Hangbin Liu <liuhangbin@xxxxxxxxx> writes: > On Wed, May 27, 2020 at 12:21:54PM +0200, Toke Høiland-Jørgensen wrote: >> > The example in patch 2 is functional, but not a lot of effort >> > has been made on performance optimisation. I did a simple test(pkt size 64) >> > with pktgen. Here is the test result with BPF_MAP_TYPE_DEVMAP_HASH >> > arrays: >> > >> > bpf_redirect_map() with 1 ingress, 1 egress: >> > generic path: ~1600k pps >> > native path: ~980k pps >> > >> > bpf_redirect_map_multi() with 1 ingress, 3 egress: >> > generic path: ~600k pps >> > native path: ~480k pps >> > >> > bpf_redirect_map_multi() with 1 ingress, 9 egress: >> > generic path: ~125k pps >> > native path: ~100k pps >> > >> > The bpf_redirect_map_multi() is slower than bpf_redirect_map() as we loop >> > the arrays and do clone skb/xdpf. The native path is slower than generic >> > path as we send skbs by pktgen. So the result looks reasonable. >> >> How are you running these tests? Still on virtual devices? We really > > I run it with the test topology in patch 2/2. The test is run on physical > machines, but I use veth interface. Do you mean use a physical NIC driver > for testing? Yes, sorry, when I said 'physical machine' I should have also 'physical NIC'. We really need to know how the performance of this is on the XDP fast path, i.e., when there are no skbs involved at all. > BTW, when using pktgen, I got an panic because the skb don't have enough > header room. The code path looks like > > do_xdp_generic() > - netif_receive_generic_xdp() > - skb_headroom(skb) < XDP_PACKET_HEADROOM > - pskb_expand_head() > - BUG_ON(skb_shared(skb)) > > So I added a draft patch for pktgen, not sure if it has any influence. Hmm, as Jesper said pktgen was really not intended to be used this way, so I guess that's why. I guess I'll let him comment on whether he thinks it's worth fixing; or you could send this as a proper patch and see if anyone complains about it ;) -Toke
