Segmentation fault in LLVM with a simple SYCL parallel reduction when running on CUDA device through pocl

4

Apologies, I’ve discovered what happened to the other post.