bjoernager/rust
bjoernager/rust
1
Fork 0
Code Activity
Empowering everyone to build reliable and efficient software. Gabriel's commits. https://www.rust-lang.org/
285734 commits 18 branches 0 tags 536 MiB
Find a file
Stuart Cook 2e4e196a5b
Rollup merge of #136457 - calder:master, r=tgross35 
Expose algebraic floating point intrinsics

# Problem

A stable Rust implementation of a simple dot product is 8x slower than C++ on modern x86-64 CPUs. The root cause is an inability to let the compiler reorder floating point operations for better vectorization.

See https://github.com/calder/dot-bench for benchmarks. Measurements below were performed on a i7-10875H.

### C++: 10us ✅

With Clang 18.1.3 and `-O2 -march=haswell`:
<table>
<tr>
    <th>C++</th>
    <th>Assembly</th>
</tr>
<tr>
<td>
<pre lang="cc">
float dot(float *a, float *b, size_t len) {
    #pragma clang fp reassociate(on)
    float sum = 0.0;
    for (size_t i = 0; i < len; ++i) {
        sum += a[i] * b[i];
    }
    return sum;
}
</pre>
</td>
<td>
<img src="https://github.com/user-attachments/assets/739573c0-380a-4d84-9fd9-141343ce7e68" />
</td>
</tr>
</table>

### Nightly Rust: 10us ✅

With rustc 1.86.0-nightly (8239a37f9) and `-C opt-level=3 -C target-feature=+avx2,+fma`:
<table>
<tr>
    <th>Rust</th>
    <th>Assembly</th>
</tr>
<tr>
<td>
<pre lang="rust">
fn dot(a: &[f32], b: &[f32]) -> f32 {
    let mut sum = 0.0;
    for i in 0..a.len() {
        sum = fadd_algebraic(sum, fmul_algebraic(a[i], b[i]));
    }
    sum
}
</pre>
</td>
<td>
<img src="https://github.com/user-attachments/assets/9dcf953a-2cd7-42f3-bc34-7117de4c5fb9" />
</td>
</tr>
</table>

### Stable Rust: 84us ❌

With rustc 1.84.1 (e71f9a9a9) and `-C opt-level=3 -C target-feature=+avx2,+fma`:
<table>
<tr>
    <th>Rust</th>
    <th>Assembly</th>
</tr>
<tr>
<td>
<pre lang="rust">
fn dot(a: &[f32], b: &[f32]) -> f32 {
    let mut sum = 0.0;
    for i in 0..a.len() {
        sum += a[i] * b[i];
    }
    sum
}
</pre>
</td>
<td>
<img src="https://github.com/user-attachments/assets/936a1f7e-33e4-4ff8-a732-c3cdfe068dca" />
</td>
</tr>
</table>

# Proposed Change

Add `core::intrinsics::f*_algebraic` wrappers to `f16`, `f32`, `f64`, and `f128` gated on a new `float_algebraic` feature.

# Alternatives Considered

https://github.com/rust-lang/rust/issues/21690 has a lot of good discussion of various options for supporting fast math in Rust, but is still open a decade later because any choice that opts in more than individual operations is ultimately contrary to Rust's design principles.

In the mean time, processors have evolved and we're leaving major performance on the table by not supporting vectorization. We shouldn't make users choose between an unstable compiler and an 8x performance hit.

# References

* https://github.com/rust-lang/rust/issues/21690
* https://github.com/rust-lang/libs-team/issues/532
* https://github.com/rust-lang/rust/issues/136469
* https://github.com/calder/dot-bench
* https://www.felixcloutier.com/x86/vfmadd132ps:vfmadd213ps:vfmadd231ps

try-job: x86_64-gnu-nopt
try-job: x86_64-gnu-aux
2025-04-05 13:18:12 +11:00
.github Add CI metadata to bootstrap metrics 2025-03-27 17:03:08 +01:00
compiler Auto merge of #139390 - matthiaskrgr:rollup-l64euwx, r=matthiaskrgr 2025-04-04 23:03:57 +00:00
library Rollup merge of #136457 - calder:master, r=tgross35 2025-04-05 13:18:12 +11:00
LICENSES
src Rollup merge of #136457 - calder:master, r=tgross35 2025-04-05 13:18:12 +11:00
tests Rollup merge of #136457 - calder:master, r=tgross35 2025-04-05 13:18:12 +11:00
.clang-format
.editorconfig Don't apply editorconfig to llvm 2025-02-09 16:21:14 -05:00
.git-blame-ignore-revs Git blame ignore recent formatting commit 2025-02-09 12:00:23 -08:00
.gitattributes Mark .pp files as Rust 2025-03-29 12:39:06 +01:00
.gitignore change config.toml to bootstrap.toml for bootstrap module 2025-03-17 12:56:41 +05:30
.gitmodules move autodiff from EnzymeAD/Enzyme to our rust-lang/Enzyme soft-fork 2025-04-01 17:17:39 -04:00
.ignore change config.toml to bootstrap.toml for bootstrap module 2025-03-17 12:56:41 +05:30
.mailmap Rollup merge of #139342 - meithecatte:mailmap, r=compiler-errors 2025-04-04 08:02:09 +02:00
bootstrap.example.toml Add new change-id option in bootstrap.example.toml and update the change-id description references 2025-03-28 08:40:24 +05:30
Cargo.lock compiler and tools dependencies 2025-04-01 20:48:17 +00:00
Cargo.toml Alphabetize the root workspace members 2025-03-13 12:21:14 -07:00
CODE_OF_CONDUCT.md
configure
CONTRIBUTING.md Fix some typos 2025-03-04 16:05:32 +08:00
COPYRIGHT
INSTALL.md change config.toml to bootstrap.toml for bootstrap module 2025-03-17 12:56:41 +05:30
LICENSE-APACHE
license-metadata.json Update license metadata 2025-02-15 16:48:37 +01:00
LICENSE-MIT
README.md Update Rust Foundation links in Readme 2025-03-16 19:03:40 -07:00
RELEASES.md Apply suggestions from code review 2025-03-30 15:45:44 -07:00
REUSE.toml change config.toml to bootstrap.toml for bootstrap module 2025-03-17 12:56:41 +05:30
rust-bors.toml
rustfmt.toml Allow formatting example/gen_block_iterate.rs 2025-04-01 14:49:15 +00:00
triagebot.toml Rollup merge of #139384 - jieyouxu:compiletest-reviewers, r=Kobzol 2025-04-04 21:55:01 +02:00
x Look for python3 first on MacOS, not py 2025-02-13 10:24:54 -05:00
x.ps1
x.py

README.md

This is the main source code repository for Rust. It contains the compiler, standard library, and documentation.

Why Rust?

  • Performance: Fast and memory-efficient, suitable for critical services, embedded devices, and easily integrated with other languages.

  • Reliability: Our rich type system and ownership model ensure memory and thread safety, reducing bugs at compile-time.

  • Productivity: Comprehensive documentation, a compiler committed to providing great diagnostics, and advanced tooling including package manager and build tool (Cargo), auto-formatter (rustfmt), linter (Clippy) and editor support (rust-analyzer).

Quick Start

Read "Installation" from The Book.

Installing from Source

If you really want to install from source (though this is not recommended), see INSTALL.md.

Getting Help

See https://www.rust-lang.org/community for a list of chat platforms and forums.

Contributing

See CONTRIBUTING.md.

License

Rust is primarily distributed under the terms of both the MIT license and the Apache License (Version 2.0), with portions covered by various BSD-like licenses.

See LICENSE-APACHE, LICENSE-MIT, and COPYRIGHT for details.

Trademark

The Rust Foundation owns and protects the Rust and Cargo trademarks and logos (the "Rust Trademarks").

If you want to use these names or brands, please read the Rust language trademark policy.

Third-party logos may be subject to third-party copyrights and trademarks. See Licenses for details.