Petr Penzin

I've done something like that before, though it is in [very rough POC stage](https://github.com/penzn/omp-wasi-parallel) and is blocked on runtime support. @ecs-commonA you are absolutely right, if 'kmp' symbols are defined...

No objections from me either. I doubt there are going to be objections in this group :)

The problem is that all current implementations of vector intrinsics (in C/C++ family) are platform-specific, while WASM is not, that's why it is an open question what syntax it should...

I think you are probably right. The main question is who would be responsible from this header, since WASM does not have a vendor in the same sense as hardware...

It just feels like the WebAssembly community is the right origin for this header, just like a hardware vendor is the right origin for corresponding hardware intrinsics header. At the...

Yes, wasm propagates NaNs, that's why they have to be of "quiet" type.

After looking into this a little more, I think there is a bigger problem than `NaN`s, specifically comparisons of -0.0 to +0.0. There are known ways to deal with (consistently...

Yeah, that is on my list as well - trying to investigate better min/max sequences, using AVX512.

(the following is an excerpt from meeting notes https://github.com/WebAssembly/relaxed-simd/issues/47) Floating-point min and max operations in WebAssembly have the following constraints: - Signed zero handling (ex.: `min(0.0, -0.0) == -0.0`) -...

SSE, Neon, and Wasm SIMD implementations. Packed shift exists on both native platforms. SSE: ``` inline static void mm_recursion(__m128i * r, __m128i a, __m128i b, __m128i c, __m128i d) {...