[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CA+mtBx9-iJjuCrL2J+jzcnztvVJsFjccs_FKO1z33HRkSw2PRg@mail.gmail.com>
Date: Tue, 17 Jun 2014 13:52:46 -0700
From: Tom Herbert <therbert@...gle.com>
To: Christopher White <chris@...icalelegance.com>
Cc: Linux Netdev List <netdev@...r.kernel.org>,
"Vina Ermagan (vermagan)" <vermagan@...co.com>,
"Lorand Jakab -X (lojakab - M SQUARED CONSULTING INC. at Cisco)"
<lojakab@...co.com>
Subject: Re: [PATCH V2 net-next] LISP: Locator/Identifier Separation Protocol
On Tue, Jun 17, 2014 at 1:41 PM, Christopher White
<chris@...icalelegance.com> wrote:
> Thanks Tom, some quick questions as I wrap this up….
>
> On Jun 5, 2014, at 3:12 PM, Tom Herbert <therbert@...gle.com> wrote:
>
>>
>>> +static inline struct rtable *find_route(struct net *net,
>>> + __be32 *saddr, __be32 daddr,
>>> + u8 ipproto, u8 tos, u32 skb_mark)
>>> +{
>>> + struct rtable *rt;
>>> +
>>> + /* Tunnel configuration keeps DSCP part of TOS bits, But Linux
>>> + * router expect RT_TOS bits only.
>>> + */
>>> + struct flowi4 fl = { .daddr = daddr,
>>> + .saddr = *saddr,
>>> + .flowi4_tos = RT_TOS(tos),
>>> + .flowi4_mark = skb_mark,
>>> + .flowi4_proto = ipproto };
>>> +
>>> + rt = ip_route_output_key(net, &fl);
>>> + *saddr = fl.saddr;
>>> + return rt;
>>> +}
>>> +
>> Looks like a function that should be in common header file?
>>
>
> I’m not certain where the right place would be for this. A new file perhaps?
>
It should be in include/net/route.h. There is already ip_route_output
and ip_route_output_gre which are very similar to this.
>>
>>> +static int create_v4_sock(struct net *net, __be16 port, struct socket **psock,
>>> + u32 flags)
>>> +{
>>> + struct sock *sk;
>>> + struct socket *sock;
>>> + struct sockaddr_in lisp_addr = {
>>> + .sin_family = AF_INET,
>>> + .sin_addr.s_addr = htonl(INADDR_ANY),
>>> + .sin_port = port,
>>> + };
>>> + int rc;
>>> +
>>> + /* Create UDP socket for encapsulation receive. */
>>> + rc = sock_create_kern(AF_INET, SOCK_DGRAM, IPPROTO_UDP, &sock);
>>> + if (rc < 0) {
>>> + pr_debug("UDP socket create failed\n");
>>> + return rc;
>>> + }
>>> +
>>> + /* Put in proper namespace */
>>> + sk = sock->sk;
>>> + sk_change_net(sk, net);
>>> +
>>> + rc = kernel_bind(sock, (struct sockaddr *)&lisp_addr,
>>> + sizeof(lisp_addr));
>>> + if (rc < 0) {
>>> + pr_debug("bind for UDP socket %pI4:%u (%d)\n",
>>> + &lisp_addr.sin_addr, ntohs(lisp_addr.sin_port), rc);
>>> + sk_release_kernel(sk);
>>> + return rc;
>>> + }
>>> +
>>> + *psock = sock;
>>> + /* Disable multicast loopback */
>>> + inet_sk(sk)->mc_loop = 0;
>>> +
>>> + if (!(flags & LISP_F_UDP_CSUM))
>>> + sock->sk->sk_no_check_tx = 1;
>>> + return 0;
>>> +}
>> Another function that is common among UDP tunneling protocols and
>> should be in common code.
>
> In udp.c perhaps? udp_tunnel_create_v4_sock()?
>
It might make sense for this to udp_offload.c (along with
skb_udp_tunnel_segment).
> Thanks!
> -Chris
>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists