I don't object to making it a personal patch, but if it's not hard to put it
back without the negative characteristics, then I say it's good to put back,
for exactly the reason Tom says.  That's pretty much how I used it - "these
results say what?  ohhhh....wrong patch applied when making the binary."

Lisa

On Wed, Jun 22, 2011 at 10:09 AM, Thomas Wenisch <[email protected]> wrote:

> FWIW, I have found it to be very valuable to have a repository revision
> number in simulation outputs.  Weeks later, when you try to figure out which
> crazy version of the code generated some random set of results, it can be
> extremely helpful for tracking things down.
>
> I would be a fan of fixing this with a compile-time option that is off by
> default as Nate suggests.
>
> Regards,
> -Thomas Wenisch
>
> On Jun 22, 2011, at 1:03 PM, nathan binkert wrote:
>
> >> Annoying, I responded to this yesterday with the wrong account and got a
> >> list rejection that I didn't notice until now.
> >>
> >> Anyhow, thanks - I'll just make my own personal patch to get it back, I
> can
> >> see how extra regressions would be annoying.
> >
> > If you (and others) like it, it could be a compile time option.  We
> > could probably also fix it if we handled it more like date.cc is
> > handled.  That shouldn't be too hard.
> >
> >  Nate
> > _______________________________________________
> > gem5-dev mailing list
> > [email protected]
> > http://m5sim.org/mailman/listinfo/gem5-dev
>
> _______________________________________________
> gem5-dev mailing list
> [email protected]
> http://m5sim.org/mailman/listinfo/gem5-dev
>
>
_______________________________________________
gem5-dev mailing list
[email protected]
http://m5sim.org/mailman/listinfo/gem5-dev

Reply via email to