eyalroz/cuda-api-wrappers

sync and async versions of copy_single disagree - pointers vs references

Opened this issue · 0 comments

We currently have:

template <typename T>
void copy_single(T* destination, const T* source)

in cuda::memory, but:

void copy_single(T& destination, const T& source, const stream_t& stream);

i cuda::memory::async. We should either have references in both signatures or in none.