|| ||Nick Piggin <firstname.lastname@example.org>|
|| ||linux-kernel <email@example.com>|
|| ||Nick's scheduler v17|
|| ||Sat, 25 Oct 2003 04:10:24 +1000|
Still working on SMP and NUMA. Some (maybe) interesting things I put in are
- Sequential CPU balancing so you don't get a big storm of balances
- Balancing is trying to err more on the side of caution, I have to start
analysing it more thoroughly though.
- Attacked the NUMA balancing code. There should now be less buslocked ops /
cache pingpongs in some fastpaths. Volanomark likes it, more realistic
won't improve so much http://www.kerneltrap.org/~npiggin/v17/volano.png
This improvement is NUMA only.
I haven't had time to reproduce Cliff's serious reaim performance
they're probably still there. I couldn't reproduce Martin's kernbench
but the 16-way I'm using only has 512K cache which might not show it up.
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to firstname.lastname@example.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/