Re: [PATCH net-next] ipv4: Support multipath hashing on inner IP pkts for GRE tunnel

[Stephen Suryaputra <ssuryaextr@gmail.com>]

On Tue, Jun 11, 2019 at 10:29:56AM +0300, Nikolay Aleksandrov wrote:
> 
> Have you considered using the flow dissector and doing something similar to the bonding ?
> It does a full flow dissect via skb_flow_dissect_flow_keys() and uses whatever headers
> it needs, but that will support any tunneling protocol which the flow dissector
> recognizes and will be improved upon automatically by people adding to it.
> Also would avoid doing dissection by yourself.
> 
> The bond commit which added that was:
>  32819dc18348 ("bonding: modify the old and add new xmit hash policies")
> 

I didn't consider it and should. Thanks for pointing me to that
direction. It's simpler.

> >  /* if skb is set it will be used and fl4 can be NULL */
> >  int fib_multipath_hash(const struct net *net, const struct flowi4 *fl4,
> >  		       const struct sk_buff *skb, struct flow_keys *flkeys)
> > @@ -1828,12 +1876,13 @@ int fib_multipath_hash(const struct net *net, const struct flowi4 *fl4,
> >  	struct flow_keys hash_keys;
> >  	u32 mhash;
> >  
> > +	memset(&hash_keys, 0, sizeof(hash_keys));
> > +
> 
> This was an optimization, it was done on purpose to avoid doing anything when we
> have L3+4 configured (1) and the skb has its hash already calculated.
> 
Will revert to the original lines.

Thanks.