El Dijous, 6 de febrer de 2014, a les 22:49:32, Gilles Chanteperdrix va 
escriure:
[...]

> > 
> >>> (CPU time limit exceeded). It's not easy for us with that.
> >> 
> >> SIGXCPU is in fact Xenomai's SIGDEBUG. See
> >> examples/native/sigdebug.c to find out why you receive SIGDEBUG.
> >> My bet would be that you are missing mlockall. That said, Xenomai
> >> should issue the mlockall automatically, you simply have to take
> >> care of putting xenomai libraries at the end of the link-edit
> >> command line, so that xenomai libraries are loaded first.
> > 
> > Ok,
> > 
> > 
> > now it works, half work. So, it could be equivalent if I call the
> > mlockall function in the entry point of the lib? I mean, if I
> > create a init function of the lib and when I load it, mlockall is
> > called.
> 
> If the library uses xenomai posix skin services (through wrapping),

no, AFAIR we don't use the posix skin. Just for the examples.

> mlockall is not sufficient, the xenomai posix skin library has an
> initialization function which must be called first. Getting a library
> written for Linux implementation to work with xenomai requires several
> changes (which the document posted in answer to your first question
> tries to document). So, if you do not need this library to run in
> primary mode, I suggest you stick to your original plan which was to
> use RTDM without the posix skin wrapping.

Ok, I will look the document. I would prefer to have a clean implementation of 
pure xenomai native.

> >>> So, can we use the xddp protocol from native? Which functions
> >>> we should
> > 
> > use to
> > 
> >>> send/receive, etc?
> >> 
> >> See:
> > http://www.xenomai.org/documentation/xenomai-2.6/html/api/group__userapi.h
> > tml> 
> >  good, done. Perfect. However we are having problems because _only_
> > 
> > one label is created. I suppose that there's no limit of the number
> > of sockets created in the RT thread, no?
> 
> Please try and reduce the problematic program to the minimal
> self-contained test case which we can run to reproduce the issue you have.

Yes, I'm almost sure that we are doing something wrong. However, I think that 
to create a modification example from the xddp-echo or label could be a good 
idea. Tomorrow more,

Thanks for all,

Leopold

-- 
--
Linux User 152692     PGP: 0xF944807E
Catalonia
-------------------------------------
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?


_______________________________________________
Xenomai mailing list
Xenomai@xenomai.org
http://www.xenomai.org/mailman/listinfo/xenomai

Reply via email to