- 10 May, 2017 3 commits
-
-
Shao authored
This patch enables perf_test_runner.py to run on Linux and MacOS. BUG=angleproject:2029 Change-Id: Ia7176a3f5fed0245ab72d5def1199623c4fb4fe2 Reviewed-on: https://chromium-review.googlesource.com/497648Reviewed-by:
Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
-
Geoff Lang authored
We already generate an INVALID_VALUE in the ValidES3InternalFormat check above. If an error is generated here, it is because the internal format and type are not valid. TEST: conformance2/textures/misc/tex-image-with-bad-args BUG=angleproject:1958 BUG=angleproject:1715 Change-Id: I50d479e64d687cc971c8f41b5691dcb123da0391 Reviewed-on: https://chromium-review.googlesource.com/499690Reviewed-by:
Jamie Madill <jmadill@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
-
Geoff Lang authored
The generated mipmaps are still not correct on all ANGLE backends. TEST=conformance/extensions/ext-sRGB BUG=angleproject:2009 Change-Id: I6299dcb93c71f48e61ec660347b4747366ecfb71 Reviewed-on: https://chromium-review.googlesource.com/491367 Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org>
-
- 09 May, 2017 1 commit
-
-
Olli Etuaho authored
Constructor argument checking rules are reorganized to make them easier to understand and constructor node creation is made simpler. This removes usage of constructor op codes from ParseContext. This paves the way for getting rid of constructor op codes entirely, which will remove duplicate information from the AST and simplify lots of code. This refactoring will make adding arrays of arrays slightly easier. BUG=angleproject:1490 TEST=angle_unittests Change-Id: I4053afec55111b629353b4ff7cb0451c1ae3511c Reviewed-on: https://chromium-review.googlesource.com/498767 Commit-Queue: Olli Etuaho <oetuaho@nvidia.com> Reviewed-by:
Corentin Wallez <cwallez@chromium.org>
-
- 08 May, 2017 6 commits
-
-
Andrei Volykhin authored
This allows to validate fragment shader outputs for shader version #300 and newer. Change-Id: I21b21ebb357eab61e36c080c328291ed20b8eb8f Reviewed-on: https://chromium-review.googlesource.com/497749Reviewed-by:
Geoff Lang <geofflang@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
-
Geoff Lang authored
BUG=angleproject:1958 Change-Id: Ia168cd1bd48b9a443da6662efeaace1f792861f3 Reviewed-on: https://chromium-review.googlesource.com/498408Reviewed-by:
Geoff Lang <geofflang@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
-
Jamie Madill authored
This makes the entry points a bit smaller and avoids redundant return statements. BUG=angleproject:1309 Change-Id: I895c26d4235900578f5acc0eb94a7deef67b946b Reviewed-on: https://chromium-review.googlesource.com/483428 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by:
Geoff Lang <geofflang@chromium.org>
-
jchen10 authored
Add checks that opaque operands can only be used with array indexing and field section, as mentioned in ESSL 3.10 section 4.1.7. BUG=angleproject:2028 Change-Id: I41b7f10785bf712dfc999f85ebff925341c51911 Reviewed-on: https://chromium-review.googlesource.com/497767 Commit-Queue: Olli Etuaho <oetuaho@nvidia.com> Reviewed-by:
Olli Etuaho <oetuaho@nvidia.com>
-
jchen10 authored
Refactor separate sampler and image validations into unified opaque type handling. This paves way for adding atomic counter, another new opaque type. BUG=angleproject:1729 Change-Id: I201d28e31c84534db43e656d518650e378bab76c Reviewed-on: https://chromium-review.googlesource.com/493618Reviewed-by:
Olli Etuaho <oetuaho@nvidia.com> Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Commit-Queue: Corentin Wallez <cwallez@chromium.org>
-
Olli Etuaho authored
It's a shader input and as such should not be writable. BUG=angleproject:1669 TEST=angle_unittests Change-Id: I05cb5c63b7272dfa6e80cad57385da02504e4d8f Reviewed-on: https://chromium-review.googlesource.com/497408Reviewed-by:
Jamie Madill <jmadill@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
-
- 05 May, 2017 3 commits
-
-
Geoff Lang authored
BUG=angleproject:1662 Change-Id: I9bd830a3021a0cfa9010c5e58d3a284cc3eeca26 Reviewed-on: https://chromium-review.googlesource.com/490346Reviewed-by:
Geoff Lang <geofflang@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
-
Geoff Lang authored
BUG=angleproject:1668 Change-Id: I93bc7584ff7d945f9efde25c3379a282ee246b53 Reviewed-on: https://chromium-review.googlesource.com/497328Reviewed-by:
Geoff Lang <geofflang@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org>
-
Jamie Madill authored
This seems to fix a crash on D3D9/Win7/Intel. BUG=chromium:704211 Change-Id: If600a6b8409774503f28ca122d87ef723e4ea2e5 Reviewed-on: https://chromium-review.googlesource.com/497649Reviewed-by:
Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
-
- 04 May, 2017 5 commits
-
-
Jamie Madill authored
BUG=angleproject:1309 Change-Id: I7817444c3ea56f932fe769a860f4a70b29262019 Reviewed-on: https://chromium-review.googlesource.com/483427Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Reviewed-by:
Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
-
Frank Henigman authored
Make all inheritance from angle::NonCopyable private so the compiler complains about this (admittedly unlikely) code: class Foo: angle::NonCopyable { virtual ~Foo() { ... } }; angle::NonCopyable *p = new Foo; delete p; In the above code ~Foo() is not called, only ~NonCopyable(), because the latter is not virtual. Making it virtual would add overhead to all derived classes which don't already have a virtual method. Also tighten access in NonCopyable, because we can. BUG=angleproject:2026 Change-Id: Id0dc4d959cfb7bb82cf49382118129abb1d3a4f0 Reviewed-on: https://chromium-review.googlesource.com/495352Reviewed-by:Jamie Madill <jmadill@chromium.org> Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Commit-Queue: Frank Henigman <fjhenigman@chromium.org>
-
Corentin Wallez authored
This commit introduces a stamping mechanism to make sure the order of build dependencies is correct. In GYP there are the following build order dependencies (between others): - A target link depends on previous target links - A target source compile depends on that target actions outputs - An action output depends on the action inputs The problem was that there were actions in vulkan_layer_utils_static that generated headers needed by vulkan_loader's sources (between others) Making a "dependencies" relationship between the two was not enough to ensure the headers were generated before the source were compiled. Instead we replicated the "stamping" mechanism used in GN using GYP actions that depend on the headers and output a dummy stamp file. This completes the build order dependencies: the sources depend on the stamp which depend on the action's input, i.e. the generated headers. BUG=angleproject:1668 Change-Id: I2de15ac2ea8f838acdbf43da1f111b28b8cc2184 Reviewed-on: https://chromium-review.googlesource.com/493892 Commit-Queue: Corentin Wallez <cwallez@chromium.org> Reviewed-by:
Frank Henigman <fjhenigman@chromium.org>
-
Geoff Lang authored
* Expose GL_CHROMIUM_color_buffer_float_rgb and GL_CHROMIUM_color_buffer_float_rgba * Fix many texture formats that were incorrectly checking the wrong extension for support or renderability. * Make all floating point texture extensions dynamically enableable. BUG=angleproject:1958 BUG=angleproject:1715 Change-Id: Iefccc8b5ae5edd97623affa9de05b1d9af5c9598 Reviewed-on: https://chromium-review.googlesource.com/468450Reviewed-by:
Jamie Madill <jmadill@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
-
Olli Etuaho authored
The shader validation now does the same checks for qualifier combinations regardless of if a declaration is empty, as is specified. Some of these checks used to be in singleDeclarationErrorCheck, and have now been moved to a new function declarationQualifierErrorCheck. The other parts of singleDeclarationErrorCheck are under a renamed nonEmptyDeclarationErrorCheck. The patch also contains another related cleanup: Unnecessary symbol nodes won't be created for empty declarations any more. BUG=angleproject:2020 TEST=angle_unittests Change-Id: I1c864a5e151c52703926d8c550450b2561bfcbb2 Reviewed-on: https://chromium-review.googlesource.com/493227 Commit-Queue: Olli Etuaho <oetuaho@nvidia.com> Reviewed-by:
Jamie Madill <jmadill@chromium.org>
-
- 03 May, 2017 5 commits
-
-
Geoff Lang authored
BUG=angleproject:1932 Change-Id: I927eb5adc73bcede220ec56a9634783edd735d18 Reviewed-on: https://chromium-review.googlesource.com/494706Reviewed-by:
Geoff Lang <geofflang@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
-
Jamie Madill authored
Templating here was causing a lot of redundant functions to be generated. We can use a bit of macro magic instead of template magic to reduce the binary size. This reduces the size of the compile file by about 25k. Note that the ASSIGN macro doesn't itself reduce size, but it does save a bit of code size and makes the file look consistent. BUG=angleproject:2022 Change-Id: I3d9a6e1739dcab1fceca3c63d2a4e35c1e847289 Reviewed-on: https://chromium-review.googlesource.com/493891Reviewed-by:
Geoff Lang <geofflang@chromium.org> Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
-
Corentin Wallez authored
dEQP tries to test specific GL ES errors, but also triggers errors in the additional WebGL validation. This commit reoders ANGLE's validation so that WebGL-specific validation for draws is done after the common GL ES checks. BUG=angleproject:2021 Change-Id: Ie06b132ef6a4378f7c82ae8d354a8343d897c478 Reviewed-on: https://chromium-review.googlesource.com/493828 Commit-Queue: Corentin Wallez <cwallez@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org>
-
Jamie Madill authored
Apparently using std::vector::push_back causes heavy inlining in MSVS. This refactor (no functionality change) reduces the binary size from ~10kb for all the gl::Determine* functions plus the holding function (setTextureExtensionSupport) to about 1.4k. The visible size of my compiled 32-bit Release libGLESv2.dll went from 4648 to 4642 kb. Note that using PoD arrays was sinificantly better than using vectors with initializer lists (size was ~4kb instead vs 1.4kb). BUG=angleproject:2022 Change-Id: I5b7efd41455d5ee35afc699bafef31ce8367934e Reviewed-on: https://chromium-review.googlesource.com/493890Reviewed-by:
Geoff Lang <geofflang@chromium.org> Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
-
Olli Etuaho authored
TIntermTyped::CreateZero can be reached with a void array type in an error case. Handle this gracefully instead of asserting. Also remove an assert that wasn't really checking anything in CreateZero. type.isScalar() || type.isVector() || type.isMatrix() can only be false in case of a struct, and struct type was being checked in the condition on the line above. BUG=chromium:717385 TEST=angle_unittests Change-Id: Iff0811d18d399d7b32b2b46deea5df172412eb8c Reviewed-on: https://chromium-review.googlesource.com/492887Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org> Commit-Queue: Olli Etuaho <oetuaho@nvidia.com>
-
- 02 May, 2017 6 commits
-
-
Jamie Madill authored
This cleans up a few things: * refactors a few remaining missed EPs * removes unnecessary includes * rename one Context entry point * moves some ES2 EP validation to validationES2 The last item makes for a significant diff, but this is a refactor change only, with no functionality change. BUG=angleproject:747 Change-Id: I7860cc4b6260b6c22faa5f2885297333c0cdb4ed Reviewed-on: https://chromium-review.googlesource.com/483426Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
-
Geoff Lang authored
This also exposed an issue in our glTexSubImage2D validation where the sized format would be used for the texture support check when the unsized format was provided. The GL_ALPHA32F format is not valid unless EXT_texture_storage is present but using GL_ALPHA + GL_FLOAT is valid. BUG=angleproject:1958 Change-Id: Ice0b7549c39559990942176481c5175df17aaf92 Reviewed-on: https://chromium-review.googlesource.com/491246 Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org>
-
Frank Henigman authored
Vulkan's cmake files check for the availability of secure_getenv() but we just assume it exists. BUG=chromium:717558 Change-Id: I22054c3967f543d163dde37b9a6b13e896da7a50 Reviewed-on: https://chromium-review.googlesource.com/494126 Commit-Queue: Frank Henigman <fjhenigman@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org>
-
Jamie Madill authored
This will make the diff for the auto-gen patch easier to review. BUG=angleproject:1309 Change-Id: I58526a29b7f538b45409c6ab48b75b41b3521321 Reviewed-on: https://chromium-review.googlesource.com/487883 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by:
Corentin Wallez <cwallez@chromium.org>
-
Jamie Madill authored
This is a first step towards auto-generating more entry point files. BUG=angleproject:1309 Change-Id: Ic75a627feb91c19b6bb398a6f9f1a37b603dea14 Reviewed-on: https://chromium-review.googlesource.com/483425 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Reviewed-by:
Geoff Lang <geofflang@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org>
-
Corentin Wallez authored
There are both compile error and missing build file dependencies in standalone builds. BUG=angleproject:1668 Change-Id: I8741bd3dd20452c94921101bbf7c5be39fb463b7 Reviewed-on: https://chromium-review.googlesource.com/493608Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Commit-Queue: Corentin Wallez <cwallez@chromium.org>
-
- 01 May, 2017 2 commits
-
-
Corentin Wallez authored
BUG=angleproject:2017 Change-Id: Id75a8c0ddf4eda4b569509e2014c20db6f5c46ff Reviewed-on: https://chromium-review.googlesource.com/490848 Commit-Queue: Corentin Wallez <cwallez@chromium.org> Reviewed-by:
Geoff Lang <geofflang@chromium.org>
-
Kenneth Russell authored
BUG=705865 Change-Id: Ife59aaf009365a51b5362a00373d1efaceba9fb8 Reviewed-on: https://chromium-review.googlesource.com/490946 Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by:
Geoff Lang <geofflang@chromium.org>
-
- 28 Apr, 2017 5 commits
-
-
Corentin Wallez authored
The ANGLE_PARAM_TYPE_INFO macro was referencing the TypeInfo static member that was only added when release asserts are enabled. BUG=angleproject:747 BUG=chromium:716323 Change-Id: I0657dde69c62ab697b3c095fe19f38c130188cef Reviewed-on: https://chromium-review.googlesource.com/490687Reviewed-by:
Jamie Madill <jmadill@chromium.org> Commit-Queue: Corentin Wallez <cwallez@chromium.org>
-
Shahmeer Esmail authored
- ClearShader made into a class that manages all required shaders and input layouts for clears - ClearShader reuses VS for all clear types. This reduces shader compilation time and memory usage significantly - Use constantBuffer for color/z values instead of VB to decouple VB & VS from clearType and allowing for the same VS to be used for multiple clear types - FL10+ Devices: Generate positions using SV_VertexID in VS to avoid having to bind VB. - FL93 Devices: Use an immutable VB containing only position data (SV_VertexID not supported) - Implement CB cache. Incoming color/Z values checked against cache and CB/cache only updated if there is a mismatch. Significantly reduces the frequency of expensive CB map/rename operations especially in common scenarios where most/all clears use the same color/z values BUG=angleproject:1935 Change-Id: I2015fbdcc135ba08b65dbecbe9c62499c2801037 Reviewed-on: https://chromium-review.googlesource.com/453882 Commit-Queue: Shahmeer Esmail <shahmeer.esmail@intel.com> Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org>
-
Jamie Madill authored
Because our ASSERT macro always references the conditional expression, not defining TypeInfo was causing a compile error on builders which had release asserts disabled. BUG=angleproject:747 BUG=chromium:716323 Change-Id: I3b5f27e39235f26b8bef5921c2d398320eb0ebaf Reviewed-on: https://chromium-review.googlesource.com/490426 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org>
-
Shao authored
This patch adds the missing return value *this in the operator= of class BitSetT BUG=angleproject:1814 Change-Id: I758d62e80e4ade92aac64315f1808ca1eabdc787 Reviewed-on: https://chromium-review.googlesource.com/487423 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org>
-
Jamie Madill authored
This should simplify the diff with the auto-gen patch. BUG=angleproject:1309 Change-Id: Ibea503f236e7dc07d0862e20f77e866d6ec1f316 Reviewed-on: https://chromium-review.googlesource.com/489081 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by:
Jamie Madill <jmadill@chromium.org> Reviewed-by:
Corentin Wallez <cwallez@chromium.org>
-
- 27 Apr, 2017 4 commits
-
-
Geoff Lang authored
This bug showed up on Mac which doesn't support the extension. The internalFormat is correctly translated to the workaround format but the format and type are not. BUG=angleproject:1958 Change-Id: I9b31f0491df1bb623bf59996cbb02d0276b6355c Reviewed-on: https://chromium-review.googlesource.com/488582Reviewed-by:
Jamie Madill <jmadill@chromium.org> Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
-
Jamie Madill authored
Some method signatures were updated. Types like GLclampf and GLvoid were replaced with other equivalents. BUG=angleproject:1309 Change-Id: I05e8e2072c5a063d87ad96a855b907424661e680 Reviewed-on: https://chromium-review.googlesource.com/475011 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by:
Geoff Lang <geofflang@chromium.org>
-
Jamie Madill authored
Our test sets (e.g. angle_perftests) were not having the config applied. BUG=angleproject:747 Change-Id: Idd3a2cba214d38a923c1936787f9dec4bd016326 Reviewed-on: https://chromium-review.googlesource.com/487730Reviewed-by:
Corentin Wallez <cwallez@chromium.org> Reviewed-by:
Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
-
Geoff Lang authored
BUG=705865 BUG=715875 Change-Id: Ib4d83a8c9cb3f65ac8517bfd3ff7c38794dd0654 Reviewed-on: https://chromium-review.googlesource.com/488661Reviewed-by:
Geoff Lang <geofflang@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
-