GPU for the Web Working Group - Activity in 2023
No matches.
TR publications
-
3 documents published.
-
WebGPU published twice as a Working Draft. History of WebGPU
-
WebGPU Shading Language published as a Working Draft. History of WebGPU Shading Language
-
GitHub
gpuweb/gpuweb
-
18 issues opened.
-
Surely it shouldn't take 5-15ms to get a trivial result back from GPU with mapAsync? gpuweb/gpuweb#4432, by @kjpgit.
-
Weird behavior while using switch code block gpuweb/gpuweb#4430, by @ch3coohlink.
-
https://developer.chrome.com/blog/webgpu-release?hl=ar#a_new_dawn_for_web_graphics gpuweb/gpuweb#4429, by @Muzaffar710.
-
add swizzles for scalar->vectors conversion too gpuweb/gpuweb#4428, by @skal65535.
-
Occlusion Query - result type not specified gpuweb/gpuweb#4427, by @aardgoose.
-
createTexture does not validate viewFormats against usage gpuweb/gpuweb#4426, by @austinEng.
-
using sample_index makes fragment shader get called for every sample gpuweb/gpuweb#4420, by @greggman.
-
Compat: Validating out `@interpolate(linear)` and `@interpolate(..., sample)` gpuweb/gpuweb#4419, by @greggman.
-
Bit shift operating missing overload for AbstractInt shiftees and `i32` shift amounts gpuweb/gpuweb#4418, by @jimblandy.
-
Compat: renderable float16, float32 textures gpuweb/gpuweb#4417, by @greggman.
-
Slight type problem with `GPURenderPipeline.[[writesDepth]]` gpuweb/gpuweb#4416, by @jimblandy.
-
Conversion from AbstractInteger to i32 appears to be underspecified gpuweb/gpuweb#4415, by @mwyrzykowski.
-
Define a `WGSLFeatureName` IDL enum. gpuweb/gpuweb#4414, by @Kangz.
-
Compat: Validate out storage texture formats that are not supported in GLES gpuweb/gpuweb#4413, by @greggman.
-
ARM64 DX Issues with Container gpuweb/gpuweb#4412, by @mehmetoguzderin.
-
Compat: Validating out viewFormats / view reinterpretation gpuweb/gpuweb#4404, by @greggman.
-
-
7 issues closed.
-
Dictionary default values referencing type members gpuweb/gpuweb#278.
-
RFC: Make sure that a minimal-garbage hot path is supported gpuweb/gpuweb#214.
-
using sample_index makes fragment shader get called for every sample gpuweb/gpuweb#4420.
-
Define a `WGSLFeatureName` IDL enum. gpuweb/gpuweb#4414.
-
39 issues received 68 new comments.
-
Conversion from AbstractInteger to i32 appears to be underspecified gpuweb/gpuweb#4415 (7 comments by @ben-clayton, @jimblandy, @mwyrzykowski)
-
Surely it shouldn't take 5-15ms to get a trivial result back from GPU with mapAsync? gpuweb/gpuweb#4432 (6 comments by @conorpo, @kjpgit)
-
Weird behavior while using switch code block gpuweb/gpuweb#4430 (5 comments by @ben-clayton, @ch3coohlink, @greggman)
-
`requiredLimits` must be an ordinary JS object gpuweb/gpuweb#4277 (3 comments by @greggman, @kainino0x)
-
Compat: Validate out storage texture formats that are not supported in GLES gpuweb/gpuweb#4413 (3 comments by @greggman, @teoxoy)
-
Define a `WGSLFeatureName` IDL enum. gpuweb/gpuweb#4414 (3 comments by @kainino0x, @Kangz, @mwyrzykowski)
-
Allow mutable function parameters with the mut keyword. gpuweb/gpuweb#4113 (2 comments by @ch3coohlink, @kdashg)
-
add a system-defined wsgl namespace for predeclared objects gpuweb/gpuweb#4308 (2 comments by @greggman, @teoxoy)
-
Bit shift operating missing overload for AbstractInt shiftees and `i32` shift amounts gpuweb/gpuweb#4418 (2 comments by @conorpo, @jimblandy)
-
Add a floored or euclidean modulo implementation to wgsl gpuweb/gpuweb#3987 (2 comments by @kdashg, @stegu)
-
Size and alignment of bool type is missing from https://gpuweb.github.io/gpuweb/wgsl/#alignment-and-size ? gpuweb/gpuweb#4406 (2 comments by @dj2, @mwyrzykowski)
-
https://developer.chrome.com/blog/webgpu-release?hl=ar#a_new_dawn_for_web_graphics gpuweb/gpuweb#4429 (1 comment by @Muzaffar710)
-
createTexture does not validate viewFormats against usage gpuweb/gpuweb#4426 (1 comment by @austinEng)
-
using sample_index makes fragment shader get called for every sample gpuweb/gpuweb#4420 (1 comment by @greggman)
-
Add extensions for member accessors and functions (maybe with traits) gpuweb/gpuweb#4337 (1 comment by @kdashg)
-
Support "uniform texel buffer" and "storage texel buffer" ? gpuweb/gpuweb#162 (1 comment by @kdashg)
-
Early Depth/Stencil test directive or flag in pipeline state creation gpuweb/gpuweb#398 (1 comment by @kdashg)
-
Missing intrinsic functions like rcp and rsqrt that are useful in other shading languages gpuweb/gpuweb#4092 (1 comment by @kdashg)
-
Compat: Validating out viewFormats / view reinterpretation gpuweb/gpuweb#4404 (1 comment by @kainino0x)
-
-
8 pull requests opened.
-
Explain the endianness of pack4xI8* and pack4xU8* builtins gpuweb/gpuweb#4425, by @dneto0.
-
Add Compat interpolation validation gpuweb/gpuweb#4424, by @greggman.
-
Add Compat rg32xxx storage texture restrictions gpuweb/gpuweb#4423, by @greggman.
-
CI check gpuweb/gpuweb#4411, by @mehmetoguzderin.
-
Hotfix1: Even more explicitly run install Puppeteer gpuweb/gpuweb#4410, by @mehmetoguzderin.
-
Hotfix: Explicitly invoke Puppeteer Install gpuweb/gpuweb#4409, by @mehmetoguzderin.
-
Reorganize WGSL tools gpuweb/gpuweb#4408, by @mehmetoguzderin.
-
minor tweaks gpuweb/gpuweb#4407, by @bbbbx.
-
-
7 pull requests merged.
-
Add subgroups proposal gpuweb/gpuweb#4368.
-
Hotfix1: Even more explicitly run install Puppeteer gpuweb/gpuweb#4410.
-
Hotfix: Explicitly invoke Puppeteer Install gpuweb/gpuweb#4409.
-
Reorganize WGSL tools gpuweb/gpuweb#4408.
-
minor tweaks gpuweb/gpuweb#4407.
-
WGSL: Fix explanation of hyperbolic angles. gpuweb/gpuweb#4351.
-
Clarify matching of pipeline constant identifiers. gpuweb/gpuweb#4382.
-
-
1 pull request closed.
-
CI check gpuweb/gpuweb#4411.
-
-
6 pull requests received 33 new comments.
-
Add subgroups proposal gpuweb/gpuweb#4368 (21 comments by @alan-baker, @dneto0, @kdashg, @raphlinus, @teoxoy)
-
gpuweb/cts
-
7 issues opened.
-
Add Uniformity Analaysis tests for syntax sugar for dereferencing composites gpuweb/cts#3226, by @amaiorano.
-
Add tests for syntax sugar for dereferencing composites gpuweb/cts#3225, by @amaiorano.
-
wgsl: Add FPInterval to the types for expected in FP unittests gpuweb/cts#3210, by @zoddicus.
-
wgsl: Change domain restriction pattern for numeric framework gpuweb/cts#3201, by @zoddicus.
-
image_copy CTS tests recently started failing for depth formats gpuweb/cts#3199, by @mwyrzykowski.
-
4 issues closed.
-
wgsl: Change domain restriction pattern for numeric framework gpuweb/cts#3201.
-
wgsl: AbstractFloat execution tests for `mix` builtin gpuweb/cts#2566.
-
wgsl: Resolve how OOB intervals behave in the numeric framework gpuweb/cts#3139.
-
image_copy CTS tests recently started failing for depth formats gpuweb/cts#3199.
-
-
5 issues received 8 new comments.
-
wgsl: Resolve how OOB intervals behave in the numeric framework gpuweb/cts#3139 (2 comments by @zoddicus)
-
image_copy CTS tests recently started failing for depth formats gpuweb/cts#3199 (2 comments by @greggman, @mwyrzykowski)
-
fixture.ts is checking for GPUPipelineError.stack to exist but it is not required on DOMException gpuweb/cts#3222 (1 comment by @mwyrzykowski)
-
-
27 pull requests opened.
-
Copy reftest html files to out/ for gh-pages gpuweb/cts#3231, by @kainino0x.
-
reftests: Allow large totalPixels for fuzzy matching gpuweb/cts#3230, by @kainino0x.
-
Test builtin(position) and interpolation gpuweb/cts#3229, by @greggman.
-
Compat: skip tests use rg32xxx formats with createBindGroupLayout gpuweb/cts#3228, by @greggman.
-
Compat: Fix tests that use @interpole(linear/sample) gpuweb/cts#3227, by @greggman.
-
Compat:Check interpolate linear/sample validation gpuweb/cts#3224, by @greggman.
-
Compat: Skip or expect error for rg32xxx texture formats gpuweb/cts#3223, by @greggman.
-
Test construction by reflection gpuweb/cts#3221, by @greggman.
-
Test that constructible objects are actually constructible gpuweb/cts#3220, by @greggman.
-
Add readonly and readwrite storage textures as bindable resources gpuweb/cts#3219, by @Jiawei-Shao.
-
Revert "Adjust fragment state blend factor and write mask validation" gpuweb/cts#3218, by @kainino0x.
-
Adjust fragment state blend factor and write mask validation gpuweb/cts#3217, by @shrekshao.
-
wgsl: Change 'abstract' -> 'abstract_float' gpuweb/cts#3216, by @zoddicus.
-
[WIP] Adding a more complex compute pass memory sync test gpuweb/cts#3215, by @toji.
-
Test readonly and readwrite storage textures in in_render_{common|misc} gpuweb/cts#3214, by @Jiawei-Shao.
-
Comapt: Fix createTexture:'viewFormats' test for compat gpuweb/cts#3213, by @greggman.
-
Compat format reinterpretation gpuweb/cts#3212, by @greggman.
-
Compat: Check viewFormats match format gpuweb/cts#3211, by @greggman.
-
Test read-only and read-write storage texture in in_pass_encoder gpuweb/cts#3209, by @Jiawei-Shao.
-
wgsl: Change *Unbounded[Interval|Vector|...] to *Infinite[Interval|Vector|...] gpuweb/cts#3208, by @zoddicus.
-
wgsl: Change the term `bounds` to `endpoints` for intervals gpuweb/cts#3207, by @zoddicus.
-
wgsl: Rework how input domain domains are limited gpuweb/cts#3206, by @zoddicus.
-
Fix ufloat_texel_data_in_shader value gpuweb/cts#3205, by @shrekshao.
-
Compat: Fix compat tests for depthOrArrayLayers restrictions gpuweb/cts#3204, by @greggman.
-
Compat: Fixes for format reinterpretation gpuweb/cts#3203, by @greggman.
-
Compat: Fix in_pass_misc.spec.ts for textureBindingViewDimension gpuweb/cts#3202, by @greggman.
-
Fix image_copy tests gpuweb/cts#3200, by @greggman.
-
-
26 pull requests merged.
-
Test builtin(position) and interpolation gpuweb/cts#3229.
-
reftests: Allow large totalPixels for fuzzy matching gpuweb/cts#3230.
-
Copy reftest html files to out/ for gh-pages gpuweb/cts#3231.
-
Test that constructible objects are actually constructible gpuweb/cts#3220.
-
Compat: skip tests use rg32xxx formats with createBindGroupLayout gpuweb/cts#3228.
-
Compat: Fix tests that use @interpole(linear/sample) gpuweb/cts#3227.
-
Compat:Check interpolate linear/sample validation gpuweb/cts#3224.
-
Compat: Skip or expect error for rg32xxx texture formats gpuweb/cts#3223.
-
Test construction by reflection gpuweb/cts#3221.
-
Add readonly and readwrite storage textures as bindable resources gpuweb/cts#3219.
-
Adjust fragment state blend factor and write mask validation gpuweb/cts#3217.
-
wgsl: Rework how input domain domains are limited gpuweb/cts#3206.
-
wgsl: Add AbstractFloat `mix` execution tests gpuweb/cts#3087.
-
Test read-only and read-write storage texture in in_pass_encoder gpuweb/cts#3209.
-
Fix ufloat_texel_data_in_shader value gpuweb/cts#3205.
-
Comapt: Fix createTexture:'viewFormats' test for compat gpuweb/cts#3213.
-
Compat format reinterpretation gpuweb/cts#3212.
-
Compat: Check viewFormats match format gpuweb/cts#3211.
-
wgsl: Change the term `bounds` to `endpoints` for intervals gpuweb/cts#3207.
-
Compat: Fix compat tests for depthOrArrayLayers restrictions gpuweb/cts#3204.
-
Compat: Fixes for format reinterpretation gpuweb/cts#3203.
-
Compat: Fix in_pass_misc.spec.ts for textureBindingViewDimension gpuweb/cts#3202.
-
Fix image_copy tests gpuweb/cts#3200.
-
-
1 pull request closed.
-
14 pull requests received 66 new comments.
-
Test builtin(position) and interpolation gpuweb/cts#3229 (17 comments by @greggman, @lokokung, @shrekshao)
-
wgsl: Change *Unbounded[Interval|Vector|...] to *Infinite[Interval|Vector|...] gpuweb/cts#3208 (12 comments by @ben-clayton, @jiangzhaoming, @zoddicus)
-
wgsl: Change the term `bounds` to `endpoints` for intervals gpuweb/cts#3207 (8 comments by @jiangzhaoming, @zoddicus)
-
Handle async pipeline creation errors more gracefully gpuweb/cts#3158 (6 comments by @ben-clayton, @kainino0x, @toji)
-
Compat: Skip or expect error for rg32xxx texture formats gpuweb/cts#3223 (6 comments by @greggman, @shrekshao)
-
Adjust fragment state blend factor and write mask validation gpuweb/cts#3217 (5 comments by @kainino0x, @shrekshao)
-
Test that constructible objects are actually constructible gpuweb/cts#3220 (2 comments by @greggman, @kainino0x)
-
validate dp4a builtins (packed_4x8_integer_dot_product language feature) gpuweb/cts#3188 (1 comment by @dneto0)
-
TR publications
-
9 documents published.
-
WebGPU Shading Language published 3 times as a Working Draft. History of WebGPU Shading Language
-
WebGPU published 6 times as a Working Draft. History of WebGPU
-
Participants
-
4 participants joined.
-
David Neto (Google LLC) joined.
-
Jim Blandy (Mozilla Foundation) joined.
-
Brandon Jones (Google LLC) joined.
-
Alan Baker (Google LLC) joined.
-
-
1 participant left.
-
Myles Maxfield (Apple Inc.) left.
-
GitHub
gpuweb/cts
-
26 issues opened.
-
wgsl: AF `degrees` execution tests failing gpuweb/cts#3194, by @zoddicus.
-
Test HDR video in cts gpuweb/cts#3193, by @shaoboyan.
-
Update `fp_primer.md` to clarify OOB behaviours for Vectors and Matrices gpuweb/cts#3191, by @zoddicus.
-
wgsl: Change usage of `abstract` to `abstract_float` gpuweb/cts#3184, by @zoddicus.
-
wgsl: test unrestricted_pointer_parameters WGSL language extension gpuweb/cts#3181, by @dneto0.
-
wgsl: test dp4a ( packed_4x8_integer_dot_product ) gpuweb/cts#3180, by @dneto0.
-
wgsl: Expose various test range generators in FPTraits gpuweb/cts#3174, by @zoddicus.
-
wgsl: Add AbstractInt `select` execution tests gpuweb/cts#3173, by @zoddicus.
-
wgsl: Execution tests for matrix accessors gpuweb/cts#3172, by @zoddicus.
-
wgsl: Execution tests for array and vector accessors gpuweb/cts#3171, by @zoddicus.
-
wgsl: Execution tests for vector accessors gpuweb/cts#3170, by @zoddicus.
-
wgsl: Execution tests for vector swizzles gpuweb/cts#3169, by @zoddicus.
-
Reduce or defer creation of strings for diagnostics gpuweb/cts#3166, by @zoddicus.
-
Reduce unnecessary logging for validation tests gpuweb/cts#3165, by @zoddicus.
-
`run:lint` warnings not showing with `npm test` gpuweb/cts#3164, by @ben-clayton.
-
Add a unhandledrejection event handler gpuweb/cts#3159, by @kainino0x.
-
Uncaught exception in shader execution tests: gpuweb/cts#3157, by @ben-clayton.
-
Put Exceptions/Validation errors at the top in standalone gpuweb/cts#3147, by @greggman.
-
texture_zero.spec.ts makes compat incompatible texture views gpuweb/cts#3145, by @greggman.
-
val: Test depthSlice from GPURenderPassColorAttachment gpuweb/cts#3144, by @haoxli.
-
image_copy.spec.ts makes compat incompatible texture views gpuweb/cts#3143, by @greggman.
-
wgsl: Resolve how OOB intervals behave in the numeric framework gpuweb/cts#3139, by @zoddicus.
-
eslint (with plugins) takes a very long time and uses a lot of memory gpuweb/cts#3135, by @kainino0x.
-
autoformat-out-wpt step broken after depedency upgrade gpuweb/cts#3128, by @kainino0x.
-
Enforce TS's `readonly` more strictly gpuweb/cts#3126, by @zoddicus.
-
-
8 issues closed.
-
wgsl: AF `degrees` execution tests failing gpuweb/cts#3194.
-
wgsl: Expose various test range generators in FPTraits gpuweb/cts#3174.
-
framework: add ability filter device based on WGSLLanguageFeatures gpuweb/cts#3057.
-
`run:lint` warnings not showing with `npm test` gpuweb/cts#3164.
-
image_copy.spec.ts makes compat incompatible texture views gpuweb/cts#3143.
-
wgsl: AbstractFloat execution test for `round` builtin gpuweb/cts#2509.
-
Add discussion of AbstractFloat behaviour to `fp_primer.md` gpuweb/cts#3111.
-
-
9 issues received 20 new comments.
-
`run:lint` warnings not showing with `npm test` gpuweb/cts#3164 (8 comments by @ben-clayton, @kainino0x)
-
image_copy.spec.ts makes compat incompatible texture views gpuweb/cts#3143 (3 comments by @greggman, @kainino0x)
-
wgsl: test unrestricted_pointer_parameters WGSL language extension gpuweb/cts#3181 (1 comment by @dneto0)
-
wgsl: Add AbstractFloat being tested to be as good as f32 to fp_primer.md gpuweb/cts#3079 (1 comment by @zoddicus)
-
Intel + DXC failure: webgpu:shader,execution,flow_control,loop:nested_loops:preventValueOptimizations=false gpuweb/cts#3063 (1 comment by @jiangzhaoming)
-
-
50 pull requests opened.
-
Compat: test depthOrArrayLayers vs textureBindingViewFormat gpuweb/cts#3198, by @greggman.
-
Remove compression on the case cache, split files to reduce transitive deps gpuweb/cts#3197, by @ben-clayton.
-
Compat: Update maxInterStageShaderVariables = 15 gpuweb/cts#3196, by @greggman.
-
wgsl: Update cache string for AF `degrees` execution tests gpuweb/cts#3195, by @zoddicus.
-
Compat image copy test depth stencil gpuweb/cts#3192, by @greggman.
-
default entry points to shader modules gpuweb/cts#3190, by @beaufortfrancois.
-
validate dp4a builtins (packed_4x8_integer_dot_product language feature) gpuweb/cts#3188, by @dneto0.
-
Pin the JavaScript runtime version for the data cache gpuweb/cts#3187, by @ben-clayton.
-
Add tests on createBindGroupLayout() with new storage texture accesses gpuweb/cts#3186, by @Jiawei-Shao.
-
wgsl: Add `sparseMatrixRange` to FPTraits API gpuweb/cts#3183, by @zoddicus.
-
wgsl: Add `scalarRange` and `sparseScalarRange` to FPTraits API gpuweb/cts#3182, by @zoddicus.
-
Add tools/server for server under ts-node, and update help gpuweb/cts#3179, by @kainino0x.
-
Add documentation for the case cache gpuweb/cts#3178, by @ben-clayton.
-
Ignore compilation error in writeTimestamp tests gpuweb/cts#3177, by @haoxli.
-
wgsl: Add `vectorRange` and `sparseVectorRange` to FPTraits API gpuweb/cts#3176, by @zoddicus.
-
Use `findFailedPixels` in `copyTextureToTexture` gpuweb/cts#3175, by @shrekshao.
-
Switch from grunt-parallel to grunt-concurrent gpuweb/cts#3168, by @kainino0x.
-
Updates for writeTimestamp removal gpuweb/cts#3167, by @kainino0x.
-
Summarize the max severity at the top of logs, star all max-severity logs gpuweb/cts#3163, by @kainino0x.
-
Add case cache to repo gpuweb/cts#3162, by @ben-clayton.
-
Reapply "Skip compressed image_copy tests in Compatibility mode. (#31… gpuweb/cts#3161, by @SenorBlanco.
-
Delete the out-node build gpuweb/cts#3160, by @kainino0x.
-
Handle async pipeline creation errors more gracefully gpuweb/cts#3158, by @toji.
-
Standalone: call out exceptions at the top of results gpuweb/cts#3156, by @greggman.
-
Compat: fix in-render-misc.spec.ts for compat gpuweb/cts#3155, by @greggman.
-
Compat: fix in_pass tests. gpuweb/cts#3154, by @greggman.
-
Compat: fix createBindGroup validation test for compat gpuweb/cts#3153, by @greggman.
-
Compat: Fix texture_zero init tests for compat gpuweb/cts#3152, by @greggman.
-
Actions: shorten timeout and enable cancel-in-progress gpuweb/cts#3151, by @kainino0x.
-
Compat fix image copy gpuweb/cts#3150, by @greggman.
-
Revert "Skip compressed image_copy tests in Compatibility mode." gpuweb/cts#3149, by @greggman.
-
Compat: copyTextureToTexture validation gpuweb/cts#3148, by @greggman.
-
Fix use of deprecated window.name. gpuweb/cts#3146, by @Kangz.
-
Make the CTS debuggable again gpuweb/cts#3142, by @greggman.
-
Fixes error scope tests for Linux/Intel gpuweb/cts#3141, by @lokokung.
-
Skip compressed image_copy tests in Compatibility mode. gpuweb/cts#3140, by @SenorBlanco.
-
Skip tests using HTMLVideoElement when it is undefined. gpuweb/cts#3138, by @Kangz.
-
Allow 3D renderable texture in validation. gpuweb/cts#3137, by @Kangz.
-
Update validation tests to allow separate read-only depth-stencil. gpuweb/cts#3136, by @Kangz.
-
Make grunt print how long each step takes gpuweb/cts#3134, by @kainino0x.
-
Fold tools/presubmit into tools/validate gpuweb/cts#3133, by @kainino0x.
-
Test errors for out-of-range limit requests gpuweb/cts#3132, by @kainino0x.
-
Remove preview workflow, do WPT lint in PR workflow gpuweb/cts#3131, by @kainino0x.
-
Parallelize Grunt build gpuweb/cts#3130, by @kainino0x.
-
Restore missing test path validation gpuweb/cts#3129, by @kainino0x.
-
Upgrade all dependencies, enable more warnings/lints gpuweb/cts#3127, by @kainino0x.
-
Speed up shader,execution,expression tests with async pipelines gpuweb/cts#3125, by @toji.
-
Prevent Double Runs and make run clearer. gpuweb/cts#3124, by @greggman.
-
Optimize some checkElements* implementations gpuweb/cts#3123, by @kainino0x.
-
-
51 pull requests merged.
-
Compat: test depthOrArrayLayers vs textureBindingViewFormat gpuweb/cts#3198.
-
Compat image copy test depth stencil gpuweb/cts#3192.
-
Standalone: call out exceptions at the top of results gpuweb/cts#3156.
-
wgsl: Update cache string for AF `degrees` execution tests gpuweb/cts#3195.
-
Compat: Update maxInterStageShaderVariables = 15 gpuweb/cts#3196.
-
default entry points to shader modules gpuweb/cts#3190.
-
Compat: Refactor memory module tests to use 4 storage buffers gpuweb/cts#3117.
-
wgsl: Add `sparseMatrixRange` to FPTraits API gpuweb/cts#3183.
-
framework: allow skipping tests based on wgslLanguageFeatures gpuweb/cts#3058.
-
wgsl: Add `scalarRange` and `sparseScalarRange` to FPTraits API gpuweb/cts#3182.
-
wgsl: Add `vectorRange` and `sparseVectorRange` to FPTraits API gpuweb/cts#3176.
-
Ignore compilation error in writeTimestamp tests gpuweb/cts#3177.
-
Add documentation for the case cache gpuweb/cts#3178.
-
Add tools/server for server under ts-node, and update help gpuweb/cts#3179.
-
Compat: fix in_pass tests. gpuweb/cts#3154.
-
Add case cache to repo gpuweb/cts#3162.
-
Use `findFailedPixels` in `copyTextureToTexture` gpuweb/cts#3175.
-
Updates for writeTimestamp removal gpuweb/cts#3167.
-
Switch from grunt-parallel to grunt-concurrent gpuweb/cts#3168.
-
Compat: fix createBindGroup validation test for compat gpuweb/cts#3153.
-
Compat: fix in-render-misc.spec.ts for compat gpuweb/cts#3155.
-
Compat: Fix texture_zero init tests for compat gpuweb/cts#3152.
-
Actions: shorten timeout and enable cancel-in-progress gpuweb/cts#3151.
-
Compat fix image copy gpuweb/cts#3150.
-
Compat: copyTextureToTexture validation gpuweb/cts#3148.
-
Fixes error scope tests for Linux/Intel gpuweb/cts#3141.
-
Fix use of deprecated window.name. gpuweb/cts#3146.
-
Allow 3D renderable texture in validation. gpuweb/cts#3137.
-
Skip tests using HTMLVideoElement when it is undefined. gpuweb/cts#3138.
-
Update validation tests to allow separate read-only depth-stencil. gpuweb/cts#3136.
-
Skip compressed image_copy tests in Compatibility mode. gpuweb/cts#3140.
-
wgsl: Add AbstractFloat `round` execution tests gpuweb/cts#3080.
-
Compat: Test textureBindingViewDimension validation gpuweb/cts#3098.
-
Parallelize Grunt build gpuweb/cts#3130.
-
Test errors for out-of-range limit requests gpuweb/cts#3132.
-
Fold tools/presubmit into tools/validate gpuweb/cts#3133.
-
Make grunt print how long each step takes gpuweb/cts#3134.
-
Speed up shader,execution,expression tests with async pipelines gpuweb/cts#3125.
-
Add details about 'atleast as good as f32' for AbstractFloats gpuweb/cts#3112.
-
Upgrade all dependencies, enable more warnings/lints gpuweb/cts#3127.
-
Audit TypedArray allocations gpuweb/cts#3114.
-
Remove preview workflow, do WPT lint in PR workflow gpuweb/cts#3131.
-
Restore missing test path validation gpuweb/cts#3129.
-
Optimize some checkElements* implementations gpuweb/cts#3123.
-
Reduce draw validation test parameterization gpuweb/cts#3122.
-
Prevent Double Runs and make run clearer. gpuweb/cts#3124.
-
wgsl: Convert `quantizeToI32/U32` to use `Math.trunc` gpuweb/cts#3120.
-
-
4 pull requests closed.
-
Rewrite the standalone test page with React gpuweb/cts#981.
-
Pin the JavaScript runtime version for the data cache gpuweb/cts#3187.
-
Revert "Skip compressed image_copy tests in Compatibility mode." gpuweb/cts#3149.
-
Generate many WPT html files for infra that doesn't support variants gpuweb/cts#1637.
-
-
32 pull requests received 225 new comments.
-
wgsl: Add `vectorRange` and `sparseVectorRange` to FPTraits API gpuweb/cts#3176 (21 comments by @jiangzhaoming, @zoddicus)
-
Test CopyEI2T and GPUExternalTexture convert video contents to display-p3 color space gpuweb/cts#3189 (18 comments by @kainino0x, @shaoboyan)
-
default entry points to shader modules gpuweb/cts#3190 (17 comments by @beaufortfrancois, @kainino0x)
-
framework: allow skipping tests based on wgslLanguageFeatures gpuweb/cts#3058 (13 comments by @ben-clayton, @dneto0, @kainino0x)
-
Compat: Refactor memory module tests to use 4 storage buffers gpuweb/cts#3117 (10 comments by @dneto0, @greggman)
-
Compat: Test textureBindingViewDimension validation gpuweb/cts#3098 (9 comments by @greggman, @kainino0x)
-
Pin the JavaScript runtime version for the data cache gpuweb/cts#3187 (8 comments by @ben-clayton, @kainino0x)
-
wgsl: Add AbstractFloat `mix` execution tests gpuweb/cts#3087 (8 comments by @ben-clayton, @zoddicus)
-
Add tests on createBindGroupLayout() with new storage texture accesses gpuweb/cts#3186 (7 comments by @ben-clayton, @jiangzhaoming, @Kangz)
-
wgsl: Add `sparseMatrixRange` to FPTraits API gpuweb/cts#3183 (7 comments by @ben-clayton, @jiangzhaoming, @zoddicus)
-
wgsl: Add `scalarRange` and `sparseScalarRange` to FPTraits API gpuweb/cts#3182 (7 comments by @jiangzhaoming, @zoddicus)
-
Reduce draw validation test parameterization gpuweb/cts#3122 (6 comments by @jiangzhaoming, @kainino0x)
-
Remove compression on the case cache, split files to reduce transitive deps gpuweb/cts#3197 (4 comments by @ben-clayton, @kainino0x)
-
Summarize the max severity at the top of logs, star all max-severity logs gpuweb/cts#3163 (4 comments by @greggman, @kainino0x)
-
Ignore compilation error in writeTimestamp tests gpuweb/cts#3177 (4 comments by @haoxli, @kainino0x)
-
validate dp4a builtins (packed_4x8_integer_dot_product language feature) gpuweb/cts#3188 (3 comments by @dneto0)
-
Add details about 'atleast as good as f32' for AbstractFloats gpuweb/cts#3112 (3 comments by @dneto0, @zoddicus)
-
Speed up shader,execution,expression tests with async pipelines gpuweb/cts#3125 (3 comments by @austinEng, @toji)
-
wgsl: Convert `quantizeToI32/U32` to use `Math.trunc` gpuweb/cts#3120 (3 comments by @jiangzhaoming, @zoddicus)
-
Actions: shorten timeout and enable cancel-in-progress gpuweb/cts#3151 (1 comment by @mehmetoguzderin)
-
Revert "Skip compressed image_copy tests in Compatibility mode." gpuweb/cts#3149 (1 comment by @greggman)
gpuweb/gpuweb
-
17 issues opened.
-
Shader source should be reflected in GPUShaderModule gpuweb/gpuweb#4399, by @greggman.
-
Clarification of behavior around GPUStoreOp.discard? gpuweb/gpuweb#4398, by @mwyrzykowski.
-
document ambiguity in tie-breaks for cube map sampling face selection gpuweb/gpuweb#4396, by @dneto0.
-
WebGPU copyTextureToBuffer auto convert YUV to RGB gpuweb/gpuweb#4394, by @huotong1212.
-
Clarify the set of attachments in render pass's colorAttachments gpuweb/gpuweb#4390, by @haoxli.
-
Make Setting canvas.width and canvas.height lighter weight gpuweb/gpuweb#4388, by @greggman.
-
Constructible types description is unclear around abstract types gpuweb/gpuweb#4385, by @jimblandy.
-
Allow `GPUBufferUsage.QUERY_RESOLVE | GPUBufferUsage.MAP_READ` buffer usage gpuweb/gpuweb#4383, by @beaufortfrancois.
-
Compat: validating out "2d" of more than 1 layer gpuweb/gpuweb#4378, by @kainino0x.
-
The `bgra8unorm-storage` feature should most likely only enable `write-only` access gpuweb/gpuweb#4377, by @teoxoy.
-
The output range of atan is wrong in the doc gpuweb/gpuweb#4376, by @shi-yan.
-
Compute shader branching depending on buffer value gpuweb/gpuweb#4373, by @h-a-n-n-e-s.
-
[FEATURE REQUEST] Pretty-print utility for GPUBufferUsage and other bitflags gpuweb/gpuweb#4372, by @tlecoz.
-
Type check for `device.queue.writeBuffer()`? gpuweb/gpuweb#4362, by @h-a-n-n-e-s.
-
Operator overloading gpuweb/gpuweb#4357, by @eliasboegel.
-
-
8 issues closed.
-
WebGPU copyTextureToBuffer auto convert YUV to RGB gpuweb/gpuweb#4394.
-
Add arrow operator as sugar for (*var). gpuweb/gpuweb#4114.
-
Default entryPoints to shader modules gpuweb/gpuweb#4342.
-
Proposal: Support DP4a as WGSL built-in functions gpuweb/gpuweb#2677.
-
Investigation: Render to 3D texture gpuweb/gpuweb#4251.
-
Compute shader branching depending on buffer value gpuweb/gpuweb#4373.
-
Type check for `device.queue.writeBuffer()`? gpuweb/gpuweb#4362.
-
Reference High Resolution Time spec gpuweb/gpuweb#4175.
-
-
2 issues reopened.
-
Uniformity annotations for global variables gpuweb/gpuweb#1791, by @kainino0x.
-
Problem measuring time deltas with writeTimestamp gpuweb/gpuweb#3952, by @haoxli.
-
-
47 issues received 125 new comments.
-
Problem measuring time deltas with writeTimestamp gpuweb/gpuweb#3952 (15 comments by @beaufortfrancois, @fintelia, @haoxli, @jimblandy, @kainino0x, @Kangz, @kdashg)
-
WebGPU Compatibility Mode gpuweb/gpuweb#4266 (14 comments by @greggman, @kainino0x, @kdashg, @mwyrzykowski, @SenorBlanco, @teoxoy)
-
The `bgra8unorm-storage` feature should most likely only enable `write-only` access gpuweb/gpuweb#4377 (7 comments by @kainino0x, @Kangz, @kdashg, @teoxoy)
-
[Feature Request] Vertex Shader storage read/write gpuweb/gpuweb#4355 (7 comments by @Kangz, @mwyrzykowski, @rconde01, @teoxoy, @tlecoz)
-
Type check for `device.queue.writeBuffer()`? gpuweb/gpuweb#4362 (5 comments by @greggman, @h-a-n-n-e-s)
-
A single shader is often split among many source files gpuweb/gpuweb#741 (5 comments by @greggman, @litherum, @Strange4)
-
Remove the need to hard-code texel format for `texture_storage_2d` in shaders gpuweb/gpuweb#4298 (4 comments by @dneto0, @kainino0x, @kdashg, @teoxoy)
-
Considerations for timing attacks using timestamp queries gpuweb/gpuweb#2218 (4 comments by @gyagp, @theb1rd)
-
Uniformity annotations for global variables gpuweb/gpuweb#1791 (3 comments by @kainino0x, @Kangz, @kdashg)
-
what happens if vertex w position is zero, or negative? gpuweb/gpuweb#4129 (3 comments by @aleino-nv, @jimblandy, @kainino0x)
-
Reference High Resolution Time spec gpuweb/gpuweb#4175 (3 comments by @fintelia, @jimblandy, @kainino0x)
-
Allow `GPUBufferUsage.QUERY_RESOLVE | GPUBufferUsage.MAP_READ` buffer usage gpuweb/gpuweb#4383 (2 comments by @kdashg, @toji)
-
`requiredLimits` must be an ordinary JS object gpuweb/gpuweb#4277 (2 comments by @kainino0x, @kdashg)
-
Shader source should be reflected in GPUShaderModule gpuweb/gpuweb#4399 (2 comments by @greggman, @kainino0x)
-
Clarification of behavior around GPUStoreOp.discard? gpuweb/gpuweb#4398 (2 comments by @kainino0x, @mwyrzykowski)
-
Clarification regarding the validation of GPUQueue.submit and submitting duplicate command buffers? gpuweb/gpuweb#4367 (2 comments by @kainino0x, @Kangz)
-
HTMLCanvasElement copyExternalImageToTexture under specifed? gpuweb/gpuweb#4255 (2 comments by @greggman, @jimblandy)
-
WGSL should forbid pointers to runtime-sized values as user-defined function parameters gpuweb/gpuweb#2268 (2 comments by @ben-clayton, @theb1rd)
-
Early Depth/Stencil test directive or flag in pipeline state creation gpuweb/gpuweb#398 (1 comment by @kdashg)
-
add a system-defined wsgl namespace for predeclared objects gpuweb/gpuweb#4308 (1 comment by @kdashg)
-
Missing intrinsic functions like rcp and rsqrt that are useful in other shading languages gpuweb/gpuweb#4092 (1 comment by @kdashg)
-
wgsl: clarify what happens for IEEE 754 "invalid operation" cases, e.g. acos(2) gpuweb/gpuweb#4219 (1 comment by @kdashg)
-
Constructible types description is unclear around abstract types gpuweb/gpuweb#4385 (1 comment by @dneto0)
-
Feature request: support one-component vertex attribute format gpuweb/gpuweb#4340 (1 comment by @Kangz)
-
Clarify (lack of) guarantees provided by timestamp queries (was: Define the clock and time origin of timestamp queries) gpuweb/gpuweb#4361 (1 comment by @kainino0x)
-
-
18 pull requests opened.
-
Add Compat Limits to Proposal gpuweb/gpuweb#4397, by @greggman.
-
[editorial] Apply component-wise to matrices gpuweb/gpuweb#4391, by @alan-baker.
-
WGSL: Using a single swizzle letter produces a scalar. gpuweb/gpuweb#4389, by @jimblandy.
-
Do not always require GPUProgrammableStage.entryPoint gpuweb/gpuweb#4387, by @beaufortfrancois.
-
Clarify matching of pipeline constant identifiers. gpuweb/gpuweb#4382, by @jimblandy.
-
Actually explain what contagious invalidity is gpuweb/gpuweb#4381, by @kainino0x.
-
Clarify comparison samplers can only be used on depth textures gpuweb/gpuweb#4380, by @kainino0x.
-
Add more details to timing attack section gpuweb/gpuweb#4379, by @kainino0x.
-
Add sample_mask consensus. gpuweb/gpuweb#4375, by @SenorBlanco.
-
Add Jim Blandy as an editor of the WebGPU API specification. gpuweb/gpuweb#4371, by @jimblandy.
-
Remove GPUCommandEncoder.writeTimestamp gpuweb/gpuweb#4370, by @beaufortfrancois.
-
Add subgroups proposal gpuweb/gpuweb#4368, by @alan-baker.
-
Specify copy rules for different canvas contexts gpuweb/gpuweb#4366, by @greggman.
-
Add the mip level and array layer consensus proposals. gpuweb/gpuweb#4365, by @SenorBlanco.
-
Add note about undefined behaviors when copying from WebGL canvas gpuweb/gpuweb#4364, by @shaoboyan.
-
editorial: clip position definition links to clip space coords gpuweb/gpuweb#4360, by @dneto0.
-
Coarsen timestamp value gpuweb/gpuweb#4359, by @beaufortfrancois.
-
Disallow requesting alignment limits that are ≥ 2**32 gpuweb/gpuweb#4358, by @kainino0x.
-
-
17 pull requests merged.
-
Add syntax sugar for dereferencing composites gpuweb/gpuweb#4311.
-
[editorial] Apply component-wise to matrices gpuweb/gpuweb#4391.
-
WGSL: Using a single swizzle letter produces a scalar. gpuweb/gpuweb#4389.
-
Do not always require GPUProgrammableStage.entryPoint gpuweb/gpuweb#4387.
-
Add sample_mask consensus. gpuweb/gpuweb#4375.
-
Remove restrictions on user-defined pointer parameters gpuweb/gpuweb#4323.
-
WGSL: Add dot4U8Packed and dot4I8Packed as new built-in functions gpuweb/gpuweb#2738.
-
Actually explain what contagious invalidity is gpuweb/gpuweb#4381.
-
Clarify comparison samplers can only be used on depth textures gpuweb/gpuweb#4380.
-
Add more details to timing attack section gpuweb/gpuweb#4379.
-
Add Jim Blandy as an editor of the WebGPU API specification. gpuweb/gpuweb#4371.
-
Add the mip level and array layer consensus proposals. gpuweb/gpuweb#4365.
-
Remove GPUCommandEncoder.writeTimestamp gpuweb/gpuweb#4370.
-
Disallow requesting alignment limits that are ≥ 2**32 gpuweb/gpuweb#4358.
-
Add note about undefined behaviors when copying from WebGL canvas gpuweb/gpuweb#4364.
-
Coarsen timestamp value gpuweb/gpuweb#4359.
-
editorial: clip position definition links to clip space coords gpuweb/gpuweb#4360.
-
-
1 pull request closed.
-
Make GPUPipelineError constructor options optional gpuweb/gpuweb#4292.
-
-
17 pull requests received 105 new comments.
-
Add subgroups proposal gpuweb/gpuweb#4368 (28 comments by @alan-baker, @ben-clayton, @dneto0, @exrook, @greggman, @Kangz, @kdashg, @Lichtso)
-
Add note about undefined behaviors when copying from WebGL canvas gpuweb/gpuweb#4364 (12 comments by @greggman, @kainino0x, @shaoboyan)
-
Add the mip level and array layer consensus proposals. gpuweb/gpuweb#4365 (11 comments by @greggman, @Kangz, @SenorBlanco)
-
WGSL: Add dot4U8Packed and dot4I8Packed as new built-in functions gpuweb/gpuweb#2738 (6 comments by @alan-baker)
-
Allow separate read-only state for depth and stencil aspects gpuweb/gpuweb#4331 (4 comments by @Kangz, @teoxoy)
-
Add syntax sugar for dereferencing composites gpuweb/gpuweb#4311 (3 comments by @alan-baker, @dneto0, @jimblandy)
-
Remove restrictions on user-defined pointer parameters gpuweb/gpuweb#4323 (3 comments by @alan-baker, @dneto0)
-
Specify copy rules for different canvas contexts gpuweb/gpuweb#4366 (3 comments by @greggman, @kainino0x)
-
Disallow requesting alignment limits that are ≥ 2**32 gpuweb/gpuweb#4358 (2 comments by @jimblandy, @mwyrzykowski)
-
Add Jim Blandy as an editor of the WebGPU API specification. gpuweb/gpuweb#4371 (1 comment by @kainino0x)
-
editorial: clip position definition links to clip space coords gpuweb/gpuweb#4360 (1 comment by @kainino0x)
-
TR publications
-
17 documents published.
-
WebGPU published 11 times as a Working Draft. History of WebGPU
-
WebGPU Shading Language published 6 times as a Working Draft. History of WebGPU Shading Language
-
Past meetings
-
1 meeting.
GitHub
gpuweb/cts
-
24 issues opened.
-
Add labels to all resources gpuweb/cts#3116, by @greggman.
-
Add discussion of AbstractFloat behaviour to `fp_primer.md` gpuweb/cts#3111, by @zoddicus.
-
val: Test TypeErrors from GPUColor/Origin/Extent sequences gpuweb/cts#3110, by @kainino0x.
-
val: Test SecurityError from importExternalTexture gpuweb/cts#3109, by @kainino0x.
-
Add tests for Adapter Capability Guarantees gpuweb/cts#3106, by @beaufortfrancois.
-
Test that DOMExceptions from WebGPU always have stacks gpuweb/cts#3103, by @kainino0x.
-
Add a case for `atomicSub(ptr, -2147483648i)` (int-min) gpuweb/cts#3100, by @ben-clayton.
-
wgsl: Implement 'heavy' and 'lite' modes for running numeric tests gpuweb/cts#3093, by @zoddicus.
-
wgsl: atan2 execution tests should filter test cases if doing const-eval gpuweb/cts#3088, by @zoddicus.
-
wgsl: Add AbstractFloat being tested to be as good as f32 to fp_primer.md gpuweb/cts#3079, by @zoddicus.
-
support read-only, read-write storage textures gpuweb/cts#3078, by @dneto0.
-
wgsl: Add AbstractFloat `select` execution tests gpuweb/cts#3076, by @zoddicus.
-
Create a kSparseMinus3PiTo3Pi and use it for calculating cartesian products gpuweb/cts#3071, by @amaiorano.
-
wgsl: test `override` inputs for various execution test gpuweb/cts#3070, by @zoddicus.
-
Test that fine derivatives are actually fine gpuweb/cts#3065, by @kainino0x.
-
framework: add ability filter device based on WGSLLanguageFeatures gpuweb/cts#3057, by @dneto0.
-
wgsl: test unrestricted_pointer_parameters feature gpuweb/cts#3056, by @dneto0.
-
wgsl: test f16 in memory model barrier tests gpuweb/cts#3053, by @dneto0.
-
wgsl: Statically allocate work data for correctly rounding gpuweb/cts#3051, by @zoddicus.
-
memory_model tests use 6 storage buffers but compat has a default limit of 4 gpuweb/cts#3049, by @greggman.
-
wgsl: scalar-matrix multiplication not correctly rounding operations gpuweb/cts#3044, by @zoddicus.
-
wgsl: Speed up AbstractFloat (especially Matrix) tests gpuweb/cts#3039, by @zoddicus.
-
-
11 issues closed.
-
Test that DOMExceptions from WebGPU always have stacks gpuweb/cts#3103.
-
Add tests for Adapter Capability Guarantees gpuweb/cts#3106.
-
wgsl: AbstractFloat execution tests for `sign` builtin gpuweb/cts#2582.
-
wgsl: Add AbstractFloat `select` execution tests gpuweb/cts#3076.
-
wgsl: scalar-matrix multiplication not correctly rounding operations gpuweb/cts#3044.
-
wgsl: test f16 in memory model barrier tests gpuweb/cts#3053.
-
wgsl: Statically allocate work data for correctly rounding gpuweb/cts#3051.
-
wgsl: test memory model for f16 types gpuweb/cts#2779.
-
-
13 issues received 30 new comments.
-
memory_model tests use 6 storage buffers but compat has a default limit of 4 gpuweb/cts#3049 (7 comments by @dneto0, @greggman, @reeselevine)
-
Investigate handling of very small (near zero) AbstractFloats and ECMAScript `numbers` gpuweb/cts#2993 (7 comments by @jiangzhaoming, @zoddicus)
-
Intel + DXC failure: webgpu:shader,execution,flow_control,loop:nested_loops:preventValueOptimizations=false gpuweb/cts#3063 (4 comments by @amaiorano, @jiangzhaoming)
-
val: Test TypeErrors from GPUColor/Origin/Extent sequences gpuweb/cts#3110 (2 comments by @kainino0x)
-
Unable to specify timing metadata for some tests due to 'run:generate-wpt-cts-html-chunked2sec' failing gpuweb/cts#2944 (1 comment by @kainino0x)
-
Should `websocket-logger` instructions use `npm ci` instead of `npm i`? gpuweb/cts#2945 (1 comment by @kainino0x)
-
Test with >= 500 subcases fails with "internal error: case is "skip" but has nonskipped subcases" gpuweb/cts#3042 (1 comment by @amaiorano)
-
-
60 pull requests opened.
-
Reduce draw validation test parameterization gpuweb/cts#3122, by @kainino0x.
-
Fix cache files being padded with trailing 0's gpuweb/cts#3121, by @ben-clayton.
-
wgsl: Convert `quantizeToI32/U32` to use `Math.trunc` gpuweb/cts#3120, by @zoddicus.
-
wgsl: Convert `quantizeToF32` to used `Math.fround` gpuweb/cts#3119, by @zoddicus.
-
wgsl: Convert `quantizeToF16` to used `hfround` gpuweb/cts#3118, by @zoddicus.
-
Compat: Refactor memory module tests to use 4 storage buffers gpuweb/cts#3117, by @greggman.
-
dev_server: Serve on localhost by default gpuweb/cts#3115, by @kainino0x.
-
Audit TypedArray allocations gpuweb/cts#3114, by @kainino0x.
-
binary_stream: Simplify away a no-op TypedArray gpuweb/cts#3113, by @kainino0x.
-
Add details about 'atleast as good as f32' for AbstractFloats gpuweb/cts#3112, by @zoddicus.
-
BinaryStream: Use little endian for f64 gpuweb/cts#3108, by @ben-clayton.
-
Add missing tests for Adapter Capability Guarantees gpuweb/cts#3107, by @beaufortfrancois.
-
Test that DOMExceptions from WebGPU always have stacks gpuweb/cts#3105, by @kainino0x.
-
Use largest maxInterStageShaderVariables in maxInterStageShaderCompon… gpuweb/cts#3104, by @Jiawei-Shao.
-
More cache deserialization micro-optimisations gpuweb/cts#3102, by @ben-clayton.
-
wgsl: Speed up `ldexp` execution tests gpuweb/cts#3101, by @zoddicus.
-
tsdoc: Remove @p with backticks gpuweb/cts#3099, by @ben-clayton.
-
Compat: Test textureBindingViewDimension validation gpuweb/cts#3098, by @greggman.
-
Make test params readonly so they can't be accidentally permanently modified gpuweb/cts#3097, by @kainino0x.
-
Fixes flaky test because parameters were changed when the test was ran. gpuweb/cts#3096, by @lokokung.
-
Remove duplicate definitions of reinterpret* gpuweb/cts#3095, by @austinEng.
-
Replace JSON case cache serialization with binary files gpuweb/cts#3094, by @ben-clayton.
-
Slim down on typed array allocations in conversion.ts gpuweb/cts#3092, by @austinEng.
-
Add back generate-cache grunt command gpuweb/cts#3091, by @amaiorano.
-
wgsl: Revert changes to `round` execution tests gpuweb/cts#3090, by @zoddicus.
-
wgsl: Filter `atan2` tests based on if const-eval or not gpuweb/cts#3089, by @zoddicus.
-
wgsl: Add AbstractFloat `mix` execution tests gpuweb/cts#3087, by @zoddicus.
-
Fix a minor issue introduced in previous stencil test fix gpuweb/cts#3086, by @toji.
-
Remove pipeline statistics query tests gpuweb/cts#3085, by @beaufortfrancois.
-
wgsl: Add AbstractFloat `floor` execution tests gpuweb/cts#3084, by @zoddicus.
-
Fix a bad slice operation in image_copy stencil tests gpuweb/cts#3083, by @toji.
-
wgsl: Cleanup cruft related to quantizeToF16 gpuweb/cts#3082, by @zoddicus.
-
wgsl: Add AbstractFloat `sign` execution tests gpuweb/cts#3081, by @zoddicus.
-
wgsl: Add AbstractFloat `round` execution tests gpuweb/cts#3080, by @zoddicus.
-
wgsl: Add AF `select` execution tests gpuweb/cts#3077, by @zoddicus.
-
wgsl: Add AF remainder (%) execution tests gpuweb/cts#3075, by @zoddicus.
-
wgsl: Add AF Division execution tests gpuweb/cts#3074, by @zoddicus.
-
Add compat moode support to cmdline.ts and server.ts gpuweb/cts#3073, by @greggman.
-
wgsl: Use sparse version of kMinus3PiTo3Pi for atan2 validation tests gpuweb/cts#3072, by @zoddicus.
-
depthCompare is not required for depth attachments if not used gpuweb/cts#3069, by @beaufortfrancois.
-
wgsl: make expectation id strats from 1 in flow_control exec tests gpuweb/cts#3068, by @jiangzhaoming.
-
wgsl: show output values in flow_control exec tests error messages gpuweb/cts#3067, by @jiangzhaoming.
-
Require depthWriteEnabled and depthCompare only for formats with depth gpuweb/cts#3066, by @beaufortfrancois.
-
Move listing_meta docs to docs/, clarify a few things gpuweb/cts#3064, by @kainino0x.
-
[Testing branch] Distinguish +/-0.0 in test cases query gpuweb/cts#3062, by @jiangzhaoming.
-
Add negative zero in float range0 gpuweb/cts#3061, by @jiangzhaoming.
-
[Reland] util: Add negative zero in full F16/F32/F64 range gpuweb/cts#3060, by @jiangzhaoming.
-
wgsl: Implement validation test for parsing f16 literal gpuweb/cts#3059, by @jiangzhaoming.
-
framework: allow skipping tests based on wgslLanguageFeatures gpuweb/cts#3058, by @dneto0.
-
memory_model: Add f16 to barrier tests gpuweb/cts#3055, by @dneto0.
-
Make Pass > Skip gpuweb/cts#3054, by @greggman.
-
wgsl: Optimize memory allocations when quantizing data gpuweb/cts#3052, by @zoddicus.
-
gen_cache: Add --nth flag gpuweb/cts#3050, by @ben-clayton.
-
Compat: make zero_init tests handle different limits gpuweb/cts#3048, by @greggman.
-
Compat: Fix attachment_compatibility color_count tests for limits gpuweb/cts#3047, by @greggman.
-
wgsl: Correctly invoke multiplicationInterval instead impl directly gpuweb/cts#3046, by @zoddicus.
-
Compat: skip unsupported formats in filter mode tests gpuweb/cts#3045, by @greggman.
-
Revert "Fix Skip/Pass Issues" gpuweb/cts#3043, by @amaiorano.
-
Fix atan and tan validation tests gpuweb/cts#3041, by @amaiorano.
-
Compat: Skip unsupported texture operations gpuweb/cts#3040, by @greggman.
-
-
54 pull requests merged.
-
Fix cache files being padded with trailing 0's gpuweb/cts#3121.
-
wgsl: Convert `quantizeToF32` to used `Math.fround` gpuweb/cts#3119.
-
wgsl: Convert `quantizeToF16` to used `hfround` gpuweb/cts#3118.
-
dev_server: Serve on localhost by default gpuweb/cts#3115.
-
Test that DOMExceptions from WebGPU always have stacks gpuweb/cts#3105.
-
Add missing tests for Adapter Capability Guarantees gpuweb/cts#3107.
-
BinaryStream: Use little endian for f64 gpuweb/cts#3108.
-
wgsl: Speed up `ldexp` execution tests gpuweb/cts#3101.
-
More cache deserialization micro-optimisations gpuweb/cts#3102.
-
tsdoc: Remove @p with backticks gpuweb/cts#3099.
-
Replace JSON case cache serialization with binary files gpuweb/cts#3094.
-
Remove duplicate definitions of reinterpret* gpuweb/cts#3095.
-
Slim down on typed array allocations in conversion.ts gpuweb/cts#3092.
-
Add back generate-cache grunt command gpuweb/cts#3091.
-
wgsl: Cleanup cruft related to quantizeToF16 gpuweb/cts#3082.
-
wgsl: Filter `atan2` tests based on if const-eval or not gpuweb/cts#3089.
-
wgsl: Revert changes to `round` execution tests gpuweb/cts#3090.
-
Remove pipeline statistics query tests gpuweb/cts#3085.
-
Fix a minor issue introduced in previous stencil test fix gpuweb/cts#3086.
-
depthCompare is not required for depth attachments if not used gpuweb/cts#3069.
-
wgsl: Add AbstractFloat `floor` execution tests gpuweb/cts#3084.
-
Fix a bad slice operation in image_copy stencil tests gpuweb/cts#3083.
-
wgsl: Add AbstractFloat `sign` execution tests gpuweb/cts#3081.
-
wgsl: Add AF `select` execution tests gpuweb/cts#3077.
-
wgsl: Add AF remainder (%) execution tests gpuweb/cts#3075.
-
wgsl: Add AF Division execution tests gpuweb/cts#3074.
-
wgsl: Implement validation test for parsing f16 literal gpuweb/cts#3059.
-
Add compat moode support to cmdline.ts and server.ts gpuweb/cts#3073.
-
wgsl: make expectation id strats from 1 in flow_control exec tests gpuweb/cts#3068.
-
Move listing_meta docs to docs/, clarify a few things gpuweb/cts#3064.
-
wgsl: Correctly invoke multiplicationInterval instead impl directly gpuweb/cts#3046.
-
wgsl: show output values in flow_control exec tests error messages gpuweb/cts#3067.
-
memory_model: Add f16 to barrier tests gpuweb/cts#3055.
-
[Reland] util: Add negative zero in full F16/F32/F64 range gpuweb/cts#3060.
-
[Testing branch] Distinguish +/-0.0 in test cases query gpuweb/cts#3062.
-
Add negative zero in float range0 gpuweb/cts#3061.
-
Make Pass > Skip gpuweb/cts#3054.
-
Compat: skip unsupported formats in filter mode tests gpuweb/cts#3045.
-
Compat: Fix attachment_compatibility color_count tests for limits gpuweb/cts#3047.
-
Compat: make zero_init tests handle different limits gpuweb/cts#3048.
-
wgsl: Optimize memory allocations when quantizing data gpuweb/cts#3052.
-
gen_cache: Add --nth flag gpuweb/cts#3050.
-
Revert "Fix Skip/Pass Issues" gpuweb/cts#3043.
-
Fix atan and tan validation tests gpuweb/cts#3041.
-
Compat: Skip unsupported texture operations gpuweb/cts#3040.
-
shader/execution/memory_model: test f16 adjacent writes gpuweb/cts#3038.
-
unittest: Fix/add f16 cases for data cache serialization unittest gpuweb/cts#3033.
-
wgsl: Add f16 validation tests for bistcast from bad const gpuweb/cts#3034.
-
Add unorm10-10-10-2 vertex format gpuweb/cts#2941.
-
-
1 pull request closed.
-
binary_stream: Simplify away a no-op TypedArray gpuweb/cts#3113.
-
-
35 pull requests received 130 new comments.
-
Require depthWriteEnabled and depthCompare only for formats with depth gpuweb/cts#3066 (23 comments by @beaufortfrancois, @kainino0x, @Kangz)
-
depthCompare is not required for depth attachments if not used gpuweb/cts#3069 (11 comments by @beaufortfrancois, @kainino0x, @Kangz)
-
Replace JSON case cache serialization with binary files gpuweb/cts#3094 (9 comments by @austinEng, @ben-clayton, @kainino0x)
-
wgsl: Add AF Division execution tests gpuweb/cts#3074 (8 comments by @dneto0, @jiangzhaoming, @zoddicus)
-
Test that DOMExceptions from WebGPU always have stacks gpuweb/cts#3105 (5 comments by @kainino0x, @lokokung)
-
Slim down on typed array allocations in conversion.ts gpuweb/cts#3092 (5 comments by @austinEng, @ben-clayton)
-
shader/execution/memory_model: test f16 adjacent writes gpuweb/cts#3038 (5 comments by @ben-clayton)
-
Add missing tests for Adapter Capability Guarantees gpuweb/cts#3107 (4 comments by @beaufortfrancois, @greggman)
-
wgsl: Add AbstractFloat `mix` execution tests gpuweb/cts#3087 (4 comments by @jiangzhaoming, @zoddicus)
-
Add details about 'atleast as good as f32' for AbstractFloats gpuweb/cts#3112 (3 comments by @kainino0x)
-
Make test params readonly so they can't be accidentally permanently modified gpuweb/cts#3097 (3 comments by @kainino0x, @lokokung)
-
wgsl: Correctly invoke multiplicationInterval instead impl directly gpuweb/cts#3046 (3 comments by @jiangzhaoming, @zoddicus)
-
[Reland] util: Add negative zero in full F16/F32/F64 range gpuweb/cts#3060 (3 comments by @dneto0, @jiangzhaoming, @zoddicus)
-
wgsl: show output values in flow_control exec tests error messages gpuweb/cts#3067 (3 comments by @ben-clayton, @jiangzhaoming)
-
binary_stream: Simplify away a no-op TypedArray gpuweb/cts#3113 (2 comments by @ben-clayton, @kainino0x)
-
wgsl: Filter `atan2` tests based on if const-eval or not gpuweb/cts#3089 (2 comments by @jiangzhaoming, @zoddicus)
-
wgsl: Implement validation test for parsing f16 literal gpuweb/cts#3059 (1 comment by @jiangzhaoming)
-
wgsl: make expectation id strats from 1 in flow_control exec tests gpuweb/cts#3068 (1 comment by @jiangzhaoming)
-
framework: allow skipping tests based on wgslLanguageFeatures gpuweb/cts#3058 (1 comment by @dneto0)
-
unittest: Fix/add f16 cases for data cache serialization unittest gpuweb/cts#3033 (1 comment by @jiangzhaoming)
-
wgsl: Add f16 validation tests for bistcast from bad const gpuweb/cts#3034 (1 comment by @jiangzhaoming)
-
gpuweb/gpuweb
-
20 issues opened.
-
[Feature Request] Vertex Shader storage read/write gpuweb/gpuweb#4355, by @rconde01.
-
reuse of an GPUAdapter to request a GPUDevice does not reject. gpuweb/gpuweb#4352, by @greggman.
-
Limitation on the precision of hex floats? gpuweb/gpuweb#4350, by @mwyrzykowski.
-
Investigation: multi-draw-indirect gpuweb/gpuweb#4349, by @kainino0x.
-
debugPrintfEXT in wgsl gpuweb/gpuweb#4348, by @PixelDust22.
-
Default entryPoints to shader modules gpuweb/gpuweb#4342, by @greggman.
-
Feature request: support one-component vertex attribute format gpuweb/gpuweb#4340, by @bbbbx.
-
Shader module compilation hint API is messy gpuweb/gpuweb#4338, by @kainino0x.
-
Add extensions for member accessors and functions (maybe with traits) gpuweb/gpuweb#4337, by @mehmetoguzderin.
-
viewport & scissor's behavior is differ from WebGL gpuweb/gpuweb#4334, by @deepkolos.
-
named swizzles gpuweb/gpuweb#4333, by @greggman.
-
Storage texture atomics gpuweb/gpuweb#4329, by @fintelia.
-
Are "Fine" derivatives guaranteed to actually be fine? gpuweb/gpuweb#4325, by @kainino0x.
-
WebGPU Compatibility mode (consensus proposal) gpuweb/gpuweb#4320, by @SenorBlanco.
-
LSP gpuweb/gpuweb#4313, by @Peyton-Spencer.
-
WGSL spec build issues on Mac gpuweb/gpuweb#4310, by @alan-baker.
-
-
17 issues closed.
-
Separate read-only state for stencil and depth aspects. gpuweb/gpuweb#3036.
-
Enable WebGPU in Safari TP gpuweb/gpuweb#4238.
-
depthWriteEnabled and depthCompare are required even for stencil-only formats gpuweb/gpuweb#3905.
-
Shader module compilation hint API is messy gpuweb/gpuweb#4338.
-
viewport & scissor's behavior is differ from WebGL gpuweb/gpuweb#4334.
-
Only one pipeline layout for each entry point can be given as compilation hint gpuweb/gpuweb#4233.
-
WebGPU Compatibility mode (consensus proposal) gpuweb/gpuweb#4320.
-
Support fuller-featured pointers (as per #3276) gpuweb/gpuweb#3546.
-
LSP gpuweb/gpuweb#4313.
-
WGSL spec build issues on Mac gpuweb/gpuweb#4310.
-
Build: tree sitter still flaking sometimes gpuweb/gpuweb#4274.
-
rgb10a2 vertex format gpuweb/gpuweb#4275.
-
-
48 issues received 159 new comments.
-
Proposal: read-write storage textures gpuweb/gpuweb#3838 (16 comments by @alan-baker, @austinEng, @dneto0, @fintelia, @kainino0x, @Kangz, @mwyrzykowski, @teoxoy)
-
Default entryPoints to shader modules gpuweb/gpuweb#4342 (12 comments by @ben-clayton, @dneto0, @floooh, @kainino0x, @Kangz, @mwyrzykowski, @teoxoy)
-
Separate read-only state for stencil and depth aspects. gpuweb/gpuweb#3036 (12 comments by @BearIvan, @kainino0x, @Kangz, @mwyrzykowski, @teoxoy)
-
[wgsl] Consider a preprocessor gpuweb/gpuweb#568 (11 comments by @greggman, @magcius, @parasyte, @theb1rd)
-
Enable WebGPU in Safari TP gpuweb/gpuweb#4238 (10 comments by @beaufortfrancois, @munrocket, @mwyrzykowski, @R00tkitSMM)
-
Dual Source Blending Investigation + Feature Proposal gpuweb/gpuweb#4283 (10 comments by @bjjones, @jimblandy, @kainino0x, @Kangz, @teoxoy)
-
Viewport rect must lie entirely within the current attachment size? gpuweb/gpuweb#373 (6 comments by @floooh, @kainino0x, @magcius)
-
add subgroups, and make them portable if possible gpuweb/gpuweb#4306 (5 comments by @alan-baker, @JMS55, @Lichtso, @magcius, @raphlinus)
-
Reference High Resolution Time spec gpuweb/gpuweb#4175 (4 comments by @jyasskin, @kainino0x, @Kangz)
-
depthWriteEnabled and depthCompare are required even for stencil-only formats gpuweb/gpuweb#3905 (4 comments by @beaufortfrancois, @kainino0x)
-
Shader module compilation hint API is messy gpuweb/gpuweb#4338 (3 comments by @beaufortfrancois, @kainino0x, @mwyrzykowski)
-
Are "Fine" derivatives guaranteed to actually be fine? gpuweb/gpuweb#4325 (3 comments by @alan-baker, @kainino0x)
-
16-bit normalized texture support gpuweb/gpuweb#3001 (3 comments by @kainino0x, @magcius, @mwyrzykowski)
-
writeBuffer/writeTexture works with shared ArrayBufferViews but not SharedArrayBuffer gpuweb/gpuweb#4186 (3 comments by @kainino0x, @Kangz, @mwyrzykowski)
-
Should we add a parameter for the number of draws to DrawIndirect so that all commands in the IndirectBuffer can be drawn at once? gpuweb/gpuweb#4344 (2 comments by @Kangz)
-
"drawIndirect" Only the first call will draw correctly, subsequent calls will have no effect. gpuweb/gpuweb#4341 (2 comments by @GoodLuckIce, @greggman)
-
Remove the need to hard-code texel format for `texture_storage_2d` in shaders gpuweb/gpuweb#4298 (2 comments by @kainino0x)
-
Double precision floats (IEEE-754 binary64) gpuweb/gpuweb#2805 (2 comments by @clickingbuttons, @fintelia)
-
WebGPU Compatibility mode (consensus proposal) gpuweb/gpuweb#4320 (2 comments by @Kangz, @SenorBlanco)
-
HTMLCanvasElement copyExternalImageToTexture under specifed? gpuweb/gpuweb#4255 (1 comment by @shaoboyan)
-
The format of "64-bit unsigned integers" is never specified gpuweb/gpuweb#4289 (1 comment by @kainino0x)
-
What happens on the device timeline after `GPUDevice.destroy` is called? gpuweb/gpuweb#4294 (1 comment by @kainino0x)
-
`AbstractInt` conversion ranking forces unfortunate compromise gpuweb/gpuweb#4291 (1 comment by @alan-baker)
-
Support for gl_PrimitiveID / SV_PrimitiveID -like construct? gpuweb/gpuweb#1786 (1 comment by @kohakukun)
-
Consider raising `maxStorageBuffersPerShaderStage` gpuweb/gpuweb#4235 (1 comment by @brendan-duncan)
-
-
22 pull requests opened.
-
Remove unnecessary anchrors for access mode gpuweb/gpuweb#4354, by @alan-baker.
-
WGSL: Fix explanation of hyperbolic angles. gpuweb/gpuweb#4351, by @jimblandy.
-
Fix nit in validating GPUDepthStencilState gpuweb/gpuweb#4345, by @beaufortfrancois.
-
Refactor `hints` record into `compilationHints` sequence gpuweb/gpuweb#4343, by @kainino0x.
-
Compatibility Mode: further items with tentative consensus. gpuweb/gpuweb#4339, by @SenorBlanco.
-
depthCompare is not required for depth attachments if not used gpuweb/gpuweb#4336, by @beaufortfrancois.
-
Allow separate read-only state for depth and stencil aspects gpuweb/gpuweb#4331, by @kainino0x.
-
editorial: Improve "Note" headings gpuweb/gpuweb#4330, by @kainino0x.
-
Allow multiple compilation hints per entry point gpuweb/gpuweb#4328, by @kainino0x.
-
editorial: fix typo in writeBuffer gpuweb/gpuweb#4327, by @kainino0x.
-
Add RO and RW storage textures gpuweb/gpuweb#4326, by @alan-baker.
-
Add bitcast from abstractint to u32 gpuweb/gpuweb#4324, by @alan-baker.
-
Remove restrictions on user-defined pointer parameters gpuweb/gpuweb#4323, by @alan-baker.
-
Revert "Remove restrictions on user-defined pointer parameters (#4312)" gpuweb/gpuweb#4322, by @alan-baker.
-
Add compatibility-mode.md proposal. gpuweb/gpuweb#4321, by @SenorBlanco.
-
Add JS example requesting device with higher limit gpuweb/gpuweb#4319, by @beaufortfrancois.
-
Require depthWriteEnabled and depthCompare only for formats with depth gpuweb/gpuweb#4318, by @beaufortfrancois.
-
[editorial] Fix typo in WGSLLanguageFeatures definition gpuweb/gpuweb#4317, by @jiangzhaoming.
-
Remove unmaskHints in GPUAdapter.requestAdapterInfo() gpuweb/gpuweb#4316, by @beaufortfrancois.
-
Simplify makefile for WGSL for cross-platform runs gpuweb/gpuweb#4314, by @mehmetoguzderin.
-
Remove restrictions on user-defined pointer parameters gpuweb/gpuweb#4312, by @alan-baker.
-
Add syntax sugar for dereferencing composites gpuweb/gpuweb#4311, by @alan-baker.
-
-
22 pull requests merged.
-
Allow separate read-only state for depth and stencil aspects gpuweb/gpuweb#4331.
-
Remove unnecessary anchrors for access mode gpuweb/gpuweb#4354.
-
depthCompare is not required for depth attachments if not used gpuweb/gpuweb#4336.
-
Compatibility Mode: further items with tentative consensus. gpuweb/gpuweb#4339.
-
Add RO and RW storage textures gpuweb/gpuweb#4326.
-
Fix nit in validating GPUDepthStencilState gpuweb/gpuweb#4345.
-
Refactor `hints` record into `compilationHints` sequence gpuweb/gpuweb#4343.
-
Add compatibility-mode.md proposal. gpuweb/gpuweb#4321.
-
Allow multiple compilation hints per entry point gpuweb/gpuweb#4328.
-
editorial: Improve "Note" headings gpuweb/gpuweb#4330.
-
editorial: fix typo in writeBuffer gpuweb/gpuweb#4327.
-
Add bitcast from abstractint to u32 gpuweb/gpuweb#4324.
-
Add depthSlice to GPURenderPassColorAttachment for 3D texture gpuweb/gpuweb#4254.
-
[editorial] Fix typo in WGSLLanguageFeatures definition gpuweb/gpuweb#4317.
-
Revert "Remove restrictions on user-defined pointer parameters (#4312)" gpuweb/gpuweb#4322.
-
Require depthWriteEnabled and depthCompare only for formats with depth gpuweb/gpuweb#4318.
-
Fix a typo gpuweb/gpuweb#4285.
-
Add JS example requesting device with higher limit gpuweb/gpuweb#4319.
-
Remove unmaskHints in GPUAdapter.requestAdapterInfo() gpuweb/gpuweb#4316.
-
Remove restrictions on user-defined pointer parameters gpuweb/gpuweb#4312.
-
Simplify makefile for WGSL for cross-platform runs gpuweb/gpuweb#4314.
-
Add unorm10-10-10-2 vertex format gpuweb/gpuweb#4288.
-
-
2 pull requests closed.
-
Make workgroup_count consistent between specs gpuweb/gpuweb#4302.
-
Add WebGPU Adapter Info Registry document gpuweb/gpuweb#2990.
-
-
18 pull requests received 119 new comments.
-
Add RO and RW storage textures gpuweb/gpuweb#4326 (40 comments by @alan-baker, @dneto0, @kainino0x, @Kangz, @teoxoy, @toji)
-
Add compatibility-mode.md proposal. gpuweb/gpuweb#4321 (27 comments by @greggman, @kainino0x, @Kangz, @SenorBlanco, @teoxoy)
-
Remove restrictions on user-defined pointer parameters gpuweb/gpuweb#4312 (6 comments by @alan-baker, @dneto0)
-
Require depthWriteEnabled and depthCompare only for formats with depth gpuweb/gpuweb#4318 (4 comments by @kainino0x, @toji)
-
Add depthSlice to GPURenderPassColorAttachment for 3D texture gpuweb/gpuweb#4254 (4 comments by @haoxli, @kainino0x)
-
depthCompare is not required for depth attachments if not used gpuweb/gpuweb#4336 (3 comments by @beaufortfrancois, @kainino0x)
-
[editorial] Fix typo in WGSLLanguageFeatures definition gpuweb/gpuweb#4317 (3 comments by @jiangzhaoming, @kainino0x)
-
WGSL: Add dot4U8Packed and dot4I8Packed as new built-in functions gpuweb/gpuweb#2738 (3 comments by @Jiawei-Shao, @Kangz)
-
Add unorm10-10-10-2 vertex format gpuweb/gpuweb#4288 (3 comments by @beaufortfrancois, @kainino0x, @Kangz)
-
Compatibility Mode: further items with tentative consensus. gpuweb/gpuweb#4339 (2 comments by @kainino0x, @SenorBlanco)
-
Remove unmaskHints in GPUAdapter.requestAdapterInfo() gpuweb/gpuweb#4316 (2 comments by @beaufortfrancois, @kainino0x)
-
Add JS example requesting device with higher limit gpuweb/gpuweb#4319 (2 comments by @beaufortfrancois, @toji)
-
Allow separate read-only state for depth and stencil aspects gpuweb/gpuweb#4331 (1 comment by @Kangz)
-
TR publications
-
4 documents published.
-
WebGPU Shading Language published 3 times as a Working Draft. History of WebGPU Shading Language
-
WebGPU published as a Working Draft. History of WebGPU
-
Participants
-
1 participant joined.
-
Victor Miura (Google LLC) joined.
-
GitHub
gpuweb/cts
-
21 issues opened.
-
Fix CTS roll into Chromium gpuweb/cts#3028, by @amaiorano.
-
Deduplicate Tests gpuweb/cts#3022, by @greggman.
-
Fix skip vs pass priority gpuweb/cts#3021, by @greggman.
-
val: test maxBindGroupsPlusVertexBuffers gpuweb/cts#2983, by @greggman.
-
npm test is taking a very long time to run gpuweb/cts#2980, by @zoddicus.
-
wgs: AbstractFloat execution tests for `clamp` gpuweb/cts#2970, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `min` builtin gpuweb/cts#2969, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `max` builtin gpuweb/cts#2968, by @zoddicus.
-
wgsl: Shorten matrix test case specification using map-reduce gpuweb/cts#2957, by @zoddicus.
-
Should `websocket-logger` instructions use `npm ci` instead of `npm i`? gpuweb/cts#2945, by @zoddicus.
-
Change subcaseMS warning to error? gpuweb/cts#2939, by @greggman.
-
Add recipe(s) for updating listing_meta.json gpuweb/cts#2938, by @zoddicus.
-
'scalar' for test names unclear for some numeric test files gpuweb/cts#2935, by @zoddicus.
-
requestDeviceGroup: make some limits optinoal gpuweb/cts#2934, by @brygsh.
-
wgsl: AbstractFloat execution test for `abs` builtin gpuweb/cts#2931, by @zoddicus.
-
Report unexpected error scope errors ahead of unexpected results gpuweb/cts#2924, by @austinEng.
-
Workaround for NV Vulkan Bug gpuweb/cts#2921, by @cwfitzgerald.
-
-
54 issues closed.
-
wgsl: f16 built-in execution tests gpuweb/cts#1248.
-
wgsl: AbstractFloat execution tests for `modf` gpuweb/cts#2569.
-
wgsl: f16 execution tests for `modf` gpuweb/cts#2568.
-
wgsl: f16 execution test for `dot` builtin gpuweb/cts#2513.
-
wgsl: f16 execution tests for `faceForward` builtin gpuweb/cts#2547.
-
wgsl: f16 execution tests for `refract` builtin gpuweb/cts#2580.
-
wgsl: f16 execution tests for `reflect` builtin gpuweb/cts#2579.
-
wgsl: f16 execution test for `determinant` builtin gpuweb/cts#2522.
-
wgsl : f16 execution test for `tanh` builtin. gpuweb/cts#2527.
-
wgsl: f16 execution test for `sinh` builtin gpuweb/cts#2535.
-
wgsl: f16 execution tests for `cosh` gpuweb/cts#2538.
-
wgsl: f16 execution tests for `acosh` builtin gpuweb/cts#2559.
-
wgsl: f16 execution tests for `asinh` builtin gpuweb/cts#2561.
-
wgsl: f16 execution tests for `atanh` builtin gpuweb/cts#2563.
-
wgsl: AbstractFloat execution tests for `fma` builtin gpuweb/cts#2548.
-
wgsl: f16 execution tests for `mix` builtin gpuweb/cts#2567.
-
wgsl: f16 execution tests for `pow` builtin gpuweb/cts#2573.
-
Add tests for `mix` with non-matching parameters gpuweb/cts#2039.
-
wgsl: f16 execution tests for `fma` builtin gpuweb/cts#2549.
-
wgsl: f16 execution test for `smoothstep` builtin gpuweb/cts#2533.
-
wgsl: f16 execution test for `tan` builtin gpuweb/cts#2507.
-
wgsl: f16 execution tests for `normalize` builtin gpuweb/cts#2571.
-
wgsl: Shorten matrix test case specification using map-reduce gpuweb/cts#2957.
-
npm test is taking a very long time to run gpuweb/cts#2980.
-
wgsl: f16 execution tests for `length` builtin gpuweb/cts#2555.
-
wgsl: f16 execution tests for `distance` builtin gpuweb/cts#2557.
-
wgsl: AbstractFloat execution tests for `cross` gpuweb/cts#2541.
-
wgsl: f16 execution tests for `cross` gpuweb/cts#2540.
-
wgsl: test saturate floating point builtin gpuweb/cts#1635.
-
wgs: AbstractFloat execution tests for `clamp` gpuweb/cts#2970.
-
wgsl: AbstractFloat execution tests for `min` builtin gpuweb/cts#2969.
-
wgsl: AbstractFloat execution tests for `max` builtin gpuweb/cts#2968.
-
wgsl: f16 execution tests for `exp` builtin gpuweb/cts#2543.
-
wgsl: f16 execution tests for `exp2` builtin gpuweb/cts#2544.
-
wgsl: AbstractFloat execution tests for `degrees` builtin gpuweb/cts#2564.
-
wgsl: f16 execution tests for `degrees` builtin gpuweb/cts#2565.
-
wgsl: AbstractFloat execution tests for `radians` builtin gpuweb/cts#2576.
-
wgsl: f16 execution tests for `radians` builtin gpuweb/cts#2577.
-
wgsl: f16 execution tests for `remainder (%)` gpuweb/cts#2586.
-
wgsl: AbstractFloat execution test for `transpose` builtin gpuweb/cts#2521.
-
wgsl: f16 execution test for `transpose` builtin gpuweb/cts#2520.
-
wgsl: f16 execution test for `trunc` builtin gpuweb/cts#2524.
-
wgsl: AbstractFloat execution test for `trunc` builtin gpuweb/cts#2525.
-
'scalar' for test names unclear for some numeric test files gpuweb/cts#2935.
-
wgsl: AbstractFloat execution test for `abs` builtin gpuweb/cts#2931.
-
Add recipe(s) for updating listing_meta.json gpuweb/cts#2938.
-
wgsl: f16 execution tests for `frexp` gpuweb/cts#2587.
-
Chunk tests for WPT using actual test time measurements gpuweb/cts#2749.
-
Add core support for f16 to floating point testing framework gpuweb/cts#2499.
-
wgsl: f16 execution test for `step` builtin gpuweb/cts#2529.
-
wgsl: f16 execution tests for `sign` builtin gpuweb/cts#2583.
-
Test cache generation on PRs gpuweb/cts#2913.
-
-
35 issues received 51 new comments.
-
npm test is taking a very long time to run gpuweb/cts#2980 (7 comments by @jiangzhaoming, @zoddicus)
-
Investigate handling of very small (near zero) AbstractFloats and ECMAScript `numbers` gpuweb/cts#2993 (4 comments by @jiangzhaoming, @zoddicus)
-
Report unexpected error scope errors ahead of unexpected results gpuweb/cts#2924 (1 comment by @kainino0x)
-
-
101 pull requests opened.
-
shader/execution/memory_model: test f16 adjacent writes gpuweb/cts#3038, by @dneto0.
-
Remove limit constants gpuweb/cts#3037, by @greggman.
-
Fix Skip/Pass Issues gpuweb/cts#3036, by @greggman.
-
Compat: Make createTexture tests handle different limits gpuweb/cts#3035, by @greggman.
-
wgsl: Add f16 validation tests for bistcast from bad const gpuweb/cts#3034, by @jiangzhaoming.
-
unittest: Fix/add f16 cases for data cache serialization unittest gpuweb/cts#3033, by @jiangzhaoming.
-
Compat: make perStageLimit tests handle different limits gpuweb/cts#3032, by @greggman.
-
Compat: make device_pool not refer to kLimitInfo gpuweb/cts#3031, by @greggman.
-
Compat: make fragment_state test handle different limits gpuweb/cts#3030, by @greggman.
-
Revert "util: Add negative zero in full F16/F32/F64 range (#2978)" gpuweb/cts#3029, by @amaiorano.
-
Compat: make createRenderBundleEncoder test handle different limits gpuweb/cts#3027, by @greggman.
-
Compat: Skip copyTextureToTexture for compressed textures gpuweb/cts#3026, by @greggman.
-
Re-organize numeric constants to be consistent gpuweb/cts#3025, by @zoddicus.
-
Update webgpu/types to add 'rgb10a2uint' gpuweb/cts#3024, by @Kangz.
-
Update subcaseMS of web_platform,copyToTexture,ImageBitmap gpuweb/cts#3023, by @jiangzhaoming.
-
wgsl: Add AbstractFloat `modf` execution tests gpuweb/cts#3020, by @zoddicus.
-
wgsl: Add f16 `modf` execution tests gpuweb/cts#3019, by @zoddicus.
-
wgsl: refactor and add f16 execution test for builtin ldexp gpuweb/cts#3018, by @jiangzhaoming.
-
wgsl: add f16 execution test for builtin fract gpuweb/cts#3017, by @jiangzhaoming.
-
unittest: FP use correctly rounded dict for frequent number constant gpuweb/cts#3016, by @jiangzhaoming.
-
wgsl: add f16 builtin execution test for faceForward gpuweb/cts#3015, by @jiangzhaoming.
-
wgsl: add f16 builtin execution test for reflect and refract gpuweb/cts#3014, by @jiangzhaoming.
-
wgsl: add f16 builtin execution test for determinant gpuweb/cts#3013, by @jiangzhaoming.
-
wgsl: add f16 execution test for builtin sinh/cosh/tanh/asinh/acosh/atanh gpuweb/cts#3012, by @jiangzhaoming.
-
Compat limit render pass desc gpuweb/cts#3011, by @greggman.
-
Compat: make pipline-output-targets tests handle different limits gpuweb/cts#3010, by @greggman.
-
Compat: make maxXXXBufferBindingSize tests handle different limits gpuweb/cts#3009, by @greggman.
-
Compat: make maxDynamicXXX tests handle different limits gpuweb/cts#3008, by @greggman.
-
Compat: make maxComputeInvocationsPerWorkgroup test handle different limits gpuweb/cts#3007, by @greggman.
-
Compat: make maxColorAttachments test handle different limits gpuweb/cts#3006, by @greggman.
-
Compat: make buffer create test handle different limits gpuweb/cts#3005, by @greggman.
-
Compat: make vertex state correctness tests handle different limits gpuweb/cts#3004, by @greggman.
-
Compat: make device allocation test handle different limits gpuweb/cts#3003, by @greggman.
-
Compat: make requestDevice tests handle different limits gpuweb/cts#3002, by @greggman.
-
Compat: make basic compute test handle different limits gpuweb/cts#3001, by @greggman.
-
Compat: make attachment compatibility test handle different limits gpuweb/cts#3000, by @greggman.
-
Compat: Make setBindGroup tests handle different limits gpuweb/cts#2999, by @greggman.
-
Compat: Make setVertexBuffer test handle different limits gpuweb/cts#2998, by @greggman.
-
Compat: Make compute pass tests handle different limits gpuweb/cts#2997, by @greggman.
-
wgsl: Add AbstractFloat `fma` execution tests gpuweb/cts#2996, by @zoddicus.
-
Compat: Make createBindGroupLayout tests handle different limits gpuweb/cts#2995, by @greggman.
-
Compat: Make createBindGroup test handle different limits gpuweb/cts#2994, by @greggman.
-
wgsl: add f16 builtin mix execution test gpuweb/cts#2992, by @jiangzhaoming.
-
wgsl: f32 execution tests for builtin mix with non-matching parameters gpuweb/cts#2991, by @jiangzhaoming.
-
wgsl: add f16 execution test for builtin fma gpuweb/cts#2990, by @jiangzhaoming.
-
wgsl: add f16 execution test for builtin smoothstep gpuweb/cts#2989, by @jiangzhaoming.
-
Compat: Refactor vertex_state tests for new limits gpuweb/cts#2988, by @greggman.
-
wgsl: add f16 execution test for builtin tan gpuweb/cts#2987, by @jiangzhaoming.
-
wgsl: add f16 execution test for builtin pow gpuweb/cts#2986, by @jiangzhaoming.
-
wgsl: add f16 execution test for builtin normalize gpuweb/cts#2985, by @jiangzhaoming.
-
Convert specifying complex case combos to map-reduce generation gpuweb/cts#2984, by @zoddicus.
-
Show parseQuery error gpuweb/cts#2982, by @greggman.
-
Change generate-cache to validate-cache gpuweb/cts#2981, by @zoddicus.
-
wgsl: F16 execution tests for built-in length and distance gpuweb/cts#2979, by @jiangzhaoming.
-
util: Add negative zero in full F16/F32/F64 range gpuweb/cts#2978, by @jiangzhaoming.
-
wgsl: f16 to bool/i32/u32 conversion expression execution tests gpuweb/cts#2977, by @jiangzhaoming.
-
wgsl: Add AbstractFloat `cross` execution tests gpuweb/cts#2976, by @zoddicus.
-
wgsl: Add f16 `cross` execution tests gpuweb/cts#2975, by @zoddicus.
-
wgsl: Add AbstractFloat `saturate` execution tests gpuweb/cts#2974, by @zoddicus.
-
wgsl: Add AbstractFloat `clamp` execution tests gpuweb/cts#2973, by @zoddicus.
-
wgsl: Add AbstractFloat `min` execution tests gpuweb/cts#2972, by @zoddicus.
-
wgsl: Add AbstractFloat `max` execution tests gpuweb/cts#2971, by @zoddicus.
-
wgsl: Add interval names to unimplemented bindings gpuweb/cts#2967, by @zoddicus.
-
wgsl: F16 unary conversion expression execution tests gpuweb/cts#2965, by @jiangzhaoming.
-
wgsl: Add f16 `saturate` tests gpuweb/cts#2964, by @zoddicus.
-
wgsl: f16 execution test for built-in exp and exp2 gpuweb/cts#2963, by @jiangzhaoming.
-
wgsl: Add AbstractFloat `degree` execution tests gpuweb/cts#2962, by @zoddicus.
-
wgsl: Add f16 `degrees` execution tests gpuweb/cts#2961, by @zoddicus.
-
wgsl: Add AbstractFloat `radians` execution tests gpuweb/cts#2960, by @zoddicus.
-
wgsl: Add f16 `radians` execution tests gpuweb/cts#2959, by @zoddicus.
-
Add a copy button gpuweb/cts#2958, by @greggman.
-
Add rgb10a2uint GPUTextureFormat testing. gpuweb/cts#2956, by @Kangz.
-
wgsl: f16 execution test for binary matrix expression and built-in dot gpuweb/cts#2955, by @jiangzhaoming.
-
Compat: Limit draw test attributes in compat gpuweb/cts#2954, by @greggman.
-
Compat: Limit max attributes gpuweb/cts#2953, by @greggman.
-
wgsl: Add AbstractFloat `transpose` execution tests gpuweb/cts#2952, by @zoddicus.
-
wgsl: Add f16 `transpose` execution tests gpuweb/cts#2951, by @zoddicus.
-
wgsl: Add f16 remainder (`%`) execution tests gpuweb/cts#2950, by @zoddicus.
-
wgsl: Add f16 `trunc` execution tests gpuweb/cts#2949, by @zoddicus.
-
wgsl: Add AbstractFloat `trunc` execution tests gpuweb/cts#2948, by @zoddicus.
-
Update websocket-logger instructions to use `npm ci` gpuweb/cts#2947, by @zoddicus.
-
Split up 'scalar' tests into 'scalar' and 'vector' tests gpuweb/cts#2946, by @zoddicus.
-
Add missing timing metadata gpuweb/cts#2943, by @zoddicus.
-
Add documentation for adding timing metadata gpuweb/cts#2942, by @zoddicus.
-
Add unorm10-10-10-2 vertex format gpuweb/cts#2941, by @beaufortfrancois.
-
Compat: Test vertex_index, instance_index limits gpuweb/cts#2940, by @greggman.
-
Fix presubmits (`npm test`) gpuweb/cts#2937, by @ben-clayton.
-
Reverts gpuweb/cts#2936, by @ben-clayton.
-
wgsl: Add AbstractFloat `abs` execution tests gpuweb/cts#2932, by @zoddicus.
-
wgsl: Add AbstractFloat simple multiplication execution tests gpuweb/cts#2930, by @zoddicus.
-
wgsl: Add AbstractFloat matrix subtraction execution tests gpuweb/cts#2929, by @zoddicus.
-
wgsl: Add non-matrix AbstractFloat subtraction execution tests gpuweb/cts#2928, by @zoddicus.
-
wgsl: Add f16 negation execution tests gpuweb/cts#2927, by @zoddicus.
-
wgsl: Add AbstractFloat matrix addition execution tests gpuweb/cts#2926, by @zoddicus.
-
wgsl: f16 built-in execution test for frexp gpuweb/cts#2925, by @jiangzhaoming.
-
Compat: Skip if copyTextureToTexture not supported gpuweb/cts#2923, by @greggman.
-
wgsl: Implement scalar/vector AbstractFloat addition execution tests gpuweb/cts#2922, by @zoddicus.
-
Reland AF negation tests gpuweb/cts#2920, by @zoddicus.
-
Add `generate-cache` step to `grunt pre` gpuweb/cts#2919, by @ben-clayton.
-
wgsl: f16 built-in execution test for clamp gpuweb/cts#2918, by @jiangzhaoming.
-
Update fp_primer.md gpuweb/cts#2917, by @ben-clayton.
-
-
97 pull requests merged.
-
Remove limit constants gpuweb/cts#3037.
-
Compat: make device_pool not refer to kLimitInfo gpuweb/cts#3031.
-
Compat: make createRenderBundleEncoder test handle different limits gpuweb/cts#3027.
-
Fix Skip/Pass Issues gpuweb/cts#3036.
-
Compat: Make createTexture tests handle different limits gpuweb/cts#3035.
-
Compat: make perStageLimit tests handle different limits gpuweb/cts#3032.
-
Compat: make fragment_state test handle different limits gpuweb/cts#3030.
-
Update subcaseMS of web_platform,copyToTexture,ImageBitmap gpuweb/cts#3023.
-
Re-organize numeric constants to be consistent gpuweb/cts#3025.
-
Revert "util: Add negative zero in full F16/F32/F64 range (#2978)" gpuweb/cts#3029.
-
Compat: Skip copyTextureToTexture for compressed textures gpuweb/cts#3026.
-
Add rgb10a2uint GPUTextureFormat testing. gpuweb/cts#2956.
-
wgsl: Add AbstractFloat `modf` execution tests gpuweb/cts#3020.
-
wgsl: Add f16 `modf` execution tests gpuweb/cts#3019.
-
unittest: FP use correctly rounded dict for frequent number constant gpuweb/cts#3016.
-
wgsl: add f16 execution test for builtin fract gpuweb/cts#3017.
-
wgsl: refactor and add f16 execution test for builtin ldexp gpuweb/cts#3018.
-
wgsl: add f16 builtin execution test for faceForward gpuweb/cts#3015.
-
wgsl: add f16 builtin execution test for reflect and refract gpuweb/cts#3014.
-
wgsl: add f16 builtin execution test for determinant gpuweb/cts#3013.
-
Compat limit render pass desc gpuweb/cts#3011.
-
Compat: make pipline-output-targets tests handle different limits gpuweb/cts#3010.
-
Compat: make maxDynamicXXX tests handle different limits gpuweb/cts#3008.
-
Compat: make maxXXXBufferBindingSize tests handle different limits gpuweb/cts#3009.
-
wgsl: Add AbstractFloat `fma` execution tests gpuweb/cts#2996.
-
Compat: Make compute pass tests handle different limits gpuweb/cts#2997.
-
Compat: Make setVertexBuffer test handle different limits gpuweb/cts#2998.
-
Compat: make device allocation test handle different limits gpuweb/cts#3003.
-
Compat: Make setBindGroup tests handle different limits gpuweb/cts#2999.
-
wgsl: add f16 builtin mix execution test gpuweb/cts#2992.
-
wgsl: add f16 execution test for builtin pow gpuweb/cts#2986.
-
wgsl: add f16 execution test for builtin fma gpuweb/cts#2990.
-
wgsl: add f16 execution test for builtin smoothstep gpuweb/cts#2989.
-
wgsl: add f16 execution test for builtin tan gpuweb/cts#2987.
-
wgsl: add f16 execution test for builtin normalize gpuweb/cts#2985.
-
util: Add negative zero in full F16/F32/F64 range gpuweb/cts#2978.
-
Compat: make attachment compatibility test handle different limits gpuweb/cts#3000.
-
Compat: make vertex state correctness tests handle different limits gpuweb/cts#3004.
-
Compat: make buffer create test handle different limits gpuweb/cts#3005.
-
Compat: make maxColorAttachments test handle different limits gpuweb/cts#3006.
-
Compat: make requestDevice tests handle different limits gpuweb/cts#3002.
-
Compat: make basic compute test handle different limits gpuweb/cts#3001.
-
Compat: Make createBindGroupLayout tests handle different limits gpuweb/cts#2995.
-
Compat: Make createBindGroup test handle different limits gpuweb/cts#2994.
-
Compat: Refactor vertex_state tests for new limits gpuweb/cts#2988.
-
Convert specifying complex case combos to map-reduce generation gpuweb/cts#2984.
-
Change generate-cache to validate-cache gpuweb/cts#2981.
-
Show parseQuery error gpuweb/cts#2982.
-
wgsl: F16 execution tests for built-in length and distance gpuweb/cts#2979.
-
wgsl: f16 to bool/i32/u32 conversion expression execution tests gpuweb/cts#2977.
-
wgsl: Add AbstractFloat `cross` execution tests gpuweb/cts#2976.
-
wgsl: Add f16 `cross` execution tests gpuweb/cts#2975.
-
wgsl: Add AbstractFloat `saturate` execution tests gpuweb/cts#2974.
-
wgsl: Add AbstractFloat `clamp` execution tests gpuweb/cts#2973.
-
wgsl: Add AbstractFloat `min` execution tests gpuweb/cts#2972.
-
wgsl: Add AbstractFloat `max` execution tests gpuweb/cts#2971.
-
wgsl: Add interval names to unimplemented bindings gpuweb/cts#2967.
-
wgsl: F16 unary conversion expression execution tests gpuweb/cts#2965.
-
wgsl: Add f16 `saturate` tests gpuweb/cts#2964.
-
wgsl: f16 execution test for built-in exp and exp2 gpuweb/cts#2963.
-
wgsl: Add AbstractFloat `degree` execution tests gpuweb/cts#2962.
-
Compat: Limit draw test attributes in compat gpuweb/cts#2954.
-
wgsl: Add f16 `degrees` execution tests gpuweb/cts#2961.
-
wgsl: Add AbstractFloat `radians` execution tests gpuweb/cts#2960.
-
wgsl: Add f16 `radians` execution tests gpuweb/cts#2959.
-
wgsl: Add f16 remainder (`%`) execution tests gpuweb/cts#2950.
-
Add a copy button gpuweb/cts#2958.
-
wgsl: Add AbstractFloat `transpose` execution tests gpuweb/cts#2952.
-
wgsl: Add f16 `transpose` execution tests gpuweb/cts#2951.
-
wgsl: Add f16 `trunc` execution tests gpuweb/cts#2949.
-
wgsl: Add AbstractFloat `trunc` execution tests gpuweb/cts#2948.
-
Split up 'scalar' tests into 'scalar' and 'vector' tests gpuweb/cts#2946.
-
Compat: Limit max attributes gpuweb/cts#2953.
-
Update websocket-logger instructions to use `npm ci` gpuweb/cts#2947.
-
wgsl: Add AbstractFloat `abs` execution tests gpuweb/cts#2932.
-
wgsl: Add AbstractFloat simple multiplication execution tests gpuweb/cts#2930.
-
wgsl: Add AbstractFloat matrix subtraction execution tests gpuweb/cts#2929.
-
wgsl: Add non-matrix AbstractFloat subtraction execution tests gpuweb/cts#2928.
-
Add documentation for adding timing metadata gpuweb/cts#2942.
-
wgsl: Add AbstractFloat matrix addition execution tests gpuweb/cts#2926.
-
Compat: Test vertex_index, instance_index limits gpuweb/cts#2940.
-
Fix presubmits (`npm test`) gpuweb/cts#2937.
-
wgsl: Add f16 negation execution tests gpuweb/cts#2927.
-
Maintain test timing metadata in-repo and auto-chunk WPT gpuweb/cts#2770.
-
wgsl: f16 built-in execution test for frexp gpuweb/cts#2925.
-
wgsl: Implement scalar/vector AbstractFloat addition execution tests gpuweb/cts#2922.
-
wgsl: f16 built-in execution test for sign and step gpuweb/cts#2911.
-
wgsl: f16 built-in execution test for clamp gpuweb/cts#2918.
-
Compat: Skip if copyTextureToTexture not supported gpuweb/cts#2923.
-
Reland AF negation tests gpuweb/cts#2920.
-
Update fp_primer.md gpuweb/cts#2917.
-
Add `generate-cache` step to `grunt pre` gpuweb/cts#2919.
-
-
3 pull requests closed.
-
Update webgpu/types to add 'rgb10a2uint' gpuweb/cts#3024.
-
Add missing timing metadata gpuweb/cts#2943.
-
Reverts gpuweb/cts#2936.
-
-
64 pull requests received 163 new comments.
-
Maintain test timing metadata in-repo and auto-chunk WPT gpuweb/cts#2770 (16 comments by @austinEng, @greggman, @kainino0x)
-
Compat: Refactor vertex_state tests for new limits gpuweb/cts#2988 (10 comments by @greggman, @Kangz)
-
Convert specifying complex case combos to map-reduce generation gpuweb/cts#2984 (7 comments by @jiangzhaoming, @zoddicus)
-
wgsl: f16 execution test for binary matrix expression and built-in dot gpuweb/cts#2955 (6 comments by @jiangzhaoming, @zoddicus)
-
Add documentation for adding timing metadata gpuweb/cts#2942 (6 comments by @austinEng, @greggman, @zoddicus)
-
Implements filtering tests for mag/min/mipmapFilters with additional float32-filterable formats. gpuweb/cts#2915 (6 comments by @lokokung)
-
wgsl: Add f16 remainder (`%`) execution tests gpuweb/cts#2950 (5 comments by @jiangzhaoming, @zoddicus)
-
WIP: Experimental reconvergence tests gpuweb/cts#2916 (4 comments by @alan-baker, @dneto0, @kainino0x)
-
wgsl: Add non-matrix AbstractFloat subtraction execution tests gpuweb/cts#2928 (4 comments by @jiangzhaoming, @zoddicus)
-
wgsl: Implement scalar/vector AbstractFloat addition execution tests gpuweb/cts#2922 (4 comments by @jiangzhaoming, @zoddicus)
-
Update subcaseMS of web_platform,copyToTexture,ImageBitmap gpuweb/cts#3023 (3 comments by @jiangzhaoming)
-
wgsl: Add AbstractFloat `modf` execution tests gpuweb/cts#3020 (3 comments by @jiangzhaoming, @zoddicus)
-
wgsl: Add AbstractFloat `fma` execution tests gpuweb/cts#2996 (3 comments by @jiangzhaoming, @zoddicus)
-
wgsl: Add AbstractFloat matrix addition execution tests gpuweb/cts#2926 (3 comments by @jiangzhaoming, @zoddicus)
-
Compat: make fragment_state test handle different limits gpuweb/cts#3030 (2 comments by @greggman, @lokokung)
-
unittest: FP use correctly rounded dict for frequent number constant gpuweb/cts#3016 (2 comments by @jiangzhaoming)
-
Compat: make vertex state correctness tests handle different limits gpuweb/cts#3004 (2 comments by @greggman, @lokokung)
-
Compat: make buffer create test handle different limits gpuweb/cts#3005 (2 comments by @greggman, @lokokung)
-
Compat: Make createBindGroupLayout tests handle different limits gpuweb/cts#2995 (2 comments by @austinEng, @greggman)
-
wgsl: F16 unary conversion expression execution tests gpuweb/cts#2965 (2 comments by @jiangzhaoming)
-
wgsl: refactor and add f16 execution test for builtin ldexp gpuweb/cts#3018 (1 comment by @jiangzhaoming)
-
wgsl: add f16 builtin execution test for reflect and refract gpuweb/cts#3014 (1 comment by @jiangzhaoming)
-
wgsl: add f16 execution test for builtin sinh/cosh/tanh/asinh/acosh/atanh gpuweb/cts#3012 (1 comment by @jiangzhaoming)
-
Compat: make pipline-output-targets tests handle different limits gpuweb/cts#3010 (1 comment by @lokokung)
-
wgsl: f32 execution tests for builtin mix with non-matching parameters gpuweb/cts#2991 (1 comment by @jiangzhaoming)
-
wgsl: F16 execution tests for built-in length and distance gpuweb/cts#2979 (1 comment by @jiangzhaoming)
-
wgsl: f16 to bool/i32/u32 conversion expression execution tests gpuweb/cts#2977 (1 comment by @jiangzhaoming)
-
wgsl: Add AbstractFloat simple multiplication execution tests gpuweb/cts#2930 (1 comment by @zoddicus)
-
gpuweb/gpuweb
-
15 issues opened.
-
add a system-defined wsgl namespace for predeclared objects gpuweb/gpuweb#4308, by @dneto0.
-
Specify swizzles as text rather than grammar gpuweb/gpuweb#4307, by @mehmetoguzderin.
-
add subgroups, and make them portable if possible gpuweb/gpuweb#4306, by @dneto0.
-
Simplify @builtin() notation gpuweb/gpuweb#4304, by @munrocket.
-
Remove the need to hard-code texel format for `texture_storage_2d` in shaders gpuweb/gpuweb#4298, by @rokcej.
-
Add diagram with memory model to WGSL spec gpuweb/gpuweb#4296, by @munrocket.
-
What happens on the device timeline after `GPUDevice.destroy` is called? gpuweb/gpuweb#4294, by @jimblandy.
-
`AbstractInt` conversion ranking forces unfortunate compromise gpuweb/gpuweb#4291, by @jimblandy.
-
Add capability to query gpu "warp" size gpuweb/gpuweb#4290, by @rconde01.
-
The format of "64-bit unsigned integers" is never specified gpuweb/gpuweb#4289, by @jimblandy.
-
Add namespaces or struct methods. gpuweb/gpuweb#4286, by @wrightwriter.
-
Raise maxVertexBuffers default to 16 gpuweb/gpuweb#4284, by @kainino0x.
-
Dual Source Blending Investigation + Feature Proposal gpuweb/gpuweb#4283, by @bjjones.
-
feature request: relaxed ftoi gpuweb/gpuweb#4280, by @starmole.
-
`requiredLimits` must be an ordinary JS object gpuweb/gpuweb#4277, by @OptimisticPeach.
-
-
4 issues closed.
-
Allow read-write storage on selected texture formats gpuweb/gpuweb#1772.
-
Dual source blending gpuweb/gpuweb#391.
-
Definition of fixed-footprint types seems to include abstract types gpuweb/gpuweb#4212.
-
`maxBindingsPerBindGroup` might be too high gpuweb/gpuweb#4270.
-
-
1 issue reopened.
-
Support "uniform texel buffer" and "storage texel buffer" ? gpuweb/gpuweb#162, by @dneto0.
-
-
47 issues received 82 new comments.
-
Proposal: read-write storage textures gpuweb/gpuweb#3838 (6 comments by @dneto0, @jimblandy, @teoxoy)
-
Support "uniform texel buffer" and "storage texel buffer" ? gpuweb/gpuweb#162 (4 comments by @devshgraphicsprogramming, @dneto0)
-
`requiredLimits` must be an ordinary JS object gpuweb/gpuweb#4277 (4 comments by @greggman, @kainino0x, @OptimisticPeach)
-
`AbstractInt` conversion ranking forces unfortunate compromise gpuweb/gpuweb#4291 (4 comments by @alan-baker, @dneto0, @jimblandy)
-
Require render bundles to be no slower than non-render bundles for the same or similar set of commands gpuweb/gpuweb#4104 (4 comments by @greggman, @kdashg, @ottoville)
-
Should there be a way to prevent loop unrolling? gpuweb/gpuweb#4110 (4 comments by @FreddyFunk, @Jasper-Bekkers, @kdashg, @myth0genesis)
-
Investigation: Programmable Blending gpuweb/gpuweb#442 (2 comments by @devshgraphicsprogramming, @Kangz)
-
Support for explicit matmul instructions / tensor core instructions gpuweb/gpuweb#4137 (2 comments by @dneto0)
-
Proposal: Support DP4a as WGSL built-in functions gpuweb/gpuweb#2677 (2 comments by @dneto0, @Jiawei-Shao)
-
Add capability to query gpu "warp" size gpuweb/gpuweb#4290 (2 comments by @dneto0, @mehmetoguzderin)
-
Problem measuring time deltas with writeTimestamp gpuweb/gpuweb#3952 (2 comments by @fintelia, @greggman)
-
Only one pipeline layout for each entry point can be given as compilation hint gpuweb/gpuweb#4233 (2 comments by @kainino0x, @ottoville)
-
Case for optional GPUBindGroupLayoutEntry entries gpuweb/gpuweb#851 (2 comments by @floooh, @kainino0x)
-
add a system-defined wsgl namespace for predeclared objects gpuweb/gpuweb#4308 (1 comment by @dneto0)
-
wgsl: clarify what happens for IEEE 754 "invalid operation" cases, e.g. acos(2) gpuweb/gpuweb#4219 (1 comment by @dneto0)
-
wgsl: describe how the name `vec4` in `vec4<f32>` is resolved gpuweb/gpuweb#3280 (1 comment by @munrocket)
-
Remove the need to hard-code texel format for `texture_storage_2d` in shaders gpuweb/gpuweb#4298 (1 comment by @mehmetoguzderin)
-
What happens on the device timeline after `GPUDevice.destroy` is called? gpuweb/gpuweb#4294 (1 comment by @Kangz)
-
The format of "64-bit unsigned integers" is never specified gpuweb/gpuweb#4289 (1 comment by @dneto0)
-
-
12 pull requests opened.
-
wgsl: clarify float-to-int conversion functions gpuweb/gpuweb#4305, by @dneto0.
-
Split grammar extract build and build in tempdir gpuweb/gpuweb#4303, by @mehmetoguzderin.
-
Make workgroup_count consistent between specs gpuweb/gpuweb#4302, by @munrocket.
-
Regen logo in Codespaces for deterministic output gpuweb/gpuweb#4301, by @mehmetoguzderin.
-
Simplify invocation id definition gpuweb/gpuweb#4300, by @munrocket.
-
Add the favicon for webgpu-notext gpuweb/gpuweb#4299, by @mehmetoguzderin.
-
webgpu favicon gpuweb/gpuweb#4297, by @dcerisano.
-
Make GPUPipelineError constructor options optional gpuweb/gpuweb#4292, by @foolip.
-
Add unorm10-10-10-2 vertex format gpuweb/gpuweb#4288, by @beaufortfrancois.
-
Fix a typo gpuweb/gpuweb#4285, by @MrBrain295.
-
Remove wgsl_unit_tests gpuweb/gpuweb#4282, by @mehmetoguzderin.
-
Clarify that *concrete* fixed-footprint types are storable gpuweb/gpuweb#4279, by @dneto0.
-
-
7 pull requests merged.
-
wgsl: clarify float-to-int conversion functions gpuweb/gpuweb#4305.
-
Split grammar extract build and build in tempdir gpuweb/gpuweb#4303.
-
Simplify invocation id definition gpuweb/gpuweb#4300.
-
Regen logo in Codespaces for deterministic output gpuweb/gpuweb#4301.
-
Add the favicon for webgpu-notext gpuweb/gpuweb#4299.
-
Remove wgsl_unit_tests gpuweb/gpuweb#4282.
-
Clarify that *concrete* fixed-footprint types are storable gpuweb/gpuweb#4279.
-
-
3 pull requests closed.
-
Add pointer-chaining expressions: ptr.field and ptr[i] gpuweb/gpuweb#1580.
-
wgsl: support arrayLength on pointer to sized array gpuweb/gpuweb#2081.
-
webgpu favicon gpuweb/gpuweb#4297.
-
-
1 pull request reopened.
-
Make workgroup_count consistent between specs gpuweb/gpuweb#4302, by @munrocket.
-
-
12 pull requests received 42 new comments.
-
Split grammar extract build and build in tempdir gpuweb/gpuweb#4303 (8 comments by @dneto0, @mehmetoguzderin, @w3cbot)
-
Add unorm10-10-10-2 vertex format gpuweb/gpuweb#4288 (7 comments by @beaufortfrancois, @Kangz, @toji)
-
Add depthSlice to GPURenderPassColorAttachment for 3D texture gpuweb/gpuweb#4254 (6 comments by @haoxli, @kainino0x, @kdashg)
-
Make GPUPipelineError constructor options optional gpuweb/gpuweb#4292 (6 comments by @beaufortfrancois, @foolip, @Kangz, @toji)
-
TR publications
-
2 documents published.
-
WebGPU Shading Language published twice as a Working Draft. History of WebGPU Shading Language
-
GitHub
gpuweb/cts
-
5 issues opened.
-
Test cache generation on PRs gpuweb/cts#2913, by @austinEng.
-
wgsl: Scalar value of -0.0 should print as a negative zero gpuweb/cts#2901, by @dneto0.
-
Destroy devices onbeforeunload gpuweb/cts#2899, by @kainino0x.
-
Add option to discard device after a certain number of test cases have used it gpuweb/cts#2898, by @kainino0x.
-
test -0.0 in vertex data format tests gpuweb/cts#2895, by @dneto0.
-
-
4 issues closed.
-
wgsl: execution test, bitcast: f16 cases remain gpuweb/cts#1609.
-
wgsl: Scalar value of -0.0 should print as a negative zero gpuweb/cts#2901.
-
test -0.0 in vertex data format tests gpuweb/cts#2895.
-
Add a pseudo-random number generator to CTS gpuweb/cts#2881.
-
-
7 issues received 7 new comments.
-
web_platform: copyExternalImageToTexture with more source types (PNG, ImageData) with color spaces gpuweb/cts#1476 (1 comment by @kainino0x)
-
val: GPUShaderModule.getCompilationInfo and destroyed devices gpuweb/cts#2361 (1 comment by @kainino0x)
-
tests for unpack*unorm and unpack*snorm builtins should not "inherit" accuracy from float division gpuweb/cts#2485 (1 comment by @munrocket)
-
30 pull requests opened.
-
WIP: Experimental reconvergence tests gpuweb/cts#2916, by @alan-baker.
-
Remove orientation from several texture copy tests gpuweb/cts#2914, by @toji.
-
Revert "wgsl: Add execution tests for AF negation" gpuweb/cts#2912, by @austinEng.
-
wgsl: f16 built-in execution test for sign and step gpuweb/cts#2911, by @jiangzhaoming.
-
Don't color tests which have not been run as skipped gpuweb/cts#2910, by @greggman.
-
wgsl: Add execution tests for AF negation gpuweb/cts#2909, by @zoddicus.
-
Fix handling of batches in case filtering gpuweb/cts#2908, by @kainino0x.
-
wgsl: Rework AF support to be two-pass gpuweb/cts#2907, by @zoddicus.
-
Prevent details toggle if user is selecting text gpuweb/cts#2906, by @greggman.
-
Add support for structs to framework and use for modf gpuweb/cts#2905, by @zoddicus.
-
Fix handling of non-primitive param values in case filtering gpuweb/cts#2904, by @kainino0x.
-
Remove TODO about f32(-0.0) printing as '-0.0f' gpuweb/cts#2903, by @dneto0.
-
Reland "Optimize loadTreeForQuery by filtering eagerly in ParamsBuilder" gpuweb/cts#2902, by @kainino0x.
-
wgsl: f16 built-in execution test for bitcast gpuweb/cts#2897, by @jiangzhaoming.
-
Test -0.0 values in vertex format test gpuweb/cts#2896, by @dneto0.
-
Simplify var decl validation code gpuweb/cts#2894, by @dneto0.
-
[wgsl] Test negative one case for firstLeadingBit gpuweb/cts#2893, by @jrprice.
-
Switch GitHub Actions to an LTS version of Node (15 -> 16) gpuweb/cts#2892, by @kainino0x.
-
Cover ImageBitmap option "colorSpaceConversion: none" gpuweb/cts#2891, by @shaoboyan.
-
Optimize loadTreeForQuery by filtering eagerly in ParamsBuilder gpuweb/cts#2890, by @kainino0x.
-
Only trigger WPT Lint test on pull_request gpuweb/cts#2889, by @kainino0x.
-
Configure vscode to use our local copy of tsc instead of its own gpuweb/cts#2888, by @kainino0x.
-
Optimize check for duplicate parameters gpuweb/cts#2887, by @kainino0x.
-
Add HTMLImageElement and ImageData to copyExternalImageToTexture gpuweb/cts#2886, by @beaufortfrancois.
-
Standalone: print the runtime of the test run (button click) to the console gpuweb/cts#2885, by @kainino0x.
-
Update floating point framework based on spec changes gpuweb/cts#2884, by @zoddicus.
-
Add a pseudo-random number generator to CTS gpuweb/cts#2883, by @zoddicus.
-
wgsl: f16 built-in execution test for select gpuweb/cts#2882, by @jiangzhaoming.
-
-
29 pull requests merged.
-
Remove orientation from several texture copy tests gpuweb/cts#2914.
-
Don't color tests which have not been run as skipped gpuweb/cts#2910.
-
Revert "wgsl: Add execution tests for AF negation" gpuweb/cts#2912.
-
wgsl: Add execution tests for AF negation gpuweb/cts#2909.
-
Fix handling of batches in case filtering gpuweb/cts#2908.
-
Prevent details toggle if user is selecting text gpuweb/cts#2906.
-
Remove TODO about f32(-0.0) printing as '-0.0f' gpuweb/cts#2903.
-
wgsl: f16 built-in execution test for bitcast gpuweb/cts#2897.
-
Fix handling of non-primitive param values in case filtering gpuweb/cts#2904.
-
Test -0.0 values in vertex format test gpuweb/cts#2896.
-
Simplify var decl validation code gpuweb/cts#2894.
-
[wgsl] Test negative one case for firstLeadingBit gpuweb/cts#2893.
-
Add HTMLImageElement and ImageData to copyExternalImageToTexture gpuweb/cts#2886.
-
Switch GitHub Actions to an LTS version of Node (15 -> 16) gpuweb/cts#2892.
-
Only trigger WPT Lint test on pull_request gpuweb/cts#2889.
-
Cover ImageBitmap option "colorSpaceConversion: none" gpuweb/cts#2891.
-
Optimize loadTreeForQuery by filtering eagerly in ParamsBuilder gpuweb/cts#2890.
-
Configure vscode to use our local copy of tsc instead of its own gpuweb/cts#2888.
-
Optimize check for duplicate parameters gpuweb/cts#2887.
-
Update floating point framework based on spec changes gpuweb/cts#2884.
-
wgsl: Implement AF execution tests for comparison operations gpuweb/cts#2878.
-
wgsl: f16 built-in execution test for select gpuweb/cts#2882.
-
Add a pseudo-random number generator to CTS gpuweb/cts#2883.
-
wgsl: Change usages of 'any' to 'indeterminate' gpuweb/cts#2877.
-
Shader execution tests for f16 built-in log, log2 gpuweb/cts#2880.
-
Shader execution tests for f16 built-in min, max gpuweb/cts#2879.
-
-
4 pull requests closed.
-
wgsl: Rework AF support to be two-pass gpuweb/cts#2907.
-
Add support for structs to framework and use for modf gpuweb/cts#2905.
-
wgsl: unpack2x16snorm has 3ULP error gpuweb/cts#2483.
-
-
19 pull requests received 79 new comments.
-
Add support for structs to framework and use for modf gpuweb/cts#2905 (16 comments by @ben-clayton, @jiangzhaoming, @zoddicus)
-
Add HTMLImageElement and ImageData to copyExternalImageToTexture gpuweb/cts#2886 (16 comments by @beaufortfrancois, @kainino0x)
-
wgsl: f16 built-in execution test for bitcast gpuweb/cts#2897 (10 comments by @dneto0, @jiangzhaoming, @zoddicus)
-
wgsl: Rework AF support to be two-pass gpuweb/cts#2907 (6 comments by @ben-clayton, @jiangzhaoming, @zoddicus)
-
Implements filtering tests for mag/min/mipmapFilters with additional float32-filterable formats. gpuweb/cts#2915 (5 comments by @austinEng)
-
wgsl: Implement AF execution tests for comparison operations gpuweb/cts#2878 (4 comments by @dneto0, @zoddicus)
-
Optimize loadTreeForQuery by filtering eagerly in ParamsBuilder gpuweb/cts#2890 (3 comments by @kainino0x)
-
Prevent details toggle if user is selecting text gpuweb/cts#2906 (2 comments by @greggman, @kainino0x)
-
Fix case generation for validation of variable declaration access modes gpuweb/cts#2834 (1 comment by @dneto0)
-
Switch GitHub Actions to an LTS version of Node (15 -> 16) gpuweb/cts#2892 (1 comment by @kainino0x)
-
gpuweb/gpuweb
-
10 issues opened.
-
rgb10a2 vertex format gpuweb/gpuweb#4275, by @kainino0x.
-
Build: tree sitter still flaking sometimes gpuweb/gpuweb#4274, by @kainino0x.
-
Add missing limits to correspondence doc gpuweb/gpuweb#4273, by @teoxoy.
-
`maxBindingsPerBindGroup` might be too high gpuweb/gpuweb#4270, by @teoxoy.
-
ErrorScopes are not useful for CommandEncoder methods gpuweb/gpuweb#4269, by @kdashg.
-
Injecting WebGPU errors gpuweb/gpuweb#4268, by @kainino0x.
-
WebGPU Compatibility Mode gpuweb/gpuweb#4266, by @SenorBlanco.
-
Luminance-clamping definition in the spec is not how it actually works gpuweb/gpuweb#4263, by @kainino0x.
-
wgsl: clarify before-runtime handling of NaN by max, min builtins gpuweb/gpuweb#4262, by @dneto0.
-
Add detail to descriptions of `XyzFine` vs `XyzCoarse` gpuweb/gpuweb#4261, by @rconde01.
-
-
2 issues closed.
-
Render to 3D texture slices gpuweb/gpuweb#2877.
-
fract() returns values out of bounds on Intel Iris Graphics 550 gpuweb/gpuweb#4144.
-
-
1 issue reopened.
-
Standard Sample Pattern gpuweb/gpuweb#4005, by @kainino0x.
-
-
24 issues received 67 new comments.
-
WebGPU Compatibility Mode gpuweb/gpuweb#4266 (12 comments by @greggman, @kdashg, @litherum, @SenorBlanco, @teoxoy)
-
Proposal: read-write storage textures gpuweb/gpuweb#3838 (8 comments by @alan-baker, @Kangz, @litherum)
-
Investigation: Render to 3D texture gpuweb/gpuweb#4251 (7 comments by @aleino-nv, @haoxli, @litherum, @magcius)
-
Extended brightness range rendering gpuweb/gpuweb#4239 (6 comments by @kainino0x, @kdashg, @litherum)
-
fract() returns values out of bounds on Intel Iris Graphics 550 gpuweb/gpuweb#4144 (5 comments by @anvaka, @litherum)
-
Feature Request: Support for Matrix Inverse Function gpuweb/gpuweb#4115 (4 comments by @ezorzin, @litherum)
-
Build: tree sitter still flaking sometimes gpuweb/gpuweb#4274 (3 comments by @kainino0x, @mehmetoguzderin)
-
HTMLImageElement width, height needs spec in GPUImageCopyExternalImage gpuweb/gpuweb#4256 (3 comments by @greggman, @kainino0x)
-
Feature Request: Some way to generate identity matrices gpuweb/gpuweb#4026 (2 comments by @ben-clayton, @litherum)
-
Consider adding a struct initialization expression with named fields: `MyStruct { foo: 1, bar: 2 }` gpuweb/gpuweb#4210 (1 comment by @kainino0x)
-
Definition of fixed-footprint types seems to include abstract types gpuweb/gpuweb#4212 (1 comment by @dneto0)
-
Make lost devices appear to function as much as possible gpuweb/gpuweb#1629 (1 comment by @kainino0x)
-
-
6 pull requests opened.
-
Adjust Makefile to Handle wgsl Dependency Issues gpuweb/gpuweb#4276, by @mehmetoguzderin.
-
Add Vulkan's `maxBufferSize` to correspondence gpuweb/gpuweb#4272, by @teoxoy.
-
Remove `maxFragmentCombinedOutputResources` from correspondence gpuweb/gpuweb#4271, by @teoxoy.
-
[editorial] minor typo in logo gpuweb/gpuweb#4264, by @munrocket.
-
Fix 'and' to 'or' in discussion of bad error cases for atan2 gpuweb/gpuweb#4260, by @dneto0.
-
Add note that user-defined functions only have one overload. gpuweb/gpuweb#4259, by @dneto0.
-
-
9 pull requests merged.
-
Adjust Makefile to Handle wgsl Dependency Issues gpuweb/gpuweb#4276.
-
Add Vulkan's `maxBufferSize` to correspondence gpuweb/gpuweb#4272.
-
Remove `maxFragmentCombinedOutputResources` from correspondence gpuweb/gpuweb#4271.
-
[editorial] minor typo in logo gpuweb/gpuweb#4264.
-
Relate fp edge cases to IEEE 754 exceptions gpuweb/gpuweb#4257.
-
Fix 'and' to 'or' in discussion of bad error cases for atan2 gpuweb/gpuweb#4260.
-
Fix many-attributes for `override` summary gpuweb/gpuweb#4211.
-
Add note that user-defined functions only have one overload. gpuweb/gpuweb#4259.
-
Rephrase atan2 "ill-defined" comment as being about error gpuweb/gpuweb#4258.
-
-
1 pull request closed.
-
Allow sequence of GPUPipelineLayouts for GPUShaderModuleCompilationHint gpuweb/gpuweb#4234.
-
-
6 pull requests received 19 new comments.
-
Add depthSlice to GPURenderPassColorAttachment for 3D texture gpuweb/gpuweb#4254 (8 comments by @haoxli, @kainino0x)
-
Adjust Makefile to Handle wgsl Dependency Issues gpuweb/gpuweb#4276 (4 comments by @kainino0x, @mehmetoguzderin)
-
Allow sequence of GPUPipelineLayouts for GPUShaderModuleCompilationHint gpuweb/gpuweb#4234 (2 comments by @kainino0x, @ottoville)
-
TR publications
-
14 documents published.
-
WebGPU Shading Language published 10 times as a Working Draft. History of WebGPU Shading Language
-
WebGPU published 4 times as a Working Draft. History of WebGPU
-
Past meetings
GitHub
gpuweb/gpuweb
-
20 issues opened.
-
HTMLImageElement width, height needs spec in GPUImageCopyExternalImage gpuweb/gpuweb#4256, by @greggman.
-
HTMLCanvasElement copyExternalImageToTexture under specifed? gpuweb/gpuweb#4255, by @greggman.
-
Should `GPUQueue.writeTexture` check `[[destroyed]]`? gpuweb/gpuweb#4253, by @jimblandy.
-
writeBuffer contents are written on device timeline gpuweb/gpuweb#4252, by @AlbinBernhardssonARM.
-
Investigation: Render to 3D texture gpuweb/gpuweb#4251, by @haoxli.
-
Transfer a GPUTexture or GPUBuffer from one Worker to another. gpuweb/gpuweb#4244, by @sucong0826.
-
Extended brightness range rendering gpuweb/gpuweb#4239, by @ccameron-chromium.
-
Enable WebGPU in Safari TP gpuweb/gpuweb#4238, by @beaufortfrancois.
-
Consider raising `maxStorageBuffersPerShaderStage` gpuweb/gpuweb#4235, by @jimblandy.
-
Only one pipeline layout for each entry point can be given as compilation hint gpuweb/gpuweb#4233, by @litherum.
-
[Feature] WebGPU Support for Edge Browsers on Xbox Hardware gpuweb/gpuweb#4229, by @Mebob207.
-
maxColorAttachmentBytesPerSample default value too small gpuweb/gpuweb#4227, by @Yuu6883.
-
weaken accuracy of unpack4x8unorm, unpack4x8snorm, unpack2x16unorm to 3 ULP gpuweb/gpuweb#4224, by @dneto0.
-
Clarify usage of GPUShaderModuleCompilationHint gpuweb/gpuweb#4222, by @ottoville.
-
wgsl: clarify what happens for IEEE 754 "invalid operation" cases, e.g. acos(2) gpuweb/gpuweb#4219, by @dneto0.
-
add support for 16unorm and 16snorm texture formats gpuweb/gpuweb#4216, by @greggman.
-
add mirror-clamp-to-edge filtering gpuweb/gpuweb#4215, by @greggman.
-
Describe behavior for realizing circular `alias` declarations gpuweb/gpuweb#4214, by @sno2.
-
Definition of fixed-footprint types seems to include abstract types gpuweb/gpuweb#4212, by @jimblandy.
-
-
13 issues closed.
-
investigate ULP bounds on unpack2x16unorm gpuweb/gpuweb#4038.
-
Accuracy for AbstractFloat builtins not specified gpuweb/gpuweb#4134.
-
frexp + subnormals needs to be fully clarified gpuweb/gpuweb#4190.
-
read-only storage buffer access from vertex shaders should be allowed gpuweb/gpuweb#4132.
-
Add `rgb10a2uint` texture format gpuweb/gpuweb#3841.
-
Unify GPUQueue.writeTexture() and GPUQueue.copyExternalImageToTexture() gpuweb/gpuweb#4206.
-
maxColorAttachmentBytesPerSample default value too small gpuweb/gpuweb#4227.
-
[Feature] WebGPU Support for Edge Browsers on Xbox Hardware gpuweb/gpuweb#4229.
-
weaken accuracy of unpack4x8unorm, unpack4x8snorm, unpack2x16unorm to 3 ULP gpuweb/gpuweb#4224.
-
add support for 16unorm and 16snorm texture formats gpuweb/gpuweb#4216.
-
Describe behavior for realizing circular `alias` declarations gpuweb/gpuweb#4214.
-
-
29 issues received 62 new comments.
-
frexp + subnormals needs to be fully clarified gpuweb/gpuweb#4190 (6 comments by @dneto0, @zoddicus)
-
Only one pipeline layout for each entry point can be given as compilation hint gpuweb/gpuweb#4233 (6 comments by @kainino0x, @litherum)
-
Transfer a GPUTexture or GPUBuffer from one Worker to another. gpuweb/gpuweb#4244 (5 comments by @Kangz, @Linchenn, @sucong0826)
-
wgsl: clarify what happens for IEEE 754 "invalid operation" cases, e.g. acos(2) gpuweb/gpuweb#4219 (3 comments by @alan-baker, @dneto0)
-
Extended brightness range rendering gpuweb/gpuweb#4239 (3 comments by @ccameron-chromium, @kainino0x, @kdashg)
-
Clarify usage of GPUShaderModuleCompilationHint gpuweb/gpuweb#4222 (3 comments by @kainino0x, @litherum, @ottoville)
-
Accuracy for AbstractFloat builtins not specified gpuweb/gpuweb#4134 (3 comments by @alan-baker, @dneto0)
-
mapSync on Workers - and possibly on the main thread gpuweb/gpuweb#2217 (2 comments by @Dinnerbone, @greggman)
-
maxColorAttachmentBytesPerSample default value too small gpuweb/gpuweb#4227 (2 comments by @greggman, @Yuu6883)
-
add support for 16unorm and 16snorm texture formats gpuweb/gpuweb#4216 (2 comments by @greggman, @teoxoy)
-
Describe behavior for realizing circular `alias` declarations gpuweb/gpuweb#4214 (2 comments by @ben-clayton, @sno2)
-
Make lost devices appear to function as much as possible gpuweb/gpuweb#1629 (1 comment by @kainino0x)
-
read-only storage buffer access from vertex shaders should be allowed gpuweb/gpuweb#4132 (1 comment by @teoxoy)
-
Unify GPUQueue.writeTexture() and GPUQueue.copyExternalImageToTexture() gpuweb/gpuweb#4206 (1 comment by @litherum)
-
Failed to execute 'importExternalTexture' on 'GPUDevice': Failed to import texture from video frame that doesn't have back resource gpuweb/gpuweb#4242 (1 comment by @Kangz)
-
[Feature] WebGPU Support for Edge Browsers on Xbox Hardware gpuweb/gpuweb#4229 (1 comment by @greggman)
-
API spec should highlight vec3f 16 byte alignment vs 12 byte size gpuweb/gpuweb#4192 (1 comment by @dneto0)
-
-
24 pull requests opened.
-
Rephrase atan2 "ill-defined" comment as being about error gpuweb/gpuweb#4258, by @dneto0.
-
Relate fp edge cases to IEEE 754 exceptions gpuweb/gpuweb#4257, by @dneto0.
-
Add depthSlice to GPURenderPassColorAttachment for 3D texture gpuweb/gpuweb#4254, by @haoxli.
-
Clarify frexp: result for denorm are indeterminate gpuweb/gpuweb#4250, by @dneto0.
-
Delete TODO: made dfn nodes for predclared types gpuweb/gpuweb#4249, by @dneto0.
-
Explain accuracy rules for AbstractFloat gpuweb/gpuweb#4248, by @dneto0.
-
Clarify fp accuracy, remove redundant overflow text gpuweb/gpuweb#4247, by @dneto0.
-
Add `rgb10a2uint` texture format gpuweb/gpuweb#4245, by @teoxoy.
-
fix(wgsl): fix doc of reference ray gpuweb/gpuweb#4243, by @georeth.
-
Add CONTRIBUTING.md (explaining Codespaces) and vscode config gpuweb/gpuweb#4241, by @kainino0x.
-
[editorial] Add missing @must_use gpuweb/gpuweb#4237, by @alan-baker.
-
Allow sequence of GPUPipelineLayouts for GPUShaderModuleCompilationHint gpuweb/gpuweb#4234, by @ottoville.
-
Modify chmod mask rather than set gpuweb/gpuweb#4232, by @mehmetoguzderin.
-
Avoid in-place sed for previews gpuweb/gpuweb#4231, by @mehmetoguzderin.
-
Add Communication to README, mention Matrix, Issues, Discussions, and email lists. gpuweb/gpuweb#4230, by @kdashg.
-
Hotfix for preview-pull-request gpuweb/gpuweb#4226, by @mehmetoguzderin.
-
Error on unpack4x8snorm unpack4x8unofrm unpack2x16norm is 3ULP gpuweb/gpuweb#4225, by @dneto0.
-
Fix minor typo: option -> optional gpuweb/gpuweb#4223, by @sno2.
-
Clarify override-expression note gpuweb/gpuweb#4221, by @teoxoy.
-
Clarify "inherited from" accuracy gpuweb/gpuweb#4220, by @dneto0.
-
Always connect the uniformity graph after a function call. gpuweb/gpuweb#4218, by @alan-baker.
-
Editorial use custom Action for validation and deployment gpuweb/gpuweb#4217, by @mehmetoguzderin.
-
Replace usage of reserved identifier gpuweb/gpuweb#4213, by @sno2.
-
Fix many-attributes for `override` summary gpuweb/gpuweb#4211, by @sno2.
-
-
22 pull requests merged.
-
Explain accuracy rules for AbstractFloat gpuweb/gpuweb#4248.
-
fix(wgsl): fix doc of reference ray gpuweb/gpuweb#4243.
-
Clarify frexp: result for denorm are indeterminate gpuweb/gpuweb#4250.
-
Add `rgb10a2uint` texture format gpuweb/gpuweb#4245.
-
Add HTMLImageElement and ImageData to copyExternalImageToTexture gpuweb/gpuweb#3430.
-
Delete TODO: made dfn nodes for predclared types gpuweb/gpuweb#4249.
-
Clarify fp accuracy, remove redundant overflow text gpuweb/gpuweb#4247.
-
Add CONTRIBUTING.md (explaining Codespaces) and vscode config gpuweb/gpuweb#4241.
-
[editorial] Add missing @must_use gpuweb/gpuweb#4237.
-
Modify chmod mask rather than set gpuweb/gpuweb#4232.
-
Avoid in-place sed for previews gpuweb/gpuweb#4231.
-
Hotfix for preview-pull-request gpuweb/gpuweb#4226.
-
Editorial use custom Action for validation and deployment gpuweb/gpuweb#4217.
-
Clarify override-expression note gpuweb/gpuweb#4221.
-
Error on unpack4x8snorm unpack4x8unofrm unpack2x16norm is 3ULP gpuweb/gpuweb#4225.
-
Fix minor typo: option -> optional gpuweb/gpuweb#4223.
-
Clarify "inherited from" accuracy gpuweb/gpuweb#4220.
-
Always connect the uniformity graph after a function call. gpuweb/gpuweb#4218.
-
Clarify rounding of an overflow intermediate value gpuweb/gpuweb#4184.
-
fix(wgsl): typo in example gpuweb/gpuweb#4209.
-
Replace usage of reserved identifier gpuweb/gpuweb#4213.
-
-
15 pull requests received 147 new comments.
-
Editorial use custom Action for validation and deployment gpuweb/gpuweb#4217 (105 comments by @dneto0, @kainino0x, @mehmetoguzderin)
-
Explain accuracy rules for AbstractFloat gpuweb/gpuweb#4248 (11 comments by @alan-baker, @dneto0, @zoddicus)
-
Add HTMLImageElement and ImageData to copyExternalImageToTexture gpuweb/gpuweb#3430 (9 comments by @ccameron-chromium, @greggman, @kainino0x)
-
Clarify override-expression note gpuweb/gpuweb#4221 (6 comments by @alan-baker, @kainino0x, @mehmetoguzderin, @teoxoy, @w3cbot)
-
Always connect the uniformity graph after a function call. gpuweb/gpuweb#4218 (3 comments by @alan-baker, @dneto0, @w3cbot)
-
Add CONTRIBUTING.md (explaining Codespaces) and vscode config gpuweb/gpuweb#4241 (1 comment by @mehmetoguzderin)
-
Allow sequence of GPUPipelineLayouts for GPUShaderModuleCompilationHint gpuweb/gpuweb#4234 (1 comment by @kainino0x)
-
gpuweb/cts
-
10 issues opened.
-
Add a pseudo-random number generator to CTS gpuweb/cts#2881, by @alan-baker.
-
WGSL test parameters are incompatible with CTS harness gpuweb/cts#2860, by @greggman.
-
wgsl: Validate diagnostics gpuweb/cts#2859, by @alan-baker.
-
Validate function calls gpuweb/cts#2833, by @alan-baker.
-
spelling of pointer types gpuweb/cts#2832, by @dneto0.
-
Function declaration validation gpuweb/cts#2830, by @alan-baker.
-
What if maxComputeWorkGroupCount is 4294967295 gpuweb/cts#2826, by @bryceshang.
-
wgsl: validate that shadowing builtin functions is permitted gpuweb/cts#2800, by @ben-clayton.
-
-
110 issues closed.
-
Support f16 as a type gpuweb/cts#1187.
-
Relaxed precision for copy tests? gpuweb/cts#2793.
-
wgsl: f16 execution test for `sqrt` builtin gpuweb/cts#2532.
-
wgsl: f16 execution tests for `asin` builtin gpuweb/cts#2553.
-
wgsl: f16 execution tests for `acos` builtin gpuweb/cts#2551.
-
WGSL test parameters are incompatible with CTS harness gpuweb/cts#2860.
-
wgsl: @must_use attribute gpuweb/cts#2306.
-
wgsl: Validate diagnostics gpuweb/cts#2859.
-
Update fp_primer.md for f16, AbstractFloat, and changes to the spec gpuweb/cts#2790.
-
test updated rules for diagnostic names gpuweb/cts#2480.
-
wgsl: `unpack4x8unorm` is failing on some platforms gpuweb/cts#2777.
-
wgsl: execution test for atomicCompareExchangeWeak builtin gpuweb/cts#1283.
-
wgsl: execution test for atomicExchange builtin gpuweb/cts#1282.
-
wgsl: execution test for atomicStore builtin gpuweb/cts#1274.
-
wgsl: execution test for atomicLoad builtin gpuweb/cts#1273.
-
forbid spelling pointer types that are not instantiable gpuweb/cts#1427.
-
spelling of pointer types gpuweb/cts#2832.
-
Validate function calls gpuweb/cts#2833.
-
wgsl: validate `@size` parsing gpuweb/cts#1455.
-
wgsl: validate `@location` parsing gpuweb/cts#1453.
-
wgsl: Validate recursion disallowed gpuweb/cts#1467.
-
wgsl: Access mode restrictions gpuweb/cts#1538.
-
Function declaration validation gpuweb/cts#2830.
-
wgsl: f32 validation tests for `clamp` gpuweb/cts#2677.
-
wgsl: f16 validation tests for `clamp` gpuweb/cts#2678.
-
wgsl: AbstractFloat validation tests for `clamp` gpuweb/cts#2676.
-
wgsl: f32 validation tests for `length` gpuweb/cts#2716.
-
wgsl: f16 validation tests for `length` gpuweb/cts#2717.
-
wgsl: AbstractFloat validation tests for `length` gpuweb/cts#2715.
-
wgsl: length(AbstractFloat) yields f32, not a vector type gpuweb/cts#1588.
-
wgsl: Validate parsing of the `@interpolate` builtin gpuweb/cts#1449.
-
wgsl: uniformity update: function returns are a reconvergence point gpuweb/cts#1720.
-
wgsl: uniformity analysis case for short-circuiting || and && gpuweb/cts#1959.
-
wgsl: shader validation tests for uniformity analysis gpuweb/cts#2788.
-
wgsl: validate function restrictions gpuweb/cts#1690.
-
wgsl: f32 validation tests for `sinh` gpuweb/cts#2736.
-
wgsl: f16 validation tests for `sinh` gpuweb/cts#2737.
-
wgsl: f32 validation tests for `cosh` gpuweb/cts#2683.
-
wgsl: f16 validation tests for `cosh` gpuweb/cts#2684.
-
wgsl: AbstractFloat validation tests for `sinh` gpuweb/cts#2735.
-
wgsl: AbstractFloat validation tests for `sqrt` gpuweb/cts#2655.
-
wgsl: AbstractFloat validation tests for `cosh` gpuweb/cts#2682.
-
wgsl: f32 validation tests for `sqrt` gpuweb/cts#2643.
-
wgsl: f16 validation tests for `sqrt` gpuweb/cts#2644.
-
wgsl: f32 validation tests for `log2` gpuweb/cts#2648.
-
wgsl: AbstractFloat validation tests for `log2` gpuweb/cts#2657.
-
wgsl: f16 validation tests for `log2` gpuweb/cts#2647.
-
wgsl: AbstractFloat validation tests for `log` gpuweb/cts#2656.
-
wgsl: f32 validation tests for `log` gpuweb/cts#2645.
-
wgsl: f16 validation tests for `log` gpuweb/cts#2646.
-
wgsl: f16 validation tests for `inverseSqrt` gpuweb/cts#2642.
-
wgsl: AbstractFloat validation tests for `inverseSqrt` gpuweb/cts#2654.
-
wgsl: f32 validation tests for `inverseSqrt` gpuweb/cts#2641.
-
wgsl: f16 validation tests for `degrees` gpuweb/cts#2690.
-
wgsl: AbstractFloat validation tests for `degrees` gpuweb/cts#2688.
-
wgsl: f32 validation tests for `degrees` gpuweb/cts#2689.
-
wgsl: f32 validation tests for `cos` gpuweb/cts#2680.
-
wgsl: f16 validation tests for `cos` gpuweb/cts#2681.
-
wgsl: AbstractFloat validation tests for `cos` gpuweb/cts#2679.
-
wgsl: f32 validation tests for `sin` gpuweb/cts#2733.
-
wgsl: f16 validation tests for `sin` gpuweb/cts#2734.
-
wgsl: AbstractFloat validation tests for `sin` gpuweb/cts#2732.
-
wgsl: f16 validation tests for `tan` gpuweb/cts#2740.
-
wgsl: f32 validation tests for `tan` gpuweb/cts#2739.
-
wgsl: AbstractFloat validation tests for `tan` gpuweb/cts#2738.
-
wgsl: f16 validation tests for `tanh` gpuweb/cts#2743.
-
wgsl: f32 validation tests for `tanh` gpuweb/cts#2742.
-
wgsl: AbstractFloat validation tests for `tanh` gpuweb/cts#2741.
-
wgsl: validate `@binding` parsing gpuweb/cts#1434.
-
wgsl: validate `@workgroup_size` parsing gpuweb/cts#1462.
-
wgsl: Validate `@group` parsing gpuweb/cts#1439.
-
wgsl: Validate `@id` parsing gpuweb/cts#1444.
-
wgsl: validate `@invariant` parsing gpuweb/cts#1447.
-
wgsl: AbstractFloat validation tests for `exp` gpuweb/cts#2697.
-
wgsl: f32 validation tests for `exp` gpuweb/cts#2698.
-
wgsl: f16 validation tests for `exp` gpuweb/cts#2699.
-
wgsl: f32 validation tests for `exp2` gpuweb/cts#2701.
-
wgsl: f16 validation tests for `exp2` gpuweb/cts#2702.
-
wgsl: validate non-finite const eval for atan gpuweb/cts#1997.
-
wgsl: AbstractFloat validation tests for `exp2` gpuweb/cts#2700.
-
wgsl: Update correctlyRounded to have 3 regions of calculations gpuweb/cts#2776.
-
wgsl: f16 validation tests for `atanh` gpuweb/cts#2672.
-
wgsl: AbstractFloat validation tests for `atanh` gpuweb/cts#2670.
-
wgsl: f32 validation tests for `atanh` gpuweb/cts#2671.
-
wgsl: f32 validation tests for `asin` gpuweb/cts#2665.
-
wgsl: f16 validation tests for `asin` gpuweb/cts#2666.
-
wgsl: f16 validation tests for `acosh` gpuweb/cts#2663.
-
wgsl: AbstractFloat validation tests for `asin` gpuweb/cts#2664.
-
wgsl: f32 validation tests for `acosh` gpuweb/cts#2662.
-
wgsl: f16 validation tests for `acos` gpuweb/cts#2660.
-
wgsl: AbstractFloat validation tests for `acosh` gpuweb/cts#2661.
-
wgsl: AbstractFloat validation tests for `acos` gpuweb/cts#2658.
-
wgsl: f32 validation tests for `acos` gpuweb/cts#2659.
-
wgsl: validate workgroup_size gpuweb/cts#1691.
-
wgsl: validate that shadowing builtin functions is permitted gpuweb/cts#2800.
-
wgsl: validate `vec` parsing gpuweb/cts#1495.
-
add requires directive gpuweb/cts#2317.
-
wgsl: validation vector access expressions gpuweb/cts#1610.
-
wgsl: updated error conditions for vector accessing gpuweb/cts#1441.
-
wgsl: use 'alias' to create a type alias instead of 'type' gpuweb/cts#2190.
-
38 issues received 64 new comments.
-
WGSL test parameters are incompatible with CTS harness gpuweb/cts#2860 (8 comments by @ben-clayton, @dneto0, @greggman, @zoddicus)
-
wgsl: f32 validation tests for `sin` gpuweb/cts#2733 (6 comments by @ben-clayton, @dneto0, @zoddicus)
-
wgsl: `unpack4x8unorm` is failing on some platforms gpuweb/cts#2777 (5 comments by @dneto0, @zoddicus)
-
tests for unpack*unorm and unpack*snorm builtins should not "inherit" accuracy from float division gpuweb/cts#2485 (4 comments by @munrocket, @zoddicus)
-
wgsl: f32 validation tests for `reflect` gpuweb/cts#2730 (4 comments by @ben-clayton, @dj2, @dneto0)
-
Validate that `GPUBuffer.unmap` succeeds even when the buffer's device is invalid gpuweb/cts#2836 (3 comments by @ErichDonGubler, @kainino0x)
-
wgsl: user-defined functions must not have formal parameters of type pointer-to-workgroup gpuweb/cts#1926 (2 comments by @dneto0)
-
behaviour case: discard prevents undef behaviour from being executed gpuweb/cts#858 (2 comments by @dneto0)
-
web_platform: copyExternalImageToTexture with more source types (PNG, ImageData) with color spaces gpuweb/cts#1476 (1 comment by @kainino0x)
-
wgsl: comparison operators overloads with AbstractInt, AbstractFloat yield bool gpuweb/cts#1587 (1 comment by @jrprice)
-
wgsl: uniformity update: function returns are a reconvergence point gpuweb/cts#1720 (1 comment by @alan-baker)
-
wgsl: uniformity analysis case for short-circuiting || and && gpuweb/cts#1959 (1 comment by @alan-baker)
-
wgsl: add workgroupUniformLoad, update uniformity analysis for pointers, references vs. their contents gpuweb/cts#2054 (1 comment by @alan-baker)
-
wgsl: uniformity distinguished between pointer param uniformity and pointer param contents uniformity gpuweb/cts#2389 (1 comment by @alan-baker)
-
wgsl: execution tests for assigning `AbstractFloat` to `AbstractFloat` gpuweb/cts#2755 (1 comment by @ben-clayton)
-
wgsl: validate you can declare things with names that are: attributes, named params for attrib, texel format names gpuweb/cts#1448 (1 comment by @dj2)
-
wgsl: validate that shadowing builtin functions is permitted gpuweb/cts#2800 (1 comment by @ben-clayton)
-
wgsl: use 'alias' to create a type alias instead of 'type' gpuweb/cts#2190 (1 comment by @ben-clayton)
-
-
77 pull requests opened.
-
Shader execution tests for f16 built-in log, log2 gpuweb/cts#2880, by @jiangzhaoming.
-
Shader execution tests for f16 built-in min, max gpuweb/cts#2879, by @jiangzhaoming.
-
wgsl: Implement AF execution tests for comparison operations gpuweb/cts#2878, by @zoddicus.
-
wgsl: Change usages of 'any' to 'indeterminate' gpuweb/cts#2877, by @zoddicus.
-
Move compat tests under webgpu gpuweb/cts#2876, by @greggman.
-
Give a visual indication for skipped tests gpuweb/cts#2875, by @greggman.
-
Compat: Make compat tests skip in non-compat mode gpuweb/cts#2874, by @greggman.
-
Remove excluded formats from image_copy tests gpuweb/cts#2873, by @greggman.
-
Change texture/float comparisons for Inf gpuweb/cts#2872, by @greggman.
-
Massively speed up zero_init workgroup memory tests on Android gpuweb/cts#2871, by @toji.
-
Shader execution tests for f16 built-in asin, acos, atan, atan2 gpuweb/cts#2869, by @jiangzhaoming.
-
Add beginSubcases() to shader/validation clamp() tests gpuweb/cts#2868, by @ben-clayton.
-
Refactor texture_ok.spec.ts to not use typedarrays as params gpuweb/cts#2867, by @greggman.
-
Refactor floating_point.spec.ts unittest params gpuweb/cts#2866, by @greggman.
-
Validate @must_use gpuweb/cts#2865, by @alan-baker.
-
shader/validation: Fix case parameter types gpuweb/cts#2863, by @ben-clayton.
-
Reject invalid test params gpuweb/cts#2862, by @greggman.
-
Validate diagnostics gpuweb/cts#2861, by @alan-baker.
-
Fix size_fp16 test gpuweb/cts#2858, by @alan-baker.
-
De-duplicate resource interface tests gpuweb/cts#2857, by @alan-baker.
-
Add F16 builtin execution tests: sin, cos gpuweb/cts#2856, by @jiangzhaoming.
-
util: Clean PowTwo and NegPowTwo in kBit gpuweb/cts#2855, by @jiangzhaoming.
-
Add F16 builtin execution tests: arrayLength gpuweb/cts#2854, by @jiangzhaoming.
-
Compat: Test bindgroup usage validation gpuweb/cts#2853, by @greggman.
-
Compat: Test copyTextureToBuffer fails on compressed textures gpuweb/cts#2852, by @greggman.
-
Compat: Test you can't use bgra8unorm-srgb gpuweb/cts#2851, by @greggman.
-
Compat: Test that sample_mask pipelines fail gpuweb/cts#2850, by @greggman.
-
shader/validation: Use `.beginSubcases()` gpuweb/cts#2849, by @ben-clayton.
-
Compat: Test mismatched color states gpuweb/cts#2848, by @greggman.
-
Compat: Skip tests that use different views of same texture gpuweb/cts#2847, by @greggman.
-
Compat: Skip sample_mask test gpuweb/cts#2846, by @greggman.
-
Compat: fragment_state:limits,maxColorAttachments gpuweb/cts#2845, by @greggman.
-
Compat: Skip bgra8unorm-srgb in validation test gpuweb/cts#2844, by @greggman.
-
Fix unpackRGB9E5UFloat usage and add tests gpuweb/cts#2843, by @greggman.
-
Convert remaining unpack intervals to 3 ULP gpuweb/cts#2842, by @zoddicus.
-
Validate function calls gpuweb/cts#2841, by @alan-baker.
-
Issue 2832 validate pointer types spelled in WGSL source gpuweb/cts#2840, by @dneto0.
-
[wgsl] Add validation tests for size attribute. gpuweb/cts#2839, by @dj2.
-
fix comment typo gpuweb/cts#2838, by @greggman.
-
validate spelling of pointers declared via let gpuweb/cts#2837, by @dneto0.
-
[wgsl] Validate no recursion for const/override gpuweb/cts#2835, by @jrprice.
-
Fix case generation for validation of variable declaration access modes gpuweb/cts#2834, by @dneto0.
-
Validate function declarations gpuweb/cts#2831, by @alan-baker.
-
Validate builtins gpuweb/cts#2829, by @ben-clayton.
-
shader/execution: Fix clamp() tests for const-eval gpuweb/cts#2828, by @ben-clayton.
-
Compat: verify compressed format copies by rendering gpuweb/cts#2827, by @greggman.
-
[wgsl] Validate no recursion in structure types gpuweb/cts#2825, by @jrprice.
-
validate access mode on var decl gpuweb/cts#2824, by @dneto0.
-
[wgsl] Validate no recursion in type aliases gpuweb/cts#2823, by @jrprice.
-
[wgsl] Add location validation tests gpuweb/cts#2822, by @dj2.
-
shader/validation: Validate length() gpuweb/cts#2821, by @ben-clayton.
-
[wgsl] Add more interpolation tests gpuweb/cts#2820, by @dj2.
-
Tests for atomic builtins: atomicLoad, atomicStore, atomicExchange, a… gpuweb/cts#2819, by @amaiorano.
-
Validate builtins gpuweb/cts#2818, by @ben-clayton.
-
Function restriction validation tests gpuweb/cts#2817, by @alan-baker.
-
Prevent server from exiting on bad suite name gpuweb/cts#2816, by @greggman.
-
[wgsl] More workgroup_size tests. gpuweb/cts#2815, by @dj2.
-
Validate builtins gpuweb/cts#2814, by @ben-clayton.
-
Tweak shader_io/id.spec.ts gpuweb/cts#2813, by @ben-clayton.
-
Fix texel view for signed values gpuweb/cts#2812, by @greggman.
-
[wgsl] More invariant parsing tests. gpuweb/cts#2811, by @dj2.
-
[wgsl] add id validation tests. gpuweb/cts#2810, by @dj2.
-
Change 'storageClass' to 'addressSpace' gpuweb/cts#2809, by @dneto0.
-
[wgsl] Validate @group and @binding parsing gpuweb/cts#2808, by @dj2.
-
Add bool cases to vector type validation gpuweb/cts#2807, by @alan-baker.
-
Validate `atan()`, `exp2()` and `exp()` gpuweb/cts#2806, by @ben-clayton.
-
Implement f16-based bitcast validation tests gpuweb/cts#2805, by @alan-baker.
-
Validation for uniformity of pointer parameters gpuweb/cts#2804, by @alan-baker.
-
wgsl: Add tests for pointer alias analysis gpuweb/cts#2803, by @jrprice.
-
Shader uniformity validation tests for functions gpuweb/cts#2802, by @alan-baker.
-
shader/validation: Expand identifier validation tests gpuweb/cts#2801, by @ben-clayton.
-
shader/validation: Validate trig builtins gpuweb/cts#2799, by @ben-clayton.
-
shader/validation: Add vector type validation tests gpuweb/cts#2798, by @ben-clayton.
-
shader/validation: Add vector validation tests gpuweb/cts#2797, by @ben-clayton.
-
Reimplement `packRGB9E5UFloat()`, add `unpackRGB9E5UFloat()` gpuweb/cts#2796, by @ben-clayton.
-
Fix image_copy tests gpuweb/cts#2795, by @greggman.
-
-
75 pull requests merged.
-
wgsl: Implement AF -> AF assignment tests gpuweb/cts#2756.
-
Give a visual indication for skipped tests gpuweb/cts#2875.
-
Move compat tests under webgpu gpuweb/cts#2876.
-
Compat: Make compat tests skip in non-compat mode gpuweb/cts#2874.
-
Massively speed up zero_init workgroup memory tests on Android gpuweb/cts#2871.
-
Remove excluded formats from image_copy tests gpuweb/cts#2873.
-
Change texture/float comparisons for Inf gpuweb/cts#2872.
-
Shader execution tests for f16 built-in asin, acos, atan, atan2 gpuweb/cts#2869.
-
Reject invalid test params gpuweb/cts#2862.
-
Add beginSubcases() to shader/validation clamp() tests gpuweb/cts#2868.
-
Refactor texture_ok.spec.ts to not use typedarrays as params gpuweb/cts#2867.
-
Refactor floating_point.spec.ts unittest params gpuweb/cts#2866.
-
shader/validation: Fix case parameter types gpuweb/cts#2863.
-
Validate @must_use gpuweb/cts#2865.
-
Validate diagnostics gpuweb/cts#2861.
-
Update FP documentation gpuweb/cts#2794.
-
De-duplicate resource interface tests gpuweb/cts#2857.
-
Fix size_fp16 test gpuweb/cts#2858.
-
Compat: Test bindgroup usage validation gpuweb/cts#2853.
-
Add F16 builtin execution tests: sin, cos gpuweb/cts#2856.
-
Add F16 builtin execution tests: arrayLength gpuweb/cts#2854.
-
util: Clean PowTwo and NegPowTwo in kBit gpuweb/cts#2855.
-
Convert remaining unpack intervals to 3 ULP gpuweb/cts#2842.
-
Compat: Test mismatched color states gpuweb/cts#2848.
-
Compat: Test copyTextureToBuffer fails on compressed textures gpuweb/cts#2852.
-
Compat: Test that sample_mask pipelines fail gpuweb/cts#2850.
-
Fix unpackRGB9E5UFloat usage and add tests gpuweb/cts#2843.
-
Compat: Test you can't use bgra8unorm-srgb gpuweb/cts#2851.
-
shader/validation: Use `.beginSubcases()` gpuweb/cts#2849.
-
Compat: verify compressed format copies by rendering gpuweb/cts#2827.
-
Issue 2832 validate pointer types spelled in WGSL source gpuweb/cts#2840.
-
Compat: Skip bgra8unorm-srgb in validation test gpuweb/cts#2844.
-
Compat: Skip tests that use different views of same texture gpuweb/cts#2847.
-
Compat: fragment_state:limits,maxColorAttachments gpuweb/cts#2845.
-
Compat: Skip sample_mask test gpuweb/cts#2846.
-
Prevent server from exiting on bad suite name gpuweb/cts#2816.
-
Validate function calls gpuweb/cts#2841.
-
[wgsl] Add validation tests for size attribute. gpuweb/cts#2839.
-
[wgsl] Add location validation tests gpuweb/cts#2822.
-
fix comment typo gpuweb/cts#2838.
-
[wgsl] Validate no recursion for const/override gpuweb/cts#2835.
-
Fix image_copy tests gpuweb/cts#2795.
-
validate access mode on var decl gpuweb/cts#2824.
-
Validate function declarations gpuweb/cts#2831.
-
Validate builtins gpuweb/cts#2829.
-
shader/execution: Fix clamp() tests for const-eval gpuweb/cts#2828.
-
[wgsl] Validate no recursion in structure types gpuweb/cts#2825.
-
[wgsl] Validate no recursion in type aliases gpuweb/cts#2823.
-
shader/validation: Validate length() gpuweb/cts#2821.
-
Add bool cases to vector type validation gpuweb/cts#2807.
-
[wgsl] Add more interpolation tests gpuweb/cts#2820.
-
Validation for uniformity of pointer parameters gpuweb/cts#2804.
-
Function restriction validation tests gpuweb/cts#2817.
-
Validate builtins gpuweb/cts#2818.
-
Validate builtins gpuweb/cts#2814.
-
[wgsl] More workgroup_size tests. gpuweb/cts#2815.
-
[wgsl] Validate @group and @binding parsing gpuweb/cts#2808.
-
Tweak shader_io/id.spec.ts gpuweb/cts#2813.
-
Fix texel view for signed values gpuweb/cts#2812.
-
[wgsl] add id validation tests. gpuweb/cts#2810.
-
[wgsl] More invariant parsing tests. gpuweb/cts#2811.
-
wgsl: Add tests for pointer alias analysis gpuweb/cts#2803.
-
Change 'storageClass' to 'addressSpace' gpuweb/cts#2809.
-
Implement f16-based bitcast validation tests gpuweb/cts#2805.
-
Validate `atan()`, `exp2()` and `exp()` gpuweb/cts#2806.
-
Reimplement `packRGB9E5UFloat()`, add `unpackRGB9E5UFloat()` gpuweb/cts#2796.
-
Add support for overflow domains to correctly rounded gpuweb/cts#2780.
-
Shader uniformity validation tests for functions gpuweb/cts#2802.
-
shader/validation: Validate trig builtins gpuweb/cts#2799.
-
[wgsl] Add workgroup_size validation tests. gpuweb/cts#2526.
-
shader/validation: Add vector type validation tests gpuweb/cts#2798.
-
shader/validation: Expand identifier validation tests gpuweb/cts#2801.
-
shader/validation: Add vector validation tests gpuweb/cts#2797.
-
-
1 pull request closed.
-
validate spelling of pointers declared via let gpuweb/cts#2837.
-
-
47 pull requests received 302 new comments.
-
Tests for atomic builtins: atomicLoad, atomicStore, atomicExchange, a… gpuweb/cts#2819 (45 comments by @alan-baker, @amaiorano, @ben-clayton, @dneto0)
-
Fix unpackRGB9E5UFloat usage and add tests gpuweb/cts#2843 (6 comments by @ben-clayton, @greggman, @kainino0x)
-
Issue 2832 validate pointer types spelled in WGSL source gpuweb/cts#2840 (6 comments by @ben-clayton, @dneto0)
-
Fix case generation for validation of variable declaration access modes gpuweb/cts#2834 (4 comments by @ben-clayton)
-
shader/validation: Add vector validation tests gpuweb/cts#2797 (4 comments by @ben-clayton, @jrprice)
-
Compat: verify compressed format copies by rendering gpuweb/cts#2827 (3 comments by @greggman, @kainino0x)
-
wgsl: Implement AF execution tests for comparison operations gpuweb/cts#2878 (2 comments by @zoddicus)
-
Compat: Make compat tests skip in non-compat mode gpuweb/cts#2874 (2 comments by @greggman, @lokokung)
-
Massively speed up zero_init workgroup memory tests on Android gpuweb/cts#2871 (2 comments by @toji)
-
Shader execution tests for f16 Add/Sub/Mul/Div and built-in sqrt/inveseSqrt gpuweb/cts#2864 (2 comments by @jiangzhaoming)
-
Compat: Test copyTextureToBuffer fails on compressed textures gpuweb/cts#2852 (2 comments by @greggman, @lokokung)
-
Validation for uniformity of pointer parameters gpuweb/cts#2804 (2 comments by @alan-baker, @jrprice)
-
Shader uniformity validation tests for functions gpuweb/cts#2802 (2 comments by @alan-baker, @jrprice)
-
shader/validation: Add vector type validation tests gpuweb/cts#2798 (2 comments by @ben-clayton, @dneto0)
-
Shader execution tests for f16 built-in asin, acos, atan, atan2 gpuweb/cts#2869 (1 comment by @jiangzhaoming)
TR publications
-
10 documents published.
-
WebGPU Shading Language published 5 times as a Working Draft. History of WebGPU Shading Language
-
WebGPU published 5 times as a Working Draft. History of WebGPU
-
Participants
-
2 participants joined.
-
Aura Munoz (Autodesk) joined.
-
Anders Leino (NVIDIA) joined.
-
GitHub
gpuweb/gpuweb
-
24 issues opened.
-
Unify GPUQueue.writeTexture() and GPUQueue.copyExternalImageToTexture() gpuweb/gpuweb#4206, by @litherum.
-
API spec build broken with latest bikeshed 3.13.1 gpuweb/gpuweb#4199, by @kainino0x.
-
API should be exposed to ServiceWorker gpuweb/gpuweb#4197, by @MiguelsPizza.
-
Subgroup matrix gpuweb/gpuweb#4195, by @alan-baker.
-
getCurrentTexture should never try to allocate an invalid size texture gpuweb/gpuweb#4193, by @greggman.
-
API spec should highlight vec3f 16 byte alignment vs 12 byte size gpuweb/gpuweb#4192, by @mikeskk.
-
frexp + subnormals needs to be fully clarified gpuweb/gpuweb#4190, by @zoddicus.
-
writeBuffer/writeTexture works with shared ArrayBufferViews but not SharedArrayBuffer gpuweb/gpuweb#4186, by @kainino0x.
-
Image uploading is insufficiently expressive to be optimized gpuweb/gpuweb#4185, by @kainino0x.
-
GPUDevice.createBuffer() may throw in content timeline gpuweb/gpuweb#4182, by @beaufortfrancois.
-
Feature request: textureQueryLod equivalent gpuweb/gpuweb#4180, by @magcius.
-
device.destroy() cannot simulate real device loss gpuweb/gpuweb#4177, by @ottoville.
-
Reference High Resolution Time spec gpuweb/gpuweb#4175, by @kainino0x.
-
Must debug groups be balanced within a single command buffer? gpuweb/gpuweb#4172, by @jimblandy.
-
Support for device pointers gpuweb/gpuweb#4168, by @holographicGirl.
-
Recursion in WGSL, like in CUDA and GLSL (Vulkan extension)? gpuweb/gpuweb#4167, by @holographicGirl.
-
WGSL bitwise operator error: shift left operation results in sign change gpuweb/gpuweb#4166, by @coramat.
-
Supported sources for copyExternalImageToTexture() gpuweb/gpuweb#4165, by @sandersdan.
-
Allow forward declaration of functions gpuweb/gpuweb#4162, by @wrightwriter.
-
GPUExternalTexture "expired" attribute is not web exposed in JS example gpuweb/gpuweb#4161, by @beaufortfrancois.
-
Clarify behavior of (norm) packing built-in functions gpuweb/gpuweb#4159, by @teoxoy.
-
Add built-in struct types for indirect calls gpuweb/gpuweb#4158, by @teoxoy.
-
14 issues closed.
-
API spec build broken with latest bikeshed 3.13.1 gpuweb/gpuweb#4199.
-
[AllowShared] BufferSource isn't well-defined gpuweb/gpuweb#4042.
-
Consider removing or reworking implicit pipeline layouts gpuweb/gpuweb#2470.
-
GPUDevice.createBuffer() may throw in content timeline gpuweb/gpuweb#4182.
-
Handling of omitted labels is unclear gpuweb/gpuweb#4083.
-
Supported sources for copyExternalImageToTexture() gpuweb/gpuweb#4165.
-
GPUExternalTexture "expired" attribute is not web exposed in JS example gpuweb/gpuweb#4161.
-
Clarify behavior of (norm) packing built-in functions gpuweb/gpuweb#4159.
-
Should navigator.gpu not be readonly? gpuweb/gpuweb#4150.
-
[editorial] Revisit the built-in values table gpuweb/gpuweb#4151.
-
[editorial] Revisit the overview gpuweb/gpuweb#4153.
-
WGSL bitwise operator error: shift left operation results in sign change gpuweb/gpuweb#4166.
-
Allow forward declaration of functions gpuweb/gpuweb#4162.
-
[editorial] GPUPipelineError constructor argument `message` should be optional gpuweb/gpuweb#3709.
-
-
1 issue reopened.
-
Make lost devices appear to function as much as possible gpuweb/gpuweb#1629, by @kainino0x.
-
-
41 issues received 153 new comments.
-
Add arrow operator as sugar for (*var). gpuweb/gpuweb#4114 (16 comments by @dneto0, @jimblandy, @kainino0x, @kdashg, @litherum)
-
Space efficient wide color support gpuweb/gpuweb#4108 (15 comments by @aleino-nv, @kainino0x, @kdashg, @litherum, @RafaelCintron)
-
Image uploading is insufficiently expressive to be optimized gpuweb/gpuweb#4185 (9 comments by @greggman, @kainino0x, @shaoboyan)
-
Supported sources for copyExternalImageToTexture() gpuweb/gpuweb#4165 (9 comments by @beaufortfrancois, @kainino0x, @kdashg, @litherum, @sandersdan, @shaoboyan)
-
Floating Point rules + const/override allow for significantly different implementation behaviours gpuweb/gpuweb#4183 (7 comments by @dneto0, @zoddicus)
-
pushDebugGroup / popDebugGroup on Queue gpuweb/gpuweb#1672 (7 comments by @jimblandy, @kainino0x, @Kangz, @magcius)
-
Clarify behavior of (norm) packing built-in functions gpuweb/gpuweb#4159 (6 comments by @dneto0, @Elabajaba, @kdashg, @magcius, @teoxoy)
-
Unify GPUQueue.writeTexture() and GPUQueue.copyExternalImageToTexture() gpuweb/gpuweb#4206 (5 comments by @emidoots, @greggman, @kainino0x, @magcius)
-
Cannot upload/download to UMA storage buffer without an unnecessary copy and unnecessary memory use gpuweb/gpuweb#2388 (4 comments by @ErichDonGubler, @kainino0x, @kdashg, @MikhailGorobets)
-
API spec build broken with latest bikeshed 3.13.1 gpuweb/gpuweb#4199 (4 comments by @JustAnotherCodemonkey, @kainino0x)
-
getCurrentTexture should never try to allocate an invalid size texture gpuweb/gpuweb#4193 (3 comments by @greggman, @kainino0x)
-
Should there be a way to prevent loop unrolling? gpuweb/gpuweb#4110 (3 comments by @kdashg, @myth0genesis, @SarkyKeshy)
-
Why GPUCommandBuffer and GPUCommandEncoder cannot be reused? gpuweb/gpuweb#4138 (3 comments by @kainino0x, @kdashg, @litherum)
-
Recursion in WGSL, like in CUDA and GLSL (Vulkan extension)? gpuweb/gpuweb#4167 (3 comments by @Kangz, @kdashg, @litherum)
-
WGSL bitwise operator error: shift left operation results in sign change gpuweb/gpuweb#4166 (3 comments by @ben-clayton, @coramat)
-
Make lost devices appear to function as much as possible gpuweb/gpuweb#1629 (2 comments by @kainino0x)
-
API spec should highlight vec3f 16 byte alignment vs 12 byte size gpuweb/gpuweb#4192 (2 comments by @greggman, @shannon)
-
Must debug groups be balanced within a single command buffer? gpuweb/gpuweb#4172 (2 comments by @jimblandy, @kainino0x)
-
Allow forward declaration of functions gpuweb/gpuweb#4162 (2 comments by @ben-clayton, @wrightwriter)
-
fract() returns values out of bounds on Intel Iris Graphics 550 gpuweb/gpuweb#4144 (1 comment by @kdashg)
-
-
27 pull requests opened.
-
fix(wgsl): typo in example gpuweb/gpuweb#4209, by @sno2.
-
fix(wgsl): fix invalid var decl example gpuweb/gpuweb#4208, by @sno2.
-
Add rgb10a2unorm as a canvas format gpuweb/gpuweb#4207, by @kainino0x.
-
Tentatively disable mermaid check gpuweb/gpuweb#4205, by @mehmetoguzderin.
-
Try generating Mermaid diagrams to match CI gpuweb/gpuweb#4204, by @mehmetoguzderin.
-
Replace explicit invocations of sh with bash gpuweb/gpuweb#4203, by @kainino0x.
-
Change all shell scripts to use bash and `set -eo pipefail` gpuweb/gpuweb#4202, by @kainino0x.
-
Update bikeshed to 3.13.1 gpuweb/gpuweb#4200, by @kainino0x.
-
Guarantee that OOB writes to storage textures will be ignored gpuweb/gpuweb#4194, by @teoxoy.
-
Remove samples gpuweb/gpuweb#4191, by @greggman.
-
Allow SharedArrayBuffer in writeBuffer/writeTexture with AllowSharedBufferSource gpuweb/gpuweb#4188, by @annevk.
-
Accept SharedArrayBuffer in writeBuffer/writeTexture gpuweb/gpuweb#4187, by @kainino0x.
-
Clarify rounding of an overflow intermediate value gpuweb/gpuweb#4184, by @dneto0.
-
Remove expired attribute from JS example gpuweb/gpuweb#4181, by @beaufortfrancois.
-
Make GPUObjectDescriptorBase.label default to the empty string. gpuweb/gpuweb#4179, by @jimblandy.
-
Add VideoFrame to GPUImageCopyExternalImage source gpuweb/gpuweb#4178, by @beaufortfrancois.
-
[editorial] Make Timelines normative gpuweb/gpuweb#4176, by @kainino0x.
-
[editorial] Fix typo in uniformity table gpuweb/gpuweb#4174, by @alan-baker.
-
[editorial] GPUPipelineError: set slots instead of setting getters gpuweb/gpuweb#4173, by @kainino0x.
-
Update WGSL reference for WebGPU gpuweb/gpuweb#4171, by @alan-baker.
-
s/importTexture/importExternalTexture/ gpuweb/gpuweb#4170, by @beaufortfrancois.
-
[editorial] Overview edits gpuweb/gpuweb#4164, by @alan-baker.
-
[editorial] Rework the built-in values table gpuweb/gpuweb#4163, by @alan-baker.
-
[editorial] Fix GPUPipelineError constructor signature gpuweb/gpuweb#4160, by @kainino0x.
-
Really fix link for localizable text gpuweb/gpuweb#4157, by @tidoust.
-
Update browser template gpuweb/gpuweb#4156, by @greggman.
-
-
24 pull requests merged.
-
fix(wgsl): fix invalid var decl example gpuweb/gpuweb#4208.
-
Tentatively disable mermaid check gpuweb/gpuweb#4205.
-
Replace explicit invocations of sh with bash gpuweb/gpuweb#4203.
-
Change all shell scripts to use bash and `set -eo pipefail` gpuweb/gpuweb#4202.
-
Update bikeshed to 3.13.1 gpuweb/gpuweb#4200.
-
Remove samples gpuweb/gpuweb#4191.
-
Allow SharedArrayBuffer in writeBuffer/writeTexture with AllowSharedBufferSource gpuweb/gpuweb#4188.
-
Allow reading storage buffers in vertex shaders (#4132). gpuweb/gpuweb#4133.
-
Make GPUObjectDescriptorBase.label default to the empty string. gpuweb/gpuweb#4179.
-
Add VideoFrame to GPUImageCopyExternalImage source gpuweb/gpuweb#4178.
-
Remove expired attribute from JS example gpuweb/gpuweb#4181.
-
[editorial] Rework the built-in values table gpuweb/gpuweb#4163.
-
[editorial] Overview edits gpuweb/gpuweb#4164.
-
[editorial] Make Timelines normative gpuweb/gpuweb#4176.
-
[editorial] GPUPipelineError: set slots instead of setting getters gpuweb/gpuweb#4173.
-
Clarify language on storage texture capabilities gpuweb/gpuweb#4075.
-
[editorial] Fix typo in uniformity table gpuweb/gpuweb#4174.
-
Update workload simulator for recent spec changes gpuweb/gpuweb#4093.
-
Update WGSL reference for WebGPU gpuweb/gpuweb#4171.
-
s/importTexture/importExternalTexture/ gpuweb/gpuweb#4170.
-
Update browser template gpuweb/gpuweb#4156.
-
Really fix link for localizable text gpuweb/gpuweb#4157.
-
[editorial] Fix GPUPipelineError constructor signature gpuweb/gpuweb#4160.
-
-
3 pull requests closed.
-
Try generating Mermaid diagrams to match CI gpuweb/gpuweb#4204.
-
Add samples gpuweb/gpuweb#3573.
-
Accept SharedArrayBuffer in writeBuffer/writeTexture gpuweb/gpuweb#4187.
-
-
1 pull request reopened.
-
Guarantee that OOB writes to storage textures will be ignored gpuweb/gpuweb#4194, by @teoxoy.
-
-
19 pull requests received 47 new comments.
-
Guarantee that OOB writes to storage textures will be ignored gpuweb/gpuweb#4194 (7 comments by @alan-baker, @ben-clayton, @jimblandy, @teoxoy)
-
WGSL: Add dot4U8Packed and dot4I8Packed as new built-in functions gpuweb/gpuweb#2738 (6 comments by @alan-baker, @Jiawei-Shao, @kdashg)
-
Allow reading storage buffers in vertex shaders (#4132). gpuweb/gpuweb#4133 (5 comments by @alan-baker, @dneto0, @w3cbot)
-
Allow SharedArrayBuffer in writeBuffer/writeTexture with AllowSharedBufferSource gpuweb/gpuweb#4188 (4 comments by @kainino0x)
-
Update workload simulator for recent spec changes gpuweb/gpuweb#4093 (3 comments by @kainino0x, @w3cbot)
-
Clarify rounding of an overflow intermediate value gpuweb/gpuweb#4184 (2 comments by @dneto0, @zoddicus)
-
Fix Mermaid Diagram Issue and Upgrade GitHub Actions Workflows for Enhanced CI System gpuweb/gpuweb#4201 (1 comment by @mehmetoguzderin)
-
gpuweb/cts
-
14 issues opened.
-
Relaxed precision for copy tests? gpuweb/cts#2793, by @austinEng.
-
Update fp_primer.md for f16, AbstractFloat, and changes to the spec gpuweb/cts#2790, by @zoddicus.
-
wgsl: shader validation tests for uniformity analysis gpuweb/cts#2788, by @alan-baker.
-
wgsl: test memory model for f16 types gpuweb/cts#2779, by @dneto0.
-
wgsl: `unpack4x8unorm` is failing on some platforms gpuweb/cts#2777, by @zoddicus.
-
wgsl: Update correctlyRounded to have 3 regions of calculations gpuweb/cts#2776, by @zoddicus.
-
wgsl: Remove assert from ULP calculations gpuweb/cts#2775, by @zoddicus.
-
Migrate regression test from Chromium gpuweb/cts#2768, by @kainino0x.
-
Test that result of `fract()` is between 0 to 1 gpuweb/cts#2766, by @ben-clayton.
-
wgsl: execution tests for assigning `AbstractFloat` to `AbstractFloat` gpuweb/cts#2755, by @zoddicus.
-
wgsl: expand coverage in const_assert.spec.ts gpuweb/cts#2750, by @zoddicus.
-
Chunk tests for WPT using actual test time measurements gpuweb/cts#2749, by @kainino0x.
-
wpt: listing.js is not available in the first run gpuweb/cts#2747, by @sagudev.
-
-
9 issues closed.
-
wgsl: Remove assert from ULP calculations gpuweb/cts#2775.
-
Change `unpack2x16snorm` acceptance intervals to 3 ULP gpuweb/cts#2451.
-
creation-time bitcast to a NaN results in a shader-creation error gpuweb/cts#1859.
-
Test that result of `fract()` is between 0 to 1 gpuweb/cts#2766.
-
wgsl: f16 execution test for `round` builtin gpuweb/cts#2510.
-
wgsl: expand coverage in const_assert.spec.ts gpuweb/cts#2750.
-
wpt: listing.js is not available in the first run gpuweb/cts#2747.
-
8 issues received 10 new comments.
-
wgsl: Update correctlyRounded to have 3 regions of calculations gpuweb/cts#2776 (1 comment by @zoddicus)
-
creation-time bitcast to a NaN results in a shader-creation error gpuweb/cts#1859 (1 comment by @dneto0)
-
34 pull requests opened.
-
Update FP documentation gpuweb/cts#2794, by @zoddicus.
-
Compat skip sample_mask gpuweb/cts#2792, by @greggman.
-
Compat skip texture view format gpuweb/cts#2791, by @greggman.
-
Refactor FP unit helpers gpuweb/cts#2789, by @zoddicus.
-
Compat: Pass in cts options to workers gpuweb/cts#2787, by @greggman.
-
Compat: Skip testing bgra8unorm-srgb in more places gpuweb/cts#2786, by @greggman.
-
Fix isCompressedTextureFormat gpuweb/cts#2785, by @greggman.
-
Disallow % in queries gpuweb/cts#2784, by @greggman.
-
Compat: skip copy t2b compressed gpuweb/cts#2783, by @greggman.
-
Compat: Skip Cube-Array gpuweb/cts#2782, by @greggman.
-
Compat: Skip bgra8uorm-srgb in test gpuweb/cts#2781, by @greggman.
-
Add support for overflow domains to correctly rounded gpuweb/cts#2780, by @zoddicus.
-
Add max_ulp entries to constants gpuweb/cts#2778, by @zoddicus.
-
Validation tests for pointer uniformity gpuweb/cts#2774, by @alan-baker.
-
Test async pipeline creation before/after device destroy gpuweb/cts#2773, by @austinEng.
-
Expression uniformity validation tests gpuweb/cts#2772, by @alan-baker.
-
Maintain test timing metadata in-repo and auto-chunk WPT gpuweb/cts#2770, by @kainino0x.
-
Fix throws/rejections of non-Error types, and add lint gpuweb/cts#2769, by @kainino0x.
-
Adding 0x80000000/0x8000 cases to `ceil`, `fract`, `floor` & `round` gpuweb/cts#2767, by @zoddicus.
-
Update Expiration cts for GPUExternalTexture imported from VideoFrame gpuweb/cts#2765, by @shaoboyan.
-
Update acceptance intervals for unpack* to 'correctly rounded' gpuweb/cts#2764, by @zoddicus.
-
Update `unpack2x16snorm` acceptance intervals gpuweb/cts#2763, by @zoddicus.
-
Add F16 builtin execution tests: abs, ceil, floor, round gpuweb/cts#2762, by @jiangzhaoming.
-
Uniformity validation tests for function variable values gpuweb/cts#2761, by @alan-baker.
-
Add VideoFrame tests for copyExternalImageToTexture gpuweb/cts#2760, by @beaufortfrancois.
-
Disable test if compat gpuweb/cts#2759, by @greggman.
-
Compat test suite gpuweb/cts#2758, by @greggman.
-
Add compatibility option gpuweb/cts#2757, by @greggman.
-
wgsl: Implement AF -> AF assignment tests gpuweb/cts#2756, by @zoddicus.
-
validation for bad bitcast cases gpuweb/cts#2754, by @dneto0.
-
Add coverage for logical ors and ands in const_assert gpuweb/cts#2753, by @zoddicus.
-
Add negative zeroes to kValue gpuweb/cts#2752, by @dneto0.
-
test bitcast {i32,u32} to f32 gpuweb/cts#2751, by @dneto0.
-
Run `generate-listings` before `build-wpt` gpuweb/cts#2748, by @sagudev.
-
-
31 pull requests merged.
-
Refactor FP unit helpers gpuweb/cts#2789.
-
Compat skip sample_mask gpuweb/cts#2792.
-
Compat skip texture view format gpuweb/cts#2791.
-
Compat: Pass in cts options to workers gpuweb/cts#2787.
-
Validation tests for pointer uniformity gpuweb/cts#2774.
-
Compat: Skip testing bgra8unorm-srgb in more places gpuweb/cts#2786.
-
Expression uniformity validation tests gpuweb/cts#2772.
-
Fix isCompressedTextureFormat gpuweb/cts#2785.
-
Disallow % in queries gpuweb/cts#2784.
-
Compat: skip copy t2b compressed gpuweb/cts#2783.
-
Test async pipeline creation before/after device destroy gpuweb/cts#2773.
-
Compat: Skip bgra8uorm-srgb in test gpuweb/cts#2781.
-
Compat: Skip Cube-Array gpuweb/cts#2782.
-
Add max_ulp entries to constants gpuweb/cts#2778.
-
Update acceptance intervals for unpack* to 'correctly rounded' gpuweb/cts#2764.
-
Update `unpack2x16snorm` acceptance intervals gpuweb/cts#2763.
-
Uniformity validation tests for function variable values gpuweb/cts#2761.
-
Fix throws/rejections of non-Error types, and add lint gpuweb/cts#2769.
-
Disable test if compat gpuweb/cts#2759.
-
Compat test suite gpuweb/cts#2758.
-
validation for bad bitcast cases gpuweb/cts#2754.
-
Add compatibility option gpuweb/cts#2757.
-
Adding 0x80000000/0x8000 cases to `ceil`, `fract`, `floor` & `round` gpuweb/cts#2767.
-
Add VideoFrame tests for copyExternalImageToTexture gpuweb/cts#2760.
-
Add F16 builtin execution tests: abs, ceil, floor, round gpuweb/cts#2762.
-
Add negative zeroes to kValue gpuweb/cts#2752.
-
Add coverage for logical ors and ands in const_assert gpuweb/cts#2753.
-
test bitcast {i32,u32} to f32 gpuweb/cts#2751.
-
Run `generate-listings` before `build-wpt` gpuweb/cts#2748.
-
Add a test for transferring a huge canvas to an imagebitmap. gpuweb/cts#2746.
-
-
1 pull request closed.
-
flake repro: 208 subcases gpuweb/cts#1821.
-
-
21 pull requests received 90 new comments.
-
Uniformity validation tests for function variable values gpuweb/cts#2761 (16 comments by @alan-baker, @jrprice, @zoddicus)
-
Test async pipeline creation before/after device destroy gpuweb/cts#2773 (15 comments by @austinEng, @kainino0x, @lokokung)
-
Update Expiration cts for GPUExternalTexture imported from VideoFrame gpuweb/cts#2765 (10 comments by @kainino0x, @shaoboyan)
-
Generate many WPT html files for infra that doesn't support variants gpuweb/cts#1637 (7 comments by @kainino0x, @sagudev)
-
Validation tests for pointer uniformity gpuweb/cts#2774 (5 comments by @alan-baker, @jrprice, @zoddicus)
-
Adding 0x80000000/0x8000 cases to `ceil`, `fract`, `floor` & `round` gpuweb/cts#2767 (4 comments by @jiangzhaoming, @zoddicus)
-
Add coverage for logical ors and ands in const_assert gpuweb/cts#2753 (3 comments by @ben-clayton, @zoddicus)
-
Add F16 builtin execution tests: abs, ceil, floor, round gpuweb/cts#2762 (2 comments by @jiangzhaoming, @zoddicus)
-
Add VideoFrame tests for copyExternalImageToTexture gpuweb/cts#2760 (2 comments by @beaufortfrancois, @Kangz)
-
Update acceptance intervals for unpack* to 'correctly rounded' gpuweb/cts#2764 (1 comment by @zoddicus)
-
TR publications
-
12 documents published.
-
WebGPU Shading Language published 6 times as a Working Draft. History of WebGPU Shading Language
-
WebGPU published 6 times as a Working Draft. History of WebGPU
-
Participants
-
1 participant joined.
-
Thomas Meier (NVIDIA) joined.
-
-
1 organization joined.
-
NVIDIA joined.
-
GitHub
gpuweb/cts
-
172 issues opened.
-
wgsl: f16 validation tests for `tanh` gpuweb/cts#2743, by @zoddicus.
-
wgsl: f32 validation tests for `tanh` gpuweb/cts#2742, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `tanh` gpuweb/cts#2741, by @zoddicus.
-
wgsl: f16 validation tests for `tan` gpuweb/cts#2740, by @zoddicus.
-
wgsl: f32 validation tests for `tan` gpuweb/cts#2739, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `tan` gpuweb/cts#2738, by @zoddicus.
-
wgsl: f16 validation tests for `sinh` gpuweb/cts#2737, by @zoddicus.
-
wgsl: f32 validation tests for `sinh` gpuweb/cts#2736, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `sinh` gpuweb/cts#2735, by @zoddicus.
-
wgsl: f16 validation tests for `sin` gpuweb/cts#2734, by @zoddicus.
-
wgsl: f32 validation tests for `sin` gpuweb/cts#2733, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `sin` gpuweb/cts#2732, by @zoddicus.
-
wgsl: f16 validation tests for `reflect` gpuweb/cts#2731, by @zoddicus.
-
wgsl: f32 validation tests for `reflect` gpuweb/cts#2730, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `reflect` gpuweb/cts#2729, by @zoddicus.
-
wgsl: f32 validation tests for `pack2x16float` gpuweb/cts#2728, by @zoddicus.
-
wgsl: f32 validation tests for `quantizeToF16` gpuweb/cts#2727, by @zoddicus.
-
wgsl: f16 validation tests for `pow` gpuweb/cts#2726, by @zoddicus.
-
wgsl: f32 validation tests for `pow` gpuweb/cts#2725, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `pow` gpuweb/cts#2724, by @zoddicus.
-
wgsl: f16 validation tests for `normalize` gpuweb/cts#2723, by @zoddicus.
-
wgsl: f32 validation tests for `normalize` gpuweb/cts#2722, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `normalize` gpuweb/cts#2721, by @zoddicus.
-
wgsl: f16 validation tests for `mix` gpuweb/cts#2720, by @zoddicus.
-
wgsl: f32 validation tests for `mix` gpuweb/cts#2719, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `mix` gpuweb/cts#2718, by @zoddicus.
-
wgsl: f16 validation tests for `length` gpuweb/cts#2717, by @zoddicus.
-
wgsl: f32 validation tests for `length` gpuweb/cts#2716, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `length` gpuweb/cts#2715, by @zoddicus.
-
wgsl: f16 validation tests for `ldexp` gpuweb/cts#2714, by @zoddicus.
-
wgsl: f32 validation tests for `ldexp` gpuweb/cts#2713, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `ldexp` gpuweb/cts#2712, by @zoddicus.
-
wgsl: f16 validation tests for `frexp` gpuweb/cts#2711, by @zoddicus.
-
wgsl: f32 validation tests for `frexp` gpuweb/cts#2710, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `frexp` gpuweb/cts#2709, by @zoddicus.
-
wgsl: f16 validation tests for `fma` gpuweb/cts#2708, by @zoddicus.
-
wgsl: f32 validation tests for `fma` gpuweb/cts#2707, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `fma` gpuweb/cts#2706, by @zoddicus.
-
wgsl: f16 validation tests for `faceForward` gpuweb/cts#2705, by @zoddicus.
-
wgsl: f32 validation tests for `faceForward` gpuweb/cts#2704, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `faceForward` gpuweb/cts#2703, by @zoddicus.
-
wgsl: f16 validation tests for `exp2` gpuweb/cts#2702, by @zoddicus.
-
wgsl: f32 validation tests for `exp2` gpuweb/cts#2701, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `exp2` gpuweb/cts#2700, by @zoddicus.
-
wgsl: f16 validation tests for `exp` gpuweb/cts#2699, by @zoddicus.
-
wgsl: f32 validation tests for `exp` gpuweb/cts#2698, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `exp` gpuweb/cts#2697, by @zoddicus.
-
wgsl: f16 validation tests for `dot` gpuweb/cts#2696, by @zoddicus.
-
wgsl: f32 validation tests for `dot` gpuweb/cts#2695, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `dot` gpuweb/cts#2694, by @zoddicus.
-
wgsl: f16 validation tests for `distance` gpuweb/cts#2693, by @zoddicus.
-
wgsl: f32 validation tests for `distance` gpuweb/cts#2692, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `distance` gpuweb/cts#2691, by @zoddicus.
-
wgsl: f16 validation tests for `degrees` gpuweb/cts#2690, by @zoddicus.
-
wgsl: f32 validation tests for `degrees` gpuweb/cts#2689, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `degrees` gpuweb/cts#2688, by @zoddicus.
-
wgsl: f16 validation tests for `cross` gpuweb/cts#2687, by @zoddicus.
-
wgsl: f32 validation tests for `cross` gpuweb/cts#2686, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `cross` gpuweb/cts#2685, by @zoddicus.
-
wgsl: f16 validation tests for `cosh` gpuweb/cts#2684, by @zoddicus.
-
wgsl: f32 validation tests for `cosh` gpuweb/cts#2683, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `cosh` gpuweb/cts#2682, by @zoddicus.
-
wgsl: f16 validation tests for `cos` gpuweb/cts#2681, by @zoddicus.
-
wgsl: f32 validation tests for `cos` gpuweb/cts#2680, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `cos` gpuweb/cts#2679, by @zoddicus.
-
wgsl: f16 validation tests for `clamp` gpuweb/cts#2678, by @zoddicus.
-
wgsl: f32 validation tests for `clamp` gpuweb/cts#2677, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `clamp` gpuweb/cts#2676, by @zoddicus.
-
wgsl: f16 validation tests for `atan2` gpuweb/cts#2675, by @zoddicus.
-
wgsl: f32 validation tests for `atan2` gpuweb/cts#2674, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `atan2` gpuweb/cts#2673, by @zoddicus.
-
wgsl: f16 validation tests for `atanh` gpuweb/cts#2672, by @zoddicus.
-
wgsl: f32 validation tests for `atanh` gpuweb/cts#2671, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `atanh` gpuweb/cts#2670, by @zoddicus.
-
F16Interval is not serializable, breaking param cache gpuweb/cts#2669, by @austinEng.
-
wgsl: f16 validation tests for `asin` gpuweb/cts#2666, by @zoddicus.
-
wgsl: f32 validation tests for `asin` gpuweb/cts#2665, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `asin` gpuweb/cts#2664, by @zoddicus.
-
wgsl: f16 validation tests for `acosh` gpuweb/cts#2663, by @zoddicus.
-
wgsl: f32 validation tests for `acosh` gpuweb/cts#2662, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `acosh` gpuweb/cts#2661, by @zoddicus.
-
wgsl: f16 validation tests for `acos` gpuweb/cts#2660, by @zoddicus.
-
wgsl: f32 validation tests for `acos` gpuweb/cts#2659, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `acos` gpuweb/cts#2658, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `log2` gpuweb/cts#2657, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `log` gpuweb/cts#2656, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `sqrt` gpuweb/cts#2655, by @zoddicus.
-
wgsl: AbstractFloat validation tests for `inverseSqrt` gpuweb/cts#2654, by @zoddicus.
-
wgsl: AbstractFloat validation tests for remainder (`%`) gpuweb/cts#2653, by @zoddicus.
-
wgsl: AbstractFloat validation tests for division (`/`) gpuweb/cts#2652, by @zoddicus.
-
wgsl: AbstractFloat validation tests for multiplication (`*`) gpuweb/cts#2651, by @zoddicus.
-
wgsl: AbstractFloat validation tests for addition (`+`) gpuweb/cts#2650, by @zoddicus.
-
wgsl: AbstractFloat validation tests for subtraction (`-`) gpuweb/cts#2649, by @zoddicus.
-
wgsl: f32 validation tests for `log2` gpuweb/cts#2648, by @zoddicus.
-
wgsl: f16 validation tests for `log2` gpuweb/cts#2647, by @zoddicus.
-
wgsl: f16 validation tests for `log` gpuweb/cts#2646, by @zoddicus.
-
wgsl: f32 validation tests for `log` gpuweb/cts#2645, by @zoddicus.
-
wgsl: f16 validation tests for `sqrt` gpuweb/cts#2644, by @zoddicus.
-
wgsl: f32 validation tests for `sqrt` gpuweb/cts#2643, by @zoddicus.
-
wgsl: f16 validation tests for `inverseSqrt` gpuweb/cts#2642, by @zoddicus.
-
wgsl: f32 validation tests for `inverseSqrt` gpuweb/cts#2641, by @zoddicus.
-
wgsl: f16 validation tests for remainder (`%`) gpuweb/cts#2640, by @zoddicus.
-
wgsl: f32 validation tests for remainder (`%`) gpuweb/cts#2639, by @zoddicus.
-
wgsl: f16 validation tests for division (`/`) gpuweb/cts#2638, by @zoddicus.
-
wgsl: f32 validation tests for division (`/`) gpuweb/cts#2637, by @zoddicus.
-
wgsl: f32 validation tests for multiplication (`*`) gpuweb/cts#2636, by @zoddicus.
-
wgsl: f16 validation tests for multiplication (`*`) gpuweb/cts#2635, by @zoddicus.
-
wgsl: f16 validation tests for subtraction (`-`) gpuweb/cts#2634, by @zoddicus.
-
wgsl: f32 validation tests for subtraction (`-`) gpuweb/cts#2633, by @zoddicus.
-
wgsl: f16 validation tests for addition `(+)` gpuweb/cts#2632, by @zoddicus.
-
wgsl: f32 validation tests for addition (`+`) gpuweb/cts#2631, by @zoddicus.
-
wpt lint errors gpuweb/cts#2628, by @sagudev.
-
wgsl: Reduce number of test cases for integer arithmetic operations gpuweb/cts#2623, by @zoddicus.
-
Invalid variants if arguments prefixes are not passed to gen_wpt_cts_html gpuweb/cts#2622, by @sagudev.
-
wgsl: execution tests for assigning `AbstractFloat` to `f16` gpuweb/cts#2618, by @zoddicus.
-
wgsl: execution tests for assigning `AbstractFloat` to `f32` gpuweb/cts#2617, by @zoddicus.
-
wgsl: Reduce number of test cases for f32 arithmetic operations gpuweb/cts#2613, by @zoddicus.
-
cts not working at gpuweb.github.io/cts/ gpuweb/cts#2612, by @greggman.
-
Handle uncaught internal errors in device_pool gpuweb/cts#2610, by @kainino0x.
-
It is too easy to introduce non-serializable Comparators gpuweb/cts#2593, by @zoddicus.
-
web_platform: tighten expectation for zero-sized transferToImageBitmap gpuweb/cts#2589, by @kainino0x.
-
wgsl: AbstractFloat execution tests for `frexp` gpuweb/cts#2588, by @zoddicus.
-
wgsl: f16 execution tests for `frexp` gpuweb/cts#2587, by @zoddicus.
-
wgsl: f16 execution tests for `remainder (%)` gpuweb/cts#2586, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `remainder (%)` gpuweb/cts#2585, by @zoddicus.
-
wgsl: AbstractInt execution tests for `sign` builtin gpuweb/cts#2584, by @zoddicus.
-
wgsl: f16 execution tests for `sign` builtin gpuweb/cts#2583, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `sign` builtin gpuweb/cts#2582, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `refract` builtin gpuweb/cts#2581, by @zoddicus.
-
wgsl: f16 execution tests for `refract` builtin gpuweb/cts#2580, by @zoddicus.
-
wgsl: f16 execution tests for `reflect` builtin gpuweb/cts#2579, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `reflect` builtin gpuweb/cts#2578, by @zoddicus.
-
wgsl: f16 execution tests for `radians` builtin gpuweb/cts#2577, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `radians` builtin gpuweb/cts#2576, by @zoddicus.
-
val: Distinguish exception type in WebGPU context configure test gpuweb/cts#2574, by @takahirox.
-
wgsl: f16 execution tests for `pow` builtin gpuweb/cts#2573, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `pow` builtin gpuweb/cts#2572, by @zoddicus.
-
wgsl: f16 execution tests for `normalize` builtin gpuweb/cts#2571, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `normalize` builtin gpuweb/cts#2570, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `modf` gpuweb/cts#2569, by @zoddicus.
-
wgsl: f16 execution tests for `modf` gpuweb/cts#2568, by @zoddicus.
-
wgsl: f16 execution tests for `mix` builtin gpuweb/cts#2567, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `mix` builtin gpuweb/cts#2566, by @zoddicus.
-
wgsl: f16 execution tests for `degrees` builtin gpuweb/cts#2565, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `degrees` builtin gpuweb/cts#2564, by @zoddicus.
-
wgsl: f16 execution tests for `atanh` builtin gpuweb/cts#2563, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `atanh` builtin gpuweb/cts#2562, by @zoddicus.
-
wgsl: f16 execution tests for `asinh` builtin gpuweb/cts#2561, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `asinh` builtin gpuweb/cts#2560, by @zoddicus.
-
wgsl: f16 execution tests for `acosh` builtin gpuweb/cts#2559, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `acosh` builtin gpuweb/cts#2558, by @zoddicus.
-
wgsl: f16 execution tests for `distance` builtin gpuweb/cts#2557, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `distance` builtin gpuweb/cts#2556, by @zoddicus.
-
wgsl: f16 execution tests for `length` builtin gpuweb/cts#2555, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `length` builtin gpuweb/cts#2554, by @zoddicus.
-
wgsl: f16 execution tests for `asin` builtin gpuweb/cts#2553, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `asin` builtin gpuweb/cts#2552, by @zoddicus.
-
wgsl: f16 execution tests for `acos` builtin gpuweb/cts#2551, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `acos` builtin gpuweb/cts#2550, by @zoddicus.
-
wgsl: f16 execution tests for `fma` builtin gpuweb/cts#2549, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `fma` builtin gpuweb/cts#2548, by @zoddicus.
-
wgsl: f16 execution tests for `faceForward` builtin gpuweb/cts#2547, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `faceForward` builtin gpuweb/cts#2546, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `exp2` builtin gpuweb/cts#2545, by @zoddicus.
-
wgsl: f16 execution tests for `exp2` builtin gpuweb/cts#2544, by @zoddicus.
-
wgsl: f16 execution tests for `exp` builtin gpuweb/cts#2543, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `exp` builtin gpuweb/cts#2542, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `cross` gpuweb/cts#2541, by @zoddicus.
-
wgsl: f16 execution tests for `cross` gpuweb/cts#2540, by @zoddicus.
-
wgsl: AbstractFloat execution tests for `cosh` gpuweb/cts#2539, by @zoddicus.
-
wgsl: f16 execution tests for `cosh` gpuweb/cts#2538, by @zoddicus.
-
-
43 issues closed.
-
wgsl: Reduce number of test cases for integer arithmetic operations gpuweb/cts#2623.
-
F16Interval is not serializable, breaking param cache gpuweb/cts#2669.
-
`npm run wpt` does not populate `out-wpt/common` fully gpuweb/cts#1092.
-
wgsl: execution tests for assigning `AbstractFloat` to `f16` gpuweb/cts#2618.
-
wgsl: execution tests for assigning `AbstractFloat` to `f32` gpuweb/cts#2617.
-
cts not working at gpuweb.github.io/cts/ gpuweb/cts#2612.
-
wgsl: Reduce number of test cases for f32 arithmetic operations gpuweb/cts#2613.
-
Ability to switch input ranges when running in const vs non-const gpuweb/cts#1979.
-
Handle uncaught internal errors in device_pool gpuweb/cts#2610.
-
It is too easy to introduce non-serializable Comparators gpuweb/cts#2593.
-
val: Distinguish exception type in WebGPU context configure test gpuweb/cts#2574.
-
wgsl: execution test for frexp for vector f16 gpuweb/cts#1211.
-
wgsl: execution test for frexp for scalar f16 gpuweb/cts#1209.
-
wgsl: test abstract versions of modf and frexp gpuweb/cts#2012.
-
wgsl: execution test for modf for vector f16 gpuweb/cts#1229.
-
wgsl: execution test for modf for scalar f16 gpuweb/cts#1227.
-
wgsl: f32 execution tests for `remainder (%)` gpuweb/cts#1763.
-
wgsl: f32 execution tests for `sign` builtin gpuweb/cts#1236.
-
wgsl: f32 execution tests for `refract` builtin gpuweb/cts#1234.
-
wgsl: f32 execution tests for `reflect` builtin gpuweb/cts#1233.
-
wgsl: f32 execution tests for `radians` builtin gpuweb/cts#1232.
-
val: Distinguish exception type of getCurrentTexture test gpuweb/cts#2518.
-
Clarify the meaning of the castable constants in constants.ts gpuweb/cts#2501.
-
wgsl: f32 execution tests for `pow` builtin gpuweb/cts#1231.
-
wgsl: f32 execution tests for `normalize` builtin gpuweb/cts#1230.
-
wgsl: f32 execution tests for modf gpuweb/cts#1226.
-
wgsl: f32 execution tests for `mix` builtin gpuweb/cts#1225.
-
wgsl: f32 execution tests for `degrees` builtin gpuweb/cts#1224.
-
wgsl: f32 execution tests for `atanh` builtin gpuweb/cts#1223.
-
wgsl: f32 execution tests for `asinh` builtin gpuweb/cts#1222.
-
wgsl: f32 execution tests for `acosh` builtin gpuweb/cts#1221.
-
wgsl: f32 execution tests for `distance` builtin gpuweb/cts#1217.
-
wgsl: f32 execution tests for `length` builtin gpuweb/cts#1216.
-
wgsl: f32 execution tests for `asin` builtin gpuweb/cts#1213.
-
wgsl: f32 execution tests for `acos` builtin gpuweb/cts#1212.
-
wgsl: f32 execution tests for `fma` builtin gpuweb/cts#1207.
-
wgsl: f32 execution tests for `faceForward` builtin gpuweb/cts#1206.
-
wgsl: f32 execution tests for `exp2` builtin gpuweb/cts#1205.
-
wgsl: f32 execution tests for `exp` builtin gpuweb/cts#1204.
-
wgsl: f32 execution tests for `cross` gpuweb/cts#1203.
-
wgsl: f32 execution tests for `cosh` gpuweb/cts#1202.
-
-
36 issues received 39 new comments.
-
`npm run wpt` does not populate `out-wpt/common` fully gpuweb/cts#1092 (2 comments by @kainino0x, @sagudev)
-
Ability to switch input ranges when running in const vs non-const gpuweb/cts#1979 (1 comment by @zoddicus)
-
val: Distinguish exception type in WebGPU context configure test gpuweb/cts#2574 (1 comment by @takahirox)
-
37 pull requests opened.
-
Add a test for transferring a huge canvas to an imagebitmap. gpuweb/cts#2746, by @Kangz.
-
Add WPT lint test workflow gpuweb/cts#2745, by @sagudev.
-
Remove console.log from a file exported to wpt gpuweb/cts#2744, by @kainino0x.
-
Include testharness.js in exposed interface tests gpuweb/cts#2668, by @kainino0x.
-
Add eslint custom rules for tabs and trailing spaces gpuweb/cts#2667, by @kainino0x.
-
Convert integer arithmetic tests to use sparse inputs gpuweb/cts#2630, by @zoddicus.
-
[WGSL] Fulfill F16 traits (step 1) and add fundamental interval unit tests gpuweb/cts#2629, by @jiangzhaoming.
-
Set `?q=` as default args prefix gpuweb/cts#2627, by @sagudev.
-
Add missing toStorage() when assigning values to outputs gpuweb/cts#2626, by @Jiawei-Shao.
-
Check for a device lost reason of 'unknown' gpuweb/cts#2625, by @toji.
-
Adds validation tests for float32-filterable gpuweb/cts#2624, by @lokokung.
-
Add missing files to out-wpt/ build gpuweb/cts#2621, by @kainino0x.
-
Add execution tests for AbstractFloat -> f16 assignment gpuweb/cts#2620, by @zoddicus.
-
Add execution tests for AbstractFloat -> f32 assignment gpuweb/cts#2619, by @zoddicus.
-
Fix floating_point AbsoluteErrorInterval unittests gpuweb/cts#2616, by @jiangzhaoming.
-
Fix gruntfile generate-listings ordering gpuweb/cts#2615, by @kainino0x.
-
Switch to using sparseF32Range in f32 arithmetic tests gpuweb/cts#2614, by @zoddicus.
-
Handle uncaught internal errors in device_pool gpuweb/cts#2611, by @austinEng.
-
expression.ts: rate-limit the maximum number of batches in flight gpuweb/cts#2609, by @austinEng.
-
Factor out and dedup 'validate' step of presubmit gpuweb/cts#2608, by @kainino0x.
-
Move texture format info to its own file gpuweb/cts#2607, by @kainino0x.
-
Split up the matrix f32 arithmetic tests gpuweb/cts#2606, by @zoddicus.
-
Split up the non-matrix f32 arithmetic tests gpuweb/cts#2605, by @zoddicus.
-
Rewrite bitcast comparators to be serializable gpuweb/cts#2603, by @zoddicus.
-
Refactor FP fundamental error intervals unittests with trait gpuweb/cts#2602, by @jiangzhaoming.
-
Make need for serialization of Comparators more obvious gpuweb/cts#2601, by @zoddicus.
-
Refactor FPInterval and utilities unittests with traits gpuweb/cts#2600, by @jiangzhaoming.
-
Fix large limit value rounding error gpuweb/cts#2599, by @jiangzhaoming.
-
Refactor texture format info table (part 1) gpuweb/cts#2598, by @kainino0x.
-
Add more uniformity validation tests gpuweb/cts#2597, by @alan-baker.
-
Updated uses of timestampWrites gpuweb/cts#2596, by @toji.
-
Adds max_dynamic_buffers_per_pipeline tests gpuweb/cts#2595, by @lokokung.
-
Remove using CaseCache in bitcast tests gpuweb/cts#2594, by @zoddicus.
-
WebGPU GPUBuffer.mapState MAP_READ mode gpuweb/cts#2592, by @takahirox.
-
Import spec files async to improve load times. gpuweb/cts#2591, by @kdashg.
-
Add TODO to last shouldThrow that doesn't expect a particular name gpuweb/cts#2590, by @kainino0x.
-
Check exception type for Webgpu context configure test gpuweb/cts#2575, by @takahirox.
-
-
38 pull requests merged.
-
Import spec files async to improve load times. gpuweb/cts#2591.
-
Add WPT lint test workflow gpuweb/cts#2745.
-
Remove console.log from a file exported to wpt gpuweb/cts#2744.
-
Convert integer arithmetic tests to use sparse inputs gpuweb/cts#2630.
-
Include testharness.js in exposed interface tests gpuweb/cts#2668.
-
Add eslint custom rules for tabs and trailing spaces gpuweb/cts#2667.
-
Set `?q=` as default args prefix gpuweb/cts#2627.
-
Add missing toStorage() when assigning values to outputs gpuweb/cts#2626.
-
Add missing files to out-wpt/ build gpuweb/cts#2621.
-
Check for a device lost reason of 'unknown' gpuweb/cts#2625.
-
Adds validation tests for float32-filterable gpuweb/cts#2624.
-
Add execution tests for AbstractFloat -> f16 assignment gpuweb/cts#2620.
-
Add execution tests for AbstractFloat -> f32 assignment gpuweb/cts#2619.
-
Fix floating_point AbsoluteErrorInterval unittests gpuweb/cts#2616.
-
Add more uniformity validation tests gpuweb/cts#2597.
-
Fix gruntfile generate-listings ordering gpuweb/cts#2615.
-
Switch to using sparseF32Range in f32 arithmetic tests gpuweb/cts#2614.
-
Factor out and dedup 'validate' step of presubmit gpuweb/cts#2608.
-
Split up the matrix f32 arithmetic tests gpuweb/cts#2606.
-
Handle uncaught internal errors in device_pool gpuweb/cts#2611.
-
Move texture format info to its own file gpuweb/cts#2607.
-
expression.ts: rate-limit the maximum number of batches in flight gpuweb/cts#2609.
-
Refactor texture format info table (part 1) gpuweb/cts#2598.
-
Split up the non-matrix f32 arithmetic tests gpuweb/cts#2605.
-
Refactor FP fundamental error intervals unittests with trait gpuweb/cts#2602.
-
Rewrite bitcast comparators to be serializable gpuweb/cts#2603.
-
Make need for serialization of Comparators more obvious gpuweb/cts#2601.
-
Refactor FPInterval and utilities unittests with traits gpuweb/cts#2600.
-
WebGPU GPUBuffer.mapState MAP_READ mode gpuweb/cts#2592.
-
Fix large limit value rounding error gpuweb/cts#2599.
-
Updated uses of timestampWrites gpuweb/cts#2596.
-
Remove using CaseCache in bitcast tests gpuweb/cts#2594.
-
Add TODO to last shouldThrow that doesn't expect a particular name gpuweb/cts#2590.
-
Check exception type for Webgpu context configure test gpuweb/cts#2575.
-
Check exception type for Webgpu getCurrentTexture test gpuweb/cts#2537.
-
Clarify that f64 castable constants are about WebIDL casting gpuweb/cts#2505.
-
test bitcast f32 to {f32,u32,i32} gpuweb/cts#2508.
-
-
1 pull request closed.
-
shader/execution: Implement quantizeToF16() tests gpuweb/cts#796.
-
-
23 pull requests received 97 new comments.
-
Add execution tests for AbstractFloat -> f32 assignment gpuweb/cts#2619 (15 comments by @ben-clayton, @kainino0x, @zoddicus)
-
[WGSL] Fulfill F16 traits (step 1) and add fundamental interval unit tests gpuweb/cts#2629 (8 comments by @jiangzhaoming, @zoddicus)
-
Add missing files to out-wpt/ build gpuweb/cts#2621 (6 comments by @ErichDonGubler, @kainino0x, @sagudev)
-
Remove console.log from a file exported to wpt gpuweb/cts#2744 (5 comments by @greggman, @kainino0x)
-
Include testharness.js in exposed interface tests gpuweb/cts#2668 (4 comments by @kainino0x, @sagudev)
-
Add execution tests for AbstractFloat -> f16 assignment gpuweb/cts#2620 (4 comments by @ben-clayton, @zoddicus)
-
expression.ts: rate-limit the maximum number of batches in flight gpuweb/cts#2609 (3 comments by @austinEng, @kainino0x)
-
Refactor FPInterval and utilities unittests with traits gpuweb/cts#2600 (3 comments by @jiangzhaoming, @Jiawei-Shao)
-
Convert integer arithmetic tests to use sparse inputs gpuweb/cts#2630 (2 comments by @amaiorano, @zoddicus)
-
Rewrite bitcast comparators to be serializable gpuweb/cts#2603 (2 comments by @ben-clayton, @zoddicus)
-
Make need for serialization of Comparators more obvious gpuweb/cts#2601 (2 comments by @ben-clayton, @zoddicus)
-
Clarify that f64 castable constants are about WebIDL casting gpuweb/cts#2505 (1 comment by @zoddicus)
-
gpuweb/gpuweb
-
46 issues opened.
-
[editorial] Revisit the overview gpuweb/gpuweb#4153, by @alan-baker.
-
[editorial] Revisit the built-in values table gpuweb/gpuweb#4151, by @alan-baker.
-
Should navigator.gpu not be readonly? gpuweb/gpuweb#4150, by @juj.
-
<deleted> gpuweb/gpuweb#4149, by @wrightwriter.
-
[editorial] WGSL Section Shuffle gpuweb/gpuweb#4145, by @alan-baker.
-
fract() returns values out of bounds on Intel Iris Graphics 550 gpuweb/gpuweb#4144, by @anvaka.
-
Add discussion about fixed size array limits gpuweb/gpuweb#4143, by @rconde01.
-
Why GPUCommandBuffer and GPUCommandEncoder cannot be reused? gpuweb/gpuweb#4138, by @microaaron.
-
Support for explicit matmul instructions / tensor core instructions gpuweb/gpuweb#4137, by @thomasdullien.
-
Accuracy for AbstractFloat builtins not specified gpuweb/gpuweb#4134, by @zoddicus.
-
read-only storage buffer access from vertex shaders should be allowed gpuweb/gpuweb#4132, by @mheistermann.
-
Can I write to a buffer in parallel using some set of compute pipelines? gpuweb/gpuweb#4130, by @dalduba.
-
what happens if vertex w position is zero, or negative? gpuweb/gpuweb#4129, by @dneto0.
-
define "clip space coordinates" gpuweb/gpuweb#4127, by @dneto0.
-
considerations for general implicit promotion of scalars to vectors gpuweb/gpuweb#4126, by @dneto0.
-
Copying depth32float to depth32float gpuweb/gpuweb#4125, by @kainino0x.
-
How Struct are handled when the shader is compiled ? gpuweb/gpuweb#4123, by @tlecoz.
-
[webgpu] Web IDL validation errors gpuweb/gpuweb#4119, by @autokagami.
-
Should we be able to unset index buffers? gpuweb/gpuweb#4118, by @Kangz.
-
No way to destroy a shader? gpuweb/gpuweb#4116, by @mrakh.
-
Feature Request: Support for Matrix Inverse Function gpuweb/gpuweb#4115, by @scratchapixel.
-
Add arrow operator as sugar for (*var). gpuweb/gpuweb#4114, by @wrightwriter.
-
Allow mutable function parameters with the mut keyword. gpuweb/gpuweb#4113, by @wrightwriter.
-
Allow some vector-float overloads for min()/max()/clamp()/step(),smoothstep() gpuweb/gpuweb#4111, by @wrightwriter.
-
Should there be a way to prevent loop unrolling? gpuweb/gpuweb#4110, by @myth0genesis.
-
View instancing gpuweb/gpuweb#4109, by @MalekiRe.
-
Space efficient wide color support gpuweb/gpuweb#4108, by @litherum.
-
WGSL u32 CRITICAL Browser / GPU FREEZE on Windows (works on Mac) gpuweb/gpuweb#4107, by @alienself.
-
Unresponsive behavior using WebGpu simple example and device warning gpuweb/gpuweb#4105, by @sameer-dudeja.
-
Add PI and TAU builtins gpuweb/gpuweb#4100, by @JMS55.
-
Range-based for loops gpuweb/gpuweb#4099, by @wrightwriter.
-
Allow modf() for scalar-vector and vector-scalar gpuweb/gpuweb#4097, by @wrightwriter.
-
Updating gpuweb/gpuweb#4096, by @lonewolf372.
-
Feature request: a way of loading row matrices gpuweb/gpuweb#4094, by @magcius.
-
Missing intrinsic functions like rcp and rsqrt that are useful in other shading languages gpuweb/gpuweb#4092, by @coramat.
-
Add global_invocation_index and workgroup_index gpuweb/gpuweb#4087, by @JMS55.
-
Simultaneous use of index_buffer in compute and render stages gpuweb/gpuweb#4085, by @afreg11.
-
black mode in addition to clamp-to-edge, etc. for sampling textures gpuweb/gpuweb#4084, by @jeancolasp.
-
Handling of omitted labels is unclear gpuweb/gpuweb#4083, by @jimblandy.
-
Readonly attributes cannot use `[EnforceRange]` gpuweb/gpuweb#4080, by @saschanaz.
-
[webgpu] Broken references in WebGPU gpuweb/gpuweb#4079, by @dontcallmedom-bot.
-
Missing validation of fragment output type gpuweb/gpuweb#4078, by @kainino0x.
-
Add note to WebGPU spec explaining what limits the number of WGSL @locations. gpuweb/gpuweb#4077, by @toji.
-
"Variable and Value Declaration Feature Summary" table missing `var<storage,write>` gpuweb/gpuweb#4076, by @rconde01.
-
-
26 issues closed.
-
[editorial] WGSL Section Shuffle gpuweb/gpuweb#4145.
-
Add discussion about fixed size array limits gpuweb/gpuweb#4143.
-
<deleted> gpuweb/gpuweb#4149.
-
use wgsl-native syntax highlighting gpuweb/gpuweb#4004.
-
define "clip space coordinates" gpuweb/gpuweb#4127.
-
Timestamp query resets gpuweb/gpuweb#4069.
-
Can I write to a buffer in parallel using some set of compute pipelines? gpuweb/gpuweb#4130.
-
[wgsl] `workgroup_size` does not list negative values as an error gpuweb/gpuweb#4071.
-
How Struct are handled when the shader is compiled ? gpuweb/gpuweb#4123.
-
Should we be able to unset index buffers? gpuweb/gpuweb#4118.
-
[webgpu] Web IDL validation errors gpuweb/gpuweb#4119.
-
Use brackets for array types gpuweb/gpuweb#854.
-
Extern declarations gpuweb/gpuweb#883.
-
WGSL u32 CRITICAL Browser / GPU FREEZE on Windows (works on Mac) gpuweb/gpuweb#4107.
-
Unresponsive behavior using WebGpu simple example and device warning gpuweb/gpuweb#4105.
-
Readonly attributes cannot use `[EnforceRange]` gpuweb/gpuweb#4080.
-
Missing validation of fragment output type gpuweb/gpuweb#4078.
-
Updating gpuweb/gpuweb#4096.
-
[webgpu] Broken references in WebGPU gpuweb/gpuweb#4079.
-
black mode in addition to clamp-to-edge, etc. for sampling textures gpuweb/gpuweb#4084.
-
Simultaneous use of index_buffer in compute and render stages gpuweb/gpuweb#4085.
-
Improve syntax authoring experience gpuweb/gpuweb#3708.
-
Pipe WGSL compilation failures in createFooPipeline. gpuweb/gpuweb#3682.
-
-
3 issues reopened.
-
<deleted> gpuweb/gpuweb#4149, by @wrightwriter.
-
Feature request: a way of loading row matrices gpuweb/gpuweb#4094, by @magcius.
-
Space efficient wide color support gpuweb/gpuweb#4108, by @kainino0x.
-
-
55 issues received 196 new comments.
-
Add arrow operator as sugar for (*var). gpuweb/gpuweb#4114 (17 comments by @alan-baker, @ben-clayton, @dneto0, @kainino0x, @litherum, @wrightwriter)
-
Allow mutable function parameters with the mut keyword. gpuweb/gpuweb#4113 (9 comments by @ben-clayton, @jimblandy, @litherum, @wrightwriter)
-
Consider matching either ECMAScript or HTML for whitespace and line break definitions gpuweb/gpuweb#4068 (8 comments by @annevk, @dneto0, @kdashg)
-
Allow some vector-float overloads for min()/max()/clamp()/step(),smoothstep() gpuweb/gpuweb#4111 (8 comments by @ben-clayton, @kdashg, @litherum, @magcius, @wrightwriter)
-
Should there be a way to prevent loop unrolling? gpuweb/gpuweb#4110 (8 comments by @ben-clayton, @kdashg, @litherum, @myth0genesis)
-
Feature Request: Some way to generate identity matrices gpuweb/gpuweb#4026 (8 comments by @ben-clayton, @greggman, @jeancolasp, @kdashg, @litherum)
-
Resurrect clamp-to-border gpuweb/gpuweb#1305 (8 comments by @greggman, @jeancolasp, @kainino0x, @magcius)
-
[editorial] WGSL Section Shuffle gpuweb/gpuweb#4145 (7 comments by @alan-baker, @dneto0, @mehmetoguzderin)
-
fract() returns values out of bounds on Intel Iris Graphics 550 gpuweb/gpuweb#4144 (6 comments by @anvaka, @ben-clayton, @dneto0, @greggman)
-
Support arrayLength for fixed-size arrays gpuweb/gpuweb#2074 (6 comments by @ben-clayton, @dj2, @greggman)
-
Feature request: a way of loading row matrices gpuweb/gpuweb#4094 (5 comments by @dneto0, @kainino0x, @kdashg, @magcius)
-
Should we be able to unset index buffers? gpuweb/gpuweb#4118 (5 comments by @greggman, @kainino0x, @Kangz)
-
Add discussion about fixed size array limits gpuweb/gpuweb#4143 (4 comments by @alan-baker, @kdashg, @rconde01)
-
Accuracy for AbstractFloat builtins not specified gpuweb/gpuweb#4134 (4 comments by @alan-baker, @kdashg, @zoddicus)
-
Why GPUCommandBuffer and GPUCommandEncoder cannot be reused? gpuweb/gpuweb#4138 (4 comments by @kainino0x, @Kangz, @magcius, @microaaron)
-
Missing intrinsic functions like rcp and rsqrt that are useful in other shading languages gpuweb/gpuweb#4092 (4 comments by @alan-baker, @coramat, @kdashg, @litherum)
-
Require render bundles to be no slower than non-render bundles for the same or similar set of commands gpuweb/gpuweb#4104 (4 comments by @greggman, @kainino0x, @litherum)
-
Readonly attributes cannot use `[EnforceRange]` gpuweb/gpuweb#4080 (4 comments by @ErichDonGubler, @kainino0x, @saschanaz)
-
Allow using depth-stencil GPUTextureView with aspects as attachment if pass doesn't use other aspect gpuweb/gpuweb#2132 (2 comments by @kainino0x)
-
read-only storage buffer access from vertex shaders should be allowed gpuweb/gpuweb#4132 (2 comments by @kainino0x, @Kangz)
-
Feature Request: Support for Matrix Inverse Function gpuweb/gpuweb#4115 (2 comments by @litherum, @magcius)
-
Would having Float16Array in JS be useful for integration with WebGPU? gpuweb/gpuweb#3848 (2 comments by @bakkot, @kainino0x)
-
Add lessThanEqual(), lessThan(), greaterThan(), greaterThanEqual(), equal(), notEqual(), any(), all(), not() gpuweb/gpuweb#4112 (2 comments by @ben-clayton, @wrightwriter)
-
Add global_invocation_index and workgroup_index gpuweb/gpuweb#4087 (2 comments by @ben-clayton, @litherum)
-
WGSL u32 CRITICAL Browser / GPU FREEZE on Windows (works on Mac) gpuweb/gpuweb#4107 (2 comments by @alienself, @greggman)
-
"Variable and Value Declaration Feature Summary" table missing `var<storage,write>` gpuweb/gpuweb#4076 (2 comments by @alan-baker, @rconde01)
-
Support for explicit matmul instructions / tensor core instructions gpuweb/gpuweb#4137 (1 comment by @alan-baker)
-
considerations for general implicit promotion of scalars to vectors gpuweb/gpuweb#4126 (1 comment by @dneto0)
-
Viewport rect must lie entirely within the current attachment size? gpuweb/gpuweb#373 (1 comment by @kainino0x)
-
black mode in addition to clamp-to-edge, etc. for sampling textures gpuweb/gpuweb#4084 (1 comment by @kainino0x)
-
-
19 pull requests opened.
-
[editorial] Reorganize WGSL spec sections gpuweb/gpuweb#4154, by @alan-baker.
-
Add WGSL syntax highlighting to more examples gpuweb/gpuweb#4152, by @jrprice.
-
Fix mat4x3 constructor gpuweb/gpuweb#4148, by @kvark.
-
Use wgsl syntax highlighting gpuweb/gpuweb#4146, by @dneto0.
-
Define ULP for values outside the finite range gpuweb/gpuweb#4141, by @dneto0.
-
ULP is for non-equal, finite numbers gpuweb/gpuweb#4140, by @alan-baker.
-
Make section title plural gpuweb/gpuweb#4139, by @alan-baker.
-
Fix workload simulator for v1 spec gpuweb/gpuweb#4136, by @kainino0x.
-
Improve the note about unexpected timestamp values gpuweb/gpuweb#4135, by @toji.
-
Allow reading storage buffers in vertex shaders (#4132). gpuweb/gpuweb#4133, by @mheistermann.
-
Define and use "clip space coordinates" gpuweb/gpuweb#4128, by @dneto0.
-
Fix more [EnforceRange]s gpuweb/gpuweb#4120, by @kainino0x.
-
More promise ordering guarantees gpuweb/gpuweb#4117, by @kainino0x.
-
Validate types of fragment outputs gpuweb/gpuweb#4103, by @kainino0x.
-
webidl fix: Remove [EnforceRange] from GPUBuffer/Texture reflection attributes gpuweb/gpuweb#4102, by @kainino0x.
-
Fix link for localizable text gpuweb/gpuweb#4095, by @toji.
-
Update workload simulator for recent spec changes gpuweb/gpuweb#4093, by @jdarpinian.
-
Add definitions to a bunch of dictionaries that were missing them gpuweb/gpuweb#4089, by @toji.
-
Clarify language on storage texture capabilities gpuweb/gpuweb#4075, by @rconde01.
-
-
20 pull requests merged.
-
[editorial] Reorganize WGSL spec sections gpuweb/gpuweb#4154.
-
Add WGSL syntax highlighting to more examples gpuweb/gpuweb#4152.
-
Fix mat4x3 constructor gpuweb/gpuweb#4148.
-
Use wgsl syntax highlighting gpuweb/gpuweb#4146.
-
Define ULP for values outside the finite range gpuweb/gpuweb#4141.
-
Define and use "clip space coordinates" gpuweb/gpuweb#4128.
-
ULP is for non-equal, finite numbers gpuweb/gpuweb#4140.
-
Make section title plural gpuweb/gpuweb#4139.
-
Improve the note about unexpected timestamp values gpuweb/gpuweb#4135.
-
Fix workload simulator for v1 spec gpuweb/gpuweb#4136.
-
More promise ordering guarantees gpuweb/gpuweb#4117.
-
Refine coordinate systems section gpuweb/gpuweb#4060.
-
Clarify workgroup size restrictions gpuweb/gpuweb#4073.
-
Fix more [EnforceRange]s gpuweb/gpuweb#4120.
-
webidl fix: Remove [EnforceRange] from GPUBuffer/Texture reflection attributes gpuweb/gpuweb#4102.
-
Validate types of fragment outputs gpuweb/gpuweb#4103.
-
Fix link for localizable text gpuweb/gpuweb#4095.
-
Add definitions to a bunch of dictionaries that were missing them gpuweb/gpuweb#4089.
-
Improve WGSL grammar authoring experience with BNF dialect source gpuweb/gpuweb#4027.
-
Specify how errors from the WGSL spec are surfaced in the API gpuweb/gpuweb#4058.
-
-
13 pull requests received 41 new comments.
-
Add definitions to a bunch of dictionaries that were missing them gpuweb/gpuweb#4089 (12 comments by @kainino0x, @toji)
-
Improve WGSL grammar authoring experience with BNF dialect source gpuweb/gpuweb#4027 (7 comments by @dneto0, @mehmetoguzderin)
-
Improve the note about unexpected timestamp values gpuweb/gpuweb#4135 (6 comments by @kainino0x, @toji)
-
Allow reading storage buffers in vertex shaders (#4132). gpuweb/gpuweb#4133 (2 comments by @alan-baker, @mheistermann)
-
Specify how errors from the WGSL spec are surfaced in the API gpuweb/gpuweb#4058 (1 comment by @kainino0x)
-
TR publications
-
18 documents published.
-
WebGPU published 8 times as a Working Draft. History of WebGPU
-
WebGPU Shading Language published 10 times as a Working Draft. History of WebGPU Shading Language
-
GitHub
gpuweb/gpuweb
-
18 issues opened.
-
WebGPU Features + Limits Survey website, did you want to link to it? gpuweb/gpuweb#4074, by @bhouston.
-
<spam> gpuweb/gpuweb#4072, by @MianoMen1.
-
[wgsl] `workgroup_size` does not list negative values as an error gpuweb/gpuweb#4071, by @dj2.
-
Timestamp query resets gpuweb/gpuweb#4069, by @kainino0x.
-
Consider matching either ECMAScript or HTML for whitespace and line break definitions gpuweb/gpuweb#4068, by @annevk.
-
[webgpu] Broken references in WebGPU gpuweb/gpuweb#4066, by @dontcallmedom-bot.
-
Feature Request: BUFFER_BINDING_ARRAY on Metal gpuweb/gpuweb#4065, by @theswerd.
-
[AllowShared] BufferSource isn't well-defined gpuweb/gpuweb#4042, by @kainino0x.
-
Extension: Scalar block layout gpuweb/gpuweb#4040, by @magcius.
-
investigate ULP bounds on unpack2x16unorm gpuweb/gpuweb#4038, by @dneto0.
-
What to do with webgpu.org gpuweb/gpuweb#4032, by @greggman.
-
Feature Request: Some way to generate identity matrices gpuweb/gpuweb#4026, by @litherum.
-
Consider simplifying limits with overlapping meaning gpuweb/gpuweb#4025, by @kainino0x.
-
Add component-wise conversion constructors? gpuweb/gpuweb#4021, by @alan-baker.
-
No abstract type overloads for zero value builtins gpuweb/gpuweb#4019, by @jimblandy.
-
Remove `maxFragmentCombinedOutputResources`? gpuweb/gpuweb#4018, by @kainino0x.
-
uncapturederror shouldn't fire after device loss gpuweb/gpuweb#4017, by @kainino0x.
-
Feature request: extending dynamic state gpuweb/gpuweb#4014, by @adrian17.
-
-
26 issues closed.
-
<spam> gpuweb/gpuweb#4072.
-
Binding aliasing algorithms don't account for dynamic offsets gpuweb/gpuweb#3802.
-
Where do we spec WebGPU-WebCodecs interaction? gpuweb/gpuweb#2498.
-
Investigation: Import VideoFrame from WebCodec to WebGPU gpuweb/gpuweb#1380.
-
Occlusion Query results seem under specified gpuweb/gpuweb#3895.
-
[webgpu] Broken references in WebGPU gpuweb/gpuweb#4066.
-
Feature Request: BUFFER_BINDING_ARRAY on Metal gpuweb/gpuweb#4065.
-
Inconsistencies in createBuffer validation gpuweb/gpuweb#3453.
-
Combined limit for max vertex stage bind groups + vertex buffer bindings gpuweb/gpuweb#2749.
-
It should be possible to unbind a bind group / vertex buffer gpuweb/gpuweb#3787.
-
fragment shader position.z is scaled to viewport mindepth -> maxdepth gpuweb/gpuweb#4003.
-
uncapturederror shouldn't fire after device loss gpuweb/gpuweb#4017.
-
maxBindingsPerBindGroup has arbitrary limit of 640 gpuweb/gpuweb#3864.
-
Fragment shader [[builtin(position)]] definition is misleading gpuweb/gpuweb#2106.
-
Referencing the Unicode Standard gpuweb/gpuweb#2829.
-
Remove `maxFragmentCombinedOutputResources`? gpuweb/gpuweb#4018.
-
No abstract type overloads for zero value builtins gpuweb/gpuweb#4019.
-
[meta] DX12 implementation friction gpuweb/gpuweb#1925.
-
Adding support for rgba16unorm texture format gpuweb/gpuweb#3333.
-
wgsl: unpack2x16snorm exhibits 3ULP error on AMD GPU on macOS Ventura gpuweb/gpuweb#3991.
-
wgsl: unrecognized diagnostics gpuweb/gpuweb#3927.
-
Possibility of SPIR-V and/or GLSL as a WebGPU extension? gpuweb/gpuweb#847.
-
add query for list of WGSL software/shader-lang extensions (e.g. do-while) gpuweb/gpuweb#3875.
-
Consider having rich math rendering in the spec gpuweb/gpuweb#1573.
-
Almost no PowerVR devices satisfy minimum Vulkan requirements gpuweb/gpuweb#3631.
-
-
1 issue reopened.
-
Support for gl_PrimitiveID / SV_PrimitiveID -like construct? gpuweb/gpuweb#1786, by @Kangz.
-
-
45 issues received 110 new comments.
-
Shader extensions are redundant gpuweb/gpuweb#3961 (9 comments by @dneto0, @greggman, @kdashg, @litherum)
-
No abstract type overloads for zero value builtins gpuweb/gpuweb#4019 (8 comments by @alan-baker, @dneto0, @jimblandy, @kdashg, @litherum)
-
Extension: Scalar block layout gpuweb/gpuweb#4040 (6 comments by @ben-clayton, @greggman, @kainino0x, @magcius)
-
Support for gl_PrimitiveID / SV_PrimitiveID -like construct? gpuweb/gpuweb#1786 (6 comments by @dneto0, @Kangz, @kdashg, @litherum)
-
Investigation: Import VideoFrame from WebCodec to WebGPU gpuweb/gpuweb#1380 (5 comments by @dalecurtis, @greggman, @kainino0x, @kdashg)
-
Filterable diagnostic names should be checked, but also permit third-party diagnostics gpuweb/gpuweb#3989 (5 comments by @ben-clayton, @dneto0, @kdashg)
-
Add a floored or euclidean modulo implementation to wgsl gpuweb/gpuweb#3987 (4 comments by @jespertheend, @kdashg, @kyoobey, @stegu)
-
Almost no PowerVR devices satisfy minimum Vulkan requirements gpuweb/gpuweb#3631 (4 comments by @JarredHDavies, @kainino0x, @Kangz)
-
Add component-wise conversion constructors? gpuweb/gpuweb#4021 (3 comments by @ben-clayton, @kdashg)
-
Possibility of SPIR-V and/or GLSL as a WebGPU extension? gpuweb/gpuweb#847 (3 comments by @FOSS-Archives, @kainino0x, @kevincrans)
-
Feature request: extending dynamic state gpuweb/gpuweb#4014 (3 comments by @adrian17, @kainino0x, @mr-mobster)
-
Make lost devices appear to function as much as possible gpuweb/gpuweb#1629 (2 comments by @kainino0x)
-
wgsl: unpack2x16snorm exhibits 3ULP error on AMD GPU on macOS Ventura gpuweb/gpuweb#3991 (2 comments by @kdashg)
-
Consider having rich math rendering in the spec gpuweb/gpuweb#1573 (2 comments by @Kangz, @mehmetoguzderin)
-
`requestAdapterInfo()` should be required to prompt (in at least some circumstances) gpuweb/gpuweb#3962 (1 comment by @kdashg)
-
Consider matching either ECMAScript or HTML for whitespace and line break definitions gpuweb/gpuweb#4068 (1 comment by @mehmetoguzderin)
-
fragment shader position.z is scaled to viewport mindepth -> maxdepth gpuweb/gpuweb#4003 (1 comment by @kainino0x)
-
Viewport rect must lie entirely within the current attachment size? gpuweb/gpuweb#373 (1 comment by @kainino0x)
-
add query for list of WGSL software/shader-lang extensions (e.g. do-while) gpuweb/gpuweb#3875 (1 comment by @kainino0x)
-
-
41 pull requests opened.
-
Clarify workgroup size restrictions gpuweb/gpuweb#4073, by @alan-baker.
-
Clean up more inline issues gpuweb/gpuweb#4070, by @kainino0x.
-
[editorial] Fix broken link gpuweb/gpuweb#4067, by @alan-baker.
-
Mini-typo gpuweb/gpuweb#4064, by @eliemichel.
-
Allow importExternalTexture from VideoFrame gpuweb/gpuweb#4063, by @kainino0x.
-
Clarify zero cases of occlusion queries gpuweb/gpuweb#4062, by @kainino0x.
-
Fix duplicated createBuffer validation, and tweak section levels gpuweb/gpuweb#4061, by @kainino0x.
-
Refine coordinate systems section gpuweb/gpuweb#4060, by @dneto0.
-
Clarify element-by-element matrix construction gpuweb/gpuweb#4059, by @dneto0.
-
Specify how errors from the WGSL spec are surfaced in the API gpuweb/gpuweb#4058, by @kainino0x.
-
Account for dynamic offsets in binding aliasing validation gpuweb/gpuweb#4057, by @kainino0x.
-
Revert "[AllowShared] ArrayBuffer or [AllowShared] ArrayBufferView" gpuweb/gpuweb#4056, by @kainino0x.
-
Small syntactic fix for the IDLs. gpuweb/gpuweb#4055, by @lokokung.
-
Define standard sample patterns gpuweb/gpuweb#4054, by @kainino0x.
-
Fix description of fragment/RasterizationPoint depth gpuweb/gpuweb#4053, by @dneto0.
-
Implementation should always issue a warning on device loss gpuweb/gpuweb#4052, by @kainino0x.
-
Stop raising errors during/after device loss gpuweb/gpuweb#4051, by @kainino0x.
-
Make maxBindingsPerBindGroup 1000 and explain it clearly gpuweb/gpuweb#4050, by @kainino0x.
-
Suggest ways implementations can use labels to enhance error messages gpuweb/gpuweb#4049, by @kainino0x.
-
Position inheritance clauses on separate lines for readability gpuweb/gpuweb#4048, by @kainino0x.
-
Update to bikeshed 3.11.21 gpuweb/gpuweb#4047, by @kainino0x.
-
Fix argumentdef layout gpuweb/gpuweb#4046, by @kainino0x.
-
[AllowShared] ArrayBuffer or [AllowShared] ArrayBufferView gpuweb/gpuweb#4045, by @kainino0x.
-
Remove maxFragmentCombinedOutputResources gpuweb/gpuweb#4044, by @kainino0x.
-
Allow alpha-to-coverage only if attachment 0 has alpha gpuweb/gpuweb#4043, by @kainino0x.
-
Clarify relation of global invocation ID to local invocation ID gpuweb/gpuweb#4041, by @dneto0.
-
Fix description of workgroup_id gpuweb/gpuweb#4039, by @dneto0.
-
Raise an exception on an unknown grammar node type gpuweb/gpuweb#4037, by @dneto0.
-
wgsl/analyzer/Grammar now understands Treesitter "REPEAT" node gpuweb/gpuweb#4036, by @dneto0.
-
Add helpful note to array sizing rules about override idents vs overr… gpuweb/gpuweb#4035, by @dneto0.
-
Fix typo in comment in first shader: "from from" -> "from the" gpuweb/gpuweb#4034, by @kenrussell.
-
Add vec2(), vec3(), vec4() gpuweb/gpuweb#4033, by @dneto0.
-
Add token.token form of diagnostic triggering rule gpuweb/gpuweb#4031, by @dneto0.
-
Use tree-sitter-cli for versioning during extraction gpuweb/gpuweb#4030, by @mehmetoguzderin.
-
Pin tree-sitter version more strictly gpuweb/gpuweb#4029, by @mehmetoguzderin.
-
Description of loop mechanic was inconsistent gpuweb/gpuweb#4028, by @EliasHasle.
-
Improve WGSL grammar authoring experience with BNF dialect source gpuweb/gpuweb#4027, by @mehmetoguzderin.
-
Fix example: must use var<private> for private variable gpuweb/gpuweb#4022, by @dneto0.
-
[editorial] Re-add splat constructors for vectors gpuweb/gpuweb#4020, by @alan-baker.
-
[editorial] Fix section scoping gpuweb/gpuweb#4016, by @alan-baker.
-
[editorial] Clarify uniformity expression categories gpuweb/gpuweb#4015, by @alan-baker.
-
-
42 pull requests merged.
-
Clean up more inline issues gpuweb/gpuweb#4070.
-
Reland "Simplify timestampWrites API" with the rest of the spec updated gpuweb/gpuweb#3979.
-
Account for dynamic offsets in binding aliasing validation gpuweb/gpuweb#4057.
-
Allow importExternalTexture from VideoFrame gpuweb/gpuweb#4063.
-
Clarify zero cases of occlusion queries gpuweb/gpuweb#4062.
-
Restore commas after final members of enums gpuweb/gpuweb#3114.
-
Suggest ways implementations can use labels to enhance error messages gpuweb/gpuweb#4049.
-
Implementation should always issue a warning on device loss gpuweb/gpuweb#4052.
-
[editorial] Fix broken link gpuweb/gpuweb#4067.
-
Mini-typo gpuweb/gpuweb#4064.
-
Fix duplicated createBuffer validation, and tweak section levels gpuweb/gpuweb#4061.
-
Clarify element-by-element matrix construction gpuweb/gpuweb#4059.
-
maxBindGroupsPlusVertexBuffers and unbinding gpuweb/gpuweb#3938.
-
Revert "[AllowShared] ArrayBuffer or [AllowShared] ArrayBufferView" gpuweb/gpuweb#4056.
-
Define standard sample patterns gpuweb/gpuweb#4054.
-
Stop raising errors during/after device loss gpuweb/gpuweb#4051.
-
Make maxBindingsPerBindGroup 1000 and explain it clearly gpuweb/gpuweb#4050.
-
Fix description of fragment/RasterizationPoint depth gpuweb/gpuweb#4053.
-
Clarify fragment shader position input gpuweb/gpuweb#3999.
-
Position inheritance clauses on separate lines for readability gpuweb/gpuweb#4048.
-
Update to bikeshed 3.11.21 gpuweb/gpuweb#4047.
-
Allow alpha-to-coverage only if attachment 0 has alpha gpuweb/gpuweb#4043.
-
Remove maxFragmentCombinedOutputResources gpuweb/gpuweb#4044.
-
[AllowShared] ArrayBuffer or [AllowShared] ArrayBufferView gpuweb/gpuweb#4045.
-
Clarify relation of global invocation ID to local invocation ID gpuweb/gpuweb#4041.
-
Add helpful note to array sizing rules about override idents vs overr… gpuweb/gpuweb#4035.
-
Fix description of workgroup_id gpuweb/gpuweb#4039.
-
Raise an exception on an unknown grammar node type gpuweb/gpuweb#4037.
-
wgsl/analyzer/Grammar now understands Treesitter "REPEAT" node gpuweb/gpuweb#4036.
-
Clarify that the second argument of getContext() is ignored gpuweb/gpuweb#4006.
-
Add token.token form of diagnostic triggering rule gpuweb/gpuweb#4031.
-
Add vec2(), vec3(), vec4() gpuweb/gpuweb#4033.
-
Fix typo in comment in first shader: "from from" -> "from the" gpuweb/gpuweb#4034.
-
Description of loop mechanic was inconsistent gpuweb/gpuweb#4028.
-
wgsl: unpack2x16snorm has 3 ULP error on some implementations gpuweb/gpuweb#3992.
-
Use tree-sitter-cli for versioning during extraction gpuweb/gpuweb#4030.
-
Pin tree-sitter version more strictly gpuweb/gpuweb#4029.
-
Fix example: must use var<private> for private variable gpuweb/gpuweb#4022.
-
[editorial] Re-add splat constructors for vectors gpuweb/gpuweb#4020.
-
Add GPU.wgslLanguageExtensions to list the supported WGSL software extensions gpuweb/gpuweb#3941.
-
[editorial] Clarify uniformity expression categories gpuweb/gpuweb#4015.
-
[editorial] Fix section scoping gpuweb/gpuweb#4016.
-
-
5 pull requests closed.
-
Define WebCodecs interaction in an extension proposal gpuweb/gpuweb#2720.
-
Small syntactic fix for the IDLs. gpuweb/gpuweb#4055.
-
Fix argumentdef layout gpuweb/gpuweb#4046.
-
Add helpful note to array sizing rules about override idents vs override exprs. gpuweb/gpuweb#3986.
-
[wgsl] Add some assert tags to the spec gpuweb/gpuweb#2783.
-
-
1 pull request reopened.
-
Add helpful note to array sizing rules about override idents vs overr… gpuweb/gpuweb#4035, by @dneto0.
-
-
28 pull requests received 140 new comments.
-
Improve WGSL grammar authoring experience with BNF dialect source gpuweb/gpuweb#4027 (60 comments by @alan-baker, @dneto0, @mehmetoguzderin)
-
Refine coordinate systems section gpuweb/gpuweb#4060 (8 comments by @dneto0, @greggman, @kainino0x, @MianoMen1)
-
maxBindGroupsPlusVertexBuffers and unbinding gpuweb/gpuweb#3938 (7 comments by @kainino0x, @litherum, @toji)
-
Small syntactic fix for the IDLs. gpuweb/gpuweb#4055 (6 comments by @ben-clayton, @kainino0x, @lokokung)
-
[editorial] Clarify uniformity expression categories gpuweb/gpuweb#4015 (4 comments by @alan-baker, @dneto0, @mehmetoguzderin)
-
Reland "Simplify timestampWrites API" with the rest of the spec updated gpuweb/gpuweb#3979 (3 comments by @kainino0x, @toji)
-
Suggest ways implementations can use labels to enhance error messages gpuweb/gpuweb#4049 (3 comments by @kainino0x, @litherum)
-
Description of loop mechanic was inconsistent gpuweb/gpuweb#4028 (3 comments by @dneto0, @EliasHasle, @w3cbot)
-
Clarify relation of global invocation ID to local invocation ID gpuweb/gpuweb#4041 (3 comments by @amaiorano, @dneto0)
-
Specify how errors from the WGSL spec are surfaced in the API gpuweb/gpuweb#4058 (2 comments by @alan-baker, @kainino0x)
-
Account for dynamic offsets in binding aliasing validation gpuweb/gpuweb#4057 (2 comments by @kainino0x, @toji)
-
Add helpful note to array sizing rules about override idents vs overr… gpuweb/gpuweb#4035 (2 comments by @alan-baker, @jrprice)
-
Add GPU.wgslLanguageExtensions to list the supported WGSL software extensions gpuweb/gpuweb#3941 (2 comments by @kainino0x)
-
Revert "[AllowShared] ArrayBuffer or [AllowShared] ArrayBufferView" gpuweb/gpuweb#4056 (1 comment by @kainino0x)
-
Add helpful note to array sizing rules about override idents vs override exprs. gpuweb/gpuweb#3986 (1 comment by @dneto0)
-
wgsl: unpack2x16snorm has 3 ULP error on some implementations gpuweb/gpuweb#3992 (1 comment by @dneto0)
-
gpuweb/cts
-
37 issues opened.
-
wgsl: AbstractFloat execution test for `sinh` builtin gpuweb/cts#2536, by @zoddicus.
-
wgsl: f16 execution test for `sinh` builtin gpuweb/cts#2535, by @zoddicus.
-
wgsl: AbstractFloat execution test for `smoothstep` builtin gpuweb/cts#2534, by @zoddicus.
-
wgsl: f16 execution test for `smoothstep` builtin gpuweb/cts#2533, by @zoddicus.
-
wgsl: f16 execution test for `sqrt` builtin gpuweb/cts#2532, by @zoddicus.
-
wgsl: AbstractFloat execution test for `sqrt` builtin gpuweb/cts#2531, by @zoddicus.
-
wgsl: AbstractFloat execution test for `step` builtin gpuweb/cts#2530, by @zoddicus.
-
wgsl: f16 execution test for `step` builtin gpuweb/cts#2529, by @zoddicus.
-
wgsl : AbstractFloat execution test for `tanh` builtin. gpuweb/cts#2528, by @zoddicus.
-
wgsl : f16 execution test for `tanh` builtin. gpuweb/cts#2527, by @zoddicus.
-
wgsl: AbstractFloat execution test for `trunc` builtin gpuweb/cts#2525, by @zoddicus.
-
wgsl: f16 execution test for `trunc` builtin gpuweb/cts#2524, by @zoddicus.
-
wgsl: AbstractFloat execution test for `determinant` builtin gpuweb/cts#2523, by @zoddicus.
-
wgsl: f16 execution test for `determinant` builtin gpuweb/cts#2522, by @zoddicus.
-
wgsl: AbstractFloat execution test for `transpose` builtin gpuweb/cts#2521, by @zoddicus.
-
wgsl: f16 execution test for `transpose` builtin gpuweb/cts#2520, by @zoddicus.
-
val: Distinguish exception type of getCurrentTexture test gpuweb/cts#2518, by @takahirox.
-
val: Make maxBindingsPerBindGroup 1000 gpuweb/cts#2517, by @takahirox.
-
wgsl: i32 execution test for `dot` builtin gpuweb/cts#2516, by @zoddicus.
-
wgsl: u32 execution test for `dot` builtin gpuweb/cts#2515, by @zoddicus.
-
wgsl: AbstractInt execution test for `dot` builtin gpuweb/cts#2514, by @zoddicus.
-
wgsl: f16 execution test for `dot` builtin gpuweb/cts#2513, by @zoddicus.
-
wgsl: AbstractFloat execution test for `dot` builtin gpuweb/cts#2512, by @zoddicus.
-
wgsl: f16 execution test for `round` builtin gpuweb/cts#2510, by @zoddicus.
-
wgsl: AbstractFloat execution test for `round` builtin gpuweb/cts#2509, by @zoddicus.
-
wgsl: f16 execution test for `tan` builtin gpuweb/cts#2507, by @zoddicus.
-
wgsl: AbstractFloat execution test for `tan` builtin gpuweb/cts#2506, by @zoddicus.
-
Clarify the meaning of the castable constants in constants.ts gpuweb/cts#2501, by @zoddicus.
-
Rename `hexToF32/F64` gpuweb/cts#2500, by @zoddicus.
-
Add core support for f16 to floating point testing framework gpuweb/cts#2499, by @zoddicus.
-
Per-aspect capability data for multi-aspect texture formats gpuweb/cts#2495, by @kainino0x.
-
op: test writeBuffer and writeTexture with all possible sources gpuweb/cts#2488, by @kainino0x.
-
unpack2x16snorm now has 3 ULP error gpuweb/cts#2482, by @dneto0.
-
add vec2(), vec3(), vec4() gpuweb/cts#2481, by @dneto0.
-
test updated rules for diagnostic names gpuweb/cts#2480, by @dneto0.
-
Double check 'rVFC' requirement in gpu_external_texture_expiration tests gpuweb/cts#2476, by @shaoboyan.
-
-
33 issues closed.
-
val: Make maxBindingsPerBindGroup 1000 gpuweb/cts#2517.
-
wgsl: f32 execution test for `sinh` builtin gpuweb/cts#1237.
-
wgsl: f32 execution test for `smoothstep` builtin gpuweb/cts#1238.
-
wgsl: f32 execution test for `sqrt` builtin gpuweb/cts#1239.
-
wgsl: Cleanup FP helper functions for implicit f32-ness gpuweb/cts#1808.
-
wgsl: f32 execution test for `step` builtin gpuweb/cts#1240.
-
wgsl : f32 execution test for `tanh` builtin. gpuweb/cts#1242.
-
Add CTS test for early-reject mapAsync gpuweb/cts#2009.
-
wgsl: f32 execution test for `trunc` builtin gpuweb/cts#1243.
-
wgsl: f32 execution test for `determinant` builtin gpuweb/cts#1245.
-
wgsl: f32 execution test for `transpose` builtin gpuweb/cts#1246.
-
wgsl: Validate program gpuweb/cts#1709.
-
val: Check GPUBuffer.mapState when mapped for read. gpuweb/cts#2359.
-
wgsl: validate `enable` gpuweb/cts#1708.
-
wgsl: wgsl source must e UTF-8 without a byte order mark gpuweb/cts#1949.
-
wgsl: f32 execution test for `dot` builtin gpuweb/cts#1247.
-
wgsl: f32 execution test for round builtin gpuweb/cts#1235.
-
wgsl: f32 execution test for `tan` builtin gpuweb/cts#1241.
-
wgsl: bitcast is not a keyword gpuweb/cts#2333.
-
Rename `hexToF32/F64` gpuweb/cts#2500.
-
Add AbstractFloat support to floating point testing framework gpuweb/cts#2416.
-
wgsl: execution test for atomicXor builtin gpuweb/cts#1281.
-
wgsl: execution test for atomicOr builtin gpuweb/cts#1280.
-
wgsl: execution test for atomicAnd builtin gpuweb/cts#1279.
-
wgsl: execution test for atomicMin builtin gpuweb/cts#1278.
-
wgsl: execution test for atomicMax builtin gpuweb/cts#1277.
-
wgsl: execution test for atomicSub builtin gpuweb/cts#1276.
-
wgsl: execution test for atomicAdd builtin gpuweb/cts#1275.
-
Debug marker validation and operation tests gpuweb/cts#1928.
-
op: commmand_buffer,occlusion_queries gpuweb/cts#888.
-
val: device_mismatch throughout CTS gpuweb/cts#912.
-
-
24 issues received 33 new comments.
-
Clarify the meaning of the castable constants in constants.ts gpuweb/cts#2501 (6 comments by @kainino0x, @shrekshao, @zoddicus)
-
val: Check GPUBuffer.mapState when mapped for read. gpuweb/cts#2359 (2 comments by @kainino0x, @takahirox)
-
Need to filter infinity/nan out of test cases when running with a const inputSource gpuweb/cts#1978 (1 comment by @zoddicus)
-
wgsl: Add documentation to Numerics Testing Framework (previously Floating Point Framework) gpuweb/cts#1846 (1 comment by @zoddicus)
-
tests for unpack*unorm and unpack*snorm builtins should not "inherit" accuracy from float division gpuweb/cts#2485 (1 comment by @dneto0)
-
Double check 'rVFC' requirement in gpu_external_texture_expiration tests gpuweb/cts#2476 (1 comment by @kainino0x)
-
-
28 pull requests opened.
-
Check exception type for Webgpu getCurrentTexture test gpuweb/cts#2537, by @takahirox.
-
[wgsl] Add workgroup_size validation tests. gpuweb/cts#2526, by @dj2.
-
Make maxBindingsPerBindGroup 1000 gpuweb/cts#2519, by @takahirox.
-
[wgsl] Validation tests for `enable` gpuweb/cts#2511, by @dj2.
-
test bitcast f32 to {f32,u32,i32} gpuweb/cts#2508, by @dneto0.
-
Clarify that f64 castable constants are about WebIDL casting gpuweb/cts#2505, by @zoddicus.
-
[wgsl] Add test for BOM. gpuweb/cts#2504, by @dj2.
-
wgsl: Add f16 comparison tests gpuweb/cts#2503, by @jiangzhaoming.
-
Replace hexToFoo function names with more accuracte names gpuweb/cts#2502, by @zoddicus.
-
Implement core abstract float support gpuweb/cts#2498, by @zoddicus.
-
Add basic uniformity validation tests gpuweb/cts#2497, by @alan-baker.
-
Execution atomic builtins: add tests for i32 workgroup/storage vars gpuweb/cts#2496, by @amaiorano.
-
Missed one instance of maxFragmentCombinedOutputResources. gpuweb/cts#2494, by @lokokung.
-
Execution tests for atomic builtins on workgroup atomic vars gpuweb/cts#2493, by @amaiorano.
-
Remove numeric framework shim gpuweb/cts#2492, by @zoddicus.
-
Migrate fundamental error interval tests gpuweb/cts#2491, by @zoddicus.
-
Replace calling Vector/Matrix constructors directly gpuweb/cts#2490, by @zoddicus.
-
Migrate bespoke floating point tests gpuweb/cts#2489, by @zoddicus.
-
Migrate Matrix -> tests gpuweb/cts#2487, by @zoddicus.
-
Removes maxFragmentCombinedOutputResources related test. gpuweb/cts#2486, by @lokokung.
-
Execution tests for atomic builtins gpuweb/cts#2484, by @amaiorano.
-
wgsl: unpack2x16snorm has 3ULP error gpuweb/cts#2483, by @dneto0.
-
Migrate various Vector -> tests gpuweb/cts#2479, by @zoddicus.
-
Migrate unpack* tests to new numeric framework gpuweb/cts#2478, by @zoddicus.
-
Migrate Scalar Triple -> Interval tests to use new numeric framework gpuweb/cts#2477, by @zoddicus.
-
Migrate Scalar Pair -> Interval tests to use new numeric framework gpuweb/cts#2475, by @zoddicus.
-
Migrate Scalar -> Interval tests to use new numeric framework gpuweb/cts#2474, by @zoddicus.
-
Migrate existing `abs` tests over to use FP.f32 directly gpuweb/cts#2473, by @zoddicus.
-
-
26 pull requests merged.
-
Make maxBindingsPerBindGroup 1000 gpuweb/cts#2519.
-
wgsl: Add f16 comparison tests gpuweb/cts#2503.
-
[wgsl] Validation tests for `enable` gpuweb/cts#2511.
-
[wgsl] Add test for BOM. gpuweb/cts#2504.
-
Add basic uniformity validation tests gpuweb/cts#2497.
-
Replace hexToFoo function names with more accuracte names gpuweb/cts#2502.
-
Implement core abstract float support gpuweb/cts#2498.
-
Execution atomic builtins: add tests for i32 workgroup/storage vars gpuweb/cts#2496.
-
Missed one instance of maxFragmentCombinedOutputResources. gpuweb/cts#2494.
-
Execution tests for atomic builtins on workgroup atomic vars gpuweb/cts#2493.
-
Remove numeric framework shim gpuweb/cts#2492.
-
Migrate fundamental error interval tests gpuweb/cts#2491.
-
Migrate bespoke floating point tests gpuweb/cts#2489.
-
Replace calling Vector/Matrix constructors directly gpuweb/cts#2490.
-
Migrate Matrix -> tests gpuweb/cts#2487.
-
Execution tests for atomic builtins gpuweb/cts#2484.
-
Removes maxFragmentCombinedOutputResources related test. gpuweb/cts#2486.
-
bitcast execution: integral 32-bit types and vectors of them gpuweb/cts#2428.
-
Migrate various Vector -> tests gpuweb/cts#2479.
-
Fix issues in LimitTestsImpl gpuweb/cts#2472.
-
Migrate unpack* tests to new numeric framework gpuweb/cts#2478.
-
Migrate Scalar Triple -> Interval tests to use new numeric framework gpuweb/cts#2477.
-
Migrate Scalar Pair -> Interval tests to use new numeric framework gpuweb/cts#2475.
-
Migrate Scalar -> Interval tests to use new numeric framework gpuweb/cts#2474.
-
Migrate existing `abs` tests over to use FP.f32 directly gpuweb/cts#2473.
-
-
2 pull requests closed.
-
Fix maxStorageTexturesPerShaderStage validation tests gpuweb/cts#2471.
-
Replace Point/Binary/Ternary with Scalar{Pair/Triple} gpuweb/cts#2470.
-
-
25 pull requests received 88 new comments.
-
Implement core abstract float support gpuweb/cts#2498 (16 comments by @ben-clayton, @dneto0, @zoddicus)
-
Clarify that f64 castable constants are about WebIDL casting gpuweb/cts#2505 (7 comments by @shrekshao, @zoddicus)
-
Execution tests for atomic builtins on workgroup atomic vars gpuweb/cts#2493 (6 comments by @amaiorano, @ben-clayton)
-
Fix maxStorageTexturesPerShaderStage validation tests gpuweb/cts#2471 (6 comments by @greggman, @haoxli, @lokokung)
-
Move "device_destroyed_before_importing_video_frame" case to "api,validation,states,device_lost,destroy" gpuweb/cts#2464 (5 comments by @kainino0x, @shaoboyan)
-
bitcast execution: integral 32-bit types and vectors of them gpuweb/cts#2428 (3 comments by @dneto0)
-
Migrate Scalar Triple -> Interval tests to use new numeric framework gpuweb/cts#2477 (1 comment by @zoddicus)
-
Migrate Scalar Pair -> Interval tests to use new numeric framework gpuweb/cts#2475 (1 comment by @zoddicus)
-
Migrate Scalar -> Interval tests to use new numeric framework gpuweb/cts#2474 (1 comment by @zoddicus)
-
TR publications
-
28 documents published.
-
WebGPU Shading Language published 16 times as a Working Draft. History of WebGPU Shading Language
-
WebGPU published 12 times as a Working Draft. History of WebGPU
-
GitHub
gpuweb/gpuweb
-
39 issues opened.
-
Associativity (and other syntax comments with reader implications) gpuweb/gpuweb#4010, by @mehmetoguzderin.
-
WebGPU dispatch z dimension limited gpuweb/gpuweb#4008, by @huyue466.
-
Standard Sample Pattern gpuweb/gpuweb#4005, by @kainino0x.
-
use wgsl-native syntax highlighting gpuweb/gpuweb#4004, by @dneto0.
-
fragment shader position.z is scaled to viewport mindepth -> maxdepth gpuweb/gpuweb#4003, by @dneto0.
-
WGSL build broken by `<xmp>` change in Bikeshed gpuweb/gpuweb#3995, by @kainino0x.
-
wgsl: unpack2x16snorm exhibits 3ULP error on AMD GPU on macOS Ventura gpuweb/gpuweb#3991, by @dneto0.
-
Filterable diagnostic names should be checked, but also permit third-party diagnostics gpuweb/gpuweb#3989, by @jimblandy.
-
Add a floored or euclidean modulo implementation to wgsl gpuweb/gpuweb#3987, by @jespertheend.
-
Validation and behavior of alpha-to-coverage gpuweb/gpuweb#3982, by @kainino0x.
-
frexp is not defined for subnormals gpuweb/gpuweb#3970, by @amaiorano.
-
Compute shader results inconsistency gpuweb/gpuweb#3968, by @afreg11.
-
Language around return values in `frexp` is potentially too subtle gpuweb/gpuweb#3967, by @zoddicus.
-
editorial: texel format, access mode are not context-dependent names anymore gpuweb/gpuweb#3963, by @dneto0.
-
`requestAdapterInfo()` should be required to prompt (in at least some circumstances) gpuweb/gpuweb#3962, by @litherum.
-
Shader extensions are redundant gpuweb/gpuweb#3961, by @litherum.
-
wgsl: define what shader "dynamic error" means gpuweb/gpuweb#3954, by @dneto0.
-
Problem measuring time deltas with writeTimestamp gpuweb/gpuweb#3952, by @fintelia.
-
Event that triggered when resources fully loaded in gpu gpuweb/gpuweb#3951, by @munrocket.
-
Considerations for subgroups gpuweb/gpuweb#3950, by @raphlinus.
-
wgsl: max(denorm1, denorm2) may return either value, in MSL/Metal gpuweb/gpuweb#3944, by @dneto0.
-
wgsl: give algorithm for number of elements of a runtime-sized array gpuweb/gpuweb#3932, by @dneto0.
-
Curation of WGSL exemplars for LaMDA, GPT-3, etc. gpuweb/gpuweb#3931, by @dcerisano.
-
Evaluation order of assignment and compound assignment are flipped gpuweb/gpuweb#3928, by @ben-clayton.
-
wgsl: unrecognized diagnostics gpuweb/gpuweb#3927, by @dneto0.
-
Reserve u8, u16, u64, u128, i8, i16, i64, i128 for future use. gpuweb/gpuweb#3923, by @rickwebiii.
-
[WGSL] Consider removing limit for maximum workgroup array size gpuweb/gpuweb#3917, by @jrprice.
-
Limit for the maximum resource size? gpuweb/gpuweb#3916, by @kainino0x.
-
Querying number of array layers for TextureCubeArray not supported in SM5.1 gpuweb/gpuweb#3913, by @jrprice.
-
wgsl: check asin, acos accuracy for f16 gpuweb/gpuweb#3909, by @dneto0.
-
wgsl: f32 asin accuracy rule is too tight gpuweb/gpuweb#3906, by @dneto0.
-
depthWriteEnabled and depthCompare are required even for stencil-only formats gpuweb/gpuweb#3905, by @kainino0x.
-
Limits that can be added post-v1 gpuweb/gpuweb#3902, by @kainino0x.
-
Occlusion Query results seem under specified gpuweb/gpuweb#3895, by @greggman.
-
wgsl: consider disabling out of bounds memory accesses to shared memories gpuweb/gpuweb#3893, by @dneto0.
-
Adapters should not be allowed to expose contradictory workgroup size limits gpuweb/gpuweb#3888, by @kainino0x.
-
-
69 issues closed.
-
Describe depth textures gpuweb/gpuweb#3824.
-
Move comment about named values and abstracts. gpuweb/gpuweb#3075.
-
wgsl: The term `override expression` is confusing gpuweb/gpuweb#2981.
-
Add a section for `constructable host-shareable types` ? gpuweb/gpuweb#3046.
-
Structure member layout offset placement doesn't not mention padding. gpuweb/gpuweb#3051.
-
Update comment about reference types not appearing to a note. gpuweb/gpuweb#3052.
-
wgsl: define what shader "dynamic error" means gpuweb/gpuweb#3954.
-
Seek horizontal reviews gpuweb/gpuweb#2531.
-
WebGPU dispatch z dimension limited gpuweb/gpuweb#4008.
-
Metal doesn't allow writing frag_depth without a depth attachment gpuweb/gpuweb#3728.
-
maxStorageBufferBindingSize should be aligned or spec updated gpuweb/gpuweb#3831.
-
Adapters should not be allowed to expose contradictory workgroup size limits gpuweb/gpuweb#3888.
-
Render pipelines with no attachments shouldn't be allowed gpuweb/gpuweb#3702.
-
wgsl: check asin, acos accuracy for f16 gpuweb/gpuweb#3909.
-
Should we be able to pass textures and samplers as function parameters? gpuweb/gpuweb#1695.
-
WGSL build broken by `<xmp>` change in Bikeshed gpuweb/gpuweb#3995.
-
frexp is not defined for subnormals gpuweb/gpuweb#3970.
-
Uniformity analysis is far too strict gpuweb/gpuweb#3479.
-
Accessing an "invalid memory reference": Trapping behavior gpuweb/gpuweb#3272.
-
[WGSL] Consider removing limit for maximum workgroup array size gpuweb/gpuweb#3917.
-
wgsl: consider disabling out of bounds memory accesses to shared memories gpuweb/gpuweb#3893.
-
"validating inter-stage interfaces" doesn't handle vertex-only pipelines gpuweb/gpuweb#3653.
-
Should "mark adapters stale" be called more quickly? gpuweb/gpuweb#1630.
-
Interpretation of `sequence<double>` as `GPUColor` is never actually spelled out gpuweb/gpuweb#3793.
-
editorial: texel format, access mode are not context-dependent names anymore gpuweb/gpuweb#3963.
-
Language around return values in `frexp` is potentially too subtle gpuweb/gpuweb#3967.
-
Compute shader results inconsistency gpuweb/gpuweb#3968.
-
maxInterStageShaderVariables with regards to the fragment stage gpuweb/gpuweb#3865.
-
Limit for the maximum resource size? gpuweb/gpuweb#3916.
-
Event that triggered when resources fully loaded in gpu gpuweb/gpuweb#3951.
-
Canvas texture expiry timing is unpredictable gpuweb/gpuweb#3295.
-
Evaluation order of assignment and compound assignment are flipped gpuweb/gpuweb#3928.
-
wgsl: max(denorm1, denorm2) may return either value, in MSL/Metal gpuweb/gpuweb#3944.
-
wgsl: give algorithm for number of elements of a runtime-sized array gpuweb/gpuweb#3932.
-
rg11b10ufloat-renderable should probably allow use as a resolve target gpuweb/gpuweb#3566.
-
Curation of WGSL exemplars for LaMDA, GPT-3, etc. gpuweb/gpuweb#3931.
-
wgsl: enable directive should support a list of enable-extensions gpuweb/gpuweb#3885.
-
Publish WGSL CI artifact to a repo for reusability gpuweb/gpuweb#3778.
-
question: can barriers synchronize memory accesses across workgroups gpuweb/gpuweb#3774.
-
Text still assumes read-only storage textures are allowed gpuweb/gpuweb#3418.
-
Querying number of array layers for TextureCubeArray not supported in SM5.1 gpuweb/gpuweb#3913.
-
Spec the texture-compression-etc and texture-compression-astc extensions gpuweb/gpuweb#2083.
-
Clarification of `stencil8` texel block size gpuweb/gpuweb#3768.
-
Missing `filterable` texture format definiton gpuweb/gpuweb#3684.
-
Designing a uniformity opt-out gpuweb/gpuweb#3554.
-
Add example in Memory Views section showing a ref type. gpuweb/gpuweb#3053.
-
should empty structs be permitted? gpuweb/gpuweb#1222.
-
Reserve u8, u16, u64, u128, i8, i16, i64, i128 for future use. gpuweb/gpuweb#3923.
-
Clarification around `host-shareable types` gpuweb/gpuweb#3035.
-
wgsl: f32 asin accuracy rule is too tight gpuweb/gpuweb#3906.
-
Defaults for depthWriteEnabled and depthCompare may be surprising gpuweb/gpuweb#3798.
-
Why does depthClearValue default to 0.0? gpuweb/gpuweb#3777.
-
Minor text changes in Memory Model section gpuweb/gpuweb#3219.
-
Optional feature request: filterable 32-bit float textures gpuweb/gpuweb#3601.
-
Accuracy not specified for `determinant` gpuweb/gpuweb#3873.
-
Explainer is marked as a "living standard" gpuweb/gpuweb#3687.
-
Spec should use [=map/exists=] for optional dictionary members gpuweb/gpuweb#1512.
-
Constant identifier expression table clarifications gpuweb/gpuweb#3147.
-
wgsl: Mark type constructors `@const`? gpuweb/gpuweb#2982.
-
Conversion expression preamble claims no conversion or promotion of numerics. gpuweb/gpuweb#2960.
-
-
3 issues reopened.
-
Canvas texture expiry timing is unpredictable gpuweb/gpuweb#3295, by @kainino0x.
-
Referencing the Unicode Standard gpuweb/gpuweb#2829, by @kainino0x.
-
Almost no PowerVR devices satisfy minimum Vulkan requirements gpuweb/gpuweb#3631, by @kainino0x.
-
-
56 issues received 171 new comments.
-
Add a floored or euclidean modulo implementation to wgsl gpuweb/gpuweb#3987 (13 comments by @ben-clayton, @jespertheend, @stegu, @teoxoy)
-
add query for list of WGSL software/shader-lang extensions (e.g. do-while) gpuweb/gpuweb#3875 (10 comments by @dj2, @greggman, @kainino0x, @Kangz, @kdashg, @toji)
-
wgsl: float to integer conversions should saturate in the destination type, not the source type. gpuweb/gpuweb#3908 (8 comments by @ben-clayton, @dneto0, @kdashg)
-
`requestAdapterInfo()` should be required to prompt (in at least some circumstances) gpuweb/gpuweb#3962 (7 comments by @annevk, @kainino0x, @Kangz, @kdashg, @litherum)
-
Curation of WGSL exemplars for LaMDA, GPT-3, etc. gpuweb/gpuweb#3931 (7 comments by @dcerisano, @dneto0, @kainino0x)
-
maxBindingsPerBindGroup has arbitrary limit of 640 gpuweb/gpuweb#3864 (6 comments by @greggman, @kainino0x, @Kangz, @mwyrzykowski)
-
Who actually wants multiple timestampWrites with the same TimestampWrite.location? gpuweb/gpuweb#3808 (6 comments by @kainino0x, @Kangz, @kdashg, @litherum)
-
wgsl: consider disabling out of bounds memory accesses to shared memories gpuweb/gpuweb#3893 (6 comments by @dneto0, @kdashg, @teoxoy)
-
Assignment of swizzled values should be permitted gpuweb/gpuweb#737 (5 comments by @johnstiles-google, @magcius)
-
Evaluation order of assignment and compound assignment are flipped gpuweb/gpuweb#3928 (5 comments by @ben-clayton, @dneto0, @kdashg)
-
Querying number of array layers for TextureCubeArray not supported in SM5.1 gpuweb/gpuweb#3913 (5 comments by @ben-clayton, @dneto0, @teoxoy)
-
wgsl: Missing specification on modf behavior for negatives. gpuweb/gpuweb#2468 (4 comments by @jespertheend, @stegu)
-
Filterable diagnostic names should be checked, but also permit third-party diagnostics gpuweb/gpuweb#3989 (3 comments by @ben-clayton, @jimblandy, @jrprice)
-
Should we be able to pass textures and samplers as function parameters? gpuweb/gpuweb#1695 (3 comments by @ben-clayton, @maierfelix)
-
Uniformity analysis is far too strict gpuweb/gpuweb#3479 (3 comments by @brendan-duncan, @dneto0, @kdashg)
-
depthWriteEnabled and depthCompare are required even for stencil-only formats gpuweb/gpuweb#3905 (3 comments by @kainino0x, @kdashg)
-
Almost no PowerVR devices satisfy minimum Vulkan requirements gpuweb/gpuweb#3631 (3 comments by @kainino0x)
-
Compute shader results inconsistency gpuweb/gpuweb#3968 (3 comments by @ben-clayton, @jrprice, @Kangz)
-
Adapters should not be allowed to expose contradictory workgroup size limits gpuweb/gpuweb#3888 (3 comments by @kainino0x, @kdashg)
-
maxInterStageShaderVariables with regards to the fragment stage gpuweb/gpuweb#3865 (3 comments by @greggman, @kainino0x, @litherum)
-
Attribute syntax should not specialize to built-in language attributes gpuweb/gpuweb#3520 (3 comments by @dneto0, @mehmetoguzderin)
-
Associativity (and other syntax comments with reader implications) gpuweb/gpuweb#4010 (2 comments by @alan-baker, @mehmetoguzderin)
-
Make lost devices appear to function as much as possible gpuweb/gpuweb#1629 (2 comments by @kainino0x)
-
Fragment shader [[builtin(position)]] definition is misleading gpuweb/gpuweb#2106 (2 comments by @dneto0)
-
Accessing an "invalid memory reference": Trapping behavior gpuweb/gpuweb#3272 (2 comments by @dneto0)
-
(GPUDeviceLostReason or undefined) seems non-standard and results in problematic feature detection gpuweb/gpuweb#3894 (2 comments by @kainino0x)
-
[WGSL] Consider removing limit for maximum workgroup array size gpuweb/gpuweb#3917 (2 comments by @kdashg)
-
Event that triggered when resources fully loaded in gpu gpuweb/gpuweb#3951 (2 comments by @kainino0x, @munrocket)
-
Publish WGSL CI artifact to a repo for reusability gpuweb/gpuweb#3778 (2 comments by @kdashg, @mehmetoguzderin)
-
wgsl: add way to get special values for a given numeric type (e.g. max, min, lowest) gpuweb/gpuweb#3431 (2 comments by @ben-clayton, @m-schuetz)
-
Validation for resolveQuerySet() doesn't require the query to have already been written to gpuweb/gpuweb#3812 (2 comments by @greggman, @kainino0x)
-
fragment shader position.z is scaled to viewport mindepth -> maxdepth gpuweb/gpuweb#4003 (1 comment by @kainino0x)
-
"validating inter-stage interfaces" doesn't handle vertex-only pipelines gpuweb/gpuweb#3653 (1 comment by @kainino0x)
-
publish-TR-webgpu failing: INE ~2043: No 'abstract-op' refs found for 'update the rendering of the WebGPU canvas'. gpuweb/gpuweb#3964 (1 comment by @dneto0)
-
wgsl: Is/should initializer of module-scope variables limited to creation-time expressions? gpuweb/gpuweb#3081 (1 comment by @dneto0)
-
wgsl: enable directive should support a list of enable-extensions gpuweb/gpuweb#3885 (1 comment by @kdashg)
-
Reserve u8, u16, u64, u128, i8, i16, i64, i128 for future use. gpuweb/gpuweb#3923 (1 comment by @ben-clayton)
-
Combined limit for max vertex stage bind groups + vertex buffer bindings gpuweb/gpuweb#2749 (1 comment by @kainino0x)
-
-
82 pull requests opened.
-
[editorial] Fill out texture types gpuweb/gpuweb#4013, by @alan-baker.
-
Specify what happens if a dynamic error occrs gpuweb/gpuweb#4011, by @dneto0.
-
WASM Aid: Less queue capacity, don't account ASCII gpuweb/gpuweb#4007, by @mehmetoguzderin.
-
Clarify that the second argument of getContext() is ignored gpuweb/gpuweb#4006, by @takahirox.
-
Expose namespaces only in secure contexts, like interfaces. gpuweb/gpuweb#4002, by @jimblandy.
-
[editorial] Suppress unused dfns using noexport gpuweb/gpuweb#4001, by @kainino0x.
-
Unpack bulitins don't have f16 counterparts gpuweb/gpuweb#4000, by @dneto0.
-
Clarify fragment shader position input gpuweb/gpuweb#3999, by @dneto0.
-
Fix when return value is ignored in a function call statement gpuweb/gpuweb#3998, by @dneto0.
-
Move all `<xmp>` to start on their own line gpuweb/gpuweb#3997, by @kainino0x.
-
Build fixes for latest Bikeshed 3.11.18 gpuweb/gpuweb#3993, by @kainino0x.
-
wgsl: unpack2x16snorm has 3 ULP error on some implementations gpuweb/gpuweb#3992, by @dneto0.
-
Weaken asin & acos accuracy rule for f16 gpuweb/gpuweb#3990, by @jiangzhaoming.
-
frexp is well defined only for zero and normal numbers gpuweb/gpuweb#3988, by @dneto0.
-
Add helpful note to array sizing rules about override idents vs override exprs. gpuweb/gpuweb#3986, by @kdashg.
-
Fix GLSL comparison in the WGSL spec gpuweb/gpuweb#3985, by @avtne.
-
Require a depth attachment if frag_depth builtin is used gpuweb/gpuweb#3984, by @kainino0x.
-
alphaToCoverage comes from output 0, not target 0 gpuweb/gpuweb#3983, by @kainino0x.
-
Make aliased bindings an error gpuweb/gpuweb#3981, by @kainino0x.
-
Require at least one attachment in GPURenderPipelineDescriptor gpuweb/gpuweb#3980, by @kainino0x.
-
Reland "Simplify timestampWrites API" with the rest of the spec updated gpuweb/gpuweb#3979, by @kainino0x.
-
Add resource-hints spec to ignored-specs gpuweb/gpuweb#3978, by @kainino0x.
-
Revert "Simplify timestampWrites API" gpuweb/gpuweb#3977, by @kainino0x.
-
GPUDeviceLostReason "unknown" instead of undefined gpuweb/gpuweb#3976, by @kainino0x.
-
Add more adapter limit guarantees gpuweb/gpuweb#3975, by @kainino0x.
-
Clarify workgroup array size limit gpuweb/gpuweb#3974, by @dneto0.
-
editorial: Change "race conditions" to "data race" gpuweb/gpuweb#3973, by @dneto0.
-
editorial: Define and use "logical texel address" gpuweb/gpuweb#3972, by @dneto0.
-
editorial: Define mathematical interval notation gpuweb/gpuweb#3971, by @dneto0.
-
Make frexp fract ranges explicit gpuweb/gpuweb#3969, by @zoddicus.
-
Fix cross-link to updating the rendering of a WebGPU canvas gpuweb/gpuweb#3966, by @dneto0.
-
editorial: Texel format and access mode are not context-dependent names gpuweb/gpuweb#3965, by @dneto0.
-
Define GPUImageCopyExternalImage source size gpuweb/gpuweb#3960, by @shaoboyan.
-
[editorial] Change redundant inter-stage variable validation to assertion gpuweb/gpuweb#3959, by @kainino0x.
-
Add accessors for GPUColor, define the sequence order gpuweb/gpuweb#3958, by @kainino0x.
-
Switch all GPUExtent/Origin accessors namespace gpuweb/gpuweb#3957, by @kainino0x.
-
Out of bounds access via reference is dynamic error gpuweb/gpuweb#3956, by @dneto0.
-
Fix up expiry timing definitions gpuweb/gpuweb#3955, by @kainino0x.
-
Add a README file to the logos folder gpuweb/gpuweb#3953, by @toji.
-
assignment evaluates from left to right gpuweb/gpuweb#3949, by @dneto0.
-
Fix .gitignore for wgsl/index.bs.pre gpuweb/gpuweb#3948, by @dneto0.
-
Variable address space restrictions gpuweb/gpuweb#3947, by @alan-baker.
-
Fix HTML validation: predeclared enumerant table only has 2 columns gpuweb/gpuweb#3946, by @dneto0.
-
Weaken min,max,clamp on denorm inputs gpuweb/gpuweb#3945, by @dneto0.
-
Specify behavior of OffscreenCanvases gpuweb/gpuweb#3943, by @kainino0x.
-
Describe the type of `GPUBindGroupLayout.[[descriptor]]`. gpuweb/gpuweb#3942, by @jimblandy.
-
Add GPU.wgslLanguageExtensions to list the supported WGSL software extensions gpuweb/gpuweb#3941, by @dneto0.
-
Adds maxFragmentCombinedOutputResources to GPUSupportedLimits. gpuweb/gpuweb#3940, by @lokokung.
-
float to integral conversion saturates in the target type gpuweb/gpuweb#3939, by @dneto0.
-
maxBindGroupsPlusVertexBuffers and unbinding gpuweb/gpuweb#3938, by @kainino0x.
-
Specify behavior of onSubmittedWorkDone/mapAsync on device loss gpuweb/gpuweb#3937, by @kainino0x.
-
Expire GPUAdapter on the first successful requestDevice gpuweb/gpuweb#3936, by @kainino0x.
-
wgsl: Regenerate recursive grammar gpuweb/gpuweb#3934, by @dneto0.
-
Give algorithm for number of elements of runtime-sized array gpuweb/gpuweb#3933, by @dneto0.
-
Simplify timestampWrites API gpuweb/gpuweb#3930, by @kainino0x.
-
[editorial] Always add vertical lines to data tables gpuweb/gpuweb#3929, by @kainino0x.
-
[editorial] Add symmetry restriction about alignment gpuweb/gpuweb#3926, by @alan-baker.
-
[editorial] Clarify memory locations gpuweb/gpuweb#3925, by @alan-baker.
-
[editorial] Rename variable_statement grammar gpuweb/gpuweb#3924, by @alan-baker.
-
[editorial] Clearly distinguish texel copy footprints vs memory costs gpuweb/gpuweb#3922, by @kainino0x.
-
[editorial] Justify GPUStorageTextureAccess when it only has one value gpuweb/gpuweb#3921, by @kainino0x.
-
Fix a "filterable" vs "blendable" reference mixup gpuweb/gpuweb#3920, by @kainino0x.
-
[editorial] Make a new section for Adapter Capability Guarantees gpuweb/gpuweb#3919, by @kainino0x.
-
Remove wrong restrictions on maxFragmentCombinedOutputResources value gpuweb/gpuweb#3918, by @kainino0x.
-
Remove textureNumLayers overloads for texture_cube_array texture_dept… gpuweb/gpuweb#3915, by @dneto0.
-
wgsl:editorial: Consistenly stye "Notes" as markers gpuweb/gpuweb#3914, by @dneto0.
-
Issue 3554 diag statement scoped generalized attrib gpuweb/gpuweb#3912, by @dneto0.
-
Generalize attribute grammar gpuweb/gpuweb#3911, by @dneto0.
-
[editorial] clarify host-shareable types gpuweb/gpuweb#3910, by @alan-baker.
-
Weaken asin accuracy rule for f32 gpuweb/gpuweb#3907, by @dneto0.
-
Ignore temporary file gpuweb/gpuweb#3904, by @mehmetoguzderin.
-
[editorial] Assorted nits gpuweb/gpuweb#3903, by @kainino0x.
-
Add ReturnIfAbrupt [=?=] on content-timeline validation gpuweb/gpuweb#3899, by @kainino0x.
-
wgsl: weaken determinant error bounds gpuweb/gpuweb#3898, by @dneto0.
-
treesitter: Add more specific unit tests for template parsing gpuweb/gpuweb#3897, by @dneto0.
-
Rename notes column to description where appropriate gpuweb/gpuweb#3896, by @alan-baker.
-
wgsl: uniform buffers don't support stores gpuweb/gpuweb#3892, by @dneto0.
-
wgsl/Makefile: fix the "clean" target gpuweb/gpuweb#3891, by @dneto0.
-
Detabularize "Predeclared Types" list gpuweb/gpuweb#3890, by @mehmetoguzderin.
-
Annotate associativity for easier reading and downstream tooling gpuweb/gpuweb#3889, by @mehmetoguzderin.
-
Add TSPath.py XPath-style expressions for Treesitter nodes gpuweb/gpuweb#3887, by @dneto0.
-
-
82 pull requests merged.
-
[editorial] Fill out texture types gpuweb/gpuweb#4013.
-
Specify what happens if a dynamic error occrs gpuweb/gpuweb#4011.
-
Fix typo gpuweb/gpuweb#4012.
-
WASM Aid: Less queue capacity, don't account ASCII gpuweb/gpuweb#4007.
-
Require a depth attachment if frag_depth builtin is used gpuweb/gpuweb#3984.
-
[editorial] Suppress unused dfns using noexport gpuweb/gpuweb#4001.
-
Add more adapter limit guarantees gpuweb/gpuweb#3975.
-
Make aliased bindings an error gpuweb/gpuweb#3981.
-
Require at least one attachment in GPURenderPipelineDescriptor gpuweb/gpuweb#3980.
-
alphaToCoverage comes from output 0, not target 0 gpuweb/gpuweb#3983.
-
Expose namespaces only in secure contexts, like interfaces. gpuweb/gpuweb#4002.
-
Unpack bulitins don't have f16 counterparts gpuweb/gpuweb#4000.
-
Fix when return value is ignored in a function call statement gpuweb/gpuweb#3998.
-
Weaken asin & acos accuracy rule for f16 gpuweb/gpuweb#3990.
-
Move all `<xmp>` to start on their own line gpuweb/gpuweb#3997.
-
Build fixes for latest Bikeshed 3.11.18 gpuweb/gpuweb#3993.
-
Define GPUImageCopyExternalImage source size gpuweb/gpuweb#3960.
-
frexp is well defined only for zero and normal numbers gpuweb/gpuweb#3988.
-
Clarify workgroup array size limit gpuweb/gpuweb#3974.
-
Out of bounds access via reference is dynamic error gpuweb/gpuweb#3956.
-
Fix GLSL comparison in the WGSL spec gpuweb/gpuweb#3985.
-
Add resource-hints spec to ignored-specs gpuweb/gpuweb#3978.
-
Revert "Simplify timestampWrites API" gpuweb/gpuweb#3977.
-
GPUDeviceLostReason "unknown" instead of undefined gpuweb/gpuweb#3976.
-
editorial: Define and use "logical texel address" gpuweb/gpuweb#3972.
-
editorial: Define mathematical interval notation gpuweb/gpuweb#3971.
-
editorial: Change "race conditions" to "data race" gpuweb/gpuweb#3973.
-
Specify behavior of onSubmittedWorkDone/mapAsync on device loss gpuweb/gpuweb#3937.
-
Make GPUPipelineError constructor `message` pseudo-optional gpuweb/gpuweb#3710.
-
Expire GPUAdapter on the first successful requestDevice gpuweb/gpuweb#3936.
-
Simplify timestampWrites API gpuweb/gpuweb#3930.
-
Fix up expiry timing definitions gpuweb/gpuweb#3955.
-
Add accessors for GPUColor, define the sequence order gpuweb/gpuweb#3958.
-
editorial: Texel format and access mode are not context-dependent names gpuweb/gpuweb#3965.
-
Make frexp fract ranges explicit gpuweb/gpuweb#3969.
-
Fix cross-link to updating the rendering of a WebGPU canvas gpuweb/gpuweb#3966.
-
[editorial] Change redundant inter-stage variable validation to assertion gpuweb/gpuweb#3959.
-
Switch all GPUExtent/Origin accessors namespace gpuweb/gpuweb#3957.
-
Specify behavior of OffscreenCanvases gpuweb/gpuweb#3943.
-
assignment evaluates from left to right gpuweb/gpuweb#3949.
-
Add a README file to the logos folder gpuweb/gpuweb#3953.
-
Weaken min,max,clamp on denorm inputs gpuweb/gpuweb#3945.
-
Fix .gitignore for wgsl/index.bs.pre gpuweb/gpuweb#3948.
-
Variable address space restrictions gpuweb/gpuweb#3947.
-
Fix HTML validation: predeclared enumerant table only has 2 columns gpuweb/gpuweb#3946.
-
Give algorithm for number of elements of runtime-sized array gpuweb/gpuweb#3933.
-
UAs MUST restrict to <=32 distinguishable [fingerprinting] buckets. gpuweb/gpuweb#3840.
-
float to integral conversion saturates in the target type gpuweb/gpuweb#3939.
-
Describe the type of `GPUBindGroupLayout.[[descriptor]]`. gpuweb/gpuweb#3942.
-
Adds maxFragmentCombinedOutputResources to GPUSupportedLimits. gpuweb/gpuweb#3940.
-
Make rg11b10ufloat-renderable also allow resolving gpuweb/gpuweb#3834.
-
Distinguish between address and contents uniformity gpuweb/gpuweb#3816.
-
Enable directive takes a list of extensions. gpuweb/gpuweb#3886.
-
[editorial] Justify GPUStorageTextureAccess when it only has one value gpuweb/gpuweb#3921.
-
wgsl: Regenerate recursive grammar gpuweb/gpuweb#3934.
-
[editorial] Clarify memory locations gpuweb/gpuweb#3925.
-
[editorial] Rename variable_statement grammar gpuweb/gpuweb#3924.
-
[editorial] Make a new section for Adapter Capability Guarantees gpuweb/gpuweb#3919.
-
[editorial] Always add vertical lines to data tables gpuweb/gpuweb#3929.
-
[editorial] Clearly distinguish texel copy footprints vs memory costs gpuweb/gpuweb#3922.
-
Fix a "filterable" vs "blendable" reference mixup gpuweb/gpuweb#3920.
-
Remove wrong restrictions on maxFragmentCombinedOutputResources value gpuweb/gpuweb#3918.
-
wgsl:editorial: Consistenly stye "Notes" as markers gpuweb/gpuweb#3914.
-
[editorial] clarify host-shareable types gpuweb/gpuweb#3910.
-
Weaken asin accuracy rule for f32 gpuweb/gpuweb#3907.
-
Detabularize "Predeclared Types" list gpuweb/gpuweb#3890.
-
treesitter: Add more specific unit tests for template parsing gpuweb/gpuweb#3897.
-
Make depthClearValue, depthWriteEnabled, depthCompare required gpuweb/gpuweb#3849.
-
Add maxFragmentCombinedOutputResources limit gpuweb/gpuweb#3829.
-
[editorial] Assorted nits gpuweb/gpuweb#3903.
-
Add ReturnIfAbrupt [=?=] on content-timeline validation gpuweb/gpuweb#3899.
-
Ignore temporary file gpuweb/gpuweb#3904.
-
Add feature for filterable *32float textures gpuweb/gpuweb#3828.
-
Rename notes column to description where appropriate gpuweb/gpuweb#3896.
-
wgsl: weaken determinant error bounds gpuweb/gpuweb#3898.
-
Improve validation language around timestampWrites gpuweb/gpuweb#3821.
-
Change explainer status to "Draft Community Group Report" gpuweb/gpuweb#3826.
-
Update checks for dictionary members to use "exists" gpuweb/gpuweb#3827.
-
wgsl: uniform buffers don't support stores gpuweb/gpuweb#3892.
-
wgsl/Makefile: fix the "clean" target gpuweb/gpuweb#3891.
-
Add TSPath.py XPath-style expressions for Treesitter nodes gpuweb/gpuweb#3887.
-
-
7 pull requests closed.
-
Working on a way to easily show issue milestones gpuweb/gpuweb#3781.
-
[editorial] Add symmetry restriction about alignment gpuweb/gpuweb#3926.
-
Add maxBindGroupsPlusVertexBuffers limit gpuweb/gpuweb#3788.
-
Remove textureNumLayers overloads for texture_cube_array texture_dept… gpuweb/gpuweb#3915.
-
Issue 3554 diag statement scoped generalized attrib gpuweb/gpuweb#3912.
-
Demo: Turbofish for generics is not ambiguous gpuweb/gpuweb#3765.
-
Rename `compilationInfo()` to `createCompilationInfo()` gpuweb/gpuweb#3805.
-
-
59 pull requests received 248 new comments.
-
Add GPU.wgslLanguageExtensions to list the supported WGSL software extensions gpuweb/gpuweb#3941 (30 comments by @dneto0, @kainino0x)
-
Out of bounds access via reference is dynamic error gpuweb/gpuweb#3956 (24 comments by @alan-baker, @dneto0)
-
maxBindGroupsPlusVertexBuffers and unbinding gpuweb/gpuweb#3938 (19 comments by @greggman, @jimblandy, @kainino0x, @Kangz, @toji)
-
Give algorithm for number of elements of runtime-sized array gpuweb/gpuweb#3933 (13 comments by @alan-baker, @dneto0, @jrprice)
-
Generalize attribute grammar gpuweb/gpuweb#3911 (13 comments by @alan-baker, @dneto0, @mehmetoguzderin)
-
Add diagnostic filtering; range diagnostics apply to control flow constructs and compound statements gpuweb/gpuweb#3713 (12 comments by @dneto0, @jimblandy)
-
Simplify timestampWrites API gpuweb/gpuweb#3930 (10 comments by @greggman, @kainino0x, @litherum, @toji)
-
Make depthClearValue, depthWriteEnabled, depthCompare required gpuweb/gpuweb#3849 (9 comments by @kainino0x, @takahirox, @toji)
-
float to integral conversion saturates in the target type gpuweb/gpuweb#3939 (7 comments by @dneto0, @mehmetoguzderin)
-
[editorial] Add symmetry restriction about alignment gpuweb/gpuweb#3926 (6 comments by @alan-baker, @ben-clayton, @dneto0, @jrprice)
-
treesitter: Add more specific unit tests for template parsing gpuweb/gpuweb#3897 (6 comments by @dneto0)
-
Define GPUImageCopyExternalImage source size gpuweb/gpuweb#3960 (5 comments by @kainino0x, @shaoboyan)
-
editorial: Define and use "logical texel address" gpuweb/gpuweb#3972 (5 comments by @alan-baker, @dneto0)
-
Specify behavior of onSubmittedWorkDone/mapAsync on device loss gpuweb/gpuweb#3937 (5 comments by @kainino0x, @kdashg, @toji)
-
Distinguish between address and contents uniformity gpuweb/gpuweb#3816 (5 comments by @dneto0, @jrprice)
-
WASM Aid: Less queue capacity, don't account ASCII gpuweb/gpuweb#4007 (4 comments by @dneto0, @mehmetoguzderin)
-
[editorial] Clearly distinguish texel copy footprints vs memory costs gpuweb/gpuweb#3922 (3 comments by @kainino0x, @teoxoy, @toji)
-
wgsl: unpack2x16snorm has 3 ULP error on some implementations gpuweb/gpuweb#3992 (2 comments by @alan-baker, @dneto0)
-
Fix HTML validation: predeclared enumerant table only has 2 columns gpuweb/gpuweb#3946 (2 comments by @dneto0)
-
Clarify that the second argument of getContext() is ignored gpuweb/gpuweb#4006 (1 comment by @kainino0x)
-
Reland "Simplify timestampWrites API" with the rest of the spec updated gpuweb/gpuweb#3979 (1 comment by @kainino0x)
-
Add helpful note to array sizing rules about override idents vs override exprs. gpuweb/gpuweb#3986 (1 comment by @dneto0)
-
Fix when return value is ignored in a function call statement gpuweb/gpuweb#3998 (1 comment by @dneto0)
-
Make GPUPipelineError constructor `message` pseudo-optional gpuweb/gpuweb#3710 (1 comment by @kainino0x)
-
Adds maxFragmentCombinedOutputResources to GPUSupportedLimits. gpuweb/gpuweb#3940 (1 comment by @greggman)
-
Expire GPUAdapter on the first successful requestDevice gpuweb/gpuweb#3936 (1 comment by @austinEng)
-
Remove textureNumLayers overloads for texture_cube_array texture_dept… gpuweb/gpuweb#3915 (1 comment by @dneto0)
-
Use task sources for device events and canvas texture expiry gpuweb/gpuweb#3632 (1 comment by @shrekshao)
-
UAs MUST restrict to <=32 distinguishable [fingerprinting] buckets. gpuweb/gpuweb#3840 (1 comment by @kainino0x)
-
Add TSPath.py XPath-style expressions for Treesitter nodes gpuweb/gpuweb#3887 (1 comment by @dneto0)
-
gpuweb/cts
-
55 issues opened.
-
depth_clip_clamp test is not actually testing unclippedDepth? gpuweb/cts#2460, by @dneto0.
-
Non-`inputSource='const'` compound assignment shouldn't go via a `var` gpuweb/cts#2453, by @ben-clayton.
-
Consider changing expression tests `inputSource` gpuweb/cts#2452, by @ben-clayton.
-
Change `unpack2x16snorm` acceptance intervals to 3 ULP gpuweb/cts#2451, by @zoddicus.
-
Adjust expections for `frexp` on subnormals gpuweb/cts#2446, by @zoddicus.
-
Replace use of two hex u32s to represent f64s with BigInt gpuweb/cts#2437, by @zoddicus.
-
ULP calculations return Scalars instead of numbers gpuweb/cts#2436, by @zoddicus.
-
`f32` conversion for `f32` inputs do no correctly account for FTZ gpuweb/cts#2435, by @zoddicus.
-
update tests for GPUDeviceLostReason change gpuweb/cts#2431, by @kainino0x.
-
op: behavior of onSubmittedWorkDone/mapAsync on device loss gpuweb/cts#2430, by @kainino0x.
-
op: device.destroy() should unmap buffers with an AbortError gpuweb/cts#2429, by @kainino0x.
-
Reorganize the tests gpuweb/cts#2427, by @greggman.
-
Add AbstractFloat support to floating point testing framework gpuweb/cts#2416, by @zoddicus.
-
getCurrentTexture on unconfigured canvas should be InvalidStateError gpuweb/cts#2410, by @kainino0x.
-
wgsl: comparison expressions. i32, u32 gpuweb/cts#2404, by @dneto0.
-
wgsl: Weaken min,max,clamp on denorm inputs gpuweb/cts#2403, by @jrprice.
-
Add and/or use additional `expect equals` helpers to the fixture gpuweb/cts#2402, by @austinEng.
-
CTS subcases should detect duplicates gpuweb/cts#2401, by @zoddicus.
-
address space restrictions vs. shader stage gpuweb/cts#2398, by @dneto0.
-
`discard` statements gpuweb/cts#2391, by @ben-clayton.
-
web_platform: canvas,getCurrentTexture gpuweb/cts#2385, by @shrekshao.
-
web_platform: different canvas types getCurrentTexture expiry test cases gpuweb/cts#2384, by @shrekshao.
-
op: adapter becomes stale after the first successful requestDevice gpuweb/cts#2380, by @austinEng.
-
val: queue,submit gpuweb/cts#2376, by @greggman.
-
val: queries,general gpuweb/cts#2375, by @greggman.
-
val: queries,begin_end gpuweb/cts#2374, by @greggman.
-
Update `acos` accuracy to match spec gpuweb/cts#2371, by @zoddicus.
-
val/op/web_platform: Check GOUColor/Extent3D/Origin2D/Origin3D reification gpuweb/cts#2370, by @Kangz.
-
reftests: Check that an "onscreen" OffscreenCanvas in a worker is tests. gpuweb/cts#2369, by @Kangz.
-
web_platform: Check the result of copying to a canvas with a different alphaMode gpuweb/cts#2368, by @Kangz.
-
web_platform: Check the result of transferToImageBitmap with a different alphaMode gpuweb/cts#2367, by @Kangz.
-
web_platform: Check the result of transferToImageBitmap after presentation gpuweb/cts#2366, by @Kangz.
-
web_platform: copyEI2T with a source rectangle OOB with the source gpuweb/cts#2365, by @Kangz.
-
reftests: setting the CSS image-rendering of the canvas gpuweb/cts#2364, by @Kangz.
-
val: GPUShaderModule.getCompilationInfo and destroyed devices gpuweb/cts#2361, by @Kangz.
-
val: createPipelineAsync and destroyed devices. gpuweb/cts#2360, by @Kangz.
-
val: Check GPUBuffer.mapState when mapped for read. gpuweb/cts#2359, by @Kangz.
-
val: check OOB setBindGroup dynamicOffsetsDataStart gpuweb/cts#2357, by @Kangz.
-
Check GPUAdapterInfo.isFallbackAdapter is present and a boolean gpuweb/cts#2356, by @Kangz.
-
op/val: requestAdapterInfo gpuweb/cts#2355, by @Kangz.
-
Check the value of GPUDeviceLost.reason gpuweb/cts#2354, by @Kangz.
-
Check the value of GPUPipelineError.reason gpuweb/cts#2353, by @Kangz.
-
Missing label coverage gpuweb/cts#2352, by @Kangz.
-
op: sampling,address_mode gpuweb/cts#2350, by @greggman.
-
@kainino0x One sad fact about `pull_request_target`: although it is a clean trigger for building PRs (and it should replace the old `continuous-integration/travis-ci/pr`, I built by taking it as reference [assuming replacing it means having the Actions as a check on PRs]), it will only propagate to new PRs (though it doesn't take the changes in the PR into consideration, making the execution still safe for secret key purposes). Sorry about the problematic execution BTW, explained in the fix PR. gpuweb/cts#2349, by @Yowyaz.
-
Update `asin` accuracy to match spec gpuweb/cts#2345, by @zoddicus.
-
wgsl: bitcast is not a keyword gpuweb/cts#2333, by @dneto0.
-
-
31 issues closed.
-
op: web_platform,external_texture,video crop, rotate, scale, flip gpuweb/cts#2269.
-
wgsl: execution test for `arrayLength` builtin gpuweb/cts#1201.
-
depth_clip_clamp test is not actually testing unclippedDepth? gpuweb/cts#2460.
-
wgsl: writes to storage buffers do not touch padding in matrices gpuweb/cts#2250.
-
wgsl: Weaken min,max,clamp on denorm inputs gpuweb/cts#2403.
-
Adjust expections for `frexp` on subnormals gpuweb/cts#2446.
-
ULP calculations return Scalars instead of numbers gpuweb/cts#2436.
-
update tests for GPUDeviceLostReason change gpuweb/cts#2431.
-
Replace use of two hex u32s to represent f64s with BigInt gpuweb/cts#2437.
-
`f32` conversion for `f32` inputs do no correctly account for FTZ gpuweb/cts#2435.
-
wgsl: comparison expressions. i32, u32 gpuweb/cts#2404.
-
f32 gpuweb/cts#2406.
-
wgsl: add signed integer overloads for the sign builtin gpuweb/cts#2057.
-
wgsl: reserve common scalar types and 'quat' gpuweb/cts#2220.
-
wgsl: execution test for frexp for scalar f32 gpuweb/cts#1208.
-
wgsl: execution test for frexp for vector f32 gpuweb/cts#1210.
-
rename frexp result significand field from .sig to .fract gpuweb/cts#1982.
-
wgsl: Add basic flow control tests gpuweb/cts#2312.
-
`discard` statements gpuweb/cts#2391.
-
Provide ability to dump WGSL shaders gpuweb/cts#1479.
-
Update `acos` accuracy to match spec gpuweb/cts#2371.
-
Update `asin` accuracy to match spec gpuweb/cts#2345.
-
Rename compilationInfo to getCompilationInfo gpuweb/cts#2290.
-
val: createPipelineAsync and destroyed devices. gpuweb/cts#2360.
-
op: sampling,address_mode gpuweb/cts#2350.
-
-
28 issues received 50 new comments.
-
Test restrictions on limits that can be exposed by adapters gpuweb/cts#2277 (5 comments by @greggman, @kainino0x)
-
web_platform: canvas,getCurrentTexture gpuweb/cts#2385 (4 comments by @greggman, @kainino0x, @shrekshao)
-
op: web_platform,external_texture,video crop, rotate, scale, flip gpuweb/cts#2269 (2 comments by @bjjones, @kainino0x)
-
depth_clip_clamp test is not actually testing unclippedDepth? gpuweb/cts#2460 (2 comments by @dneto0, @kainino0x)
-
op: shader_module,compilation_info with utf16 messages gpuweb/cts#2266 (2 comments by @greggman, @kainino0x)
-
Add and/or use additional `expect equals` helpers to the fixture gpuweb/cts#2402 (1 comment by @kainino0x)
-
web_platform: copyEI2T with a source rectangle OOB with the source gpuweb/cts#2365 (1 comment by @shaoboyan)
-
web_platform: different canvas types getCurrentTexture expiry test cases gpuweb/cts#2384 (1 comment by @shrekshao)
-
val: state,device_lost,destroy:{createRenderPipelineAsync, createComputePipelineAsync} gpuweb/cts#2265 (1 comment by @Kangz)
-
87 pull requests opened.
-
Fix issues in LimitTestsImpl gpuweb/cts#2472, by @greggman.
-
Fix maxStorageTexturesPerShaderStage validation tests gpuweb/cts#2471, by @haoxli.
-
Replace Point/Binary/Ternary with Scalar{Pair/Triple} gpuweb/cts#2470, by @zoddicus.
-
Fix fragment shader in maxInterStageShaderComponents validation gpuweb/cts#2469, by @haoxli.
-
Refactor how floating point intervals are calculated gpuweb/cts#2468, by @zoddicus.
-
Fix Occlusion Query Op Test gpuweb/cts#2467, by @greggman.
-
Slightly expand sample_mask tests gpuweb/cts#2466, by @kainino0x.
-
[wgsl] Add execution tests for arrayLength gpuweb/cts#2465, by @jrprice.
-
Fix non-zero queryOffset in occlusionQuery.spec.ts gpuweb/cts#2463, by @haoxli.
-
Set depthClearValue when depthLoadOp is clear gpuweb/cts#2462, by @Jiawei-Shao.
-
Fix unexpected failure in api,validation,queue,copyToTexture,CopyExte… gpuweb/cts#2461, by @shaoboyan.
-
Add more increment and decrement tests. gpuweb/cts#2459, by @dj2.
-
Improve requestAdapterInfo error messages. gpuweb/cts#2458, by @bryceshang.
-
Fixes remaining tests that hit deprecated paths. gpuweb/cts#2457, by @lokokung.
-
Use (number | IntervalBounds) for specifying unittests gpuweb/cts#2456, by @zoddicus.
-
Update tests for min/max spec changes gpuweb/cts#2455, by @zoddicus.
-
[wgsl] Test preservation of matrix padding bytes gpuweb/cts#2454, by @jrprice.
-
Adjust acceptance intervals for frexp gpuweb/cts#2450, by @zoddicus.
-
Cleanup ULP related code gpuweb/cts#2449, by @zoddicus.
-
[wgsl] Add shadow execution tests. gpuweb/cts#2448, by @dj2.
-
[wgsl] Add some float parsing tests. gpuweb/cts#2447, by @dj2.
-
Set depthWriteEnabled to false when depth stencil format is stencil8 gpuweb/cts#2445, by @Jiawei-Shao.
-
Add scalar-vector and vector-scalar f32 arithmetic tests gpuweb/cts#2444, by @zoddicus.
-
[wgsl] More matrix compound expression tests. gpuweb/cts#2443, by @dj2.
-
Validate that atomic builtins are only valid in compute and fragment … gpuweb/cts#2442, by @amaiorano.
-
Replace u32 pairs for representing f64 hex values with BigInt gpuweb/cts#2439, by @zoddicus.
-
Return correctlyRoundedInterval for f32->f32 conversions gpuweb/cts#2438, by @zoddicus.
-
Update for GPUDeviceLostReason spec change gpuweb/cts#2434, by @kainino0x.
-
Fix a missing "enable f16" in maxComputeWorkgroupStorageSize test gpuweb/cts#2433, by @jiangzhaoming.
-
bitcast execution: integral 32-bit types and vectors of them gpuweb/cts#2428, by @dneto0.
-
[wgsl] Make compound operators explicit. gpuweb/cts#2426, by @dj2.
-
[wgsl] Compound assignment for boolean operations. gpuweb/cts#2425, by @dj2.
-
[wgsl] Cleanup compound statement tests gpuweb/cts#2424, by @dj2.
-
webgpu/shader/execution: Code refactoring gpuweb/cts#2423, by @ben-clayton.
-
[wgsl] Add compound operator tests. gpuweb/cts#2422, by @dj2.
-
Fix render_pass,store_op_and_read_only test gpuweb/cts#2421, by @greggman.
-
Update target in tsconfig.json to 'es2020' gpuweb/cts#2420, by @zoddicus.
-
More flow ctrl gpuweb/cts#2419, by @ben-clayton.
-
shader/execution: Add a continue in switch flow control test gpuweb/cts#2418, by @ben-clayton.
-
Bump to es2020 gpuweb/cts#2417, by @zoddicus.
-
Clean up enums gpuweb/cts#2415, by @greggman.
-
Fix for maxColorAttachmentsBytesPerSample test gpuweb/cts#2414, by @greggman.
-
Fixes bug in maxColorAtthacmentBytesPerSample limits test gpuweb/cts#2413, by @greggman.
-
Make all the limits test have no sub cases gpuweb/cts#2412, by @greggman.
-
fix(gen_wpt_cts_html): use non-zero exit code on incorrect arg count gpuweb/cts#2411, by @ErichDonGubler.
-
Add External Texture VideoFrame visibileRect Tests gpuweb/cts#2409, by @bjjones.
-
Test sign(i32) builtin gpuweb/cts#2408, by @dneto0.
-
Rename f32 comparisons gpuweb/cts#2407, by @dneto0.
-
wgsl: add i32, u32 comparisons gpuweb/cts#2405, by @dneto0.
-
Add ImageBitmapRenderingContext canvas for copyToTexture tests gpuweb/cts#2400, by @shaoboyan.
-
Bitwise shift validation gpuweb/cts#2399, by @amaiorano.
-
Add bitwise shift tests gpuweb/cts#2397, by @amaiorano.
-
Validate maxFragmentCombinedOutputResources gpuweb/cts#2396, by @greggman.
-
Add tests for unary~ i32, u32, and vectors of them gpuweb/cts#2395, by @dneto0.
-
[wgsl] Add increment and decrement execution tests. gpuweb/cts#2394, by @dj2.
-
Add sparse color attachment cases to sample_mask alphaToCoverage test gpuweb/cts#2393, by @shrekshao.
-
Modify zero init tests to pollute workgroup memory gpuweb/cts#2392, by @alan-baker.
-
shader/execution: Flip eval order of assignment gpuweb/cts#2390, by @ben-clayton.
-
Test destroyed buffers and textures on submit gpuweb/cts#2388, by @greggman.
-
Adds stale adapter tests gpuweb/cts#2387, by @lokokung.
-
Get current texture expiry gpuweb/cts#2386, by @shrekshao.
-
Sync keyword/reserved word list. gpuweb/cts#2383, by @dj2.
-
Implement `f32` builtin tests gpuweb/cts#2382, by @zoddicus.
-
shader/execution: Add more complex flow control tests gpuweb/cts#2381, by @ben-clayton.
-
Implement `i32` builtin tests gpuweb/cts#2379, by @zoddicus.
-
shader/execution: Add 'preventValueOptimizations' case parameterization gpuweb/cts#2378, by @ben-clayton.
-
Start adding queue validation tests for buffers and textures gpuweb/cts#2377, by @greggman.
-
shader/execution: Port evaluation order tests to flow_control gpuweb/cts#2373, by @ben-clayton.
-
Implement absolute error for `acos` intervals gpuweb/cts#2372, by @zoddicus.
-
Test import GPUExternalTexture from destroyed device gpuweb/cts#2351, by @shaoboyan.
-
Add validation tests on BGRA8Unorm canvas with StorageBinding usage gpuweb/cts#2348, by @Jiawei-Shao.
-
Test Occlusion Queries gpuweb/cts#2347, by @greggman.
-
Implement absolute error for `asin` intervals gpuweb/cts#2346, by @zoddicus.
-
Implement `u32` builtin tests gpuweb/cts#2344, by @zoddicus.
-
Fix wrong conditions and missing formats for maxAttachmentState tests. gpuweb/cts#2343, by @lokokung.
-
Changes for depthWriteEnabled/depthCompare/depthClearValue being required gpuweb/cts#2342, by @kainino0x.
-
Require npm >=8.5.2 gpuweb/cts#2340, by @kainino0x.
-
Updates max*BufferBindingSize limit tests to adhere to maxBufferSize gpuweb/cts#2339, by @lokokung.
-
Add isZeroCopy cases for ExternalTexture gpuweb/cts#2338, by @jchen10.
-
Generate OOM error using createTexture gpuweb/cts#2337, by @kainino0x.
-
Implement `frexp` f32 tests gpuweb/cts#2336, by @zoddicus.
-
Check maxBindingsPerBindGroup limit matches spec limits gpuweb/cts#2335, by @greggman.
-
Roll @webgpu/types 48dafa2 -> d1b649b gpuweb/cts#2334, by @takahirox.
-
Depend on github.com/gpuweb/types instead of npm @webgpu/types gpuweb/cts#2332, by @kainino0x.
-
Removes OOM errors in error scope tests since we can't cause OOM. gpuweb/cts#2331, by @lokokung.
-
-
87 pull requests merged.
-
Fix Occlusion Query Op Test gpuweb/cts#2467.
-
Fix fragment shader in maxInterStageShaderComponents validation gpuweb/cts#2469.
-
Refactor how floating point intervals are calculated gpuweb/cts#2468.
-
Slightly expand sample_mask tests gpuweb/cts#2466.
-
Use (number | IntervalBounds) for specifying unittests gpuweb/cts#2456.
-
[wgsl] Add execution tests for arrayLength gpuweb/cts#2465.
-
Set depthClearValue when depthLoadOp is clear gpuweb/cts#2462.
-
Fixes remaining tests that hit deprecated paths. gpuweb/cts#2457.
-
Add more increment and decrement tests. gpuweb/cts#2459.
-
[wgsl] Test preservation of matrix padding bytes gpuweb/cts#2454.
-
[wgsl] Add shadow execution tests. gpuweb/cts#2448.
-
[wgsl] Add some float parsing tests. gpuweb/cts#2447.
-
Improve requestAdapterInfo error messages. gpuweb/cts#2458.
-
Update tests for min/max spec changes gpuweb/cts#2455.
-
Adjust acceptance intervals for frexp gpuweb/cts#2450.
-
Cleanup ULP related code gpuweb/cts#2449.
-
Update for GPUDeviceLostReason spec change gpuweb/cts#2434.
-
Set depthWriteEnabled to false when depth stencil format is stencil8 gpuweb/cts#2445.
-
Add scalar-vector and vector-scalar f32 arithmetic tests gpuweb/cts#2444.
-
[wgsl] More matrix compound expression tests. gpuweb/cts#2443.
-
Replace u32 pairs for representing f64 hex values with BigInt gpuweb/cts#2439.
-
Return correctlyRoundedInterval for f32->f32 conversions gpuweb/cts#2438.
-
Fix render_pass,store_op_and_read_only test gpuweb/cts#2421.
-
Fix a missing "enable f16" in maxComputeWorkgroupStorageSize test gpuweb/cts#2433.
-
[wgsl] Make compound operators explicit. gpuweb/cts#2426.
-
[wgsl] Compound assignment for boolean operations. gpuweb/cts#2425.
-
[wgsl] Cleanup compound statement tests gpuweb/cts#2424.
-
webgpu/shader/execution: Code refactoring gpuweb/cts#2423.
-
[wgsl] Add compound operator tests. gpuweb/cts#2422.
-
Clean up enums gpuweb/cts#2415.
-
Update target in tsconfig.json to 'es2020' gpuweb/cts#2420.
-
More flow ctrl gpuweb/cts#2419.
-
shader/execution: Add a continue in switch flow control test gpuweb/cts#2418.
-
Bump to es2020 gpuweb/cts#2417.
-
Fix for maxColorAttachmentsBytesPerSample test gpuweb/cts#2414.
-
fix(gen_wpt_cts_html): use non-zero exit code on incorrect arg count gpuweb/cts#2411.
-
Fixes bug in maxColorAtthacmentBytesPerSample limits test gpuweb/cts#2413.
-
Test Occlusion Queries gpuweb/cts#2347.
-
Get current texture expiry gpuweb/cts#2386.
-
Add ImageBitmapRenderingContext canvas for copyToTexture tests gpuweb/cts#2400.
-
wgsl: add i32, u32 comparisons gpuweb/cts#2405.
-
Rename f32 comparisons gpuweb/cts#2407.
-
Test sign(i32) builtin gpuweb/cts#2408.
-
Make all the limits test have no sub cases gpuweb/cts#2412.
-
Add External Texture VideoFrame visibileRect Tests gpuweb/cts#2409.
-
Modify zero init tests to pollute workgroup memory gpuweb/cts#2392.
-
Add test case about copying HTMLVideoElement to texture gpuweb/cts#2018.
-
Bitwise shift validation gpuweb/cts#2399.
-
Add tests for unary~ i32, u32, and vectors of them gpuweb/cts#2395.
-
Add bitwise shift tests gpuweb/cts#2397.
-
Validate maxFragmentCombinedOutputResources gpuweb/cts#2396.
-
[wgsl] Add increment and decrement execution tests. gpuweb/cts#2394.
-
Implement `frexp` f32 tests gpuweb/cts#2336.
-
Implement `determinant` f32 tests gpuweb/cts#2321.
-
Adds stale adapter tests gpuweb/cts#2387.
-
Test destroyed buffers and textures on submit gpuweb/cts#2388.
-
Implement absolute error for `acos` intervals gpuweb/cts#2372.
-
Implement absolute error for `asin` intervals gpuweb/cts#2346.
-
shader/execution: Flip eval order of assignment gpuweb/cts#2390.
-
shader/execution: Port evaluation order tests to flow_control gpuweb/cts#2373.
-
Add isZeroCopy cases for ExternalTexture gpuweb/cts#2338.
-
shader/execution: Add more complex flow control tests gpuweb/cts#2381.
-
Implement `bool` builtin tests gpuweb/cts#2327.
-
Implement `i32` builtin tests gpuweb/cts#2379.
-
Implement `u32` builtin tests gpuweb/cts#2344.
-
Test import GPUExternalTexture from destroyed device gpuweb/cts#2351.
-
Sync keyword/reserved word list. gpuweb/cts#2383.
-
Implement `f32` builtin tests gpuweb/cts#2382.
-
Start adding queue validation tests for buffers and textures gpuweb/cts#2377.
-
Put the same URL in all "Open" links gpuweb/cts#2316.
-
Updates max*BufferBindingSize limit tests to adhere to maxBufferSize gpuweb/cts#2339.
-
Add validation tests on BGRA8Unorm canvas with StorageBinding usage gpuweb/cts#2348.
-
Add validation tests for maxComputeXXX gpuweb/cts#2300.
-
Rename maximumLimit to adapterLimit and related changes gpuweb/cts#2330.
-
Check maxBindingsPerBindGroup limit matches spec limits gpuweb/cts#2335.
-
Require npm >=8.5.2 gpuweb/cts#2340.
-
Generate OOM error using createTexture gpuweb/cts#2337.
-
Update external_texture,expire test case gpuweb/cts#2283.
-
Removes OOM errors in error scope tests since we can't cause OOM. gpuweb/cts#2331.
-
Rename to GPUShaderModule.getCompilationInfo() gpuweb/cts#2291.
-
Depend on github.com/gpuweb/types instead of npm @webgpu/types gpuweb/cts#2332.
-
-
2 pull requests closed.
-
Fix non-zero queryOffset in occlusionQuery.spec.ts gpuweb/cts#2463.
-
Roll @webgpu/types 48dafa2 -> d1b649b gpuweb/cts#2334.
-
-
55 pull requests received 346 new comments.
-
Get current texture expiry gpuweb/cts#2386 (36 comments by @austinEng, @greggman, @kainino0x, @shrekshao)
-
Use (number | IntervalBounds) for specifying unittests gpuweb/cts#2456 (21 comments by @ben-clayton, @zoddicus)
-
[wgsl] Add increment and decrement execution tests. gpuweb/cts#2394 (17 comments by @ben-clayton, @dj2, @zoddicus)
-
Start adding queue validation tests for buffers and textures gpuweb/cts#2377 (13 comments by @greggman, @lokokung)
-
Validate that atomic builtins are only valid in compute and fragment … gpuweb/cts#2442 (11 comments by @amaiorano, @dj2)
-
Add ImageBitmapRenderingContext canvas for copyToTexture tests gpuweb/cts#2400 (11 comments by @kainino0x, @shaoboyan)
-
Add test case about copying HTMLVideoElement to texture gpuweb/cts#2018 (10 comments by @kainino0x, @shaoboyan)
-
Refactor how floating point intervals are calculated gpuweb/cts#2468 (9 comments by @ben-clayton, @zoddicus)
-
Removes OOM errors in error scope tests since we can't cause OOM. gpuweb/cts#2331 (9 comments by @austinEng, @kainino0x, @Kangz, @lokokung)
-
Add External Texture VideoFrame visibileRect Tests gpuweb/cts#2409 (7 comments by @austinEng, @bjjones, @kainino0x)
-
Add tests for unary~ i32, u32, and vectors of them gpuweb/cts#2395 (7 comments by @dneto0, @zoddicus)
-
Updates max*BufferBindingSize limit tests to adhere to maxBufferSize gpuweb/cts#2339 (7 comments by @greggman, @lokokung)
-
Move "device_destroyed_before_importing_video_frame" case to "api,validation,states,device_lost,destroy" gpuweb/cts#2464 (5 comments by @kainino0x, @shaoboyan)
-
shader/execution: Port evaluation order tests to flow_control gpuweb/cts#2373 (5 comments by @amaiorano, @ben-clayton)
-
bitcast execution: integral 32-bit types and vectors of them gpuweb/cts#2428 (4 comments by @zoddicus)
-
Fix writable binding aliasing in webgpu:api,validation,resource_usages,texture,in_pass_encoder gpuweb/cts#2341 (4 comments by @kainino0x, @shrekshao)
-
Fix non-zero queryOffset in occlusionQuery.spec.ts gpuweb/cts#2463 (3 comments by @greggman, @haoxli)
-
Add validation tests on BGRA8Unorm canvas with StorageBinding usage gpuweb/cts#2348 (3 comments by @jiangzhaoming)
-
Depend on github.com/gpuweb/types instead of npm @webgpu/types gpuweb/cts#2332 (3 comments by @kainino0x, @takahirox)
-
[wgsl] Test preservation of matrix padding bytes gpuweb/cts#2454 (2 comments by @ben-clayton, @jrprice)
-
Add scalar-vector and vector-scalar f32 arithmetic tests gpuweb/cts#2444 (2 comments by @amaiorano, @zoddicus)
-
[wgsl] Compound assignment for boolean operations. gpuweb/cts#2425 (2 comments by @ben-clayton, @dj2)
-
Fix fragment shader in maxInterStageShaderComponents validation gpuweb/cts#2469 (1 comment by @greggman)
-
Fix unexpected failure in api,validation,queue,copyToTexture,CopyExte… gpuweb/cts#2461 (1 comment by @kainino0x)
-
Fix a missing "enable f16" in maxComputeWorkgroupStorageSize test gpuweb/cts#2433 (1 comment by @jiangzhaoming)
-
Changes for depthWriteEnabled/depthCompare/depthClearValue being required gpuweb/cts#2342 (1 comment by @kainino0x)
-
TR publications
-
18 documents published.
-
WebGPU published 6 times as a Working Draft. History of WebGPU
-
WebGPU Shading Language published 12 times as a Working Draft. History of WebGPU Shading Language
-
Participants
-
1 participant left.
-
Jerran Schmidt (Autodesk) left.
-
GitHub
gpuweb/cts
-
45 issues opened.
-
rename LimitTestImpl.maximumLimit to adapterLimit gpuweb/cts#2328, by @greggman.
-
wpt: rendering to hidden canvas gpuweb/cts#2326, by @austinEng.
-
stress: bind group allocation gpuweb/cts#2324, by @Kangz.
-
op:canvas Try getting textures on a destroyed / lost device gpuweb/cts#2323, by @Kangz.
-
op:viewport Add tests for various ways to set the viewport and scissor size to 0 gpuweb/cts#2322, by @Kangz.
-
add requires directive gpuweb/cts#2317, by @dneto0.
-
wgsl: Add basic flow control tests gpuweb/cts#2312, by @ben-clayton.
-
wgsl: array element count is an unrestricted expression gpuweb/cts#2308, by @dneto0.
-
wgsl: types, type-generators are not keywords gpuweb/cts#2307, by @dneto0.
-
wgsl: @must_use attribute gpuweb/cts#2306, by @dneto0.
-
wgsl: conversion expressions gpuweb/cts#2302, by @zoddicus.
-
Test that GPUPipelineError's name is GPUPipelineError gpuweb/cts#2297, by @Kangz.
-
`Expected` output should also display value in hex gpuweb/cts#2295, by @amaiorano.
-
Support `vertex` and `fragment` shaders for builtin tests gpuweb/cts#2294, by @zoddicus.
-
Rename compilationInfo to getCompilationInfo gpuweb/cts#2290, by @takahirox.
-
val: texture capabilities (blending, filtering) gpuweb/cts#2282, by @kainino0x.
-
op: update color_target_state tests kBlendableFormats gpuweb/cts#2281, by @kainino0x.
-
Test restrictions on limits that can be exposed by adapters gpuweb/cts#2277, by @kainino0x.
-
Refactor floating point Case generation to follow integer patterns gpuweb/cts#2276, by @zoddicus.
-
val: bgra8unorm-storage gpuweb/cts#2274, by @Jiawei-Shao.
-
op: rendering,storage gpuweb/cts#2271, by @austinEng.
-
op: pipeline,layout:visibility_none gpuweb/cts#2270, by @austinEng.
-
op: web_platform,external_texture,video crop, rotate, scale, flip gpuweb/cts#2269, by @austinEng.
-
wgsl: add validation and execution tests for all storage texture formats gpuweb/cts#2268, by @austinEng.
-
op: queue,writeTexture gpuweb/cts#2267, by @austinEng.
-
op: shader_module,compilation_info with utf16 messages gpuweb/cts#2266, by @austinEng.
-
val: buffer,mapping:mapAsync,state,{destroyed, mappedAtCreation} gpuweb/cts#2264, by @austinEng.
-
val: encoding,cmds,setBindGroup:minBindingSize gpuweb/cts#2263, by @austinEng.
-
op: sampling,compare gpuweb/cts#2262, by @austinEng.
-
op: render_pass,clear_value gpuweb/cts#2261, by @austinEng.
-
op: rendering,scissor_rect gpuweb/cts#2260, by @austinEng.
-
op: sampling,address_mode gpuweb/cts#2258, by @austinEng.
-
op: render_pass,depth_stencil:{depthReadOnly,stencilReadOnly} gpuweb/cts#2257, by @austinEng.
-
op: command_buffer,clear_buffer:defaults gpuweb/cts#2256, by @austinEng.
-
Is it correct to both finish and execute in createEncoder WRT validation tests? gpuweb/cts#2255, by @greggman.
-
`F32Vector` and `F32Matrix` could store if they are contain only finite values gpuweb/cts#2253, by @zoddicus.
-
Export or move the WebGPU CTS into WPT gpuweb/cts#2252, by @kainino0x.
-
wgsl: writes to storage buffers do not touch padding in matrices gpuweb/cts#2250, by @dneto0.
-
val: test allowed lengths of color/origin/extent sequences gpuweb/cts#2225, by @kainino0x.
-
-
28 issues closed.
-
op: render pipeline culling gpuweb/cts#2185.
-
op: labels gpuweb/cts#1006.
-
val: createTexture: sampleCount is invalid with 1d/3d textures gpuweb/cts#1078.
-
val: copyImageBitmapToTexture should only allow 2d textures gpuweb/cts#2108.
-
reftests: test canvas colorSpace gpuweb/cts#918.
-
reftest: rendering via transferControlToOffscreen gpuweb/cts#941.
-
reftest: canvas sizing with ResizeObserver and checkerboard pattern gpuweb/cts#1115.
-
reftest: canvas scaling gpuweb/cts#1116.
-
op: shader module sourceMap gpuweb/cts#2209.
-
Test that GPUPipelineError's name is GPUPipelineError gpuweb/cts#2297.
-
`Expected` output should also display value in hex gpuweb/cts#2295.
-
val: bgra8unorm-storage gpuweb/cts#2274.
-
wgsl: Validate `@const` is not permitted gpuweb/cts#1438.
-
wgsl: validate `break` statement gpuweb/cts#1671.
-
wgsl: rename static_assert to const_assert gpuweb/cts#2178.
-
wsgl: disallow repeated use of builtin inputs and builtin outputs gpuweb/cts#1712.
-
wgsl: rename staticAssert to static_assert gpuweb/cts#1714.
-
wgsl: disallow duplicated pipeline builtin inputs, and outputs gpuweb/cts#1745.
-
wgsl: 'invariant' is no longer a reserved word gpuweb/cts#1817.
-
wgsl: validate pipeline stage attributes gpuweb/cts#1464.
-
op: render_pipeline,sample_mask gpuweb/cts#2146.
-
wgsl: some @attrib parameters can be expressions gpuweb/cts#1716.
-
wgsl: Validate `@align` parsing gpuweb/cts#1432.
-
web_platform: context_creation from transferControlToOffscreen gpuweb/cts#908.
-
Add support for Matrices to Type/Value system gpuweb/cts#2029.
-
-
25 issues received 34 new comments.
-
Test restrictions on limits that can be exposed by adapters gpuweb/cts#2277 (5 comments by @greggman, @kainino0x)
-
Track all the buffers and textures with TrackForCleanup() gpuweb/cts#1752 (2 comments by @Jiawei-Shao, @kainino0x)
-
reftest: rendering via transferControlToOffscreen gpuweb/cts#941 (2 comments by @greggman, @kainino0x)
-
op: web_platform,external_texture,video crop, rotate, scale, flip gpuweb/cts#2269 (1 comment by @bjjones)
-
Is it correct to both finish and execute in createEncoder WRT validation tests? gpuweb/cts#2255 (1 comment by @kainino0x)
-
op: pipeline,default_layout:layout - default layout is computed correctly gpuweb/cts#924 (1 comment by @austinEng)
-
`F32Vector` and `F32Matrix` could store if they are contain only finite values gpuweb/cts#2253 (1 comment by @zoddicus)
-
wsgl: disallow repeated use of builtin inputs and builtin outputs gpuweb/cts#1712 (1 comment by @dj2)
-
web_platform: context_creation from transferControlToOffscreen gpuweb/cts#908 (1 comment by @kainino0x)
-
-
64 pull requests opened.
-
Rename maximumLimit to adapterLimit and related changes gpuweb/cts#2330, by @greggman.
-
Updates BGL sampleType when using multisampling for validation errors. gpuweb/cts#2329, by @lokokung.
-
Implement `bool` builtin tests gpuweb/cts#2327, by @zoddicus.
-
Implement `determinant` f32 tests gpuweb/cts#2321, by @zoddicus.
-
OperationError -> GPUPipelineError gpuweb/cts#2320, by @greggman.
-
Fix WGSL warning for maxStorageTexturesPerShaderStage.spec.ts gpuweb/cts#2319, by @greggman.
-
Less leaving prompt gpuweb/cts#2318, by @greggman.
-
Put the same URL in all "Open" links gpuweb/cts#2316, by @greggman.
-
createXXXPiplelineAsync should return GPUPipelineError gpuweb/cts#2315, by @greggman.
-
Convert more OperationError to GPUPipelineError. gpuweb/cts#2314, by @Kangz.
-
shader/execution: Begin flow control tests gpuweb/cts#2313, by @ben-clayton.
-
Only pass depth-stencil load/store values when needed. gpuweb/cts#2311, by @lokokung.
-
Test minStorageBufferOffsetAlignment and minUniformBufferOffsetAlignment gpuweb/cts#2310, by @greggman.
-
Test maxDynamic(Uniform|Storage)BuffersPerPipelineLayout gpuweb/cts#2309, by @greggman.
-
Test maxInterStageShaderComponents and maxInterStageShaderVariables gpuweb/cts#2305, by @greggman.
-
Add f32 matrix-vector and vector-matrix multiplication tests gpuweb/cts#2304, by @zoddicus.
-
Fix normalization checks gpuweb/cts#2303, by @greggman.
-
Add f32 matrix-matrix multiplication tests gpuweb/cts#2301, by @zoddicus.
-
Add validation tests for maxComputeXXX gpuweb/cts#2300, by @greggman.
-
Add ExternalTexture Rotatation Tests gpuweb/cts#2299, by @bjjones.
-
GPUPipelineError.name is GPUPipelineError and not OperationError. gpuweb/cts#2298, by @Kangz.
-
Emit `f32` values for F32Interval.toString() gpuweb/cts#2296, by @zoddicus.
-
Move `static_assert` and `type` to reserved words. gpuweb/cts#2293, by @dj2.
-
Update tests to use `alias`. gpuweb/cts#2292, by @dj2.
-
Rename to GPUShaderModule.getCompilationInfo() gpuweb/cts#2291, by @takahirox.
-
Test rg11b10ufloat-renderable makes the format resolvable and MSAAable gpuweb/cts#2289, by @Kangz.
-
Validate that case input parameter types match those provided to test runner gpuweb/cts#2288, by @zoddicus.
-
Validate maxColorAttachments and maxColorAttachmentBytesPerSample gpuweb/cts#2287, by @greggman.
-
Add f32 matrix-scalar multiplication tests gpuweb/cts#2286, by @zoddicus.
-
webgpu/shader/validation: Remove types as keywords gpuweb/cts#2284, by @ben-clayton.
-
Update external_texture,expire test case gpuweb/cts#2283, by @shaoboyan.
-
add maxXXXBufferBindingSize limit tests gpuweb/cts#2280, by @greggman.
-
Add f32 matrix subtraction tests gpuweb/cts#2279, by @zoddicus.
-
Add f32 matrix addition tests gpuweb/cts#2278, by @zoddicus.
-
Add validation tests about feature 'bgra8unorm-storage' gpuweb/cts#2275, by @Jiawei-Shao.
-
Upgrade gpuweb/types to 0.1.27 gpuweb/cts#2273, by @shaoboyan.
-
Validate maxVertexBuffers, maxVertexAttributes, maxVertexBufferArrayStride gpuweb/cts#2272, by @greggman.
-
Cleanup Validate binding groups gpuweb/cts#2259, by @greggman.
-
Add more tests for end twice gpuweb/cts#2251, by @Richard-Yunchao.
-
Implement `transpose` f32 tests gpuweb/cts#2249, by @zoddicus.
-
[wgsl] Add placement tests for @builtin. gpuweb/cts#2248, by @dj2.
-
Add validation tests for unary not and binary and/or gpuweb/cts#2247, by @amaiorano.
-
[wgsl] Tests `@const` is not allowed to be used as an attribute. gpuweb/cts#2246, by @dj2.
-
Update builtin parse test to not use code as URL. gpuweb/cts#2245, by @dj2.
-
Update `align` parsing tests. gpuweb/cts#2244, by @dj2.
-
Update to use `keysOf` helper. gpuweb/cts#2243, by @dj2.
-
Use new test video files for ExternalTexture gpuweb/cts#2242, by @jchen10.
-
[wgsl] Add validation tests for `break` placement gpuweb/cts#2240, by @dj2.
-
[wgsl] Add validation of `discard` placement. gpuweb/cts#2239, by @dj2.
-
add maxXXXPerShaderStage limit tests gpuweb/cts#2238, by @greggman.
-
[wgsl] Update `static_assert` to `const_assert`. gpuweb/cts#2237, by @dj2.
-
add maxBufferSize limit test gpuweb/cts#2236, by @greggman.
-
Validate binding groups gpuweb/cts#2235, by @greggman.
-
Validate texture array layers gpuweb/cts#2234, by @greggman.
-
[wgsl] Add validation tests for pipeline stage attributes. gpuweb/cts#2233, by @dj2.
-
Add a few more constant/override tests gpuweb/cts#2231, by @greggman.
-
Avoid writable storage buffer binding aliasing gpuweb/cts#2230, by @shrekshao.
-
[wgsl] Validate expressions passed to attributes. gpuweb/cts#2229, by @dj2.
-
Add ExternalTexture Flip/Rotate Tests gpuweb/cts#2228, by @bjjones.
-
Add limit tests for maxTextureDimension1/2/3D gpuweb/cts#2227, by @greggman.
-
Remove tests of end pass twice in encoder_state.spec.ts gpuweb/cts#2226, by @Richard-Yunchao.
-
Fix a few spelling errors gpuweb/cts#2224, by @greggman.
-
Test destroying an invalid queryset and an invalid texture gpuweb/cts#2223, by @greggman.
-
CSS, make button colors match gpuweb/cts#2221, by @greggman.
-
-
60 pull requests merged.
-
Less leaving prompt gpuweb/cts#2318.
-
OperationError -> GPUPipelineError gpuweb/cts#2320.
-
Fix WGSL warning for maxStorageTexturesPerShaderStage.spec.ts gpuweb/cts#2319.
-
Validate maxColorAttachments and maxColorAttachmentBytesPerSample gpuweb/cts#2287.
-
Test maxInterStageShaderComponents and maxInterStageShaderVariables gpuweb/cts#2305.
-
Test maxDynamic(Uniform|Storage)BuffersPerPipelineLayout gpuweb/cts#2309.
-
Convert more OperationError to GPUPipelineError. gpuweb/cts#2314.
-
Only pass depth-stencil load/store values when needed. gpuweb/cts#2311.
-
shader/execution: Begin flow control tests gpuweb/cts#2313.
-
Updates error type for create oom buffers. gpuweb/cts#2217.
-
add maxXXXPerShaderStage limit tests gpuweb/cts#2238.
-
add maxXXXBufferBindingSize limit tests gpuweb/cts#2280.
-
Add f32 matrix-vector and vector-matrix multiplication tests gpuweb/cts#2304.
-
GPUPipelineError.name is GPUPipelineError and not OperationError. gpuweb/cts#2298.
-
Add f32 matrix-matrix multiplication tests gpuweb/cts#2301.
-
Fix normalization checks gpuweb/cts#2303.
-
Add ExternalTexture Rotatation Tests gpuweb/cts#2299.
-
Use new test video files for ExternalTexture gpuweb/cts#2242.
-
Emit `f32` values for F32Interval.toString() gpuweb/cts#2296.
-
Move `static_assert` and `type` to reserved words. gpuweb/cts#2293.
-
Update tests to use `alias`. gpuweb/cts#2292.
-
Add f32 matrix-scalar multiplication tests gpuweb/cts#2286.
-
webgpu/shader/validation: Remove types as keywords gpuweb/cts#2284.
-
Add f32 matrix subtraction tests gpuweb/cts#2279.
-
Add f32 matrix addition tests gpuweb/cts#2278.
-
Add validation tests about feature 'bgra8unorm-storage' gpuweb/cts#2275.
-
Upgrade gpuweb/types to 0.1.27 gpuweb/cts#2273.
-
Cleanup Validate binding groups gpuweb/cts#2259.
-
Validate binding groups gpuweb/cts#2235.
-
Implement `transpose` f32 tests gpuweb/cts#2249.
-
Add more tests for end twice gpuweb/cts#2251.
-
Update to use `keysOf` helper. gpuweb/cts#2243.
-
[wgsl] Tests `@const` is not allowed to be used as an attribute. gpuweb/cts#2246.
-
[wgsl] Add placement tests for @builtin. gpuweb/cts#2248.
-
Add validation tests for unary not and binary and/or gpuweb/cts#2247.
-
Update builtin parse test to not use code as URL. gpuweb/cts#2245.
-
Update `align` parsing tests. gpuweb/cts#2244.
-
[wgsl] Add validation tests for `break` placement gpuweb/cts#2240.
-
Validate texture array layers gpuweb/cts#2234.
-
add maxBufferSize limit test gpuweb/cts#2236.
-
[wgsl] Add validation of `discard` placement. gpuweb/cts#2239.
-
[wgsl] Update `static_assert` to `const_assert`. gpuweb/cts#2237.
-
[wgsl] Add validation tests for pipeline stage attributes. gpuweb/cts#2233.
-
Add limit tests for maxTextureDimension1/2/3D gpuweb/cts#2227.
-
Avoid writable storage buffer binding aliasing gpuweb/cts#2230.
-
[wgsl] Validate expressions passed to attributes. gpuweb/cts#2229.
-
Add a few more constant/override tests gpuweb/cts#2231.
-
[wgsl] Alignment attribute tests. gpuweb/cts#2180.
-
Remove tests of end pass twice in encoder_state.spec.ts gpuweb/cts#2226.
-
Test destroying an invalid queryset and an invalid texture gpuweb/cts#2223.
-
Fix a few spelling errors gpuweb/cts#2224.
-
CSS, make button colors match gpuweb/cts#2221.
-
Culling: Test triangle-strips, depth, and stencil results gpuweb/cts#2186.
-
Add tests for i32 binary scalar-vector ops gpuweb/cts#2219.
-
Implement Matrix type for numeric testing gpuweb/cts#2204.
-
2 pull requests closed.
-
createXXXPiplelineAsync should return GPUPipelineError gpuweb/cts#2315.
-
Add ExternalTexture Flip/Rotate Tests gpuweb/cts#2228.
-
-
38 pull requests received 240 new comments.
-
Add limit tests for maxTextureDimension1/2/3D gpuweb/cts#2227 (28 comments by @greggman, @kainino0x)
-
Use new test video files for ExternalTexture gpuweb/cts#2242 (20 comments by @bjjones, @jchen10, @kainino0x, @shaoboyan)
-
Validate maxColorAttachments and maxColorAttachmentBytesPerSample gpuweb/cts#2287 (15 comments by @austinEng, @greggman)
-
Updates error type for create oom buffers. gpuweb/cts#2217 (14 comments by @austinEng, @kainino0x, @lokokung)
-
Test minStorageBufferOffsetAlignment and minUniformBufferOffsetAlignment gpuweb/cts#2310 (6 comments by @greggman, @Kangz)
-
Test maxInterStageShaderComponents and maxInterStageShaderVariables gpuweb/cts#2305 (6 comments by @greggman, @Kangz)
-
Remove tests of end pass twice in encoder_state.spec.ts gpuweb/cts#2226 (6 comments by @kainino0x, @Richard-Yunchao)
-
Avoid writable storage buffer binding aliasing gpuweb/cts#2230 (3 comments by @kainino0x, @shrekshao)
-
Validate maxVertexBuffers, maxVertexAttributes, maxVertexBufferArrayStride gpuweb/cts#2272 (2 comments by @greggman, @kainino0x)
-
Add validation tests about feature 'bgra8unorm-storage' gpuweb/cts#2275 (2 comments by @jiangzhaoming)
-
Fix WGSL warning for maxStorageTexturesPerShaderStage.spec.ts gpuweb/cts#2319 (1 comment by @greggman)
-
Test rg11b10ufloat-renderable makes the format resolvable and MSAAable gpuweb/cts#2289 (1 comment by @Kangz)
-
gpuweb/gpuweb
-
37 issues opened.
-
wgsl: enable directive should support a list of enable-extensions gpuweb/gpuweb#3885, by @dneto0.
-
Typo in "Texture View Creation" docs gpuweb/gpuweb#3882, by @rconde01.
-
add query for list of WGSL software/shader-lang extensions (e.g. do-while) gpuweb/gpuweb#3875, by @dneto0.
-
Accuracy not specified for `determinant` gpuweb/gpuweb#3873, by @zoddicus.
-
Accuracy not specified for `transpose` gpuweb/gpuweb#3871, by @zoddicus.
-
maxInterStageShaderVariables with regards to the fragment stage gpuweb/gpuweb#3865, by @greggman.
-
maxBindingsPerBindGroup has arbitrary limit of 640 gpuweb/gpuweb#3864, by @mwyrzykowski.
-
HDR ASTC texture feature "texture-compression-astc-hdr" gpuweb/gpuweb#3856, by @mwyrzykowski.
-
wgsl: Allow unrestricted expression for array element count gpuweb/gpuweb#3852, by @dneto0.
-
Would having Float16Array in JS be useful for integration with WebGPU? gpuweb/gpuweb#3848, by @bakkot.
-
wgsl:editorial: "type constructors" are actually "value constructors" gpuweb/gpuweb#3846, by @dneto0.
-
Investigation: `32bit-format-{multisample,resolve}` extensions gpuweb/gpuweb#3844, by @teoxoy.
-
Add `rgb10a2uint` texture format gpuweb/gpuweb#3841, by @teoxoy.
-
Proposal: `norm16-{filterable,resolve}` extensions gpuweb/gpuweb#3839, by @teoxoy.
-
Proposal: read-write storage textures gpuweb/gpuweb#3838, by @teoxoy.
-
Proposal: `formats-tier-1` extension gpuweb/gpuweb#3837, by @teoxoy.
-
`workgroupUniformLoad` not included in "collective operations" list gpuweb/gpuweb#3836, by @jimblandy.
-
TR build broken on latest Bikeshed: duplicate infopanel gpuweb/gpuweb#3835, by @kainino0x.
-
maxStorageBufferBindingSize should be aligned or spec updated gpuweb/gpuweb#3831, by @greggman.
-
treesitter custom scanner: fix TODOs and issues gpuweb/gpuweb#3830, by @dneto0.
-
Describe depth textures gpuweb/gpuweb#3824, by @alan-baker.
-
Should type constructors and bitcasts be moved to built-in functions? gpuweb/gpuweb#3823, by @alan-baker.
-
Reorganize memory sections gpuweb/gpuweb#3820, by @alan-baker.
-
wgsl: Statement must not be only a type-constructor. gpuweb/gpuweb#3819, by @dneto0.
-
wgsl: limit validation gpuweb/gpuweb#3815, by @greggman.
-
wgsl: type constructor expressions should have same execution rules as function calls gpuweb/gpuweb#3813, by @dneto0.
-
Validation for resolveQuerySet() doesn't require the query to have already been written to gpuweb/gpuweb#3812, by @litherum.
-
wgsl CI: add unit tests for tree-sitter grammar gpuweb/gpuweb#3810, by @dneto0.
-
clarify whether commands in a command buffer execute in order gpuweb/gpuweb#3809, by @dneto0.
-
Who actually wants multiple timestampWrites with the same TimestampWrite.location? gpuweb/gpuweb#3808, by @litherum.
-
Binding aliasing algorithms don't account for dynamic offsets gpuweb/gpuweb#3802, by @kainino0x.
-
Support STORAGE usage on BGRA8Unorm as extension instead of core feature gpuweb/gpuweb#3799, by @Jiawei-Shao.
-
Defaults for depthWriteEnabled and depthCompare may be surprising gpuweb/gpuweb#3798, by @kainino0x.
-
What should be the error name for GPUPipelineError gpuweb/gpuweb#3796, by @Kangz.
-
-
30 issues closed.
-
add diagnostic for unrecognized diagnostic rules gpuweb/gpuweb#3714.
-
Should type constructors and bitcasts be moved to built-in functions? gpuweb/gpuweb#3823.
-
Typo in "Texture View Creation" docs gpuweb/gpuweb#3882.
-
consider adding matrix determinant gpuweb/gpuweb#1623.
-
Is it intended that keywords can not be identifiers? gpuweb/gpuweb#3692.
-
wgsl: Allow unrestricted expression for array element count gpuweb/gpuweb#3852.
-
Accuracy not specified for `transpose` gpuweb/gpuweb#3871.
-
wgsl: Statement must not be only a type-constructor. gpuweb/gpuweb#3819.
-
Add extent/origin/color "reification" step gpuweb/gpuweb#1838.
-
Reorganize memory sections gpuweb/gpuweb#3820.
-
What should be the error name for GPUPipelineError gpuweb/gpuweb#3796.
-
wgsl:editorial: "type constructors" are actually "value constructors" gpuweb/gpuweb#3846.
-
Would having Float16Array in JS be useful for integration with WebGPU? gpuweb/gpuweb#3848.
-
`workgroupUniformLoad` not included in "collective operations" list gpuweb/gpuweb#3836.
-
D3D12 does not support copying between textures of different dimensionality gpuweb/gpuweb#3723.
-
WebGPU: Forbid null code point '\0' (U+0000) within strings gpuweb/gpuweb#3576.
-
TR build broken on latest Bikeshed: duplicate infopanel gpuweb/gpuweb#3835.
-
Concretization of abstract expressions can produce surprising types gpuweb/gpuweb#3716.
-
Support STORAGE usage on BGRA8Unorm as extension instead of core feature gpuweb/gpuweb#3799.
-
Lookahead disambiguation of less-than vs template argument list (v2) gpuweb/gpuweb#3770.
-
Are padding bytes between matrix columns preserved for storage buffer writes? gpuweb/gpuweb#3753.
-
wgsl: limit validation gpuweb/gpuweb#3815.
-
abs(INT_MIN) should be well-defined on integral types gpuweb/gpuweb#3785.
-
How do we specify the timing of external texture expiration? gpuweb/gpuweb#3324.
-
-
1 issue reopened.
-
Accuracy not specified for `determinant` gpuweb/gpuweb#3873, by @dneto0.
-
-
60 issues received 185 new comments.
-
Support STORAGE usage on BGRA8Unorm as extension instead of core feature gpuweb/gpuweb#3799 (18 comments by @ben-clayton, @greggman, @Jiawei-Shao, @kainino0x, @teoxoy)
-
Defaults for depthWriteEnabled and depthCompare may be surprising gpuweb/gpuweb#3798 (14 comments by @greggman, @kainino0x, @Kangz, @kdashg, @magcius)
-
It should be possible to unbind a bind group / vertex buffer gpuweb/gpuweb#3787 (10 comments by @kainino0x, @Kangz, @litherum)
-
WebGPU: Forbid null code point '\0' (U+0000) within strings gpuweb/gpuweb#3576 (9 comments by @greggman, @kainino0x)
-
Who actually wants multiple timestampWrites with the same TimestampWrite.location? gpuweb/gpuweb#3808 (8 comments by @kainino0x, @Kangz, @kdashg, @litherum)
-
Attribute syntax should not specialize to built-in language attributes gpuweb/gpuweb#3520 (6 comments by @dneto0, @kdashg, @mehmetoguzderin)
-
wgsl: Statement must not be only a type-constructor. gpuweb/gpuweb#3819 (6 comments by @alan-baker, @ben-clayton, @dneto0, @kdashg)
-
Accuracy not specified for `determinant` gpuweb/gpuweb#3873 (5 comments by @alan-baker, @dneto0, @zoddicus)
-
Combined limit for max vertex stage bind groups + vertex buffer bindings gpuweb/gpuweb#2749 (5 comments by @kainino0x, @kdashg, @magcius)
-
wgsl: language evolution: managing rollout of core (essentially sugar) language features gpuweb/gpuweb#3149 (5 comments by @ben-clayton, @dneto0, @kdashg)
-
Validation for resolveQuerySet() doesn't require the query to have already been written to gpuweb/gpuweb#3812 (4 comments by @kainino0x, @Kangz, @kdashg)
-
Publish WGSL CI artifact to a repo for reusability gpuweb/gpuweb#3778 (4 comments by @kdashg, @mehmetoguzderin)
-
Would having Float16Array in JS be useful for integration with WebGPU? gpuweb/gpuweb#3848 (4 comments by @bakkot, @ben-clayton, @grovesNL, @zoddicus)
-
`workgroupUniformLoad` not included in "collective operations" list gpuweb/gpuweb#3836 (4 comments by @alan-baker, @jimblandy)
-
add query for list of WGSL software/shader-lang extensions (e.g. do-while) gpuweb/gpuweb#3875 (3 comments by @dneto0)
-
Should type constructors and bitcasts be moved to built-in functions? gpuweb/gpuweb#3823 (3 comments by @alan-baker, @mehmetoguzderin)
-
Should the API enforce that writable storage buffer bindings don't alias each other? gpuweb/gpuweb#1842 (2 comments by @kainino0x, @Kangz)
-
wgsl: enable directive should support a list of enable-extensions gpuweb/gpuweb#3885 (2 comments by @litherum, @mehmetoguzderin)
-
Are padding bytes between matrix columns preserved for storage buffer writes? gpuweb/gpuweb#3753 (2 comments by @kdashg, @litherum)
-
wgsl: template list discovery doesn't handle relational expressions in templates, when they have '=' in the token gpuweb/gpuweb#3868 (2 comments by @dneto0)
-
WebGPU Explainer: Example code for retrieving data from a texture rendered on the GPU has issues gpuweb/gpuweb#3858 (2 comments by @ghost, @Kangz)
-
TR build broken on latest Bikeshed: duplicate infopanel gpuweb/gpuweb#3835 (2 comments by @kainino0x)
-
rg11b10ufloat-renderable should probably allow use as a resolve target gpuweb/gpuweb#3566 (2 comments by @kainino0x)
-
wgsl: type constructor expressions should have same execution rules as function calls gpuweb/gpuweb#3813 (2 comments by @ben-clayton, @dneto0)
-
wgsl: Allow unrestricted expression for array element count gpuweb/gpuweb#3852 (1 comment by @dneto0)
-
D3D12 does not support copying between textures of different dimensionality gpuweb/gpuweb#3723 (1 comment by @kainino0x)
-
Optional feature request: filterable 32-bit float textures gpuweb/gpuweb#3601 (1 comment by @kainino0x)
-
Concretization of abstract expressions can produce surprising types gpuweb/gpuweb#3716 (1 comment by @dneto0)
-
Distinguish between pointer and pointer contents for uniformity requirements on parameters gpuweb/gpuweb#3677 (1 comment by @dneto0)
-
publish-TR-webgpu: Attribute “xlink” not allowed on element “svg” at this point gpuweb/gpuweb#3222 (1 comment by @kainino0x)
-
clarify whether commands in a command buffer execute in order gpuweb/gpuweb#3809 (1 comment by @Kangz)
-
Binding aliasing algorithms don't account for dynamic offsets gpuweb/gpuweb#3802 (1 comment by @kainino0x)
-
Interpretation of `sequence<double>` as `GPUColor` is never actually spelled out gpuweb/gpuweb#3793 (1 comment by @Kangz)
-
-
52 pull requests opened.
-
Enable directive takes a list of extensions. gpuweb/gpuweb#3886, by @dneto0.
-
Describe accuracy for determinant gpuweb/gpuweb#3884, by @dneto0.
-
fix `GPUTextureDescriptor` to `GPUTextureViewDescriptor` gpuweb/gpuweb#3883, by @rconde01.
-
Small typo in section 8.3.1. Pipeline Layout Creation gpuweb/gpuweb#3881, by @abdo643-HULK.
-
Fix incorrectly named variable in example 3.2.2 gpuweb/gpuweb#3880, by @typio.
-
Fix GPURenderPassDepthStencilAttachment validation gpuweb/gpuweb#3879, by @takahirox.
-
Fix template list discovery, to recognize embedded comparisons and shifts gpuweb/gpuweb#3878, by @dneto0.
-
Treesitter scanner: fix handling of '=' and shifts in template args gpuweb/gpuweb#3877, by @dneto0.
-
Define "software extensions", and the "requires" directive gpuweb/gpuweb#3874, by @dneto0.
-
Define floating point accuracy for `transpose` gpuweb/gpuweb#3872, by @zoddicus.
-
Fix case of `Hullshader` in unreserve list. gpuweb/gpuweb#3870, by @dj2.
-
wgsl: Fix cross-reference to GPUExternalTexture gpuweb/gpuweb#3869, by @dneto0.
-
Fix tabs in wgsl/index.bs gpuweb/gpuweb#3863, by @dneto0.
-
Add @must_use attribute gpuweb/gpuweb#3862, by @dneto0.
-
Fix intro to "Memory" section, to refer to Memory Views gpuweb/gpuweb#3861, by @dneto0.
-
Make reserved words list more consistent gpuweb/gpuweb#3860, by @dj2.
-
Rename to getCompilationInfo() gpuweb/gpuweb#3857, by @takahirox.
-
Remove some fixed or unnecessary inline issues. gpuweb/gpuweb#3855, by @Kangz.
-
Validate the shape of Extent/Origin/Color sequence variants. gpuweb/gpuweb#3854, by @Kangz.
-
[editorial] Move value constructors and bitcasts to built-in functions gpuweb/gpuweb#3853, by @alan-baker.
-
[editorial] Reorganize memory sections gpuweb/gpuweb#3851, by @alan-baker.
-
Add wgsl to install-dependencies.sh args in spec/README.md gpuweb/gpuweb#3850, by @takahirox.
-
Make depthClearValue, depthWriteEnabled, depthCompare required gpuweb/gpuweb#3849, by @takahirox.
-
Rename "type constructor" -> "value constructor" gpuweb/gpuweb#3847, by @dneto0.
-
wgsl: Concretization must be concrete gpuweb/gpuweb#3845, by @dneto0.
-
Clarify all sync built-in functions must only be in uniform control flow gpuweb/gpuweb#3843, by @alan-baker.
-
Add longer WGSL example to top of WGSL spec gpuweb/gpuweb#3842, by @toji.
-
UAs MUST restrict to <=32 distinguishable [fingerprinting] buckets. gpuweb/gpuweb#3840, by @kdashg.
-
Make rg11b10ufloat-renderable also allow resolving gpuweb/gpuweb#3834, by @kainino0x.
-
Add column for whether format is blendable gpuweb/gpuweb#3833, by @kainino0x.
-
[editorial] Chapters 8+ gpuweb/gpuweb#3832, by @alan-baker.
-
Add maxFragmentCombinedOutputResources limit gpuweb/gpuweb#3829, by @kainino0x.
-
Add feature for filterable *32float textures gpuweb/gpuweb#3828, by @kainino0x.
-
Update checks for dictionary members to use "exists" gpuweb/gpuweb#3827, by @kainino0x.
-
Change explainer status to "Draft Community Group Report" gpuweb/gpuweb#3826, by @kainino0x.
-
[editorial] Chapters 1-7 gpuweb/gpuweb#3825, by @alan-baker.
-
Merge dfns in headings with the headings themselves gpuweb/gpuweb#3822, by @kainino0x.
-
Improve validation language around timestampWrites gpuweb/gpuweb#3821, by @kainino0x.
-
Prevent matrix padding from being modified gpuweb/gpuweb#3818, by @alan-baker.
-
wgsl: Fix capitalization of section headers gpuweb/gpuweb#3817, by @dneto0.
-
Distinguish between address and contents uniformity gpuweb/gpuweb#3816, by @alan-baker.
-
Say that type constructors act like function calls gpuweb/gpuweb#3814, by @dneto0.
-
wgsl: Fix description of truncate and trunc gpuweb/gpuweb#3811, by @dneto0.
-
wgsl: Regenerate HTML when diagram sources change gpuweb/gpuweb#3807, by @dneto0.
-
Make regex single-line comment gpuweb/gpuweb#3806, by @mehmetoguzderin.
-
Rename `compilationInfo()` to `createCompilationInfo()` gpuweb/gpuweb#3805, by @litherum.
-
Rename `getBindGroupLayout()` to `createBindGroupLayout()` gpuweb/gpuweb#3804, by @litherum.
-
Fix outdated `maxTextureArrayLayers` description gpuweb/gpuweb#3801, by @teoxoy.
-
Support STORAGE usage on bgra8unorm as an extension gpuweb/gpuweb#3800, by @Jiawei-Shao.
-
grammar analyzer: Fix first and follow sets for degenerate rules gpuweb/gpuweb#3797, by @dneto0.
-
-
41 pull requests merged.
-
Fix GPURenderPassDepthStencilAttachment validation gpuweb/gpuweb#3879.
-
Treesitter scanner: fix handling of '=' and shifts in template args gpuweb/gpuweb#3877.
-
Describe accuracy for determinant gpuweb/gpuweb#3884.
-
Define "software extensions", and the "requires" directive gpuweb/gpuweb#3874.
-
[editorial] Move value constructors and bitcasts to built-in functions gpuweb/gpuweb#3853.
-
Fix template list discovery, to recognize embedded comparisons and shifts gpuweb/gpuweb#3878.
-
fix `GPUTextureDescriptor` to `GPUTextureViewDescriptor` gpuweb/gpuweb#3883.
-
Small typo in section 8.3.1. Pipeline Layout Creation gpuweb/gpuweb#3881.
-
Fix incorrectly named variable in example 3.2.2 gpuweb/gpuweb#3880.
-
Add wgsl to install-dependencies.sh args in spec/README.md gpuweb/gpuweb#3850.
-
Define floating point accuracy for `transpose` gpuweb/gpuweb#3872.
-
Add @must_use attribute gpuweb/gpuweb#3862.
-
Fix case of `Hullshader` in unreserve list. gpuweb/gpuweb#3870.
-
wgsl: Fix cross-reference to GPUExternalTexture gpuweb/gpuweb#3869.
-
Fix intro to "Memory" section, to refer to Memory Views gpuweb/gpuweb#3861.
-
Fix tabs in wgsl/index.bs gpuweb/gpuweb#3863.
-
Make reserved words list more consistent gpuweb/gpuweb#3860.
-
Validate the shape of Extent/Origin/Color sequence variants. gpuweb/gpuweb#3854.
-
Remove some fixed or unnecessary inline issues. gpuweb/gpuweb#3855.
-
Rename to getCompilationInfo() gpuweb/gpuweb#3857.
-
Add longer WGSL example to top of WGSL spec gpuweb/gpuweb#3842.
-
[editorial] Reorganize memory sections gpuweb/gpuweb#3851.
-
Rename "type constructor" -> "value constructor" gpuweb/gpuweb#3847.
-
wgsl: Concretization must be concrete gpuweb/gpuweb#3845.
-
[editorial] Chapters 1-7 gpuweb/gpuweb#3825.
-
Clarify all sync built-in functions must only be in uniform control flow gpuweb/gpuweb#3843.
-
[editorial] Chapters 8+ gpuweb/gpuweb#3832.
-
Add column for whether format is blendable gpuweb/gpuweb#3833.
-
Merge dfns in headings with the headings themselves gpuweb/gpuweb#3822.
-
Support STORAGE usage on bgra8unorm as an extension gpuweb/gpuweb#3800.
-
Prevent matrix padding from being modified gpuweb/gpuweb#3818.
-
wgsl: Fix capitalization of section headers gpuweb/gpuweb#3817.
-
Say that type constructors act like function calls gpuweb/gpuweb#3814.
-
wgsl: Fix description of truncate and trunc gpuweb/gpuweb#3811.
-
Make regex single-line comment gpuweb/gpuweb#3806.
-
Fix outdated `maxTextureArrayLayers` description gpuweb/gpuweb#3801.
-
grammar analyzer: Fix first and follow sets for degenerate rules gpuweb/gpuweb#3797.
-
Use automatic expiry for imported textures, remove .expired gpuweb/gpuweb#3633.
-
grammar analyzer: Print first and follow sets for each rule gpuweb/gpuweb#3795.
-
-
8 pull requests closed.
-
Implement a tree-sitter scanner for template disambiguation gpuweb/gpuweb#3794.
-
Rename `getBindGroupLayout()` to `createBindGroupLayout()` gpuweb/gpuweb#3804.
-
Add wgsl_version_at_least builtin function. gpuweb/gpuweb#3792.
-
Make the explainer a "Living Document" gpuweb/gpuweb#3705.
-
Make use of [=map/exists=] where appropriate gpuweb/gpuweb#1812.
-
Add diagnostics, derivative uniformity problems trigger diagnostics gpuweb/gpuweb#3683.
-
wgsl: Regenerate HTML when diagram sources change gpuweb/gpuweb#3807.
-
wgsl analyzer: test toy template language gpuweb/gpuweb#3748.
-
-
1 pull request reopened.
-
Fix incorrectly named variable in example 3.2.2 gpuweb/gpuweb#3880, by @typio.
-
-
45 pull requests received 236 new comments.
-
wgsl: discover template lists early (Lookahead disambiguation of less-than vs template argument list (v2)) gpuweb/gpuweb#3803 (31 comments by @alan-baker, @ben-clayton, @dneto0, @mehmetoguzderin)
-
Define "software extensions", and the "requires" directive gpuweb/gpuweb#3874 (20 comments by @alan-baker, @dneto0, @mehmetoguzderin)
-
Add longer WGSL example to top of WGSL spec gpuweb/gpuweb#3842 (19 comments by @alan-baker, @ben-clayton, @dneto0, @toji)
-
Add diagnostic filtering; range diagnostics apply to control flow constructs and compound statements gpuweb/gpuweb#3713 (13 comments by @dneto0, @jimblandy, @jrprice)
-
types, type-generators, enumerants are ordinary predeclared identifiers; not keywords or context-dependent names gpuweb/gpuweb#3859 (11 comments by @alan-baker, @dneto0, @mehmetoguzderin)
-
WGSL: Add dot4U8Packed and dot4I8Packed as new built-in functions gpuweb/gpuweb#2738 (10 comments by @dneto0, @Jiawei-Shao)
-
Add wgsl_version_at_least builtin function. gpuweb/gpuweb#3792 (9 comments by @dneto0, @kdashg, @litherum)
-
Treesitter scanner: fix handling of '=' and shifts in template args gpuweb/gpuweb#3877 (8 comments by @ben-clayton, @dneto0)
-
[editorial] Move value constructors and bitcasts to built-in functions gpuweb/gpuweb#3853 (8 comments by @alan-baker, @dneto0)
-
Rename `compilationInfo()` to `createCompilationInfo()` gpuweb/gpuweb#3805 (8 comments by @kainino0x, @Kangz, @litherum, @takahirox, @toji)
-
Add maxFragmentCombinedOutputResources limit gpuweb/gpuweb#3829 (7 comments by @greggman, @kainino0x, @kdashg)
-
Rename `getBindGroupLayout()` to `createBindGroupLayout()` gpuweb/gpuweb#3804 (6 comments by @kainino0x, @Kangz, @kdashg, @litherum, @toji)
-
Make depthClearValue, depthWriteEnabled, depthCompare required gpuweb/gpuweb#3849 (6 comments by @Kangz, @takahirox, @toji)
-
Support STORAGE usage on bgra8unorm as an extension gpuweb/gpuweb#3800 (5 comments by @Jiawei-Shao, @kainino0x, @Kangz)
-
Add feature for filterable *32float textures gpuweb/gpuweb#3828 (4 comments by @kainino0x, @kdashg, @teoxoy)
-
Validate the shape of Extent/Origin/Color sequence variants. gpuweb/gpuweb#3854 (3 comments by @toji)
-
Improve validation language around timestampWrites gpuweb/gpuweb#3821 (3 comments by @haoxli, @kainino0x)
-
Merge dfns in headings with the headings themselves gpuweb/gpuweb#3822 (3 comments by @kainino0x, @toji)
-
Distinguish between address and contents uniformity gpuweb/gpuweb#3816 (3 comments by @alan-baker, @dneto0)
-
Fix template list discovery, to recognize embedded comparisons and shifts gpuweb/gpuweb#3878 (1 comment by @dneto0)
-
gpuweb/gpuweb#3866 (1 comment by @w3cbot)
-
Implement a tree-sitter scanner for template disambiguation gpuweb/gpuweb#3794 (1 comment by @ben-clayton)
-
Add diagnostics, derivative uniformity problems trigger diagnostics gpuweb/gpuweb#3683 (1 comment by @dneto0)
-
Use automatic expiry for imported textures, remove .expired gpuweb/gpuweb#3633 (1 comment by @kainino0x)
-
grammar analyzer: Fix first and follow sets for degenerate rules gpuweb/gpuweb#3797 (1 comment by @dneto0)
-
TR publications
-
15 documents published.
-
WebGPU Shading Language published 12 times as a Working Draft. History of WebGPU Shading Language
-
WebGPU published 3 times as a Working Draft. History of WebGPU
-
Participants
-
1 participant joined.
-
Antoine Dumas (Preset Computing) joined.
-
-
1 organization joined.
-
Preset Computing joined.
-
GitHub
gpuweb/gpuweb
-
40 issues opened.
-
Interpretation of `sequence<double>` as `GPUColor` is never actually spelled out gpuweb/gpuweb#3793, by @jimblandy.
-
It should be possible to unbind a bind group / vertex buffer gpuweb/gpuweb#3787, by @litherum.
-
WGSL fails to clearly define the target user, and fails to meet user needs gpuweb/gpuweb#3786, by @jwatte.
-
abs(INT_MIN) should be well-defined on integral types gpuweb/gpuweb#3785, by @litherum.
-
No public access to GPU Web meetings raw minutes gpuweb/gpuweb#3780, by @vincentisambart.
-
wgsl: add a note saying function prototypes are not required gpuweb/gpuweb#3779, by @dneto0.
-
Publish WGSL CI artifact to a repo for reusability gpuweb/gpuweb#3778, by @mehmetoguzderin.
-
Why does depthClearValue default to 0.0? gpuweb/gpuweb#3777, by @mwyrzykowski.
-
question: can barriers synchronize memory accesses across workgroups gpuweb/gpuweb#3774, by @dneto0.
-
WGSL TR spec has unreadable text when `prefers-color-scheme: dark` gpuweb/gpuweb#3773, by @ben-clayton.
-
Bytes per texel of `depth24plus-stencil8` could be 4 - 5 gpuweb/gpuweb#3771, by @teoxoy.
-
Lookahead disambiguation of less-than vs template argument list (v2) gpuweb/gpuweb#3770, by @ben-clayton.
-
wgsl/analyze/Grammar.py doesn't compute full lookaheads when token strings are inlined gpuweb/gpuweb#3769, by @dneto0.
-
Clarification of `stencil8` texel block size gpuweb/gpuweb#3768, by @teoxoy.
-
Simplify resolveTarget validation language gpuweb/gpuweb#3767, by @teoxoy.
-
document use cases for onSubmittedWorkDone gpuweb/gpuweb#3762, by @greggman.
-
Remove rule about `n` in `@align(n)` needing to be a multiple of the `RequiredAlignOf` gpuweb/gpuweb#3756, by @ben-clayton.
-
Are padding bytes between matrix columns preserved for storage buffer writes? gpuweb/gpuweb#3753, by @jrprice.
-
Permission Policy for WebGPU power usage gpuweb/gpuweb#3751, by @kainino0x.
-
How to write NaN in a shader? gpuweb/gpuweb#3749, by @unconed.
-
Should WGSL have a ternary operator? gpuweb/gpuweb#3747, by @greggman.
-
Lookahead disambiguation of less-than vs template argument list gpuweb/gpuweb#3746, by @ben-clayton.
-
wgsl: fix crossreference to "GPU command" in WebGPU spec gpuweb/gpuweb#3744, by @dneto0.
-
for to loop transformation example seems incorrect gpuweb/gpuweb#3740, by @vincentisambart.
-
wgsl: stop using "type" to create type aliases. gpuweb/gpuweb#3738, by @dneto0.
-
wgsl: rename static_assert to const_assert gpuweb/gpuweb#3737, by @dneto0.
-
Remove remaining mentions of IO-shareable gpuweb/gpuweb#3735, by @dj2.
-
Usage of depthReadOnly / stencilReadOnly is not clear in the specification gpuweb/gpuweb#3733, by @mwyrzykowski.
-
Language around `atan2`'s accuracy is unclear gpuweb/gpuweb#3731, by @zoddicus.
-
copyTextureToBuffer, should we add native API limits or workaround them? gpuweb/gpuweb#3729, by @greggman.
-
Metal doesn't allow writing frag_depth without a depth attachment gpuweb/gpuweb#3728, by @kainino0x.
-
CF_return in uniformity example 1 is confusing gpuweb/gpuweb#3726, by @dneto0.
-
D3D12 does not support copying between textures of different dimensionality gpuweb/gpuweb#3723, by @kainino0x.
-
editorial: ensure call expressions are type checked gpuweb/gpuweb#3717, by @dneto0.
-
Concretization of abstract expressions can produce surprising types gpuweb/gpuweb#3716, by @ben-clayton.
-
Add feature to specify occlusion query precision gpuweb/gpuweb#3715, by @toji.
-
add diagnostic for unrecognized diagnostic rules gpuweb/gpuweb#3714, by @dneto0.
-
-
22 issues closed.
-
Uniform control flow vs textures for flat interpolated varyings gpuweb/gpuweb#3668.
-
Shader compilation is fragile because FXC gpuweb/gpuweb#3691.
-
WGSL fails to clearly define the target user, and fails to meet user needs gpuweb/gpuweb#3786.
-
wgsl: fix crossreference to "GPU command" in WebGPU spec gpuweb/gpuweb#3744.
-
No public access to GPU Web meetings raw minutes gpuweb/gpuweb#3780.
-
copyTextureToBuffer, should we add native API limits or workaround them? gpuweb/gpuweb#3729.
-
WGSL TR spec has unreadable text when `prefers-color-scheme: dark` gpuweb/gpuweb#3773.
-
wgsl: stop using "type" to create type aliases. gpuweb/gpuweb#3738.
-
Lookahead disambiguation of less-than vs template argument list gpuweb/gpuweb#3746.
-
Permissions Policy for WebGPU and for unmaskHints gpuweb/gpuweb#3483.
-
wgsl: rename static_assert to const_assert gpuweb/gpuweb#3737.
-
Language around `atan2`'s accuracy is unclear gpuweb/gpuweb#3731.
-
Define a MIME type gpuweb/gpuweb#1682.
-
for to loop transformation example seems incorrect gpuweb/gpuweb#3740.
-
editorial: ensure call expressions are type checked gpuweb/gpuweb#3717.
-
shaders: semantics of "discard" gpuweb/gpuweb#361.
-
Make GPURenderPassDecriptor.colorAttachments optional gpuweb/gpuweb#3701.
-
Support boolean varyings gpuweb/gpuweb#1379.
-
-
3 issues reopened.
-
WebGPU: Forbid null code point '\0' (U+0000) within strings gpuweb/gpuweb#3576, by @greggman.
-
How to write NaN in a shader? gpuweb/gpuweb#3749, by @dneto0.
-
Usage of depthReadOnly / stencilReadOnly is not clear in the specification gpuweb/gpuweb#3733, by @kainino0x.
-
-
53 issues received 231 new comments.
-
Lookahead disambiguation of less-than vs template argument list (v2) gpuweb/gpuweb#3770 (28 comments by @ben-clayton, @dj2, @dneto0, @jimblandy, @mehmetoguzderin)
-
wgsl: vec2 shouldn't be a keyword; disambiguate parsing of type construction expr `vec2<i32>(0,1)` gpuweb/gpuweb#3739 (23 comments by @ben-clayton, @dneto0, @greggman, @kdashg, @litherum, @mehmetoguzderin, @munrocket, @teoxoy, @vincentisambart)
-
Usage of depthReadOnly / stencilReadOnly is not clear in the specification gpuweb/gpuweb#3733 (13 comments by @kainino0x, @Kangz, @litherum, @mwyrzykowski)
-
Should WGSL have a ternary operator? gpuweb/gpuweb#3747 (12 comments by @dneto0, @ErichDonGubler, @greggman, @kdashg)
-
Publish WGSL CI artifact to a repo for reusability gpuweb/gpuweb#3778 (9 comments by @dneto0, @Kangz, @litherum, @mehmetoguzderin)
-
document use cases for onSubmittedWorkDone gpuweb/gpuweb#3762 (9 comments by @greggman, @Kangz, @shmookey, @tidoust)
-
Add feature to specify occlusion query precision gpuweb/gpuweb#3715 (9 comments by @kainino0x, @toji)
-
How to write NaN in a shader? gpuweb/gpuweb#3749 (8 comments by @ben-clayton, @dneto0, @jrprice, @kainino0x, @unconed)
-
Should the API enforce that writable storage buffer bindings don't alias each other? gpuweb/gpuweb#1842 (7 comments by @austinEng, @kainino0x, @Kangz, @kdashg, @litherum, @shrekshao)
-
Bytes per texel of `depth24plus-stencil8` could be 4 - 5 gpuweb/gpuweb#3771 (6 comments by @kainino0x, @teoxoy)
-
Lookahead disambiguation of less-than vs template argument list gpuweb/gpuweb#3746 (6 comments by @ben-clayton, @dneto0)
-
wgsl: language evolution: managing rollout of core (essentially sugar) language features gpuweb/gpuweb#3149 (5 comments by @kdashg, @litherum, @mehmetoguzderin)
-
It should be possible to unbind a bind group / vertex buffer gpuweb/gpuweb#3787 (5 comments by @greggman, @Kangz, @litherum)
-
WGSL TR spec has unreadable text when `prefers-color-scheme: dark` gpuweb/gpuweb#3773 (5 comments by @ben-clayton, @mehmetoguzderin)
-
Remove rule about `n` in `@align(n)` needing to be a multiple of the `RequiredAlignOf` gpuweb/gpuweb#3756 (5 comments by @ben-clayton, @dneto0)
-
WebGPU bikeshed error: LINE ~6751: No 'dfn' refs found for 'serializable objects' that are marked for export. gpuweb/gpuweb#3752 (5 comments by @kainino0x, @tidoust)
-
Fingerprinting Surface gpuweb/gpuweb#3101 (5 comments by @kainino0x, @kdashg, @litherum, @yoavweiss)
-
copyTextureToBuffer, should we add native API limits or workaround them? gpuweb/gpuweb#3729 (4 comments by @greggman, @kainino0x, @Kangz, @litherum)
-
D3D12 does not support copying between textures of different dimensionality gpuweb/gpuweb#3723 (4 comments by @kainino0x, @Kangz, @shmookey)
-
question: can barriers synchronize memory accesses across workgroups gpuweb/gpuweb#3774 (3 comments by @dneto0, @kdashg)
-
WGSL fails to clearly define the target user, and fails to meet user needs gpuweb/gpuweb#3786 (3 comments by @jimblandy, @kdashg)
-
Why does depthClearValue default to 0.0? gpuweb/gpuweb#3777 (3 comments by @kainino0x, @Kangz, @mwyrzykowski)
-
Permissions Policy for WebGPU and for unmaskHints gpuweb/gpuweb#3483 (3 comments by @kainino0x, @kdashg)
-
wgsl: stop using "type" to create type aliases. gpuweb/gpuweb#3738 (3 comments by @ben-clayton, @kdashg, @mehmetoguzderin)
-
Metal doesn't allow writing frag_depth without a depth attachment gpuweb/gpuweb#3728 (3 comments by @kainino0x, @mwyrzykowski)
-
Concretization of abstract expressions can produce surprising types gpuweb/gpuweb#3716 (2 comments by @dneto0, @kdashg)
-
No public access to GPU Web meetings raw minutes gpuweb/gpuweb#3780 (2 comments by @Kangz, @vincentisambart)
-
Improve color and display-size readability of diagrams and code blocks and other elements in WGSL spec gpuweb/gpuweb#3649 (2 comments by @mehmetoguzderin)
-
Accessing an "invalid memory reference": Trapping behavior gpuweb/gpuweb#3272 (2 comments by @dneto0, @kdashg)
-
for to loop transformation example seems incorrect gpuweb/gpuweb#3740 (2 comments by @munrocket, @vincentisambart)
-
Make GPURenderPassDecriptor.colorAttachments optional gpuweb/gpuweb#3701 (2 comments by @kainino0x, @kdashg)
-
Make lost devices appear to function as much as possible gpuweb/gpuweb#1629 (2 comments by @litherum)
-
Uniform control flow vs textures for flat interpolated varyings gpuweb/gpuweb#3668 (1 comment by @kdashg)
-
Combined limit for max vertex stage bind groups + vertex buffer bindings gpuweb/gpuweb#2749 (1 comment by @litherum)
-
WebGPU: Forbid null code point '\0' (U+0000) within strings gpuweb/gpuweb#3576 (1 comment by @greggman)
-
Both the WebGPU spec and the WGSL spec have sections that are not readable when the user prefers dark color scheme gpuweb/gpuweb#3776 (1 comment by @greggman)
-
Distinguish between pointer and pointer contents for uniformity requirements on parameters gpuweb/gpuweb#3677 (1 comment by @kdashg)
-
wgsl: refine range diagnostic filters: where they can be applied, and associated affected ranges gpuweb/gpuweb#3689 (1 comment by @dneto0)
-
-
41 pull requests opened.
-
grammar analyzer: Print first and follow sets for each rule gpuweb/gpuweb#3795, by @dneto0.
-
Implement a tree-sitter scanner for template disambiguation gpuweb/gpuweb#3794, by @ben-clayton.
-
Add wgsl_version_at_least builtin function. gpuweb/gpuweb#3792, by @dneto0.
-
Fix wgsl spec example gpuweb/gpuweb#3789, by @IcanDivideBy0.
-
Add maxBindGroupsPlusVertexBuffers limit gpuweb/gpuweb#3788, by @litherum.
-
Reserve common scalar types, and quat gpuweb/gpuweb#3784, by @dneto0.
-
grammar analyzer: fix first and follow sets when terminals are inlined gpuweb/gpuweb#3783, by @dneto0.
-
wgsl tree-sitter: only invoke "npm install" when needed gpuweb/gpuweb#3782, by @dneto0.
-
Working on a way to easily show issue milestones gpuweb/gpuweb#3781, by @toji.
-
Tentative quick fix for readability gpuweb/gpuweb#3775, by @mehmetoguzderin.
-
Fix handling of comments gpuweb/gpuweb#3772, by @DJMcNab.
-
NFKC normal WGSL spec sources and introduce check gpuweb/gpuweb#3766, by @mehmetoguzderin.
-
Demo: Turbofish for generics is not ambiguous gpuweb/gpuweb#3765, by @mehmetoguzderin.
-
Use 'alias' to create type aliases instead of 'type' gpuweb/gpuweb#3764, by @dneto0.
-
Incorporate the restrictions on division into `atan2` gpuweb/gpuweb#3763, by @zoddicus.
-
wgsl: align attribute constraint is technically redundant gpuweb/gpuweb#3761, by @dneto0.
-
Fix cross-refs for new Bikeshed changes gpuweb/gpuweb#3760, by @kainino0x.
-
Reserve static_assert from C++ gpuweb/gpuweb#3759, by @dneto0.
-
Rename static_assert to const_assert gpuweb/gpuweb#3758, by @alan-baker.
-
grammar analyzer: simplify lookahead test gpuweb/gpuweb#3757, by @dneto0.
-
Fix comment in Alias analysis example gpuweb/gpuweb#3755, by @vincentisambart.
-
grammar analyzer: fix computation of lookahead sets gpuweb/gpuweb#3754, by @dneto0.
-
Update `atan2` accuracy to be 4096 ULP for f32 gpuweb/gpuweb#3750, by @zoddicus.
-
wgsl analyzer: test toy template language gpuweb/gpuweb#3748, by @dneto0.
-
Use IANA-MEDIA-TYPES as informative reference gpuweb/gpuweb#3745, by @dneto0.
-
Fix "For to Loop Transformation" example, with fixed headings gpuweb/gpuweb#3743, by @dneto0.
-
Fix "For to Loop transformation" example gpuweb/gpuweb#3742, by @vincentisambart.
-
Finalize text/wgsl media type registration gpuweb/gpuweb#3741, by @dneto0.
-
Remove static_assert from reserved words gpuweb/gpuweb#3736, by @dneto0.
-
[wgsl] Change remainder to say Inherited from. gpuweb/gpuweb#3734, by @dj2.
-
Update text/wgsl media registration with privacy note gpuweb/gpuweb#3732, by @dneto0.
-
Remove static_assert from reserved words. gpuweb/gpuweb#3730, by @dj2.
-
Fix sort order in recursive-descent grammar gpuweb/gpuweb#3727, by @dneto0.
-
Make bikeshed definitions for uniformity nodes gpuweb/gpuweb#3725, by @dneto0.
-
Allow abstract array element access via const-expression index gpuweb/gpuweb#3724, by @alan-baker.
-
Define/document behavior of missing channels in blending gpuweb/gpuweb#3720, by @kainino0x.
-
Formal parameter scope is the function body gpuweb/gpuweb#3719, by @dneto0.
-
Introduce "call_expression" grammar rule gpuweb/gpuweb#3718, by @dneto0.
-
Add missing line breaks in shift preconditions gpuweb/gpuweb#3712, by @jrprice.
-
Dedup state gpuweb/gpuweb#3711, by @toji.
-
-
33 pull requests merged.
-
Reserve common scalar types, and quat gpuweb/gpuweb#3784.
-
grammar analyzer: fix first and follow sets when terminals are inlined gpuweb/gpuweb#3783.
-
wgsl tree-sitter: only invoke "npm install" when needed gpuweb/gpuweb#3782.
-
wgsl: align attribute constraint is technically redundant gpuweb/gpuweb#3761.
-
NFKC normal WGSL spec sources and introduce check gpuweb/gpuweb#3766.
-
Tentative quick fix for readability gpuweb/gpuweb#3775.
-
Fix handling of comments gpuweb/gpuweb#3772.
-
Use 'alias' to create type aliases instead of 'type' gpuweb/gpuweb#3764.
-
Incorporate the restrictions on division into `atan2` gpuweb/gpuweb#3763.
-
Fix cross-refs for new Bikeshed changes gpuweb/gpuweb#3760.
-
Use IANA-MEDIA-TYPES as informative reference gpuweb/gpuweb#3745.
-
grammar analyzer: simplify lookahead test gpuweb/gpuweb#3757.
-
Reserve static_assert from C++ gpuweb/gpuweb#3759.
-
Rename static_assert to const_assert gpuweb/gpuweb#3758.
-
grammar analyzer: fix computation of lookahead sets gpuweb/gpuweb#3754.
-
Fix comment in Alias analysis example gpuweb/gpuweb#3755.
-
Update `atan2` accuracy to be 4096 ULP for f32 gpuweb/gpuweb#3750.
-
Remove static_assert from reserved words gpuweb/gpuweb#3736.
-
Finalize text/wgsl media type registration gpuweb/gpuweb#3741.
-
Fix "For to Loop Transformation" example, with fixed headings gpuweb/gpuweb#3743.
-
Formal parameter scope is the function body gpuweb/gpuweb#3719.
-
Introduce "call_expression" grammar rule gpuweb/gpuweb#3718.
-
[wgsl] Change remainder to say Inherited from. gpuweb/gpuweb#3734.
-
Update text/wgsl media registration with privacy note gpuweb/gpuweb#3732.
-
Allow abstract array element access via const-expression index gpuweb/gpuweb#3724.
-
Make bikeshed definitions for uniformity nodes gpuweb/gpuweb#3725.
-
Fix sort order in recursive-descent grammar gpuweb/gpuweb#3727.
-
Define/document behavior of missing channels in blending gpuweb/gpuweb#3720.
-
WGSL: allow bgra8unorm as texel format gpuweb/gpuweb#3640.
-
Complete several algorithms that enqueue commands gpuweb/gpuweb#3699.
-
Remove spurious note on depthStoreOp gpuweb/gpuweb#3704.
-
Add missing line breaks in shift preconditions gpuweb/gpuweb#3712.
-
Dedup state gpuweb/gpuweb#3711.
-
-
4 pull requests closed.
-
Fix wgsl spec example gpuweb/gpuweb#3789.
-
Fix "For to Loop transformation" example gpuweb/gpuweb#3742.
-
Remove static_assert from reserved words. gpuweb/gpuweb#3730.
-
Add convenience type aliases [fiub]vecN and fmatNxM. gpuweb/gpuweb#1976.
-
-
37 pull requests received 219 new comments.
-
Add diagnostic filtering; range diagnostics apply to control flow constructs and compound statements gpuweb/gpuweb#3713 (78 comments by @alan-baker, @ben-clayton, @dneto0, @jimblandy, @jrprice)
-
Complete several algorithms that enqueue commands gpuweb/gpuweb#3699 (12 comments by @kainino0x, @Kangz, @toji)
-
Formal parameter scope is the function body gpuweb/gpuweb#3719 (10 comments by @alan-baker, @dneto0, @kdashg)
-
Fix handling of comments gpuweb/gpuweb#3772 (9 comments by @dj2, @DJMcNab, @mehmetoguzderin, @w3cbot)
-
NFKC normal WGSL spec sources and introduce check gpuweb/gpuweb#3766 (8 comments by @dneto0, @mehmetoguzderin)
-
Demo: Turbofish for generics is not ambiguous gpuweb/gpuweb#3765 (8 comments by @ben-clayton, @Kangz, @mehmetoguzderin, @teoxoy)
-
WGSL: allow bgra8unorm as texel format gpuweb/gpuweb#3640 (8 comments by @Jiawei-Shao, @kainino0x, @kdashg, @litherum, @teoxoy)
-
Fix "For to Loop transformation" example gpuweb/gpuweb#3742 (7 comments by @dneto0, @vincentisambart, @w3cbot)
-
Introduce "call_expression" grammar rule gpuweb/gpuweb#3718 (6 comments by @alan-baker, @dneto0, @mehmetoguzderin)
-
Add wgsl_version_at_least builtin function. gpuweb/gpuweb#3792 (5 comments by @alan-baker, @dneto0, @litherum)
-
Implement a tree-sitter scanner for template disambiguation gpuweb/gpuweb#3794 (5 comments by @ben-clayton, @mehmetoguzderin)
-
Add diagnostics, derivative uniformity problems trigger diagnostics gpuweb/gpuweb#3683 (5 comments by @dneto0, @jimblandy)
-
wgsl tree-sitter: only invoke "npm install" when needed gpuweb/gpuweb#3782 (3 comments by @dneto0, @greggman)
-
Working on a way to easily show issue milestones gpuweb/gpuweb#3781 (3 comments by @kainino0x, @toji)
-
Make GPUPipelineError constructor `message` pseudo-optional gpuweb/gpuweb#3710 (3 comments by @kainino0x)
-
Update text/wgsl media registration with privacy note gpuweb/gpuweb#3732 (3 comments by @dneto0, @mehmetoguzderin)
-
wgsl: align attribute constraint is technically redundant gpuweb/gpuweb#3761 (2 comments by @dneto0, @mehmetoguzderin)
-
Fix "For to Loop Transformation" example, with fixed headings gpuweb/gpuweb#3743 (2 comments by @dneto0, @mehmetoguzderin)
-
grammar analyzer: fix first and follow sets when terminals are inlined gpuweb/gpuweb#3783 (1 comment by @dneto0)
-
gpuweb/cts
-
30 issues opened.
-
wgsl: reserve common scalar types and 'quat' gpuweb/cts#2220, by @dneto0.
-
op: buffer size reflection with OOM gpuweb/cts#2218, by @austinEng.
-
val: Allow bgra8unorm as storage textures gpuweb/cts#2211, by @Jiawei-Shao.
-
op: shader module sourceMap gpuweb/cts#2209, by @greggman.
-
val? op? test for unusual objects as descriptors needed? gpuweb/cts#2205, by @greggman.
-
op: render_pipeline,alpha_to_coverage gpuweb/cts#2201, by @shrekshao.
-
val: render_pipeline,fragment_state:max_color_attachment_bytes_per_sample gpuweb/cts#2199, by @lokokung.
-
val: render_pass,render_pass_descriptor:max_color_attachment_bytes_per_sample gpuweb/cts#2198, by @lokokung.
-
val: capability_checks,limits gpuweb/cts#2195, by @greggman.
-
val: buffer,create:limit gpuweb/cts#2191, by @lokokung.
-
wgsl: use 'alias' to create a type alias instead of 'type' gpuweb/cts#2190, by @dneto0.
-
Should CTS assert instead of quantizing when generating u32/i32 cases? gpuweb/cts#2189, by @zoddicus.
-
op: render pipeline culling gpuweb/cts#2185, by @greggman.
-
wgsl: rename static_assert to const_assert gpuweb/cts#2178, by @dneto0.
-
wgsl: corner cases for scoping of formal parameters to user-declared functions gpuweb/cts#2157, by @dneto0.
-
test bgra8unorm storage texel format gpuweb/cts#2156, by @dneto0.
-
wgsl: `dot` CTS broken for vec3 and vec4 gpuweb/cts#2155, by @zoddicus.
-
op: render_pipeline,sample_mask gpuweb/cts#2146, by @shrekshao.
-
val: test that createRenderPipeline does not match entryPoints with `\0` gpuweb/cts#2135, by @greggman.
-
Differing ordering of operations for builtins, i.e. `dot`, need to be handled gpuweb/cts#2134, by @zoddicus.
-
reftest: test canvas with CSS background color gpuweb/cts#2132, by @greggman.
-
op: Add Buffer.mapState test gpuweb/cts#2124, by @takahirox.
-
op: render_pipeline,pipeline_output_targets gpuweb/cts#2115, by @shrekshao.
-
op: reflections:query_set,* gpuweb/cts#2114, by @greggman.
-
op: reflections:texture,* gpuweb/cts#2113, by @greggman.
-
op: reflections:buffer,* gpuweb/cts#2112, by @greggman.
-
op: default_layout:getBindGroupLayout_js_object gpuweb/cts#2109, by @greggman.
-
val: copyImageBitmapToTexture should only allow 2d textures gpuweb/cts#2108, by @greggman.
-
op: pipeline override values in shader, any type casts gpuweb/cts#2105, by @shrekshao.
-
-
16 issues closed.
-
val: buffer,create:limit gpuweb/cts#2191.
-
Fix wgsl validation tests for IO-shareables gpuweb/cts#656.
-
Enhance the wgsl type generator gpuweb/cts#763.
-
wgsl: validate `static_assert` statement gpuweb/cts#1686.
-
wgsl: validate `;` a valid statement gpuweb/cts#1685.
-
wgsl: execution test for pack2x16float builtin gpuweb/cts#1288.
-
wgsl: value `(@group, @binding)` pair must be unique gpuweb/cts#1707.
-
wgsl: `dot` CTS broken for vec3 and vec4 gpuweb/cts#2155.
-
Support Dark Mode gpuweb/cts#2097.
-
-
1 issue reopened.
-
Update expectation values in webgpu:web_platform,external_texture,video:* gpuweb/cts#2042, by @bjjones.
-
-
25 issues received 42 new comments.
-
reftest: canvas sizing with ResizeObserver and checkerboard pattern gpuweb/cts#1115 (7 comments by @greggman, @kainino0x)
-
web_platform: import from 100 different video streams simultaneously gpuweb/cts#1398 (3 comments by @greggman, @kainino0x)
-
Update expectation values in webgpu:web_platform,external_texture,video:* gpuweb/cts#2042 (2 comments by @bjjones)
-
Differing ordering of operations for builtins, i.e. `dot`, need to be handled gpuweb/cts#2134 (2 comments by @zoddicus)
-
reftests: test more ways to write into canvas texture gpuweb/cts#916 (2 comments by @greggman, @kainino0x)
-
val: render_pass,render_pass_descriptor:max_color_attachment_bytes_per_sample gpuweb/cts#2198 (1 comment by @lokokung)
-
val: render_pipeline,fragment_state:max_color_attachment_bytes_per_sample gpuweb/cts#2199 (1 comment by @lokokung)
-
web_platform: context_creation from transferControlToOffscreen gpuweb/cts#908 (1 comment by @greggman)
-
add 'create' method for ScalarType WGSL builtins testing framework gpuweb/cts#829 (1 comment by @ben-clayton)
-
val: copyImageBitmapToTexture should only allow 2d textures gpuweb/cts#2108 (1 comment by @kainino0x)
-
-
87 pull requests opened.
-
Add tests for i32 binary scalar-vector ops gpuweb/cts#2219, by @amaiorano.
-
Updates error type for create oom buffers. gpuweb/cts#2217, by @lokokung.
-
Add test for u32 scalar-vector remainder gpuweb/cts#2216, by @amaiorano.
-
Add tests for u32 scalar-vector division gpuweb/cts#2215, by @amaiorano.
-
Fix package-lock.json gpuweb/cts#2214, by @greggman.
-
Make labels pass new eslint rules gpuweb/cts#2213, by @greggman.
-
Allow bgra8unorm as storage textures gpuweb/cts#2212, by @Jiawei-Shao.
-
Added sourcmaps to webgpu:api,operation,shader_module,compilation_info gpuweb/cts#2210, by @greggman.
-
Add tests for u32 scalar-vector addition and subtraction gpuweb/cts#2208, by @amaiorano.
-
Adds typescript-eslint/require-await and fixes existing warnings. gpuweb/cts#2207, by @lokokung.
-
Add a `--coverage` flag to cmdline.ts gpuweb/cts#2206, by @ben-clayton.
-
Implement Matrix type for numeric testing gpuweb/cts#2204, by @zoddicus.
-
Add transferControlToOffscreen cases to colorSpace ref test gpuweb/cts#2203, by @greggman.
-
op,render_pipeline,alpha_to_coverage gpuweb/cts#2202, by @shrekshao.
-
Implements and updates maxColorAttachmentBytesPerSample tests gpuweb/cts#2200, by @lokokung.
-
Add a few more debug markers gpuweb/cts#2197, by @greggman.
-
Test labels in descriptors appear on objects created. gpuweb/cts#2196, by @greggman.
-
Make fuzzy test match wpt fuzzy specs gpuweb/cts#2194, by @greggman.
-
Use screenshot library gpuweb/cts#2193, by @greggman.
-
Adds val,buffer,create,limit tests gpuweb/cts#2192, by @lokokung.
-
Adjust fuzzy value on canvas_colorspace_rgba16float ref test gpuweb/cts#2188, by @lokokung.
-
Make canvas_colorspace_rgba16float ref test more fuzzy gpuweb/cts#2187, by @greggman.
-
Culling: Test triangle-strips, depth, and stencil results gpuweb/cts#2186, by @greggman.
-
Texture utility refactoring. gpuweb/cts#2184, by @lokokung.
-
Implement the division param restrictions in atan2 gpuweb/cts#2183, by @zoddicus.
-
Clear pages between pages. gpuweb/cts#2182, by @greggman.
-
Add tests for binary multiplication of scalars with vectors for u32 gpuweb/cts#2181, by @amaiorano.
-
[wgsl] Alignment attribute tests. gpuweb/cts#2180, by @dj2.
-
op,render_pipeline,sample_mask depth stencil multisampled tests gpuweb/cts#2179, by @shrekshao.
-
Fix name of binary/u32_arithmetic cache gpuweb/cts#2177, by @amaiorano.
-
Test colorSpace in readback tests gpuweb/cts#2176, by @greggman.
-
Fix u32 and i32 non-const divide by 0 expecting lhs (not 0) gpuweb/cts#2175, by @amaiorano.
-
[wgsl] Add align size requirement test. gpuweb/cts#2174, by @dj2.
-
Add test for u32 binary arithmetic operations gpuweb/cts#2173, by @amaiorano.
-
Replace map+filter with reduce in case generators gpuweb/cts#2172, by @amaiorano.
-
Add additional unit testing for builtins that depend on `dot` gpuweb/cts#2171, by @zoddicus.
-
Add tests for i32 binary arithmetic operations gpuweb/cts#2170, by @amaiorano.
-
Convert remaining dispatch() to dispatchWorkgroups() gpuweb/cts#2169, by @Kangz.
-
webgpu/shader/validation: Add static_assert tests gpuweb/cts#2168, by @ben-clayton.
-
runtime: Add server '/load' RPC gpuweb/cts#2167, by @ben-clayton.
-
webgpu/shader: Validate semicolons gpuweb/cts#2166, by @ben-clayton.
-
Remove the canvas_size_different_with_back_buffer_size reftest gpuweb/cts#2165, by @Kangz.
-
webgpu/shader/validation: Add resource binding tests gpuweb/cts#2164, by @ben-clayton.
-
val,render_pipeline,depth_stencil_state,depth_write,frag_depth gpuweb/cts#2163, by @shrekshao.
-
Add unary i32 execution tests gpuweb/cts#2162, by @amaiorano.
-
Test GPUCanvasContext.configure colorSpace gpuweb/cts#2161, by @greggman.
-
op,render_pipeline,sample_mask: depth stencil attachment tests gpuweb/cts#2160, by @shrekshao.
-
Remove io-shareable test, move more types into locations test. gpuweb/cts#2159, by @dj2.
-
Convert `dot` to calculating permutations for vec3 and vec4 gpuweb/cts#2158, by @zoddicus.
-
Fix local ref-test runner gpuweb/cts#2154, by @greggman.
-
Make objectEquals(NaN, NaN) return true gpuweb/cts#2153, by @kainino0x.
-
Remove phantom dependency on reftest-wait.js, copy code into CTS gpuweb/cts#2152, by @kainino0x.
-
Maximize the acceptance interval size for `dot` gpuweb/cts#2151, by @zoddicus.
-
Add subrect copy tests for copyToTexture gpuweb/cts#2150, by @shaoboyan.
-
Make it possible to run the wpt ref tests from this repo gpuweb/cts#2149, by @greggman.
-
Optimize `webgpu:shader,execution,expression,*` tests gpuweb/cts#2148, by @ben-clayton.
-
operation,render_pipeline,sample_mask gpuweb/cts#2147, by @shrekshao.
-
Fix Buffer.mapState op tests gpuweb/cts#2145, by @takahirox.
-
If subcase can't be added to .message, use .stack gpuweb/cts#2144, by @kainino0x.
-
Apply error to final result of `atan2` calculation, not intermediates gpuweb/cts#2143, by @zoddicus.
-
[wgs] Sync keyword and reserved word list to spec. gpuweb/cts#2141, by @dj2.
-
Remove class=reftest-wait from a -ref.html gpuweb/cts#2140, by @kainino0x.
-
op: Add infinity bias slope scale params to 'depth_bias' test gpuweb/cts#2139, by @Gyuyoung.
-
Remove class=reftest-wait from reftests gpuweb/cts#2138, by @kainino0x.
-
Remove <link rel=match> from -ref.html gpuweb/cts#2137, by @kainino0x.
-
Check null char create render pipeline gpuweb/cts#2136, by @greggman.
-
Add more states to canvas composite ref tests. gpuweb/cts#2133, by @greggman.
-
Transfer to image bitmap npe gpuweb/cts#2131, by @Kangz.
-
Add Buffer.mapState test gpuweb/cts#2130, by @takahirox.
-
shader/expression: Reduce 'const' shader size gpuweb/cts#2129, by @ben-clayton.
-
Add a ResizeObserve test gpuweb/cts#2128, by @greggman.
-
Add HTML5 DOCTYPE to all reftests gpuweb/cts#2127, by @kainino0x.
-
Test texture descriptor validation in GPUCanvasContext gpuweb/cts#2126, by @Kangz.
-
Unclutter the URL gpuweb/cts#2123, by @greggman.
-
Revise External Texture Test Expectations gpuweb/cts#2122, by @bjjones.
-
Add wpt ref test docs gpuweb/cts#2121, by @greggman.
-
Fix const-eval failures for ldexp gpuweb/cts#2120, by @ben-clayton.
-
Add a canvas image-rendering wpt ref test gpuweb/cts#2119, by @greggman.
-
Add flag to unroll const-eval loops gpuweb/cts#2118, by @ben-clayton.
-
math: Remove pointless lerp() in biasedRange gpuweb/cts#2117, by @ben-clayton.
-
Add reflection tests for buffers. gpuweb/cts#2116, by @greggman.
-
Test you can not create 1d, 2d-array, and 3d multisampled textures gpuweb/cts#2111, by @greggman.
-
Check that getBindGroupLayout returns unique JS objects. gpuweb/cts#2110, by @greggman.
-
operation,render_pipeline_overrides tests gpuweb/cts#2106, by @shrekshao.
-
Update kMaxQueryCount to 4096 gpuweb/cts#2104, by @Kangz.
-
-
80 pull requests merged.
-
Texture utility refactoring. gpuweb/cts#2184.
-
op,render_pipeline,alpha_to_coverage gpuweb/cts#2202.
-
Add test for u32 scalar-vector remainder gpuweb/cts#2216.
-
Allow bgra8unorm as storage textures gpuweb/cts#2212.
-
Fix package-lock.json gpuweb/cts#2214.
-
Add tests for u32 scalar-vector division gpuweb/cts#2215.
-
Add transferControlToOffscreen cases to colorSpace ref test gpuweb/cts#2203.
-
Make labels pass new eslint rules gpuweb/cts#2213.
-
Add tests for u32 scalar-vector addition and subtraction gpuweb/cts#2208.
-
Test labels in descriptors appear on objects created. gpuweb/cts#2196.
-
Add a few more debug markers gpuweb/cts#2197.
-
Adds typescript-eslint/require-await and fixes existing warnings. gpuweb/cts#2207.
-
Adds val,buffer,create,limit tests gpuweb/cts#2192.
-
Add a `--coverage` flag to cmdline.ts gpuweb/cts#2206.
-
Add tests for binary multiplication of scalars with vectors for u32 gpuweb/cts#2181.
-
Implements and updates maxColorAttachmentBytesPerSample tests gpuweb/cts#2200.
-
Make fuzzy test match wpt fuzzy specs gpuweb/cts#2194.
-
Use screenshot library gpuweb/cts#2193.
-
Adjust fuzzy value on canvas_colorspace_rgba16float ref test gpuweb/cts#2188.
-
Make canvas_colorspace_rgba16float ref test more fuzzy gpuweb/cts#2187.
-
op,render_pipeline,sample_mask depth stencil multisampled tests gpuweb/cts#2179.
-
Implement the division param restrictions in atan2 gpuweb/cts#2183.
-
math: Remove pointless lerp() in biasedRange gpuweb/cts#2117.
-
Test GPUCanvasContext.configure colorSpace gpuweb/cts#2161.
-
Clear pages between pages. gpuweb/cts#2182.
-
Test colorSpace in readback tests gpuweb/cts#2176.
-
[wgsl] Add align size requirement test. gpuweb/cts#2174.
-
Fix name of binary/u32_arithmetic cache gpuweb/cts#2177.
-
Transfer to image bitmap npe gpuweb/cts#2131.
-
Fix u32 and i32 non-const divide by 0 expecting lhs (not 0) gpuweb/cts#2175.
-
Add test for u32 binary arithmetic operations gpuweb/cts#2173.
-
Replace map+filter with reduce in case generators gpuweb/cts#2172.
-
Add additional unit testing for builtins that depend on `dot` gpuweb/cts#2171.
-
Add tests for i32 binary arithmetic operations gpuweb/cts#2170.
-
Remove the canvas_size_different_with_back_buffer_size reftest gpuweb/cts#2165.
-
Convert remaining dispatch() to dispatchWorkgroups() gpuweb/cts#2169.
-
Remove io-shareable test, move more types into locations test. gpuweb/cts#2159.
-
Add unary i32 execution tests gpuweb/cts#2162.
-
webgpu/shader/validation: Add static_assert tests gpuweb/cts#2168.
-
runtime: Add server '/load' RPC gpuweb/cts#2167.
-
webgpu/shader: Validate semicolons gpuweb/cts#2166.
-
webgpu/shader/validation: Add resource binding tests gpuweb/cts#2164.
-
Fix local ref-test runner gpuweb/cts#2154.
-
val,render_pipeline,depth_stencil_state,depth_write,frag_depth gpuweb/cts#2163.
-
Add subrect copy tests for copyToTexture gpuweb/cts#2150.
-
Convert `dot` to calculating permutations for vec3 and vec4 gpuweb/cts#2158.
-
operation,render_pipeline,sample_mask gpuweb/cts#2147.
-
Make objectEquals(NaN, NaN) return true gpuweb/cts#2153.
-
Make it possible to run the wpt ref tests from this repo gpuweb/cts#2149.
-
Fix Buffer.mapState op tests gpuweb/cts#2145.
-
Remove phantom dependency on reftest-wait.js, copy code into CTS gpuweb/cts#2152.
-
Maximize the acceptance interval size for `dot` gpuweb/cts#2151.
-
If subcase can't be added to .message, use .stack gpuweb/cts#2144.
-
Add more states to canvas composite ref tests. gpuweb/cts#2133.
-
Optimize `webgpu:shader,execution,expression,*` tests gpuweb/cts#2148.
-
Unclutter the URL gpuweb/cts#2123.
-
Add Buffer.mapState test gpuweb/cts#2130.
-
[wgs] Sync keyword and reserved word list to spec. gpuweb/cts#2141.
-
Add a ResizeObserve test gpuweb/cts#2128.
-
Test you can not create 1d, 2d-array, and 3d multisampled textures gpuweb/cts#2111.
-
Remove class=reftest-wait from a -ref.html gpuweb/cts#2140.
-
Remove <link rel=match> from -ref.html gpuweb/cts#2137.
-
Test texture descriptor validation in GPUCanvasContext gpuweb/cts#2126.
-
shader/expression: Reduce 'const' shader size gpuweb/cts#2129.
-
Add HTML5 DOCTYPE to all reftests gpuweb/cts#2127.
-
Add a canvas image-rendering wpt ref test gpuweb/cts#2119.
-
Revise External Texture Test Expectations gpuweb/cts#2122.
-
Improve Buffer.mapAsync tests gpuweb/cts#2101.
-
Add wpt ref test docs gpuweb/cts#2121.
-
Check that getBindGroupLayout returns unique JS objects. gpuweb/cts#2110.
-
Fix const-eval failures for ldexp gpuweb/cts#2120.
-
Add a powerPreference option gpuweb/cts#2084.
-
operation,render_pipeline_overrides tests gpuweb/cts#2106.
-
Add flag to unroll const-eval loops gpuweb/cts#2118.
-
Fix silencing of DOMExceptions gpuweb/cts#2102.
-
Update kMaxQueryCount to 4096 gpuweb/cts#2104.
-
-
7 pull requests closed.
-
op,render_pipeline,sample_mask: depth stencil attachment tests gpuweb/cts#2160.
-
Put subcase in the stack instead of the message gpuweb/cts#2103.
-
op: Add infinity bias slope scale params to 'depth_bias' test gpuweb/cts#2139.
-
Check null char create render pipeline gpuweb/cts#2136.
-
Remove class=reftest-wait from reftests gpuweb/cts#2138.
-
Make things work on Node 12.x, and test it on presubmit. gpuweb/cts#1158.
-
Add reflection tests for buffers. gpuweb/cts#2116.
-
-
64 pull requests received 308 new comments.
-
Texture utility refactoring. gpuweb/cts#2184 (28 comments by @austinEng, @kainino0x, @lokokung, @shrekshao)
-
op,render_pipeline,sample_mask depth stencil multisampled tests gpuweb/cts#2179 (18 comments by @kainino0x, @shrekshao)
-
Adds typescript-eslint/require-await and fixes existing warnings. gpuweb/cts#2207 (12 comments by @kainino0x, @lokokung)
-
Implements and updates maxColorAttachmentBytesPerSample tests gpuweb/cts#2200 (12 comments by @lokokung, @shrekshao)
-
Add tests for binary multiplication of scalars with vectors for u32 gpuweb/cts#2181 (11 comments by @amaiorano, @zoddicus)
-
Test labels in descriptors appear on objects created. gpuweb/cts#2196 (9 comments by @greggman, @kainino0x)
-
webgpu/shader/validation: Add static_assert tests gpuweb/cts#2168 (9 comments by @ben-clayton, @dj2, @zoddicus)
-
Add more states to canvas composite ref tests. gpuweb/cts#2133 (8 comments by @greggman, @kainino0x)
-
Culling: Test triangle-strips, depth, and stencil results gpuweb/cts#2186 (6 comments by @kainino0x)
-
Make it possible to run the wpt ref tests from this repo gpuweb/cts#2149 (6 comments by @austinEng, @greggman, @kainino0x)
-
Add tests for i32 binary arithmetic operations gpuweb/cts#2170 (5 comments by @amaiorano, @zoddicus)
-
Convert `dot` to calculating permutations for vec3 and vec4 gpuweb/cts#2158 (5 comments by @ben-clayton, @zoddicus)
-
Put subcase in the stack instead of the message gpuweb/cts#2103 (5 comments by @austinEng, @kainino0x)
-
Remove io-shareable test, move more types into locations test. gpuweb/cts#2159 (4 comments by @ben-clayton, @dj2)
-
Test you can not create 1d, 2d-array, and 3d multisampled textures gpuweb/cts#2111 (4 comments by @greggman, @kainino0x)
-
op: Add 'srcFactor' and 'dstFactor' to the parameters in 'blending,clamping' test gpuweb/cts#2107 (4 comments by @Gyuyoung, @kainino0x)
-
Adjust fuzzy value on canvas_colorspace_rgba16float ref test gpuweb/cts#2188 (3 comments by @greggman, @lokokung)
-
Apply error to final result of `atan2` calculation, not intermediates gpuweb/cts#2143 (3 comments by @zoddicus)
-
op: Add infinity bias slope scale params to 'depth_bias' test gpuweb/cts#2139 (3 comments by @Gyuyoung, @kainino0x)
-
Check that getBindGroupLayout returns unique JS objects. gpuweb/cts#2110 (3 comments by @bjjones, @greggman, @toji)
-
Add additional unit testing for builtins that depend on `dot` gpuweb/cts#2171 (2 comments by @ben-clayton, @zoddicus)
-
op,render_pipeline,sample_mask: depth stencil attachment tests gpuweb/cts#2160 (2 comments by @kainino0x, @shrekshao)
-
Optimize `webgpu:shader,execution,expression,*` tests gpuweb/cts#2148 (2 comments by @ben-clayton, @dneto0)
-
Make things work on Node 12.x, and test it on presubmit. gpuweb/cts#1158 (2 comments by @greggman, @kainino0x)
-
op: Implement 'filter_mode,minFilter' test in filter_mode.spec.ts gpuweb/cts#2090 (2 comments by @Gyuyoung, @kainino0x)
-
Added sourcmaps to webgpu:api,operation,shader_module,compilation_info gpuweb/cts#2210 (1 comment by @lokokung)
-
Add transferControlToOffscreen cases to colorSpace ref test gpuweb/cts#2203 (1 comment by @kainino0x)
-
op: Add 'index_format,setIndexBuffer_before_setPipeline' test in index_format.spec.ts gpuweb/cts#2125 (1 comment by @Gyuyoung)
-
Test beginComputePass/RenderPass after encoder.finish() in call_after_successful_finish gpuweb/cts#2093 (1 comment by @shrekshao)
-