SSE 4.1 SrcOver blits: color32, blitmask.

SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

I've punted for now on pmaddubsw for lerping. I do intend to try that,
but I want this (relatively simple) code as my basis for comparison.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1526883004
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot

Committed: https://skia.googlesource.com/skia/+/78e0aef610d762009ef7f4eb51a83771443be665

[mtklein_C, 2015-12-15 17:47:18]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: TBD

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

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: TBD

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

[mtklein, 2015-12-15 17:58:31]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: TBD

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.23ms  (0.93x)   (That's BlitMask_D32_A8.)

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

[mtklein, 2015-12-15 18:01:46]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.23ms  (0.93x)   (That's BlitMask_D32_A8.)

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.23ms  (0.93x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) also show speedups in the 0.95x
(black) to 0.85x (other opaque) range.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

[mtklein, 2015-12-15 18:03:07]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.23ms  (0.93x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) also show speedups in the 0.95x
(black) to 0.85x (other opaque) range.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.23ms  (0.93x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) also show speedups in the 0.95x
(black) to 0.85x (other opaque) range.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

[mtklein, 2015-12-15 18:18:40]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.23ms  (0.93x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) also show speedups in the 0.95x
(black) to 0.85x (other opaque) range.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.13ms  (0.90x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) show speedups similar to
SrcOver_aa.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

[mtklein, 2015-12-15 19:12:15]

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

[commit-bot: I haz the power, 2015-12-15 19:12:21]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1526883004/120001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1526883004/120001

[mtklein, 2015-12-15 19:13:36]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.13ms  (0.90x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) show speedups similar to
SrcOver_aa.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) show speedups similar to
SrcOver_aa,
in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

[mtklein, 2015-12-15 19:13:54]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) show speedups similar to
SrcOver_aa,
in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) show speedups similar to
SrcOver_aa,
in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

[commit-bot: I haz the power, 2015-12-15 19:31:36]

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

[commit-bot: I haz the power, 2015-12-15 19:31:37]

Dry run: This issue passed the CQ dry run.

[mtklein, 2015-12-15 19:55:13]

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

[commit-bot: I haz the power, 2015-12-15 19:55:19]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1526883004/160001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1526883004/160001

[mtklein, 2015-12-15 19:57:23]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) show speedups similar to
SrcOver_aa,
in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) show speedups similar to
SrcOver_aa,
in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

[mtklein, 2015-12-15 19:59:27]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

Opaque AA text blits (text_16_AA_{FF,WT,BK}) show speedups similar to
SrcOver_aa,
in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

[mtklein_C, 2015-12-15 20:03:43]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot

COMMIT=false
Don't land until 1523363002 does.
==========

[mtklein_C, 2015-12-15 20:09:42]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot

COMMIT=false
Don't land until 1523363002 does.
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

I've punted for now on pmaddubsw for lerping. I do intend to try that, but I
want this (relatively simple) code as my basis for comparison.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot

COMMIT=false
Don't land until 1523363002 does.
==========

[mtklein_C, 2015-12-15 20:09:51]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

I've punted for now on pmaddubsw for lerping. I do intend to try that, but I
want this (relatively simple) code as my basis for comparison.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot

COMMIT=false
Don't land until 1523363002 does.
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

I've punted for now on pmaddubsw for lerping. I do intend to try that,
but I want this (relatively simple) code as my basis for comparison.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot

COMMIT=false
Don't land until 1523363002 does.
==========

[commit-bot: I haz the power, 2015-12-15 20:12:36]

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

[commit-bot: I haz the power, 2015-12-15 20:12:36]

Dry run: This issue passed the CQ dry run.

[mtklein_C, 2015-12-15 20:13:06]

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

[mtklein_C, 2015-12-15 20:13:06]

This is a start at capitalizing on all that hacking we did the last couple
weeks.

[herb_g, 2015-12-16 19:08:50]

lgtm

[mtklein, 2015-12-16 19:15:12]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

I've punted for now on pmaddubsw for lerping. I do intend to try that,
but I want this (relatively simple) code as my basis for comparison.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot

COMMIT=false
Don't land until 1523363002 does.
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

I've punted for now on pmaddubsw for lerping. I do intend to try that,
but I want this (relatively simple) code as my basis for comparison.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

[mtklein, 2015-12-16 19:15:15]

The CQ bit was checked by mtklein@google.com

[commit-bot: I haz the power, 2015-12-16 19:15:22]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1526883004/160001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1526883004/160001

[commit-bot: I haz the power, 2015-12-16 19:16:00]

Description was changed from

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

I've punted for now on pmaddubsw for lerping. I do intend to try that,
but I want this (relatively simple) code as my basis for comparison.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
==========

to

==========
SSE 4.1 SrcOver blits: color32, blitmask.

This is mainly warmup for an AVX2 version.
The machine I'm typing this on just doesn't support AVX2.
This strategy should translate easily down to SSSE3 and SSE2.

Xfermode_SrcOver:    2.73ms -> 2.62ms  (0.96x)   (That's Color32.)
Xfermode_SrcOver_aa: 3.48ms -> 3.09ms  (0.89x)   (That's BlitMask_D32_A8.)

AA text blits (text_16_AA_{88,FF,WT,BK}) show speedups in the range of 5 to 20%.

Unlike previous versions of this code, all the div255() are exactly (x+127)/255.
This won't fix any major bugs, but it does correct our bias in the middle.
There will be many diffs, all minor.

I've punted for now on pmaddubsw for lerping. I do intend to try that,
but I want this (relatively simple) code as my basis for comparison.

BUG=skia:
GOLD_TRYBOT_URL=
https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&is...
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot

Committed:
https://skia.googlesource.com/skia/+/78e0aef610d762009ef7f4eb51a83771443be665
==========

[commit-bot: I haz the power, 2015-12-16 19:16:01]

Committed patchset #9 (id:160001) as
https://skia.googlesource.com/skia/+/78e0aef610d762009ef7f4eb51a83771443be665