Use scratch keys for stencil buffers.

Use scratch keys for stencil buffers.

BUG=skia:2889
Committed: https://skia.googlesource.com/skia/+/91175f19664a62851da4ca4e0984a7c7c45b258f

Committed: https://skia.googlesource.com/skia/+/10e23caea3106be125acea10a637789e5a15c728

[bsalomon, 2014-11-24 14:30:31]

bsalomon@google.com changed reviewers:
+ robertphillips@google.com

[robertphillips, 2014-11-24 14:39:14]

lgtm + some comment suggestions & an oddity

https://codereview.chromium.org/747043004/diff/20001/bench/GrResourceCacheBen...
File bench/GrResourceCacheBench.cpp (right):

https://codereview.chromium.org/747043004/diff/20001/bench/GrResourceCacheBen...
bench/GrResourceCacheBench.cpp:23: 
// Test resource that emulates a stencil buffer's behavior ?

https://codereview.chromium.org/747043004/diff/20001/bench/GrResourceCacheBen...
bench/GrResourceCacheBench.cpp:53: 
// Test resource that emulates a texture's behavior ?

https://codereview.chromium.org/747043004/diff/20001/bench/GrResourceCacheBen...
bench/GrResourceCacheBench.cpp:86: *h = i * 2 % 1024;
% 1 again ?

[bsalomon, 2014-11-24 14:49:59]

I'll cleanup this bench separately by splitting it into a bench and unit test.

https://codereview.chromium.org/747043004/diff/20001/bench/GrResourceCacheBen...
File bench/GrResourceCacheBench.cpp (right):

https://codereview.chromium.org/747043004/diff/20001/bench/GrResourceCacheBen...
bench/GrResourceCacheBench.cpp:86: *h = i * 2 % 1024;
On 2014/11/24 14:39:14, robertphillips wrote:
> % 1 again ?

yikes, changed to 2.

[bsalomon, 2014-11-24 14:50:02]

The CQ bit was checked by bsalomon@google.com

[commit-bot: I haz the power, 2014-11-24 14:50:22]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/747043004/40001

[commit-bot: I haz the power, 2014-11-24 15:05:20]

Committed patchset #3 (id:40001) as
https://skia.googlesource.com/skia/+/91175f19664a62851da4ca4e0984a7c7c45b258f

[Tom Hudson, 2014-11-24 19:22:30]

A revert of this CL (patchset #3 id:40001) has been created in
https://codereview.chromium.org/752213002/ by tomhudson@chromium.org.

The reason for reverting is: Bisecting points to this as the culprit in
persistent DM crashes, e.g.:

> 2033 tasks left	 344M peak	  843ms	tabl-cuteoverload_skp
> Signal 11:
> _sigtramp (+0x1a)
> create_surface(CanvasConfig const&, GrContext*) (+0x1b2)
> test_WritePixels(skiatest::Reporter*, GrContextFactory*) (+0x8d4)
> skiatest::WritePixelsClass::onRun(skiatest::Reporter*) (+0x21)
> skiatest::Test::run() (+0x7c)
> DM::GpuTestTask::draw(GrContextFactory*) (+0x8c)
> DM::GpuTask::run(GrContextFactory*) (+0xa6)
> DM::TaskRunner::wait() (+0x7f)
> dm_main() (+0x33f)
> main (+0x27)
 
 
On all MacMini platforms (the following URL with -Mac{10.6,10.7,10.8}- and
-{Debug,Release}):

http://build.chromium.org/p/client.skia/builders/Test-Mac10.6-MacMini4.1-GeFo....

[bsalomon, 2014-11-24 22:01:08]

Rob, can you take another look?

This fixes a bug where a SB is created but can't be used because of format
incompatibility. We wind up leaving it in the cache and finding it later. The
fix is to clear out its scratch key.

This also contains a revert-revert of the changes related to the original. Those
were rebaselines and the grresourcecache bench refactor.

The original patch that landed was patchset 3.

[robertphillips, 2014-11-25 13:34:47]

lgtm + a typo

https://codereview.chromium.org/747043004/diff/70001/src/gpu/gl/GrGpuGL.cpp
File src/gpu/gl/GrGpuGL.cpp (right):

https://codereview.chromium.org/747043004/diff/70001/src/gpu/gl/GrGpuGL.cpp#n...
src/gpu/gl/GrGpuGL.cpp:1181: sb->cacheAccess().removeScratchKey();
it off _to_ the ?

[bsalomon, 2014-11-25 13:41:37]

The CQ bit was checked by bsalomon@google.com

[commit-bot: I haz the power, 2014-11-25 13:41:56]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/747043004/90001

[commit-bot: I haz the power, 2014-11-25 13:52:10]

Committed patchset #6 (id:90001) as
https://skia.googlesource.com/skia/+/10e23caea3106be125acea10a637789e5a15c728