Lay groundwork for SkOpts.

Lay groundwork for SkOpts.

This doesn't really do anything yet.  It's just the CPU detection code, skeleton new .cpp files, and a few little .gyp tweaks.

BUG=skia:4117
Committed: https://skia.googlesource.com/skia/+/ce2c5055cee5d5d3c9fc84c1b3eeed4b4d84a827

Committed: https://skia.googlesource.com/skia/+/8317a1832f55e175531ee7ae7ccd12a3a15e3c75

[mtklein_C, 2015-07-27 16:12:16]

The CQ bit was checked by mtklein@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2015-07-27 16:12:27]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1255193002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1255193002/1

[mtklein_C, 2015-07-27 16:12:28]

mtklein@chromium.org changed reviewers:
+ djsollen@google.com

[mtklein_C, 2015-07-27 17:02:43]

The CQ bit was checked by mtklein@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2015-07-27 17:02:54]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1255193002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1255193002/20001

[commit-bot: I haz the power, 2015-07-27 17:10:33]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-07-27 17:10:34]

Dry run: This issue passed the CQ dry run.

[djsollen, 2015-07-27 17:38:02]

lgtm with nit

https://codereview.chromium.org/1255193002/diff/20001/src/core/SkOpts.cpp
File src/core/SkOpts.cpp (right):

https://codereview.chromium.org/1255193002/diff/20001/src/core/SkOpts.cpp#new...
src/core/SkOpts.cpp:44: // We just assume NEON for non-Android ARM platforms.
instead of assuming neon for 32-bit arm devices why not just check for
SK_ARM_HAS_NEON or do you plan on trying to delete that as well?

[mtklein_C, 2015-07-27 17:44:10]

On 2015/07/27 17:38:02, djsollen wrote:
> lgtm with nit
> 
> https://codereview.chromium.org/1255193002/diff/20001/src/core/SkOpts.cpp
> File src/core/SkOpts.cpp (right):
> 
>
https://codereview.chromium.org/1255193002/diff/20001/src/core/SkOpts.cpp#new...
> src/core/SkOpts.cpp:44: // We just assume NEON for non-Android ARM platforms.
> instead of assuming neon for 32-bit arm devices why not just check for
> SK_ARM_HAS_NEON or do you plan on trying to delete that as well?

Good idea.  Done.

[mtklein_C, 2015-07-27 17:44:37]

The CQ bit was checked by mtklein@chromium.org

[mtklein_C, 2015-07-27 17:44:38]

The patchset sent to the CQ was uploaded after l-g-t-m from djsollen@google.com
Link to the patchset: https://codereview.chromium.org/1255193002/#ps40001
(title: "good idea")

[commit-bot: I haz the power, 2015-07-27 17:44:45]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1255193002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1255193002/40001

[commit-bot: I haz the power, 2015-07-27 17:52:35]

Committed patchset #3 (id:40001) as
https://skia.googlesource.com/skia/+/ce2c5055cee5d5d3c9fc84c1b3eeed4b4d84a827

[mtklein_C, 2015-07-27 19:03:06]

A revert of this CL (patchset #3 id:40001) has been created in
https://codereview.chromium.org/1261743002/ by mtklein@chromium.org.

The reason for reverting is: Chromium doesn't call SkGraphics::Init().  This
setup won't work..

[jbudorick, 2015-07-27 20:33:46]

jbudorick@chromium.org changed reviewers:
+ jbudorick@chromium.org

[jbudorick, 2015-07-27 20:33:47]

https://codereview.chromium.org/1255193002/diff/40001/src/core/SkOpts.cpp
File src/core/SkOpts.cpp (right):

https://codereview.chromium.org/1255193002/diff/40001/src/core/SkOpts.cpp#new...
src/core/SkOpts.cpp:19: #include <cpu-features.h>
From
http://build.chromium.org/p/chromium.linux/builders/Android%20Arm64%20Builder...:

../../third_party/skia/src/core/SkOpts.cpp:19:30: fatal error: cpu-features.h:
No such file or directory
     #include <cpu-features.h>

[mtklein, 2015-07-27 20:44:25]

On 2015/07/27 20:33:47, jbudorick wrote:
> https://codereview.chromium.org/1255193002/diff/40001/src/core/SkOpts.cpp
> File src/core/SkOpts.cpp (right):
> 
>
https://codereview.chromium.org/1255193002/diff/40001/src/core/SkOpts.cpp#new...
> src/core/SkOpts.cpp:19: #include <cpu-features.h>
> From
>
http://build.chromium.org/p/chromium.linux/builders/Android%20Arm64%20Builder...:
> 
> ../../third_party/skia/src/core/SkOpts.cpp:19:30: fatal error: cpu-features.h:
> No such file or directory
>      #include <cpu-features.h>

Funky.  That works on our ARMv8 bots.  Wonder why that's not found.  Thanks for
reverting the roll.

[mtklein_C, 2015-07-28 13:33:33]

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

[mtklein_C, 2015-07-28 13:33:33]

Would you mind (re-) reviewing this CL?  I think we can make the original
pleasant design work with some tweaks.  PS4 has these changes over yesterday:

   - ARMv8 build problem ought to be fixed.
   - SkOpts::Init() is called automatically if we allow global constructors,
and/or by SkGraphics::Init(), all guarded by SkOnce to be thread-safe and
idempotent.
   - Removed some old cruft from SkGraphics::Init() to make it clearer.

My plan is to have Chrome's renderer threads call SkGraphics::Init() at
startup*, and perhaps the browser process too if that's necessary.  Android's
calling SkGraphics::Init() and gets the automatic call to SkOpts::Init(), and
anyone else who's not calling SkGraphics::Init() will still get the automatic
SkOpts::Init() call if they use any code that depends transitively on the SkOpts
function pointers.

* Right around here:
https://code.google.com/p/chromium/codesearch#chromium/src/content/renderer/r...

[mtklein_C, 2015-07-28 13:33:43]

The CQ bit was checked by mtklein@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2015-07-28 13:33:48]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1255193002/60001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1255193002/60001

[commit-bot: I haz the power, 2015-07-28 13:42:20]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-07-28 13:42:21]

Dry run: This issue passed the CQ dry run.

[mtklein, 2015-07-28 13:43:49]

The CQ bit was checked by mtklein@google.com to run a CQ dry run

[commit-bot: I haz the power, 2015-07-28 13:43:58]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1255193002/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1255193002/80001

[commit-bot: I haz the power, 2015-07-28 13:51:09]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-07-28 13:51:10]

Dry run: This issue passed the CQ dry run.

[mtklein_C, 2015-07-30 14:20:44]

The CQ bit was checked by mtklein@chromium.org

[mtklein_C, 2015-07-30 14:20:44]

The patchset sent to the CQ was uploaded after l-g-t-m from djsollen@google.com
Link to the patchset: https://codereview.chromium.org/1255193002/#ps100001
(title: "rebase")

[commit-bot: I haz the power, 2015-07-30 14:20:48]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1255193002/100001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1255193002/100001

[commit-bot: I haz the power, 2015-07-30 14:30:19]

Committed patchset #6 (id:100001) as
https://skia.googlesource.com/skia/+/8317a1832f55e175531ee7ae7ccd12a3a15e3c75