xxx_poll callbacks and multiple event loops

Marc Lehmann schmorp at schmorp.de
Tue Nov 8 19:45:10 CET 2011


On Tue, Nov 08, 2011 at 06:39:46PM +0000, Paddy Byers <paddy.byers at gmail.com> wrote:
> Thanks for your considered response. I will have a look at your proposal. I
> wanted to avoid the extra thread context-switch but I guess the next step
> is to measure what that impact is, and have a think about the best way
> forward.

You should definitely consider using the MMU also - using separate
processes allows you to use (usually) faster global variables without
having to use pointer indirection (and/or slow tls variables) for every
access, and the MMU in your cpu will do the job at no extra cost: the MMU
overhead is there in any case, but for multi-core systems, you can take
advantage of having more registers available, which is very important on
x86, but also on less register-starved machines such as amd64.

-- 
                The choice of a       Deliantra, the free code+content MORPG
      -----==-     _GNU_              http://www.deliantra.net
      ----==-- _       generation
      ---==---(_)__  __ ____  __      Marc Lehmann
      --==---/ / _ \/ // /\ \/ /      schmorp at schmorp.de
      -=====/_/_//_/\_,_/ /_/\_\



More information about the libev mailing list