kc3-lang/angle/src/compiler

src/compiler

Log

Author Commit Date CI Message
Shahbaz Youssefi 2aa4f7e5 2021-01-08T16:35:29 Fix translator fuzzer Since some ASTs are now disabled on certain conditions, this change lets the fuzzer know when it has generated invalid translator options. Bug: chromium:1164448 Change-Id: I1f1b120c33fb70f9776df858db033ec914d7ad89 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2618203 Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi e2a8a69a 2021-01-04T23:52:02 Vulkan: Move xfb emulation offset calculation to translator A new function is added by the translator to calculate the transform feedback offsets in the emulation path. This function makes the generated code for transform feedback smaller. Bug: angleproject:3606 Change-Id: I01460f907e20e2887cb720bddad96697fdcb0cf3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2607492 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi 5e280545 2021-01-06T14:13:29 Move RegenerateStructNames AST transformation to gl/ This AST transformation is used by the chromium validating command decoder on Linux/AMD/GL as well as MacOS. Bug: chromium:1161513 Change-Id: I67e40215ae5576819ba5a58987ceb1776c86c731 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2613195 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Shahbaz Youssefi 382bf288 2020-12-24T23:56:39 Organize AST transforms per backend Most of the AST transforms are written as a workaround to an issue that affects a single backend. This change identifies such transforms and organizes them by backend. They are then only built if the respective backend is. Additionally, about half of the GL transforms are due to mac workarounds, including the large RewriteRowMajorMatrices transform. Mac-specific workarounds are additionally only built on said platform. This change reduces the ANGLE binary size: - 106KB in a Vulkan-only build on Linux - 27KB in a GL-only build on Android (60KB on Linux) Bug: chromium:1084580 Bug: chromium:1161513 Change-Id: I64b334332c0d4f848756c6538af0d8d96864c7e9 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2601346 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Shahbaz Youssefi 8065aa82 2020-12-22T23:38:42 Front-end support for xfb capture of I/O block members Validation and generation of transform feedback varyings that specify an I/O block member are implemented in this change. The GL backend is able to pass the added tests. Bug: angleproject:3606 Change-Id: I66d02bed8ca9161555d0d1e7a32ae9ef4d9e813f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2599952 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Charlie Lao 64508f44 2020-12-12T18:00:10 Vulkan: Reduce driver uniform data to minimum if specConst is used If specialization constant is used, driver uniform data structure should be reduced to minimum to increase cache locality. Bug: b/175479076 Change-Id: I1fc50666542c6763c60bfe011cde5bc77ccc08e1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2588549 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Mohan Maiya 8f9e17d5 2020-12-14T14:11:11 Add EXT_clip_cull_distance extension autogen code Addition of shader autogen code for EXT_clip_cull_distance extension. Bug: angleproject:5458 Change-Id: I299528957c0e747fd84987c420588c314f54aae0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2590989 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 8326b26a 2020-12-11T13:59:40 Fix link validation with ambiguous instanceless interface blocks The following interface blocks should fail link: VS: layout(binding=0) buffer BufferBlockNameA { mediump float variable; }; FS: layout(binding=1) buffer BufferBlockNameB { mediump float variable; }; Because `variable` is ambiguous. Bug: angleproject:3580 Change-Id: I29576a6f152780819af0e9fb63249dbee7d9f2fc Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2587450 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao c75473c2 2020-11-23T14:26:32 Vulkan: Generalize FlipRotationSpecConst to SpecializationConstant Specialization constant are used not just for flip/rotation. It also used for other things. This CL merges all specialization constant usage (lineRasterEmulation, flip, rotation, halfRenderArea) into one class and rename FlipRotationSpecConst to SpecConst. Bug: b/173800146 Change-Id: I8dc3354b6caedbb183cec29855fc1c301ec8872a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2555812 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Jamie Madill f6df8692 2020-12-09T12:46:35 Vulkan: Support XFB in non-Vertex stages. This updates the code in several places to support Geometry and Tessellation Evaluation shaders to output transform feedback. Does not turn on any new tests but enables support for XFB when we turn on GS/TS. Bug: angleproject:3571 Bug: angleproject:3572 Change-Id: I6dcb768f2df4eeee81a4a500e999fcf16716f58f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2581941 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Charlie Lao 067a12f0 2020-11-19T16:47:25 Reland "Vulkan: Use specialization constant for halfRenderAreaWidth" This is a reland of ff38106d62fca2e36ec2db1cd580b546fd53ae43 Original change's description: > Vulkan: Use specialization constant for halfRenderAreaWidth > > halfRenderArea is used by fragment shader to adjust gl_FragCoord for > rotation and yflip compensation. This CL bakes halfRenderAreaWidth into > shader via specialization constant, thus allow compiler to consolidate > the calculation into one MAD instruction. > > Bug: b/173800146 > Change-Id: Id66301278e3389e2582369b695825e632bccecee > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2551541 > Commit-Queue: Charlie Lao <cclao@google.com> > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Reviewed-by: Tim Van Patten <timvp@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> Bug: b/173800146 Change-Id: Ifc36086f4c5d6a44fb5456aa6e31bc8a783bba71 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2579648 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Xinghua Cao 378653f8 2020-11-20T12:24:41 D3D: throw a perf warning for uniform block We had translated an uniform block only containing a large array member into StructuredBuffer instead of cbuffer on D3D backend for slow fxc compile performance issue with dynamic uniform indexing. This patch throw a warning if a uniform block containing a large array member fails to hit the optimization. Bug: angleproject:3682 Change-Id: I33459b559923f16a8dfb70c6f46ec52f68d96e06 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2552365 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jiajia Qin <jiajia.qin@intel.com>
Shahbaz Youssefi 8b851562 2020-12-08T16:51:51 Don't output matrix qualifiers in I/O blocks row_major and column_major only apply to uniform and buffer interface blocks per GLSL ES 3.2 spec. Bug: angleproject:3580 Change-Id: Iad0afeffc7ddb7bff2ebaec0ea0ec4eda26171a7 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2580191 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi f691b3b5 2020-12-02T13:11:54 Vulkan: Support PrimitivesGenerated query This query uses the Vulkan transform feedback extension. In GL, GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN and GL_PRIMITIVES_GENERATED queries can be independently begun/ended. However, Vulkan requires that queries from pools of the same type can only be active one at a time. This forbids the two GL queries from being handled by two VK queries when they are simultaneously begun. This change makes these queries share their QueryHelper objects. The Vulkan transform feedback queries unconditionally retrieve both results anyway, so this is just a matter of making sure the two GL queries are merged as one when they are simultaneously used. The change fixes a number of issues as collateral: - TransformFeedbackPrimitivesWritten queries when !emulated were not released - Stashed queries were never released - If no render pass is open when a query ends, then getResult(no_wait) ended up waiting Bug: angleproject:5404 Change-Id: I8ce13ea76ffd31b3152ded7c713c6466d0315504 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2573580 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi 158089fd 2020-12-07T16:58:26 Fix location validation for I/O blocks I/O blocks can specify location for each member of the block separately, in arbitrary fashion. This change fixes up the code that validates varying locations to take this into account. Bug: angleproject:3580 Change-Id: If883347fc5db9f18722e41938d1b61fa64650d0c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2578047 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Mohan Maiya <m.maiya@samsung.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 73d4d9fa 2020-12-07T23:53:55 Refactor varying location count and validation In preparation for support for mid-block location qualifiers. Bug: angleproject:3580 Change-Id: Ia63f20c8dc26e8f0e90371c4f00913585a22267a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2577983 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi e59d8716 2020-12-07T16:41:35 GL: Expose OES_shader_io_blocks As extra validation that new tests are indeed correct. Bug: angleproject:3580 Change-Id: I4792bd8e26f96ab6c988f2605f053ce5789a782f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2577818 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Jamie Madill 2c90a90a 2020-12-04T12:39:50 Geometry Shaders: Don't require 'flat' for integer varyings. The 3.2 spec and the dEQP tests are quite clear in only requiring the 'flat' on fragment shader inputs. This restriction remains on older shader versions. Bug: angleproject:5409 Change-Id: Icc4b113b404280c2867a3e0bf6f77256a4316bff Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2574826 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 7354c67b 2020-12-04T12:21:10 Vulkan: Support for geometry shader with max_vertices=0 This is simply done by specifying max_vertices=1. Bug: angleproject:5411 Change-Id: I50318128cca0e96a8b7c7e5d01165b9ef46b4db8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2574823 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Tim Van Patten c859c0ac 2020-11-24T17:21:38 Batch replace std::unordered_map with angle::HashMap in src/ There are a few places that will remain std::unordered_map due to build or run-time errors, which will need to be evaluated more closely to determine if they should remain std::unordered_map or if there is another Abseil data structure that would be more efficient while still working correctly. Bug: angleproject:4873 Change-Id: Ib04253e3ad6398e63f4cc2bfe12c0f9e57cb112b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2558873 Commit-Queue: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com>
Jamie Madill 173f1c94 2020-12-03T13:53:16 Translator: Collect gl_in as a varying. This unifies the variable handling for IO blocks. Previously the gl_in block we treated as an "in block" similar to an interface block. Bug: angleproject:5423 Change-Id: Idf2db4c209055fa9ffca567dab39c89464eba156 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2571966 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Ian Elliott 1ad5791d 2020-12-03T22:05:51 Revert "Vulkan: Use specialization constant for halfRenderAreaWidth" This reverts commit ff38106d62fca2e36ec2db1cd580b546fd53ae43. Reason for revert: Suspect that this CL may be causing win-angle-rel-32 bot failures Original change's description: > Vulkan: Use specialization constant for halfRenderAreaWidth > > halfRenderArea is used by fragment shader to adjust gl_FragCoord for > rotation and yflip compensation. This CL bakes halfRenderAreaWidth into > shader via specialization constant, thus allow compiler to consolidate > the calculation into one MAD instruction. > > Bug: b/173800146 > Change-Id: Id66301278e3389e2582369b695825e632bccecee > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2551541 > Commit-Queue: Charlie Lao <cclao@google.com> > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Reviewed-by: Tim Van Patten <timvp@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> TBR=timvp@google.com,syoussefi@chromium.org,jmadill@chromium.org,cclao@google.com # Not skipping CQ checks because original CL landed > 1 day ago. Bug: b/173800146 Change-Id: I7bc9966ebbe949f3999d7927afa629dd2e3a4187 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2572801 Reviewed-by: Ian Elliott <ianelliott@google.com> Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Commit-Queue: Ian Elliott <ianelliott@google.com>
Shahbaz Youssefi abe96578 2020-11-30T14:09:11 Vulkan: I/O blocks: Support unsized arrays Geometry shader inputs have an extra array dimension, which can be unsized as it can be derived from the primitive type. This change fixes the grammar to support such arrays. Additionally, it enables EXT_shader_io_blocks automatically with EXT_geometry_shader and EXT_tessellation_shader per spec. Bug: angleproject:3580 Change-Id: Ia7eb3e8be28c2eef2072dbe2a546fa34973104ab Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2568242 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi dfdf90a7 2020-12-01T16:09:41 Create helper for identifying builtin names Bug: angleproject:5405 Change-Id: I652f2d6d3af4468b9662cdea5d2706d36e118239 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2568550 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 8797714c 2020-12-01T11:39:25 Vulkan: Support OES_shader_io_blocks Enables OES/EXT_shader_io_blocks extensions in Vulkan backends. With shader I/O blocks, the varyings can now be an array of struct (the block itself) of struct (nested in the block). This change is missing a number of features. In particular, if the shader I/O block has a location decoration in the middle of the block, that is not handled yet. Based on changes from m.maiya@samsung.com and jmadill@chromium.org. Bug: angleproject:3580 Tests: dEQP-GLES31.functional.shaders.linkage.es31.io_block.* dEQP-GLES31.functional.separate_shader.validation.es31.io_blocks.* dEQP-GLES31.functional.program_interface_query.program_input.* dEQP-GLES31.functional.program_interface_query.program_output.* Change-Id: I593840475d2365ff6c9ce7b2290f5ee462a30dfb Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2567645 Reviewed-by: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Jamie Madill 720b85e6 2020-12-01T20:45:29 Reland "Vulkan: Use specialization constant for halfRenderAreaWidth" This reverts commit b1917fe1a314709cd9f7f53b2923d6be0e8bbf6c. Reason for revert: Not the root cause. See bug. Original change's description: > Revert "Vulkan: Use specialization constant for halfRenderAreaWidth" > > This reverts commit ff38106d62fca2e36ec2db1cd580b546fd53ae43. > > Reason for revert: Speculative revert to clear up AMD. See bug. > > Bug: chromium:1154356 > > Original change's description: > > Vulkan: Use specialization constant for halfRenderAreaWidth > > > > halfRenderArea is used by fragment shader to adjust gl_FragCoord for > > rotation and yflip compensation. This CL bakes halfRenderAreaWidth into > > shader via specialization constant, thus allow compiler to consolidate > > the calculation into one MAD instruction. > > > > Bug: b/173800146 > > Change-Id: Id66301278e3389e2582369b695825e632bccecee > > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2551541 > > Commit-Queue: Charlie Lao <cclao@google.com> > > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > > Reviewed-by: Tim Van Patten <timvp@google.com> > > Reviewed-by: Jamie Madill <jmadill@chromium.org> > > TBR=timvp@google.com,syoussefi@chromium.org,jmadill@chromium.org,cclao@google.com > > Change-Id: Idab612bcad4f179857c41ffc4bf8fa2355fae946 > No-Presubmit: true > No-Tree-Checks: true > No-Try: true > Bug: b/173800146 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2568244 > Reviewed-by: Jamie Madill <jmadill@chromium.org> > Commit-Queue: Jamie Madill <jmadill@chromium.org> TBR=timvp@google.com,syoussefi@chromium.org,jmadill@chromium.org,cclao@google.com # Not skipping CQ checks because this is a reland. Bug: chromium:1154356 Bug: b/173800146 Change-Id: I5b86bbee8600c6597620f31d8d7b08ca8b3d2d57 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2568246 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Jamie Madill b1917fe1 2020-12-01T20:42:36 Revert "Vulkan: Use specialization constant for halfRenderAreaWidth" This reverts commit ff38106d62fca2e36ec2db1cd580b546fd53ae43. Reason for revert: Speculative revert to clear up AMD. See bug. Bug: chromium:1154356 Original change's description: > Vulkan: Use specialization constant for halfRenderAreaWidth > > halfRenderArea is used by fragment shader to adjust gl_FragCoord for > rotation and yflip compensation. This CL bakes halfRenderAreaWidth into > shader via specialization constant, thus allow compiler to consolidate > the calculation into one MAD instruction. > > Bug: b/173800146 > Change-Id: Id66301278e3389e2582369b695825e632bccecee > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2551541 > Commit-Queue: Charlie Lao <cclao@google.com> > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Reviewed-by: Tim Van Patten <timvp@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> TBR=timvp@google.com,syoussefi@chromium.org,jmadill@chromium.org,cclao@google.com Change-Id: Idab612bcad4f179857c41ffc4bf8fa2355fae946 No-Presubmit: true No-Tree-Checks: true No-Try: true Bug: b/173800146 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2568244 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Charlie Lao ff38106d 2020-11-19T16:47:25 Vulkan: Use specialization constant for halfRenderAreaWidth halfRenderArea is used by fragment shader to adjust gl_FragCoord for rotation and yflip compensation. This CL bakes halfRenderAreaWidth into shader via specialization constant, thus allow compiler to consolidate the calculation into one MAD instruction. Bug: b/173800146 Change-Id: Id66301278e3389e2582369b695825e632bccecee Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2551541 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Xinghua Cao da1e4122 2020-11-24T14:36:49 D3D: Fix wrong visit mode UniformBlockTranslatedToStructuredBufferTraverser needs InVisit step when calling visitBinary. Bug: angleproject:3682 Change-Id: I5c6cec1df52f94de24e1ea0ed13e4fe1e5674faf Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2557557 Commit-Queue: Xinghua Cao <xinghua.cao@intel.com> Reviewed-by: Jiajia Qin <jiajia.qin@intel.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Charlie Lao 24f64249 2020-11-16T17:55:02 Vulkan: Track specialization constant usage bit and feedback to ctx Right now context does not know which specialization constant is used by the shader. Whenever a specialization constant changes, we assume shader program is using it, we always reach into vulkan driver to ask for a new program. Instead we can track shader's usage of specialization constant so that context can utilize this information to avoid recompile pipeline program if an unused specialization constant has changed. This CL implements the plumbing the usage bits form compiler to program object, it does not actually utilize the usage bits to avoid unnecessary compilation yet. Bug: b/173461931 Change-Id: Iebc9d0638c17b1a282c8b6093ce6bae154246e57 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2542866 Reviewed-by: Ian Elliott <ianelliott@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Mohan Maiya c4ca12e3 2020-11-20T09:02:54 Vulkan: Support the single-sampled targets Add support for gl_SampleMask for single-sampled targets. When dealing with a single-sampled target, gl_SampleMask is always be set to 0xFFFFFFFF. And when the target is single-sampled, sample shading is disabled to enable Bresenham line rasterization. Bug: angleproject:3588 Tests: dEQP-GLES31.functional.shaders.sample_variables. sample_mask.discard_half_per_pixel.* sample_mask.discard_half_per_sample.* sample_pos.correctness.* sample_mask_in.bit*_per_two_samples.* sample_mask.discard_half_per_two_samples.* sample_mask.inverse_per_* Change-Id: Ibb471261b8451ff01fab3dc43f2e965ae2999610 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2477909 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Mohan Maiya <m.maiya@samsung.com>
Geoff Lang e0619360 2020-11-18T11:13:11 Skip tracking non-preprocessor tokens in disabled if-blocks ESSL1 states that it is an error to have a non-preprocessor token before an extension pragma but ANGLE would mark a non-preprocessor token as seen even if it was in a disabled if-block. Bug: b/161716126 Change-Id: I782224bee211bc7214a2d1d98d4829b4c94101bd Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2547205 Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Mohan Maiya cc5f7989 2020-11-19T07:59:01 Vulkan: Support gl_SampleMask with non-const index Add support for access to gl_SampleMask or gl_SampleMaskIn with non-constant index. Replace gl_SampleMask with ANGLESampleMask and reassign ANGLESampleMask to gl_SampleMask in the end. Bug: angleproject:3588 Tests: dEQP-GLES31.functional.shaders.sample_variables.sample_mask. discard_half_per_pixel.* discard_half_per_sample.* Change-Id: Ie668c3ccf4d40352574db57f5e43b2f17ef6d8b8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2477908 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Mohan Maiya 8a401051 2020-11-19T07:58:05 Vulkan: Support OES_sample_variables extension Enable OES_sample_variables extension on Vulkan backend. Add support for built-in variables, uniform state and constants - gl_SampleId, gl_SamplePosition, gl_SampleMask, gl_SampleMaskIn - gl_NumSamples - gl_MaxSamples Bug: angleproject:3588 Tests: dEQP-GLES31.functional.shaders.sample_variables.* KHR-GLES31.core.sample_variables.* Change-Id: Idf37c7b4ccb0331dbda7acd6389ff4e1022ba959 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2477907 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Shahbaz Youssefi 8d63a7fc 2020-11-17T15:15:44 Basic support for running deqp khr gles32 tests Bug: angleproject:3647 Change-Id: Ia707ffcd41926662d1513f1a3bf49149b76f6b9c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2544920 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Yuly Novikov <ynovikov@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Charlie Lao 995493cc 2020-11-18T20:33:33 Revert "Vulkan: Remove rotation related data from driver uniform" This reverts commit fd97c3343e9ff23677124b7528ec945bca65796a. Reason for revert: The specialization constant for rotation causes performance regression with older qualcomm driver. We need to keep the driver uniform code path for the older driver. Original change's description: > Vulkan: Remove rotation related data from driver uniform > > Rotation is now handled in the shader compiler with specialization > constant, it should be removed from driver uniforms. Since Metal is > using the flipXY, flipXY/negFlipXY are still kept in the shader side > implementation, but have moved to TranslatorMetal in this CL. > > Bug: b/171750979 > Change-Id: Ie8d15ef227cb52a6e19e4319ecc9f09bda42e667 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2519863 > Reviewed-by: Ian Elliott <ianelliott@google.com> > Reviewed-by: Tim Van Patten <timvp@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> > Commit-Queue: Charlie Lao <cclao@google.com> TBR=ianelliott@google.com,timvp@google.com,jmadill@chromium.org,cclao@google.com Bug: b/171750979 Change-Id: Iff9cffb28851ade1d9c5cd23fde73910a19867ce Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2547808 Reviewed-by: Ian Elliott <ianelliott@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com> Commit-Queue: Charlie Lao <cclao@google.com>
Mohan Maiya 738092ae 2020-11-18T07:58:53 Add auto generated code for GL_EXT_tessellation_shader To support tessellation, code and auto generated code are added. Entry points function, ShaderTypes, built in variables and constants, builtin function barrier and patch keyword are added. Bug: angleproject:3572 Change-Id: Ia5fe473e884466cb88cea7138e13377a1d7b4fa0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2538393 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao 2eb85a5b 2020-11-12T18:37:26 Vulkan: Experimental: forcing highp to mediump in fragment shader. Handy code for debug/experiment that forces highp in fragment shader to mediump. Bug: b/173140243 Change-Id: I1841b67862343df8a12bf302066d7eefd7709604 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2537072 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Courtney Goeltzenleuchter <courtneygo@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi 86d7e4d8 2020-11-10T15:55:34 Vulkan: Support texture buffers Texture buffers are placed in the same descriptor set with the rest of the textures. However, the different code paths that handle textures have special cases for texture buffers as they create a different descriptor type (texel buffer instead of combined image sampler). Image view serials are used to track the buffer view serials as well so the texture descriptor cache can handle texture buffers as well. This CL is missing storage texel buffer support. Bug: angleproject:3573 Change-Id: Iff80ca22ff9b9957a0c9a3c7aaada1fa54b24ec8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2532653 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Charlie Lao fd97c334 2020-11-12T11:05:40 Vulkan: Remove rotation related data from driver uniform Rotation is now handled in the shader compiler with specialization constant, it should be removed from driver uniforms. Since Metal is using the flipXY, flipXY/negFlipXY are still kept in the shader side implementation, but have moved to TranslatorMetal in this CL. Bug: b/171750979 Change-Id: Ie8d15ef227cb52a6e19e4319ecc9f09bda42e667 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2519863 Reviewed-by: Ian Elliott <ianelliott@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Charlie Lao <cclao@google.com>
Geoff Lang 5b419533 2020-11-03T13:33:44 GL: Implement EXT_YUV_target Add test coverage of YUV format sampling as RGB or directly as YUV and rendering as YUV using layout(yuv). Initializing YUV AHardwareBuffers requires Android API 29 so ANGLE must be compiled with: android32_ndk_api_level = 29 android64_ndk_api_level = 29 The following tests can still run with Android API 26 because they don't need to initialize the buffer: ImageTestES3.ClearYUVAHB ImageTestES3.YUVValidation Bug: angleproject:4852 Bug: b/172649538 Change-Id: I4fe9afb2a68fb827dc5a5732b213b5eb60d585ac Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2517562 Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Charlie Lao 3a1838a1 2020-11-10T20:30:43 Vulkan: Let RewriteInterpolateAtOffset use FlipRotateSpecConst Instead of using uniform, use flip/rotation specialization constant when available. This allows matrix multiplication calculated at compile time instead of shader execution time. Bug: angleproject:4901 Bug: angleproject:3589 Change-Id: I5a8a82b1cffbbbeba2617cb7e9a0ec38f9984d50 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2530457 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Ian Elliott <ianelliott@google.com>
Charlie Lao 15ead5d5 2020-11-02T12:25:32 Reland "Vulkan: Let shader use rotation specialized constant" This is a reland of 47b3db22be33213eea4ad58f2453ee1088324ceb Original change's description: > Vulkan: Let shader use rotation specialized constant > > If use rotation specialized constant is enabled via compiler options, > this CL will use rotation specialized constant to generate flipXY, > rotation matrix and negFlipXY. This allows the driver to optimize for > the minimum instructions for rotation. > > Bug: b/171750979 > Change-Id: I9851ac999d4d35b9f230f796e5445bca0dcb1e77 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2514773 > Commit-Queue: Charlie Lao <cclao@google.com> > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Reviewed-by: Ian Elliott <ianelliott@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> > Reviewed-by: Tim Van Patten <timvp@google.com> Bug: b/171750979 Change-Id: I9efde40eb138bacb161cce483a30ec01fee0983b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2533933 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Charlie Lao 3f18ad09 2020-11-11T14:12:18 Vulkan: Factor out DriverUniform code to tree_utils/DriverUniform.cpp The driver uniform code is used by various tree_ops functions. But all driver uniform code are defined and implemented in the TranslatorVulkan.cpp file. There is dependency rule that tree_ops code can not call into vulkan specific function. Right now we are working around this problem by always having TranslatorVulkan creates uniform and pass it down to tree_ops. This creates inefficiency for cases that dFdy/dFdx where we don't know if we will need the driver uniform or not, until we walk the tree and see the dFdx node. This CL refactors driver uniform code into its own file and class under tree_utils so that everybody can use. Mean time we can also make it much easier for metal to expand it to have its own uniform structure. Bug: b/173047182 Change-Id: I06bd9a005ccd6dc0a21321a3010dda1eab9d6fdb Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2533443 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Tim Van Patten c3607098 2020-11-11T16:06:51 Revert "Vulkan: Let shader use rotation specialized constant" This reverts commit 47b3db22be33213eea4ad58f2453ee1088324ceb. Reason for revert: Broke the ANGLE roll into Chromium: https://chromium-review.googlesource.com/c/chromium/src/+/2531069 [1875/27343] SOLINK ./libGLESv2.so FAILED: libGLESv2.so libGLESv2.so.TOC /b/s/w/ir/cipd_bin_packages/cpython/bin/python "../../build/toolchain/gcc_solink_wrapper.py" --readelf="readelf" --nm="nm" --sofile="./libGLESv2.so" --tocfile="./libGLESv2.so.TOC" --output="./libGLESv2.so" -- ../../third_party/llvm-build/Release+Asserts/bin/clang++ -shared -Wl,-soname="libGLESv2.so" -Wl,--fatal-warnings -Wl,--build-id -fPIC -Wl,-z,noexecstack -Wl,-z,relro -Wl,-z,defs -Wl,--as-needed -fuse-ld=lld -Wl,--color-diagnostics -Wl,--no-call-graph-profile-sort -m64 -Werror -nostdlib++ --sysroot=../../build/linux/debian_sid_amd64-sysroot -L../../build/linux/debian_sid_amd64-sysroot/usr/local/lib/x86_64-linux-gnu -L../../build/linux/debian_sid_amd64-sysroot/lib/x86_64-linux-gnu -L../../build/linux/debian_sid_amd64-sysroot/usr/lib/x86_64-linux-gnu -Wl,-rpath=\$ORIGIN -Wl,--export-dynamic -Wl,-rpath=\$ORIGIN -o "./libGLESv2.so" @"./libGLESv2.so.rsp" ld.lld: error: undefined symbol: sh::TranslatorVulkan::GetDriverUniformFlipXYRef(sh::TVariable const*) >>> referenced by RewriteDfdy.cpp:134 (../../third_party/angle/src/compiler/translator/tree_ops/RewriteDfdy.cpp:134) >>> translator/RewriteDfdy.o:(sh::(anonymous namespace)::Traverser::visitUnaryWithRotation(sh::Visit, sh::TIntermUnary*)) in archive obj/third_party/angle/libtranslator.a >>> referenced by RewriteDfdy.cpp:200 (../../third_party/angle/src/compiler/translator/tree_ops/RewriteDfdy.cpp:200) >>> translator/RewriteDfdy.o:(sh::(anonymous namespace)::Traverser::visitUnaryWithoutRotation(sh::Visit, sh::TIntermUnary*)) in archive obj/third_party/angle/libtranslator.a ld.lld: error: undefined symbol: sh::TranslatorVulkan::GetDriverUniformFragRotationMatrixRef(sh::TVariable const*) >>> referenced by RewriteDfdy.cpp:136 (../../third_party/angle/src/compiler/translator/tree_ops/RewriteDfdy.cpp:136) >>> translator/RewriteDfdy.o:(sh::(anonymous namespace)::Traverser::visitUnaryWithRotation(sh::Visit, sh::TIntermUnary*)) in archive obj/third_party/angle/libtranslator.a clang: error: linker command failed with exit code 1 (use -v to see invocation) Original change's description: > Vulkan: Let shader use rotation specialized constant > > If use rotation specialized constant is enabled via compiler options, > this CL will use rotation specialized constant to generate flipXY, > rotation matrix and negFlipXY. This allows the driver to optimize for > the minimum instructions for rotation. > > Bug: b/171750979 > Change-Id: I9851ac999d4d35b9f230f796e5445bca0dcb1e77 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2514773 > Commit-Queue: Charlie Lao <cclao@google.com> > Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> > Reviewed-by: Ian Elliott <ianelliott@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> > Reviewed-by: Tim Van Patten <timvp@google.com> TBR=ianelliott@google.com,timvp@google.com,syoussefi@chromium.org,jmadill@chromium.org,cclao@google.com Change-Id: Id3976259d1f4049aa62b679ad5b25c2ac337c92b No-Presubmit: true No-Tree-Checks: true No-Try: true Bug: b/171750979 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2532702 Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Tim Van Patten <timvp@google.com>
Charlie Lao 47b3db22 2020-11-02T12:25:32 Vulkan: Let shader use rotation specialized constant If use rotation specialized constant is enabled via compiler options, this CL will use rotation specialized constant to generate flipXY, rotation matrix and negFlipXY. This allows the driver to optimize for the minimum instructions for rotation. Bug: b/171750979 Change-Id: I9851ac999d4d35b9f230f796e5445bca0dcb1e77 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2514773 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Ian Elliott <ianelliott@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Mohan Maiya 34130c2e 2020-11-07T15:11:14 Vulkan: Rewrite interpolateAtOffset to account for Y-flip In order to fully suppot OES_shader_multisample_interpolation rewrite interpolateAtOffset's offset parameter to account for Y-coordinate flip. Bug: angleproject:3589 Tests: dEQP-GLES31.functional.shaders.multisample_interpolation. interpolate_at_offset.*.default_framebuffer Change-Id: I6bf72ceb1c0466ff5d4a92b72d1ec9e2552d6f2a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2524711 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Jamie Madill aca3d408 2020-11-05T20:53:37 Fix return value of cube map array imageStore. Should be void according to the spec. Noticed while working on texture buffer. Bug: angleproject:3584 Change-Id: I79a671c82fa117bab69bc2e6b35c840aa30429c5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2521241 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Jamie Madill 43acf3ba 2020-07-31T15:43:57 GL: Compiler changes for GL_OES_texture_buffer This extension is core in GLES 3.2 This CL enables the extension, added 412 new dEQP tests for texture buffer. 410 of these tests pass, but there are two tests that fail, detailed in anglebug.com/4933 Also includes a fix for completeness checks. Based on a CL by Jonah Ryan-Davis. Bug: angleproject:3573 Bug: angleproject:4933 Change-Id: I45759d765c88c64f21b592eab01d910dc4bc8da3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2521239 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com>
Jamie Madill f2a78926 2020-11-05T14:01:39 Symbol Table: Error on duplicate group keys. This prevents newer group keys from overwriting old ones. Prevents bugs from creeping in like happened during texture buffer development. Bug: angleproject:3573 Change-Id: Iea69cac988e2cf4802a9f64c54cb694d32f1d3e1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2521238 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Xinghua Cao dc1c1cb5 2020-08-12T13:30:26 Restrict to translate uniform block to StructuredBuffer We had translated an uniform block only containing a large array member into StructuredBuffer instead of cbuffer on D3D backend for slow fxc compile performance issue with dynamic uniform indexing. Now we add more conditions to restrict the translation. Only indexing operator is allowed to operate on this uniform block variable. And we also restrict the types of uniform block's member. Bug: angleproject:3682 Change-Id: I992b7890d84fcaa6169722af6d7e14785526d48a Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2351728 Commit-Queue: Xinghua Cao <xinghua.cao@intel.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jiajia Qin <jiajia.qin@intel.com>
Mohan Maiya e895ced6 2020-10-29T09:06:00 Vulkan: Add OES_shader_image_atomic support Add support for shader builtins and enable the extension. Bug: angleproject:3578 Tests: dEQP-GLES31.functional.*image_atomic* Change-Id: Idd45b2ee62efe1474c6c5947c77da64ff2221bf6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2505540 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Mohan Maiya 5d0458fa 2020-10-23T12:52:34 Vulkan: Add OES_shader_multisample_interpolation support Support OES_shader_multisample_interpolation extension if maxInterpolationOffset >= 0.5 Bug: angleproject:3589 Tests: dEQP-GLES31.functional.shaders.multisample_interpolation.* dEQP-GLES31.functional.state_query.multisample_interpolation.* Change-Id: I42997f10be82e3be8b63c56833cbbf791bf4be9b Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2477905 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Mohan Maiya 553726a4 2020-10-20T11:29:17 Vulkan: Add OES_shader_multisample_interpolation extension Addition of 'sample' qualifier keyword and 'interplateAt*' fragment shader builtin functions with autogen Bug: angleproject:3589 Change-Id: If358eb371fbcefffa715c8da4ba5e96eefaf6f52 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2477904 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill e5590288 2020-10-18T12:40:14 Rename "texelFetchInvoked" to "texelFetchStaticUse". Technically more precise because something could be statically used and never invoked. Also more consistent with the spec. Bug: angleproject:5176 Change-Id: I70dd0787d67d9f046b7359abb24107cb430f5cae Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2483465 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Courtney Goeltzenleuchter <courtneygo@google.com>
Mohan Maiya 7bbe497a 2020-10-16T14:46:45 Vulkan: Implement EXT_texture_sRGB_decode Implement EXT_texture_sRGB_decode. This builds on the existing functionality from EXT_texture_sRGB_override, with 2 major edge cases: 1. sRGB_decode allows the texture state to be overridden by sampler state, which is implemented by forcing a a texture state sync during updateActiveTextures if a texture is bound to the same unit as a sampler with that state 2. texelFetch calls require us to reenable decoding, regardless of decode state. We add a new compiler pass (FlagSamplersWithTexelFetch) to mark samplers that are used with texelFetch in order to support this. This change also re-enables EXT_texture_sRGB_R8, which was disabled due to a dEQP bug that this change will bypass. Bug: angleproject:3609 Bug: angleproject:4503 Test: dEQP.GLES31/functional_srgb_texture_decode_skip_decode_* Test: GLES31/functional_state_query_texture_*_srgb_decode_* Test: GLES31/functional_state_query_sampler_*_srgb_decode_* Test: GLES31/functional_debug_negative_coverage_*_srgb_decode_* Test: GLES31/functional_android_extension_pack_extensions_ext_texture_srgb_decode Test: angle_end2end_tests --gtest_filter=SRGBTextureTest.*Vulkan* Change-Id: I4a67e487dc82e2f57c8c87d4bcd8ef442b6fe220 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2359481 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com>
Trevor David Black e815afbf 2020-09-07T22:09:22 First pass at increasing inclusivity Link to the inclusivity rules https://source.android.com/setup/contribute/respectful-code Bug: b/162834212 Bug: chromium:1097198 Change-Id: Ied5a9e3879d72bff3f77ea6fcda9b82f30c32c2f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2396737 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Trevor Black <vantablack@google.com>
Le Hoang Quyen 6136cbcb 2020-09-23T21:31:05 Metal: Implement transform feedback - XFB is currently emulated by writing to storage buffers. - Metal doesn't allow vertex shader to both write to storage buffers and to stage output (i.e clip position). So if GL_RASTERIZER_DISCARD is NOT enabled, the draw with XFB enabled will have 2 passes: + First pass: vertex shader writes to XFB buffers + not write to stage output + disable rasterizer. + Second pass: vertex shader writes to stage output (i.e. [[position]]) + enable rasterizer. If GL_RASTERIZER_DISCARD is enabled, the second pass is omitted. + This effectively executes the same vertex shader twice. TODO: possible improvement is writing vertex outputs to buffer in first pass then re-use that buffer as input for second pass which has a passthrough vertex shader. - If GL_RASTERIZER_DISCARD is enabled, and XFB is enabled: + Only first pass above will be executed, and the render pass will use an empty 1x1 texture attachment since rasterization is not needed. - If GL_RASTERIZER_DISCARD is enabled, but XFB is NOT enabled: + we still enable Metal rasterizer. + but vertex shader must emulate the discard by writing gl_Position = (-3, -3, -3, 1). This effectively moves the vertex out of clip space's visible area. + This is because GLSL still allows vertex shader to write to stage output when rasterizer is disabled. However, Metal doesn't allow that. In Metal, if rasterizer is disabled, then vertex shader must not write to stage output. - See src/libANGLE/renderer/metal/doc/TransformFeedback.md for more details. Bug: angleproject:2634 Change-Id: I6c700e031052560326b7f660ee7597202d38e6aa Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2408594 Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jonah Ryan-Davis <jonahr@google.com>
Jamie Madill 7a0faa82 2020-09-24T20:15:26 Revert "Pass #pragma optimize setting down to compilation." This reverts commit 499173de1c91932ba272269cab6918bf7e8d7c11. Reason for revert: Causes unexpected HLSL compiler errors in some cases. See bug. Bug: angleproject:5094 Original change's description: > Pass #pragma optimize setting down to compilation. > > This will allow us to disable optimizations in the back-end. This can > be useful both for developers and for ANGLE to disable very slow > shader compilation on D3D11. > > Also apply this pragma to VerifyMaxVertexUniformVectorsWithSamplers. > Reduces compilation time by half in local testing. > > Bug: angleproject:5076 > Change-Id: I64ad576e11b9cee5b41f8af0d3621570304d65c2 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2420749 > Commit-Queue: Jamie Madill <jmadill@chromium.org> > Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> > Reviewed-by: Geoff Lang <geofflang@chromium.org> TBR=geofflang@chromium.org,jonahr@google.com,jmadill@chromium.org # Not skipping CQ checks because original CL landed > 1 day ago. Bug: angleproject:5076 Change-Id: I733e788fe8e9421ae0af662c0eb51af1ed79dde3 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2429517 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Tim Van Patten e78d9a61 2020-07-24T11:58:45 Convert unordered_map to absl::flat_hash_map for select files This is the initial CL to start migrating to abseil in various places: - formatutils.h - FramebufferVk.h - Program.h - ProgramExecutableVk.h - RewriteRowMajorMatrices.cpp This intentionally hits a couple different places in the code to make sure the abseil dependencies are added to the required targets. Bug: angleproject:4873 Change-Id: Idd6084dff2ebce47833f304c605bbf3151b97414 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2402382 Commit-Queue: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com>
Le Hoang Quyen 9ff58ddd 2020-08-26T01:08:00 Metal: Support array of matrices varying. MSL doesn't support array of matrices as varying variable. This CL converts array of matrices varying to array of vectors. This fixes WebGL test: conformance/glsl/misc/shader-varying-packing-restrictions.html Bug: angleproject:4846 Bug: angleproject:2634 Change-Id: I3cfc638dc01092ddf1add7df02c131d1a5c8ba1d Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2392172 Commit-Queue: Le Hoang Quyen <le.hoang.q@gmail.com> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jamie Madill 499173de 2020-09-20T10:42:56 Pass #pragma optimize setting down to compilation. This will allow us to disable optimizations in the back-end. This can be useful both for developers and for ANGLE to disable very slow shader compilation on D3D11. Also apply this pragma to VerifyMaxVertexUniformVectorsWithSamplers. Reduces compilation time by half in local testing. Bug: angleproject:5076 Change-Id: I64ad576e11b9cee5b41f8af0d3621570304d65c2 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2420749 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Nicolas Capens 07c03b6d 2020-09-11T14:43:37 Validate GLSL attribute location range The location index of a vertex shader input's layout qualifier must be less than GL_MAX_VERTEX_ATTRIBS to link successfully. We can check for this during shader compilation to not rely on other layers to handle pathological cases. While strictly speaking only 'active' attributes are considered during shader linking, this is merely intended to allow for 'uber' shaders to declare more inputs than GL_MAX_VERTEX_ATTRIBS but only use a subset of them. There is no known reasonable use case for a manually specified location to exceed the valid range. Note that according to http://opengl.gpuinfo.org, the highest GL_MAX_VERTEX_ATTRIBS value at the time of writing is 32. Also, D3D12_VS_INPUT_REGISTER_COUNT = 32. Hence the unit test's value of 1000 should be sufficiently future proof. Also address the case where the uniform location might be close to INT_MAX and not be detected as out-of-range due to numeric overflow. Bug: chromium:1110800 Change-Id: I9985c8eab3bb8a2a59b8f985e8f5b6884756383c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2405368 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Nicolas Capens <capn@chromium.org>
Xinyi He 377e7487 2020-08-27T17:11:35 Vulkan: Support array of array image type Implement supporting the array of array of image type in uniform. Add a new end2end test for it. Bug: angleproject:3881 Change-Id: Idd757ae1d0ed34d585ae1ca5e0b6577459a0acb7 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2379335 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Geoff Lang 7e815c77 2020-07-08T13:43:40 Implement EXT_shadow_samplers Bug: angleproject:4863,b/161716126 Change-Id: I6beb45d91f59a851787c9f40b40266fb985198d5 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2288330 Commit-Queue: Geoff Lang <geofflang@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: back sept 10 - Jamie Madill <jmadill@chromium.org>
Jamie Madill 8d412db2 2020-08-29T22:25:18 Revert "Convert unordered_map to absl::flat_hash_map for select files" This reverts commit 1acaf4eced42a26f09b4d9180a9e5db0f37bc9a4. Reason for revert: Fails when is_component_build is not set: lld-link: error: <root>: undefined symbol: public: __cdecl absl::Condition::Condition<struct std::__1::atomic<bool> const>(bool (__cdecl *)(struct std::__1::atomic<bool> const *), struct std::__1::atomic<bool> const *) <snip> Original change's description: > Convert unordered_map to absl::flat_hash_map for select files > > This is the initial CL to start migrating to abseil in various places: > - formatutils.h > - FramebufferVk.h > - Program.h > - ProgramExecutableVk.h > - RewriteRowMajorMatrices.cpp > > This intentionally hits a couple different places in the code to make > sure the abseil dependencies are added to the required targets. > > Bug: angleproject:4873 > Change-Id: I68c7d067b6912b0cc0ecde231501dbed92f0b189 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2321735 > Commit-Queue: Tim Van Patten <timvp@google.com> > Reviewed-by: Jamie Madill <jmadill@chromium.org> > Reviewed-by: Charlie Lao <cclao@google.com> TBR=timvp@google.com,jmadill@chromium.org,cclao@google.com Change-Id: I75041532fc5126b4c7cc5e0d4529883fb357e05b No-Presubmit: true No-Tree-Checks: true No-Try: true Bug: angleproject:4873 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2383870 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Tim Van Patten 1acaf4ec 2020-07-24T11:58:45 Convert unordered_map to absl::flat_hash_map for select files This is the initial CL to start migrating to abseil in various places: - formatutils.h - FramebufferVk.h - Program.h - ProgramExecutableVk.h - RewriteRowMajorMatrices.cpp This intentionally hits a couple different places in the code to make sure the abseil dependencies are added to the required targets. Bug: angleproject:4873 Change-Id: I68c7d067b6912b0cc0ecde231501dbed92f0b189 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2321735 Commit-Queue: Tim Van Patten <timvp@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Charlie Lao <cclao@google.com>
Jamie Madill 5c56f228 2020-08-28T14:45:14 Enable two override suggestion warnings. -Wsuggest-destructor-override and -Wsuggest-override. Bug: skia:7647 Change-Id: Iaac1baa8f34fdf210baf2fdbe811a582b3ac2d14 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2376717 Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Shahbaz Youssefi e3988722 2020-08-06T18:23:28 Fix an OOB access bug with RewriteRowMajorMatrices Bug: angleproject:2273 Change-Id: I27109b873a290133cf3d016f612e1fd1de9d354c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2341763 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Le Hoang Quyen cbd5bee8 2020-07-13T20:31:29 Disable shader's pre-rotation code on Metal & non-Android. Pre-rotation code were added to transform gl_Position, gl_FragCoords, gl_PointCoords, dFdX, dFdY in shader. However, it is only useful for android's surface pre-rotation and completely un-needed in Metal back-end. This CL disables these pre-rotation code if the platform is not android. Bug: angleproject:4678 Change-Id: I89c42fcf24b49896f4ed9c2f9465da521beaf25f Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2295000 Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Reviewed-by: Ian Elliott <ianelliott@google.com>
Le Hoang Quyen 08b9ad57 2020-07-08T12:50:00 Metal: Prevent vertex shader's inactive varying from being removed If vertex shader declares a varying but doesn't use it, and this varying is being used in fragment shader, linking will fail because the varying would be removed by spirv-cross. Fix: initilize the unused varying with zero(s). New test: GLSLTest.InactiveVaryingInVertexActiveInFragment. Bug: angleproject:2634 Change-Id: Ia37acd8eea85d5da4ddbca6457d6cddb1b41662c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2281796 Commit-Queue: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Le Hoang Quyen 9277ee74 2020-07-06T12:40:58 Metal: Implement MSAA default framebuffer. GL_SAMPLE_COVERAGE_VALUE is implemented by inserting gl_SampleMask writing logic to fragment shader. New test added: MultisampleTest.ContentPresevedAfterInterruption. - Skip on D3D11 (Bug: angleproject:4609) Bug: angleproject:2634 Change-Id: Ib44daf0baccc36ea320596d81713156047da059c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2281783 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Xinghua Cao d1bf23d7 2020-06-23T16:57:05 Restrict member types for StructuredBuffer Std140 storage layout has some rules for uniform block, must pad StructuredBuffer element type explicitly to match std140 layout rules when translating std140 uniform block to StructuredBuffer. In this CL, restrict member types which are not needed to add paddings according std140 layout rules. Bug: angleproject:4783 Change-Id: I50130da64ad6865083849bf55b76525f3d2835a8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2259226 Commit-Queue: Xinghua Cao <xinghua.cao@intel.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jiajia Qin <jiajia.qin@intel.com>
Stuart Morgan f4e6ae91 2020-06-15T15:56:43 Fixes missing newlines in headers Bug: angleproject:4757 Change-Id: Ic5d9bb62834cbd66463086c298ebcaeb3eb44ad0 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2247486 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Lingfeng Yang 98d0f30c 2020-06-10T22:04:03 Add to names map even if it's a prefixed name Previously, we skipped adding to the names map if it was a prefixed name. This prevents the mapping of glsl es to glsl names from being accessible by an outside client. Bug: b/158533223 Change-Id: Ib1d1e229ceb6d836df8dd0c30ea848b081a40cad Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2240637 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Lingfeng Yang <lfy@google.com> Commit-Queue: Lingfeng Yang <lfy@google.com>
Shahbaz Youssefi b131d25f 2020-06-03T10:53:39 Vulkan: Remove validation error suppressions for VVL bug These warnings were generated due to a validation layer bug, which is presumably fixed now. Bug: angleproject:4572 Bug: angleproject:4577 Bug: angleproject:4578 Bug: angleproject:4579 Bug: angleproject:4580 Bug: angleproject:4583 Change-Id: I8799c4e33bc38f83e0fdd364cf6e7eb607c39890 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2227810 Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Ian Elliott 11d7b952 2020-06-02T16:58:26 Vulkan: Pre-rotate dFdx() & dFdy() for Android Test: angle_deqp_gles3_tests --gtest_filter=dEQP.GLES3/functional_shaders_derivate Test: angle_end2end_tests --gtest_filter=*EGLPreRotationSurfaceTest.OrientedWindow Bug: b/157476241 Change-Id: I7b60b7a33313ba91731755d289792bd0968515c6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2222825 Commit-Queue: Ian Elliott <ianelliott@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Philip Lamoureux 9515707b 2020-05-29T11:23:44 Move ShaderVariable utility function from utilities.cpp ShaderVars.cpp. Fixes an implied dependency cycle between translator and angle_common. This will also allow us to be more strict about declaring dependencies on GLSLANG headers. Tested by building and running angle_unit_tests on Linux. Bug: angleproject:4672 Change-Id: I331230d2cf179ccea140ee7a0d5a3c8768c58cb1 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2222682 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jamie Madill <jmadill@chromium.org>
Ian Elliott 4d7a3993 2020-05-27T14:08:20 Vulkan: Rotate gl_PointCoord for Android pre-rotation This extends the current y-flip that's done for gl_PointCoord. Now, gl_PointCoord is rotated and then flips both the x and y-axis (similar to gl_FragCoord). The tests used to reproduce the problem and the fix are: angle_deqp_gles2_tests --local-output --gtest_filter=dEQP.GLES2/functional_shaders_builtin_variable_pointcoord* --use-angle=vulkan angle_end2end_tests --local-output --gtest_filter=*GLSLTest.PointCoordConsistency*Vulkan* Bug: b/157476696 Change-Id: Iada8680eda8322f7382ff242e4a9422a66114f07 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2216700 Commit-Queue: Ian Elliott <ianelliott@google.com> Reviewed-by: Courtney Goeltzenleuchter <courtneygo@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Jonah Ryan-Davis eadc387c 2020-04-01T16:53:31 GL: Support GL_OES_texture_cube_map_array in shader translator GL_OES_texture_cube_map_array is core in 3.2. This CL adds support for this extension in the shader translator. Now passes dEQP.KHR_GLES31/core_texture_cube_map_array.*, dEQP.GLES31/functional_fbo_color_texcubearray_*, and dEQP.GLES31/functional_opaque_type_indexing*samplercubearray* with GL backend. Bug: angleproject:3584 Change-Id: I1a8485c7d9d3fffb9b5109d292b35b0c56a3c665 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2133086 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Jonah Ryan-Davis <jonahr@google.com>
Ian Elliott 901045cc 2020-05-26T10:41:18 Vulkan: Rotate gl_FragCoord for Android pre-rotation This extends the current y-flip that's done for gl_FragCoord. Now, gl_FragCoord is rotated and then flips both the x and y-axis. Bug: angleproject:4643 Bug: b/156395519 Change-Id: Iada8680eda8322f7382ff242e4a9422a66114f05 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2210936 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Ian Elliott <ianelliott@google.com>
George Burgess IV ca092f01 2020-05-26T11:45:08 translator: explicitly specify underlying types for enums Some of these enums are stored in const structs and such. Since we don't need a full `sizeof(int)` bytes each, don't use it. This shaves 16KB off of libGLESv2.so for eve. Bug: chromium:1084580 Change-Id: Ib6eca49203a74c094661f890e32160974c0c531e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2216431 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Brian Osman 25e21f9f 2020-05-07T16:57:37 Suppress HLSL warnings 3556 and 3571 These are both fairly common, and rarely actionable. Bug: chromium:765576 Change-Id: Ib067fcbb1d4ffe7fcd47b69dbe5defac81430a70 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2188952 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Geoff Lang <geofflang@chromium.org> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Geoff Lang 36c938e0 2020-03-31T12:42:16 GL: Ignore warnings about requested extensions not being present. Some drivers emit warnings about enabled extensions not being present dispite the 'enable' flag only being a request. Supress these warnings to clean up the Chrome console. Also request GL_EXT_gpu_shader5 in addition to GL_ARB_gpu_shader5 to maximize reach. BUG=chromium:1066212 Change-Id: If48cc60b8533f6ffedb727e845a4ca784d98c07c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2129111 Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Jonah Ryan-Davis <jonahr@google.com> Commit-Queue: Geoff Lang <geofflang@chromium.org>
Xinghua Cao d93b9559 2020-04-16T16:16:27 Translate uniform block with large matrix array to StructuredBuffer There is a slow compiling issue if translate a uniform block with a large matrix array member to cbuffer, so translate the uniform block to StructuredBuffer. But pack_matrix pragma directive is invalid for matrix element of StructuredBuffer, so pack the matrix into a structure firstly, and then specifies packing alignment for the matrix in this structure, last the structure is used as StructuredBuffer element. Bug: angleproject:4206 Change-Id: I5a2a512e57b37dcdb88c3060d0d078b46f72eb5e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2154525 Commit-Queue: Xinghua Cao <xinghua.cao@intel.com> Reviewed-by: Geoff Lang <geofflang@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org>
Le Hoang Quyen dadd1986 2020-04-21T01:50:00 Implement GL_APPLE_clip_distance - Built-in variable gl_ClipDistance has been added to compiler. - Desktop GL: gl_ClipDistance is supported since GL 3.0. Enable/Disable each gl_ClipDistances[i] works out of the box via glEnable(). - Vulkan/Metal: Use uniform variable to control writing to each gl_ClipDistance. One bit flag controls one element in the gl_ClipDistance array. The writing to the disabled element in vertex shader will be ignored, and turned into zero assignment instead. - Direct3D/Mobile GL: Not implemented yet. - Added ClipDistanceTest to gl_tests and compiler unittests. - GL_APPLE_clip_distance is a subset of GL_EXT_clip_cull_distance, so GL_EXT_clip_cull_distance could be implemented in future if needed. Bug: angleproject:4452 Change-Id: I571ac8b56826989808a680226a04bec4cf59988e Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2084324 Commit-Queue: Jamie Madill <jmadill@chromium.org> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tim Van Patten 33b58ebb 2019-11-25T16:31:43 Vulkan: Give unsized arrays at least 1 entry The function ShaderVariable::getNestedArraySize needs to return at least 1 entry for unsized arrays to ensure the shader buffer sizes are reported correctly. This also allows ANGLE to treat unsized arrays of structs and basic types the same way, allowing for the special treatment of unsized arrays in TraverseStructArrayVariable() to be removed. Bug: angleproject:3596 Test: dEQP-GLES31.functional.program_interface_query.shader_storage_block.buffer_data_size.* Change-Id: I3b2a3a68c25e0913b79e989d7c719b34ce9b75fd Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/1934952 Commit-Queue: Tim Van Patten <timvp@google.com> Reviewed-by: Cody Northrop <cnorthrop@google.com>
Charlie Lao dee4d7a5 2020-04-10T10:22:56 Vulkan: Early fragment tests optimization Checks if early fragment tests as an optimization is feasible and enable it if we can. In the link time, if context state diagrees with optimization (in rare case), then remove the ExecutionModeEarlyFragmentTests sprv op code. Bug: angleproject:4508 Change-Id: Ifbb06c0ffb050a9f3ddb16ab50362e908b4b9cf6 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2136490 Commit-Queue: Charlie Lao <cclao@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Reviewed-by: Courtney Goeltzenleuchter <courtneygo@google.com>
Xinghua Cao 2caf7ffd 2020-03-23T12:12:42 Address a TODO for instance uniform block This patch resolves one TODO that Support to translate instance uniform block containing only a large array member to Structured buffer on D3D backend when necessary. Bug: angleproject:4205 Change-Id: If2cd6cf633080820ea33e52269d7d86cd587c9ee Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2114912 Commit-Queue: Xinghua Cao <xinghua.cao@intel.com> Reviewed-by: Jiajia Qin <jiajia.qin@intel.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Mohan Maiya 96c26c68 2020-04-03T07:52:52 Add support for NV_shader_noperspective_interpolation Added support for GL_NV_shader_noperspective_interpolation on the Vulkan and Desktop OpenGL backends Bug: angleproject:4388 Test: angle_end2end_tests --gtest_filter=ShaderInterpTest.NoPerspective/* Change-Id: I12473830c0ea8b4fffeae9c4a8ec92d979c8e18c Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2107234 Commit-Queue: Mohan Maiya <m.maiya@samsung.com> Reviewed-by: Geoff Lang <geofflang@chromium.org>
Mohan Maiya ca2b6e1f 2020-03-19T16:30:21 Vulkan: Implement EXT_shader_non_constant_global_initializers Enables the translator to handle non-constant global initializers in all essl versions iff the shader enables the extension to do so. Bug: angleproject:4468 Test: angle_end2end_tests --gtest_filter=ShaderNonConstGlobalInitializerTest.* Change-Id: I8f138c12fc83d2f38ff8f45ca9133222b01e4087 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2102959 Reviewed-by: Tobin Ehlis <tobine@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Mohan Maiya <m.maiya@samsung.com>
Cody Northrop a2f9ad39 2020-03-09T14:28:39 Vulkan: Remove unused atomic counter builtins Atomic counters are not supported by Vulkan. Most are already converted by the RewriteAtomicCounters traversal, but that is only invoked when atomic counters are active. This CL introduces another pass that removes any atomic counter builtin that was not handled by the previous pass. It also will assert if it sees any atomic counters active, thus ensuring it is only used when needed. Test: KHR-GLES31.core.compute_shader.shared-struct Test: angle_end2end_tests.exe --gtest_filter="*AtomicCounter*" Bug: angleproject:4189 Bug: b:150310216 Change-Id: I61d10e954886dc94fede8b344f5a0ede3b689adb Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2095688 Commit-Queue: Cody Northrop <cnorthrop@google.com> Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org>
Tobin Ehlis 1a5c7a16 2020-02-25T12:02:31 Reland "Vulkan:Include precision qualifier in GLSL" Currently still ignoring precision qualifiers for Vulkan shaders by default, but have added feature "enablePrecisionQualifiers" that can be enabled in order to include precision qualifiers. With this initial implementation, it's possible to get precision qualifier mis-matches in the generated GLSL 4.50. According to the spec this is allowed. From GLSLangSpec 4.50 section 4.7 "Precision and Precision Qualifiers": For the purposes of determining if an output from one shader stage matches an input of the next stage, the precision qualifier need not match. However, when converted to SPIR-V and run through the shader validation any mismatches will cause shader validation errors. Initially just ignoring those errors with this commit. Bug: angleproject:3078 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2057749 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tobin Ehlis <tobine@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Tobin Ehlis <tobine@google.com> Change-Id: Ieecca604bb2c834c9b1c2bcab85279d1f8755dfa Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2086280
Ian Elliott 3797fc85 2020-03-05T11:44:28 Generate new compiler errors for array-of-arrays The GLES GLSL 3.1 spec states that it is an error to use array-of-arrays in the following cases: - Vertex shader outputs - Fragment shader inputs - Fragment shader outputs ANGLE currently generates errors in the case of vertex shader inputs. It needs to add the above cases. Reproduce by running "angle_deqp_khr_gles31_tests" with: --gtest_filter=dEQP.KHR_GLES31/core_arrays_of_arrays_InteractionInterfaceArrays2* --use-angle=swiftshader Bug: angleproject:4148 Change-Id: I1147fc7d4755e5b6c07d2ab5d14566d919dec530 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2087647 Reviewed-by: Cody Northrop <cnorthrop@google.com> Commit-Queue: Ian Elliott <ianelliott@google.com>
Tobin Ehlis a60e4973 2020-02-27T20:11:20 Revert "Vulkan:Include precision qualifier in GLSL" This reverts commit 93e72f5f0eb80b89334ea8fe577b32901a00b5fb. Reason for revert: This is having a negative impact on gfxbench Manhattan 3.1 tests (20-30%) with only a minor improvement (~4%) in TRex offscreen. Pulling out for now and need to re-evaluate best solution here. Original change's description: > Vulkan:Include precision qualifier in GLSL > > With this initial implementation, it's possible to get precision > qualifier mis-matches in the generated GLSL 4.50. According to the > spec this is allowed. From GLSLangSpec 4.50 section 4.7 "Precision and > Precision Qualifiers": > > For the purposes of determining if an output from one shader stage > matches an input of the next stage, the precision qualifier need not > match. > > However, when converted to SPIR-V and run through the shader validation > any mismatches will cause shader validation errors. Initially just > ignoring those errors with this commit. > > Bug: angleproject:3078 > Change-Id: Ia2b04f90f0a7b6b1302c1b1b3e32bcfd8db9ed49 > Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2057749 > Reviewed-by: Jamie Madill <jmadill@chromium.org> > Reviewed-by: Tobin Ehlis <tobine@google.com> > Reviewed-by: Tim Van Patten <timvp@google.com> > Commit-Queue: Tobin Ehlis <tobine@google.com> TBR=tobine@google.com,timvp@google.com,jmadill@chromium.org # Not skipping CQ checks because original CL landed > 1 day ago. Bug: angleproject:3078 Change-Id: I84984e2a963bde573f0f1de585001ceb83623c89 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2079334 Reviewed-by: Tobin Ehlis <tobine@google.com> Commit-Queue: Yuly Novikov <ynovikov@chromium.org>
Ian Elliott e488d8b8 2020-02-26T16:11:29 Vulkan: Implement Android pre-rotation As an Android GLES driver on top of Vulkan, ANGLE must pre-rotate rendering on behalf of the application. This involves modifying the vertex shader to multiply gl_Position with a mat2 "rotation matrix". Not doing so means that SurfaceFlinger (SF) will perform a costly rotation blit before presenting every image. Setting WindowSurfaceVk::mPreTransform to mCurrentTransform tells SF to not do the blit. When the surface is rotated 90 or 270 degrees, the width and height must be swapped for: - The swapchain images, and for any depth, stencil, and/or multisample attachments used with the swapchain image. - The viewport, scissor, and render area. Because the Metal back-end shares the TranslatorVulkan, it will define the same preRotation (mat2) DriverUniform that is used for Vulkan. Bug: angleproject:3502 Change-Id: I968dbe8869ba0f50de18dd41f1195e847c06b545 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2038272 Commit-Queue: Ian Elliott <ianelliott@google.com> Reviewed-by: Jamie Madill <jmadill@chromium.org>
Tobin Ehlis 93e72f5f 2020-02-25T12:02:31 Vulkan:Include precision qualifier in GLSL With this initial implementation, it's possible to get precision qualifier mis-matches in the generated GLSL 4.50. According to the spec this is allowed. From GLSLangSpec 4.50 section 4.7 "Precision and Precision Qualifiers": For the purposes of determining if an output from one shader stage matches an input of the next stage, the precision qualifier need not match. However, when converted to SPIR-V and run through the shader validation any mismatches will cause shader validation errors. Initially just ignoring those errors with this commit. Bug: angleproject:3078 Change-Id: Ia2b04f90f0a7b6b1302c1b1b3e32bcfd8db9ed49 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2057749 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tobin Ehlis <tobine@google.com> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Tobin Ehlis <tobine@google.com>
Tobin Ehlis a1ceb87b 2020-02-25T11:58:19 Vulkan:Refactor precision emulation code GLSL, ESSL, and HLSL translators all had the same code to emulate precision so migrate that code to base class. This is a structural change to prepare for honoring precision qualifiers in the Vulkan translator as well, which is done in a follow-on commit. Bug: angleproject:3078 Change-Id: Id85bba83543189e3bbc4002be92b056e63afc8de Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2072644 Reviewed-by: Jamie Madill <jmadill@chromium.org> Reviewed-by: Tim Van Patten <timvp@google.com> Commit-Queue: Tobin Ehlis <tobine@google.com>
Shahbaz Youssefi 1f5f7ea3 2020-02-14T23:39:11 Vulkan: Fix SPIR-V transformation name-info association Prior to this commit, when "OpName %id name" was encountered, the info corresponding to "name" was immediately associated with %id. This is not necessarily correct because there could be multiple ids with the same name. For example a sampler declaration and an unrelated function argument could have the same name. In this case, the sampler declaration and function argument name don't even need to be in the same shader stage. This change modifies the SPIR-V transformation such that the name-id mapping is tracked until the OpVariable instruction that actually declares the variable is visited. The mapping to variable info is only done if the storage class specified in this instruction corresponds to a shader interface variable. Bug: angleproject:3394 Change-Id: I35a1f6f8278e4b1ad81c9955a55e1b72d6f2e4ea Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2057248 Reviewed-by: Shahbaz Youssefi <syoussefi@chromium.org> Commit-Queue: Shahbaz Youssefi <syoussefi@chromium.org>
Yuly Novikov bb7534ee 2020-02-14T16:23:28 Use ASSERT() instead of assert() when possible. Since ASSERT() works in Release builds. Left assert() in a few places where calling a function may not be safe. Bug: angleproject:4396 Change-Id: Ic75a3d41d846e327097f8c37fe2336dcd3be6cb8 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/2057745 Reviewed-by: Jamie Madill <jmadill@chromium.org> Commit-Queue: Yuly Novikov <ynovikov@chromium.org>
kmx.io     mail     discord kmxgit v0.4.0