diff options
author | Cedric Nugteren <web@cedricnugteren.nl> | 2017-03-05 15:06:14 +0100 |
---|---|---|
committer | Cedric Nugteren <web@cedricnugteren.nl> | 2017-03-05 15:06:14 +0100 |
commit | b114ea49a9228ee7a8bf1b00c092324c0ce972c3 (patch) | |
tree | 5aa34777226bdb11c6349c9ce43de160478d9b41 /include/clblast_c.h | |
parent | cdf354f89524ed88d4f9358004e5a8eabd9ce286 (diff) |
Added first naive version of the batched AXPY routine
Diffstat (limited to 'include/clblast_c.h')
-rw-r--r-- | include/clblast_c.h | 21 |
1 files changed, 11 insertions, 10 deletions
diff --git a/include/clblast_c.h b/include/clblast_c.h index f933ef6c..5c84b5d7 100644 --- a/include/clblast_c.h +++ b/include/clblast_c.h @@ -96,6 +96,7 @@ typedef enum CLBlastStatusCode_ { CLBlastInsufficientMemoryY = -1007, // Vector Y's OpenCL buffer is too small // Custom additional status codes for CLBlast + CLBlastInvalidBatchCount = -2049, // The batch count needs to be positive CLBlastInvalidOverrideKernel = -2048, // Trying to override parameters for an invalid kernel CLBlastMissingOverrideParameter = -2047, // Missing override parameter(s) for the target kernel CLBlastInvalidLocalMemUsage = -2046, // Not enough local memory available on this device @@ -1330,32 +1331,32 @@ CLBlastStatusCode PUBLIC_API CLBlastHomatcopy(const CLBlastLayout layout, const // Batched version of AXPY: SAXPYBATCHED/DAXPYBATCHED/CAXPYBATCHED/ZAXPYBATCHED/HAXPYBATCHED CLBlastStatusCode PUBLIC_API CLBlastSaxpyBatched(const size_t n, const float *alphas, - const cl_mem *x_buffers, const size_t x_offset, const size_t x_inc, - cl_mem *y_buffers, const size_t y_offset, const size_t y_inc, + const cl_mem *x_buffers, const size_t x_inc, + cl_mem *y_buffers, const size_t y_inc, const size_t batch_count, cl_command_queue* queue, cl_event* event); CLBlastStatusCode PUBLIC_API CLBlastDaxpyBatched(const size_t n, const double *alphas, - const cl_mem *x_buffers, const size_t x_offset, const size_t x_inc, - cl_mem *y_buffers, const size_t y_offset, const size_t y_inc, + const cl_mem *x_buffers, const size_t x_inc, + cl_mem *y_buffers, const size_t y_inc, const size_t batch_count, cl_command_queue* queue, cl_event* event); CLBlastStatusCode PUBLIC_API CLBlastCaxpyBatched(const size_t n, const cl_float2 *alphas, - const cl_mem *x_buffers, const size_t x_offset, const size_t x_inc, - cl_mem *y_buffers, const size_t y_offset, const size_t y_inc, + const cl_mem *x_buffers, const size_t x_inc, + cl_mem *y_buffers, const size_t y_inc, const size_t batch_count, cl_command_queue* queue, cl_event* event); CLBlastStatusCode PUBLIC_API CLBlastZaxpyBatched(const size_t n, const cl_double2 *alphas, - const cl_mem *x_buffers, const size_t x_offset, const size_t x_inc, - cl_mem *y_buffers, const size_t y_offset, const size_t y_inc, + const cl_mem *x_buffers, const size_t x_inc, + cl_mem *y_buffers, const size_t y_inc, const size_t batch_count, cl_command_queue* queue, cl_event* event); CLBlastStatusCode PUBLIC_API CLBlastHaxpyBatched(const size_t n, const cl_half *alphas, - const cl_mem *x_buffers, const size_t x_offset, const size_t x_inc, - cl_mem *y_buffers, const size_t y_offset, const size_t y_inc, + const cl_mem *x_buffers, const size_t x_inc, + cl_mem *y_buffers, const size_t y_inc, const size_t batch_count, cl_command_queue* queue, cl_event* event); |