Re: Hugh's alternate page fault scalability approach on 512p Altix

[Pavel Machek]

Hi!

> 2. The use rss and anon_rss deltas instead of atomic incs brings small 
>    performance enhancements in the lower cpu ranges (1-32) but hurt (%50 
>    performance drop at 512 processors) in the high range. The hurting may
>    be due to the percularities of SGIs NUMA router architecture and the 

I'd say that 50% drop at 512 CPUs is acceptable. I did not even
know we support that many CPUs.
				Pavel

-- 
64 bytes from 195.113.31.123: icmp_seq=28 ttl=51 time=448769.1 ms         

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/