-
Notifications
You must be signed in to change notification settings - Fork 39
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Leverage CTAD with [MD]RangePolicy #1141
Conversation
I may indeed misunderstand some things. I do wonder why it compiled everywhere CPU, though, and not universally failed. |
Because only one execution space was enabled and these types were equivalent/compatible. |
6dd6798
to
58b77a5
Compare
I stopped and didn't actually go through the rest of the changes. I expect you will find other occurrences if any |
Thank you. I used regex and should have been much more careful about the changes. I'm sorry. |
58b77a5
to
825c2b9
Compare
Update: this reproducer fails on CUDA-11.1 with Kokkos 4.3: #include <Kokkos_Core.hpp>
template <typename ExecutionSpace, typename ViewType>
void iota(ExecutionSpace const &space, ViewType const &v,
typename ViewType::value_type value = 0)
{
Kokkos::parallel_for(
"ArborX::Algorithms::iota", Kokkos::RangePolicy(space, 0, v.extent(0)),
KOKKOS_LAMBDA(int i) { v(i) = value + i; });
}
int main(int argc, char *argv[])
{
Kokkos::ScopeGuard guard(argc, argv);
Kokkos::View<int*> v("view", 10);
iota(Kokkos::DefaultExecutionSpace{}, v);
return EXIT_SUCCESS;
} The failure:
It passes with added |
d7fdf31
to
274f5af
Compare
Rebased. |
No description provided.