Skip to content

[NFC][SYCL] Replace reduCGFuncAtomic64 with reduCGFuncForNDRangeBothFastReduceAndAtomics #7272

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.

Sign up for GitHub

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

Merged
merged 1 commit into from
Nov 3, 2022
Merged

[NFC][SYCL] Replace reduCGFuncAtomic64 with reduCGFuncForNDRangeBothFastReduceAndAtomics #7272

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
43 changes: 2 additions & 41 deletions sycl/include/sycl/reduction.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1932,45 +1932,6 @@ void reduCGFuncMulti(handler &CGH, KernelType KernelFunc,
Rest(createReduOutAccs<false>(NWorkGroups, CGH, ReduTuple, ReduIndices));
}

namespace reduction {
namespace main_krn {
template <class KernelName> struct NDRangeAtomic64;
} // namespace main_krn
} // namespace reduction

// Specialization for devices with the atomic64 aspect, which guarantees 64 bit
// floating point support for atomic reduction operation.
template <typename KernelName, typename KernelType, int Dims,
typename PropertiesT, class Reduction>
void reduCGFuncAtomic64(handler &CGH, KernelType KernelFunc,
const nd_range<Dims> &Range, PropertiesT Properties,
Reduction &Redu) {
auto Out = Redu.getReadWriteAccessorToInitializedMem(CGH);
static_assert(
Reduction::has_float64_atomics,
"Only suitable for reductions that have FP64 atomic operations.");
size_t NElements = Reduction::num_elements;
using Name =
__sycl_reduction_kernel<reduction::main_krn::NDRangeAtomic64, KernelName>;
CGH.parallel_for<Name>(Range, Properties, [=](nd_item<Dims> NDIt) {
// Call user's function. Reducer.MValue gets initialized there.
typename Reduction::reducer_type Reducer;
KernelFunc(NDIt, Reducer);

// If there are multiple values, reduce each separately
// reduce_over_group is only defined for each T, not for span<T, ...>
for (int E = 0; E < NElements; ++E) {
typename Reduction::binary_operation BOp;
Reducer.getElement(E) =
reduce_over_group(NDIt.get_group(), Reducer.getElement(E), BOp);
}

if (NDIt.get_local_linear_id() == 0) {
Reducer.atomic_combine(&Out[0]);
}
});
}

template <typename... Reductions, size_t... Is>
void associateReduAccsWithHandler(handler &CGH,
std::tuple<Reductions...> &ReduTuple,
Expand Down Expand Up @@ -2386,8 +2347,8 @@ void reduction_parallel_for(handler &CGH,
device D = detail::getDeviceFromHandler(CGH);

if (D.has(aspect::atomic64)) {
reduCGFuncAtomic64<KernelName>(CGH, KernelFunc, Range, Properties,
Redu);
reduCGFuncForNDRangeBothFastReduceAndAtomics<KernelName>(
CGH, KernelFunc, Range, Properties, Redu);
} else {
// Resort to basic implementation as well.
reduction_parallel_for_basic_impl<KernelName>(
Expand Down