| /external/tensorflow/tensorflow/lite/delegates/gpu/gl/workgroups/ |
| D | calculator.cc | 65 uint3 workgroup_size = shader_code.recommended_workgroup; in Calculate() local
66 if (workgroup_size == kEmptyWorkgroupSize) { in Calculate()
67 workgroup_size = CalculateInternal(shader_code); in Calculate()
69 MaybeShrinkWorkgroup(gpu_info_, &workgroup_size); in Calculate()
70 return workgroup_size; in Calculate()
|
| /external/tensorflow/tensorflow/lite/delegates/gpu/gl/ |
| D | api.cc | 177 uint3 workgroup_size; member
203 uint3 workgroup_size = workgroup_calculator.Calculate(code); in Add() local
204 uint3 num_workgroups = DivideRoundUp(code.workload, workgroup_size); in Add()
215 AddFullShader(code.source_code, workgroup_size, &shader_idx)); in Add()
219 workgroup_size, in Add()
229 const uint3& workgroup_size, size_t* size) { in AddFullShader() argument
230 std::string shader_src = GetShaderHeader(workgroup_size) + partial_shader; in AddFullShader()
284 const uint3& workgroup_size, in OnProgram() argument
295 workgroup_size, &shader_idx)); in OnProgram()
299 workgroup_size, in OnProgram()
[all …]
|
| D | serialization_test.cc | 42 uint3 workgroup_size; member
55 const uint3& workgroup_size, in OnProgram()
59 {parameters, objects, workgroup_size, num_workgroups, shader_index}); in OnProgram()
185 uint3 workgroup_size(1, 2, 3); in TEST() local
186 builder.AddProgram(parameters, objects, workgroup_size, num_workgroups, 1); in TEST()
194 EXPECT_EQ(workgroup_size.data_, handler.programs[0].workgroup_size.data_); in TEST()
|
| D | serialization.h | 51 const uint3& workgroup_size, const uint3& num_workgroups,
74 const uint3& workgroup_size,
|
| D | serialization.cc | 327 const uint3& workgroup_size, const uint3& num_workgroups, in AddProgram() argument
330 Offset<data::Uint3> fb_workgroup_size = Encode(workgroup_size, &builder_); in AddProgram()
604 uint3 workgroup_size(program->workgroup_size()->x(), in DeserializeCompiledModel() local
605 program->workgroup_size()->y(), in DeserializeCompiledModel()
606 program->workgroup_size()->z()); in DeserializeCompiledModel()
610 RETURN_IF_ERROR(handler->OnProgram(parameters, objects, workgroup_size, in DeserializeCompiledModel()
|
| D | compiled_model.fbs | 36 workgroup_size:Uint3;
|
| /external/tensorflow/tensorflow/lite/delegates/gpu/gl/converters/ |
| D | phwc4_to_bhwc.cc | 35 uint3 workgroup_size = uint3(4, 4, 4); in Create() local
36 std::string shader_source = GetShaderHeader(workgroup_size) + R"( in Create()
64 *converter = ConverterPhwc4ToBhwc(std::move(program), workgroup_size); in Create()
|
| D | bhwc_to_phwc4.cc | 35 uint3 workgroup_size = uint3(4, 4, 4); in Create() local
36 std::string shader_source = GetShaderHeader(workgroup_size) + R"( in Create()
71 *converter = ConverterBhwcToPhwc4(std::move(program), workgroup_size); in Create()
|
| D | bhwc_to_phwc4.h | 42 explicit ConverterBhwcToPhwc4(GlProgram program, const uint3& workgroup_size) in ConverterBhwcToPhwc4() argument
43 : program_(std::move(program)), workgroup_size_(workgroup_size) {} in ConverterBhwcToPhwc4()
|
| D | phwc4_to_bhwc.h | 42 explicit ConverterPhwc4ToBhwc(GlProgram program, const uint3& workgroup_size) in ConverterPhwc4ToBhwc() argument
43 : program_(std::move(program)), workgroup_size_(workgroup_size) {} in ConverterPhwc4ToBhwc()
|
| /external/mesa3d/src/amd/compiler/ |
| D | aco_instruction_selection_setup.cpp | 1108 program->workgroup_size = program->wave_size; in setup_isel_context()
1111 program->workgroup_size = shaders[0]->info.cs.local_size[0] * in setup_isel_context()
1116 program->workgroup_size = program->wave_size; in setup_isel_context()
1122 uint32_t workgroup_size = MAX2(es_verts_per_subgrp, gs_instr_prims_in_subgrp); in setup_isel_context() local
1123 program->workgroup_size = MAX2(MIN2(workgroup_size, 256), 1); in setup_isel_context()
1126 …program->workgroup_size = UINT_MAX; /* TODO: probably tcs_num_patches * tcs_vertices_in, but those… in setup_isel_context()
1130 program->workgroup_size = ctx.tcs_num_patches * shaders[0]->info.tess.tcs_vertices_out; in setup_isel_context()
1134 …program->workgroup_size = ctx.tcs_num_patches * MAX2(shaders[1]->info.tess.tcs_vertices_out, ctx.a… in setup_isel_context()
1148 program->workgroup_size = MAX4(max_esverts, max_gs_input_prims, max_out_vtx, max_out_prm); in setup_isel_context()
|
| D | aco_live_var_analysis.cpp | 259 unsigned workgroup_size = program->workgroup_size == UINT_MAX in calc_waves_per_workgroup() local
261 : program->workgroup_size; in calc_waves_per_workgroup()
263 return align(workgroup_size, program->wave_size) / program->wave_size; in calc_waves_per_workgroup()
|
| D | aco_insert_waitcnt.cpp | 453 …sync_scope subgroup_scope = ctx.program->workgroup_size <= ctx.program->wave_size ? scope_workgrou… in perform_barrier()
|
| D | aco_ir.h | 1646 unsigned workgroup_size; /* if known; otherwise UINT_MAX */ variable
|
| D | aco_lower_to_hw_instr.cpp | 1998 program->workgroup_size > program->wave_size; in lower_to_hw_instr()
|
| D | aco_instruction_selection.cpp | 3993 if (ctx->program->workgroup_size <= ctx->program->wave_size) in thread_id_in_threadgroup()
11363 if (ctx->program->workgroup_size <= ctx->program->wave_size) in ngg_gs_workgroup_reduce_and_scan()
11970 program->workgroup_size = 1; /* XXX */ in select_trap_handler_shader()
|
| /external/tensorflow/tensorflow/lite/delegates/gpu/gl/kernels/ |
| D | converter.cc | 56 absl::Status InitializeProgram(const uint3& workgroup_size, in InitializeProgram() argument
58 workgroup_size_ = workgroup_size; in InitializeProgram()
61 GL_COMPUTE_SHADER, GetShaderHeader(workgroup_size) + shader_source, in InitializeProgram()
|
| /external/llvm-project/mlir/test/Conversion/GPUToSPIRV/ |
| D | builtins.mlir | 89 // We may want to define gpu.workgroup_size and convert it to the entry
|
| /external/OpenCL-CTS/test_conformance/subgroups/ |
| D | subhelpers.h | 1429 int subgroup_size, int &workgroup_size,
1434 workgroup_size = non_uniform_size;
|
| /external/mesa3d/src/intel/compiler/ |
| D | brw_fs.h | 440 unsigned workgroup_size() const;
|
| D | brw_fs_nir.cpp | 3749 workgroup_size() <= dispatch_width) { in nir_emit_cs_intrinsic()
4351 slm_fence && workgroup_size() <= dispatch_width) in nir_emit_intrinsic()
|
| D | brw_fs.cpp | 9451 fs_visitor::workgroup_size() const in workgroup_size() function in fs_visitor
|
| /external/mesa3d/src/amd/vulkan/ |
| D | radv_shader.c | 1604 unsigned workgroup_size = local_size[0] * local_size[1] * local_size[2]; in radv_GetShaderInfoAMD() local
1607 ceil((double)workgroup_size / statistics.numPhysicalVgprs); in radv_GetShaderInfoAMD()
|
| /external/llvm-project/llvm/docs/ |
| D | AMDGPUUsage.rst | 4289 workgroup_size.x - 1) /
4290 workgroup_size.x).
4297 workgroup_size.y - 1) /
4308 workgroup_size.z - 1) /
|
| /external/mesa3d/docs/relnotes/ |
| D | 20.0.0.rst | 647 - intel/fs: Add workgroup_size() helper
|