Home
last modified time | relevance | path

Searched refs:kernel_info (Results 1 – 25 of 49) sorted by relevance

12

/external/tensorflow/tensorflow/core/profiler/utils/
Dkernel_stats_utils_test.cc71 KernelDetails kernel_info; in TEST() local
72 kernel_info.registers_per_thread = 10; in TEST()
73 kernel_info.static_shared_memory_usage = 128; in TEST()
74 kernel_info.dynamic_shared_memory_usage = 256; in TEST()
75 kernel_info.block_x = 32; in TEST()
76 kernel_info.block_y = 8; in TEST()
77 kernel_info.block_z = 4; in TEST()
78 kernel_info.grid_x = 3; in TEST()
79 kernel_info.grid_y = 2; in TEST()
80 kernel_info.grid_z = 1; in TEST()
[all …]
/external/tensorflow/tensorflow/lite/delegates/gpu/common/task/
Dwork_group_picking.cc36 int multiplier, int3 grid, const KernelInfo& kernel_info, in GenerateWorkGroupSizesXYMultipleOf() argument
43 for (int x = 1; x <= kernel_info.max_work_group_size; x *= 2) { in GenerateWorkGroupSizesXYMultipleOf()
44 for (int y = 1; y <= kernel_info.max_work_group_size; y *= 2) { in GenerateWorkGroupSizesXYMultipleOf()
47 work_group_size_xy > kernel_info.max_work_group_size) { in GenerateWorkGroupSizesXYMultipleOf()
51 if (work_group_size_xy * z > kernel_info.max_work_group_size) { in GenerateWorkGroupSizesXYMultipleOf()
66 int multiplier, int3 grid, const KernelInfo& kernel_info, in GenerateWorkGroupSizesXMultipleOf() argument
76 x <= kernel_info.max_work_group_size && x < grid.x + multiplier; in GenerateWorkGroupSizesXMultipleOf()
83 x * y * z <= kernel_info.max_work_group_size) { in GenerateWorkGroupSizesXMultipleOf()
93 const KernelInfo& kernel_info, const int3& grid, in GetWorkGroupsAlignedToGrid() argument
100 grid, max_wg_size, kernel_info.max_work_group_size, work_groups); in GetWorkGroupsAlignedToGrid()
[all …]
Dwork_group_picking.h32 const KernelInfo& kernel_info,
38 const KernelInfo& kernel_info,
51 const KernelInfo& kernel_info, const int3& grid,
55 const KernelInfo& kernel_info, const int3& grid,
Dgpu_operation.h115 const KernelInfo& kernel_info, std::vector<int3>* work_groups) const;
120 const KernelInfo& kernel_info) { in PostCompileCheck() argument
Dgpu_operation.cc211 const KernelInfo& kernel_info, std::vector<int3>* work_groups) const { in GetPossibleKernelWorkGroups() argument
212 GetPossibleWorkGroups(tuning_type, gpu_info, kernel_info, grid_size_, in GetPossibleKernelWorkGroups()
/external/tensorflow/tensorflow/core/profiler/internal/gpu/
Dcupti_collector.h98 inline std::string ToXStat(const KernelDetails& kernel_info, in ToXStat() argument
101 "regs:", kernel_info.registers_per_thread, in ToXStat()
102 " static_shared:", kernel_info.static_shared_memory_usage, in ToXStat()
103 " dynamic_shared:", kernel_info.dynamic_shared_memory_usage, in ToXStat()
104 " grid:", kernel_info.grid_x, ",", kernel_info.grid_y, ",", in ToXStat()
105 kernel_info.grid_z, " block:", kernel_info.block_x, ",", in ToXStat()
106 kernel_info.block_y, ",", kernel_info.block_z, in ToXStat()
171 KernelDetails kernel_info; member
Dcupti_collector.cc182 static_cast<int>(event.kernel_info.registers_per_thread); in CreateXEvent()
184 event.kernel_info.static_shared_memory_usage; in CreateXEvent()
188 params.block_size = static_cast<int>(event.kernel_info.block_x * in CreateXEvent()
189 event.kernel_info.block_y * in CreateXEvent()
190 event.kernel_info.block_z); in CreateXEvent()
192 params.dynamic_smem_size = event.kernel_info.dynamic_shared_memory_usage; in CreateXEvent()
210 event.kernel_info, occ_stats.occupancy_pct))); in CreateXEvent()
379 kernel_name, " regs:", event.kernel_info.registers_per_thread, in Flush()
380 " shm:", event.kernel_info.static_shared_memory_usage, in Flush()
381 " grid: ", event.kernel_info.grid_x, ",", in Flush()
[all …]
/external/ImageMagick/MagickCore/
Deffect.c777 *kernel_info; in BlurImage() local
795 kernel_info=AcquireKernelInfo(geometry,exception); in BlurImage()
796 if (kernel_info == (KernelInfo *) NULL) in BlurImage()
798 blur_image=ConvolveImage(image,kernel_info,exception); in BlurImage()
799 kernel_info=DestroyKernelInfo(kernel_info); in BlurImage()
1171 const KernelInfo *kernel_info,ExceptionInfo *exception) in ConvolveImage() argument
1177 convolve_image=AccelerateConvolveImage(image,kernel_info,exception); in ConvolveImage()
1182 convolve_image=MorphologyImage(image,ConvolveMorphology,1,kernel_info, in ConvolveImage()
1532 *kernel_info; in EdgeImage() local
1547 kernel_info=AcquireKernelInfo((const char *) NULL,exception); in EdgeImage()
[all …]
Dcompare.c1379 *kernel_info; in GetStructuralSimilarityDistortion() local
1408 kernel_info=AcquireKernelInfo(geometry,exception); in GetStructuralSimilarityDistortion()
1409 if (kernel_info == (KernelInfo *) NULL) in GetStructuralSimilarityDistortion()
1445 p=GetCacheViewVirtualPixels(image_view,-((ssize_t) kernel_info->width/2L),y- in GetStructuralSimilarityDistortion()
1446 ((ssize_t) kernel_info->height/2L),columns+kernel_info->width, in GetStructuralSimilarityDistortion()
1447 kernel_info->height,exception); in GetStructuralSimilarityDistortion()
1448 q=GetCacheViewVirtualPixels(reconstruct_view,-((ssize_t) kernel_info->width/ in GetStructuralSimilarityDistortion()
1449 2L),y-((ssize_t) kernel_info->height/2L),columns+kernel_info->width, in GetStructuralSimilarityDistortion()
1450 kernel_info->height,exception); in GetStructuralSimilarityDistortion()
1489 k=kernel_info->values; in GetStructuralSimilarityDistortion()
[all …]
/external/tensorflow/tensorflow/lite/delegates/gpu/common/tasks/
Dwinograd.h36 const KernelInfo& kernel_info, in GetPossibleKernelWorkGroups() argument
70 const KernelInfo& kernel_info,
89 int3 SelectBestWorkGroup(const KernelInfo& kernel_info) const;
103 const KernelInfo& kernel_info, in GetPossibleKernelWorkGroups() argument
136 const KernelInfo& kernel_info,
156 int3 SelectBestWorkGroup(const KernelInfo& kernel_info) const;
Dwinograd.cc440 const KernelInfo& kernel_info) const { in SelectBestWorkGroup()
444 return GetFirstSuitableWorkGroup(wgs, kernel_info.max_work_group_size); in SelectBestWorkGroup()
469 const KernelInfo& kernel_info, std::vector<int3>* work_groups) const { in GetPossibleKernelWorkGroups() argument
476 GetPossibleWorkGroups(tuning_type, gpu_info, kernel_info, grid_size_, in GetPossibleKernelWorkGroups()
481 work_groups->push_back(SelectBestWorkGroup(kernel_info)); in GetPossibleKernelWorkGroups()
659 const KernelInfo& kernel_info) const { in SelectBestWorkGroup()
663 return GetFirstSuitableWorkGroup(wgs, kernel_info.max_work_group_size); in SelectBestWorkGroup()
683 const KernelInfo& kernel_info, std::vector<int3>* work_groups) const { in GetPossibleKernelWorkGroups() argument
690 GetPossibleWorkGroups(tuning_type, gpu_info, kernel_info, grid_size_, in GetPossibleKernelWorkGroups()
695 work_groups->push_back(SelectBestWorkGroup(kernel_info)); in GetPossibleKernelWorkGroups()
Ddepthwise_conv_3x3_stride_h2.cc199 const KernelInfo& kernel_info, std::vector<int3>* work_groups) const { in GetPossibleKernelWorkGroups() argument
203 GetPossibleWorkGroups(tuning_type, gpu_info, kernel_info, grid_size_, in GetPossibleKernelWorkGroups()
Dmean_stddev_normalization.h36 const KernelInfo& kernel_info, in GetPossibleKernelWorkGroups() argument
Dsoftmax1x1.h31 const KernelInfo& kernel_info, in GetPossibleKernelWorkGroups() argument
Dreduce.h38 const KernelInfo& kernel_info,
Dconvolution_transposed_3x3.h42 const KernelInfo& kernel_info,
Dconvolution_transposed_4x4.h42 const KernelInfo& kernel_info, in GetPossibleKernelWorkGroups() argument
Ddepthwise_conv_3x3.cc293 const KernelInfo& kernel_info, std::vector<int3>* work_groups) const { in GetPossibleKernelWorkGroups() argument
297 GetPossibleWorkGroups(tuning_type, gpu_info, kernel_info, grid_size_, in GetPossibleKernelWorkGroups()
Dconv_metal.h59 const KernelInfo& kernel_info, in GetPossibleKernelWorkGroups() argument
/external/tensorflow/tensorflow/lite/delegates/gpu/metal/
Dcompute_task.cc244 KernelInfo kernel_info; in Tune() local
245 kernel_info.max_work_group_size = [program_ maxTotalThreadsPerThreadgroup]; in Tune()
246 kernel_info.private_memory_size = 0; in Tune()
248 kernel_info, &possible_work_groups); in Tune()
/external/llvm-project/openmp/libomptarget/plugins/amdgpu/impl/
Datmi_interop_hsa.cpp50 hsa_executable_symbol_info_t kernel_info, uint32_t *value) { in atmi_interop_hsa_get_kernel_info() argument
74 switch (kernel_info) { in atmi_interop_hsa_get_kernel_info()
/external/ImageMagick/MagickWand/
Dconvert.c1050 *kernel_info; in ConvertImageCommand() local
1057 kernel_info=AcquireKernelInfo(argv[i],exception); in ConvertImageCommand()
1058 if (kernel_info == (KernelInfo *) NULL) in ConvertImageCommand()
1060 kernel_info=DestroyKernelInfo(kernel_info); in ConvertImageCommand()
1189 *kernel_info; in ConvertImageCommand() local
1196 kernel_info=AcquireKernelInfo(argv[i],exception); in ConvertImageCommand()
1197 if (kernel_info == (KernelInfo *) NULL) in ConvertImageCommand()
1199 kernel_info=DestroyKernelInfo(kernel_info); in ConvertImageCommand()
2263 *kernel_info; in ConvertImageCommand() local
2279 kernel_info=AcquireKernelInfo(argv[i],exception); in ConvertImageCommand()
[all …]
/external/arm-trusted-firmware/plat/mediatek/common/
Dmtk_plat_common.h41 struct kernel_info { struct
Dmtk_plat_common.c40 static struct kernel_info k_info;
/external/ImageMagick/Magick++/lib/
DImage.cpp2421 *kernel_info; in colorMatrix() local
2424 kernel_info=AcquireKernelInfo((const char *) NULL,exceptionInfo); in colorMatrix()
2425 if (kernel_info != (KernelInfo *) NULL) in colorMatrix()
2427 kernel_info->width=order_; in colorMatrix()
2428 kernel_info->height=order_; in colorMatrix()
2429 kernel_info->values=(MagickRealType *) AcquireAlignedMemory(order_, in colorMatrix()
2430 order_*sizeof(*kernel_info->values)); in colorMatrix()
2431 if (kernel_info->values != (MagickRealType *) NULL) in colorMatrix()
2437 kernel_info->values[i]=color_matrix_[i]; in colorMatrix()
2438 newImage=ColorMatrixImage(image(),kernel_info,exceptionInfo); in colorMatrix()
[all …]

12