Re: [RFC PATCHv2 net-next 1/3] net/udp: Add a new struct for hash2 slot

[Philo Lu]

On 2024/9/24 20:51, Gur Stavi wrote:
> > @@ -224,13 +224,12 @@ struct sock *__udp6_lib_lookup(const struct net *net,
> >   			       struct sk_buff *skb)
> >   {
> >   	unsigned short hnum = ntohs(dport);
> > -	unsigned int hash2, slot2;
> >   	struct udp_hslot *hslot2;
> >   	struct sock *result, *sk;
> > +	unsigned int hash2;
> >
> >   	hash2 = ipv6_portaddr_hash(net, daddr, hnum);
> > -	slot2 = hash2 & udptable->mask;
> > -	hslot2 = &udptable->hash2[slot2];
> > +	hslot2 = udp_hashslot2(udptable, hash2);
>
> Why not minimize the code change by using udptable->hash2[slot2].hslot?
> Especially since later you do it in __udp6_lib_mcast_deliver.
> I think that many developers would find usage of C primitives more
> readable.

Yes I can use udptable->hash2[slot2].hslot. But I prefer the 
udp_hashslot2() helper, because I found I have to add ".hslot" at many 
places, while most of them look exactly what udp_hashslot2() does. I 
think replacing them may get the codes simpler and benefit future coding.

Some places like __udp6_lib_mcast_deliver() get hslot2 differently from 
udp_hashslot2(), so just appending ".hslot" to avoid functionality change.

Thanks.
--
Philo