Results 545 comments of David Neto

I have a followup which greatly expands coverage of loops. I'll land this now as it is still valid and does fix the #4476 issue

https://github.com/gpuweb/cts/pull/4332 oh, I didn't now about that! It's a good start. * I'd also like to see loading from an atomic inside a larger workgroup variable. The #4332 only tests...

@beaufortfrancois FYI in case you wanted to pursue this further

Please keep this open for the memory model stress tests. Otherwise I'd open a new one for the stress tests... :-)

Yes, this is a real problem. This also comes up internal to our compiler (Tint) where we (used to) do a bunch of AST transformations. It's annoying and extra bookkeeping...

I've added branch protections requiring CI core compilation tests to pass. Once this is fixed, add a ruleset for CI sanitizer tests to pass.

MSL and Vulkan wording is simliar to what exists in WGSL: MSL: (5.4 sampling and interpolation attributes) > If a centroid attribute variant is specified (centroid_perspective and centroid_no_perspective), the interpolation...

I agree with @jimblandy this seems bound to the mechanics of the `@if` feature. They should be considered together. As the discussion points out, there are many ways compilation can...

# Editorial feedback ## Grammar for translate-time expression includes too much? Right now the way things are worded, translate-time expressions include logical expressions and parenthesized expressions. But formally those pull...