cuda.cu 1.3 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243
  1. /* StarPU --- Runtime system for heterogeneous multicore architectures.
  2. *
  3. * Copyright (C) 2012 INRIA
  4. *
  5. * StarPU is free software; you can redistribute it and/or modify
  6. * it under the terms of the GNU Lesser General Public License as published by
  7. * the Free Software Foundation; either version 2.1 of the License, or (at
  8. * your option) any later version.
  9. *
  10. * StarPU is distributed in the hope that it will be useful, but
  11. * WITHOUT ANY WARRANTY; without even the implied warranty of
  12. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
  13. *
  14. * See the GNU Lesser General Public License in COPYING.LGPL for more details.
  15. */
  16. #include <starpu.h>
  17. #include "custom_types.h"
  18. #include "custom_interface.h"
  19. static __global__ void scal_cuda(unsigned n,
  20. float *x,
  21. float *y)
  22. {
  23. unsigned i = blockIdx.x*blockDim.x + threadIdx.x;
  24. if (i < n)
  25. x[i] *= y[i];
  26. }
  27. extern "C" void custom_scal_cuda_func(void *buffers[], void *_args)
  28. {
  29. (void) _args;
  30. unsigned int n = CUSTOM_GET_NX(buffers[0]);
  31. float *x = (float*) CUSTOM_GET_X_PTR(buffers[0]);
  32. float *y = (float*) CUSTOM_GET_Y_PTR(buffers[0]);
  33. unsigned threads_per_block = 64;
  34. unsigned nblocks = (n + threads_per_block-1) / threads_per_block;
  35. scal_cuda<<<nblocks,threads_per_block,2,starpu_cuda_get_local_stream()>>>(n, x, y);
  36. }