Skip to content

fix: enable opt for cutlass sources to avoid infinite compile time #2595

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
timmoon10 merged 1 commit into from
Jan 15, 2026
Merged

fix: enable opt for cutlass sources to avoid infinite compile time #2595

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
16 changes: 14 additions & 2 deletions transformer_engine/common/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -230,12 +230,24 @@ add_library(transformer_engine SHARED ${transformer_engine_SOURCES})
target_include_directories(transformer_engine PUBLIC
"${CMAKE_CURRENT_SOURCE_DIR}/include")

# CUTLASS kernels require SM90a and cause hang in debug build
# Grouped GEMM kernels require SM90a
set_property(
SOURCE gemm/cutlass_grouped_gemm.cu
APPEND
PROPERTY
COMPILE_OPTIONS "--generate-code=arch=compute_90a,code=sm_90a;-g0")
COMPILE_OPTIONS "--generate-code=arch=compute_90a,code=sm_90a")

# CUTLASS kernels could cause hang in debug build
set(CUTLASS_KERNEL_SOURCES
gemm/cutlass_grouped_gemm.cu
hadamard_transform/group_hadamard_transform_cast_fusion.cu
hadamard_transform/group_row_cast_col_hadamard_transform_cast_fusion.cu
hadamard_transform/hadamard_transform_cast_fusion.cu)
set_property(
SOURCE ${CUTLASS_KERNEL_SOURCES}
APPEND
PROPERTY
COMPILE_OPTIONS "-g0;-dopt=on")

# Configure dependencies
target_link_libraries(transformer_engine PUBLIC
Expand Down
Loading