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
|
#include "Halide.h"
#include "halide_benchmark.h"
#include <stdio.h>
using namespace Halide;
using namespace Halide::Tools;
Expr expensive(Expr x, int c) {
if (c <= 0) {
return x;
} else {
return expensive(x * (x + 1), c - 1);
}
}
int main(int argc, char **argv) {
Target target = get_jit_target_from_environment();
if (target.arch == Target::WebAssembly) {
printf("[SKIP] Performance tests are meaningless and/or misleading under WebAssembly interpreter.\n");
return 0;
}
if (target.has_feature(Target::Vulkan)) {
printf("[SKIP] Skipping test for Vulkan. Async performance needs to be improved before this test will pass.\n");
return 0;
}
if (!target.has_gpu_feature()) {
printf("[SKIP] No GPU target enabled.\n");
return 0;
}
double times[2];
uint32_t correct = 0;
for (int use_async = 0; use_async < 2; use_async++) {
Var x, y, t, xi, yi;
ImageParam in(UInt(32), 3);
Func cpu("cpu"), gpu("gpu");
// We have a two-stage pipeline that processes frames. We want
// to run the first stage on the GPU and the second stage on
// the CPU. We'd like to get the CPU and GPU running at the
// same time using async. The amount of math we do here
// doesn't matter much - the important thing is that we
// overlap CPU computation with the GPU buffer copies.
gpu(x, y, t) = expensive(in(x, y, t), 16);
cpu(x, y, t) = expensive(gpu(x, y, t), 16);
cpu.parallel(y, 16).vectorize(x, 8);
// Assume GPU memory is limited, and compute the GPU stage one
// frame at a time. Hoist the allocation to the top level.
gpu.compute_at(gpu.in(), Var::outermost()).store_root().gpu_tile(x, y, xi, yi, 8, 8);
// Stage the copy-back of the GPU result into a host-side
// double-buffer.
gpu.in().copy_to_host().compute_at(cpu, t).store_root().fold_storage(t, 2);
if (use_async) {
// gpu.async();
gpu.in().async();
}
Buffer<uint32_t> in_buf(800, 800, 16);
in_buf.fill(17);
in.set(in_buf);
Buffer<uint32_t> out(800, 800, 16);
cpu.compile_jit();
times[use_async] = benchmark(10, 1, [&]() {
cpu.realize(out);
});
if (!use_async) {
correct = out(0, 0, 0);
} else {
for (int t = 0; t < out.dim(2).extent(); t++) {
for (int y = 0; y < out.dim(1).extent(); y++) {
for (int x = 0; x < out.dim(0).extent(); x++) {
if (out(x, y, t) != correct) {
printf("Async output at (%d, %d, %d) is %u instead of %u\n",
x, y, t, out(x, y, t), correct);
return 1;
}
}
}
}
}
printf("%s: %f\n",
use_async ? "with async" : "without async",
times[use_async]);
}
if (times[1] > 1.2 * times[0]) {
printf("Using async should have been faster\n");
return 1;
}
printf("Success!\n");
return 0;
}
|
