On Thu, Sep 20, 2012 at 5:48 PM, Tritium Cat <[email protected]> wrote:
>
> 1743:inline u_int32_t hash_pkt_header(struct pfring_pkthdr * hdr, u_char
> mask_src, u_char mask_dst,
> 1744: u_int8_t mask_port, u_int8_t mask_proto,
> u_int8_t mask_vlan)
> 1745:{
> 1746: if(hdr->extended_hdr.pkt_hash == 0) {
> 1747: u_int8_t use_tunneled_peers =
> hdr->extended_hdr.parsed_pkt.tunnel.tunnel_id == NO_TUNNEL_ID ? 0 : 1;
> 1749: hdr->extended_hdr.pkt_hash = hash_pkt(
> 1750: mask_vlan ? 0 : hdr->extended_hdr.parsed_pkt.vlan_id,
>
I verify with "enable_debugging" that the vlan_id is sent to hash_pkt
regardless of cluster_mode.
[PF_RING] hash_pkt(vlan_id=88, proto=6, port_peer_a=XX, port_peer_b=XX)
[PF_RING] -> lockless_list_get_first=xxxxxxxxxxxxxx [num
elements=1][last_list_idx=1]
[PF_RING] --> add_skb_to_ring(len=64) [channel_id=2/48][active=0][eth5]
[PF_RING] -> lockless_list_get_next= (null) [num
elements=1][last_list_idx=1]
[PF_RING] kfree_skb()
For now I "patch" line 1750 with "0" to fix the problem. Something with
ternary operator ?
/tc
_______________________________________________
Ntop-misc mailing list
[email protected]
http://listgateway.unipi.it/mailman/listinfo/ntop-misc
