Use a distinct ToString implementation for u128 and i128

[GuillaumeGomez]

Part of #135543.

Follow-up of #136264.

When working on #142098, I realized that i128 and u128 could also benefit from a distinct ToString implementation so here it.

The last commit is just me realizing that I forgot to add the format tests for usize and isize.

Here is the bench comparison:

bench name	last nightly	with this PR	diff
bench_i128	29.25 ns/iter (+/- 0.66)	17.52 ns/iter (+/- 0.7)	-40.1%
bench_u128	34.06 ns/iter (+/- 0.21)	16.1 ns/iter (+/- 0.6)	-52.7%

I used this code to test:

#![feature(test)]

extern crate test;

use test::{Bencher, black_box};

#[inline(always)]
fn convert_to_string<T: ToString>(n: T) -> String {
    n.to_string()
}

macro_rules! decl_benches {
    ($($name:ident: $ty:ident,)+) => {
        $(
	    #[bench]
            fn $name(c: &mut Bencher) {
                c.iter(|| convert_to_string(black_box({ let nb: $ty = 20; nb })));
            }
	)+
    }
}

decl_benches! {
    bench_u128: u128,
    bench_i128: i128,
}

[rustbot]

r? @workingjubilee

rustbot has assigned @workingjubilee.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[bors]

☔ The latest upstream changes (presumably #136594) made this pull request unmergeable. Please resolve the merge conflicts.

[GuillaumeGomez]

Fixed merge conflicts.

[workingjubilee]

Having done code review of things that use Rust's feature(min_specialization), I now consider "specialization" a dirty word.

[GuillaumeGomez]

Noted, thanks for the title update then. :)

[tgross35]

If you've tested locally, is there any measurable perf/size impact here?

[GuillaumeGomez]

@tgross35 I added the comparison in the first comment with the code I use to get the performance comparison.

[tgross35]

Thanks!

I don't think this needs a pre-merge perf run since we didn't see any perf-visible impact in #136594, but it wouldn't hurt to get one after so

@bors r+ rollup=never

Fyi @pascaldekloe since you did the last round of perf boosts for these methods

[bors]

📌 Commit 9b09948 has been approved by tgross35

It is now in the queue for this repository.

[bors]

⌛ Testing commit 9b09948 with merge 5b74275...

[bors]

☀️ Test successful - checks-actions
Approved by: tgross35
Pushing 5b74275 to master...

[github-actions]

What is this?

This is an experimental post-merge analysis report that shows differences in test outcomes between the merged PR and its parent PR.

Comparing 255aa22 (parent) -> 5b74275 (this PR)

Test differences

Show 26 test diffs

Stage 1

• num::test_isize_to_string: [missing] -> pass (J0)
• num::test_usize_to_string: [missing] -> pass (J0)

Additionally, 24 doctest diffs were found. These are ignored, as they are noisy.

Job group index

• J0: aarch64-apple, aarch64-gnu, arm-android, armhf-gnu, dist-i586-gnu-i586-i686-musl, i686-gnu-1, i686-gnu-nopt-1, i686-msvc-1, mingw-check-2, test-various, x86_64-apple-1, x86_64-gnu, x86_64-gnu-aux, x86_64-gnu-llvm-19-2, x86_64-gnu-llvm-19-3, x86_64-gnu-llvm-20-2, x86_64-gnu-llvm-20-3, x86_64-gnu-nopt, x86_64-gnu-stable, x86_64-mingw-1, x86_64-msvc-1

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard 5b74275f89b6041bf2e9dc2abcf332e206d4cfca --output-dir test-dashboard

And then open test-dashboard/index.html in your browser to see an overview of all executed tests.

Job duration changes

1. dist-aarch64-apple: 4689.2s -> 6379.9s (36.1%)
2. x86_64-apple-2: 3543.9s -> 4546.3s (28.3%)
3. dist-x86_64-apple: 9048.3s -> 7868.7s (-13.0%)
4. mingw-check-tidy: 77.1s -> 69.0s (-10.5%)
5. x86_64-apple-1: 6181.5s -> 6732.2s (8.9%)
6. x86_64-rust-for-linux: 2603.2s -> 2819.3s (8.3%)
7. i686-gnu-1: 7677.0s -> 7174.4s (-6.5%)
8. x86_64-gnu-llvm-20-3: 6777.2s -> 6353.4s (-6.3%)
9. dist-aarch64-linux: 8178.3s -> 7691.0s (-6.0%)
10. dist-x86_64-netbsd: 4763.9s -> 5035.2s (5.7%)

How to interpret the job duration changes?

Job durations can vary a lot, based on the actual runner instance
that executed the job, system noise, invalidated caches, etc. The table above is provided
mostly for t-infra members, for simpler debugging of potential CI slow-downs.

[rust-timer]

Finished benchmarking commit (5b74275): comparison URL.

Overall result: ✅ improvements - no action needed

@rustbot label: -perf-regression

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.2%	[-0.2%, -0.2%]	1
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary -1.9%, secondary -2.4%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	2.8%	[2.8%, 2.8%]	1
Improvements ✅ (primary)	-1.9%	[-2.4%, -1.3%]	5
Improvements ✅ (secondary)	-2.6%	[-6.8%, -1.0%]	23
All ❌✅ (primary)	-1.9%	[-2.4%, -1.3%]	5

Cycles

Results (secondary 4.3%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	5.2%	[2.0%, 10.5%]	9
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-3.9%	[-3.9%, -3.9%]	1
All ❌✅ (primary)	-	-	0

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 692.261s -> 692.705s (0.06%)
Artifact size: 371.96 MiB -> 372.00 MiB (0.01%)