diff -Nurp linux-2.6.22-592/drivers/net/gre.c linux-2.6.22-593/drivers/net/gre.c --- linux-2.6.22-592/drivers/net/gre.c 1969-12-31 19:00:00.000000000 -0500 +++ linux-2.6.22-593/drivers/net/gre.c 2008-03-18 16:10:13.000000000 -0400 @@ -0,0 +1,1632 @@ +/* + * Linux NET3: GRE over IP protocol decoder. + * + * Authors: Alexey Kuznetsov (kuznet@ms2.inr.ac.ru) + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include /**XXX added XXX */ +#include +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#ifdef CONFIG_IPV6 +#include +#include +#include +#endif + +//#define GRE_DEBUG 1 + +/* + Problems & solutions + -------------------- + + 1. The most important issue is detecting local dead loops. + They would cause complete host lockup in transmit, which + would be "resolved" by stack overflow or, if queueing is enabled, + with infinite looping in net_bh. + + We cannot track such dead loops during route installation, + it is infeasible task. The most general solutions would be + to keep skb->encapsulation counter (sort of local ttl), + and silently drop packet when it expires. It is the best + solution, but it supposes maintaing new variable in ALL + skb, even if no tunneling is used. + + Current solution: t->recursion lock breaks dead loops. It looks + like dev->tbusy flag, but I preferred new variable, because + the semantics is different. One day, when hard_start_xmit + will be multithreaded we will have to use skb->encapsulation. + + + + 2. Networking dead loops would not kill routers, but would really + kill network. IP hop limit plays role of "t->recursion" in this case, + if we copy it from packet being encapsulated to upper header. + It is very good solution, but it introduces two problems: + + - Routing protocols, using packets with ttl=1 (OSPF, RIP2), + do not work over tunnels. + - traceroute does not work. I planned to relay ICMP from tunnel, + so that this problem would be solved and traceroute output + would even more informative. This idea appeared to be wrong: + only Linux complies to rfc1812 now (yes, guys, Linux is the only + true router now :-)), all routers (at least, in neighbourhood of mine) + return only 8 bytes of payload. It is the end. + + Hence, if we want that OSPF worked or traceroute said something reasonable, + we should search for another solution. + + One of them is to parse packet trying to detect inner encapsulation + made by our node. It is difficult or even impossible, especially, + taking into account fragmentation. TO be short, tt is not solution at all. + + Current solution: The solution was UNEXPECTEDLY SIMPLE. + We force DF flag on tunnels with preconfigured hop limit, + that is ALL. :-) Well, it does not remove the problem completely, + but exponential growth of network traffic is changed to linear + (branches, that exceed pmtu are pruned) and tunnel mtu + fastly degrades to value <68, where looping stops. + Yes, it is not good if there exists a router in the loop, + which does not force DF, even when encapsulating packets have DF set. + But it is not our problem! Nobody could accuse us, we made + all that we could make. Even if it is your gated who injected + fatal route to network, even if it were you who configured + fatal static route: you are innocent. :-) + + + + 3. Really, ipv4/ipip.c, ipv4/ip_gre.c and ipv6/sit.c contain + practically identical code. It would be good to glue them + together, but it is not very evident, how to make them modular. + sit is integral part of IPv6, ipip and gre are naturally modular. + We could extract common parts (hash table, ioctl etc) + to a separate module (ip_tunnel.c). + + Alexey Kuznetsov. + */ + +static int ipgre_tunnel_init(struct net_device *dev); +static void ipgre_ip_tunnel_setup(struct net_device *dev); +static void ipgre_eth_tunnel_setup(struct net_device *dev); + +/* Fallback tunnel: no source, no destination, no key, no options */ + +static int ipgre_fb_tunnel_init(struct net_device *dev); + +static struct net_device *ipgre_fb_tunnel_dev; + +/* Tunnel hash table */ + +/* + 4 hash tables: + + 3: (remote,local) + 2: (remote,*) + 1: (*,local) + 0: (*,*) + + We require exact key match i.e. if a key is present in packet + it will match only tunnel with the same key; if it is not present, + it will match only keyless tunnel. + + All keysless packets, if not matched configured keyless tunnels + will match fallback tunnel. + */ + +#define HASH_SIZE 1024 +#define HASH(addr) (ntohl(addr)&1023) + +static struct ip_tunnel *tunnels[4][HASH_SIZE]; + +#define tunnels_r_l (tunnels[3]) +#define tunnels_r (tunnels[2]) +#define tunnels_l (tunnels[1]) +#define tunnels_wc (tunnels[0]) + +static DEFINE_RWLOCK(ipgre_lock); + +/* Given src, dst and key, find appropriate for input tunnel. */ + +static struct ip_tunnel * ipgre_tunnel_lookup(__be32 remote, __be32 local, __be32 key) +{ + /* HACK */ + unsigned hash_value = HASH(key); + struct ip_tunnel *t; + + t = tunnels_r_l[hash_value]; + + if (t && (t->parms.i_key == key) && (t->dev->flags&IFF_UP)) { + return t; + } + + t = tunnels_r[hash_value]; + if (t && (t->parms.i_key == key) && (t->dev->flags&IFF_UP)) + return t; + + t = tunnels_l[hash_value]; + if (t && (t->parms.i_key == key) && (t->dev->flags&IFF_UP)) + return t; + t = tunnels_wc[hash_value]; + if (t && (t->parms.i_key == key) && (t->dev->flags&IFF_UP)) + return t; + if (ipgre_fb_tunnel_dev->flags&IFF_UP) + return netdev_priv(ipgre_fb_tunnel_dev); + return NULL; +} + +static struct ip_tunnel **ipgre_bucket(struct ip_tunnel *t) +{ + __be32 remote = t->parms.iph.daddr; + __be32 local = t->parms.iph.saddr; + __be32 key = t->parms.i_key; + unsigned h = HASH(key); + int prio = 0; + + if (local) + prio |= 1; + if (remote && !MULTICAST(remote)) { + prio |= 2; + //h ^= HASH(remote); + } + + return &tunnels[prio][h]; +} + +static void ipgre_tunnel_link(struct ip_tunnel *t) +{ + struct ip_tunnel **tp = ipgre_bucket(t); + + t->next = *tp; + write_lock_bh(&ipgre_lock); + *tp = t; + write_unlock_bh(&ipgre_lock); +} + +static void ipgre_tunnel_unlink(struct ip_tunnel *t) +{ + struct ip_tunnel **tp; + + for (tp = ipgre_bucket(t); *tp; tp = &(*tp)->next) { + if (t == *tp) { + write_lock_bh(&ipgre_lock); + *tp = t->next; + write_unlock_bh(&ipgre_lock); + break; + } + } +} + +static struct ip_tunnel * ipgre_tunnel_locate(struct ip_tunnel_parm *parms, int create) +{ + __be32 remote = parms->iph.daddr; + __be32 local = parms->iph.saddr; + __be32 key = parms->i_key; + __be16 proto = parms->proto_type; + struct ip_tunnel *t, **tp, *nt; + struct net_device *dev; + unsigned h = HASH(key); + int prio = 0; + char name[IFNAMSIZ]; + + if (local) + prio |= 1; + if (remote && !MULTICAST(remote)) { + prio |= 2; + //h ^= HASH(remote); + } + for (tp = &tunnels[prio][h]; (t = *tp) != NULL; tp = &t->next) { + if (local == t->parms.iph.saddr && remote == t->parms.iph.daddr) { + if (key == t->parms.i_key) + return t; + } + } + if (!create) + return NULL; + + printk(KERN_CRIT "Adding tunnel %s with key %d\n", parms->name, ntohl(key)); + + if (parms->name[0]) + strlcpy(name, parms->name, IFNAMSIZ); + else { + int i; + for (i=1; i<100; i++) { + sprintf(name, "gre%d", i); + if (__dev_get_by_name(&init_net, name) == NULL) + break; + } + if (i==100) + goto failed; + } + + /* Tunnel creation: check payload type and call appropriate + * function */ + switch (proto) + { + case ETH_P_IP: + dev = alloc_netdev(sizeof(*t), name, ipgre_ip_tunnel_setup); + break; + case ETH_P_ETH: + dev = alloc_netdev(sizeof(*t), name, ipgre_eth_tunnel_setup); + break; + default: + return NULL; + } + + if (!dev) + return NULL; + + dev->init = ipgre_tunnel_init; + nt = netdev_priv(dev); + nt->parms = *parms; + + if (register_netdevice(dev) < 0) { + free_netdev(dev); + goto failed; + } + + dev_hold(dev); + ipgre_tunnel_link(nt); + return nt; + +failed: + return NULL; +} + +static void ipgre_tunnel_uninit(struct net_device *dev) +{ + ipgre_tunnel_unlink(netdev_priv(dev)); + dev_put(dev); +} + + +static void ipgre_err(struct sk_buff *skb, u32 info) +{ +#ifndef I_WISH_WORLD_WERE_PERFECT + +/* It is not :-( All the routers (except for Linux) return only + 8 bytes of packet payload. It means, that precise relaying of + ICMP in the real Internet is absolutely infeasible. + + Moreover, Cisco "wise men" put GRE key to the third word + in GRE header. It makes impossible maintaining even soft state for keyed + GRE tunnels with enabled checksum. Tell them "thank you". + + Well, I wonder, rfc1812 was written by Cisco employee, + what the hell these idiots break standrads established + by themself??? + */ + + struct iphdr *iph = (struct iphdr*)skb->data; + __be16 *p = (__be16*)(skb->data+(iph->ihl<<2)); + int grehlen = (iph->ihl<<2) + 4; + int type = icmp_hdr(skb)->type; + int code = icmp_hdr(skb)->code; + struct ip_tunnel *t; + __be16 flags; + + flags = p[0]; + if (flags&(GRE_CSUM|GRE_KEY|GRE_SEQ|GRE_ROUTING|GRE_VERSION)) { + if (flags&(GRE_VERSION|GRE_ROUTING)) + return; + if (flags&GRE_KEY) { + grehlen += 4; + if (flags&GRE_CSUM) + grehlen += 4; + } + } + + /* If only 8 bytes returned, keyed message will be dropped here */ + if (skb_headlen(skb) < grehlen) + return; + + switch (type) { + default: + case ICMP_PARAMETERPROB: + return; + + case ICMP_DEST_UNREACH: + switch (code) { + case ICMP_SR_FAILED: + case ICMP_PORT_UNREACH: + /* Impossible event. */ + return; + case ICMP_FRAG_NEEDED: + /* Soft state for pmtu is maintained by IP core. */ + return; + default: + /* All others are translated to HOST_UNREACH. + rfc2003 contains "deep thoughts" about NET_UNREACH, + I believe they are just ether pollution. --ANK + */ + break; + } + break; + case ICMP_TIME_EXCEEDED: + if (code != ICMP_EXC_TTL) + return; + break; + } + + read_lock(&ipgre_lock); + t = ipgre_tunnel_lookup(iph->daddr, iph->saddr, (flags&GRE_KEY) ? *(((__be32*)p) + (grehlen>>2) - 1) : 0); + if (t == NULL || t->parms.iph.daddr == 0 || MULTICAST(t->parms.iph.daddr)) + goto out; + + if (t->parms.iph.ttl == 0 && type == ICMP_TIME_EXCEEDED) + goto out; + + if (jiffies - t->err_time < IPTUNNEL_ERR_TIMEO) + t->err_count++; + else + t->err_count = 1; + t->err_time = jiffies; +out: + read_unlock(&ipgre_lock); + return; +#else + struct iphdr *iph = (struct iphdr*)dp; + struct iphdr *eiph; + __be16 *p = (__be16*)(dp+(iph->ihl<<2)); + int type = skb->h.icmph->type; + int code = skb->h.icmph->code; + int rel_type = 0; + int rel_code = 0; + __be32 rel_info = 0; + __u32 n = 0; + __be16 flags; + int grehlen = (iph->ihl<<2) + 4; + struct sk_buff *skb2; + struct flowi fl; + struct rtable *rt; + + if (skb->dev->nd_net != &init_net) + return; + + if (p[1] != htons(ETH_P_IP)) + return; + + flags = p[0]; + if (flags&(GRE_CSUM|GRE_KEY|GRE_SEQ|GRE_ROUTING|GRE_VERSION)) { + if (flags&(GRE_VERSION|GRE_ROUTING)) + return; + if (flags&GRE_CSUM) + grehlen += 4; + if (flags&GRE_KEY) + grehlen += 4; + if (flags&GRE_SEQ) + grehlen += 4; + } + if (len < grehlen + sizeof(struct iphdr)) + return; + eiph = (struct iphdr*)(dp + grehlen); + + switch (type) { + default: + return; + case ICMP_PARAMETERPROB: + n = ntohl(skb->h.icmph->un.gateway) >> 24; + if (n < (iph->ihl<<2)) + return; + + /* So... This guy found something strange INSIDE encapsulated + packet. Well, he is fool, but what can we do ? + */ + rel_type = ICMP_PARAMETERPROB; + n -= grehlen; + rel_info = htonl(n << 24); + break; + + case ICMP_DEST_UNREACH: + switch (code) { + case ICMP_SR_FAILED: + case ICMP_PORT_UNREACH: + /* Impossible event. */ + return; + case ICMP_FRAG_NEEDED: + /* And it is the only really necessary thing :-) */ + n = ntohs(skb->h.icmph->un.frag.mtu); + if (n < grehlen+68) + return; + n -= grehlen; + /* BSD 4.2 MORE DOES NOT EXIST IN NATURE. */ + if (n > ntohs(eiph->tot_len)) + return; + rel_info = htonl(n); + break; + default: + /* All others are translated to HOST_UNREACH. + rfc2003 contains "deep thoughts" about NET_UNREACH, + I believe, it is just ether pollution. --ANK + */ + rel_type = ICMP_DEST_UNREACH; + rel_code = ICMP_HOST_UNREACH; + break; + } + break; + case ICMP_TIME_EXCEEDED: + if (code != ICMP_EXC_TTL) + return; + break; + } + + /* Prepare fake skb to feed it to icmp_send */ + skb2 = skb_clone(skb, GFP_ATOMIC); + if (skb2 == NULL) + return; + dst_release(skb2->dst); + skb2->dst = NULL; + skb_pull(skb2, skb->data - (u8*)eiph); + skb_reset_network_header(skb2); + + /* Try to guess incoming interface */ + memset(&fl, 0, sizeof(fl)); --- trellis20/include/linux/if_tunnel.h.orig 2009-02-26 01:10:20.000000000 -0500 +++ trellis20/include/linux/if_tunnel.h 2009-02-26 00:19:32.000000000 -0500 @@ -29,6 +29,7 @@ __be16 o_flags; __be32 i_key; __be32 o_key; + __be16 proto_type; /*Added*/ struct iphdr iph; }; --- trellis20/include/linux/if_ether.h.orig 2009-02-26 01:10:06.000000000 -0500 +++ trellis20/include/linux/if_ether.h 2009-02-26 00:19:32.000000000 -0500 @@ -56,6 +56,7 @@ #define ETH_P_DIAG 0x6005 /* DEC Diagnostics */ #define ETH_P_CUST 0x6006 /* DEC Customer use */ #define ETH_P_SCA 0x6007 /* DEC Systems Comms Arch */ +#define ETH_P_ETH 0x6558 /* Ethernet in Ethernet */ #define ETH_P_RARP 0x8035 /* Reverse Addr Res packet */ #define ETH_P_ATALK 0x809B /* Appletalk DDP */ #define ETH_P_AARP 0x80F3 /* Appletalk AARP */ --- trellis20/net/ipv4/ip_gre.c.orig 2009-02-26 01:09:40.000000000 -0500 +++ trellis20/net/ipv4/ip_gre.c 2009-02-26 00:54:20.000000000 -0500 @@ -25,6 +25,7 @@ #include #include #include +#include /**XXX added XXX */ #include #include #include @@ -48,6 +49,8 @@ #include #endif +//#define GRE_DEBUG 1 + /* Problems & solutions -------------------- @@ -118,7 +121,8 @@ */ static int ipgre_tunnel_init(struct net_device *dev); -static void ipgre_tunnel_setup(struct net_device *dev); +static void ipgre_ip_tunnel_setup(struct net_device *dev); +static void ipgre_eth_tunnel_setup(struct net_device *dev); /* Fallback tunnel: no source, no destination, no key, no options */ @@ -255,6 +259,7 @@ __be32 remote = parms->iph.daddr; __be32 local = parms->iph.saddr; __be32 key = parms->i_key; + __be16 proto = parms->proto_type; struct ip_tunnel *t, **tp, *nt; struct net_device *dev; char name[IFNAMSIZ]; @@ -269,12 +274,29 @@ if (!create) return NULL; + printk(KERN_CRIT "Adding tunnel %s with key %d\n", parms->name, ntohl(key)); + if (parms->name[0]) strlcpy(name, parms->name, IFNAMSIZ); else sprintf(name, "gre%%d"); dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup); + + /* Tunnel creation: check payload type and call appropriate + * function */ + switch (proto) + { + case ETH_P_IP: + dev = alloc_netdev(sizeof(*t), name, ipgre_ip_tunnel_setup); + break; + case ETH_P_ETH: + dev = alloc_netdev(sizeof(*t), name, ipgre_eth_tunnel_setup); + break; + default: + return NULL; + } + if (!dev) return NULL; @@ -431,6 +453,7 @@ u32 seqno = 0; struct ip_tunnel *tunnel; int offset = 4; + __be16 proto; if (!pskb_may_pull(skb, 16)) goto drop_nolock; @@ -439,6 +462,11 @@ h = skb->data; flags = *(__be16*)h; +#ifdef GRE_DEBUG + printk(KERN_DEBUG "gre.c [601] src:%x dst:%x proto:%d %p", iph->saddr, iph->daddr, iph->protocol, skb->data); +#endif + proto = ntohs(*(__be16*)(h+2)); /* XXX added XXX */ + if (flags&(GRE_CSUM|GRE_KEY|GRE_ROUTING|GRE_SEQ|GRE_VERSION)) { /* - Version must be 0. - We do not support routing headers. @@ -493,7 +521,28 @@ __pskb_pull(skb, offset); skb_reset_network_header(skb); skb_postpull_rcsum(skb, skb_transport_header(skb), offset); - skb->pkt_type = PACKET_HOST; + if(proto == ETH_P_ETH) + { +#ifdef GRE_DEBUG + unsigned char* tmp_hdr = skb->data; + printk(KERN_DEBUG "gre.c [658] %x %x %x %x %x %x\tskb %p\n", tmp_hdr[0], tmp_hdr[1], tmp_hdr[2], tmp_hdr[3], tmp_hdr[4], tmp_hdr[5], skb->data); +#endif + skb->protocol = eth_type_trans(skb, tunnel->dev); + + /* XXX added these lines to make arp work? XXX */ + /*skb->mac.raw = skb->data;*/ + skb->network_header = skb->network_header + ETH_HLEN; + /* XXX added these lines to make arp work? XXX */ + +#ifdef GRE_DEBUG + tmp_hdr = skb->data; + printk(KERN_DEBUG "gre.c [669] %x %x %x %x %x %x\tskb %p\n", tmp_hdr[0], tmp_hdr[1], tmp_hdr[2], tmp_hdr[3], tmp_hdr[4], tmp_hdr[5], skb->data); + printk(KERN_ALERT "gre.c [671] received ethernet on gre %x\n",skb->protocol); +#endif + memset(&(IPCB(skb)->opt), 0, sizeof(struct ip_options)); + } + else + skb->pkt_type = PACKET_HOST; #ifdef CONFIG_NET_IPGRE_BROADCAST if (ipv4_is_multicast(iph->daddr)) { /* Looped back packet, drop it! */ @@ -539,7 +588,7 @@ return(0); } -static int ipgre_tunnel_xmit(struct sk_buff *skb, struct net_device *dev) +static int ipgre_ip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev) { struct ip_tunnel *tunnel = netdev_priv(dev); struct net_device_stats *stats = &tunnel->dev->stats; @@ -799,7 +848,15 @@ tdev = rt->u.dst.dev; ip_rt_put(rt); } - dev->flags |= IFF_POINTOPOINT; + if (tunnel->parms.proto_type == ETH_P_ETH) + { + dev->flags |= IFF_BROADCAST; + } + else + { + dev->flags |= IFF_POINTOPOINT; + } + } if (!tdev && tunnel->parms.link) @@ -822,10 +879,235 @@ } dev->hard_header_len = hlen + addend; dev->mtu = mtu - addend; + if (tunnel->parms.proto_type == ETH_P_ETH) + dev->mtu -= ETH_HLEN; + tunnel->hlen = addend; } +static int ipgre_eth_tunnel_xmit(struct sk_buff *skb, struct net_device *dev) +{ + struct ip_tunnel *tunnel = netdev_priv(dev); + struct net_device_stats *stats = &tunnel->stat; + struct iphdr *old_iph = ip_hdr(skb); + struct iphdr *tiph = &tunnel->parms.iph; + u8 tos; + __be16 df; + struct rtable *rt; /* Route to the other host */ + struct net_device *tdev; /* Device to other host */ + int gre_hlen = tunnel->hlen; /* XXX changed XXX*/ + //struct etheriphdr *ethiph; + struct iphdr *iph; /* Our new IP header */ + int max_headroom; /* The extra header space needed */ + int mtu; + +#ifdef GRE_DEBUG + printk(KERN_ALERT "gre.c:972 Starting xmit\n"); +#endif + + if (tunnel->recursion++) { + stats->collisions++; + goto tx_error; + } + + /* Need valid non-multicast daddr. */ + if (tiph->daddr == 0 || MULTICAST(tiph->daddr)) + goto tx_error; + + tos = tiph->tos; + if (tos&1) { + if (skb->protocol == htons(ETH_P_IP)) + tos = old_iph->tos; + tos &= ~1; + } +#ifdef GRE_DEBUG + printk(KERN_ALERT "gre.c:991 Passed tos assignment.\n"); +#endif + + + { + struct flowi fl = { .fl_net = &init_net, + .oif = tunnel->parms.link, + .nl_u = { .ip4_u = + { .daddr = tiph->daddr, + .saddr = tiph->saddr, + .tos = RT_TOS(tos) } }, + .proto = IPPROTO_GRE }; + if (ip_route_output_key(&rt, &fl)) { + stats->tx_carrier_errors++; + goto tx_error_icmp; + } + } + tdev = rt->u.dst.dev; +#ifdef GRE_DEBUG + printk(KERN_ALERT "gre.c:1006 Passed the route retrieval\n"); +#endif + if (tdev == dev) { + ip_rt_put(rt); + stats->collisions++; + goto tx_error; + } +#ifdef GRE_DEBUG + printk(KERN_ALERT "gre.c:1018 Passed tdev collision check.\n"); +#endif + + /* Check MTU stuff if kernel panic */ + df = tiph->frag_off; + if (df) + mtu = dst_mtu(&rt->u.dst) - tunnel->hlen; + else + mtu = skb->dst ? dst_mtu(skb->dst) : dev->mtu; +/* + if (skb->dst) + skb->dst->ops->update_pmtu(skb->dst, mtu); + XXX */ +#ifdef GRE_DEBUG + printk(KERN_ALERT "gre.c:1032 Passed the pmtu setting.\n"); +#endif + + if (skb->protocol == htons(ETH_P_IP)) { + df |= (old_iph->frag_off&htons(IP_DF)); + + if ((old_iph->frag_off & htons(IP_DF)) && + mtu < ntohs(old_iph->tot_len)) { + icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu)); + ip_rt_put(rt); + goto tx_error; + } + } +#ifdef CONFIG_IPV6 + else if (skb->protocol == htons(ETH_P_IPV6)) { + struct rt6_info *rt6 = (struct rt6_info*)skb->dst; + + if (rt6 && mtu < dst_mtu(skb->dst) && mtu >= IPV6_MIN_MTU) { + if (tiph->daddr || rt6->rt6i_dst.plen == 128) { + rt6->rt6i_flags |= RTF_MODIFIED; + skb->dst->metrics[RTAX_MTU-1] = mtu; + } + } + + /* @@@ Is this correct? */ + if (mtu >= IPV6_MIN_MTU && mtu < skb->len - tunnel->hlen + gre_hlen) { + icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, dev); + ip_rt_put(rt); + goto tx_error; + } + } +#endif +#ifdef GRE_DEBUG + printk(KERN_ALERT "gre.c:1065 Passed the fragmentation check.\n"); +#endif + + if (tunnel->err_count > 0) { + if (jiffies - tunnel->err_time < IPTUNNEL_ERR_TIMEO) { + tunnel->err_count--; + dst_link_failure(skb); + } else + tunnel->err_count = 0; + } + + max_headroom = LL_RESERVED_SPACE(tdev) + gre_hlen; + + if (skb_headroom(skb) < max_headroom || skb_cloned(skb) || skb_shared(skb)) { + struct sk_buff *new_skb = skb_realloc_headroom(skb, max_headroom); + if (!new_skb) { + ip_rt_put(rt); + stats->tx_dropped++; + dev_kfree_skb(skb); + tunnel->recursion--; + return 0; + } + if (skb->sk) + skb_set_owner_w(new_skb, skb->sk); + dev_kfree_skb(skb); + skb = new_skb; + old_iph = ip_hdr(skb); + } +#ifdef GRE_DEBUG + printk(KERN_ALERT "gre.c:1094 Passed the headroom calculation\n"); +#endif + + skb->transport_header = skb->data; + skb_push(skb, gre_hlen); + skb_reset_network_header(skb); + memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt)); + dst_release(skb->dst); + skb->dst = &rt->u.dst; + + /* + * Push down and install the etherip header. + */ + + iph = ip_hdr(skb); + iph->version = 4; + iph->ihl = sizeof(struct iphdr) >> 2; + iph->frag_off = df; + iph->protocol = IPPROTO_GRE; + iph->tos = ipgre_ecn_encapsulate(tos, old_iph, skb); + iph->daddr = rt->rt_dst; + iph->saddr = rt->rt_src; + +/* ethiph->version = htons(ETHERIP_VERSION); */ +#ifdef GRE_DEBUG + printk(KERN_ALERT "gre.c:1121 Passed outer IP header construction.\n"); +#endif + + if ((iph->ttl = tiph->ttl) == 0) { + if (skb->protocol == htons(ETH_P_IP)) + iph->ttl = old_iph->ttl; +#ifdef CONFIG_IPV6 + else if (skb->protocol == htons(ETH_P_IPV6)) + iph->ttl = ((struct ipv6hdr*)old_iph)->hop_limit; +#endif + else + iph->ttl = dst_metric(&rt->u.dst, RTAX_HOPLIMIT); + } +#ifdef GRE_DEBUG + printk(KERN_ALERT "gre.c:1006 Passed the TTL check.\n"); +#endif + + ((__be16*)(iph+1))[0] = tunnel->parms.o_flags; + ((__be16*)(iph+1))[1] = htons(tunnel->parms.proto_type); + + if (tunnel->parms.o_flags&(GRE_KEY|GRE_CSUM|GRE_SEQ)) { + __be32 *ptr = (__be32*)(((u8*)iph) + tunnel->hlen - 4); + + if (tunnel->parms.o_flags&GRE_SEQ) { + ++tunnel->o_seqno; + *ptr = htonl(tunnel->o_seqno); + ptr--; + } + if (tunnel->parms.o_flags&GRE_KEY) { + *ptr = tunnel->parms.o_key; + ptr--; + } + if (tunnel->parms.o_flags&GRE_CSUM) { + *ptr = 0; + *(__sum16*)ptr = ip_compute_csum((void*)(iph+1), skb->len - sizeof(struct iphdr)); + } + } +#ifdef GRE_DEBUG + printk(KERN_ALERT "gre.c:1006 Passed the tunnel transmit.\n"); +#endif + + nf_reset(skb); + + IPTUNNEL_XMIT(); + tunnel->recursion--; + return 0; + +tx_error_icmp: + dst_link_failure(skb); + +tx_error: + stats->tx_errors++; + dev_kfree_skb(skb); + tunnel->recursion--; + return 0; +} + + static int ipgre_tunnel_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd) { @@ -835,6 +1117,8 @@ struct net *net = dev_net(dev); struct ipgre_net *ign = net_generic(net, ipgre_net_id); + printk(KERN_ALERT "1174 GRE: entering gre ioctl. command is: %d\n", cmd); + switch (cmd) { case SIOCGETTUNNEL: t = NULL; @@ -876,6 +1160,7 @@ p.o_key = 0; t = ipgre_tunnel_locate(net, &p, cmd == SIOCADDTUNNEL); + if (t) printk(KERN_ALERT "1174 GRE: proto %s %x\n", p.name, p.proto_type); if (dev != ign->fb_tunnel_dev && cmd == SIOCCHGTUNNEL) { if (t != NULL) { @@ -892,6 +1177,12 @@ nflags = IFF_BROADCAST; else if (p.iph.daddr) nflags = IFF_POINTOPOINT; + + /* XXX:Set back IFF_BROADCAST if + * transporting ethernet */ + printk(KERN_ALERT "1193 GRE: proto %s %d\n", p.name, p.proto_type); + if (p.proto_type == ETH_P_ETH) + nflags = IFF_BROADCAST; if ((dev->flags^nflags)&(IFF_POINTOPOINT|IFF_BROADCAST)) { err = -EINVAL; @@ -902,8 +1193,13 @@ t->parms.iph.daddr = p.iph.daddr; t->parms.i_key = p.i_key; t->parms.o_key = p.o_key; - memcpy(dev->dev_addr, &p.iph.saddr, 4); - memcpy(dev->broadcast, &p.iph.daddr, 4); + /* XXX:Copy in the protocol field */ + t->parms.proto_type = p.proto_type; + if (t->parms.proto_type != ETH_P_ETH) { + memcpy(dev->dev_addr, &p.iph.saddr, 4); + memcpy(dev->broadcast, &p.iph.daddr, 4); + } + ipgre_tunnel_link(ign, t); netdev_state_change(dev); } @@ -1076,11 +1372,11 @@ #endif -static void ipgre_tunnel_setup(struct net_device *dev) +static void ipgre_ip_tunnel_setup(struct net_device *dev) { dev->uninit = ipgre_tunnel_uninit; dev->destructor = free_netdev; - dev->hard_start_xmit = ipgre_tunnel_xmit; + dev->hard_start_xmit = ipgre_ip_tunnel_xmit; dev->do_ioctl = ipgre_tunnel_ioctl; dev->change_mtu = ipgre_tunnel_change_mtu; @@ -1093,6 +1389,35 @@ dev->features |= NETIF_F_NETNS_LOCAL; } +/* Tunnel setup for ipgre_eth */ +static void ipgre_eth_tunnel_setup(struct net_device *dev) +{ + SET_MODULE_OWNER(dev); + + // Set default values for Ethernet device + ether_setup(dev); + + dev->uninit = ipgre_tunnel_uninit; + dev->destructor = free_netdev; + dev->hard_start_xmit = ipgre_eth_tunnel_xmit; + dev->get_stats = ipgre_tunnel_get_stats; + dev->do_ioctl = ipgre_tunnel_ioctl; + dev->change_mtu = ipgre_tunnel_change_mtu; + + dev->hard_header_len = LL_MAX_HEADER + ETH_HLEN + sizeof(struct iphdr) + 4; + dev->mtu = ETH_DATA_LEN - ETH_HLEN - sizeof(struct iphdr) - 4; + dev->tx_queue_len = 0; + dev->iflink = 0; + + random_ether_addr(dev->dev_addr); + +#ifdef GRE_DEBUG + { unsigned char* d = dev->dev_addr; + printk(KERN_ALERT "Here is the address we got:%x%x%x%x%x%x\n",d[0],d[1],d[2],d[3],d[4],d[5]); } +#endif +} + + static int ipgre_tunnel_init(struct net_device *dev) { struct ip_tunnel *tunnel; @@ -1104,8 +1429,10 @@ tunnel->dev = dev; strcpy(tunnel->parms.name, dev->name); - memcpy(dev->dev_addr, &tunnel->parms.iph.saddr, 4); - memcpy(dev->broadcast, &tunnel->parms.iph.daddr, 4); + if (tunnel->parms.proto_type != ETH_P_ETH) { + memcpy(dev->dev_addr, &tunnel->parms.iph.saddr, 4); + memcpy(dev->broadcast, &tunnel->parms.iph.daddr, 4); + } ipgre_tunnel_bind_dev(dev); @@ -1181,7 +1508,7 @@ goto err_assign; ign->fb_tunnel_dev = alloc_netdev(sizeof(struct ip_tunnel), "gre0", - ipgre_tunnel_setup); + ipgre_ip_tunnel_setup); if (!ign->fb_tunnel_dev) { err = -ENOMEM; goto err_alloc_dev; --- trellis20/drivers/net/Makefile.orig 2009-02-26 01:10:33.000000000 -0500 +++ trellis20/drivers/net/Makefile 2009-02-26 00:20:33.000000000 -0500 @@ -2,6 +2,7 @@ # Makefile for the Linux network (ethercard) device drivers. # +obj-m +=gre.o obj-$(CONFIG_E1000) += e1000/ obj-$(CONFIG_E1000E) += e1000e/ obj-$(CONFIG_IBM_NEW_EMAC) += ibm_newemac/