[PATCH] Re: Whats up with _POSIX_CPUTIME

[Roland McGrath]

> On Wed, Dec 01, 2004 at 12:00:06PM -0800, Roland McGrath wrote:
> > > Did you just mean the s/#if #defined/#if defined/ in posix/tst-regex2.c?
> > > That's the only thing I found during testing...
> > 
> > That's the only thing I noticed, but it indicated the patch wasn't tested.
> > 
> > I don't like the __HAVE_CPUTIME stuff.  Already on ia64 it is a runtime
> > determination if the clocks are available.  
> 
> So should I redo that just to call __sysconf (_SC_{THREAD_,}CPUTIME)?

I think so, yes.  Or anyway, cope with runtime failures somehow.  I'd have
to look into the code closer to see if there is an obvious cheaper way to
catch it.