1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
|
#include "Halide.h"
namespace {
using Halide::float16_t;
enum class SomeEnum { Foo,
Bar };
class MetadataTester : public Halide::Generator<MetadataTester> {
public:
Input<Func> input{"input"}; // must be overridden to {UInt(8), 3}
Input<Buffer<uint8_t, 3>> typed_input_buffer{"typed_input_buffer"};
Input<Buffer<void, 3>> dim_only_input_buffer{"dim_only_input_buffer"}; // must be overridden to type=UInt(8)
Input<Buffer<>> untyped_input_buffer{"untyped_input_buffer"}; // must be overridden to {UInt(8), 3}
Input<int32_t> no_default_value{"no_default_value"};
Input<bool> b{"b", true};
Input<int8_t> i8{"i8", 8, -8, 127};
Input<int16_t> i16{"i16", 16, -16, 127};
Input<int32_t> i32{"i32", 32, -32, 127};
Input<int64_t> i64{"i64", 64, -64, 127};
Input<uint8_t> u8{"u8", 80, 8, 255};
Input<uint16_t> u16{"u16", 160, 16, 2550};
Input<uint32_t> u32{"u32", 320, 32, 2550};
Input<uint64_t> u64{"u64", 640, 64, 2550};
Input<float> f32{"f32", 32.1234f, -3200.1234f, 3200.1234f};
Input<double> f64{"f64", 64.25f, -6400.25f, 6400.25f};
Input<void *> h{"h", nullptr};
Input<Func> input_not_nod{"input_not_nod"}; // must be overridden to type=uint8 dim=3
Input<Func> input_nod{"input_nod", UInt(8)}; // must be overridden to type=uint8 dim=3
Input<Func> input_not{"input_not", 3}; // must be overridden to type=uint8
Input<Func[]> array_input{"array_input", UInt(8), 3}; // must be overridden to size=2
Input<Func[2]> array2_input{"array2_input", UInt(8), 3};
Input<int8_t[]> array_i8{"array_i8"}; // must be overridden to size=2
Input<int8_t[2]> array2_i8{"array2_i8"};
Input<int16_t[]> array_i16{"array_i16", 16}; // must be overridden to size=2
Input<int16_t[2]> array2_i16{"array2_i16", 16};
Input<int32_t[]> array_i32{"array_i32", 32, -32, 127}; // must be overridden to size=2
Input<int32_t[2]> array2_i32{"array2_i32", 32, -32, 127};
Input<void *[]> array_h{"array_h", nullptr}; // must be overridden to size=2
Input<Buffer<float, 3>[2]> buffer_array_input1{"buffer_array_input1"};
Input<Buffer<float>[2]> buffer_array_input2{"buffer_array_input2"}; // buffer_array_input2.dim must be set
Input<Buffer<void, 3>[2]> buffer_array_input3{"buffer_array_input3"}; // buffer_array_input2.type must be set
Input<Buffer<>[2]> buffer_array_input4{"buffer_array_input4"}; // dim and type must be set
// .size must be specified for all of these
Input<Buffer<float, 3>[]> buffer_array_input5{"buffer_array_input5"};
Input<Buffer<float>[]> buffer_array_input6{"buffer_array_input6"}; // buffer_array_input2.dim must be set
Input<Buffer<void, 3>[]> buffer_array_input7{"buffer_array_input7"}; // buffer_array_input2.type must be set
Input<Buffer<>[]> buffer_array_input8{"buffer_array_input8"}; // dim and type must be set
Input<Buffer<float16_t, 1>> buffer_f16_typed{"buffer_f16_typed"};
Input<Buffer<void, 1>> buffer_f16_untyped{"buffer_f16_untyped"};
Input<Expr> untyped_scalar_input{"untyped_scalar_input"}; // untyped_scalar_input.type must be set to uint8
Output<Func> output{"output"}; // must be overridden to {{Float(32), Float(32)}, 3}
Output<Buffer<float, 3>> typed_output_buffer{"typed_output_buffer"};
Output<Buffer<float>> type_only_output_buffer{"type_only_output_buffer"}; // untyped outputs can have type and/or dimensions inferred
Output<Buffer<void, 3>> dim_only_output_buffer{"dim_only_output_buffer"}; // untyped outputs can have type and/or dimensions inferred
Output<Buffer<>> untyped_output_buffer{"untyped_output_buffer"}; // untyped outputs can have type and/or dimensions inferred
Output<Buffer<void, 3>> tupled_output_buffer{"tupled_output_buffer", {Float(32), Int(32)}};
Output<float> output_scalar{"output_scalar"};
Output<Func[]> array_outputs{"array_outputs", Float(32), 3}; // must be overridden to size=2
Output<Func[2]> array_outputs2{"array_outputs2", {Float(32), Float(32)}, 3};
Output<float[2]> array_outputs3{"array_outputs3"};
Output<Buffer<float, 3>[2]> array_outputs4{"array_outputs4"};
Output<Buffer<float>[2]> array_outputs5{"array_outputs5"}; // dimensions will be inferred by usage
Output<Buffer<>[2]> array_outputs6{"array_outputs6"}; // dimensions and type will be inferred by usage
// .size must be specified for all of these
Output<Buffer<float, 3>[]> array_outputs7{"array_outputs7"};
Output<Buffer<float>[]> array_outputs8{"array_outputs8"};
Output<Buffer<>[]> array_outputs9{"array_outputs9"};
// Output<void> untyped_scalar_output{"untyped_scalar_output"}; // untyped_scalar_output.type must be set
void generate() {
Var x("x"), y("y"), c("c");
assert(buffer_f16_untyped.type() == Float(16));
assert(untyped_scalar_input.type() == UInt(8));
// These should all be zero; they are here to exercise the operator[] overloads
Expr zero1 = array_input[1](x, y, c) - array_input[0](x, y, c);
Expr zero2 = array_i32[1] - array_i32[0];
Expr bzero1 = buffer_array_input1[1](x, y, c) - buffer_array_input1[0](x, y, c);
Expr bzero2 = buffer_array_input2[1](x, y, c) - buffer_array_input2[0](x, y, c);
Expr bzero3 = buffer_array_input3[1](x, y, c) - buffer_array_input3[0](x, y, c);
Expr bzero4 = buffer_array_input4[1](x, y, c) - buffer_array_input4[0](x, y, c);
Expr bzero5 = buffer_array_input5[1](x, y, c) - buffer_array_input5[0](x, y, c);
Expr bzero6 = buffer_array_input6[1](x, y, c) - buffer_array_input6[0](x, y, c);
Expr bzero7 = buffer_array_input7[1](x, y, c) - buffer_array_input7[0](x, y, c);
Expr bzero8 = buffer_array_input8[1](x, y, c) - buffer_array_input8[0](x, y, c);
Expr zero = zero1 + zero2 + bzero1 + bzero2 + bzero3 + bzero4 + bzero5 + bzero6 + bzero7 + bzero8;
assert(output.types().size() == 2);
Type output_type = output.types().at(0);
Func f1("f1"), f2("f2");
f1(x, y, c) = cast(output_type, input(x, y, c) + zero + untyped_scalar_input);
f2(x, y, c) = cast<float>(f1(x, y, c) + 1);
Func t1("t1");
t1(x, y, c) = Tuple(f1(x, y, c), f2(x, y, c));
output = t1;
typed_output_buffer(x, y, c) = f1(x, y, c);
type_only_output_buffer(x, y, c) = f1(x, y, c);
dim_only_output_buffer(x, y, c) = f1(x, y, c);
tupled_output_buffer(x, y, c) = Tuple(f2(x, y, c), cast<int32_t>(f2(x, y, c) + 1.5f));
// verify that we can assign a Func to an Output<Buffer<>>
untyped_output_buffer = f2;
output_scalar() = 1234.25f;
for (size_t i = 0; i < array_outputs.size(); ++i) {
array_outputs[i](x, y, c) = (i + 1) * 1.5f;
Func z1("z1");
z1(x, y, c) = Tuple((i + 1) * 1.5f, 42.f);
array_outputs2[i] = z1;
array_outputs3[i]() = 42.f;
array_outputs4[i](x, y, c) = cast<float>(x + y + c + (int)i);
array_outputs5[i](x, y, c) = cast<float>(x + y + c + (int)i);
array_outputs6[i](x, y, c) = cast<float>(x + y + c + (int)i);
array_outputs7[i](x, y, c) = cast<float>(x + y + c + (int)i);
array_outputs8[i](x, y, c) = cast<float>(x + y + c + (int)i);
array_outputs9[i](x, y, c) = cast<float>(x + y + c + (int)i);
// Verify compute_with works for Output<Func>
array_outputs2[i].compute_with(array_outputs[i], x);
}
// Verify compute_with works for Output<Buffer>
dim_only_output_buffer.compute_with(Func(typed_output_buffer), x);
// Provide some bounds estimates for a Buffer input
typed_input_buffer.set_estimate(Halide::_0, 0, 2592);
typed_input_buffer.dim(1).set_estimate(42, 1968);
// Provide some bounds estimates for a Func input
input
.set_estimate(Halide::_0, 10, 2592)
.set_estimate(Halide::_1, 20, 1968)
.set_estimate(Halide::_2, 0, 3);
// Provide some scalar estimates.
b.set_estimate(false);
i8.set_estimate(3);
f32.set_estimate(48.5f);
array2_i8.set_estimate(0, 42);
// Provide some bounds estimates for an Output<Func>.
// Note that calling bound() implicitly calls estimate() as well.
output
.set_estimate(x, 10, 2592)
.set_estimate(y, 20, 1968)
.bound(c, 0, 3);
// Provide partial bounds estimates for an Output<Buffer>
typed_output_buffer.set_estimate(x, 10, 2592);
typed_output_buffer.dim(1).set_estimate(20, 1968);
// Note that calling set_bounds()/bound() implicitly calls set_estimate()/estimate() as well.
type_only_output_buffer.dim(1).set_bounds(0, 32);
type_only_output_buffer.bound(c, 0, 3);
// Verify that we can call schedule methods on arrays-of-Buffer
for (int i = 0; i < 2; i++) {
array_outputs4[i].compute_root();
}
}
void schedule() {
// empty
}
};
} // namespace
HALIDE_REGISTER_GENERATOR(MetadataTester, metadata_tester)
|
