fix race: this kCacheSize is probably not meant to be static

fix race: this kCacheSize is probably not meant to be static

TSAN sees us racing to intialize the static const kCacheSize.

This static const value depends on the runtime value image->getSize(), so there
is a race to set it.  I think this is unintentionally copy-paste from the other
tests that use a constant kCacheSize.  It's weird to intialize a constant based
on the first call to test_internal_purge(), though luckily we're always calling
it with same sized images today.

See these TSAN failures:
    https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-TSAN/builds/4937
    https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-TSAN/builds/4940/steps/dm/logs/stdio

TBR=robertphillips@google.com


BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1720743002

Committed: https://skia.googlesource.com/skia/+/03762fea75b0fe34ac0375407b83b6ae52e453b4

[mtklein_C, 2016-02-21 21:14:11]

Description was changed from

==========
fix race: this kCacheSize is probably not meant to be static

TSAN sees us racing to intialize the static const kCacheSize.

This static const value depends on the runtime value image->getSize(), so there
is a race to set it.  I think this is unintentionally copy-paste from the other
tests that use a constant kCacheSize.  It's weird to intialize a constant based
on the first call to test_internal_purge(), though luckily we're always calling
it with same sized images today.

See these TSAN failures:
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...


BUG=skia:
==========

to

==========
fix race: this kCacheSize is probably not meant to be static

TSAN sees us racing to intialize the static const kCacheSize.

This static const value depends on the runtime value image->getSize(), so there
is a race to set it.  I think this is unintentionally copy-paste from the other
tests that use a constant kCacheSize.  It's weird to intialize a constant based
on the first call to test_internal_purge(), though luckily we're always calling
it with same sized images today.

See these TSAN failures:
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...


BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
==========

[mtklein, 2016-02-21 21:14:44]

Description was changed from

==========
fix race: this kCacheSize is probably not meant to be static

TSAN sees us racing to intialize the static const kCacheSize.

This static const value depends on the runtime value image->getSize(), so there
is a race to set it.  I think this is unintentionally copy-paste from the other
tests that use a constant kCacheSize.  It's weird to intialize a constant based
on the first call to test_internal_purge(), though luckily we're always calling
it with same sized images today.

See these TSAN failures:
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...


BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
==========

to

==========
fix race: this kCacheSize is probably not meant to be static

TSAN sees us racing to intialize the static const kCacheSize.

This static const value depends on the runtime value image->getSize(), so there
is a race to set it.  I think this is unintentionally copy-paste from the other
tests that use a constant kCacheSize.  It's weird to intialize a constant based
on the first call to test_internal_purge(), though luckily we're always calling
it with same sized images today.

See these TSAN failures:
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...

TBR=robertphillips@google.com


BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
==========

[mtklein, 2016-02-21 21:14:57]

The CQ bit was checked by mtklein@google.com

[commit-bot: I haz the power, 2016-02-21 21:15:08]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1720743002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1720743002/1

[mtklein_C, 2016-02-21 21:21:25]

mtklein@chromium.org changed reviewers:
+ robertphillips@google.com
- mtklein@google.com

[commit-bot: I haz the power, 2016-02-21 21:36:53]

Description was changed from

==========
fix race: this kCacheSize is probably not meant to be static

TSAN sees us racing to intialize the static const kCacheSize.

This static const value depends on the runtime value image->getSize(), so there
is a race to set it.  I think this is unintentionally copy-paste from the other
tests that use a constant kCacheSize.  It's weird to intialize a constant based
on the first call to test_internal_purge(), though luckily we're always calling
it with same sized images today.

See these TSAN failures:
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...

TBR=robertphillips@google.com


BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
==========

to

==========
fix race: this kCacheSize is probably not meant to be static

TSAN sees us racing to intialize the static const kCacheSize.

This static const value depends on the runtime value image->getSize(), so there
is a race to set it.  I think this is unintentionally copy-paste from the other
tests that use a constant kCacheSize.  It's weird to intialize a constant based
on the first call to test_internal_purge(), though luckily we're always calling
it with same sized images today.

See these TSAN failures:
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...
   
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-G...

TBR=robertphillips@google.com


BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...

Committed:
https://skia.googlesource.com/skia/+/03762fea75b0fe34ac0375407b83b6ae52e453b4
==========

[commit-bot: I haz the power, 2016-02-21 21:36:54]

Committed patchset #1 (id:1) as
https://skia.googlesource.com/skia/+/03762fea75b0fe34ac0375407b83b6ae52e453b4

[robertphillips, 2016-02-22 12:39:47]

lgtm - it was not intended to be static.