`example.nvexec.maxwell_gpu_m` failing at runtime with `cudaErrorSymbolNotFound`

NVIDIA / stdexec

`std::execution`, the proposed C++ framework for asynchronous and parallel programming.

Apache License 2.0

1.56k stars 159 forks source link

`example.nvexec.maxwell_gpu_m` failing at runtime with `cudaErrorSymbolNotFound` #888

Closed ericniebler closed 1 year ago

ericniebler commented 1 year ago

CUDA error cudaErrorSymbolNotFound [/home/eniebler/Code/stdexec/include/nvexec/stream/bulk.cuh:239]: named symbol not found

gevtushenko commented 1 year ago

I can't reproduce the issue locally. @ericniebler where is this happening? Is there any chance that the problematic system has multiple GPUs with different architectures but you specify -DCMAKE_CUDA_ARCHITECTURES only for one of those?

ericniebler commented 1 year ago

Aha. Yes indeed. Ty.