Hi there,
in the past, I reported about my multithreaded benchmark being faster with
HZ=3D400 than with HZ=3D100 on ix86.
=46rom that, a defficiency in the linux SMP scheduling algorithm could be
guessed.
I created a HZ=3D400 patch (configurable) for ix86, which can be found
on my web page http://www.garloff.de/kurt/linux/=20
The benchmark is on http://pebbles.e-technik.uni-dortmund.de/numlib/
Now, with the scheduler changes in 2.2.8, my benchmark runs at about the
same speed with 100 or 400Hz, so I think the problem is cured.
Regards,
--=20
Kurt Garloff <garloff@suse.de> SuSE GmbH, N=FCrnberg, FRG
Linux kernel development; SCSI driver: DC390 (tmscsim/AM53C974)
--WfZ7S8PLGjBY9Voh
Content-Type: application/pgp-signature
-----BEGIN PGP SIGNATURE-----
Version: 2.6.3in
iQCVAwUBN1ZjoBaQN/7O/JIVAQF1iwP9GjzGNcklk5n6Ng9Ws4pDc0ZIICt3JVw6
j0NKxxzI8Ha6F5uJ0gmH+mhXvdeRYKYSjPCAxlVhdhpBQt4K91ukJZyb00GL3K1j
Wdaq9GNBKRhiSelFPWcW+YPUGDf1A1I8IvpH9ZouskiuDQVEc6O9dYU0lYHDYT7a
S8E9KhWUlVA=
=FGkp
-----END PGP SIGNATURE-----
--WfZ7S8PLGjBY9Voh--
-
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/