AnyEvent::Impl::Perl - Pure-Perl event loop and AnyEvent adaptor for itself
use AnyEvent; # use AnyEvent::Impl::Perl; # this module gets loaded automatically as required # Explicit use: use AnyEvent::Impl::Perl; use AnyEvent; ... AnyEvent::Impl::Perl::loop; # run the event loop
This module provides transparent support for AnyEvent in case no other event loop could be found or loaded. You don't have to do anything to make it work with AnyEvent except by possibly loading it before creating the first AnyEvent watcher.
If you want to use this module instead of autoloading another event loop you can simply load it before creating the first watcher.
As for performance, this module is on par with (and usually faster than) most select/poll-based C event modules such as Event or Glib (it does not even come close to EV, though), with respect to I/O watchers. Timers are handled less optimally, but for many common tasks, it's still on par with event loops written in C.
This event loop has been optimised for the following use cases:
- monotonic clock is available
This module will use the POSIX monotonic clock option (if it can be detected at runtime) or the POSIX
timesfunction (if the resolution is at least 100Hz), in which case it will not suffer adversely from time jumps.
If no monotonic clock is available, this module will not attempt to correct for time jumps in any way.
The clock chosen will be reported if the environment variable
$PERL_ANYEVENT_VERBOSEis set to 8 or higher.
- any number of watchers on one fd
Supporting a large number of watchers per fd is purely a dirty benchmark optimisation not relevant in practise. The more common case of having one watcher per fd/poll combo is special-cased, however, and therefore fast, too.
- relatively few active fds per
This module expects that only a tiny amount of fds is active at any one time. This is relatively typical of larger servers (but not the case where
selecttraditionally is fast), at the expense of the "dense activity case" where most of the fds are active (which suits
The optimal implementation of the "dense" case is not much faster, though, so the module should behave very well in most cases, subject to the bad scalability of
- lots of timer changes/iteration, or none at all
This module sorts the timer list using perl's
sort, even though a total ordering is not required for timers.
This sorting is expensive, but means sorting can be avoided unless the timer list has changed in a way that requires a new sort.
This means that adding lots of timers is very efficient, as well as not changing the timers. Advancing timers (e.g. recreating a timeout watcher on activity) is also relatively efficient, for example, if you have a large number of timeout watchers that time out after 10 seconds, then the timer list will be sorted only once every 10 seconds.
This should not have much of an impact unless you have hundreds or thousands of timers, though, or your timers have very small timeouts.
The only user-visible function provided by this module is the
Run the event loop, usually the last thing done in the main program when you want to use the pure-perl backend.
Marc Lehmann <firstname.lastname@example.org> http://home.schmorp.de/