Merge bk://kernel.bkbits.net/davem/net-2.5

into home.transmeta.com:/home/torvalds/v2.5/linux

include/linux/igmp.h

@@ -191,7 +191,7 @@ struct ip_mc_list
          (IGMPV3_MASK((value) >> (nbmant), nbexp) + (nbexp))))
 
 #define IGMPV3_QQIC(value) IGMPV3_EXP(0x80, 4, 3, value)
-#define IGMPV3_MRC(value) IGMPV3_EXP(0x8000, 12, 3, value)
+#define IGMPV3_MRC(value) IGMPV3_EXP(0x80, 4, 3, value)
 
 extern int ip_check_mc(struct in_device *dev, u32 mc_addr, u32 src_addr, u16 proto);
 extern int igmp_rcv(struct sk_buff *);
@@ -199,10 +199,12 @@ extern int ip_mc_join_group(struct sock *sk, struct ip_mreqn *imr);
 extern int ip_mc_leave_group(struct sock *sk, struct ip_mreqn *imr);
 extern void ip_mc_drop_socket(struct sock *sk);
 extern int ip_mc_source(int add, int omode, struct sock *sk,
-		struct ip_mreq_source *mreqs);
-extern int ip_mc_msfilter(struct sock *sk, struct ip_msfilter *msf);
+		struct ip_mreq_source *mreqs, int ifindex);
+extern int ip_mc_msfilter(struct sock *sk, struct ip_msfilter *msf,int ifindex);
 extern int ip_mc_msfget(struct sock *sk, struct ip_msfilter *msf,
 		struct ip_msfilter *optval, int *optlen);
+extern int ip_mc_gsfget(struct sock *sk, struct group_filter *gsf,
+		struct group_filter *optval, int *optlen);
 extern int ip_mc_sf_allow(struct sock *sk, u32 local, u32 rmt, int dif);
 extern void ip_mr_init(void);
 extern void ip_mc_init_dev(struct in_device *);

include/linux/net.h

@@ -140,6 +140,9 @@ struct net_proto_family {
 	struct module	*owner;
 };
 
+extern int	     net_family_get(int family);
+extern void	     net_family_put(int family);
+
 struct iovec;
 
 extern int	     sock_wake_async(struct socket *sk, int how, int band);

include/net/ip6_route.h

@@ -59,7 +59,7 @@ extern struct rt6_info		*rt6_lookup(struct in6_addr *daddr,
 					    struct in6_addr *saddr,
 					    int oif, int flags);
 
-extern struct rt6_info		*ndisc_get_dummy_rt(void);
+extern struct rt6_info		*ip6_dst_alloc(void);
 
 /*
  *	support functions for ND

net/core/dev.c

@@ -2582,6 +2582,16 @@ int register_netdevice(struct net_device *dev)
 	if ((ret = kobject_register(&dev->kobj)))
 		goto out_err;
 	
+	/* Fix illegal SG+CSUM combinations. */
+	if ((dev->features & NETIF_F_SG) &&
+	    !(dev->features & (NETIF_F_IP_CSUM |
+			       NETIF_F_NO_CSUM |
+			       NETIF_F_HW_CSUM))) {
+		printk("%s: Dropping NETIF_F_SG since no checksum feature.\n",
+		       dev->name);
+		dev->features &= ~NETIF_F_SG;
+	}
+
 	/*
 	 *	nil rebuild_header routine,
 	 *	that should be never called and used as just bug trap.

net/core/sock.c

@@ -589,8 +589,10 @@ static kmem_cache_t *sk_cachep;
  */
 struct sock *sk_alloc(int family, int priority, int zero_it, kmem_cache_t *slab)
 {
-	struct sock *sk;
+	struct sock *sk = NULL;
 
+	if (!net_family_get(family))
+		goto out;
 	if (!slab)
 		slab = sk_cachep;
 	sk = kmem_cache_alloc(slab, priority);
@@ -602,14 +604,16 @@ struct sock *sk_alloc(int family, int priority, int zero_it, kmem_cache_t *slab)
 			sock_lock_init(sk);
 		}
 		sk->slab = slab;
-	}
-
+	} else
+		net_family_put(family);
+out:
 	return sk;
 }
 
 void sk_free(struct sock *sk)
 {
 	struct sk_filter *filter;
+	const int family = sk->family;
 
 	if (sk->destruct)
 		sk->destruct(sk);
@@ -624,6 +628,7 @@ void sk_free(struct sock *sk)
 		printk(KERN_DEBUG "sk_free: optmem leakage (%d bytes) detected.\n", atomic_read(&sk->omem_alloc));
 
 	kmem_cache_free(sk->slab, sk);
+	net_family_put(family);
 }
 
 void __init sk_init(void)

net/ipv4/ip_output.c

@@ -1312,6 +1312,6 @@ void __init ip_init(void)
 
 #ifdef CONFIG_IP_MULTICAST
 	proc_net_create("igmp", 0, ip_mc_procinfo);
-	proc_net_create("mcfilter", 0, ip_mcf_procinfo);
 #endif
+	proc_net_create("mcfilter", 0, ip_mcf_procinfo);
 }

net/ipv4/ip_sockglue.c

@@ -631,7 +631,7 @@ int ip_setsockopt(struct sock *sk, int level, int optname, char *optval, int opt
 				kfree(msf);
 				break;
 			}
-			err = ip_mc_msfilter(sk, msf);
+			err = ip_mc_msfilter(sk, msf, 0);
 			kfree(msf);
 			break;
 		}
@@ -670,7 +670,142 @@ int ip_setsockopt(struct sock *sk, int level, int optname, char *optval, int opt
 				omode = MCAST_INCLUDE;
 				add = 0;
 			}
-			err = ip_mc_source(add, omode, sk, &mreqs);
+			err = ip_mc_source(add, omode, sk, &mreqs, 0);
+			break;
+		}
+		case MCAST_JOIN_GROUP:
+		case MCAST_LEAVE_GROUP: 
+		{
+			struct group_req greq;
+			struct sockaddr_in *psin;
+			struct ip_mreqn mreq;
+
+			if (optlen < sizeof(struct group_req))
+				goto e_inval;
+			err = -EFAULT;
+			if(copy_from_user(&greq, optval, sizeof(greq)))
+				break;
+			psin = (struct sockaddr_in *)&greq.gr_group;
+			if (psin->sin_family != AF_INET)
+				goto e_inval;
+			memset(&mreq, 0, sizeof(mreq));
+			mreq.imr_multiaddr = psin->sin_addr;
+			mreq.imr_ifindex = greq.gr_interface;
+
+			if (optname == MCAST_JOIN_GROUP)
+				err = ip_mc_join_group(sk, &mreq);
+			else
+				err = ip_mc_leave_group(sk, &mreq);
+			break;
+		}
+		case MCAST_JOIN_SOURCE_GROUP:
+		case MCAST_LEAVE_SOURCE_GROUP:
+		case MCAST_BLOCK_SOURCE:
+		case MCAST_UNBLOCK_SOURCE:
+		{
+			struct group_source_req greqs;
+			struct ip_mreq_source mreqs;
+			struct sockaddr_in *psin;
+			int omode, add;
+
+			if (optlen != sizeof(struct group_source_req))
+				goto e_inval;
+			if (copy_from_user(&greqs, optval, sizeof(greqs))) {
+				err = -EFAULT;
+				break;
+			}
+			if (greqs.gsr_group.ss_family != AF_INET ||
+			    greqs.gsr_source.ss_family != AF_INET) {
+				err = -EADDRNOTAVAIL;
+				break;
+			}
+			psin = (struct sockaddr_in *)&greqs.gsr_group;
+			mreqs.imr_multiaddr = psin->sin_addr.s_addr;
+			psin = (struct sockaddr_in *)&greqs.gsr_source;
+			mreqs.imr_sourceaddr = psin->sin_addr.s_addr;
+			mreqs.imr_interface = 0; /* use index for mc_source */
+
+			if (optname == MCAST_BLOCK_SOURCE) {
+				omode = MCAST_EXCLUDE;
+				add = 1;
+			} else if (optname == MCAST_UNBLOCK_SOURCE) {
+				omode = MCAST_EXCLUDE;
+				add = 0;
+			} else if (optname == MCAST_JOIN_SOURCE_GROUP) {
+				struct ip_mreqn mreq;
+
+				psin = (struct sockaddr_in *)&greqs.gsr_group;
+				mreq.imr_multiaddr = psin->sin_addr;
+				mreq.imr_address.s_addr = 0;
+				mreq.imr_ifindex = greqs.gsr_interface;
+				err = ip_mc_join_group(sk, &mreq);
+				if (err)
+					break;
+				omode = MCAST_INCLUDE;
+				add = 1;
+			} else /* MCAST_LEAVE_SOURCE_GROUP */ {
+				omode = MCAST_INCLUDE;
+				add = 0;
+			}
+			err = ip_mc_source(add, omode, sk, &mreqs,
+				greqs.gsr_interface);
+			break;
+		}
+		case MCAST_MSFILTER:
+		{
+			struct sockaddr_in *psin;
+			struct ip_msfilter *msf = 0;
+			struct group_filter *gsf = 0;
+			int msize, i, ifindex;
+
+			if (optlen < GROUP_FILTER_SIZE(0))
+				goto e_inval;
+			gsf = (struct group_filter *)kmalloc(optlen,GFP_KERNEL);
+			if (gsf == 0) {
+				err = -ENOBUFS;
+				break;
+			}
+			err = -EFAULT;
+			if (copy_from_user(gsf, optval, optlen)) {
+				goto mc_msf_out;
+			}
+			if (GROUP_FILTER_SIZE(gsf->gf_numsrc) < optlen) {
+				err = EINVAL;
+				goto mc_msf_out;
+			}
+			msize = IP_MSFILTER_SIZE(gsf->gf_numsrc);
+			msf = (struct ip_msfilter *)kmalloc(msize,GFP_KERNEL);
+			if (msf == 0) {
+				err = -ENOBUFS;
+				goto mc_msf_out;
+			}
+			ifindex = gsf->gf_interface;
+			psin = (struct sockaddr_in *)&gsf->gf_group;
+			if (psin->sin_family != AF_INET) {
+				err = -EADDRNOTAVAIL;
+				goto mc_msf_out;
+			}
+			msf->imsf_multiaddr = psin->sin_addr.s_addr;
+			msf->imsf_interface = 0;
+			msf->imsf_fmode = gsf->gf_fmode;
+			msf->imsf_numsrc = gsf->gf_numsrc;
+			err = -EADDRNOTAVAIL;
+			for (i=0; i<gsf->gf_numsrc; ++i) {
+				psin = (struct sockaddr_in *)&gsf->gf_slist[i];
+
+				if (psin->sin_family != AF_INET)
+					goto mc_msf_out;
+				msf->imsf_slist[i] = psin->sin_addr.s_addr;
+			}
+			kfree(gsf);
+			gsf = 0;
+
+			err = ip_mc_msfilter(sk, msf, ifindex);
+mc_msf_out:
+			if (msf)
+				kfree(msf);
+			if (gsf)
+				kfree(gsf);
 			break;
 		}
 		case IP_ROUTER_ALERT:	
@@ -826,15 +961,37 @@ int ip_getsockopt(struct sock *sk, int level, int optname, char *optval, int *op
 			struct ip_msfilter msf;
 			int err;
 
-			if (len < IP_MSFILTER_SIZE(0))
+			if (len < IP_MSFILTER_SIZE(0)) {
+				release_sock(sk);
 				return -EINVAL;
-			if (copy_from_user(&msf, optval, IP_MSFILTER_SIZE(0)))
+			}
+			if (copy_from_user(&msf, optval, IP_MSFILTER_SIZE(0))) {
+				release_sock(sk);
 				return -EFAULT;
+			}
 			err = ip_mc_msfget(sk, &msf,
 				(struct ip_msfilter *)optval, optlen);
 			release_sock(sk);
 			return err;
 		}
+		case MCAST_MSFILTER:
+		{
+			struct group_filter gsf;
+			int err;
+
+			if (len < GROUP_FILTER_SIZE(0)) {
+				release_sock(sk);
+				return -EINVAL;
+			}
+			if (copy_from_user(&gsf, optval, GROUP_FILTER_SIZE(0))) {
+				release_sock(sk);
+				return -EFAULT;
+			}
+			err = ip_mc_gsfget(sk, &gsf,
+				(struct group_filter *)optval, optlen);
+			release_sock(sk);
+			return err;
+		}
 		case IP_PKTOPTIONS:		
 		{
 			struct msghdr msg;

net/ipv4/netfilter/Kconfig

@@ -48,7 +48,7 @@ config IP_NF_IRC
 	  <file:Documentation/modules.txt>.  If unsure, say `Y'.
 
 config IP_NF_TFTP
-	tristate "TFTP prtocol support"
+	tristate "TFTP protocol support"
 	depends on IP_NF_CONNTRACK
 	help
 	  TFTP connection tracking helper, this is required depending

net/ipv6/ipv6_sockglue.c

@@ -120,6 +120,12 @@ int ip6_ra_control(struct sock *sk, int sel, void (*destructor)(struct sock *))
 	return 0;
 }
 
+extern int ip6_mc_source(int add, int omode, struct sock *sk,
+	struct group_source_req *pgsr);
+extern int ip6_mc_msfilter(struct sock *sk, struct group_filter *gsf);
+extern int ip6_mc_msfget(struct sock *sk, struct group_filter *gsf,
+	struct group_filter *optval, int *optlen);
+
 
 int ipv6_setsockopt(struct sock *sk, int level, int optname, char *optval, 
 		    int optlen)
@@ -393,12 +399,13 @@ int ipv6_setsockopt(struct sock *sk, int level, int optname, char *optval,
 			break;
 		}
 		psin6 = (struct sockaddr_in6 *)&greq.gr_group;
-		if (optname == IPV6_ADD_MEMBERSHIP)
+		if (optname == MCAST_JOIN_GROUP)
 			retv = ipv6_sock_mc_join(sk, greq.gr_interface,
 				&psin6->sin6_addr);
 		else
 			retv = ipv6_sock_mc_drop(sk, greq.gr_interface,
 				&psin6->sin6_addr);
+		break;
 	}
 	case MCAST_JOIN_SOURCE_GROUP:
 	case MCAST_LEAVE_SOURCE_GROUP:
@@ -414,7 +421,8 @@ int ipv6_setsockopt(struct sock *sk, int level, int optname, char *optval,
 			retv = -EFAULT;
 			break;
 		}
-		if (greqs.gsr_group.ss_family != AF_INET6) {
+		if (greqs.gsr_group.ss_family != AF_INET6 ||
+		    greqs.gsr_source.ss_family != AF_INET6) {
 			retv = -EADDRNOTAVAIL;
 			break;
 		}

net/ipv6/mcast.c

@@ -768,7 +768,7 @@ static void mld_clear_delrec(struct inet6_dev *idev)
 		psf = pmc->mca_tomb;
 		pmc->mca_tomb = 0;
 		spin_unlock_bh(&pmc->mca_lock);
-		for (psf=pmc->mca_tomb; psf; psf=psf_next) {
+		for (; psf; psf=psf_next) {
 			psf_next = psf->sf_next;
 			kfree(psf);
 		}
@@ -1042,6 +1042,8 @@ int igmp6_event_query(struct sk_buff *skb)
 		mld_clear_delrec(idev);
 	} else if (len >= 28) {
 		max_delay = MLDV2_MRC(ntohs(mlh2->mrc))*(HZ/10);
+		if (!max_delay)
+			max_delay = 1;
 		idev->mc_maxdelay = max_delay;
 		if (mlh2->qrv)
 			idev->mc_qrv = mlh2->qrv;
@@ -2096,8 +2098,6 @@ static int ip6_mcf_read_proc(char *buffer, char **start, off_t offset,
 			unsigned long icount, xcount, i;
 
 			spin_lock_bh(&imc->mca_lock);
-			icount = imc->mca_sfcount[MCAST_INCLUDE];
-			xcount = imc->mca_sfcount[MCAST_EXCLUDE];
 			for (psf=imc->mca_sources; psf; psf=psf->sf_next) {
 				if (first) {
 					len += sprintf(buffer+len, "%3s %6s "
@@ -2130,36 +2130,6 @@ static int ip6_mcf_read_proc(char *buffer, char **start, off_t offset,
 					in6_dev_put(idev);
 					goto done;
 				}
-				icount -= psf->sf_count[MCAST_INCLUDE];
-				xcount -= psf->sf_count[MCAST_EXCLUDE];
-			}
-			if (icount > 0 || xcount > 0) {
-				if (first) {
-					len += sprintf(buffer+len, "%3s %6s "
-						"%32s %32s %6s %6s\n", "Idx",
-						"Device", "Multicast Address",
-						"Source Address", "INC", "EXC");
-					first = 0;
-				}
-				len += sprintf(buffer+len,"%3d %6.6s ",
-					dev->ifindex, dev->name);
-
-				for (i=0; i<16; i++)
-					len += sprintf(buffer+len, "%02x",
-						imc->mca_addr.s6_addr[i]);
-				len += sprintf(buffer+len, " %32s %6lu %6lu\n",
-					"NONE", icount, xcount);
-				pos = begin+len;
-				if (pos < offset) {
-					len=0;
-					begin=pos;
-				}
-				if (pos > offset+length) {
-					spin_unlock_bh(&imc->mca_lock);
-					read_unlock_bh(&idev->lock);
-					in6_dev_put(idev);
-					goto done;
-				}
 			}
 			spin_unlock_bh(&imc->mca_lock);
 		}

net/ipv6/ndisc.c

@@ -427,7 +427,7 @@ static void ndisc_send_na(struct net_device *dev, struct neighbour *neigh,
 
 	len = sizeof(struct icmp6hdr) + sizeof(struct in6_addr);
 
-	rt = ndisc_get_dummy_rt();
+	rt = ip6_dst_alloc();
 	if (!rt) 
 		return;
 
@@ -524,7 +524,7 @@ void ndisc_send_ns(struct net_device *dev, struct neighbour *neigh,
 		saddr = &addr_buf;
 	}
 
-	rt = ndisc_get_dummy_rt();
+	rt = ip6_dst_alloc();
 	if (!rt) 
 		return;
 
@@ -595,7 +595,7 @@ void ndisc_send_rs(struct net_device *dev, struct in6_addr *saddr,
         int len;
 	int err;
 
-	rt = ndisc_get_dummy_rt();
+	rt = ip6_dst_alloc();
 	if (!rt) 
 		return;
 

net/ipv6/route.c

@@ -130,12 +130,17 @@ struct fib6_node ip6_routing_table = {
 rwlock_t rt6_lock = RW_LOCK_UNLOCKED;
 
 
-/*	Dummy rt for ndisc */
-struct rt6_info *ndisc_get_dummy_rt()
+/* allocate dst with ip6_dst_ops */
+static __inline__ struct rt6_info *__ip6_dst_alloc(void)
 {
 	return dst_alloc(&ip6_dst_ops);
 }
 
+struct rt6_info *ip6_dst_alloc(void)
+{
+	return __ip6_dst_alloc();
+}
+
 /*
  *	Route lookup. Any rt6_lock is implied.
  */
@@ -640,7 +645,7 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh)
 	if (rtmsg->rtmsg_metric == 0)
 		rtmsg->rtmsg_metric = IP6_RT_PRIO_USER;
 
-	rt = dst_alloc(&ip6_dst_ops);
+	rt = __ip6_dst_alloc();
 
 	if (rt == NULL)
 		return -ENOMEM;
@@ -1035,9 +1040,7 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
 
 static struct rt6_info * ip6_rt_copy(struct rt6_info *ort)
 {
-	struct rt6_info *rt;
-
-	rt = dst_alloc(&ip6_dst_ops);
+	struct rt6_info *rt = __ip6_dst_alloc();
 
 	if (rt) {
 		rt->u.dst.input = ort->u.dst.input;
@@ -1181,9 +1184,8 @@ int ip6_pkt_discard(struct sk_buff *skb)
 
 int ip6_rt_addr_add(struct in6_addr *addr, struct net_device *dev)
 {
-	struct rt6_info *rt;
+	struct rt6_info *rt = __ip6_dst_alloc();
 
-	rt = dst_alloc(&ip6_dst_ops);
 	if (rt == NULL)
 		return -ENOMEM;
 

net/netlink/af_netlink.c

@@ -1052,6 +1052,7 @@ struct proto_ops netlink_ops = {
 struct net_proto_family netlink_family_ops = {
 	.family = PF_NETLINK,
 	.create = netlink_create,
+	.owner	= THIS_MODULE,	/* for consistency 8) */
 };
 
 static int __init netlink_proto_init(void)
@@ -1065,6 +1066,11 @@ static int __init netlink_proto_init(void)
 	sock_register(&netlink_family_ops);
 #ifdef CONFIG_PROC_FS
 	create_proc_read_entry("net/netlink", 0, 0, netlink_read_proc, NULL);
+#endif
+	/* The netlink device handler may be needed early. */ 
+	rtnetlink_init();
+#ifdef CONFIG_NETLINK_DEV
+	init_netlink();
 #endif
 	return 0;
 }
@@ -1075,7 +1081,7 @@ static void __exit netlink_proto_exit(void)
        remove_proc_entry("net/netlink", NULL);
 }
 
-module_init(netlink_proto_init);
+subsys_initcall(netlink_proto_init);
 module_exit(netlink_proto_exit);
 
 MODULE_LICENSE("GPL");

net/socket.c

@@ -69,8 +69,6 @@
 #include <linux/proc_fs.h>
 #include <linux/seq_file.h>
 #include <linux/wanrouter.h>
-#include <linux/netlink.h>
-#include <linux/rtnetlink.h>
 #include <linux/if_bridge.h>
 #include <linux/init.h>
 #include <linux/poll.h>
@@ -143,6 +141,36 @@ static struct file_operations socket_file_ops = {
 
 static struct net_proto_family *net_families[NPROTO];
 
+static __inline__ void net_family_bug(int family)
+{
+	printk(KERN_ERR "%d is not yet sock_registered!\n", family);
+	BUG();
+}
+
+int net_family_get(int family)
+{
+	struct net_proto_family *prot = net_families[family];
+	int rc = 1;
+
+	barrier();
+	if (likely(prot != NULL))
+		rc = try_module_get(prot->owner);
+	else
+		net_family_bug(family);
+	return rc;
+}
+
+void net_family_put(int family)
+{
+	struct net_proto_family *prot = net_families[family];
+
+	barrier();
+	if (likely(prot != NULL))
+		module_put(prot->owner);
+	else
+		net_family_bug(family);
+}
+
 #if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT)
 static atomic_t net_family_lockct = ATOMIC_INIT(0);
 static spinlock_t net_family_lock = SPIN_LOCK_UNLOCKED;
@@ -511,7 +539,7 @@ void sock_release(struct socket *sock)
 
 		sock->ops->release(sock);
 		sock->ops = NULL;
-		module_put(net_families[family]->owner);
+		net_family_put(family);
 	}
 
 	if (sock->fasync_list)
@@ -1064,7 +1092,7 @@ int sock_create(int family, int type, int protocol, struct socket **res)
 	sock->type  = type;
 
 	i = -EBUSY;
-	if (!try_module_get(net_families[family]->owner))
+	if (!net_family_get(family))
 		goto out_release;
 
 	if ((i = net_families[family]->create(sock, protocol)) < 0) 
@@ -1953,17 +1981,6 @@ void __init sock_init(void)
 	 *  do_initcalls is run.  
 	 */
 
-
-	/*
-	 * The netlink device handler may be needed early.
-	 */
-
-#ifdef CONFIG_NET
-	rtnetlink_init();
-#endif
-#ifdef CONFIG_NETLINK_DEV
-	init_netlink();
-#endif
 #ifdef CONFIG_NETFILTER
 	netfilter_init();
 #endif