[cmake-developers] CUDA as a compiler instead FindCUDA?

Adam Strzelecki ono at java.pl
Fri Jan 16 06:54:13 EST 2015

Hi. It makes me curious why CUDA support is not just implemented as compiler via proper CMakeDetermineCUDACompiler and other files defining how to handle .cu files, but rather it uses just awkward 3rd party functions, cmake script wrappers, etc, which basically obscure whole process and instead using "ninja -v" to see invocation of "nvcc" you need CUDA_VERBOSE_BUILD=YES which makes me more and more sick every time I use it.

Is it some kind of legacy reasons? If so is CMake open to accept patch that will make CUDA support via new compiler definition?


More information about the cmake-developers mailing list