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
|
#include "Halide.h"
#include "halide_benchmark.h"
using namespace Halide;
using namespace Halide::Tools;
double run_test(bool auto_schedule) {
int W = 1920;
int H = 1024;
Buffer<uint8_t> in(W, H, 3);
for (int y = 0; y < in.height(); y++) {
for (int x = 0; x < in.width(); x++) {
for (int c = 0; c < 3; c++) {
in(x, y, c) = rand() & 0xfff;
}
}
}
Var x("x"), y("y"), c("c");
Func Y("Y");
Y(x, y) = 0.299f * in(x, y, 0) + 0.587f * in(x, y, 1) + 0.114f * in(x, y, 2);
Func Cr("Cr");
Expr R = in(x, y, 0);
Cr(x, y) = (R - Y(x, y)) * 0.713f + 128;
Func Cb("Cb");
Expr B = in(x, y, 2);
Cb(x, y) = (B - Y(x, y)) * 0.564f + 128;
Func hist_rows("hist_rows");
hist_rows(x, y) = 0;
RDom rx(0, in.width());
Expr bin = cast<uint8_t>(clamp(Y(rx, y), 0, 255));
hist_rows(bin, y) += 1;
Func hist("hist");
hist(x) = 0;
RDom ry(0, in.height());
hist(x) += hist_rows(x, ry);
Func cdf("cdf");
cdf(x) = hist(0);
RDom b(1, 255);
cdf(b.x) = cdf(b.x - 1) + hist(b.x);
Func eq("equalize");
Expr cdf_bin = cast<uint8_t>(clamp(Y(x, y), 0, 255));
eq(x, y) = clamp(cdf(cdf_bin) * (255.0f / (in.height() * in.width())), 0, 255);
Func color("color");
Expr red = cast<uint8_t>(clamp(eq(x, y) + (Cr(x, y) - 128) * 1.4f, 0, 255));
Expr green = cast<uint8_t>(clamp(eq(x, y) - 0.343f * (Cb(x, y) - 128) - 0.711f * (Cr(x, y) - 128), 0, 255));
Expr blue = cast<uint8_t>(clamp(eq(x, y) + 1.765f * (Cb(x, y) - 128), 0, 255));
color(x, y, c) = mux(c, {red, green, blue});
Target target = get_jit_target_from_environment();
Pipeline p(color);
if (auto_schedule) {
// Provide estimates on the pipeline output
color.set_estimates({{0, 1920}, {0, 1024}, {0, 3}});
// Auto-schedule the pipeline
p.auto_schedule(target);
} else if (target.has_gpu_feature()) {
Var xi("xi"), yi("yi");
Y.compute_root().gpu_tile(x, y, xi, yi, 16, 16);
hist_rows.compute_root().gpu_tile(y, yi, 16).update().gpu_tile(y, yi, 16);
hist.compute_root().gpu_tile(x, xi, 16).update().gpu_tile(x, xi, 16);
cdf.compute_root().gpu_single_thread();
Cr.compute_at(color, xi);
Cb.compute_at(color, xi);
eq.compute_at(color, xi);
color.compute_root()
.reorder(c, x, y)
.bound(c, 0, 3)
.unroll(c)
.gpu_tile(x, y, xi, yi, 16, 16);
} else {
Y.compute_root().parallel(y, 8).vectorize(x, 8);
hist_rows.compute_root()
.vectorize(x, 8)
.parallel(y, 8)
.update()
.parallel(y, 8);
hist.compute_root()
.vectorize(x, 8)
.update()
.reorder(x, ry)
.vectorize(x, 8)
.unroll(x, 4)
.parallel(x)
.reorder(ry, x);
cdf.compute_root();
eq.compute_at(color, x).unroll(x);
Cb.compute_at(color, x).vectorize(x);
Cr.compute_at(color, x).vectorize(x);
color.reorder(c, x, y)
.bound(c, 0, 3)
.unroll(c)
.parallel(y, 8)
.vectorize(x, 8);
}
p.compile_to_lowered_stmt("histogram.html", {in}, HTML, target);
color.print_loop_nest();
Buffer<uint8_t> out(in.width(), in.height(), in.channels());
double t = benchmark(3, 10, [&]() {
p.realize(out);
});
return t * 1000;
}
int main(int argc, char **argv) {
if (get_jit_target_from_environment().arch == Target::WebAssembly) {
printf("[SKIP] Autoschedulers do not support WebAssembly.\n");
return 0;
}
if (argc != 2) {
fprintf(stderr, "Usage: %s <autoscheduler-lib>\n", argv[0]);
return 1;
}
load_plugin(argv[1]);
double manual_time = run_test(false);
double auto_time = run_test(true);
std::cout << "======================" << std::endl;
std::cout << "Manual time: " << manual_time << "ms" << std::endl;
std::cout << "Auto time: " << auto_time << "ms" << std::endl;
std::cout << "======================" << std::endl;
if (auto_time > manual_time * 3) {
fprintf(stderr, "Warning: Auto-scheduler is much much slower than it should be.\n");
}
printf("Success!\n");
return 0;
}
|
