Adding thrust headers & Visual Studio 2022 Version 17.14.21 (November 2025) throws CUDACOMPILE : nvcc error : ‘cudafe++’ died with status 0xC0000409
This is with CUDA Runtime 13.0
The path to cccl is correctly set.
Same setup works fine with CUDA Runtime 12.9.
Any clues?
This seems to be an old bug in a previous CUDA Runtime a long time ago. But seems to be recurring now.