opencv/modules/core/src
Chuanbo Weng 2d8c89c40b Remove unnecessary kercn limitation of 4.
When accessing global memory by DWORD4, memory bandwidth
can be fully utilized on Intel platform. This patch will
make more image format(e.g. 8UC4) be processed in DWORD4
by work-item. After applying this patch, 3 subcase of
./opencv_perf_core --gtest_filter=OCL_RepeatFixture_Repeat.Repeat/*
can be speedup on HD4000 graphics card with Beignet:
OCL_RepeatFixture_Repeat.Repeat/2, 64% improvement.
OCL_RepeatFixture_Repeat.Repeat/6, 50% improvement.
OCL_RepeatFixture_Repeat.Repeat/8, 56% improvement.

Signed-off-by: Chuanbo Weng <chuanbo.weng@intel.com>
2014-12-04 11:15:13 +08:00
..
cuda datastart must be non-const in CUDA. 2014-08-13 11:11:33 +04:00
opencl Removed incorrect using of rootn() and powr() in ocl_pow 2014-11-06 16:23:02 +03:00
algorithm.cpp quickly corrected the previous refactoring of features2d: moved from set(SOME_PROP, val) to setSomeProp(val) 2014-10-18 20:44:26 +04:00
alloc.cpp Merge remote-tracking branch 'origin/2.4' 2013-08-06 13:56:49 +04:00
arithm.cpp Implementation detector and selector for IPP and OpenCL; 2014-10-15 14:24:41 +04:00
array.cpp Merge remote-tracking branch 'origin/2.4' into merge-2.4 2014-03-11 17:20:01 +04:00
bufferpool.impl.hpp core/ocl: OpenCLBufferPool 2014-02-04 16:48:35 +04:00
command_line_parser.cpp fixed defects from coverity.com 2014-02-23 20:21:08 +04:00
conjugate_gradient.cpp removed optim module; moved its functionality to core and photo modules; moved drawing functions from core to imgproc. Removed FilterEngine etc. from public API 2014-08-14 12:50:07 +04:00
convert.cpp Optimize OpenCL version of conversScaleAbs function 2014-10-21 19:20:20 +04:00
copy.cpp Remove unnecessary kercn limitation of 4. 2014-12-04 11:15:13 +08:00
cuda_buffer_pool.cpp modified default stream initialization to allow concurrent calls modified cuda surf.cuda.cpp to allow concurrent call 2014-07-23 14:21:21 +02:00
cuda_gpu_mat.cpp datastart must be non-const in CUDA. 2014-08-13 11:11:33 +04:00
cuda_host_mem.cpp datastart must be non-const in CUDA. 2014-08-13 11:11:33 +04:00
cuda_info.cpp renamed gpu* source to cuda* in core module 2013-09-02 14:00:42 +04:00
cuda_stream.cpp trailing whitespace 2014-07-23 14:34:22 +02:00
datastructs.cpp removed optim module; moved its functionality to core and photo modules; moved drawing functions from core to imgproc. Removed FilterEngine etc. from public API 2014-08-14 12:50:07 +04:00
directx.cpp cleaned up super_resolution sample from old OCL 2014-02-05 00:28:10 +04:00
directx.inc.hpp core: added DirectX interop implementation (OpenCL) with samples 2013-12-13 19:02:41 +04:00
downhill_simplex.cpp removed optim module; moved its functionality to core and photo modules; moved drawing functions from core to imgproc. Removed FilterEngine etc. from public API 2014-08-14 12:50:07 +04:00
dxt.cpp Implementation detector and selector for IPP and OpenCL; 2014-10-15 14:24:41 +04:00
gl_core_3_1.cpp Merge remote-tracking branch 'origin/2.4' into merge-2.4 2014-02-17 15:24:05 +04:00
gl_core_3_1.hpp updated license header in whole gpu module 2013-03-22 14:03:15 +04:00
glob.cpp deleted excess semicolons, commas 2014-01-18 01:31:33 +04:00
kmeans.cpp removed optim module; moved its functionality to core and photo modules; moved drawing functions from core to imgproc. Removed FilterEngine etc. from public API 2014-08-14 12:50:07 +04:00
lapack.cpp Several type of formal refactoring: 2014-08-13 15:21:35 +04:00
lda.cpp added some basic functionality needed by the new face module (moved from the old "contrib") 2014-08-10 01:54:16 +04:00
lpsolver.cpp trying to fix builds 2014-08-14 13:18:04 +04:00
mathfuncs.cpp Removed incorrect using of rootn() and powr() in ocl_pow 2014-11-06 16:23:02 +03:00
matmul.cpp Some small fixes 2014-10-27 14:38:22 +03:00
matop.cpp the first draft of transparent API and new UMat class. more files 2013-10-22 14:05:15 +04:00
matrix.cpp Implementation detector and selector for IPP and OpenCL; 2014-10-15 14:24:41 +04:00
ocl.cpp Fix incorrect size computation 2014-11-26 12:24:53 +03:00
opengl.cpp unused function in case of unavailable OpenGL 2014-02-01 18:14:10 +04:00
out.cpp Fix for (Bug #2789) 2014-04-16 10:56:27 +09:00
parallel.cpp Fixing iOS clang warnings, part 2 2014-10-17 18:14:54 +04:00
pca.cpp removed optim module; moved its functionality to core and photo modules; moved drawing functions from core to imgproc. Removed FilterEngine etc. from public API 2014-08-14 12:50:07 +04:00
persistence.cpp Fixing iOS clang warnings, part 2 2014-10-17 18:14:54 +04:00
precomp.hpp Implementation detector and selector for IPP and OpenCL; 2014-10-15 14:24:41 +04:00
rand.cpp Several type of formal refactoring: 2014-08-13 15:21:35 +04:00
stat.cpp IPPICV: disable NormDiff_L1_16s_C1R for IPP/ICV 8.2/8.2.1 2014-11-05 13:26:23 +03:00
stl.cpp War on Whitespace, master edition: file endings. 2013-08-27 14:23:26 +04:00
system.cpp Implementation detector and selector for IPP and OpenCL; 2014-10-15 14:24:41 +04:00
tables.cpp removed duplicated Hershey fonts 2014-08-15 12:09:16 +04:00
types.cpp War on Whitespace, master edition: file endings. 2013-08-27 14:23:26 +04:00
umatrix.cpp Implementation detector and selector for IPP and OpenCL; 2014-10-15 14:24:41 +04:00