Skip to content

wrong compiler flags to nvcc for icx in debug mode #729

@valassi

Description

@valassi

wrong compiler flags to nvcc for icx in debug mode

[avalassi@itscrd90 icx2023/cvmfs] /data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gg_tt.sa/SubProcesses/P1_Sigma_sm_gg_ttx> ccache /usr/local/cuda-12.0/bin/nvcc  -g -O0 -G -O0 -I. -I../../src -I/usr/local/cuda-12.0/include/ -DUSE_NVTX -gencode arch=compute_70,code=compute_70 -gencode arch=compute_70,code=sm_70 -use_fast_math -std=c++17  -ccbin /cvmfs/projects.cern.ch/intelsw/oneAPI/linux/x86_64/2023/compiler/2023.1.0/linux/bin/icpx -allow-unsupported-compiler -DMGONGPU_FPTYPE_DOUBLE -DMGONGPU_FPTYPE2_DOUBLE -Xcompiler -Wno-deprecated-builtins -Xcompiler -fPIC -c gcheck_sa.cu -o gcheck_sa.o
icpx: remark: Note that use of '-g' without any optimization-level option will turn off most compiler optimizations similar to use of '-O0'; use '-Rno-debug-disables-optimization' to disable this remark [-Rdebug-disables-optimization]

This is better, we need to add -Xcompiler

[avalassi@itscrd90 icx2023/cvmfs] /data/avalassi/GPU2023/madgraph4gpuX/epochX/cudacpp/gg_tt.sa/SubProcesses/P1_Sigma_sm_gg_ttx> ccache /usr/local/cuda-12.0/bin/nvcc  -Xcompiler -g -Xcompiler -O0 -G -I. -I../../src -I/usr/local/cuda-12.0/include/ -DUSE_NVTX -gencode arch=compute_70,code=compute_70 -gencode arch=compute_70,code=sm_70 -use_fast_math -std=c++17  -ccbin /cvmfs/projects.cern.ch/intelsw/oneAPI/linux/x86_64/2023/compiler/2023.1.0/linux/bin/icpx -allow-unsupported-compiler -DMGONGPU_FPTYPE_DOUBLE -DMGONGPU_FPTYPE2_DOUBLE -Xcompiler -Wno-deprecated-builtins -Xcompiler -fPIC -c gcheck_sa.cu -o gcheck_sa.o

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions