Issue with cuda vs cudav2

This issue has been reported by Narasinga Rao Miniskar:

Dear Chameleon team, I am trying to use Chameleon for heterogeneous GPUs. However, I am getting cublas conflict error as it includes both cublas.h and cublas_v2.h I am using CUDA 12.0 for Nvidia A100 GPUs.

make[2]: *** [gpucublas/compute/CMakeFiles/gpucublas.dir/build.make:1434: gpucublas/compute/CMakeFiles/gpucublas.dir/cuda_strsm.c.o] Error 1
make[2]: *** [gpucublas/compute/CMakeFiles/gpucublas.dir/build.make:1364: gpucublas/compute/CMakeFiles/gpucublas.dir/cuda_ctrmm.c.o] Error 1
In file included from /software/chameleon/gpucublas/include/gpucublas.h:40,
                 from /software/chameleon/build/gpucublas/compute/cuda_zunmlqt.c:21:
/usr/local/cuda-12.0/include/cublas_v2.h:59:2: error: #error "It is an error to include both cublas.h and cublas_v2.h"
   59 | #error "It is an error to include both cublas.h and cublas_v2.h"
      |  ^~~~~
make[2]: *** [gpucublas/compute/CMakeFiles/gpucublas.dir/build.make:1490: gpucublas/compute/CMakeFiles/gpucublas.dir/cuda_stsmlq.c.o] Error 1
In file included from /software/chameleon/gpucublas/include/gpucublas.h:40,
                 from /software/chameleon/build/gpucublas/compute/cuda_cunmqrt.c:21:
/usr/local/cuda-12.0/include/cublas_v2.h:59:2: error: #error "It is an error to include both cublas.h and cublas_v2.h"
   59 | #error "It is an error to include both cublas.h and cublas_v2.h"
      |  ^~~~~

--

With Regards, Narasinga Rao Miniskar, Ph.D. Oak Ridge National Laboratory, Oak Ridge, TN 37831.