Re: [PATCH] MI: new timing command

[Daniel Jacobowitz <drow@false.org>]

On Sun, Dec 31, 2006 at 07:20:48PM +0300, Vladimir Prus wrote:
> On Sunday 31 December 2006 19:09, Mark Kettenis wrote:
> 
> > > Is this important? This timing is entirely for diagnostic purposes, 
> > > so why try to make it work on every possible platform. We need to document
> > > that -enable-timing may fail, and that's it.
> > 
> > The point is to use get_run_time() from -liberty and never worry about
> > portability again.
> 
> And lose user/system/elapse time distinction? For me, having all those times
> on Linux is more valuable than having any time at all on Windows. No because
> Windows is not important, but because I doubt there are any plantform specific
> MI slowdowns.

I really don't think the divided times are going to be useful.  But
then, I probably won't be the one using them.  If you would prefer to
maintain the distinction between system time and user time, then I
withdraw my objection.

-- 
Daniel Jacobowitz
CodeSourcery