BUG=487341
Change-Id: I792e0c9419566facb0bec0ad93f3646294e5a2a7
Reviewed-on: https://chromium-review.googlesource.com/293740Reviewed-by: Nico Weber <thakis@chromium.org>
Tested-by: Nico Weber <thakis@chromium.org>

Under certain situations an integer overflow could lead to ANGLE
writing to places where it shouldn't.

BUG=518206

Change-Id: I9217685daecb160a4072fbf79c26e5bee9f4621e
Reviewed-on: https://chromium-review.googlesource.com/293200Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

The order of the scoped histogram init got clobbered from a bad
merge.

BUG=516027

Change-Id: Id5319c5a3a586b21fb64c0f161aebcfb3366722a
Reviewed-on: https://chromium-review.googlesource.com/290860Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

BUG=516027

Change-Id: Ibbf45c863570581ea79e198604cd4c5cc1836620
Reviewed-on: https://chromium-review.googlesource.com/290841Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

These macros replace a lot of repeated timing code, and mirror how
Chromium operates.

BUG=516027

Change-Id: Ib03e35d0f3b8638b9b0e8785c397c8381f16b041
Reviewed-on: https://chromium-review.googlesource.com/290840Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

When the static buffer has just been invalidated we stream from the buffer
directly for some time. This streaming forgot to offset the starting
point, causing a bug.

BUG=510585

Change-Id: I8321e462841eb0bd7ed97ab78197914cbdb68c55
Reviewed-on: https://chromium-review.googlesource.com/287525Reviewed-by: Hendrik Wagenaar <hendrikw@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/289461Tested-by: Corentin Wallez <cwallez@chromium.org>

No intended behavior change.

BUG=505317
Change-Id: I5dfa1d67715b18133f3373ca00a6d1d96b678043
Reviewed-on: https://chromium-review.googlesource.com/284850Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Nico Weber <thakis@chromium.org>

Added new construction operations for
non square matrices, along with the
required changes to the related
translation functions.

Change-Id: I04ae7d4b2d1bb363b35d088cea45c0e7c4bc8a13
Reviewed-on: https://chromium-review.googlesource.com/277729Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Geoff Lang <geofflang@chromium.org>

BUG=angleproject:1057

Change-Id: Ic52c385cb19f6ddccaf3d46a88a56670da0cf67d
Reviewed-on: https://chromium-review.googlesource.com/283987Tested-by: Cooper Partin <coopp@microsoft.com>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

This change saves us ~70KiB per compiler with just ~3KiB of cache.

BUG=492725

Change-Id: I7b8053c032c9c68f646162baf61abf6db3afe52e
Reviewed-on: https://chromium-review.googlesource.com/284003Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

Using the same design as for the Framebuffer::Data helper, we can
use a struct to share between the object and the Impl. This also
gives the Impl access to the maxEnabledAttrib, and saves some
duplicated storage.

BUG=angleproject:1040
TEST=WebGL CTS, end2end_tests, unittests

Change-Id: I55c91e8a5f3dcae302cab441182320aafd5375ef
Reviewed-on: https://chromium-review.googlesource.com/283930Tested-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

This seems to have broken some dEQP tests in debug build, possibly
uncovering a previously undetected bug.

TEST=dEQP-GLES3.functional.shaders.linkage.varying.basic_types.*

This reverts commit d2328a52.

Change-Id: I9f082d7a6fca38b4c076fecc342ac40d5416ebef
Reviewed-on: https://chromium-review.googlesource.com/284780Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Olli Etuaho <oetuaho@nvidia.com>

Uniform block members may be structs containing matrices, and there's
no explicit mention in the spec that the matrix qualifiers would only
be valid for matrix fields. Only generate a warning when an extraneous
matrix layout qualifier is used.

TEST=dEQP-GLES3.functional.shaders.*layout* (all pass with this patch)
BUG=angleproject:1058

Change-Id: Ifb2f2392d84100d997cc2d68fd55e9ad17a17cea
Reviewed-on: https://chromium-review.googlesource.com/284112Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Tested-by: Olli Etuaho <oetuaho@nvidia.com>

..\..\third_party\angle\src\libANGLE\renderer\d3d\d3d11\SwapChain11.cpp(30,7)
: error: field 'mDepthStencilRenderTarget' will be initialized after
field 'mPassThroughResourcesInit' [-Werror,-Wreorder]
      mDepthStencilRenderTarget(this, renderer, true),
      ^

While here, also reorder the fields that are initialized in the
constructor to match the order the fields are declared in, to make it
easier to see if any are missing.

No intended behavior change.

BUG=505304

Change-Id: If48195c84161868d33d4361088d3ad22b7a0fc31
Reviewed-on: https://chromium-review.googlesource.com/284682Reviewed-by: Nico Weber <thakis@chromium.org>
Tested-by: Nico Weber <thakis@chromium.org>

Unsure why this is failing:

functional.shaders.struct.local.dynamic_loop_struct_array_fragment

BUG=angleproject:1059

Change-Id: I72e05269f831769a763c44354b95a7224edb72fa
Reviewed-on: https://chromium-review.googlesource.com/284636Tested-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

Using the new vertex format type enum, we can shrink the size
of the input layout tables and reduce draw call overhead
further.

BUG=angleproject:959

Change-Id: Iea595fa64edbbd91f669138dfdeb9d2543b83929
Reviewed-on: https://chromium-review.googlesource.com/277291Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Kenneth Russell <kbr@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

Use switch statements based on the format type enum, instead of
a map from the old VertexFormat structure to the info payload.
This saves on lookup and initialization time.

BUG=angleproject:959

Change-Id: I6d1bf282c88f67272ca067cb64bc71e77824a12e
Reviewed-on: https://chromium-review.googlesource.com/277290Tested-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

This will be included in isolates for targets which really need it;
initially, angle_end2end_tests.

BUG=angleproject:892

Change-Id: I466455b0cc50a70980ddc09a6db2211f13774abc
Reviewed-on: https://chromium-review.googlesource.com/284361Tested-by: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Antoine Labour <piman@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

Do this by simply disallowing all function calls in global initializers
- all built-in math ops are handled as other ops, not function calls.

This change was tested extensively with popular WebGL content, with no
regressions found.

TEST=angle_unittests
BUG=angleproject:988

Change-Id: Id1107fa294ae4012d5dd3949539d0b7b4cd21943
Reviewed-on: https://chromium-review.googlesource.com/283703Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Tested-by: Olli Etuaho <oetuaho@nvidia.com>

BUG=angleproject:516

Change-Id: I7f969435a89c0063b504628d2ec3321cc02df7ce
Reviewed-on: https://chromium-review.googlesource.com/283243Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>

Before this patch, index buffers where always mapped in case we needed the
index data for the indexed point sprites workaround.
This patch makes it so the index buffer is only mapped in this case, when
we need to stream index data or when we need to fill the static copies of
the index buffers.

This make the memory usage of
http://alteredqualia.com/xg/examples/mammoth.html go down from 41MB to
28MB.

BUG=angleproject:516

Change-Id: I937506d06fd6f074ef2120469dbd235e20245fca
Reviewed-on: https://chromium-review.googlesource.com/283626Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>

This saves us ~15KiB per compiler instance on 32-bit Android.

BUG=492725

Change-Id: I9db2998482941b5ab2eb5cb4925cbd1fb58ab4be
Reviewed-on: https://chromium-review.googlesource.com/283584Reviewed-by: Zhenyao Mo <zmo@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

WindowTest.cpp was missing <algorithm> for std::min and std::max.

BUG=angleproject:892

Change-Id: I9bf6e7b8e66cd4ffc34ee639f495b469dabde91a
Reviewed-on: https://chromium-review.googlesource.com/284590Tested-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>

This change adds constant folding support for following derivative
functions:
    - dFdx, dFdy and fwidth.
As per spec, derivatives of constant arguments should be 0, so this
change just sets result components to zero when the above operations
are performed on constant argument.

BUG=angleproject:913
TEST=dEQP Tests
dEQP-GLES3.functional.shaders.constant_expressions.builtin_functions.
fragment_processing.*
(All 12 tests started passing with this change.)

Change-Id: I67b7c5abd970b740e15e3bd1ee56721e81bbecc4
Reviewed-on: https://chromium-review.googlesource.com/283759Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Tested-by: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

This change adds const folding support for below mentioned common
built-ins:
    - isnan, ininf, floatBitsToInt, floatBitsToUint,
      intBitsToFloat and uintBitsToFloat

BUG=angleproject:913
TEST=angle_unittests(new: MathUtilTest.isNan/inInf), dEQP Tests
dEQP-GLES3.functional.shaders.constant_expressions.builtin_functions.common.*
(20 more tests started passing with this change)

Change-Id: Ifdedb251cd8b183b4999314c0f5de857bc20702f
Reviewed-on: https://chromium-review.googlesource.com/283767Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Tested-by: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

BUG=angleproject:516

Change-Id: I907ec41ce5691300d17ca5b08a62b0d4579b01e5
Reviewed-on: https://chromium-review.googlesource.com/283244Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>

This will simplify the work on follow up CLs a lot

BUG=angleproject:516

Change-Id: Ie8374ec28fce2e01e76c711a4d280e1ed004a21f
Reviewed-on: https://chromium-review.googlesource.com/283380Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

BUG=angleproject:892

Change-Id: I35288ff43fb7a50a327047eacb8f41ca91e2b441
Reviewed-on: https://chromium-review.googlesource.com/283471Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>

This will help port OSWindow to other platforms.

BUG=angleproject:892

Change-Id: If9884d310d80421e2fc09cffd03bcf01327b1eee
Reviewed-on: https://chromium-review.googlesource.com/283900Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>

Otherwise glGetString(GL_RENDERER) returns "ANGLE" and doesn't provide
information on the vendor.

BUG=angleproject:892

Change-Id: I01899ca78656aaa8db860829c78bf3c61c355eae
Reviewed-on: https://chromium-review.googlesource.com/284051Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Corentin Wallez <cwallez@chromium.org>

Instead of storing a vertex format as a struct with the full info,
instead use an enum, and look up the info when we need it. This
saves a lot of constructor initialization time, operator comparison
time, and storage. It also will allow us to look up D3D format info
more quickly.

BUG=angleproject:959

Change-Id: I202fd1ea96981073bc1b5b232b1ec3efa91485cb
Reviewed-on: https://chromium-review.googlesource.com/277289Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

On some configs, this would use an RGBA4 16-bit backbuffer, which is
not enough resolution for the test. Fix this by requiring 32-bit
backbuffers. We might also want to figure out if we want to default
to 16-bit on Windows 8 desktop, which seems less than ideal.

BUG=angleproject:959

Change-Id: I53b3a9ef6cbecf76c7c764623e5eb1f454a6b72d
Reviewed-on: https://chromium-review.googlesource.com/283773Tested-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Austin Kinross <austin.kinross@gmail.com>
Reviewed-by: Cooper Partin <coopp@microsoft.com>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

Clang warns if there are missing braces around a subobject
initializer. The most common idiom that triggers this is:
STRUCT s = {0};
which can be more simply written as:
STRUCT s = {};

BUG=505297

Change-Id: Ie9f9037a1ba4b6303daa30424cdc24c1ecf18896
Reviewed-on: https://chromium-review.googlesource.com/284160Reviewed-by: Geoff Lang <geofflang@chromium.org>
Tested-by: Geoff Lang <geofflang@chromium.org>

This change splits TIntermConstantUnion::foldUnary into two
functions:
    - foldUnaryWithDifferentReturnType:
        - Handles constant folding of operations where the return
          type has a different number of components compared to the
          operand type.
    - foldUnaryWithSameReturnType:
        - Handles constant folding of unary operations where the
          return type is the same as operand type.

BUG=angleproject:913
TEST=angle_unittests, dEQP Tests
(dEQP-GLES3.functional.shaders.constant_expressions.builtin_functions.*)

Change-Id: I675891138d4e17fd2390c03e9f710e0be0b7c7b6
Reviewed-on: https://chromium-review.googlesource.com/283464Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Tested-by: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

This change adds constant folding support for following floating
point pack and unpack functions:
    - packSnorm2x16, unpackSnorm2x16, packUnorm2x16, unpackUnorm2x16,
      packHalf2x16, and unpackHalf2x16.

BUG=angleproject:913
TEST=angle_unittests(new: MathUtilTest.packAndUnpack*), dEQP Tests
dEQP-GLES3.functional.shaders.constant_expressions.builtin_functions.float_pack_unpack.*
(all 10 tests started passing with this change)

Change-Id: I2b69faebee4127e5e13900b3a9485b7145950277
Reviewed-on: https://chromium-review.googlesource.com/282683Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Tested-by: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>

It should not be possible to use globals as l-values in global initializers.

This change was tested extensively with popular WebGL content, with no
regressions found.

TEST=angle_unittests
BUG=angleproject:988

Change-Id: I21ab731eb1d92aeae25795856ccae280792ad1f0
Reviewed-on: https://chromium-review.googlesource.com/283910Tested-by: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Zhenyao Mo <zmo@chromium.org>

Verified the NVIDIA suppression locally; assuming the AMD suppression
will take effect on the bot. Thanks to geofflang@ and cwallez@ for the
suggestion on the approach to take.

BUG=angleproject:892

Change-Id: Ida4b6c1395185b0dfb0ea890395e35cb6db7ccd5
Reviewed-on: https://chromium-review.googlesource.com/283974Tested-by: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Brandon Jones <bajones@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>

Lots of compile errors.

BUG=492725

This reverts commit 5377720a.

Change-Id: I64889b99b1f1f48d39b87ebb668f6a32a3abac45
Reviewed-on: https://chromium-review.googlesource.com/283945Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Jamie Madill <jmadill@chromium.org>

This is to help with the FYI bot and pending a refactor to use
our test instantiation.

BUG=angleproject:892

Change-Id: I56e772d7988a2172ad2384bdf408a8eaeab7129a
Reviewed-on: https://chromium-review.googlesource.com/283971Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Cooper Partin <coopp@microsoft.com>
Tested-by: Corentin Wallez <cwallez@chromium.org>

This change saves us ~70KiB per compiler with just ~3KiB of cache.

BUG=492725

Change-Id: I2ee748f7271b1afe0834083cbff6f0b84104aa53
Reviewed-on: https://chromium-review.googlesource.com/281411Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tested-by: Dmitry Skiba <dskiba@google.com>