| /external/tensorflow/tensorflow/lite/delegates/gpu/gl/workgroups/ |
| D | ideal_workgroup_picker.cc | 47 uint3 ideal_workgroup;
58 uint3 ideal_workgroup;
73 {OperationType::CONVOLUTION_2D, uint3(4, 8, 4)},
74 {OperationType::DEPTHWISE_CONVOLUTION, uint3(4, 4, 8)},
79 {OperationType::CONVOLUTION_2D, HW(1, 1), HW(1, 1), uint3(4, 8, 4)},
80 {OperationType::CONVOLUTION_2D, HW(3, 3), HW(2, 2), uint3(8, 4, 4)},
82 uint3(8, 4, 4)},
84 uint3(4, 4, 4)},
90 {OperationType::CONVOLUTION_2D, uint3(8, 2, 2)},
91 {OperationType::DEPTHWISE_CONVOLUTION, uint3(8, 8, 2)},
[all …]
|
| D | default_calculator.cc | 31 uint3 CalculateInternal(const ShaderCode& shader_code) const final { in CalculateInternal()
34 return uint3(4, 4, 64); in CalculateInternal()
37 return uint3(8, 4, 32); in CalculateInternal()
40 return uint3(8, 8, 16); in CalculateInternal()
43 return uint3(16, 8, 8); in CalculateInternal()
46 return uint3(16, 16, 4); in CalculateInternal()
49 return uint3(32, 16, 2); in CalculateInternal()
51 return uint3(32, 32, 1); in CalculateInternal()
59 uint3 CalculateInternal(const ShaderCode& shader_code) const final { in CalculateInternal()
62 return uint3(2, 2, 32); in CalculateInternal()
[all …]
|
| D | calculator.cc | 27 uint64_t CalculateProduct(const uint3& value) { in CalculateProduct()
31 void MaybeShrinkWorkgroup(const GpuInfo& gpu_info, uint3* wg) { in MaybeShrinkWorkgroup()
64 uint3 WorkgroupsCalculator::Calculate(const ShaderCode& shader_code) const { in Calculate()
65 uint3 workgroup_size = shader_code.recommended_workgroup; in Calculate()
|
| D | calculator.h | 29 constexpr uint3 kEmptyWorkgroupSize(0, 0, 0);
45 uint3 Calculate(const ShaderCode& shader_code) const;
48 virtual uint3 CalculateInternal(const ShaderCode& shader_code) const = 0;
|
| D | ideal_workgroup_picker.h | 31 uint3 GetIdealWorkgroupIfPossible(const GpuInfo& gpu_info,
38 uint3 GetIdealWorkgroupIfPossible(const GpuInfo& gpu_info,
40 uint3 default_wg, OHWI workload);
|
| /external/tensorflow/tensorflow/lite/delegates/gpu/common/ |
| D | memory_management_test.cc | 344 std::vector<TensorUsageRecord<uint3>> usage_records{ in TEST()
345 {/*size=*/uint3(1, 2, 8), /*first=*/0, /*last=*/1}, in TEST()
346 {/*size=*/uint3(4, 3, 2), /*first=*/1, /*last=*/2}, in TEST()
347 {/*size=*/uint3(1, 1, 1), /*first=*/2, /*last=*/4}, in TEST()
348 {/*size=*/uint3(2, 4, 1), /*first=*/3, /*last=*/5}, in TEST()
349 {/*size=*/uint3(2, 2, 2), /*first=*/4, /*last=*/5}, in TEST()
350 {/*size=*/uint3(8, 1, 2), /*first=*/5, /*last=*/7}, in TEST()
351 {/*size=*/uint3(1, 2, 1), /*first=*/6, /*last=*/8}, in TEST()
352 {/*size=*/uint3(1, 1, 1), /*first=*/7, /*last=*/8}, in TEST()
353 {/*size=*/uint3(2, 2, 2), /*first=*/8, /*last=*/9}}; in TEST()
[all …]
|
| D | util.h | 34 inline uint3 DivideRoundUp(uint3 n, uint3 divisor) { in DivideRoundUp()
35 return uint3(DivideRoundUp(n.x, divisor.x), DivideRoundUp(n.y, divisor.y), in DivideRoundUp()
|
| /external/tensorflow/tensorflow/lite/delegates/gpu/gl/kernels/ |
| D | add.cc | 62 /*workload=*/uint3(), in GenerateCode()
63 /*workgroup=*/uint3(), in GenerateCode()
85 /*workload=*/uint3(), in GenerateCode()
86 /*workgroup=*/uint3(), in GenerateCode()
99 /*workload=*/uint3(), in GenerateCode()
100 /*workgroup=*/uint3(), in GenerateCode()
112 uint3(ctx.input_shapes[0][2], ctx.input_shapes[0][1], in GenerateCode()
114 /*workgroup=*/uint3(), in GenerateCode()
|
| D | mul.cc | 73 /*workload=*/uint3(), in GenerateApplyMaskCode()
74 /*workgroup=*/uint3(), in GenerateApplyMaskCode()
91 /*workload=*/uint3(), in GenerateMultiplyScalarCode()
92 /*workgroup=*/uint3(), in GenerateMultiplyScalarCode()
110 uint3(static_cast<int>(ctx.input_shapes[0][2]), in GenerateMultiplyScalarCode()
113 /*workgroup=*/uint3(), in GenerateMultiplyScalarCode()
127 uint3(static_cast<int>(ctx.input_shapes[0][2]), in GenerateMultiplyScalarCode()
135 uint3(static_cast<int>(ctx.input_shapes[0][2]), in GenerateMultiplyScalarCode()
138 /*workgroup=*/uint3(), in GenerateMultiplyScalarCode()
|
| D | prelu.cc | 56 /*workload=*/uint3(), in GenerateCode()
57 /*workgroup=*/uint3(), in GenerateCode()
70 uint3(static_cast<int>(ctx.output_shapes[0][2]), in GenerateCode()
74 /*workgroup=*/uint3(), in GenerateCode()
102 uint3(static_cast<int>(ctx.output_shapes[0][2]), in GenerateCode()
117 uint3(static_cast<int>(ctx.output_shapes[0][2]), in GenerateCode()
121 /*workgroup=*/uint3(), in GenerateCode()
137 uint3(static_cast<int>(ctx.output_shapes[0][2]), in GenerateCode()
141 /*workgroup=*/uint3(), in GenerateCode()
|
| D | conv.cc | 134 /*workload=*/uint3(), in GenerateCode()
138 HW(weights.h, weights.w), attr.strides, uint3(0, 0, 0), in GenerateCode()
196 MakeReadonlyObject(uint3(4, DivideRoundUp(attr.weights.shape.i, 4), in GenerateCode()
235 uint3 workgroup = uint3(16, 16, 1); in GenerateCode()
238 workgroup = uint3(8, 8, 2); in GenerateCode()
241 workgroup = uint3(4, 8, 4); in GenerateCode()
244 workgroup = uint3(4, 4, 8); in GenerateCode()
247 workgroup = uint3(4, 4, 16); in GenerateCode()
250 workgroup = uint3(2, 8, 16); in GenerateCode()
254 workgroup = uint3(16, 8, 2); in GenerateCode()
[all …]
|
| D | resize.cc | 58 /*workload=*/uint3(), in GenerateCode()
59 /*workgroup=*/uint3(), in GenerateCode()
130 /*workload=*/uint3(), in GenerateCode()
131 /*workgroup=*/uint3(), in GenerateCode()
|
| D | softmax.cc | 156 /*workload=*/uint3(depth, 1, 1), in GenerateCodeFor1x1()
157 /*workgroup=*/uint3(32, 1, 1), in GenerateCodeFor1x1()
203 uint3(static_cast<int>(ctx.output_shapes[0][2]), in GenerateCodeGeneral()
205 /*workgroup=*/uint3(), in GenerateCodeGeneral()
|
| /external/tensorflow/tensorflow/compiler/xla/service/gpu/llvm_gpu_backend/tests_data/ |
| D | saxpy.ll | 4 %struct.uint3 = type { i32, i32, i32 }
7 @blockIdx = external addrspace(1) global %struct.uint3
9 @threadIdx = external addrspace(1) global %struct.uint3
35 …etelementptr inbounds (%struct.uint3, %struct.uint3* addrspacecast (%struct.uint3 addrspace(1)* @b…
38 …telementptr inbounds (%struct.uint3, %struct.uint3* addrspacecast (%struct.uint3 addrspace(1)* @th…
87 …etelementptr inbounds (%struct.uint3, %struct.uint3* addrspacecast (%struct.uint3 addrspace(1)* @b…
90 …telementptr inbounds (%struct.uint3, %struct.uint3* addrspacecast (%struct.uint3 addrspace(1)* @th…
|
| /external/angle/third_party/vulkan-deps/glslang/src/Test/ |
| D | hlsl.promotions.frag | 10 uniform uint3 u3;
21 void Fn_U3(uint3 x) { }
36 uint3 Fn_R_U3I(out uint3 p) { p = i3; return i3; }
37 uint3 Fn_R_U3F(out uint3 p) { p = f3; return f3; }
38 uint3 Fn_R_U3B(out uint3 p) { p = b3; return b3; }
39 uint3 Fn_R_U3D(out uint3 p) { p = d3; return d3; } // valid, but loss of precision on downconver…
64 uint3 r20 = b3;
65 uint3 r21 = i3;
66 uint3 r22 = f3;
67 uint3 r23 = d3; // valid, but loss of precision on downconversion.
|
| /external/deqp-deps/glslang/Test/ |
| D | hlsl.promotions.frag | 10 uniform uint3 u3;
21 void Fn_U3(uint3 x) { }
36 uint3 Fn_R_U3I(out uint3 p) { p = i3; return i3; }
37 uint3 Fn_R_U3F(out uint3 p) { p = f3; return f3; }
38 uint3 Fn_R_U3B(out uint3 p) { p = b3; return b3; }
39 uint3 Fn_R_U3D(out uint3 p) { p = d3; return d3; } // valid, but loss of precision on downconver…
64 uint3 r20 = b3;
65 uint3 r21 = i3;
66 uint3 r22 = f3;
67 uint3 r23 = d3; // valid, but loss of precision on downconversion.
|
| /external/tensorflow/tensorflow/lite/delegates/gpu/gl/converters/ |
| D | phwc4_to_bhwc.cc | 35 uint3 workgroup_size = uint3(4, 4, 4); in Create()
85 uint3 workload = uint3(shape.w, shape.h, shape.c); in Convert()
86 uint3 num_workgroups = DivideRoundUp(workload, workgroup_size_); in Convert()
|
| D | bhwc_to_phwc4.cc | 35 uint3 workgroup_size = uint3(4, 4, 4); in Create()
91 uint3 workload = uint3(shape.w, shape.h, DivideRoundUp(shape.c, 4)); in Convert()
92 uint3 num_workgroups = DivideRoundUp(workload, workgroup_size_); in Convert()
|
| /external/tensorflow/tensorflow/lite/delegates/gpu/gl/compiler/ |
| D | shader_code.h | 34 const std::vector<Object>& in_objects, const uint3& in_workload, in ShaderCode()
35 const uint3& in_recommended_workgroup, in ShaderCode()
51 uint3 workload;
54 uint3 recommended_workgroup;
|
| D | object_accessor_test.cc | 71 "obj", MakeReadonlyBuffer(uint3(1, 2, 3), std::vector<float>{1.0}))); in TEST()
82 "obj", MakeReadonlyBuffer(uint3(1, 2, 3), std::vector<float>{1.0}))); in TEST()
96 "obj", MakeReadonlyTexture(uint3(1, 2, 3), {1.0, 2.0, 3.0, 4.0}))); in TEST()
131 "obj", MakeReadonlyBuffer(uint3(1, 2, 3), {1.0, 2.0, 3.0, 4.0}))); in TEST()
145 "obj", MakeReadonlyTexture(uint3(1, 1, 1), {1.0, 2.0, 3.0, 4.0}))); in TEST()
179 "obj", MakeReadonlyTexture(uint3(1, 1, 1), {1.0, 2.0, 3.0, 4.0}))); in TEST()
189 "obj", MakeReadonlyTexture(uint3(1, 1, 1), {1.0, 2.0, 3.0, 4.0}))); in TEST()
|
| /external/llvm-project/clang/lib/Headers/ |
| D | __clang_cuda_builtin_vars.h | 14 struct uint3;
59 __attribute__((device)) operator uint3() const;
72 __attribute__((device)) operator uint3() const;
85 __attribute__((device)) operator uint3() const;
98 __attribute__((device)) operator uint3() const;
|
| /external/clang/test/SemaOpenCL/ |
| D | vector_conv_invalid.cl | 6 typedef unsigned uint3 __attribute((ext_vector_type(3)));
13 …uint3 u4 = (uint3)u; // expected-error{{invalid conversion between ext-vector type 'uint3' (vector…
|
| /external/tensorflow/tensorflow/lite/delegates/gpu/gl/ |
| D | serialization.h | 51 const uint3& workgroup_size, const uint3& num_workgroups,
74 const uint3& workgroup_size,
75 const uint3& num_workgroups,
|
| /external/angle/src/libANGLE/renderer/d3d/d3d11/shaders/ |
| D | Passthrough3D11.hlsl | 67 uint3 size;
75 uint3 size;
88 uint3 size;
96 uint3 size;
109 uint3 size;
117 uint3 size;
130 uint3 size;
138 uint3 size;
|
| /external/tensorflow/tensorflow/lite/delegates/gpu/common/memory_management/ |
| D | internal.h | 57 bool IsCoveringObject(const uint3& first_object, const uint3& second_object);
66 size_t AbsDiffInElements(const uint3& first_size, const uint3& second_size);
|