CUDA and autoconf

Accelerated Computing CUDA CUDA Programming and Performance
1

Hi,

I am trying to test some typical cuda functions during the configure process. How can I write it in my configure.ac? Something like:

AC_TRY_COMPILE([],

[

__global__ static void test_cuda() {

	const int tid = threadIdx.x;

	const int bid = blockIdx.x;

	__syncthreads();

}

],

[cuda_comp=ok],[cuda_comp=no])

But nvcc is not defined in AC_LANG. Must I create my own m4 macros?

2

Here my solution. This is not too hard to find this solution, as this is more autoconf relevant than CUDA, but CUDA coders may not be autoconf experts.

AC_MSG_CHECKING([for nvcc compilation])

ac_compile_nvcc=no

	cat>conftest.cu<<EOF

	__global__ static void test_cuda() {

	const int tid = threadIdx.x;

	const int bid = blockIdx.x;

	__syncthreads();

	}

EOF

	if $NVCC -c $NVCCFLAGS conftest.cu &> /dev/null

	then

		ac_compile_nvcc=yes

	fi

	AC_MSG_RESULT([$ac_compile_nvcc])

if test "x$ac_compile_nvcc" = "xno" 

then

	AC_MSG_ERROR([CUDA compiler has problems.])

fi