David Neto

Fixed by https://github.com/gpuweb/gpuweb/pull/5394

Preview here: https://github.com/dneto0/shaderc/tree/issue-1429

> _ No description provided. _ What version of CMake are you using? How did you try configuring CMake, and building the project?

The way this would get exposed would have to fit in with the limit on the number of configuration buckets exposed by the API. See https://github.com/gpuweb/gpuweb/pull/3840

For concreteness, here are some cases that I used to get my head straight on this. Where "today" is after #4978 lands. It shows that scalar block layout is still...

Now that #4978 has landed, the spec change for this is: - first determine if this is to be available everywhere. See emulation considerations at https://github.com/gpuweb/gpuweb/issues/4973#issuecomment-2491471946 - change `RequiredAlignOf(vecN)` to...

With recent fixes for [validating GPUProgrammableStage](https://gpuweb.github.io/gpuweb/#abstract-opdef-validating-gpuprogrammablestage), if a sampled texture or depth texture is used with a filtering sampler, then the texture must have a 'float' sampleType. The first gist...

For the record., using the gist at https://jsgist.org/?src=e60d7e7967962f9b3119a0fd3d622f9a and using powerPreference:'high-performance', on my Mac AMD GPU Macbook Pro 2017, AMD Radeon Pro 560 Vendor: AMD 0x1002 Device ID 0x6ef Revision...

Still need to address: - Apple GPU texel buffer width limits. https://github.com/gpuweb/gpuweb/pull/4912#discussion_r1806963919 - Add a note to get info for metal GPUs prior to Apple silicon. - Add @teoxoy 's...

I'll land now, and submit a new PR to address the last items.