Home
last modified time | relevance | path

Searched refs:block_x (Results 1 – 25 of 31) sorted by relevance

12

/external/tensorflow/tensorflow/lite/delegates/gpu/common/tasks/
Dfully_connected.h72 for (int block_x = 0; 4 * block_x < padded_src_channels; block_x++) { in RearrangeFCWeightsToIOO4I4() local
75 int x = 4 * block_x + x_in_block; in RearrangeFCWeightsToIOO4I4()
78 int dst_index = block_x * padded_dst_channels * 4 + block_y * 16 + in RearrangeFCWeightsToIOO4I4()
/external/ComputeLibrary/src/core/CL/cl_kernels/nhwc/
Dbatch_to_space.cl67 const int block_x = *((__global int *)vector_offset(&block, 0));
70 const int r = (BATCH_SIZE / (block_x * block_y));
76 const int out_x = x * block_x + (batch_id / r) % block_x;
77 const int out_y = y * block_y + (batch_id / r) / block_x;
117 const int block_x = BLOCK_SHAPE_X;
120 const int r = (BATCH_SIZE / (block_x * block_y));
126 const int out_x = x * block_x + (batch_id / r) % block_x;
127 const int out_y = y * block_y + (batch_id / r) / block_x;
Dspace_to_batch.cl77 int block_x = *((__global int *)vector_offset(&block, 0));
84 const int pos_x = out_x * block_x + ((batch_id / BATCH_IN) % block_x);
85 const int pos_y = out_y * block_y + ((batch_id / BATCH_IN) / block_x);
136 int block_x = BLOCK_SHAPE_X;
143 const int pos_x = out_x * block_x + ((batch_id / BATCH_IN) % block_x);
144 const int pos_y = out_y * block_y + ((batch_id / BATCH_IN) / block_x);
/external/ComputeLibrary/src/core/CL/cl_kernels/nchw/
Dbatch_to_space.cl67 const int block_x = *((__global int *)vector_offset(&block, 0));
70 const int r = (BATCH_SIZE / (block_x * block_y));
76 const int out_x = x * block_x + (batch_id / r) % block_x;
77 const int out_y = y * block_y + (batch_id / r) / block_x;
117 const int block_x = BLOCK_SHAPE_X;
120 const int r = (BATCH_SIZE / (block_x * block_y));
126 const int out_x = x * block_x + (batch_id / r) % block_x;
127 const int out_y = y * block_y + (batch_id / r) / block_x;
Dspace_to_batch.cl77 int block_x = *((__global int *)vector_offset(&block, 0));
84 const int pos_x = out_x * block_x + ((batch_id / BATCH_IN) % block_x);
85 const int pos_y = out_y * block_y + ((batch_id / BATCH_IN) / block_x);
137 int block_x = BLOCK_SHAPE_X;
144 const int pos_x = out_x * block_x + ((batch_id / BATCH_IN) % block_x);
145 const int pos_y = out_y * block_y + ((batch_id / BATCH_IN) / block_x);
/external/tensorflow/tensorflow/lite/delegates/gpu/common/tasks/special/
Dfc_fc_add.h53 for (int block_x = 0; 4 * block_x < padded_src_channels; block_x++) { in RearrangeFCWeightsToIOO4I4() local
56 int x = 4 * block_x + x_in_block; in RearrangeFCWeightsToIOO4I4()
57 int dst_index = block_x * padded_dst_channels * 4 + block_y * 16 + in RearrangeFCWeightsToIOO4I4()
/external/pdfium/third_party/libopenjpeg/
Dsparse_array.c127 OPJ_UINT32 x, block_x; in opj_sparse_array_int32_read_or_write() local
134 block_x = x0 / block_width; in opj_sparse_array_int32_read_or_write()
135 for (x = x0; x < x1; block_x ++, x += x_incr) { in opj_sparse_array_int32_read_or_write()
142 src_block = sa->data_blocks[block_y * sa->block_count_hor + block_x]; in opj_sparse_array_int32_read_or_write()
242 sa->data_blocks[block_y * sa->block_count_hor + block_x] = src_block; in opj_sparse_array_int32_read_or_write()
/external/tensorflow/tensorflow/compiler/xla/mlir_hlo/tests/Dialect/gml_st/
Dlegacy_tiling.mlir2 // RUN: -test-gml-st-loop-tiling="tile-sizes=2,3,4 distribution-types=block_x,block_y,none" \
35 // CHECK-SAME: distribution["block_x", "block_y", "none"] {
92 // CHECK-SAME: distribution["block_x", "block_y", "none"] {
/external/ComputeLibrary/src/core/CL/cl_kernels/common/
Dgemm_reshaped_only_rhs_mmul.cl101 uint block_x = thread_id % MMUL_N0;
105 uint dst_x = min(block_x * N0 + block_id * MMUL_N0 * N0, (uint)(N - 1));
112 uint lhs_x = block_x;
116 uint rhs_x = block_y * N0 * MMUL_N0 + block_x * N0;
161 if(block_x * N0 + block_id * MMUL_N0 * N0 >= N)
354 uint block_x = thread_id % MMUL_N0;
358 uint dst_x = min(block_x * N0 + block_id * MMUL_N0 * N0, (uint)(N - 1));
365 uint lhs_x = block_x;
369 uint rhs_x = block_y * N0 * MMUL_N0 + block_x * N0;
410 if(block_x * N0 + block_id * MMUL_N0 * N0 >= N)
Dgemmlowp_reshaped_only_rhs_mmul.cl122 uint block_x = thread_id % MMUL_N0;
126 uint dst_x = min(block_x * N0 + block_id * MMUL_N0 * N0, (uint)(N - 1));
129 uint lhs_x = VEC_SIZE * block_x;
133 uint rhs_y = 4 * block_id + block_x;
168 if(block_x * N0 + block_id * MMUL_N0 * N0 >= N)
/external/tensorflow/tensorflow/core/profiler/backends/gpu/
Dcupti_collector.h94 uint32 block_x; member
119 kernel_info.grid_z, " block:", kernel_info.block_x, ",", in ToXStat()
Drocm_tracer.cc463 event.kernel_info.block_x = data->args.hipModuleLaunchKernel.blockDimX; in AddKernelEventUponApiExit()
487 event.kernel_info.block_x = blockDimX; in AddKernelEventUponApiExit()
506 event.kernel_info.block_x = data->args.hipHccModuleLaunchKernel.blockDimX; in AddKernelEventUponApiExit()
511 event.kernel_info.block_x; in AddKernelEventUponApiExit()
532 event.kernel_info.block_x = data->args.hipLaunchKernel.dimBlocks.x; in AddKernelEventUponApiExit()
Drocm_tracer.h63 uint32_t block_x; member
Ddevice_tracer_rocm.cc604 kernel_info.grid_z, " block:", kernel_info.block_x, ",", in ToXStat()
717 params.block_size = static_cast<int>(event.kernel_info.block_x * in CreateXEvent()
/external/ComputeLibrary/arm_compute/core/utils/misc/
DShapeCalculator.h1081 inline TensorShape compute_batch_to_space_shape(const ITensorInfo *input, const int block_x, const … in compute_batch_to_space_shape() argument
1083 ARM_COMPUTE_ERROR_ON(block_x <= 0 || block_y <= 0); in compute_batch_to_space_shape()
1091 output_shape.set(idx_width, input->tensor_shape()[idx_width] * block_x); in compute_batch_to_space_shape()
1093 output_shape.set(idx_batch, input->tensor_shape()[idx_batch] / (block_x * block_y)); in compute_batch_to_space_shape()
1165 inline TensorShape compute_space_to_batch_shape(const ITensorInfo *input, const int block_x, const … in compute_space_to_batch_shape() argument
1174 …ERROR_ON((input->tensor_shape()[idx_width] + padding_left.x() + padding_right.x()) % block_x != 0); in compute_space_to_batch_shape()
1177 …et(idx_width, (input->tensor_shape()[idx_width] + padding_left.x() + padding_right.x()) / block_x); in compute_space_to_batch_shape()
1179 output_shape.set(idx_batch, input->tensor_shape()[idx_batch] * block_x * block_y); in compute_space_to_batch_shape()
/external/tensorflow/tensorflow/core/profiler/utils/
Dkernel_stats_utils_test.cc75 kernel_info.block_x = 32; in TEST()
/external/ComputeLibrary/cl_kernels/nhwc/
Dbatch_to_space.clembed1259 const int block_x = *((__global int *)vector_offset(&block, 0));
1262 const int r = (BATCH_SIZE / (block_x * block_y));
1268 const int out_x = x * block_x + (batch_id / r) % block_x;
1269 const int out_y = y * block_y + (batch_id / r) / block_x;
1285 const int block_x = BLOCK_SHAPE_X;
1288 const int r = (BATCH_SIZE / (block_x * block_y));
1294 const int out_x = x * block_x + (batch_id / r) % block_x;
1295 const int out_y = y * block_y + (batch_id / r) / block_x;
Dspace_to_batch.clembed1266 int block_x = *((__global int *)vector_offset(&block, 0));
1273 const int pos_x = out_x * block_x + ((batch_id / BATCH_IN) % block_x);
1274 const int pos_y = out_y * block_y + ((batch_id / BATCH_IN) / block_x);
1297 int block_x = BLOCK_SHAPE_X;
1304 const int pos_x = out_x * block_x + ((batch_id / BATCH_IN) % block_x);
1305 const int pos_y = out_y * block_y + ((batch_id / BATCH_IN) / block_x);
/external/ComputeLibrary/cl_kernels/nchw/
Dbatch_to_space.clembed1259 const int block_x = *((__global int *)vector_offset(&block, 0));
1262 const int r = (BATCH_SIZE / (block_x * block_y));
1268 const int out_x = x * block_x + (batch_id / r) % block_x;
1269 const int out_y = y * block_y + (batch_id / r) / block_x;
1285 const int block_x = BLOCK_SHAPE_X;
1288 const int r = (BATCH_SIZE / (block_x * block_y));
1294 const int out_x = x * block_x + (batch_id / r) % block_x;
1295 const int out_y = y * block_y + (batch_id / r) / block_x;
Dspace_to_batch.clembed1266 int block_x = *((__global int *)vector_offset(&block, 0));
1273 const int pos_x = out_x * block_x + ((batch_id / BATCH_IN) % block_x);
1274 const int pos_y = out_y * block_y + ((batch_id / BATCH_IN) / block_x);
1298 int block_x = BLOCK_SHAPE_X;
1305 const int pos_x = out_x * block_x + ((batch_id / BATCH_IN) % block_x);
1306 const int pos_y = out_y * block_y + ((batch_id / BATCH_IN) / block_x);
/external/tensorflow/tensorflow/compiler/xla/mlir_hlo/include/mlir-hlo/Dialect/gml_st/IR/
Dgml_st_legacy_ops.td55 distribution("block_x") {
86 distribution("block_x") {
/external/tensorflow/tensorflow/core/kernels/
Dsegment_reduction_ops_gpu.cu.h372 int block_x = std::min(ninner_pow2, static_cast<Tindex>(max_block_size));
378 block_x,
379 std::min(static_cast<Tindex>(Eigen::divup(min_block_size, block_x)),
/external/llvm/test/CodeGen/ARM/
D2007-01-19-InfiniteLoop.ll12 define fastcc i32 @dct_luma_sp(i32 %block_x, i32 %block_y, i32* %coeff_cost) {
/external/mesa3d/src/gallium/drivers/softpipe/
Dsp_setup.c190 block_x(int x) in block_x() function
209 const int minleft = block_x(MIN2(xleft0, xleft1)); in flush_spans()
/external/ComputeLibrary/cl_kernels/common/
Dgemm_reshaped_only_rhs_mmul.clembed3436 uint block_x = thread_id % MMUL_N0;
3440 uint dst_x = min(block_x * N0 + block_id * MMUL_N0 * N0, (uint)(N - 1));
3447 uint lhs_x = block_x;
3451 uint rhs_x = block_y * N0 * MMUL_N0 + block_x * N0;
3496 if(block_x * N0 + block_id * MMUL_N0 * N0 >= N)
3640 uint block_x = thread_id % MMUL_N0;
3644 uint dst_x = min(block_x * N0 + block_id * MMUL_N0 * N0, (uint)(N - 1));
3651 uint lhs_x = block_x;
3655 uint rhs_x = block_y * N0 * MMUL_N0 + block_x * N0;
3696 if(block_x * N0 + block_id * MMUL_N0 * N0 >= N)

12