#parallel #performance

rustc-rayon

Simple work-stealing parallelism for Rust - fork for rustc

9 unstable releases

0.5.0 Mar 24, 2023
0.4.0 May 27, 2022
0.3.2 Jan 10, 2022
0.3.1 Mar 11, 2021
0.1.1 Jun 6, 2018

#123 in Concurrency

Download history 10316/week @ 2024-06-18 10925/week @ 2024-06-25 9718/week @ 2024-07-02 11238/week @ 2024-07-09 10518/week @ 2024-07-16 10350/week @ 2024-07-23 11793/week @ 2024-07-30 9917/week @ 2024-08-06 10327/week @ 2024-08-13 10567/week @ 2024-08-20 9792/week @ 2024-08-27 10152/week @ 2024-09-03 12014/week @ 2024-09-10 9514/week @ 2024-09-17 9312/week @ 2024-09-24 6695/week @ 2024-10-01

39,339 downloads per month
Used in 61 crates (7 directly)

MIT/Apache

1MB
22K SLoC

rustc-rayon

rustc-rayon is a fork of the Rayon crate. It adds a few "in progress" features that rustc is using, mostly around deadlock detection. These features are not stable and should not be used by others -- though they may find their way into rayon proper at some point. In general, if you are not rustc, you should be using the real rayon crate, not rustc-rayon. =)

License

rustc-rayon is a fork of rayon. rayon is distributed under the terms of both the MIT license and the Apache License (Version 2.0). See LICENSE-APACHE and LICENSE-MIT for details. Opening a pull request is assumed to signal agreement with these licensing terms.


lib.rs:

Data-parallelism library that makes it easy to convert sequential computations into parallel

Rayon is lightweight and convenient for introducing parallelism into existing code. It guarantees data-race free executions and takes advantage of parallelism when sensible, based on work-load at runtime.

How to use Rayon

There are two ways to use Rayon:

  • High-level parallel constructs are the simplest way to use Rayon and also typically the most efficient.
    • Parallel iterators make it easy to convert a sequential iterator to execute in parallel.
    • The par_sort method sorts &mut [T] slices (or vectors) in parallel.
    • par_extend can be used to efficiently grow collections with items produced by a parallel iterator.
  • Custom tasks let you divide your work into parallel tasks yourself.
    • join is used to subdivide a task into two pieces.
    • scope creates a scope within which you can create any number of parallel tasks.
    • ThreadPoolBuilder can be used to create your own thread pools or customize the global one.

Basic usage and the Rayon prelude

First, you will need to add rayon to your Cargo.toml.

Next, to use parallel iterators or the other high-level methods, you need to import several traits. Those traits are bundled into the module rayon::prelude. It is recommended that you import all of these traits at once by adding use rayon::prelude::* at the top of each module that uses Rayon methods.

These traits give you access to the par_iter method which provides parallel implementations of many iterative functions such as map, for_each, filter, fold, and more.

Crate Layout

Rayon extends many of the types found in the standard library with parallel iterator implementations. The modules in the rayon crate mirror std itself: so, e.g., the option module in Rayon contains parallel iterators for the Option type, which is found in the option module of std. Similarly, the collections module in Rayon offers parallel iterator types for the collections from std. You will rarely need to access these submodules unless you need to name iterator types explicitly.

Targets without threading

Rayon has limited support for targets without std threading implementations. See the rayon_core documentation for more information about its global fallback.

Other questions?

See the Rayon FAQ.

Dependencies

~685KB
~11K SLoC