We can use this instead of ImmutableString("").

Bug: angleproject:2665
Change-Id: I8b3d5d3075838b9f2caa1627071202e48a5fdc83
Reviewed-on: https://chromium-review.googlesource.com/1108085
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>

The contextVk color mask needed to be updated when we
were syncing the state of the framebufferVk to make sure
the right mask gets applied in the test described in
ClearTest.cpp

Bug: angleproject:2685
Bug: angleproject:2597
Change-Id: I575028aad4cd495bf4ec1484cc870940dcde92d5
Reviewed-on: https://chromium-review.googlesource.com/1112038
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

Bug: angleproject:2685
Bug: angleproject:2597
Bug: angleproject:2687
Change-Id: I4cab6b70c3eae26b827cbbd00ecbfa8b4eddcd7c
Reviewed-on: https://chromium-review.googlesource.com/1114779Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>

Caused timeouts on the Skia bots when rolling ANGLE.

BUG=angleproject:2686
BUG=angleproject:2464

Change-Id: I311b075522b86410d055e0b742342423eaa7eeb5
Reviewed-on: https://chromium-review.googlesource.com/1115216Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

Nested structs are handled similarly as to non-nested samplers in
structs. They are extracted and named according to the same
pattern.

Also enables functional.shaders.random.all_features.fragment*

The remaining work in the samplers-in-structs implementation is to
translate function arguments.

Bug: angleproject:2494
Bug: angleproject:2595
Change-Id: If8170feb71137d4036d352b2b0078518647d48a1
Reviewed-on: https://chromium-review.googlesource.com/1101569
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>

Samplers are extracted from arrays of struct uniforms similarly as
with non-arrays. They are named according to the variable name,
array element and sampler field name.

Nested structs to come later.

Bug: angleproject:2494
Change-Id: Ie2f5f7bd1f747e73b3c6b505b2b1043cfe1073b5
Reviewed-on: https://chromium-review.googlesource.com/1101568
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>

Change-Id: Ia44413aa94906b89d4981063600cd3de0edacee8
Reviewed-on: https://chromium-review.googlesource.com/1108454
Commit-Queue: Ian Elliott <ianelliott@google.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

BUG=768943

Change-Id: I8ed69c70d1914b56145c52ffe26f13193fb55e9e
Reviewed-on: https://chromium-review.googlesource.com/685278
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

This patch implements 'shared' variables in compute shader on D3D11
back-ends. GLSL shared variables are translated into 'groupshared'
ones in HLSL.

Note that although HLSL allows initializing the variables with
'groupshared' qualifier, currently we do not initialize them because:
1. It is very slow to for d3d11 drivers to compile the compute shader if
   we add the code to initialize a shared variable with large array size.
2. It seems unnecessary to do so and in GLSL it is not allowed to
   initialize a shared variable in the declaration. (ESSL 3.1, Chapter
   4.3.8)

BUG=angleproject:2682
TEST=angle_end2end_tests

Change-Id: Ica8247e1b98059968612a36e369718ef113a598c
Reviewed-on: https://chromium-review.googlesource.com/1109587Reviewed-by: Jiajia Qin <jiajia.qin@intel.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jiawei Shao <jiawei.shao@intel.com>

Forgot to remove it from the header in
https://chromium-review.googlesource.com/577922

Bug: angleproject:1644
Change-Id: I89f9c975b3fc3b7161edd2c2907549eaa672c223
Reviewed-on: https://chromium-review.googlesource.com/1112647Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Commit-Queue: Yuly Novikov <ynovikov@chromium.org>

Bug: angleproject:2675
Change-Id: I3774ca4602e9eea7c191e99a95939043ac0d8777
Reviewed-on: https://chromium-review.googlesource.com/1111901
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

Bug: angleproject:2647
Change-Id: I3bf78c617a39191e725e0fa73befff4d42ac5962
Reviewed-on: https://chromium-review.googlesource.com/1111903Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>

BUG=angleproject:2464

Change-Id: Iaa4d075af6e92fd603f3dfaab0946daed2af64ab
Reviewed-on: https://chromium-review.googlesource.com/1111909Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

This helper method will also transition the Image to the correct read
layout. We will need to revisit the implementation when working on
simulatenous read.

Bug: angleproject:2539
Change-Id: Id61404460f3ef0dbb054e6ac2dfc0b59adb78402
Reviewed-on: https://chromium-review.googlesource.com/1108378
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>

Updated from https://raw.githubusercontent.com/KhronosGroup/OpenGL-Registry/master/xml/gl.xml.

BUG=chromium:849576

Change-Id: I92520d53844f7fff7e68ca49207151d4fa69c79e
Reviewed-on: https://chromium-review.googlesource.com/1111469Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

When updating the Uniforms and Textures pools at different rates we
could trigger an ANGLE bug. We would sometimes reset and free the
currently bound descriptors. We can fix this by using separate
descriptor pools for each descriptor set.

Bug: angleproject:2678
Change-Id: I605b558531e7745484e16156a3af5eac40ffcc79
Reviewed-on: https://chromium-review.googlesource.com/1110662
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>

The bug would only manifest after a glUseProgram(0) at a specific time.
Fix the bug by releasing the uniform buffer to the Renderer instead.

Bug: angleproject:2397
Change-Id: Ia071da5af00dfc740e0fc20864d41f7b3e0afda4
Reviewed-on: https://chromium-review.googlesource.com/1110712
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>

Changes the set counting to use a free count by descriptor type. Also
changes the init to take a pool size count that more closely mirrors
Vulkan.

Will lead to a fix for descriptor set recreation.

Bug: angleproject:2678
Change-Id: I26816befec28ebb866fb0f3aaba27d2bc1d3be43
Reviewed-on: https://chromium-review.googlesource.com/1110661Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

Knowing the share context at native context creation time is required if
we want to honor the requested share group instead of virtualizing
contexts or using global share groups.

BUG=angleproject:2464

Change-Id: I1fb4d71de266b3191986b1754e73d474e49445bb
Reviewed-on: https://chromium-review.googlesource.com/1108743
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Yuly Novikov <ynovikov@chromium.org>

When using unvirtualized contexts, DisplayWGL still creates a Renderer for
managing any internal GL resources such as emulated back buffers for
DXGISwapChainWindowSurfaceWGL or D3DTextureSurfaceWGL but also creates a new
Renderer for each GL context.  All created contexts share resources.

BUG=angleproject:2464

Change-Id: I945502514079368e062beef70bed49c61ed44403
Reviewed-on: https://chromium-review.googlesource.com/1097459
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Yuly Novikov <ynovikov@chromium.org>

Vulkan does not support separate depth and stencil attachments.
Adding this check to checkStatus skips the relevant tests to that.

There was also a bug in FramebufferVk::clear where we were not
checking the number of bits for depth/stencil attachments before
setting the aspects for clearing.

Bug: angleproject:2597
Change-Id: Iabe9f77a51fc7aca08b8faf3cecc3df9f99a7f1e
Reviewed-on: https://chromium-review.googlesource.com/1107847
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

When casting types to one another in C++, the weaker the cast,
the better.

This change replaces instances of reinterpret_cast with static_cast
or no cast where it safe and correct to do so.

BUG=angleproject:2683

Change-Id: I99c9033614a65282ae1d78cf0f4b80fabd75877a
Reviewed-on: https://chromium-review.googlesource.com/1109396
Commit-Queue: Rafael Cintron <rafael.cintron@microsoft.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

When the angleFormat we have has no alpha channel, but the actual
texture underneath has one, we shouldn't be drawing over the alpha
channel, so we apply a mask on it when we're in this situation.

Bug: angleproject:2597
Change-Id: Ia7110709e6ee32bb61988d08f5049e4e80e7e24e
Reviewed-on: https://chromium-review.googlesource.com/1106759
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

This avoids unnecessary pool allocations when generating errors. Most
of the returned strings are static.

BUG=angleproject:2267
TEST=angle_unittests

Change-Id: Ia9b26898d499985d61e629fddde7aded4714eddf
Reviewed-on: https://chromium-review.googlesource.com/885818Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

Vulkan requires all uniforms to be declared in a block. Uniform
blocks can't have store samplers. Thus we can't use structs in
samplers with Vulkan GLSL.

To work around this limitation we extract samplers from structs
and move them into standalone types. The samplers are named
according to the variable and struct fields.

Arrays of structs and nested structs to come later.

Bug: angleproject:2494
Change-Id: I83a94ab082c6ce7ee68ec1290751ecee18820683
Reviewed-on: https://chromium-review.googlesource.com/1101567Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

Also known as nameless structs. Uniform structs without a struct name
would not be parsed correctly. This fixes the bug by adding a tree
transformation. The transformation gives an internally scoped name to
the embedded struct.

Bug: angleproject:2665
Change-Id: I43e4dad7d9ad64a40e382066bb136e4f8f719797
Reviewed-on: https://chromium-review.googlesource.com/1101566Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

Bug: angleproject:2675
Change-Id: Idbf2590a763fd7b81d2bcc5850180a1bb931360e
Reviewed-on: https://chromium-review.googlesource.com/1108490Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>

libnativewindow doesn't have all the functions angle_util requires,
namely: AConfiguration_*, AInputEvent_*, AInputQueue_*, ALooper_*.
It should be safe to always use libandroid, since angle_util is only
used by tests, thus it won't interfere with AngleLibraries.apk

Bug: angleproject:2418
Change-Id: I8343924629205bfb5d65dafe34db8e2baa987bae
Reviewed-on: https://chromium-review.googlesource.com/1108747Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Reviewed-by: Courtney Goeltzenleuchter <courtneygo@google.com>
Commit-Queue: Yuly Novikov <ynovikov@chromium.org>

We were calling context->releaseSurface on a context that was no longer
current, this somtimes caused incorrect framebuffers to be deleted.

BUG=angleproject:2464

Change-Id: I6859b5739a509c1343901f8c0eef1fdd1598cea1
Reviewed-on: https://chromium-review.googlesource.com/1108087Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

Bug: angleproject:2397
Change-Id: I847772aa5ce1c27f7fd232d50d39c15f10135900
Reviewed-on: https://chromium-review.googlesource.com/1108515Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

Bug: angleproject:2681
Change-Id: I728e57cee4c56da049a27c2d86a8899c44cf4917
Reviewed-on: https://chromium-review.googlesource.com/1108494Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>

Use the Vulkan format table to look up the Vulkan format for vertex data.
This will let us support more vertex formats by adding them to the table.
It also eliminates one usage of gl::VertexFormatType.
No functional change.

BUG=angleproject:2405
BUG=angleproject:2531

Change-Id: I73eb69ccac50d427de3e7d5479f92bb17c49aed3
Reviewed-on: https://chromium-review.googlesource.com/1051028Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>
Commit-Queue: Frank Henigman <fjhenigman@chromium.org>

This is a reland of 4004ae0e

Fix: Put the missing early-out in ValidatePointParameterCommon

Original change's description:
> GLES1: Point rasterization (partial implementation)
>
> - Not included: Smooth points
>
> - GL_OES_point_sprite
> - Update test expectations. Note: due to different random sampling,
> edge cases were hit in UserClip. Disabling that test for now.
>
> BUG=angleproject:2306
>
> Change-Id: If8367bc3321804b3299d3bc381d6a8e236754baa
> Reviewed-on: https://chromium-review.googlesource.com/1101910
> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
> Commit-Queue: Lingfeng Yang <lfy@google.com>

Bug: angleproject:2306
Change-Id: Id8e71352a77ff0ce71cb604965effbfb8aca613e
Reviewed-on: https://chromium-review.googlesource.com/1108458Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Lingfeng Yang <lfy@google.com>

Vulkan prefers to use a single binding for each texture array.
Previously we were reserving binding space for array elements.

Doesn't immediately fix any tests but allows more tests to
pass in subsequent CLs.

Bug: angleproject:2494
Change-Id: I238d4e6f491bfa6daf04dde24f8111c8418cd59c
Reviewed-on: https://chromium-review.googlesource.com/1101570
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>

This reverts commit 4004ae0e.

Crash in PointParameterTest.NegativeEnum/ES1_OPENGL.
https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Mac%20FYI%20GPU%20ASAN%20Release/1178
angle_end2end_tests on Intel GPU on Mac on Mac-10.12.6 
angle_end2end_tests on ATI GPU on Mac Retina on Mac-10.12.6

Original change's description:
> GLES1: Point rasterization (partial implementation)
> 
> - Not included: Smooth points
> 
> - GL_OES_point_sprite
> - Update test expectations. Note: due to different random sampling,
> edge cases were hit in UserClip. Disabling that test for now.
> 
> BUG=angleproject:2306
> 
> Change-Id: If8367bc3321804b3299d3bc381d6a8e236754baa
> Reviewed-on: https://chromium-review.googlesource.com/1101910
> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
> Commit-Queue: Lingfeng Yang <lfy@google.com>

TBR=geofflang@chromium.org,jmadill@chromium.org,cwallez@chromium.org,lfy@google.com

Change-Id: I776ce0506d349382b3af035c962aa2c3f6826b99
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: angleproject:2306
Bug: angleproject:2680
Reviewed-on: https://chromium-review.googlesource.com/1108457Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
Reviewed-by: Lingfeng Yang <lfy@google.com>
Commit-Queue: Frank Henigman <fjhenigman@chromium.org>

Updated from http://khronos.org/registry/OpenGL/api/GLES2/gl2ext.h

BUG=chromium:849576

Change-Id: I80e07fad80899b96b037a978855b70ff25bb5bd9
Reviewed-on: https://chromium-review.googlesource.com/1107494
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

Instead of passing around strings from TType::getCompleteString(),
add a stream operator to InfoSinkBase that takes a TType. This makes
the compiler executable a few kilobytes smaller and will help with
getting rid of TString altogether.

BUG=angleproject:2267
TEST=angle_unittests

Change-Id: I31a6693b40a28824b3959e19ad3c0a2ce0f0a35f
Reviewed-on: https://chromium-review.googlesource.com/1107712Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

Fixes the behavior of clearWithDraw to keep the alpha channel if
we should not be clearing it.

Bug: angleproject:2667
Change-Id: Id49c1d2ca30ecb5b5bdd8abe00825a6210cfacf6
Reviewed-on: https://chromium-review.googlesource.com/1106052Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>

This patch is the last patch of storing shader resources into
ShaderMap in Statemanager11.

This patch also splits several large functions into smaller
one to make the code structure clearer.

BUG=angleproject:2169

Change-Id: Id6d89976de0376b2479bd11d7551fc6f5b521c13
Reviewed-on: https://chromium-review.googlesource.com/1092511Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jiawei Shao <jiawei.shao@intel.com>

- Not included: Smooth points

- GL_OES_point_sprite
- Update test expectations. Note: due to different random sampling,
edge cases were hit in UserClip. Disabling that test for now.

BUG=angleproject:2306

Change-Id: If8367bc3321804b3299d3bc381d6a8e236754baa
Reviewed-on: https://chromium-review.googlesource.com/1101910Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Lingfeng Yang <lfy@google.com>