Re: Realtime safe signalling?

[Igor Gorbounov <igorbounov topazelectro ru>]

Jan Hudec wrote:

> On Tue, Feb 01, 2005 at 17:30:51 -0500, Paul Davis wrote:
>  
>
> > > > no. none of these mechanisms are safe.
> > > > =20
> > > > you need to to use some kernel facility (FIFOs are my favorite) to do
> > > >        
> > > And what do you think Glib::Dispatcher::emit() is using?? Right, FIFOs.
> > > Be sure not to share one dispatcher between several threads though as it
> > > does some locking.
> > >      
> > no locking is acceptable at all for RT usage. My implementation of the
> > same idea stuffs Slots into a lock-free FIFO, and then writes a single
> > bytes to the FIFO to wake up the GUI. If you lock, its not RT-safe.
> >    
>
> Yes. Sure. Glib::Dispatcher is no good for RT.
>  
Do you mean this quote from 
http://www.gtkmm.org/docs/glibmm-2.4/docs/reference/html/classGlib_1_1Dispatcher.html:
"After instantiation, Glib::Dispatcher will never lock any mutexes on 
its own. The interaction with the GLib main loop might involve locking 
on the /receiver/ side. The /sender/ side, however, is guaranteed not to 
lock, except for internal locking in the |write()| system call." ?

But can FIFO, created wth the use of mkfifo, be of a proper use for RT 
signals? There is a special file created in
a filesystem - is it good?
   Igor Gorbounov