On Wed, Apr 29, 2015 at 06:16:35PM -0700, Alexei Starovoitov wrote: > On Thu, Apr 30, 2015 at 02:48:39AM +0200, Patrick McHardy wrote: > > Netfilter is based on hook chains. The cost when only using a single hook > > is minimal (as Pablo showed in his numbers), but even if only using > > TC and a single netfilter classifier chain, there has to be some relative > > ordering and the hooks provide this in a generic way. > > Unfortunately the numbers that Pablo shown are not measuring > the right thing. The numbers I've shown were measuring the critical ingress path. I never posted numbers on generic hook infrastructure. > > 840203pps 403Mb/sec > > this is 20 times less than what they should be. It's ~10% less, because obviously we run more code in the generic hook infrastructure, as they provide more features. So far it was only possible to register one single chain at ingress. Eric Dumazet already indicated that this benchmark only measures *one single CPU* and icache is fully populated. The numbers with smp will go worse because of the central lock contention. These are the numbers I got banging *one single CPU*: * Without patches + qdisc ingress: Result: OK: 16298126(c16298125+d0) usec, 10000000 (60byte,0frags) 613567pps 294Mb/sec (294512160bps) errors: 10000000 * With patches + qdisc ingress on top of hooks: Result: OK: 18339281(c18339280+d0) usec, 10000000 (60byte,0frags) 545277pps 261Mb/sec (261732960bps) errors: 10000000 * With patches + nftables ingress chain: Result: OK: 17118167(c17118167+d0) usec, 10000000 (60byte,0frags) 584174pps 280Mb/sec (280403520bps) errors: 10000000 -- To unsubscribe from this list: send the line "unsubscribe netfilter-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
