Fix Plus

Fix Plus

This makes Plus with AA ~3% slower.  Seems like a good deal.

GMs affected: mixed_xfermodes, the new one.

Based on https://codereview.chromium.org/1150833003/

Still TODO: NEON.  The new GM should show this.

BUG=skia:3852
Committed: https://skia.googlesource.com/skia/+/160d08cb01a395e584a767b00f45e7fcf996b890

[mtklein_C, 2015-05-21 16:11:23]

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

[f(malita), 2015-05-21 17:17:28]

Peanut gallery LGTM :)

[reed1, 2015-05-21 21:11:21]

lgtm

[f(malita), 2015-05-21 21:25:52]

Chromium guard landed: https://codereview.chromium.org/1155663002/

[mtklein, 2015-05-21 21:31:11]

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

[mtklein, 2015-05-21 21:31:12]

Oh wow, thank you.  How efficient!

[mtklein, 2015-05-21 21:31:23]

The CQ bit was checked by mtklein@google.com

[mtklein, 2015-05-21 21:31:23]

The patchset sent to the CQ was uploaded after l-g-t-m from fmalita@chromium.org
Link to the patchset: https://codereview.chromium.org/1156453002/#ps20001
(title: "rebase")

[commit-bot: I haz the power, 2015-05-21 21:31:57]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1156453002/20001

[f(malita), 2015-05-21 21:35:13]

On 2015/05/21 21:31:12, mtklein wrote:
> Oh wow, thank you.  How efficient!

Let's just say kPlus_Mode is my new best friend :)

Re. your NEON comment: does that mean that the issue is not fixed for
Chrome/Android yet?

[mtklein, 2015-05-21 21:39:18]

On 2015/05/21 21:35:13, f(malita) wrote:
> On 2015/05/21 21:31:12, mtklein wrote:
> > Oh wow, thank you.  How efficient!
> 
> Let's just say kPlus_Mode is my new best friend :)
> 
> Re. your NEON comment: does that mean that the issue is not fixed for
> Chrome/Android yet?

Yeah.  I've been staring at it this afternoon trying to think of a good way to
fix it.  (Well, a good way is to turn on NEON in Chrome unconditionally, which
would simply activate this code.)  The NEON code is even less geared to be able
to treat AA specially than this code was.  If all goes well with these CLs, I'm
hoping to fix NEON tomorrow and then whine a bit more about how Chrome should
really turn NEON on globally if only to make us more productive.

[f(malita), 2015-05-21 21:41:50]

On 2015/05/21 21:39:18, mtklein wrote:
> On 2015/05/21 21:35:13, f(malita) wrote:
> > On 2015/05/21 21:31:12, mtklein wrote:
> > > Oh wow, thank you.  How efficient!
> > 
> > Let's just say kPlus_Mode is my new best friend :)
> > 
> > Re. your NEON comment: does that mean that the issue is not fixed for
> > Chrome/Android yet?
> 
> Yeah.  I've been staring at it this afternoon trying to think of a good way to
> fix it.  (Well, a good way is to turn on NEON in Chrome unconditionally, which
> would simply activate this code.)  The NEON code is even less geared to be
able
> to treat AA specially than this code was.  If all goes well with these CLs,
I'm
> hoping to fix NEON tomorrow and then whine a bit more about how Chrome should
> really turn NEON on globally if only to make us more productive.

Thanks!

[commit-bot: I haz the power, 2015-05-21 22:47:47]

Committed patchset #2 (id:20001) as
https://skia.googlesource.com/skia/+/160d08cb01a395e584a767b00f45e7fcf996b890