cpp20

We're seeing a lot of warnings on Linux, host compiler GCC 9.3.0 with -Wconversion.

For a basic example, compile:

#include <thrust/device_vector.h>
int main() {
  thrust::device_vector<int> a;
  return 0;
}

With: nvcc main.cu -Xcompiler=-Wconversion.

This will result in around 1000 lines of output warnings.
Tested with the Thrust versions in CUDA 11.0, 11.3, and the la

When compiling example/device/example_device_radix_sort.cu with

clang++-11 -DTHRUST_IGNORE_CUB_VERSION_CHECK -I../.. --cuda-gpu-arch=sm_61 -std=c++17 -x cuda -c example_device_radix_sort.cu

it errors:

clang: warning: Unknown CUDA version 11.0. Assuming the latest supported version 10.1 [-Wunknown-cuda-version]
In file included from example_device_radix_sort.cu:46:
In file inc

I've recently come to realize that the 'one-size-fits-all' mega-snippets on marzer.github.io/tomlplusplus aren't necessarily useful for everyone, and I should break them up into smaller sub-examples. I'd be totally happy to accept help with this, though that may be wishful thinking!

A custom generator for systems documentation. It should parse the system definition and generate a custom page according to our requirements.

The output should list:

• dimension definition and a coherent unit
• quantity alias with a default representation type
• all the units with their UDLs and constants

The problem is that the OpenCL types in https://github.com/triSYCL/triSYCL/blob/master/include/triSYCL/opencl_types.hpp are defined on the host according to the x86-64 Linux ABI which depends on the CPU & OS instead of using the description from https://www.khronos.org/registry/OpenCL/specs/2.2/html/OpenCL_C.html#built-in-scalar-data-types

Note that the system-wide cl_size_t has been removed