This reverts commit dc7bffd0.

Reason for revert: Causes a memory leak, detected by ASAN bot:

https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_asan_rel_ng/builds/494713

Direct leak of 24 byte(s) in 1 object(s) allocated from:
    #0 0x847aa2 in operator new(unsigned long) /b/build/slave/linux_upload_clang/build/src/third_party/llvm/compiler-rt/lib/asan/asan_new_delete.cc:92:3
    #1 0x193a833 in sh::TType::buildMangledName() const third_party/angle/src/compiler/translator/Types.cpp:545:21
    #2 0x193d2e8 in getMangledName third_party/angle/src/compiler/translator/Types.cpp:751:24
    #3 0x193d2e8 in sh::TType::realize() third_party/angle/src/compiler/translator/Types.cpp:759
    #4 0x1834474 in sh::TCache::getType(sh::TBasicType, sh::TPrecision, sh::TQualifier, unsigned char, unsigned char) third_party/angle/src/compiler/translator/Cache.cpp:89:11
    #5 0x1859ac7 in getType third_party/angle/src/compiler/translator/Cache.h:36:16
    #6 0x1859ac7 in sh::InsertBuiltInFunctions(unsigned int, ShShaderSpec, ShBuiltInResources const&, sh::TSymbolTable&) third_party/angle/src/compiler/translator/Initialize.cpp:28

Bug: angleproject:1432

Original change's description:
> Make TType store a const char * for mangled name.
> 
> We would only ever use the c_str value from the mangled name. This
> makes it easier to make constexpr TTypes.
> 
> Bug: angleproject:1432
> Change-Id: I147b3a85f9b8b2453e2d7f4a713d767b22036cc9
> Reviewed-on: https://chromium-review.googlesource.com/776277
> Commit-Queue: Jamie Madill <jmadill@chromium.org>
> Reviewed-by: Kai Ninomiya <kainino@chromium.org>

TBR=jmadill@chromium.org,kainino@chromium.org

# Not skipping CQ checks because original CL landed > 1 day ago.

Bug: angleproject:1432
Change-Id: Ib112a2ce9871a4f4afc53101ac1a3ddd166008cf
Reviewed-on: https://chromium-review.googlesource.com/780420Reviewed-by: Jamie Madill <jmadill@chromium.org>

Move all packing related logic from Program to VaryingPacking.
Do linkValidateTransformFeedback check first to simplify packing.
Make getPackedVaryings collect all required VaryingPackings in one
run pass, so that packUserVaryings can focus on packing merely.

BUG=angleproject:2241
Change-Id: I5e84e6e4688c80ba7a77a265b8ffaf79f4ddbb87
Reviewed-on: https://chromium-review.googlesource.com/764821Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

This makes it a bit simpler to work with the prepareIndexData method.
Also removes the mRendererClass member of the index data manager
since it was no longer necessary.

BUG=angleproject:2229

Change-Id: I03768b104da21f57c499239bbbb5f716efb0fd48
Reviewed-on: https://chromium-review.googlesource.com/765261
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

Some places would compare with "== GL_TRUE" and others with "!= GL_FALSE".
This behaviour is not in the OpenGL spec but "!= GL_FALSE" is the most
standard and follows the same rules as C and C++.

Remove un-necessary validation that params are either GL_TRUE or
GL_FALSE.

Update some internal storage from GLboolean to bool.

BUG=angleproject:2258

Change-Id: I12adbe2d24318a206521ca6ad1099ee7e2bf677e
Reviewed-on: https://chromium-review.googlesource.com/779799Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

These replace the ::create methods, to facilitate constexpr versions.
Also adds a constructor of WorkGroupSize that can be made to be
constexpr.

Bug: angleproject:1432
Change-Id: I413d1a56102c94d6a383775110a2b850a7532ea0
Reviewed-on: https://chromium-review.googlesource.com/776279
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>

By sizing the vector to at least one element, it is not possible to pass
invalid pointers to the driver or binary stream even if the binary is
zero-sized.

BUG=angleproject:2257

Change-Id: Ie0e42bff1192207e0e069934b03dfd49e8d34824
Reviewed-on: https://chromium-review.googlesource.com/779739Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

These were actually always required, but only was caught by the newer
versions of the SPIR-V toolchain.

BUG=angleproject:2237

Change-Id: I43fef179e8792e46a602b39a6decafcab03b19df
Reviewed-on: https://chromium-review.googlesource.com/760638Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

It is possible for some hardware to not support multisampling on any
texture formats which results in an empty sample set for all formats.

BUG=angleproject:2242

Change-Id: I8148c781ec22ec02562b6a8b5d1b966c6f13915e
Reviewed-on: https://chromium-review.googlesource.com/766568Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

Permits depth-stencil attachment points for WebGL.

BUG=angleproject:2090

Change-Id: I7f5a7c63f2a4a76116ce5639833e5fd8d7f50ffb
Reviewed-on: https://chromium-review.googlesource.com/764591Reviewed-by: Bryan Bernhart <bryan.bernhart@intel.com>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

Warnings about implicit type conversions were showing up as errors on
MSVS 2017.

TEST=build with MSVS 2017

Change-Id: I5f173a6838da16e96fad134756d31f56890c147b
Reviewed-on: https://chromium-review.googlesource.com/778841
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

We would only ever use the c_str value from the mangled name. This
makes it easier to make constexpr TTypes.

Bug: angleproject:1432
Change-Id: I147b3a85f9b8b2453e2d7f4a713d767b22036cc9
Reviewed-on: https://chromium-review.googlesource.com/776277
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>

This is a reland of 3f286cd1
Original change's description:
> Clean up the TType class.
>
> Move more methods into the cpp file, and rename member variables to
> start with the "m" prefix.
>
> Also move most of the TPublicType methods into the cpp.
>
> Bug: angleproject:1432
> Change-Id: Ib11a3c8c6ace654fd52077a317814665f81a7261
> Reviewed-on: https://chromium-review.googlesource.com/776276
> Reviewed-by: Kai Ninomiya <kainino@chromium.org>
> Commit-Queue: Jamie Madill <jmadill@chromium.org>

Bug: angleproject:1432
Bug: chromium:786603
Change-Id: I7f91557d9f9065f4aa90130ce30dc27e9f796065
Reviewed-on: https://chromium-review.googlesource.com/777715
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

Now SimpleStateChangeTest.RedefineBufferInUse can pass on Win/Intel
with https://chromium-review.googlesource.com/756959

BUG=angleproject:2221
TEST=angle_end2end_test

Change-Id: Ibdb619a61fcaa35c72ff5a22dc6905edf9f0ec7b
Reviewed-on: https://chromium-review.googlesource.com/774679
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

This reverts commit 3f286cd1.

Reason for revert: crbug.com/786603

Original change's description:
> Clean up the TType class.
> 
> Move more methods into the cpp file, and rename member variables to
> start with the "m" prefix.
> 
> Also move most of the TPublicType methods into the cpp.
> 
> Bug: angleproject:1432
> Change-Id: Ib11a3c8c6ace654fd52077a317814665f81a7261
> Reviewed-on: https://chromium-review.googlesource.com/776276
> Reviewed-by: Kai Ninomiya <kainino@chromium.org>
> Commit-Queue: Jamie Madill <jmadill@chromium.org>

TBR=jmadill@chromium.org,oetuaho@nvidia.com,kainino@chromium.org

Change-Id: Ib3062a71118095a3e58e75bfee2474c2e4ae9431
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: angleproject:1432
Reviewed-on: https://chromium-review.googlesource.com/777801Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Commit-Queue: Zhenyao Mo <zmo@chromium.org>

Move more methods into the cpp file, and rename member variables to
start with the "m" prefix.

Also move most of the TPublicType methods into the cpp.

Bug: angleproject:1432
Change-Id: Ib11a3c8c6ace654fd52077a317814665f81a7261
Reviewed-on: https://chromium-review.googlesource.com/776276Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

Unreferenced local and global variables are now pruned from the AST.
They will be removed unless their initializer has side effects.

The CollectVariables step needs to be run after the pruning, as the
pruning may affect which interface variables are statically used.

It's also good to gather built-ins that need to be emulated after the
pruning, so unnecessary built-in emulation functions are not added to
the translator output.

This will help handle some dEQP tests for arrays of arrays that have
extremely large local arrays that are only used in an array length
query. By constant folding the length and pruning unused variables we
will avoid adding a large amount of array initialization code to the
generated shaders.

BUG=angleproject:2166
TEST=angle_unittests, angle_end2end_tests

Change-Id: Ic918bfe8f16460bcd6101d73a7a674145f5aeecd
Reviewed-on: https://chromium-review.googlesource.com/766434
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

Previously, creating nodes for initializing each single array element
could result in memory bloat during translation when dealing with
large arrays. The resulting shader could also end up very long.
Initialize most arrays using a simple for loop instead. The loop is
compatible with ESSL 1.00 Appendix A limitations.

An exception is made for fragment outputs, so that they are not
indexed by non-constant values.

On some platforms using the a loop to initialize variables can cause
problems, so we also have a compiler flag for turning this behavior
off. The flag was already added earlier for a staggered rollout of
this functionality.

BUG=chromium:735497
TEST=angle_unittests, angle_end2end_tests, WebGL conformance tests

Change-Id: Iec727821d8137db56b440ddbe007879b1b55f61f
Reviewed-on: https://chromium-review.googlesource.com/707195
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

The traverser must avoid inserting two statements to the same position
on a single traversal, so it doesn't trigger an assert.

BUG=chromium:784078
TEST=angle_unittests

Change-Id: I855054e62cc1b1cf4e6bb02af527954151c7d0e7
Reviewed-on: https://chromium-review.googlesource.com/771611
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

This fixes a very odd use case where an app would try to copy from an
uninitialized buffer. I didn't search the spec too closely, but it's
likely a valid operation that produces undefined buffer contents.
Previously to this change we would genearte an OOM error.

Also includes an unrelated fix to ensure the latest buffer storage is
never nullptr when we have any data.

Bug: angleproject:1155
Change-Id: I4292bd302cc2b84d125a7d3e8d28e4d2b0210e53
Reviewed-on: https://chromium-review.googlesource.com/774991Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

BUG=angleproject:2255

Change-Id: I986ee0e674d0f3fc2425c922ca92a3f8d86efb8c
Reviewed-on: https://chromium-review.googlesource.com/775816Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Commit-Queue: Yuly Novikov <ynovikov@chromium.org>

As a follow up to commit "8d6af086 Use dedicated ALPHA passthrough shader",
this change adds a regression test that verifies RGB channels are set to
zero when performing an ALPHA copy.

BUG=angleproject:2239
Change-Id: Icc08deb523c388a0f2cbdcadd1eec69a97fbe136
Reviewed-on: https://chromium-review.googlesource.com/774579
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

This avoids the possibility of initializing globals twice in Chromium.
Now we also never initialize variables redundantly in case we're
writing HLSL output. This was already the intent of the code before,
but a mistake had slipped in and the code didn't actually check the
output type properly.

This also simplifies DeferGlobalInitializers by running it after
SeparateDeclarations.

BUG=chromium:735497
TEST=WebGL conformance tests

Change-Id: I95036a24ac8cf18113755510376a2fca286b3ee6
Reviewed-on: https://chromium-review.googlesource.com/771555Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

The code path in the D3D11 back-end is sigificantly different for
unsigned short indices vs unsigned int. Because of the workaround we
have for the primitive restart index in D3D11, ushort rendering can
be slower.

BUG=angleproject:2229

Change-Id: I303dcc55b0314ec45508044995ba47b250cbb87d
Reviewed-on: https://chromium-review.googlesource.com/767149Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

Also force ANGLE_SKIP_TEST_IF() to require a semicolon.

BUG=angleproject:2220
TEST=angle_end2end_tests

Change-Id: I2dc777ec75babb8fbb1a2dd8949636c5c05c5767
Reviewed-on: https://chromium-review.googlesource.com/754337
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

BUG=angleproject:1442
TEST=angle_unittests, angle_end2end_tests
     dEQP-GLES31.functional.compute.shared_var.atomic*
     dEQP-GLES31.functional.compute.basic.shared_atomic_op*
     dEQP-GLES31.functional.ssbo.atomic.*

Change-Id: I82b54fde3a852d3bd917b1e19680baa1c28fce4d
Reviewed-on: https://chromium-review.googlesource.com/765061
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

This will require a Chrome-side patch to the build/ folder to enable
the new path, with appropriate Chrome-side changes.

Bug: chromium:727437
Change-Id: Iae8d8f4caeead703b322abce75c97c7c9d9e1619
Reviewed-on: https://chromium-review.googlesource.com/773021Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

BUG=anglebug:2253

Change-Id: I4c2fc6904f88a6d61c5d847b812a4272a99c1906
Reviewed-on: https://chromium-review.googlesource.com/773063Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Commit-Queue: Yuly Novikov <ynovikov@chromium.org>

None of these deps are used in the Chromium builds, so this only
affects the standalone version.

Includes a change in the build repo necessary to enable fixing the
Clang build with dEQP.

Bug: chromium:727437
Change-Id: Iea5f6299e9f17555891c812229a54a850776f117
Reviewed-on: https://chromium-review.googlesource.com/773020Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

This uses the existing dirty bit mDirtyImages to early return in the
TextureD3D::updateStorage methods.

Also fixes a couple missing cases for setting dirty images. The code
could be improved in the future by using a bitset of dirty images per
mip level instead of a single bit.

BUG=angleproject:1392

Change-Id: I58df2be455b6ee51c3e3f76a6fa44ad585d1480d
Reviewed-on: https://chromium-review.googlesource.com/761243Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

This prevents possible size count overflow when in 32-bit mode with
a lot of very large resources.

Issue originally reported via Mozilla:
https://bugzilla.mozilla.org/show_bug.cgi?id=1412554

BUG=angleproject:2234

Change-Id: Idacfda98356c13bf22bc493ef7aa2bff1a287f97
Reviewed-on: https://chromium-review.googlesource.com/769749
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

The newer validation layers picked up this missing bit.

BUG=angleproject:2237

Change-Id: I5bb2f1db3a4e284c003de0f5b23f7f613056af3a
Reviewed-on: https://chromium-review.googlesource.com/760637
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Frank Henigman <fjhenigman@chromium.org>

This will be needed outside of InitializeVariables when removing
unreferenced variables is added. Named struct type declarations
cannot be pruned as easily as they might be referenced later.

BUG=angleproject:2166
TEST=angle_unittests

Change-Id: If1462abe67e62ae19bde97de3c8f3d15e99ae9ea
Reviewed-on: https://chromium-review.googlesource.com/771790Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

NVIDIA 387.* drivers do not reproduce the issue anymore, so the test
expectations for the previously failing indirect draw tests can be
lifted.

BUG=angleproject:1659
TEST=dEQP-GLES31.functional.draw_indirect.random.*

Change-Id: Iba69821d19bfad8d2f2e81045658f3ef08695c80
Reviewed-on: https://chromium-review.googlesource.com/744021
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

When using a transfer destination, we should be using the transfer
access mask, not bottom of pipe or other bits.

Uncovered by using a newer version of the layers SDK.

BUG=angleproject:2237

Change-Id: I3ebcca707ac8889d2ece099e3765f5b0ca3dfdbc
Reviewed-on: https://chromium-review.googlesource.com/760636Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

BUG=angleproject:1155

Change-Id: I7bc5c196d10ae040487ecd1137d9b18c3000775a
Reviewed-on: https://chromium-review.googlesource.com/769730Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

This is too difficult to implement without mocking new/delete.

BUG=angleproject:875

Change-Id: I7cdafe11d26939f14a43bc515454dd5cf70c4a2a
Reviewed-on: https://chromium-review.googlesource.com/769849
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

From the spec (2.3.1 Data Conversion For State-Setting Commands):
"Validation of values performed by state-setting commands is performed
after conversion, unless specified otherwise for a specific command."

BUG=783574

Change-Id: I9edf585a17489ad284bc85a1c3c2236b53ee34d9
Reviewed-on: https://chromium-review.googlesource.com/766569Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>

The shader translator is intended to prune unreferenced variables in
the future. To maintain test coverage when this is done, most tests
that used to have unreferenced GLSL locals and globals are edited to
use built-ins or reference the necessary variables instead.

BUG=angleproject:2166
TEST=angle_unittests, angle_end2end_tests

Change-Id: I3964e7a80d52fc04a95f57e73da6095e433095e8
Reviewed-on: https://chromium-review.googlesource.com/768740Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>

This should improve ANGLE startup time by avoiding creating a large
std::map filled with the entry points.

BUG=chromium:781460

Change-Id: I20cfdb10b99844d0f60759dda73b729991dc60fe
Reviewed-on: https://chromium-review.googlesource.com/768209Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Antoine Labour <piman@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>

BUG=angleproject:1523

Change-Id: I5d6df35d2e65be6d73ec6100e3351ba5f9ff53a2
Reviewed-on: https://chromium-review.googlesource.com/688639
Commit-Queue: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>