OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-06 12:20:52 +01:00
Code Issues Actions 38 Packages Projects Releases Wiki Activity

Fix aarch64 debug build with GCC (#136990)

Browse Source 
Fixes #136440

**Issue:**
When building PyTorch in debug mode on aarch64 architecture using GCC, we encounter relocation errors due to the R_AARCH64_CALL26 relocation limit. This occurs because debug builds with -O0 optimization generate larger code sizes, potentially exceeding the range limit for these relocations.

**Fix:**
Apply -Og optimization instead of -O0 for aarch64 GCC debug builds. This slightly reduces code size while maintaining debuggability, bringing function calls back within the range of R_AARCH64_CALL26 relocations.

The fix is implemented by conditionally setting compiler and linker flags in CMakeLists.txt:
- For aarch64 GCC debug builds: use -Og
- For all other debug builds: retain -O0

This change affects only debug builds on aarch64 with GCC, leaving other configurations unchanged.

**Testing:**
Verified that the build succeeds without relocation errors on aarch64 systems with GCC in debug mode. Ensured that debugging information is still available and useful for debugging purposes.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/136990
Approved by: https://github.com/malfet

Co-authored-by: Nikita Shulga <2453524+malfet@users.noreply.github.com>
This commit is contained in:
abhishek-fujitsu 2024-09-30 21:11:48 +00:00 committed by PyTorch MergeBot
parent 866a64ce9a
commit b825848d85
1 changed files with 10 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
CMakeLists.txt
View File

@ -1083,8 +1083,16 @@ if(NOT MSVC)
append_cxx_flag_if_supported("-Wno-unused-but-set-variable" CMAKE_CXX_FLAGS) append_cxx_flag_if_supported("-Wno-unused-but-set-variable" CMAKE_CXX_FLAGS)
append_cxx_flag_if_supported("-Wno-maybe-uninitialized" CMAKE_CXX_FLAGS) append_cxx_flag_if_supported("-Wno-maybe-uninitialized" CMAKE_CXX_FLAGS)
append_cxx_flag_if_supported("-fstandalone-debug" CMAKE_CXX_FLAGS_DEBUG) append_cxx_flag_if_supported("-fstandalone-debug" CMAKE_CXX_FLAGS_DEBUG)
string(APPEND CMAKE_CXX_FLAGS_DEBUG " -fno-omit-frame-pointer -O0") if(CMAKE_SYSTEM_PROCESSOR MATCHES "aarch64" AND CMAKE_CXX_COMPILER_ID MATCHES "GNU")
string(APPEND CMAKE_LINKER_FLAGS_DEBUG " -fno-omit-frame-pointer -O0") if(CMAKE_BUILD_TYPE MATCHES Debug)
message(Warning "Applying -Og optimization for aarch64 GCC debug build to workaround ICE")
endif()
string(APPEND CMAKE_CXX_FLAGS_DEBUG " -fno-omit-frame-pointer -Og")
string(APPEND CMAKE_LINKER_FLAGS_DEBUG " -fno-omit-frame-pointer -Og")
else()
string(APPEND CMAKE_CXX_FLAGS_DEBUG " -fno-omit-frame-pointer -O0")
string(APPEND CMAKE_LINKER_FLAGS_DEBUG " -fno-omit-frame-pointer -O0")
endif()
append_cxx_flag_if_supported("-fno-math-errno" CMAKE_CXX_FLAGS) append_cxx_flag_if_supported("-fno-math-errno" CMAKE_CXX_FLAGS)
append_cxx_flag_if_supported("-fno-trapping-math" CMAKE_CXX_FLAGS) append_cxx_flag_if_supported("-fno-trapping-math" CMAKE_CXX_FLAGS)
append_cxx_flag_if_supported("-Werror=format" CMAKE_CXX_FLAGS) append_cxx_flag_if_supported("-Werror=format" CMAKE_CXX_FLAGS)