[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20171106005921.GA13972@leo.usersys.redhat.com>
Date: Mon, 6 Nov 2017 08:59:21 +0800
From: Hangbin Liu <liuhangbin@...il.com>
To: Eric Dumazet <edumazet@...gle.com>
Cc: netdev <netdev@...r.kernel.org>, Paolo Abeni <pabeni@...hat.com>,
"David S. Miller" <davem@...emloft.net>
Subject: Re: [PATCH net] bonding: discard lowest hash bit for 802.3ad layer3+4
On Sun, Nov 05, 2017 at 01:38:47PM -0800, Eric Dumazet wrote:
> > diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
> > index c99dc59..728fa08 100644
> > --- a/drivers/net/bonding/bond_main.c
> > +++ b/drivers/net/bonding/bond_main.c
> > @@ -3237,7 +3237,7 @@ u32 bond_xmit_hash(struct bonding *bond, struct sk_buff *skb)
> >
> > if (bond->params.xmit_policy == BOND_XMIT_POLICY_ENCAP34 &&
> > skb->l4_hash)
> > - return skb->hash;
> > + return skb->hash >> 1;
>
> Why are you changing this part ?
>
> The l4 hash provided by local TCP stack does not use a pathological
> XOR based on ports/addresses,
> but a random value with pretty good entropy.
>
Oh, my bad. I only tested the patch works, but not carefully check the
skb->hash when skb->l4_hash is 1. Will send a v2 patch to fix this.
Regards
Hangbin
Powered by blists - more mailing lists