ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-17 11:30:02 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add -gsplit-dwarf flag to reduce debug section size and fix ckProfiler link errors (#2611)

Browse Source 
Resolves R_X86_64_32 relocation out of range errors in grouped conv2d instances
by splitting debug information into separate .dwo files.

Add explicit cast to avoid signed/unsigned comparison warning.

[ROCm/composable_kernel commit: 15eb493152]
This commit is contained in:
Jinchao Xu
2025-08-05 02:26:08 +08:00
committed by GitHub
parent c903850b65
commit d2525548c6
3 changed files with 11 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
library/src/tensor_operation_instance/gpu/CMakeLists.txt
View File

@@ -175,6 +175,10 @@ function(add_instance_library INSTANCE_NAME)
target_compile_features(${INSTANCE_NAME} PUBLIC)
# splits debug information into separate .dwo files to reduce debug section size
if(CMAKE_BUILD_TYPE STREQUAL "Debug" OR CMAKE_BUILD_TYPE STREQUAL "RelWithDebInfo")
target_compile_options(${INSTANCE_NAME} PRIVATE -gsplit-dwarf)
endif()
# flags to compress the library
if(NOT DISABLE_OFFLOAD_COMPRESS AND NOT WIN32 AND ${hip_VERSION_FLAT} GREATER 600241132)
message(DEBUG "Adding --offload-compress flag for ${INSTANCE_NAME}")