Have SkRasterPipelineBlitter take over for F16 when it can.

Have SkRasterPipelineBlitter take over for F16 when it can.

3 minor diffs to a couple GMs from fixed transfermodes (arithmetic and plus).

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2181033002

Committed: https://skia.googlesource.com/skia/+/4e90e3eb5e43ffa8f347570578f5a538793b0cc7

[mtklein_C, 2016-07-25 19:53:55]

Description was changed from

==========
Have SkRasterPipelineBlitter take over for F16 when it can.

3 minor diffs to a couple GMs from fixed transfermodes (arithmetic and plus).

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

to

==========
Have SkRasterPipelineBlitter take over for F16 when it can.

3 minor diffs to a couple GMs from fixed transfermodes (arithmetic and plus).

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2181033002
==========

[mtklein_C, 2016-07-25 20:09:04]

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

[mtklein_C, 2016-07-25 20:09:05]

The CQ bit was checked by mtklein@chromium.org

[commit-bot: I haz the power, 2016-07-25 20:09:16]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[mtklein_C, 2016-07-25 21:27:35]

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

[mtklein_C, 2016-07-25 21:27:35]

+Mike

[reed1, 2016-07-25 21:34:01]

seems like we might already have a nanobench for some of these cases. Have we
looked at the perf comparison?

code-change is nice. lgtm

[commit-bot: I haz the power, 2016-07-25 21:35:33]

Description was changed from

==========
Have SkRasterPipelineBlitter take over for F16 when it can.

3 minor diffs to a couple GMs from fixed transfermodes (arithmetic and plus).

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2181033002
==========

to

==========
Have SkRasterPipelineBlitter take over for F16 when it can.

3 minor diffs to a couple GMs from fixed transfermodes (arithmetic and plus).

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2181033002

Committed:
https://skia.googlesource.com/skia/+/4e90e3eb5e43ffa8f347570578f5a538793b0cc7
==========

[commit-bot: I haz the power, 2016-07-25 21:35:34]

Committed patchset #1 (id:1) as
https://skia.googlesource.com/skia/+/4e90e3eb5e43ffa8f347570578f5a538793b0cc7

[mtklein, 2016-07-25 21:42:51]

On 2016/07/25 21:34:01, reed1 wrote:
> seems like we might already have a nanobench for some of these cases. Have we
> looked at the perf comparison?
> 
> code-change is nice. lgtm

Yeah, it's on my radar to make sure I've got a good handle on how perf is
changing.
We do already for 565 (which of course got slower now that we're taking it up to
float and back).