quick hackup of SkNx SkMatrixConvolutionImageFilter

quick hackup of SkNx SkMatrixConvolutionImageFilter

bench times drop on my laptop from ~12ms to ~6ms.

The !convolveAlpha path seems very slightly different in precision.
Not sure why yet.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1886553002

[mtklein_C, 2016-04-12 20:37:14]

Description was changed from

==========
quick hackup of SkNx SkMatrixConvolutionImageFilter

this only vectorizes the innermost loop

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

to

==========
quick hackup of SkNx SkMatrixConvolutionImageFilter

this only vectorizes the innermost loop

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

[mtklein, 2016-04-12 21:10:08]

Description was changed from

==========
quick hackup of SkNx SkMatrixConvolutionImageFilter

this only vectorizes the innermost loop

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

to

==========
quick hackup of SkNx SkMatrixConvolutionImageFilter

bench times drop on my laptop from ~12ms to ~6ms.

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

[mtklein, 2016-04-12 21:12:10]

Description was changed from

==========
quick hackup of SkNx SkMatrixConvolutionImageFilter

bench times drop on my laptop from ~12ms to ~6ms.

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

to

==========
quick hackup of SkNx SkMatrixConvolutionImageFilter

bench times drop on my laptop from ~12ms to ~6ms.

The !convolveAlpha path seems very slightly different in precision.
Not sure why yet.

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

[Stephen White, 2016-04-13 12:08:49]

LGTM as long as you mark the appropriate Blink tests for rebaseline (if any)

[reed1, 2016-04-13 13:28:24]

reed@google.com changed reviewers:
+ reed@google.com

[reed1, 2016-04-13 13:28:24]

great speed gain, lgtm

[mtklein, 2016-05-05 17:52:45]

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

[commit-bot: I haz the power, 2016-05-05 17:52:55]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1886553002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1886553002/40001

[commit-bot: I haz the power, 2016-05-05 18:08:44]

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

[commit-bot: I haz the power, 2016-05-05 18:08:45]

Dry run: This issue passed the CQ dry run.