Closed iomaganaris closed 2 months ago
It should be fixed in #1789. Can you test that PR (maybe after adding the __CUDA_ARCH__
check).
Hi there, this is jenkins continuous integration... Do you want me to verify this patch?
It should be fixed in #1789. Can you test that PR (maybe after adding the
__CUDA_ARCH__
check).
Missed this PR.. It works as expected so this PR can be closed.
It seems like the
nvcc
compiler for device code doesn't support the check__has_builtin(__builtin_assume)
even if__builtin_assume
is supported since CUDA 11.2. Adding a check for__CUDA_ARCH__
and the compiler version fixes the issue. See godbolt example.