Architecture 30 was removed with CUDA 11, so most of the CUDA tests fail with
it.
Remove setting the architecture and bump the minimum version to 3.18, so
CMP0104 takes effect and we can rely on the default architecture, which is
guaranteed to be compilable.
Use of __ldg() in ProperLinkFlags was removed as it only affects performance
and is available only on sm_35 and above.
Testing the functionality of CUDA_ARCHITECTURES is already covered by
CudaOnly.Architecture and CudaOnly.CompileFlags.
Previously the CMake didn't compute the required set of libraries
needed to properly device link a static library when
CUDA_RESOLVE_DEVICE_SYMBOLS was enabled.
The previous version had two bugs that caused the JIT runtime errors.
1. It was building the executable without separable compilation enabled
2. All kernel launches will fail if any kernel is missing a symbol, that
is why the call to file2_launch_kernel had to be removed
If a static library has the property CUDA_RESOLVE_DEVICE_SYMBOLS enabled
it will now perform the device link step. The normal behavior is
to delay calling device link until the static library is consumed by
a shared library or an executable.
