xsimd

xsimd copied to clipboard

dispatch assumes that target architectures are supported

6

First, sorry for the long issue, or if it's wildly off-base. Compiler architecture flags change both the set of instructions available (through intrinsics) and the set of instructions used by...

tomjnixon

Error compiling xtensor tests with latest xsimd on MSVC 2019 v16.11.5

10

I get a number of errors when building latest xtensor tests against xsimd 8.0.3 on MSVC with latest Visual Studio v16.11.5. The problem is here reported for test_xdatesupport, but appears...

emmenlau

support for padding?

2

Does xsimd provide support for padding dimension when specific sizes are not available for SIMD operations?

den-run-ai

Feature/all inline

5

serge-sans-paille

Support RISC-V SIMD/Vector extension

2

It could be interesting to support RISC-V SIMD or Vector extension. The latter looks very promising as the vector length and data type can be changed at run-time, and is...

davidbrochart

Implement an inline macro to have maximum performance on MSVC

8

:wave: While debugging the still bad codegen of MSVC, I realised that on reasonably big functions, it doesn't inline any of the xsimd functions even when marked as `inline`. The...

amyspark

👋 Vc has a "Scalar" architecture that allows the library to be used without targeting any particular ISA, hence doing all computations on individual values. xsimd does have something similar...

amyspark

Enable compiling arm/neon with MSVC for windows on arm64

4

This patch enables building arm/neon with MSVC compiler for windows on arm64 target and contains following changes, - Replace the dispatcher mechanism and use an explicit function selection using scalar...

niyas-sait

MSVC compile error for NEON (win/arm64)

3

XSIMD library cannot be built using MSVC compiler on a windows/arm64 target. ### Steps to reproduce Visual Studio 2019 or higher is required for compilation. Compilation can be done natively...

niyas-sait

The lgamma test with negative inputs have been disabled for architecture supporting AVX but not AVX2, because they lead to small differences for a few inputs: ```bash idx = 1533...

JohanMabille