Re: stdint.h

["Joel Sherrill <joel at OARcorp dot com>" <joel dot sherrill at OARcorp dot com>]

Jeff Johnston wrote:
> Corinna Vinschen wrote:
>
> > On Sep 20 11:32, Jeff Johnston wrote:
> >
> > > Corinna Vinschen wrote:
> > >
> > > > First, Jeff, would you mind to wait a few hours before changing stuff
> > > > which affects all targets like that?  The whole process of checkin plus
> > > > discussion took place in one single night from my timezone's point 
> > > > of view.
> > >
> > >
> > > There should be no problem with a change like this.  Every system 
> > > should be supplying and potentially overriding any header file 
> > > specific to its system.
> >
> >
> >
> > True, but a bit of preliminary discussion won't hurt, right?  We will
> > keep the Cygwin version for the next time, but I hope it wasn't wrong
> > to point out potential problems with the header.
>
> Not wrong at all.  Your points were well appreciated and this has 
> sparked some discussion and changes.

I've just watched so far. :)

I see nothing wrong with the discussion.  The RTEMS stdint.h file isn't 
perfect but at the same time no one had complained or submitted anything
else for newlib.  As Jeff said, newlib needed one in the shared include
files and the RTEMS one was the only one in newlib.

The RTEMS community would be extremely pleased to have a better 
implementation of this.  Sharing any .h file between Cygwin and
RTEMS and other newlib targets only helps improve all of us.

I don't see why stdint.h is tied to anything but gcc and eventually 
there should be one version for all newlib targets.

So submit problems, patches, etc.


> -- Jeff J.


--
Joel Sherrill, Ph.D.             Director of Research & Development
joel@OARcorp.com                 On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
   Support Available             (256) 722-9985