Commit d8b192c8 authored by yao's avatar yao

Fix the mismatch on NV GPUs

parent 4162ebfa
...@@ -277,9 +277,15 @@ __kernel void arithm_mul_D6 (__global double *src1, int src1_step, int src1_offs ...@@ -277,9 +277,15 @@ __kernel void arithm_mul_D6 (__global double *src1, int src1_step, int src1_offs
} }
#endif #endif
#ifdef DOUBLE_SUPPORT
#define SCALAR_TYPE double
#else
#define SCALAR_TYPE float
#endif
__kernel void arithm_muls_D5 (__global float *src1, int src1_step, int src1_offset, __kernel void arithm_muls_D5 (__global float *src1, int src1_step, int src1_offset,
__global float *dst, int dst_step, int dst_offset, __global float *dst, int dst_step, int dst_offset,
int rows, int cols, int dst_step1, float scalar) int rows, int cols, int dst_step1, SCALAR_TYPE scalar)
{ {
int x = get_global_id(0); int x = get_global_id(0);
int y = get_global_id(1); int y = get_global_id(1);
......
...@@ -472,4 +472,8 @@ void ocl_tvl1flow::warpBackward(const oclMat &I0, const oclMat &I1, oclMat &I1x, ...@@ -472,4 +472,8 @@ void ocl_tvl1flow::warpBackward(const oclMat &I0, const oclMat &I1, oclMat &I1x,
args.push_back( make_pair( sizeof(cl_int), (void*)&u2_offset_y)); args.push_back( make_pair( sizeof(cl_int), (void*)&u2_offset_y));
openCLExecuteKernel(clCxt, &tvl1flow, kernelName, globalThread, localThread, args, -1, -1); openCLExecuteKernel(clCxt, &tvl1flow, kernelName, globalThread, localThread, args, -1, -1);
releaseTexture(I1_tex);
releaseTexture(I1x_tex);
releaseTexture(I1y_tex);
} }
\ No newline at end of file
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment