Is it necessary to create an interface in CUDA Fortran to call cudaMallocPitch

Accelerated Computing HPC Compilers nvc, nvc++ and nvfortran
1

Hello CUDA Forum, I would like to inquire whether it is necessary to write a structure to call cudaMallocPitch from C++ when writing CUDA Fortran code?

2

I assume you mean interface like the title of the post?

If so, then no. You just need to use the cudafor module.

For it’s use, please see: CUDA Fortran Programming Guide Version 24.11 for ARM, OpenPower, x86