Re: [PATCH] CPU-time clocks

[Joel Sherrill <joel dot sherrill at oarcorp dot com>]

On 05/16/2011 07:04 AM, Ralf Corsepius wrote:
> On 05/16/2011 01:02 PM, Corinna Vinschen wrote:
> > On May 15 14:35, Yaakov (Cygwin/X) wrote:
> > > On Sun, 2011-05-15 at 13:49 -0500, Yaakov Selkowitz wrote:
> > > > I have a patch pending for Cygwin to support POSIX CPU-time clocks, but
> > > > newlib needs to be patched first.
> > > >
> > > > Unfortunately, newlib's clock ID definitions are not POSIX-compliant;
> > > > the names are missing the ending "_ID".  Fixing this will affect RTEMS,
> > > > as they have been using the misnamed definitions for some time.  I have
> > > > a patch for RTEMS' CVS HEAD, but the ramifications may be larger than
> > > > that, I don't know.
> > > >
> > > > Patch attached.
> > Thanks.  Maybe we should just define CLOCK_PROCESS_CPUTIME and
> > CLOCK_THREAD_CPUTIME for rtems as well.  So the rtems maintainers
> > can decide when this can go away:
> RTEMS prefers to stay in-line with POSIX.
>
> I.e. a radical switch from *_CPUTIME_ID for RTEMS would be OK with me.
I agree with Ralf.  If anyone is impacted, we will just have
to explain it. It isn't a huge deal for folks to change.  Please
just leave the POSIX constants.


> Ralf


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