> >So the problem really seems to be link somehow with Pentium PRO and
> >Pentium II processors !
>
> Sound really strange. Can you run the same binary image of the Pentium
> (that works fine) on the PII/PPROs (just to remove all software variables
> from the equation?)?
>
> Andrea[s] Arcangeli
I took the binary compiled on Pentium 100 MHz with gcc 2.7.2 + glibc 1
and I tried it on our IBM dual PPro 200 MHz (256 kb cache): the
problem is still there. The performance loss does not seem to be
linked with gcc/egcs or glibc version. We have reproduced
(unfortunately) the problem with 2 different dual PII systems, single
and dual PPro systems that had a variety of gcc/egcs/glibc versions.
I also disabled the L2 cache to see if that would make any
differences: the system is slower and there is still a performance
loss. So it does not seem linked with L2 cache either.
We will keep investigating, (we should try Andrea profiling test this
afternoon)
Claude
-- Claude Gamache, CAE Electronique Ltee, 8585 Cote-de-Liesse Saint-Laurent, Quebec, Canada H4T 1G6 Email: cgamache@cae.ca Tel.: (514) 341-2000 x3194- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/faq.html