CPU Cost of having Posix API enabled if not used.

Tom Bojsen Tom.Bojsen at man.eu
Mon Mar 4 09:41:34 UTC 2013


Thank you for your answer, Sebastian.
I think it make sense that we disable posix to improve task switch performance on our current version.
We will eventually update our RTEMS though.

/Tom

> -----Original Message-----
> From: rtems-users-bounces at rtems.org [mailto:rtems-users-
> bounces at rtems.org] On Behalf Of Sebastian Huber
> Sent: 28. februar 2013 17:38
> To: rtems-users at rtems.org
> Subject: Re: CPU Cost of having Posix API enabled if not used.
> 
> Hello,
> 
> On 02/28/2013 05:26 PM, Tom Bojsen wrote:
> > Hi,
> >
> > Does enabling the Posix interface cost extra CPU cycles if it is not
> used?
> 
> in your version, yes.  In the latest Git master branch, no.  The RTEMS
> and
> POSIX signals need a post thread switch extension.  See end of
> _Thread_Dispatch() (_API_extensions_Run_post_switch( executing )):
> 
> http://git.rtems.org/rtems/tree/cpukit/score/src/threaddispatch.c
> 
> In earlier RTEMS versions the post switch extension was installed
> during system
> start.  Now it is installed on demand.  On systems which frequently
> switch
> threads the post switch overhead is measurable.
> 
> http://git.rtems.org/rtems/commit/?id=ba3163131770e730cbf5168333c84daa5
> 7bc14e5
> http://git.rtems.org/rtems/commit/?id=9465fca287e9417e05d2efe572d770fd6
> 0ad61db
> 
> An unused POSIX API adds also some dead code.
> 
> --
> Sebastian Huber, embedded brains GmbH
> 
> Address : Dornierstr. 4, D-82178 Puchheim, Germany
> Phone   : +49 89 189 47 41-16
> Fax     : +49 89 189 47 41-09
> E-Mail  : sebastian.huber at embedded-brains.de
> PGP     : Public key available on request.
> 
> Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.
> _______________________________________________
> rtems-users mailing list
> rtems-users at rtems.org
> http://www.rtems.org/mailman/listinfo/rtems-users




More information about the users mailing list