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

use std::{
    sync::{
        atomic::{AtomicBool, Ordering},
        Arc,
    },
    time::Instant,
};

const EXPENSE: u64 = 1_000_000_000;

#[inline(never)]
fn expensive() {
    let mut i = 0;
    for _ in 0..EXPENSE {
        i = std::hint::black_box(i + 1);
    }
}

fn main() {
    let thread_count = num_cpus::get();
    println!("thread amount of cores: {thread_count}\n");

    let mut w0 = vec![];
    let mut w1 = vec![];
    let mut w2 = vec![];
    let mut w3 = vec![];

    // thread count workers
    w0.push((thread_count, perform_tasks(thread_count, 25)));
    w1.push((thread_count, perform_tasks(thread_count, 50)));
    w2.push((thread_count, perform_tasks(thread_count, 75)));
    w3.push((thread_count, perform_tasks(thread_count, 100)));

    // thread count * 2 workers
    w0.push((thread_count * 2, perform_tasks(thread_count * 2, 25)));
    w1.push((thread_count * 2, perform_tasks(thread_count * 2, 50)));
    w2.push((thread_count * 2, perform_tasks(thread_count * 2, 75)));
    w3.push((thread_count * 2, perform_tasks(thread_count * 2, 100)));

    // 4 workers
    w0.push((4, perform_tasks(4, 25)));
    w1.push((4, perform_tasks(4, 50)));
    w2.push((4, perform_tasks(4, 75)));
    w3.push((4, perform_tasks(4, 100)));

    // 8 workers
    w0.push((8, perform_tasks(8, 25)));
    w1.push((8, perform_tasks(8, 50)));
    w2.push((8, perform_tasks(8, 75)));
    w3.push((8, perform_tasks(8, 100)));

    // 16 workers
    w0.push((16, perform_tasks(16, 25)));
    w1.push((16, perform_tasks(16, 50)));
    w2.push((16, perform_tasks(16, 75)));
    w3.push((16, perform_tasks(16, 100)));

    // 32 workers
    w0.push((32, perform_tasks(32, 25)));
    w1.push((32, perform_tasks(32, 50)));
    w2.push((32, perform_tasks(32, 75)));
    w3.push((32, perform_tasks(32, 100)));

    // 128 workers
    w0.push((128, perform_tasks(128, 25)));
    w1.push((128, perform_tasks(128, 50)));
    w2.push((128, perform_tasks(128, 75)));
    w3.push((128, perform_tasks(128, 100)));

    // 256 workers
    w0.push((256, perform_tasks(256, 25)));
    w1.push((256, perform_tasks(256, 50)));
    w2.push((256, perform_tasks(256, 75)));
    w3.push((256, perform_tasks(256, 100)));

    // 1024 workers
    w0.push((1024, perform_tasks(1024, 25)));
    w1.push((1024, perform_tasks(1024, 50)));
    w2.push((1024, perform_tasks(1024, 75)));
    w3.push((1024, perform_tasks(1024, 100)));

    // best total, avg for 25 work
    let tot = w0
        .iter()
        .min_by(|&&e, &&o| e.1 .0.total_cmp(&o.1 .0))
        .unwrap();
    let avg = w0
        .iter()
        .min_by(|&&e, &&o| e.1 .1.total_cmp(&o.1 .1))
        .unwrap();
    println!(
        "work: 25, best total: [{}] {:.02}, best avg: [{}] {:.05}",
        tot.0, tot.1 .0, avg.0, avg.1 .1
    );

    // best total, avg for 50 work
    let tot = w1
        .iter()
        .min_by(|&&e, &&o| e.1 .0.total_cmp(&o.1 .0))
        .unwrap();
    let avg = w1
        .iter()
        .min_by(|&&e, &&o| e.1 .1.total_cmp(&o.1 .1))
        .unwrap();
    println!(
        "work: 50, best total: [{}] {:.02}, best avg: [{}] {:.05}",
        tot.0, tot.1 .0, avg.0, avg.1 .1
    );

    // best total, avg for 75 work
    let tot = w2
        .iter()
        .min_by(|&&e, &&o| e.1 .0.total_cmp(&o.1 .0))
        .unwrap();
    let avg = w2
        .iter()
        .min_by(|&&e, &&o| e.1 .1.total_cmp(&o.1 .1))
        .unwrap();
    println!(
        "work: 75, best total: [{}] {:.02}, best avg: [{}] {:.05}",
        tot.0, tot.1 .0, avg.0, avg.1 .1
    );

    // best total, avg for 100 work
    let tot = w3
        .iter()
        .min_by(|&&e, &&o| e.1 .0.total_cmp(&o.1 .0))
        .unwrap();
    let avg = w3
        .iter()
        .min_by(|&&e, &&o| e.1 .1.total_cmp(&o.1 .1))
        .unwrap();
    println!(
        "work: 100, best total: [{}] {:.02}, best avg: [{}] {:.05}",
        tot.0, tot.1 .0, avg.0, avg.1 .1
    );
}

fn perform_tasks(workers: usize, work: usize) -> (f64, f64) {
    println!("workers: {workers}, work: {work}");

    let ctx = tasc::StdContext::new_blocking(workers);
    let start = Arc::new(AtomicBool::new(false));

    let mut handlers = Vec::with_capacity(workers);

    let work_per_worker = work / workers;
    let mut remainder = work % workers;
    for _ in 0..workers {
        let worker_load = work_per_worker + if remainder > 0 { 1 } else { 0 };
        let start = start.clone();
        handlers.push(
            tasc::TaskBuilder::<_, tasc::global::Signal>::from_ctx(&ctx).spawn_blocking(
                move |_| {
                    while !start.load(Ordering::Acquire) {
                        // we want it to busy wait
                        std::hint::black_box(())
                    }
                    let start = Instant::now();
                    for _ in 0..worker_load {
                        #[allow(clippy::unit_arg)]
                        std::hint::black_box(expensive());
                    }
                    Instant::now() - start
                },
            ),
        );
        remainder = remainder.saturating_sub(1);
    }
    let total_time = Instant::now();
    start.store(true, Ordering::Release);
    let all_workers_time = handlers
        .into_iter()
        .map(|h| h.wait().unwrap().as_secs_f64())
        .sum::<f64>();

    let total_time = (Instant::now() - total_time).as_secs_f64();
    let avg_time = all_workers_time / workers as f64;

    println!("total time: {total_time:.02}, avg time: {avg_time:.05}\n");
    (total_time, avg_time)
}