0

I am trying to compile some CUDA/GPU code on Ubuntu 16.04 with CUDA 8.0, but I am receiving the following error. I'm not sure how to debug this.

nvcc -I/usr/local/cuda/include  -gencode arch=compute_61,code=sm_61 -I/home/xxx/software/pr/include -I/usr/include/glib-2.0 -I/usr/lib/x86_64-linux-gnu/glib-2.0/include  -m64 -dc  -I/home/xxx/software/include -DUSEFFTW -DUSEMMAP -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64  -Xcompiler -Wno-unused-result  -Xcompiler -Wno-comment  -Xcompiler -Wno-unused-but-set-variable  -Xcompiler -Wno-unused-but-set-parameter  -Xcompiler -Wno-unused-variable  -Xcompiler -Wno-unused-parameter  -Xcompiler -Wno-unused-function -DNVVP -O3 -DFLOAT -DFORUM -use_fast_math -prec-div=false -ftz=true -prec-sqrt=false --fmad=true -lineinfo -Xptxas -dlcm=ca -Xcompiler -fopenmp -DWITHOMP -DCUDA -Xcompiler -fPIC -o build/release/cuda_accel_IN.cu.o -c cuda_accel_IN.cu

/usr/include/glib-2.0/glib/gtypes.h(418): error: identifier "__builtin_uadd_overflow" is undefined

/usr/include/glib-2.0/glib/gtypes.h(420): error: identifier "__builtin_umul_overflow" is undefined

/usr/include/glib-2.0/glib/gtypes.h(423): error: identifier "__builtin_uaddll_overflow" is undefined

/usr/include/glib-2.0/glib/gtypes.h(425): error: identifier "__builtin_umulll_overflow" is undefined

4 errors detected in the compilation of "/tmp/tmpxft_0000235c_00000000-7_cuda_accel_IN.cpp1.ii".
Makefile:657: recipe for target 'build/release/cuda_accel_IN.cu.o' failed
make: *** [build/release/cuda_accel_IN.cu.o] Error 2
  • Likely you are missing an include file that gtypes.h needs. – Mark Stewart Jan 17 at 20:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.