On Dec 14, 11:57 pm, "William Stein" <[email protected]> wrote:
> On Sun, Dec 14, 2008 at 11:20 PM, mabshoff
<SNIP>
> > This indicates that the Intel CPU you use doesn't support AMD specific
> > extensions :). I don't think we use any 3dnow[ext] or mmxext
> > instructions, so feel free to delete local/lib/sage-flags.txt to get
> > rid of this message. There is still a small chance that you might see
> > an "illegal instruction - aborting" situation, but I doubt it will
> > happen.
>
> Just for the record, on Saturday I was anxious and I took exactly that
> same Opteron binary and ran it on our new Intel Xeon-based server.
> Things mostly worked, but I *did* get segfaults when I created
> matrices then pressed [tab] to do tab completion via readline. I
> built sage from scratch on that machine, and the version built from
> scratch did *not* have these problems. Thus it's entirely possible
> that the user will find that their sage is not working right.
Yes, but I would blame those problems probably on the fact that it was
build with a different Ubuntu release, but obviously there are
potential problems there related to CPU instructions.
> > Yeah, building from source will avoid this altogether and you will get
> > optimum performance for your specific CPU this way.
>
> Definitely do that if you can. It's not supposed to be difficult (it
> just takes a lot of cputime), and you can let us know if it doesn't
> work, since we want to know.
Yep, nothing goes over build a custom tuned Sage for your specific CPU
- after all, this is Open Source :)
> -- William
Cheers,
Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/sage-support
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
