scal_cuda.cu 1.4 KB

12345678910111213141516171819202122232425262728293031323334353637383940
  1. /* StarPU --- Runtime system for heterogeneous multicore architectures.
  2. *
  3. * Copyright (C) 2010, 2011 Centre National de la Recherche Scientifique
  4. * Copyright (C) 2010-2011 Université de Bordeaux 1
  5. *
  6. * StarPU is free software; you can redistribute it and/or modify
  7. * it under the terms of the GNU Lesser General Public License as published by
  8. * the Free Software Foundation; either version 2.1 of the License, or (at
  9. * your option) any later version.
  10. *
  11. * StarPU is distributed in the hope that it will be useful, but
  12. * WITHOUT ANY WARRANTY; without even the implied warranty of
  13. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
  14. *
  15. * See the GNU Lesser General Public License in COPYING.LGPL for more details.
  16. */
  17. #include <starpu.h>
  18. #include <starpu_cuda.h>
  19. static __global__ void vector_mult_cuda(unsigned *val, unsigned n)
  20. {
  21. unsigned i = blockIdx.x*blockDim.x + threadIdx.x;
  22. if (i < n)
  23. val[i] *= 2;
  24. }
  25. extern "C" void scal_func_cuda(void *buffers[], void *_args)
  26. {
  27. unsigned n = STARPU_VECTOR_GET_NX(buffers[0]);
  28. unsigned *val = (unsigned *)STARPU_VECTOR_GET_PTR(buffers[0]);
  29. unsigned threads_per_block = 64;
  30. unsigned nblocks = (n + threads_per_block-1) / threads_per_block;
  31. vector_mult_cuda<<<nblocks,threads_per_block,0,starpu_cuda_get_local_stream()>>>(val, n);
  32. cudaStreamSynchronize(starpu_cuda_get_local_stream());
  33. }