On Tue, 2017-10-31 at 11:10 +0900, David Miller wrote: > From: Tom Herbert <t...@quantonium.net> > Date: Mon, 30 Oct 2017 14:16:00 -0700 > > > I wrote a quick test program that floods a whole bunch of these > > packets to a host and sure enough there is substantial time spent > > in ip6_parse_tlv. > ... > > 25.38% [kernel] [k] __fib6_clean_all > > 21.63% [kernel] [k] ip6_parse_tlv > > Yet the routing code still dominates the cost.
I am guessing the ip6_rt_max_size defaulting to 4096 needs to be revisited, after all per-cpu added stuff.
