Remove static effects from the effect memory pool.

Remove static effects from the effect memory pool.
Committed: https://code.google.com/p/skia/source/detail?r=8828

[bsalomon, 2013-04-22 20:54:45]

:(

[robertphillips, 2013-04-22 22:05:13]

lgtm + 2 nits

What symptom were we seeing with the old method?

https://codereview.chromium.org/14081016/diff/5001/include/core/SkTemplates.h
File include/core/SkTemplates.h (right):

https://codereview.chromium.org/14081016/diff/5001/include/core/SkTemplates.h...
include/core/SkTemplates.h:126: ~SkAutoTDestroy() {
NULL !=

https://codereview.chromium.org/14081016/diff/5001/include/gpu/GrEffect.h
File include/gpu/GrEffect.h (right):

https://codereview.chromium.org/14081016/diff/5001/include/gpu/GrEffect.h#new...
include/gpu/GrEffect.h:30: *
pools -> pool

[bsalomon, 2013-04-23 13:19:29]

On 2013/04/22 22:05:13, robertphillips wrote:
> lgtm + 2 nits
> 
> What symptom were we seeing with the old method?

I hit asserts in gm on android. We're tearing down the pool but there are still
active effects in the pool (yikes). With this change I don't get the asserts.

[jvanverth1, 2013-04-23 15:10:39]

On 2013/04/23 13:19:29, bsalomon wrote:
> On 2013/04/22 22:05:13, robertphillips wrote:
> > lgtm + 2 nits
> > 
> > What symptom were we seeing with the old method?
> 
> I hit asserts in gm on android. We're tearing down the pool but there are
still
> active effects in the pool (yikes). With this change I don't get the asserts.

LGTM, though I wish there were another way. But I can't think of one myself.

[bsalomon, 2013-04-23 15:38:03]

Committed patchset #2 manually as r8828 (presubmit successful).