nvcc -ptx -gencode=arch=compute_20,code=sm_20 utils.cu
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
In file included from /opt/cuda/bin/..//include/cuda_runtime.h:78:0,
from <command-line>:0:
/opt/cuda/bin/..//include/host_config.h:119:2: error: #error -- unsupported GNU version! gcc versions later than 5 are not supported!
#error -- unsupported GNU version! gcc versions later than 5 are not supported!
^~~~~
Makefile:15: recipe for target 'utils.ptx' failed
make: *** [utils.ptx] Error 1
Luckily, I've already implemented such a mechanism as part of CUDAdrv. Maybe we should move this functionality, and the entire compilation example, over to CUDArt (since CUDAdrv doesn't require nvcc to be installed except for that example)?
Plainly invoking
nvcc
isn't guaranteed to work:Luckily, I've already implemented such a mechanism as part of CUDAdrv. Maybe we should move this functionality, and the entire compilation example, over to CUDArt (since CUDAdrv doesn't require
nvcc
to be installed except for that example)?