Update libjpeg-turbo to 1.4.90

Google3, Chrome, and Android are considering updates.  Let's start
testing Skia on the newest version.
https://critique.corp.google.com/#review/116667446
https://codereview.chromium.org/1934113002
https://buganizer.corp.google.com/u/0/issues/28268702

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1940903002

Review-Url: https://codereview.chromium.org/1940903002

remove non-static uses of SkOncePtr

Still slowly working through all the SK_DECLARE_STATIC_FOO macros.

SkOncePtr is complicating things by having SkOncePtr delete its pointer
and SkBaseOncePtr not.  Simplify things by removing SkOncePtr, leaving
only the leaky SkBaseOncePtr.

We replace SkOncePtr<T> instead with SkOnce and T.  In most cases this
did not need to be a pointer, and in some cases here we're even saving
a few bytes by replacing SkOncePtr<T> with SkOnce and a T.

The dependency map of SK_DECLARE_STATIC_FOO is:
  SkBaseMutex -> SkBaseSemaphore -> SkBaseOncePtr

They're intertwined enough that I think I've got to do all three in one
next CL.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1939503002

Review-Url: https://codereview.chromium.org/1939503002

Remove unused encodedString API on SkWriteBuffer/SkReadBuffer

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1936103002

Review-Url: https://codereview.chromium.org/1936103002

Unify DescriptorPools for vulkan uniform descriptor sets

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1927943002

Review-Url: https://codereview.chromium.org/1927943002

Disable yet another GCC warning in Release_Developer mode.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1934293002

Review-Url: https://codereview.chromium.org/1934293002

Add CMake path to gradle

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1921553009

Review-Url: https://codereview.chromium.org/1921553009

Enable generating SkImage with the same uniqueID in SkImageGenerator subclass

Enable reusing uniqueID when instantiating SkImageGenerator subclasses enables
using uniqueID in client code to cache generated bitmaps with no need to keep
the reference to SkImageGenerator.

This is a bug fix for out of memory cause in chromium and 100% CPU usage
described in issue 165750#13:
- cache uses SkImage::uniqueID() to cache decoded bitmaps.
- every animation loop creates new SkImage instances.
- after decoding, bitmap copies are added to cache, filling it up with
duplicates of previous loops frames.

BUG=165750

Blink patch that depends on this:
https://codereview.chromium.org/1925533003/
"High CPU and increased memory usage fix for high-res (GIF, WEBP...) animations."

Review-Url: https://codereview.chromium.org/1928403002

Patch in internal cl/121138374.

TBR=mtklein
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1939483002

Review-Url: https://codereview.chromium.org/1939483002

remove SK_SUPPORT_LEGACY_IMAGEFILTER_CTM flag

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1938823002

Review-Url: https://codereview.chromium.org/1938823002

Update SKP version

Automatic commit by the RecreateSKPs bot.

TBR=
NO_MERGE_BUILDS
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1938813002

Review-Url: https://codereview.chromium.org/1938813002

Fix gcc warnings in Vulkan backend

BUG=skia:
TBR=jvanverth@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1935903002

Review-Url: https://codereview.chromium.org/1935903002

Fix race in SkTypeface_FreeType::onCountGlyphs.

Found by TSAN. This makes the operation slightly slower, but it is
not on any user's hot path.

BUG=skia:5238

Review-Url: https://codereview.chromium.org/1940613002

Remove SkWriter32::contiguousArray().

This method requires SkWriter32 have a contiguous array.
It does, and I plan to keep it that way (last time we checked
it's faster), but this turns that feature back into an
implementation detail.

This feature is only used by another unused feature, deep copies of
ops arrays in SkPictureData.  Kill that, kill this.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1932223003

Depends on https://codereview.chromium.org/1936563002/

Only deleting API
TBR=reed@google.com

Review-Url: https://codereview.chromium.org/1932223003

Fix old debugger
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1935623002

Review-Url: https://codereview.chromium.org/1935623002

Delete ChromeOS code

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1926163002

Review-Url: https://codereview.chromium.org/1926163002

Add reminders that these classes have constexpr default constructors.

We can do this now that we can write "constexpr".

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1936653002

Review-Url: https://codereview.chromium.org/1936653002

remove SkWriteBuffer::getWriter32().

SkWriteBuffer exposes its lower-level implementation SkWriter32 through
this one call.  It's not currently used in any interesting way:
   - write_encoded_bitmap() uses it to manually re-create writeDataAsByteArray();
   - unit tests use it incidentally as a quick way to read the serialized bytes.

This should be SkWriteBuffer no longer necessarily needs to have an SkWriter32.

Landing this will let us then remove SkWriter32::contiguousArray().

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1936563002

TBR=reed@google.com
Just deleting API

Review-Url: https://codereview.chromium.org/1936563002

Add mipmap loading to Vulkan.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1925303002

Review-Url: https://codereview.chromium.org/1925303002

Fix warnings that were exposed when running with SK_DEBUG enable on an optimized release build.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1917223006

Review-Url: https://codereview.chromium.org/1917223006

Unbreak vulkan after stencil constant name changes

TBR=egdaniel@google.com,jvanverth@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1930323003

NOTREECHECKS=true

Review-Url: https://codereview.chromium.org/1930323003

Revert of Make stencils be attachable to render targets created via SkSurface::MakeFromBackendTextureAsRender… (patchset #5 id:80001 of https://codereview.chromium.org/1924183003/ )

Reason for revert:
Breaking the DEPS roll.

Also seems to be causing failure in DM on some Windows bots:
c:\0\build\slave\workdir\build\skia\tests\surfacetest.cpp:963 kShapeColor == bitmap.getColor(kW / 2, kH / 2)

Original issue's description:
> Make stencils be attachable to render targets created via SkSurface::MakeFromBackendTextureAsRenderTarget
>
> This is a regression from "Refactor to separate backend object lifecycle
> and GpuResource budget decision".
>
> GrGLRenderTarget::CreateWrapped creates only render targets that wrap
> the FBO.
> GrGLRenderTargetTexture::CreateWrapped creates render targets that
> wrap the texture.
>
> Use the latter as the implementation for
> SkSurface::MakeFromBackendTextureAsRenderTarget.
>
> The test contains disabled code. The
> MakeFromBackendTextureAsRenderTarget does not copy the existing texture
> contents to the FBO render buffer.
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1924183003
>
> Committed: https://skia.googlesource.com/skia/+/fe635fd76bbc375d527d1884af23cc617ca364ee

TBR=bsalomon@google.com,kkinnunen@nvidia.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review-Url: https://codereview.chromium.org/1931293003

Revert of Introduce SkGammas type to represent ICC gamma curves (patchset #7 id:40002 of https://codereview.chromium.org/1928123002/ )

Reason for revert:
Breaks TSAN
https://build.chromium.org/p/client.skia/builders/Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-TSAN/builds/6387/steps/dm/logs/stdio

Original issue's description:
> Introduce SkGammas type to represent ICC gamma curves
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1928123002
>
> Committed: https://skia.googlesource.com/skia/+/7b2c6dd8c918209cb92e1338905d511c68da3eb2

TBR=scroggo@google.com,reed@google.com,brianosman@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review-Url: https://codereview.chromium.org/1933863002

Introduce SkGammas type to represent ICC gamma curves

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1928123002

Review-Url: https://codereview.chromium.org/1928123002

Fix Vulkan Build in Android

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1935523002

Review-Url: https://codereview.chromium.org/1935523002

Don't mark mipmaps dirty if uploading > 1 levels

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1933293002

Review-Url: https://codereview.chromium.org/1933293002

Add ColorCodecSrc for testing/comparison on color corrected decodes

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1933753002

Review-Url: https://codereview.chromium.org/1933753002

remove SkWriteBuffer::reserve()

We're not using it for anything that we can't do using the clearer writeFoo()/readFoo() methods.

The deletions in SkPictureFlat.{h,cpp} and BitmapHeapTest.cpp clean up dead code.
This code has not been used for a long time.  I happened to notice it because it
was another caller of reserve() (in a terrifyingly hacking/interesting way that
I bet we'll find I authored...)

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1930103003

Deleting APIs unused publicly.
TBR=reed@google.com

Review-Url: https://codereview.chromium.org/1930103003

Use constexpr constructors of GrStencilSettings rather than hacky macro system.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1933053002

Review-Url: https://codereview.chromium.org/1933053002

Remove skstd::is_function.

The VC++ STL with 2015u2 now provides a complete std::is_function.
Also, Skia is no longer using skstd::is_function.

Review-Url: https://codereview.chromium.org/1929343002

Require Android SDK copy of CMake to build Vulkan on Android.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1930423002

Review-Url: https://codereview.chromium.org/1930423002

Revert of Adjust tolerance to fix bezier glitches issue (patchset #1 id:1 of https://codereview.chromium.org/1926063002/ )

Reason for revert:
This is causing undesirable image failures in a large number of Skia GMs.

Original issue's description:
> Adjust tolerance to fix bezier glitches issue
>
> In a previous CL: https://codereview.chromium.org/1126793003, the value
> of gDegenerateToLineTol was changed from 1.0 to 0.25 which caused the
> bezier glitches.
>
> This CL changes the value to SK_ScalarRoot2Over2, with this change the
> bezier curve looks fine, and the result of the layout tests pointed out
> in this bug: https://bugs.chromium.org/p/chromium/issues/detail?id=472847
> is also reasonable.
>
> BUG=605439
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1926063002
>
> Committed: https://skia.googlesource.com/skia/+/45a6f49ef9ff276e08025d0d2ba6a821a96d3f34

TBR=robertphillips@google.com,xidachen@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=605439

Review-Url: https://codereview.chromium.org/1933733002

Update running Chromium trybots on Skia CLs docs.

NOTRY=true
DOCS_PREVIEW= https://skia.org/?cl=1930773005

Review-Url: https://codereview.chromium.org/1930773005

Unify implementations of stroking radius calculationsns
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1928133002

Review-Url: https://codereview.chromium.org/1928133002

Add sk_sp to SkSurface_Gpu and SkGpuDevice

Split off of https://codereview.chromium.org/1930013002/ ((Mostly) Retract GrRenderTarget from SkGpuDevice)

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1925803004

Review-Url: https://codereview.chromium.org/1925803004

Repurpose Release_Developer BUILDTYPE and remove SK_DEVELOPER.

The Release_Developer build type is now used to build Skia with
release level optimizations but still enable SK_DEBUG. This in turn
means that SkASSERTS and SkRTConf are available in this mode.

Further we can then remove SK_DEVELOPER as a define as it is true
iff SK_DEBUG is true.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1931903002
DOCS_PREVIEW= https://skia.org/?cl=1931903002

Review-Url: https://codereview.chromium.org/1931903002

Make stencils be attachable to render targets created via SkSurface::MakeFromBackendTextureAsRenderTarget

This is a regression from "Refactor to separate backend object lifecycle
and GpuResource budget decision".

GrGLRenderTarget::CreateWrapped creates only render targets that wrap
the FBO.
GrGLRenderTargetTexture::CreateWrapped creates render targets that
wrap the texture.

Use the latter as the implementation for
SkSurface::MakeFromBackendTextureAsRenderTarget.

The test contains disabled code. The
MakeFromBackendTextureAsRenderTarget does not copy the existing texture
contents to the FBO render buffer.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1924183003

Review-Url: https://codereview.chromium.org/1924183003

Whitespace change to test bots

NOTREECHECKS=true
TBR=rmistry
BUG=skia:4763
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1936543002

Review-Url: https://codereview.chromium.org/1936543002

Adjust tolerance to fix bezier glitches issue

In a previous CL: https://codereview.chromium.org/1126793003, the value
of gDegenerateToLineTol was changed from 1.0 to 0.25 which caused the
bezier glitches.

This CL changes the value to SK_ScalarRoot2Over2, with this change the
bezier curve looks fine, and the result of the layout tests pointed out
in this bug: https://bugs.chromium.org/p/chromium/issues/detail?id=472847
is also reasonable.

BUG=605439
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1926063002

Review-Url: https://codereview.chromium.org/1926063002

Tighten up SkSpecialSurface factory functions

This is split out of https://codereview.chromium.org/1930013002/ ((Mostly) Retract GrRenderTarget from SkGpuDevice)

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1925313002

Review-Url: https://codereview.chromium.org/1925313002

Clean up test drawContext usage

The general idea is to provide access to SkGpuDevice's drawContext rather than its GrRenderTarget. That is usually what the testing framework actually wants.

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1930623003

Review-Url: https://codereview.chromium.org/1930623003

Remove skstd::is_convertible.

We should now be able to use std::is_convertible.

TBR=reed@google.com
This doesn't change any API, just where we get a type trait from.

Review-Url: https://codereview.chromium.org/1931003002

Update Win toolchain

I believe this now correctly updates us to VS 2015 update 2.

CQ_EXTRA_TRYBOTS=client.skia.compile:Build-Win-MSVC-x86-Debug-VS2015-Trybot

https://isolateserver.appspot.com/browse?namespace=default-gzip&digest=3c710fd7f4df8ec24fde62681536fad21be2751c

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1926813007

Review-Url: https://codereview.chromium.org/1926813007

Refactor drawContext/RenderTarget creation
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1914883002

Committed: https://skia.googlesource.com/skia/+/2f1c42e8448bbbadeb3df1c626faa90aa33f8907

Review-Url: https://codereview.chromium.org/1914883002

Delete unused code from SkColorSpace

IMO, a lot of this code reflects an oversimplified view of
handling color profiles.  And a lot of this code is valid
math/checks/enums that we will eventually need.

Right now, it's all unused and is adding a bit of clutter
and confusion as I try to redesign this object.  I think
it makes sense to start simpler and build this back up as
necessary.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1932743002

Review-Url: https://codereview.chromium.org/1932743002

Parse A2B0 tag on ICC profiles

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1925753002

Review-Url: https://codereview.chromium.org/1925753002

Stop printing warnings about ADB version

BUG=skia:5156
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1927973002

Review-Url: https://codereview.chromium.org/1927973002

remove (now unused) SkDivBits

needs https://codereview.chromium.org/1927823002/ to land first

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1925913002

Review-Url: https://codereview.chromium.org/1925913002

Detect empty (r)rects in GrShape.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1929643002

Review-Url: https://codereview.chromium.org/1929643002

Fix vulkan build on linux and android

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1921963003

Review-Url: https://codereview.chromium.org/1921963003

Revert of Repurpose Release_Developer BUILDTYPE and remove SK_DEVELOPER. (patchset #2 id:20001 of https://codereview.chromium.org/1927583002/ )

Reason for revert:
still blocking DEPS roll

Original issue's description:
> Repurpose Release_Developer BUILDTYPE and remove SK_DEVELOPER.
>
> The Release_Developer build type is now used to build Skia with
> release level optimizations but still enable SK_DEBUG. This in turn
> means that SkASSERTS and SkRTConf are available in this mode.
>
> Further we can then remove SK_DEVELOPER as a define as it is true
> iff SK_DEBUG is true.
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1927583002
> DOCS_PREVIEW= https://skia.org/?cl=1927583002
>
> Committed: https://skia.googlesource.com/skia/+/48d6f88479e72efaf9b9d8f9c0769acb1ba773db

TBR=bungeman@gmail.com,bungeman@google.com,benjaminwagner@google.com,djsollen@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review-Url: https://codereview.chromium.org/1918253006

Revert of fix build by using ifdef instead of if for SK_DEBUG (patchset #1 id:1 of https://codereview.chromium.org/1924943003/ )

Reason for revert:
need to revert main change for now

Original issue's description:
> fix build by using ifdef instead of if for SK_DEBUG
>
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1924943003
>
> TBR=
> NOTRY=True
>
> Committed: https://skia.googlesource.com/skia/+/b8498825b54718cdd90c30c39323cfc433695f23

TBR=djsollen@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review-Url: https://codereview.chromium.org/1927113002

Revert of Refactor drawContext/RenderTarget creation (patchset #8 id:140001 of https://codereview.chromium.org/1914883002/ )

Reason for revert:
Experimental revert to see if this is blocking the DEPS roll.

Original issue's description:
> Refactor drawContext/RenderTarget creation
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1914883002
>
> Committed: https://skia.googlesource.com/skia/+/2f1c42e8448bbbadeb3df1c626faa90aa33f8907

TBR=bsalomon@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review-Url: https://codereview.chromium.org/1929833004

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1924943003

TBR=
NOTRY=True

Review-Url: https://codereview.chromium.org/1924943003

remove SK_SUPPORT_LEGACY_DIVBITS_UB -- chrome no longer defines this

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1927823002

TBR=mtklein@google.com

Review-Url: https://codereview.chromium.org/1927823002

Enable flattening of SkRecordedDrawable

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1913843002

Committed: https://skia.googlesource.com/skia/+/d636950e8ecba89c0021ce4e76037c67aa6ee2cd

Review-Url: https://codereview.chromium.org/1913843002

skip unneeded header from writeFlattenable when building font-cache-key. Saves ~30% in size

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1924033002

TBR=mtklein

Review-Url: https://codereview.chromium.org/1924033002

adb_wait_for_charge: Don't download toolchains

BUG=skia:5156
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1926833003

Review-Url: https://codereview.chromium.org/1926833003

Make the default android build be release_developer.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1929663002

Review-Url: https://codereview.chromium.org/1929663002

Repurpose Release_Developer BUILDTYPE and remove SK_DEVELOPER.

The Release_Developer build type is now used to build Skia with
release level optimizations but still enable SK_DEBUG. This in turn
means that SkASSERTS and SkRTConf are available in this mode.

Further we can then remove SK_DEVELOPER as a define as it is true
iff SK_DEBUG is true.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1927583002
DOCS_PREVIEW= https://skia.org/?cl=1927583002

Review-Url: https://codereview.chromium.org/1927583002

Refactor drawContext/RenderTarget creation
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1914883002

Review-Url: https://codereview.chromium.org/1914883002

Add tests for volatile paths to GrShapeTest

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1923403003

Review-Url: https://codereview.chromium.org/1923403003

Add tests for empty GrShapes.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1927673002

Review-Url: https://codereview.chromium.org/1927673002

Simplify style handling in custom font manager.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1915103002

Review-Url: https://codereview.chromium.org/1915103002

SkDocument::CreateXPS fix roll

NOTRY=true
TBR=reed@google.com

Review-Url: https://codereview.chromium.org/1918713006

Remove AttachmentInfo from VulkanViewer setup

It's not necessary to pass this information back up to the client;
we can store it in the TestContext if needed for Surface creation.

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1920163004

Review-Url: https://codereview.chromium.org/1920163004

Add work around for vulkan copies on the 5X

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1923963002

Review-Url: https://codereview.chromium.org/1923963002

Image filters: de-nest SkImageFilter::Cache and Cache::Key.

This reduces the public header, and also allows us to foward-
declare them, so we can remove an #include from all the
SkDevice subclass headers.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1919063002

Review-Url: https://codereview.chromium.org/1919063002

SkPDF/None: fix compilation

TBR=stephana@google.com

Review-Url: https://codereview.chromium.org/1929593002

Bring sk_sp to GrDrawContext

This is split out of: https://codereview.chromium.org/1914883002/ (Refactor drawContext/RenderTarget creation)

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1918003003

Review-Url: https://codereview.chromium.org/1918003003

Add oblique as a slant.

Some fonts have italic and oblique in the same family, see

http://lucidafonts.com/fonts/family/lucida-sans
http://www.gust.org.pl/projects/e-foundry/latin-modern

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1921903002

Chromium side change at https://crrev.com/1921503006/ .

Review-Url: https://codereview.chromium.org/1921903002

Add dry_run_access_list to Skia's cq.proto

This will give all users who have access to our trybots access to our dry run CQ as well.

BUG=591183
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1926783002

Review-Url: https://codereview.chromium.org/1926783002

Revert of Enable flattening of SkRecordedDrawable (patchset #8 id:140001 of https://codereview.chromium.org/1913843002/ )

Reason for revert:
Release mode bots appear to be crashing while loading .skp files in nanobench.

Original issue's description:
> Enable flattening of SkRecordedDrawable
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1913843002
>
> Committed: https://skia.googlesource.com/skia/+/d636950e8ecba89c0021ce4e76037c67aa6ee2cd

TBR=djsollen@google.com,reed@google.com,msarett@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review-Url: https://codereview.chromium.org/1923393002

Enable flattening of SkRecordedDrawable

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1913843002

Review-Url: https://codereview.chromium.org/1913843002

Upgrade Android Gradle plugins
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1926743002

Review-Url: https://codereview.chromium.org/1926743002

remove 'deprecated' region from SkDraw

Most call-sites that used it just took its bounds, so it was trivial to convert them
to get the bounds of the RasterClip. Two clients wanted the actual region:
1. layeriter for android
2. pdf

Android already only has BW clips, so should be safe.
PDF now overrides its clip methods to ensure that all clips are BW.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1925693002

Review URL: https://codereview.chromium.org/1925693002

SkDocument/PDF: new API

has a pdf/a switch.
sets metadata in factory.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1916093002

Review URL: https://codereview.chromium.org/1916093002

tools/pdf-comparison.sh: stop relying on /usr/bin/timeout

NOTRY=true
TBR=

Review URL: https://codereview.chromium.org/1921943006

Treat kIncompleteInput as successes in get_images_from_skps

Context:
https://bugs.chromium.org/p/skia/issues/detail?id=5226
and
https://codereview.chromium.org/1844713003/diff/160001/tools/get_images_from_skps.cpp#newcode83

BUG=skia:5226
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1917413003

Review URL: https://codereview.chromium.org/1917413003

Disable ReadWriteAlpha on TSAN

BUG=skia:
R=borenet
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1927573002

Review URL: https://codereview.chromium.org/1927573002

Add support for building GrShape from SkPath and more tests
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1922713002

Review URL: https://codereview.chromium.org/1922713002

Add comments and sanity checks to parsing of ICC headers

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1922073003

Review URL: https://codereview.chromium.org/1922073003

Don't isolate the Skia repo for Swarming compiles

BUG=skia:
NOTREECHECKS=true
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1886343002

Review URL: https://codereview.chromium.org/1886343002

Change default window size on Windows.

The OS tends to pick a very large size if you let it. Instead, choose a
smaller (fixed) window (matching what we currently use on Linux).

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1919133004

Review URL: https://codereview.chromium.org/1919133004

Ensure we're operating on the correct texture when generating mipmaps.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1915403004

Review URL: https://codereview.chromium.org/1915403004

Revert of Remove gpu-specialization of SkSpecialImage (patchset #5 id:80001 of https://codereview.chromium.org/1918993002/ )

Reason for revert:
N6 bot has gone red

Original issue's description:
> Remove gpu-specialization of SkSpecialImage
>
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1918993002
>
> Committed: https://skia.googlesource.com/skia/+/0533d6e4142c2787d6d3c325ef0d8d1962292cad

TBR=reed@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1921393003

Remove unmatched restore call in gamma GM

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1917303002

Review URL: https://codereview.chromium.org/1917303002

Remove gpu-specialization of SkSpecialImage

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1918993002

Review URL: https://codereview.chromium.org/1918993002

Don't buffer stdout in DM

TBR=egdaniel@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1922223002

Review URL: https://codereview.chromium.org/1922223002

pathops: Split loop type cubics only when there is a self-intersection.

The ComplexBreak code comes from Loop and Blinn, which requires loops to be
split if either double point is visible.  However for intersection loops only
need to be split when there is actually a self-intersection (when both double
points are in [0, 1]).  This patch splits cubics much less often so the output
doesn't have extra segments unless the input had a self-intersecting loop.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1920663002

Review URL: https://codereview.chromium.org/1920663002

pathops: Cosmetic only change, update comments around ComplexBreak().

Rename loopT to splitT and make it clearer that it is not just
self intersecting loop cubics that are split.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1921173003

Review URL: https://codereview.chromium.org/1921173003

pathops: Add missing header guards in two files.

Guards were missing in SkIntersectionHelper.h and
SkPathOpsTSect.h.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1914873002

Review URL: https://codereview.chromium.org/1914873002

Update SK_IMAGE_VERSION, run nanobench on its own image set

I've added a new directory to the cloud
(gs://chromium-skia-gm/skimage/v4/nanobench) which will
allow us to run dm and nanobench on different sets of
images.

The dm image set will stay the same, and includes
incomplete images, edge cases images, etc.

The nanobench set is much smaller, and includes jpegs and
pngs that we've found useful to bench in the past, plus
a couple of each for the other formats.  We can refine
this set as we move forward (and may also want to monitor
these images on perf.skia.org).

For now the goal is to reduce nanobench runtime by not
testing on many, many obscure images that are not
particularly interesting to time.

The actual change to use nanobench images is in:
https://codereview.chromium.org/1917903003/

BUG=skia:5181
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1913813004

Review URL: https://codereview.chromium.org/1913813004

Remove redundant tests to make DM a little faster

This should help with some of the slow down caused by
https://codereview.chromium.org/1907593004.

BUG=skia:5181
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1923533002

Review URL: https://codereview.chromium.org/1923533002

Remove redundant test

FlattenDrawable test now exercises the use of custom unflatten procs.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1917183002

Review URL: https://codereview.chromium.org/1917183002

Add initial implementation of GrShape and GrStyle classes and tests

The initial intent is to use GrShape to simplify the mask blur code paths. However, I also want to use this to explore a more unified drawing code flow for different geometry types. The goal is to have a single representation for geometries+styling that attempts to always keep the geometry in the simplest form (e.g. preferring rrects to paths). It also allows for converting styling information into modified geometry and for computing consistent keys.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1822723003

Committed: https://skia.googlesource.com/skia/+/c885dacfe4625af8b0e2e5c6e8a8ae8dc2d620a8

Committed: https://skia.googlesource.com/skia/+/824e58fc6df8fc149d9675f577f7deeaba698b09

Review URL: https://codereview.chromium.org/1822723003

Add runhooks to multi_repo_trybots

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1924573002
NOTRY=true
DOCS_PREVIEW= https://skia.org/?cl=1924573002

Review URL: https://codereview.chromium.org/1924573002

pdf comparison tool

NOTRY=true

Review URL: https://codereview.chromium.org/1924513002

Update android docs for how to sync source code
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1921023002
NOTRY=true
DOCS_PREVIEW= https://skia.org/?cl=1921023002

Review URL: https://codereview.chromium.org/1921023002

bin/fetch-skps: print instruction on error

NOTRY=true

Review URL: https://codereview.chromium.org/1919133003

Whitespace change to trigger bots after recipe fix

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1922783003
NOTREECHECKS=true

Review URL: https://codereview.chromium.org/1922783003