Switch SkMergeImageFilter over to new onFilterImage interface

This relies on CL: https://codereview.chromium.org/1762013002/ (Swap over to using SkImageFilter::filterImage instead of filterImageDeprecated)

TBR=bsalomon@google.com

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1763223002

Review URL: https://codereview.chromium.org/1763223002

Restore SkImage::NewTextureFromPixmap changes.

6ceeebd37a43d879c120b6ba100ae1febdd67a18 contained the fix for the issues that prompted the revert (index8 bitmaps weren't drawing).

Revert "TBR=robertphillips@google.com"

This reverts commit 236640872fa8cad105273055fdcf6829d5c025a1.

TBR=
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1775263004

Review URL: https://codereview.chromium.org/1775263004

un-restore test, as bug is not fixed

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1783633002

TBR=

Review URL: https://codereview.chromium.org/1783633002

Don't build CG on Linux CMake

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1784663002
CQ_EXTRA_TRYBOTS=client.skia.compile:Build-Ubuntu-GCC-x86_64-Release-CMake-Trybot,Build-Mac-Clang-x86_64-Release-CMake-Trybot

Review URL: https://codereview.chromium.org/1784663002

Create explicit microhttpd.gyp

Committed: https://skia.googlesource.com/skia/+/80993639dfb8bcd26633ba1c3b7334f3ef7f6eb8
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1779903002

Review URL: https://codereview.chromium.org/1779903002

TBR=robertphillips@google.com

Revert "Add SkImage::NewTextureFromPixmap"

This reverts commit 045afea97902936869d71389e656a9c6917fa32f.

Revert "Fix Mac bots because of failing readback of 565"

This reverts commit e344b26a7d3c0efaab21bb17fa5b2d96dfc92edf.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1777863003

Review URL: https://codereview.chromium.org/1777863003

SkImage now has makeShader to return sk_sp

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1785473002

Review URL: https://codereview.chromium.org/1785473002

unify peekPixels around pixmap parameter

requires this chrome CL to land first
https://codereview.chromium.org/1775393003/

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1784563002

Review URL: https://codereview.chromium.org/1784563002

Revert of Create explicit microhttpd.gyp (patchset #6 id:100001 of https://codereview.chromium.org/1779903002/ )

Reason for revert:
breaking per commit housekeeper

Original issue's description:
> Create explicit microhttpd.gyp
>
> Committed: https://skia.googlesource.com/skia/+/80993639dfb8bcd26633ba1c3b7334f3ef7f6eb8

TBR=bungeman@google.com,jcgregorio@google.com,ethannicholas@google.com,joshualitt@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1781853002

Add an SkImageGeneratorCG

This will serve as a replacement for SkImageDecoder_CG.

BUG=skia:4914
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1718273004

Review URL: https://codereview.chromium.org/1718273004

Add a bit more to the python test driver

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1773263002

Review URL: https://codereview.chromium.org/1773263002

Create explicit microhttpd.gyp

Review URL: https://codereview.chromium.org/1779903002

Limit the maximum buffer size of SkRawBufferedStream

BUG=b/27475341
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1780643003

Review URL: https://codereview.chromium.org/1780643003

Fix Mac bots because of failing readback of 565
TBR=robertphillips@google.com

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1784553003

Review URL: https://codereview.chromium.org/1784553003 .

Fix Windows clang compiler complaint

https://codereview.chromium.org/1740533003 ( Initialize font fallback when creating DWrite font manager) seems to have broken the clang/win builds with:

..\..\third_party\skia\src\ports\SkFontMgr_win_dw.cpp(272,11) : error: field
'fLocaleName' will be initialized after field 'fFontFallback'
[-Werror,-Wreorder]
, fLocaleName(localeNameLength)
^
1 error generated.

I have also created skbug.com/5078 (Add Windows clang bot)

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1784533002

Review URL: https://codereview.chromium.org/1784533002

Add SkImage::NewTextureFromPixmap
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1776913003

Review URL: https://codereview.chromium.org/1776913003

experimental: coreGraphicsPdf2png

motivation: used for testing on my macbook.

TBR=

Review URL: https://codereview.chromium.org/1131643004

SkPDF/Bench becnchmark new SkPDFUtils::FloatToDecimal function.

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1735063004

Review URL: https://codereview.chromium.org/1735063004

Switch SkPaintImageFilter over to new onFilterImage interface

This CL relies on: https://codereview.chromium.org/1762013002/ (Swap over to using SkImageFilter::filterImage instead of filterImageDeprecated)

TBR=bsalomon@google.com

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1761373002

Review URL: https://codereview.chromium.org/1761373002

Fully specify libmicrohttpd.a path in build.

This is more uniform and also makes things work with the cmake gyp
generator.

Review URL: https://codereview.chromium.org/1777573005

Remove uses SkImageDecoder in imgblur and imgslice

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1777753002

Review URL: https://codereview.chromium.org/1777753002

Fix up picture clip bounds in SkiaServe

BUG=skia:5067
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1781653002

Review URL: https://codereview.chromium.org/1781653002

Fix some issues for Linux Vulkan build

lifecycle in GrVkGpu was tagged as uninitialized before use, and err in GrVkImage as unused.

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1782453004

Review URL: https://codereview.chromium.org/1782453004

Reenable dev bounds clip reduction

TBR=joshualitt@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1781663002

Review URL: https://codereview.chromium.org/1781663002

Fix nullptr access in GrTextureProvider::createMipMappedTexture

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1777013002

Review URL: https://codereview.chromium.org/1777013002

partial switch over to sp usage of shaders

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1776973003

Review URL: https://codereview.chromium.org/1776973003

Revert of don't create zero length intervals (patchset #1 id:1 of https://codereview.chromium.org/1766243004/ )

Reason for revert:
This may, or may not, be blocking the DEPS roll with:

svg/W3C-SVG-1.1/painting-stroke-04-t.svg

-      LayoutSVGPath {path} at (50,127) size 380x26 [stroke={[type=SOLID] [color=#000000] [stroke width=25.00] [dash offset=10.00] [dash array={10.00, 10.00}]}] [data="M 50 140 L 430 140"]
+      LayoutSVGPath {path} at (60,127) size 370x26 [stroke={[type=SOLID] [color=#000000] [stroke width=25.00] [dash offset=10.00] [dash array={10.00, 10.00}]}] [data="M 50 140 L 430 140"]

Original issue's description:
> don't create zero length intervals
>
> Dashing a pattern without zero-length intervals should
> not create them if the end of the on interval coincides
> with the beginning of the initial dash offset.
>
> R=reed@google.com
> BUG=591993
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1766243004
>
> Committed: https://skia.googlesource.com/skia/+/18bbd00190623fb6cdb119df4a118ac3c1aed52a

TBR=reed@google.com,fmalita@chromium.org,caryclark@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=591993

Review URL: https://codereview.chromium.org/1779803002

SkPDF: use sk_ref_sp

    s/sk_sp<[^>]*>(SkRef(\([^)]*\)))/sk_ref_sp(\1)/

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1777923002

Review URL: https://codereview.chromium.org/1777923002

Add SkDrawPosTextHCommand JSON, fix skiaserve build.

Adds SkDrawPosTextHCommand ::fromJSON and ::toJSON.

Both SkDrawPosTextCommand and SkDrawPosTextHCommand's
::toJSON write the correct number of positions, preventing reading
uninitialized memory.

The microhttpd build is now done in the build tree as opposed to in a
temporary directory.

The microhttpd build script uses os.path.join so that absolute paths
do not confuse the build. This allows compatibility with the cmake
gyp generator as CMake likes to pass absolute paths.

The microhttpd gyp target is now marked as 'none' since it is not
a 'static_library' target (which directs gyp to compile sources into
a static library). The dependencies to the action are updated to the
minimum required for sane re-building.

The everything gyp target now depends on the skiaserve gyp target.
This means that when using skia_build_server=1, building 'most' will
build skiaserve, but when skia_build_server is not defined the
skiaserve target will still be available if specified manually.

The old json.gyp is removed as it currently does not build anything.
All of the files currently referenced by it as sources no longer exist.

Review URL: https://codereview.chromium.org/1775203002

Deprecate SkImageDecoder

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1776153003

Review URL: https://codereview.chromium.org/1776153003

Don't allow nullptr in texels array params (unless using a transfer buffer).

Require all levels in writePixels to have a non-nullptr.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1765633002

Committed: https://skia.googlesource.com/skia/+/8ee78f31b2a29a5f76403755ea17bad9be74a3ec

Review URL: https://codereview.chromium.org/1765633002

Remove usage of SkImageDecoder in debugger

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1774313003

Review URL: https://codereview.chromium.org/1774313003

don't create zero length intervals

Dashing a pattern without zero-length intervals should
not create them if the end of the on interval coincides
with the beginning of the initial dash offset.

R=reed@google.com
BUG=591993
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1766243004

Review URL: https://codereview.chromium.org/1766243004

Switch SkImageSource image filter over to new onFilterImage interface

This CL relies on: https://codereview.chromium.org/1762013002/ (Swap over to using SkImageFilter::filterImage instead of filterImageDeprecated)

TBR=bsalomon@google.com

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1772933002

Review URL: https://codereview.chromium.org/1772933002

use Make instead of Create to return a shared shader

Partially updated call sites. Undefine the flag in SkSHader.h to convert the remaining sites.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1772463002

Review URL: https://codereview.chromium.org/1772463002

Initialize the font fallback when the direct write font
manager is created.

This ensures that the fallback is initialized prior to
sandbox startup in Chromium.

This enables the use of font fallback when combined with
Chromium change https://codereview.chromium.org/1740593002/

Review URL: https://codereview.chromium.org/1740533003

SkPDF: Add sk_sp setters; .release() becomes std::move()

Note to reviewers:  Start with changes to SkPDFTypes.h

Many places that had a bare pointer owning a reference are refactored to
use a sk_sp.

There remain several places where a non-owning pointer `T*` should be
replaced with `const sk_sp<T>&` to eliminate the common pattern
`sk_sp<T>(SkRef(x))`.

Committed: https://skia.googlesource.com/skia/+/9904c9212074279380e21f96575078734dbbd308
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1775043002

Review URL: https://codereview.chromium.org/1775043002

Revert of nanobench: turn on f16 and sRGB on the bots (patchset #1 id:1 of https://codereview.chromium.org/1769163006/ )

Reason for revert:
Debug bots failing at
   Running writepix_RGBA_UPM f16

../../../src/core/SkBitmap.cpp:558: fatal error: ""false && \"Can't return addr for config\"""

Original issue's description:
> nanobench: turn on f16 and sRGB on the bots
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1769163006
>
> Committed: https://skia.googlesource.com/skia/+/6b5c5cc7b7bdffb4882706b915b544aba8714961

TBR=reed@google.com,mtklein@chromium.org
# 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/1780463003

Fix skiaserve in debug mode

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1774263002

Review URL: https://codereview.chromium.org/1774263002

Revert of SkPDF: Add sk_sp setters; .release() becomes std::move() (patchset #2 id:20001 of https://codereview.chromium.org/1775043002/ )

Reason for revert:
https://build.chromium.org/p/client.skia/builders/Linux%20Builder/builds/6405/steps/compile/logs/stdio

Original issue's description:
> SkPDF: Add sk_sp setters; .release() becomes std::move()
>
> Note to reviewers:  Start with changes to SkPDFTypes.h
>
> Many places that had a bare pointer owning a reference are refactored to
> use a sk_sp.
>
> There remain several places where a non-owning pointer `T*` should be
> replaced with `const sk_sp<T>&` to eliminate the common pattern
> `sk_sp<T>(SkRef(x))`.
>
> Committed: https://skia.googlesource.com/skia/+/9904c9212074279380e21f96575078734dbbd308

TBR=bungeman@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/1775143002

Add sk_ref_sp helper function.

Review URL: https://codereview.chromium.org/1773943004

SkPDF: Add sk_sp setters; .release() becomes std::move()

Note to reviewers:  Start with changes to SkPDFTypes.h

Many places that had a bare pointer owning a reference are refactored to
use a sk_sp.

There remain several places where a non-owning pointer `T*` should be
replaced with `const sk_sp<T>&` to eliminate the common pattern
`sk_sp<T>(SkRef(x))`.

Review URL: https://codereview.chromium.org/1775043002

nanobench: turn on f16 and sRGB on the bots

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1769163006

Review URL: https://codereview.chromium.org/1769163006

Fill out more information in GrVkCaps

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1768083003

Review URL: https://codereview.chromium.org/1768083003

Remove SkImageGeneratorUtils. None of this code is used.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1778613002

Review URL: https://codereview.chromium.org/1778613002

Buffer fix for VkProgram caching

When caching VkPrograms, we can end up with situations where we
overwrite uniform buffers while they're in use (or being prepared for
use) in the command buffer. This fix will address that. This also
addresses the rare but similar case of overwriting vertex or index
buffers when they're in flight.

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1774963003

Review URL: https://codereview.chromium.org/1774963003

Switch SkComposeImageFilter over to new onFilterImage interface

This CL depends on https://codereview.chromium.org/1762013002/ (Swap over to using SkImageFilter::filterImage instead of filterImageDeprecated)

TBR=bsalomon@google.com

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1768283002

Review URL: https://codereview.chromium.org/1768283002

Expand coverage of vertices gm

Review URL: https://codereview.chromium.org/1361173003

Update DNG SDK version.

* Integrate integer overflow changes from security branch

BUG=b/26535130
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1748363004

Committed: https://skia.googlesource.com/skia/+/be2924ad17d47f4ed49666894d87571a28f89567

Committed: https://skia.googlesource.com/skia/+/62c0f75159674087079140e09ef1396316577754

Review URL: https://codereview.chromium.org/1748363004

add f16 and srgb --configs to nanobench

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1775093002

Review URL: https://codereview.chromium.org/1775093002

4377:  drawVertices crashes when shader+colorfilter in the paint

Fix memory corruption that is result of static_casting SkFilterShaderContext*
 to SkTriColorShaderContext* and then changing it's state using setup()

SkTriColorShader and it's Context could be wrapped by Compose, 3DShader and
 SkFilterShader, but the link SkTriColorShaderContext -> SkTriColorShader
 (fShader) always remain - so this way, data about triangle is propagated.

BUG=4377

Didn't add new unit test - it would be redundant since the GM vertices update here: https://codereview.chromium.org/1361173003/ is covering the bug. With this patch it passes.

Review URL: https://codereview.chromium.org/1431713003

Create a small python driver for skiaserve

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1754213005

Review URL: https://codereview.chromium.org/1754213005

Rename SkSpecialImage::getBitmap to getBitmapDeprecated

Review URL: https://codereview.chromium.org/1780433002

Fix up GrAuditTrail to allow arbitrary reordering

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1765123002

Review URL: https://codereview.chromium.org/1765123002

Fix 4f gradient swizzle post crrev.com/1774523002

* store interval colors in pm4f-natural/RGBA order
* swizzle to dst order during interval advance

Also remove an unused Interval ctor.

R=reed@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1770153002

Review URL: https://codereview.chromium.org/1770153002

make pm4f be RGBA always, not pmcolor order

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1774523002

Review URL: https://codereview.chromium.org/1774523002

Lazy init batch unique ID

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1772023003

Review URL: https://codereview.chromium.org/1772023003

Small GrTessellator refactor and cleanup.

Implement a VertexList, and use it.
Rename insert<> -> list_insert<>.
Rename remove<> -> list_remove<>.
Remove some spurious returns.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1771373002

Review URL: https://codereview.chromium.org/1771373002

work around leftshift for negative values

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1774963002

Review URL: https://codereview.chromium.org/1774963002

DM: verbose and quiet logging

If you pass no flags, output is pretty much unchanged from today: calls to
info() go to the console, calls to vlog() are ignored.

If you pass --verbose and -w dir, all calls to info(), vlog() and all of stderr
are redirected to dir/verbose.log.  info() calls also go to the console,
unless you pass --quiet, in which case nothing goes to the console at all.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1771143002

Review URL: https://codereview.chromium.org/1771143002

Switch SkOffsetImageFilter over to new onFilterImage interface

This CL relies on: https://codereview.chromium.org/1762013002/ (Swap over to using SkImageFilter::filterImage instead of filterImageDeprecated)

TBR=bsalomon@google.com

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1766743003

Review URL: https://codereview.chromium.org/1766743003

Fix caching bug in new SkImageFilter::filterImage path

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1777573002

Review URL: https://codereview.chromium.org/1777573002

Add element_type, swap, operators, fix reset on sk_sp.

The 'element_type' typedef is to play nice with std::pointer_traits.

The full complement of operators and swap to match unique_ptr so that
sk_sp can be properly compared to nullptr and used with standard
containers.

Update to 'reset' so that calling 'unref' is the last operation.

This also adds tests for these changes, and sets the fPtr to nullptr
in debug for easier bug finding.

Review URL: https://codereview.chromium.org/1773453002

SkPDF: use sk_make_sp<T> when it makes sense.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1773033002

Review URL: https://codereview.chromium.org/1773033002

Circle GP/batch housecleaning
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1764923003

Review URL: https://codereview.chromium.org/1764923003

Swarming bots: fix trybots

Requires a recipe-side change as well.

NOTRY=true
BUG=skia:4763
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1768333002

Review URL: https://codereview.chromium.org/1768333002

Hide new methods from non-Android.

Review URL: https://codereview.chromium.org/1769933004

Disable color space test for old versions of libpng

Should fix the Google3 build.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1769953002

Review URL: https://codereview.chromium.org/1769953002

Archive SkRemote-related code.  I want to start fresh.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1774013002

Review URL: https://codereview.chromium.org/1774013002

Add support for 2x2 matrices

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1755483002

Review URL: https://codereview.chromium.org/1755483002

SkPDF: s/SkAutoTUnref/sk_sp/g

Review URL: https://codereview.chromium.org/1771073002

Add "sample locations" feature to GrProcessor

Adds a "sample locations" feature to GrProcessor. When enabled, this
allows a processor to know inside the shader where all the samples are
located. Also adds various infastructure to query, cache, and identify
multisample data.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1717393002

Review URL: https://codereview.chromium.org/1717393002

dm: log log

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1771993002

Review URL: https://codereview.chromium.org/1771993002

Swap over to using SkImageFilter::filterImage instead of filterImageDeprecated

This CL relies on https://codereview.chromium.org/1757983002/ (Add SkSpecialImage-based methods to SkImageFilter)

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1762013002

TBR=bsalomon@google.com

Review URL: https://codereview.chromium.org/1762013002

Add advance[XY]Fixed getters for Android.

This prepares for changing the type of these fields to float.

BUG=skia:4632

Review URL: https://codereview.chromium.org/1772863002

SkPDF Create working move constructor for inner classes

BUG=chromium:592330,chromium:592702

Review URL: https://codereview.chromium.org/1774633002

Update android_make to support both --gcc and --clang options.

This CL also fixes all the errors that resulted from compiling
with clang and ccache.

BUG=skia:4948
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1774503004

Review URL: https://codereview.chromium.org/1774503004

add virtual ~Effect()

Today we make an sk_sp<Effect> from an sk_sp<EffectImpl>.
But when the sk_sp<Effect> dies, it calls ~Effect(), not ~EffectImpl().
Making ~Effect() virtual fixes this.

This should make our Google3 tests sized-delete clean, unblocking those folks.

Review URL: https://codereview.chromium.org/1769093002

Attempt to combine batches in forward direction before flush
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1763883005

Review URL: https://codereview.chromium.org/1763883005

link with -rdynamic on Linux

This lets our tools get symbols when they backtrace themselves.

Here's an example improved stack trace.
Notice there are symbols for some Skia methods now.

Stack trace:
    out/Release/dm() [0x5ec19b]
    out/Release/dm() [0x5ec209]
    /lib/x86_64-linux-gnu/libc.so.6(+0x36d40) [0x7f6ae0ce2d40]
    /lib/x86_64-linux-gnu/libc.so.6(gsignal+0x39) [0x7f6ae0ce2cc9]
    /lib/x86_64-linux-gnu/libc.so.6(abort+0x148) [0x7f6ae0ce60d8]
    out/Release/dm() [0xcd19f9]
    out/Release/dm() [0x8ca959]
    out/Release/dm(_ZN11XfermodesGM16onOnceBeforeDrawEv+0x70) [0x8cb640]
    out/Release/dm(_ZN6skiagm2GM4drawEP8SkCanvas+0x1c) [0x5fc81c]
    out/Release/dm(_ZNK2DM5GMSrc4drawEP8SkCanvas+0x73) [0x5f3563]
    out/Release/dm(_ZNK2DM10RasterSink4drawERKNS_3SrcEP8SkBitmapP9SkWStreamP8SkString+0xcc) [0x5f9eac]
    out/Release/dm(_ZN4Task3RunERKS_+0xd5) [0x5f1685]
    out/Release/dm() [0xa4d320]
    out/Release/dm() [0xcb66ba]
    /lib/x86_64-linux-gnu/libpthread.so.0(+0x8182) [0x7f6ae2667182]
    /lib/x86_64-linux-gnu/libc.so.6(clone+0x6d) [0x7f6ae0da647d]

Today the same crash looks more like this.

Stack trace:
    out/Release/dm() [0x5ec19b]
    out/Release/dm() [0x5ec209]
    /lib/x86_64-linux-gnu/libc.so.6(+0x36d40) [0x7f6ae0ce2d40]
    /lib/x86_64-linux-gnu/libc.so.6(gsignal+0x39) [0x7f6ae0ce2cc9]
    /lib/x86_64-linux-gnu/libc.so.6(abort+0x148) [0x7f6ae0ce60d8]
    out/Release/dm() [0xcd19f9]
    out/Release/dm() [0x8ca959]
    out/Release/dm() [0x8cb640]
    out/Release/dm() [0x5fc81c]
    out/Release/dm() [0x5f3563]
    out/Release/dm() [0x5f9eac]
    out/Release/dm() [0x5f1685]
    out/Release/dm() [0xa4d320]
    out/Release/dm() [0xcb66ba]
    /lib/x86_64-linux-gnu/libpthread.so.0(+0x8182) [0x7f6ae2667182]
    /lib/x86_64-linux-gnu/libc.so.6(clone+0x6d) [0x7f6ae0da647d]

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1774593002

Review URL: https://codereview.chromium.org/1774593002

Swarming bots: add build products whitelist

Should reduce output isolate size (and therefore upload time) for
compile task by ~10x

NOTRY=true
BUG=skia:4763
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1768273002

Review URL: https://codereview.chromium.org/1768273002

Rework files and add test cases.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1757193002

Review URL: https://codereview.chromium.org/1757193002

Revert of Update Skia's YUV API (patchset #5 id:160001 of https://codereview.chromium.org/1716523002/ )

Reason for revert:
Blimp failures with manual roll in Chrome.

Original issue's description:
> Update Skia's YUV API
>
> We should match the recently designed API in SkCodec.
> https://codereview.chromium.org/1549473003/
>
> This requires changes in Chromium as well.
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1716523002
>
> Committed: https://skia.googlesource.com/skia/+/095d31c8a0eeb5d491febf064bc3c8a44e22b94f

TBR=scroggo@google.com,reed@google.com,bsalomon@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/1775493002

Revert of Add BigRectGM. (patchset #2 id:20001 of https://codereview.chromium.org/1758113005/ )

Reason for revert:
Causing ASAN failures. (I had expected float-cast ASAN failures, which are disabled right now, but we're also seeing integer overflows.)

Original issue's description:
> Add BigRectGM.
>
> BUG=skia:5060
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1758113005
>
> Committed: https://skia.googlesource.com/skia/+/a9101eef5ffc5f329a5edc75fd67441d868b0e9b

TBR=mtklein@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:5060

Review URL: https://codereview.chromium.org/1773593005

Update Skia's YUV API

We should match the recently designed API in SkCodec.
https://codereview.chromium.org/1549473003/

This requires changes in Chromium as well.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1716523002

Review URL: https://codereview.chromium.org/1716523002

Add BigRectGM.

BUG=skia:5060
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1758113005

Review URL: https://codereview.chromium.org/1758113005

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=1769023002

Review URL: https://codereview.chromium.org/1769023002

Use a smart pointer for SkColorSpace factories

This should fix master-skia in Android because we no longer need
to include src/core for SkCodec.h.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1766413002

Review URL: https://codereview.chromium.org/1766413002

Add SkSpecialImage-based methods to SkImageFilter

This is calved off of https://codereview.chromium.org/1695823002/ (Get OffsetImageFilter really working with SkSpecialImages)

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1757983002

Review URL: https://codereview.chromium.org/1757983002

Add variants of the setters on SkPaint which take a sk_sp<effect>. At the same time, change the internal storage to be sk_sp<effect>.

Follow-on CL might try to use = default for the constructors and assignment operators.

This reverts commit 992854d62e179a589aa7366e443246e3672c3248.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1770723002

Review URL: https://codereview.chromium.org/1770723002

SkPDF: AutoTUnref<T> changes in use

    ::detach()      -> ::release()
    ::operator T*() -> ::get()

This makes all use of AutoTUnref work the same as sk_sp.
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1772493002

Review URL: https://codereview.chromium.org/1772493002

Fix behavior of sk_sp::reset(T*) and add unittest.

Previously, sk_sp::reset(T* t) did not release its own reference
if its internal pointer was the same as 't'. This leaks a reference.
Now always release the current reference when non-nullptr.

Review URL: https://codereview.chromium.org/1767983002

Revert of add setter on SkPaint that takes sk_sp (patchset #1 id:1 of https://codereview.chromium.org/1769803002/ )

Reason for revert:
investigate leak

Original issue's description:
> add setter on SkPaint that takes sk_sp
>
> BUG=skia:
> GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1769803002
>
> TBR=
>
> Committed: https://skia.googlesource.com/skia/+/a917eba6ea8a6936f2c9271e487b14d14b99c98e

TBR=
# 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/1771613002

add setter on SkPaint that takes sk_sp

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1769803002

TBR=

Review URL: https://codereview.chromium.org/1769803002

sk_sp: fix contravariant constructors

TBR=reed
This is obviously correct and needs to be fixed.

Review URL: https://codereview.chromium.org/1771583002

remove unneeded flag for annotations

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1767783004

TBR=halcanary

Review URL: https://codereview.chromium.org/1767783004

Move annotations to canvas virtual (patchset #8 id:140001 of https://codereview.chromium.org/1744103002/ )"

Need to land chrome change first https://codereview.chromium.org/1766723003/

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1758023003

TBR=

Review URL: https://codereview.chromium.org/1758023003

Fix stall in Vk clear

TBR=egdaniel@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1765863003

Review URL: https://codereview.chromium.org/1765863003

Check libpng version before reading color space

TBR=scroggo@google.com

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1767723003

Review URL: https://codereview.chromium.org/1767723003

SkPDF: PDFDevice use SkTArray<T> rather than SkTDArray<T*>

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=1767713002

Review URL: https://codereview.chromium.org/1767713002