Revert "Rm makeRenderTargetContext in favor of deferred version"

This reverts commit d83ec0441279b79551b4ecb4598696249ea55145.

Reason for revert: various bot failures

Original change's description:
> Rm makeRenderTargetContext in favor of deferred version
>
> Change-Id: Ifdf67453607256ba7cdb0837263377bd83907212
> Reviewed-on: https://skia-review.googlesource.com/13001
> Reviewed-by: Brian Salomon <bsalomon@google.com>
> Commit-Queue: Robert Phillips <robertphillips@google.com>
>

TBR=bsalomon@google.com,robertphillips@google.com,reviews@skia.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Change-Id: I4e67a20cbe658b24725dcf386ac8bfdd127647ea
Reviewed-on: https://skia-review.googlesource.com/13131
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>

Rm makeRenderTargetContext in favor of deferred version

Change-Id: Ifdf67453607256ba7cdb0837263377bd83907212
Reviewed-on: https://skia-review.googlesource.com/13001
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>

remove kReturnNullForEmpty_FinishFlag feature

It's unused, and doesn't appear to be correct, returning nullptr too
early when we should return a non-empty SkMiniPicture.

The mini-recorder path will return a zero-allocation SkEmptyPicture anyway.

Change-Id: I1be538049e731acfc2b0f8b4f30cafee434a91dc
Reviewed-on: https://skia-review.googlesource.com/12626
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>

SkImageFilter::onMakeColorSpace() - more overrides

Fills out most of the simple implementations.
Improves 22 gms in gbr-8888.

Bug: skia:
Change-Id: I881ade140993568263de75be51aed240d2de8cc6
Reviewed-on: https://skia-review.googlesource.com/13126
Commit-Queue: Matt Sarett <msarett@google.com>
Reviewed-by: Mike Klein <mtklein@chromium.org>

Add Galaxy S6 to CQ as an experimental bot.

BUG=skia:

Change-Id: I8455d5595c0d0700d23fde6171c578fcf3be9270
Reviewed-on: https://skia-review.googlesource.com/13100
Reviewed-by: Ravi Mistry <rmistry@google.com>
Commit-Queue: Ben Wagner <benjaminwagner@google.com>

getDeferredTextureImageData(): preserve color space in legacy mode

Bug: skia:
Change-Id: Ib205f6104827b734635df6932d6dcfdc2248d091
Reviewed-on: https://skia-review.googlesource.com/12103
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Matt Sarett <msarett@google.com>

SkImageCacherator: Remove unneeded heuristic

This was the old way of handling color spaces.

This is now handled by chooseCacheFormat() / buildCacheInfo().

Bug: skia:
Change-Id: Ifab1ab25d547a00296af237578dd8befec9c86c6
Reviewed-on: https://skia-review.googlesource.com/13004
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Matt Sarett <msarett@google.com>

PM/UPM conversion improvements

Snap input values to N / 255 before doing PM/UPM conversions, and add
a new round-to-nearest conversion pair. With the pre-snapping, that
pair is chosen by (almost?) every GPU we have. This now lets us round
trip perfectly on Mali GPUs (which were falling back to SW before).

Bug: skia:
Change-Id: I0cebf1382e1c829aedfef4ec1614d7fa980057b5
Reviewed-on: https://skia-review.googlesource.com/11520
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>

fix too-dark 565 (really, all legacy) gradients

This CL reverts legacy destinations back to lerping bytes.
Lerping linear is only the correct behavior for gradients
drawing into colorspace-aware destinations.

As written we're lerping between de-sRGB'd colors but never
transforming the outputs back.  That leaves us in a weird
halfway-right-is-worse-than-wrong spot for legacy.

Change-Id: I79b85552b6913649afd2414205cf57108a8b93c6
Reviewed-on: https://skia-review.googlesource.com/13064
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Florin Malita <fmalita@chromium.org>

remove unused lockPixelsAreWritable

Bug: skia:4328
Change-Id: I7271a95a52fe6e5de781f83e0d48af5b69892a95
Reviewed-on: https://skia-review.googlesource.com/13066
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Mike Reed <reed@google.com>

Add MIPS Ci20 jobs

Bug: skia:6423
NOTRY=true
Change-Id: I2193c49151359a174ffa071650e1dbd366c91b28
Reviewed-on: https://skia-review.googlesource.com/11783
Commit-Queue: Kevin Lubick <kjlubick@google.com>
Reviewed-by: Mike Klein <mtklein@chromium.org>

Remove GeForce 320M bots

Bug: skia:
Change-Id: I3ae75d4edde1b3c8cc44413db74e5ab9789a9020
Reviewed-on: https://skia-review.googlesource.com/13065
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

fix -Fast bot

N=8 on that bot.

CQ_INCLUDE_TRYBOTS=skia.primary:Build-Ubuntu-Clang-x86_64-Release-Fast

Change-Id: If54ae800b50d9dffb9f983b23ff6f522657943b1
Reviewed-on: https://skia-review.googlesource.com/13061
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>

constexprify GrBlend.h and GrPorterDuffXferProcessor.cpp

Also remove some unused functions from GrBlend.h and related unit test.

Bug: skia:
Change-Id: Id8ad0057a02f65a9e19dc75e4b88709a762f4139
Reviewed-on: https://skia-review.googlesource.com/12623
Reviewed-by: Chris Dalton <csmartdalton@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>

Remove Index_8 handling in cacherator heuristic code

Constructor changes the color type to N32 now, so this
code is never needed.

Bug: skia:
Change-Id: If464440fed799122fed810d870f5d6d9a932bf12
Reviewed-on: https://skia-review.googlesource.com/13000
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Matt Sarett <msarett@google.com>

Recipes: Remove depot_tools path hack

Bug: skia:6473
Change-Id: I722059844bb0bc9d58347b4f6002ee62043b9e20
Reviewed-on: https://skia-review.googlesource.com/13006
Reviewed-by: Ravi Mistry <rmistry@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

Fix BundleRecipes depending on Release-Shared build

It doesn't create an actual dependency, but it does cause the
Release-Shared build to be added to the tasks DAG, where it would be
orphaned if not for the other housekeeper.

Bug: skia:
Change-Id: Ia2974728ec353c7f209db45474e572d83cacbda5
Reviewed-on: https://skia-review.googlesource.com/13003
Reviewed-by: Ravi Mistry <rmistry@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

Add multi-stop SkJumper stage.

CQ_INCLUDE_TRYBOTS=skia.primary:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD

Change-Id: I954d02638a785bec284d2fdf8f46abfccd474e7a
Reviewed-on: https://skia-review.googlesource.com/10211
Commit-Queue: Herb Derby <herb@google.com>
Reviewed-by: Mike Klein <mtklein@chromium.org>
Reviewed-by: Florin Malita <fmalita@chromium.org>

Move ref counting out of GrProcessor and into subclasses.

This will allow different subclasses to use different models for lifetime management.

GrXferProcessor moves to simple ref counting since they don't own GrGpuResources.

This also constifies GrXferProcessor factories.

Change-Id: I6bea0ea8de718874063224232f9da50887868b16
Reviewed-on: https://skia-review.googlesource.com/11792
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>

Clean up no-longer-used recipes stuff

- Remove buildbot logic
- Remove non-Gerrit logic
- Remove master, slave, and build number properties

Bug: skia:6473
Change-Id: I203d523f3a0e03f933c0c48d29c7aabac503b2b9
Reviewed-on: https://skia-review.googlesource.com/12625
Reviewed-by: Ravi Mistry <rmistry@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

Add SkImageFilter::makeColorSpace(SkColorSpaceXformer*)

Plus two sample implementations.  Will fill out the
rest if the model looks ok.

Bug: skia:
Change-Id: Ie24ecd0cbc36263220cb1e506f7812c582554e50
Reviewed-on: https://skia-review.googlesource.com/12100
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Matt Sarett <msarett@google.com>

Fix infinite loop in SkRTree::insert()

When aspectRatio is not finite, insert() can fall into an infinite loop.
This happens if you pass SkRect::MakeLargest() to the factory as bounds.

BUG=skia:5974

Change-Id: Ibcc9e5c5943c718608d4c1448305f7b8f11413bc
Reviewed-on: https://skia-review.googlesource.com/11784
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>

Move recipe content from modules back into recipes

At one point I moved the contents of the recipes into modules so that
they could be shared between repos. It turns out that we don't need
that, and it adds complexity.

Bug: skia:6473
Change-Id: I75a920b6a8474dcdd8b37ee9edd52aac801d1ab0
Reviewed-on: https://skia-review.googlesource.com/12622
Reviewed-by: Ravi Mistry <rmistry@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

consolidate read/writePixels in GrSurfaceContext

Change-Id: I118fcd49990597d4dfea92efd3f9d99e52fdbfab
Reviewed-on: https://skia-review.googlesource.com/11481
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>

Rename recipes: swarm_[name] -> [name]

Bug: skia:6473
Change-Id: Id78f0b5cb50ae2243b9a5aed8e552f7ed8202228
Reviewed-on: https://skia-review.googlesource.com/12060
Commit-Queue: Eric Boren <borenet@google.com>
Reviewed-by: Ravi Mistry <rmistry@google.com>

Update SKP version

Automatic commit by the RecreateSKPs bot.

TBR=update-skps@skia.org
NO_MERGE_BUILDS

Change-Id: Ia2240c06c3d778874de32cb1feba672f6e557aa2
Reviewed-on: https://skia-review.googlesource.com/12952
Reviewed-by: update-skps <update-skps@skia.org>
Commit-Queue: update-skps <update-skps@skia.org>

Update SKP version

Automatic commit by the RecreateSKPs bot.

TBR=update-skps@skia.org
NO_MERGE_BUILDS

Change-Id: Ib8926f5626f9d99aafc305b7172179f3d0bc573f
Reviewed-on: https://skia-review.googlesource.com/12950
Reviewed-by: update-skps <update-skps@skia.org>
Commit-Queue: update-skps <update-skps@skia.org>

Fix GrGLSync definition

By fixing this some ugly casts are avoided in setting up Chrome's GrGLInterface

Change-Id: Ie540fc41b3cbb27f976d784cad90ae10c2cef9ed
Reviewed-on: https://skia-review.googlesource.com/12621
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>

Revert "Revert "Create GrXferProcessor while doing GrProcessorSet analysis.""

This reverts commit 5dac9b3b5bf7e9c06c207cb92e257535c7d9ec95.

Bug: skia:
Change-Id: I3db2ec6776eb10c1f863b3992a2c8048c54b130f
Reviewed-on: https://skia-review.googlesource.com/12620
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>

Revert "Create GrXferProcessor while doing GrProcessorSet analysis."

This reverts commit 0ae0e23696f2ef08503040f8c02765eb58b26ddf.

Reason for revert: static assert failure on ios build

Original change's description:
> Create GrXferProcessor while doing GrProcessorSet analysis.
>
>
> Bug: skia:
> Change-Id: I62a628f9c0536ffb05c8f9d0c9ded5657f93b48e
> Reviewed-on: https://skia-review.googlesource.com/11482
> Reviewed-by: Greg Daniel <egdaniel@google.com>
> Commit-Queue: Brian Salomon <bsalomon@google.com>
>

TBR=egdaniel@google.com,bsalomon@google.com,reviews@skia.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Change-Id: Ifd57367b0326dd2b53c622705a868e246645d589
Reviewed-on: https://skia-review.googlesource.com/12104
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>

Create GrXferProcessor while doing GrProcessorSet analysis.

Bug: skia:
Change-Id: I62a628f9c0536ffb05c8f9d0c9ded5657f93b48e
Reviewed-on: https://skia-review.googlesource.com/11482
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>

remove obsolete SK_SUPPORT_LEGACY_SCALARMUL code

Bug: skia:
Change-Id: Ifdff1af7fc6b8cd6db77e4afd9794b11011cc31e
Reviewed-on: https://skia-review.googlesource.com/12242
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Reed <reed@google.com>

Set the kMultiMaster_FontFlag in DirectWrite.

DirectWrite can now produce system font typefaces which are variations,
so mark these as multiple masters so printing knows what to do.

BUG=chromium:697916

Change-Id: Idf09ebba3c7002a09ff2e4a2dbae13dbce4e79d4
Reviewed-on: https://skia-review.googlesource.com/12101
Commit-Queue: Ben Wagner <bungeman@google.com>
Reviewed-by: Hal Canary <halcanary@google.com>

Make SkColorSetA() a constexpr function.

This gives compilers a strong hint that SkColor constants to be computed
at compile time.

Change-Id: I09c8a7fd23e4488f8b101f53d67a155f64d5d423
Reviewed-on: https://skia-review.googlesource.com/12056
Commit-Queue: Lei Zhang <thestig@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Mike Klein <mtklein@chromium.org>

Convert Sk4fGradientInterval to (bias, factor) format

And rename some fileds:

 * p0, p1   -> t0, t1
 * fC0, fDc -> fCb, fCg

Change-Id: I94b520f83d8f21d62c9c24740b6d3e2dbc3a9ff7
Reviewed-on: https://skia-review.googlesource.com/11799
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>

Allow Vulkan to upload to main mip level without uploading to all levels.

It should be allowed for the client to upload original data to a texture via
writePixels and then we just regenerate the mipmaps. I think it also resonable
to limit this to either writting to all levels or just the top level.

Bug: skia:
Change-Id: I66943cca54c2a7187a781788653948fb69c17c68
Reviewed-on: https://skia-review.googlesource.com/11798
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>

make SkAutoTMalloc self-move safe

We were just combing through possible changes that might have affected
the attached bug, this popped out as pretty obviously unsafe.

This doesn't explain the Chrome bug...
SkAutoTMalloc and SKAutoSTMalloc are separate types.  :(

The new test fails and crashes before, passes after.

Change-Id: I033f488a7f644b7a70e612c8535fedfac35c76db
Reviewed-on: https://skia-review.googlesource.com/11797
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Herb Derby <herb@google.com>

SkTypeface:: populate_glyph_to_unicode uniformly across platforms

Change-Id: I7edce3c38b4dc0261d425c26a443df9a1f25b32e
Reviewed-on: https://skia-review.googlesource.com/11786
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Hal Canary <halcanary@google.com>

jumper, byte_tables + byte_tables_rgb

Factors out a function F from_byte(U8) too.

Change-Id: Ib739ccbd509ddf25d2bfb7751ba6eaf51b16c12f
Reviewed-on: https://skia-review.googlesource.com/11791
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>

cacherator upscales colortables to unify caching

Bug: skia:
Change-Id: Ib63f96b83d696743bbe4335c998acd4d2ea8acdb
Reviewed-on: https://skia-review.googlesource.com/11787
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Mike Reed <reed@google.com>

SkMultiPictureDocument : clean up MultiPictureDocumentReader interface

Motivation:  We may want to make SkMultiPictureDocument.h public in the
future.

Change-Id: Ie97b88d51a179c2283155d65bcadee32178115ca
Reviewed-on: https://skia-review.googlesource.com/11402
Commit-Queue: Hal Canary <halcanary@google.com>
Reviewed-by: Herb Derby <herb@google.com>

Make GrCopySurfaceOp friendlier to GrSurfaceProxy (take 2)

Reland of: https://skia-review.googlesource.com/c/11325/ (Make GrCopySurfaceOp friendlier to GrSurfaceProxy)

Change-Id: Ibe7312cebe7793af5803353a1499fbe7874c558f
Reviewed-on: https://skia-review.googlesource.com/11780
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>

Simplify subset computation in SkGifImageReader

If xOffset or yOffset was not zero, these computations would not
be evaluated anyway, so no need to add zero for these comparisons.

(Split off from CL 9810)

Change-Id: I8481eb257d96766696945cef4eafaff4281fb7d9
Reviewed-on: https://skia-review.googlesource.com/11782
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Leon Scroggins <scroggo@google.com>

stop disabling threadsafe statics

Chrome's got threadsafe statics, Android too, Flutter too, and heck, we even have them on all our Windows builds.  No one responded to my mail about them.  Let's stop turning them off on not-Windows!

Change-Id: Iadf00336aa73df3f694ac019c9d54a3f2257a3bd
Reviewed-on: https://skia-review.googlesource.com/11683
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>

Fix missing valgrind.supp in recipe-bundled tasks

Bug: skia:5813
Change-Id: I734c716d953091e84d30ad36c55a6c74727f376c
Reviewed-on: https://skia-review.googlesource.com/11781
Reviewed-by: Eric Boren <borenet@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

Add ChromeOS Test/Perf jobs and recipe changes

Minor cleanup of left over Tab3 tests.

Bug: skia:6442
Change-Id: I03bb4a034101a2773619abd2be7a300a70f6ac3d
Reviewed-on: https://skia-review.googlesource.com/11414
Reviewed-by: Eric Boren <borenet@google.com>

SkTypeface_win_dw: populate_glyph_to_unicode uses lowest unichar

BUG=chromium:538133
BUG=chromium:359065
Change-Id: I8208efc63579e0f1ebb77680beae2cd08cdba547
Reviewed-on: https://skia-review.googlesource.com/11681
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Hal Canary <halcanary@google.com>

simplify SkRRect serialization

I think this happens to fix the particular issue in the attached bug.
memcpy() is kind of the swiss army knife as far as strict aliasing is
concerned... you're always allowed to use it.

The generated code for writeToMemory() is unchanged, and
readFromMemory() gets a bit better looking.

BUG=skia:5105

Change-Id: Ib5bf96600f1138650c004ced2d696e9a4ba83ca7
Reviewed-on: https://skia-review.googlesource.com/11682
Reviewed-by: Cary Clark <caryclark@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>

Try disabling MSAN workarounds in SkJpegCodec

I don't think we've tried this since switching to GN,
where we decided to just go with jsimd_none.c on all x86 bots.

CQ_INCLUDE_TRYBOTS=skia.primary:Test-Ubuntu-Clang-GCE-CPU-AVX2-x86_64-Debug-MSAN,Perf-Ubuntu-Clang-GCE-CPU-AVX2-x86_64-Debug-MSAN

BUG=skia:4550

Change-Id: Iff4be3b640fb77d1094565a5db4d32778e7f2b4f
Reviewed-on: https://skia-review.googlesource.com/11680
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>

Reland: Use bundled recipes for all test/perf tasks

It won't save as much time for faster machines, but it should still
shave a few seconds, and it'll be nice to be consistent.

Bug: skia:
Change-Id: I0b2aa010579b527c66325cb5b691bfb549b31316
Reviewed-on: https://skia-review.googlesource.com/11418
Reviewed-by: Kevin Lubick <kjlubick@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

jumper, gather_f16

Here we use 64-bit gather instructions for HSW,
which I think we haven't done before.

Change-Id: I7b22b3cc0b7a151952518bb9afb90624ebdb4a22
Reviewed-on: https://skia-review.googlesource.com/11602
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>

jumper, gather_i8

Change-Id: Iefa8044bac0555c5fff370217a6270b4f3c64300
Reviewed-on: https://skia-review.googlesource.com/11582
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>

Update stencil ops to have a parallel proxyID

Change-Id: I842c7f2011c8b9d4487cec4349de8b77f326a381
Reviewed-on: https://skia-review.googlesource.com/11360
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>

jumper, more gathers

This is all the gathers except index 8 and f16, which aren't conceptually hard
but I want to land separately.

Change-Id: I525f2496e55451041bd6ea07985858fda7b56a40
Reviewed-on: https://skia-review.googlesource.com/11524
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>

Revert "Make GrCopySurfaceOp friendlier to GrSurfaceProxy"

This reverts commit d7cda9a6f050c30e65b0744b3cfb45cecdc40834.

Reason for revert: Tegra3!!!!

Original change's description:
> Make GrCopySurfaceOp friendlier to GrSurfaceProxy
>
> Change-Id: Iba8c068bd96ab154d466e485a9db31dd552138de
> Reviewed-on: https://skia-review.googlesource.com/11325
> Commit-Queue: Robert Phillips <robertphillips@google.com>
> Reviewed-by: Brian Salomon <bsalomon@google.com>
>

TBR=bsalomon@google.com,robertphillips@google.com,reviews@skia.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Change-Id: Idc31d469d94e6e2772ee5714987b3a05bb902d88
Reviewed-on: https://skia-review.googlesource.com/11580
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>

Specialize boxBlur(Interp) based on 'transpose' parameter

boxBlurInterp showed up hot when running Speedometer. I noticed that
BoxBlur calls boxBlurInterp repeatedly, the call does not get inlined,
and the function could be simplified a bit if 'transpose' is a
compile-time constant.

In particular, dst_x_stride or dst_y_stride now become a constant 1,
which simplifies the dptr updates.

Bug: skia:none
Change-Id: I8782e5b133a2195328e13878aca8a749af150b86
Reviewed-on: https://skia-review.googlesource.com/11362
Reviewed-by: Stephen White <senorblanco@chromium.org>
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>

Revert "Manual revert: Add support for writing ICC profiles to webp encoder"

This reverts commit 4293a1e5f259cf12560d6e794c1a05ad4dbc0b32.

Reason for revert: Relanding, libwebp updated in Android master

Original change's description:
> Manual revert: Add support for writing ICC profiles to webp encoder
>
> Bug: skia:
> Change-Id: I4e70bee8c2ea8dbd5ae1e84aa097f5a7e2e62721
> Reviewed-on: https://skia-review.googlesource.com/11444
> Reviewed-by: Matt Sarett <msarett@google.com>
> Reviewed-by: Leon Scroggins <scroggo@google.com>
> Commit-Queue: Matt Sarett <msarett@google.com>
>

TBR=msarett@google.com,scroggo@google.com,reviews@skia.org,jzern@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Change-Id: I73aa9e8183241ada4ec6451567bce3d3d18995cc
Reviewed-on: https://skia-review.googlesource.com/11523
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Matt Sarett <msarett@google.com>

jumper, gather_8888

Change-Id: I70bd64d114a2460534bcb51d356e13d9bc3b8603
Reviewed-on: https://skia-review.googlesource.com/11491
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>

Make GrCopySurfaceOp friendlier to GrSurfaceProxy

Change-Id: Iba8c068bd96ab154d466e485a9db31dd552138de
Reviewed-on: https://skia-review.googlesource.com/11325
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>

tweaks to make gather_* easier in SkJumper

This moves all the values that gather_8888, gather_a8, etc. need to
the front of SkImageShaderContext, and dereferences the color table.

This should be a no-op, but will make these stages easier to write
in SkJumper.

Change-Id: I0dff97d5113d14e941e7b717cd85f0036764eb88
Reviewed-on: https://skia-review.googlesource.com/11492
Reviewed-by: Herb Derby <herb@google.com>
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>

jumper, turn off a few fancy features

This doesn't change any of the generated .S files, but it does cut a few
misc. sections from the intermediate .o files.  It's nice to get those
sections out of the way, and one day we might be able to find ways to
cut everything but .text... that'd allow us to switch the supicious
section sniffing code from a blacklist (no .const, no .literal, etc.) to
a more foolproof whitelist (.text or bust).

The remaining sections are only in ELF objects (aarch64.o, vfp4.o):
   .comment         (notes the version of Clang/LLVM that compiled it)
   .note.GNU-stack  (we manually add this back in build_stages.py)
and vfp4.o has two more sections that I don't understand yet:
   .ARM.exidx       (I'd have thought -fno-unwind-tables would cut this)
   .ARM.attributes

While doing this, I've tried to make the ARM flags a bit more compact.

Change-Id: I30ef6acb2a917ec938c5358c3f970fe04b6d7afa
Reviewed-on: https://skia-review.googlesource.com/11485
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>

make copyTo_with_allocator android only

Bug: skia:6464

Change-Id: I3d95daae2ae854671e6231e62f200913ce3586d0
Reviewed-on: https://skia-review.googlesource.com/11486
Commit-Queue: Mike Reed <reed@google.com>
Reviewed-by: Matt Sarett <msarett@google.com>

Commit the patch before bundling recipes

This was part of https://skia-review.googlesource.com/c/11340/ but split
out before relanding.

Bug: skia:5813
Change-Id: I5cc97a588300a8656989a8d73e1231bdd606ef39
Reviewed-on: https://skia-review.googlesource.com/11483
Reviewed-by: Eric Boren <borenet@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

Factor out common work in readSurfacePixels and writeSurfacePixels

Use more consistent structure between the two functions.

BUG=skia:5853

Change-Id: I82465bee30025574d0d6fad9e287b187d75fff8d
Reviewed-on: https://skia-review.googlesource.com/10042
Reviewed-by: Robert Phillips <robertphillips@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

handle odd-index-count in encode/decode

Bug: skia:
Change-Id: Iffb123001a77049c6581f63bbc69c62f241a87f8
Reviewed-on: https://skia-review.googlesource.com/11405
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

fix 36605271

typo in function header makes it appear
that a buffer overflow may occur

R=djsollen@google.com

Bug: b/36605271
Change-Id: Iab6d08ae71f49c46736dc025791faefdeb74df5e
Reviewed-on: https://skia-review.googlesource.com/11454
Reviewed-by: Derek Sollenberger <djsollen@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

Revert "Remove CPU dimension for Win GCE bots"

This reverts commit 25bf350d1096622e10cb9d244705c91ed309f8e6.

Reason for revert: Speculatively reverting to see if this unsticks try jobs.

Original change's description:
> Remove CPU dimension for Win GCE bots
>
> This is mainly to collapse the two groups on the capacity stats page.
>
> Bug: skia:
> Change-Id: Icc125f6fa773ebf11909b0606bc7958146da5983
> Reviewed-on: https://skia-review.googlesource.com/11417
> Reviewed-by: Kevin Lubick <kjlubick@google.com>
> Commit-Queue: Eric Boren <borenet@google.com>
>

TBR=borenet@google.com,kjlubick@google.com,reviews@skia.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Change-Id: I52f01f6a915fa0ac4446bd492b9e455a07841ee9
Reviewed-on: https://skia-review.googlesource.com/11484
Reviewed-by: Eric Boren <borenet@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

jumper, add load_f32()

Change-Id: I71d85ffe29bc11678ff1e696fa4a2c93d0b4fcbe
Reviewed-on: https://skia-review.googlesource.com/11446
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>

jumper, kill off F4

Its alignment (sometimes 4, sometimes 16) has proven to be error-prone.

This also means we don't really need LazyCtx::load().

I think I only had it there to make sure we were doing unaligned loads
of F4; the better way is to just never declare the data as aligned...

The generated code isn't quite as good, but I can live with it.

Change-Id: I5d57a580ca12c94ca84a5e8b72a66cf8d0c829eb
Reviewed-on: https://skia-review.googlesource.com/11406
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>

jumper, only print non-zero counts

This makes it easier for me to read.

Change-Id: I6936c31981a2af6b7d3d7db9e60d36010fefa7ec
Reviewed-on: https://skia-review.googlesource.com/11445
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>

Manual revert: Add support for writing ICC profiles to webp encoder

Bug: skia:
Change-Id: I4e70bee8c2ea8dbd5ae1e84aa097f5a7e2e62721
Reviewed-on: https://skia-review.googlesource.com/11444
Reviewed-by: Matt Sarett <msarett@google.com>
Reviewed-by: Leon Scroggins <scroggo@google.com>
Commit-Queue: Matt Sarett <msarett@google.com>

Add Windows tester to the CQ

This will likely be followed up by an increase in number
of Win VMs.

Bug: skia:6462
Change-Id: If22d934c43c57d4ad3a20730e91caed0d4848cd4
Reviewed-on: https://skia-review.googlesource.com/11440
Reviewed-by: Mike Klein <mtklein@chromium.org>
Reviewed-by: Ravi Mistry <rmistry@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

Remove CPU dimension for Win GCE bots

This is mainly to collapse the two groups on the capacity stats page.

Bug: skia:
Change-Id: Icc125f6fa773ebf11909b0606bc7958146da5983
Reviewed-on: https://skia-review.googlesource.com/11417
Reviewed-by: Kevin Lubick <kjlubick@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

Do nondeterministic FP testing on iOS

Bug: skia:6395
Change-Id: Icc2ba15c6152b8e68bbad656c3ebadb12b56d55e
Reviewed-on: https://skia-review.googlesource.com/11403
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>

Revert "SpecialImageTest no longer fails on GalaxyS6."

This reverts commit 093cb35807b20c6571874e5f8621bef5e9524b81.

Reason for revert: Not sure why, but it's failing again today.

Original change's description:
> SpecialImageTest no longer fails on GalaxyS6.
>
> BUG=skia:6338
> NOTRY=true
>
> Change-Id: Ie7305f654e9f818dc31cb49eba2e7bf4000a36f9
> Reviewed-on: https://skia-review.googlesource.com/11411
> Reviewed-by: Robert Phillips <robertphillips@google.com>
> Commit-Queue: Ben Wagner <benjaminwagner@google.com>
>

TBR=benjaminwagner@google.com,robertphillips@google.com,reviews@skia.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=skia:6338

Change-Id: I1e07c74d37682df42023e3b52924f9a29da2cccb
Reviewed-on: https://skia-review.googlesource.com/11441
Reviewed-by: Ben Wagner <benjaminwagner@google.com>
Commit-Queue: Ben Wagner <benjaminwagner@google.com>

remove legacy vertices flag and code

Bug: skia:6366
Change-Id: Ibea5f71c5ff570e4f3ccec7d064893602c3335cf
Reviewed-on: https://skia-review.googlesource.com/11359
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Reed <reed@google.com>

remove SK_SUPPORT_LEGACY_BITMAP_SETPIXELREF flag and code

Bug: skia:
Change-Id: I234d221d685a6feab891cbca84869177b4a954bb
Reviewed-on: https://skia-review.googlesource.com/11404
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Reed <reed@google.com>

Revert "Use bundled recipes on all Test/Perf tasks"

This reverts commit 43b9c6bbf66b0927a99062c68dff9ea8358f82db.

Reason for revert: Windows failing

Original change's description:
> Use bundled recipes on all Test/Perf tasks
>
> It won't save as much time for faster machines, but it should still
> shave a few seconds, and it'll be nice to be consistent.
>
> Bug: skia:5813
> Change-Id: I5724e7aae83851edff3129265a9ffee5c3f95825
> Reviewed-on: https://skia-review.googlesource.com/11340
> Commit-Queue: Eric Boren <borenet@google.com>
> Reviewed-by: Kevin Lubick <kjlubick@google.com>
>

TBR=borenet@google.com,kjlubick@google.com,reviews@skia.org,iannucci@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Change-Id: I2a9cc1c12079755969879e37da5cae9f93921ddd
Reviewed-on: https://skia-review.googlesource.com/11416
Reviewed-by: Eric Boren <borenet@google.com>
Commit-Queue: Eric Boren <borenet@google.com>

add guard for obsolete (non-functioning) replayClips

replayClips has been disabled (broken) for a while.
This CL just attempts to hide the api (will remove
once android's callsite is removed)

Bug: skia:
Change-Id: I35b412addfc0a08ea888a62609888b9b54dce2a6
Reviewed-on: https://skia-review.googlesource.com/11401
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Reed <reed@google.com>

Rm readPixels from GrSurface & move read/writeSurfacePixels to GrContextPriv (take 3)

This is in service of: https://skia-review.googlesource.com/c/11125/ (Add parallel proxyID to StencilOps & RenderTargetOpList) where I want a better choke point for texture creation to improve discard handling.

This is a re-reland of: https://skia-review.googlesource.com/c/11200/ (Rm readPixels from GrSurface & move read/writeSurfacePixels to GrContextPriv)

Change-Id: Icfb9dd223418dd460405efd2bfd9d1c356beed1a
Reviewed-on: https://skia-review.googlesource.com/11412
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>

Use bundled recipes on all Test/Perf tasks

It won't save as much time for faster machines, but it should still
shave a few seconds, and it'll be nice to be consistent.

Bug: skia:5813
Change-Id: I5724e7aae83851edff3129265a9ffee5c3f95825
Reviewed-on: https://skia-review.googlesource.com/11340
Commit-Queue: Eric Boren <borenet@google.com>
Reviewed-by: Kevin Lubick <kjlubick@google.com>

SpecialImageTest no longer fails on GalaxyS6.

BUG=skia:6338
NOTRY=true

Change-Id: Ie7305f654e9f818dc31cb49eba2e7bf4000a36f9
Reviewed-on: https://skia-review.googlesource.com/11411
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Ben Wagner <benjaminwagner@google.com>

Update MotoG4 to NPJ25.93-14.

NOTRY=true
BUG=skia:6453

Change-Id: Ie8c35643c7e038f64f03d95e1f8d58ba748d4a52
Reviewed-on: https://skia-review.googlesource.com/10754
Reviewed-by: Kevin Lubick <kjlubick@google.com>
Commit-Queue: Ben Wagner <benjaminwagner@google.com>

Fix the size check for the drawBitmap fast-path in SkBitmapDevice::drawBitmapRect. It would fail when the source rectangle had a non-zero offset, in which case it would compare the source rectangle with the offset to the extracted bitmap size, which always fails. The only thing that should matter is that the source rectangle and extract bitmap have the same size, since the offset gets added onto the matrix.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2089583002

Review-Url: https://codereview.chromium.org/2089583002
Committed: https://skia.googlesource.com/skia/+/ea9bc0c07b5dae78a9a449d7d7a07fc79262d41a
Review-Url: https://codereview.chromium.org/2089583002

Fix gather_uninteresting_hashes()

It can crash when the file length is an exact multiple of the page size.

When the file size is not an exact page size multiple,
SkData::MakeFromFileName() (i.e. mmap) happens to fill the rest with \0,
giving us a terminating \0 for free.  SkStrSplit() uses this to stop.

When the file size is an exact page size multiple, there's no guaranteed
\0.  We might find one on the next page immediately, eventually, or we
might just segfault.  Whoops.

To fix, copy to an SkString which ought to plop in a \0 for us.

Change-Id: I51bbfdd85dfbb1c2276249d0255cf1c410ef9999
Reviewed-on: https://skia-review.googlesource.com/11409
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Mike Klein <mtklein@chromium.org>

SkImageEncoder: Be more lenient on inputs

(1) Some clients want us to write ICC profiles, even though they
    have not opted into linear unpremultiplication.  This CL allows
    that behavior.

(2) We should not assert that the transfer function must be linear
    or srgb.  Particularly in non-linear blending modes, skia is
    willing to support a larger set of transfer functions.

(3) We still need to require linear or srgb when in kRespect transfer
    function mode.  We have not yet implemented linear unpremultiplies
    for arbitrary transfer functions.

Bug: skia:
Change-Id: Idce9f07c3d36eca4d78ede5e2650b2cab412904c
Reviewed-on: https://skia-review.googlesource.com/11349
Commit-Queue: Matt Sarett <msarett@google.com>
Reviewed-by: Leon Scroggins <scroggo@google.com>

Revert "Revert "Add support for writing ICC profiles to webp encoder""

This reverts commit e87d7781db75a7ec22a103eda0886c2ca8dc4d9d.

Reason for revert: Not the cause of the crashes.

Original change's description:
> Revert "Add support for writing ICC profiles to webp encoder"
>
> This reverts commit 0c9d0b4e03b6c10778329e995b5dfdcb97954b28.
>
> Reason for revert: Looks like it's breaking a number of bots.
>
> Original change's description:
> > Add support for writing ICC profiles to webp encoder
> >
> > Bug: skia:
> > Change-Id: If0a8f84ed88da96924370b841f2283c0ff8e32ab
> > Reviewed-on: https://skia-review.googlesource.com/10212
> > Commit-Queue: Matt Sarett <msarett@google.com>
> > Reviewed-by: Leon Scroggins <scroggo@google.com>
> >
>
> TBR=msarett@google.com,scroggo@google.com,reviews@skia.org,jzern@google.com
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
>
> Change-Id: Ic06ad9f19a4d743b34f8d3bd33f171b9d74badcb
> Reviewed-on: https://skia-review.googlesource.com/11408
> Reviewed-by: Jim Van Verth <jvanverth@google.com>
> Commit-Queue: Jim Van Verth <jvanverth@google.com>
>

TBR=jvanverth@google.com,msarett@google.com,scroggo@google.com,reviews@skia.org,jzern@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Change-Id: I42f6ddefaf87c87b155640950b52a456952130ec
Reviewed-on: https://skia-review.googlesource.com/11410
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Matt Sarett <msarett@google.com>

Fix MSAN error in test

Bug: skia:
Change-Id: I27afd02b3383e259101a856630d04566fdf7a994
Reviewed-on: https://skia-review.googlesource.com/11407
Reviewed-by: Matt Sarett <msarett@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Matt Sarett <msarett@google.com>

Revert "Add support for writing ICC profiles to webp encoder"

This reverts commit 0c9d0b4e03b6c10778329e995b5dfdcb97954b28.

Reason for revert: Looks like it's breaking a number of bots.

Original change's description:
> Add support for writing ICC profiles to webp encoder
>
> Bug: skia:
> Change-Id: If0a8f84ed88da96924370b841f2283c0ff8e32ab
> Reviewed-on: https://skia-review.googlesource.com/10212
> Commit-Queue: Matt Sarett <msarett@google.com>
> Reviewed-by: Leon Scroggins <scroggo@google.com>
>

TBR=msarett@google.com,scroggo@google.com,reviews@skia.org,jzern@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Change-Id: Ic06ad9f19a4d743b34f8d3bd33f171b9d74badcb
Reviewed-on: https://skia-review.googlesource.com/11408
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>

restore old api for android to unblock roll

Bug: skia:
Change-Id: Ib115666503784b928f44f04878ed06a5aa844e6e
NOTRY=True
Reviewed-on: https://skia-review.googlesource.com/11420
Commit-Queue: Mike Reed <reed@google.com>
Reviewed-by: Mike Reed <reed@google.com>

Add support for writing ICC profiles to webp encoder

Bug: skia:
Change-Id: If0a8f84ed88da96924370b841f2283c0ff8e32ab
Reviewed-on: https://skia-review.googlesource.com/10212
Commit-Queue: Matt Sarett <msarett@google.com>
Reviewed-by: Leon Scroggins <scroggo@google.com>

jumper, to_2dot2 and from_2dot2

Nothing too tricky here.

Change-Id: I2a10548efc75a6fd875fcb242790880d9b9a28fd
Reviewed-on: https://skia-review.googlesource.com/11388
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Matt Sarett <msarett@google.com>

Fix various SkBitmap::copyTo() bugs

I believe that the best fix is to move this entire function into the
framework and call SkPixmap::readPixels() from there.  Given that,
we have already branched, I think this is the simplest fix.

(1) We never added F16 support.  Enable F16.  This involves assigning
    color spaces to src and dst pixmaps so that the conversion makes
    sense.
(2) SkPixmap::readPixels() does not support parametric transfer
    functions.  Fortunately, the framework never uses copyTo() to
    do color space conversions.  We should at least support the
    case where the src color space matches the dst color space.

Bug: skia:
Change-Id: I129b7c02249eed34a9ec4aa0ca736aadccc19777
Reviewed-on: https://skia-review.googlesource.com/11387
Commit-Queue: Matt Sarett <msarett@google.com>
Reviewed-by: Mike Klein <mtklein@chromium.org>

remove trace and registers stages

These can't really be done with SkJumper.

Change-Id: Ic357f00695eacd2766f6dfb9a3be13b0c07c3650
Reviewed-on: https://skia-review.googlesource.com/11386
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>

jumper, load_u16_be and store_u16_be

Change-Id: I2d58538ab071b217d8dbbf2d802493d9045eabf2
Reviewed-on: https://skia-review.googlesource.com/11384
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>

Revert "Rm readPixels from GrSurface & move read/writeSurfacePixels to GrContextPriv (take 2)"

This reverts commit aaee31f18c0845417103d84285e365575def3c40.

Reason for revert: possible valgrind leak

Original change's description:
> Rm readPixels from GrSurface & move read/writeSurfacePixels to GrContextPriv (take 2)
>
> This is in service of: https://skia-review.googlesource.com/c/11125/ (Add parallel proxyID to StencilOps & RenderTargetOpList) where I want a better choke point for texture creation to improve discard handling.
>
> This is a reland of: https://skia-review.googlesource.com/c/11200/ (Rm readPixels from GrSurface & move read/writeSurfacePixels to GrContextPriv)
>
> Change-Id: Icd0a90d2beb483dc24ed87c3bace9c817019e148
> Reviewed-on: https://skia-review.googlesource.com/11326
> Reviewed-by: Brian Salomon <bsalomon@google.com>
> Commit-Queue: Robert Phillips <robertphillips@google.com>
>

TBR=bsalomon@google.com,robertphillips@google.com,reviews@skia.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Change-Id: Ia0b92bf6402cb5f9607310d356f43bff2e3e75eb
Reviewed-on: https://skia-review.googlesource.com/11361
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>

Delete SkPixelRef::readPixels()

Nobody calls this or overrides this.

Bug: skia:
Change-Id: I6235e4c7bcdc45523ce7f63842635c79e00d2dc4
Reviewed-on: https://skia-review.googlesource.com/11357
Commit-Queue: Matt Sarett <msarett@google.com>
Reviewed-by: Mike Reed <reed@google.com>

remove android legacy flags

relies on https://googleplex-android-review.git.corp.google.com/#/c/2060523/ landing first

Bug: skia:
Change-Id: I603c8e0835235e9d81bd0a5a559cb5ab5b5cfd56
Reviewed-on: https://skia-review.googlesource.com/11350
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Reed <reed@google.com>

Refactor Vulkan support to support Fuchsia

- Refactor runtime lookup of Vulkan functions
- Support Magma surfaces
- Add build flags to toggle Vulkan support

BUG=skia:

Change-Id: Ic71965c453507882521443006c4155c7481a9a3f
Reviewed-on: https://skia-review.googlesource.com/9342
Commit-Queue: Greg Daniel <egdaniel@google.com>
Commit-Queue: Mike Klein <mtklein@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Reviewed-by: Michael Jurka <mikejurka@google.com>
Reviewed-by: Mike Klein <mtklein@chromium.org>

Use the rounded text size for gasp.

When comparing against gasp ranges always use the rounded (gdi) size.
Previously the truncated size was used, but it appears other
implementations are using the rounded size.

BUG=chromium:706693

Change-Id: I185cdf5b905261038e5150a04eef1b99bf73d875
Reviewed-on: https://skia-review.googlesource.com/11354
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>