Make sure that only one sequence operator is split on one iteration
of SplitSequenceOperator. This prevents multiple successive PostVisit
calls to nested sequence operator nodes from adding duplicate nodes
to the AST. The sequence operators are split starting from the
outermost one to preserve execution order.

Note that the shader translator somewhat unexpectedly generates nested
sequence operators in the AST when there is a sequence operator with
more than two operands, so this bug ended up affecting shaders in the
wild. The code around parsing sequence operators could be clarified
separately.

BUG=638313
TEST=angle_end2end_tests

Change-Id: Ic6400a484ceff0c790c2290f7b4b80980f87cd88
Reviewed-on: https://chromium-review.googlesource.com/376678Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-on: https://chromium-review.googlesource.com/387173

Add some missing functions,
modify preprocessor conditions and build files
to include some other missing functions when building for Android.

BUG=angleproject:1471

Change-Id: Iadc0a0b9fed2444b8bc9a894ee65c8b66ea7f3c9
Reviewed-on: https://chromium-review.googlesource.com/368982
Commit-Queue: Yuly Novikov <ynovikov@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

Change-Id: Ibf041e74d81ef266791e6b6a3bb43bf504b035b3
Reviewed-on: https://chromium-review.googlesource.com/374040
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

Change-Id: Ib1f911d1a30921e4d74d405b77f930754a647d17
Reviewed-on: https://chromium-review.googlesource.com/374048Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

ANGLE has moved onto the Windows 10 SDK, and is using new compiler
features that aren't supported by the old Windows 8.1 toolchain. Support
for the 8.1 projects has been broken in master for some time now.

Since more and more developers are moving towards Windows 10, we
are going to take a snapshot of ANGLE that includes 8.1 support and
freeze it in a branch on github.com/microsoft/angle. If developers
wish to compile for use ANGLE in 8.1 apps then they should use that
branch going forward.

Change-Id: Ifee2d8a8cc0332500e0bd338911d831e7624fa82
Reviewed-on: https://chromium-review.googlesource.com/374039
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

These cases were unreachable.

BUG=angleproject:1455

Change-Id: I5f3b969de3aca3862944ac9ed510c9f5e9fd77e2
Reviewed-on: https://chromium-review.googlesource.com/367695
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

The indexing of the table data was mismatched with the table,
causing confusion with no appreciable gain. Fix this, reducing
the lines in the script and table at the same time.

BUG=angleproject:1455

Change-Id: I59095ba11a416575fd2310765eaac8d368b90cab
Reviewed-on: https://chromium-review.googlesource.com/367694Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

The 'LoadToNative' copy is the only one that should be considered a
non-conversion. For depth formats, we should be clamping on SetData
and CopyImage calls, so this changes the flags for two of the
depth formats, but won't affect behaviour since the conversion bit
is never check for depth/stencil formats.

This allows us to remove the 'requiresConversion' bit from the data.

BUG=angleproject:1455

Change-Id: I739ecc54406d65049caad7851ffd51d7cb777946
Reviewed-on: https://chromium-review.googlesource.com/367693Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

This default format mapping serves as a base for a Renderer format
mapping. Any formats that require emulation or other representation
in a Renderer can be overridden in the renderer layer.

BUG=angleproject:1455

Change-Id: I47f070c2ea132bada9cc0fc4353b89b74b8ee8f3
Reviewed-on: https://chromium-review.googlesource.com/367692Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

BUG=None

Change-Id: I3b697c3b721b5be4888d039a5cefd160ab5b23f8
Reviewed-on: https://chromium-review.googlesource.com/367877Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

This is a followup CL of
https://chromium-review.googlesource.com/366720. Unittests is added to
check invariant status does not leak across shaders.

This CL also moves mInvariantVaryings and mGlobalInvariant from
TSymbolTable to TSymbolTableLevel. So at the end of a compilation, the
levels pop, and the settings will be cleared and will not affect the
next compilation.

Change-Id: I1199fade7a637276ab149ab9a599621b9977298b
Reviewed-on: https://chromium-review.googlesource.com/370844
Commit-Queue: Zhenyao Mo <zmo@chromium.org>
Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Reviewed-by: Kenneth Russell <kbr@chromium.org>

1. Simplify code by using asserts instead of adding internal errors to
log.

2. Add a TIntermBinary constructor that takes left and right operand
nodes as parameters.

3. Remove TIntermediate functions with trivial functionality.

BUG=angleproject:952
TEST=angle_unittests

Change-Id: I2e0e52160c9377d8efcf15f14fd59f01cb41bd83
Reviewed-on: https://chromium-review.googlesource.com/372720
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

This will make it easier to generate errors and warnings with a
consistent format outside of ParseContext, for example in
TIntermBinary::fold() which warns about division by zero.

BUG=angleproject:952
TEST=angle_unittests

Change-Id: I25999d7bdc77efafe77785a0d6f4d63417dfaab9
Reviewed-on: https://chromium-review.googlesource.com/372719Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

The shader translator used to accept some invalid struct operations,
like struct += struct and struct == struct with a different type. Fix
this.

BUG=angleproject:1476
TEST=angle_unittests

Change-Id: Ia2303fc1f740da4d78242e094ee6004b07364973
Reviewed-on: https://chromium-review.googlesource.com/372718Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

This is implemented as a compiler option which is enabled by default
when outputting to desktop GLSL version 130 and greater, which does
not support this #pragma in fragment shaders. As a workaround, and for
better compatibility on desktop OpenGL drivers, this pragma is also
flattened into the outputs of vertex shaders, and the inputs of ESSL
1.00 fragment shaders.

TEST=conformance/glsl/misc/shaders-with-invariance.html with --enable-unsafe-es3-apis
BUG=629622, angleproject:1293

Change-Id: Ib040230915e639971505ed496d26e804c9d64e68
Reviewed-on: https://chromium-review.googlesource.com/361792Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Commit-Queue: Kenneth Russell <kbr@chromium.org>

This CL was causing inverted rendering in a WebGL application.

This reverts commit 7da98506.

Vectors or matrices that are dynamically indexed as a part of an
l-value generate new statements in the RemoveDynamicIndexing AST
transformation step. SplitSequenceOperator needs to detect this case
and split the sequence operator before statements are generated from
its operands to ensure the correct order of execution.

BUG=angleproject:1341
TEST=angle_end2end_tests

Change-Id: I854f8cce2d46107afa62f48edf3d32c6d5c97eda
Reviewed-on: https://chromium-review.googlesource.com/371643Reviewed-by: Kenneth Russell <kbr@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

A slight bug in the problem case detection applied the workaround
overly-broadly. Also included a much more thorough test.

BUG=angleproject:851

Change-Id: I5c09d67eee3622c144401769af85958f23b60c57
Reviewed-on: https://chromium-review.googlesource.com/371380Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

These failures were exposed with the new R7 240 card.

BUG=chromium:637037

Change-Id: I1fe5add2fc08d5c93e338250efb2af62712d0e53
Reviewed-on: https://chromium-review.googlesource.com/371318Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Kenneth Russell <kbr@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

Instead of returning false when a check succeeds in ParseContext,
return true. This is more intuitive and in line with conventions used
elsewhere inside ANGLE.

Also includes some minor other cleanup in ParseContext, like improved
variable names and code structure especially when checking array
element properties. This will make introducing arrays of arrays easier
in the future.

BUG=angleproject:911
TEST=angle_unittests

Change-Id: I68233c01ccfbfef9529341af588f615efc2b503a
Reviewed-on: https://chromium-review.googlesource.com/371238Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

Change-Id: I2a35899bffd28767b8bc638f415857636d4ad8ef
Reviewed-on: https://chromium-review.googlesource.com/366720
Commit-Queue: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Kenneth Russell <kbr@chromium.org>

enabled.

BUG=angleproject:1467,635433
TEST=WebGL conformance, angle_unittests

Change-Id: I9eb4ce715732087a3786da886f42243716f2b9b2
Reviewed-on: https://chromium-review.googlesource.com/367532
Commit-Queue: Zhenyao Mo <zmo@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

Most error checking functions in ParseContext used to follow a format
like <property>ErrorCheck. Sometimes this function would check that
the node/value would have <property>, sometimes it would check that
the node/value would not have it, which was confusing. Change most of
these functions to use a lot more descriptive names, which clearly
communicate what they are checking for.

Also includes a bit of refactoring in constructorErrorCheck(), so that
the function only checks for errors rather than also setting the type
of the constructor node.

Also make TType::arraySize unsigned, and return a sanitized size from
checkIsValidArraySize() instead of using an output parameter.

BUG=angleproject:911
TEST=angle_unittests

Change-Id: Id9767b8c79594ad3f782f801ea68eb96df721a31
Reviewed-on: https://chromium-review.googlesource.com/367070Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

These cases weren't actually ever hit within ANGLE.

BUG=angleproject:1455

Change-Id: I60da3fb9eae91598ddaed5572a4bab0a66345694
Reviewed-on: https://chromium-review.googlesource.com/367691Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

These tables were duplicated in D3D11 and D3D9, and would have to be
further duplicated into Vulkan.

BUG=angleproject:1455

Change-Id: Ice1b81417d7b14f933b61861c4a9997c260ef72e
Reviewed-on: https://chromium-review.googlesource.com/367690Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

ETC1 data doesn't need to be fully decoded before encoding to BC1. Saves
~50% computation.

BUG=angleproject:1285

Change-Id: I0cb479a08159ccd0472e5fdf40dd84323151883c
Reviewed-on: https://chromium-review.googlesource.com/361553
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

BUG=angleproject:1455

Change-Id: I896b3ed2d5e4ff3ad72de9a3a4b554841129e4e9
Reviewed-on: https://chromium-review.googlesource.com/367093Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

BUG=angleproject:1455

Change-Id: I7fc2640ec9d73528feb526b4d8b89cabccda4d17
Reviewed-on: https://chromium-review.googlesource.com/367092Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

d3d11::TextureFormat and d3d11::FormatSet don't really do different
things, so it would be best to merge them.

BUG=angleproject:1455

Change-Id: I8c09ae763e4cb284bb611fed9d74b9048fb9f13e
Reviewed-on: https://chromium-review.googlesource.com/367091Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

Mention our good ES 3.0 support. Also mention our multiplatform and
Vulkan support.

BUG=None

Change-Id: Ib2a3f94b56ecf3a3b20b960b3270ebbc8948a628
Reviewed-on: https://chromium-review.googlesource.com/367755
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Shannon Woods <shannonwoods@chromium.org>

The only piece of data left in the d3d11::TextureFormat is the
initializer function, which could just as easily be in the
ANGLEFormatSet.

BUG=angleproject:1455

Change-Id: Iee8eb84f71807575c2a46bf4fcbf283aa1b7f565
Reviewed-on: https://chromium-review.googlesource.com/367090Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

BUG=angleproject:1455

Change-Id: I263719cc77ff80580a551683d062e862dee1bdab
Reviewed-on: https://chromium-review.googlesource.com/365826Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

This will unify the way we call into PackPixels and also cleans u
the code. It required adding a few D3D9-specific formats to the table.

BUG=angleproject:1455

Change-Id: Ic8282f40f52cabdb2925e2bb305c26582284082d
Reviewed-on: https://chromium-review.googlesource.com/365825
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Zhenyao Mo <zmo@chromium.org>

Some bits of information, like the copy functions and mipmap gen
functions, can be shared across back-ends in the angle::Format
class.

Also name the info struct angle::Format, and use an enum class
angle::Format::ID to identify the particular format.

This patch introduces a new table generator for angle formats
and updates the D3D11 generator accordingly.

BUG=angleproject:1455

Change-Id: I13b8b98822b1186c6a9e436dc232c18fef50980c
Reviewed-on: https://chromium-review.googlesource.com/365824Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

This general enum can encapsulate different formats between
GL/Vulkan/D3D9/D3D11/etc so we can use them in common routines like
PackPixels. It also can help us get rid of the ANGLEX enums which
we use to represent internal formats not present in GL.

It is currently used for Textures/Renderbuffers/Surfaces, but can
also be extended in the future to cover things like vertex formats.
It mirrors something like a DXGI_FORMAT in D3D11 or VkFormat.

BUG=angleproject:1455

Change-Id: I467d7b36d8fc92bb45239d56b9243d06f4e29378
Reviewed-on: https://chromium-review.googlesource.com/365413Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

This call is a no-op. The shader parser is intended to almost always
recover from errors, so including it doesn't clarify the code either.
It's simpler to remove it entirely.

BUG=angleproject:911
TEST=angle_unittests

Change-Id: I0feae097c2807c8e9559672e7a3d50a2fc4fbdea
Reviewed-on: https://chromium-review.googlesource.com/367040
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

Support is added for the compute shader special variables given in
OpenGL GLSL ES 3.1 Revision 4, 7.1.3 Compute Shader Special Variables.
Unit tests are added for legal and illegal usage of the special
variables.

BUG=angleproject:1442

TEST=angle_unittests

Change-Id: Idb25811c15c4044c55c611c0e73ef26eb5b3e9d7
Reviewed-on: https://chromium-review.googlesource.com/366661
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

Support is added for compute shader compilation. There is a small
extension to the parser so that 'local_size_x = ', 'local_size_y = '
and 'local_size_z = ' are supported as layout qualifiers.

A few shader compilation tests are added and one which checks the AST
whether the layout qualifiers are properly parsed.

BUG=angleproject:1442

TEST=angle_unittests
TEST=angle_end2end_tests

Change-Id: I67283797d1cf13fa4ac47faa2a6e66d93a2db867
Reviewed-on: https://chromium-review.googlesource.com/362300Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

This CL stores the d3d11::TextureFormat by reference in the texture
storage. Adding the internalFormat to the TextureFormat allows us
to store a single ref instead of three per TextureStorage11.

Also store the format sets in a d3d11::TextureFormat by-ref instead
of by-pointer, making the code a bit cleaner.

BUG=angleproject:1455

Change-Id: I3c0e966d948c694435577d7d45dc0cd156480cdb
Reviewed-on: https://chromium-review.googlesource.com/365412Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

Previously we were just dumping any unexpected errors here. Fix this
by passing in the RT explicitly in the constructor, although we still
store the EGLImage pointer in case the RT changes.

BUG=angleproject:1455

Change-Id: Ie4d4a07c3a7b3a5ccc0a43c4e80bfc6ee43a676b
Reviewed-on: https://chromium-review.googlesource.com/365411Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

It's better to maintain the code outside of a generator script. It
gives easier formatting and updating.

BUG=angleproject:1455

Change-Id: I2b4383f1ed01545004de10024d03879201e2bf41
Reviewed-on: https://chromium-review.googlesource.com/365410Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>