Dave, Please consider pulling the following changesets available on the 2.6.20-net-2.6-20070214-FOR_DAVEM branch at <git://git.linux-ipv6.org/gitroot/yoshfuji/linux-2.6-dev.git> (on top of commit 928ba4169dc1d82c83105831f5ddb5472379b440).
Thank you. HEADLINES --------- [IPV6] ADDRCONF: Manage prefix route corresponding to address manually added. [IPV6] IP6TUNNEL: Use update_pmtu() of dst on xmit. [IPV6] IP6TUNNEL: Split out generic rutine in ip6ip6_err(). [IPV6] IP6TUNNEL: Split out generic rutine in ip6ip6_rcv(). [IPV6] IP6TUNNEL: Split out generic rutine in ip6ip6_xmit(). [IPV6] IP6TUNNEL: Add support to IPv4 over IPv6 tunnel. [IPV6] IP6TUNNEL: Rename functions ip6ip6_* to ip6_tnl_*. [IPV6] IP6TUNNEL: Enable to control the handled inner protocol. [IPV6] ADDRCONF: Register inet6_dev earlier. DIFFSTAT -------- net/ipv6/addrconf.c | 30 ++ net/ipv6/ip6_tunnel.c | 630 +++++++++++++++++++++++++++++++++++-------------- 2 files changed, 478 insertions(+), 182 deletions(-) CHANGESETS ---------- commit bda786c68d05984f5d4a1812d3ebd2290d5bea57 Author: YOSHIFUJI Hideaki <[EMAIL PROTECTED]> Date: Thu Feb 15 01:09:10 2007 +0900 [IPV6] ADDRCONF: Manage prefix route corresponding to address manually added. It is more natural to manage prefix routes corresponding to address which is being added manually. With help from Masafumi Aramoto <[EMAIL PROTECTED]>. Signed-off-by: YOSHIFUJI Hideaki <[EMAIL PROTECTED]> diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c index ea0755b..c6683ea 100644 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@ -1911,6 +1911,7 @@ static int inet6_addr_add(int ifindex, struct in6_addr *pfx, int plen, struct inet6_dev *idev; struct net_device *dev; int scope; + u32 flags = RTF_EXPIRES; ASSERT_RTNL(); @@ -1926,9 +1927,10 @@ static int inet6_addr_add(int ifindex, struct in6_addr *pfx, int plen, scope = ipv6_addr_scope(pfx); - if (valid_lft == INFINITY_LIFE_TIME) + if (valid_lft == INFINITY_LIFE_TIME) { ifa_flags |= IFA_F_PERMANENT; - else if (valid_lft >= 0x7FFFFFFF/HZ) + flags = 0; + } else if (valid_lft >= 0x7FFFFFFF/HZ) valid_lft = 0x7FFFFFFF/HZ; if (prefered_lft == 0) @@ -1946,6 +1948,8 @@ static int inet6_addr_add(int ifindex, struct in6_addr *pfx, int plen, ifp->tstamp = jiffies; spin_unlock_bh(&ifp->lock); + addrconf_prefix_route(&ifp->addr, ifp->prefix_len, dev, + jiffies_to_clock_t(valid_lft * HZ), flags); addrconf_dad_start(ifp, 0); in6_ifa_put(ifp); addrconf_verify(0); @@ -2125,6 +2129,7 @@ static void addrconf_add_linklocal(struct inet6_dev *idev, struct in6_addr *addr ifp = ipv6_add_addr(idev, addr, 64, IFA_LINK, IFA_F_PERMANENT); if (!IS_ERR(ifp)) { + addrconf_prefix_route(&ifp->addr, ifp->prefix_len, idev->dev, 0, 0); addrconf_dad_start(ifp, 0); in6_ifa_put(ifp); } @@ -2539,10 +2544,6 @@ static void addrconf_dad_start(struct inet6_ifaddr *ifp, u32 flags) addrconf_join_solict(dev, &ifp->addr); - if (ifp->prefix_len != 128 && (ifp->flags&IFA_F_PERMANENT)) - addrconf_prefix_route(&ifp->addr, ifp->prefix_len, dev, 0, - flags); - net_srandom(ifp->addr.s6_addr32[3]); read_lock_bh(&idev->lock); @@ -2973,12 +2974,15 @@ inet6_rtm_deladdr(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg) static int inet6_addr_modify(struct inet6_ifaddr *ifp, u8 ifa_flags, u32 prefered_lft, u32 valid_lft) { + u32 flags = RTF_EXPIRES; + if (!valid_lft || (prefered_lft > valid_lft)) return -EINVAL; - if (valid_lft == INFINITY_LIFE_TIME) + if (valid_lft == INFINITY_LIFE_TIME) { ifa_flags |= IFA_F_PERMANENT; - else if (valid_lft >= 0x7FFFFFFF/HZ) + flags = 0; + } else if (valid_lft >= 0x7FFFFFFF/HZ) valid_lft = 0x7FFFFFFF/HZ; if (prefered_lft == 0) @@ -2997,6 +3001,8 @@ static int inet6_addr_modify(struct inet6_ifaddr *ifp, u8 ifa_flags, if (!(ifp->flags&IFA_F_TENTATIVE)) ipv6_ifa_notify(0, ifp); + addrconf_prefix_route(&ifp->addr, ifp->prefix_len, ifp->idev->dev, + jiffies_to_clock_t(valid_lft * HZ), flags); addrconf_verify(0); return 0; --- commit b5ea2d4352fab08edd0cbceb280c15c764063f0d Author: Yasuyuki Kozakai <[EMAIL PROTECTED]> Date: Thu Feb 15 01:09:11 2007 +0900 [IPV6] IP6TUNNEL: Use update_pmtu() of dst on xmit. Signed-off-by: Yasuyuki Kozakai <[EMAIL PROTECTED]> Signed-off-by: YOSHIFUJI Hideaki <[EMAIL PROTECTED]> diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 662edb8..08d9442 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -727,11 +727,8 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) } if (mtu < IPV6_MIN_MTU) mtu = IPV6_MIN_MTU; - if (skb->dst && mtu < dst_mtu(skb->dst)) { - struct rt6_info *rt = (struct rt6_info *) skb->dst; - rt->rt6i_flags |= RTF_MODIFIED; - rt->u.dst.metrics[RTAX_MTU-1] = mtu; - } + if (skb->dst) + skb->dst->ops->update_pmtu(skb->dst, mtu); if (skb->len > mtu) { icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, dev); goto tx_err_dst_release; --- commit c4b2af7ef485929b75ce4f8e76b402c25b078b21 Author: Yasuyuki Kozakai <[EMAIL PROTECTED]> Date: Thu Feb 15 01:09:12 2007 +0900 [IPV6] IP6TUNNEL: Split out generic rutine in ip6ip6_err(). This enables to add IPv4/IPv6 specific error handling later, Signed-off-by: Yasuyuki Kozakai <[EMAIL PROTECTED]> Signed-off-by: YOSHIFUJI Hideaki <[EMAIL PROTECTED]> diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 08d9442..6022fc5 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -372,16 +372,16 @@ parse_tlv_tnl_enc_lim(struct sk_buff *skb, __u8 * raw) } /** - * ip6ip6_err - tunnel error handler + * ip6_tnl_err - tunnel error handler * * Description: - * ip6ip6_err() should handle errors in the tunnel according + * ip6_tnl_err() should handle errors in the tunnel according * to the specifications in RFC 2473. **/ static int -ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, - int type, int code, int offset, __be32 info) +ip6_tnl_err(struct sk_buff *skb, struct inet6_skb_parm *opt, + int *type, int *code, int *msg, __be32 *info, int offset) { struct ipv6hdr *ipv6h = (struct ipv6hdr *) skb->data; struct ip6_tnl *t; @@ -402,7 +402,7 @@ ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, err = 0; - switch (type) { + switch (*type) { __u32 teli; struct ipv6_tlv_tnl_enc_lim *tel; __u32 mtu; @@ -414,7 +414,7 @@ ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, rel_msg = 1; break; case ICMPV6_TIME_EXCEED: - if (code == ICMPV6_EXC_HOPLIMIT) { + if ((*code) == ICMPV6_EXC_HOPLIMIT) { if (net_ratelimit()) printk(KERN_WARNING "%s: Too small hop limit or " @@ -425,10 +425,10 @@ ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, break; case ICMPV6_PARAMPROB: teli = 0; - if (code == ICMPV6_HDR_FIELD) + if ((*code) == ICMPV6_HDR_FIELD) teli = parse_tlv_tnl_enc_lim(skb, skb->data); - if (teli && teli == ntohl(info) - 2) { + if (teli && teli == ntohl(*info) - 2) { tel = (struct ipv6_tlv_tnl_enc_lim *) &skb->data[teli]; if (tel->encap_limit == 0) { if (net_ratelimit()) @@ -445,7 +445,7 @@ ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, } break; case ICMPV6_PKT_TOOBIG: - mtu = ntohl(info) - offset; + mtu = ntohl(*info) - offset; if (mtu < IPV6_MIN_MTU) mtu = IPV6_MIN_MTU; t->dev->mtu = mtu; @@ -458,12 +458,38 @@ ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, } break; } - if (rel_msg && pskb_may_pull(skb, offset + sizeof (*ipv6h))) { + + *type = rel_type; + *code = rel_code; + *info = rel_info; + *msg = rel_msg; + +out: + read_unlock(&ip6ip6_lock); + return err; +} + +static int +ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, + int type, int code, int offset, __u32 info) +{ + int rel_msg = 0; + int rel_type = type; + int rel_code = code; + __u32 rel_info = info; + int err; + + err = ip6_tnl_err(skb, opt, &rel_type, &rel_code, &rel_msg, &rel_info, + offset); + if (err < 0) + return err; + + if (rel_msg && pskb_may_pull(skb, offset + sizeof(struct ipv6hdr))) { struct rt6_info *rt; struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC); if (!skb2) - goto out; + return 0; dst_release(skb2->dst); skb2->dst = NULL; @@ -483,9 +509,8 @@ ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, kfree_skb(skb2); } -out: - read_unlock(&ip6ip6_lock); - return err; + + return 0; } static inline void ip6ip6_ecn_decapsulate(struct ipv6hdr *outer_iph, --- commit 6755e45a94433858930b8ec5d13e7fc1c5835899 Author: Yasuyuki Kozakai <[EMAIL PROTECTED]> Date: Thu Feb 15 01:09:13 2007 +0900 [IPV6] IP6TUNNEL: Split out generic rutine in ip6ip6_rcv(). This enables to add IPv4/IPv6 specific handling later, Signed-off-by: Yasuyuki Kozakai <[EMAIL PROTECTED]> Signed-off-by: YOSHIFUJI Hideaki <[EMAIL PROTECTED]> diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 6022fc5..985d106 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -513,14 +513,17 @@ ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, return 0; } -static inline void ip6ip6_ecn_decapsulate(struct ipv6hdr *outer_iph, - struct sk_buff *skb) +static void ip6ip6_dscp_ecn_decapsulate(struct ip6_tnl *t, + struct ipv6hdr *ipv6h, + struct sk_buff *skb) { - struct ipv6hdr *inner_iph = skb->nh.ipv6h; + if (t->parms.flags & IP6_TNL_F_RCV_DSCP_COPY) + ipv6_copy_dscp(ipv6h, skb->nh.ipv6h); - if (INET_ECN_is_ce(ipv6_get_dsfield(outer_iph))) - IP6_ECN_set_ce(inner_iph); + if (INET_ECN_is_ce(ipv6_get_dsfield(ipv6h))) + IP6_ECN_set_ce(skb->nh.ipv6h); } + static inline int ip6_tnl_rcv_ctl(struct ip6_tnl *t) { struct ip6_tnl_parm *p = &t->parms; @@ -546,12 +549,16 @@ static inline int ip6_tnl_rcv_ctl(struct ip6_tnl *t) /** * ip6ip6_rcv - decapsulate IPv6 packet and retransmit it locally * @skb: received socket buffer + * @protocol: ethernet protocol ID + * @dscp_ecn_decapsulate: the function to decapsulate DSCP code and ECN * * Return: 0 **/ -static int -ip6ip6_rcv(struct sk_buff *skb) +static int ip6_tnl_rcv(struct sk_buff *skb, __u16 protocol, + void (*dscp_ecn_decapsulate)(struct ip6_tnl *t, + struct ipv6hdr *ipv6h, + struct sk_buff *skb)) { struct ipv6hdr *ipv6h; struct ip6_tnl *t; @@ -574,16 +581,16 @@ ip6ip6_rcv(struct sk_buff *skb) secpath_reset(skb); skb->mac.raw = skb->nh.raw; skb->nh.raw = skb->data; - skb->protocol = htons(ETH_P_IPV6); + skb->protocol = htons(protocol); skb->pkt_type = PACKET_HOST; memset(skb->cb, 0, sizeof(struct inet6_skb_parm)); skb->dev = t->dev; dst_release(skb->dst); skb->dst = NULL; nf_reset(skb); - if (t->parms.flags & IP6_TNL_F_RCV_DSCP_COPY) - ipv6_copy_dscp(ipv6h, skb->nh.ipv6h); - ip6ip6_ecn_decapsulate(ipv6h, skb); + + dscp_ecn_decapsulate(t, ipv6h, skb); + t->stat.rx_packets++; t->stat.rx_bytes += skb->len; netif_rx(skb); @@ -598,6 +605,11 @@ discard: return 0; } +static int ip6ip6_rcv(struct sk_buff *skb) +{ + return ip6_tnl_rcv(skb, ETH_P_IPV6, ip6ip6_dscp_ecn_decapsulate); +} + struct ipv6_tel_txoption { struct ipv6_txoptions ops; __u8 dst_opt[8]; --- commit 2df8f28648dfe839060b1e0fe4e4871bee72efd5 Author: Yasuyuki Kozakai <[EMAIL PROTECTED]> Date: Thu Feb 15 01:09:13 2007 +0900 [IPV6] IP6TUNNEL: Split out generic rutine in ip6ip6_xmit(). This enables to add IPv4/IPv6 specific handling later, Signed-off-by: Yasuyuki Kozakai <[EMAIL PROTECTED]> Signed-off-by: YOSHIFUJI Hideaki <[EMAIL PROTECTED]> diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 985d106..4546bb9 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -678,9 +678,13 @@ static inline int ip6_tnl_xmit_ctl(struct ip6_tnl *t) return ret; } /** - * ip6ip6_tnl_xmit - encapsulate packet and send + * ip6_tnl_xmit2 - encapsulate packet and send * @skb: the outgoing socket buffer * @dev: the outgoing tunnel device + * @dsfield: dscp code for outer header + * @fl: flow of tunneled packet + * @encap_limit: encapsulation limit + * @pmtu: Path MTU is stored if packet is too big * * Description: * Build new header and do some sanity checks on the packet before sending @@ -688,62 +692,35 @@ static inline int ip6_tnl_xmit_ctl(struct ip6_tnl *t) * * Return: * 0 + * -1 fail + * %-EMSGSIZE message too big. return mtu in this case. **/ -static int -ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) +static int ip6_tnl_xmit2(struct sk_buff *skb, + struct net_device *dev, + __u8 dsfield, + struct flowi *fl, + int encap_limit, + __u32 *pmtu) { struct ip6_tnl *t = netdev_priv(dev); struct net_device_stats *stats = &t->stat; struct ipv6hdr *ipv6h = skb->nh.ipv6h; - int encap_limit = -1; struct ipv6_tel_txoption opt; - __u16 offset; - struct flowi fl; struct dst_entry *dst; struct net_device *tdev; int mtu; int max_headroom = sizeof(struct ipv6hdr); u8 proto; - int err; + int err = -1; int pkt_len; - int dsfield; - - if (t->recursion++) { - stats->collisions++; - goto tx_err; - } - if (skb->protocol != htons(ETH_P_IPV6) || - !ip6_tnl_xmit_ctl(t) || ip6ip6_tnl_addr_conflict(t, ipv6h)) - goto tx_err; - - if ((offset = parse_tlv_tnl_enc_lim(skb, skb->nh.raw)) > 0) { - struct ipv6_tlv_tnl_enc_lim *tel; - tel = (struct ipv6_tlv_tnl_enc_lim *) &skb->nh.raw[offset]; - if (tel->encap_limit == 0) { - icmpv6_send(skb, ICMPV6_PARAMPROB, - ICMPV6_HDR_FIELD, offset + 2, skb->dev); - goto tx_err; - } - encap_limit = tel->encap_limit - 1; - } else if (!(t->parms.flags & IP6_TNL_F_IGN_ENCAP_LIMIT)) - encap_limit = t->parms.encap_limit; - - memcpy(&fl, &t->fl, sizeof (fl)); - proto = fl.proto; - - dsfield = ipv6_get_dsfield(ipv6h); - if ((t->parms.flags & IP6_TNL_F_USE_ORIG_TCLASS)) - fl.fl6_flowlabel |= (*(__be32 *) ipv6h & IPV6_TCLASS_MASK); - if ((t->parms.flags & IP6_TNL_F_USE_ORIG_FLOWLABEL)) - fl.fl6_flowlabel |= (*(__be32 *) ipv6h & IPV6_FLOWLABEL_MASK); if ((dst = ip6_tnl_dst_check(t)) != NULL) dst_hold(dst); else { - dst = ip6_route_output(NULL, &fl); + dst = ip6_route_output(NULL, fl); - if (dst->error || xfrm_lookup(&dst, &fl, NULL, 0) < 0) + if (dst->error || xfrm_lookup(&dst, fl, NULL, 0) < 0) goto tx_err_link_failure; } @@ -767,7 +744,8 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) if (skb->dst) skb->dst->ops->update_pmtu(skb->dst, mtu); if (skb->len > mtu) { - icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, dev); + *pmtu = mtu; + err = -EMSGSIZE; goto tx_err_dst_release; } @@ -793,20 +771,21 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) skb->h.raw = skb->nh.raw; + proto = fl->proto; if (encap_limit >= 0) { init_tel_txopt(&opt, encap_limit); ipv6_push_nfrag_opts(skb, &opt.ops, &proto, NULL); } skb->nh.raw = skb_push(skb, sizeof(struct ipv6hdr)); ipv6h = skb->nh.ipv6h; - *(__be32*)ipv6h = fl.fl6_flowlabel | htonl(0x60000000); + *(__be32*)ipv6h = fl->fl6_flowlabel | htonl(0x60000000); dsfield = INET_ECN_encapsulate(0, dsfield); ipv6_change_dsfield(ipv6h, ~INET_ECN_MASK, dsfield); ipv6h->payload_len = htons(skb->len - sizeof(struct ipv6hdr)); ipv6h->hop_limit = t->parms.hop_limit; ipv6h->nexthdr = proto; - ipv6_addr_copy(&ipv6h->saddr, &fl.fl6_src); - ipv6_addr_copy(&ipv6h->daddr, &fl.fl6_dst); + ipv6_addr_copy(&ipv6h->saddr, &fl->fl6_src); + ipv6_addr_copy(&ipv6h->daddr, &fl->fl6_dst); nf_reset(skb); pkt_len = skb->len; err = NF_HOOK(PF_INET6, NF_IP6_LOCAL_OUT, skb, NULL, @@ -820,13 +799,87 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) stats->tx_aborted_errors++; } ip6_tnl_dst_store(t, dst); - t->recursion--; return 0; tx_err_link_failure: stats->tx_carrier_errors++; dst_link_failure(skb); tx_err_dst_release: dst_release(dst); + return err; +} + +static inline int +ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) +{ + struct ip6_tnl *t = netdev_priv(dev); + struct ipv6hdr *ipv6h = skb->nh.ipv6h; + int encap_limit = -1; + __u16 offset; + struct flowi fl; + __u8 dsfield; + __u32 mtu; + int err; + + if (!ip6_tnl_xmit_ctl(t) || ip6ip6_tnl_addr_conflict(t, ipv6h)) + return -1; + + if ((offset = parse_tlv_tnl_enc_lim(skb, skb->nh.raw)) > 0) { + struct ipv6_tlv_tnl_enc_lim *tel; + tel = (struct ipv6_tlv_tnl_enc_lim *) &skb->nh.raw[offset]; + if (tel->encap_limit == 0) { + icmpv6_send(skb, ICMPV6_PARAMPROB, + ICMPV6_HDR_FIELD, offset + 2, skb->dev); + return -1; + } + encap_limit = tel->encap_limit - 1; + } else if (!(t->parms.flags & IP6_TNL_F_IGN_ENCAP_LIMIT)) + encap_limit = t->parms.encap_limit; + + memcpy(&fl, &t->fl, sizeof (fl)); + fl.proto = IPPROTO_IPV6; + + dsfield = ipv6_get_dsfield(ipv6h); + if ((t->parms.flags & IP6_TNL_F_USE_ORIG_TCLASS)) + fl.fl6_flowlabel |= (*(__be32 *) ipv6h & IPV6_TCLASS_MASK); + if ((t->parms.flags & IP6_TNL_F_USE_ORIG_FLOWLABEL)) + fl.fl6_flowlabel |= (*(__be32 *) ipv6h & IPV6_FLOWLABEL_MASK); + + err = ip6_tnl_xmit2(skb, dev, dsfield, &fl, encap_limit, &mtu); + if (err != 0) { + if (err == -EMSGSIZE) + icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, dev); + return -1; + } + + return 0; +} + +static int +ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) +{ + struct ip6_tnl *t = netdev_priv(dev); + struct net_device_stats *stats = &t->stat; + int ret; + + if (t->recursion++) { + t->stat.collisions++; + goto tx_err; + } + + switch (skb->protocol) { + case __constant_htons(ETH_P_IPV6): + ret = ip6ip6_tnl_xmit(skb, dev); + break; + default: + goto tx_err; + } + + if (ret < 0) + goto tx_err; + + t->recursion--; + return 0; + tx_err: stats->tx_errors++; stats->tx_dropped++; @@ -1088,7 +1141,7 @@ static void ip6ip6_tnl_dev_setup(struct net_device *dev) SET_MODULE_OWNER(dev); dev->uninit = ip6ip6_tnl_dev_uninit; dev->destructor = free_netdev; - dev->hard_start_xmit = ip6ip6_tnl_xmit; + dev->hard_start_xmit = ip6_tnl_xmit; dev->get_stats = ip6ip6_tnl_get_stats; dev->do_ioctl = ip6ip6_tnl_ioctl; dev->change_mtu = ip6ip6_tnl_change_mtu; --- commit c635a40acaeccb97a9df3c38ac5bea1fca8c5411 Author: Yasuyuki Kozakai <[EMAIL PROTECTED]> Date: Thu Feb 15 01:09:14 2007 +0900 [IPV6] IP6TUNNEL: Add support to IPv4 over IPv6 tunnel. Some notes - Protocol number IPPROTO_IPIP is used for IPv4 over IPv6 packets. - If IP6_TNL_F_USE_ORIG_TCLASS is set, TOS in IPv4 header is copied to Traffic Class in outer IPv6 header on xmit. - IP6_TNL_F_USE_ORIG_FLOWLABEL is ignored on xmit of IPv4 packets, because IPv4 header does not have flow label. - Kernel sends ICMP error if IPv4 packet is too big on xmit, even if DF flag is not set. Signed-off-by: Yasuyuki Kozakai <[EMAIL PROTECTED]> Signed-off-by: YOSHIFUJI Hideaki <[EMAIL PROTECTED]> diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 4546bb9..a654149 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -1,14 +1,15 @@ /* - * IPv6 over IPv6 tunnel device + * IPv6 tunneling device * Linux INET6 implementation * * Authors: * Ville Nuorvala <[EMAIL PROTECTED]> + * Yasuyuki Kozakai <[EMAIL PROTECTED]> * * $Id$ * * Based on: - * linux/net/ipv6/sit.c + * linux/net/ipv6/sit.c and linux/net/ipv4/ipip.c * * RFC 2473 * @@ -24,6 +25,7 @@ #include <linux/errno.h> #include <linux/types.h> #include <linux/sockios.h> +#include <linux/icmp.h> #include <linux/if.h> #include <linux/in.h> #include <linux/ip.h> @@ -41,6 +43,7 @@ #include <asm/uaccess.h> #include <asm/atomic.h> +#include <net/icmp.h> #include <net/ip.h> #include <net/ipv6.h> #include <net/ip6_route.h> @@ -51,7 +54,7 @@ #include <net/inet_ecn.h> MODULE_AUTHOR("Ville Nuorvala"); -MODULE_DESCRIPTION("IPv6-in-IPv6 tunnel"); +MODULE_DESCRIPTION("IPv6 tunneling device"); MODULE_LICENSE("GPL"); #define IPV6_TLV_TEL_DST_SIZE 8 @@ -63,6 +66,7 @@ MODULE_LICENSE("GPL"); #endif #define IPV6_TCLASS_MASK (IPV6_FLOWINFO_MASK & ~IPV6_FLOWLABEL_MASK) +#define IPV6_TCLASS_SHIFT 20 #define HASH_SIZE 32 @@ -470,6 +474,104 @@ out: } static int +ip4ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, + int type, int code, int offset, __u32 info) +{ + int rel_msg = 0; + int rel_type = type; + int rel_code = code; + __u32 rel_info = info; + int err; + struct sk_buff *skb2; + struct iphdr *eiph; + struct flowi fl; + struct rtable *rt; + + err = ip6_tnl_err(skb, opt, &rel_type, &rel_code, &rel_msg, &rel_info, + offset); + if (err < 0) + return err; + + if (rel_msg == 0) + return 0; + + switch (rel_type) { + case ICMPV6_DEST_UNREACH: + if (rel_code != ICMPV6_ADDR_UNREACH) + return 0; + rel_type = ICMP_DEST_UNREACH; + rel_code = ICMP_HOST_UNREACH; + break; + case ICMPV6_PKT_TOOBIG: + if (rel_code != 0) + return 0; + rel_type = ICMP_DEST_UNREACH; + rel_code = ICMP_FRAG_NEEDED; + break; + default: + return 0; + } + + if (!pskb_may_pull(skb, offset + sizeof(struct iphdr))) + return 0; + + skb2 = skb_clone(skb, GFP_ATOMIC); + if (!skb2) + return 0; + + dst_release(skb2->dst); + skb2->dst = NULL; + skb_pull(skb2, offset); + skb2->nh.raw = skb2->data; + eiph = skb2->nh.iph; + + /* Try to guess incoming interface */ + memset(&fl, 0, sizeof(fl)); + fl.fl4_dst = eiph->saddr; + fl.fl4_tos = RT_TOS(eiph->tos); + fl.proto = IPPROTO_IPIP; + if (ip_route_output_key(&rt, &fl)) + goto out; + + skb2->dev = rt->u.dst.dev; + + /* route "incoming" packet */ + if (rt->rt_flags & RTCF_LOCAL) { + ip_rt_put(rt); + rt = NULL; + fl.fl4_dst = eiph->daddr; + fl.fl4_src = eiph->saddr; + fl.fl4_tos = eiph->tos; + if (ip_route_output_key(&rt, &fl) || + rt->u.dst.dev->type != ARPHRD_TUNNEL) { + ip_rt_put(rt); + goto out; + } + } else { + ip_rt_put(rt); + if (ip_route_input(skb2, eiph->daddr, eiph->saddr, eiph->tos, + skb2->dev) || + skb2->dst->dev->type != ARPHRD_TUNNEL) + goto out; + } + + /* change mtu on this route */ + if (rel_type == ICMP_DEST_UNREACH && rel_code == ICMP_FRAG_NEEDED) { + if (rel_info > dst_mtu(skb2->dst)) + goto out; + + skb2->dst->ops->update_pmtu(skb2->dst, rel_info); + rel_info = htonl(rel_info); + } + + icmp_send(skb2, rel_type, rel_code, rel_info); + +out: + kfree_skb(skb2); + return 0; +} + +static int ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, int type, int code, int offset, __u32 info) { @@ -513,6 +615,19 @@ ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, return 0; } +static void ip4ip6_dscp_ecn_decapsulate(struct ip6_tnl *t, + struct ipv6hdr *ipv6h, + struct sk_buff *skb) +{ + __u8 dsfield = ipv6_get_dsfield(ipv6h) & ~INET_ECN_MASK; + + if (t->parms.flags & IP6_TNL_F_RCV_DSCP_COPY) + ipv4_change_dsfield(skb->nh.iph, INET_ECN_MASK, dsfield); + + if (INET_ECN_is_ce(dsfield)) + IP_ECN_set_ce(skb->nh.iph); +} + static void ip6ip6_dscp_ecn_decapsulate(struct ip6_tnl *t, struct ipv6hdr *ipv6h, struct sk_buff *skb) @@ -605,6 +720,11 @@ discard: return 0; } +static int ip4ip6_rcv(struct sk_buff *skb) +{ + return ip6_tnl_rcv(skb, ETH_P_IP, ip4ip6_dscp_ecn_decapsulate); +} + static int ip6ip6_rcv(struct sk_buff *skb) { return ip6_tnl_rcv(skb, ETH_P_IPV6, ip6ip6_dscp_ecn_decapsulate); @@ -691,7 +811,7 @@ static inline int ip6_tnl_xmit_ctl(struct ip6_tnl *t) * it. * * Return: - * 0 + * 0 on success * -1 fail * %-EMSGSIZE message too big. return mtu in this case. **/ @@ -809,6 +929,44 @@ tx_err_dst_release: } static inline int +ip4ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) +{ + struct ip6_tnl *t = netdev_priv(dev); + struct iphdr *iph = skb->nh.iph; + int encap_limit = -1; + struct flowi fl; + __u8 dsfield; + __u32 mtu; + int err; + + if (!ip6_tnl_xmit_ctl(t)) + return -1; + + if (!(t->parms.flags & IP6_TNL_F_IGN_ENCAP_LIMIT)) + encap_limit = t->parms.encap_limit; + + memcpy(&fl, &t->fl, sizeof (fl)); + fl.proto = IPPROTO_IPIP; + + dsfield = ipv4_get_dsfield(iph); + + if ((t->parms.flags & IP6_TNL_F_USE_ORIG_TCLASS)) + fl.fl6_flowlabel |= ntohl(((__u32)iph->tos << IPV6_TCLASS_SHIFT) + & IPV6_TCLASS_MASK); + + err = ip6_tnl_xmit2(skb, dev, dsfield, &fl, encap_limit, &mtu); + if (err != 0) { + /* XXX: send ICMP error even if DF is not set. */ + if (err == -EMSGSIZE) + icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, + htonl(mtu)); + return -1; + } + + return 0; +} + +static inline int ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) { struct ip6_tnl *t = netdev_priv(dev); @@ -867,6 +1025,9 @@ ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) } switch (skb->protocol) { + case __constant_htons(ETH_P_IP): + ret = ip4ip6_tnl_xmit(skb, dev); + break; case __constant_htons(ETH_P_IPV6): ret = ip6ip6_tnl_xmit(skb, dev); break; @@ -1199,6 +1360,12 @@ ip6ip6_fb_tnl_dev_init(struct net_device *dev) return 0; } +static struct xfrm6_tunnel ip4ip6_handler = { + .handler = ip4ip6_rcv, + .err_handler = ip4ip6_err, + .priority = 1, +}; + static struct xfrm6_tunnel ip6ip6_handler = { .handler = ip6ip6_rcv, .err_handler = ip6ip6_err, @@ -1215,9 +1382,16 @@ static int __init ip6_tunnel_init(void) { int err; + if (xfrm6_tunnel_register(&ip4ip6_handler, AF_INET)) { + printk(KERN_ERR "ip4ip6 init: can't register tunnel\n"); + err = -EAGAIN; + goto out; + } + if (xfrm6_tunnel_register(&ip6ip6_handler, AF_INET6)) { printk(KERN_ERR "ip6ip6 init: can't register tunnel\n"); - return -EAGAIN; + err = -EAGAIN; + goto unreg_ip4ip6; } ip6ip6_fb_tnl_dev = alloc_netdev(sizeof(struct ip6_tnl), "ip6tnl0", ip6ip6_tnl_dev_setup); @@ -1235,6 +1409,9 @@ static int __init ip6_tunnel_init(void) return 0; fail: xfrm6_tunnel_deregister(&ip6ip6_handler, AF_INET6); +unreg_ip4ip6: + xfrm6_tunnel_deregister(&ip4ip6_handler, AF_INET); +out: return err; } @@ -1258,6 +1435,9 @@ static void __exit ip6ip6_destroy_tunnels(void) static void __exit ip6_tunnel_cleanup(void) { + if (xfrm6_tunnel_deregister(&ip4ip6_handler, AF_INET)) + printk(KERN_INFO "ip4ip6 close: can't deregister tunnel\n"); + if (xfrm6_tunnel_deregister(&ip6ip6_handler, AF_INET6)) printk(KERN_INFO "ip6ip6 close: can't deregister tunnel\n"); --- commit 2ef4c54a4510772fbc3353068c61926f6bd2b4ae Author: Yasuyuki Kozakai <[EMAIL PROTECTED]> Date: Thu Feb 15 01:09:15 2007 +0900 [IPV6] IP6TUNNEL: Rename functions ip6ip6_* to ip6_tnl_*. Signed-off-by: Yasuyuki Kozakai <[EMAIL PROTECTED]> Signed-off-by: YOSHIFUJI Hideaki <[EMAIL PROTECTED]> diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index a654149..70684e0 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -74,12 +74,12 @@ MODULE_LICENSE("GPL"); (addr)->s6_addr32[2] ^ (addr)->s6_addr32[3]) & \ (HASH_SIZE - 1)) -static int ip6ip6_fb_tnl_dev_init(struct net_device *dev); -static int ip6ip6_tnl_dev_init(struct net_device *dev); -static void ip6ip6_tnl_dev_setup(struct net_device *dev); +static int ip6_fb_tnl_dev_init(struct net_device *dev); +static int ip6_tnl_dev_init(struct net_device *dev); +static void ip6_tnl_dev_setup(struct net_device *dev); /* the IPv6 tunnel fallback device */ -static struct net_device *ip6ip6_fb_tnl_dev; +static struct net_device *ip6_fb_tnl_dev; /* lists for storing tunnels in use */ @@ -88,7 +88,7 @@ static struct ip6_tnl *tnls_wc[1]; static struct ip6_tnl **tnls[2] = { tnls_wc, tnls_r_l }; /* lock for the tunnel lists */ -static DEFINE_RWLOCK(ip6ip6_lock); +static DEFINE_RWLOCK(ip6_tnl_lock); static inline struct dst_entry *ip6_tnl_dst_check(struct ip6_tnl *t) { @@ -119,7 +119,7 @@ static inline void ip6_tnl_dst_store(struct ip6_tnl *t, struct dst_entry *dst) } /** - * ip6ip6_tnl_lookup - fetch tunnel matching the end-point addresses + * ip6_tnl_lookup - fetch tunnel matching the end-point addresses * @remote: the address of the tunnel exit-point * @local: the address of the tunnel entry-point * @@ -130,7 +130,7 @@ static inline void ip6_tnl_dst_store(struct ip6_tnl *t, struct dst_entry *dst) **/ static struct ip6_tnl * -ip6ip6_tnl_lookup(struct in6_addr *remote, struct in6_addr *local) +ip6_tnl_lookup(struct in6_addr *remote, struct in6_addr *local) { unsigned h0 = HASH(remote); unsigned h1 = HASH(local); @@ -149,18 +149,18 @@ ip6ip6_tnl_lookup(struct in6_addr *remote, struct in6_addr *local) } /** - * ip6ip6_bucket - get head of list matching given tunnel parameters + * ip6_tnl_bucket - get head of list matching given tunnel parameters * @p: parameters containing tunnel end-points * * Description: - * ip6ip6_bucket() returns the head of the list matching the + * ip6_tnl_bucket() returns the head of the list matching the * &struct in6_addr entries laddr and raddr in @p. * * Return: head of IPv6 tunnel list **/ static struct ip6_tnl ** -ip6ip6_bucket(struct ip6_tnl_parm *p) +ip6_tnl_bucket(struct ip6_tnl_parm *p) { struct in6_addr *remote = &p->raddr; struct in6_addr *local = &p->laddr; @@ -175,36 +175,36 @@ ip6ip6_bucket(struct ip6_tnl_parm *p) } /** - * ip6ip6_tnl_link - add tunnel to hash table + * ip6_tnl_link - add tunnel to hash table * @t: tunnel to be added **/ static void -ip6ip6_tnl_link(struct ip6_tnl *t) +ip6_tnl_link(struct ip6_tnl *t) { - struct ip6_tnl **tp = ip6ip6_bucket(&t->parms); + struct ip6_tnl **tp = ip6_tnl_bucket(&t->parms); t->next = *tp; - write_lock_bh(&ip6ip6_lock); + write_lock_bh(&ip6_tnl_lock); *tp = t; - write_unlock_bh(&ip6ip6_lock); + write_unlock_bh(&ip6_tnl_lock); } /** - * ip6ip6_tnl_unlink - remove tunnel from hash table + * ip6_tnl_unlink - remove tunnel from hash table * @t: tunnel to be removed **/ static void -ip6ip6_tnl_unlink(struct ip6_tnl *t) +ip6_tnl_unlink(struct ip6_tnl *t) { struct ip6_tnl **tp; - for (tp = ip6ip6_bucket(&t->parms); *tp; tp = &(*tp)->next) { + for (tp = ip6_tnl_bucket(&t->parms); *tp; tp = &(*tp)->next) { if (t == *tp) { - write_lock_bh(&ip6ip6_lock); + write_lock_bh(&ip6_tnl_lock); *tp = t->next; - write_unlock_bh(&ip6ip6_lock); + write_unlock_bh(&ip6_tnl_lock); break; } } @@ -241,12 +241,12 @@ static struct ip6_tnl *ip6_tnl_create(struct ip6_tnl_parm *p) if (i == IP6_TNL_MAX) goto failed; } - dev = alloc_netdev(sizeof (*t), name, ip6ip6_tnl_dev_setup); + dev = alloc_netdev(sizeof (*t), name, ip6_tnl_dev_setup); if (dev == NULL) goto failed; t = netdev_priv(dev); - dev->init = ip6ip6_tnl_dev_init; + dev->init = ip6_tnl_dev_init; t->parms = *p; if ((err = register_netdevice(dev)) < 0) { @@ -254,19 +254,19 @@ static struct ip6_tnl *ip6_tnl_create(struct ip6_tnl_parm *p) goto failed; } dev_hold(dev); - ip6ip6_tnl_link(t); + ip6_tnl_link(t); return t; failed: return NULL; } /** - * ip6ip6_tnl_locate - find or create tunnel matching given parameters + * ip6_tnl_locate - find or create tunnel matching given parameters * @p: tunnel parameters * @create: != 0 if allowed to create new tunnel if no match found * * Description: - * ip6ip6_tnl_locate() first tries to locate an existing tunnel + * ip6_tnl_locate() first tries to locate an existing tunnel * based on @parms. If this is unsuccessful, but @create is set a new * tunnel device is created and registered for use. * @@ -274,13 +274,13 @@ failed: * matching tunnel or NULL **/ -static struct ip6_tnl *ip6ip6_tnl_locate(struct ip6_tnl_parm *p, int create) +static struct ip6_tnl *ip6_tnl_locate(struct ip6_tnl_parm *p, int create) { struct in6_addr *remote = &p->raddr; struct in6_addr *local = &p->laddr; struct ip6_tnl *t; - for (t = *ip6ip6_bucket(p); t; t = t->next) { + for (t = *ip6_tnl_bucket(p); t; t = t->next) { if (ipv6_addr_equal(local, &t->parms.laddr) && ipv6_addr_equal(remote, &t->parms.raddr)) return t; @@ -291,24 +291,24 @@ static struct ip6_tnl *ip6ip6_tnl_locate(struct ip6_tnl_parm *p, int create) } /** - * ip6ip6_tnl_dev_uninit - tunnel device uninitializer + * ip6_tnl_dev_uninit - tunnel device uninitializer * @dev: the device to be destroyed * * Description: - * ip6ip6_tnl_dev_uninit() removes tunnel from its list + * ip6_tnl_dev_uninit() removes tunnel from its list **/ static void -ip6ip6_tnl_dev_uninit(struct net_device *dev) +ip6_tnl_dev_uninit(struct net_device *dev) { struct ip6_tnl *t = netdev_priv(dev); - if (dev == ip6ip6_fb_tnl_dev) { - write_lock_bh(&ip6ip6_lock); + if (dev == ip6_fb_tnl_dev) { + write_lock_bh(&ip6_tnl_lock); tnls_wc[0] = NULL; - write_unlock_bh(&ip6ip6_lock); + write_unlock_bh(&ip6_tnl_lock); } else { - ip6ip6_tnl_unlink(t); + ip6_tnl_unlink(t); } ip6_tnl_dst_reset(t); dev_put(dev); @@ -400,8 +400,8 @@ ip6_tnl_err(struct sk_buff *skb, struct inet6_skb_parm *opt, in trouble since we might need the source address for further processing of the error. */ - read_lock(&ip6ip6_lock); - if ((t = ip6ip6_tnl_lookup(&ipv6h->daddr, &ipv6h->saddr)) == NULL) + read_lock(&ip6_tnl_lock); + if ((t = ip6_tnl_lookup(&ipv6h->daddr, &ipv6h->saddr)) == NULL) goto out; err = 0; @@ -469,7 +469,7 @@ ip6_tnl_err(struct sk_buff *skb, struct inet6_skb_parm *opt, *msg = rel_msg; out: - read_unlock(&ip6ip6_lock); + read_unlock(&ip6_tnl_lock); return err; } @@ -662,7 +662,7 @@ static inline int ip6_tnl_rcv_ctl(struct ip6_tnl *t) } /** - * ip6ip6_rcv - decapsulate IPv6 packet and retransmit it locally + * ip6_tnl_rcv - decapsulate IPv6 packet and retransmit it locally * @skb: received socket buffer * @protocol: ethernet protocol ID * @dscp_ecn_decapsulate: the function to decapsulate DSCP code and ECN @@ -680,17 +680,17 @@ static int ip6_tnl_rcv(struct sk_buff *skb, __u16 protocol, ipv6h = skb->nh.ipv6h; - read_lock(&ip6ip6_lock); + read_lock(&ip6_tnl_lock); - if ((t = ip6ip6_tnl_lookup(&ipv6h->saddr, &ipv6h->daddr)) != NULL) { + if ((t = ip6_tnl_lookup(&ipv6h->saddr, &ipv6h->daddr)) != NULL) { if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb)) { - read_unlock(&ip6ip6_lock); + read_unlock(&ip6_tnl_lock); goto discard; } if (!ip6_tnl_rcv_ctl(t)) { t->stat.rx_dropped++; - read_unlock(&ip6ip6_lock); + read_unlock(&ip6_tnl_lock); goto discard; } secpath_reset(skb); @@ -709,10 +709,10 @@ static int ip6_tnl_rcv(struct sk_buff *skb, __u16 protocol, t->stat.rx_packets++; t->stat.rx_bytes += skb->len; netif_rx(skb); - read_unlock(&ip6ip6_lock); + read_unlock(&ip6_tnl_lock); return 0; } - read_unlock(&ip6ip6_lock); + read_unlock(&ip6_tnl_lock); return 1; discard: @@ -750,7 +750,7 @@ static void init_tel_txopt(struct ipv6_tel_txoption *opt, __u8 encap_limit) } /** - * ip6ip6_tnl_addr_conflict - compare packet addresses to tunnel's own + * ip6_tnl_addr_conflict - compare packet addresses to tunnel's own * @t: the outgoing tunnel device * @hdr: IPv6 header from the incoming packet * @@ -764,7 +764,7 @@ static void init_tel_txopt(struct ipv6_tel_txoption *opt, __u8 encap_limit) **/ static inline int -ip6ip6_tnl_addr_conflict(struct ip6_tnl *t, struct ipv6hdr *hdr) +ip6_tnl_addr_conflict(struct ip6_tnl *t, struct ipv6hdr *hdr) { return ipv6_addr_equal(&t->parms.raddr, &hdr->saddr); } @@ -978,7 +978,7 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) __u32 mtu; int err; - if (!ip6_tnl_xmit_ctl(t) || ip6ip6_tnl_addr_conflict(t, ipv6h)) + if (!ip6_tnl_xmit_ctl(t) || ip6_tnl_addr_conflict(t, ipv6h)) return -1; if ((offset = parse_tlv_tnl_enc_lim(skb, skb->nh.raw)) > 0) { @@ -1068,7 +1068,7 @@ static void ip6_tnl_set_cap(struct ip6_tnl *t) } } -static void ip6ip6_tnl_link_config(struct ip6_tnl *t) +static void ip6_tnl_link_config(struct ip6_tnl *t) { struct net_device *dev = t->dev; struct ip6_tnl_parm *p = &t->parms; @@ -1121,17 +1121,17 @@ static void ip6ip6_tnl_link_config(struct ip6_tnl *t) } /** - * ip6ip6_tnl_change - update the tunnel parameters + * ip6_tnl_change - update the tunnel parameters * @t: tunnel to be changed * @p: tunnel configuration parameters * @active: != 0 if tunnel is ready for use * * Description: - * ip6ip6_tnl_change() updates the tunnel parameters + * ip6_tnl_change() updates the tunnel parameters **/ static int -ip6ip6_tnl_change(struct ip6_tnl *t, struct ip6_tnl_parm *p) +ip6_tnl_change(struct ip6_tnl *t, struct ip6_tnl_parm *p) { ipv6_addr_copy(&t->parms.laddr, &p->laddr); ipv6_addr_copy(&t->parms.raddr, &p->raddr); @@ -1141,18 +1141,18 @@ ip6ip6_tnl_change(struct ip6_tnl *t, struct ip6_tnl_parm *p) t->parms.flowinfo = p->flowinfo; t->parms.link = p->link; ip6_tnl_dst_reset(t); - ip6ip6_tnl_link_config(t); + ip6_tnl_link_config(t); return 0; } /** - * ip6ip6_tnl_ioctl - configure ipv6 tunnels from userspace + * ip6_tnl_ioctl - configure ipv6 tunnels from userspace * @dev: virtual device associated with tunnel * @ifr: parameters passed from userspace * @cmd: command to be performed * * Description: - * ip6ip6_tnl_ioctl() is used for managing IPv6 tunnels + * ip6_tnl_ioctl() is used for managing IPv6 tunnels * from userspace. * * The possible commands are the following: @@ -1174,7 +1174,7 @@ ip6ip6_tnl_change(struct ip6_tnl *t, struct ip6_tnl_parm *p) **/ static int -ip6ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) +ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) { int err = 0; struct ip6_tnl_parm p; @@ -1182,12 +1182,12 @@ ip6ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) switch (cmd) { case SIOCGETTUNNEL: - if (dev == ip6ip6_fb_tnl_dev) { + if (dev == ip6_fb_tnl_dev) { if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof (p))) { err = -EFAULT; break; } - t = ip6ip6_tnl_locate(&p, 0); + t = ip6_tnl_locate(&p, 0); } if (t == NULL) t = netdev_priv(dev); @@ -1207,8 +1207,8 @@ ip6ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) err = -EINVAL; if (p.proto != IPPROTO_IPV6) break; - t = ip6ip6_tnl_locate(&p, cmd == SIOCADDTUNNEL); - if (dev != ip6ip6_fb_tnl_dev && cmd == SIOCCHGTUNNEL) { + t = ip6_tnl_locate(&p, cmd == SIOCADDTUNNEL); + if (dev != ip6_fb_tnl_dev && cmd == SIOCCHGTUNNEL) { if (t != NULL) { if (t->dev != dev) { err = -EEXIST; @@ -1217,9 +1217,9 @@ ip6ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) } else t = netdev_priv(dev); - ip6ip6_tnl_unlink(t); - err = ip6ip6_tnl_change(t, &p); - ip6ip6_tnl_link(t); + ip6_tnl_unlink(t); + err = ip6_tnl_change(t, &p); + ip6_tnl_link(t); netdev_state_change(dev); } if (t) { @@ -1235,15 +1235,15 @@ ip6ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) if (!capable(CAP_NET_ADMIN)) break; - if (dev == ip6ip6_fb_tnl_dev) { + if (dev == ip6_fb_tnl_dev) { err = -EFAULT; if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof (p))) break; err = -ENOENT; - if ((t = ip6ip6_tnl_locate(&p, 0)) == NULL) + if ((t = ip6_tnl_locate(&p, 0)) == NULL) break; err = -EPERM; - if (t->dev == ip6ip6_fb_tnl_dev) + if (t->dev == ip6_fb_tnl_dev) break; dev = t->dev; } @@ -1257,20 +1257,20 @@ ip6ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) } /** - * ip6ip6_tnl_get_stats - return the stats for tunnel device + * ip6_tnl_get_stats - return the stats for tunnel device * @dev: virtual device associated with tunnel * * Return: stats for device **/ static struct net_device_stats * -ip6ip6_tnl_get_stats(struct net_device *dev) +ip6_tnl_get_stats(struct net_device *dev) { return &(((struct ip6_tnl *)netdev_priv(dev))->stat); } /** - * ip6ip6_tnl_change_mtu - change mtu manually for tunnel device + * ip6_tnl_change_mtu - change mtu manually for tunnel device * @dev: virtual device associated with tunnel * @new_mtu: the new mtu * @@ -1280,7 +1280,7 @@ ip6ip6_tnl_get_stats(struct net_device *dev) **/ static int -ip6ip6_tnl_change_mtu(struct net_device *dev, int new_mtu) +ip6_tnl_change_mtu(struct net_device *dev, int new_mtu) { if (new_mtu < IPV6_MIN_MTU) { return -EINVAL; @@ -1290,22 +1290,22 @@ ip6ip6_tnl_change_mtu(struct net_device *dev, int new_mtu) } /** - * ip6ip6_tnl_dev_setup - setup virtual tunnel device + * ip6_tnl_dev_setup - setup virtual tunnel device * @dev: virtual device associated with tunnel * * Description: * Initialize function pointers and device parameters **/ -static void ip6ip6_tnl_dev_setup(struct net_device *dev) +static void ip6_tnl_dev_setup(struct net_device *dev) { SET_MODULE_OWNER(dev); - dev->uninit = ip6ip6_tnl_dev_uninit; + dev->uninit = ip6_tnl_dev_uninit; dev->destructor = free_netdev; dev->hard_start_xmit = ip6_tnl_xmit; - dev->get_stats = ip6ip6_tnl_get_stats; - dev->do_ioctl = ip6ip6_tnl_ioctl; - dev->change_mtu = ip6ip6_tnl_change_mtu; + dev->get_stats = ip6_tnl_get_stats; + dev->do_ioctl = ip6_tnl_ioctl; + dev->change_mtu = ip6_tnl_change_mtu; dev->type = ARPHRD_TUNNEL6; dev->hard_header_len = LL_MAX_HEADER + sizeof (struct ipv6hdr); @@ -1316,12 +1316,12 @@ static void ip6ip6_tnl_dev_setup(struct net_device *dev) /** - * ip6ip6_tnl_dev_init_gen - general initializer for all tunnel devices + * ip6_tnl_dev_init_gen - general initializer for all tunnel devices * @dev: virtual device associated with tunnel **/ static inline void -ip6ip6_tnl_dev_init_gen(struct net_device *dev) +ip6_tnl_dev_init_gen(struct net_device *dev) { struct ip6_tnl *t = netdev_priv(dev); t->fl.proto = IPPROTO_IPV6; @@ -1330,31 +1330,31 @@ ip6ip6_tnl_dev_init_gen(struct net_device *dev) } /** - * ip6ip6_tnl_dev_init - initializer for all non fallback tunnel devices + * ip6_tnl_dev_init - initializer for all non fallback tunnel devices * @dev: virtual device associated with tunnel **/ static int -ip6ip6_tnl_dev_init(struct net_device *dev) +ip6_tnl_dev_init(struct net_device *dev) { struct ip6_tnl *t = netdev_priv(dev); - ip6ip6_tnl_dev_init_gen(dev); - ip6ip6_tnl_link_config(t); + ip6_tnl_dev_init_gen(dev); + ip6_tnl_link_config(t); return 0; } /** - * ip6ip6_fb_tnl_dev_init - initializer for fallback tunnel device + * ip6_fb_tnl_dev_init - initializer for fallback tunnel device * @dev: fallback device * * Return: 0 **/ static int -ip6ip6_fb_tnl_dev_init(struct net_device *dev) +ip6_fb_tnl_dev_init(struct net_device *dev) { struct ip6_tnl *t = netdev_priv(dev); - ip6ip6_tnl_dev_init_gen(dev); + ip6_tnl_dev_init_gen(dev); dev_hold(dev); tnls_wc[0] = t; return 0; @@ -1383,27 +1383,27 @@ static int __init ip6_tunnel_init(void) int err; if (xfrm6_tunnel_register(&ip4ip6_handler, AF_INET)) { - printk(KERN_ERR "ip4ip6 init: can't register tunnel\n"); + printk(KERN_ERR "ip6_tunnel init: can't register ip4ip6\n"); err = -EAGAIN; goto out; } if (xfrm6_tunnel_register(&ip6ip6_handler, AF_INET6)) { - printk(KERN_ERR "ip6ip6 init: can't register tunnel\n"); + printk(KERN_ERR "ip6_tunnel init: can't register ip6ip6\n"); err = -EAGAIN; goto unreg_ip4ip6; } - ip6ip6_fb_tnl_dev = alloc_netdev(sizeof(struct ip6_tnl), "ip6tnl0", - ip6ip6_tnl_dev_setup); + ip6_fb_tnl_dev = alloc_netdev(sizeof(struct ip6_tnl), "ip6tnl0", + ip6_tnl_dev_setup); - if (!ip6ip6_fb_tnl_dev) { + if (!ip6_fb_tnl_dev) { err = -ENOMEM; goto fail; } - ip6ip6_fb_tnl_dev->init = ip6ip6_fb_tnl_dev_init; + ip6_fb_tnl_dev->init = ip6_fb_tnl_dev_init; - if ((err = register_netdev(ip6ip6_fb_tnl_dev))) { - free_netdev(ip6ip6_fb_tnl_dev); + if ((err = register_netdev(ip6_fb_tnl_dev))) { + free_netdev(ip6_fb_tnl_dev); goto fail; } return 0; @@ -1415,7 +1415,7 @@ out: return err; } -static void __exit ip6ip6_destroy_tunnels(void) +static void __exit ip6_tnl_destroy_tunnels(void) { int h; struct ip6_tnl *t; @@ -1436,13 +1436,13 @@ static void __exit ip6ip6_destroy_tunnels(void) static void __exit ip6_tunnel_cleanup(void) { if (xfrm6_tunnel_deregister(&ip4ip6_handler, AF_INET)) - printk(KERN_INFO "ip4ip6 close: can't deregister tunnel\n"); + printk(KERN_INFO "ip6_tunnel close: can't deregister ip4ip6\n"); if (xfrm6_tunnel_deregister(&ip6ip6_handler, AF_INET6)) - printk(KERN_INFO "ip6ip6 close: can't deregister tunnel\n"); + printk(KERN_INFO "ip6_tunnel close: can't deregister ip6ip6\n"); rtnl_lock(); - ip6ip6_destroy_tunnels(); + ip6_tnl_destroy_tunnels(); rtnl_unlock(); } --- commit cf83f0fa3e9dafab6e9f547a1106bd2540732a1f Author: Yasuyuki Kozakai <[EMAIL PROTECTED]> Date: Thu Feb 15 01:09:15 2007 +0900 [IPV6] IP6TUNNEL: Enable to control the handled inner protocol. ip6_tunnel before supporting IPv4/IPv6 tunnel allows only IPPROTO_IPV6 in configurations from userland. This allows userland to set IPPROTO_IPIP and 0(wildcard). ip6_tunnel only handles allowed inner protocols. Signed-off-by: Yasuyuki Kozakai <[EMAIL PROTECTED]> Signed-off-by: YOSHIFUJI Hideaki <[EMAIL PROTECTED]> diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 70684e0..d8c84d8 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -384,7 +384,7 @@ parse_tlv_tnl_enc_lim(struct sk_buff *skb, __u8 * raw) **/ static int -ip6_tnl_err(struct sk_buff *skb, struct inet6_skb_parm *opt, +ip6_tnl_err(struct sk_buff *skb, __u8 ipproto, struct inet6_skb_parm *opt, int *type, int *code, int *msg, __be32 *info, int offset) { struct ipv6hdr *ipv6h = (struct ipv6hdr *) skb->data; @@ -404,6 +404,9 @@ ip6_tnl_err(struct sk_buff *skb, struct inet6_skb_parm *opt, if ((t = ip6_tnl_lookup(&ipv6h->daddr, &ipv6h->saddr)) == NULL) goto out; + if (t->parms.proto != ipproto && t->parms.proto != 0) + goto out; + err = 0; switch (*type) { @@ -487,8 +490,8 @@ ip4ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, struct flowi fl; struct rtable *rt; - err = ip6_tnl_err(skb, opt, &rel_type, &rel_code, &rel_msg, &rel_info, - offset); + err = ip6_tnl_err(skb, IPPROTO_IPIP, opt, &rel_type, &rel_code, + &rel_msg, &rel_info, offset); if (err < 0) return err; @@ -581,8 +584,8 @@ ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, __u32 rel_info = info; int err; - err = ip6_tnl_err(skb, opt, &rel_type, &rel_code, &rel_msg, &rel_info, - offset); + err = ip6_tnl_err(skb, IPPROTO_IPV6, opt, &rel_type, &rel_code, + &rel_msg, &rel_info, offset); if (err < 0) return err; @@ -671,6 +674,7 @@ static inline int ip6_tnl_rcv_ctl(struct ip6_tnl *t) **/ static int ip6_tnl_rcv(struct sk_buff *skb, __u16 protocol, + __u8 ipproto, void (*dscp_ecn_decapsulate)(struct ip6_tnl *t, struct ipv6hdr *ipv6h, struct sk_buff *skb)) @@ -683,6 +687,11 @@ static int ip6_tnl_rcv(struct sk_buff *skb, __u16 protocol, read_lock(&ip6_tnl_lock); if ((t = ip6_tnl_lookup(&ipv6h->saddr, &ipv6h->daddr)) != NULL) { + if (t->parms.proto != ipproto && t->parms.proto != 0) { + read_unlock(&ip6_tnl_lock); + goto discard; + } + if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb)) { read_unlock(&ip6_tnl_lock); goto discard; @@ -722,12 +731,14 @@ discard: static int ip4ip6_rcv(struct sk_buff *skb) { - return ip6_tnl_rcv(skb, ETH_P_IP, ip4ip6_dscp_ecn_decapsulate); + return ip6_tnl_rcv(skb, ETH_P_IP, IPPROTO_IPIP, + ip4ip6_dscp_ecn_decapsulate); } static int ip6ip6_rcv(struct sk_buff *skb) { - return ip6_tnl_rcv(skb, ETH_P_IPV6, ip6ip6_dscp_ecn_decapsulate); + return ip6_tnl_rcv(skb, ETH_P_IPV6, IPPROTO_IPV6, + ip6ip6_dscp_ecn_decapsulate); } struct ipv6_tel_txoption { @@ -939,7 +950,8 @@ ip4ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) __u32 mtu; int err; - if (!ip6_tnl_xmit_ctl(t)) + if ((t->parms.proto != IPPROTO_IPIP && t->parms.proto != 0) || + !ip6_tnl_xmit_ctl(t)) return -1; if (!(t->parms.flags & IP6_TNL_F_IGN_ENCAP_LIMIT)) @@ -978,7 +990,8 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) __u32 mtu; int err; - if (!ip6_tnl_xmit_ctl(t) || ip6_tnl_addr_conflict(t, ipv6h)) + if ((t->parms.proto != IPPROTO_IPV6 && t->parms.proto != 0) || + !ip6_tnl_xmit_ctl(t) || ip6_tnl_addr_conflict(t, ipv6h)) return -1; if ((offset = parse_tlv_tnl_enc_lim(skb, skb->nh.raw)) > 0) { @@ -1140,6 +1153,7 @@ ip6_tnl_change(struct ip6_tnl *t, struct ip6_tnl_parm *p) t->parms.encap_limit = p->encap_limit; t->parms.flowinfo = p->flowinfo; t->parms.link = p->link; + t->parms.proto = p->proto; ip6_tnl_dst_reset(t); ip6_tnl_link_config(t); return 0; @@ -1205,7 +1219,8 @@ ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof (p))) break; err = -EINVAL; - if (p.proto != IPPROTO_IPV6) + if (p.proto != IPPROTO_IPV6 && p.proto != IPPROTO_IPIP && + p.proto != 0) break; t = ip6_tnl_locate(&p, cmd == SIOCADDTUNNEL); if (dev != ip6_fb_tnl_dev && cmd == SIOCCHGTUNNEL) { @@ -1324,7 +1339,6 @@ static inline void ip6_tnl_dev_init_gen(struct net_device *dev) { struct ip6_tnl *t = netdev_priv(dev); - t->fl.proto = IPPROTO_IPV6; t->dev = dev; strcpy(t->parms.name, dev->name); } @@ -1355,6 +1369,7 @@ ip6_fb_tnl_dev_init(struct net_device *dev) { struct ip6_tnl *t = netdev_priv(dev); ip6_tnl_dev_init_gen(dev); + t->parms.proto = IPPROTO_IPV6; dev_hold(dev); tnls_wc[0] = t; return 0; --- commit bd292ad1ccbf15c72351f79d4a130dd50da81613 Author: YOSHIFUJI Hideaki <[EMAIL PROTECTED]> Date: Thu Feb 15 02:07:27 2007 +0900 [IPV6] ADDRCONF: Register inet6_dev earlier. Allocate inet6_dev earlier to allow users to set up per-interface variables. Signed-off-by: YOSHIFUJI Hideaki <[EMAIL PROTECTED]> diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c index c6683ea..42b5cc5 100644 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@ -2246,6 +2246,14 @@ static int addrconf_notify(struct notifier_block *this, unsigned long event, int run_pending = 0; switch(event) { + case NETDEV_REGISTER: + if (!idev) { + idev = ipv6_add_dev(dev); + if (!idev) + printk(KERN_WARNING "IPv6: add_dev failed for %s\n", + dev->name); + } + break; case NETDEV_UP: case NETDEV_CHANGE: if (event == NETDEV_UP) { --- -- YOSHIFUJI Hideaki @ USAGI Project <[EMAIL PROTECTED]> GPG-FP : 9022 65EB 1ECF 3AD1 0BDF 80D8 4807 F894 E062 0EEA - To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html