Skip to content

Skip alignment check for sourceless epilogues #555

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

Draft
wants to merge 1 commit into
base: main
Choose a base branch
from
Draft

Skip alignment check for sourceless epilogues #555

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
25 changes: 19 additions & 6 deletions include/cutlass/epilogue/collective/xe_epilogue.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -238,12 +238,21 @@ class CollectiveEpilogue<
}
}

bool is_beta_zero = (args.thread.beta == decltype(args.thread.beta)(0));
bool beta_implementable = true;

if (!is_source_supported || args.ptr_C == nullptr) {
beta_implementable = is_beta_zero;
}

if constexpr (is_source_supported) {
constexpr int min_aligned_elements_C = copy_alignment_bits / sizeof_bits<ElementC>::value;
implementable &= cutlass::detail::check_alignment<min_aligned_elements_C>(cute::make_shape(M,N,L), args.dC);
if (L > 1) {
constexpr int min_batch_aligned_elements_C = batch_alignment_bits / sizeof_bits<ElementC>::value;
implementable &= get<2>(args.dC) % min_batch_aligned_elements_C == 0;
if (!is_beta_zero) {
constexpr int min_aligned_elements_C = copy_alignment_bits / sizeof_bits<ElementC>::value;
implementable &= cutlass::detail::check_alignment<min_aligned_elements_C>(cute::make_shape(M,N,L), args.dC);
if (L > 1) {
constexpr int min_batch_aligned_elements_C = batch_alignment_bits / sizeof_bits<ElementC>::value;
implementable &= get<2>(args.dC) % min_batch_aligned_elements_C == 0;
}
}
}

Expand All @@ -257,7 +266,11 @@ class CollectiveEpilogue<
CUTLASS_TRACE_HOST(" CAN IMPLEMENT: Problem Size doesn't meet the minimum requirements for FusionCallbacks.\n");
}

return implementable && fusion_implementable;
if (!beta_implementable) {
CUTLASS_TRACE_HOST(" CAN IMPLEMENT: Beta was set, but epilogue is sourceless (void-C).\n");
}

return implementable && beta_implementable && fusion_implementable;
}

CUTLASS_HOST_DEVICE
Expand Down