krzikalla

[FR] Add a version number

1

I tried to figure out the version of my installed google/benchmark. It is either 1.5.0 or 1.5.1, but I am unsure. I'd like to have a version.h file or something...

If you change https://github.com/VcDevel/Vc/blob/b84dcd0a65d8dc5de6a2bd4d367882b3748f812c/Vc/common/simdarrayfwd.h#L49-L54 to ```` Vc_INTRINSIC Vector(const Vector &x) = default; Vc_INTRINSIC Vector &operator=(const Vector &x) = default; ```` `Vector` and in turn `SimdArray` becomes trivially copyable. Currently we...

I experience crashes since some weeks. During the last days it is happening so frequently, that I can't consider it a productive environment anymore. Often it happens after I change...

Make CUDA annotated functions available in usual std namespace

3

What are the chances that one day I will be able to write the following: ``` #include #ifdef __NVCC__ #define FN_PREFIX __host__ __device__ #else #define FN_PREFIX #endif FN_PREFIX void flexible_foo(std::complex...

Add support for other multi-threading APIs

3

Support the benchmarking of code, which relies on other multi-threading APIs, e.g. OpenMP.

Is the reference returned by operator[] to restrictive?

1

By disallowing a binding of simd::reference to an lvalue the code below (ref. https://godbolt.org/z/7qW8bffe5) becomes clumsy: ``` #include #include namespace stdx = std::experimental; void foo(auto&& x) { std::forward(x) = .2;...

**Describe the bug** This is a follow-up to #764 . That one is closed, but I fail to see a fix there. ``` #include struct X { X() : value{}...

[FR] Compare benchmarks of the same run

1

**Is your feature request related to a problem? Please describe.** Often I have different implementations of one algorithm (e.g. a scalar and a vectorized version). I'd like to know the...

clang error with OpenMP atomic capture compare

4

I'd like to draw your attention on https://github.com/llvm/llvm-project/issues/127466#issue-2857364223 It looks like clang has still serious problems with the `#pragma omp atomic capture compare` construct used in include/alpaka/atomic/AtomicOmpBuiltIn.hpp. I wonder if...

Intrinsic implementation prevents instruction shuffling

1

Just consider the generated code for the two `foo` variants: https://godbolt.org/z/P7TMe4vax The code using `double` interleaves multiplications for the first and second summand to achieve more ILP. The code using...