simde
simde copied to clipboard
simd-everywhere
Implementations of SIMD instruction sets for systems which don't natively support them.
This PR contains a `riscv64-cross.txt` file for cross compiling for riscv64 architecture on meson, a `cloud-v-pipeline` file for CI flow and a Makefile for running tests on RISC-V architecture (since...
Hello. As you may know, Mozilla bundles an older version of simde with their sources, and it works. I'm trying to use the latest version (0.8.2) installed by the [FreeBSD...
On one hand, the prefetch macros are originally defined `_MM_HINT_T0` (resp, `T2`) by Intel as: ``` xmmintrin.h:#define _MM_HINT_T0 3 ... xmmintrin.h:#define _MM_HINT_T2 1 ``` On the other hand, SIMDE is...
This PR translates `vst4_u8` to Wasm SIMD. We change the header to match the Arm `vmovl` Wasm translations: https://github.com/simd-everywhere/simde/blob/master/simde/arm/neon/movl.h#L43 .
Simple test case: ```cpp void test(double* dst, double* a) { __m256d b = _mm256_loadu_pd(a); __m256d c = _mm256_add_pd(b, b); _mm256_storeu_pd(dst, c); } ``` On arm64, with gcc 11 (`-O2 -fno-stack-protector...
The qdmlsl instructions were implemented without any saturation. This has been fixed by utilising existing saturating instructions which are implemented correctly. Unit tests have also been updated to test for...
qdmlal instructions were implemented without saturation. This has been fixed by utilising existing SIMDe saturating mult and add instructions. Unit tests have been updated to test for all possible saturation...
I did a little poking around and didn't find the source of this issue. It *seems* like it should be using the gcc builtins for this, but instead are looking...
The masked functions should not touch memory where the mask is set to 0 at all. Instead, the simde_mm512_mask_storeu_* functions write a 0 if the mask is set to 0....
Metadata
Owner
Metadata
Implementations of SIMD instruction sets for systems which don't natively support them.