Revert of df generation: single allocation with calloc (patchset #2 id:20001 of https://codereview.chromium.org/1544983004/ )

Reason for revert:
Crashing multiple bots, e.g.
https://uberchromegw.corp.google.com/i/client.skia.android/builders/Perf-Android-GCC-Nexus7-GPU-Tegra3-Arm7-Release/builds/3525

Original issue's description:
> df generation: single allocation with calloc
>
> The dfStorage DFData allocation can never fit in its stack space: 5px padding on each side always implies at least a 10x10 DFData allocation, but the stack space only fits 64 DFData.
>
> So we've always been spilling to the heap.
>
> If we're going to spill to the heap, we might as well allocate/free all our temporary memory in one block, and since we want it zeroed, might as well calloc.
>
> So in practice this replaces 1-2 malloc, 1-2 free, and 2 bzeros with 1 calloc and 1 free.
>
>
> BUG=skia:4729
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1544983004
>
> Committed: https://skia.googlesource.com/skia/+/12204d90337656542a42fa0fcccb7bec13af0cce

TBR=jvanverth@google.com,mtklein@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:4729

Review URL: https://codereview.chromium.org/1556263002

Two malloc+bzero -> calloc.

I was profiling DM and noticed a couple spots where we malloc then bzero.
These might as well call calloc instead:
   - any time DM itself allocates bitmaps for raster drawing;
   - any time Skia allocates memory for a raster SkSurface.

We could use malloc for opaque surfaces, but it seems simpler to always calloc.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1557713003

Review URL: https://codereview.chromium.org/1557713003

remove use of malloc.h

It's non-portable, commented out, and those asserts are not important.
Let's just blow it away.

BUG=skia:4736
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1551673002

Review URL: https://codereview.chromium.org/1551673002

df generation: single allocation with calloc

The dfStorage DFData allocation can never fit in its stack space: 5px padding on each side always implies at least a 10x10 DFData allocation, but the stack space only fits 64 DFData.

So we've always been spilling to the heap.

If we're going to spill to the heap, we might as well allocate/free all our temporary memory in one block, and since we want it zeroed, might as well calloc.

So in practice this replaces 1-2 malloc, 1-2 free, and 2 bzeros with 1 calloc and 1 free.

BUG=skia:4729
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1544983004

Review URL: https://codereview.chromium.org/1544983004

Revert of Try using std::call_once (patchset #1 id:1 of https://codereview.chromium.org/1550893002/ )

Reason for revert:
Can't use on XP.  :(

Original issue's description:
> Try using std::call_once
>
> Now that we've got std library support, perhaps we should start using it.
> This CL acts as a little canary, and may help fix the linked bug.
>
> I'm not really sure what's going on in the linked bug, but using
> std::call_once over homegrown atomics has to be the right answer...
>
> BUG=chromium:418041
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1550893002
> CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot
>
> Going to land this ahead of review while the tree is quiet to see how it rolls.
> TBR=herb@google.com
>
> Committed: https://skia.googlesource.com/skia/+/8895b72f789e5dc8bb99cb9727875439005fc919

TBR=herb@google.com,mtklein@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=chromium:418041

Review URL: https://codereview.chromium.org/1552333003

handle halfway case in scan converter

Scan edges that start at exactly -0.5 aren't trimmed by
clipping or by rounding, triggering a debug assert.

One way to fix this is to round the top and left
down instead of up.

Also, move the path initialization of gm/composeshader.cpp
to make debugging other path problems easier.

R=reed@google.com
BUG=skia:2715
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1544873002

Review URL: https://codereview.chromium.org/1544873002

remove now defunct get_paint_offset

This used to be used when we wanted to know where the paint was, so we could perform peep-hole optimizations.

We no longer do that on this (older) picture format, so now this function only serves to clog up the file with asserts.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1557123002

Review URL: https://codereview.chromium.org/1557123002

Android bots: set battery threshold to 80%

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1559913003

Review URL: https://codereview.chromium.org/1559913003

On Valgrind bots, print a message every 20 minutes

I believe that the timeout failure on the Valgrind bot occurs
because we skip many tests consecutively without printing any
output.  Skipping these tests requires non-trivial work.

Printing a message every 20 minutes will avoid timeouts.

BUG=skia:4740
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1554193002

Review URL: https://codereview.chromium.org/1554193002

Add unit test for crbug.com/570232

This is the promised followup to https://codereview.chromium.org/1539553002/ (Add default ctor to SkMask) and would've caught the bug before it got out into the wild.

BUG=570232
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1552303002

Review URL: https://codereview.chromium.org/1552303002

Android devices: wait for 100% charge

Some devices are getting pretty low by the end of the build (3%).  No reason to cut it close, bump to 100% charge.  This will cause some devices to sit for a while as they charge up.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1556023002

Review URL: https://codereview.chromium.org/1556023002

[Reland] Repeating SkBitmapProcState rounding bias

1) observe the bias in repeat matrix procs also.
2) add utility for device space -> bitmap space mapping.
3) remove unneeded filter bias

This is a reland of https://codereview.chromium.org/1529833003/. The
main difference (and the fix) vs. the prev version is increased
precision: the mapper now operates with SkFractionalInts.

R=reed@google.com
BUG=skia:4680, skia:4649
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1539083002

Review URL: https://codereview.chromium.org/1539083002

remove 565 effects shaders

This removes the 565 optimization for radial,
linear, sweep, and picture shaders. The
optimization gets little real-world use and
has a reasonable fallback. This also permits
removing the square root table for the radial
gradient case.

R=reed@google.com
BUG=skia:1947
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1553103003

Review URL: https://codereview.chromium.org/1553103003

Move path fallback for AA rect out of SkGpuDevice

The AA rect batches currently in Skia have view matrix restrictions. If
a view matrix is not supported, we have to draw the rect as a path
instead. This changes moves the decision to draw as a path out of
SkGpuDevice and into GrDrawContext.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1551883003

Review URL: https://codereview.chromium.org/1551883003

Don't preserve buggy Android behavior of ignoring the paint's alpha.

BUG=skia:4296
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1532573002

Review URL: https://codereview.chromium.org/1532573002

Change use of LoadLibrary to LoadLibraryA

This allows Skia to be built with UNICODE defined.

BUG=skia:4735
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1552203003

Review URL: https://codereview.chromium.org/1552203003

[SkDebugger] Display run glyph positioning modes

R=robertphillips@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1559883002

Review URL: https://codereview.chromium.org/1559883002

Add conversion from SaveLayerFlags to SaveFlags

Needed for SkiaAndroidProxy::getSaveLayerStrategy.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1559873003

Review URL: https://codereview.chromium.org/1559873003

move SkDither.h to private, remove unused shadeSpan16

Since perlin noise shader never set the kHasShad16 flag, its shadeSpan16 method was never called.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1558933002

TBR=

Review URL: https://codereview.chromium.org/1558933002

remove forth experiment

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1559823002

TBR=

Review URL: https://codereview.chromium.org/1559823002

move SkTDict and SkTSearch into private

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1558923002

TBR=

Review URL: https://codereview.chromium.org/1558923002

Fix -Wunused-function warnings in chrome/ios builds.

BUG=chromium:573779
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1559783002

Review URL: https://codereview.chromium.org/1559783002

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=1551273002

Review URL: https://codereview.chromium.org/1551273002

replace older/slower/scaler version of conic evalAt

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1552863005

TBR=

Review URL: https://codereview.chromium.org/1552863005

change factory to method, to parallel pattern in imagefilters. no functionality change.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1553743002

TBR=

Review URL: https://codereview.chromium.org/1553743002

Use device-independent text for nvprmsaa4/16 in nanobench and dm

Use device-independent text rendering in nanobench and dm when
using configs nvprmsaa4 or nvprmsaa16.

This was turned on for dm in commit:
"Re-enable nvpr text rendering in DM"
https://codereview.chromium.org/1306733007

This regressed in commit:
"Add config options to run different GPU APIs to dm and nanobench"
https://codereview.chromium.org/1490113005/

BUG=skia:2992
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1553453004

Review URL: https://codereview.chromium.org/1553453004

Try using std::call_once

Now that we've got std library support, perhaps we should start using it.
This CL acts as a little canary, and may help fix the linked bug.

I'm not really sure what's going on in the linked bug, but using
std::call_once over homegrown atomics has to be the right answer...

BUG=chromium:418041
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1550893002
CQ_EXTRA_TRYBOTS=client.skia:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD-Trybot

Going to land this ahead of review while the tree is quiet to see how it rolls.
TBR=herb@google.com

Review URL: https://codereview.chromium.org/1550893002

initialize offscreen in StrokeZeroGM

Valgrind's complaining that we're drawing uninitialized source pixels
from that offscreen via drawBitmap():
https://uberchromegw.corp.google.com/i/client.skia/builders/Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-Valgrind/builds/736/steps/dm/logs/stdio

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1546363002

Review URL: https://codereview.chromium.org/1546363002

Disable ES2 on Linux SDL

TBR=borenet@google.com
BUG=skia:4737
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1550123002

Review URL: https://codereview.chromium.org/1550123002

Reland of Make draw* lines responsible for cache management (patchset #1 id:1 of https://codereview.chromium.org/1552513003/ )

Reason for revert:
might have been premature

Original issue's description:
> Revert of Make draw* lines responsible for cache management (patchset #1 id:1 of https://codereview.chromium.org/1521613002/ )
>
> Reason for revert:
> seeing crashes on android and linux
>
> Original issue's description:
> > Make draw* lines responsible for cache management
> >
> > TBR=bsalomon@google.com
> > BUG=skia:
> >
> > Committed: https://skia.googlesource.com/skia/+/66957946c0c418e71c0f36bb2904cc38c334eba4
>
> TBR=bsalomon@google.com,jvanverth@google.com,joshualitt@chromium.org
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/90e293dae4990a86cb952663ccde8fae55112d7d

TBR=bsalomon@google.com,jvanverth@google.com,joshualitt@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/1551563003

Revert of Make draw* lines responsible for cache management (patchset #1 id:1 of https://codereview.chromium.org/1521613002/ )

Reason for revert:
seeing crashes on android and linux

Original issue's description:
> Make draw* lines responsible for cache management
>
> TBR=bsalomon@google.com
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/66957946c0c418e71c0f36bb2904cc38c334eba4

TBR=bsalomon@google.com,jvanverth@google.com,joshualitt@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/1552513003

remove cruft from SkTypes.h, including SkBool

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1548223002

TBR=

Review URL: https://codereview.chromium.org/1548223002

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=1550663002

Review URL: https://codereview.chromium.org/1550663002

remove deprecated annotation helper

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1552443003

TBR=

Review URL: https://codereview.chromium.org/1552443003

make GrRect.h private

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1548213002

TBR=bsalomon

Review URL: https://codereview.chromium.org/1548213002

remove unused SkAutoROCanvasPixels

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1550583002

TBR=

Review URL: https://codereview.chromium.org/1550583002

remove dead flags

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1543323002

TBR=

Review URL: https://codereview.chromium.org/1543323002

remove unused SkCubicInterval

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1533393005

TBR=scroggo

Review URL: https://codereview.chromium.org/1533393005

Add config options to run different GPU APIs to dm and nanobench

Add extended config specification form that can be used to run different
gpu backend with different APIs.

The configs can be specified with the form:
gpu(api=string,dit=bool,nvpr=bool,samples=int)

This replaces and removes the --gpuAPI flag.

All existing configs should still work.

Adds following documentation:

out/Debug/dm --help config

Flags:
    --config: type: string default: 565 8888 gpu nonrendering
        Options: 565 8888 debug gpu gpudebug gpudft gpunull msaa16 msaa4
        nonrendering null nullgpu nvprmsaa16 nvprmsaa4 pdf pdf_poppler skp svg
        xps or use extended form 'backend(option=value,...)'.

        Extended form: 'backend(option=value,...)'

        Possible backends and options:

        gpu(api=string,dit=bool,nvpr=bool,samples=int) GPU backend
         api type: string default: native.
             Select graphics API to use with gpu backend.
             Options:
         native Use platform default OpenGL or OpenGL ES backend.
         gl     Use OpenGL.
         gles   Use OpenGL ES.
         debug  Use debug OpenGL.
         null   Use null OpenGL.
         dit type: bool default: false.
             Use device independent text.
         nvpr type: bool default: false.
             Use NV_path_rendering OpenGL and OpenGL ES extension.
         samples type: int default: 0.
             Use multisampling with N samples.

        Predefined configs:

         gpu       = gpu()
         msaa4     = gpu(samples=4)
         msaa16    = gpu(samples=16)
         nvprmsaa4 = gpu(nvpr=true,samples=4)
         nvprmsaa16 = gpu(nvpr=true,samples=16)
         gpudft     = gpu(dit=true)
         gpudebug   = gpu(api=debug)
         gpunull    = gpu(api=null)
         debug      = gpu(api=debug)
         nullgpu    = gpu(api=null)

BUG=skia:2992

Committed: https://skia.googlesource.com/skia/+/e13ca329fca4c28cf4e078561f591ab27b743d23
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1490113005

Committed: https://skia.googlesource.com/skia/+/c8b4336444e7b90382e04e33665fb3b8490b825b

Committed: https://skia.googlesource.com/skia/+/9ebc3f0ee6db215dde461dc4777d85988cf272dd

Review URL: https://codereview.chromium.org/1490113005

Append '\n' character to nanobench error message

This prevents the error message from inserting itself on a line before
a bench result, which can mess up parsing.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1540923003

Review URL: https://codereview.chromium.org/1540923003

Revert of Add config options to run different GPU APIs to dm and nanobench (patchset #21 id:400001 of https://codereview.chromium.org/1490113005/ )

Reason for revert:
The Test-Win8-MSVC-ShuttleB-GPU-HD4600-x86_64-Debug builder fails after this CL.

Links to specific builds:
http://build.chromium.org/p/client.skia/builders/Test-Win8-MSVC-ShuttleB-GPU-HD4600-x86_64-Debug/builds/1689
http://build.chromium.org/p/client.skia/builders/Test-Win8-MSVC-ShuttleB-GPU-HD4600-x86_64-Debug/builds/1690
http://build.chromium.org/p/client.skia/builders/Test-Win8-MSVC-ShuttleB-GPU-HD4600-x86_64-Debug/builds/1691

Original issue's description:
> Add config options to run different GPU APIs to dm and nanobench
>
> Add extended config specification form that can be used to run different
> gpu backend with different APIs.
>
> The configs can be specified with the form:
> gpu(api=string,dit=bool,nvpr=bool,samples=int)
>
> This replaces and removes the --gpuAPI flag.
>
> All existing configs should still work.
>
> Adds following documentation:
>
> out/Debug/dm --help config
>
> Flags:
>     --config: type: string default: 565 8888 gpu nonrendering
>         Options: 565 8888 debug gpu gpudebug gpudft gpunull msaa16 msaa4
>         nonrendering null nullgpu nvprmsaa16 nvprmsaa4 pdf pdf_poppler skp svg
>         xps or use extended form 'backend(option=value,...)'.
>
>         Extended form: 'backend(option=value,...)'
>
>         Possible backends and options:
>
>         gpu(api=string,dit=bool,nvpr=bool,samples=int) GPU backend
>          api type: string default: native.
>              Select graphics API to use with gpu backend.
>              Options:
>          native Use platform default OpenGL or OpenGL ES backend.
>          gl     Use OpenGL.
>          gles   Use OpenGL ES.
>          debug  Use debug OpenGL.
>          null   Use null OpenGL.
>          dit type: bool default: false.
>              Use device independent text.
>          nvpr type: bool default: false.
>              Use NV_path_rendering OpenGL and OpenGL ES extension.
>          samples type: int default: 0.
>              Use multisampling with N samples.
>
>         Predefined configs:
>
>          gpu       = gpu()
>          msaa4     = gpu(samples=4)
>          msaa16    = gpu(samples=16)
>          nvprmsaa4 = gpu(nvpr=true,samples=4)
>          nvprmsaa16 = gpu(nvpr=true,samples=16)
>          gpudft     = gpu(dit=true)
>          gpudebug   = gpu(api=debug)
>          gpunull    = gpu(api=null)
>          debug      = gpu(api=debug)
>          nullgpu    = gpu(api=null)
>
> BUG=skia:2992
>
> Committed: https://skia.googlesource.com/skia/+/e13ca329fca4c28cf4e078561f591ab27b743d23
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1490113005
>
> Committed: https://skia.googlesource.com/skia/+/c8b4336444e7b90382e04e33665fb3b8490b825b
>
> Committed: https://skia.googlesource.com/skia/+/9ebc3f0ee6db215dde461dc4777d85988cf272dd

TBR=mtklein@google.com,bsalomon@google.com,joshualitt@google.com,scroggo@google.com,kkinnunen@nvidia.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:2992

Review URL: https://codereview.chromium.org/1548683002

revised strokerect gm

In service of the bug below, revise the gm to show what zero
widths draw, and what they should draw (by including rects
with fractionally tiny widths).

TBR=halcanary@google.com,reed@google.com
BUG=1294
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1548643002

Review URL: https://codereview.chromium.org/1548643002

Documentation: FAQ TOC

NOTRY=true
TBR=

Review URL: https://codereview.chromium.org/1542853003

fix hair fuzz

If the end and control points of a quad, conic, or cubic are the same,
adjust all of them when stretching the curve to account for a square
or round end cap. If all of the points are the same, move all but the
last.

Enlarge the clip check to account for the cap.

The clip bug was detected by ASAN.

R=reed@google.com, msarett@google.com
BUG=571214
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1547483003

Review URL: https://codereview.chromium.org/1547483003

move SkParsePaint.h into views

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1547653002
TBR=

Review URL: https://codereview.chromium.org/1547653002

Add config options to run different GPU APIs to dm and nanobench

Add extended config specification form that can be used to run different
gpu backend with different APIs.

The configs can be specified with the form:
gpu(api=string,dit=bool,nvpr=bool,samples=int)

This replaces and removes the --gpuAPI flag.

All existing configs should still work.

Adds following documentation:

out/Debug/dm --help config

Flags:
    --config: type: string default: 565 8888 gpu nonrendering
        Options: 565 8888 debug gpu gpudebug gpudft gpunull msaa16 msaa4
        nonrendering null nullgpu nvprmsaa16 nvprmsaa4 pdf pdf_poppler skp svg
        xps or use extended form 'backend(option=value,...)'.

        Extended form: 'backend(option=value,...)'

        Possible backends and options:

        gpu(api=string,dit=bool,nvpr=bool,samples=int) GPU backend
         api type: string default: native.
             Select graphics API to use with gpu backend.
             Options:
         native Use platform default OpenGL or OpenGL ES backend.
         gl     Use OpenGL.
         gles   Use OpenGL ES.
         debug  Use debug OpenGL.
         null   Use null OpenGL.
         dit type: bool default: false.
             Use device independent text.
         nvpr type: bool default: false.
             Use NV_path_rendering OpenGL and OpenGL ES extension.
         samples type: int default: 0.
             Use multisampling with N samples.

        Predefined configs:

         gpu       = gpu()
         msaa4     = gpu(samples=4)
         msaa16    = gpu(samples=16)
         nvprmsaa4 = gpu(nvpr=true,samples=4)
         nvprmsaa16 = gpu(nvpr=true,samples=16)
         gpudft     = gpu(dit=true)
         gpudebug   = gpu(api=debug)
         gpunull    = gpu(api=null)
         debug      = gpu(api=debug)
         nullgpu    = gpu(api=null)

BUG=skia:2992

Committed: https://skia.googlesource.com/skia/+/e13ca329fca4c28cf4e078561f591ab27b743d23
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1490113005

Committed: https://skia.googlesource.com/skia/+/c8b4336444e7b90382e04e33665fb3b8490b825b

Review URL: https://codereview.chromium.org/1490113005

remove unused SkCullPoints

need to remove file references from chrome before landing this
https://codereview.chromium.org/1547603002#

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1541933004

TBR=scroggo

Review URL: https://codereview.chromium.org/1541933004

Fold color attachment verification bit into GrGLCaps::ConfigInfo
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1536033003

Review URL: https://codereview.chromium.org/1536033003

Don't ref/unref the static src-over xp
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1540363002

bug=chromium:570301

Review URL: https://codereview.chromium.org/1540363002

hide old SaveFlags, but keep them available (for now) for Android

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1535993003

requires https://codereview.chromium.org/1537203002/# to land in chrome first (android should be safe)

Review URL: https://codereview.chromium.org/1535993003

Move config table to GrGLCaps from GrGLGpu.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1535153002

Review URL: https://codereview.chromium.org/1535153002

Reland of change all factories to return their base-class (patchset #1 id:1 of https://codereview.chromium.org/1540203002/ )

Reason for revert:
chrome changes have landed

Original issue's description:
> Revert of change all factories to return their base-class (patchset #1 id:1 of https://codereview.chromium.org/1535353002/ )
>
> Reason for revert:
> need to update some chrome/blink call-sites
>
> Original issue's description:
> > change all factories to return their base-class
> >
> > will watch DEPS roll to see if there are chrome sites needing updates
> >
> > BUG=skia:
> > GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1535353002
> >
> > TBR=
> >
> > Committed: https://skia.googlesource.com/skia/+/d63f60a36327e9580861205ebb35cade8c49bd34
>
> TBR=
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=skia:
>
> Committed: https://skia.googlesource.com/skia/+/2d6ba6690f8951e152d8e793191b14afd52f5506

TBR=
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/1533373002

fix stroked text underline / strikethrough

Pass 1 to DrawRect for underline and strikethrough since it will
scale by the text size later.

R=reed@google.com
BUG=skia:971
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1535793004

Review URL: https://codereview.chromium.org/1535793004

Create a hash table from id<-->key in SkImageFilter::CacheImpl

There is memory leak in the SkImageFilter::Cache. There are two sources
of memory leak:
1. The cache filling up quickly.
2. A slow small leak that never stops.
This CL solves the first issue, which prevents the cache filling up quickly.
This CL creates a new hash table that index the
SkImageFilter::uniqueID to an array of keys, and with the existing
key<-->Value hash table, we can have SkImageFilters proactively
purge content derived cached content when destroyed.

BUG=489543
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1514893003

Committed: https://skia.googlesource.com/skia/+/f5d1f8dcc841516d7ea63c151b13059af40ca76d

Review URL: https://codereview.chromium.org/1514893003

check in direct quad length measure

Add code so that it at minimum won't bit-rot.
Next: add tests to see if it works.

R=reed@google.com
BUG=skia:1036
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1541523002

Review URL: https://codereview.chromium.org/1541523002

Enable NVPR with command buffer

Set Chromium command-line flag --enable-gl-path-rendering
before creating command buffer context. This is needed for
the command buffer to expose NVPR at the moment.

BUG=skia:2992
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1538863002

Review URL: https://codereview.chromium.org/1538863002

Revert of change all factories to return their base-class (patchset #1 id:1 of https://codereview.chromium.org/1535353002/ )

Reason for revert:
need to update some chrome/blink call-sites

Original issue's description:
> change all factories to return their base-class
>
> will watch DEPS roll to see if there are chrome sites needing updates
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1535353002
>
> TBR=
>
> Committed: https://skia.googlesource.com/skia/+/d63f60a36327e9580861205ebb35cade8c49bd34

TBR=
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/1540203002

change all factories to return their base-class

will watch DEPS roll to see if there are chrome sites needing updates

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1535353002

TBR=

Review URL: https://codereview.chromium.org/1535353002

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=1535373002

Review URL: https://codereview.chromium.org/1535373002

remove SK_SUPPORT_LEGACY_SAVELAYERPARAMS code

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1537193002

TBR=

Review URL: https://codereview.chromium.org/1537193002

Remove EGL_config from vb bot

TBR=borenet@google.com

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1540583003

Review URL: https://codereview.chromium.org/1540583003

Rejigger distance field path rect calcs to avoid underflow

BUG=skia:4716
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1535173002

Review URL: https://codereview.chromium.org/1535173002

Add Drawing Manager guards against re-entrant flushes

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1538013002

Review URL: https://codereview.chromium.org/1538013002

Fix UB function problems for shaders and mask.

BUG=skia:4634
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1530743002

Committed: https://skia.googlesource.com/skia/+/dafd044e4aea529f88d899247678d4549f776388

Review URL: https://codereview.chromium.org/1530743002

Revert of Repeating SkBitmapProcState rounding bias (patchset #6 id:100001 of https://codereview.chromium.org/1529833003/ )

Reason for revert:
We need a SkFractionalInt auto mapper flavor, otherwise we're losing precision for some procs => seaming artifacs

(https://gold.skia.org/diff?test=giantbitmap_mirror_point_rotate&left=0dd7a412fce81586e6a142e039ec2de9&top=a706d4bb2b8e994b936445b87115b07e)

Original issue's description:
> Repeating SkBitmapProcState rounding bias
>
> Observe the bias in repeat matrix procs also.
>
> Introduce a utility class to handle device space -> bitmap space
> mapping.
>
> BUG=skia:4680,skia:4649
> R=reed@google.com
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1529833003
>
> Committed: https://skia.googlesource.com/skia/+/5ae7fdcc3d7712da3193c39a751e88b092aa82db

TBR=reed@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:4680,skia:4649

Review URL: https://codereview.chromium.org/1531423003

Move SkAndroidSDKCanvas to tools and ensure that it is built on all Android builds
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1536013003

Review URL: https://codereview.chromium.org/1536013003

Revert of Fix UB function problems for shaders and mask. (patchset #3 id:40001 of https://codereview.chromium.org/1530743002/ )

Reason for revert:
Breaks man gold tests.

Original issue's description:
> Fix UB function problems for shaders and mask.
>
> BUG=skia:4634
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1530743002
>
> Committed: https://skia.googlesource.com/skia/+/dafd044e4aea529f88d899247678d4549f776388

TBR=mtklein@google.com,reed@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:4634

Review URL: https://codereview.chromium.org/1534243002

Add visualbench as valid target to Makefile

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1530233005

Committed: https://skia.googlesource.com/skia/+/13221b83663778a2f5cbd90e51e253268755877e

Review URL: https://codereview.chromium.org/1530233005

Repeating SkBitmapProcState rounding bias

Observe the bias in repeat matrix procs also.

Introduce a utility class to handle device space -> bitmap space
mapping.

BUG=skia:4680,skia:4649
R=reed@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1529833003

Review URL: https://codereview.chromium.org/1529833003

Remove GrGLGpu::configToGLFormats
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1533903004

Review URL: https://codereview.chromium.org/1533903004

update androidsdk to new savelayer signature

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1540583002

NOTRY=True
NOTREECHECKS=True
TBR=

Review URL: https://codereview.chromium.org/1540583002

Revert of Add visualbench as valid target to Makefile (patchset #2 id:20001 of https://codereview.chromium.org/1530233005/ )

Reason for revert:
might be breaking cmakebot

Original issue's description:
> Add visualbench as valid target to Makefile
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1530233005
>
> Committed: https://skia.googlesource.com/skia/+/13221b83663778a2f5cbd90e51e253268755877e

TBR=borenet@google.com,joshualitt@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:

Review URL: https://codereview.chromium.org/1537903003

Reland of Add test for previously unflattenables (patchset #1 id:1 of https://codereview.chromium.org/1532753002/ )

Reason for revert:
The test should now be blacklisted.

Original issue's description:
> Revert of Add test for previously unflattenables (patchset #1 id:1 of https://codereview.chromium.org/1514373003/ )
>
> Reason for revert:
> Speculative fix for skbug.com/4709
>
> Original issue's description:
> > Add test for previously unflattenables
> >
> > BUG=skia:4613
> >
> > Committed: https://skia.googlesource.com/skia/+/061aaa79f7d8a2e93962e8296abaae13f0a7a715
>
> TBR=halcanary@google.com
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=skia:4613
>
> Committed: https://skia.googlesource.com/skia/+/c8f969309cafebeb16ad057f766b61bdc406a8b8

TBR=halcanary@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:4613

Review URL: https://codereview.chromium.org/1535143002

Add visualbench as valid target to Makefile

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1530233005

Review URL: https://codereview.chromium.org/1530233005

change signature for virtual related to saveLayer, passing SaveLayerRec

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1533953002

NOTREECHECKS=True

Review URL: https://codereview.chromium.org/1533953002

fix large dashed paths

Paths with lots of points exceed the 32767 point index.
Make the index larger, add a test, and allow the
t value to use the extra bits.

R=reed@google.com,fmalita@chromium.org
BUG=skia:3501
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1534223002

Review URL: https://codereview.chromium.org/1534223002

Remove SkImageMinRowBytes

As stated in the comments in crrev.com/1379193002, this method name is
misleading. It returns a larger number than the minimum rowbytes - it
increases to the next four-byte alignment. This has the effect that the
one place that calls it does not support 565 (which is not already
four-byte aligned), but does not serve any other purpose. Remove it and
the only call-site.

BUG=skia:4396
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1528383004

Review URL: https://codereview.chromium.org/1528383004

Blacklist GLPrograms on Angle Debug bot

BUG=skia:4717
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1527323003

Review URL: https://codereview.chromium.org/1527323003

Update dm_flags.json to match dm_flags.py

dm_flags.json verifies that the changes to dm_flags.py made sense. It
was not updated in crrev.com/1537793002, making me confused why my
.py changes seemed to affect WritePixelsTest in .json.

TBR=jvanverth@google.com
BUG=skia:2733
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1534043003

Review URL: https://codereview.chromium.org/1534043003

Revert of Add config options to run different GPU APIs to dm and nanobench (patchset #18 id:340001 of https://codereview.chromium.org/1490113005/ )

Reason for revert:
This CL changed 1200 images on gold, when I wouldn't expect any diffs from the description.

Original issue's description:
> Add config options to run different GPU APIs to dm and nanobench
>
> Add extended config specification form that can be used to run different
> gpu backend with different APIs.
>
> The configs can be specified with the form:
> gpu(api=string,dit=bool,nvpr=bool,samples=int)
>
> This replaces and removes the --gpuAPI flag.
>
> All existing configs should still work.
>
> Adds following documentation:
>
> out/Debug/dm --help config
>
> Flags:
>     --config: type: string default: 565 8888 gpu nonrendering
>         Options: 565 8888 debug gpu gpudebug gpudft gpunull msaa16 msaa4
>         nonrendering null nullgpu nvprmsaa16 nvprmsaa4 pdf pdf_poppler skp svg
>         xps or use extended form 'backend(option=value,...)'.
>
>         Extended form: 'backend(option=value,...)'
>
>         Possible backends and options:
>
>         gpu(api=string,dit=bool,nvpr=bool,samples=int) GPU backend
>          api type: string default: native.
>              Select graphics API to use with gpu backend.
>              Options:
>          native Use platform default OpenGL or OpenGL ES backend.
>          gl     Use OpenGL.
>          gles   Use OpenGL ES.
>          debug  Use debug OpenGL.
>          null   Use null OpenGL.
>          dit type: bool default: false.
>              Use device independent text.
>          nvpr type: bool default: false.
>              Use NV_path_rendering OpenGL and OpenGL ES extension.
>          samples type: int default: 0.
>              Use multisampling with N samples.
>
>         Predefined configs:
>
>          gpu       = gpu()
>          msaa4     = gpu(samples=4)
>          msaa16    = gpu(samples=16)
>          nvprmsaa4 = gpu(nvpr=true,samples=4)
>          nvprmsaa16 = gpu(nvpr=true,samples=16)
>          gpudft     = gpu(dit=true)
>          gpudebug   = gpu(api=debug)
>          gpunull    = gpu(api=null)
>          debug      = gpu(api=debug)
>          nullgpu    = gpu(api=null)
>
> BUG=skia:2992
>
> Committed: https://skia.googlesource.com/skia/+/e13ca329fca4c28cf4e078561f591ab27b743d23
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1490113005
>
> Committed: https://skia.googlesource.com/skia/+/c8b4336444e7b90382e04e33665fb3b8490b825b

TBR=mtklein@google.com,bsalomon@google.com,scroggo@google.com,kkinnunen@nvidia.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:2992

Review URL: https://codereview.chromium.org/1536963002

Revert of Create a hash table from id<-->key in SkImageFilter::CacheImpl (patchset #7 id:120001 of https://codereview.chromium.org/1514893003/ )

Reason for revert:
speculative revert to try to unblock DEPS roll

https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_asan_rel_ng/builds/94665

Original issue's description:
> Create a hash table from id<-->key in SkImageFilter::CacheImpl
>
> There is memory leak in the SkImageFilter::Cache. There are two sources
> of memory leak:
> 1. The cache filling up quickly.
> 2. A slow small leak that never stops.
> This CL solves the first issue, which prevents the cache filling up quickly.
> This CL creates a new hash table that index the
> SkImageFilter::uniqueID to an array of keys, and with the existing
> key<-->Value hash table, we can have SkImageFilters proactively
> purge content derived cached content when destroyed.
>
> BUG=489543
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1514893003
>
> Committed: https://skia.googlesource.com/skia/+/f5d1f8dcc841516d7ea63c151b13059af40ca76d

TBR=mtklein@google.com,junov@chromium.org,xidachen@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=489543

Review URL: https://codereview.chromium.org/1537923002

fix bugs in path contains

Pull out the logic to check to see if the point is on the edge
so all curve types can share.

Reorder cubic to be like conic and quad so that mixed types
consider the curves consistently.

Don't count on curve points twice if they are on the end
and compute a zero cross product.

Remove logic that checks, when there are no roots, if the
point is closer to the top or the bottom (it's always the top).

Initialize the iterator correctly when it is accessing
the list of on point curves.

Use 'multiply' instead of 'subtract' to see if the vectors
are pointing in opposite directions.

Add more test cases.

R=reed@google.com,fs@opera.com
BUG=skia:4265
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1532003004

Review URL: https://codereview.chromium.org/1532003004

Add config options to run different GPU APIs to dm and nanobench

Add extended config specification form that can be used to run different
gpu backend with different APIs.

The configs can be specified with the form:
gpu(api=string,dit=bool,nvpr=bool,samples=int)

This replaces and removes the --gpuAPI flag.

All existing configs should still work.

Adds following documentation:

out/Debug/dm --help config

Flags:
    --config: type: string default: 565 8888 gpu nonrendering
        Options: 565 8888 debug gpu gpudebug gpudft gpunull msaa16 msaa4
        nonrendering null nullgpu nvprmsaa16 nvprmsaa4 pdf pdf_poppler skp svg
        xps or use extended form 'backend(option=value,...)'.

        Extended form: 'backend(option=value,...)'

        Possible backends and options:

        gpu(api=string,dit=bool,nvpr=bool,samples=int) GPU backend
         api type: string default: native.
             Select graphics API to use with gpu backend.
             Options:
         native Use platform default OpenGL or OpenGL ES backend.
         gl     Use OpenGL.
         gles   Use OpenGL ES.
         debug  Use debug OpenGL.
         null   Use null OpenGL.
         dit type: bool default: false.
             Use device independent text.
         nvpr type: bool default: false.
             Use NV_path_rendering OpenGL and OpenGL ES extension.
         samples type: int default: 0.
             Use multisampling with N samples.

        Predefined configs:

         gpu       = gpu()
         msaa4     = gpu(samples=4)
         msaa16    = gpu(samples=16)
         nvprmsaa4 = gpu(nvpr=true,samples=4)
         nvprmsaa16 = gpu(nvpr=true,samples=16)
         gpudft     = gpu(dit=true)
         gpudebug   = gpu(api=debug)
         gpunull    = gpu(api=null)
         debug      = gpu(api=debug)
         nullgpu    = gpu(api=null)

BUG=skia:2992

Committed: https://skia.googlesource.com/skia/+/e13ca329fca4c28cf4e078561f591ab27b743d23
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1490113005

Review URL: https://codereview.chromium.org/1490113005

Remove some unneeded calls to GrGLGpu::configToGLFormats
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1530313005

Review URL: https://codereview.chromium.org/1530313005

Add field for internalFormat param to glTexImage to GrGLGpu::ConfigEntry.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1532273002

Review URL: https://codereview.chromium.org/1532273002

Remove dead code for glTexStorage

This is untested and therefore would be hard to maintain as this code is modified. We can always add it back later if we decide to use it.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1531013005

Review URL: https://codereview.chromium.org/1531013005

Make SkMatrix::get*Scale[s]() fail on NaN

BUG=skia:4718
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1532963002

Review URL: https://codereview.chromium.org/1532963002

Fix UB function problems for shaders and mask.

BUG=skia:4634
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1530743002

Review URL: https://codereview.chromium.org/1530743002

Add fExternalFormatForTexImage to GrGLGpu::ConfigEntry. Use it in GrGLGpu::uploadTexData.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1531343002

Review URL: https://codereview.chromium.org/1531343002

Re-enable WritePixels test for iOS

BUG=skia:2733
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1537793002

Review URL: https://codereview.chromium.org/1537793002

SkMatrix::preScale() is too conservative

SkMatrix::preScale() always sets the kScale bit, which means something
like

  m = SkMatrix::MakeScale(2, 2);
  m.preScale(0.5, 0.5);

leaves m.getType() == kScale_Mask, and can throw off the bitmap proc
heuristics.

We could detect the inverse case and clear the scale bit instead.

R=reed@google.com,caryclark@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1531323002

Review URL: https://codereview.chromium.org/1531323002

Create a hash table from id<-->key in SkImageFilter::CacheImpl

There is memory leak in the SkImageFilter::Cache. There are two sources
of memory leak:
1. The cache filling up quickly.
2. A slow small leak that never stops.
This CL solves the first issue, which prevents the cache filling up quickly.
This CL creates a new hash table that index the
SkImageFilter::uniqueID to an array of keys, and with the existing
key<-->Value hash table, we can have SkImageFilters proactively
purge content derived cached content when destroyed.

BUG=489543
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1514893003

Review URL: https://codereview.chromium.org/1514893003

Add default ctor to SkMask

The minimal fix here seems to be handling BoxBlur's return value in SkBlurMaskFilter.cpp::GrRRectBlurEffect::Create. We seem to do enough special handling of the fImage field though that always initialializing it may not be a bad idea.

BUG=570232
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1539553002

Review URL: https://codereview.chromium.org/1539553002

Fixup GL instanced rendering commands

We've ended up with a flag in both GrCaps and GrGLCaps to indicate
the same instanced rendering functionality. This change removes the
GrGLCaps version of the flag and also also fixes the interface assembler
to use the proper suffix when loading the EXT version of the extension.

BUG=skia:

Review URL: https://codereview.chromium.org/1537483002

Added concave paths to aaxfermodes GM
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1534823004

Review URL: https://codereview.chromium.org/1534823004

update CMake and Google3 builds for AVX/AVX2

The _avx.cpp code was working in a degraded (SSE2/SSSE3) mode silently.
This CL will make the _avx.cpp code start using AVX.

There is currently no _avx2.cpp code, but I'm writing some now,
and this change will make it work properly from the start.

Chrome and Android builds should already be okay.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1535443003
CQ_EXTRA_TRYBOTS=client.skia.compile:Build-Ubuntu-GCC-x86_64-Release-CMake-Trybot,Build-Mac10.9-Clang-x86_64-Release-CMake-Trybot

Review URL: https://codereview.chromium.org/1535443003

Move the GrGLGpu config stencil format index into ConfigTable
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1526253007

Review URL: https://codereview.chromium.org/1526253007

Add 2 new SKPS to the buildbot repository

BUG=skia:4713
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1534483004
NOTRY=true

Review URL: https://codereview.chromium.org/1534483004

Don't allow BGRA rendertargets on iOS w/ES 3.0+.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1531283002

Review URL: https://codereview.chromium.org/1531283002

Use the unswapped end point y for early out case in winding_line

The x-coordinates are not swapped, so using the swapped y will result
in a comparison with the wrong (end) point.

BUG=skia:4265
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1533873002

Review URL: https://codereview.chromium.org/1533873002