GPU for the Web Working Group - Activity in 2021
No matches.
TR publications
-
38 documents published.
-
WebGPU published 19 times as a Working Draft. History of WebGPU
-
WebGPU Shading Language published 19 times as a Working Draft. History of WebGPU Shading Language
-
GitHub
gpuweb/gpuweb
-
33 issues opened.
-
Supporting MSAA without rendering capability? gpuweb/gpuweb#2465, by @kvark.
-
Behavior of endPass() and pass commands after pass has ended gpuweb/gpuweb#2464, by @kainino0x.
-
device.queue.writeBuffer with sharedArrayBuffer not working gpuweb/gpuweb#2460, by @lslzl3000.
-
MSAA and resolve capabilities need to be specified gpuweb/gpuweb#2459, by @kvark.
-
Allow float modulo for scalar-vector and vector-scalar gpuweb/gpuweb#2450, by @amaiorano.
-
The link of "WebGPU Playground" in wiki Implementation-Status is not correct gpuweb/gpuweb#2448, by @kylexu.
-
Should GPUCompliationIssue.messages be an ObservableArray? gpuweb/gpuweb#2447, by @greggman.
-
wgsl editorial: Builtin variables are not "variables" gpuweb/gpuweb#2445, by @dneto0.
-
wgsl editorial: "entry point IO types" are not types gpuweb/gpuweb#2442, by @dneto0.
-
"Storage class" term is unnecessarily confusing gpuweb/gpuweb#2440, by @kvark.
-
Make WebGPU shuddup about compilation errors gpuweb/gpuweb#2438, by @greggman.
-
Dealing with CRLF and error line numbers gpuweb/gpuweb#2435, by @greggman.
-
behaviour of empty statement is {Next} gpuweb/gpuweb#2431, by @dneto0.
-
image format names, 'read', 'read_write' should not be keywords gpuweb/gpuweb#2428, by @dneto0.
-
Binding model in WGSL does not respect encapsulation gpuweb/gpuweb#2426, by @litherum.
-
spec: forbid infinite loops gpuweb/gpuweb#2414, by @amaiorano.
-
Allow destroying a map-pending buffer gpuweb/gpuweb#2410, by @kainino0x.
-
Do we need a separate RESOLVE texture usage bit? gpuweb/gpuweb#2408, by @kvark.
-
GPUQueue.readBuffer() gpuweb/gpuweb#2405, by @litherum.
-
reassociation is always allowed gpuweb/gpuweb#2402, by @dneto0.
-
maxInterStageShaderComponents is not enough gpuweb/gpuweb#2400, by @kvark.
-
wgsl: Preferred way to extract mat3x3 from mat4x4 gpuweb/gpuweb#2399, by @Axiverse.
-
Need to specify the order of dynamicOffsets in setBindGroup gpuweb/gpuweb#2392, by @Kangz.
-
spec: Cannot bind "null" texture. gpuweb/gpuweb#2389, by @Axiverse.
-
Dynamic Resolution Rendering with WebGPU? gpuweb/gpuweb#2379, by @juj.
-
Should unreachable-code be an error or warning? gpuweb/gpuweb#2378, by @ben-clayton.
-
Atomic<T> vs Untyped Atomics gpuweb/gpuweb#2377, by @cwfitzgerald.
-
Allow copying from any canvas that toDataURL can be called on gpuweb/gpuweb#2373, by @kainino0x.
-
Add gpuCommandEncoder.fillBuffer8() and gpuCommandEncoder.fillBuffer32() gpuweb/gpuweb#2372, by @litherum.
-
wgsl: switch (e1) {...} should remove Next from e1 gpuweb/gpuweb#2366, by @dneto0.
-
-
56 issues closed.
-
MSAA and resolve capabilities need to be specified gpuweb/gpuweb#2459.
-
[wgsl] Consider using multiple scalars to construct matrix. gpuweb/gpuweb#1342.
-
Specification unclear on weather rendering into 3D textures is allowed gpuweb/gpuweb#2303.
-
Validation errors after command encoder finishes gpuweb/gpuweb#2264.
-
beginRenderPass should validate that the attachment don't alias gpuweb/gpuweb#1358.
-
Adding invertY and preMultiplyAlpha options to copyExternalImageToTexture gpuweb/gpuweb#2324.
-
device.queue.writeBuffer with sharedArrayBuffer not working gpuweb/gpuweb#2460.
-
Document cubemap face orientation gpuweb/gpuweb#1946.
-
Validate that requested alignment limits are power of two gpuweb/gpuweb#2099.
-
Make GPUDevice.lost behave like it has [SameObject] gpuweb/gpuweb#2147.
-
Unable to use [SameObject] on "readonly attribute GPUError error" gpuweb/gpuweb#1225.
-
Describe the privacy implications of shader / pipeline caching gpuweb/gpuweb#1311.
-
copyToTexture for color-managed image sources gpuweb/gpuweb#1483.
-
allow both 0x and 0X as hexadecimal literal prefix gpuweb/gpuweb#1453.
-
The link of "WebGPU Playground" in wiki Implementation-Status is not correct gpuweb/gpuweb#2448.
-
behaviour of empty statement is {Next} gpuweb/gpuweb#2431.
-
WGSL requires dynamic matrix indexing, which requires surprising SPIR-V gpuweb/gpuweb#1782.
-
WGSL should specify the execution order gpuweb/gpuweb#2261.
-
Add integer increment and decrement statements gpuweb/gpuweb#2320.
-
Restrictions on function parameters gpuweb/gpuweb#1139.
-
Error conventions for extensions gpuweb/gpuweb#701.
-
Copies of depth values outside [0, 1] gpuweb/gpuweb#789.
-
Binding to zero-sized range of a buffer gpuweb/gpuweb#686.
-
Out-of-bounds indirect dispatch handling gpuweb/gpuweb#323.
-
Case for empty usages gpuweb/gpuweb#1001.
-
Characterize the performance benefit of GPUBufferUsage and GPUTextureUsage gpuweb/gpuweb#150.
-
Consider limiting buffer sizes to multiples of 4 bytes gpuweb/gpuweb#1999.
-
Problem with the default queue gpuweb/gpuweb#1977.
-
Support webgpu context canvas for copyExternalImageToTexture gpuweb/gpuweb#2350.
-
Should it be possible to create 3D depth textures? gpuweb/gpuweb#2304.
-
GPUExternalTexture: Deal with Chroma Reconstruction gpuweb/gpuweb#2098.
-
Allow destroying a map-pending buffer gpuweb/gpuweb#2410.
-
Can value of visibility attribute in GPUBindGroupLayoutBinding be NONE ? gpuweb/gpuweb#405.
-
Missing limits in GPULimits gpuweb/gpuweb#1343.
-
Metal Validation Layers require sourceBytesPerImage >= 512 bytes gpuweb/gpuweb#537.
-
Uniform buffer block size issue on Metal gpuweb/gpuweb#425.
-
Investigation: Compressed Texture Formats gpuweb/gpuweb#144.
-
wgsl: Preferred way to extract mat3x3 from mat4x4 gpuweb/gpuweb#2399.
-
Dynamic Resolution Rendering with WebGPU? gpuweb/gpuweb#2379.
-
GPUQueue.readBuffer() gpuweb/gpuweb#2405.
-
reassociation is always allowed gpuweb/gpuweb#2402.
-
Dynamically sized threadgroup memory gpuweb/gpuweb#2024.
-
What is the toDataURL behavior for GPUCanvasContext? gpuweb/gpuweb#1548.
-
generated spec lacks version info gpuweb/gpuweb#776.
-
spec: Cannot bind "null" texture. gpuweb/gpuweb#2389.
-
wgsl: switch (e1) {...} should remove Next from e1 gpuweb/gpuweb#2366.
-
Consider validating stripIndexFormat at draw time. gpuweb/gpuweb#2199.
-
Method of ensuring GPUShaderModules can contain MTLLibraries gpuweb/gpuweb#1064.
-
wgsl: support texture gather operations gpuweb/gpuweb#2327.
-
wgsl: add degrees, radians conversion functions gpuweb/gpuweb#2358.
-
Document optional depth/stencil formats gpuweb/gpuweb#1479.
-
Not possible to read multisampled depth texture's resolved output gpuweb/gpuweb#2354.
-
Ensure extending fillBuffer is feature detectable gpuweb/gpuweb#2278.
-
-
90 issues received 319 new comments.
-
Atomic<T> vs Untyped Atomics gpuweb/gpuweb#2377 (39 comments by @alan-baker, @AmesingFlank, @cwfitzgerald, @dneto0, @jrmuizel, @Kangz, @kvark, @litherum, @m-schuetz, @raphlinus)
-
maxInterStageShaderComponents is not enough gpuweb/gpuweb#2400 (13 comments by @Jiawei-Shao, @kainino0x, @Kangz, @kvark)
-
Cannot upload/download to UMA storage buffer without an unnecessary copy and unnecessary memory use gpuweb/gpuweb#2388 (13 comments by @benvanik, @kainino0x, @Kangz, @kvark, @litherum)
-
GPUAdapter.name considered harmful gpuweb/gpuweb#2191 (11 comments by @doug-moen, @greggman, @kainino0x, @kenrussell, @krogovin)
-
Storage variables shouldn't be required to be structs gpuweb/gpuweb#2188 (10 comments by @cwfitzgerald, @dneto0, @kainino0x, @kvark, @litherum, @magcius)
-
mapSync on Workers - and possibly on the main thread gpuweb/gpuweb#2217 (8 comments by @benvanik, @kainino0x)
-
Exposing GPU hardware information to web developers in WebGPU gpuweb/gpuweb#2195 (8 comments by @beaufortfrancois, @gyagp, @kainino0x, @krogovin, @toji)
-
Dealing with CRLF and error line numbers gpuweb/gpuweb#2435 (7 comments by @ben-clayton, @dneto0, @greggman, @kainino0x)
-
spec: forbid infinite loops gpuweb/gpuweb#2414 (7 comments by @alan-baker, @ben-clayton, @dneto0, @kdashg, @kvark)
-
Need to specify the order of dynamicOffsets in setBindGroup gpuweb/gpuweb#2392 (6 comments by @kainino0x, @Kangz, @kvark)
-
Do we need a separate RESOLVE texture usage bit? gpuweb/gpuweb#2408 (6 comments by @kainino0x, @Kangz, @kvark)
-
Can value of visibility attribute in GPUBindGroupLayoutBinding be NONE ? gpuweb/gpuweb#405 (6 comments by @kainino0x, @Kangz, @kvark)
-
Adding invertY and preMultiplyAlpha options to copyExternalImageToTexture gpuweb/gpuweb#2324 (6 comments by @kainino0x, @kenrussell, @shaoboyan)
-
copyExternalImageToTexture/importExternalTexture with destination in linear color space gpuweb/gpuweb#1715 (5 comments by @donmccurdy, @kainino0x, @svgeesus)
-
Dealing with holes in the pipeline layout. gpuweb/gpuweb#2043 (5 comments by @Kangz, @kvark, @magcius)
-
Add gpuCommandEncoder.fillBuffer8() and gpuCommandEncoder.fillBuffer32() gpuweb/gpuweb#2372 (5 comments by @kainino0x, @kvark, @magcius)
-
Should WGSL have different types of extensions? gpuweb/gpuweb#2310 (5 comments by @dneto0, @kdashg, @litherum)
-
Proposal: Add a usage bit to allow creating a view with different format when creating a texture gpuweb/gpuweb#168 (5 comments by @fintelia, @kainino0x, @kvark)
-
Should unreachable-code be an error or warning? gpuweb/gpuweb#2378 (4 comments by @kdashg, @kvark, @litherum)
-
Binding to zero-sized range of a buffer gpuweb/gpuweb#686 (4 comments by @kainino0x, @Kangz, @kvark)
-
Should it be possible to create 3D depth textures? gpuweb/gpuweb#2304 (4 comments by @kainino0x, @Kangz)
-
GPUExternalTexture: Deal with Chroma Reconstruction gpuweb/gpuweb#2098 (4 comments by @bjjones, @Kangz, @shaoboyan)
-
Metal Validation Layers require sourceBytesPerImage >= 512 bytes gpuweb/gpuweb#537 (4 comments by @kainino0x, @Kangz, @litherum)
-
WGSL requires dynamic matrix indexing, which requires surprising SPIR-V gpuweb/gpuweb#1782 (4 comments by @dneto0, @kvark, @litherum)
-
Should WGSL support Unicode identifiers? gpuweb/gpuweb#1160 (4 comments by @greggman, @Kangz, @litherum, @mehmetoguzderin)
-
spec: Last vertex buffer element should not be required to have the full stride size, instead only size of attribs. gpuweb/gpuweb#2359 (4 comments by @kainino0x, @Kangz, @kvark)
-
Consider validating stripIndexFormat at draw time. gpuweb/gpuweb#2199 (4 comments by @Kangz, @litherum)
-
texture_2d_depth for loading depth values might be unnecessary? gpuweb/gpuweb#2094 (3 comments by @kainino0x, @kvark)
-
Adding DataView to WGSL to read various types of 1-byte-aligned values from a storage buffer. gpuweb/gpuweb#2429 (3 comments by @kvark, @m-schuetz, @magcius)
-
Binding model in WGSL does not respect encapsulation gpuweb/gpuweb#2426 (3 comments by @kdashg, @kvark, @magcius)
-
Should objects have pointers back to their parents? gpuweb/gpuweb#1497 (3 comments by @Kangz, @kdashg, @litherum)
-
wgsl: Preferred way to extract mat3x3 from mat4x4 gpuweb/gpuweb#2399 (3 comments by @alan-baker, @Axiverse, @kdashg)
-
Select operator arguments are in a non-intuitive order gpuweb/gpuweb#2242 (3 comments by @ben-clayton, @kainino0x)
-
Support webgpu context canvas for copyExternalImageToTexture gpuweb/gpuweb#2350 (3 comments by @kainino0x, @shaoboyan)
-
device.queue.writeBuffer with sharedArrayBuffer not working gpuweb/gpuweb#2460 (2 comments by @kainino0x, @lslzl3000)
-
Fragment output <-> attachment interface matching rules def gpuweb/gpuweb#2013 (2 comments by @kainino0x, @Kangz)
-
Do pipeline blend states need to be validated against the pipline render targets? gpuweb/gpuweb#2025 (2 comments by @kainino0x, @Kangz)
-
Allow read-write storage on selected texture formats gpuweb/gpuweb#1772 (2 comments by @apbodnar, @kainino0x)
-
spec: It is useful to allow GPUVertexBufferLayout.arrayStride to be less than offset + sizeof(attrib.format) gpuweb/gpuweb#2349 (2 comments by @Kangz, @kvark)
-
[wgsl] Consider using multiple scalars to construct matrix. gpuweb/gpuweb#1342 (1 comment by @dneto0)
-
Behavior of endPass() and pass commands after pass has ended gpuweb/gpuweb#2464 (1 comment by @kainino0x)
-
Describe the privacy implications of shader / pipeline caching gpuweb/gpuweb#1311 (1 comment by @kainino0x)
-
Should GPUCompliationIssue.messages be an ObservableArray? gpuweb/gpuweb#2447 (1 comment by @kainino0x)
-
The link of "WebGPU Playground" in wiki Implementation-Status is not correct gpuweb/gpuweb#2448 (1 comment by @kvark)
-
image format names, 'read', 'read_write' should not be keywords gpuweb/gpuweb#2428 (1 comment by @kvark)
-
GPURenderBundleEncoder shouldn't have a [[command_encoder]] gpuweb/gpuweb#1270 (1 comment by @kainino0x)
-
Unable to use [SameObject] on "readonly attribute GPUError error" gpuweb/gpuweb#1225 (1 comment by @kainino0x)
-
Should objects expose the descriptor they were created with? gpuweb/gpuweb#1498 (1 comment by @brendan-duncan)
-
Not possible to read multisampled depth texture's resolved output gpuweb/gpuweb#2354 (1 comment by @Kangz)
-
Allow copying from any canvas that toDataURL can be called on gpuweb/gpuweb#2373 (1 comment by @kainino0x)
-
Consider copyToTexture interactions with high-bit-depth source images gpuweb/gpuweb#1482 (1 comment by @kainino0x)
-
wgsl: for texture coordinates, allow array index be appended to the coordinate vector gpuweb/gpuweb#2340 (1 comment by @kdashg)
-
wgsl: If a function call argument has behaviour {Discard} then the function call behaviour should be {Discard} gpuweb/gpuweb#2367 (1 comment by @dneto0)
-
-
62 pull requests opened.
-
wsgl: Add bit-finding functions. gpuweb/gpuweb#2467, by @dneto0.
-
wgsl: Add extractBits, insertBits gpuweb/gpuweb#2466, by @dneto0.
-
Texture format caps for MSAA and resolve gpuweb/gpuweb#2463, by @kvark.
-
Fully describe validation of render attachments gpuweb/gpuweb#2458, by @kainino0x.
-
Make GPUDevice.lost return the same Promise object gpuweb/gpuweb#2457, by @kainino0x.
-
Require alignment limits to be powers of 2 gpuweb/gpuweb#2456, by @kainino0x.
-
Define GPUTextureViewDimension values gpuweb/gpuweb#2455, by @kainino0x.
-
Clarify that attachments may not alias gpuweb/gpuweb#2454, by @kainino0x.
-
Restore the box around algorithm divs gpuweb/gpuweb#2453, by @kainino0x.
-
Rework encoder state and mixins gpuweb/gpuweb#2452, by @kainino0x.
-
Specify compilation message order/locations are impl-defined gpuweb/gpuweb#2451, by @kainino0x.
-
Disallow pipe for hex literals and allow capital gpuweb/gpuweb#2449, by @mehmetoguzderin.
-
wgsl: Allow 0X for hex prefix gpuweb/gpuweb#2446, by @dneto0.
-
wgsl: fold expression semantics text into definition of type rule gpuweb/gpuweb#2444, by @dneto0.
-
wgsl: Fix "Entry Point" section TODOs gpuweb/gpuweb#2443, by @dneto0.
-
Disallow aliasing writable resources gpuweb/gpuweb#2441, by @kainino0x.
-
Remove [[block]] attribute from sample gpuweb/gpuweb#2439, by @maxammann.
-
wgsl: float to integer conversion saturates gpuweb/gpuweb#2434, by @dneto0.
-
wgsl: Fix TODOs in Execution section gpuweb/gpuweb#2433, by @dneto0.
-
Behaviour of empty statement is {Next} gpuweb/gpuweb#2432, by @dneto0.
-
Behaviors: Ban obviously infinite loops gpuweb/gpuweb#2430, by @dneto0.
-
Enable dynamic indexing on matrix and array values gpuweb/gpuweb#2427, by @dneto0.
-
Specify that filtering with comparison is implementation-defined gpuweb/gpuweb#2425, by @kainino0x.
-
[Process] Add RequirementsForAdditionalFunctionality.md gpuweb/gpuweb#2424, by @litherum.
-
Remove [SameObject] from GPUUncapturedErrorEvent.error gpuweb/gpuweb#2423, by @kainino0x.
-
Disallow empty buffer/texture usages gpuweb/gpuweb#2422, by @kainino0x.
-
Require depth clear value in 0.0-1.0 gpuweb/gpuweb#2421, by @kainino0x.
-
Fix error conventions for optional features gpuweb/gpuweb#2420, by @kainino0x.
-
Require buffer bindings to have non-zero size gpuweb/gpuweb#2419, by @kainino0x.
-
Officially add Brandon as a spec editor gpuweb/gpuweb#2418, by @kainino0x.
-
Specify dispatchIndirect behavior when exceeding limit gpuweb/gpuweb#2417, by @kainino0x.
-
Rely on canvas size instead of allowing size to be configured gpuweb/gpuweb#2416, by @kainino0x.
-
Fix typo in requestDevice and clarify error cases gpuweb/gpuweb#2415, by @kainino0x.
-
Specify order of evaluation for expressions gpuweb/gpuweb#2413, by @alan-baker.
-
Fix classes of examples, globals, and previews gpuweb/gpuweb#2412, by @mehmetoguzderin.
-
Fix GPUBuffer.destroy when "mapping pending" gpuweb/gpuweb#2411, by @Kangz.
-
Fix size of copyBufferToBuffer command in explainer example. gpuweb/gpuweb#2406, by @wwwtyro.
-
Add override declarations gpuweb/gpuweb#2404, by @alan-baker.
-
Always allow reassociation gpuweb/gpuweb#2403, by @dneto0.
-
Store type for buffer does not have to be structure gpuweb/gpuweb#2401, by @dneto0.
-
Fix inplace gpuweb/gpuweb#2398, by @mehmetoguzderin.
-
Use correct substitution gpuweb/gpuweb#2397, by @mehmetoguzderin.
-
Test CI gpuweb/gpuweb#2396, by @mehmetoguzderin.
-
Fix preview version gpuweb/gpuweb#2395, by @mehmetoguzderin.
-
Reduce patterns gpuweb/gpuweb#2394, by @mehmetoguzderin.
-
Validate that stripIndexFormat must be undefined for non-strip topologies gpuweb/gpuweb#2393, by @toji.
-
Tiny fixes gpuweb/gpuweb#2391, by @mehmetoguzderin.
-
Only require/allow depth/stencil load/store ops when they actually have an effect gpuweb/gpuweb#2387, by @toji.
-
Separate loadOp and clear values gpuweb/gpuweb#2386, by @toji.
-
Validate stripIndexFormat at draw-call time. gpuweb/gpuweb#2385, by @Kangz.
-
Remove Next from switch condition when body doesn't have Next gpuweb/gpuweb#2382, by @dneto0.
-
Describe samplers gpuweb/gpuweb#2381, by @alan-baker.
-
Add source image orientation to copyExternalImageToTexture gpuweb/gpuweb#2376, by @shaoboyan.
-
CopyExternalImageToTexture: Support copy from webgpu context canvas/offscreenCanvas gpuweb/gpuweb#2375, by @shaoboyan.
-
Refactored attachment load/store ops and clear value gpuweb/gpuweb#2371, by @toji.
-
[editorial] Describe barycentric coordinates gpuweb/gpuweb#2370, by @kvark.
-
Include container option gpuweb/gpuweb#2369, by @mehmetoguzderin.
-
Rename fillBuffer to clearBuffer gpuweb/gpuweb#2368, by @toji.
-
clarify scope of the for loop initializer gpuweb/gpuweb#2365, by @dneto0.
-
wgsl: Fix no-break loop behavior gpuweb/gpuweb#2364, by @ben-clayton.
-
-
58 pull requests merged.
-
Texture format caps for MSAA and resolve gpuweb/gpuweb#2463.
-
Fully describe validation of render attachments gpuweb/gpuweb#2458.
-
Rework encoder state and mixins gpuweb/gpuweb#2452.
-
Fix classes of examples, globals, and previews gpuweb/gpuweb#2412.
-
Clarify that attachments may not alias gpuweb/gpuweb#2454.
-
Add source image orientation to copyExternalImageToTexture gpuweb/gpuweb#2376.
-
Restore the box around algorithm divs gpuweb/gpuweb#2453.
-
Define GPUTextureViewDimension values gpuweb/gpuweb#2455.
-
Require alignment limits to be powers of 2 gpuweb/gpuweb#2456.
-
Make GPUDevice.lost return the same Promise object gpuweb/gpuweb#2457.
-
Remove [SameObject] from GPUUncapturedErrorEvent.error gpuweb/gpuweb#2423.
-
Disallow pipe for hex literals and allow capital gpuweb/gpuweb#2449.
-
Specify compilation message order/locations are impl-defined gpuweb/gpuweb#2451.
-
wgsl: Allow 0X for hex prefix gpuweb/gpuweb#2446.
-
wgsl: Fix "Entry Point" section TODOs gpuweb/gpuweb#2443.
-
wgsl: Fix TODOs in Execution section gpuweb/gpuweb#2433.
-
Behaviour of empty statement is {Next} gpuweb/gpuweb#2432.
-
Enable dynamic indexing on matrix and array values gpuweb/gpuweb#2427.
-
Remove [[block]] attribute from sample gpuweb/gpuweb#2439.
-
Specify order of evaluation for expressions gpuweb/gpuweb#2413.
-
Add increment and decrement statements gpuweb/gpuweb#2342.
-
Fix error conventions for optional features gpuweb/gpuweb#2420.
-
Require depth clear value in 0.0-1.0 gpuweb/gpuweb#2421.
-
Require buffer bindings to have non-zero size gpuweb/gpuweb#2419.
-
Specify dispatchIndirect behavior when exceeding limit gpuweb/gpuweb#2417.
-
Disallow empty buffer/texture usages gpuweb/gpuweb#2422.
-
Fix typo in requestDevice and clarify error cases gpuweb/gpuweb#2415.
-
Fix GPUBuffer.destroy when "mapping pending" gpuweb/gpuweb#2411.
-
Always allow reassociation gpuweb/gpuweb#2403.
-
Validate that stripIndexFormat must be undefined for non-strip topologies gpuweb/gpuweb#2393.
-
Fix inplace gpuweb/gpuweb#2398.
-
Use correct substitution gpuweb/gpuweb#2397.
-
Reduce patterns gpuweb/gpuweb#2394.
-
Fix preview version gpuweb/gpuweb#2395.
-
Add git version info to wgsl spec gpuweb/gpuweb#2362.
-
wgsl: Add vector and matrix constructor overloads that infer component type gpuweb/gpuweb#2305.
-
Tiny fixes gpuweb/gpuweb#2391.
-
Remove Next from switch condition when body doesn't have Next gpuweb/gpuweb#2382.
-
Fix typo gpuweb/gpuweb#2390.
-
wgsl: Fix no-break loop behavior gpuweb/gpuweb#2364.
-
Validate stripIndexFormat at draw-call time. gpuweb/gpuweb#2385.
-
Updates specs regarding error buffers gpuweb/gpuweb#2314.
-
Allow GPUShaderModules to contain MTLLibraries gpuweb/gpuweb#2363.
-
[editorial] Describe barycentric coordinates gpuweb/gpuweb#2370.
-
Add texture gather builtin functions gpuweb/gpuweb#2341.
-
Add degrees, radians builtin functions gpuweb/gpuweb#2360.
-
Add compound assignment (e.g. +=), updated gpuweb/gpuweb#2345.
-
Describe samplers gpuweb/gpuweb#2381.
-
Add memory model links in the atomic type section gpuweb/gpuweb#2356.
-
[editorial] Simplify links to Vulkan memory model gpuweb/gpuweb#2361.
-
Copyedit depth/stencil format wording and related bits gpuweb/gpuweb#2344.
-
fix typo gpuweb/gpuweb#2374.
-
Rename fillBuffer to clearBuffer gpuweb/gpuweb#2368.
-
Include container option gpuweb/gpuweb#2369.
-
clarify scope of the for loop initializer gpuweb/gpuweb#2365.
-
[editorial] Describe the fragment stage of the rendering algorithm gpuweb/gpuweb#2162.
-
-
10 pull requests closed.
-
wgsl: Remove level from 1D texture load intrinsic gpuweb/gpuweb#1814.
-
[WIP] Read-only GPUTextureView gpuweb/gpuweb#860.
-
Give fillBuffer() a value argument gpuweb/gpuweb#2290.
-
Added byte value to fillBuffer gpuweb/gpuweb#2289.
-
Add complex assignments (e.g. +=) gpuweb/gpuweb#2241.
-
Describe resource aliasing rules gpuweb/gpuweb#2065.
-
Define intrinsic size / natural dimensions of WebGPU canvases gpuweb/gpuweb#2300.
-
Fix size of copyBufferToBuffer command in explainer example. gpuweb/gpuweb#2406.
-
Test CI gpuweb/gpuweb#2396.
-
Refactored attachment load/store ops and clear value gpuweb/gpuweb#2371.
-
-
59 pull requests received 225 new comments.
-
Add source image orientation to copyExternalImageToTexture gpuweb/gpuweb#2376 (17 comments by @kainino0x, @shaoboyan)
-
Only require/allow depth/stencil load/store ops when they actually have an effect gpuweb/gpuweb#2387 (12 comments by @kainino0x, @kvark, @toji)
-
[Process] Add RequirementsForAdditionalFunctionality.md gpuweb/gpuweb#2424 (7 comments by @Kangz, @kvark, @litherum)
-
Add GPUTextureDescriptor option for srgb reinterpretation gpuweb/gpuweb#2336 (6 comments by @kainino0x, @kvark)
-
Rely on canvas size instead of allowing size to be configured gpuweb/gpuweb#2416 (5 comments by @kainino0x, @kvark, @toji)
-
Allow GPUShaderModules to contain MTLLibraries gpuweb/gpuweb#2363 (5 comments by @kainino0x, @kvark, @litherum)
-
Add compound assignment (e.g. +=), updated gpuweb/gpuweb#2345 (5 comments by @dneto0, @jrprice, @kdashg)
-
Fully describe validation of render attachments gpuweb/gpuweb#2458 (4 comments by @kainino0x, @kvark)
-
Store type for buffer does not have to be structure gpuweb/gpuweb#2401 (4 comments by @dneto0, @Kangz, @kvark)
-
Validate stripIndexFormat at draw-call time. gpuweb/gpuweb#2385 (4 comments by @kainino0x, @kvark, @toji)
-
Refactored attachment load/store ops and clear value gpuweb/gpuweb#2371 (4 comments by @kainino0x, @toji)
-
Specify that filtering with comparison is implementation-defined gpuweb/gpuweb#2425 (3 comments by @kainino0x, @kvark)
-
Fix size of copyBufferToBuffer command in explainer example. gpuweb/gpuweb#2406 (3 comments by @kvark, @w3cbot, @wwwtyro)
-
Allow arrays with pipeline-overridable element count, as store type for workgroup variable gpuweb/gpuweb#2347 (3 comments by @dneto0, @kdashg)
-
Add increment and decrement statements gpuweb/gpuweb#2342 (3 comments by @alan-baker, @dneto0, @kvark)
-
wgsl: fold expression semantics text into definition of type rule gpuweb/gpuweb#2444 (2 comments by @dneto0, @kvark)
-
CopyExternalImageToTexture: Support copy from webgpu context canvas/offscreenCanvas gpuweb/gpuweb#2375 (2 comments by @kainino0x, @Kangz)
-
wgsl: Add vector and matrix constructor overloads that infer component type gpuweb/gpuweb#2305 (2 comments by @kdashg)
-
Renames GPUProgrammablePassEncoder to GPUProgrammableEncoder gpuweb/gpuweb#2312 (1 comment by @kainino0x)
-
Define intrinsic size / natural dimensions of WebGPU canvases gpuweb/gpuweb#2300 (1 comment by @kainino0x)
-
Add proposals/ directory and pipeline statistics extension doc gpuweb/gpuweb#2301 (1 comment by @kainino0x)
-
Validate that stripIndexFormat must be undefined for non-strip topologies gpuweb/gpuweb#2393 (1 comment by @Kangz)
-
[editorial] Describe the fragment stage of the rendering algorithm gpuweb/gpuweb#2162 (1 comment by @kainino0x)
-
gpuweb/cts
-
2 issues opened.
-
behaviour case: discard prevents undef behaviour from being executed gpuweb/cts#858, by @dneto0.
-
We need additional unittests for parseQuery gpuweb/cts#855, by @Kangz.
-
-
1 issue closed.
-
We need additional unittests for parseQuery gpuweb/cts#855.
-
-
18 pull requests opened.
-
Track all test-coverage TODOs in descriptions gpuweb/cts#864, by @kainino0x.
-
Improve PR template gpuweb/cts#863, by @kainino0x.
-
Triage testing tasks gpuweb/cts#862, by @kainino0x.
-
Test copy from webgpu context canvas to webgpu texture gpuweb/cts#861, by @shaoboyan.
-
Add api,operation,command_buffer,render,state_tracking:* - Part I gpuweb/cts#860, by @Jiawei-Shao.
-
Remove the parameter 'opts' in makeBufferWithContents gpuweb/cts#859, by @Jiawei-Shao.
-
wgsl: Remove [[block]] attribute from all tests gpuweb/cts#857, by @jrprice.
-
Fix handling of magic values in parseQuery. gpuweb/cts#856, by @Kangz.
-
webgpu: Fix for compilation_info tests gpuweb/cts#854, by @ben-clayton.
-
Implement `correctlyRounded` and update tests to use it gpuweb/cts#853, by @zoddicus.
-
Add api,operation,rendering,indirect_draw tests gpuweb/cts#852, by @haoxli.
-
Implement index buffer OOB validation test for drawIndexed/Indirect gpuweb/cts#851, by @jiangzhaoming.
-
Expand robust_access.spec.ts with dynamicOffset param gpuweb/cts#850, by @austinEng.
-
Roll to the lastest WebGPU type definitions gpuweb/cts#849, by @toji.
-
Fix desynchronized @webgpu/types dependency gpuweb/cts#848, by @kainino0x.
-
Add validation and operations tests for clearBuffer gpuweb/cts#847, by @toji.
-
Add `nextAfter` utility and tests gpuweb/cts#846, by @zoddicus.
-
Remove mipLevelCount > 1 cases of storage texture view gpuweb/cts#845, by @jiangzhaoming.
-
-
17 pull requests merged.
-
Test copy from webgpu context canvas to webgpu texture gpuweb/cts#861.
-
wgsl: Test that [[interpolate]] requires [[location]] gpuweb/cts#842.
-
Remove the parameter 'opts' in makeBufferWithContents gpuweb/cts#859.
-
wgsl: Remove [[block]] attribute from all tests gpuweb/cts#857.
-
Add api,operation,rendering,indirect_draw tests gpuweb/cts#852.
-
Fix handling of magic values in parseQuery. gpuweb/cts#856.
-
webgpu: Fix for compilation_info tests gpuweb/cts#854.
-
Implement `correctlyRounded` and update tests to use it gpuweb/cts#853.
-
Expand robust_access.spec.ts with dynamicOffset param gpuweb/cts#850.
-
Add validation and operations tests for clearBuffer gpuweb/cts#847.
-
Roll to the lastest WebGPU type definitions gpuweb/cts#849.
-
Add `nextAfter` utility and tests gpuweb/cts#846.
-
Fix desynchronized @webgpu/types dependency gpuweb/cts#848.
-
Remove mipLevelCount > 1 cases of storage texture view gpuweb/cts#845.
-
Implement `floor` tests gpuweb/cts#843.
-
-
10 pull requests received 67 new comments.
-
Implement index buffer OOB validation test for drawIndexed/Indirect gpuweb/cts#851 (22 comments by @jiangzhaoming, @kainino0x)
-
Test copy from webgpu context canvas to webgpu texture gpuweb/cts#861 (11 comments by @kainino0x, @shaoboyan)
-
Add validation and operations tests for clearBuffer gpuweb/cts#847 (10 comments by @kainino0x, @toji)
-
wgsl: Test that [[interpolate]] requires [[location]] gpuweb/cts#842 (2 comments by @dneto0, @jrprice)
-
Add api,operation,command_buffer,render,state_tracking:* - Part I gpuweb/cts#860 (1 comment by @kainino0x)
-
TR publications
-
26 documents published.
-
WebGPU published 13 times as a Working Draft. History of WebGPU
-
WebGPU Shading Language published 13 times as a Working Draft. History of WebGPU Shading Language
-
Participants
-
1 participant joined.
-
Van Nguyen (Vidbase, Inc) joined.
-
-
1 organization joined.
-
Vidbase, Inc joined.
-
GitHub
gpuweb/gpuweb
-
45 issues opened.
-
wgsl: add degrees, radians conversion functions gpuweb/gpuweb#2358, by @dneto0.
-
wgsl: on-demand interpolation gpuweb/gpuweb#2357, by @dneto0.
-
Not possible to read multisampled depth texture's resolved output gpuweb/gpuweb#2354, by @ottoville.
-
Support webgpu context canvas for copyExternalImageToTexture gpuweb/gpuweb#2350, by @qjia7.
-
wgsl: add 'array' overload to construct an array from components, infering type and count gpuweb/gpuweb#2346, by @dneto0.
-
describe the "sample footprint" gpuweb/gpuweb#2343, by @dneto0.
-
wgsl: for texture coordinates, allow array index be appended to the coordinate vector gpuweb/gpuweb#2340, by @dneto0.
-
Whether webGPU can render Yuv data directly to canvas gpuweb/gpuweb#2338, by @yangfangfang1204.
-
Module not found: Can't resolve 'stanford-dragon/4' gpuweb/gpuweb#2337, by @yangfangfang1204.
-
wgsl: clarify that attributes can statically-access a declaration gpuweb/gpuweb#2333, by @dneto0.
-
Depth copies between depth32float and depth32float-stencil8 gpuweb/gpuweb#2328, by @kainino0x.
-
wgsl: support texture gather operations gpuweb/gpuweb#2327, by @dneto0.
-
Adding invertY and preMultiplyAlpha options to copyExternalImageToTexture gpuweb/gpuweb#2324, by @Popov72.
-
Annotation for the uniformity analysis gpuweb/gpuweb#2323, by @Kangz.
-
Allow format changes in copyTextureToTexture gpuweb/gpuweb#2322, by @kainino0x.
-
Uniformity analysis: find a way to load from a storage buffer uniformly. gpuweb/gpuweb#2321, by @Kangz.
-
Add integer increment and decrement statements gpuweb/gpuweb#2320, by @dneto0.
-
Shader resources reflection from ShaderModule is needed. gpuweb/gpuweb#2316, by @JasonHuang3D.
-
Should WGSL have different types of extensions? gpuweb/gpuweb#2310, by @alan-baker.
-
Remove isNormal()? gpuweb/gpuweb#2309, by @litherum.
-
Expose WGSL errors in the API gpuweb/gpuweb#2308, by @kainino0x.
-
Should it be possible to create 3D depth textures? gpuweb/gpuweb#2304, by @Wumpf.
-
Specification unclear on weather rendering into 3D textures is allowed gpuweb/gpuweb#2303, by @Wumpf.
-
textureLoad with sampled textures. gpuweb/gpuweb#2299, by @mr-ariel.
-
Allow identifiers to start with _ gpuweb/gpuweb#2298, by @mr-ariel.
-
wgsl: should NUL characters be permitted in source code? gpuweb/gpuweb#2294, by @dneto0.
-
wgsl: add unary plus? gpuweb/gpuweb#2279, by @dneto0.
-
Ensure extending fillBuffer is feature detectable gpuweb/gpuweb#2278, by @toji.
-
It's not clear what popErrorScope() returns gpuweb/gpuweb#2274, by @litherum.
-
Is it okay that `isInf()` and `isNan()` can return false for infinities / nans? gpuweb/gpuweb#2270, by @ben-clayton.
-
WGSL should forbid pointers to runtime-sized values as user-defined function parameters gpuweb/gpuweb#2268, by @jimblandy.
-
Validation errors after command encoder finishes gpuweb/gpuweb#2264, by @kainino0x.
-
WGSL should specify the execution order gpuweb/gpuweb#2261, by @RobinMorisset.
-
wgsl: add feature to enable more IEEE 754 floating point behaviours gpuweb/gpuweb#2259, by @dneto0.
-
Support for Wasm memory64? gpuweb/gpuweb#2252, by @juj.
-
should WGSL handle unicode bidirectional overrides specially? gpuweb/gpuweb#2245, by @dneto0.
-
Are double brackets necessary for attributes? gpuweb/gpuweb#2243, by @mehmetoguzderin.
-
Select operator arguments are in a non-intuitive order gpuweb/gpuweb#2242, by @pow2clk.
-
[proposal] Changes to overridable constants gpuweb/gpuweb#2238, by @alan-baker.
-
Describe more clearly who can contribute in README.md gpuweb/gpuweb#2237, by @Kangz.
-
How to measure the shader compilation time in js gpuweb/gpuweb#2236, by @qjia7.
-
support dot(vecN<i32>, vecN<i32>) in WGSL gpuweb/gpuweb#2231, by @axinging.
-
69 issues closed.
-
Do we need [[block]] annotations? gpuweb/gpuweb#1008.
-
what is the initial value of a workgroup variable? gpuweb/gpuweb#1137.
-
WGSL: function parameters act like const declarations, not var declarations gpuweb/gpuweb#1171.
-
specify rules for 'block' decoration gpuweb/gpuweb#1123.
-
call_statement vs. a call expression gpuweb/gpuweb#1412.
-
wgsl: an entry point must not be the target of a function call gpuweb/gpuweb#1663.
-
Consider improving variable array indexing for 'let' arrays gpuweb/gpuweb#2086.
-
WebGPU should validate that storageTexture bindings have a single mip level. gpuweb/gpuweb#1576.
-
[wgsl] functions with return values + return + control flow gpuweb/gpuweb#1897.
-
Copying of depth24plus formats gpuweb/gpuweb#652.
-
Whether webGPU can render Yuv data directly to canvas gpuweb/gpuweb#2338.
-
Module not found: Can't resolve 'stanford-dragon/4' gpuweb/gpuweb#2337.
-
wgsl: clarify that attributes can statically-access a declaration gpuweb/gpuweb#2333.
-
Depth copies between depth32float and depth32float-stencil8 gpuweb/gpuweb#2328.
-
Allow identifiers to start with _ gpuweb/gpuweb#2298.
-
Remove isNormal()? gpuweb/gpuweb#2309.
-
Atomic concerns gpuweb/gpuweb#2229.
-
Is it okay that `isInf()` and `isNan()` can return false for infinities / nans? gpuweb/gpuweb#2270.
-
Task: Define atomic operations gpuweb/gpuweb#673.
-
wgsl: should NUL characters be permitted in source code? gpuweb/gpuweb#2294.
-
Add [Exposed] to every interface gpuweb/gpuweb#1158.
-
Are implementations allowed to crush DOMString labels down to ASCII? gpuweb/gpuweb#1211.
-
Assignment operators (e.g. `+=`) in wgsl gpuweb/gpuweb#1518.
-
wgsl: add unary plus? gpuweb/gpuweb#2279.
-
Allow f suffix for float literals gpuweb/gpuweb#2210.
-
[proposal] Location attribute assignment gpuweb/gpuweb#1504.
-
Disallow implicit LOD texture methods in Compute shaders. gpuweb/gpuweb#1221.
-
sized arrays must have at least one element gpuweb/gpuweb#1195.
-
what does ...snorm mean, exactly gpuweb/gpuweb#1114.
-
define the term "statically reachable" gpuweb/gpuweb#1091.
-
Specify the stdlib methods gpuweb/gpuweb#1088.
-
Naming of texture types. gpuweb/gpuweb#923.
-
Vertex and Instance ID and base ID builtins. gpuweb/gpuweb#901.
-
Should all expressions have known types? gpuweb/gpuweb#723.
-
Proposal: Permit redundant as-casts gpuweb/gpuweb#715.
-
Interface matching rules gpuweb/gpuweb#644.
-
Limit the binding number for resources gpuweb/gpuweb#1783.
-
The set of functions which have to be imported seems arbitrary gpuweb/gpuweb#640.
-
[wgsl] Textures gpuweb/gpuweb#573.
-
[wgsl] Samplers gpuweb/gpuweb#574.
-
[wgsl] Implicit types for variable declarations with expressions. gpuweb/gpuweb#560.
-
Pipeline statistics clarifications gpuweb/gpuweb#794.
-
textureLoad with sampled textures. gpuweb/gpuweb#2299.
-
Development-only API features gpuweb/gpuweb#814.
-
Storage buffer access: compatibility with the pipeline layout gpuweb/gpuweb#1837.
-
should WGSL handle unicode bidirectional overrides specially? gpuweb/gpuweb#2245.
-
It's not clear what popErrorScope() returns gpuweb/gpuweb#2274.
-
Consider replacing elseif with else if gpuweb/gpuweb#2212.
-
Operator Precedence gpuweb/gpuweb#1146.
-
Interpolate attribute clarification gpuweb/gpuweb#2017.
-
Clarify the rule about discard in a continuing block gpuweb/gpuweb#2219.
-
Parallels between WGSL and SPIR-V don't justify `let` gpuweb/gpuweb#2207.
-
samples need updating gpuweb/gpuweb#1873.
-
Task: Memory model gpuweb/gpuweb#670.
-
The order of top-level declarations should be irrelevant gpuweb/gpuweb#875.
-
support dot(vecN<i32>, vecN<i32>) in WGSL gpuweb/gpuweb#2231.
-
add a combined image sampler as Vulkan gpuweb/gpuweb#770.
-
Investigation: Storage Texture gpuweb/gpuweb#513.
-
MSL does not support address-of vector component gpuweb/gpuweb#1931.
-
Implement "depth_clamp" on D3D12 gpuweb/gpuweb#2100.
-
Consider removing GPUCommandBuffer.executionTime gpuweb/gpuweb#2156.
-
[WGSL] Web IDL validation errors gpuweb/gpuweb#2224.
-
[webgpu] Web IDL validation errors gpuweb/gpuweb#2223.
-
-
124 issues received 370 new comments.
-
Determine GPUExternalTexture lifetime gpuweb/gpuweb#2124 (21 comments by @kainino0x, @Kangz, @kdashg, @litherum, @shaoboyan)
-
Adding invertY and preMultiplyAlpha options to copyExternalImageToTexture gpuweb/gpuweb#2324 (14 comments by @kainino0x, @kenrussell, @Popov72, @shaoboyan)
-
[proposal] Changes to overridable constants gpuweb/gpuweb#2238 (11 comments by @alan-baker, @ben-clayton, @dneto0, @Kangz, @kdashg, @kvark, @litherum)
-
Streaming implementations and indirect draws/dispatches gpuweb/gpuweb#2189 (9 comments by @kainino0x, @Kangz, @kvark, @litherum)
-
GPUAdapter.name considered harmful gpuweb/gpuweb#2191 (9 comments by @gyagp, @juj, @Kangz, @kenrussell, @krogovin, @raphlinus, @toji)
-
Validation errors after command encoder finishes gpuweb/gpuweb#2264 (9 comments by @kainino0x, @Kangz)
-
Proposal: Add a usage bit to allow creating a view with different format when creating a texture gpuweb/gpuweb#168 (9 comments by @Jiawei-Shao, @kainino0x, @Kangz, @magcius)
-
support dot(vecN<i32>, vecN<i32>) in WGSL gpuweb/gpuweb#2231 (9 comments by @ben-clayton, @bjornbytes, @dneto0, @Jiawei-Shao, @kdashg)
-
Select operator arguments are in a non-intuitive order gpuweb/gpuweb#2242 (8 comments by @ben-clayton, @dneto0, @kainino0x, @kvark)
-
Ensure extending fillBuffer is feature detectable gpuweb/gpuweb#2278 (7 comments by @kainino0x, @Kangz, @kvark, @litherum, @toji)
-
Allow format changes in copyTextureToTexture gpuweb/gpuweb#2322 (7 comments by @fintelia, @kainino0x, @kvark)
-
Allow identifiers to start with _ gpuweb/gpuweb#2298 (7 comments by @ben-clayton, @dneto0, @kdashg, @mr-ariel)
-
Exposing GPU hardware information to web developers in WebGPU gpuweb/gpuweb#2195 (7 comments by @gyagp, @Kangz, @krogovin, @kvark)
-
Are double brackets necessary for attributes? gpuweb/gpuweb#2243 (7 comments by @aloucks, @dneto0, @Kangz, @kdashg, @mehmetoguzderin)
-
WGSL should specify the execution order gpuweb/gpuweb#2261 (7 comments by @dneto0, @kdashg, @litherum, @RobinMorisset)
-
spec: Last vertex buffer element should not be required to have the full stride size, instead only size of attribs. gpuweb/gpuweb#2359 (6 comments by @Axiverse, @Kangz, @kvark)
-
☂️ Initial internal feedback gpuweb/gpuweb#2213 (6 comments by @archo5, @jimblandy, @kvark, @magcius)
-
Uniformity analysis: find a way to load from a storage buffer uniformly. gpuweb/gpuweb#2321 (6 comments by @Kangz, @kvark)
-
Investigation: Import VideoFrame from WebCodec to WebGPU gpuweb/gpuweb#1380 (6 comments by @alexkarpenko, @kainino0x)
-
WGSL should forbid pointers to runtime-sized values as user-defined function parameters gpuweb/gpuweb#2268 (6 comments by @jimblandy, @jrprice)
-
How to measure the shader compilation time in js gpuweb/gpuweb#2236 (6 comments by @kainino0x, @Kangz, @litherum, @qjia7)
-
wgsl: for texture coordinates, allow array index be appended to the coordinate vector gpuweb/gpuweb#2340 (5 comments by @ben-clayton, @dneto0, @kvark, @litherum)
-
Shader resources reflection from ShaderModule is needed. gpuweb/gpuweb#2316 (5 comments by @JasonHuang3D, @Kangz)
-
Add integer increment and decrement statements gpuweb/gpuweb#2320 (4 comments by @dneto0, @Kangz, @kdashg, @kvark)
-
Validate that [[builtin fragDepth]] can't be written if depthWriteEnabled = false? gpuweb/gpuweb#1013 (4 comments by @kainino0x)
-
Is it okay that `isInf()` and `isNan()` can return false for infinities / nans? gpuweb/gpuweb#2270 (4 comments by @dneto0, @kdashg, @RobinMorisset)
-
Clarify the rule about discard in a continuing block gpuweb/gpuweb#2219 (4 comments by @dneto0, @kdashg, @RobinMorisset)
-
Whether webGPU can render Yuv data directly to canvas gpuweb/gpuweb#2338 (3 comments by @shaoboyan, @yangfangfang1204)
-
Case for optional GPUBindGroupLayoutEntry entries gpuweb/gpuweb#851 (3 comments by @kainino0x, @Kangz, @kvark)
-
wgsl: should NUL characters be permitted in source code? gpuweb/gpuweb#2294 (3 comments by @dneto0, @kdashg, @RobinMorisset)
-
The order of top-level declarations should be irrelevant gpuweb/gpuweb#875 (3 comments by @dneto0, @kdashg, @kvark)
-
should WGSL handle unicode bidirectional overrides specially? gpuweb/gpuweb#2245 (3 comments by @dneto0, @kdashg)
-
Support webgpu context canvas for copyExternalImageToTexture gpuweb/gpuweb#2350 (2 comments by @kainino0x, @shaoboyan)
-
Consider validating stripIndexFormat at draw time. gpuweb/gpuweb#2199 (2 comments by @kainino0x, @kvark)
-
wgsl: add 'array' overload to construct an array from components, infering type and count gpuweb/gpuweb#2346 (2 comments by @dneto0)
-
Module not found: Can't resolve 'stanford-dragon/4' gpuweb/gpuweb#2337 (2 comments by @Jiawei-Shao, @yangfangfang1204)
-
[wgsl] Support compile-time-constants (constexpr) gpuweb/gpuweb#1272 (2 comments by @alan-baker, @dneto0)
-
Depth copies between depth32float and depth32float-stencil8 gpuweb/gpuweb#2328 (2 comments by @kainino0x)
-
Investigate copying between textures with different formats. gpuweb/gpuweb#842 (2 comments by @kainino0x)
-
Are implementations allowed to crush DOMString labels down to ASCII? gpuweb/gpuweb#1211 (2 comments by @kainino0x, @kdashg)
-
[wgsl] functions with return values + return + control flow gpuweb/gpuweb#1897 (2 comments by @Kangz, @RobinMorisset)
-
context-aware tokenization: Sometimes >> is better parsed as two copies of > (similar for ]] ) gpuweb/gpuweb#2092 (2 comments by @dneto0)
-
Should it be possible to create 3D depth textures? gpuweb/gpuweb#2304 (2 comments by @Kangz, @kvark)
-
mapSync on Workers - and possibly on the main thread gpuweb/gpuweb#2217 (2 comments by @kainino0x, @shanumante-sc)
-
Storage buffer access: compatibility with the pipeline layout gpuweb/gpuweb#1837 (2 comments by @Kangz)
-
Consider not requiring braces in some situations gpuweb/gpuweb#2211 (2 comments by @dneto0, @litherum)
-
copyToTexture for color-managed image sources gpuweb/gpuweb#1483 (2 comments by @kainino0x, @svgeesus)
-
How should WGSL map to the availability/visibility parts of the Vulkan Memory Model? gpuweb/gpuweb#2228 (2 comments by @alan-baker, @RobinMorisset)
-
constructor expressions for arrays and structs should probably only be for storable types gpuweb/gpuweb#982 (1 comment by @dneto0)
-
WGSL: function parameters act like const declarations, not var declarations gpuweb/gpuweb#1171 (1 comment by @dneto0)
-
wgsl: an entry point must not be the target of a function call gpuweb/gpuweb#1663 (1 comment by @dneto0)
-
spec: It is useful to allow GPUVertexBufferLayout.arrayStride to be less than offset + sizeof(attrib.format) gpuweb/gpuweb#2349 (1 comment by @Axiverse)
-
Spec the texture-compression-etc and texture-compression-astc extensions gpuweb/gpuweb#2083 (1 comment by @kainino0x)
-
Can value of visibility attribute in GPUBindGroupLayoutBinding be NONE ? gpuweb/gpuweb#405 (1 comment by @kvark)
-
wgsl: clarify that attributes can statically-access a declaration gpuweb/gpuweb#2333 (1 comment by @dneto0)
-
texture_2d_depth for loading depth values might be unnecessary? gpuweb/gpuweb#2094 (1 comment by @kainino0x)
-
WGSL requires dynamic matrix indexing, which requires surprising SPIR-V gpuweb/gpuweb#1782 (1 comment by @kvark)
-
Task: missing rules for stage-in and stage-out variables with conflicting index gpuweb/gpuweb#779 (1 comment by @Kangz)
-
var declaration without storage class: spell out rules, infer Function storage class inside a function gpuweb/gpuweb#654 (1 comment by @Kangz)
-
The set of functions which have to be imported seems arbitrary gpuweb/gpuweb#640 (1 comment by @Kangz)
-
[wgsl] Implicit types for variable declarations with expressions. gpuweb/gpuweb#560 (1 comment by @Kangz)
-
wgsl: finding most-significant bit and least-significant bit gpuweb/gpuweb#2130 (1 comment by @fintelia)
-
Specify behavior of GPUCommandBuffer.executionTime on devices without timestamp queries. gpuweb/gpuweb#1773 (1 comment by @Kangz)
-
Metal Validation Layers require sourceBytesPerImage >= 512 bytes gpuweb/gpuweb#537 (1 comment by @Kangz)
-
-
84 pull requests opened.
-
Allow GPUShaderModules to contain MTLLibraries gpuweb/gpuweb#2363, by @litherum.
-
Add git version info to wgsl spec gpuweb/gpuweb#2362, by @dneto0.
-
[editorial] Simplify links to Vulkan memory model gpuweb/gpuweb#2361, by @alan-baker.
-
Add degrees, radians builtin functions gpuweb/gpuweb#2360, by @dneto0.
-
Add memory model links in the atomic type section gpuweb/gpuweb#2356, by @alan-baker.
-
Workgroup variables are zero-initialized gpuweb/gpuweb#2355, by @dneto0.
-
Update wording of texture usage to allow parameter gpuweb/gpuweb#2353, by @alan-baker.
-
Cross reference WGSL errors in WebGPU spec gpuweb/gpuweb#2352, by @alan-baker.
-
Fill out scope explanations and collective operations gpuweb/gpuweb#2351, by @alan-baker.
-
Allow arrays with pipeline-overridable element count, as store type for workgroup variable gpuweb/gpuweb#2347, by @dneto0.
-
Add compound assignment (e.g. +=), updated gpuweb/gpuweb#2345, by @dneto0.
-
Copyedit depth/stencil format wording and related bits gpuweb/gpuweb#2344, by @kainino0x.
-
Add increment and decrement statements gpuweb/gpuweb#2342, by @dneto0.
-
Add texture gather builtin functions gpuweb/gpuweb#2341, by @dneto0.
-
Add GPUTextureDescriptor option for srgb reinterpretation gpuweb/gpuweb#2336, by @kainino0x.
-
Explain why minBindingSize is "optional" gpuweb/gpuweb#2335, by @kainino0x.
-
wgsl: Clarify "static accessed" gpuweb/gpuweb#2334, by @dneto0.
-
Fix typo, per code review gpuweb/gpuweb#2332, by @dneto0.
-
Describe tight-packing of image-copy ops, including depth/stencil gpuweb/gpuweb#2331, by @kainino0x.
-
Conversion from float to integral rounds toward zero gpuweb/gpuweb#2330, by @dneto0.
-
Specify copyTextureToTexture format compatibility gpuweb/gpuweb#2329, by @kainino0x.
-
Allow identifiers to start with underscore, with exceptions gpuweb/gpuweb#2326, by @dneto0.
-
Remove isFinite, isNormal gpuweb/gpuweb#2325, by @dneto0.
-
Indirect first instance cleanup gpuweb/gpuweb#2319, by @rconde01.
-
Clarify behaviour of a function gpuweb/gpuweb#2318, by @dneto0.
-
Do not require function with return value to end with return. gpuweb/gpuweb#2317, by @Kangz.
-
Add multi draw indirect feature gpuweb/gpuweb#2315, by @rconde01.
-
Updates specs regarding error buffers gpuweb/gpuweb#2314, by @lokokung.
-
WGSL source must not include a null character gpuweb/gpuweb#2313, by @dneto0.
-
Renames GPUProgrammablePassEncoder to GPUProgrammableEncoder gpuweb/gpuweb#2312, by @lokokung.
-
Remove isInf, isNan gpuweb/gpuweb#2311, by @dneto0.
-
Allow 'f' suffix on floats gpuweb/gpuweb#2307, by @dneto0.
-
Typing of literals gpuweb/gpuweb#2306, by @RobinMorisset.
-
wgsl: Add vector and matrix constructor overloads that infer component type gpuweb/gpuweb#2305, by @dneto0.
-
Extend lifetime of GPUExternalTexture imported from video element gpuweb/gpuweb#2302, by @kainino0x.
-
Add proposals/ directory and pipeline statistics extension doc gpuweb/gpuweb#2301, by @kainino0x.
-
Define intrinsic size / natural dimensions of WebGPU canvases gpuweb/gpuweb#2300, by @kainino0x.
-
Fixes to memory model for barriers and atomics gpuweb/gpuweb#2297, by @alan-baker.
-
Delete pipeline statistics queries gpuweb/gpuweb#2296, by @litherum.
-
A block comment must affirmatively end before end of source gpuweb/gpuweb#2295, by @dneto0.
-
Remove reset method and exemplify usage gpuweb/gpuweb#2293, by @mehmetoguzderin.
-
Extend references for external scanner section gpuweb/gpuweb#2292, by @mehmetoguzderin.
-
Support nesting block comments in CI grammar gpuweb/gpuweb#2291, by @mehmetoguzderin.
-
Give fillBuffer() a value argument gpuweb/gpuweb#2290, by @litherum.
-
Added byte value to fillBuffer gpuweb/gpuweb#2289, by @toji.
-
Describe nesting block comments gpuweb/gpuweb#2288, by @dneto0.
-
wgsl: Remove TODO about default rounding mode gpuweb/gpuweb#2287, by @dneto0.
-
Limits binding numbers to 65535 gpuweb/gpuweb#2286, by @Kangz.
-
Require storage texture views to have mipLevelCount == 1 gpuweb/gpuweb#2285, by @Kangz.
-
Make invalidity less-magic gpuweb/gpuweb#2284, by @kainino0x.
-
Resolve issue: No other characters count as blankspace gpuweb/gpuweb#2282, by @dneto0.
-
wgsl: all directives appear before any declarations gpuweb/gpuweb#2281, by @dneto0.
-
Fix conflict between the PRs: 'else if' & 'Behavior analysis' gpuweb/gpuweb#2280, by @RobinMorisset.
-
wgsl: describe type aliases gpuweb/gpuweb#2276, by @dneto0.
-
Clarify inline issue about minBindingSize in draw-time validation gpuweb/gpuweb#2273, by @kainino0x.
-
Minor clarifications to GPUImageCopyTextureTagged language gpuweb/gpuweb#2272, by @kainino0x.
-
Tweak behaviour analysis gpuweb/gpuweb#2271, by @dneto0.
-
wgsl: Fix a wrong padding size in the structures layout example gpuweb/gpuweb#2269, by @jiangzhaoming.
-
Update operator precedence rules gpuweb/gpuweb#2267, by @RobinMorisset.
-
Remove obsolete attribute_list in variable_ident_decl gpuweb/gpuweb#2266, by @RobinMorisset.
-
Restrict the lhs of assignments to fix a grammar conflict gpuweb/gpuweb#2265, by @RobinMorisset.
-
Go from `attribute_list *` to `attribute_list ?` in the grammar gpuweb/gpuweb#2263, by @RobinMorisset.
-
Remove stray extra column in storage class table gpuweb/gpuweb#2262, by @dneto0.
-
Fix setIndex/VertexBuffer size defaulting allowing any offset. gpuweb/gpuweb#2260, by @Kangz.
-
Describe the effective size of a buffer binding. gpuweb/gpuweb#2258, by @Kangz.
-
Update samples to the latest spec. gpuweb/gpuweb#2257, by @Kangz.
-
Neither GPUBuffer nor GPUShaderModule is actually serializable gpuweb/gpuweb#2256, by @litherum.
-
Secure context gpuweb/gpuweb#2255, by @litherum.
-
Whitespace nits gpuweb/gpuweb#2254, by @litherum.
-
Explicitly describe declaration cycles gpuweb/gpuweb#2253, by @dneto0.
-
wgsl: Require [[location]] for [[interpolate]] gpuweb/gpuweb#2251, by @jrprice.
-
minor fix gpuweb/gpuweb#2250, by @CatCode79.
-
Support dot() on vecN<i32> and vecN<u32> gpuweb/gpuweb#2249, by @Jiawei-Shao.
-
Add missing brackets for double brackets gpuweb/gpuweb#2248, by @mehmetoguzderin.
-
Comment wasm build gpuweb/gpuweb#2247, by @mehmetoguzderin.
-
Use pip from python3 as a subcommand gpuweb/gpuweb#2246, by @mehmetoguzderin.
-
Allow out of order module scope declarations gpuweb/gpuweb#2244, by @alan-baker.
-
Add complex assignments (e.g. +=) gpuweb/gpuweb#2241, by @RobinMorisset.
-
Add validation rules for RenderPassEncoder.executeBundles() gpuweb/gpuweb#2240, by @Richard-Yunchao.
-
Replace the "elseif" syntax by "else if" gpuweb/gpuweb#2239, by @RobinMorisset.
-
Use trailing commas in enums gpuweb/gpuweb#2235, by @litherum.
-
Style changes about command encoder state gpuweb/gpuweb#2234, by @Richard-Yunchao.
-
Remove TODO about "lazy-decay" gpuweb/gpuweb#2233, by @dneto0.
-
Style: make whitespace consistent gpuweb/gpuweb#2232, by @litherum.
-
-
68 pull requests merged.
-
Remove block attribute gpuweb/gpuweb#2104.
-
Workgroup variables are zero-initialized gpuweb/gpuweb#2355.
-
Make invalidity less-magic gpuweb/gpuweb#2284.
-
Cross reference WGSL errors in WebGPU spec gpuweb/gpuweb#2352.
-
Update wording of texture usage to allow parameter gpuweb/gpuweb#2353.
-
Fill out scope explanations and collective operations gpuweb/gpuweb#2351.
-
Require storage texture views to have mipLevelCount == 1 gpuweb/gpuweb#2285.
-
Explain why minBindingSize is "optional" gpuweb/gpuweb#2335.
-
Conversion from float to integral rounds toward zero gpuweb/gpuweb#2330.
-
Do not require function with return value to end with return. gpuweb/gpuweb#2317.
-
Describe tight-packing of image-copy ops, including depth/stencil gpuweb/gpuweb#2331.
-
Fix typo, per code review gpuweb/gpuweb#2332.
-
wgsl: Clarify "static accessed" gpuweb/gpuweb#2334.
-
Specify copyTextureToTexture format compatibility gpuweb/gpuweb#2329.
-
Indirect first instance cleanup gpuweb/gpuweb#2319.
-
Allow identifiers to start with underscore, with exceptions gpuweb/gpuweb#2326.
-
Remove isFinite, isNormal gpuweb/gpuweb#2325.
-
Minor clarifications to GPUImageCopyTextureTagged language gpuweb/gpuweb#2272.
-
Clarify inline issue about minBindingSize in draw-time validation gpuweb/gpuweb#2273.
-
Remove isInf, isNan gpuweb/gpuweb#2311.
-
WGSL source must not include a null character gpuweb/gpuweb#2313.
-
Clarify behaviour of a function gpuweb/gpuweb#2318.
-
Fixes to memory model for barriers and atomics gpuweb/gpuweb#2297.
-
A block comment must affirmatively end before end of source gpuweb/gpuweb#2295.
-
Resolve issue: No other characters count as blankspace gpuweb/gpuweb#2282.
-
Allow 'f' suffix on floats gpuweb/gpuweb#2307.
-
Limits binding numbers to 65535 gpuweb/gpuweb#2286.
-
Delete pipeline statistics queries gpuweb/gpuweb#2296.
-
wgsl: Remove TODO about default rounding mode gpuweb/gpuweb#2287.
-
Remove reset method and exemplify usage gpuweb/gpuweb#2293.
-
Support nesting block comments in CI grammar gpuweb/gpuweb#2291.
-
Describe nesting block comments gpuweb/gpuweb#2288.
-
wgsl: all directives appear before any declarations gpuweb/gpuweb#2281.
-
Neither GPUBuffer nor GPUShaderModule is actually serializable gpuweb/gpuweb#2256.
-
Add validation rules for RenderPassEncoder.executeBundles() gpuweb/gpuweb#2240.
-
Fix conflict between the PRs: 'else if' & 'Behavior analysis' gpuweb/gpuweb#2280.
-
Replace the "elseif" syntax by "else if" gpuweb/gpuweb#2239.
-
Restrict the lhs of assignments to fix a grammar conflict gpuweb/gpuweb#2265.
-
wgsl: describe type aliases gpuweb/gpuweb#2276.
-
Update operator precedence rules gpuweb/gpuweb#2267.
-
Remove obsolete attribute_list in variable_ident_decl gpuweb/gpuweb#2266.
-
wgsl: Require [[location]] for [[interpolate]] gpuweb/gpuweb#2251.
-
Tweak behaviour analysis gpuweb/gpuweb#2271.
-
Add indirect-first-instance feature. gpuweb/gpuweb#2022.
-
Secure context gpuweb/gpuweb#2255.
-
Describe the effective size of a buffer binding. gpuweb/gpuweb#2258.
-
Fix setIndex/VertexBuffer size defaulting allowing any offset. gpuweb/gpuweb#2260.
-
Remove stray extra column in storage class table gpuweb/gpuweb#2262.
-
Behaviours analysis gpuweb/gpuweb#2216.
-
wgsl: Fix a wrong padding size in the structures layout example gpuweb/gpuweb#2269.
-
Update samples to the latest spec. gpuweb/gpuweb#2257.
-
Add a memory model mapping to Vulkan gpuweb/gpuweb#2055.
-
Explicitly describe declaration cycles gpuweb/gpuweb#2253.
-
Whitespace nits gpuweb/gpuweb#2254.
-
Add missing brackets for double brackets gpuweb/gpuweb#2248.
-
Allow out of order module scope declarations gpuweb/gpuweb#2244.
-
Support dot() on vecN<i32> and vecN<u32> gpuweb/gpuweb#2249.
-
minor fix gpuweb/gpuweb#2250.
-
Comment wasm build gpuweb/gpuweb#2247.
-
Use pip from python3 as a subcommand gpuweb/gpuweb#2246.
-
wgsl: Disallow taking the address of a vector component gpuweb/gpuweb#2225.
-
Use trailing commas in enums gpuweb/gpuweb#2235.
-
Style changes about command encoder state gpuweb/gpuweb#2234.
-
Replace depth-clamp feature with depth-clip feature gpuweb/gpuweb#2139.
-
Remove GPUCommandBuffer.executionTime gpuweb/gpuweb#2214.
-
Define 'overload' more generally gpuweb/gpuweb#2226.
-
Style: make whitespace consistent gpuweb/gpuweb#2232.
-
Remove TODO about "lazy-decay" gpuweb/gpuweb#2233.
-
-
7 pull requests closed.
-
Renames GPUProgrammablePassEncoder to GPUProgrammableEncoder gpuweb/gpuweb#2312.
-
Refactor descriptors for attachment load/store/read-only gpuweb/gpuweb#2215.
-
Go from `attribute_list *` to `attribute_list ?` in the grammar gpuweb/gpuweb#2263.
-
Extend references for external scanner section gpuweb/gpuweb#2292.
-
Add /* */ block comments. gpuweb/gpuweb#1470.
-
wgsl: more restrictions on control flow interruption gpuweb/gpuweb#2096.
-
Trailing commas in WebIDL enums are not allowed gpuweb/gpuweb#2230.
-
-
70 pull requests received 396 new comments.
-
[editorial] Describe the fragment stage of the rendering algorithm gpuweb/gpuweb#2162 (27 comments by @kainino0x, @kvark)
-
Add validation rules for RenderPassEncoder.executeBundles() gpuweb/gpuweb#2240 (17 comments by @kainino0x, @Kangz, @Richard-Yunchao)
-
Fixes to memory model for barriers and atomics gpuweb/gpuweb#2297 (14 comments by @alan-baker, @dneto0, @kdashg, @raphlinus, @reeselevine)
-
Add the uniformity analysis to the WGSL spec gpuweb/gpuweb#1571 (14 comments by @Kangz, @RobinMorisset)
-
Add complex assignments (e.g. +=) gpuweb/gpuweb#2241 (13 comments by @dneto0, @jrprice, @Kangz, @kdashg, @kvark, @RobinMorisset)
-
Literal unification, bottom-up gpuweb/gpuweb#2227 (13 comments by @alan-baker, @dneto0, @kdashg, @mehmetoguzderin)
-
Add a memory model mapping to Vulkan gpuweb/gpuweb#2055 (11 comments by @alan-baker, @dneto0, @kdashg, @RobinMorisset)
-
Updates specs regarding error buffers gpuweb/gpuweb#2314 (10 comments by @kainino0x, @kvark, @lokokung)
-
Renames GPUProgrammablePassEncoder to GPUProgrammableEncoder gpuweb/gpuweb#2312 (9 comments by @austinEng, @juj, @kainino0x, @Kangz, @kvark, @lokokung)
-
Add GPUTextureDescriptor option for srgb reinterpretation gpuweb/gpuweb#2336 (8 comments by @kainino0x, @kvark)
-
Extend lifetime of GPUExternalTexture imported from video element gpuweb/gpuweb#2302 (8 comments by @kainino0x, @Kangz, @toji)
-
Allow identifiers to start with underscore, with exceptions gpuweb/gpuweb#2326 (7 comments by @dneto0, @litherum, @mehmetoguzderin)
-
Go from `attribute_list *` to `attribute_list ?` in the grammar gpuweb/gpuweb#2263 (7 comments by @dneto0, @kainino0x, @litherum, @mehmetoguzderin, @RobinMorisset)
-
Support nesting block comments in CI grammar gpuweb/gpuweb#2291 (7 comments by @dneto0, @mehmetoguzderin)
-
Describe the effective size of a buffer binding. gpuweb/gpuweb#2258 (7 comments by @kainino0x, @Kangz)
-
Add proposals/ directory and pipeline statistics extension doc gpuweb/gpuweb#2301 (6 comments by @kainino0x, @kvark, @litherum)
-
Allow out of order module scope declarations gpuweb/gpuweb#2244 (6 comments by @dneto0, @RobinMorisset)
-
Fill out scope explanations and collective operations gpuweb/gpuweb#2351 (5 comments by @alan-baker, @dneto0)
-
A block comment must affirmatively end before end of source gpuweb/gpuweb#2295 (5 comments by @dneto0, @kdashg, @mehmetoguzderin)
-
Add indirect-first-instance feature. gpuweb/gpuweb#2022 (5 comments by @Kangz, @mrshannon, @tidoust)
-
Refactor descriptors for attachment load/store/read-only gpuweb/gpuweb#2215 (4 comments by @kainino0x, @kvark, @toji)
-
Conversion from float to integral rounds toward zero gpuweb/gpuweb#2330 (4 comments by @dneto0, @litherum)
-
Define intrinsic size / natural dimensions of WebGPU canvases gpuweb/gpuweb#2300 (4 comments by @kainino0x, @toji)
-
Restrict the lhs of assignments to fix a grammar conflict gpuweb/gpuweb#2265 (4 comments by @dneto0, @kdashg, @RobinMorisset)
-
wgsl: Add vector and matrix constructor overloads that infer component type gpuweb/gpuweb#2305 (3 comments by @ben-clayton, @dneto0)
-
Describe tight-packing of image-copy ops, including depth/stencil gpuweb/gpuweb#2331 (3 comments by @kainino0x, @kvark)
-
Neither GPUBuffer nor GPUShaderModule is actually serializable gpuweb/gpuweb#2256 (3 comments by @kainino0x, @litherum)
-
Fix setIndex/VertexBuffer size defaulting allowing any offset. gpuweb/gpuweb#2260 (3 comments by @kainino0x, @Kangz)
-
Update WGSL README: instructions about tree-sitter gpuweb/gpuweb#2126 (3 comments by @ben-clayton, @mehmetoguzderin)
-
Allow arrays with pipeline-overridable element count, as store type for workgroup variable gpuweb/gpuweb#2347 (2 comments by @dneto0, @kvark)
-
Require storage texture views to have mipLevelCount == 1 gpuweb/gpuweb#2285 (2 comments by @kainino0x, @kvark)
-
Resolve issue: No other characters count as blankspace gpuweb/gpuweb#2282 (2 comments by @dneto0, @kdashg)
-
wgsl: Disallow taking the address of a vector component gpuweb/gpuweb#2225 (2 comments by @dneto0, @jrprice)
-
Do not require function with return value to end with return. gpuweb/gpuweb#2317 (1 comment by @dneto0)
-
wgsl: more restrictions on control flow interruption gpuweb/gpuweb#2096 (1 comment by @RobinMorisset)
gpuweb/cts
-
3 issues opened.
-
test builtins and expressions in all shader stages gpuweb/cts#841, by @dneto0.
-
Gracefully handle ULP calculations when value is zero-like gpuweb/cts#837, by @zoddicus.
-
add 'create' method for ScalarType WGSL builtins testing framework gpuweb/cts#829, by @dneto0.
-
-
1 issue closed.
-
Gracefully handle ULP calculations when value is zero-like gpuweb/cts#837.
-
-
3 issues received 6 new comments.
-
How should accuracy of floating point builtins be evaluated? gpuweb/cts#792 (4 comments by @dneto0, @zoddicus)
-
add 'create' method for ScalarType WGSL builtins testing framework gpuweb/cts#829 (1 comment by @dneto0)
-
-
32 pull requests opened.
-
Fix typos noticed in conversion utilities gpuweb/cts#844, by @zoddicus.
-
Implement `floor` tests gpuweb/cts#843, by @zoddicus.
-
wgsl: Test that [[interpolate]] requires [[location]] gpuweb/cts#842, by @jrprice.
-
Add test for dpdx/dpdy in loop with dynamic iterations gpuweb/cts#840, by @amaiorano.
-
Remove tests for isNan, isInf gpuweb/cts#839, by @dneto0.
-
Handle zero-like cases with ULP accuracy constraints gpuweb/cts#838, by @zoddicus.
-
Fix bug in testing copyExternalImageToTexture with *-srgb dst formats gpuweb/cts#836, by @shaoboyan.
-
Reland "Adds ETC2 and ASTC types to enable testing when relevant." gpuweb/cts#835, by @lokokung.
-
Fix api,operation,vertex_state,correctness,* gpuweb/cts#834, by @Jiawei-Shao.
-
wgsl: Use 1.0 instead of 0.5 for color tests gpuweb/cts#833, by @jrprice.
-
Use boundary values as numbers instead of bits in `atan` tests gpuweb/cts#832, by @zoddicus.
-
Revert "Adds ETC2 and ASTC types to enable testing when relevant." gpuweb/cts#831, by @kainino0x.
-
Test select with vector condition gpuweb/cts#830, by @dneto0.
-
Test select with scalar condition gpuweb/cts#828, by @dneto0.
-
wgsl: Test built-in variables with indirect dispatch gpuweb/cts#827, by @jrprice.
-
Allow isinf to return false for infinities gpuweb/cts#826, by @dneto0.
-
Bump maxUniformBufferBindingSize from 2^14 to 2^16 gpuweb/cts#825, by @Kangz.
-
wgsl: Add test for [[interpolate]] parameters gpuweb/cts#824, by @jrprice.
-
Update kValidTextureFormatsForCopyE2T gpuweb/cts#823, by @shaoboyan.
-
Implement `asin` tests gpuweb/cts#822, by @zoddicus.
-
Implement `atan` tests gpuweb/cts#821, by @zoddicus.
-
wgsl: Test that vertex shaders must return builtin(position) gpuweb/cts#820, by @jrprice.
-
wgsl: Add tests for duplicated built-in variables gpuweb/cts#819, by @jrprice.
-
wgsl: Add validation tests for nested built-ins gpuweb/cts#818, by @jrprice.
-
wgsl: Add tests for duplicate user-defined IO gpuweb/cts#817, by @jrprice.
-
test isInf builtin function gpuweb/cts#816, by @ben-clayton.
-
Implement `min()` tests for integer types gpuweb/cts#815, by @ben-clayton.
-
wgsl: Add invariant attribute validation tests gpuweb/cts#814, by @jrprice.
-
Implement `any` tests gpuweb/cts#813, by @zoddicus.
-
wgsl: Add tests for shared entry point IO structs gpuweb/cts#812, by @jrprice.
-
shader/execution/builtin: Add `cos()` test gpuweb/cts#811, by @ben-clayton.
-
shader/execution/builtin: Overhaul builtin tests gpuweb/cts#810, by @ben-clayton.
-
-
34 pull requests merged.
-
Fix typos noticed in conversion utilities gpuweb/cts#844.
-
wgsl: Add test for [[interpolate]] parameters gpuweb/cts#824.
-
Handle zero-like cases with ULP accuracy constraints gpuweb/cts#838.
-
Add test for dpdx/dpdy in loop with dynamic iterations gpuweb/cts#840.
-
Remove tests for isNan, isInf gpuweb/cts#839.
-
Test select with vector condition gpuweb/cts#830.
-
Fix api,operation,vertex_state,correctness,* gpuweb/cts#834.
-
wgsl: Use 1.0 instead of 0.5 for color tests gpuweb/cts#833.
-
wgsl: Test built-in variables with indirect dispatch gpuweb/cts#827.
-
Use boundary values as numbers instead of bits in `atan` tests gpuweb/cts#832.
-
Implement `atan` tests gpuweb/cts#821.
-
Test select with scalar condition gpuweb/cts#828.
-
Update kValidTextureFormatsForCopyE2T gpuweb/cts#823.
-
Test isnan gpuweb/cts#800.
-
Bump maxUniformBufferBindingSize from 2^14 to 2^16 gpuweb/cts#825.
-
wgsl: Test that vertex shaders must return builtin(position) gpuweb/cts#820.
-
wgsl: Add validation tests for nested built-ins gpuweb/cts#818.
-
wgsl: Add tests for duplicated built-in variables gpuweb/cts#819.
-
wgsl: Add tests for duplicate user-defined IO gpuweb/cts#817.
-
common/runtime: Improve error handling gpuweb/cts#803.
-
test isInf builtin function gpuweb/cts#816.
-
Implement `min()` tests for integer types gpuweb/cts#815.
-
wgsl: Add tests for shared entry point IO structs gpuweb/cts#812.
-
wgsl: Add tests for missing shader IO attributes gpuweb/cts#801.
-
wgsl: Add invariant attribute validation tests gpuweb/cts#814.
-
shader/execution/builtin: Add `cos()` test gpuweb/cts#811.
-
Implement `any` tests gpuweb/cts#813.
-
Adds ETC2 and ASTC types to enable testing when relevant. gpuweb/cts#805.
-
wgsl: Add built-in variable validation tests gpuweb/cts#775.
-
shader/execution/builtin: Overhaul builtin tests gpuweb/cts#810.
-
Change .html to .https.html, add tests for [SecureContext] gpuweb/cts#809.
-
-
6 pull requests closed.
-
Test isfinite gpuweb/cts#802.
-
Test isnormal gpuweb/cts#804.
-
Allow isinf to return false for infinities gpuweb/cts#826.
-
Implement `asin` tests gpuweb/cts#822.
-
test isInf builtin function gpuweb/cts#799.
-
-
31 pull requests received 198 new comments.
-
Reland "Adds ETC2 and ASTC types to enable testing when relevant." gpuweb/cts#835 (28 comments by @kainino0x, @lokokung)
-
Add test for dpdx/dpdy in loop with dynamic iterations gpuweb/cts#840 (21 comments by @amaiorano, @ben-clayton, @kainino0x)
-
shader/execution/builtin: Overhaul builtin tests gpuweb/cts#810 (18 comments by @ben-clayton, @dneto0)
-
Handle zero-like cases with ULP accuracy constraints gpuweb/cts#838 (13 comments by @ben-clayton, @dneto0, @zoddicus)
-
Implement api,validation,encoding,programmable,pipeline_bind_group_compat gpuweb/cts#777 (3 comments by @haoxli, @kainino0x, @Kangz)
-
Adds ETC2 and ASTC types to enable testing when relevant. gpuweb/cts#805 (3 comments by @kainino0x, @lokokung)
-
wgsl: Test built-in variables with indirect dispatch gpuweb/cts#827 (2 comments by @ben-clayton, @jrprice)
-
Bump maxUniformBufferBindingSize from 2^14 to 2^16 gpuweb/cts#825 (2 comments by @ben-clayton, @Kangz)
-
wgsl: Add validation tests for nested built-ins gpuweb/cts#818 (2 comments by @ben-clayton, @jrprice)
-
Fix bug in testing copyExternalImageToTexture with *-srgb dst formats gpuweb/cts#836 (1 comment by @shaoboyan)
-
Use boundary values as numbers instead of bits in `atan` tests gpuweb/cts#832 (1 comment by @zoddicus)
-
TR publications
-
26 documents published.
-
WebGPU published 13 times as a Working Draft. History of WebGPU
-
WebGPU Shading Language published 13 times as a Working Draft. History of WebGPU Shading Language
-
GitHub
gpuweb/cts
-
3 issues opened.
-
How should accuracy of floating point builtins be evaluated? gpuweb/cts#792, by @zoddicus.
-
WebGPU devices are never destroyed? gpuweb/cts#784, by @ben-clayton.
-
-
1 issue closed.
-
1 issue received 1 new comment.
-
float abs: If e evaluates to the largest negative value, then the result is e. gpuweb/cts#769 (1 comment by @sarahM0)
-
-
39 pull requests opened.
-
Change .html to .https.html, add tests for [SecureContext] gpuweb/cts#809, by @kainino0x.
-
Add Chrome Origin Trial tokens gpuweb/cts#808, by @kainino0x.
-
Serve directory listings to make it easier to find html files gpuweb/cts#807, by @kainino0x.
-
Change .html to .https.html, add tests for [SecureContext] gpuweb/cts#806, by @kainino0x.
-
Adds ETC2 and ASTC types to enable testing when relevant. gpuweb/cts#805, by @lokokung.
-
Test isnormal gpuweb/cts#804, by @dneto0.
-
common/runtime: Improve error handling gpuweb/cts#803, by @ben-clayton.
-
Test isfinite gpuweb/cts#802, by @dneto0.
-
wgsl: Add tests for missing shader IO attributes gpuweb/cts#801, by @jrprice.
-
Test isnan gpuweb/cts#800, by @dneto0.
-
test isInf builtin function gpuweb/cts#799, by @dneto0.
-
Implement `all` tests gpuweb/cts#798, by @zoddicus.
-
Fix extension of `float_built_functions.spec.ts` gpuweb/cts#797, by @ben-clayton.
-
shader/execution: Implement quantizeToF16() tests gpuweb/cts#796, by @ben-clayton.
-
Add `diffULP` utility function gpuweb/cts#795, by @ben-clayton.
-
shader/execution: Add tests for pipeline overrideable constants gpuweb/cts#794, by @ben-clayton.
-
Implement api,validation,capability_checks,features,texture_formats:* gpuweb/cts#793, by @Jiawei-Shao.
-
common/util: Fix subtle bug introduced in #785 gpuweb/cts#791, by @ben-clayton.
-
Add "--gpu-provider-flag <flag>=<value>" gpuweb/cts#790, by @amaiorano.
-
Add a commandline server test runner gpuweb/cts#789, by @ben-clayton.
-
Gruntfile: Incrementally build the node target gpuweb/cts#788, by @ben-clayton.
-
robust_access_vertex: allow invalid draws to be completely noop-ed gpuweb/cts#787, by @austinEng.
-
Remove 'access' parameter from builtins tests gpuweb/cts#786, by @dneto0.
-
src/common/util: Fix 5 second stall at end of test run gpuweb/cts#785, by @ben-clayton.
-
webgpu/shader/exection/builtin: Colorize failure messages gpuweb/cts#783, by @ben-clayton.
-
Add src/common/utils/colors.ts gpuweb/cts#782, by @ben-clayton.
-
webgpu/shader/exection/builtin: Fix checkExpectation() gpuweb/cts#781, by @ben-clayton.
-
webgpu/shader/execution/builtin: Implement `ceil` tests gpuweb/cts#780, by @ben-clayton.
-
Add and use `ansi-colors` as a dependency, fix `checkExpectation()` gpuweb/cts#779, by @ben-clayton.
-
wgsl: Add user-defined IO validation tests gpuweb/cts#778, by @jrprice.
-
Implement api,validation,encoding,programmable,pipeline_bind_group_compat gpuweb/cts#777, by @haoxli.
-
Clamp limits between 0 and their maximum IDL value gpuweb/cts#776, by @austinEng.
-
wgsl: Add built-in variable validation tests gpuweb/cts#775, by @jrprice.
-
CopyExternalImageToTexture should always copy source from top-left gpuweb/cts#774, by @shaoboyan.
-
Filter out non-depth formats from depth_test_input_clamped gpuweb/cts#773, by @kainino0x.
-
Add no-op function .url() to class TestBuilder gpuweb/cts#772, by @sarahM0.
-
Fix maxStorageTexturesPerShaderStage expectations gpuweb/cts#771, by @ben-clayton.
-
Change uniqueId() parameter to string. Update the test plans. gpuweb/cts#770, by @sarahM0.
-
Remove 'read-only' from storageTextureBindingEntries() gpuweb/cts#768, by @ben-clayton.
-
-
26 pull requests merged.
-
Add Chrome Origin Trial tokens gpuweb/cts#808.
-
Serve directory listings to make it easier to find html files gpuweb/cts#807.
-
wgsl: Add user-defined IO validation tests gpuweb/cts#778.
-
Implement `all` tests gpuweb/cts#798.
-
Fix extension of `float_built_functions.spec.ts` gpuweb/cts#797.
-
Add `diffULP` utility function gpuweb/cts#795.
-
Add "--gpu-provider-flag <flag>=<value>" gpuweb/cts#790.
-
Remove 'access' parameter from builtins tests gpuweb/cts#786.
-
common/util: Fix subtle bug introduced in #785 gpuweb/cts#791.
-
Add a commandline server test runner gpuweb/cts#789.
-
Gruntfile: Incrementally build the node target gpuweb/cts#788.
-
webgpu/shader/exection/builtin: Colorize failure messages gpuweb/cts#783.
-
src/common/util: Fix 5 second stall at end of test run gpuweb/cts#785.
-
webgpu/shader/execution/builtin: Implement `ceil` tests gpuweb/cts#780.
-
webgpu/shader/exection/builtin: Fix checkExpectation() gpuweb/cts#781.
-
Add src/common/utils/colors.ts gpuweb/cts#782.
-
Clamp limits between 0 and their maximum IDL value gpuweb/cts#776.
-
Add no-op function .url() to class TestBuilder gpuweb/cts#772.
-
Change uniqueId() parameter to string. Update the test plans. gpuweb/cts#770.
-
Fix maxStorageTexturesPerShaderStage expectations gpuweb/cts#771.
-
Remove 'read-only' from storageTextureBindingEntries() gpuweb/cts#768.
-
Add wgsl how to contribute doc and readme gpuweb/cts#761.
-
-
3 pull requests closed.
-
Change .html to .https.html, add tests for [SecureContext] gpuweb/cts#806.
-
wgsl: sections 16 and 3 proposed test plan gpuweb/cts#727.
-
-
25 pull requests received 114 new comments.
-
shader/execution: Add tests for pipeline overrideable constants gpuweb/cts#794 (19 comments by @ben-clayton, @jrprice, @kainino0x, @shrekshao)
-
Implement api,validation,encoding,programmable,pipeline_bind_group_compat gpuweb/cts#777 (14 comments by @austinEng, @haoxli, @Kangz)
-
Adds ETC2 and ASTC types to enable testing when relevant. gpuweb/cts#805 (10 comments by @kainino0x, @lokokung)
-
wgsl: Add user-defined IO validation tests gpuweb/cts#778 (8 comments by @ben-clayton, @jrprice, @kainino0x)
-
wgsl: Add test plan Value-testing, float, and integer built-in functions gpuweb/cts#746 (6 comments by @kainino0x, @sarahM0)
-
webgpu/api/validation: Fix vertex_shader_input_location_limit test gpuweb/cts#767 (5 comments by @ben-clayton, @kainino0x)
-
CopyExternalImageToTexture should always copy source from top-left gpuweb/cts#774 (2 comments by @kainino0x, @shaoboyan)
-
Clamp limits between 0 and their maximum IDL value gpuweb/cts#776 (2 comments by @austinEng, @Kangz)
-
Serve directory listings to make it easier to find html files gpuweb/cts#807 (1 comment by @austinEng)
-
Add and use `ansi-colors` as a dependency, fix `checkExpectation()` gpuweb/cts#779 (1 comment by @ben-clayton)
-
gpuweb/gpuweb
-
33 issues opened.
-
Atomic concerns gpuweb/gpuweb#2229, by @raphlinus.
-
How should WGSL map to the availability/visibility parts of the Vulkan Memory Model? gpuweb/gpuweb#2228, by @RobinMorisset.
-
[WGSL] Web IDL validation errors gpuweb/gpuweb#2224, by @autokagami.
-
[webgpu] Web IDL validation errors gpuweb/gpuweb#2223, by @autokagami.
-
C-style casts gpuweb/gpuweb#2220, by @litherum.
-
Clarify the rule about discard in a continuing block gpuweb/gpuweb#2219, by @RobinMorisset.
-
Considerations for timing attacks using timestamp queries gpuweb/gpuweb#2218, by @litherum.
-
mapSync on Workers - and possibly on the main thread gpuweb/gpuweb#2217, by @kainino0x.
-
☂️ Initial internal feedback gpuweb/gpuweb#2213, by @litherum.
-
Consider replacing elseif with else if gpuweb/gpuweb#2212, by @litherum.
-
Consider not requiring braces in some situations gpuweb/gpuweb#2211, by @litherum.
-
Allow f suffix for float literals gpuweb/gpuweb#2210, by @litherum.
-
Parallels between WGSL and SPIR-V don't justify `let` gpuweb/gpuweb#2207, by @jimblandy.
-
Literal unification gpuweb/gpuweb#2200, by @alan-baker.
-
Consider validating stripIndexFormat at draw time. gpuweb/gpuweb#2199, by @Kangz.
-
Removing the COPY* texture usages gpuweb/gpuweb#2196, by @Kangz.
-
Exposing GPU hardware information to web developers in WebGPU gpuweb/gpuweb#2195, by @beaufortfrancois.
-
GPUAdapter.name considered harmful gpuweb/gpuweb#2191, by @litherum.
-
Streaming implementations and indirect draws/dispatches gpuweb/gpuweb#2189, by @litherum.
-
Storage variables shouldn't be required to be structs gpuweb/gpuweb#2188, by @litherum.
-
wgsl: plural form `attribute_lists` should be explicitly explained somewhere gpuweb/gpuweb#2184, by @cloudhan.
-
how GPUBuffer becomes mapping again? gpuweb/gpuweb#2181, by @sensetimeXsy.
-
wgsl: builtin functions described with type assertions and with function prototypes gpuweb/gpuweb#2176, by @dneto0.
-
wgsl: enable_directive grammar and example mismatch gpuweb/gpuweb#2173, by @cloudhan.
-
Specify cube texture filtering as seamless gpuweb/gpuweb#2171, by @mrshannon.
-
wgsl: 0x12345 is an integer literal, not a hex float gpuweb/gpuweb#2169, by @dneto0.
-
wgsl: 0x12345p12 should be a hex float (missing case from the token grammar) gpuweb/gpuweb#2168, by @dneto0.
-
Adapter preference for "not used for display" gpuweb/gpuweb#2167, by @kainino0x.
-
Atomic operations support gpuweb/gpuweb#2166, by @AlexandruPopovici.
-
Disambiguate "storage" in WGSL spec gpuweb/gpuweb#2163, by @kvark.
-
Listing all available adapters gpuweb/gpuweb#2160, by @ghost.
-
wgsl: description of how a function call executes needs to accommodate builtin functions gpuweb/gpuweb#2157, by @dneto0.
-
Consider removing GPUCommandBuffer.executionTime gpuweb/gpuweb#2156, by @Kangz.
-
-
22 issues closed.
-
Request for a GPUCommandEncoder fillBuffer operation gpuweb/gpuweb#1968.
-
requestDevice and device properties gpuweb/gpuweb#474.
-
requestAdapter and adapter properties gpuweb/gpuweb#473.
-
wgsl: 0x12345 is an integer literal, not a hex float gpuweb/gpuweb#2169.
-
Add spec text for "Navigator"/"WorkerNavigator.gpu" gpuweb/gpuweb#471.
-
Recursion in WGSL? gpuweb/gpuweb#1471.
-
wgsl: plural form `attribute_lists` should be explicitly explained somewhere gpuweb/gpuweb#2184.
-
Implicit flat interpolation on integers is confusing gpuweb/gpuweb#2061.
-
how GPUBuffer becomes mapping again? gpuweb/gpuweb#2181.
-
Packed vector types gpuweb/gpuweb#780.
-
WGSL: Drop requirement to ignore() atomic operations? gpuweb/gpuweb#2090.
-
wgsl: 0x12345p12 should be a hex float (missing case from the token grammar) gpuweb/gpuweb#2168.
-
Add magic `_` identifier (and remove `ignore()`) gpuweb/gpuweb#2115.
-
Should hex exponents be optional? gpuweb/gpuweb#2036.
-
Atomic operations support gpuweb/gpuweb#2166.
-
Writing to different components of a vector from different invocations gpuweb/gpuweb#2112.
-
Listing all available adapters gpuweb/gpuweb#2160.
-
Which corner is the origin of a copyExternalImageToTexture? gpuweb/gpuweb#2110.
-
Require a render pipeline flag to enable sample-rate shading gpuweb/gpuweb#2133.
-
Hardware limits related API design gpuweb/gpuweb#2151.
-
-
56 issues received 252 new comments.
-
GPUAdapter.name considered harmful gpuweb/gpuweb#2191 (25 comments by @beaufortfrancois, @juj, @kainino0x, @Kangz, @litherum, @magcius, @mehmetoguzderin, @mrshannon, @RafaelCintron, @toji)
-
WGSL requires dynamic matrix indexing, which requires surprising SPIR-V gpuweb/gpuweb#1782 (20 comments by @ben-clayton, @jimblandy, @krogovin, @kvark, @litherum)
-
Consider removing GPUCommandBuffer.executionTime gpuweb/gpuweb#2156 (16 comments by @kainino0x, @Kangz, @kvark, @litherum)
-
MSL does not support address-of vector component gpuweb/gpuweb#1931 (15 comments by @dneto0, @jrprice, @kdashg, @litherum)
-
Proposal: Add a usage bit to allow creating a view with different format when creating a texture gpuweb/gpuweb#168 (14 comments by @Jiawei-Shao, @kainino0x, @Kangz, @litherum, @shaoboyan)
-
Literal unification gpuweb/gpuweb#2200 (13 comments by @ben-clayton, @dneto0, @kdashg, @mehmetoguzderin, @RobinMorisset)
-
wgsl: 0x12345 is an integer literal, not a hex float gpuweb/gpuweb#2169 (8 comments by @cloudhan, @dneto0, @mehmetoguzderin)
-
Streaming implementations and indirect draws/dispatches gpuweb/gpuweb#2189 (7 comments by @kainino0x, @Kangz, @litherum)
-
Consider replacing elseif with else if gpuweb/gpuweb#2212 (7 comments by @ben-clayton, @dneto0, @litherum, @RobinMorisset)
-
wgsl: plural form `attribute_lists` should be explicitly explained somewhere gpuweb/gpuweb#2184 (7 comments by @cloudhan, @dneto0, @mehmetoguzderin)
-
Define a MIME type gpuweb/gpuweb#1682 (7 comments by @annevk, @dneto0, @kainino0x, @kdashg, @litherum)
-
Consider not requiring braces in some situations gpuweb/gpuweb#2211 (6 comments by @alan-baker, @dneto0, @Kangz, @kdashg, @litherum, @magcius)
-
Make GPUDevice.lost behave like it has [SameObject] gpuweb/gpuweb#2147 (5 comments by @greggman, @kainino0x)
-
Parallels between WGSL and SPIR-V don't justify `let` gpuweb/gpuweb#2207 (5 comments by @dneto0, @jimblandy, @kvark)
-
wgsl: enable_directive grammar and example mismatch gpuweb/gpuweb#2173 (5 comments by @cloudhan, @dneto0, @kdashg)
-
Which corner is the origin of a copyExternalImageToTexture? gpuweb/gpuweb#2110 (5 comments by @kainino0x, @kenrussell)
-
What OSes and hardware are we targeting? gpuweb/gpuweb#1069 (4 comments by @juj, @kainino0x, @Kangz)
-
The order of top-level declarations should be irrelevant gpuweb/gpuweb#875 (4 comments by @dneto0, @kdashg, @RobinMorisset)
-
Storage variables shouldn't be required to be structs gpuweb/gpuweb#2188 (4 comments by @ben-clayton, @jrprice, @kdashg, @litherum)
-
Consider making ShaderModule compilation errors special gpuweb/gpuweb#2119 (3 comments by @greggman, @kainino0x)
-
Clarify the rule about discard in a continuing block gpuweb/gpuweb#2219 (3 comments by @alan-baker, @litherum)
-
Refactor API for load/store OP and read-only attachments gpuweb/gpuweb#2131 (3 comments by @kainino0x, @kvark, @toji)
-
texture_2d_depth for loading depth values might be unnecessary? gpuweb/gpuweb#2094 (3 comments by @kainino0x, @Kangz, @litherum)
-
wgsl: builtin functions described with type assertions and with function prototypes gpuweb/gpuweb#2176 (3 comments by @dneto0)
-
Require a render pipeline flag to enable sample-rate shading gpuweb/gpuweb#2133 (3 comments by @kainino0x, @litherum, @magcius)
-
Storage buffer access: compatibility with the pipeline layout gpuweb/gpuweb#1837 (2 comments by @litherum)
-
Consider validating stripIndexFormat at draw time. gpuweb/gpuweb#2199 (2 comments by @Kangz, @magcius)
-
Requiring integral-valued float literals to have .0 seems gratuitous gpuweb/gpuweb#739 (1 comment by @litherum)
-
Consider improving variable array indexing for 'let' arrays gpuweb/gpuweb#2086 (1 comment by @litherum)
-
Can value of visibility attribute in GPUBindGroupLayoutBinding be NONE ? gpuweb/gpuweb#405 (1 comment by @Kangz)
-
wgsl: 0x12345p12 should be a hex float (missing case from the token grammar) gpuweb/gpuweb#2168 (1 comment by @dneto0)
-
Specify behavior of GPUCommandBuffer.executionTime on devices without timestamp queries. gpuweb/gpuweb#1773 (1 comment by @kvark)
-
-
34 pull requests opened.
-
Trailing commas in WebIDL enums are not allowed gpuweb/gpuweb#2230, by @litherum.
-
Literal unification, bottom-up gpuweb/gpuweb#2227, by @dneto0.
-
Define 'overload' more generally gpuweb/gpuweb#2226, by @dneto0.
-
wgsl: Disallow taking the address of a vector component gpuweb/gpuweb#2225, by @jrprice.
-
Replace ptr/ref todo with link to memory views gpuweb/gpuweb#2222, by @alan-baker.
-
Literal unification gpuweb/gpuweb#2221, by @alan-baker.
-
Behaviours analysis gpuweb/gpuweb#2216, by @RobinMorisset.
-
Refactor descriptors for attachment load/store/read-only gpuweb/gpuweb#2215, by @kainino0x.
-
Remove GPUCommandBuffer.executionTime gpuweb/gpuweb#2214, by @toji.
-
Add GPUComandEncoder.fillBuffer() gpuweb/gpuweb#2208, by @litherum.
-
Remove pipeline compability section gpuweb/gpuweb#2206, by @alan-baker.
-
Remove statement TODOs gpuweb/gpuweb#2205, by @alan-baker.
-
Remove TODOs from expression sections gpuweb/gpuweb#2202, by @alan-baker.
-
wgsl: decimal point is optional in float literal having an exponent gpuweb/gpuweb#2201, by @dneto0.
-
Add validation rules for RenderBundleEncoder.finish() gpuweb/gpuweb#2198, by @Richard-Yunchao.
-
Minor typo "a memory allocations" gpuweb/gpuweb#2194, by @OmarShehata.
-
timestamp-query is unimplementable on TBDR architectures gpuweb/gpuweb#2190, by @litherum.
-
wgsl: OpXMod -> OpXRem for integers gpuweb/gpuweb#2187, by @munrocket.
-
Add validation rules for RenderBundleEncoderDescriptor gpuweb/gpuweb#2186, by @Richard-Yunchao.
-
Use consistent repeat expression gpuweb/gpuweb#2185, by @mehmetoguzderin.
-
Integral pipeline IO must be specified interpolate(flat) gpuweb/gpuweb#2183, by @dneto0.
-
Add ETC2 and ASTC texture formats and feature name documentation gpuweb/gpuweb#2180, by @lokokung.
-
wgsl: Function call section accommodates built-in functions gpuweb/gpuweb#2179, by @dneto0.
-
describe builtin functions as prototypes gpuweb/gpuweb#2178, by @dneto0.
-
Fix render by closing table gpuweb/gpuweb#2177, by @mehmetoguzderin.
-
wgsl: Rename `smoothStep` to `smoothstep` gpuweb/gpuweb#2175, by @ben-clayton.
-
wgsl: Add scalar overloads of any() and all() gpuweb/gpuweb#2174, by @ben-clayton.
-
Remove COPY buffer usages gpuweb/gpuweb#2172, by @kainino0x.
-
wgsl: 0x123p0 is a hex float gpuweb/gpuweb#2170, by @dneto0.
-
Make hex exponent suffix optional for floats gpuweb/gpuweb#2164, by @mehmetoguzderin.
-
[editorial] Describe the fragment stage of the rendering algorithm gpuweb/gpuweb#2162, by @kvark.
-
Raise maxUniformBufferBindingSize to 64k gpuweb/gpuweb#2159, by @kainino0x.
-
Make copyExternalImageToTexture orientation configurable gpuweb/gpuweb#2158, by @kainino0x.
-
wgsl: editorial cleanup: memory vs. storage gpuweb/gpuweb#2155, by @dneto0.
-
-
27 pull requests merged.
-
Add validation rules for RenderBundleEncoder.finish() gpuweb/gpuweb#2198.
-
Replace ptr/ref todo with link to memory views gpuweb/gpuweb#2222.
-
Remove pipeline compability section gpuweb/gpuweb#2206.
-
Remove TODOs from expression sections gpuweb/gpuweb#2202.
-
Remove statement TODOs gpuweb/gpuweb#2205.
-
Add GPUComandEncoder.fillBuffer() gpuweb/gpuweb#2208.
-
wgsl: decimal point is optional in float literal having an exponent gpuweb/gpuweb#2201.
-
Extend color spaces; more formats in copyExternalImageToTexture gpuweb/gpuweb#2142.
-
Raise maxUniformBufferBindingSize to 64k gpuweb/gpuweb#2159.
-
timestamp-query is unimplementable on TBDR architectures gpuweb/gpuweb#2190.
-
wgsl: Add scalar overloads of any() and all() gpuweb/gpuweb#2174.
-
wgsl: Function call section accommodates built-in functions gpuweb/gpuweb#2179.
-
Minor typo "a memory allocations" gpuweb/gpuweb#2194.
-
Add validation rules for RenderBundleEncoderDescriptor gpuweb/gpuweb#2186.
-
Use consistent repeat expression gpuweb/gpuweb#2185.
-
Add ETC2 and ASTC texture formats and feature name documentation gpuweb/gpuweb#2180.
-
describe builtin functions as prototypes gpuweb/gpuweb#2178.
-
Integral pipeline IO must be specified interpolate(flat) gpuweb/gpuweb#2183.
-
Fix render by closing table gpuweb/gpuweb#2177.
-
Allow syntactic navigation and styling gpuweb/gpuweb#2143.
-
wgsl: remove ignore, add phony-assignment gpuweb/gpuweb#2127.
-
wgsl: editorial cleanup: memory vs. storage gpuweb/gpuweb#2155.
-
Make hex exponent suffix optional for floats gpuweb/gpuweb#2164.
-
wgsl: Add limits section, and "spurious" failure gpuweb/gpuweb#1997.
-
Clarify memory locations accessed when writing a vector component gpuweb/gpuweb#2152.
-
Fix inequality symbol in compute shader grid bound gpuweb/gpuweb#2153.
-
Fix some misplaced characters in the `select` description. gpuweb/gpuweb#2154.
-
-
3 pull requests closed.
-
Add round_up_size attribute gpuweb/gpuweb#1909.
-
wgsl: 0x123p0 is a hex float gpuweb/gpuweb#2170.
-
Make copyExternalImageToTexture orientation configurable gpuweb/gpuweb#2158.
-
-
33 pull requests received 237 new comments.
-
Add validation rules for RenderBundleEncoder.finish() gpuweb/gpuweb#2198 (18 comments by @kainino0x, @Richard-Yunchao)
-
Allow syntactic navigation and styling gpuweb/gpuweb#2143 (15 comments by @dneto0, @kdashg, @mehmetoguzderin)
-
wgsl: Function call section accommodates built-in functions gpuweb/gpuweb#2179 (13 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
Raise maxUniformBufferBindingSize to 64k gpuweb/gpuweb#2159 (11 comments by @kainino0x, @Kangz, @kvark, @RafaelCintron)
-
Add a memory model mapping to Vulkan gpuweb/gpuweb#2055 (10 comments by @alan-baker, @dneto0, @RobinMorisset)
-
wgsl: remove ignore, add phony-assignment gpuweb/gpuweb#2127 (10 comments by @dneto0, @jrprice, @kdashg, @kvark)
-
wgsl: editorial cleanup: memory vs. storage gpuweb/gpuweb#2155 (10 comments by @alan-baker, @dneto0, @kdashg)
-
Replace depth-clamp feature with depth-clip feature gpuweb/gpuweb#2139 (8 comments by @kainino0x, @Kangz, @kvark)
-
Add ETC2 and ASTC texture formats and feature name documentation gpuweb/gpuweb#2180 (7 comments by @kainino0x, @Kangz, @kvark, @lokokung)
-
wgsl: Add limits section, and "spurious" failure gpuweb/gpuweb#1997 (7 comments by @dneto0, @kdashg, @kvark)
-
wgsl: replace stride attribute with array type parameter gpuweb/gpuweb#2103 (6 comments by @dneto0, @kdashg, @kvark)
-
wgsl: Rename `smoothStep` to `smoothstep` gpuweb/gpuweb#2175 (4 comments by @dneto0, @kdashg, @kvark)
-
Make copyExternalImageToTexture orientation configurable gpuweb/gpuweb#2158 (3 comments by @kainino0x, @kvark)
-
wgsl: decimal point is optional in float literal having an exponent gpuweb/gpuweb#2201 (2 comments by @dneto0, @kdashg)
-
timestamp-query is unimplementable on TBDR architectures gpuweb/gpuweb#2190 (2 comments by @Kangz, @litherum)
-
Clarify memory locations accessed when writing a vector component gpuweb/gpuweb#2152 (2 comments by @dneto0, @kdashg)
-
wgsl: detailed semantics of integer division and remainder gpuweb/gpuweb#1830 (1 comment by @kdashg)
-
Add validation rules for RenderBundleEncoderDescriptor gpuweb/gpuweb#2186 (1 comment by @Richard-Yunchao)
TR publications
-
14 documents published.
-
WebGPU published 8 times as a Working Draft. History of WebGPU
-
WebGPU Shading Language published 6 times as a Working Draft. History of WebGPU Shading Language
-
GitHub
gpuweb/gpuweb
-
36 issues opened.
-
Hardware limits related API design gpuweb/gpuweb#2151, by @gyagp.
-
wgsl: Literals are context dependent gpuweb/gpuweb#2149, by @cloudhan.
-
Make GPUDevice.lost behave like it has [SameObject] gpuweb/gpuweb#2147, by @kainino0x.
-
Question on how to make and use library. gpuweb/gpuweb#2144, by @cloudhan.
-
Remove requirement that function / builtin return values are always consumed gpuweb/gpuweb#2138, by @ben-clayton.
-
Request for compute: anyInvocation() and allInvocation() gpuweb/gpuweb#2137, by @krogovin.
-
Consider allowing null entries in binding groups gpuweb/gpuweb#2134, by @magcius.
-
Require a render pipeline flag to enable sample-rate shading gpuweb/gpuweb#2133, by @kainino0x.
-
Refactor API for load/store OP and read-only attachments gpuweb/gpuweb#2131, by @kvark.
-
wgsl: finding most-significant bit and least-significant bit gpuweb/gpuweb#2130, by @dneto0.
-
wgsl: support bitfield insertion and extraction gpuweb/gpuweb#2129, by @dneto0.
-
Determine GPUExternalTexture lifetime gpuweb/gpuweb#2124, by @Kangz.
-
Consider adding GLSL's uaddCarry and friends. gpuweb/gpuweb#2123, by @Kangz.
-
Consider having allowing separators in integer literals. gpuweb/gpuweb#2122, by @Kangz.
-
Consider making ShaderModule compilation errors special gpuweb/gpuweb#2119, by @Kangz.
-
wgsl: limit layout size of a type gpuweb/gpuweb#2118, by @dneto0.
-
Add magic `_` identifier (and remove `ignore()`) gpuweb/gpuweb#2115, by @kdashg.
-
Develop a visual language for the syntactic rules in WGSL specification gpuweb/gpuweb#2114, by @mehmetoguzderin.
-
Writing to different components of a vector from different invocations gpuweb/gpuweb#2112, by @jrprice.
-
Compute shader piplelines are not cached gpuweb/gpuweb#2111, by @qjia7.
-
Which corner is the origin of a copyExternalImageToTexture? gpuweb/gpuweb#2110, by @kainino0x.
-
Inquiry about force running on NIVIDA GPU not on built-in one of motherboard gpuweb/gpuweb#2107, by @supcn.
-
Fragment shader [[builtin(position)]] definition is misleading gpuweb/gpuweb#2106, by @kainino0x.
-
wgsl: Sketch design for arrays of buffer-bindings gpuweb/gpuweb#2105, by @dneto0.
-
Support arrays of top-level structures in storage buffers gpuweb/gpuweb#2102, by @kvark.
-
Implement "depth_clamp" on D3D12 gpuweb/gpuweb#2100, by @Jiawei-Shao.
-
Validate that requested alignment limits are power of two gpuweb/gpuweb#2099, by @Kangz.
-
GPUExternalTexture: Deal with Chroma Reconstruction gpuweb/gpuweb#2098, by @shaoboyan.
-
texture_2d_depth for loading depth values might be unnecessary? gpuweb/gpuweb#2094, by @magcius.
-
further restrict where a control-flow interruption can be placed gpuweb/gpuweb#2093, by @dneto0.
-
WGSL: Drop requirement to ignore() atomic operations? gpuweb/gpuweb#2090, by @m-schuetz.
-
Consider improving variable array indexing for 'let' arrays gpuweb/gpuweb#2086, by @brendan-duncan.
-
Spec the texture-compression-etc and texture-compression-astc extensions gpuweb/gpuweb#2083, by @Kangz.
-
Consider adding a `textureLuminance` builtin function for `external_texture` gpuweb/gpuweb#2082, by @Kangz.
-
-
108 issues closed.
-
[wgsl] validate the grammar on CI gpuweb/gpuweb#578.
-
Remove requirement that function / builtin return values are always consumed gpuweb/gpuweb#2138.
-
wgsl: Literals are context dependent gpuweb/gpuweb#2149.
-
Ability to issue command and create resources from workers gpuweb/gpuweb#9.
-
Out-of-bounds behavior in shaders gpuweb/gpuweb#35.
-
Question on how to make and use library. gpuweb/gpuweb#2144.
-
Feedback after porting a few hundred GLSL lines to WGSL gpuweb/gpuweb#1364.
-
Simplifications and improvements to WGSL syntax gpuweb/gpuweb#1466.
-
Frag output uniformity requirements? gpuweb/gpuweb#1515.
-
Restrict where entry point IO types can be used gpuweb/gpuweb#1526.
-
wgsl: Restrict `modf` and `frexp`pointer parameter storage classes gpuweb/gpuweb#1846.
-
Why is `block` the only keyword attribute? gpuweb/gpuweb#1752.
-
Brain dump on all things wrong with WGSL gpuweb/gpuweb#2048.
-
atomicCompareExchangeWeak is not idiomatic gpuweb/gpuweb#2021.
-
Swapchain texture usage gpuweb/gpuweb#437.
-
Compute shader piplelines are not cached gpuweb/gpuweb#2111.
-
Implicit GPUTextureSampleType for texture_2d<f32> gpuweb/gpuweb#2064.
-
pr-preview doesn't run for WGSL spec PRs gpuweb/gpuweb#1405.
-
Using structured buffers in HLSL gpuweb/gpuweb#1961.
-
Consider adding GLSL's uaddCarry and friends. gpuweb/gpuweb#2123.
-
Allow emulated capabilities for testing? gpuweb/gpuweb#432.
-
Inquiry about force running on NIVIDA GPU not on built-in one of motherboard gpuweb/gpuweb#2107.
-
Support arrays of top-level structures in storage buffers gpuweb/gpuweb#2102.
-
Should all the constant-only interfaces be namespaces instead? gpuweb/gpuweb#304.
-
Metal is unable to copy depth formats into colors gpuweb/gpuweb#1692.
-
Inconsistency between bind groups and vertex buffers gpuweb/gpuweb#251.
-
Sampler address modes gpuweb/gpuweb#348.
-
Have a plan for planar format like YUV 420 and 422. gpuweb/gpuweb#625.
-
Restrictions on range of zeroing-out-buffer method? gpuweb/gpuweb#2009.
-
WebGL - WebGPU interoperability gpuweb/gpuweb#18.
-
WebGL - WebGPU interop: WebGPU computation API with existing WebGL gpuweb/gpuweb#17.
-
Possibility to optimize global app-specific and frame-specific state binding? gpuweb/gpuweb#1414.
-
First WebGPU Stable Release / Prognosis ? gpuweb/gpuweb#1276.
-
Thoughts on uploading contents from ImageBitmap to WebGPU gpuweb/gpuweb#700.
-
Can GPUExtent3D have fewer than 3 entries? gpuweb/gpuweb#781.
-
Feedback on GPUCommandEncoder missing a transfer pass encoder gpuweb/gpuweb#1969.
-
Is Timestamp Query not supported on Apple M1? gpuweb/gpuweb#1956.
-
Fix Y2K bugs in WebGPU gpuweb/gpuweb#2000.
-
Copy Validation of Depth/Stencil textures gpuweb/gpuweb#997.
-
Publish WebGPU as First Public Working Draft gpuweb/gpuweb#1210.
-
Texture dimensions should be required to be >=1 (forbid zero-sized textures) gpuweb/gpuweb#1200.
-
Summary of current robust buffer access opinions gpuweb/gpuweb#955.
-
Barriers? gpuweb/gpuweb#862.
-
Sampler/image not-really feedback in compute shader gpuweb/gpuweb#858.
-
Texture usage validation constraints gpuweb/gpuweb#808.
-
Compressed texture operations need to be well-defined gpuweb/gpuweb#807.
-
validation rules for resource usage tracking gpuweb/gpuweb#769.
-
Add "depth32float_stencil8" extension for GPUTextureFormat gpuweb/gpuweb#755.
-
Should various unused resource usages be counted toward usage scopes? gpuweb/gpuweb#657.
-
Investigation: Query API gpuweb/gpuweb#614.
-
Using buffer with multiple usages in the same compute pass gpuweb/gpuweb#547.
-
Maximum uniform buffer size/alignment in GPULimits gpuweb/gpuweb#487.
-
Asynchronous pipeline creation gpuweb/gpuweb#486.
-
[meta] Vertex Input gpuweb/gpuweb#484.
-
Add error code to GPUValidationError gpuweb/gpuweb#464.
-
Validation rules for GPUVertexAttributeDescriptor offset gpuweb/gpuweb#463.
-
Coordinate systems gpuweb/gpuweb#416.
-
Maximum and minimum total resources in GPUBindGroup/Layout? gpuweb/gpuweb#409.
-
Max number for dynamic resources gpuweb/gpuweb#406.
-
GPUVertexBufferDescriptor 's stride size needs to be a multiple of 4 bytes gpuweb/gpuweb#404.
-
Rendering to OffscreenCanvas on non-yielding workers gpuweb/gpuweb#403.
-
Add render sub-passes gpuweb/gpuweb#393.
-
Investigation: Copy With Compressed Texture Format gpuweb/gpuweb#363.
-
Out-of-bounds clamping may require additional draw-call validation gpuweb/gpuweb#311.
-
Investigation: Multisample Coverage gpuweb/gpuweb#267.
-
Determine limits for point size gpuweb/gpuweb#263.
-
Consider exposing WebGPU as an ES6 module? gpuweb/gpuweb#243.
-
shader programming model: signed integer representation gpuweb/gpuweb#233.
-
shader float bit widths: restrict to 32 bits gpuweb/gpuweb#230.
-
shader int bit widths: restrict to 32 bits gpuweb/gpuweb#229.
-
shaders: equivalent of gl_PointCoord and gl_PointSize don't exist in HLSL gpuweb/gpuweb#332.
-
determine data layout rules inside buffers gpuweb/gpuweb#166.
-
What image formats are supported? What sampling options? gpuweb/gpuweb#163.
-
Proposals for buffer operations (immediate uploads, buffer mapping) gpuweb/gpuweb#138.
-
Investigation: Rasterization State gpuweb/gpuweb#137.
-
Question: shall we set default values for parameters of descriptors gpuweb/gpuweb#136.
-
Investigation: Viewport and Scissor Test gpuweb/gpuweb#120.
-
Investigation: Stencil Test gpuweb/gpuweb#118.
-
Proposal: Run all index buffers through a compute shader validator gpuweb/gpuweb#117.
-
Proposal: Dynamic uniform and storage buffer offsets gpuweb/gpuweb#116.
-
Investigation: Multisampled Render Targets and Resolve Operations gpuweb/gpuweb#108.
-
"Either" attachment load and store ops gpuweb/gpuweb#97.
-
Investigation: Texture View gpuweb/gpuweb#79.
-
Occlusion Queries gpuweb/gpuweb#72.
-
Copies investigation (+ proposals) gpuweb/gpuweb#69.
-
Investigation: Cube dimensionality gpuweb/gpuweb#68.
-
Use callbacks directly? gpuweb/gpuweb#52.
-
[meta] Big questions gpuweb/gpuweb#48.
-
Resource Capabilities gpuweb/gpuweb#37.
-
SPIR-V sources of undefined behavior gpuweb/gpuweb#34.
-
SPIR-V robust resource access gpuweb/gpuweb#33.
-
Depth test state gpuweb/gpuweb#32.
-
Indirect draw/dispatch commands investigation gpuweb/gpuweb#31.
-
Memory barriers investigations gpuweb/gpuweb#27.
-
Pipeline states investigation gpuweb/gpuweb#26.
-
Render target / render pass investigation gpuweb/gpuweb#23.
-
Allow GPUFence.onCompletion to be called with any value gpuweb/gpuweb#377.
-
Require [[block]] on nested structures in WGSL gpuweb/gpuweb#1398.
-
-
175 issues received 425 new comments.
-
Implement "depth_clamp" on D3D12 gpuweb/gpuweb#2100 (30 comments by @jenatali, @Jiawei-Shao, @kainino0x, @Kangz, @kvark, @litherum, @magcius)
-
GPUExternalTexture: Deal with Chroma Reconstruction gpuweb/gpuweb#2098 (19 comments by @bjjones, @kainino0x, @kdashg, @litherum, @shaoboyan)
-
Consider making ShaderModule compilation errors special gpuweb/gpuweb#2119 (17 comments by @dneto0, @kainino0x, @Kangz, @kvark, @litherum, @magcius)
-
Add magic `_` identifier (and remove `ignore()`) gpuweb/gpuweb#2115 (16 comments by @ben-clayton, @dneto0, @kainino0x, @kvark, @m-schuetz, @magcius)
-
timestamp-query is unimplementable on TBDR architectures gpuweb/gpuweb#2046 (14 comments by @haoxli, @kainino0x, @Kangz, @kvark, @litherum, @mrshannon)
-
texture_2d_depth for loading depth values might be unnecessary? gpuweb/gpuweb#2094 (12 comments by @kainino0x, @Kangz, @kdashg, @kvark, @magcius, @magicus)
-
Error with literal indexing in a shader gpuweb/gpuweb#1803 (11 comments by @ben-clayton, @dneto0, @jrprice, @kdashg, @kvark, @RobinMorisset)
-
FMod vs FRem semantics gpuweb/gpuweb#1696 (11 comments by @axinging, @ben-clayton, @kainino0x, @kvark, @munrocket)
-
Add method to disable fast-math on a per-shader basis. gpuweb/gpuweb#2076 (10 comments by @dneto0, @kainino0x, @kvark, @mrshannon, @munrocket, @pow2clk)
-
Change ArrayBuffer arguments to ArrayBuffer + srcOffset gpuweb/gpuweb#244 (9 comments by @domenic, @juj, @kainino0x, @Kangz, @litherum)
-
Develop a visual language for the syntactic rules in WGSL specification gpuweb/gpuweb#2114 (8 comments by @dneto0, @mehmetoguzderin)
-
Require a render pipeline flag to enable sample-rate shading gpuweb/gpuweb#2133 (6 comments by @kainino0x, @Kangz, @kvark, @magcius)
-
Support "uniform texel buffer" and "storage texel buffer" ? gpuweb/gpuweb#162 (6 comments by @Kangz, @kvark, @magcius)
-
Consider improving variable array indexing for 'let' arrays gpuweb/gpuweb#2086 (6 comments by @ben-clayton, @jrprice, @Kangz, @kvark)
-
Request for compute: anyInvocation() and allInvocation() gpuweb/gpuweb#2137 (5 comments by @dneto0, @kdashg, @krogovin, @litherum)
-
Fragment shader framebuffer fetch gpuweb/gpuweb#396 (5 comments by @devshgraphicsprogramming, @Kangz, @krogovin, @tklajnscek)
-
Inquiry about force running on NIVIDA GPU not on built-in one of motherboard gpuweb/gpuweb#2107 (5 comments by @Jiawei-Shao, @Kangz, @kvark, @supcn)
-
Refactor API for load/store OP and read-only attachments gpuweb/gpuweb#2131 (5 comments by @kainino0x, @Kangz, @kvark)
-
Fragment shader [[builtin(position)]] definition is misleading gpuweb/gpuweb#2106 (5 comments by @kainino0x, @kvark)
-
Support arrays of top-level structures in storage buffers gpuweb/gpuweb#2102 (5 comments by @dneto0, @kainino0x, @Kangz, @kvark)
-
copyToTexture for color-managed image sources gpuweb/gpuweb#1483 (5 comments by @kainino0x, @Kangz, @shaoboyan)
-
Support arrayLength for fixed-size arrays gpuweb/gpuweb#2074 (5 comments by @ben-clayton, @jrprice, @munrocket)
-
RFC: Make sure that a minimal-garbage hot path is supported gpuweb/gpuweb#214 (4 comments by @Kangz, @kdashg, @magcius)
-
wgsl: finding most-significant bit and least-significant bit gpuweb/gpuweb#2130 (3 comments by @dneto0, @kdashg)
-
Should depth-stencil render attachments require views to have aspect = "all" for combined depth-stencil formats? gpuweb/gpuweb#2062 (3 comments by @austinEng, @kainino0x, @kvark)
-
Out-of-bounds clamping brings big perf regression on TFJS ML demos. gpuweb/gpuweb#1202 (3 comments by @ben-clayton, @kvark, @qjia7)
-
wgsl: Sketch design for arrays of buffer-bindings gpuweb/gpuweb#2105 (3 comments by @kainino0x, @kvark)
-
Dictionary default values referencing type members gpuweb/gpuweb#278 (3 comments by @kainino0x, @Kangz, @kvark)
-
WGSL: Drop requirement to ignore() atomic operations? gpuweb/gpuweb#2090 (3 comments by @ben-clayton, @kainino0x, @m-schuetz)
-
Should all the constant-only interfaces be namespaces instead? gpuweb/gpuweb#304 (3 comments by @ExE-Boss, @Kangz)
-
Remove requirement that function / builtin return values are always consumed gpuweb/gpuweb#2138 (2 comments by @dneto0, @kdashg)
-
Fragment shader interlock/ordering gpuweb/gpuweb#395 (2 comments by @devshgraphicsprogramming, @krogovin)
-
Allow using depth-stencil GPUTextureView with aspects as attachment if pass doesn't use other aspect gpuweb/gpuweb#2132 (2 comments by @kainino0x, @kvark)
-
Which corner is the origin of a copyExternalImageToTexture? gpuweb/gpuweb#2110 (2 comments by @kainino0x, @kdashg)
-
Consider adding a "gpuadapterschanged" event to the Web platform gpuweb/gpuweb#221 (2 comments by @kainino0x, @Kangz)
-
Contradictions in the validation rules of copyTextureToTexture() with depth stencil formats gpuweb/gpuweb#1464 (2 comments by @kainino0x)
-
further restrict where a control-flow interruption can be placed gpuweb/gpuweb#2093 (2 comments by @dneto0)
-
Investigation: Validation about 2d-array texture view upon 3D texture and vice versa, and more gpuweb/gpuweb#1372 (2 comments by @Kangz, @kvark)
-
shorthands for var and const declarations, and infer types gpuweb/gpuweb#1289 (2 comments by @Kangz, @kdashg)
-
Make error scope objects have same shape as DOMExceptions? gpuweb/gpuweb#1884 (2 comments by @juj, @Kangz)
-
Have a plan for planar format like YUV 420 and 422. gpuweb/gpuweb#625 (2 comments by @Kangz, @shaoboyan)
-
Possibility to optimize global app-specific and frame-specific state binding? gpuweb/gpuweb#1414 (2 comments by @juj, @Kangz)
-
Ability to issue command and create resources from workers gpuweb/gpuweb#9 (1 comment by @kainino0x)
-
wgsl: Restrict `modf` and `frexp`pointer parameter storage classes gpuweb/gpuweb#1846 (1 comment by @kdashg)
-
Consider adding a 32- by 32-bit multiplication function resulting in two u32s, high and low gpuweb/gpuweb#2023 (1 comment by @kdashg)
-
context-aware tokenization: Sometimes >> is better parsed as two copies of > (similar for ]] ) gpuweb/gpuweb#2092 (1 comment by @kdashg)
-
wgsl: support extended integer arithmetic for add, subtract, multiply gpuweb/gpuweb#1565 (1 comment by @Kangz)
-
Writing to different components of a vector from different invocations gpuweb/gpuweb#2112 (1 comment by @alan-baker)
-
copyExternalImageToTexture/importExternalTexture with destination in linear color space gpuweb/gpuweb#1715 (1 comment by @kainino0x)
-
Consider copyToTexture interactions with high-bit-depth source images gpuweb/gpuweb#1482 (1 comment by @kainino0x)
-
wgsl: can hex floats explicitly represent infinity and NaN as float literals? gpuweb/gpuweb#1769 (1 comment by @dneto0)
-
WebGL - WebGPU interop: WebGPU computation API with existing WebGL gpuweb/gpuweb#17 (1 comment by @Kangz)
-
Can begin/endPipelineStatisticsQuery and writeTimestamp be moved to GPUProgrammablePassEncoder? gpuweb/gpuweb#1430 (1 comment by @Kangz)
-
Feedback on GPUCommandEncoder missing a transfer pass encoder gpuweb/gpuweb#1969 (1 comment by @Kangz)
-
Do pipeline blend states need to be validated against the pipline render targets? gpuweb/gpuweb#2025 (1 comment by @Kangz)
-
Texture dimensions should be required to be >=1 (forbid zero-sized textures) gpuweb/gpuweb#1200 (1 comment by @Kangz)
-
CopyImageBitmapToTexture: Texture usage of the dest texture may prevent blit using internal shaders. gpuweb/gpuweb#1150 (1 comment by @Kangz)
-
In GPULimits there is no parameter maxVaryingVectors/maxVertexOutputComponents gpuweb/gpuweb#721 (1 comment by @Kangz)
-
Should various unused resource usages be counted toward usage scopes? gpuweb/gpuweb#657 (1 comment by @Kangz)
-
Sampler compareEnable needs to be specified at sampler creation, and validated against the shader gpuweb/gpuweb#552 (1 comment by @Kangz)
-
GPUVertexBufferDescriptor 's stride size needs to be a multiple of 4 bytes gpuweb/gpuweb#404 (1 comment by @Kangz)
-
Out-of-bounds clamping may require additional draw-call validation gpuweb/gpuweb#311 (1 comment by @Kangz)
-
shader programming model: use little-endian layouts for scalar data gpuweb/gpuweb#234 (1 comment by @Kangz)
-
shaders: equivalent of gl_PointCoord and gl_PointSize don't exist in HLSL gpuweb/gpuweb#332 (1 comment by @Kangz)
-
Position: Data races and Lacking Synchronisation for GPU Memory Objects are not a robustness concern* gpuweb/gpuweb#119 (1 comment by @Kangz)
-
Proposal: BufferSubRanges to solve buffer-tracking, streaming and descriptor set explosion. gpuweb/gpuweb#156 (1 comment by @Kangz)
-
Characterize the performance benefit of GPUBufferUsage and GPUTextureUsage gpuweb/gpuweb#150 (1 comment by @Kangz)
-
Proposals for buffer operations (immediate uploads, buffer mapping) gpuweb/gpuweb#138 (1 comment by @Kangz)
-
Question: shall we set default values for parameters of descriptors gpuweb/gpuweb#136 (1 comment by @Kangz)
-
Proposal: Run all index buffers through a compute shader validator gpuweb/gpuweb#117 (1 comment by @Kangz)
-
Investigation: Multisampled Render Targets and Resolve Operations gpuweb/gpuweb#108 (1 comment by @Kangz)
-
Should objects expose the descriptor they were created with? gpuweb/gpuweb#1498 (1 comment by @Kangz)
-
-
36 pull requests opened.
-
Fix some misplaced characters in the `select` description. gpuweb/gpuweb#2154, by @jimblandy.
-
Fix inequality symbol in compute shader grid bound gpuweb/gpuweb#2153, by @jrprice.
-
Clarify memory locations accessed when writing a vector component gpuweb/gpuweb#2152, by @alan-baker.
-
wgsl: Update README.md gpuweb/gpuweb#2150, by @ben-clayton.
-
wgsl: Function call statement may call non-void functions gpuweb/gpuweb#2148, by @dneto0.
-
Clarify copies into -srgb formats gpuweb/gpuweb#2146, by @kainino0x.
-
Allow syntactic navigation and styling gpuweb/gpuweb#2143, by @mehmetoguzderin.
-
Extend color spaces; more formats in copyExternalImageToTexture gpuweb/gpuweb#2142, by @kainino0x.
-
Remove no-op minBindingSize text gpuweb/gpuweb#2141, by @kainino0x.
-
Сlarify rounding gpuweb/gpuweb#2140, by @munrocket.
-
Replace depth-clamp feature with depth-clip feature gpuweb/gpuweb#2139, by @kainino0x.
-
Enable Assume Explicit For gpuweb/gpuweb#2128, by @mehmetoguzderin.
-
wgsl: remove ignore, add phony-assignment gpuweb/gpuweb#2127, by @dneto0.
-
Update WGSL README: instructions about tree-sitter gpuweb/gpuweb#2126, by @dneto0.
-
Clarify line-ending comment gpuweb/gpuweb#2125, by @dneto0.
-
Fix a few typos and styling issues in RenderPassDescriptor validation gpuweb/gpuweb#2121, by @Richard-Yunchao.
-
update WGSL style guide: algorithms, code samples gpuweb/gpuweb#2120, by @dneto0.
-
Make CI sad with a bad syntax gpuweb/gpuweb#2117, by @mehmetoguzderin.
-
Make CI sad with a bad example gpuweb/gpuweb#2116, by @mehmetoguzderin.
-
Make atomicCompareExchangeWeak return a structure gpuweb/gpuweb#2113, by @kvark.
-
Fix typos in normalize(), mix(), cross() gpuweb/gpuweb#2109, by @munrocket.
-
Allow non-overriable constants in texel offsets gpuweb/gpuweb#2108, by @kvark.
-
Remove block attribute gpuweb/gpuweb#2104, by @dneto0.
-
wgsl: replace stride attribute with array type parameter gpuweb/gpuweb#2103, by @dneto0.
-
Syntactic CI for Grammar gpuweb/gpuweb#2101, by @mehmetoguzderin.
-
wgsl: describe sequenced execution gpuweb/gpuweb#2097, by @dneto0.
-
wgsl: more restrictions on control flow interruption gpuweb/gpuweb#2096, by @dneto0.
-
wgsl: editorial fixes: control flow statements gpuweb/gpuweb#2095, by @dneto0.
-
Fix error in optional chaning description gpuweb/gpuweb#2089, by @samsamson33.
-
Fill out "Textual structure", eliminating several TODOs gpuweb/gpuweb#2087, by @dneto0.
-
Rewrite select() for scalar condition and add missing algorithm tags gpuweb/gpuweb#2084, by @sarahM0.
-
wgsl: support arrayLength on pointer to sized array gpuweb/gpuweb#2081, by @dneto0.
-
precise_math attribute on functions gpuweb/gpuweb#2080, by @kvark.
-
Require GPURenderPassDepthStencilAttachments to reference all aspects gpuweb/gpuweb#2079, by @austinEng.
-
Add defaults for GPUTextureSampleType in the default pipeline layout gpuweb/gpuweb#2078, by @austinEng.
-
Expose fast-math gpuweb/gpuweb#2077, by @kvark.
-
-
21 pull requests merged.
-
wgsl: Function call statement may call non-void functions gpuweb/gpuweb#2148.
-
wgsl: Update README.md gpuweb/gpuweb#2150.
-
Remove no-op minBindingSize text gpuweb/gpuweb#2141.
-
Clarify copies into -srgb formats gpuweb/gpuweb#2146.
-
Enable Assume Explicit For gpuweb/gpuweb#2128.
-
Make atomicCompareExchangeWeak return a structure gpuweb/gpuweb#2113.
-
Fix a few typos and styling issues in RenderPassDescriptor validation gpuweb/gpuweb#2121.
-
Add defaults for GPUTextureSampleType in the default pipeline layout gpuweb/gpuweb#2078.
-
Require GPURenderPassDepthStencilAttachments to reference all aspects gpuweb/gpuweb#2079.
-
update WGSL style guide: algorithms, code samples gpuweb/gpuweb#2120.
-
Clarify line-ending comment gpuweb/gpuweb#2125.
-
Rewrite select() for scalar condition and add missing algorithm tags gpuweb/gpuweb#2084.
-
Allow non-overriable constants in texel offsets gpuweb/gpuweb#2108.
-
Fix typos in normalize(), mix(), cross() gpuweb/gpuweb#2109.
-
Syntactic CI for Grammar gpuweb/gpuweb#2101.
-
wgsl: editorial fixes: control flow statements gpuweb/gpuweb#2095.
-
Use namespaces for flags instead of interfaces gpuweb/gpuweb#1709.
-
Fix error in optional chaning description gpuweb/gpuweb#2089.
-
Allow indirection gpuweb/gpuweb#2052.
-
Match let declarations for inferred types gpuweb/gpuweb#2053.
-
Fill out "Textual structure", eliminating several TODOs gpuweb/gpuweb#2087.
-
-
4 pull requests closed.
-
Сlarify rounding gpuweb/gpuweb#2140.
-
Make CI sad with a bad syntax gpuweb/gpuweb#2117.
-
Make CI sad with a bad example gpuweb/gpuweb#2116.
-
Fix module-scope constant declarations wrt type inference gpuweb/gpuweb#1944.
-
-
41 pull requests received 255 new comments.
-
precise_math attribute on functions gpuweb/gpuweb#2080 (50 comments by @dneto0, @kainino0x, @kdashg, @kvark, @litherum, @mrshannon, @munrocket)
-
wgsl: Add limits section, and "spurious" failure gpuweb/gpuweb#1997 (14 comments by @alan-baker, @dneto0, @kdashg, @kvark, @litherum, @pow2clk)
-
Syntactic CI for Grammar gpuweb/gpuweb#2101 (14 comments by @dneto0, @kdashg, @kvark, @mehmetoguzderin)
-
Fix a few typos and styling issues in RenderPassDescriptor validation gpuweb/gpuweb#2121 (14 comments by @kainino0x, @kvark, @Richard-Yunchao)
-
Allow syntactic navigation and styling gpuweb/gpuweb#2143 (10 comments by @dneto0, @mehmetoguzderin)
-
Rewrite select() for scalar condition and add missing algorithm tags gpuweb/gpuweb#2084 (10 comments by @ben-clayton, @dneto0, @jrprice, @kdashg, @sarahM0)
-
wgsl: more restrictions on control flow interruption gpuweb/gpuweb#2096 (10 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
Fix typos in normalize(), mix(), cross() gpuweb/gpuweb#2109 (9 comments by @dneto0, @munrocket, @w3cbot)
-
Replace depth-clamp feature with depth-clip feature gpuweb/gpuweb#2139 (8 comments by @kainino0x, @kvark)
-
wgsl: replace stride attribute with array type parameter gpuweb/gpuweb#2103 (7 comments by @alan-baker, @dneto0, @kvark)
-
Add defaults for GPUTextureSampleType in the default pipeline layout gpuweb/gpuweb#2078 (7 comments by @austinEng, @kvark)
-
Extend color spaces; more formats in copyExternalImageToTexture gpuweb/gpuweb#2142 (6 comments by @kainino0x, @kdashg)
-
wgsl: support arrayLength on pointer to sized array gpuweb/gpuweb#2081 (6 comments by @alan-baker, @ben-clayton, @dneto0, @kdashg, @kvark)
-
wgsl: remove ignore, add phony-assignment gpuweb/gpuweb#2127 (5 comments by @ben-clayton, @dneto0, @kdashg)
-
Add a memory model mapping to Vulkan gpuweb/gpuweb#2055 (4 comments by @alan-baker, @kdashg, @litherum, @pow2clk)
-
Allow non-overriable constants in texel offsets gpuweb/gpuweb#2108 (4 comments by @ben-clayton, @kdashg, @kvark)
-
Make atomicCompareExchangeWeak return a structure gpuweb/gpuweb#2113 (4 comments by @ben-clayton, @kdashg, @kvark)
-
Match let declarations for inferred types gpuweb/gpuweb#2053 (4 comments by @dneto0, @kdashg, @kvark)
-
Fill out "Textual structure", eliminating several TODOs gpuweb/gpuweb#2087 (4 comments by @dneto0, @kdashg, @mehmetoguzderin)
-
wgsl: editorial fixes: control flow statements gpuweb/gpuweb#2095 (3 comments by @alan-baker, @dneto0)
-
Require [SecureContext] for using WebGPU. gpuweb/gpuweb#1363 (3 comments by @AshleyScirra, @kainino0x, @munrocket)
-
Require GPURenderPassDepthStencilAttachments to reference all aspects gpuweb/gpuweb#2079 (2 comments by @austinEng, @kainino0x)
-
Clarify memory locations accessed when writing a vector component gpuweb/gpuweb#2152 (1 comment by @alan-baker)
-
Update WGSL README: instructions about tree-sitter gpuweb/gpuweb#2126 (1 comment by @mehmetoguzderin)
-
gpuweb/cts
-
2 issues opened.
-
Enhance the wgsl type generator gpuweb/cts#763, by @sarahM0.
-
wgsl: validate identifier token pattern gpuweb/cts#748, by @dneto0.
-
-
1 issue closed.
-
Adding wgsl validation folder in CTS automatically gpuweb/cts#674.
-
-
1 issue received 1 new comment.
-
35 pull requests opened.
-
webgpu/api/validation: Fix vertex_shader_input_location_limit test gpuweb/cts#767, by @ben-clayton.
-
Add test for compute shader input builtin values gpuweb/cts#766, by @jrprice.
-
webgpu/idl: Initialize the GPU gpuweb/cts#765, by @ben-clayton.
-
Replace grunt-http-server with http-server gpuweb/cts#764, by @kainino0x.
-
webgpu/shader/execution: Fix bad WGSL generation gpuweb/cts#762, by @ben-clayton.
-
Add wgsl how to contribute doc and readme gpuweb/cts#761, by @sarahM0.
-
cmdline: Add new flag to allow overriding of the GPU provider gpuweb/cts#760, by @ben-clayton.
-
Adding test for vertex-only render pipeline: Part 1 gpuweb/cts#759, by @jiangzhaoming.
-
Do not require grunt to be installed globally. gpuweb/cts#758, by @greggman.
-
Add --list flag to the command line runner gpuweb/cts#757, by @ben-clayton.
-
Add operation tests on the T2T copies with multisampled depth textures gpuweb/cts#756, by @Jiawei-Shao.
-
draft for wgsl validation gpuweb/cts#755, by @munrocket.
-
Use vertex-only pipeline when initializing depth aspect gpuweb/cts#754, by @Jiawei-Shao.
-
Depth clip/clamp test gpuweb/cts#753, by @kainino0x.
-
require indirect-first-instance for indirect draws with firstInstance > 0 gpuweb/cts#752, by @krockot.
-
ShaderValidationTest: remove string overload gpuweb/cts#751, by @dneto0.
-
ShaderValidationTest: remove substring overload gpuweb/cts#750, by @dneto0.
-
Fix a bug about max color attachment gpuweb/cts#749, by @Richard-Yunchao.
-
sketch a NumberRepr class that represents f/i/u 64/32/16/8 gpuweb/cts#747, by @kainino0x.
-
wgsl: Add test plan Value-testing, float, and integer built-in functions gpuweb/cts#746, by @sarahM0.
-
Update default limits to match the spec gpuweb/cts#745, by @austinEng.
-
wgsl: add 32-bit limits and other constants gpuweb/cts#744, by @sarahM0.
-
Test compute shader variables are zero initialized gpuweb/cts#743, by @austinEng.
-
Update destinationOffset alignment to 256 in resolveQuerySet tests gpuweb/cts#742, by @haoxli.
-
Implement api,validation,state,device_mismatched: Part Ⅰ gpuweb/cts#741, by @haoxli.
-
Update all uses of 'gpupresent' to 'webgpu' gpuweb/cts#740, by @toji.
-
Add operation tests on T2T copy with multisampled color textures gpuweb/cts#739, by @Jiawei-Shao.
-
Remove CopyImageBitmapToTexture gpuweb/cts#738, by @shaoboyan.
-
Add always fail test for dry run test gpuweb/cts#737, by @jiangzhaoming.
-
Dry run test gpuweb/cts#736, by @jiangzhaoming.
-
Include tools in the deploy gpuweb/cts#735, by @kvark.
-
Fix tools/deno gpuweb/cts#734, by @lucacasonato.
-
Add an option to automatically batch large lists of subcases gpuweb/cts#733, by @austinEng.
-
Move shader-related type enums and helpers to webgpu/shader/types.ts gpuweb/cts#732, by @austinEng.
-
wgsl type parameter generator gpuweb/cts#731, by @sarahM0.
-
-
31 pull requests merged.
-
Add test for compute shader input builtin values gpuweb/cts#766.
-
webgpu/idl: Initialize the GPU gpuweb/cts#765.
-
Adding test for vertex-only render pipeline: Part 1 gpuweb/cts#759.
-
webgpu/shader/execution: Fix bad WGSL generation gpuweb/cts#762.
-
Replace grunt-http-server with http-server gpuweb/cts#764.
-
cmdline: Add new flag to allow overriding of the GPU provider gpuweb/cts#760.
-
wgsl: add 32-bit limits and other constants gpuweb/cts#744.
-
Depth clip/clamp test gpuweb/cts#753.
-
Do not require grunt to be installed globally. gpuweb/cts#758.
-
Add --list flag to the command line runner gpuweb/cts#757.
-
Implement api,validation,state,device_mismatched: Part Ⅰ gpuweb/cts#741.
-
Use vertex-only pipeline when initializing depth aspect gpuweb/cts#754.
-
sketch a NumberRepr class that represents f/i/u 64/32/16/8 gpuweb/cts#747.
-
Fix a bug about max color attachment gpuweb/cts#749.
-
Test buffer related validations for texture copies gpuweb/cts#690.
-
Add operation tests on T2T copy with multisampled color textures gpuweb/cts#739.
-
Test compute shader variables are zero initialized gpuweb/cts#743.
-
Update default limits to match the spec gpuweb/cts#745.
-
Add an option to automatically batch large lists of subcases gpuweb/cts#733.
-
wgsl abs() built-in function execution test gpuweb/cts#730.
-
Remove CopyImageBitmapToTexture gpuweb/cts#738.
-
wgsl type parameter generator gpuweb/cts#731.
-
Fix tools/deno gpuweb/cts#734.
-
Include tools in the deploy gpuweb/cts#735.
-
Update all uses of 'gpupresent' to 'webgpu' gpuweb/cts#740.
-
Add always fail test for dry run test gpuweb/cts#737.
-
Dry run test gpuweb/cts#736.
-
Add uniqueId() to class TestBuilder gpuweb/cts#729.
-
-
2 pull requests closed.
-
ShaderValidationTest: remove substring overload gpuweb/cts#750.
-
-
24 pull requests received 136 new comments.
-
wgsl: add 32-bit limits and other constants gpuweb/cts#744 (18 comments by @ben-clayton, @kainino0x, @sarahM0)
-
Add operation tests on T2T copy with multisampled color textures gpuweb/cts#739 (15 comments by @Jiawei-Shao, @kainino0x, @Kangz)
-
Test buffer related validations for texture copies gpuweb/cts#690 (13 comments by @kainino0x, @Richard-Yunchao)
-
Add test for compute shader input builtin values gpuweb/cts#766 (9 comments by @jrprice, @kainino0x)
-
cmdline: Add new flag to allow overriding of the GPU provider gpuweb/cts#760 (8 comments by @austinEng, @ben-clayton, @kainino0x)
-
Implement api,validation,state,device_mismatched: Part Ⅰ gpuweb/cts#741 (5 comments by @haoxli, @kainino0x)
-
wgsl: Add test plan Value-testing, float, and integer built-in functions gpuweb/cts#746 (5 comments by @ben-clayton, @kainino0x, @sarahM0)
-
Add an option to automatically batch large lists of subcases gpuweb/cts#733 (5 comments by @austinEng, @kainino0x)
-
Test compute shader variables are zero initialized gpuweb/cts#743 (4 comments by @austinEng, @kainino0x)
-
webgpu/api/validation: Fix vertex_shader_input_location_limit test gpuweb/cts#767 (3 comments by @ben-clayton, @kainino0x)
-
require indirect-first-instance for indirect draws with firstInstance > 0 gpuweb/cts#752 (3 comments by @kainino0x, @krockot)
-
Update destinationOffset alignment to 256 in resolveQuerySet tests gpuweb/cts#742 (3 comments by @haoxli, @kainino0x)
-
Move shader-related type enums and helpers to webgpu/shader/types.ts gpuweb/cts#732 (3 comments by @austinEng, @kainino0x)
-
webgpu/shader/execution: Fix bad WGSL generation gpuweb/cts#762 (2 comments by @ben-clayton, @sarahM0)
-
sketch a NumberRepr class that represents f/i/u 64/32/16/8 gpuweb/cts#747 (1 comment by @ben-clayton)
-
TR publications
-
26 documents published.
-
WebGPU published 13 times as a Working Draft. History of WebGPU
-
WebGPU Shading Language published 13 times as a Working Draft. History of WebGPU Shading Language
-
Participants
-
1 organization joined.
-
China Mobile Communications Corporation joined.
-
GitHub
gpuweb/cts
-
43 pull requests opened.
-
wgsl abs() built-in function execution test gpuweb/cts#730, by @sarahM0.
-
Add uniqueId() to class TestBuilder gpuweb/cts#729, by @sarahM0.
-
Test postMessage on ArrayBuffer from getMappedRange gpuweb/cts#728, by @kainino0x.
-
wgsl: sections 16 and 3 proposed test plan gpuweb/cts#727, by @sarahM0.
-
fix async race condition in expectCompileResult gpuweb/cts#726, by @kainino0x.
-
Remove stenci8 and depth16unorm references gpuweb/cts#725, by @zoddicus.
-
wgsl: test builtin operand types gpuweb/cts#724, by @sarahM0.
-
Introduce expectGPUBufferRepeatsSingleValue gpuweb/cts#723, by @krockot.
-
some stress tests for shaders and render passes gpuweb/cts#722, by @krockot.
-
wgsl: check if abs(e: T)->T works gpuweb/cts#721, by @sarahM0.
-
wgsl cts for builtin fucntions gpuweb/cts#720, by @sarahM0.
-
Revise validation tests for depth/stencil formats for copies gpuweb/cts#719, by @Richard-Yunchao.
-
Constrain type to guard against future bugs like #716 gpuweb/cts#718, by @kainino0x.
-
Implement api,validation,state,device_mismatched: helper functions gpuweb/cts#717, by @haoxli.
-
Fix a bug in api/validation/image_copy/image_copy.ts gpuweb/cts#716, by @Richard-Yunchao.
-
Test with PR gpuweb/cts#715, by @mehmetoguzderin.
-
Fix project name gpuweb/cts#714, by @mehmetoguzderin.
-
Test a PR gpuweb/cts#713, by @mehmetoguzderin.
-
Use split account for previews gpuweb/cts#712, by @mehmetoguzderin.
-
Change depth/stencil sample type to 'depth' gpuweb/cts#711, by @toji.
-
Updating deprecated TextureUsage flags gpuweb/cts#710, by @toji.
-
Implement some initial stress tests gpuweb/cts#709, by @krockot.
-
Remove some alpha blending factor validation gpuweb/cts#708, by @shrekshao.
-
Remove log message deduplication, hide more log details gpuweb/cts#707, by @kainino0x.
-
Add operation tests on copyTextureToTexture with depth/stencil - Part II gpuweb/cts#706, by @Jiawei-Shao.
-
api,validation,state,device_mismatched plan gpuweb/cts#705, by @haoxli.
-
Make reftest slightly fuzzy to account for apparent differences gpuweb/cts#704, by @kainino0x.
-
Catch promise rejection from video.play() gpuweb/cts#703, by @kainino0x.
-
Fix api,operation,rendering,draw:vertex_attributes,basic:* gpuweb/cts#702, by @Jiawei-Shao.
-
Pipeline_layout_compat: Add buffer and sampler binding equivalent checks gpuweb/cts#701, by @shaoboyan.
-
Roll @webgpu/types to 0.1.6 gpuweb/cts#699, by @toji.
-
Implement veretx_access validation test: Part 1 gpuweb/cts#698, by @jiangzhaoming.
-
api,validation,encoding,programmable,pipeline_bind_group_compat plan gpuweb/cts#697, by @haoxli.
-
Add tests for requiredLimits in requestDevice gpuweb/cts#695, by @toji.
-
Add operation tests on the T2B copies with depth aspects gpuweb/cts#694, by @Jiawei-Shao.
-
Fix bugs introduced with memcpy gpuweb/cts#693, by @kainino0x.
-
Implement External Texture Follow Up Test Changes gpuweb/cts#692, by @bjjones.
-
Validate texture format for copies gpuweb/cts#691, by @Richard-Yunchao.
-
Test buffer related validations for texture copies gpuweb/cts#690, by @Richard-Yunchao.
-
Add refttests for different size of canvas and swapchain gpuweb/cts#689, by @shaoboyan.
-
Add operation tests on T2B copies with stencil aspects gpuweb/cts#688, by @Jiawei-Shao.
-
-
40 pull requests merged.
-
Test postMessage on ArrayBuffer from getMappedRange gpuweb/cts#728.
-
Introduce expectGPUBufferRepeatsSingleValue gpuweb/cts#723.
-
some stress tests for shaders and render passes gpuweb/cts#722.
-
fix async race condition in expectCompileResult gpuweb/cts#726.
-
Implement veretx_access validation test: Part 1 gpuweb/cts#698.
-
Revise validation tests for depth/stencil formats for copies gpuweb/cts#719.
-
Validate texture format for copies gpuweb/cts#691.
-
Constrain type to guard against future bugs like #716 gpuweb/cts#718.
-
Fix a bug in api/validation/image_copy/image_copy.ts gpuweb/cts#716.
-
Change depth/stencil sample type to 'depth' gpuweb/cts#711.
-
Fix project name gpuweb/cts#714.
-
Use split account for previews gpuweb/cts#712.
-
Implement some initial stress tests gpuweb/cts#709.
-
Updating deprecated TextureUsage flags gpuweb/cts#710.
-
Remove some alpha blending factor validation gpuweb/cts#708.
-
Remove log message deduplication, hide more log details gpuweb/cts#707.
-
api,validation,state,device_mismatched plan gpuweb/cts#705.
-
Add type validation tests for fragment outputs and attachments gpuweb/cts#676.
-
Fix api,operation,rendering,draw:vertex_attributes,basic:* gpuweb/cts#702.
-
Make reftest slightly fuzzy to account for apparent differences gpuweb/cts#704.
-
Catch promise rejection from video.play() gpuweb/cts#703.
-
Add tests for requiredLimits in requestDevice gpuweb/cts#695.
-
Roll @webgpu/types to 0.1.6 gpuweb/cts#699.
-
Add operation tests on the T2B copies with depth aspects gpuweb/cts#694.
-
Add refttests for different size of canvas and swapchain gpuweb/cts#689.
-
Implement External Texture Follow Up Test Changes gpuweb/cts#692.
-
Fix bugs introduced with memcpy gpuweb/cts#693.
-
More detailed plans for stress tests gpuweb/cts#671.
-
Track resources for cleanup at the end of the test gpuweb/cts#683.
-
Add operation tests on T2B copies with stencil aspects gpuweb/cts#688.
-
Add ImportExternalTexture Video Test gpuweb/cts#675.
-
Add memcpy helper gpuweb/cts#616.
-
Add tests for 3D texture copies gpuweb/cts#679.
-
6 pull requests closed.
-
Remove stenci8 and depth16unorm references gpuweb/cts#725.
-
wgsl: check if abs(e: T)->T works gpuweb/cts#721.
-
wgsl cts for builtin fucntions gpuweb/cts#720.
-
Test with PR gpuweb/cts#715.
-
Test a PR gpuweb/cts#713.
-
-
37 pull requests received 399 new comments.
-
Implement veretx_access validation test: Part 1 gpuweb/cts#698 (89 comments by @jiangzhaoming, @kainino0x)
-
Pipeline_layout_compat: Add buffer and sampler binding equivalent checks gpuweb/cts#701 (35 comments by @kainino0x, @shaoboyan)
-
Add refttests for different size of canvas and swapchain gpuweb/cts#689 (23 comments by @kainino0x, @shaoboyan)
-
Reftests expand with formats, write canvas methods, and compositingAlphaMode gpuweb/cts#696 (22 comments by @kainino0x, @shrekshao)
-
Test buffer related validations for texture copies gpuweb/cts#690 (19 comments by @kainino0x, @Richard-Yunchao)
-
Add type validation tests for fragment outputs and attachments gpuweb/cts#676 (18 comments by @kainino0x, @shrekshao)
-
api,validation,encoding,programmable,pipeline_bind_group_compat plan gpuweb/cts#697 (17 comments by @haoxli, @kainino0x, @shaoboyan)
-
Implement api,operation,command_buffer,programmable,state_tracking:* gpuweb/cts#614 (16 comments by @kainino0x, @toji)
-
Add ImportExternalTexture Video Test gpuweb/cts#675 (13 comments by @bjjones, @kainino0x, @shaoboyan)
-
Implement External Texture Follow Up Test Changes gpuweb/cts#692 (12 comments by @bjjones, @kainino0x)
-
Introduce expectGPUBufferRepeatsSingleValue gpuweb/cts#723 (11 comments by @austinEng, @kainino0x, @krockot)
-
Implement api,validation,state,device_mismatched: helper functions gpuweb/cts#717 (10 comments by @haoxli, @kainino0x)
-
Revise validation tests for depth/stencil formats for copies gpuweb/cts#719 (8 comments by @Jiawei-Shao, @kainino0x, @Richard-Yunchao)
-
Add operation tests on the T2B copies with depth aspects gpuweb/cts#694 (8 comments by @Jiawei-Shao, @kainino0x)
-
Add operation tests on copyTextureToTexture with depth/stencil - Part II gpuweb/cts#706 (4 comments by @Jiawei-Shao, @kainino0x)
-
Make reftest slightly fuzzy to account for apparent differences gpuweb/cts#704 (4 comments by @kainino0x, @shaoboyan)
-
Fix a bug in api/validation/image_copy/image_copy.ts gpuweb/cts#716 (2 comments by @Richard-Yunchao)
-
Track resources for cleanup at the end of the test gpuweb/cts#683 (2 comments by @kainino0x, @shaoboyan)
-
Fix api,operation,rendering,draw:vertex_attributes,basic:* gpuweb/cts#702 (1 comment by @Jiawei-Shao)
-
Add operation tests about copyTextureToTexture with depth/stencil - Part I gpuweb/cts#700 (1 comment by @Jiawei-Shao)
-
gpuweb/gpuweb
-
34 issues opened.
-
Add method to disable fast-math on a per-shader basis. gpuweb/gpuweb#2076, by @mrshannon.
-
wgsl: add a note showing that leading zeroes are invalid for number literals, eg. 012345 gpuweb/gpuweb#2075, by @dneto0.
-
Support arrayLength for fixed-size arrays gpuweb/gpuweb#2074, by @ben-clayton.
-
Triage milestones for all issues gpuweb/gpuweb#2073, by @kainino0x.
-
Fake transfers for getMappedRange-created ArrayBuffers? gpuweb/gpuweb#2072, by @kainino0x.
-
wgsl: -0 is not a valid token gpuweb/gpuweb#2071, by @dneto0.
-
[meta] Wiki Matrix link is outdated? gpuweb/gpuweb#2070, by @ShadowJonathan.
-
Implicit GPUTextureSampleType for texture_2d<f32> gpuweb/gpuweb#2064, by @austinEng.
-
Implicit flat interpolation on integers is confusing gpuweb/gpuweb#2061, by @kvark.
-
Feature request: ignore shader writes to color attachments gpuweb/gpuweb#2060, by @MatusT.
-
Brain dump on all things wrong with WGSL gpuweb/gpuweb#2048, by @morganbarrett.
-
timestamp-query is unimplementable on TBDR architectures gpuweb/gpuweb#2046, by @litherum.
-
Creating arrays/structs in expressions gpuweb/gpuweb#2045, by @litherum.
-
Binding a zero-sized range of a vertex buffer gpuweb/gpuweb#2044, by @austinEng.
-
Dealing with holes in the pipeline layout. gpuweb/gpuweb#2043, by @Kangz.
-
Shader stage input-output subsetting rules gpuweb/gpuweb#2038, by @litherum.
-
Should hex exponents be optional? gpuweb/gpuweb#2036, by @mehmetoguzderin.
-
Typo "fingeprinting" gpuweb/gpuweb#2033, by @esoterra.
-
Rename forceSoftware and isSoftware gpuweb/gpuweb#2030, by @litherum.
-
Do pipeline blend states need to be validated against the pipline render targets? gpuweb/gpuweb#2025, by @toji.
-
Dynamically sized threadgroup memory gpuweb/gpuweb#2024, by @litherum.
-
Consider adding a 32- by 32-bit multiplication function resulting in two u32s, high and low gpuweb/gpuweb#2023, by @hcsch.
-
atomicCompareExchangeWeak is not idiomatic gpuweb/gpuweb#2021, by @kvark.
-
Missing atomicSub function gpuweb/gpuweb#2019, by @kvark.
-
Handling limits which are not unsigned longs. gpuweb/gpuweb#2018, by @toji.
-
Interpolate attribute clarification gpuweb/gpuweb#2017, by @sarahM0.
-
Fragment output <-> attachment interface matching rules def gpuweb/gpuweb#2013, by @shrekshao.
-
WGSL spec is not clear on whether array size can be an unsigned literal gpuweb/gpuweb#2012, by @afd.
-
Pad implicit mapping sizes up to multiples of 4 bytes? gpuweb/gpuweb#2010, by @kainino0x.
-
Restrictions on range of zeroing-out-buffer method? gpuweb/gpuweb#2009, by @litherum.
-
Why do block-compressed textures have to be aligned to a multiple of the texel block size? gpuweb/gpuweb#2006, by @magcius.
-
-
23 issues closed.
-
[meta] Wiki Matrix link is outdated? gpuweb/gpuweb#2070.
-
Binding a zero-sized range of a vertex buffer gpuweb/gpuweb#2044.
-
Implicit BGL conflicts with de-duplication gpuweb/gpuweb#1806.
-
Missing atomicSub function gpuweb/gpuweb#2019.
-
Add opaque region or isOpaque hint gpuweb/gpuweb#1871.
-
Sparse array of GPUBindGroupLayout in GPUPipelineLayoutDescriptor? gpuweb/gpuweb#502.
-
Should the alignment of destinationOffset in resolveQuerySet be 256 on Metal? gpuweb/gpuweb#1845.
-
Creating arrays/structs in expressions gpuweb/gpuweb#2045.
-
Rename forceSoftware and isSoftware gpuweb/gpuweb#2030.
-
bindings naming should (maybe) match attributesSet gpuweb/gpuweb#279.
-
Matrix initialization gpuweb/gpuweb#1964.
-
Typo "fingeprinting" gpuweb/gpuweb#2033.
-
Query API: How to understand the execution time of the Dispatch command gpuweb/gpuweb#992.
-
Buffer indices should be unsigned gpuweb/gpuweb#1135.
-
WGSL spec is not clear on whether array size can be an unsigned literal gpuweb/gpuweb#2012.
-
Invalid Web IDL: maxComputeWorkgroupSize cannot be a dictionary type gpuweb/gpuweb#1936.
-
Handling limits which are not unsigned longs. gpuweb/gpuweb#2018.
-
Tweak workgroup limit names gpuweb/gpuweb#1929.
-
Data races and security gpuweb/gpuweb#1788.
-
stencil gpuweb/gpuweb#2007.
-
-
69 issues received 320 new comments.
-
timestamp-query is unimplementable on TBDR architectures gpuweb/gpuweb#2046 (23 comments by @haoxli, @kainino0x, @Kangz, @kvark, @litherum, @m-schuetz, @mrshannon)
-
Fragment output <-> attachment interface matching rules def gpuweb/gpuweb#2013 (19 comments by @kainino0x, @kvark, @shrekshao)
-
Do pipeline blend states need to be validated against the pipline render targets? gpuweb/gpuweb#2025 (14 comments by @kainino0x, @Kangz, @litherum, @RafaelCintron, @shrekshao)
-
Rename forceSoftware and isSoftware gpuweb/gpuweb#2030 (14 comments by @dneto0, @kainino0x, @kvark, @litherum, @sminnee, @toji)
-
Use brackets for array types gpuweb/gpuweb#854 (12 comments by @dneto0, @jimblandy, @kainino0x, @kdashg, @kvark, @litherum)
-
Shader stage input-output subsetting rules gpuweb/gpuweb#2038 (11 comments by @kainino0x, @Kangz, @kenrussell, @kvark, @litherum)
-
Dealing with holes in the pipeline layout. gpuweb/gpuweb#2043 (11 comments by @frguthmann, @kainino0x, @Kangz, @kvark)
-
Investigation: Storage Texture gpuweb/gpuweb#513 (11 comments by @fintelia, @kainino0x, @Kangz, @kvark, @vorg)
-
[wgsl] Support compile-time-constants (constexpr) gpuweb/gpuweb#1272 (9 comments by @dneto0, @Kangz, @kdashg, @litherum, @RobinMorisset)
-
[meta] Wiki Matrix link is outdated? gpuweb/gpuweb#2070 (9 comments by @dneto0, @kainino0x, @kvark, @ShadowJonathan)
-
Implicit GPUTextureSampleType for texture_2d<f32> gpuweb/gpuweb#2064 (9 comments by @austinEng, @fintelia, @kainino0x, @kvark)
-
Should depth-stencil render attachments require views to have aspect = "all" for combined depth-stencil formats? gpuweb/gpuweb#2062 (8 comments by @austinEng, @kainino0x, @Kangz, @kvark)
-
Fake transfers for getMappedRange-created ArrayBuffers? gpuweb/gpuweb#2072 (8 comments by @kainino0x, @m-schuetz, @pythonesque)
-
Brain dump on all things wrong with WGSL gpuweb/gpuweb#2048 (8 comments by @esoterra, @kainino0x, @morganbarrett, @munrocket)
-
Binding a zero-sized range of a vertex buffer gpuweb/gpuweb#2044 (7 comments by @austinEng, @kainino0x, @Kangz, @kvark)
-
Why do block-compressed textures have to be aligned to a multiple of the texel block size? gpuweb/gpuweb#2006 (7 comments by @Jiawei-Shao, @Kangz, @magcius, @RafaelCintron)
-
Support for attach-less render passes gpuweb/gpuweb#503 (6 comments by @Jiawei-Shao, @kainino0x, @Kangz, @kvark)
-
List of issues to discuss with TC39, WASM, JS engine teams gpuweb/gpuweb#747 (6 comments by @kainino0x, @Kangz, @pythonesque)
-
Feature request: ignore shader writes to color attachments gpuweb/gpuweb#2060 (6 comments by @jobtalle, @kvark, @litherum, @magcius)
-
wgsl: can hex floats explicitly represent infinity and NaN as float literals? gpuweb/gpuweb#1769 (6 comments by @dneto0, @jimblandy, @kdashg, @litherum, @RobinMorisset)
-
WGSL requires dynamic matrix indexing, which requires surprising SPIR-V gpuweb/gpuweb#1782 (5 comments by @kainino0x, @kvark, @munrocket)
-
Creating arrays/structs in expressions gpuweb/gpuweb#2045 (5 comments by @alan-baker, @jrprice, @kdashg, @litherum)
-
Consider limiting buffer sizes to multiples of 4 bytes gpuweb/gpuweb#1999 (5 comments by @austinEng, @kainino0x, @kvark, @litherum)
-
atomicCompareExchangeWeak is not idiomatic gpuweb/gpuweb#2021 (5 comments by @ben-clayton, @dneto0, @kvark)
-
Add method to disable fast-math on a per-shader basis. gpuweb/gpuweb#2076 (4 comments by @kainino0x, @kvark)
-
Should the alignment of destinationOffset in resolveQuerySet be 256 on Metal? gpuweb/gpuweb#1845 (4 comments by @haoxli, @kainino0x, @litherum)
-
Request for a GPUQueue readBuffer API gpuweb/gpuweb#1972 (4 comments by @benvanik, @Kangz, @kdashg, @kvark)
-
Handling limits which are not unsigned longs. gpuweb/gpuweb#2018 (4 comments by @kainino0x, @kvark, @toji)
-
Efficient Per-Frame/Transient Bind Groups gpuweb/gpuweb#915 (4 comments by @benvanik, @kvark, @magcius)
-
Viewport rect must lie entirely within the current attachment size? gpuweb/gpuweb#373 (3 comments by @kainino0x, @Kangz, @magcius)
-
Pad implicit mapping sizes up to multiples of 4 bytes? gpuweb/gpuweb#2010 (3 comments by @kainino0x, @kvark)
-
[WGSL] WGSL needs a plan for expansion gpuweb/gpuweb#600 (3 comments by @krogovin, @litherum, @magcius)
-
Missing clarification on malicious use of GPU resources (e.g. mining) against intent of browser user gpuweb/gpuweb#2049 (2 comments by @kuro68k, @kvark)
-
WGSL spec is not clear on whether array size can be an unsigned literal gpuweb/gpuweb#2012 (2 comments by @kdashg, @kvark)
-
Consider adding a 32- by 32-bit multiplication function resulting in two u32s, high and low gpuweb/gpuweb#2023 (2 comments by @dneto0, @hcsch)
-
Request for a GPUCommandEncoder fillBuffer operation gpuweb/gpuweb#1968 (2 comments by @benvanik, @kainino0x)
-
wgsl: add a note showing that leading zeroes are invalid for number literals, eg. 012345 gpuweb/gpuweb#2075 (1 comment by @dneto0)
-
Investigate possible solution for cheaper opaque canvases gpuweb/gpuweb#1988 (1 comment by @kainino0x)
-
Sparse array of GPUBindGroupLayout in GPUPipelineLayoutDescriptor? gpuweb/gpuweb#502 (1 comment by @kainino0x)
-
Does compositingAlphaMode affect using the canvas as an image source? gpuweb/gpuweb#1847 (1 comment by @kainino0x)
-
Should we constrain the location of user input-output stage variables WGSL gpuweb/gpuweb#1962 (1 comment by @Jiawei-Shao)
-
-
35 pull requests opened.
-
Fix description of mix builtin; fix fma (works on vectors too) gpuweb/gpuweb#2069, by @dneto0.
-
Prevent PipelineLayouts from using implicitly created BindGroupLayouts gpuweb/gpuweb#2068, by @toji.
-
fix mix and fma builtins to match GLSLstd450 gpuweb/gpuweb#2067, by @sarahM0.
-
Changing Float32Array to GPUBuffer in copyBufferToBuffer gpuweb/gpuweb#2066, by @lsh.
-
Describe resource aliasing rules gpuweb/gpuweb#2065, by @litherum.
-
Add atomicSub to WGSL gpuweb/gpuweb#2063, by @kvark.
-
Decrease the minimum required maxTextureArrayLayers to 256 gpuweb/gpuweb#2059, by @Kangz.
-
Fix atan2() params gpuweb/gpuweb#2058, by @munrocket.
-
Make implictly created BGLs incompatible with all other pipelines. gpuweb/gpuweb#2057, by @toji.
-
editorial: move "operation" first in GPUBlendComponent gpuweb/gpuweb#2056, by @kainino0x.
-
Add a memory model mapping to Vulkan gpuweb/gpuweb#2055, by @alan-baker.
-
Queries: Update destinationOffset alignment to 256 gpuweb/gpuweb#2054, by @haoxli.
-
Match let declarations for inferred types gpuweb/gpuweb#2053, by @mehmetoguzderin.
-
Allow indirection gpuweb/gpuweb#2052, by @mehmetoguzderin.
-
Fix example with single closing bracket gpuweb/gpuweb#2051, by @mehmetoguzderin.
-
Restrict statements from matching empty string gpuweb/gpuweb#2050, by @mehmetoguzderin.
-
Added several variants of the WebGPU logo gpuweb/gpuweb#2047, by @toji.
-
Fixing VB OOB validation in draw and drawIndexed for zero stride gpuweb/gpuweb#2042, by @jiangzhaoming.
-
Revert "Fixing VB OOB validation in draw and drawIndexed for zero stride" gpuweb/gpuweb#2041, by @Kangz.
-
Fixing VB OOB validation in draw and drawIndexed for zero stride gpuweb/gpuweb#2040, by @jiangzhaoming.
-
Changed isSoftware/forceSoftware to use the term 'fallback' gpuweb/gpuweb#2039, by @toji.
-
Match attribute grammar with attributes gpuweb/gpuweb#2037, by @mehmetoguzderin.
-
[WGSL] Escape RegEx characters gpuweb/gpuweb#2035, by @mehmetoguzderin.
-
Clarify the types that are valid for ignore() gpuweb/gpuweb#2034, by @jrprice.
-
Clarify out-of-bounds RMW atomics gpuweb/gpuweb#2032, by @alan-baker.
-
array element count can be a module-scope constant, but not overridable gpuweb/gpuweb#2031, by @dneto0.
-
Tweak the description of adapters to not mention accelerators or GPUs explicitly gpuweb/gpuweb#2029, by @litherum.
-
Remove all mentions of 'Metal' by name gpuweb/gpuweb#2028, by @litherum.
-
Mark notes and examples as informative gpuweb/gpuweb#2027, by @litherum.
-
Clarify the uniform array element alignment rule. gpuweb/gpuweb#2026, by @dneto0.
-
Add indirect-first-instance feature. gpuweb/gpuweb#2022, by @mrshannon.
-
Updating limits to only use GPUSize32/GPUSize64 gpuweb/gpuweb#2020, by @toji.
-
Restrict out-of-bounds behaviour for RMW gpuweb/gpuweb#2016, by @alan-baker.
-
A 0x1 literal is ambiguously either a HEX_FLOAT_LITERAL or an INT_LITERAL gpuweb/gpuweb#2014, by @litherum.
-
The reference to Metal's coordinate system should be non-normative gpuweb/gpuweb#2011, by @litherum.
-
-
34 pull requests merged.
-
Fix description of mix builtin; fix fma (works on vectors too) gpuweb/gpuweb#2069.
-
Prevent PipelineLayouts from using implicitly created BindGroupLayouts gpuweb/gpuweb#2068.
-
Fix atan2() params gpuweb/gpuweb#2058.
-
Require [SecureContext] for using WebGPU. gpuweb/gpuweb#1363.
-
Make implictly created BGLs incompatible with all other pipelines. gpuweb/gpuweb#2057.
-
Changing Float32Array to GPUBuffer in copyBufferToBuffer gpuweb/gpuweb#2066.
-
Decrease the minimum required maxTextureArrayLayers to 256 gpuweb/gpuweb#2059.
-
Add atomicSub to WGSL gpuweb/gpuweb#2063.
-
editorial: move "operation" first in GPUBlendComponent gpuweb/gpuweb#2056.
-
Require storage texture access gpuweb/gpuweb#1996.
-
Restrict statements from matching empty string gpuweb/gpuweb#2050.
-
Fix example with single closing bracket gpuweb/gpuweb#2051.
-
Queries: Update destinationOffset alignment to 256 gpuweb/gpuweb#2054.
-
Added several variants of the WebGPU logo gpuweb/gpuweb#2047.
-
Match attribute grammar with attributes gpuweb/gpuweb#2037.
-
Fixing VB OOB validation in draw and drawIndexed for zero stride gpuweb/gpuweb#2042.
-
Changed isSoftware/forceSoftware to use the term 'fallback' gpuweb/gpuweb#2039.
-
Clarify out-of-bounds RMW atomics gpuweb/gpuweb#2032.
-
Add new matrix constructors gpuweb/gpuweb#2001.
-
[WGSL] Escape RegEx characters gpuweb/gpuweb#2035.
-
Revert "Fixing VB OOB validation in draw and drawIndexed for zero stride" gpuweb/gpuweb#2041.
-
Fixing VB OOB validation in draw and drawIndexed for zero stride gpuweb/gpuweb#2040.
-
Remove readable texture storage example gpuweb/gpuweb#1993.
-
Clarify the types that are valid for ignore() gpuweb/gpuweb#2034.
-
Tweak the description of adapters to not mention accelerators or GPUs explicitly gpuweb/gpuweb#2029.
-
Remove all mentions of 'Metal' by name gpuweb/gpuweb#2028.
-
Clarify IO location overlap gpuweb/gpuweb#1998.
-
array element count can be a module-scope constant, but not overridable gpuweb/gpuweb#2031.
-
Clarify the uniform array element alignment rule. gpuweb/gpuweb#2026.
-
Updating limits to only use GPUSize32/GPUSize64 gpuweb/gpuweb#2020.
-
Stencil reads are S,X,X,X. (but expect S,S,S,S or S,0,0,1) gpuweb/gpuweb#1978.
-
Rename compute size limits for clarity gpuweb/gpuweb#1991.
-
wgsl: s/with |T| is a vector/when |T| is a vector gpuweb/gpuweb#2002.
-
The reference to Metal's coordinate system should be non-normative gpuweb/gpuweb#2011.
-
-
8 pull requests closed.
-
fix mix and fma builtins to match GLSLstd450 gpuweb/gpuweb#2067.
-
Associate command encoding with a queue gpuweb/gpuweb#1278.
-
[wgsl]: Add `fmod()`, `frem()` builtins gpuweb/gpuweb#1913.
-
wgsl: array size may be module-scope constant (possibly overridable) gpuweb/gpuweb#1792.
-
Restrict out-of-bounds behaviour for RMW gpuweb/gpuweb#2016.
-
Mark notes and examples as informative gpuweb/gpuweb#2027.
-
A 0x1 literal is ambiguously either a HEX_FLOAT_LITERAL or an INT_LITERAL gpuweb/gpuweb#2014.
-
wgsl: explicitly restrict storage classes for frexp, modf gpuweb/gpuweb#1851.
-
-
43 pull requests received 173 new comments.
-
Add a memory model mapping to Vulkan gpuweb/gpuweb#2055 (13 comments by @alan-baker, @dneto0, @kvark)
-
Require [SecureContext] for using WebGPU. gpuweb/gpuweb#1363 (11 comments by @AshleyScirra, @kvark, @litherum)
-
Updating limits to only use GPUSize32/GPUSize64 gpuweb/gpuweb#2020 (11 comments by @kainino0x, @kdashg, @toji)
-
wgsl: Add limits section, and "spurious" failure gpuweb/gpuweb#1997 (9 comments by @kainino0x, @kdashg, @kvark)
-
Add indirect-first-instance feature. gpuweb/gpuweb#2022 (8 comments by @kainino0x, @kdashg, @mrshannon, @tidoust)
-
[wgsl]: Add `fmod()`, `frem()` builtins gpuweb/gpuweb#1913 (7 comments by @ben-clayton, @kdashg, @kvark, @munrocket)
-
fix mix and fma builtins to match GLSLstd450 gpuweb/gpuweb#2067 (6 comments by @ben-clayton, @dneto0, @sarahM0)
-
array element count can be a module-scope constant, but not overridable gpuweb/gpuweb#2031 (6 comments by @alan-baker, @dneto0, @jrprice, @kvark)
-
Remove all mentions of 'Metal' by name gpuweb/gpuweb#2028 (6 comments by @kainino0x, @kvark, @litherum)
-
Fixing VB OOB validation in draw and drawIndexed for zero stride gpuweb/gpuweb#2042 (5 comments by @jiangzhaoming, @kainino0x, @kvark)
-
Fixing VB OOB validation in draw and drawIndexed for zero stride gpuweb/gpuweb#2040 (5 comments by @jiangzhaoming, @Kangz)
-
wgsl: detailed semantics of integer division and remainder gpuweb/gpuweb#1830 (5 comments by @dneto0, @kdashg, @kvark)
-
Add `multi-draw-indirect` feature. gpuweb/gpuweb#1949 (5 comments by @kainino0x, @litherum, @mrshannon)
-
Decrease the minimum required maxTextureArrayLayers to 256 gpuweb/gpuweb#2059 (4 comments by @fintelia, @kainino0x, @Kangz)
-
wgsl: array size may be module-scope constant (possibly overridable) gpuweb/gpuweb#1792 (4 comments by @dneto0, @kdashg, @litherum)
-
Restrict out-of-bounds behaviour for RMW gpuweb/gpuweb#2016 (4 comments by @alan-baker, @dneto0, @kdashg)
-
Stencil reads are S,X,X,X. (but expect S,S,S,S or S,0,0,1) gpuweb/gpuweb#1978 (4 comments by @kainino0x, @kdashg, @litherum)
-
Fix description of mix builtin; fix fma (works on vectors too) gpuweb/gpuweb#2069 (3 comments by @kdashg, @kvark)
-
Tweak the description of adapters to not mention accelerators or GPUs explicitly gpuweb/gpuweb#2029 (3 comments by @litherum, @toji)
-
Make implictly created BGLs incompatible with all other pipelines. gpuweb/gpuweb#2057 (2 comments by @kvark)
-
editorial: move "operation" first in GPUBlendComponent gpuweb/gpuweb#2056 (2 comments by @Kangz, @kdashg)
-
Changed isSoftware/forceSoftware to use the term 'fallback' gpuweb/gpuweb#2039 (2 comments by @kainino0x, @litherum)
-
A 0x1 literal is ambiguously either a HEX_FLOAT_LITERAL or an INT_LITERAL gpuweb/gpuweb#2014 (1 comment by @alan-baker)
TR publications
-
36 documents published.
-
WebGPU published 18 times as a Working Draft. History of WebGPU
-
WebGPU Shading Language published 18 times as a Working Draft. History of WebGPU Shading Language
-
Participants
-
1 participant joined.
-
lingxiao zhu (ByteDance) joined.
-
GitHub
gpuweb/cts
-
3 issues opened.
-
Adding wgsl validation folder in CTS automatically gpuweb/cts#674, by @munrocket.
-
Fix wgsl validation tests for IO-shareables gpuweb/cts#656, by @sarahM0.
-
Simulate swap for capture tool support gpuweb/cts#635, by @bbernhar.
-
-
2 issues closed.
-
expectCompileResult should check for validation error gpuweb/cts#556.
-
Make expectContents take a "validator" function gpuweb/cts#388.
-
-
4 issues received 7 new comments.
-
Adding wgsl validation folder in CTS automatically gpuweb/cts#674 (3 comments by @kainino0x, @munrocket)
-
webgpu:api,operation,buffers,map_oom:mappedAtCreation,smaller_getMappedRange might be wrong gpuweb/cts#504 (1 comment by @kainino0x)
-
-
73 pull requests opened.
-
Add tests for behavior on lost devices gpuweb/cts#687, by @kainino0x.
-
Add support for resources/ directory gpuweb/cts#686, by @kainino0x.
-
port startPlayingAndWaitForVideo from WebGL CTS gpuweb/cts#685, by @kainino0x.
-
Add validation tests on writeTexture() with depth/stencil formats gpuweb/cts#684, by @Jiawei-Shao.
-
Track resources for cleanup at the end of the test gpuweb/cts#683, by @kainino0x.
-
Limiting instanceCount in robust_access_vertex to 10**4 gpuweb/cts#682, by @jiangzhaoming.
-
Add tests for reading back contents from WebGPU canvas gpuweb/cts#681, by @shaoboyan.
-
Copy to or from a multisampled texture should be allowed gpuweb/cts#680, by @Richard-Yunchao.
-
Add tests for 3D texture copies gpuweb/cts#679, by @Richard-Yunchao.
-
Fix validation errors in operation,vertex_state,index_format gpuweb/cts#678, by @Kangz.
-
Add type validation tests for fragment outputs and attachments gpuweb/cts#676, by @shrekshao.
-
Add ImportExternalTexture Video Test gpuweb/cts#675, by @bjjones.
-
Fix another test failing fragment output validation gpuweb/cts#673, by @austinEng.
-
Review/rewrite createView validation tests gpuweb/cts#672, by @kainino0x.
-
More detailed plans for stress tests gpuweb/cts#671, by @krockot.
-
Add 3 new var duplications wgsl tests gpuweb/cts#670, by @munrocket.
-
Fix var duplication wgsl test gpuweb/cts#669, by @munrocket.
-
Allow splitting slow vertex_state tests gpuweb/cts#668, by @Kangz.
-
Fix validation,image_copy,texture_related:size_alignment:* gpuweb/cts#667, by @Jiawei-Shao.
-
Fixing the vertex buffer robust access test gpuweb/cts#666, by @jiangzhaoming.
-
Fix vertex shader in api,operation,rendering,depth:depth_compare_func gpuweb/cts#665, by @Jiawei-Shao.
-
Fix readme link gpuweb/cts#664, by @munrocket.
-
New shader test: let duplication gpuweb/cts#663, by @munrocket.
-
Fix shaders: const -> let gpuweb/cts#662, by @munrocket.
-
Use min(4, formatSize) for vertex attrib offset alignment gpuweb/cts#661, by @Kangz.
-
vertex_state,correctness: test up to maxVertexAttributes gpuweb/cts#660, by @Kangz.
-
Fix vertex_buffer_used_multiple_times_interleaved vertex OOB error gpuweb/cts#659, by @Kangz.
-
Fix WGSL usage in linear_memory:storageClass="uniform";* gpuweb/cts#657, by @austinEng.
-
Update tests following changes to GPUFragmentState validation gpuweb/cts#655, by @austinEng.
-
Fix writeBuffer in robust_access_vertex gpuweb/cts#653, by @kainino0x.
-
Fix expectCompileResult and variable_and_const tests gpuweb/cts#652, by @kainino0x.
-
Fix Index Buffer OOB validation for DrawIndexed gpuweb/cts#651, by @jiangzhaoming.
-
Add shouldWebGPUCTSFailOnWarnings to template gpuweb/cts#650, by @austinEng.
-
flip select() conditions in large-draw CTS gpuweb/cts#649, by @krockot.
-
Depth/stencil formats should be invalid for 3D textures gpuweb/cts#648, by @Richard-Yunchao.
-
Add control case to timestamp_query,invalid_query_set gpuweb/cts#647, by @kainino0x.
-
Do validation checks inside createEncoder's finish() gpuweb/cts#646, by @kainino0x.
-
StoreOp 'clear' -> 'discard' gpuweb/cts#645, by @kainino0x.
-
WPT: Add an option to fail on warnings gpuweb/cts#644, by @austinEng.
-
Visually indicate test results of entire subtrees gpuweb/cts#643, by @toji.
-
Remove multisampled texture case from in_pass_encoder:replaced_binding gpuweb/cts#642, by @austinEng.
-
Fix filtering sampler validation in createBindGroup.spec.ts gpuweb/cts#641, by @austinEng.
-
Fix robust_access_vertex.spec.ts gpuweb/cts#639, by @austinEng.
-
Disallow zero-sized buffer bind group bindings gpuweb/cts#638, by @austinEng.
-
Plan a stress-test version of createBuffer_invalid_and_oom gpuweb/cts#637, by @kainino0x.
-
Some minor visual updates for the standalone page gpuweb/cts#636, by @toji.
-
Fix in_pass_encoder.ts bindings_in_bundle test gpuweb/cts#634, by @austinEng.
-
createBuffer simultaneous validation error and OOM gpuweb/cts#633, by @kainino0x.
-
More changes to match spec getMappedRange/unmap behavior gpuweb/cts#632, by @kainino0x.
-
Fix WGSL syntax in in_pass_encoder.spec.ts gpuweb/cts#631, by @austinEng.
-
Add a stress test for mapAtCreation+getMappedRange gpuweb/cts#630, by @kainino0x.
-
Fix bugs about texture view descriptor's default values gpuweb/cts#629, by @Richard-Yunchao.
-
implement api,operation,rendering,basic:large_draw:* gpuweb/cts#628, by @krockot.
-
map_oom.spec.ts fixes gpuweb/cts#627, by @shrekshao.
-
api,operation,rendering,basic:large_draw plan gpuweb/cts#626, by @krockot.
-
Raise up blending.spec.ts tolerance gpuweb/cts#625, by @shrekshao.
-
Add Validation tests for CopyExternalImageToTexture() gpuweb/cts#624, by @shaoboyan.
-
Fix a bug about texture and view dimension type compatibility gpuweb/cts#623, by @Richard-Yunchao.
-
Update api,validation,buffer,mapping:* gpuweb/cts#621, by @shrekshao.
-
Add test plan CopyTextureToBuffer() with depth aspect operation tests gpuweb/cts#620, by @Jiawei-Shao.
-
Implement the rest of the vertex_state correctness tests. gpuweb/cts#619, by @Kangz.
-
Implement api,validation,encoding,cmds,render,other,* gpuweb/cts#618, by @haoxli.
-
Add new limits to DefaultLimits, update compute tests gpuweb/cts#617, by @kainino0x.
-
Add memcpy helper gpuweb/cts#616, by @kainino0x.
-
operation/vertex_state: split runTest into smaller functions gpuweb/cts#615, by @Kangz.
-
Implement api,operation,command_buffer,programmable,state_tracking:* gpuweb/cts#614, by @toji.
-
Test that mapped buffers can be read with a variety of TypedArrays gpuweb/cts#613, by @toji.
-
Port and refactor robust access tests gpuweb/cts#612, by @kainino0x.
-
-
66 pull requests merged.
-
Add support for resources/ directory gpuweb/cts#686.
-
port startPlayingAndWaitForVideo from WebGL CTS gpuweb/cts#685.
-
Add tests for reading back contents from WebGPU canvas gpuweb/cts#681.
-
Review/rewrite createView validation tests gpuweb/cts#672.
-
Limiting instanceCount in robust_access_vertex to 10**4 gpuweb/cts#682.
-
Fix validation errors in operation,vertex_state,index_format gpuweb/cts#678.
-
Fixing the vertex buffer robust access test gpuweb/cts#666.
-
Add Validation tests for CopyExternalImageToTexture() gpuweb/cts#624.
-
Fix another test failing fragment output validation gpuweb/cts#673.
-
Fix var duplication wgsl test gpuweb/cts#669.
-
Add 3 new var duplications wgsl tests gpuweb/cts#670.
-
Fix validation,image_copy,texture_related:size_alignment:* gpuweb/cts#667.
-
Allow splitting slow vertex_state tests gpuweb/cts#668.
-
Fix shaders: const -> let gpuweb/cts#662.
-
Fix readme link gpuweb/cts#664.
-
Fix WGSL usage in linear_memory:storageClass="uniform";* gpuweb/cts#657.
-
Update tests following changes to GPUFragmentState validation gpuweb/cts#655.
-
Use min(4, formatSize) for vertex attrib offset alignment gpuweb/cts#661.
-
Fix expectCompileResult and variable_and_const tests gpuweb/cts#652.
-
Add control case to timestamp_query,invalid_query_set gpuweb/cts#647.
-
vertex_state,correctness: test up to maxVertexAttributes gpuweb/cts#660.
-
Fix Index Buffer OOB validation for DrawIndexed gpuweb/cts#651.
-
Fix writeBuffer in robust_access_vertex gpuweb/cts#653.
-
Do validation checks inside createEncoder's finish() gpuweb/cts#646.
-
Depth/stencil formats should be invalid for 3D textures gpuweb/cts#648.
-
Implement the rest of the vertex_state correctness tests. gpuweb/cts#619.
-
Visually indicate test results of entire subtrees gpuweb/cts#643.
-
Add shouldWebGPUCTSFailOnWarnings to template gpuweb/cts#650.
-
WPT: Add an option to fail on warnings gpuweb/cts#644.
-
Disallow zero-sized buffer bind group bindings gpuweb/cts#638.
-
Plan a stress-test version of createBuffer_invalid_and_oom gpuweb/cts#637.
-
flip select() conditions in large-draw CTS gpuweb/cts#649.
-
StoreOp 'clear' -> 'discard' gpuweb/cts#645.
-
Fix filtering sampler validation in createBindGroup.spec.ts gpuweb/cts#641.
-
Fix robust_access_vertex.spec.ts gpuweb/cts#639.
-
Add new limits to DefaultLimits, update compute tests gpuweb/cts#617.
-
Some minor visual updates for the standalone page gpuweb/cts#636.
-
Fix in_pass_encoder.ts bindings_in_bundle test gpuweb/cts#634.
-
createBuffer simultaneous validation error and OOM gpuweb/cts#633.
-
implement api,operation,rendering,basic:large_draw:* gpuweb/cts#628.
-
Add a stress test for mapAtCreation+getMappedRange gpuweb/cts#630.
-
More changes to match spec getMappedRange/unmap behavior gpuweb/cts#632.
-
Fix WGSL syntax in in_pass_encoder.spec.ts gpuweb/cts#631.
-
map_oom.spec.ts fixes gpuweb/cts#627.
-
Fix bugs about texture view descriptor's default values gpuweb/cts#629.
-
Fix a bug about texture and view dimension type compatibility gpuweb/cts#623.
-
Implement api,validation,encoding,cmds,render,other,* gpuweb/cts#618.
-
api,operation,rendering,basic:large_draw plan gpuweb/cts#626.
-
Raise up blending.spec.ts tolerance gpuweb/cts#625.
-
Update api,validation,buffer,mapping:* gpuweb/cts#621.
-
operation/vertex_state: split runTest into smaller functions gpuweb/cts#615.
-
api,validation,encoding,cmds,render,other plans gpuweb/cts#603.
-
Port and refactor robust access tests gpuweb/cts#612.
-
-
3 pull requests closed.
-
Copy to or from a multisampled texture should be allowed gpuweb/cts#680.
-
New shader test: let duplication gpuweb/cts#663.
-
Port the robust access tests from GLSL to WGSL gpuweb/cts#380.
-
-
50 pull requests received 402 new comments.
-
Fixing the vertex buffer robust access test gpuweb/cts#666 (65 comments by @jiangzhaoming, @kainino0x, @Kangz)
-
Add Validation tests for CopyExternalImageToTexture() gpuweb/cts#624 (35 comments by @kainino0x, @shaoboyan)
-
Add operation tests on writeTexture() and B2T copies with stencil aspects gpuweb/cts#606 (25 comments by @Jiawei-Shao, @kainino0x, @Kangz)
-
implement api,operation,rendering,basic:large_draw:* gpuweb/cts#628 (22 comments by @austinEng, @kainino0x, @krockot)
-
Add type validation tests for fragment outputs and attachments gpuweb/cts#676 (16 comments by @kainino0x, @Kangz, @shrekshao)
-
Review/rewrite createView validation tests gpuweb/cts#672 (16 comments by @kainino0x, @Richard-Yunchao)
-
Depth/stencil formats should be invalid for 3D textures gpuweb/cts#648 (15 comments by @kainino0x, @Richard-Yunchao)
-
Visually indicate test results of entire subtrees gpuweb/cts#643 (13 comments by @austinEng, @kainino0x, @toji)
-
Fix Index Buffer OOB validation for DrawIndexed gpuweb/cts#651 (12 comments by @jiangzhaoming, @kainino0x, @Kangz, @shrekshao)
-
Implement the rest of the vertex_state correctness tests. gpuweb/cts#619 (12 comments by @austinEng, @kainino0x, @Kangz)
-
Copy to or from a multisampled texture should be allowed gpuweb/cts#680 (11 comments by @Jiawei-Shao, @kainino0x, @Kangz, @Richard-Yunchao)
-
Implement api,validation,encoding,cmds,render,other,* gpuweb/cts#618 (11 comments by @haoxli, @kainino0x)
-
Do validation checks inside createEncoder's finish() gpuweb/cts#646 (7 comments by @egalli, @kainino0x)
-
operation/vertex_state: split runTest into smaller functions gpuweb/cts#615 (7 comments by @kainino0x, @Kangz)
-
Add tests for reading back contents from WebGPU canvas gpuweb/cts#681 (5 comments by @kainino0x, @shaoboyan)
-
Fix in_pass_encoder:subresources_and_binding_types_combination_for_aspect:* gpuweb/cts#658 (5 comments by @austinEng, @kainino0x)
-
Change blending test target texture format to rgba16float and add helpers for float16 gpuweb/cts#622 (5 comments by @kainino0x, @shrekshao)
-
More changes to match spec getMappedRange/unmap behavior gpuweb/cts#632 (3 comments by @kainino0x, @shrekshao)
-
Add test plan CopyTextureToBuffer() with depth aspect operation tests gpuweb/cts#620 (3 comments by @Kangz)
-
Fix WGSL usage in linear_memory:storageClass="uniform";* gpuweb/cts#657 (2 comments by @austinEng, @kainino0x)
-
Fix vertex_buffer_used_multiple_times_interleaved vertex OOB error gpuweb/cts#659 (2 comments by @Kangz)
-
Fix a bug about texture and view dimension type compatibility gpuweb/cts#623 (2 comments by @Richard-Yunchao)
-
Test that buffer writes are preserved after another range is mapped. gpuweb/cts#610 (2 comments by @kainino0x, @toji)
-
Fix validation errors in operation,vertex_state,index_format gpuweb/cts#678 (1 comment by @austinEng)
-
Fix bugs about texture view descriptor's default values gpuweb/cts#629 (1 comment by @Richard-Yunchao)
-
Implement api,operation,command_buffer,programmable,state_tracking:* gpuweb/cts#614 (1 comment by @kainino0x)
-
gpuweb/gpuweb
-
40 issues opened.
-
Fix Y2K bugs in WebGPU gpuweb/gpuweb#2000, by @kainino0x.
-
Consider limiting buffer sizes to multiples of 4 bytes gpuweb/gpuweb#1999, by @kainino0x.
-
Consider adding a [[sample_position]] gpuweb/gpuweb#1994, by @Kangz.
-
wsgl:editorial: missing access mode in a few descriptions gpuweb/gpuweb#1992, by @dneto0.
-
Investigate possible solution for cheaper opaque canvases gpuweb/gpuweb#1988, by @kainino0x.
-
Require loops to terminate gpuweb/gpuweb#1987, by @ben-clayton.
-
Do we have a plan for non-32bit literals? gpuweb/gpuweb#1986, by @ben-clayton.
-
builtin attributes on struct members gpuweb/gpuweb#1981, by @sarahM0.
-
Problem with the default queue gpuweb/gpuweb#1977, by @kvark.
-
Add to text: Global vars/consts declaration initializers must be const_expr gpuweb/gpuweb#1975, by @sarahM0.
-
Request for a GPUQueue readBuffer API gpuweb/gpuweb#1972, by @benvanik.
-
Are command buffers reusable? gpuweb/gpuweb#1971, by @benvanik.
-
[editorial] Make object creation to respect the device timeline gpuweb/gpuweb#1970, by @kvark.
-
Feedback on GPUCommandEncoder missing a transfer pass encoder gpuweb/gpuweb#1969, by @benvanik.
-
Request for a GPUCommandEncoder fillBuffer operation gpuweb/gpuweb#1968, by @benvanik.
-
unpack4x8unorm give "no definition in scope for identifier: 'unpack4x8unorm' " gpuweb/gpuweb#1966, by @pwouik.
-
Matrix initialization gpuweb/gpuweb#1964, by @munrocket.
-
Should we constrain the location of user input-output stage variables WGSL gpuweb/gpuweb#1962, by @Kangz.
-
Using structured buffers in HLSL gpuweb/gpuweb#1961, by @kvark.
-
[wgsl] Limit number of function parameters to 255 gpuweb/gpuweb#1959, by @ben-clayton.
-
Invariant attributes usage gpuweb/gpuweb#1958, by @sarahM0.
-
Is Timestamp Query not supported on Apple M1? gpuweb/gpuweb#1956, by @haoxli.
-
Clarify the vertex stage output / fragment stage input matching rules gpuweb/gpuweb#1953, by @Jiawei-Shao.
-
cleanup - textureSample*() offset argument limitations gpuweb/gpuweb#1948, by @sarahM0.
-
Document cubemap face orientation gpuweb/gpuweb#1946, by @kvark.
-
Vector component assignment gpuweb/gpuweb#1937, by @munrocket.
-
Invalid Web IDL: maxComputeWorkgroupSize cannot be a dictionary type gpuweb/gpuweb#1936, by @tidoust.
-
Consider allowing access::read on function and private storage classes gpuweb/gpuweb#1935, by @jrprice.
-
MSL does not support address-of vector component gpuweb/gpuweb#1931, by @jrprice.
-
Tweak workgroup limit names gpuweb/gpuweb#1929, by @kainino0x.
-
Pipeline layout changes invalidate all bindings on DX12 gpuweb/gpuweb#1926, by @kvark.
-
[meta] DX12 implementation friction gpuweb/gpuweb#1925, by @kvark.
-
Reading from MSAA depth textures gpuweb/gpuweb#1924, by @kvark.
-
Consider passing the stride as an argument to `setVertexBuffer` gpuweb/gpuweb#1921, by @kvark.
-
WGSL: Allow global let-declarations initialized with constant expressions gpuweb/gpuweb#1911, by @Jiawei-Shao.
-
WGSL: Support type conversion from bool to int/uint/float gpuweb/gpuweb#1910, by @Jiawei-Shao.
-
[meta] Draw-time validation gpuweb/gpuweb#1907, by @kvark.
-
Lifetime of GPUBindGroupLayout and GPURender/ComputePipeline objects? gpuweb/gpuweb#1904, by @juj.
-
Why does creating a GPUTexture require a format field? gpuweb/gpuweb#1902, by @juj.
-
Allow render pipelines to have color targets not written to by fragment outputs gpuweb/gpuweb#1900, by @austinEng.
-
-
27 issues closed.
-
[wgsl] Proposal: Remove pointer out parameters from modf, frexp gpuweb/gpuweb#1480.
-
We should forbid pointers to handles gpuweb/gpuweb#1832.
-
Task: Define accesses into vectors and matrices gpuweb/gpuweb#671.
-
Support mix(vec, vec, f32) gpuweb/gpuweb#1588.
-
Dispatch usage scopes contain unused bindgroups gpuweb/gpuweb#1708.
-
Allow render pipelines to have color targets not written to by fragment outputs gpuweb/gpuweb#1900.
-
unpack4x8unorm give "no definition in scope for identifier: 'unpack4x8unorm' " gpuweb/gpuweb#1966.
-
wgsl editorial: tighten IO-shareable gpuweb/gpuweb#1877.
-
cleanup - textureSample*() offset argument limitations gpuweb/gpuweb#1948.
-
Invariant attributes usage gpuweb/gpuweb#1958.
-
Pipeline layout changes invalidate all bindings on DX12 gpuweb/gpuweb#1926.
-
WGSL: Support type conversion from bool to int/uint/float gpuweb/gpuweb#1910.
-
Consider whether we need a quantize-f32-value-to-f16 gpuweb/gpuweb#704.
-
Vector component assignment gpuweb/gpuweb#1937.
-
Define WebGPURenderingContext contextType DomString gpuweb/gpuweb#131.
-
"Sampled" is ambiguous in the spec gpuweb/gpuweb#1785.
-
Consider passing the stride as an argument to `setVertexBuffer` gpuweb/gpuweb#1921.
-
OOB behaviour for storage textures gpuweb/gpuweb#1895.
-
need saturating conversions between signed/unsigned int scalars and vectors gpuweb/gpuweb#720.
-
Remove the requirement for Vulkan's drawIndirectFirstInstance feature. gpuweb/gpuweb#1878.
-
"workgroup_size" builtin is not supported on some Apple platforms gpuweb/gpuweb#1590.
-
API should be exposed to DedicatedWorker gpuweb/gpuweb#1882.
-
Operations happening in GPUVertexState are underspecified gpuweb/gpuweb#1880.
-
wgsl: Consider reserving `vec` and `mat` gpuweb/gpuweb#1862.
-
Lifetime of GPUBindGroupLayout and GPURender/ComputePipeline objects? gpuweb/gpuweb#1904.
-
Why does creating a GPUTexture require a format field? gpuweb/gpuweb#1902.
-
-
72 issues received 274 new comments.
-
Matrix initialization gpuweb/gpuweb#1964 (21 comments by @alan-baker, @dneto0, @kainino0x, @Kangz, @kdashg, @kvark, @litherum, @munrocket, @pow2clk)
-
when does applying an attribute to a type result in a different type gpuweb/gpuweb#1534 (16 comments by @ben-clayton, @dneto0, @jrprice, @kainino0x, @kdashg, @kvark)
-
[wgsl] Proposal: Remove pointer out parameters from modf, frexp gpuweb/gpuweb#1480 (14 comments by @ben-clayton, @dneto0, @kdashg, @kvark, @litherum)
-
Buffer indices should be unsigned gpuweb/gpuweb#1135 (14 comments by @dneto0, @kdashg, @kvark, @litherum)
-
Using structured buffers in HLSL gpuweb/gpuweb#1961 (10 comments by @ben-clayton, @dj2, @Kangz, @kvark)
-
Request for a GPUQueue readBuffer API gpuweb/gpuweb#1972 (9 comments by @benvanik, @fintelia, @kainino0x, @Kangz, @kvark, @litherum)
-
builtin attributes on struct members gpuweb/gpuweb#1981 (8 comments by @alan-baker, @dneto0, @kvark, @sarahM0)
-
Request for a GPUCommandEncoder fillBuffer operation gpuweb/gpuweb#1968 (8 comments by @benvanik, @kainino0x, @Kangz, @litherum)
-
Is Timestamp Query not supported on Apple M1? gpuweb/gpuweb#1956 (8 comments by @haoxli, @Kangz, @kvark)
-
Reading from MSAA depth textures gpuweb/gpuweb#1924 (7 comments by @austinEng, @kainino0x, @Kangz, @kdashg, @kvark, @litherum)
-
Should the API enforce that writable storage buffer bindings don't alias each other? gpuweb/gpuweb#1842 (7 comments by @benvanik, @kainino0x, @Kangz, @kvark, @litherum)
-
Should we constrain the location of user input-output stage variables WGSL gpuweb/gpuweb#1962 (6 comments by @Jiawei-Shao, @Kangz, @kvark)
-
2nd, 3rd, and 4th components of sampled depth/stencil textures are undefined in Metal gpuweb/gpuweb#1198 (6 comments by @jimblandy, @kainino0x, @kdashg, @kvark)
-
Consider allowing access::read on function and private storage classes gpuweb/gpuweb#1935 (6 comments by @alan-baker, @jrprice, @kvark)
-
Pipeline layout changes invalidate all bindings on DX12 gpuweb/gpuweb#1926 (6 comments by @Kangz, @kvark, @magcius)
-
can pipeline creation fail in the driver compiler? particularly due to shader compilation failure gpuweb/gpuweb#1872 (6 comments by @dneto0, @kvark)
-
Invalid Web IDL: maxComputeWorkgroupSize cannot be a dictionary type gpuweb/gpuweb#1936 (5 comments by @kainino0x, @Kangz, @kvark)
-
Investigate possible solution for cheaper opaque canvases gpuweb/gpuweb#1988 (4 comments by @kainino0x, @Kangz, @kvark)
-
[wgsl] Limit number of function parameters to 255 gpuweb/gpuweb#1959 (4 comments by @dneto0, @kainino0x, @kdashg)
-
Remove the requirement for Vulkan's drawIndirectFirstInstance feature. gpuweb/gpuweb#1878 (4 comments by @kainino0x, @kvark, @mrshannon)
-
Efficient Per-Frame/Transient Bind Groups gpuweb/gpuweb#915 (3 comments by @benvanik, @Kangz, @kvark)
-
Feedback on GPUCommandEncoder missing a transfer pass encoder gpuweb/gpuweb#1969 (3 comments by @benvanik, @kainino0x, @kvark)
-
Allow render pipelines to have color targets not written to by fragment outputs gpuweb/gpuweb#1900 (3 comments by @austinEng, @kvark, @vorg)
-
WGSL: Support type conversion from bool to int/uint/float gpuweb/gpuweb#1910 (3 comments by @dneto0, @kdashg)
-
Consider passing the stride as an argument to `setVertexBuffer` gpuweb/gpuweb#1921 (3 comments by @austinEng, @kvark, @magcius)
-
WGSL: Allow global let-declarations initialized with constant expressions gpuweb/gpuweb#1911 (3 comments by @dneto0, @kdashg)
-
Define WebGPURenderingContext contextType DomString gpuweb/gpuweb#131 (3 comments by @kainino0x, @kvark, @toji)
-
Consider limiting buffer sizes to multiples of 4 bytes gpuweb/gpuweb#1999 (2 comments by @kainino0x, @Kangz)
-
wgsl: can hex floats explicitly represent infinity and NaN as float literals? gpuweb/gpuweb#1769 (2 comments by @dneto0, @kdashg)
-
Clarify the vertex stage output / fragment stage input matching rules gpuweb/gpuweb#1953 (2 comments by @alan-baker, @Kangz)
-
MSL does not support address-of vector component gpuweb/gpuweb#1931 (2 comments by @alan-baker, @kvark)
-
Consider whether we need a quantize-f32-value-to-f16 gpuweb/gpuweb#704 (2 comments by @dneto0, @kdashg)
-
need saturating conversions between signed/unsigned int scalars and vectors gpuweb/gpuweb#720 (2 comments by @alan-baker, @dneto0)
-
Make error scope objects have same shape as DOMExceptions? gpuweb/gpuweb#1884 (2 comments by @juj, @kainino0x)
-
wgsl: Restrict `modf` and `frexp`pointer parameter storage classes gpuweb/gpuweb#1846 (2 comments by @ben-clayton, @kvark)
-
[wgsl] functions with return values + return + control flow gpuweb/gpuweb#1897 (2 comments by @dneto0)
-
[wgsl] Define the interface of an entry point in a WGSL program gpuweb/gpuweb#774 (1 comment by @kdashg)
-
Add to text: Global vars/consts declaration initializers must be const_expr gpuweb/gpuweb#1975 (1 comment by @munrocket)
-
[editorial] Make object creation to respect the device timeline gpuweb/gpuweb#1970 (1 comment by @kvark)
-
unpack4x8unorm give "no definition in scope for identifier: 'unpack4x8unorm' " gpuweb/gpuweb#1966 (1 comment by @kainino0x)
-
"workgroup_size" builtin is not supported on some Apple platforms gpuweb/gpuweb#1590 (1 comment by @kdashg)
-
WebGPU should validate that storageTexture bindings have a single mip level. gpuweb/gpuweb#1576 (1 comment by @Kangz)
-
Investigate remaining GLSL.std.450 methods for inclusion. gpuweb/gpuweb#1087 (1 comment by @jiangzhaoming)
-
-
63 pull requests opened.
-
wgsl: s/with |T| is a vector/when |T| is a vector gpuweb/gpuweb#2002, by @525c1e21-bd67-4735-ac99-b4b0e5262290.
-
Add new matrix constructors gpuweb/gpuweb#2001, by @munrocket.
-
Clarify IO location overlap gpuweb/gpuweb#1998, by @dneto0.
-
wgsl: Add limits section, and "spurious" failure gpuweb/gpuweb#1997, by @dneto0.
-
Require storage texture access gpuweb/gpuweb#1996, by @kvark.
-
Fixes buffer storage access modes gpuweb/gpuweb#1995, by @kvark.
-
Remove readable texture storage example gpuweb/gpuweb#1993, by @kvark.
-
Rename compute size limits for clarity gpuweb/gpuweb#1991, by @kainino0x.
-
Change lodMaxClamp default to 32 gpuweb/gpuweb#1990, by @kainino0x.
-
Rename texture usage flags to bindings gpuweb/gpuweb#1989, by @kvark.
-
wgsl: minor fix for unused variable gpuweb/gpuweb#1985, by @RvanderLaan.
-
Fix typo: "independent" gpuweb/gpuweb#1984, by @dneto0.
-
Disallow taking the address of a handle gpuweb/gpuweb#1983, by @alan-baker.
-
wgsl: minor fix gpuweb/gpuweb#1982, by @munrocket.
-
Fix GPUVertexAttribute.offset alignment requirements for D3D12 gpuweb/gpuweb#1980, by @Kangz.
-
Refactor the pipeline-pass compatibility and check RODS gpuweb/gpuweb#1979, by @kvark.
-
Stencil reads are S,X,X,X. (but expect S,S,S,S or S,0,0,1) gpuweb/gpuweb#1978, by @kdashg.
-
Add convenience type aliases [fiub]vecN and fmatNxM. gpuweb/gpuweb#1976, by @kdashg.
-
Add mix(vec, vec, f32) overload gpuweb/gpuweb#1974, by @kvark.
-
wgsl: Change `modf` and `frexp` to return a structure gpuweb/gpuweb#1973, by @ben-clayton.
-
Describe pipeline-renderpass compatibility gpuweb/gpuweb#1967, by @kvark.
-
[wgsl] Fix definition of length and distance gpuweb/gpuweb#1965, by @jrprice.
-
Rename SHADER_READ back to SAMPLED gpuweb/gpuweb#1963, by @Kangz.
-
Add explicit validation between sampleType/etc and WGSL gpuweb/gpuweb#1960, by @kainino0x.
-
[wgsl] Fix parameter order of `select()` gpuweb/gpuweb#1957, by @ben-clayton.
-
Require copyExternalImageToTexture canvases to be 2d/webgl/webgl2 gpuweb/gpuweb#1955, by @kainino0x.
-
Clarify dispatch usage scope gpuweb/gpuweb#1954, by @kainino0x.
-
wgsl: type checking applies to const_expr too gpuweb/gpuweb#1952, by @dneto0.
-
wgsl: tighten constraints on 'offset' param to texture sampling gpuweb/gpuweb#1951, by @dneto0.
-
Simplify IO-shareable gpuweb/gpuweb#1950, by @dneto0.
-
Add `multi-draw-indirect` feature. gpuweb/gpuweb#1949, by @mrshannon.
-
Editorial: clarify unmap() validation rules; fix typos gpuweb/gpuweb#1947, by @kainino0x.
-
wgsl: Change `operator%(f32,32)` to use `OpFRem` gpuweb/gpuweb#1945, by @ben-clayton.
-
Fix module-scope constant declarations wrt type inference gpuweb/gpuweb#1944, by @kvark.
-
Issue 1534's array types differ on stride gpuweb/gpuweb#1943, by @dneto0.
-
Allow unsigned wgsl array sizes. gpuweb/gpuweb#1942, by @kdashg.
-
Use 64bit limits for max uniform/storage buffer binding sizes. gpuweb/gpuweb#1941, by @kdashg.
-
Add quantizeToF16 builtin function gpuweb/gpuweb#1940, by @dneto0.
-
wgsl: Properly reference IEEE 754 gpuweb/gpuweb#1939, by @dneto0.
-
[wgsl] Add `textureDimensions(t: texture_1d<T>, level: i32)` gpuweb/gpuweb#1938, by @ben-clayton.
-
wgsl: vector bool conversion to numeric vectors gpuweb/gpuweb#1934, by @dneto0.
-
wgsl: add conversion from bool to other scalars gpuweb/gpuweb#1933, by @dneto0.
-
Clarify depth/stencil copy validation gpuweb/gpuweb#1932, by @kvark.
-
Change context type to 'webgpu' gpuweb/gpuweb#1930, by @toji.
-
Require indirect draw calls use a firstInstance of 0 gpuweb/gpuweb#1928, by @toji.
-
Rename GPUInputStepMode to GPUVertexStepMode gpuweb/gpuweb#1927, by @kvark.
-
[wgsl] Add `textureNumLevels` overload for texture_1d gpuweb/gpuweb#1923, by @ben-clayton.
-
wgsl: fix typo on zero-valued vectors example gpuweb/gpuweb#1922, by @Arcnor.
-
Fix smoothstep semantics gpuweb/gpuweb#1920, by @kvark.
-
Validate configured context width/height is > 0 gpuweb/gpuweb#1919, by @toji.
-
Loosen pipeline fragment output validation gpuweb/gpuweb#1918, by @austinEng.
-
Fix inconsistency in draw call parameter name gpuweb/gpuweb#1917, by @krockot.
-
Add canvas to GPUPresentationContext gpuweb/gpuweb#1916, by @toji.
-
[wgsl] Improve the Memory Layout sections gpuweb/gpuweb#1915, by @ben-clayton.
-
[wgsl]: Remove scalar overloads of `faceForward` and `reflect` gpuweb/gpuweb#1914, by @ben-clayton.
-
[wgsl]: Add `fmod()`, `frem()` builtins gpuweb/gpuweb#1913, by @ben-clayton.
-
wgsl: specify the input type of reflect to be vector gpuweb/gpuweb#1912, by @jiangzhaoming.
-
Add round_up_size attribute gpuweb/gpuweb#1909, by @alan-baker.
-
Define vertex data conversions and validation gpuweb/gpuweb#1908, by @kvark.
-
Clarify out of bounds texture operations gpuweb/gpuweb#1906, by @alan-baker.
-
wgsl: add refract as a built-in function gpuweb/gpuweb#1901, by @jiangzhaoming.
-
Add GPUExternalTexture Destroy gpuweb/gpuweb#1899, by @bjjones.
-
Fix the compute limits, adds maxComputeWorkgroupSize gpuweb/gpuweb#1898, by @Kangz.
-
-
56 pull requests merged.
-
wgsl: Change `modf` and `frexp` to return a structure gpuweb/gpuweb#1973.
-
Fixes buffer storage access modes gpuweb/gpuweb#1995.
-
wgsl: refine pointer function call restrictions gpuweb/gpuweb#1855.
-
Allow unsigned wgsl array sizes. gpuweb/gpuweb#1942.
-
Disallow taking the address of a handle gpuweb/gpuweb#1983.
-
Change lodMaxClamp default to 32 gpuweb/gpuweb#1990.
-
copyExternalImageToTexture: add validation, make validation async gpuweb/gpuweb#1737.
-
Rename texture usage flags to bindings gpuweb/gpuweb#1989.
-
wgsl: minor fix for unused variable gpuweb/gpuweb#1985.
-
wgsl: minor fix gpuweb/gpuweb#1982.
-
Fix typo: "independent" gpuweb/gpuweb#1984.
-
Refactor the pipeline-pass compatibility and check RODS gpuweb/gpuweb#1979.
-
Add mix(vec, vec, f32) overload gpuweb/gpuweb#1974.
-
Fix GPUVertexAttribute.offset alignment requirements for D3D12 gpuweb/gpuweb#1980.
-
Define index/vertex buffer OOB behavior gpuweb/gpuweb#1835.
-
Add explicit validation between sampleType/etc and WGSL gpuweb/gpuweb#1960.
-
wgsl: Change `operator%(f32,32)` to use `OpFRem` gpuweb/gpuweb#1945.
-
Describe pipeline-renderpass compatibility gpuweb/gpuweb#1967.
-
Add texture_depth_multisampled_2d gpuweb/gpuweb#1793.
-
[wgsl] Fix definition of length and distance gpuweb/gpuweb#1965.
-
Loosen pipeline fragment output validation gpuweb/gpuweb#1918.
-
wgsl: type checking applies to const_expr too gpuweb/gpuweb#1952.
-
Simplify IO-shareable gpuweb/gpuweb#1950.
-
Require copyExternalImageToTexture canvases to be 2d/webgl/webgl2 gpuweb/gpuweb#1955.
-
Clarify dispatch usage scope gpuweb/gpuweb#1954.
-
wgsl: tighten constraints on 'offset' param to texture sampling gpuweb/gpuweb#1951.
-
[wgsl] Fix parameter order of `select()` gpuweb/gpuweb#1957.
-
Editorial: clarify unmap() validation rules; fix typos gpuweb/gpuweb#1947.
-
wgsl: vector bool conversion to numeric vectors gpuweb/gpuweb#1934.
-
Use 64bit limits for max uniform/storage buffer binding sizes. gpuweb/gpuweb#1941.
-
[wgsl] Add `textureDimensions(t: texture_1d<T>, level: i32)` gpuweb/gpuweb#1938.
-
wgsl: add conversion from bool to other scalars gpuweb/gpuweb#1933.
-
Add quantizeToF16 builtin function gpuweb/gpuweb#1940.
-
wgsl: Properly reference IEEE 754 gpuweb/gpuweb#1939.
-
[wgsl] Add `textureNumLevels` overload for texture_1d gpuweb/gpuweb#1923.
-
Change context type to 'webgpu' gpuweb/gpuweb#1930.
-
Remove the read-only texture storage bindings gpuweb/gpuweb#1794.
-
Clarify depth/stencil copy validation gpuweb/gpuweb#1932.
-
Match introduction with title gpuweb/gpuweb#1886.
-
Clarify out of bounds texture operations gpuweb/gpuweb#1906.
-
[wgsl] Improve the Memory Layout sections gpuweb/gpuweb#1915.
-
wgsl: fix typo on zero-valued vectors example gpuweb/gpuweb#1922.
-
Fix the compute limits, adds maxComputeWorkgroupSize gpuweb/gpuweb#1898.
-
Require indirect draw calls use a firstInstance of 0 gpuweb/gpuweb#1928.
-
Rename GPUInputStepMode to GPUVertexStepMode gpuweb/gpuweb#1927.
-
Add canvas to GPUPresentationContext gpuweb/gpuweb#1916.
-
Expose all interfaces to DedicatedWorker gpuweb/gpuweb#1883.
-
Fix smoothstep semantics gpuweb/gpuweb#1920.
-
Validate configured context width/height is > 0 gpuweb/gpuweb#1919.
-
[wgsl]: Remove scalar overloads of `faceForward` and `reflect` gpuweb/gpuweb#1914.
-
Fix inconsistency in draw call parameter name gpuweb/gpuweb#1917.
-
Define vertex data conversions and validation gpuweb/gpuweb#1908.
-
wgsl: add refract as a built-in function gpuweb/gpuweb#1901.
-
wgsl: Reserve `vec` and `mat` keywords gpuweb/gpuweb#1896.
-
Update size range of "stencil8" format gpuweb/gpuweb#1888.
-
Reduce maxStorageTexturePerStage from 8 to 4. gpuweb/gpuweb#1879.
-
-
4 pull requests closed.
-
Rename SHADER_READ back to SAMPLED gpuweb/gpuweb#1963.
-
Change RO texture usage to be restricted gpuweb/gpuweb#1864.
-
wgsl: specify the input type of reflect to be vector gpuweb/gpuweb#1912.
-
Add GPUExternalTexture Destroy gpuweb/gpuweb#1899.
-
-
62 pull requests received 294 new comments.
-
Add `multi-draw-indirect` feature. gpuweb/gpuweb#1949 (20 comments by @kainino0x, @Kangz, @kvark, @litherum, @mrshannon)
-
Fix the compute limits, adds maxComputeWorkgroupSize gpuweb/gpuweb#1898 (17 comments by @alan-baker, @kainino0x, @Kangz, @kvark)
-
wgsl: array size may be module-scope constant (possibly overridable) gpuweb/gpuweb#1792 (16 comments by @alan-baker, @dneto0, @kainino0x, @kdashg, @kvark, @litherum)
-
Describe pipeline-renderpass compatibility gpuweb/gpuweb#1967 (13 comments by @kainino0x, @Kangz, @kvark)
-
Add explicit validation between sampleType/etc and WGSL gpuweb/gpuweb#1960 (12 comments by @kainino0x, @kvark, @toji)
-
[wgsl] Improve the Memory Layout sections gpuweb/gpuweb#1915 (12 comments by @ben-clayton, @dneto0, @kdashg, @kvark)
-
Add convenience type aliases [fiub]vecN and fmatNxM. gpuweb/gpuweb#1976 (11 comments by @ben-clayton, @dneto0, @kainino0x, @kdashg, @kvark)
-
[wgsl]: Add `fmod()`, `frem()` builtins gpuweb/gpuweb#1913 (11 comments by @ben-clayton, @dneto0, @kvark)
-
Fix module-scope constant declarations wrt type inference gpuweb/gpuweb#1944 (10 comments by @ben-clayton, @dneto0, @jimblandy, @kainino0x, @kvark)
-
wgsl: add refract as a built-in function gpuweb/gpuweb#1901 (10 comments by @jiangzhaoming, @Kangz, @kvark)
-
Rename SHADER_READ back to SAMPLED gpuweb/gpuweb#1963 (9 comments by @kainino0x, @Kangz, @kvark, @toji)
-
wgsl: Change `modf` and `frexp` to return a structure gpuweb/gpuweb#1973 (8 comments by @ben-clayton, @dneto0, @litherum, @RobinMorisset)
-
Define index/vertex buffer OOB behavior gpuweb/gpuweb#1835 (8 comments by @kainino0x, @Kangz, @kvark)
-
Remove the read-only texture storage bindings gpuweb/gpuweb#1794 (8 comments by @kainino0x, @Kangz, @kvark)
-
Clarify depth/stencil copy validation gpuweb/gpuweb#1932 (8 comments by @austinEng, @kainino0x, @kvark, @toji)
-
Clarify out of bounds texture operations gpuweb/gpuweb#1906 (8 comments by @alan-baker, @ben-clayton, @jimblandy, @Kangz, @kdashg)
-
Rename compute size limits for clarity gpuweb/gpuweb#1991 (7 comments by @kainino0x, @Kangz, @kvark)
-
Stencil reads are S,X,X,X. (but expect S,S,S,S or S,0,0,1) gpuweb/gpuweb#1978 (6 comments by @kainino0x, @kdashg, @kvark)
-
Fix GPUVertexAttribute.offset alignment requirements for D3D12 gpuweb/gpuweb#1980 (5 comments by @kainino0x, @Kangz, @kvark)
-
Loosen pipeline fragment output validation gpuweb/gpuweb#1918 (5 comments by @austinEng, @kainino0x, @kvark)
-
wgsl: specify the input type of reflect to be vector gpuweb/gpuweb#1912 (5 comments by @ben-clayton, @jiangzhaoming)
-
Define vertex data conversions and validation gpuweb/gpuweb#1908 (5 comments by @alan-baker, @Kangz, @kvark)
-
wgsl: Add limits section, and "spurious" failure gpuweb/gpuweb#1997 (4 comments by @alan-baker, @dneto0)
-
Refactor the pipeline-pass compatibility and check RODS gpuweb/gpuweb#1979 (4 comments by @kainino0x, @Kangz, @kvark)
-
wgsl: Change `operator%(f32,32)` to use `OpFRem` gpuweb/gpuweb#1945 (4 comments by @ben-clayton, @dneto0, @kdashg)
-
Require copyExternalImageToTexture canvases to be 2d/webgl/webgl2 gpuweb/gpuweb#1955 (4 comments by @kainino0x, @shaoboyan)
-
[wgsl]: Remove scalar overloads of `faceForward` and `reflect` gpuweb/gpuweb#1914 (2 comments by @ben-clayton, @kvark)
-
copyExternalImageToTexture: add validation, make validation async gpuweb/gpuweb#1737 (1 comment by @kainino0x)
-
Use 64bit limits for max uniform/storage buffer binding sizes. gpuweb/gpuweb#1941 (1 comment by @kdashg)
-
TR publications
-
39 documents published.
-
WebGPU published 20 times as a Working Draft. History of WebGPU
-
WebGPU Shading Language published 19 times as a Working Draft. History of WebGPU Shading Language
-
Participants
-
3 participants joined.
-
Ming Yu (ByteDance) joined.
-
Jerran Schmidt (Autodesk) joined.
-
Henrik Edstrom (Autodesk) joined.
-
-
2 organizations joined.
-
ByteDance joined.
-
Autodesk joined.
-
GitHub
gpuweb/cts
-
1 issue opened.
-
builtin(vertex_index) should be u32 not i32 gpuweb/cts#597, by @zoddicus.
-
-
2 issues closed.
-
builtin(vertex_index) should be u32 not i32 gpuweb/cts#597.
-
Test that the alpha:true/false of canvas is honored gpuweb/cts#460.
-
-
3 issues received 3 new comments.
-
42 pull requests opened.
-
Fix non-filtering sampler + filtering sampler binding validation test gpuweb/cts#611, by @austinEng.
-
Test that buffer writes are preserved after another range is mapped. gpuweb/cts#610, by @toji.
-
Test that array values are preserved across multiple WRITE mappings gpuweb/cts#609, by @toji.
-
wgsl: add workgroup_size to all stage(compute) gpuweb/cts#608, by @Kangz.
-
wgsl: [[set(n)]] is now [[group(n)]] gpuweb/cts#607, by @ben-clayton.
-
Fix byte vs element units bug; allow buffer checking by mapping gpuweb/cts#605, by @kainino0x.
-
Review api,operation,buffers,map_detach:* gpuweb/cts#604, by @toji.
-
api,validation,encoding,cmds,render,other plans gpuweb/cts#603, by @haoxli.
-
vertex_state correctness: test more attrib.offset gpuweb/cts#602, by @Kangz.
-
Add operation tests about buffer zero initialization - Part VI gpuweb/cts#601, by @Jiawei-Shao.
-
CopyToTexture,canvas basic tests : Add premultiplied alpha tests gpuweb/cts#600, by @shaoboyan.
-
Add operation tests about buffer zero initialization - Part V (#598) gpuweb/cts#599, by @Jiawei-Shao.
-
Add operation tests about buffer zero initialization - Part IV gpuweb/cts#598, by @Jiawei-Shao.
-
wgsl: Fix datatype of [[builtin(vertex_index)]] gpuweb/cts#596, by @ben-clayton.
-
Add operation tests about buffer zero initialization - Part III gpuweb/cts#595, by @Jiawei-Shao.
-
Trim redundant message from stack (for Chrome), clarify stack hiding gpuweb/cts#594, by @kainino0x.
-
add ErrorWithExtra for extra debug data gpuweb/cts#593, by @kainino0x.
-
Refactor ArrayBuffer/GPUBuffer content checking helpers gpuweb/cts#592, by @kainino0x.
-
A bit more comments for map.spec.ts for range/mapRegion gpuweb/cts#591, by @shrekshao.
-
Update nonGuaranteed->required gpuweb/cts#590, by @kainino0x.
-
Review and expand api,validation,buffer,destroy:* gpuweb/cts#589, by @toji.
-
Roll @webgpu/types to 0.1.4 gpuweb/cts#588, by @toji.
-
eslint tweaks gpuweb/cts#587, by @kainino0x.
-
Add operation tests about buffer zero initialization - Part II gpuweb/cts#586, by @Jiawei-Shao.
-
Validation tests for queue.writeBuffer gpuweb/cts#585, by @haoxli.
-
Modify plan to test the setVertex/IndexBuffer offset alignment gpuweb/cts#584, by @Kangz.
-
Add vertex_state operation tests for all GPUVertexState params gpuweb/cts#583, by @Kangz.
-
Add operation tests about buffer zero initialization - Part I gpuweb/cts#582, by @Jiawei-Shao.
-
CopyToTexture,canvas basic tests gpuweb/cts#581, by @shaoboyan.
-
wgsl: Fixes for new validation failures gpuweb/cts#580, by @ben-clayton.
-
createRenderPipelineAsync validation gpuweb/cts#579, by @jiangzhaoming.
-
Document capability_info, refactor texture info slightly gpuweb/cts#578, by @kainino0x.
-
Fix unintended compilation and validation failures gpuweb/cts#577, by @ben-clayton.
-
wgsl: Remove access control decorations gpuweb/cts#576, by @ben-clayton.
-
Remove expected error codes from WGSL validation tests gpuweb/cts#575, by @jrprice.
-
Document helpers gpuweb/cts#574, by @kainino0x.
-
Migrate helper documentation to tsdoc gpuweb/cts#573, by @kainino0x.
-
Update texture zero init tests for 3D textures gpuweb/cts#572, by @austinEng.
-
Add test plan for the operation tests about stencil aspect gpuweb/cts#571, by @Jiawei-Shao.
-
Update the maxStorageBuffer/Texture limits to 8. gpuweb/cts#570, by @Kangz.
-
Add validation for creating compute pipeline gpuweb/cts#569, by @jiangzhaoming.
-
-
41 pull requests merged.
-
Fix byte vs element units bug; allow buffer checking by mapping gpuweb/cts#605.
-
Review api,operation,buffers,map_detach:* gpuweb/cts#604.
-
wgsl: add workgroup_size to all stage(compute) gpuweb/cts#608.
-
wgsl: [[set(n)]] is now [[group(n)]] gpuweb/cts#607.
-
CopyToTexture,canvas basic tests : Add premultiplied alpha tests gpuweb/cts#600.
-
vertex_state correctness: test more attrib.offset gpuweb/cts#602.
-
Add operation tests about buffer zero initialization - Part VI gpuweb/cts#601.
-
Validation tests for queue.writeBuffer gpuweb/cts#585.
-
Add operation tests about buffer zero initialization - Part IV gpuweb/cts#598.
-
Refactor ArrayBuffer/GPUBuffer content checking helpers gpuweb/cts#592.
-
Add operation tests about buffer zero initialization - Part III gpuweb/cts#595.
-
add ErrorWithExtra for extra debug data gpuweb/cts#593.
-
Review and expand api,validation,buffer,destroy:* gpuweb/cts#589.
-
wgsl: Fix datatype of [[builtin(vertex_index)]] gpuweb/cts#596.
-
A bit more comments for map.spec.ts for range/mapRegion gpuweb/cts#591.
-
Update nonGuaranteed->required gpuweb/cts#590.
-
Roll @webgpu/types to 0.1.4 gpuweb/cts#588.
-
Add operation tests about buffer zero initialization - Part II gpuweb/cts#586.
-
eslint tweaks gpuweb/cts#587.
-
CopyToTexture,canvas basic tests gpuweb/cts#581.
-
Add operation tests about buffer zero initialization - Part I gpuweb/cts#582.
-
Modify plan to test the setVertex/IndexBuffer offset alignment gpuweb/cts#584.
-
Add vertex_state operation tests for all GPUVertexState params gpuweb/cts#583.
-
createRenderPipelineAsync validation gpuweb/cts#579.
-
wgsl: Fixes for new validation failures gpuweb/cts#580.
-
Document capability_info, refactor texture info slightly gpuweb/cts#578.
-
Document helpers gpuweb/cts#574.
-
Fix unintended compilation and validation failures gpuweb/cts#577.
-
wgsl: Remove access control decorations gpuweb/cts#576.
-
Remove expected error codes from WGSL validation tests gpuweb/cts#575.
-
Migrate helper documentation to tsdoc gpuweb/cts#573.
-
Update texture zero init tests for 3D textures gpuweb/cts#572.
-
Add test plan for the operation tests about stencil aspect gpuweb/cts#571.
-
Add {Case,Subcase}ParamsBuilder to replace .cases().subcases() gpuweb/cts#566.
-
Check features support when requesting device gpuweb/cts#563.
-
Update the maxStorageBuffer/Texture limits to 8. gpuweb/cts#570.
-
Add validation for creating compute pipeline gpuweb/cts#569.
-
Update validation tests of copyImageBitmapToTexture gpuweb/cts#568.
-
1 pull request closed.
-
Update texture zero init tests for 3D textures gpuweb/cts#557.
-
-
27 pull requests received 197 new comments.
-
CopyToTexture,canvas basic tests : Add premultiplied alpha tests gpuweb/cts#600 (24 comments by @kainino0x, @shaoboyan)
-
Add vertex_state operation tests for all GPUVertexState params gpuweb/cts#583 (19 comments by @kainino0x, @Kangz)
-
Refactor ArrayBuffer/GPUBuffer content checking helpers gpuweb/cts#592 (16 comments by @kainino0x, @Kangz)
-
api,validation,encoding,cmds,render,other plans gpuweb/cts#603 (11 comments by @haoxli, @Jiawei-Shao, @kainino0x)
-
Add operation tests about buffer zero initialization - Part I gpuweb/cts#582 (11 comments by @Jiawei-Shao, @Kangz)
-
Add {Case,Subcase}ParamsBuilder to replace .cases().subcases() gpuweb/cts#566 (10 comments by @austinEng, @kainino0x)
-
Add operation tests on writeTexture() and B2T copies with stencil aspects gpuweb/cts#606 (7 comments by @kainino0x, @Kangz)
-
Fix unintended compilation and validation failures gpuweb/cts#577 (6 comments by @ben-clayton, @kainino0x, @Kangz)
-
Add operation tests about buffer zero initialization - Part III gpuweb/cts#595 (5 comments by @Jiawei-Shao, @Kangz)
-
Document capability_info, refactor texture info slightly gpuweb/cts#578 (5 comments by @kainino0x, @shrekshao)
-
Add operation tests about buffer zero initialization - Part VI gpuweb/cts#601 (3 comments by @Jiawei-Shao, @Kangz)
-
Add operation tests about buffer zero initialization - Part II gpuweb/cts#586 (3 comments by @Jiawei-Shao, @Kangz)
-
Update validation tests of copyImageBitmapToTexture gpuweb/cts#568 (2 comments by @kainino0x, @shaoboyan)
-
Add operation tests about buffer zero initialization - Part V (#598) gpuweb/cts#599 (1 comment by @kainino0x)
-
gpuweb/gpuweb
-
44 issues opened.
-
[wgsl] functions with return values + return + control flow gpuweb/gpuweb#1897, by @magcius.
-
OOB behaviour for storage textures gpuweb/gpuweb#1895, by @ben-clayton.
-
Add query for memory footprint of unsized depth/stencil formats gpuweb/gpuweb#1887, by @kainino0x.
-
Make error scope objects have same shape as DOMExceptions? gpuweb/gpuweb#1884, by @juj.
-
API should be exposed to DedicatedWorker gpuweb/gpuweb#1882, by @Kaiido.
-
Operations happening in GPUVertexState are underspecified gpuweb/gpuweb#1880, by @Kangz.
-
Remove the requirement for Vulkan's drawIndirectFirstInstance feature. gpuweb/gpuweb#1878, by @Kangz.
-
wgsl editorial: tighten IO-shareable gpuweb/gpuweb#1877, by @dneto0.
-
samples need updating gpuweb/gpuweb#1873, by @jack1232.
-
Add opaque region or isOpaque hint gpuweb/gpuweb#1871, by @rmader.
-
Structs vs out params gpuweb/gpuweb#1870, by @litherum.
-
Out of bounds behaviour and atomics gpuweb/gpuweb#1868, by @alan-baker.
-
Rules about "break" inside "continuing" are strange gpuweb/gpuweb#1867, by @kvark.
-
Some limits don't need to be configurable gpuweb/gpuweb#1865, by @kainino0x.
-
wgsl: Consider reserving `vec` and `mat` gpuweb/gpuweb#1862, by @ben-clayton.
-
wgsl: can't form type-constructor or zero-value for array or struct containing an atomic gpuweb/gpuweb#1859, by @dneto0.
-
wgsl: allow workgroup size dimensions to be unsigned as well as signed gpuweb/gpuweb#1857, by @dneto0.
-
Spec doesn't compile gpuweb/gpuweb#1852, by @Kangz.
-
wgsl: fix contradiction between function pointer argument rules and builtin functions gpuweb/gpuweb#1850, by @dneto0.
-
No alignment constraints for setIndex/VertexBuffer offsets. gpuweb/gpuweb#1849, by @Kangz.
-
Does compositingAlphaMode affect using the canvas as an image source? gpuweb/gpuweb#1847, by @kainino0x.
-
wgsl: Restrict `modf` and `frexp`pointer parameter storage classes gpuweb/gpuweb#1846, by @ben-clayton.
-
Should the alignment of destinationOffset in resolveQuerySet be 256 on Metal? gpuweb/gpuweb#1845, by @haoxli.
-
wgsl: limiting undefined behaviour; subdividing "dynamic error" gpuweb/gpuweb#1844, by @dneto0.
-
Should the API enforce that writable storage buffer bindings don't alias each other? gpuweb/gpuweb#1842, by @Kangz.
-
[wgsl]: `atomic` missing from keywords and grammar gpuweb/gpuweb#1840, by @ben-clayton.
-
Add extent/origin/color "reification" step gpuweb/gpuweb#1838, by @kainino0x.
-
Storage buffer access: compatibility with the pipeline layout gpuweb/gpuweb#1837, by @kvark.
-
Figure out whether indirectBuffer OOB is a validation error gpuweb/gpuweb#1834, by @kainino0x.
-
We should forbid pointers to handles gpuweb/gpuweb#1832, by @alan-baker.
-
Web IDL hiccup in GPUPresentationContext: Dictionary arguments cannot be nullable gpuweb/gpuweb#1822, by @tidoust.
-
Questions about copyExternalImageToTexture gpuweb/gpuweb#1821, by @Jiawei-Shao.
-
wgsl: frexp: what happens if the exponent is negative but the second operand is ptr-to-u32 gpuweb/gpuweb#1820, by @dneto0.
-
WGSL code containing \0 gpuweb/gpuweb#1816, by @kainino0x.
-
WGSL: Standardized error codes gpuweb/gpuweb#1813, by @ben-clayton.
-
Implicit BGL conflicts with de-duplication gpuweb/gpuweb#1806, by @kvark.
-
WGSL grammar needs update for var type inference gpuweb/gpuweb#1805, by @jrprice.
-
Error with literal indexing in a shader gpuweb/gpuweb#1803, by @kvark.
-
wgsl: derivative functions should only be usable in the fragment shader stage gpuweb/gpuweb#1795, by @dneto0.
-
Uniformity annotations for global variables gpuweb/gpuweb#1791, by @RobinMorisset.
-
Data races and security gpuweb/gpuweb#1788, by @alan-baker.
-
-
31 issues closed.
-
Need to define alpha premultiplication for external images gpuweb/gpuweb#1762.
-
Document vertex expansion rules gpuweb/gpuweb#861.
-
Structs vs out params gpuweb/gpuweb#1870.
-
wgsl: allow workgroup size dimensions to be unsigned as well as signed gpuweb/gpuweb#1857.
-
Figure out whether indirectBuffer OOB is a validation error gpuweb/gpuweb#1834.
-
No alignment constraints for setIndex/VertexBuffer offsets. gpuweb/gpuweb#1849.
-
Spec doesn't compile gpuweb/gpuweb#1852.
-
Define behavior of reading from WebGPU canvases using other Web APIs gpuweb/gpuweb#1781.
-
[wgsl]: `atomic` missing from keywords and grammar gpuweb/gpuweb#1840.
-
Missing mapping for OpFRem, and MODULO token missing gpuweb/gpuweb#702.
-
Questions about copyExternalImageToTexture gpuweb/gpuweb#1821.
-
WGSL code containing \0 gpuweb/gpuweb#1816.
-
Need a proposal for specialization constants on the API side. gpuweb/gpuweb#1608.
-
Consider merging GPUSwapChain into GPUCanvasContext gpuweb/gpuweb#1780.
-
Reconsider the name of storeOp: "clear" gpuweb/gpuweb#1761.
-
What binding numbers are used by an external texture binding? gpuweb/gpuweb#1721.
-
Describe Buffer/Texture destroy() semantics gpuweb/gpuweb#472.
-
How to add default offsets to setVertexBuffers? gpuweb/gpuweb#421.
-
getBindGroupLayout gpuweb/gpuweb#446.
-
wgsl: derivative functions should only be usable in the fragment shader stage gpuweb/gpuweb#1795.
-
What happens in configureSwapChain and on canvas resize? gpuweb/gpuweb#1691.
-
Determine how errors during createReady*Pipeline are exposed gpuweb/gpuweb#1049.
-
wgsl: compute shaders should have a workgroup_size attribute gpuweb/gpuweb#1777.
-
Validation rules for WGSL modules: binding collisions gpuweb/gpuweb#889.
-
-
63 issues received 315 new comments.
-
2nd, 3rd, and 4th components of sampled depth/stencil textures are undefined in Metal gpuweb/gpuweb#1198 (31 comments by @austinEng, @dneto0, @kainino0x, @Kangz, @kdashg, @kvark, @litherum)
-
Investigation: Compressed Texture Formats gpuweb/gpuweb#144 (25 comments by @alecazam, @fintelia, @kainino0x, @Kangz, @kdashg, @magcius)
-
Buffer indices should be unsigned gpuweb/gpuweb#1135 (19 comments by @dneto0, @Kangz, @kdashg, @kvark, @litherum, @RafaelCintron, @tex3d)
-
Implicit BGL conflicts with de-duplication gpuweb/gpuweb#1806 (14 comments by @kainino0x, @Kangz, @kvark, @magcius)
-
signed integer division: unspecified rounding mode for the quotient; and overflow gpuweb/gpuweb#1774 (13 comments by @dneto0, @kdashg, @kvark)
-
Error with literal indexing in a shader gpuweb/gpuweb#1803 (11 comments by @dneto0, @kdashg, @kvark)
-
Define WebGPURenderingContext contextType DomString gpuweb/gpuweb#131 (10 comments by @domenic, @Kaiido, @kainino0x, @metaeaux, @toji)
-
Web IDL hiccup in GPUPresentationContext: Dictionary arguments cannot be nullable gpuweb/gpuweb#1822 (10 comments by @kainino0x, @saschanaz, @tidoust, @toji)
-
Define behavior of reading from WebGPU canvases using other Web APIs gpuweb/gpuweb#1781 (9 comments by @kainino0x, @Kangz, @kvark, @RafaelCintron)
-
Consider merging GPUSwapChain into GPUCanvasContext gpuweb/gpuweb#1780 (9 comments by @kainino0x, @Kangz, @kvark, @toji)
-
Spec should use [=map/exists=] for optional dictionary members gpuweb/gpuweb#1512 (9 comments by @kainino0x, @toji)
-
why postMessage GPUDevice to worker cause DataCloneError when cross-origin isolated capability is true? gpuweb/gpuweb#1817 (8 comments by @Kangz, @yyc-git)
-
Add opaque region or isOpaque hint gpuweb/gpuweb#1871 (7 comments by @jrmuizel, @kainino0x, @Kangz, @kvark, @magcius, @rmader)
-
WGSL requires dynamic matrix indexing, which requires surprising SPIR-V gpuweb/gpuweb#1782 (7 comments by @ben-clayton, @dneto0, @kdashg, @krogovin, @kvark)
-
Should the API enforce that writable storage buffer bindings don't alias each other? gpuweb/gpuweb#1842 (6 comments by @Kangz, @kdashg, @kvark, @litherum)
-
can pipeline creation fail in the driver compiler? particularly due to shader compilation failure gpuweb/gpuweb#1872 (6 comments by @alan-baker, @dneto0, @Kangz, @kvark)
-
Rules about "break" inside "continuing" are strange gpuweb/gpuweb#1867 (6 comments by @dneto0, @jimblandy, @kvark)
-
[wgsl] Proposal: Remove pointer out parameters from modf, frexp gpuweb/gpuweb#1480 (6 comments by @ben-clayton, @dneto0, @kvark, @litherum)
-
Assignment operators (e.g. `+=`) in wgsl gpuweb/gpuweb#1518 (4 comments by @dneto0, @kainino0x, @kdashg)
-
Storage buffer access: compatibility with the pipeline layout gpuweb/gpuweb#1837 (4 comments by @Kangz, @kvark)
-
Does compositingAlphaMode affect using the canvas as an image source? gpuweb/gpuweb#1847 (3 comments by @kainino0x)
-
Figure out whether indirectBuffer OOB is a validation error gpuweb/gpuweb#1834 (3 comments by @kainino0x, @Kangz)
-
Investigation: Import VideoFrame from WebCodec to WebGPU gpuweb/gpuweb#1380 (3 comments by @kainino0x, @Kangz, @kvark)
-
Questions about copyExternalImageToTexture gpuweb/gpuweb#1821 (3 comments by @Jiawei-Shao, @kainino0x, @Kangz)
-
Support for gl_PrimitiveID / SV_PrimitiveID -like construct? gpuweb/gpuweb#1786 (3 comments by @darzu, @martinken)
-
wgsl: compute shaders should have a workgroup_size attribute gpuweb/gpuweb#1777 (3 comments by @dneto0, @Kangz, @kdashg)
-
Make error scope objects have same shape as DOMExceptions? gpuweb/gpuweb#1884 (2 comments by @juj, @kainino0x)
-
Operations happening in GPUVertexState are underspecified gpuweb/gpuweb#1880 (2 comments by @kainino0x, @Kangz)
-
Remove the requirement for Vulkan's drawIndirectFirstInstance feature. gpuweb/gpuweb#1878 (2 comments by @kainino0x, @Kangz)
-
WebGPU should validate that storageTexture bindings have a single mip level. gpuweb/gpuweb#1576 (2 comments by @ben-clayton, @Kangz)
-
wgsl: fix contradiction between function pointer argument rules and builtin functions gpuweb/gpuweb#1850 (2 comments by @alan-baker, @ben-clayton)
-
In a single shader instance, can one render target or resource alias the storage of another? gpuweb/gpuweb#287 (2 comments by @Kangz)
-
Requiring integral-valued float literals to have .0 seems gratuitous gpuweb/gpuweb#739 (1 comment by @kdashg)
-
Add query for memory footprint of unsized depth/stencil formats gpuweb/gpuweb#1887 (1 comment by @kainino0x)
-
unsized arrays have a zero value, but you shouldn't be able to write it down in WGSL gpuweb/gpuweb#981 (1 comment by @dneto0)
-
wgsl: allow workgroup size dimensions to be unsigned as well as signed gpuweb/gpuweb#1857 (1 comment by @dneto0)
-
The semicolons are an attempt to better separate the conditions. These complex rule sets are visually crammed in a narrow box. gpuweb/gpuweb#1856 (1 comment by @kainino0x)
-
wgsl: frexp: what happens if the exponent is negative but the second operand is ptr-to-u32 gpuweb/gpuweb#1820 (1 comment by @dneto0)
-
What binding numbers are used by an external texture binding? gpuweb/gpuweb#1721 (1 comment by @kainino0x)
-
-
60 pull requests opened.
-
wgsl: Reserve `vec` and `mat` keywords gpuweb/gpuweb#1896, by @ben-clayton.
-
wgsl: Fix typo on description of structure type gpuweb/gpuweb#1889, by @dneto0.
-
Update size range of "stencil8" format gpuweb/gpuweb#1888, by @kainino0x.
-
Match introduction with title gpuweb/gpuweb#1886, by @mehmetoguzderin.
-
Remove sentinel value in setVertex/IndexBuffer gpuweb/gpuweb#1885, by @kainino0x.
-
Expose all interfaces to DedicatedWorker gpuweb/gpuweb#1883, by @kainino0x.
-
wgsl: clarify interpolation defaulting gpuweb/gpuweb#1881, by @dneto0.
-
Reduce maxStorageTexturePerStage from 8 to 4. gpuweb/gpuweb#1879, by @Kangz.
-
wgsl: define 'constructible' types, and apply it gpuweb/gpuweb#1876, by @dneto0.
-
wgsl: cleanup old text, discuss conversions/constructions gpuweb/gpuweb#1875, by @dneto0.
-
fix << operator: first operand can be unsigned gpuweb/gpuweb#1874, by @dneto0.
-
Fix italics in explainer gpuweb/gpuweb#1869, by @surma.
-
Add OOB checks to draw and drawIndexed gpuweb/gpuweb#1866, by @toji.
-
Change RO texture usage to be restricted gpuweb/gpuweb#1864, by @kvark.
-
Add a bunch of limits. gpuweb/gpuweb#1863, by @Kangz.
-
wgsl: Fix channel transfer functions for `8uint` and `8sint`. gpuweb/gpuweb#1861, by @jimblandy.
-
wgsl: can't type-construct or zero-value-construct aggregates containing atomics gpuweb/gpuweb#1860, by @dneto0.
-
wgsl: allow unsigned workgroup_size components gpuweb/gpuweb#1858, by @dneto0.
-
wgsl: refine pointer function call restrictions gpuweb/gpuweb#1855, by @dneto0.
-
Alignment for index and vertex buffers gpuweb/gpuweb#1854, by @kvark.
-
Add setVertex/IndexBuffer offset alignment rules. gpuweb/gpuweb#1853, by @Kangz.
-
wgsl: explicitly restrict storage classes for frexp, modf gpuweb/gpuweb#1851, by @dneto0.
-
Resolve missing color space and alpha issues gpuweb/gpuweb#1848, by @kainino0x.
-
Add unconfigure() to GPUPresentationContext gpuweb/gpuweb#1843, by @toji.
-
Add atomic type to grammar and keywords gpuweb/gpuweb#1841, by @alan-baker.
-
Define WebGPU canvas read behavior gpuweb/gpuweb#1839, by @kainino0x.
-
Expand allowable type for ignore() gpuweb/gpuweb#1836, by @alan-baker.
-
Define index/vertex buffer OOB behavior gpuweb/gpuweb#1835, by @kainino0x.
-
Add note that sized and runtime-sized arrays are distinct types gpuweb/gpuweb#1833, by @dneto0.
-
Describe integer negation corner case gpuweb/gpuweb#1831, by @alan-baker.
-
wgsl: detailed semantics of integer division and remainder gpuweb/gpuweb#1830, by @dneto0.
-
Split adapter/device section into parts gpuweb/gpuweb#1829, by @kainino0x.
-
wgsl: type rule for parenthesized expr gpuweb/gpuweb#1828, by @dneto0.
-
wgsl: clarify and fix wording for function calls gpuweb/gpuweb#1827, by @dneto0.
-
wgsl: fix typo in var type inference example gpuweb/gpuweb#1826, by @dneto0.
-
wgsl: remove unsigned case for frexp 2nd arg gpuweb/gpuweb#1823, by @dneto0.
-
wgsl: Add notes reinforcing where plain types go gpuweb/gpuweb#1819, by @dneto0.
-
wgsl: fix front_facing type in example gpuweb/gpuweb#1818, by @dneto0.
-
Rename GPUStoreOp "clear" to "discard" gpuweb/gpuweb#1815, by @kainino0x.
-
wgsl: Remove level from 1D texture load intrinsic gpuweb/gpuweb#1814, by @jrprice.
-
Make use of [=map/exists=] where appropriate gpuweb/gpuweb#1812, by @toji.
-
setBindGroup: make start+length check synchronous gpuweb/gpuweb#1811, by @kainino0x.
-
Combine CanvasContext and SwapChain into a single interface gpuweb/gpuweb#1810, by @toji.
-
wgsl: explain internal layout of atomic type gpuweb/gpuweb#1809, by @dneto0.
-
wgsl: briefly explain storable and IO-shareable gpuweb/gpuweb#1808, by @dneto0.
-
wgsl: fix grammar: access mode is optional gpuweb/gpuweb#1807, by @dneto0.
-
Rename misleading nonGuaranteed* names gpuweb/gpuweb#1804, by @kainino0x.
-
Increase maxStorageBuffer/TexturePerShaderStage to 8. gpuweb/gpuweb#1802, by @Kangz.
-
wgsl: Limit dynamic indexes to references to matrices and arrays. gpuweb/gpuweb#1801, by @jimblandy.
-
wgsl: Fix formatting of assignment rule. gpuweb/gpuweb#1800, by @jimblandy.
-
Fix typos in comparison tables gpuweb/gpuweb#1799, by @dneto0.
-
wgsl: assignment only works on atomic-free plain type gpuweb/gpuweb#1798, by @dneto0.
-
wgsl: textureSampleLevel on texture_external samples level 0 gpuweb/gpuweb#1797, by @dneto0.
-
wgsl: derivatives can only be invoked in fragment stage gpuweb/gpuweb#1796, by @dneto0.
-
Remove the read-only texture storage bindings gpuweb/gpuweb#1794, by @kvark.
-
Add texture_depth_multisampled_2d gpuweb/gpuweb#1793, by @kvark.
-
wgsl: array size may be module-scope constant (possibly overridable) gpuweb/gpuweb#1792, by @dneto0.
-
wgsl: workgroup_size must be applied to compute entry point gpuweb/gpuweb#1790, by @dneto0.
-
wgsl: fix description of correctly rounded gpuweb/gpuweb#1789, by @dneto0.
-
create_PipelineAsync rejects when failure gpuweb/gpuweb#1787, by @jiangzhaoming.
-
-
52 pull requests merged.
-
wgsl: Fix typo on description of structure type gpuweb/gpuweb#1889.
-
Remove sentinel value in setVertex/IndexBuffer gpuweb/gpuweb#1885.
-
Resolve missing color space and alpha issues gpuweb/gpuweb#1848.
-
Add a bunch of limits. gpuweb/gpuweb#1863.
-
wgsl: clarify interpolation defaulting gpuweb/gpuweb#1881.
-
wgsl: cleanup old text, discuss conversions/constructions gpuweb/gpuweb#1875.
-
wgsl: define 'constructible' types, and apply it gpuweb/gpuweb#1876.
-
fix << operator: first operand can be unsigned gpuweb/gpuweb#1874.
-
wgsl: allow unsigned workgroup_size components gpuweb/gpuweb#1858.
-
Add OOB checks to draw and drawIndexed gpuweb/gpuweb#1866.
-
Fix italics in explainer gpuweb/gpuweb#1869.
-
wgsl: Fix channel transfer functions for `8uint` and `8sint`. gpuweb/gpuweb#1861.
-
wgsl: briefly explain storable and IO-shareable gpuweb/gpuweb#1808.
-
Add setVertex/IndexBuffer offset alignment rules. gpuweb/gpuweb#1853.
-
Define WebGPU canvas read behavior gpuweb/gpuweb#1839.
-
Expand allowable type for ignore() gpuweb/gpuweb#1836.
-
Add unconfigure() to GPUPresentationContext gpuweb/gpuweb#1843.
-
wgsl: remove unsigned case for frexp 2nd arg gpuweb/gpuweb#1823.
-
copyExternalImageToTexture: add colorSpace and premultipliedAlpha gpuweb/gpuweb#1768.
-
Add atomic type to grammar and keywords gpuweb/gpuweb#1841.
-
Rename misleading nonGuaranteed* names gpuweb/gpuweb#1804.
-
Describe integer negation corner case gpuweb/gpuweb#1831.
-
Add note that sized and runtime-sized arrays are distinct types gpuweb/gpuweb#1833.
-
wgsl: clarify and fix wording for function calls gpuweb/gpuweb#1827.
-
wgsl: explain internal layout of atomic type gpuweb/gpuweb#1809.
-
Clarify out of bounds behavior gpuweb/gpuweb#1722.
-
Split adapter/device section into parts gpuweb/gpuweb#1829.
-
wgsl: type rule for parenthesized expr gpuweb/gpuweb#1828.
-
wgsl: fix typo in var type inference example gpuweb/gpuweb#1826.
-
Add API surface for pipeline-overridable constants gpuweb/gpuweb#1766.
-
Combine CanvasContext and SwapChain into a single interface gpuweb/gpuweb#1810.
-
wgsl: Add notes reinforcing where plain types go gpuweb/gpuweb#1819.
-
Remove workgroup_size builtin gpuweb/gpuweb#1771.
-
wgsl: textureSampleLevel on texture_external samples level 0 gpuweb/gpuweb#1797.
-
wgsl: fix front_facing type in example gpuweb/gpuweb#1818.
-
Rename GPUStoreOp "clear" to "discard" gpuweb/gpuweb#1815.
-
setBindGroup: make start+length check synchronous gpuweb/gpuweb#1811.
-
Add forceSoftware adapter option and isSoftware adapter attribute gpuweb/gpuweb#1734.
-
wgsl: fix grammar: access mode is optional gpuweb/gpuweb#1807.
-
wgsl: Limit dynamic indexes to references to matrices and arrays. gpuweb/gpuweb#1801.
-
Increase maxStorageBuffer/TexturePerShaderStage to 8. gpuweb/gpuweb#1802.
-
wgsl: derivatives can only be invoked in fragment stage gpuweb/gpuweb#1796.
-
wgsl: assignment only works on atomic-free plain type gpuweb/gpuweb#1798.
-
wgsl: Fix formatting of assignment rule. gpuweb/gpuweb#1800.
-
Fix typos in comparison tables gpuweb/gpuweb#1799.
-
Document new resizing logic for GPUSwapChain gpuweb/gpuweb#1738.
-
wgsl: define texture type parameterization gpuweb/gpuweb#1776.
-
create_PipelineAsync rejects when failure gpuweb/gpuweb#1787.
-
wgsl: workgroup_size must be applied to compute entry point gpuweb/gpuweb#1790.
-
wgsl: fix description of correctly rounded gpuweb/gpuweb#1789.
-
Remove colliding [[device]] slot from GPUDevice gpuweb/gpuweb#1758.
-
Update GPUAdapterLimits to GPUSupportedLimits gpuweb/gpuweb#1765.
-
-
6 pull requests closed.
-
wgsl: can't type-construct or zero-value-construct aggregates containing atomics gpuweb/gpuweb#1860.
-
Alignment for index and vertex buffers gpuweb/gpuweb#1854.
-
Fix a bug in GPUExtent3D gpuweb/gpuweb#1712.
-
Remove % for signed integral gpuweb/gpuweb#1763.
-
Add swapchain size to the configuration gpuweb/gpuweb#1779.
-
Add GPURequestAdapterOptions.allowSoftware gpuweb/gpuweb#1634.
-
-
54 pull requests received 276 new comments.
-
Add a bunch of limits. gpuweb/gpuweb#1863 (33 comments by @jrprice, @kainino0x, @Kangz, @kdashg, @kvark, @toji)
-
Combine CanvasContext and SwapChain into a single interface gpuweb/gpuweb#1810 (17 comments by @kainino0x, @kvark, @toji)
-
wgsl: Remove level from 1D texture load intrinsic gpuweb/gpuweb#1814 (16 comments by @dneto0, @jrprice, @kdashg, @kvark)
-
Add OOB checks to draw and drawIndexed gpuweb/gpuweb#1866 (16 comments by @kainino0x, @Kangz, @kvark, @toji)
-
Make use of [=map/exists=] where appropriate gpuweb/gpuweb#1812 (12 comments by @kainino0x, @kvark, @toji)
-
Remove the read-only texture storage bindings gpuweb/gpuweb#1794 (10 comments by @kainino0x, @Kangz, @kvark)
-
wgsl: can't type-construct or zero-value-construct aggregates containing atomics gpuweb/gpuweb#1860 (9 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
wgsl: explicitly restrict storage classes for frexp, modf gpuweb/gpuweb#1851 (9 comments by @dneto0, @kdashg, @kvark, @litherum)
-
Rename misleading nonGuaranteed* names gpuweb/gpuweb#1804 (8 comments by @kainino0x, @Kangz, @kvark)
-
Add API surface for pipeline-overridable constants gpuweb/gpuweb#1766 (8 comments by @kainino0x, @kvark)
-
Define WebGPU canvas read behavior gpuweb/gpuweb#1839 (7 comments by @Juanmihd, @kainino0x, @sunnyps, @toji)
-
wgsl: detailed semantics of integer division and remainder gpuweb/gpuweb#1830 (6 comments by @dneto0, @kdashg, @kvark, @litherum)
-
wgsl: refine pointer function call restrictions gpuweb/gpuweb#1855 (6 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
wgsl: Limit dynamic indexes to references to matrices and arrays. gpuweb/gpuweb#1801 (6 comments by @dneto0, @jimblandy, @kvark)
-
Expose all interfaces to DedicatedWorker gpuweb/gpuweb#1883 (4 comments by @kainino0x, @kdashg, @toji)
-
copyExternalImageToTexture: add validation, make validation async gpuweb/gpuweb#1737 (4 comments by @Kaiido, @kainino0x)
-
wgsl: allow unsigned workgroup_size components gpuweb/gpuweb#1858 (4 comments by @alan-baker, @dneto0, @kdashg)
-
copyExternalImageToTexture: add colorSpace and premultipliedAlpha gpuweb/gpuweb#1768 (4 comments by @kainino0x, @Kangz, @toji)
-
Increase maxStorageBuffer/TexturePerShaderStage to 8. gpuweb/gpuweb#1802 (4 comments by @kainino0x, @Kangz, @kvark)
-
wgsl: cleanup old text, discuss conversions/constructions gpuweb/gpuweb#1875 (3 comments by @alan-baker, @dneto0)
-
Alignment for index and vertex buffers gpuweb/gpuweb#1854 (3 comments by @kainino0x, @Kangz, @kvark)
-
Use namespaces for flags instead of interfaces gpuweb/gpuweb#1709 (3 comments by @kainino0x, @Kangz)
-
Add note that sized and runtime-sized arrays are distinct types gpuweb/gpuweb#1833 (1 comment by @dneto0)
-
Add forceSoftware adapter option and isSoftware adapter attribute gpuweb/gpuweb#1734 (1 comment by @toji)
-
wgsl: textureSampleLevel on texture_external samples level 0 gpuweb/gpuweb#1797 (1 comment by @dneto0)
-
TR publications
-
10 documents published.
-
WebGPU published 4 times as a Working Draft. History of WebGPU
-
WebGPU Shading Language published 4 times as a Working Draft. History of WebGPU Shading Language
-
WebGPU published as a First Public Working Draft. History of WebGPU
-
WebGPU Shading Language published as a First Public Working Draft. History of WebGPU Shading Language
-
GitHub
gpuweb/gpuweb
-
28 issues opened.
-
Support for gl_PrimitiveID / SV_PrimitiveID -like construct? gpuweb/gpuweb#1786, by @darzu.
-
"Sampled" is ambiguous in the spec gpuweb/gpuweb#1785, by @kvark.
-
xor doesn't work for vec gpuweb/gpuweb#1784, by @Ciebiada.
-
Limit the binding number for resources gpuweb/gpuweb#1783, by @kvark.
-
WGSL requires dynamic matrix indexing, which requires surprising SPIR-V gpuweb/gpuweb#1782, by @jimblandy.
-
Define behavior of reading from WebGPU canvases using other Web APIs gpuweb/gpuweb#1781, by @kainino0x.
-
Consider merging GPUSwapChain into GPUCanvasContext gpuweb/gpuweb#1780, by @kainino0x.
-
wgsl: compute shaders should have a workgroup_size attribute gpuweb/gpuweb#1777, by @dneto0.
-
signed integer division: unspecified rounding mode for the quotient; and overflow gpuweb/gpuweb#1774, by @dneto0.
-
Specify behavior of GPUCommandBuffer.executionTime on devices without timestamp queries. gpuweb/gpuweb#1773, by @toji.
-
Allow read-write storage on selected texture formats gpuweb/gpuweb#1772, by @kvark.
-
SPIR-V doesn't grant reconvergence for multiple returns gpuweb/gpuweb#1770, by @kdashg.
-
wgsl: can hex floats explicitly represent infinity and NaN as float literals? gpuweb/gpuweb#1769, by @dneto0.
-
Need to define alpha premultiplication for external images gpuweb/gpuweb#1762, by @kainino0x.
-
Reconsider the name of storeOp: "clear" gpuweb/gpuweb#1761, by @kainino0x.
-
Wrong certificate on https://webgpu.io gpuweb/gpuweb#1757, by @sinnpi.
-
Why is `block` the only keyword attribute? gpuweb/gpuweb#1752, by @elijaharita.
-
Are padding bytes of structures intialized with zero value expressions? gpuweb/gpuweb#1747, by @dj2.
-
`location` decoration example in wrong spot gpuweb/gpuweb#1746, by @dj2.
-
Errant reference to `break if` in WGSL spec gpuweb/gpuweb#1745, by @dj2.
-
Shader-visible entities occupying multiple locations gpuweb/gpuweb#1733, by @kvark.
-
What binding numbers are used by an external texture binding? gpuweb/gpuweb#1721, by @kainino0x.
-
[proposal] Numerical compliance gpuweb/gpuweb#1716, by @alan-baker.
-
copyExternalImageToTexture/importExternalTexture with destination in linear color space gpuweb/gpuweb#1715, by @kainino0x.
-
specify behaviour of floating point operations on NaNs gpuweb/gpuweb#1711, by @dneto0.
-
Automate publication of WebGPU / WGSL to /TR gpuweb/gpuweb#1710, by @tidoust.
-
Dispatch usage scopes contain unused bindgroups gpuweb/gpuweb#1708, by @Kangz.
-
Is rendering to many different canvas elements supposed to be slow? gpuweb/gpuweb#1704, by @ziriax.
-
-
22 issues closed.
-
xor doesn't work for vec gpuweb/gpuweb#1784.
-
[proposal] Numerical compliance gpuweb/gpuweb#1716.
-
Errant reference to `break if` in WGSL spec gpuweb/gpuweb#1745.
-
consider pulling 'access' into the type of a pointer/reference gpuweb/gpuweb#1604.
-
Automate publication of WebGPU / WGSL to /TR gpuweb/gpuweb#1710.
-
Wrong certificate on https://webgpu.io gpuweb/gpuweb#1757.
-
[wgsl]: textureDimensions() returns vec3 for cube textures gpuweb/gpuweb#1345.
-
var: infer type from initializer gpuweb/gpuweb#1550.
-
`location` decoration example in wrong spot gpuweb/gpuweb#1746.
-
GPUCommandEncoder should be called GPUCommandBufferEncoder. gpuweb/gpuweb#1624.
-
Should we have a core "stencil8" format? gpuweb/gpuweb#306.
-
does textureSampleCompare assume level-of-detail 0 in non-fragment shaders? gpuweb/gpuweb#1319.
-
[wgsl] support readonly and writeonly on storage buffers gpuweb/gpuweb#935.
-
Proposal for a GPUVideoTexture object. gpuweb/gpuweb#1415.
-
[wgsl] Intra-scope shadowing? gpuweb/gpuweb#1473.
-
Would love to see optional Geometry Shaders gpuweb/gpuweb#1239.
-
Require that vertex shaders return a position? gpuweb/gpuweb#1567.
-
wgsl: storable types include atomics, textures, samplers gpuweb/gpuweb#1686.
-
Invariant qualifier gpuweb/gpuweb#893.
-
Is rendering to many different canvas elements supposed to be slow? gpuweb/gpuweb#1704.
-
-
65 issues received 225 new comments.
-
Buffer indices should be unsigned gpuweb/gpuweb#1135 (24 comments by @alan-baker, @dneto0, @kainino0x, @Kangz, @kdashg, @kvark, @litherum, @RafaelCintron)
-
consider pulling 'access' into the type of a pointer/reference gpuweb/gpuweb#1604 (19 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
What happens in configureSwapChain and on canvas resize? gpuweb/gpuweb#1691 (16 comments by @austinEng, @kainino0x, @Kangz, @kdashg, @magcius, @toji)
-
Investigation: Import VideoFrame from WebCodec to WebGPU gpuweb/gpuweb#1380 (13 comments by @kainino0x, @Kangz, @kdashg, @kvark, @shaoboyan)
-
Shader-visible entities occupying multiple locations gpuweb/gpuweb#1733 (11 comments by @Kangz, @kvark, @litherum)
-
Out-of-bounds clamping brings big perf regression on TFJS ML demos. gpuweb/gpuweb#1202 (10 comments by @kainino0x, @Kangz, @krogovin, @qjia7)
-
FMod vs FRem semantics gpuweb/gpuweb#1696 (9 comments by @alan-baker, @dneto0, @kdashg, @kvark, @litherum, @RobinMorisset)
-
What binding numbers are used by an external texture binding? gpuweb/gpuweb#1721 (7 comments by @dneto0, @kainino0x, @Kangz, @kvark, @shaoboyan)
-
[proposal] Numerical compliance gpuweb/gpuweb#1716 (7 comments by @alan-baker, @kdashg, @litherum, @mmgeorge)
-
SPIR-V doesn't grant reconvergence for multiple returns gpuweb/gpuweb#1770 (6 comments by @dneto0, @kainino0x, @kdashg, @kvark, @litherum)
-
Investigation: Compressed Texture Formats gpuweb/gpuweb#144 (6 comments by @alecazam, @fintelia, @kainino0x, @kdashg, @limingchina)
-
The order of top-level declarations should be irrelevant gpuweb/gpuweb#875 (5 comments by @fintelia, @kvark, @litherum)
-
Are padding bytes of structures intialized with zero value expressions? gpuweb/gpuweb#1747 (4 comments by @alan-baker, @dneto0, @litherum)
-
Need to define alpha premultiplication for external images gpuweb/gpuweb#1762 (4 comments by @kainino0x, @shaoboyan)
-
[wgsl]: textureDimensions() returns vec3 for cube textures gpuweb/gpuweb#1345 (4 comments by @ben-clayton, @dneto0, @kdashg)
-
Should all the constant-only interfaces be namespaces instead? gpuweb/gpuweb#304 (4 comments by @domenic, @Kangz, @toji)
-
Support for gl_PrimitiveID / SV_PrimitiveID -like construct? gpuweb/gpuweb#1786 (3 comments by @Kangz, @kvark)
-
wgsl: can hex floats explicitly represent infinity and NaN as float literals? gpuweb/gpuweb#1769 (3 comments by @dneto0, @litherum)
-
"workgroup_size" builtin is not supported on some Apple platforms gpuweb/gpuweb#1590 (3 comments by @jrprice, @kvark)
-
does textureSampleCompare assume level-of-detail 0 in non-fragment shaders? gpuweb/gpuweb#1319 (3 comments by @dneto0, @kainino0x, @kdashg)
-
[wgsl] support readonly and writeonly on storage buffers gpuweb/gpuweb#935 (3 comments by @dneto0, @kdashg)
-
Consider merging GPUSwapChain into GPUCanvasContext gpuweb/gpuweb#1780 (2 comments by @Kangz, @kvark)
-
Consider having rich math rendering in the spec gpuweb/gpuweb#1573 (2 comments by @kainino0x, @litherum)
-
Specify behavior of GPUCommandBuffer.executionTime on devices without timestamp queries. gpuweb/gpuweb#1773 (2 comments by @kainino0x, @Kangz)
-
Should we be able to pass textures and samplers as function parameters? gpuweb/gpuweb#1695 (2 comments by @kdashg, @litherum)
-
Should WGSL expose the concept of coherent buffers? gpuweb/gpuweb#1621 (2 comments by @dneto0, @raphlinus)
-
Define behavior of reading from WebGPU canvases using other Web APIs gpuweb/gpuweb#1781 (1 comment by @Kangz)
-
signed integer division: unspecified rounding mode for the quotient; and overflow gpuweb/gpuweb#1774 (1 comment by @kvark)
-
editorial: I'd like to write WGSL instead of [SHORTNAME] gpuweb/gpuweb#1589 (1 comment by @litherum)
-
wgsl: an entry point must not be the target of a function call gpuweb/gpuweb#1663 (1 comment by @litherum)
-
[wgsl] Redundant [[layout]] decorations in addition to [[offset]] decorations gpuweb/gpuweb#621 (1 comment by @kainino0x)
-
wgsl: support extended integer arithmetic for add, subtract, multiply gpuweb/gpuweb#1565 (1 comment by @kdashg)
-
specify behaviour of floating point operations on NaNs gpuweb/gpuweb#1711 (1 comment by @alan-baker)
-
List of issues to discuss with TC39, WASM, JS engine teams gpuweb/gpuweb#747 (1 comment by @kainino0x)
-
Is rendering to many different canvas elements supposed to be slow? gpuweb/gpuweb#1704 (1 comment by @ziriax)
-
-
54 pull requests opened.
-
Add swapchain size to the configuration gpuweb/gpuweb#1779, by @kvark.
-
wsgl: remove dead section gpuweb/gpuweb#1778, by @dneto0.
-
wgsl: define texture type parameterization gpuweb/gpuweb#1776, by @dneto0.
-
wgsl: Fix typo for modf pointer type gpuweb/gpuweb#1775, by @dneto0.
-
Remove workgroup_size builtin gpuweb/gpuweb#1771, by @jrprice.
-
copyExternalImageToTexture: add colorSpace and premultipliedAlpha gpuweb/gpuweb#1768, by @kainino0x.
-
Add `void` to the reserved keywords list gpuweb/gpuweb#1767, by @jrprice.
-
Add API surface for pipeline-overridable constants gpuweb/gpuweb#1766, by @kainino0x.
-
Update GPUAdapterLimits to GPUSupportedLimits gpuweb/gpuweb#1765, by @toji.
-
Fix description of unsigned integer mod gpuweb/gpuweb#1764, by @dneto0.
-
Remove % for signed integral gpuweb/gpuweb#1763, by @dneto0.
-
publish-specs: run on all PRs, remove special syntax from matrix gpuweb/gpuweb#1760, by @kainino0x.
-
Make "valid to use with" check for device validity gpuweb/gpuweb#1759, by @toji.
-
Remove colliding [[device]] slot from GPUDevice gpuweb/gpuweb#1758, by @toji.
-
wgsl: infer var store type from initializer gpuweb/gpuweb#1756, by @dneto0.
-
wgsl: Clarify coordinate use by cube textures gpuweb/gpuweb#1755, by @dneto0.
-
textureDimensions on cube returns 2-elem vec gpuweb/gpuweb#1754, by @dneto0.
-
wgsl: Remove stale "location" decoration example gpuweb/gpuweb#1753, by @dneto0.
-
Restrict barriers to compute shader stage gpuweb/gpuweb#1751, by @alan-baker.
-
Clarify that buffer.offset in WriteTexture has no alignment requirement gpuweb/gpuweb#1750, by @Richard-Yunchao.
-
[editorial] Remove type shorthands gpuweb/gpuweb#1749, by @alan-baker.
-
Add some text to describe function variables going out of scope gpuweb/gpuweb#1748, by @alan-baker.
-
Remove redundant type info on executionTime gpuweb/gpuweb#1744, by @kainino0x.
-
Add floating point execution rules gpuweb/gpuweb#1743, by @alan-baker.
-
Allow non-overridable constants in workgroup_size gpuweb/gpuweb#1742, by @jrprice.
-
Remove texture_multisampled_2d_array builtins gpuweb/gpuweb#1741, by @jrprice.
-
Re-introduce floating point comparisons gpuweb/gpuweb#1740, by @alan-baker.
-
Automate publication of specs to /TR gpuweb/gpuweb#1739, by @tidoust.
-
Document new resizing logic for GPUSwapChain gpuweb/gpuweb#1738, by @toji.
-
copyExternalImageToTexture: add validation, make validation async gpuweb/gpuweb#1737, by @kainino0x.
-
copyExternalImageToTexture: disallow detached ImageBitmap, add todo gpuweb/gpuweb#1736, by @kainino0x.
-
wgsl: access mode is no longer an attribute gpuweb/gpuweb#1735, by @dneto0.
-
Add forceSoftware adapter option and isSoftware adapter attribute gpuweb/gpuweb#1734, by @toji.
-
wgsl editorial: clean up IO-shareable, storable gpuweb/gpuweb#1732, by @dneto0.
-
wgsl: Workgroup storage holds plain types gpuweb/gpuweb#1731, by @dneto0.
-
[editorial] Leverage component-wise to reduce number of table entries gpuweb/gpuweb#1730, by @alan-baker.
-
wgsl: add textureSampleCompareLevel gpuweb/gpuweb#1729, by @dneto0.
-
wgsl: describe derivatives, related uniformity gpuweb/gpuweb#1728, by @dneto0.
-
wgsl: mipmap level -> mip level gpuweb/gpuweb#1727, by @dneto0.
-
wgsl: Add missing w3cid, fix invalid markup and link to /TR document gpuweb/gpuweb#1726, by @tidoust.
-
Add w3cid to editors and fix new invalid markup gpuweb/gpuweb#1725, by @tidoust.
-
s/const/let, remove whitespace before : in `name: type` gpuweb/gpuweb#1724, by @ben-clayton.
-
wgsl: fix cross-references to WebGPU spec gpuweb/gpuweb#1723, by @dneto0.
-
Clarify out of bounds behavior gpuweb/gpuweb#1722, by @alan-baker.
-
Limit pipeline constant IDs to between 0 and 65535 gpuweb/gpuweb#1720, by @jrprice.
-
[editorial] Refactor validation section gpuweb/gpuweb#1719, by @alan-baker.
-
Remove IN and OUT from storage class grammar gpuweb/gpuweb#1718, by @alan-baker.
-
wgsl: fix typo in module scope variables example gpuweb/gpuweb#1717, by @dwbrite.
-
Clarify function parameter and return types gpuweb/gpuweb#1714, by @alan-baker.
-
explainer: Put the security section after the JS section gpuweb/gpuweb#1713, by @Kangz.
-
Fix a bug in GPUExtent3D gpuweb/gpuweb#1712, by @Richard-Yunchao.
-
Use namespaces for flags instead of interfaces gpuweb/gpuweb#1709, by @kainino0x.
-
wgsl: Fix markup, broken links, adjust metadata for publication to /TR gpuweb/gpuweb#1707, by @tidoust.
-
WebGPU: Fix markup, broken links, adjust metadata for publication to /TR gpuweb/gpuweb#1706, by @tidoust.
-
-
51 pull requests merged.
-
wsgl: remove dead section gpuweb/gpuweb#1778.
-
Make "valid to use with" check for device validity gpuweb/gpuweb#1759.
-
Add `void` to the reserved keywords list gpuweb/gpuweb#1767.
-
wgsl: Fix typo for modf pointer type gpuweb/gpuweb#1775.
-
Add floating point execution rules gpuweb/gpuweb#1743.
-
wgsl: access mode is no longer an attribute gpuweb/gpuweb#1735.
-
wgsl: mipmap level -> mip level gpuweb/gpuweb#1727.
-
Allow non-overridable constants in workgroup_size gpuweb/gpuweb#1742.
-
Fix description of unsigned integer mod gpuweb/gpuweb#1764.
-
Add some text to describe function variables going out of scope gpuweb/gpuweb#1748.
-
publish-specs: run on all PRs, remove special syntax from matrix gpuweb/gpuweb#1760.
-
Automate publication of specs to /TR gpuweb/gpuweb#1739.
-
[editorial] Remove type shorthands gpuweb/gpuweb#1749.
-
Clarify that buffer.offset in WriteTexture has no alignment requirement gpuweb/gpuweb#1750.
-
textureDimensions on cube returns 2-elem vec gpuweb/gpuweb#1754.
-
wgsl: Clarify coordinate use by cube textures gpuweb/gpuweb#1755.
-
wgsl: infer var store type from initializer gpuweb/gpuweb#1756.
-
wgsl: Remove stale "location" decoration example gpuweb/gpuweb#1753.
-
Restrict barriers to compute shader stage gpuweb/gpuweb#1751.
-
Remove texture_multisampled_2d_array builtins gpuweb/gpuweb#1741.
-
Remove redundant type info on executionTime gpuweb/gpuweb#1744.
-
wgsl: describe derivatives, related uniformity gpuweb/gpuweb#1728.
-
Limit pipeline constant IDs to between 0 and 65535 gpuweb/gpuweb#1720.
-
Re-introduce floating point comparisons gpuweb/gpuweb#1740.
-
wgsl: Fix typo in vertex_index gpuweb/gpuweb#1687.
-
wgsl editorial: clean up IO-shareable, storable gpuweb/gpuweb#1732.
-
Add w3cid to editors and fix new invalid markup gpuweb/gpuweb#1725.
-
[editorial] Leverage component-wise to reduce number of table entries gpuweb/gpuweb#1730.
-
wgsl: Workgroup storage holds plain types gpuweb/gpuweb#1731.
-
wgsl: add textureSampleCompareLevel gpuweb/gpuweb#1729.
-
wgsl: Add missing w3cid, fix invalid markup and link to /TR document gpuweb/gpuweb#1726.
-
s/const/let, remove whitespace before : in `name: type` gpuweb/gpuweb#1724.
-
wgsl: fix cross-references to WebGPU spec gpuweb/gpuweb#1723.
-
wgsl: Fix markup, broken links, adjust metadata for publication to /TR gpuweb/gpuweb#1707.
-
GPUExternalTexture: API for video; WGSL type and functions gpuweb/gpuweb#1666.
-
Add initial color spaces gpuweb/gpuweb#1690.
-
[editorial] Refactor validation section gpuweb/gpuweb#1719.
-
wgsl: fix typo in module scope variables example gpuweb/gpuweb#1717.
-
Remove IN and OUT from storage class grammar gpuweb/gpuweb#1718.
-
explainer: Put the security section after the JS section gpuweb/gpuweb#1713.
-
Clarify function parameter and return types gpuweb/gpuweb#1714.
-
Add function restrictions gpuweb/gpuweb#1674.
-
wgsl: Fix headings on three examples gpuweb/gpuweb#1680.
-
Refactor binding limit validation spec gpuweb/gpuweb#1689.
-
[editorial] Unify built-in function tables gpuweb/gpuweb#1701.
-
Merge texture's internal slots gpuweb/gpuweb#1694.
-
Add invariant attribute gpuweb/gpuweb#1673.
-
Clarify that CompilationMessage line/pos numbers should be 1-based gpuweb/gpuweb#1700.
-
WebGPU: Fix markup, broken links, adjust metadata for publication to /TR gpuweb/gpuweb#1706.
-
wgsl: "Structure Layout Rules": Remove forbidden [[block]] attrs. gpuweb/gpuweb#1703.
-
-
5 pull requests closed.
-
Add external_texture luminance functions gpuweb/gpuweb#1681.
-
copyExternalImageToTexture: disallow detached ImageBitmap, add todo gpuweb/gpuweb#1736.
-
Pluralize requestAdapters() and add GPUAdapter.isSoftware gpuweb/gpuweb#1660.
-
textureSampleCompare always uses mip level 0 gpuweb/gpuweb#1320.
-
copyExternalImageToTexture for HTMLVideoElement gpuweb/gpuweb#1647.
-
-
52 pull requests received 309 new comments.
-
Automate publication of specs to /TR gpuweb/gpuweb#1739 (29 comments by @dneto0, @kainino0x, @kvark, @tidoust)
-
Clarify out of bounds behavior gpuweb/gpuweb#1722 (24 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
wgsl: access mode is no longer an attribute gpuweb/gpuweb#1735 (19 comments by @alan-baker, @dneto0, @kvark)
-
Add API surface for pipeline-overridable constants gpuweb/gpuweb#1766 (18 comments by @alan-baker, @kainino0x, @kvark, @toji)
-
GPUExternalTexture: API for video; WGSL type and functions gpuweb/gpuweb#1666 (18 comments by @kainino0x, @Kangz, @kdashg, @kvark, @shaoboyan)
-
Document new resizing logic for GPUSwapChain gpuweb/gpuweb#1738 (14 comments by @kainino0x, @kvark, @toji)
-
Clarify that buffer.offset in WriteTexture has no alignment requirement gpuweb/gpuweb#1750 (13 comments by @kainino0x, @Kangz, @Richard-Yunchao)
-
copyExternalImageToTexture: add validation, make validation async gpuweb/gpuweb#1737 (12 comments by @kainino0x, @kvark, @shaoboyan, @toji)
-
[editorial] Remove type shorthands gpuweb/gpuweb#1749 (12 comments by @alan-baker, @dneto0, @kainino0x, @kvark)
-
wgsl: Fix markup, broken links, adjust metadata for publication to /TR gpuweb/gpuweb#1707 (8 comments by @dneto0, @kvark, @tidoust)
-
copyExternalImageToTexture: add colorSpace and premultipliedAlpha gpuweb/gpuweb#1768 (7 comments by @kainino0x, @kvark, @toji)
-
Pluralize requestAdapters() and add GPUAdapter.isSoftware gpuweb/gpuweb#1660 (6 comments by @litherum, @toji)
-
Introduce Subgroup Operations Extension gpuweb/gpuweb#1459 (6 comments by @m-schuetz, @mehmetoguzderin)
-
Add some text to describe function variables going out of scope gpuweb/gpuweb#1748 (5 comments by @alan-baker, @dneto0)
-
Add external_texture luminance functions gpuweb/gpuweb#1681 (5 comments by @kainino0x, @Kangz, @litherum)
-
Use namespaces for flags instead of interfaces gpuweb/gpuweb#1709 (5 comments by @domenic, @ExE-Boss, @kainino0x, @kvark)
-
explainer: Put the security section after the JS section gpuweb/gpuweb#1713 (5 comments by @kainino0x, @Kangz, @kvark)
-
add atomic type, texture type, store type to storable; adjust rules for var and let. issue #1686 gpuweb/gpuweb#1693 (5 comments by @dneto0, @kdashg, @litherum)
-
WebGPU: Fix markup, broken links, adjust metadata for publication to /TR gpuweb/gpuweb#1706 (5 comments by @kvark, @tidoust)
-
Remove texture_multisampled_2d_array builtins gpuweb/gpuweb#1741 (3 comments by @alan-baker, @kvark)
-
[editorial] Leverage component-wise to reduce number of table entries gpuweb/gpuweb#1730 (2 comments by @dneto0)
-
Clarify that CompilationMessage line/pos numbers should be 1-based gpuweb/gpuweb#1700 (2 comments by @toji)
-
wgsl: "Structure Layout Rules": Remove forbidden [[block]] attrs. gpuweb/gpuweb#1703 (2 comments by @ben-clayton, @w3cbot)
-
publish-specs: run on all PRs, remove special syntax from matrix gpuweb/gpuweb#1760 (1 comment by @kainino0x)
-
copyExternalImageToTexture: disallow detached ImageBitmap, add todo gpuweb/gpuweb#1736 (1 comment by @kainino0x)
-
Add forceSoftware adapter option and isSoftware adapter attribute gpuweb/gpuweb#1734 (1 comment by @kainino0x)
-
gpuweb/cts
-
2 issues opened.
-
expectCompileResult should check for validation error gpuweb/cts#556, by @kainino0x.
-
-
1 issue received 5 new comments.
-
18 pull requests opened.
-
Update validation tests of copyImageBitmapToTexture gpuweb/cts#568, by @shaoboyan.
-
Remove last WGSL `const` and `-> void` usages gpuweb/cts#567, by @jrprice.
-
Add {Case,Subcase}ParamsBuilder to replace .cases().subcases() gpuweb/cts#566, by @kainino0x.
-
Simplify TestBuilder, generalize ParamArgument, rename CaseParams gpuweb/cts#565, by @kainino0x.
-
Make WPT output prettier gpuweb/cts#564, by @kainino0x.
-
Check features support when requesting device gpuweb/cts#563, by @haoxli.
-
Document cases/subcases and ParamsBuilder better gpuweb/cts#562, by @kainino0x.
-
Use more subcases gpuweb/cts#561, by @kainino0x.
-
Fix sampledTextureBindingTypeInfo gpuweb/cts#560, by @kainino0x.
-
Review createBindGroupLayout spec gpuweb/cts#559, by @toji.
-
Add DrawIndirect test gpuweb/cts#558, by @shrekshao.
-
Update texture zero init tests for 3D textures gpuweb/cts#557, by @austinEng.
-
Add compute,basic,large_dispatch test. gpuweb/cts#555, by @kenrussell.
-
Reviewed createBindGroup spec gpuweb/cts#554, by @toji.
-
Update variable_and_const.spec.ts shader to match current WGSL spec gpuweb/cts#553, by @zoddicus.
-
api,operation,buffers,map: describe, clean notes, add cases gpuweb/cts#552, by @kainino0x.
-
Update/cleanup map_oom.spec.ts gpuweb/cts#551, by @austinEng.
-
docs: nits, add pointer to a good reference test gpuweb/cts#550, by @kainino0x.
-
-
17 pull requests merged.
-
Remove last WGSL `const` and `-> void` usages gpuweb/cts#567.
-
Make WPT output prettier gpuweb/cts#564.
-
Use more subcases gpuweb/cts#561.
-
Document cases/subcases and ParamsBuilder better gpuweb/cts#562.
-
Review createBindGroupLayout spec gpuweb/cts#559.
-
Add DrawIndirect test gpuweb/cts#558.
-
Fix sampledTextureBindingTypeInfo gpuweb/cts#560.
-
Expand operation tests for copyToTexture,ImageBitmap gpuweb/cts#545.
-
api,operation,buffers,map: describe, clean notes, add cases gpuweb/cts#552.
-
Reviewed createBindGroup spec gpuweb/cts#554.
-
Added some CompilationInfo tests. gpuweb/cts#548.
-
gen_wpt_cts_html: always split down to individual tests gpuweb/cts#544.
-
Add compute,basic,large_dispatch test. gpuweb/cts#555.
-
Update/cleanup map_oom.spec.ts gpuweb/cts#551.
-
docs: nits, add pointer to a good reference test gpuweb/cts#550.
-
-
1 pull request reopened.
-
Check features support when requesting device gpuweb/cts#563, by @haoxli.
-
-
20 pull requests received 120 new comments.
-
Expand operation tests for copyToTexture,ImageBitmap gpuweb/cts#545 (21 comments by @kainino0x, @shaoboyan)
-
Update variable_and_const.spec.ts shader to match current WGSL spec gpuweb/cts#553 (11 comments by @jrprice, @kainino0x, @zoddicus)
-
Update texture zero init tests for 3D textures gpuweb/cts#557 (10 comments by @austinEng, @kainino0x, @Richard-Yunchao)
-
Update validation tests of copyImageBitmapToTexture gpuweb/cts#568 (6 comments by @kainino0x, @shaoboyan)
-
Check features support when requesting device gpuweb/cts#563 (6 comments by @haoxli, @kainino0x, @Kangz)
-
Simplify TestBuilder, generalize ParamArgument, rename CaseParams gpuweb/cts#565 (4 comments by @austinEng, @kainino0x)
-
Add {Case,Subcase}ParamsBuilder to replace .cases().subcases() gpuweb/cts#566 (2 comments by @kainino0x)
-
Document cases/subcases and ParamsBuilder better gpuweb/cts#562 (2 comments by @kainino0x, @shrekshao)
-
api,operation,buffers,map: describe, clean notes, add cases gpuweb/cts#552 (1 comment by @kainino0x)
-
Participants
-
1 chair appointed.
-
Kelsey Gilbert became a chair.
-
-
1 chair stepped down.
-
Dean Jackson stepped down as a chair.
-
-
1 participant joined.
-
Kelsey Gilbert (W3C Invited Experts) joined.
-
GitHub
gpuweb/cts
-
33 pull requests opened.
-
Added some CompilationInfo tests. gpuweb/cts#548, by @toji.
-
[CI] Reduce expire field to 10 days gpuweb/cts#547, by @mehmetoguzderin.
-
Fix WGSL storage_buffer->storage in by_sampling.ts gpuweb/cts#546, by @austinEng.
-
Expand operation tests for copyToTexture,ImageBitmap gpuweb/cts#545, by @shaoboyan.
-
gen_wpt_cts_html: always split down to individual tests gpuweb/cts#544, by @kainino0x.
-
Use new WGSL entry point IO syntax gpuweb/cts#543, by @jrprice.
-
Add tests for GCing GPUBuffer before their mappings. gpuweb/cts#542, by @Kangz.
-
operation,vertex_state: add todo gpuweb/cts#541, by @kainino0x.
-
Replace builtin(position) with builtin(frag_coord) gpuweb/cts#540, by @jrprice.
-
Add validation tests for occlusion query gpuweb/cts#539, by @haoxli.
-
Update bindings API usage gpuweb/cts#538, by @kainino0x.
-
Update WGSL: const -> let and remove '-> void' gpuweb/cts#537, by @shrekshao.
-
Replace fence tests with new onSubmittedWorkDone tests gpuweb/cts#536, by @kainino0x.
-
More deprecations gpuweb/cts#535, by @kainino0x.
-
Fix some storage buffers gpuweb/cts#534, by @alan-baker.
-
Fix types deprecations gpuweb/cts#533, by @kainino0x.
-
roll types, fixes for Iterable color/extent/origin gpuweb/cts#532, by @kainino0x.
-
Roll types gpuweb/cts#531, by @kainino0x.
-
Update README.md to mention WPT gpuweb/cts#530, by @Kangz.
-
Add selectDeviceForQueryTypeOrSkipTestCase to gpu_test.ts gpuweb/cts#529, by @haoxli.
-
Add tests for mipmap with various formats gpuweb/cts#528, by @Richard-Yunchao.
-
Fix bugs for texture creation validation gpuweb/cts#527, by @Richard-Yunchao.
-
Add tests for GPUBlendComponent gpuweb/cts#526, by @austinEng.
-
Much more verbose checklists gpuweb/cts#525, by @kainino0x.
-
Convert primitive topology tests to a single parameterized test gpuweb/cts#523, by @austinEng.
-
Implement api,operation,rendering,draw:vertex_attributes,basic gpuweb/cts#522, by @austinEng.
-
Fix a bug in createTexture validation for multisample gpuweb/cts#521, by @Richard-Yunchao.
-
node.tsconfig.json: include more files gpuweb/cts#520, by @austinEng.
-
Some depth related tests gpuweb/cts#519, by @shrekshao.
-
checklist.ts: Split lines by either \r\n or \n gpuweb/cts#518, by @austinEng.
-
gen_wpt_cts_html: Split lines by either \r\n or \n gpuweb/cts#517, by @austinEng.
-
Fix multicase test expectation filtering gpuweb/cts#516, by @austinEng.
-
-
30 pull requests merged.
-
[CI] Reduce expire field to 10 days gpuweb/cts#547.
-
Fix WGSL storage_buffer->storage in by_sampling.ts gpuweb/cts#546.
-
operation,vertex_state: add todo gpuweb/cts#541.
-
Add validation tests for occlusion query gpuweb/cts#539.
-
More deprecations gpuweb/cts#535.
-
Use new WGSL entry point IO syntax gpuweb/cts#543.
-
Replace fence tests with new onSubmittedWorkDone tests gpuweb/cts#536.
-
Add tests for GCing GPUBuffer before their mappings. gpuweb/cts#542.
-
Update bindings API usage gpuweb/cts#538.
-
Replace builtin(position) with builtin(frag_coord) gpuweb/cts#540.
-
Update WGSL: const -> let and remove '-> void' gpuweb/cts#537.
-
Fix some storage buffers gpuweb/cts#534.
-
Fix types deprecations gpuweb/cts#533.
-
roll types, fixes for Iterable color/extent/origin gpuweb/cts#532.
-
Roll types gpuweb/cts#531.
-
Update README.md to mention WPT gpuweb/cts#530.
-
Add selectDeviceForQueryTypeOrSkipTestCase to gpu_test.ts gpuweb/cts#529.
-
Add tests for mipmap with various formats gpuweb/cts#528.
-
Fix bugs for texture creation validation gpuweb/cts#527.
-
Add tests for GPUBlendComponent gpuweb/cts#526.
-
Much more verbose checklists gpuweb/cts#525.
-
Implement api,operation,rendering,draw:vertex_attributes,basic gpuweb/cts#522.
-
Convert primitive topology tests to a single parameterized test gpuweb/cts#523.
-
Some depth related tests gpuweb/cts#519.
-
Fix a bug in createTexture validation for multisample gpuweb/cts#521.
-
node.tsconfig.json: include more files gpuweb/cts#520.
-
gen_wpt_cts_html: Split lines by either \r\n or \n gpuweb/cts#517.
-
checklist.ts: Split lines by either \r\n or \n gpuweb/cts#518.
-
Fix multicase test expectation filtering gpuweb/cts#516.
-
-
1 pull request closed.
-
22 pull requests received 100 new comments.
-
gen_wpt_cts_html: always split down to individual tests gpuweb/cts#544 (10 comments by @austinEng, @kainino0x)
-
Add validation tests for occlusion query gpuweb/cts#539 (8 comments by @haoxli, @kainino0x, @shaoboyan)
-
Fix bugs for texture creation validation gpuweb/cts#527 (8 comments by @kainino0x, @Richard-Yunchao)
-
Expand operation tests for copyToTexture,ImageBitmap gpuweb/cts#545 (6 comments by @haoxli, @shaoboyan)
-
Fix a bug in createTexture validation for multisample gpuweb/cts#521 (5 comments by @austinEng, @kainino0x, @Richard-Yunchao)
-
Implement api,operation,rendering,draw:vertex_attributes,basic gpuweb/cts#522 (4 comments by @austinEng, @kainino0x)
-
Add selectDeviceForQueryTypeOrSkipTestCase to gpu_test.ts gpuweb/cts#529 (4 comments by @haoxli, @Jiawei-Shao, @kainino0x)
-
Add tests for mipmap with various formats gpuweb/cts#528 (4 comments by @kainino0x, @Kangz, @Richard-Yunchao)
-
Replace fence tests with new onSubmittedWorkDone tests gpuweb/cts#536 (2 comments by @kainino0x, @Kangz)
-
Implement api,operation,render_pipeline,primitive_topology:unaligned_vertex_count gpuweb/cts#524 (1 comment by @kainino0x)
-
Convert primitive topology tests to a single parameterized test gpuweb/cts#523 (1 comment by @austinEng)
-
Split createEncoder finish into three parts for strict validation gpuweb/cts#489 (1 comment by @kainino0x)
gpuweb/gpuweb
-
30 issues opened.
-
FMod vs FRem semantics gpuweb/gpuweb#1696, by @kvark.
-
Should we be able to pass textures and samplers as function parameters? gpuweb/gpuweb#1695, by @alan-baker.
-
Metal is unable to copy depth formats into colors gpuweb/gpuweb#1692, by @kvark.
-
What happens in configureSwapChain and on canvas resize? gpuweb/gpuweb#1691, by @kainino0x.
-
wgsl: storable types include atomics, textures, samplers gpuweb/gpuweb#1686, by @dneto0.
-
Define a MIME type gpuweb/gpuweb#1682, by @annevk.
-
pushDebugGroup / popDebugGroup on Queue gpuweb/gpuweb#1672, by @magcius.
-
Support NumWorkgroups builtin gpuweb/gpuweb#1671, by @kvark.
-
Allow importing canvases as `GPUExternalTexture`s gpuweb/gpuweb#1670, by @kainino0x.
-
wgsl: an entry point must not be the target of a function call gpuweb/gpuweb#1663, by @dneto0.
-
Allow mat2/3/4 as attribute types gpuweb/gpuweb#1652, by @Popov72.
-
Relax the out-of-bounds specification gpuweb/gpuweb#1649, by @kvark.
-
webgpu gpuweb/gpuweb#1644, by @M-Hiro-tex.
-
GPUImageCopyTexture misses arrayLayer gpuweb/gpuweb#1642, by @Reon90.
-
Reduce blocking on shader compilation gpuweb/gpuweb#1641, by @jdarpinian.
-
Should "mark adapters stale" be called more quickly? gpuweb/gpuweb#1630, by @kainino0x.
-
Make lost devices appear to function as much as possible gpuweb/gpuweb#1629, by @kainino0x.
-
Add transpose() builtin gpuweb/gpuweb#1625, by @kvark.
-
GPUCommandEncoder should be called GPUCommandBufferEncoder. gpuweb/gpuweb#1624, by @ghost.
-
consider adding matrix determinant gpuweb/gpuweb#1623, by @dneto0.
-
consider adding: acosh, asinh, atanh gpuweb/gpuweb#1622, by @dneto0.
-
Should WGSL expose the concept of coherent buffers? gpuweb/gpuweb#1621, by @alan-baker.
-
Use the same interface for both GPUAdapter features and GPUDevice features gpuweb/gpuweb#1613, by @toji.
-
Need a proposal for specialization constants on the API side. gpuweb/gpuweb#1608, by @Kangz.
-
consider pulling 'access' into the type of a pointer/reference gpuweb/gpuweb#1604, by @dneto0.
-
GPUCompilationMessage: Specify linePos units, consider span gpuweb/gpuweb#1597, by @kainino0x.
-
Performance problems / best usage of the API gpuweb/gpuweb#1596, by @Popov72.
-
"workgroup_size" builtin is not supported on some Apple platforms gpuweb/gpuweb#1590, by @kvark.
-
editorial: I'd like to write WGSL instead of [SHORTNAME] gpuweb/gpuweb#1589, by @dneto0.
-
28 issues closed.
-
Support NumWorkgroups builtin gpuweb/gpuweb#1671.
-
Validation rules and issues of createTexture gpuweb/gpuweb#1522.
-
[proposal] Interface Matching gpuweb/gpuweb#1563.
-
GPUCompilationMessage: Specify linePos units, consider span gpuweb/gpuweb#1597.
-
Proposal for importing Web platform images in WebGPU gpuweb/gpuweb#1154.
-
Reduce blocking on shader compilation gpuweb/gpuweb#1641.
-
Uniform structure sizes have to be aligned to 16 bytes? gpuweb/gpuweb#1558.
-
GPUImageCopyTexture misses arrayLayer gpuweb/gpuweb#1642.
-
Add transpose() builtin gpuweb/gpuweb#1625.
-
Removing the need "non-filtering" samplers gpuweb/gpuweb#1562.
-
Specify what happens to mappings when the buffer is GCed. gpuweb/gpuweb#1544.
-
webgpu gpuweb/gpuweb#1644.
-
Support for vector splats (scalar operations on vectors) gpuweb/gpuweb#1520.
-
Removing `ptr` and adding `ref` gpuweb/gpuweb#1334.
-
specify rules about duplicate and maybe conflicting attributes gpuweb/gpuweb#1125.
-
Can WebGPU canvas alpha be configured? gpuweb/gpuweb#1425.
-
Performance problems / best usage of the API gpuweb/gpuweb#1596.
-
Queries spec inconsistent with ErrorConventions.md gpuweb/gpuweb#1532.
-
Proposal: `GPUAdapter.gpuApiPreferenceRating` gpuweb/gpuweb#1439.
-
Use the same interface for both GPUAdapter features and GPUDevice features gpuweb/gpuweb#1613.
-
Query API: What is the GPU timestamp unit on Metal? gpuweb/gpuweb#1325.
-
Atomics proposal gpuweb/gpuweb#1360.
-
[wgsl] Define interpolation decorations/qualifiers gpuweb/gpuweb#802.
-
spell out when an implementation must reject a shader module gpuweb/gpuweb#1241.
-
Have "var" as syntactic sugar for explicit pointer refs/derefs gpuweb/gpuweb#1456.
-
WGSL: add fn void(sink:T) -> void builtin function gpuweb/gpuweb#1499.
-
-
57 issues received 191 new comments.
-
consider pulling 'access' into the type of a pointer/reference gpuweb/gpuweb#1604 (14 comments by @alan-baker, @ben-clayton, @dneto0, @Kangz, @kvark)
-
Removing the need "non-filtering" samplers gpuweb/gpuweb#1562 (11 comments by @kainino0x, @Kangz, @kvark)
-
Should "mark adapters stale" be called more quickly? gpuweb/gpuweb#1630 (9 comments by @jdarpinian, @kainino0x, @kvark)
-
[wgsl] Encoding for shader text needs to be specified gpuweb/gpuweb#565 (8 comments by @annevk, @kainino0x, @kvark)
-
GPUCompilationMessage: Specify linePos units, consider span gpuweb/gpuweb#1597 (8 comments by @kainino0x, @Kangz)
-
Investigation: D3D12 does not support SRC_COLOR et al used in SrcBlendAlpha slot gpuweb/gpuweb#65 (7 comments by @kainino0x, @Kangz, @kvark)
-
Performance problems / best usage of the API gpuweb/gpuweb#1596 (7 comments by @Kangz, @kvark, @magcius, @Popov72)
-
What happens in configureSwapChain and on canvas resize? gpuweb/gpuweb#1691 (6 comments by @austinEng, @Juanmihd, @kainino0x, @toji)
-
Metal is unable to copy depth formats into colors gpuweb/gpuweb#1692 (5 comments by @austinEng, @kainino0x, @Kangz, @kvark)
-
The order of top-level declarations should be irrelevant gpuweb/gpuweb#875 (5 comments by @Jamesernator, @kvark, @litherum, @magcius)
-
pushDebugGroup / popDebugGroup on Queue gpuweb/gpuweb#1672 (5 comments by @kainino0x, @Kangz, @kvark)
-
Need a proposal for specialization constants on the API side. gpuweb/gpuweb#1608 (5 comments by @alan-baker, @kainino0x, @kdashg)
-
Relax the out-of-bounds specification gpuweb/gpuweb#1649 (5 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
Method of ensuring GPUShaderModules can contain MTLLibraries gpuweb/gpuweb#1064 (5 comments by @Kangz, @kdashg, @kvark, @litherum)
-
Support mix(vec, vec, f32) gpuweb/gpuweb#1588 (5 comments by @ben-clayton, @dneto0, @kdashg, @magcius)
-
Require that vertex shaders return a position? gpuweb/gpuweb#1567 (5 comments by @jimblandy, @kainino0x, @kdashg, @kvark)
-
Should objects expose the descriptor they were created with? gpuweb/gpuweb#1498 (5 comments by @almarklein, @kainino0x, @Kangz, @kvark)
-
Proposal for a GPUVideoTexture object. gpuweb/gpuweb#1415 (5 comments by @kainino0x, @RafaelCintron)
-
Should WGSL expose the concept of coherent buffers? gpuweb/gpuweb#1621 (4 comments by @alan-baker, @kvark)
-
[wgsl] Define interpolation decorations/qualifiers gpuweb/gpuweb#802 (4 comments by @alan-baker, @kdashg, @kvark, @pow2clk)
-
Allow importing canvases as `GPUExternalTexture`s gpuweb/gpuweb#1670 (3 comments by @kainino0x, @Kangz, @shaoboyan)
-
WebGPU should validate that storageTexture bindings have a single mip level. gpuweb/gpuweb#1576 (3 comments by @Kangz, @kvark)
-
Make lost devices appear to function as much as possible gpuweb/gpuweb#1629 (2 comments by @jdarpinian)
-
What is the toDataURL behavior for GPUCanvasContext? gpuweb/gpuweb#1548 (2 comments by @kainino0x, @kdashg)
-
Specify what happens to mappings when the buffer is GCed. gpuweb/gpuweb#1544 (2 comments by @kainino0x, @Kangz)
-
[wgsl] Proposal: Remove pointer out parameters from modf, frexp gpuweb/gpuweb#1480 (2 comments by @ben-clayton, @kdashg)
-
GPUCommandEncoder should be called GPUCommandBufferEncoder. gpuweb/gpuweb#1624 (2 comments by @Kangz, @kvark)
-
WGSL says [[block]] structs can't be members of other structs, but shows an example of this gpuweb/gpuweb#1702 (1 comment by @ben-clayton)
-
Should we be able to pass textures and samplers as function parameters? gpuweb/gpuweb#1695 (1 comment by @dneto0)
-
wgsl: an entry point must not be the target of a function call gpuweb/gpuweb#1663 (1 comment by @kvark)
-
arrayLength should operate on pointer-to-runtime-sized-array instead of runtime-array gpuweb/gpuweb#1329 (1 comment by @kdashg)
-
specify rules about duplicate and maybe conflicting attributes gpuweb/gpuweb#1125 (1 comment by @dneto0)
-
Should all the constant-only interfaces be namespaces instead? gpuweb/gpuweb#304 (1 comment by @domenic)
-
Use the same interface for both GPUAdapter features and GPUDevice features gpuweb/gpuweb#1613 (1 comment by @kvark)
-
Are values in GPUVertexAttribute.shaderLocation arbitrary? gpuweb/gpuweb#1418 (1 comment by @kainino0x)
-
does textureSampleCompare assume level-of-detail 0 in non-fragment shaders? gpuweb/gpuweb#1319 (1 comment by @litherum)
-
"workgroup_size" builtin is not supported on some Apple platforms gpuweb/gpuweb#1590 (1 comment by @dneto0)
-
call_statement: should only be able to call a function that yields void? gpuweb/gpuweb#1413 (1 comment by @kvark)
-
-
82 pull requests opened.
-
wgsl: "Structure Layout Rules": Remove forbidden [[block]] attrs. gpuweb/gpuweb#1703, by @jimblandy.
-
[editorial] Unify built-in function tables gpuweb/gpuweb#1701, by @alan-baker.
-
Clarify that CompilationMessage line/pos numbers should be 1-based gpuweb/gpuweb#1700, by @toji.
-
[CI] Reduce expire field to 10 days gpuweb/gpuweb#1699, by @mehmetoguzderin.
-
Merge texture's internal slots gpuweb/gpuweb#1694, by @Richard-Yunchao.
-
Add initial color spaces gpuweb/gpuweb#1690, by @kainino0x.
-
Refactor binding limit validation spec gpuweb/gpuweb#1689, by @kainino0x.
-
wgsl: only atomic modifications are mutually ordered gpuweb/gpuweb#1688, by @dneto0.
-
wgsl: Fix typo in vertex_index gpuweb/gpuweb#1687, by @egalli.
-
Clarify type matching for var initializer gpuweb/gpuweb#1685, by @dneto0.
-
Add num_workgroups built-in variable gpuweb/gpuweb#1684, by @alan-baker.
-
Add vertex and fragment interface matching rules gpuweb/gpuweb#1683, by @alan-baker.
-
Add external_texture luminance functions gpuweb/gpuweb#1681, by @kainino0x.
-
wgsl: Fix headings on three examples gpuweb/gpuweb#1680, by @dneto0.
-
wgsl: fill out var-and-let gpuweb/gpuweb#1679, by @dneto0.
-
Rename constant_id to override gpuweb/gpuweb#1678, by @alan-baker.
-
Explainer: edit motivation section gpuweb/gpuweb#1677, by @kainino0x.
-
Reorder sections gpuweb/gpuweb#1676, by @Kangz.
-
Update grammar for constexpr type constructors gpuweb/gpuweb#1675, by @jrprice.
-
Add function restrictions gpuweb/gpuweb#1674, by @alan-baker.
-
Add invariant attribute gpuweb/gpuweb#1673, by @alan-baker.
-
Deploy samples to GitHub pages. gpuweb/gpuweb#1669, by @jdarpinian.
-
Fix hello-cube.html sample gpuweb/gpuweb#1668, by @kainino0x.
-
Change GPUSupportedFeatures entry type to DOMString gpuweb/gpuweb#1667, by @toji.
-
GPUExternalTexture: API for video; WGSL type and functions gpuweb/gpuweb#1666, by @kainino0x.
-
Explainer: bit more editing gpuweb/gpuweb#1665, by @kainino0x.
-
Adding workload simulator as a sample. gpuweb/gpuweb#1664, by @jdarpinian.
-
Explainer: Add a more general introduction / motivation section. gpuweb/gpuweb#1662, by @Kangz.
-
Fix a few tiny issues about references gpuweb/gpuweb#1661, by @Richard-Yunchao.
-
Pluralize requestAdapters() and add GPUAdapter.isSoftware gpuweb/gpuweb#1660, by @toji.
-
Make GPUSupportedFeatures `setlike<DOMString>` gpuweb/gpuweb#1659, by @kainino0x.
-
Add spec for texture creation gpuweb/gpuweb#1658, by @Richard-Yunchao.
-
[editorial] Remove vector short convenience names gpuweb/gpuweb#1657, by @alan-baker.
-
Add sections on User Agent State for Privacy gpuweb/gpuweb#1656, by @Kangz.
-
Explainer: editing/clarifications gpuweb/gpuweb#1655, by @kainino0x.
-
Fix description of a variable's reference type gpuweb/gpuweb#1654, by @dneto0.
-
Add the definition of renderable format gpuweb/gpuweb#1653, by @Richard-Yunchao.
-
Describe function calls gpuweb/gpuweb#1651, by @alan-baker.
-
Describe if statements gpuweb/gpuweb#1650, by @alan-baker.
-
Add transpose built-in function gpuweb/gpuweb#1648, by @alan-baker.
-
copyExternalImageToTexture for HTMLVideoElement gpuweb/gpuweb#1647, by @kainino0x.
-
UTF-8 is always a valid encoding for WGSL gpuweb/gpuweb#1646, by @dneto0.
-
Move <Buffer Usage> to a subsection of <Buffer Creation> gpuweb/gpuweb#1645, by @Richard-Yunchao.
-
Added offset, length to GPUCompilationMessage gpuweb/gpuweb#1643, by @toji.
-
Add derivatives on vector. gpuweb/gpuweb#1639, by @Kangz.
-
Explainer: image input gpuweb/gpuweb#1638, by @kainino0x.
-
Explainer: command encoding gpuweb/gpuweb#1637, by @kainino0x.
-
Explainer: initialization gpuweb/gpuweb#1636, by @kainino0x.
-
Require srcFactor=dstFactor="one" for min/max blend op gpuweb/gpuweb#1635, by @kainino0x.
-
Add GPURequestAdapterOptions.allowSoftware gpuweb/gpuweb#1634, by @kainino0x.
-
GPUBlendFactor: remove -component gpuweb/gpuweb#1633, by @kainino0x.
-
Document texture dimension limits in Limit.md explainer gpuweb/gpuweb#1632, by @Richard-Yunchao.
-
Explainer: Add the security and privacy questionnaire gpuweb/gpuweb#1631, by @Kangz.
-
TypeError on usage of enums from unavailable features gpuweb/gpuweb#1628, by @kainino0x.
-
rename setBlendColor -> setBlendConstant gpuweb/gpuweb#1627, by @kainino0x.
-
Specify WGSL as UTF-16-encoded text. gpuweb/gpuweb#1626, by @kdashg.
-
Use a setlike for GPUDevice features gpuweb/gpuweb#1620, by @toji.
-
arrayLength acts on poitner-to-runtime-array gpuweb/gpuweb#1619, by @dneto0.
-
Explainer: Swap Chain gpuweb/gpuweb#1618, by @kainino0x.
-
Make [[currentTexture]] nullable gpuweb/gpuweb#1617, by @kainino0x.
-
Explainer: Multithreading gpuweb/gpuweb#1616, by @kainino0x.
-
Explainer: mini-section on bitflags gpuweb/gpuweb#1615, by @kainino0x.
-
Naming changes for GPUBlendFactor gpuweb/gpuweb#1614, by @kainino0x.
-
WGSL: Fix wording gpuweb/gpuweb#1612, by @dneto0.
-
Vector-matrix arithmetic extended with scalar "splats" gpuweb/gpuweb#1611, by @dneto0.
-
[editorial] Detailed operations section with the rendering algorithm gpuweb/gpuweb#1610, by @kvark.
-
Explainer: better explain the need for error scopes gpuweb/gpuweb#1609, by @kainino0x.
-
Explainer: section on errors, start on device loss gpuweb/gpuweb#1607, by @kainino0x.
-
Rasterization of points, more general wording gpuweb/gpuweb#1606, by @kvark.
-
Add interpolation section and attribute gpuweb/gpuweb#1605, by @alan-baker.
-
[editorial] Separate sections for the vertex processing and primitive assembly gpuweb/gpuweb#1603, by @kvark.
-
Explainer: Use <object> instead of <img> for SVG gpuweb/gpuweb#1602, by @kainino0x.
-
Deploy imgs to gh-pages gpuweb/gpuweb#1601, by @kainino0x.
-
Add the buffer mapping explainer gpuweb/gpuweb#1600, by @Kangz.
-
Replace error_monad_timeline_diagram's PNG with an SVG gpuweb/gpuweb#1599, by @Kangz.
-
Fix the use of clampDepth gpuweb/gpuweb#1598, by @kvark.
-
Explainer: Object validity and destroyedness gpuweb/gpuweb#1595, by @Kangz.
-
redirect wgsl.html to wgsl/ gpuweb/gpuweb#1594, by @kainino0x.
-
WIP: Shader lifecycle, and kinds of errors gpuweb/gpuweb#1593, by @dneto0.
-
WGSL: add Technical Overview gpuweb/gpuweb#1592, by @dneto0.
-
Editorial: Align with Web IDL specification gpuweb/gpuweb#1591, by @autokagami.
-
-
79 pull requests merged.
-
Adding workload simulator as a sample. gpuweb/gpuweb#1664.
-
[CI] Reduce expire field to 10 days gpuweb/gpuweb#1699.
-
Rename constant_id to override gpuweb/gpuweb#1678.
-
wgsl: only atomic modifications are mutually ordered gpuweb/gpuweb#1688.
-
Add num_workgroups built-in variable gpuweb/gpuweb#1684.
-
Clarify type matching for var initializer gpuweb/gpuweb#1685.
-
Add vertex and fragment interface matching rules gpuweb/gpuweb#1683.
-
wgsl: fill out var-and-let gpuweb/gpuweb#1679.
-
Add spec for texture creation gpuweb/gpuweb#1658.
-
Explainer: edit motivation section gpuweb/gpuweb#1677.
-
Update grammar for constexpr type constructors gpuweb/gpuweb#1675.
-
Reorder sections gpuweb/gpuweb#1676.
-
copyExternalImageToTexture for ImageBitmap and canvases gpuweb/gpuweb#1581.
-
Fix hello-cube.html sample gpuweb/gpuweb#1668.
-
Added offset, length to GPUCompilationMessage gpuweb/gpuweb#1643.
-
Explainer: Add a more general introduction / motivation section. gpuweb/gpuweb#1662.
-
Explainer: bit more editing gpuweb/gpuweb#1665.
-
Deploy samples to GitHub pages. gpuweb/gpuweb#1669.
-
Change GPUSupportedFeatures entry type to DOMString gpuweb/gpuweb#1667.
-
Fix a few tiny issues about references gpuweb/gpuweb#1661.
-
Allow trailing commas in the argument lists, calls, and cases gpuweb/gpuweb#1493.
-
Fix uniform struct size requirement gpuweb/gpuweb#1560.
-
Describe function calls gpuweb/gpuweb#1651.
-
Add sections on User Agent State for Privacy gpuweb/gpuweb#1656.
-
[editorial] Remove vector short convenience names gpuweb/gpuweb#1657.
-
Add the definition of renderable format gpuweb/gpuweb#1653.
-
Explainer: editing/clarifications gpuweb/gpuweb#1655.
-
Fix description of a variable's reference type gpuweb/gpuweb#1654.
-
Describe if statements gpuweb/gpuweb#1650.
-
UTF-8 is always a valid encoding for WGSL gpuweb/gpuweb#1646.
-
Add transpose built-in function gpuweb/gpuweb#1648.
-
Move <Buffer Usage> to a subsection of <Buffer Creation> gpuweb/gpuweb#1645.
-
Add GPUSwapChainAlphaMode = opaque/premultiplied gpuweb/gpuweb#1474.
-
Explainer: initialization gpuweb/gpuweb#1636.
-
Explainer: image input gpuweb/gpuweb#1638.
-
Add derivatives on vector. gpuweb/gpuweb#1639.
-
Explainer: command encoding gpuweb/gpuweb#1637.
-
Require srcFactor=dstFactor="one" for min/max blend op gpuweb/gpuweb#1635.
-
GPUBlendFactor: remove -component gpuweb/gpuweb#1633.
-
Explainer: Add the security and privacy questionnaire gpuweb/gpuweb#1631.
-
Document texture dimension limits in Limit.md explainer gpuweb/gpuweb#1632.
-
rename setBlendColor -> setBlendConstant gpuweb/gpuweb#1627.
-
TypeError on usage of enums from unavailable features gpuweb/gpuweb#1628.
-
Explainer: Swap Chain gpuweb/gpuweb#1618.
-
Add adapter.[[current]] gpuweb/gpuweb#1523.
-
Explainer: Multithreading gpuweb/gpuweb#1616.
-
Naming changes for GPUBlendFactor gpuweb/gpuweb#1614.
-
Use a setlike for GPUDevice features gpuweb/gpuweb#1620.
-
Vector-matrix arithmetic extended with scalar "splats" gpuweb/gpuweb#1611.
-
Explainer: mini-section on bitflags gpuweb/gpuweb#1615.
-
Make [[currentTexture]] nullable gpuweb/gpuweb#1617.
-
Add GPUQuerySet state gpuweb/gpuweb#1531.
-
arrayLength acts on poitner-to-runtime-array gpuweb/gpuweb#1619.
-
WGSL: Fix wording gpuweb/gpuweb#1612.
-
Add atomics to WGSL gpuweb/gpuweb#1448.
-
Explainer: better explain the need for error scopes gpuweb/gpuweb#1609.
-
[editorial] Detailed operations section with the rendering algorithm gpuweb/gpuweb#1610.
-
Explainer: section on errors, start on device loss gpuweb/gpuweb#1607.
-
Add the buffer mapping explainer gpuweb/gpuweb#1600.
-
Rasterization of points, more general wording gpuweb/gpuweb#1606.
-
Add WGSL spec writing style guide gpuweb/gpuweb#1585.
-
Add interpolation section and attribute gpuweb/gpuweb#1605.
-
[editorial] Separate sections for the vertex processing and primitive assembly gpuweb/gpuweb#1603.
-
WIP: Shader lifecycle, and kinds of errors gpuweb/gpuweb#1593.
-
WGSL: describe ref ptr gpuweb/gpuweb#1569.
-
WGSL: add Technical Overview gpuweb/gpuweb#1592.
-
Add section for "Constant Identifier Expression" gpuweb/gpuweb#1586.
-
Add ignore builtin function gpuweb/gpuweb#1587.
-
Deploy imgs to gh-pages gpuweb/gpuweb#1601.
-
Explainer: Use <object> instead of <img> for SVG gpuweb/gpuweb#1602.
-
Replace error_monad_timeline_diagram's PNG with an SVG gpuweb/gpuweb#1599.
-
Fix the use of clampDepth gpuweb/gpuweb#1598.
-
Move clampDepth to GPUPrimitiveState. gpuweb/gpuweb#1583.
-
Editorial: Align with Web IDL specification gpuweb/gpuweb#1591.
-
Explainer: Object validity and destroyedness gpuweb/gpuweb#1595.
-
redirect wgsl.html to wgsl/ gpuweb/gpuweb#1594.
-
Specify the primitive clipping and rasterization stages gpuweb/gpuweb#1564.
-
Define 'composite' type, pin down "plain" type. gpuweb/gpuweb#1578.
-
Add Explainer bikeshed document gpuweb/gpuweb#1582.
-
-
13 pull requests closed.
-
Add validation for GPUDevice.createTexture() gpuweb/gpuweb#799.
-
Make GPUSupportedFeatures `setlike<DOMString>` gpuweb/gpuweb#1659.
-
Pluralize GPU.requestAdapters gpuweb/gpuweb#1314.
-
Specify WGSL as UTF-16-encoded text. gpuweb/gpuweb#1626.
-
Allow comma at the end of the parameter list gpuweb/gpuweb#1584.
-
arrayLength acts on pointer-to-runtime-sized-array gpuweb/gpuweb#1330.
-
Make ptr load/store explicit with unary `*`. gpuweb/gpuweb#1455.
-
Add GPURequestAdapterOptions.majorPerformanceCaveat gpuweb/gpuweb#1302.
-
Revert "Add an array access expression for runtime arrays" gpuweb/gpuweb#1490.
-
Add arithmatic-operators section. gpuweb/gpuweb#1575.
-
WGSL: describe pointers gpuweb/gpuweb#1368.
-
Assignable expression gpuweb/gpuweb#1549.
-
Split first few sections into files; tweak to match planned TOC gpuweb/gpuweb#833.
-
-
82 pull requests received 506 new comments.
-
Add the uniformity analysis to the WGSL spec gpuweb/gpuweb#1571 (51 comments by @alan-baker, @kainino0x, @kdashg, @kvark, @pow2clk, @RobinMorisset)
-
copyExternalImageToTexture for ImageBitmap and canvases gpuweb/gpuweb#1581 (34 comments by @kainino0x, @Kangz, @kenrussell, @kvark, @shaoboyan)
-
Add spec for texture creation gpuweb/gpuweb#1658 (29 comments by @kainino0x, @kvark, @Richard-Yunchao)
-
GPUExternalTexture: API for video; WGSL type and functions gpuweb/gpuweb#1666 (25 comments by @dneto0, @kainino0x, @Kangz, @shaoboyan)
-
Adding workload simulator as a sample. gpuweb/gpuweb#1664 (20 comments by @jdarpinian, @kainino0x, @kvark)
-
Add pointer-chaining expressions: ptr.field and ptr[i] gpuweb/gpuweb#1580 (13 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
Add arithmatic-operators section. gpuweb/gpuweb#1575 (13 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
Add interpolation section and attribute gpuweb/gpuweb#1605 (12 comments by @alan-baker, @dneto0, @kvark)
-
Explainer: Add a more general introduction / motivation section. gpuweb/gpuweb#1662 (11 comments by @Kangz, @kvark)
-
Add the definition of renderable format gpuweb/gpuweb#1653 (11 comments by @kainino0x, @kvark, @Richard-Yunchao)
-
Specify WGSL as UTF-16-encoded text. gpuweb/gpuweb#1626 (9 comments by @ben-clayton, @dneto0, @kdashg, @kvark, @litherum)
-
WIP: Shader lifecycle, and kinds of errors gpuweb/gpuweb#1593 (8 comments by @alan-baker, @dneto0, @kainino0x)
-
Add section for "Constant Identifier Expression" gpuweb/gpuweb#1586 (8 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
Add function restrictions gpuweb/gpuweb#1674 (6 comments by @alan-baker, @dneto0, @jrprice, @kdashg, @RobinMorisset)
-
Add GPURequestAdapterOptions.majorPerformanceCaveat gpuweb/gpuweb#1302 (6 comments by @kainino0x, @Kangz, @kvark)
-
Specify the primitive clipping and rasterization stages gpuweb/gpuweb#1564 (6 comments by @kainino0x, @kvark)
-
copyExternalImageToTexture for HTMLVideoElement gpuweb/gpuweb#1647 (5 comments by @kainino0x, @shaoboyan)
-
Allow trailing commas in the argument lists, calls, and cases gpuweb/gpuweb#1493 (5 comments by @alan-baker, @kdashg, @kvark)
-
Add GPUSwapChainAlphaMode = opaque/premultiplied gpuweb/gpuweb#1474 (5 comments by @kainino0x, @Kangz, @kenrussell, @kvark)
-
add atomic type, texture type, store type to storable; adjust rules for var and let. issue #1686 gpuweb/gpuweb#1693 (4 comments by @alan-baker)
-
Add ignore builtin function gpuweb/gpuweb#1587 (4 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
Replace error_monad_timeline_diagram's PNG with an SVG gpuweb/gpuweb#1599 (4 comments by @kainino0x, @Kangz, @kvark)
-
Clarify that CompilationMessage line/pos numbers should be 1-based gpuweb/gpuweb#1700 (3 comments by @kainino0x, @kdashg, @toji)
-
Require srcFactor=dstFactor="one" for min/max blend op gpuweb/gpuweb#1635 (3 comments by @kainino0x, @Kangz)
-
Explainer: Add the security and privacy questionnaire gpuweb/gpuweb#1631 (3 comments by @Kangz, @kvark)
-
rename setBlendColor -> setBlendConstant gpuweb/gpuweb#1627 (3 comments by @kainino0x, @Kangz, @kvark)
-
Vector-matrix arithmetic extended with scalar "splats" gpuweb/gpuweb#1611 (3 comments by @dneto0, @kvark)
-
Explainer: section on errors, start on device loss gpuweb/gpuweb#1607 (3 comments by @kainino0x, @Kangz)
-
Pluralize requestAdapters() and add GPUAdapter.isSoftware gpuweb/gpuweb#1660 (2 comments by @kainino0x, @Kangz)
-
[editorial] Separate sections for the vertex processing and primitive assembly gpuweb/gpuweb#1603 (2 comments by @kainino0x)
-
Document texture dimension limits in Limit.md explainer gpuweb/gpuweb#1632 (1 comment by @kainino0x)
-
Revert "Add an array access expression for runtime arrays" gpuweb/gpuweb#1490 (1 comment by @dneto0)
-
[editorial] Detailed operations section with the rendering algorithm gpuweb/gpuweb#1610 (1 comment by @kvark)
-
Split first few sections into files; tweak to match planned TOC gpuweb/gpuweb#833 (1 comment by @kainino0x)
-
GitHub
gpuweb/gpuweb
-
48 issues opened.
-
Support mix(vec, vec, f32) gpuweb/gpuweb#1588, by @kvark.
-
Uniformity of Private-class global variables gpuweb/gpuweb#1579, by @kvark.
-
Replace the arrow token for functions with a colon gpuweb/gpuweb#1577, by @jespertheend.
-
WebGPU should validate that storageTexture bindings have a single mip level. gpuweb/gpuweb#1576, by @Kangz.
-
Consider having rich math rendering in the spec gpuweb/gpuweb#1573, by @kvark.
-
Equality comparisons on structs gpuweb/gpuweb#1572, by @litherum.
-
Require that vertex shaders return a position? gpuweb/gpuweb#1567, by @jrprice.
-
wgsl: support extended integer arithmetic for add, subtract, multiply gpuweb/gpuweb#1565, by @dneto0.
-
[proposal] Interface Matching gpuweb/gpuweb#1563, by @alan-baker.
-
Removing the need "non-filtering" samplers gpuweb/gpuweb#1562, by @Kangz.
-
Uniform structure sizes have to be aligned to 16 bytes? gpuweb/gpuweb#1558, by @kvark.
-
wgsl: support inter-scope shadowing gpuweb/gpuweb#1556, by @dneto0.
-
editorial: Make a "invocation" a defined term, and cross-link it gpuweb/gpuweb#1555, by @dneto0.
-
add syntax for very short const declarations inside a function gpuweb/gpuweb#1552, by @dneto0.
-
var: infer type from initializer gpuweb/gpuweb#1550, by @dneto0.
-
What is the toDataURL behavior for GPUCanvasContext? gpuweb/gpuweb#1548, by @kdashg.
-
MSL question: can mem_flags be combined in threadgroup_barrier? gpuweb/gpuweb#1547, by @dneto0.
-
Support for line-list width gpuweb/gpuweb#1546, by @jespertheend.
-
Specify what happens to mappings when the buffer is GCed. gpuweb/gpuweb#1544, by @Kangz.
-
[Suggestion] Allow converting format with copyTextureToTexture gpuweb/gpuweb#1543, by @AshleyScirra.
-
update description of shader-output mask gpuweb/gpuweb#1539, by @dneto0.
-
editorial: do we need a definition of "object"? gpuweb/gpuweb#1538, by @dneto0.
-
Taking advantage of Vulkan built-in YCbCr conversion gpuweb/gpuweb#1537, by @kvark.
-
when does applying an attribute to a type result in a different type gpuweb/gpuweb#1534, by @dneto0.
-
Queries spec inconsistent with ErrorConventions.md gpuweb/gpuweb#1532, by @kainino0x.
-
Pointer chaining in WGSL gpuweb/gpuweb#1530, by @kvark.
-
Restrict where entry point IO types can be used gpuweb/gpuweb#1526, by @jrprice.
-
Validation rules and issues of createTexture gpuweb/gpuweb#1522, by @Richard-Yunchao.
-
Support for vector splats (scalar operations on vectors) gpuweb/gpuweb#1520, by @jespertheend.
-
Assignment operators (e.g. `+=`) in wgsl gpuweb/gpuweb#1518, by @jespertheend.
-
Frag output uniformity requirements? gpuweb/gpuweb#1515, by @kdashg.
-
Spec should use [=map/exists=] for optional dictionary members gpuweb/gpuweb#1512, by @kainino0x.
-
Editorial: attributes vs. decorations gpuweb/gpuweb#1510, by @dneto0.
-
[proposal] Location attribute assignment gpuweb/gpuweb#1504, by @alan-baker.
-
alignment must be a power of two gpuweb/gpuweb#1503, by @dneto0.
-
Should array stride be a property of the type? gpuweb/gpuweb#1502, by @kvark.
-
WGSL: add fn void(sink:T) -> void builtin function gpuweb/gpuweb#1499, by @dneto0.
-
Should objects expose the descriptor they were created with? gpuweb/gpuweb#1498, by @kainino0x.
-
Should objects have pointers back to their parents? gpuweb/gpuweb#1497, by @kainino0x.
-
wgsl editorial: fix inconsistent usage: memory location vs. "byte" gpuweb/gpuweb#1491, by @dneto0.
-
[WGSL]: Spec should state that decorations MUST only be applied once gpuweb/gpuweb#1488, by @ben-clayton.
-
GPUAdapterLimits uses GPUSize32 for read only attributes gpuweb/gpuweb#1484, by @smaug----.
-
copyToTexture for color-managed image sources gpuweb/gpuweb#1483, by @kainino0x.
-
Consider copyToTexture interactions with high-bit-depth source images gpuweb/gpuweb#1482, by @kainino0x.
-
Multi-level errors gpuweb/gpuweb#1481, by @kvark.
-
[wgsl] Proposal: Remove pointer out parameters from modf, frexp gpuweb/gpuweb#1480, by @ben-clayton.
-
Document optional depth/stencil formats gpuweb/gpuweb#1479, by @kainino0x.
-
Is the existence of the WGSL legitimate? gpuweb/gpuweb#1475, by @onlymaresia.
-
-
21 issues closed.
-
add syntax for very short const declarations inside a function gpuweb/gpuweb#1552.
-
Replace the arrow token for functions with a colon gpuweb/gpuweb#1577.
-
Follow-up: Using the void type gpuweb/gpuweb#743.
-
wgsl: support inter-scope shadowing gpuweb/gpuweb#1556.
-
[Suggestion] Allow converting format with copyTextureToTexture gpuweb/gpuweb#1543.
-
Considering making `ARROW function_type_decl` optional for void functions gpuweb/gpuweb#1219.
-
[WGSL]: Spec should state that decorations MUST only be applied once gpuweb/gpuweb#1488.
-
Barriers proposal gpuweb/gpuweb#1374.
-
getCurrentTexture has side effects but doesn't sound like it gpuweb/gpuweb#1424.
-
Require storeOp when resolveTarget is specified? gpuweb/gpuweb#1376.
-
Multi-level errors gpuweb/gpuweb#1481.
-
alignment must be a power of two gpuweb/gpuweb#1503.
-
Should array stride be a property of the type? gpuweb/gpuweb#1502.
-
[wgsl] Offset decorations on Structs gpuweb/gpuweb#561.
-
"Multisampled" textures with sampleCount=1? gpuweb/gpuweb#1226.
-
[WGSL] Proposal: Default Struct Layout gpuweb/gpuweb#1393.
-
Forbid --x gpuweb/gpuweb#801.
-
Are values in GPUVertexAttribute.shaderLocation arbitrary? gpuweb/gpuweb#1418.
-
T2B/B2T copies should require whole-subresource copies for depth/stencil gpuweb/gpuweb#1140.
-
GPUAdapterLimits uses GPUSize32 for read only attributes gpuweb/gpuweb#1484.
-
Is the existence of the WGSL legitimate? gpuweb/gpuweb#1475.
-
-
70 issues received 307 new comments.
-
Have "var" as syntactic sugar for explicit pointer refs/derefs gpuweb/gpuweb#1456 (28 comments by @alan-baker, @ben-clayton, @dneto0, @kvark, @litherum, @RobinMorisset)
-
Support for vector splats (scalar operations on vectors) gpuweb/gpuweb#1520 (20 comments by @11mhg, @dneto0, @jespertheend, @kdashg, @kvark)
-
Method of ensuring GPUShaderModules can contain MTLLibraries gpuweb/gpuweb#1064 (17 comments by @frguthmann, @kainino0x, @Kangz, @kvark, @litherum, @magcius, @toji)
-
Can WebGPU canvas alpha be configured? gpuweb/gpuweb#1425 (11 comments by @juj, @kainino0x, @kenrussell, @kvark, @magcius, @RafaelCintron)
-
Restrict where entry point IO types can be used gpuweb/gpuweb#1526 (11 comments by @alan-baker, @dneto0, @jrprice, @kdashg, @kvark)
-
Query API: What is the GPU timestamp unit on Metal? gpuweb/gpuweb#1325 (10 comments by @haoxli, @Kangz, @kvark)
-
Support for arrays of textures gpuweb/gpuweb#822 (10 comments by @jozanza, @kainino0x, @Kangz, @kvark, @litherum)
-
What is the toDataURL behavior for GPUCanvasContext? gpuweb/gpuweb#1548 (8 comments by @kainino0x, @kenrussell, @RafaelCintron, @sunnyps)
-
Zero initializing textures can not be done efficiently gpuweb/gpuweb#1422 (8 comments by @frguthmann, @kainino0x, @kdashg, @kvark)
-
call_statement: should only be able to call a function that yields void? gpuweb/gpuweb#1413 (7 comments by @ben-clayton, @dneto0, @kvark)
-
WGSL: add fn void(sink:T) -> void builtin function gpuweb/gpuweb#1499 (7 comments by @dneto0, @kdashg, @kvark)
-
Validation rules and issues of createTexture gpuweb/gpuweb#1522 (7 comments by @Kangz, @kvark, @Richard-Yunchao)
-
Proposal for a GPUVideoTexture object. gpuweb/gpuweb#1415 (7 comments by @kainino0x, @Kangz, @RafaelCintron, @shaoboyan)
-
Assignment operators (e.g. `+=`) in wgsl gpuweb/gpuweb#1518 (7 comments by @alan-baker, @ben-clayton, @dneto0, @jespertheend, @kvark, @litherum)
-
add syntax for very short const declarations inside a function gpuweb/gpuweb#1552 (6 comments by @dneto0, @kdashg, @kvark)
-
Specify what happens to mappings when the buffer is GCed. gpuweb/gpuweb#1544 (6 comments by @austinEng, @kainino0x, @Kangz, @kvark)
-
when does applying an attribute to a type result in a different type gpuweb/gpuweb#1534 (6 comments by @alan-baker, @dneto0, @kvark)
-
getCurrentTexture has side effects but doesn't sound like it gpuweb/gpuweb#1424 (6 comments by @juj, @Kangz, @kvark, @litherum)
-
"Multisampled" textures with sampleCount=1? gpuweb/gpuweb#1226 (6 comments by @kainino0x, @Kangz, @kvark)
-
WebGPU should validate that storageTexture bindings have a single mip level. gpuweb/gpuweb#1576 (5 comments by @alan-baker, @Kangz, @kvark)
-
Uniform structure sizes have to be aligned to 16 bytes? gpuweb/gpuweb#1558 (5 comments by @alan-baker, @ben-clayton, @dneto0, @kvark)
-
First WebGPU Stable Release / Prognosis ? gpuweb/gpuweb#1276 (5 comments by @jespertheend, @kainino0x, @v-fox)
-
Queries spec inconsistent with ErrorConventions.md gpuweb/gpuweb#1532 (4 comments by @kainino0x, @Kangz, @kvark)
-
Taking advantage of Vulkan built-in YCbCr conversion gpuweb/gpuweb#1537 (4 comments by @Kangz, @kvark, @sfricke-samsung, @shaoboyan)
-
Considering making `ARROW function_type_decl` optional for void functions gpuweb/gpuweb#1219 (4 comments by @dneto0, @kvark, @litherum)
-
GPUAdapterLimits uses GPUSize32 for read only attributes gpuweb/gpuweb#1484 (4 comments by @kainino0x, @Kangz, @kvark)
-
Replace the arrow token for functions with a colon gpuweb/gpuweb#1577 (3 comments by @dneto0, @frzi, @jespertheend)
-
Consider having rich math rendering in the spec gpuweb/gpuweb#1573 (3 comments by @dneto0, @Kangz, @kdashg)
-
[Suggestion] Allow converting format with copyTextureToTexture gpuweb/gpuweb#1543 (3 comments by @AshleyScirra, @kvark)
-
Proposal for importing Web platform images in WebGPU gpuweb/gpuweb#1154 (3 comments by @kainino0x, @Kangz)
-
Should array stride be a property of the type? gpuweb/gpuweb#1502 (3 comments by @dneto0, @kainino0x, @kvark)
-
copyToTexture for color-managed image sources gpuweb/gpuweb#1483 (2 comments by @kainino0x, @kenrussell)
-
Is the existence of the WGSL legitimate? gpuweb/gpuweb#1475 (2 comments by @kainino0x, @mehmetoguzderin)
-
wgsl: support extended integer arithmetic for add, subtract, multiply gpuweb/gpuweb#1565 (1 comment by @alan-baker)
-
Should objects expose the descriptor they were created with? gpuweb/gpuweb#1498 (1 comment by @AshleyScirra)
-
does textureSampleCompare assume level-of-detail 0 in non-fragment shaders? gpuweb/gpuweb#1319 (1 comment by @dneto0)
-
[wgsl] Proposal: Remove pointer out parameters from modf, frexp gpuweb/gpuweb#1480 (1 comment by @dneto0)
-
[WGSL]: Spec should state that decorations MUST only be applied once gpuweb/gpuweb#1488 (1 comment by @dneto0)
-
shaders: equivalent of gl_PointCoord and gl_PointSize don't exist in HLSL gpuweb/gpuweb#332 (1 comment by @potree)
-
wgsl editorial: fix inconsistent usage: memory location vs. "byte" gpuweb/gpuweb#1491 (1 comment by @dneto0)
-
T2B/B2T copies should require whole-subresource copies for depth/stencil gpuweb/gpuweb#1140 (1 comment by @kainino0x)
-
-
65 pull requests opened.
-
Add ignore builtin function gpuweb/gpuweb#1587, by @dneto0.
-
Add section for "Constant Identifier Expression" gpuweb/gpuweb#1586, by @dneto0.
-
Add WGSL spec writing style guide gpuweb/gpuweb#1585, by @dneto0.
-
Allow comma at the end of the parameter list gpuweb/gpuweb#1584, by @kvark.
-
Move clampDepth to GPUPrimitiveState. gpuweb/gpuweb#1583, by @Kangz.
-
Add Explainer bikeshed document gpuweb/gpuweb#1582, by @kainino0x.
-
copyExternalImageToTexture for ImageBitmap and canvases gpuweb/gpuweb#1581, by @kainino0x.
-
Add pointer-chaining expressions: ptr.field and ptr[i] gpuweb/gpuweb#1580, by @dneto0.
-
Define 'composite' type, pin down "plain" type. gpuweb/gpuweb#1578, by @dneto0.
-
Add arithmatic-operators section. gpuweb/gpuweb#1575, by @kdashg.
-
Replace `const` with `let`. gpuweb/gpuweb#1574, by @kdashg.
-
Add the uniformity analysis to the WGSL spec gpuweb/gpuweb#1571, by @RobinMorisset.
-
function return type must be plain (or void) gpuweb/gpuweb#1570, by @dneto0.
-
WGSL: describe ref ptr gpuweb/gpuweb#1569, by @dneto0.
-
Rename "Value types" to "Plain types" gpuweb/gpuweb#1568, by @dneto0.
-
Fix typo in compound statement gpuweb/gpuweb#1566, by @alan-baker.
-
Specify the primitive clipping and rasterization stages gpuweb/gpuweb#1564, by @kvark.
-
Allow inter-scope shadowing gpuweb/gpuweb#1561, by @alan-baker.
-
Fix uniform struct size requirement gpuweb/gpuweb#1560, by @alan-baker.
-
Remove the block attribute on structures gpuweb/gpuweb#1559, by @kvark.
-
Fix grammar to make function return type optional gpuweb/gpuweb#1557, by @dneto0.
-
Clarify discard statement gpuweb/gpuweb#1554, by @dneto0.
-
Fix double "the the" gpuweb/gpuweb#1553, by @dneto0.
-
Fix some examples in the spec gpuweb/gpuweb#1551, by @alan-baker.
-
Assignable expression gpuweb/gpuweb#1549, by @kvark.
-
Overhaul description of types, in the abstract gpuweb/gpuweb#1545, by @dneto0.
-
Define the extent of scopes for various statements gpuweb/gpuweb#1542, by @alan-baker.
-
Disallow specifying the same attribute multiple times gpuweb/gpuweb#1541, by @alan-baker.
-
Fix link to storage-texture section gpuweb/gpuweb#1540, by @dneto0.
-
Fix storage class constraint links gpuweb/gpuweb#1536, by @alan-baker.
-
Fix inconsistency in barrier names gpuweb/gpuweb#1535, by @alan-baker.
-
[wgsl] Remove rule on arrays needing stride decorations to be host-sharable gpuweb/gpuweb#1533, by @ben-clayton.
-
Add GPUQuerySet state gpuweb/gpuweb#1531, by @haoxli.
-
Infer constant types in WGSL gpuweb/gpuweb#1529, by @kvark.
-
Clarify the check on buffer offset in buffer-texture copy with depth stencil formats gpuweb/gpuweb#1528, by @Jiawei-Shao.
-
Add vector constructor from scalar gpuweb/gpuweb#1527, by @dneto0.
-
Fix array [[stride]] rule gpuweb/gpuweb#1525, by @ben-clayton.
-
Refactor attributes gpuweb/gpuweb#1524, by @alan-baker.
-
Add adapter.[[current]] gpuweb/gpuweb#1523, by @kainino0x.
-
Make storeOp required gpuweb/gpuweb#1521, by @kvark.
-
Describe location requirements gpuweb/gpuweb#1519, by @alan-baker.
-
Fix broken vulkan links in Limits.md gpuweb/gpuweb#1517, by @jespertheend.
-
Describe functions, high level gpuweb/gpuweb#1516, by @dneto0.
-
[EDIT] Use `<pre>` instead of markdown for code blocks gpuweb/gpuweb#1514, by @darionco.
-
[EDIT] navigator.gpu gpuweb/gpuweb#1513, by @darionco.
-
Write "Declaration and scope" gpuweb/gpuweb#1511, by @dneto0.
-
Fix greater than descriptions gpuweb/gpuweb#1509, by @amaiorano.
-
[wgsl] Allow `builtin` and `location` as struct member decorations gpuweb/gpuweb#1508, by @jrprice.
-
Fix e2 type for component-wise logical shift right gpuweb/gpuweb#1507, by @amaiorano.
-
Alignment is a power of two. gpuweb/gpuweb#1505, by @dneto0.
-
Add stage restriction to some texture built-ins gpuweb/gpuweb#1501, by @alan-baker.
-
Allow sample functions in integer images gpuweb/gpuweb#1500, by @alan-baker.
-
Added `HTMLCanvasElement.getContext` definition gpuweb/gpuweb#1496, by @darionco.
-
Add enable directive gpuweb/gpuweb#1495, by @dneto0.
-
Add validation rules for Occlusion Query gpuweb/gpuweb#1494, by @haoxli.
-
Allow trailing commas in the argument lists, calls, and cases gpuweb/gpuweb#1493, by @kvark.
-
Rename sample_mask_out to sample_mask gpuweb/gpuweb#1492, by @kvark.
-
Revert "Add an array access expression for runtime arrays" gpuweb/gpuweb#1490, by @dneto0.
-
[wgsl] Update grammar rules for entry point IO gpuweb/gpuweb#1489, by @jrprice.
-
Make read_write access the default for storage buffers gpuweb/gpuweb#1487, by @alan-baker.
-
De-duplicate value testing tables gpuweb/gpuweb#1486, by @alan-baker.
-
Remove [EnforceRange] in GPUAdapterLimits gpuweb/gpuweb#1485, by @kainino0x.
-
Remove GPUDevice.adapter gpuweb/gpuweb#1478, by @kainino0x.
-
Add GPU.[[adapters_active]] gpuweb/gpuweb#1477, by @kainino0x.
-
Describe discard statement gpuweb/gpuweb#1476, by @alan-baker.
-
-
54 pull requests merged.
-
Replace `const` with `let`. gpuweb/gpuweb#1574.
-
Rename "Value types" to "Plain types" gpuweb/gpuweb#1568.
-
function return type must be plain (or void) gpuweb/gpuweb#1570.
-
Remove the requirement to have entry points gpuweb/gpuweb#1340.
-
Remove the void type gpuweb/gpuweb#1460.
-
Describe location requirements gpuweb/gpuweb#1519.
-
Fix typo in compound statement gpuweb/gpuweb#1566.
-
Allow inter-scope shadowing gpuweb/gpuweb#1561.
-
Fix grammar to make function return type optional gpuweb/gpuweb#1557.
-
Fix some examples in the spec gpuweb/gpuweb#1551.
-
Describe functions, high level gpuweb/gpuweb#1516.
-
Infer constant types in WGSL gpuweb/gpuweb#1529.
-
Overhaul description of types, in the abstract gpuweb/gpuweb#1545.
-
Disallow specifying the same attribute multiple times gpuweb/gpuweb#1541.
-
Clarify discard statement gpuweb/gpuweb#1554.
-
Fix double "the the" gpuweb/gpuweb#1553.
-
Define the extent of scopes for various statements gpuweb/gpuweb#1542.
-
Fix link to storage-texture section gpuweb/gpuweb#1540.
-
Rename sample_mask_out to sample_mask gpuweb/gpuweb#1492.
-
Refactor attributes gpuweb/gpuweb#1524.
-
Fix storage class constraint links gpuweb/gpuweb#1536.
-
Fix inconsistency in barrier names gpuweb/gpuweb#1535.
-
[wgsl] Remove rule on arrays needing stride decorations to be host-sharable gpuweb/gpuweb#1533.
-
Add vector constructor from scalar gpuweb/gpuweb#1527.
-
Add enable directive gpuweb/gpuweb#1495.
-
Add workgroupBarrier to WGSL gpuweb/gpuweb#1449.
-
Fix array [[stride]] rule gpuweb/gpuweb#1525.
-
[EDIT] navigator.gpu gpuweb/gpuweb#1513.
-
Make storeOp required gpuweb/gpuweb#1521.
-
[wgsl] Allow `builtin` and `location` as struct member decorations gpuweb/gpuweb#1508.
-
Fix broken vulkan links in Limits.md gpuweb/gpuweb#1517.
-
Write "Declaration and scope" gpuweb/gpuweb#1511.
-
[EDIT] Use `<pre>` instead of markdown for code blocks gpuweb/gpuweb#1514.
-
Added `HTMLCanvasElement.getContext` definition gpuweb/gpuweb#1496.
-
Fix e2 type for component-wise logical shift right gpuweb/gpuweb#1507.
-
Fix greater than descriptions gpuweb/gpuweb#1509.
-
Add validation rules for Occlusion Query gpuweb/gpuweb#1494.
-
Alignment is a power of two. gpuweb/gpuweb#1505.
-
Document trade-off of GPULoadOp "load" gpuweb/gpuweb#1451.
-
Add stage restriction to some texture built-ins gpuweb/gpuweb#1501.
-
Reserve -- and ++ gpuweb/gpuweb#1458.
-
[WGSL] Default struct memory layout gpuweb/gpuweb#1447.
-
Rename frag_coord builtin to position gpuweb/gpuweb#1436.
-
Describe discard statement gpuweb/gpuweb#1476.
-
[wgsl] Update grammar rules for entry point IO gpuweb/gpuweb#1489.
-
Remove GPUDevice.adapter gpuweb/gpuweb#1478.
-
Remove in/out storage classes gpuweb/gpuweb#1437.
-
De-duplicate value testing tables gpuweb/gpuweb#1486.
-
Add an array access expression for runtime arrays gpuweb/gpuweb#1468.
-
[wgsl] Describe memory locations gpuweb/gpuweb#1467.
-
Remove [EnforceRange] in GPUAdapterLimits gpuweb/gpuweb#1485.
-
Add more clarifications in "Depth/stencil formats" gpuweb/gpuweb#1463.
-
Make requestDevice not return null gpuweb/gpuweb#1465.
-
-
7 pull requests closed.
-
Remove the block attribute on structures gpuweb/gpuweb#1559.
-
Support basic inter-scope shadowing. gpuweb/gpuweb#1472.
-
Make read_write access the default for storage buffers gpuweb/gpuweb#1487.
-
Add GPU.[[adapters_active]] gpuweb/gpuweb#1477.
-
Allow sample functions in integer images gpuweb/gpuweb#1500.
-
Remove GPUDevice.{adapter,features,limits} gpuweb/gpuweb#1309.
-
Change `offset` to `span`. gpuweb/gpuweb#1339.
-
-
52 pull requests received 540 new comments.
-
Add the uniformity analysis to the WGSL spec gpuweb/gpuweb#1571 (64 comments by @alan-baker, @dneto0, @kdashg, @litherum, @RobinMorisset)
-
WGSL: describe ref ptr gpuweb/gpuweb#1569 (53 comments by @alan-baker, @ben-clayton, @dneto0, @kdashg, @kvark)
-
[WGSL] Default struct memory layout gpuweb/gpuweb#1447 (53 comments by @alan-baker, @ben-clayton, @dneto0, @kdashg)
-
Specify the primitive clipping and rasterization stages gpuweb/gpuweb#1564 (41 comments by @kainino0x, @Kangz, @kvark)
-
Describe location requirements gpuweb/gpuweb#1519 (25 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
Add validation rules for Occlusion Query gpuweb/gpuweb#1494 (19 comments by @haoxli, @Kangz, @kvark)
-
Overhaul description of types, in the abstract gpuweb/gpuweb#1545 (17 comments by @alan-baker, @dneto0, @kvark)
-
Infer constant types in WGSL gpuweb/gpuweb#1529 (16 comments by @dneto0, @jespertheend, @kainino0x, @kvark, @RobinMorisset)
-
Define the extent of scopes for various statements gpuweb/gpuweb#1542 (16 comments by @alan-baker, @dneto0, @kvark)
-
Added `HTMLCanvasElement.getContext` definition gpuweb/gpuweb#1496 (16 comments by @darionco, @kainino0x, @kvark, @w3cbot)
-
Replace `const` with `let`. gpuweb/gpuweb#1574 (15 comments by @alan-baker, @dneto0, @kainino0x, @kdashg, @kvark)
-
Describe discard statement gpuweb/gpuweb#1476 (11 comments by @alan-baker, @dneto0, @kvark, @litherum)
-
Make read_write access the default for storage buffers gpuweb/gpuweb#1487 (9 comments by @alan-baker, @kdashg, @kvark, @zoddicus)
-
copyExternalImageToTexture for ImageBitmap and canvases gpuweb/gpuweb#1581 (8 comments by @kainino0x, @Kangz, @shaoboyan)
-
Remove the block attribute on structures gpuweb/gpuweb#1559 (7 comments by @alan-baker, @ben-clayton, @Kangz, @kvark)
-
[wgsl] Update grammar rules for entry point IO gpuweb/gpuweb#1489 (7 comments by @dneto0, @jrprice, @kvark)
-
Clarify the check on buffer offset in buffer-texture copy with depth stencil formats gpuweb/gpuweb#1528 (6 comments by @Jiawei-Shao, @kainino0x, @Kangz, @kvark)
-
Add GPUSwapChainAlphaMode = opaque/premultiplied gpuweb/gpuweb#1474 (5 comments by @kainino0x, @kdashg, @kvark)
-
Fix broken vulkan links in Limits.md gpuweb/gpuweb#1517 (5 comments by @jespertheend, @kvark, @w3cbot)
-
Revert "Add an array access expression for runtime arrays" gpuweb/gpuweb#1490 (3 comments by @alan-baker, @kvark)
-
Add pointer-chaining expressions: ptr.field and ptr[i] gpuweb/gpuweb#1580 (2 comments by @alan-baker)
-
Fix e2 type for component-wise logical shift right gpuweb/gpuweb#1507 (2 comments by @ben-clayton, @w3cbot)
-
createBindGroup: Require a superset of the layout's bindings gpuweb/gpuweb#1061 (1 comment by @kvark)
-
gpuweb/cts
-
2 issues opened.
-
webgpu:api,operation,sampling,anisotropy:* uses invalid WGSL gpuweb/cts#513, by @austinEng.
-
-
3 issues closed.
-
webgpu:api,operation,sampling,anisotropy:* uses invalid WGSL gpuweb/cts#513.
-
Remove descriptions from listing.js? gpuweb/cts#391.
-
Add new formats to capability_info gpuweb/cts#467.
-
-
4 issues received 14 new comments.
-
webgpu:api,operation,buffers,map_oom:mappedAtCreation,smaller_getMappedRange might be wrong gpuweb/cts#504 (7 comments by @austinEng, @kainino0x, @Kangz, @lucacasonato)
-
webgpu:api,operation,sampling,anisotropy:* uses invalid WGSL gpuweb/cts#513 (4 comments by @austinEng, @shrekshao)
-
-
36 pull requests opened.
-
WPT: Pass expectations to worker.run gpuweb/cts#515, by @austinEng.
-
Fix wgsl .0 suffix to indicate f32 type gpuweb/cts#514, by @shrekshao.
-
Fix check of `loadWebGPUExpectations` for undefined gpuweb/cts#512, by @austinEng.
-
Clean up TODOs for queries validation gpuweb/cts#511, by @haoxli.
-
Add stubs for new tests and organize a few old ones gpuweb/cts#510, by @kainino0x.
-
Support passing skip/fail test expectations in wpt or by command line gpuweb/cts#509, by @austinEng.
-
capability_checks tests should expect exceptions not validation errors gpuweb/cts#508, by @kainino0x.
-
Don't validate test modules in `makeListing` gpuweb/cts#507, by @austinEng.
-
Adding Context Creation Test for OffscreenCanvas gpuweb/cts#505, by @Juanmihd.
-
add tools/deno runner gpuweb/cts#503, by @lucacasonato.
-
Fix page loading on Safari gpuweb/cts#502, by @kainino0x.
-
Make the intent of the expand/collapse UI clearer using details elements gpuweb/cts#501, by @toji.
-
Updating CTS tests to use the new RenderPipelineDescriptor layout gpuweb/cts#500, by @toji.
-
Remove the description from the listing gpuweb/cts#499, by @austinEng.
-
Implement api,operation,rendering,draw:arguments:* gpuweb/cts#498, by @austinEng.
-
Add a tool to verify tracking spreadsheet gpuweb/cts#497, by @kainino0x.
-
Make reftests work with dev server, add npm start gpuweb/cts#496, by @kainino0x.
-
Clean code for createTexture validation tests gpuweb/cts#494, by @Richard-Yunchao.
-
Add validation tests for setBindGroup gpuweb/cts#493, by @egalli.
-
Break gen_listings dep on ts-node / fast-glob. Add src_dir/out_dir args gpuweb/cts#492, by @austinEng.
-
Add validation tests plan for Occlusion Query gpuweb/cts#491, by @haoxli.
-
Add test for zero sized index buffer out of bounds draw gpuweb/cts#490, by @shrekshao.
-
Split createEncoder finish into three parts for strict validation gpuweb/cts#489, by @kainino0x.
-
Reuse interface in dataBytesForCopyOrOverestimate gpuweb/cts#488, by @kainino0x.
-
Remove unnecessary step from wpt runtime gpuweb/cts#487, by @kainino0x.
-
Deduplicate texture table header gpuweb/cts#486, by @kainino0x.
-
Vertex state validation gpuweb/cts#485, by @Kangz.
-
Add validation tests for bufferSize in B2T T2B copies on depth stenci… gpuweb/cts#484, by @Jiawei-Shao.
-
Validate compressed texture gpuweb/cts#483, by @Richard-Yunchao.
-
Update to webgpu types 0.0.43 gpuweb/cts#482, by @Kangz.
-
Add texture usage validation test for createTexture gpuweb/cts#481, by @Richard-Yunchao.
-
Update plan for rendering/draw.spec.ts and indirect_draw.spec.ts gpuweb/cts#480, by @austinEng.
-
Add texture size tests for 1D/2D/3D textures gpuweb/cts#479, by @Richard-Yunchao.
-
Add sample count validation test for createTexture gpuweb/cts#478, by @Richard-Yunchao.
-
-
37 pull requests merged.
-
WPT: Pass expectations to worker.run gpuweb/cts#515.
-
Fix wgsl .0 suffix to indicate f32 type gpuweb/cts#514.
-
Clean up TODOs for queries validation gpuweb/cts#511.
-
Add validation tests for setBindGroup gpuweb/cts#493.
-
Fix check of `loadWebGPUExpectations` for undefined gpuweb/cts#512.
-
Add stubs for new tests and organize a few old ones gpuweb/cts#510.
-
Implement api,operation,rendering,draw:arguments:* gpuweb/cts#498.
-
Add validation tests plan for Occlusion Query gpuweb/cts#491.
-
Add a tool to verify tracking spreadsheet gpuweb/cts#497.
-
Don't validate test modules in `makeListing` gpuweb/cts#507.
-
Adding Context Creation Test for OffscreenCanvas gpuweb/cts#505.
-
add tools/deno runner gpuweb/cts#503.
-
Fix page loading on Safari gpuweb/cts#502.
-
Clean code for createTexture validation tests gpuweb/cts#494.
-
Make reftests work with dev server, add npm start gpuweb/cts#496.
-
Vertex state validation gpuweb/cts#485.
-
Add texture usage validation test for createTexture gpuweb/cts#481.
-
Reuse interface in dataBytesForCopyOrOverestimate gpuweb/cts#488.
-
Add test for zero sized index buffer out of bounds draw gpuweb/cts#490.
-
Remove unnecessary step from wpt runtime gpuweb/cts#487.
-
Update plan for rendering/draw.spec.ts and indirect_draw.spec.ts gpuweb/cts#480.
-
Validate compressed texture gpuweb/cts#483.
-
Add validation tests for Timestamp Query gpuweb/cts#469.
-
Add sample count validation test for createTexture gpuweb/cts#478.
-
Vertex state validation gpuweb/cts#465.
-
Update to webgpu types 0.0.43 gpuweb/cts#482.
-
Add texture size tests for 1D/2D/3D textures gpuweb/cts#479.
-
Bug fixes for new texture formats gpuweb/cts#477.
-
-
2 pull requests closed.
-
Remove the description from the listing gpuweb/cts#499.
-
Deduplicate texture table header gpuweb/cts#486.
-
-
29 pull requests received 334 new comments.
-
Add texture usage validation test for createTexture gpuweb/cts#481 (32 comments by @kainino0x, @Kangz, @Richard-Yunchao)
-
Support passing skip/fail test expectations in wpt or by command line gpuweb/cts#509 (30 comments by @austinEng, @kainino0x, @lucacasonato)
-
Update plan for rendering/draw.spec.ts and indirect_draw.spec.ts gpuweb/cts#480 (24 comments by @austinEng, @kainino0x)
-
Clean code for createTexture validation tests gpuweb/cts#494 (19 comments by @kainino0x, @Richard-Yunchao)
-
Implement api,operation,rendering,draw:arguments:* gpuweb/cts#498 (17 comments by @austinEng, @kainino0x, @Kangz)
-
Add sample count validation test for createTexture gpuweb/cts#478 (16 comments by @kainino0x, @Richard-Yunchao)
-
Add validation tests on buffer offset in B2T and T2B copies with depth stencil formats gpuweb/cts#506 (15 comments by @Jiawei-Shao, @kainino0x)
-
Break gen_listings dep on ts-node / fast-glob. Add src_dir/out_dir args gpuweb/cts#492 (12 comments by @austinEng, @kainino0x)
-
Add texture size tests for 1D/2D/3D textures gpuweb/cts#479 (11 comments by @kainino0x, @Richard-Yunchao)
-
Add validation tests for bufferSize in B2T T2B copies on depth stenci… gpuweb/cts#484 (10 comments by @Jiawei-Shao, @kainino0x, @Kangz)
-
capability_checks tests should expect exceptions not validation errors gpuweb/cts#508 (3 comments by @kainino0x)
-
Updating CTS tests to use the new RenderPipelineDescriptor layout gpuweb/cts#500 (3 comments by @Kangz, @toji)
-
Add test for zero sized index buffer out of bounds draw gpuweb/cts#490 (3 comments by @kainino0x, @shrekshao)
-
Add validation tests on B2T and T2B copies with depth stencil formats - Part I gpuweb/cts#472 (2 comments by @Jiawei-Shao, @kainino0x)
-
Add validation tests about default values and lower bound for texture… gpuweb/cts#495 (1 comment by @Richard-Yunchao)
Participants
-
4 participants joined.
-
Maciej Stachowiak (Apple Inc.) joined.
-
Theresa O'Connor (Apple Inc.) joined.
-
Ping Wu (Baidu, Inc.) joined.
-
Jiawei Shao (Intel Corporation) joined.
-
-
1 organization joined.
-
Baidu, Inc. joined.
-
GitHub
gpuweb/gpuweb
-
36 issues opened.
-
[wgsl] Intra-scope shadowing? gpuweb/gpuweb#1473, by @kdashg.
-
Recursion in WGSL? gpuweb/gpuweb#1471, by @kdashg.
-
Simplifications and improvements to WGSL syntax gpuweb/gpuweb#1466, by @lain-dono.
-
WGSL pointer aliasing rules gpuweb/gpuweb#1457, by @litherum.
-
Have "var" as syntactic sugar for explicit pointer refs/derefs gpuweb/gpuweb#1456, by @kvark.
-
allow both 0x and 0X as hexadecimal literal prefix gpuweb/gpuweb#1453, by @dneto0.
-
WGSL references a 1D-array texture dimension, but WebGPU does not define one gpuweb/gpuweb#1450, by @zoddicus.
-
HLSL requires the number of samples for a MS texture as part of the type? gpuweb/gpuweb#1445, by @ben-clayton.
-
Does this work on Linux because I can't really get it to work gpuweb/gpuweb#1443, by @romulasry.
-
allow workgroup size components to be pipeline-overrideable constants gpuweb/gpuweb#1442, by @dneto0.
-
Proposal: `GPUAdapter.gpuApiPreferenceRating` gpuweb/gpuweb#1439, by @kdashg.
-
Using x1 in vertex format variants gpuweb/gpuweb#1434, by @kvark.
-
Float hex literal notation? gpuweb/gpuweb#1433, by @kdashg.
-
Float literal scientific notation? gpuweb/gpuweb#1432, by @kdashg.
-
allow array size that is a pipeline-overrideable constant gpuweb/gpuweb#1431, by @dneto0.
-
Can begin/endPipelineStatisticsQuery and writeTimestamp be moved to GPUProgrammablePassEncoder? gpuweb/gpuweb#1430, by @toji.
-
Two Blend-Color Related Blend Factors are Missing gpuweb/gpuweb#1429, by @Lichtso.
-
Can WebGPU canvas alpha be configured? gpuweb/gpuweb#1425, by @kainino0x.
-
getCurrentTexture has side effects but doesn't sound like it gpuweb/gpuweb#1424, by @kainino0x.
-
Zero initializing textures can not be done efficiently gpuweb/gpuweb#1422, by @kvark.
-
access decorations for textures (text/example inconsistency) gpuweb/gpuweb#1419, by @jrprice.
-
Are values in GPUVertexAttribute.shaderLocation arbitrary? gpuweb/gpuweb#1418, by @Kangz.
-
How to convert mat3(mvi) in GLSL into WGSL? gpuweb/gpuweb#1417, by @jack1232.
-
Proposal for a GPUVideoTexture object. gpuweb/gpuweb#1415, by @Kangz.
-
Possibility to optimize global app-specific and frame-specific state binding? gpuweb/gpuweb#1414, by @juj.
-
call_statement: should only be able to call a function that yields void? gpuweb/gpuweb#1413, by @dneto0.
-
call_statement vs. a call expression gpuweb/gpuweb#1412, by @dneto0.
-
WGSL: limit types that can go into arrays and structs gpuweb/gpuweb#1409, by @dneto0.
-
pr-preview doesn't run for WGSL spec PRs gpuweb/gpuweb#1405, by @kdashg.
-
Mechanism for opting into shading language extensions gpuweb/gpuweb#1404, by @litherum.
-
`MissingBlockDecoration` in a glsl shader. What does that mean? gpuweb/gpuweb#1401, by @nikitablack.
-
Consider making some optional members also nullable gpuweb/gpuweb#1400, by @kainino0x.
-
Require [[block]] on nested structures in WGSL gpuweb/gpuweb#1398, by @kvark.
-
Missing specification for GPUTexture.createView gpuweb/gpuweb#1397, by @Kangz.
-
Proposal: turn "const" at global level into a real storage class gpuweb/gpuweb#1394, by @kvark.
-
-
23 issues closed.
-
Using x1 in vertex format variants gpuweb/gpuweb#1434.
-
Release WebGPU 1.0.0.0 Specification gpuweb/gpuweb#1000.
-
Which storage classes do we need? gpuweb/gpuweb#1105.
-
Query size of texture gpuweb/gpuweb#1107.
-
Sampling non-float textures is forbidden in HLSL gpuweb/gpuweb#1162.
-
Proposal to enhance defining input/output variables. gpuweb/gpuweb#1155.
-
allow workgroup size components to be pipeline-overrideable constants gpuweb/gpuweb#1442.
-
HLSL requires the number of samples for a MS texture as part of the type? gpuweb/gpuweb#1445.
-
WGSL references a 1D-array texture dimension, but WebGPU does not define one gpuweb/gpuweb#1450.
-
Proposal: Support C99 hex float literals to accurately capture float values gpuweb/gpuweb#709.
-
Does this work on Linux because I can't really get it to work gpuweb/gpuweb#1443.
-
WGSL: limit types that can go into arrays and structs gpuweb/gpuweb#1409.
-
Provide shader inputs as arguments to entry point functions gpuweb/gpuweb#1315.
-
Is Input/Output access one-way? gpuweb/gpuweb#1113.
-
Float hex literal notation? gpuweb/gpuweb#1433.
-
Remove offset. gpuweb/gpuweb#1349.
-
Replace [[offset]] with [[span]]? gpuweb/gpuweb#1303.
-
How to convert mat3(mvi) in GLSL into WGSL? gpuweb/gpuweb#1417.
-
Define how .5 rounds when using the `round` method. gpuweb/gpuweb#1381.
-
Missing specification for GPUTexture.createView gpuweb/gpuweb#1397.
-
`MissingBlockDecoration` in a glsl shader. What does that mean? gpuweb/gpuweb#1401.
-
Typo in storage write restriction in vertex bindings gpuweb/gpuweb#1338.
-
-
68 issues received 352 new comments.
-
Can WebGPU canvas alpha be configured? gpuweb/gpuweb#1425 (31 comments by @greggman, @juj, @kainino0x, @Kangz, @kenrussell, @kvark, @magcius, @mstange, @RafaelCintron)
-
[WGSL] Proposal: Default Struct Layout gpuweb/gpuweb#1393 (23 comments by @ben-clayton, @dneto0, @kainino0x, @kvark, @litherum)
-
Method of ensuring GPUShaderModules can contain MTLLibraries gpuweb/gpuweb#1064 (22 comments by @dneto0, @kainino0x, @Kangz, @kenrussell, @kvark, @litherum, @magcius, @toji)
-
Have "var" as syntactic sugar for explicit pointer refs/derefs gpuweb/gpuweb#1456 (22 comments by @dneto0, @kainino0x, @kdashg, @kvark, @litherum)
-
vec4<f32> is too hard to type on the keyboard gpuweb/gpuweb#736 (20 comments by @aloucks, @dj2, @dneto0, @frzi, @juj, @kainino0x, @kdashg, @kvark, @lain-dono, @litherum, @parasyte)
-
getCurrentTexture has side effects but doesn't sound like it gpuweb/gpuweb#1424 (16 comments by @juj, @kainino0x, @Kangz, @kvark, @litherum, @toji)
-
Proposal for a GPUVideoTexture object. gpuweb/gpuweb#1415 (13 comments by @kainino0x, @Kangz, @kdashg, @kvark, @litherum, @shaoboyan)
-
Are values in GPUVertexAttribute.shaderLocation arbitrary? gpuweb/gpuweb#1418 (11 comments by @kainino0x, @Kangz, @kvark, @litherum)
-
Mechanism for opting into shading language extensions gpuweb/gpuweb#1404 (11 comments by @dneto0, @kvark, @litherum)
-
Development-only API features gpuweb/gpuweb#814 (11 comments by @kainino0x, @Kangz, @kdashg, @kvark, @litherum)
-
[WGSL] WGSL needs a plan for expansion gpuweb/gpuweb#600 (10 comments by @ben-clayton, @dneto0, @kdashg, @kvark, @litherum)
-
does textureSampleCompare assume level-of-detail 0 in non-fragment shaders? gpuweb/gpuweb#1319 (9 comments by @dneto0, @kvark, @litherum)
-
Proposal: turn "const" at global level into a real storage class gpuweb/gpuweb#1394 (9 comments by @dneto0, @kvark)
-
[wgsl] Define interpolation decorations/qualifiers gpuweb/gpuweb#802 (7 comments by @alan-baker, @dneto0, @kvark)
-
Require [[block]] on nested structures in WGSL gpuweb/gpuweb#1398 (6 comments by @dneto0, @kvark, @litherum)
-
allow workgroup size components to be pipeline-overrideable constants gpuweb/gpuweb#1442 (6 comments by @dneto0, @kvark, @tex3d)
-
spell out when an implementation must reject a shader module gpuweb/gpuweb#1241 (5 comments by @dneto0, @kainino0x, @kvark)
-
Enforce calling GPUSwapChain.getCurrentTexture() should return the same binding within same event callback gpuweb/gpuweb#1388 (5 comments by @juj, @kainino0x, @Kangz, @toji)
-
Investigation: Compressed Texture Formats gpuweb/gpuweb#144 (4 comments by @alecazam, @kainino0x, @magcius)
-
Simplifications and improvements to WGSL syntax gpuweb/gpuweb#1466 (4 comments by @dj2, @kainino0x, @kvark)
-
HLSL requires the number of samples for a MS texture as part of the type? gpuweb/gpuweb#1445 (4 comments by @ben-clayton, @dj2, @kvark, @magcius)
-
Two Blend-Color Related Blend Factors are Missing gpuweb/gpuweb#1429 (4 comments by @Kangz, @Lichtso, @litherum)
-
Require storeOp when resolveTarget is specified? gpuweb/gpuweb#1376 (4 comments by @kainino0x, @kvark, @litherum)
-
Does this work on Linux because I can't really get it to work gpuweb/gpuweb#1443 (3 comments by @dneto0, @romulasry)
-
Can begin/endPipelineStatisticsQuery and writeTimestamp be moved to GPUProgrammablePassEncoder? gpuweb/gpuweb#1430 (3 comments by @kainino0x, @Kangz)
-
Possibility to optimize global app-specific and frame-specific state binding? gpuweb/gpuweb#1414 (3 comments by @juj, @kainino0x, @kvark)
-
The order of top-level declarations should be irrelevant gpuweb/gpuweb#875 (2 comments by @kvark, @litherum)
-
scientific notation is not permitted in float literals (and A-5 problem) gpuweb/gpuweb#775 (2 comments by @dneto0)
-
call_statement: should only be able to call a function that yields void? gpuweb/gpuweb#1413 (2 comments by @dneto0, @kvark)
-
Investigation: kill vs demoteToHelper behavior in native APIs gpuweb/gpuweb#921 (2 comments by @dneto0, @kvark)
-
Contradictions in the validation rules of copyTextureToTexture() with depth stencil formats gpuweb/gpuweb#1464 (1 comment by @kainino0x)
-
allow array size that is a pipeline-overrideable constant gpuweb/gpuweb#1431 (1 comment by @litherum)
-
[wgsl] Define the interface of an entry point in a WGSL program gpuweb/gpuweb#774 (1 comment by @kdashg)
-
Provide shader inputs as arguments to entry point functions gpuweb/gpuweb#1315 (1 comment by @kdashg)
-
`MissingBlockDecoration` in a glsl shader. What does that mean? gpuweb/gpuweb#1401 (1 comment by @Kangz)
-
-
39 pull requests opened.
-
Add GPUSwapChainAlphaMode = opaque/premultiplied gpuweb/gpuweb#1474, by @kdashg.
-
Support basic inter-scope shadowing. gpuweb/gpuweb#1472, by @kdashg.
-
Add /* */ block comments. gpuweb/gpuweb#1470, by @kdashg.
-
Rename single element vertex formats to remove the x1 postfix gpuweb/gpuweb#1469, by @toji.
-
Add an array access expression for runtime arrays gpuweb/gpuweb#1468, by @alan-baker.
-
[wgsl] Describe memory locations gpuweb/gpuweb#1467, by @alan-baker.
-
Make requestDevice not return null gpuweb/gpuweb#1465, by @kainino0x.
-
Add more clarifications in "Depth/stencil formats" gpuweb/gpuweb#1463, by @Jiawei-Shao.
-
Make Justin a former editor gpuweb/gpuweb#1462, by @kvark.
-
Set dsinclair as a former editor for wgsl gpuweb/gpuweb#1461, by @dj2.
-
Remove the void type gpuweb/gpuweb#1460, by @kvark.
-
Introduce Subgroup Operations Extension gpuweb/gpuweb#1459, by @mehmetoguzderin.
-
Reserve -- and ++ gpuweb/gpuweb#1458, by @dneto0.
-
Make ptr load/store explicit with unary `*`. gpuweb/gpuweb#1455, by @kdashg.
-
Rewrite explanation of number tokens gpuweb/gpuweb#1454, by @dneto0.
-
WGSL: remove 1d_array textures gpuweb/gpuweb#1452, by @dneto0.
-
Document trade-off of GPULoadOp "load" gpuweb/gpuweb#1451, by @kainino0x.
-
Add workgroupBarrier to WGSL gpuweb/gpuweb#1449, by @alan-baker.
-
Add atomics to WGSL gpuweb/gpuweb#1448, by @alan-baker.
-
[WGSL] Default struct memory layout gpuweb/gpuweb#1447, by @ben-clayton.
-
chore: add comma gpuweb/gpuweb#1446, by @ahaoboy.
-
Workgroup size dimensions can be pipeline-overrideable constants gpuweb/gpuweb#1444, by @dneto0.
-
Make GPUExtent3DDict's width member required gpuweb/gpuweb#1441, by @foolip.
-
Make dictionary arguments optional with default value gpuweb/gpuweb#1440, by @foolip.
-
Remove STRING_LITERAL gpuweb/gpuweb#1438, by @dneto0.
-
Remove in/out storage classes gpuweb/gpuweb#1437, by @kvark.
-
Rename frag_coord builtin to position gpuweb/gpuweb#1436, by @kvark.
-
Add dneto@google.com as WGSL editor gpuweb/gpuweb#1435, by @dneto0.
-
Entry point arguments and return values gpuweb/gpuweb#1426, by @kvark.
-
Restrict GPUVertexAttribute.shaderLocation < maxAttribs gpuweb/gpuweb#1423, by @Kangz.
-
Add support for hex float literals gpuweb/gpuweb#1420, by @alan-baker.
-
Further improvements to createView() gpuweb/gpuweb#1416, by @toji.
-
Restrict contents of array and struct gpuweb/gpuweb#1411, by @dneto0.
-
indexing is 0-based, and array size is at least 1 gpuweb/gpuweb#1410, by @dneto0.
-
Better describe how swap chains update the canvas gpuweb/gpuweb#1408, by @toji.
-
WGSL: Explain vector and matrix gpuweb/gpuweb#1407, by @dneto0.
-
Add algorithm for GPUTexture.createView() gpuweb/gpuweb#1406, by @toji.
-
WGSL: round break ties toward even gpuweb/gpuweb#1403, by @dneto0.
-
Fix "storage" restriction for VERTEX visible bindings gpuweb/gpuweb#1399, by @toji.
-
-
25 pull requests merged.
-
Rename single element vertex formats to remove the x1 postfix gpuweb/gpuweb#1469.
-
Make Justin a former editor gpuweb/gpuweb#1462.
-
Set dsinclair as a former editor for wgsl gpuweb/gpuweb#1461.
-
Workgroup size dimensions can be pipeline-overrideable constants gpuweb/gpuweb#1444.
-
WGSL: remove 1d_array textures gpuweb/gpuweb#1452.
-
Add support for hex float literals gpuweb/gpuweb#1420.
-
Make GPUExtent3DDict's width member required gpuweb/gpuweb#1441.
-
chore: add comma gpuweb/gpuweb#1446.
-
WGSL: Explain vector and matrix gpuweb/gpuweb#1407.
-
Remove STRING_LITERAL gpuweb/gpuweb#1438.
-
Restrict contents of array and struct gpuweb/gpuweb#1411.
-
indexing is 0-based, and array size is at least 1 gpuweb/gpuweb#1410.
-
Add dneto@google.com as WGSL editor gpuweb/gpuweb#1435.
-
Entry point arguments and return values gpuweb/gpuweb#1426.
-
Rename GPUVertexFormat members to match GPUTextureFormat gpuweb/gpuweb#1322.
-
Further improvements to createView() gpuweb/gpuweb#1416.
-
Restrict GPUVertexAttribute.shaderLocation < maxAttribs gpuweb/gpuweb#1423.
-
Rename GPUExtent3DDict.depth to .depthOrArrayLayers gpuweb/gpuweb#1390.
-
Better describe how swap chains update the canvas gpuweb/gpuweb#1408.
-
WGSL: round break ties toward even gpuweb/gpuweb#1403.
-
Add algorithm for GPUTexture.createView() gpuweb/gpuweb#1406.
-
Add type rules for composite value decomposition gpuweb/gpuweb#1370.
-
Change [Buffer/Texture/ImageBitmap]CopyView to "ImageCopy" names gpuweb/gpuweb#1375.
-
Fix "storage" restriction for VERTEX visible bindings gpuweb/gpuweb#1399.
-
Remove obsolete note about matrix layout gpuweb/gpuweb#1386.
-
-
7 pull requests closed.
-
Introduce Subgroup Operations Extension gpuweb/gpuweb#954.
-
Add GPULoadOp "zero" gpuweb/gpuweb#1389.
-
Make dictionary arguments optional with default value gpuweb/gpuweb#1440.
-
[wgsl] Fix type constructor matching in primary_expression gpuweb/gpuweb#1382.
-
Make the default value of addressMode{U,V,W} align with WebGL gpuweb/gpuweb#1297.
-
Replace [[block]] attribute by [[layout(...)]] gpuweb/gpuweb#1361.
-
Add validation steps for createTexture and createTextureView. gpuweb/gpuweb#1237.
-
-
45 pull requests received 295 new comments.
-
[WGSL] Default struct memory layout gpuweb/gpuweb#1447 (40 comments by @alan-baker, @ben-clayton, @dj2, @kvark)
-
Better describe how swap chains update the canvas gpuweb/gpuweb#1408 (19 comments by @juj, @kainino0x, @Kangz, @kvark, @toji)
-
Add algorithm for GPUTexture.createView() gpuweb/gpuweb#1406 (17 comments by @kainino0x, @Kangz, @kvark, @toji)
-
Replace [[block]] attribute by [[layout(...)]] gpuweb/gpuweb#1361 (13 comments by @dneto0, @kainino0x, @kvark, @litherum)
-
Rewrite explanation of number tokens gpuweb/gpuweb#1454 (12 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
Rename GPUVertexFormat members to match GPUTextureFormat gpuweb/gpuweb#1322 (11 comments by @kainino0x, @Kangz, @kvark, @litherum, @toji)
-
Remove the requirement to have entry points gpuweb/gpuweb#1340 (10 comments by @dj2, @dneto0, @kainino0x, @kdashg, @kvark, @litherum)
-
Add an array access expression for runtime arrays gpuweb/gpuweb#1468 (7 comments by @alan-baker, @kvark)
-
Require [SecureContext] for using WebGPU. gpuweb/gpuweb#1363 (6 comments by @annevk, @christhompson, @kainino0x, @Kangz, @litherum)
-
Make ptr load/store explicit with unary `*`. gpuweb/gpuweb#1455 (5 comments by @alan-baker, @dneto0, @kdashg, @kvark)
-
Rename GPUExtent3DDict.depth to .depthOrArrayLayers gpuweb/gpuweb#1390 (5 comments by @kainino0x, @Kangz, @kvark)
-
Change [Buffer/Texture/ImageBitmap]CopyView to "ImageCopy" names gpuweb/gpuweb#1375 (5 comments by @kainino0x, @kvark, @litherum, @toji)
-
Workgroup size dimensions can be pipeline-overrideable constants gpuweb/gpuweb#1444 (4 comments by @dneto0, @kvark, @litherum)
-
Make dictionary arguments optional with default value gpuweb/gpuweb#1440 (4 comments by @foolip, @kvark)
-
Add GPURequestAdapterOptions.majorPerformanceCaveat gpuweb/gpuweb#1302 (4 comments by @kainino0x, @kdashg, @kvark, @litherum)
-
Add more clarifications in "Depth/stencil formats" gpuweb/gpuweb#1463 (3 comments by @kainino0x, @Kangz)
-
Add validation steps for createTexture and createTextureView. gpuweb/gpuweb#1237 (3 comments by @kainino0x, @Kangz)
-
indexing is 0-based, and array size is at least 1 gpuweb/gpuweb#1410 (2 comments by @dneto0, @kvark)
-
Make the default value of addressMode{U,V,W} align with WebGL gpuweb/gpuweb#1297 (1 comment by @Kangz)
-
createBindGroup: Require a superset of the layout's bindings gpuweb/gpuweb#1061 (1 comment by @Kangz)
-
gpuweb/cts
-
2 issues opened.
-
Add new formats to capability_info gpuweb/cts#467, by @kainino0x.
-
Test that the alpha:true/false of canvas is honored gpuweb/cts#460, by @Kangz.
-
-
1 issue closed.
-
3 issues received 6 new comments.
-
Test that the alpha:true/false of canvas is honored gpuweb/cts#460 (2 comments by @kainino0x, @Kangz)
-
26 pull requests opened.
-
Bug fixes for new texture formats gpuweb/cts#477, by @Richard-Yunchao.
-
Update to types 0.0.41 gpuweb/cts#476, by @kainino0x.
-
Fix incorrect indexing bug in resource_init/check_texture/by_sampling.ts gpuweb/cts#474, by @austinEng.
-
update webgpu/types to 0.0.38 gpuweb/cts#473, by @Richard-Yunchao.
-
Make ValidationTest.createEncoder work on any union of EncoderTypes gpuweb/cts#471, by @kainino0x.
-
Remove babel transforms for ?? .? &&= ||= gpuweb/cts#470, by @kainino0x.
-
Add validation tests for Timestamp Query gpuweb/cts#469, by @haoxli.
-
Fix copyBetweenLinearDataAndTexture operation tests gpuweb/cts#468, by @kainino0x.
-
Some more cases for index_access.spec.ts gpuweb/cts#466, by @shrekshao.
-
Vertex state validation gpuweb/cts#465, by @Kangz.
-
Remove explicit storage class from samplers/textures gpuweb/cts#464, by @jrprice.
-
WIP image copy operation test work gpuweb/cts#463, by @kainino0x.
-
partially update image copy operation tests; assorted small refactors gpuweb/cts#462, by @kainino0x.
-
Implement some of api:validation:vertex_state gpuweb/cts#461, by @Kangz.
-
Add operation tests for GPUQueue.writeBuffer() gpuweb/cts#459, by @egalli.
-
Update the test plan for validation/vertex_state gpuweb/cts#458, by @Kangz.
-
Add button to print creation stack to console in /standalone/ gpuweb/cts#457, by @egalli.
-
Update wgsl comments from # to // gpuweb/cts#456, by @shrekshao.
-
Add validation tests for mipLevels for texture creation gpuweb/cts#455, by @Richard-Yunchao.
-
fix "INFO: subcase:" prints gpuweb/cts#454, by @kainino0x.
-
Use subcases in copyImageBitmapToTexture tests gpuweb/cts#453, by @kainino0x.
-
mipLevelCount and arrayLayerCount should not be 0 gpuweb/cts#451, by @Richard-Yunchao.
-
add todos in createView tests gpuweb/cts#450, by @kainino0x.
-
-
24 pull requests merged.
-
Update to types 0.0.41 gpuweb/cts#476.
-
Fix copyBetweenLinearDataAndTexture operation tests gpuweb/cts#468.
-
update webgpu/types to 0.0.38 gpuweb/cts#473.
-
Add operation tests for GPUQueue.writeBuffer() gpuweb/cts#459.
-
Remove babel transforms for ?? .? &&= ||= gpuweb/cts#470.
-
Some more cases for index_access.spec.ts gpuweb/cts#466.
-
Add validation tests for mipLevels for texture creation gpuweb/cts#455.
-
Implement some of api:validation:vertex_state gpuweb/cts#461.
-
Remove explicit storage class from samplers/textures gpuweb/cts#464.
-
Add button to print creation stack to console in /standalone/ gpuweb/cts#457.
-
Update the test plan for validation/vertex_state gpuweb/cts#458.
-
Update wgsl comments from # to // gpuweb/cts#456.
-
Use subcases in copyImageBitmapToTexture tests gpuweb/cts#453.
-
fix "INFO: subcase:" prints gpuweb/cts#454.
-
Loosen precision in multisample resolve operation test gpuweb/cts#446.
-
Add sampler anisotropy tests gpuweb/cts#437.
-
add todos in createView tests gpuweb/cts#450.
-
-
2 pull requests closed.
-
WIP image copy operation test work gpuweb/cts#463.
-
mipLevelCount and arrayLayerCount should not be 0 gpuweb/cts#451.
-
-
24 pull requests received 266 new comments.
-
Add validation tests for mipLevels for texture creation gpuweb/cts#455 (77 comments by @austinEng, @kainino0x, @Richard-Yunchao)
-
Add operation tests for GPUQueue.writeBuffer() gpuweb/cts#459 (26 comments by @egalli, @kainino0x, @Kangz)
-
update webgpu/types to 0.0.40 in order to add missing texture formats gpuweb/cts#475 (25 comments by @austinEng, @kainino0x, @Richard-Yunchao)
-
Fix copyBetweenLinearDataAndTexture operation tests gpuweb/cts#468 (17 comments by @kainino0x, @Kangz)
-
mipLevelCount and arrayLayerCount should not be 0 gpuweb/cts#451 (12 comments by @kainino0x, @Kangz, @Richard-Yunchao)
-
Some more cases for index_access.spec.ts gpuweb/cts#466 (10 comments by @austinEng, @kainino0x, @shrekshao)
-
partially update image copy operation tests; assorted small refactors gpuweb/cts#462 (8 comments by @kainino0x, @Kangz)
-
Add validation tests on B2T and T2B copies with depth stencil formats - Part I gpuweb/cts#472 (5 comments by @Jiawei-Shao, @kainino0x, @Kangz)
-
Update copyBetweenLinearDataAndTexture tests for new bytesPerRow/rowsPerImage rules gpuweb/cts#452 (4 comments by @kainino0x, @Kangz)
-
Make ValidationTest.createEncoder work on any union of EncoderTypes gpuweb/cts#471 (3 comments by @egalli)
-
Add button to print creation stack to console in /standalone/ gpuweb/cts#457 (1 comment by @kainino0x)
-
Update defaultQueue->queue, OUTPUT_ATTACHMENT->RENDER_ATTACHMENT to match latest spec gpuweb/cts#449 (1 comment by @toji)
-
Participants
-
1 participant joined.
-
Narifumi Iwamoto (Intel Corporation) joined.
-
-
1 participant left.
-
Ping Wu (Baidu, Inc.) left.
-
GitHub
gpuweb/gpuweb
-
30 issues opened.
-
[WGSL] Proposal: Default Struct Layout gpuweb/gpuweb#1393, by @ben-clayton.
-
WGSL type rules need to say something about aliases, and about attributes gpuweb/gpuweb#1387, by @dneto0.
-
WGSL is irrational gpuweb/gpuweb#1385, by @jpryne.
-
Define how .5 rounds when using the `round` method. gpuweb/gpuweb#1381, by @dj2.
-
Investigation: Import VideoFrame from WebCodec to WebGPU gpuweb/gpuweb#1380, by @shaoboyan.
-
Support boolean varyings gpuweb/gpuweb#1379, by @kvark.
-
Named parameters for texture functions gpuweb/gpuweb#1378, by @ben-clayton.
-
Reconsider GPULoadOp "clear"/"zero" gpuweb/gpuweb#1377, by @kainino0x.
-
Require storeOp when resolveTarget is specified? gpuweb/gpuweb#1376, by @kainino0x.
-
Barriers proposal gpuweb/gpuweb#1374, by @alan-baker.
-
Limit for the maximum buffer size gpuweb/gpuweb#1371, by @kvark.
-
Typo on Division operation in the specs gpuweb/gpuweb#1365, by @baptistemanson.
-
Feedback after porting a few hundred GLSL lines to WGSL gpuweb/gpuweb#1364, by @baptistemanson.
-
Atomics proposal gpuweb/gpuweb#1360, by @alan-baker.
-
beginRenderPass should validate that the attachment don't alias gpuweb/gpuweb#1358, by @Kangz.
-
GPUVertexAttributeDescriptor.offset validation gpuweb/gpuweb#1357, by @andrewvarga.
-
vector of int/float -> vector of bool casts gpuweb/gpuweb#1355, by @litherum.
-
DrawIndirectCount gpuweb/gpuweb#1354, by @frustum.
-
Consider marking feature-dependent APIs with custom IDL Extended Attributes gpuweb/gpuweb#1353, by @toji.
-
Remove offset. gpuweb/gpuweb#1349, by @dj2.
-
[wgsl]: textureDimensions() returns vec3 for cube textures gpuweb/gpuweb#1345, by @ben-clayton.
-
Missing limits in GPULimits gpuweb/gpuweb#1343, by @Kangz.
-
[wgsl] Consider using multiple scalars to construct matrix. gpuweb/gpuweb#1342, by @shaoboyan.
-
Typo in storage write restriction in vertex bindings gpuweb/gpuweb#1338, by @kainino0x.
-
Removing `ptr` and adding `ref` gpuweb/gpuweb#1334, by @dj2.
-
getExecutionTime requires encoding commands into an otherwise finished command buffer gpuweb/gpuweb#1331, by @kvark.
-
arrayLength should operate on pointer-to-runtime-sized-array instead of runtime-array gpuweb/gpuweb#1329, by @dneto0.
-
Maximum values of Texture Dimension gpuweb/gpuweb#1327, by @Richard-Yunchao.
-
-
29 issues closed.
-
vector of int/float -> vector of bool casts gpuweb/gpuweb#1355.
-
Placement of read_only attributes gpuweb/gpuweb#1159.
-
WGSL is irrational gpuweb/gpuweb#1385.
-
update IO-shareable to include bool gpuweb/gpuweb#1244.
-
Consider marking feature-dependent APIs with custom IDL Extended Attributes gpuweb/gpuweb#1353.
-
Typo on Division operation in the specs gpuweb/gpuweb#1365.
-
conflict between mat2x2 layout in uniform buffer and storage buffer gpuweb/gpuweb#1258.
-
Scalar -> Bool casts gpuweb/gpuweb#753.
-
GPUVertexAttributeDescriptor.offset validation gpuweb/gpuweb#1357.
-
Workgroup Size builtin decoration restrictions. gpuweb/gpuweb#750.
-
WGSL is missing a `sample_mask` builtin. gpuweb/gpuweb#1259.
-
Texture views cause unnecessary memory bandwidth use gpuweb/gpuweb#744.
-
Expose whether WebGPU is supported or not gpuweb/gpuweb#132.
-
add sample_index and sample_mask builtin variables gpuweb/gpuweb#1317.
-
Support for outerProduct() gpuweb/gpuweb#1286.
-
Query API: Maximum limit on query count gpuweb/gpuweb#1324.
-
Portability concern about adapter extensions and limits gpuweb/gpuweb#684.
-
Maximum values of Texture Dimension gpuweb/gpuweb#1327.
-
Relationship between setVertexBuffers() and setBindGroup() is unclear gpuweb/gpuweb#145.
-
Resource Round-Robining gpuweb/gpuweb#38.
-
Case for existing GLSL on the Web gpuweb/gpuweb#219.
-
STORAGE_READ usages for buffers and textures gpuweb/gpuweb#541.
-
Write-only storages? gpuweb/gpuweb#542.
-
Disallow fragment atomic ops on texture storages gpuweb/gpuweb#728.
-
Separate linear sampler binding type? gpuweb/gpuweb#1034.
-
Primitive cut values for non-strip topologies gpuweb/gpuweb#1220.
-
-
64 issues received 271 new comments.
-
Query API: What is the GPU timestamp unit on Metal? gpuweb/gpuweb#1325 (33 comments by @cwfitzgerald, @haoxli, @kainino0x, @Kangz, @kvark, @litherum, @potree)
-
Feedback after porting a few hundred GLSL lines to WGSL gpuweb/gpuweb#1364 (14 comments by @baptistemanson, @dj2, @kainino0x, @kvark, @litherum)
-
Investigation: Validation about 2d-array texture view upon 3D texture and vice versa, and more gpuweb/gpuweb#1372 (11 comments by @kainino0x, @Kangz, @kvark, @Richard-Yunchao)
-
does textureSampleCompare assume level-of-detail 0 in non-fragment shaders? gpuweb/gpuweb#1319 (11 comments by @ben-clayton, @dneto0, @kdashg, @kvark, @pow2clk)
-
GPUVertexAttributeDescriptor.offset validation gpuweb/gpuweb#1357 (11 comments by @andrewvarga, @kainino0x, @kvark)
-
Maximum values of Texture Dimension gpuweb/gpuweb#1327 (10 comments by @fintelia, @jsheard, @kainino0x, @Kangz, @kvark, @Richard-Yunchao)
-
Provide shader inputs as arguments to entry point functions gpuweb/gpuweb#1315 (9 comments by @dj2, @dneto0, @kdashg, @kvark)
-
Reconsider GPULoadOp "clear"/"zero" gpuweb/gpuweb#1377 (8 comments by @kainino0x, @Kangz, @kdashg, @kvark, @magcius)
-
[wgsl]: textureDimensions() returns vec3 for cube textures gpuweb/gpuweb#1345 (7 comments by @ben-clayton, @kdashg, @kvark)
-
arrayLength should operate on pointer-to-runtime-sized-array instead of runtime-array gpuweb/gpuweb#1329 (7 comments by @dj2, @dneto0, @kvark)
-
Investigation: Import VideoFrame from WebCodec to WebGPU gpuweb/gpuweb#1380 (6 comments by @kainino0x, @Kangz, @shaoboyan)
-
Consider marking feature-dependent APIs with custom IDL Extended Attributes gpuweb/gpuweb#1353 (6 comments by @kainino0x, @kvark, @toji)
-
Would love to see optional Geometry Shaders gpuweb/gpuweb#1239 (6 comments by @kainino0x, @krogovin)
-
[wgsl] Consider allowing `_` ident for struct members gpuweb/gpuweb#1310 (5 comments by @dj2, @dneto0, @kainino0x, @kvark, @litherum)
-
[wgsl] Consider using multiple scalars to construct matrix. gpuweb/gpuweb#1342 (5 comments by @dj2, @dneto0, @litherum, @shaoboyan)
-
Require storeOp when resolveTarget is specified? gpuweb/gpuweb#1376 (4 comments by @kainino0x, @kdashg, @kvark)
-
add sample_index and sample_mask builtin variables gpuweb/gpuweb#1317 (4 comments by @dj2, @dneto0, @kvark)
-
Investigation: kill vs demoteToHelper behavior in native APIs gpuweb/gpuweb#921 (3 comments by @dneto0, @litherum)
-
Proposal for importing Web platform images in WebGPU gpuweb/gpuweb#1154 (2 comments by @AshleyScirra, @kainino0x)
-
conflict between mat2x2 layout in uniform buffer and storage buffer gpuweb/gpuweb#1258 (2 comments by @dneto0, @kdashg)
-
beginRenderPass should validate that the attachment don't alias gpuweb/gpuweb#1358 (2 comments by @Kangz, @kvark)
-
Re-organize GPURenderPipelineDescriptor according to the structure of the GPU pipeline gpuweb/gpuweb#936 (2 comments by @kvark, @litherum)
-
Enforce calling GPUSwapChain.getCurrentTexture() should return the same binding within same event callback gpuweb/gpuweb#1388 (1 comment by @kainino0x)
-
texture builtins: For most texture builtins, Vulkan only allows compile-time-constant offset, and with with bounded range gpuweb/gpuweb#1235 (1 comment by @kdashg)
-
[wgsl] Redundant [[layout]] decorations in addition to [[offset]] decorations gpuweb/gpuweb#621 (1 comment by @kvark)
-
getExecutionTime requires encoding commands into an otherwise finished command buffer gpuweb/gpuweb#1331 (1 comment by @kvark)
-
Out-of-bounds clamping brings big perf regression on TFJS ML demos. gpuweb/gpuweb#1202 (1 comment by @krogovin)
-
Proposal: allow storage to be unsynchronized within a compute pass, like it is for render passes gpuweb/gpuweb#59 (1 comment by @kvark)
-
-
34 pull requests opened.
-
Rename GPUExtent3DDict.depth to .depthOrArrayLayers gpuweb/gpuweb#1390, by @kainino0x.
-
Add GPULoadOp "zero" gpuweb/gpuweb#1389, by @kainino0x.
-
Remove obsolete note about matrix layout gpuweb/gpuweb#1386, by @dneto0.
-
Fix example shader in 8.2 Entry Point Declaration gpuweb/gpuweb#1383, by @dneto0.
-
[wgsl] Fix type constructor matching in primary_expression gpuweb/gpuweb#1382, by @jrprice.
-
Change [Buffer/Texture/ImageBitmap]CopyView to "ImageCopy" names gpuweb/gpuweb#1375, by @toji.
-
Reset state even on empty executeBundles. gpuweb/gpuweb#1373, by @Kangz.
-
Add type rules for composite value decomposition gpuweb/gpuweb#1370, by @dneto0.
-
Clarify: Structure members have names gpuweb/gpuweb#1369, by @dneto0.
-
WGSL: describe pointers gpuweb/gpuweb#1368, by @dneto0.
-
GPUDevice: s/defaultQueue/queue/ gpuweb/gpuweb#1367, by @Kangz.
-
Fixup SPIR-V for float division. gpuweb/gpuweb#1366, by @dj2.
-
Require [SecureContext] for using WebGPU. gpuweb/gpuweb#1363, by @kdashg.
-
Add bool vector coercion gpuweb/gpuweb#1362, by @dneto0.
-
Replace [[block]] attribute by [[layout(...)]] gpuweb/gpuweb#1361, by @kvark.
-
Explain matrix type also in terms of column vectors gpuweb/gpuweb#1359, by @dneto0.
-
Add bool coercion for numeric scalars gpuweb/gpuweb#1356, by @dneto0.
-
Refactor the GPURenderPipelineDescriptor according to aspects gpuweb/gpuweb#1352, by @kvark.
-
Tighten vertex state validation. gpuweb/gpuweb#1351, by @Kangz.
-
Clarify some text related to GPUVertexState.indexFormat gpuweb/gpuweb#1350, by @toji.
-
[wgsl]: Add textureNum[Elements,Levels,Samples] gpuweb/gpuweb#1348, by @ben-clayton.
-
[wgsl] Sort texture built-ins alphabetically gpuweb/gpuweb#1347, by @ben-clayton.
-
Rename sample_mask to sample_mask_out gpuweb/gpuweb#1346, by @dneto0.
-
[wgsl] Further update access qualifiers for storage textures gpuweb/gpuweb#1344, by @ben-clayton.
-
mat2x2 is laid out tightly gpuweb/gpuweb#1341, by @dneto0.
-
Remove the requirement to have entry points gpuweb/gpuweb#1340, by @kvark.
-
Change `offset` to `span`. gpuweb/gpuweb#1339, by @dj2.
-
[wgsl] Add textureDimensions() gpuweb/gpuweb#1337, by @ben-clayton.
-
Change the name of CreateReady*Pipeline to Create*PipelineAsync gpuweb/gpuweb#1336, by @Jiawei-Shao.
-
Only allow 2d textures for copyImageBitmapToTexture for now gpuweb/gpuweb#1335, by @kainino0x.
-
Fix mistakes in examples gpuweb/gpuweb#1333, by @ben-clayton.
-
wgsl: Markup examples with language and scope gpuweb/gpuweb#1332, by @ben-clayton.
-
arrayLength acts on pointer-to-runtime-sized-array gpuweb/gpuweb#1330, by @dneto0.
-
Add texture dimension limits gpuweb/gpuweb#1328, by @Richard-Yunchao.
-
-
33 pull requests merged.
-
Add bool vector coercion gpuweb/gpuweb#1362.
-
Rendering algorithm (vertex and primitive assembly) gpuweb/gpuweb#1293.
-
Fix example shader in 8.2 Entry Point Declaration gpuweb/gpuweb#1383.
-
Bools are IO-shareable, but only for builtins gpuweb/gpuweb#1246.
-
Make all textureLoad functions on sampled textures require the level gpuweb/gpuweb#1301.
-
Reset state even on empty executeBundles. gpuweb/gpuweb#1373.
-
GPUDevice: s/defaultQueue/queue/ gpuweb/gpuweb#1367.
-
Change the name of CreateReady*Pipeline to Create*PipelineAsync gpuweb/gpuweb#1336.
-
Clarify: Structure members have names gpuweb/gpuweb#1369.
-
Fixup SPIR-V for float division. gpuweb/gpuweb#1366.
-
Refactor the GPURenderPipelineDescriptor according to aspects gpuweb/gpuweb#1352.
-
Add bool coercion for numeric scalars gpuweb/gpuweb#1356.
-
Explain matrix type also in terms of column vectors gpuweb/gpuweb#1359.
-
Tighten vertex state validation. gpuweb/gpuweb#1351.
-
Add workgroup size and ID builtins gpuweb/gpuweb#1295.
-
wsgl: Add rule that offsets must be compile time constant gpuweb/gpuweb#1238.
-
abs only works on signed integer scalar or vector gpuweb/gpuweb#1282.
-
Add GPUDevice.destroy() gpuweb/gpuweb#1316.
-
Clarify some text related to GPUVertexState.indexFormat gpuweb/gpuweb#1350.
-
[wgsl]: Add textureNum[Elements,Levels,Samples] gpuweb/gpuweb#1348.
-
[wgsl] Sort texture built-ins alphabetically gpuweb/gpuweb#1347.
-
Rename sample_mask to sample_mask_out gpuweb/gpuweb#1346.
-
WGSL: add sample_index, sample_mask_in, sample_mask_out gpuweb/gpuweb#1318.
-
mat2x2 is laid out tightly gpuweb/gpuweb#1341.
-
Remove outerProduct gpuweb/gpuweb#1290.
-
Rename GPUDeviceDescriptor features/limits to nonGuaranteed* gpuweb/gpuweb#1313.
-
Only allow 2d textures for copyImageBitmapToTexture for now gpuweb/gpuweb#1335.
-
Add texture dimension limits gpuweb/gpuweb#1328.
-
[wgsl] Add textureDimensions() gpuweb/gpuweb#1337.
-
Update sampler anisotropic related validations in spec gpuweb/gpuweb#1323.
-
[wgsl] Update storage textures to use access qualifiers gpuweb/gpuweb#1182.
-
Fix mistakes in examples gpuweb/gpuweb#1333.
-
wgsl: Markup examples with language and scope gpuweb/gpuweb#1332.
-
-
8 pull requests closed.
-
Introduce operator precedence table gpuweb/gpuweb#1111.
-
Reorder expression sections gpuweb/gpuweb#1136.
-
Refactor GPURenderPipelineDescriptor API gpuweb/gpuweb#1307.
-
Add depth16unorm to WebGPU SPEC gpuweb/gpuweb#508.
-
[WGSL] Update goals section gpuweb/gpuweb#599.
-
[wgsl] Update Goals section of spec gpuweb/gpuweb#588.
-
[wgsl] Add FAQ for direction on shader language gpuweb/gpuweb#562.
-
[wgsl] Further update access qualifiers for storage textures gpuweb/gpuweb#1344.
-
-
47 pull requests received 349 new comments.
-
Refactor the GPURenderPipelineDescriptor according to aspects gpuweb/gpuweb#1352 (37 comments by @kainino0x, @Kangz, @kvark, @toji)
-
Remove the requirement to have entry points gpuweb/gpuweb#1340 (32 comments by @ben-clayton, @dj2, @kainino0x, @kvark)
-
Rename GPUVertexFormat members to match GPUTextureFormat gpuweb/gpuweb#1322 (22 comments by @grorg, @jespertheend, @kainino0x, @Kangz, @kdashg, @kvark)
-
Rendering algorithm (vertex and primitive assembly) gpuweb/gpuweb#1293 (20 comments by @kainino0x, @kvark, @toji)
-
Add texture dimension limits gpuweb/gpuweb#1328 (19 comments by @kainino0x, @kvark, @Richard-Yunchao)
-
Make all textureLoad functions on sampled textures require the level gpuweb/gpuweb#1301 (9 comments by @ben-clayton, @dneto0, @kdashg, @kvark)
-
Change the name of CreateReady*Pipeline to Create*PipelineAsync gpuweb/gpuweb#1336 (9 comments by @kainino0x, @Kangz, @kdashg, @kvark)
-
Change [Buffer/Texture/ImageBitmap]CopyView to "ImageCopy" names gpuweb/gpuweb#1375 (8 comments by @kainino0x, @Kangz, @kvark, @toji)
-
Only allow 2d textures for copyImageBitmapToTexture for now gpuweb/gpuweb#1335 (8 comments by @kainino0x, @Kangz, @kdashg, @shaoboyan)
-
[wgsl] Further update access qualifiers for storage textures gpuweb/gpuweb#1344 (7 comments by @ben-clayton, @dj2, @kvark)
-
Require [SecureContext] for using WebGPU. gpuweb/gpuweb#1363 (6 comments by @annevk, @bholley, @Kangz, @kdashg)
-
[wgsl]: Add textureNum[Elements,Levels,Samples] gpuweb/gpuweb#1348 (4 comments by @ben-clayton, @kvark)
-
[wgsl] Fix type constructor matching in primary_expression gpuweb/gpuweb#1382 (3 comments by @dj2, @kvark)
-
Clarify some text related to GPUVertexState.indexFormat gpuweb/gpuweb#1350 (3 comments by @mcclure, @toji)
-
Rename GPUDeviceDescriptor features/limits to nonGuaranteed* gpuweb/gpuweb#1313 (3 comments by @kainino0x)
-
[wgsl] Update storage textures to use access qualifiers gpuweb/gpuweb#1182 (3 comments by @dj2, @kvark)
-
Allow e.g. `if (cond) break/continue/return/kill;` without brackets. gpuweb/gpuweb#705 (2 comments by @kdashg)
gpuweb/cts
-
24 pull requests opened.
-
Simplify how debug=1 affects log messages, fixes gpuweb/cts#447, by @kainino0x.
-
Loosen precision in multisample resolve operation test gpuweb/cts#446, by @kainino0x.
-
Adding subcases() to TestGroupBuilder gpuweb/cts#444, by @egalli.
-
Add ShaderValidationTest fixture gpuweb/cts#443, by @kainino0x.
-
Add validation tests on the member 'aspect' in GPUTextureCopyView gpuweb/cts#442, by @Jiawei-Shao.
-
Update gpuweb/types to include GPUTextureCopyView.aspect gpuweb/cts#441, by @Jiawei-Shao.
-
Update vertex_idx to vertex_index in WGSL. gpuweb/cts#440, by @dj2.
-
Update group to set in WGSL. gpuweb/cts#439, by @dj2.
-
Add fix to package.json scripts gpuweb/cts#438, by @shrekshao.
-
Add sampler anisotropy tests gpuweb/cts#437, by @shrekshao.
-
Usability improvements for tree in standalone gpuweb/cts#435, by @kainino0x.
-
Update texture storage syntax. gpuweb/cts#434, by @dj2.
-
Organize more and add more test stubs gpuweb/cts#433, by @kainino0x.
-
Update WGSL tests to use // comments. gpuweb/cts#432, by @dj2.
-
Added range validation tests for GPUQueue.writeBuffer gpuweb/cts#431, by @toji.
-
Add plans for texture view reads/writes gpuweb/cts#430, by @kainino0x.
-
Organize and rename some tests gpuweb/cts#429, by @kainino0x.
-
Add validation tests for the extensions of Query API gpuweb/cts#427, by @haoxli.
-
Add validation tests for createQuerySet gpuweb/cts#426, by @haoxli.
-
27 pull requests merged.
-
Simplify how debug=1 affects log messages, fixes gpuweb/cts#447.
-
Adding subcases() to TestGroupBuilder gpuweb/cts#444.
-
Add validation tests for createQuerySet gpuweb/cts#426.
-
Add validation tests for the extensions of Query API gpuweb/cts#427.
-
Add ShaderValidationTest fixture gpuweb/cts#443.
-
Update vertex_idx to vertex_index in WGSL. gpuweb/cts#440.
-
Update gpuweb/types to include GPUTextureCopyView.aspect gpuweb/cts#441.
-
Update group to set in WGSL. gpuweb/cts#439.
-
Add fix to package.json scripts gpuweb/cts#438.
-
Usability improvements for tree in standalone gpuweb/cts#435.
-
Organize more and add more test stubs gpuweb/cts#433.
-
Update texture storage syntax. gpuweb/cts#434.
-
Added range validation tests for GPUQueue.writeBuffer gpuweb/cts#431.
-
Update WGSL tests to use // comments. gpuweb/cts#432.
-
Organize and rename some tests gpuweb/cts#429.
-
Add plans for texture view reads/writes gpuweb/cts#430.
-
1 pull request closed.
-
23 pull requests received 191 new comments.
-
Added range validation tests for GPUQueue.writeBuffer gpuweb/cts#431 (29 comments by @kainino0x, @toji)
-
[wgsl] validation v-0033 - the initializer’s type must match the store type of the variable. gpuweb/cts#445 (22 comments by @dj2, @kainino0x, @sarahM0)
-
Add validation tests on the member 'aspect' in GPUTextureCopyView gpuweb/cts#442 (21 comments by @Jiawei-Shao, @kainino0x, @Kangz)
-
Add validation tests for the extensions of Query API gpuweb/cts#427 (7 comments by @haoxli, @kainino0x)
-
Add validation test for createSampler with different maxAnisotropy values gpuweb/cts#425 (7 comments by @kainino0x, @shrekshao)
-
[wgsl] Add validation test - v-0038: A module scope variable in the in or out storage class must be IO-shareable. gpuweb/cts#448 (4 comments by @dj2, @kainino0x)
-
[wgsl] Add validation test - v-0032: constraints on module variables with an initializers gpuweb/cts#417 (4 comments by @dj2, @sarahM0)
-
Update defaultQueue->queue, OUTPUT_ATTACHMENT->RENDER_ATTACHMENT to match latest spec gpuweb/cts#449 (3 comments by @kainino0x, @Kangz, @toji)
-
[wgsl] Add validation test - v-0038: A module scope variable in the in or out storage class must be IO-shareable. gpuweb/cts#424 (1 comment by @sarahM0)
-
copyTextureToTexture:color_textures,*: format first for the benefit of splitting run time gpuweb/cts#436 (1 comment by @Jiawei-Shao)
-
[wgsl] add validation test - v-0035: The access decoration is required for variables in the storage storage class. gpuweb/cts#421 (1 comment by @dj2)
-
[wgsl] validation v-0040: uniform and storage buffer are declared with group and binding decorations. gpuweb/cts#428 (1 comment by @dj2)