Skip to content
Projects
Groups
Snippets
Help
Loading...
Sign in / Register
Toggle navigation
O
opencv
Project
Project
Details
Activity
Cycle Analytics
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Charts
Issues
0
Issues
0
List
Board
Labels
Milestones
Merge Requests
0
Merge Requests
0
CI / CD
CI / CD
Pipelines
Jobs
Schedules
Charts
Packages
Packages
Wiki
Wiki
Snippets
Snippets
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Charts
Create a new issue
Jobs
Commits
Issue Boards
Open sidebar
submodule
opencv
Commits
c8a54f67
Commit
c8a54f67
authored
Aug 22, 2012
by
marina.kolpakova
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
minor warning fix
parent
d395bf15
Hide whitespace changes
Inline
Side-by-side
Showing
10 changed files
with
24 additions
and
24 deletions
+24
-24
bf_radius_match.cu
modules/gpu/src/cuda/bf_radius_match.cu
+2
-2
canny.cu
modules/gpu/src/cuda/canny.cu
+2
-2
column_filter.cu
modules/gpu/src/cuda/column_filter.cu
+1
-1
fast.cu
modules/gpu/src/cuda/fast.cu
+2
-2
hist.cu
modules/gpu/src/cuda/hist.cu
+1
-1
integral_image.cu
modules/gpu/src/cuda/integral_image.cu
+2
-2
matrix_reductions.cu
modules/gpu/src/cuda/matrix_reductions.cu
+8
-8
pyrlk.cu
modules/gpu/src/cuda/pyrlk.cu
+4
-4
row_filter.cu
modules/gpu/src/cuda/row_filter.cu
+1
-1
NPP_staging.cu
modules/gpu/src/nvidia/NPP_staging/NPP_staging.cu
+1
-1
No files found.
modules/gpu/src/cuda/bf_radius_match.cu
View file @
c8a54f67
...
...
@@ -56,7 +56,7 @@ namespace cv { namespace gpu { namespace device
__global__ void matchUnrolled(const DevMem2D_<T> query, int imgIdx, const DevMem2D_<T> train, float maxDistance, const Mask mask,
PtrStepi bestTrainIdx, PtrStepi bestImgIdx, PtrStepf bestDistance, unsigned int* nMatches, int maxCount)
{
#if
__CUDA_ARCH__ >= 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 110)
extern __shared__ int smem[];
...
...
@@ -168,7 +168,7 @@ namespace cv { namespace gpu { namespace device
__global__ void match(const DevMem2D_<T> query, int imgIdx, const DevMem2D_<T> train, float maxDistance, const Mask mask,
PtrStepi bestTrainIdx, PtrStepi bestImgIdx, PtrStepf bestDistance, unsigned int* nMatches, int maxCount)
{
#if
__CUDA_ARCH__ >= 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 110)
extern __shared__ int smem[];
...
...
modules/gpu/src/cuda/canny.cu
View file @
c8a54f67
...
...
@@ -261,7 +261,7 @@ namespace cv { namespace gpu { namespace device
__global__ void edgesHysteresisLocal(PtrStepi map, ushort2* st, int rows, int cols)
{
#if
__CUDA_ARCH__ >= 120
#if
defined (__CUDA_ARCH__) && (__CUDA_ARCH__ >= 120)
__shared__ int smem[18][18];
...
...
@@ -358,7 +358,7 @@ namespace cv { namespace gpu { namespace device
__global__ void edgesHysteresisGlobal(PtrStepi map, ushort2* st1, ushort2* st2, int rows, int cols, int count)
{
#if __CUDA_ARCH__ >= 120
#if
defined (__CUDA_ARCH__) &&
__CUDA_ARCH__ >= 120
const int stack_size = 512;
...
...
modules/gpu/src/cuda/column_filter.cu
View file @
c8a54f67
...
...
@@ -64,7 +64,7 @@ namespace cv { namespace gpu { namespace device
template <int KSIZE, typename T, typename D, typename B>
__global__ void linearColumnFilter(const DevMem2D_<T> src, PtrStep<D> dst, const int anchor, const B brd)
{
#if
__CUDA_ARCH__ >= 200
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 200)
const int BLOCK_DIM_X = 16;
const int BLOCK_DIM_Y = 16;
const int PATCH_PER_BLOCK = 4;
...
...
modules/gpu/src/cuda/fast.cu
View file @
c8a54f67
...
...
@@ -223,7 +223,7 @@ namespace cv { namespace gpu { namespace device
template <bool calcScore, class Mask>
__global__ void calcKeypoints(const DevMem2Db img, const Mask mask, short2* kpLoc, const unsigned int maxKeypoints, PtrStepi score, const int threshold)
{
#if
__CUDA_ARCH__ >= 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 110)
const int j = threadIdx.x + blockIdx.x * blockDim.x + 3;
const int i = threadIdx.y + blockIdx.y * blockDim.y + 3;
...
...
@@ -325,7 +325,7 @@ namespace cv { namespace gpu { namespace device
__global__ void nonmaxSupression(const short2* kpLoc, int count, const DevMem2Di scoreMat, short2* locFinal, float* responseFinal)
{
#if
__CUDA_ARCH__ >= 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 110)
const int kpIdx = threadIdx.x + blockIdx.x * blockDim.x;
...
...
modules/gpu/src/cuda/hist.cu
View file @
c8a54f67
...
...
@@ -63,7 +63,7 @@ namespace cv { namespace gpu { namespace device
#define MERGE_THREADBLOCK_SIZE 256
#define USE_SMEM_ATOMICS (
__CUDA_ARCH__ >= 120
)
#define USE_SMEM_ATOMICS (
defined (__CUDA_ARCH__) && (__CUDA_ARCH__ >= 120)
)
namespace hist
{
...
...
modules/gpu/src/cuda/integral_image.cu
View file @
c8a54f67
...
...
@@ -59,7 +59,7 @@ namespace cv { namespace gpu { namespace device
__global__ void shfl_integral_horizontal(const PtrStep_<uint4> img, PtrStep_<uint4> integral)
{
#if
__CUDA_ARCH__ >= 300
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 300)
__shared__ int sums[128];
const int id = threadIdx.x;
...
...
@@ -299,7 +299,7 @@ namespace cv { namespace gpu { namespace device
// block sums.
__global__ void shfl_integral_vertical(DevMem2D_<unsigned int> integral)
{
#if
__CUDA_ARCH__ >= 300
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 300)
__shared__ unsigned int sums[32][9];
const int tidx = blockIdx.x * blockDim.x + threadIdx.x;
...
...
modules/gpu/src/cuda/matrix_reductions.cu
View file @
c8a54f67
...
...
@@ -215,7 +215,7 @@ namespace cv { namespace gpu { namespace device
maxval[blockIdx.y * gridDim.x + blockIdx.x] = (T)smaxval[0];
}
#if
__CUDA_ARCH__ >= 110
#if
defined (__CUDA_ARCH__) && (__CUDA_ARCH__ >= 110)
__shared__ bool is_last;
if (tid == 0)
...
...
@@ -535,7 +535,7 @@ namespace cv { namespace gpu { namespace device
findMinMaxLocInSmem<nthreads, best_type>(sminval, smaxval, sminloc, smaxloc, tid);
#if
__CUDA_ARCH__ >= 110
#if
defined (__CUDA_ARCH__) && (__CUDA_ARCH__ >= 110)
__shared__ bool is_last;
if (tid == 0)
...
...
@@ -841,7 +841,7 @@ namespace cv { namespace gpu { namespace device
sumInSmem<nthreads, uint>(scount, tid);
#if
__CUDA_ARCH__ >= 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 110)
__shared__ bool is_last;
if (tid == 0)
...
...
@@ -1034,7 +1034,7 @@ namespace cv { namespace gpu { namespace device
sumInSmem<nthreads, R>(smem, tid);
#if
__CUDA_ARCH__ >= 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 110)
__shared__ bool is_last;
if (tid == 0)
...
...
@@ -1115,7 +1115,7 @@ namespace cv { namespace gpu { namespace device
sumInSmem<nthreads, R>(smem, tid);
sumInSmem<nthreads, R>(smem + nthreads, tid);
#if
__CUDA_ARCH__ >= 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 110)
__shared__ bool is_last;
if (tid == 0)
...
...
@@ -1222,7 +1222,7 @@ namespace cv { namespace gpu { namespace device
sumInSmem<nthreads, R>(smem + nthreads, tid);
sumInSmem<nthreads, R>(smem + 2 * nthreads, tid);
#if __CUDA_ARCH__ >= 110
#if
defined(__CUDA_ARCH__) &&
__CUDA_ARCH__ >= 110
__shared__ bool is_last;
if (tid == 0)
...
...
@@ -1339,7 +1339,7 @@ namespace cv { namespace gpu { namespace device
sumInSmem<nthreads, R>(smem + 2 * nthreads, tid);
sumInSmem<nthreads, R>(smem + 3 * nthreads, tid);
#if
__CUDA_ARCH__ >= 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 110)
__shared__ bool is_last;
if (tid == 0)
...
...
@@ -1975,7 +1975,7 @@ namespace cv { namespace gpu { namespace device
for (int c = 0; c < cn; ++c)
myVal[c] = op.startValue();
#if __CUDA_ARCH__ >= 200
#if
defined (__CUDA_ARCH__) &&
__CUDA_ARCH__ >= 200
// For cc >= 2.0 prefer L1 cache
for (int x = threadIdx.x; x < src.cols; x += 256)
...
...
modules/gpu/src/cuda/pyrlk.cu
View file @
c8a54f67
...
...
@@ -82,7 +82,7 @@ namespace cv { namespace gpu { namespace device
smem3[tid] = val3;
__syncthreads();
#if
__CUDA_ARCH__ > 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ > 110)
if (tid < 128)
{
smem1[tid] = val1 += smem1[tid + 128];
...
...
@@ -138,7 +138,7 @@ namespace cv { namespace gpu { namespace device
smem2[tid] = val2;
__syncthreads();
#if
__CUDA_ARCH__ > 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ > 110)
if (tid < 128)
{
smem1[tid] = val1 += smem1[tid + 128];
...
...
@@ -184,7 +184,7 @@ namespace cv { namespace gpu { namespace device
smem1[tid] = val1;
__syncthreads();
#if
__CUDA_ARCH__ > 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ > 110)
if (tid < 128)
{
smem1[tid] = val1 += smem1[tid + 128];
...
...
@@ -271,7 +271,7 @@ namespace cv { namespace gpu { namespace device
template <int cn, int PATCH_X, int PATCH_Y, bool calcErr>
__global__ void lkSparse(const float2* prevPts, float2* nextPts, uchar* status, float* err, const int level, const int rows, const int cols)
{
#if
__CUDA_ARCH__ <= 110
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ <= 110)
__shared__ float smem1[128];
__shared__ float smem2[128];
__shared__ float smem3[128];
...
...
modules/gpu/src/cuda/row_filter.cu
View file @
c8a54f67
...
...
@@ -64,7 +64,7 @@ namespace cv { namespace gpu { namespace device
template <int KSIZE, typename T, typename D, typename B>
__global__ void linearRowFilter(const DevMem2D_<T> src, PtrStep<D> dst, const int anchor, const B brd)
{
#if
__CUDA_ARCH__ >= 200
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 200)
const int BLOCK_DIM_X = 32;
const int BLOCK_DIM_Y = 8;
const int PATCH_PER_BLOCK = 4;
...
...
modules/gpu/src/nvidia/NPP_staging/NPP_staging.cu
View file @
c8a54f67
...
...
@@ -2070,7 +2070,7 @@ NCVStatus nppiStInterpolateFrames(const NppStInterpolationState *pState)
//==============================================================================
#if
__CUDA_ARCH__ < 200
#if
defined(__CUDA_ARCH__) && (__CUDA_ARCH__ < 200)
// FP32 atomic add
static __forceinline__ __device__ float _atomicAdd(float *addr, float val)
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment