On Thu, Oct 29, 2015 at 8:00 AM, Paolo Abeni <pab...@redhat.com> wrote: > On Thu, 2015-10-29 at 07:31 -0700, Eric Dumazet wrote: >> On Wed, 2015-10-28 at 18:39 +0100, Paolo Abeni wrote: >> > This patch changes how the multipath hash is computed for locally >> > generated UDP or TCP flows: now the hash comprises also l4 information >> > (source and destination port). >> > >> > This allows better utilization of the available paths when the existing >> > flows have the same source IP and the same destination IP: with l3 hash, >> > even when multiple connections are in place simultaneously, a single path >> > will be used, while with l4 hash we can use all the available paths. >> > >> > Signed-off-by: Paolo Abeni <pab...@redhat.com> >> >> Please note that many skb already have a l4 hash. No need to perform yet >> another dissection. > > This patch do not add dissection code: it use the information provided > by the available flowi4 structure. Moreover the skb is not available on > the calling site (in __ip_route_output_key_hash) and pushing it all the > way will require a lot of intrusive changes. Do you think it's the > better option ? > If there is an associated skb then skb_get_hash should be called to get the hash. If you only have the flow structure then get_hash_from_flowi4 (or skb_get_hash_flowi6 for IPv6 should be called).
Tom > Paolo > > -- > To unsubscribe from this list: send the line "unsubscribe netdev" in > the body of a message to majord...@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html