shuttle 0.9.1

A library for testing concurrent Rust code
Documentation 
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

use criterion::{criterion_group, criterion_main, BenchmarkId, Criterion, SamplingMode};
use shuttle::scheduler::{PctScheduler, RandomScheduler, Scheduler};
use shuttle::sync::Mutex;
use shuttle::{thread, Runner};
use std::sync::Arc;
use std::time::Duration;

const TOTAL_EVENTS: u32 = 10000;
const NARROW_TASKS: u32 = 5;
const WIDE_TASKS: u32 = 100;
const ITERATIONS: usize = 1;
const WIDE_EVENTS_PER_TASK: u32 = TOTAL_EVENTS / WIDE_TASKS;
const NARROW_EVENTS_PER_TASK: u32 = TOTAL_EVENTS / NARROW_TASKS;

/// A benchmark that runs threads incrementing a lock. This is a stress test of our `Execution` logic,
/// since the threads spend basically all their time taking and dropping the lock.
fn lock_benchmark(scheduler: impl Scheduler + 'static, num_tasks: u32, num_events_per_task: u32) {
    let runner = Runner::new(scheduler, Default::default());
    runner.run(move || {
        let lock = Arc::new(Mutex::new(0usize));

        let thds: Vec<_> = (0..num_tasks)
            .map(|_| {
                let lock = Arc::clone(&lock);
                thread::spawn(move || {
                    for _ in 0..num_events_per_task {
                        *lock.lock().unwrap() += 1;
                    }
                })
            })
            .collect();

        for thd in thds {
            thd.join().unwrap();
        }
    });
}

pub fn lock_sync_benchmark(c: &mut Criterion) {
    let mut g = c.benchmark_group("lock sync");
    g.warm_up_time(Duration::from_secs(1));

    g.bench_function("pct-narrow", |b| {
        b.iter(|| {
            let scheduler = PctScheduler::new_from_seed(0x12345678, 2, ITERATIONS);
            lock_benchmark(scheduler, NARROW_TASKS, NARROW_EVENTS_PER_TASK);
        });
    });

    g.bench_function("random-narrow", |b| {
        b.iter(|| {
            let scheduler = RandomScheduler::new_from_seed(0x12345678, ITERATIONS);
            lock_benchmark(scheduler, NARROW_TASKS, NARROW_EVENTS_PER_TASK);
        });
    });

    g.bench_function("pct-wide", |b| {
        b.iter(|| {
            let scheduler = PctScheduler::new_from_seed(0x12345678, 2, ITERATIONS);
            lock_benchmark(scheduler, WIDE_TASKS, WIDE_EVENTS_PER_TASK);
        });
    });

    g.bench_function("random-wide", |b| {
        b.iter(|| {
            let scheduler = RandomScheduler::new_from_seed(0x12345678, ITERATIONS);
            lock_benchmark(scheduler, WIDE_TASKS, WIDE_EVENTS_PER_TASK);
        });
    });
}

const SCALING_TOTAL_EVENTS: [u32; 3] = [1000, 10000, 100000];
const SCALING_TASKS: [u32; 6] = [4, 16, 32, 64, 128, 1024];

pub fn lock_scaling_sync_benchmark(c: &mut Criterion) {
    let mut g = c.benchmark_group("lock scaling sync");
    g.sample_size(20)
        .warm_up_time(Duration::from_millis(100))
        .sampling_mode(SamplingMode::Flat);

    for num_tasks in SCALING_TASKS {
        for num_total_events in SCALING_TOTAL_EVENTS {
            // skip configs without meaningful number of events per thread
            if num_tasks * 10 >= num_total_events {
                continue;
            }
            let num_events_per_task = num_total_events / num_tasks;
            let input = (num_tasks, num_events_per_task);
            let parameter_string = format!("{{tasks:{num_tasks},events:{num_total_events}}}");
            g.bench_with_input(
                BenchmarkId::new("RW", parameter_string),
                &input,
                |b, (num_tasks, num_events_per_task)| {
                    b.iter(|| {
                        let scheduler = RandomScheduler::new_from_seed(0x12345678, ITERATIONS);
                        lock_benchmark(scheduler, *num_tasks, *num_events_per_task);
                    })
                },
            );
        }
    }
}

criterion_group!(benches, lock_sync_benchmark, lock_scaling_sync_benchmark);
criterion_main!(benches);