45:35, junov wrote:
> On 2014/10/16 20:28:53, Stephen White wrote:
> > LGTM, but will leave for Justin.
> 
> FWIW, logic behind this is that if we know a canvas is going out of use
> permanently, the likelihood of requiring another texture of the exact same
size
> is low, so no need to return it to cache.  This strategy helped improve
scratch
> pool performance in cases where the window is being resized.
> 
> So what happens now?

There is probably a regression in that case. I think we will have to have
another way of handling the use case where there are lots of exact match scratch
textures sizes that aren't reused or add another mechanism for the client to
signal this.

I'm going to go ahead and land this since the flag isn't used anywhere in Skia
code anymore (setting it has no effect).

[bsalomon, 2014-10-22 14:30:16]

The CQ bit was checked by bsalomon@google.com

[commit-bot: I haz the power, 2014-10-22 14:30:44]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/660083002/1

[commit-bot: I haz the power, 2014-10-22 15:36:09]

Committed patchset #1 (id:1) as 184191