Disable bigblurs GM (for fe689c46a76)

BUG=skia:
R=humper@google.com
TBR=humper@google.com
NOTRY=True
NOTREECHECKS=True

Author: jvanverth@google.com

Review URL: https://codereview.chromium.org/325453002

one channel texture fix for blur shader

BUG=skia:
R=bsalomon@google.com
TBR=bsalomon
NOTRY=True
NOTREECHECKS=True

Author: humper@google.com

Review URL: https://codereview.chromium.org/319543007

disable GPU/CPU blur match test (possibly temporary)

BUG=skia:
TBR=jvanverth
NOTRY=True
NOTREECHECKS=True

Author: humper@google.com

Review URL: https://codereview.chromium.org/322593002

fix blur test

BUG=skia:
TBR=jvanverth
NOTRY=True
NOTREECHECKS=True

Author: humper@google.com

Review URL: https://codereview.chromium.org/323583002

Revert of Revert of another GM to ignore for blur rebaseline (https://codereview.chromium.org/319943004/)

Reason for revert:
Hopefully working now

Original issue's description:
> Revert of another GM to ignore for blur rebaseline (https://codereview.chromium.org/319203002/)
>
> Reason for revert:
> Follow-on to fe689c46a76, which needs to be reverted due to bot failures.
>
> Original issue's description:
> > another GM to ignore for blur rebaseline
> >
> > BUG=skia:
> > TBR=jvanverth
> > NOTRY=True
> > NOTREECHECKS=True
> >
> > Committed: https://skia.googlesource.com/skia/+/f131055929afb9c9671aee3e575c380d48c5ee4d
>
> TBR=humper@google.com
> NOTREECHECKS=true
> NOTRY=true
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/f255de0fe160e33b340301859be1d9374c5607fe

R=jvanverth@google.com
TBR=jvanverth@google.com
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Author: humper@google.com

Review URL: https://codereview.chromium.org/324453004

Revert of another GM to ignore for blur rebaseline (https://codereview.chromium.org/319203002/)

Reason for revert:
Follow-on to fe689c46a76, which needs to be reverted due to bot failures.

Original issue's description:
> another GM to ignore for blur rebaseline
>
> BUG=skia:
> TBR=jvanverth
> NOTRY=True
> NOTREECHECKS=True
>
> Committed: https://skia.googlesource.com/skia/+/f131055929afb9c9671aee3e575c380d48c5ee4d

R=humper@google.com
TBR=humper@google.com
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Author: jvanverth@google.com

Review URL: https://codereview.chromium.org/319943004

another GM to ignore for blur rebaseline

BUG=skia:
TBR=jvanverth
NOTRY=True
NOTREECHECKS=True

Author: humper@google.com

Review URL: https://codereview.chromium.org/319203002

Faster GPU rect blur that doesn't require computing vertical and horizontal
scanlines on the CPU first.  Should make extremely large drop shadows fast
again.

BUG=skia:
R=bsalomon@google.com

Author: humper@google.com

Review URL: https://codereview.chromium.org/316273004

DM: SKP source / PDF backend

Removed expectations code for PDF backend for now, given that we don't have any, and refactored a little to make that cleaner.

We can now test .skp -> .pdf -> .png in DM.  Neat eh?

BUG=skia:2598
R=halcanary@google.com, mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/316643003

remove SK_SUPPORT_LEGACY_SETCONFIG_INFO and SK_SUPPORT_LEGACY_INSTALLPIXELSPARAMS from chrome

TBR=
NOTRY=True

Author: reed@google.com

Review URL: https://codereview.chromium.org/320583003

ARM Skia NEON patches - 39 - arm64 565 blitters

This enables all 565 blitters except S32A_D565_Opaque.

Here are some performance results:

S32_D565_Opaque:
================

+-------+------------+------------+
| count | Cortex-A53 | Cortex-A57 |
+-------+------------+------------+
|     1 |    -18.37% |    -13.04% |
+-------+------------+------------+
|     2 |     -9.90% |    -13.78% |
+-------+------------+------------+
|     4 |     -8.28% |     -6.77% |
+-------+------------+------------+
|     8 |    157.63% |     78.15% |
+-------+------------+------------+
|    16 |     72.67% |     44.81% |
+-------+------------+------------+
|    64 |     76.78% |     40.89% |
+-------+------------+------------+
|   256 |     73.85% |     36.05% |
+-------+------------+------------+
|  1024 |     75.73% |     36.70% |
+-------+------------+------------+

S32_D565_Blend:
===============

+-------+------------+------------+
| count | Cortex-A53 | Cortex-A57 |
+-------+------------+------------+
|     1 |     -9.99% |    -13.79% |
+-------+------------+------------+
|     2 |     -9.17% |     -6.74% |
+-------+------------+------------+
|     4 |     -6.73% |     -4.42% |
+-------+------------+------------+
|     8 |    163.31% |    112.82% |
+-------+------------+------------+
|    16 |     55.21% |     44.68% |
+-------+------------+------------+
|    64 |     54.09% |     41.99% |
+-------+------------+------------+
|   256 |     52.63% |     40.64% |
+-------+------------+------------+
|  1024 |     52.46% |     40.45% |
+-------+------------+------------+

S32A_D565_Blend:
================

+-------+------------+------------+
| count | Cortex-A53 | Cortex-A57 |
+-------+------------+------------+
|     1 |     -5.88% |     -6.06% |
+-------+------------+------------+
|     2 |     -4.74% |     -0.01% |
+-------+------------+------------+
|     4 |     -5.42% |     -3.03% |
+-------+------------+------------+
|     8 |     78.78% |     77.96% |
+-------+------------+------------+
|    16 |     98.19% |     79.61% |
+-------+------------+------------+
|    64 |    111.56% |     72.60% |
+-------+------------+------------+
|   256 |    113.80% |     69.96% |
+-------+------------+------------+
|  1024 |    114.42% |     70.85% |
+-------+------------+------------+

S32_D565_Opaque_Dither:
=======================

+-------+------------+------------+
| count | Cortex-A53 | Cortex-A57 |
+-------+------------+------------+
|     1 |     -4.18% |     -0.93% |
+-------+------------+------------+
|     2 |     -2.43% |     -2.04% |
+-------+------------+------------+
|     4 |     -1.09% |     -1.23% |
+-------+------------+------------+
|     8 |    184.89% |    136.53% |
+-------+------------+------------+
|    16 |    128.64% |     89.11% |
+-------+------------+------------+
|    64 |    132.68% |    100.98% |
+-------+------------+------------+
|   256 |    157.02% |    100.86% |
+-------+------------+------------+
|  1024 |    163.85% |    103.62% |
+-------+------------+------------+

S32_D565_Blend_Dither:
======================

+-------+------------+------------+
| count | Cortex-A53 | Cortex-A57 |
+-------+------------+------------+
|     1 |     -4.87% |      0.01% |
+-------+------------+------------+
|     2 |     -2.71% |      2.97% |
+-------+------------+------------+
|     4 |     -2.20% |      0.28% |
+-------+------------+------------+
|     8 |    149.76% |    146.80% |
+-------+------------+------------+
|    16 |     85.69% |     95.77% |
+-------+------------+------------+
|    64 |     88.81% |    101.39% |
+-------+------------+------------+
|   256 |     97.32% |    107.22% |
+-------+------------+------------+
|  1024 |     98.08% |    115.71% |
+-------+------------+------------+

S32A_D565_Opaque_Dither:
========================

+-------+------------+------------+
| count | Cortex-A53 | Cortex-A57 |
+-------+------------+------------+
|     1 |     -1.86% |      0.02% |
+-------+------------+------------+
|     2 |     -0.58% |     -1.52% |
+-------+------------+------------+
|     4 |     -0.75% |      1.16% |
+-------+------------+------------+
|     8 |    240.74% |    155.16% |
+-------+------------+------------+
|    16 |    181.97% |    132.15% |
+-------+------------+------------+
|    64 |    203.11% |    136.48% |
+-------+------------+------------+
|   256 |    223.45% |    133.05% |
+-------+------------+------------+
|  1024 |    225.96% |    134.05% |
+-------+------------+------------+

Signed-off-by: Kévin PETIT <kevin.petit@arm.com>
BUG=skia:
R=djsollen@google.com, mtklein@google.com

Author: kevin.petit@arm.com

Review URL: https://codereview.chromium.org/317193003

Effectively unlimited linking.

Ninja will still cap concurrency with it's default value (from -j), so we just
need a big number equal or greater than the number of cores in any machine.

BUG=skia:
R=borenet@google.com, mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/319153002

Revert of Fix Assert for gpu ConicalTwoPointGradient edgecase to be correct bounds. (https://codereview.chromium.org/311373003/)

Reason for revert:
changed assert is failing

Original issue's description:
> Fix Assert for gpu ConicalTwoPointGradient edgecase to be correct bounds.
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/be4fd4f7349b97a7c34a4177ec26d8aea9616628

R=bsalomon@google.com
TBR=bsalomon@google.com
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Author: egdaniel@google.com

Review URL: https://codereview.chromium.org/320573002

Pull in GYP_LINK_CONCURRENCY=N support.

BUG=skia:
R=borenet@google.com, mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/315183003

Fix Assert for gpu ConicalTwoPointGradient edgecase to be correct bounds.

BUG=skia:
R=bsalomon@google.com

Author: egdaniel@google.com

Review URL: https://codereview.chromium.org/311373003

Fixing another clusterfuzz issue

This was introduced by removing SkValidatingReadBuffer::readBitmap in https://codereview.chromium.org/295793002/

Since SkReadBuffer::skip wasn't virtual, it was using the unsafe SkReadBuffer::skip within SkReadBuffer::readBitmap rather than using SkValidatingReadBuffer::skip. I also removed direct uses of fReader within SkReadBuffer::readBitmap so that it can use the virtual readInt / readFixed functions that have a version in SkValidatingReadBuffer.

Also, I changed SkReadBuffer::readPoint so that it uses the virtual readScalar, that way, it becomes redundant with SkValidatingReadBuffer::readPoint, which can then be removed.

BUG=380723
R=reed@google.com, mtklein@google.com, sugoi@google.com

Author: sugoi@chromium.org

Review URL: https://codereview.chromium.org/317003003

Fix android_run_skia script running with command line options

This script cannot recognize command line options when running bench:
e.g. android_run_skia bench --config 8888.

BUG=skia:
R=mtklein@google.com, djsollen@google.com

Author: qiankun.miao@intel.com

Review URL: https://codereview.chromium.org/314903003

Remove GrIsPow2 in favor of SkIsPow2.

Looks like there is no good reason to have two copies of this function
doing the same thing with different name.

BUG=None
TEST=make tests
R=bsalomon@google.com

Author: tfarina@chromium.org

Review URL: https://codereview.chromium.org/318873002

Add metrics reporting to SkFiddle.

Also update the headers for template.cpp while we're here.

BUG=skia:
R=mtklein@google.com

Author: jcgregorio@google.com

Review URL: https://codereview.chromium.org/312373002

Whitespace change to test new CQ trybots

BUG=skia:2203
TBR=

(SkipBuildbotRuns)

Author: rmistry@google.com

Review URL: https://codereview.chromium.org/319123002

Adding kernel offset validation

Fixing bug found by clusterfuzz in SkMatrixConvolutionImageFilter.

BUG=381244
R=senorblanco@google.com, senorblanco@chromium.org, sugoi@google.com

Author: sugoi@chromium.org

Review URL: https://codereview.chromium.org/319003002

Add SkBarriers_tsan.h.

Slight counterproposal to crrev.com/310663002.

BUG=skia:

No API changes.

R=bungeman@google.com, mtklein@google.com, reed@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/306373002

Use vertex attribute for text color in GrBitmapTextContext

R=jvanverth@google.com

Author: bsalomon@google.com

Review URL: https://codereview.chromium.org/316063002

Initial KTX encoder

The encoder comes with tests to check that the encoding/decoding
operations between ETC encoded bitmaps and ARGB bitmaps are sane.

R=bsalomon@google.com, robertphillips@google.com

Author: krajcevski@google.com

Review URL: https://codereview.chromium.org/312353003

Fix SK_API for DirectWrite port.

The declaration and definition should have the same calling convention.

TBR=reed@google.com

Review URL: https://codereview.chromium.org/317993002

Rename SkFontHost_win_dw.cpp SkScalerContext.

This is a continuation of https://codereview.chromium.org/314193002/ .

TBR=reed@google.com

Review URL: https://codereview.chromium.org/315173004

Split SkFontHost_win_dw.

Split SkFontHost_win_dw into FontMgr, Typeface, and ScalerContext.
This makes working on these files easier, and moves away from the legacy
FontHost naming.

R=reed@google.com

Author: bungeman@google.com

Review URL: https://codereview.chromium.org/314193002

SkClipStack::Element tweaks.

(
This is intended to facilitate efficient kMatrix_SaveFlags emulation
on restore():

  * collect all clip stack elements for the current save count into a
    side clip stack
  * canvas.restore(everything)
  * replay the collected clip ops to restore the initial clip state
  => we restored the matrix but the clip state is unchanged
)

Two main changes:

  * expose the save count for SkClipStack::Element
  * expose a replay method for the same (logic relocated from
    SkCanvas::replayClips)

The SkCanvas::ClipVisitor shuffling is to enable forward decl
in SkClipStack.h (cannot fwdecl a nested class).

R=reed@google.com, robertphillips@google.com
TBR=reed@google.com

Author: fmalita@chromium.org

Review URL: https://codereview.chromium.org/269693003

Update gyp for msvs-ninja clean.

BUG=skia:2439
R=borenet@google.com, epoger@google.com

Author: bungeman@google.com

Review URL: https://codereview.chromium.org/317833005

Fixes for SkAlphaTypes.

When using allocPixels to set the info and create a new pixelRef in one
shot, create the pixelRef with the corrected info of the bitmap (i.e.
alphatype).

Simplify the check to determine whether to copy the genID in SkBitmap::
copyTo's same configs/same sizes case. The old logic (only checking the
height and assuming the others must be the same) relied on a deep
(mis?)understanding of the rest of the function, and I cannot convince
myself that it has to be true, given that readPixels may have switched
src to a different SkBitmap. Instead, just compare the SkImageInfos,
which is much clearer. (This also caught a bug where a 565 bitmap had
the wrong alphatype, leading to the allocPixels fixes.)

If readPixels succeeds but the source is unpremultiplied, fail out
of copyTo, since readPixels assumes premultiplied.

When copying from 8888 to 4444 or in the fallback case of drawing,
return false if the alphatype is unpremul, which would have failed
anyway.

BUG=skia:2012
R=reed@google.com, bsalomon@google.com

Author: scroggo@google.com

Review URL: https://codereview.chromium.org/317053003

reland "rebaseline_server: download actual-results.json files from GCS instead of SVN"

relands https://codereview.chromium.org/310093003 with modifications.

BUG=skia:2641
R=jcgregorio@google.com

Author: epoger@google.com

Review URL: https://codereview.chromium.org/313343003

Add mips64 to the makefile.

Currently we have no optimizations for mips64, so it correctly
contains the various _none.cpps.

Corresponds to https://android-review.googlesource.com/#/c/96536/1

R=halcanary@google.com, djsollen@google.com

Author: scroggo@google.com

Review URL: https://codereview.chromium.org/311163003

add SK_SUPPORT_LEGACY_COMPUTE_CONFIG_SIZE

TBR=scroggo
NOTRY=True

Author: reed@google.com

Review URL: https://codereview.chromium.org/315263002

Rename skimage expectations for ASAN bot to match bot renaming.

BUG=skia:
R=borenet@google.com, mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/316063004

remove SkBitmap::allocConfigPixels and update dox

I see no callers in android

R=scroggo@google.com

Author: reed@google.com

Review URL: https://codereview.chromium.org/310283004

Revert of Add SSE4 optimization of S32A_Opaque_Blitrow (https://codereview.chromium.org/289473009/)

Reason for revert:
Buildbot failures on Mac 10.6 and Mac 10.7.

R=reed@google.com, mtklein@google.com, tomhudson@google.com, djsollen@google.com, joakim.landberg@intel.com, henrik.smiding@intel.com
TBR=reed@google.com
NOTRY=True

Original issue's description:
> Add SSE4 optimization of S32A_Opaque_Blitrow
>
> Adds optimization of Skia S32A_Opaque_Blitrow blitter using SSE4.2 SIMD
> instruction set. Special case for when alpha is zero or opaque.
>
> Performance increase of 10%-400% compared to the existing SSE2
> optimization (measured on Silvermont architecture).
> Noticeable in ~25 different skia bench subtests, especially in
> bitmap_8888_*, repeatTile_*, and morph_*.
>
> bitmap_8888_A - 100% faster
> bitmap_8888_A_source_transparent - 250% faster
> bitmap_8888_A_source_opaque - 25% faster
> bitmap_8888_A_scale_bicubic - 75% faster
>
> Signed-off-by: Henrik Smiding <henrik.smiding@intel.com>
>
> Committed: https://skia.googlesource.com/skia/+/e2527b147679b0c43019fae7d59cc3777d2d097e

Author: jvanverth@google.com

Review URL: https://codereview.chromium.org/311053009

remove ashmem from skia

R=scroggo@google.com, djsollen@google.com

Author: halcanary@google.com

Review URL: https://codereview.chromium.org/319533004

Rebaselines for drrect GM on GPU
TBR=

Author: bsalomon@google.com

Review URL: https://codereview.chromium.org/318773007

Add SSE4 optimization of S32A_Opaque_Blitrow

Adds optimization of Skia S32A_Opaque_Blitrow blitter using SSE4.2 SIMD
instruction set. Special case for when alpha is zero or opaque.

Performance increase of 10%-400% compared to the existing SSE2
optimization (measured on Silvermont architecture).
Noticeable in ~25 different skia bench subtests, especially in
bitmap_8888_*, repeatTile_*, and morph_*.

bitmap_8888_A - 100% faster
bitmap_8888_A_source_transparent - 250% faster
bitmap_8888_A_source_opaque - 25% faster
bitmap_8888_A_scale_bicubic - 75% faster

Signed-off-by: Henrik Smiding <henrik.smiding@intel.com>
R=reed@google.com, mtklein@google.com, tomhudson@google.com, djsollen@google.com, joakim.landberg@intel.com

Author: henrik.smiding@intel.com

Review URL: https://codereview.chromium.org/289473009

Use the correct homophone of 'there'.

R=reed@google.com
TBR=reed@google.com

Author: scroggo@google.com

Review URL: https://codereview.chromium.org/319493004

Fix submit_try

BUG=skia:2643
R=epoger@google.com, tfarina@chromium.org

Author: borenet@google.com

Review URL: https://codereview.chromium.org/317823003

Add bench for rectanizers

R=bsalomon@google.com, jvanverth@google.com

Author: robertphillips@google.com

Review URL: https://codereview.chromium.org/319523003

Change default generators on windows to include msvs-ninja rather than msvs

BUG=skia:2439
R=epoger@google.com
TBR=epoger@google.com

Author: bsalomon@google.com

Review URL: https://codereview.chromium.org/317903002

Two new benches for ETC1 bitmaps.

First bench: Simply render from a compressed ETC1 bitmap. This is roughly
equal between compressed and uncompressed data.

Second bench: Notify the bitmap that the pixels have changed (that's right,
like a liar) to force a re-upload to the GPU. On this bench, decompressed
textures seem to inexplicably do an order of magnitude better than
compressed textures when it should be the other way around, but that
investigation is reserved for a future CL.

R=robertphillips@google.com

Author: krajcevski@google.com

Review URL: https://codereview.chromium.org/317023002

On Windows, fail if there are missing files specified in the project.

BUG=skia:2638
R=robertphillips@google.com, borenet@google.com, rmistry@google.com, epoger@google.com

Author: jvanverth@google.com

Review URL: https://codereview.chromium.org/319503003

Remove SK_IGNORE_DWRITE_BITMAP_FIX.

This define has been removed from Blink and is no longer used anywhere.

R=reed@google.com
TBR=reed1

Author: bungeman@google.com

Review URL: https://codereview.chromium.org/319513006

Revert "rebaseline_server: download actual-results.json files from GCS instead of SVN"

This reverts commit f4666045f378dee22ce40108e90bf5195634ebfc.

BUG=skia:2641
NOTREECHECKS=True
NOTRY=True
R=jcgregorio@google.com
TBR=jcgregorio

Author: epoger@google.com

Review URL: https://codereview.chromium.org/317783004

Update expectations for gyp change.

It appears that with the gyp update, floating point operations
on x86_32 are ever so slightly different when drawing some paths.

R=jvanverth@google.com
TBR=jvanverth@google.com
NOTRY=True

Author: bungeman@google.com

Review URL: https://codereview.chromium.org/315043006

Remove Android flag for legacy SkCanvas::drawPicture API

Android should no longer need this flag

R=scroggo@google.com, djsollen@google.com

Author: robertphillips@google.com

Review URL: https://codereview.chromium.org/316933005

Remove SkImageRef and related functionality.

From here on out, use SkDecodingImageGenerator+SkDiscardablePixelRef.

R=scroggo@google.com

Review URL: https://codereview.chromium.org/100183002

Update gyp for msvs-ninja.

BUG=skia:2439
R=borenet@google.com, epoger@google.com

Author: bungeman@google.com

Review URL: https://codereview.chromium.org/311213002

Add utils gypi file

BUG=skia:
R=bsalomon@google.com, robertphillips@google.com

Author: egdaniel@google.com

Review URL: https://codereview.chromium.org/317733003

Relocate SK_IGNORE_CORRECT_HIGH_QUALITY_IMAGE_SCALE

Temporarily move the flag into Chromium's repo, to facilitate
atomic remove-and-suppress-failures changes.

BUG=380686
R=humper@google.com
TBR=humper@google.com

Author: fmalita@chromium.org

Review URL: https://codereview.chromium.org/319533002

rebaseline_server: download actual-results.json files from GCS instead of SVN

BUG=skia:553
R=borenet@google.com

Author: epoger@google.com

Review URL: https://codereview.chromium.org/310093003

Revert of Move Dashing filterPath to a dashing utils file (https://codereview.chromium.org/314623004/)

Reason for revert:
Need to create utils.gypi file and update chrome

Original issue's description:
> Move Dashing filterPath to a dashing utils file
>
> From inside GrContext, we have a need to create an SkPath an original path and
> some dashing info. We do not have access to the original path effect so we need
> a way to make the FilterPath function accessible outside of the effect. So I moved
> the core filterPath code (and all need helper functions) out of SkDashPathEffect
> and created a SkDashPath in utils to store these helper functions.
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/576dcdc793a762ec63fbecdbfd5768066b548fe5

R=bsalomon@google.com, reed@google.com, rmistry@google.com
TBR=bsalomon@google.com, reed@google.com, rmistry@google.com
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Author: egdaniel@google.com

Review URL: https://codereview.chromium.org/317663003

Add non power of two gm for ETC1 bitmap

R=robertphillips@google.com

Author: krajcevski@google.com

Review URL: https://codereview.chromium.org/316813005

Use tests tag for Skia tools on Android.

When generating the makefiles for Skia testing tools on Android,
use the tag 'tests' and do not use the tags 'optional' or 'eng'.

R=djsollen@google.com

Author: scroggo@google.com

Review URL: https://codereview.chromium.org/315013003

Use Identity matrix for SkTransparentShader.

The SkTransparentShader is just a placeholder that should not
actually be used, so its CTM is not meaningful. However, it needs
to be invertible in order to not assert in SkShader::Context's
constructor. So use SkMatrix::I().

This fixes a CTS test on Android.

R=reed@google.com

Author: scroggo@google.com

Review URL: https://codereview.chromium.org/315743003

Is this SkDebug important?  It drives me nuts.

BUG=skia:
R=bsalomon@google.com, mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/310113006

Whitespace change to test bugdroid

(SkipBuildbotRuns)

BUG=skia:2139
BUG=chromium:380617

Move Dashing filterPath to a dashing utils file

From inside GrContext, we have a need to create an SkPath an original path and
some dashing info. We do not have access to the original path effect so we need
a way to make the FilterPath function accessible outside of the effect. So I moved
the core filterPath code (and all need helper functions) out of SkDashPathEffect
and created a SkDashPath in utils to store these helper functions.

BUG=skia:
R=bsalomon@google.com, reed@google.com, rmistry@google.com

Author: egdaniel@google.com

Review URL: https://codereview.chromium.org/314623004

Add bench to measure drawing a dashed grid

BUG=skia:
R=bsalomon@google.com

Author: egdaniel@google.com

Review URL: https://codereview.chromium.org/310083005

Alter SkCanvas::drawPicture (devirtualize, take const SkPicture, take pointer)

R=reed@google.com, bsalomon@google.com, mtklein@google.com

Author: robertphillips@google.com

Review URL: https://codereview.chromium.org/313613004

Whitespace change to test bugdroid

(SkipBuildbotRuns)

BUG=skia:2139,chromium:380617

Update SKP version to 17

Automatic commit by the RecreateSKPs bot.

TBR=

Author: borenet@google.com

Review URL: https://codereview.chromium.org/312043004

readd SK_SUPPORT_LEGACY_SETCONFIG_INFO -- missed a place in a blink test

R=jvanverth@google.com
TBR=jvanverth
NOTRY=True

Author: reed@google.com

Review URL: https://codereview.chromium.org/316753002

Disable PDF in dm for tonight.  It's crashy.

BUG=skia:2598
R=mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/317533003

Added grid data to logging output

BUG=skia:
R=bensong@google.com, jcgregorio@google.com

Author: kelvinly@google.com

Review URL: https://codereview.chromium.org/318433003

remove unneeded flags SK_SUPPORT_LEGACY_SETCONFIG_INFO and SK_SUPPORT_LEGACY_DEVICE_CONFIG

TBR=scroggo

Author: reed@google.com

Review URL: https://codereview.chromium.org/315733002

Android: remove -lcutils dependency

I believe we no longer depend on this after we ported our atomics to __sync.

Tested by running android_ninja.

BUG=skia:
R=scroggo@google.com, mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/318493002

DM: add pdf

BUG=skia:2598
R=halcanary@google.com, mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/312873002

DM: go back to memcmp for BitmapsEqual

Even when autovectorized, using MaxComponentDifference is slower than memcmp.
And debug builds (most runs of DM) will never even be autovectorized.

DM::MaxComponentDifference is the top function on DM profile, and memcmp moves
to ~20th.

BUG=skia:
R=halcanary@google.com, mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/315743002

Whitespace change to test the new cron.chrome bugdroid.

(SkipBuildbotRuns)
BUG=skia:2139

Review URL: https://codereview.chromium.org/309333003

Fix neglected cleanup of SkTileGridPicture.h

R=bsalomon@google.com

Author: robertphillips@google.com

Review URL: https://codereview.chromium.org/312803003

Return NULL when building empty LayerRasterizer.

In SkLayerRasterizer::snapshotRasterizer() and ::detachRasterizer(),
if no layers have been added, do not attempt to create an
SkLayerRasterizer. Instead, return NULL.

This fixes an error when running tests on Android.

Update dox to state that NULL may be returned.

Add tests.

R=reed@google.com

Author: scroggo@google.com

Review URL: https://codereview.chromium.org/313653006

Initial KTX file decoder

R=bsalomon@google.com, robertphillips@google.com, halcanary@google.com, reed@google.com

Author: krajcevski@google.com

Review URL: https://codereview.chromium.org/302333002

Add LICENSE & README.google files to third_party/etc1

R=djsollen@google.com, krajcevski@google.com

Author: robertphillips@google.com

Review URL: https://codereview.chromium.org/309343004

SK_CPU_ARM --> SK_CPU_ARM32

That's what it means.  It keeps confusing us as named today.

BUG=skia:
R=djsollen@google.com, mtklein@google.com, reed@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/314643004

manual bench rebase after 55ada06

R=robertphillips@google.com
TBR=robertphillips@google.com

Bypassing trybots:
NOTRY=true

Author: jvanverth@google.com

Review URL: https://codereview.chromium.org/315683004

remove SkBounder -- unused and unloved

BUG=skia:
R=scroggo@google.com

Author: reed@google.com

Review URL: https://codereview.chromium.org/312553006

download_actuals.py: download JSON files from Google Storage instead of skia-autogen

uses google-api-python-client instead of gsutil binary to interact with Google Storage

BUG=skia:553
R=rmistry@google.com, bensong@google.com

Author: epoger@google.com

Review URL: https://codereview.chromium.org/309653005

a few rebaseline stragglers from image scaling change

BUG=skia:
NOTRY=True
NOTREECHECKS=True
TBR=jvanverth

Author: humper@google.com

Review URL: https://codereview.chromium.org/311853004

ARM Skia NEON patches - 38 - arm64 8888 blitters

Enable NEON on arm64 for most 8888 blitters

This patch enables NEON optimisation for the Color32, S32_Blend,
S32A_Opaque blitters on arm64.

Here are the perf improvements vs the existing code:

Color32:
========

+-------+------------+------------+
| count | Cortex-A53 | Cortex-A57 |
+-------+------------+------------+
|     1 |     -2.39% |     23.78% |
+-------+------------+------------+
|     2 |     -5.46% |      8.88% |
+-------+------------+------------+
|     4 |     -4.74% |      4.89% |
+-------+------------+------------+
|     8 |     67.74% |    107.12% |
+-------+------------+------------+
|    16 |     40.03% |    101.20% |
+-------+------------+------------+
|    64 |     11.09% |     98.40% |
+-------+------------+------------+
|   256 |     -2.20% |     74.81% |
+-------+------------+------------+
|  1024 |     -4.28% |     78.90% |
+-------+------------+------------+

S32_Blend:
==========

+-------+------------+------------+
| count | Cortex-A53 | Cortex-A57 |
+-------+------------+------------+
|     1 |      7.84% |     -6.75% |
+-------+------------+------------+
|     2 |     28.95% |     39.77% |
+-------+------------+------------+
|     4 |      5.80% |      8.26% |
+-------+------------+------------+
|     8 |      1.35% |     33.80% |
+-------+------------+------------+
|    16 |     -2.13% |     41.13% |
+-------+------------+------------+
|    64 |     -4.91% |     42.84% |
+-------+------------+------------+
|   256 |     -6.53% |     48.72% |
+-------+------------+------------+
|  1024 |     -6.65% |     46.66% |
+-------+------------+------------+

S32A_Opaque:
============

+-------+------------+------------+
| count | Cortex-A53 | Cortex-A57 |
+-------+------------+------------+
|     1 |     -7.51% |    -19.06% |
+-------+------------+------------+
|     2 |     -5.02% |    -27.70% |
+-------+------------+------------+
|     4 |     15.38% |    -21.66% |
+-------+------------+------------+
|     8 |     -0.98% |      1.05% |
+-------+------------+------------+
|    16 |     -7.35% |      3.34% |
+-------+------------+------------+
|    64 |     50.53% |     94.63% |
+-------+------------+------------+
|   256 |     71.17% |    164.10% |
+-------+------------+------------+
|  1024 |     79.58% |    197.60% |
+-------+------------+------------+

Signed-off-by: Kevin PETIT <kevin.petit@arm.com>
BUG=skia:
R=djsollen@google.com, mtklein@google.com

Author: kevin.petit@arm.com

Review URL: https://codereview.chromium.org/302283003

Remove legacy picture recording

This is unblocked now that Android no longer uses the old interface.

This is just the first step in cleaning this up. Future CLs will constify SkPicture access in SkCanvas and split up the SkPicture/SkPicturePlayback/SkPictureRecord trio.

R=bsalomon@google.com, reed@google.com, mtklein@google.com

Author: robertphillips@google.com

Review URL: https://codereview.chromium.org/313613002

Update SKP version to 16

Automatic commit by the RecreateSKPs bot.

TBR=

Author: borenet@google.com

Review URL: https://codereview.chromium.org/313593002

Fix etc1 support for SkFiddle.

Also update design doc to fix MySQL notation for auto incrementing columns.

BUG=skia:
R=mtklein@google.com

Author: jcgregorio@google.com

Review URL: https://codereview.chromium.org/315643002

Fix memory leak in etc1bitmap GM.

We never drop our first ref on fileData. (All our ref counts start at one).
SkAutoTUnref<T> is almost always the right thing to use for refcounted types.

BUG=skia:2635
R=krajcevski@google.com, mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/310793003

Allow chrome to use old version of installPixels.

R=jvanverth@google.com, reed@google.com

Author: scroggo@google.com

Review URL: https://codereview.chromium.org/312703003

Update link to Doxygen.

BUG=skia:
R=mtklein@google.com

Author: jcgregorio@google.com

Review URL: https://codereview.chromium.org/302543007

Rebase a few GMs and remove some tests from ignored-tests.txt

Review URL: https://codereview.chromium.org/311813003

Remove brackets from gm names.

BUG=skia:2634
R=epoger@google.com, borenet@google.com

Author: scroggo@google.com

Review URL: https://codereview.chromium.org/307373003

Fall back to using clip effect for outer rect in drawdrrect

R=jvanverth@google.com

Author: bsalomon@google.com

Review URL: https://codereview.chromium.org/310483010

Whitespace change to test multiple master support in CQ

TBR=
BUG=skia:2624

(SkipBuildbotRuns)

Author: rmistry@google.com

Review URL: https://codereview.chromium.org/315633002

remove unneeded flags

TBR=

Author: reed@google.com

Review URL: https://codereview.chromium.org/316463002

Whitespace change to test new CQ trybot

TBR=
BUG=skia:2203
(SkipBuildbotRuns)

Author: rmistry@google.com

Review URL: https://codereview.chromium.org/311813002

Revert of Try TSAN blacklist for SkGLContextHelper::init. (https://codereview.chromium.org/309823004/)

Reason for revert:
the bot's clang doesn't know -fsanitize-blacklist

Original issue's description:
> Try TSAN blacklist for SkGLContextHelper::init.
>
> Submitting to test this; this particular problem only happens on the bot.
>
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/ab2ec19b975a898d4ee2278ddad7d4268f134478

R=mtklein@chromium.org
TBR=mtklein@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Author: mtklein@google.com

Review URL: https://codereview.chromium.org/314543002

Try TSAN blacklist for SkGLContextHelper::init.

Submitting to test this; this particular problem only happens on the bot.

BUG=skia:
R=mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/309823004

Clean up another silly race in benches when run concurrently.

BUG=skia:
R=mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/309193003

Tweak TSAN suppressions.

BUG=skia:
R=mtklein@google.com

Author: mtklein@chromium.org

Review URL: https://codereview.chromium.org/312613002

remove unneeded inval

TBR=

Author: reed@google.com

Review URL: https://codereview.chromium.org/307993012