Struct pond::Pool

pub struct Pool { /* private fields */ }

Holds a number of threads that you can run tasks on

Implementations

impl Pool

pub fn new() -> Pool

Spawn a pool with the native number of threads and a fixed backlog

The native number of threads is the number of logical cpus according to the crate num_cpus

The backlog, 4 times the number of threads, is the number of unprocessed jobs that are allowed to accumulate from Scope::execute

pub fn unbounded() -> Pool

Spawn a pool with the native number of threads an un unbounded backlog

The native number of threads is the number of logical cpus according to the crate num_cpus

Scope::execute will never block and will accumulate any task you give it. This may require a lot of memory.

pub fn new_threads(nthreads: usize, backlog: usize) -> Pool

Spawn a number of threads. The pool’s queue of pending jobs is limited.

backlog is the number of jobs that haven’t been run. If specified, Scope::execute will block until a job completes.

pub fn new_threads_unbounded(nthreads: usize) -> Pool

Spawn a number of threads. The pool’s queue of pending jobs is limited.

the backlog is unbounded as in unbounded.

pub fn scoped<'pool, 'scope, F>(&'pool mut self, f: F)where
    F: FnOnce(Scope<'pool, 'scope>),

Store the current scope so that you can run jobs in it.

This function panics if the given closure or any threads panic.

Does not return until all executed jobs complete.

Trait Implementations

impl Drop for Pool

fn drop(&mut self)

terminates all threads