Performance benchmark of context() vs with_context()

benchmark.rs

use anyhow::{Context, Result};
use criterion::{black_box, criterion_group, criterion_main, Criterion};
use std::error::Error;
use std::fmt;

#[derive(Debug)]
struct FooError {}

fn foo() -> Result<(), FooError> {
    Err(FooError {})
}

impl Error for FooError {}

impl fmt::Display for FooError {
    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
        write!(f, "FooError happened")
    }
}

fn run() -> Result<()> {
    foo().with_context(|| format!("Now calling: {}", "foo()"))?;
    Ok(())
}

fn criterion_benchmark(c: &mut Criterion) {
    c.bench_function("count_words", |b| b.iter(|| black_box(run())));
}

criterion_group!(benches, criterion_benchmark);
criterion_main!(benches);

diff

@@ -19,7 +19,7 @@ impl fmt::Display for FooError {
 }

 fn run() -> Result<()> {
-    foo().context(format!("Now calling: {}", "foo()"))?;
+    foo().with_context(|| format!("Now calling: {}", "foo()"))?;
     Ok(())
 }

first run output

➔  cargo bench
   Compiling rust-wordcount v0.1.0 (/home/work/Workspace/zoni/rust-wordcount)
    Finished bench [optimized] target(s) in 1.74s
     Running target/release/deps/wordcount-03e3bc23db692a6a

running 4 tests
test wordcounter::tests::count_multiple_words ... ignored
test wordcounter::tests::count_single_word ... ignored
test wordcounter::tests::empty_input ... ignored
test wordcounter::tests::read_broken_pipe ... ignored

test result: ok. 0 passed; 0 failed; 4 ignored; 0 measured; 0 filtered out

     Running target/release/deps/rwc-0b748e838683aeca

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out

     Running target/release/deps/context_vs_with_context-aed1599f929f47c2
count_words             time:   [40.922 ns 41.101 ns 41.311 ns]
Found 12 outliers among 100 measurements (12.00%)
  1 (1.00%) low mild
  8 (8.00%) high mild
  3 (3.00%) high severe

second run (after patch) output

➔  cargo bench
   Compiling rust-wordcount v0.1.0 (/home/work/Workspace/zoni/rust-wordcount)
    Finished bench [optimized] target(s) in 1.69s
     Running target/release/deps/wordcount-03e3bc23db692a6a

running 4 tests
test wordcounter::tests::count_multiple_words ... ignored
test wordcounter::tests::count_single_word ... ignored
test wordcounter::tests::empty_input ... ignored
test wordcounter::tests::read_broken_pipe ... ignored

test result: ok. 0 passed; 0 failed; 4 ignored; 0 measured; 0 filtered out

     Running target/release/deps/rwc-0b748e838683aeca

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out

     Running target/release/deps/context_vs_with_context-aed1599f929f47c2
count_words             time:   [40.949 ns 41.086 ns 41.233 ns]
                        change: [-1.9642% -0.5919% +0.4932%] (p = 0.39 > 0.05)
                        No change in performance detected.
Found 8 outliers among 100 measurements (8.00%)
  2 (2.00%) low mild
  3 (3.00%) high mild
  3 (3.00%) high severe