No we don't know what the bug is. I would guess that absolutely nothing stops the tuning program from generating invalid values (except that for the two algorithms in question there are no "invalid values").
The bug is in *all likelihood* caused by the tuning program never having been run for those arches. But if you want to be quite paranoid about it, check the two values in question are set to 0 by the tuning program and that the bug is no longer present after you run it. This will ensure the more recent algorithm is not run. By the law of averages, the more recent algorithm is more likely to contain a real bug than the old one. But again, this has nothing to do with "invalid values". You only get "invalid values" if you don't run the tuning program, which we didn't. Unless the tuning program also has a bug, which is possible. But that would be a completely independent bug since we know the tuning program has never been run on these arches. Bill. On 30 September 2015 at 13:12, Jeroen Demeyer <jdeme...@cage.ugent.be> wrote: > On 2015-09-30 12:49, 'Bill Hart' via mpir-devel wrote: > >> I'm not entirely sure the values 11 on their own were bad. That in >> combination with other missing tuning values, was bad. >> >> Let's just fix the bug instead of speculating shall we? >> > Well, do we know what the bug really is? It seems that you are not sure > that the tuning program always generates valid values, so I think that > needs to be clarified first. > > > -- > You received this message because you are subscribed to the Google Groups > "mpir-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to mpir-devel+unsubscr...@googlegroups.com. > To post to this group, send email to mpir-devel@googlegroups.com. > Visit this group at http://groups.google.com/group/mpir-devel. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "mpir-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to mpir-devel+unsubscr...@googlegroups.com. To post to this group, send email to mpir-devel@googlegroups.com. Visit this group at http://groups.google.com/group/mpir-devel. For more options, visit https://groups.google.com/d/optout.
