Merge nuts.davemloft.net:/disk1/BK/net-exp-2.6

into nuts.davemloft.net:/disk1/BK/net-2.6

crypto/Kconfig

@@ -182,7 +182,7 @@ config CRYPTO_TEA
 	  many rounds for security.  It is very fast and uses
 	  little memory.
 
-	  Xtendend Tiny Encryption Algorithm is a modifcation to
+	  Xtendend Tiny Encryption Algorithm is a modification to
 	  the TEA algorithm to address a potential key weakness
 	  in the TEA algorithm.
 

crypto/aes.c

@@ -160,7 +160,7 @@ gen_tabs (void)
 	u8 p, q;
 
 	/* log and power tables for GF(2**8) finite field with
-	   0x011b as modular polynomial - the simplest prmitive
+	   0x011b as modular polynomial - the simplest primitive
 	   root is 0x03, used here to generate the tables */
 
 	for (i = 0, p = 1; i < 256; ++i) {

crypto/arc4.c

@@ -3,7 +3,7 @@
  *
  * ARC4 Cipher Algorithm
  *
- * Jon Oberheide <jon@focalhost.com>
+ * Jon Oberheide <jon@oberheide.org>
  *
  * 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
@@ -100,4 +100,4 @@ module_exit(arc4_exit);
 
 MODULE_LICENSE("GPL");
 MODULE_DESCRIPTION("ARC4 Cipher Algorithm");
-MODULE_AUTHOR("Jon Oberheide <jon@focalhost.com>");
+MODULE_AUTHOR("Jon Oberheide <jon@oberheide.org>");

crypto/blowfish.c

@@ -3,9 +3,9 @@
  *
  * Blowfish Cipher Algorithm, by Bruce Schneier.
  * http://www.counterpane.com/blowfish.html
- * 
- * Adapated from Kerneli implementation.
- * 
+ *
+ * Adapted from Kerneli implementation.
+ *
  * Copyright (c) Herbert Valerio Riedel <hvr@hvrlab.org>
  * Copyright (c) Kyle McMartin <kyle@debian.org>
  * Copyright (c) 2002 James Morris <jmorris@intercode.com.au>

crypto/scatterwalk.c

@@ -70,7 +70,7 @@ static void scatterwalk_pagedone(struct scatter_walk *walk, int out,
 {
 	/* walk->data may be pointing the first byte of the next page;
 	   however, we know we transfered at least one byte.  So,
-	   walk->data - 1 will be a virutual address in the mapped page. */
+	   walk->data - 1 will be a virtual address in the mapped page. */
 
 	if (out)
 		flush_dcache_page(walk->page);

crypto/tcrypt.h

@@ -1186,7 +1186,7 @@ struct cipher_testvec tf_cbc_dec_tv_template[] = {
 
 /*
  * Serpent test vectors.  These are backwards because Serpent writes
- * octect sequences in right-to-left mode.
+ * octet sequences in right-to-left mode.
  */
 #define SERPENT_ENC_TEST_VECTORS	4
 #define SERPENT_DEC_TEST_VECTORS	4

crypto/twofish.c

 /*
  * Twofish for CryptoAPI
  *
- * Originaly Twofish for GPG
+ * Originally Twofish for GPG
  * By Matthew Skala <mskala@ansuz.sooke.bc.ca>, July 26, 1998
  * 256-bit key length added March 20, 1999
  * Some modifications to reduce the text size by Werner Koch, April, 1998
@@ -514,7 +514,7 @@ static const u8 calc_sb_tbl[512] = {
  * preprocessed through q0 and q1 respectively; for longer keys they are the
  * output of previous stages.  j is the index of the first key byte to use.
  * CALC_K computes a pair of subkeys for 128-bit Twofish, by calling CALC_K_2
- * twice, doing the Psuedo-Hadamard Transform, and doing the necessary
+ * twice, doing the Pseudo-Hadamard Transform, and doing the necessary
  * rotations.  Its parameters are: a, the array to write the results into,
  * j, the index of the first output entry, k and l, the preprocessed indices
  * for index 2i, and m and n, the preprocessed indices for index 2i+1.

drivers/net/wan/syncppp.c

@@ -50,6 +50,7 @@
 #include <linux/random.h>
 #include <linux/pkt_sched.h>
 #include <linux/spinlock.h>
+#include <linux/rcupdate.h>
 
 #include <net/syncppp.h>
 
@@ -767,9 +768,9 @@ static void sppp_cisco_input (struct sppp *sp, struct sk_buff *skb)
 		struct in_ifaddr *ifa;
 		u32 addr = 0, mask = ~0; /* FIXME: is the mask correct? */
 #ifdef CONFIG_INET
-		if ((in_dev=in_dev_get(dev)) != NULL)
+		rcu_read_lock();
+		if ((in_dev = __in_dev_get(dev)) != NULL)
 		{
-			read_lock(&in_dev->lock);
 			for (ifa=in_dev->ifa_list; ifa != NULL;
 				ifa=ifa->ifa_next) {
 				if (strcmp(dev->name, ifa->ifa_label) == 0) 
@@ -779,9 +780,8 @@ static void sppp_cisco_input (struct sppp *sp, struct sk_buff *skb)
 					break;
 				}
 			}
-			read_unlock(&in_dev->lock);
-			in_dev_put(in_dev);
 		}
+		rcu_read_unlock();
 #endif		
 		/* I hope both addr and mask are in the net order */
 		sppp_cisco_send (sp, CISCO_ADDR_REPLY, addr, mask);

drivers/net/wireless/strip.c

@@ -106,6 +106,7 @@ static const char StripVersion[] = "1.3A-STUART.CHESHIRE";
 #include <linux/seq_file.h>
 #include <linux/serial.h>
 #include <linux/serialP.h>
+#include <linux/rcupdate.h>
 #include <net/arp.h>
 
 #include <linux/ip.h>
@@ -1348,14 +1349,17 @@ static unsigned char *strip_make_packet(unsigned char *buffer,
 	 */
 	if (haddr.c[0] == 0xFF) {
 		u32 brd = 0;
-		struct in_device *in_dev = in_dev_get(strip_info->dev);
-		if (in_dev == NULL)
+		struct in_device *in_dev;
+
+		rcu_read_lock();
+		in_dev = __in_dev_get(strip_info->dev);
+		if (in_dev == NULL) {
+			rcu_read_unlock();
 			return NULL;
-		read_lock(&in_dev->lock);
+		}
 		if (in_dev->ifa_list)
 			brd = in_dev->ifa_list->ifa_broadcast;
-		read_unlock(&in_dev->lock);
-		in_dev_put(in_dev);
+		rcu_read_unlock();
 
 		/* arp_query returns 1 if it succeeds in looking up the address, 0 if it fails */
 		if (!arp_query(haddr.c, brd, strip_info->dev)) {
@@ -1500,17 +1504,18 @@ static void strip_send(struct strip *strip_info, struct sk_buff *skb)
 	}
 
 	if (1) {
-		struct in_device *in_dev = in_dev_get(strip_info->dev);
+		struct in_device *in_dev;
+
 		brd = addr = 0;
+		rcu_read_lock();
+		in_dev = __in_dev_get(strip_info->dev);
 		if (in_dev) {
-			read_lock(&in_dev->lock);
 			if (in_dev->ifa_list) {
 				brd = in_dev->ifa_list->ifa_broadcast;
 				addr = in_dev->ifa_list->ifa_local;
 			}
-			read_unlock(&in_dev->lock);
-			in_dev_put(in_dev);
 		}
+		rcu_read_unlock();
 	}
 
 

drivers/s390/net/lcs.c

@@ -1002,7 +1002,7 @@ lcs_register_mc_addresses(void *data)
 	in4_dev = in_dev_get(card->dev);
 	if (in4_dev == NULL)
 		return 0;
-	read_lock(&in4_dev->lock);
+	read_lock(&in4_dev->mc_list_lock);
 	spin_lock(&card->ipm_lock);
 	/* Check for multicast addresses to be removed. */
 	list_for_each(l, &card->ipm_list) {
@@ -1046,7 +1046,7 @@ lcs_register_mc_addresses(void *data)
 		list_add(&ipm->list, &card->ipm_list);
 	}
 	spin_unlock(&card->ipm_lock);
-	read_unlock(&in4_dev->lock);
+	read_unlock(&in4_dev->mc_list_lock);
 	in_dev_put(in4_dev);
 	lcs_fix_multicast_list(card);
 	return 0;

drivers/s390/net/qeth_main.c

@@ -73,6 +73,7 @@ qeth_eyecatcher(void)
 #include <linux/reboot.h>
 #include <asm/qeth.h>
 #include <linux/mii.h>
+#include <linux/rcupdate.h>
 
 #include "qeth.h"
 #include "qeth_mpc.h"
@@ -4733,9 +4734,10 @@ qeth_free_vlan_addresses4(struct qeth_card *card, unsigned short vid)
 	QETH_DBF_TEXT(trace, 4, "frvaddr4");
 	if (!card->vlangrp)
 		return;
-	in_dev = in_dev_get(card->vlangrp->vlan_devices[vid]);
+	rcu_read_lock();
+	in_dev = __in_dev_get(card->vlangrp->vlan_devices[vid]);
 	if (!in_dev)
-		return;
+		goto out;
 	for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next){
 		addr = qeth_get_addr_buffer(QETH_PROT_IPV4);
 		if (addr){
@@ -4746,7 +4748,8 @@ qeth_free_vlan_addresses4(struct qeth_card *card, unsigned short vid)
 				kfree(addr);
 		}
 	}
-	in_dev_put(in_dev);
+out:
+	rcu_read_unlock();
 }
 
 static void
@@ -4918,9 +4921,9 @@ qeth_add_vlan_mc(struct qeth_card *card)
 		in_dev = in_dev_get(vg->vlan_devices[i]);
 		if (!in_dev)
 			continue;
-		read_lock(&in_dev->lock);
+		read_lock(&in_dev->mc_list_lock);
 		qeth_add_mc(card,in_dev);
-		read_unlock(&in_dev->lock);
+		read_unlock(&in_dev->mc_list_lock);
 		in_dev_put(in_dev);
 	}
 #endif
@@ -4935,10 +4938,10 @@ qeth_add_multicast_ipv4(struct qeth_card *card)
 	in4_dev = in_dev_get(card->dev);
 	if (in4_dev == NULL)
 		return;
-	read_lock(&in4_dev->lock);
+	read_lock(&in4_dev->mc_list_lock);
 	qeth_add_mc(card, in4_dev);
 	qeth_add_vlan_mc(card);
-	read_unlock(&in4_dev->lock);
+	read_unlock(&in4_dev->mc_list_lock);
 	in_dev_put(in4_dev);
 }
 

include/linux/inetdevice.h

@@ -3,6 +3,8 @@
 
 #ifdef __KERNEL__
 
+#include <linux/rcupdate.h>
+
 struct ipv4_devconf
 {
 	int	accept_redirects;
@@ -31,13 +33,13 @@ extern struct ipv4_devconf ipv4_devconf;
 
 struct in_device
 {
-	struct net_device		*dev;
+	struct net_device	*dev;
 	atomic_t		refcnt;
-	rwlock_t		lock;
 	int			dead;
 	struct in_ifaddr	*ifa_list;	/* IP ifaddr chain		*/
+	rwlock_t		mc_list_lock;
 	struct ip_mc_list	*mc_list;	/* IP multicast filter chain    */
-	rwlock_t		mc_lock;	/* for mc_tomb */
+	spinlock_t		mc_tomb_lock;
 	struct ip_mc_list	*mc_tomb;
 	unsigned long		mr_v1_seen;
 	unsigned long		mr_v2_seen;
@@ -50,6 +52,7 @@ struct in_device
 
 	struct neigh_parms	*arp_parms;
 	struct ipv4_devconf	cnf;
+	struct rcu_head		rcu_head;
 };
 
 #define IN_DEV_FORWARD(in_dev)		((in_dev)->cnf.forwarding)
@@ -80,6 +83,7 @@ struct in_ifaddr
 {
 	struct in_ifaddr	*ifa_next;
 	struct in_device	*ifa_dev;
+	struct rcu_head		rcu_head;
 	u32			ifa_local;
 	u32			ifa_address;
 	u32			ifa_mask;
@@ -133,19 +137,16 @@ static __inline__ int bad_mask(u32 mask, u32 addr)
 
 #define endfor_ifa(in_dev) }
 
-extern rwlock_t inetdev_lock;
-
-
 static __inline__ struct in_device *
 in_dev_get(const struct net_device *dev)
 {
 	struct in_device *in_dev;
 
-	read_lock(&inetdev_lock);
+	rcu_read_lock();
 	in_dev = dev->ip_ptr;
 	if (in_dev)
 		atomic_inc(&in_dev->refcnt);
-	read_unlock(&inetdev_lock);
+	rcu_read_unlock();
 	return in_dev;
 }
 
@@ -157,8 +158,7 @@ __in_dev_get(const struct net_device *dev)
 
 extern void in_dev_finish_destroy(struct in_device *idev);
 
-static __inline__ void
-in_dev_put(struct in_device *idev)
+static inline void in_dev_put(struct in_device *idev)
 {
 	if (atomic_dec_and_test(&idev->refcnt))
 		in_dev_finish_destroy(idev);

include/linux/net.h

@@ -169,6 +169,7 @@ extern struct socket *sockfd_lookup(int fd, int *err);
 extern int	     net_ratelimit(void);
 extern unsigned long net_random(void);
 extern void	     net_srandom(unsigned long);
+extern void	     net_random_init(void);
 
 extern int   	     kernel_sendmsg(struct socket *sock, struct msghdr *msg,
 				    struct kvec *vec, size_t num, size_t len);

include/net/ip_fib.h

@@ -242,7 +242,6 @@ extern u32 fib_rules_map_destination(u32 daddr, struct fib_result *res);
 #ifdef CONFIG_NET_CLS_ROUTE
 extern u32 fib_rules_tclass(struct fib_result *res);
 #endif
-extern u32 fib_rules_policy(u32 saddr, struct fib_result *res, unsigned *flags);
 extern void fib_rules_init(void);
 #endif
 

include/net/route.h

@@ -73,11 +73,6 @@ struct rtable
 	/* Miscellaneous cached information */
 	__u32			rt_spec_dst; /* RFC1122 specific destination */
 	struct inet_peer	*peer; /* long-living peer info */
-
-#ifdef CONFIG_IP_ROUTE_NAT
-	__u32			rt_src_map;
-	__u32			rt_dst_map;
-#endif
 };
 
 struct ip_rt_acct

net/bridge/netfilter/ebt_vlan.c

@@ -21,13 +21,14 @@
 #include <linux/if_ether.h>
 #include <linux/if_vlan.h>
 #include <linux/module.h>
+#include <linux/moduleparam.h>
 #include <linux/netfilter_bridge/ebtables.h>
 #include <linux/netfilter_bridge/ebt_vlan.h>
 
-static unsigned char debug;
+static int debug;
 #define MODULE_VERS "0.6"
 
-MODULE_PARM(debug, "0-1b");
+module_param(debug, int, 0);
 MODULE_PARM_DESC(debug, "debug=1 is turn on debug messages");
 MODULE_AUTHOR("Nick Fedchik <nick@fedchik.org.ua>");
 MODULE_DESCRIPTION("802.1Q match module (ebtables extension), v"

net/core/dev.c

@@ -3280,6 +3280,8 @@ static int __init net_dev_init(void)
 
 	BUG_ON(!dev_boot_phase);
 
+	net_random_init();
+
 	if (dev_proc_init())
 		goto out;
 

net/core/netpoll.c

@@ -18,6 +18,7 @@
 #include <linux/interrupt.h>
 #include <linux/netpoll.h>
 #include <linux/sched.h>
+#include <linux/rcupdate.h>
 #include <net/tcp.h>
 #include <net/udp.h>
 
@@ -572,16 +573,18 @@ int netpoll_setup(struct netpoll *np)
 		memcpy(np->local_mac, ndev->dev_addr, 6);
 
 	if (!np->local_ip) {
-		in_dev = in_dev_get(ndev);
+		rcu_read_lock();
+		in_dev = __in_dev_get(ndev);
 
 		if (!in_dev) {
+			rcu_read_unlock();
 			printk(KERN_ERR "%s: no IP address for %s, aborting\n",
 			       np->name, np->dev_name);
 			goto release;
 		}
 
 		np->local_ip = ntohl(in_dev->ifa_list->ifa_local);
-		in_dev_put(in_dev);
+		rcu_read_unlock();
 		printk(KERN_INFO "%s: local IP %d.%d.%d.%d\n",
 		       np->name, HIPQUAD(np->local_ip));
 	}

net/core/pktgen.c

@@ -70,6 +70,7 @@
 #include <linux/delay.h>
 #include <linux/init.h>
 #include <linux/inet.h>
+#include <linux/rcupdate.h>
 #include <asm/byteorder.h>
 #include <asm/bitops.h>
 #include <asm/io.h>
@@ -263,14 +264,17 @@ static struct net_device *setup_inject(struct pktgen_info* info)
 	info->saddr_min = 0;
 	info->saddr_max = 0;
 	if (strlen(info->src_min) == 0) {
-		struct in_device *in_dev = in_dev_get(odev);
+		struct in_device *in_dev;
+
+		rcu_read_lock();
+		in_dev = __in_dev_get(odev);
 		if (in_dev) {
 			if (in_dev->ifa_list) {
 				info->saddr_min = in_dev->ifa_list->ifa_address;
 				info->saddr_max = info->saddr_min;
 			}
-			in_dev_put(in_dev);
 		}
+		rcu_read_unlock();
 	}
 	else {
 		info->saddr_min = in_aton(info->src_min);

net/core/utils.c

@@ -19,22 +19,116 @@
 #include <linux/mm.h>
 #include <linux/string.h>
 #include <linux/types.h>
+#include <linux/random.h>
+#include <linux/percpu.h>
 
 #include <asm/system.h>
 #include <asm/uaccess.h>
 
-static unsigned long net_rand_seed = 152L;
+
+/*
+  This is a maximally equidistributed combined Tausworthe generator
+  based on code from GNU Scientific Library 1.5 (30 Jun 2004)
+
+   x_n = (s1_n ^ s2_n ^ s3_n) 
+
+   s1_{n+1} = (((s1_n & 4294967294) <<12) ^ (((s1_n <<13) ^ s1_n) >>19))
+   s2_{n+1} = (((s2_n & 4294967288) << 4) ^ (((s2_n << 2) ^ s2_n) >>25))
+   s3_{n+1} = (((s3_n & 4294967280) <<17) ^ (((s3_n << 3) ^ s3_n) >>11))
+
+   The period of this generator is about 2^88.
+
+   From: P. L'Ecuyer, "Maximally Equidistributed Combined Tausworthe
+   Generators", Mathematics of Computation, 65, 213 (1996), 203--213.
+
+   This is available on the net from L'Ecuyer's home page,
+
+   http://www.iro.umontreal.ca/~lecuyer/myftp/papers/tausme.ps
+   ftp://ftp.iro.umontreal.ca/pub/simulation/lecuyer/papers/tausme.ps 
+
+   There is an erratum in the paper "Tables of Maximally
+   Equidistributed Combined LFSR Generators", Mathematics of
+   Computation, 68, 225 (1999), 261--269:
+   http://www.iro.umontreal.ca/~lecuyer/myftp/papers/tausme2.ps
+
+        ... the k_j most significant bits of z_j must be non-
+        zero, for each j. (Note: this restriction also applies to the 
+        computer code given in [4], but was mistakenly not mentioned in
+        that paper.)
+   
+   This affects the seeding procedure by imposing the requirement
+   s1 > 1, s2 > 7, s3 > 15.
+
+*/
+struct nrnd_state {
+	u32 s1, s2, s3;
+};
+
+static DEFINE_PER_CPU(struct nrnd_state, net_rand_state);
+
+static u32 __net_random(struct nrnd_state *state)
+{
+#define TAUSWORTHE(s,a,b,c,d) ((s&c)<<d) ^ (((s <<a) ^ s)>>b)
+
+	state->s1 = TAUSWORTHE(state->s1, 13, 19, 4294967294UL, 12);
+	state->s2 = TAUSWORTHE(state->s2, 2, 25, 4294967288UL, 4);
+	state->s3 = TAUSWORTHE(state->s3, 3, 11, 4294967280UL, 17);
+
+	return (state->s1 ^ state->s2 ^ state->s3);
+}
+
+static void __net_srandom(struct nrnd_state *state, unsigned long entropy)
+{
+	u32 s = state->s1 ^ entropy;
+
+	if (s == 0)
+		s = 1;      /* default seed is 1 */
+
+#define LCG(n) (69069 * n)
+	state->s1 = LCG(s);
+	state->s2 = LCG(state->s1);
+	state->s3 = LCG(state->s2);
+
+	/* "warm it up" */
+	__net_random(state);
+	__net_random(state);
+	__net_random(state);
+	__net_random(state);
+	__net_random(state);
+	__net_random(state);
+}
+
 
 unsigned long net_random(void)
 {
-	net_rand_seed=net_rand_seed*69069L+1;
-        return net_rand_seed^jiffies;
+	unsigned long r;
+	struct nrnd_state *state = &get_cpu_var(net_rand_state);
+	r = __net_random(state);
+	put_cpu_var(state);
+	return r;
 }
 
+
 void net_srandom(unsigned long entropy)
 {
-	net_rand_seed ^= entropy;
-	net_random();
+	struct nrnd_state *state = &get_cpu_var(net_rand_state);
+	__net_srandom(state, entropy);
+	put_cpu_var(state);
+}
+
+void __init net_random_init(void)
+{
+	int i;
+	unsigned long seed[NR_CPUS];
+
+	get_random_bytes(seed, sizeof(seed));
+
+	for (i = 0; i < NR_CPUS; i++) {
+		struct nrnd_state *state = &per_cpu(net_rand_state,i);
+
+		memset(state, 0, sizeof(*state));
+		__net_srandom(state, seed[i]);
+	}
 }
 
 int net_msg_cost = 5*HZ;

net/econet/af_econet.c

@@ -39,6 +39,7 @@
 #include <net/udp.h>
 #include <net/ip.h>
 #include <linux/spinlock.h>
+#include <linux/rcupdate.h>
 
 #include <asm/uaccess.h>
 #include <asm/system.h>
@@ -401,16 +402,17 @@ static int econet_sendmsg(struct kiocb *iocb, struct socket *sock,
 	   y.x maps to IP a.b.c.x.  This should be replaced with something
 	   more flexible and more aware of subnet masks.  */
 	{
-		struct in_device *idev = in_dev_get(dev);
+		struct in_device *idev;
 		unsigned long network = 0;
+
+		rcu_read_lock();
+		idev = __in_dev_get(dev);
 		if (idev) {
-			read_lock(&idev->lock);
 			if (idev->ifa_list)
 				network = ntohl(idev->ifa_list->ifa_address) & 
 					0xffffff00;		/* !!! */
-			read_unlock(&idev->lock);
-			in_dev_put(idev);
 		}
+		rcu_read_unlock();
 		udpdest.sin_addr.s_addr = htonl(network | addr.station);
 	}
 

net/ipv4/Kconfig

@@ -82,16 +82,6 @@ config IP_ROUTE_FWMARK
 	  If you say Y here, you will be able to specify different routes for
 	  packets with different mark values (see iptables(8), MARK target).
 
-config IP_ROUTE_NAT
-	bool "IP: fast network address translation"
-	depends on IP_MULTIPLE_TABLES
-	help
-	  If you say Y here, your router will be able to modify source and
-	  destination addresses of packets that pass through it, in a manner
-	  you specify.  General information about Network Address Translation
-	  can be gotten from the document
-	  <http://www.hasenstein.com/linux-ip-nat/diplom/nat.html>.
-
 config IP_ROUTE_MULTIPATH
 	bool "IP: equal cost multipath"
 	depends on IP_ADVANCED_ROUTER

net/ipv4/devinet.c

@@ -88,31 +88,31 @@ static void devinet_sysctl_register(struct in_device *in_dev,
 static void devinet_sysctl_unregister(struct ipv4_devconf *p);
 #endif
 
-int inet_ifa_count;
-int inet_dev_count;
-
 /* Locks all the inet devices. */
 
-rwlock_t inetdev_lock = RW_LOCK_UNLOCKED;
-
 static struct in_ifaddr *inet_alloc_ifa(void)
 {
 	struct in_ifaddr *ifa = kmalloc(sizeof(*ifa), GFP_KERNEL);
 
 	if (ifa) {
 		memset(ifa, 0, sizeof(*ifa));
-		inet_ifa_count++;
+		INIT_RCU_HEAD(&ifa->rcu_head);
 	}
 
 	return ifa;
 }
 
-static __inline__ void inet_free_ifa(struct in_ifaddr *ifa)
+static void inet_rcu_free_ifa(struct rcu_head *head)
 {
+	struct in_ifaddr *ifa = container_of(head, struct in_ifaddr, rcu_head);
 	if (ifa->ifa_dev)
-		__in_dev_put(ifa->ifa_dev);
+		in_dev_put(ifa->ifa_dev);
 	kfree(ifa);
-	inet_ifa_count--;
+}
+
+static inline void inet_free_ifa(struct in_ifaddr *ifa)
+{
+	call_rcu(&ifa->rcu_head, inet_rcu_free_ifa);
 }
 
 void in_dev_finish_destroy(struct in_device *idev)
@@ -129,7 +129,6 @@ void in_dev_finish_destroy(struct in_device *idev)
 	if (!idev->dead)
 		printk("Freeing alive in_device %p\n", idev);
 	else {
-		inet_dev_count--;
 		kfree(idev);
 	}
 }
@@ -144,24 +143,24 @@ struct in_device *inetdev_init(struct net_device *dev)
 	if (!in_dev)
 		goto out;
 	memset(in_dev, 0, sizeof(*in_dev));
-	in_dev->lock = RW_LOCK_UNLOCKED;
+	INIT_RCU_HEAD(&in_dev->rcu_head);
 	memcpy(&in_dev->cnf, &ipv4_devconf_dflt, sizeof(in_dev->cnf));
 	in_dev->cnf.sysctl = NULL;
 	in_dev->dev = dev;
 	if ((in_dev->arp_parms = neigh_parms_alloc(dev, &arp_tbl)) == NULL)
 		goto out_kfree;
-	inet_dev_count++;
 	/* Reference in_dev->dev */
 	dev_hold(dev);
 #ifdef CONFIG_SYSCTL
 	neigh_sysctl_register(dev, in_dev->arp_parms, NET_IPV4,
 			      NET_IPV4_NEIGH, "ipv4", NULL);
 #endif
-	write_lock_bh(&inetdev_lock);
-	dev->ip_ptr = in_dev;
+
 	/* Account for reference dev->ip_ptr */
 	in_dev_hold(in_dev);
-	write_unlock_bh(&inetdev_lock);
+	smp_wmb();
+	dev->ip_ptr = in_dev;
+
 #ifdef CONFIG_SYSCTL
 	devinet_sysctl_register(in_dev, &in_dev->cnf);
 #endif
@@ -176,6 +175,12 @@ struct in_device *inetdev_init(struct net_device *dev)
 	goto out;
 }
 
+static void in_dev_rcu_put(struct rcu_head *head)
+{
+	struct in_device *idev = container_of(head, struct in_device, rcu_head);
+	in_dev_put(idev);
+}
+
 static void inetdev_destroy(struct in_device *in_dev)
 {
 	struct in_ifaddr *ifa;
@@ -194,30 +199,28 @@ static void inetdev_destroy(struct in_device *in_dev)
 #ifdef CONFIG_SYSCTL
 	devinet_sysctl_unregister(&in_dev->cnf);
 #endif
-	write_lock_bh(&inetdev_lock);
+
 	in_dev->dev->ip_ptr = NULL;
-	/* in_dev_put following below will kill the in_device */
-	write_unlock_bh(&inetdev_lock);
 
 #ifdef CONFIG_SYSCTL
 	neigh_sysctl_unregister(in_dev->arp_parms);
 #endif
 	neigh_parms_release(&arp_tbl, in_dev->arp_parms);
-	in_dev_put(in_dev);
+	call_rcu(&in_dev->rcu_head, in_dev_rcu_put);
 }
 
 int inet_addr_onlink(struct in_device *in_dev, u32 a, u32 b)
 {
-	read_lock(&in_dev->lock);
+	rcu_read_lock();
 	for_primary_ifa(in_dev) {
 		if (inet_ifa_match(a, ifa)) {
 			if (!b || inet_ifa_match(b, ifa)) {
-				read_unlock(&in_dev->lock);
+				rcu_read_unlock();
 				return 1;
 			}
 		}
 	} endfor_ifa(in_dev);
-	read_unlock(&in_dev->lock);
+	rcu_read_unlock();
 	return 0;
 }
 
@@ -241,9 +244,8 @@ static void inet_del_ifa(struct in_device *in_dev, struct in_ifaddr **ifap,
 				ifap1 = &ifa->ifa_next;
 				continue;
 			}
-			write_lock_bh(&in_dev->lock);
+
 			*ifap1 = ifa->ifa_next;
-			write_unlock_bh(&in_dev->lock);
 
 			rtmsg_ifa(RTM_DELADDR, ifa);
 			notifier_call_chain(&inetaddr_chain, NETDEV_DOWN, ifa);
@@ -253,9 +255,7 @@ static void inet_del_ifa(struct in_device *in_dev, struct in_ifaddr **ifap,
 
 	/* 2. Unlink it */
 
-	write_lock_bh(&in_dev->lock);
 	*ifap = ifa1->ifa_next;
-	write_unlock_bh(&in_dev->lock);
 
 	/* 3. Announce address deletion */
 
@@ -317,9 +317,7 @@ static int inet_insert_ifa(struct in_ifaddr *ifa)
 	}
 
 	ifa->ifa_next = *ifap;
-	write_lock_bh(&in_dev->lock);
 	*ifap = ifa;
-	write_unlock_bh(&in_dev->lock);
 
 	/* Send message first, then call notifier.
 	   Notifier will trigger FIB update, so that
@@ -771,12 +769,11 @@ u32 inet_select_addr(const struct net_device *dev, u32 dst, int scope)
 	u32 addr = 0;
 	struct in_device *in_dev;
 
-	read_lock(&inetdev_lock);
+	rcu_read_lock();
 	in_dev = __in_dev_get(dev);
 	if (!in_dev)
-		goto out_unlock_inetdev;
+		goto no_in_dev;
 
-	read_lock(&in_dev->lock);
 	for_primary_ifa(in_dev) {
 		if (ifa->ifa_scope > scope)
 			continue;
@@ -787,8 +784,8 @@ u32 inet_select_addr(const struct net_device *dev, u32 dst, int scope)
 		if (!addr)
 			addr = ifa->ifa_local;
 	} endfor_ifa(in_dev);
-	read_unlock(&in_dev->lock);
-	read_unlock(&inetdev_lock);
+no_in_dev:
+	rcu_read_unlock();
 
 	if (addr)
 		goto out;
@@ -798,30 +795,24 @@ u32 inet_select_addr(const struct net_device *dev, u32 dst, int scope)
 	   in dev_base list.
 	 */
 	read_lock(&dev_base_lock);
-	read_lock(&inetdev_lock);
+	rcu_read_lock();
 	for (dev = dev_base; dev; dev = dev->next) {
 		if ((in_dev = __in_dev_get(dev)) == NULL)
 			continue;
 
-		read_lock(&in_dev->lock);
 		for_primary_ifa(in_dev) {
 			if (ifa->ifa_scope != RT_SCOPE_LINK &&
 			    ifa->ifa_scope <= scope) {
-				read_unlock(&in_dev->lock);
 				addr = ifa->ifa_local;
 				goto out_unlock_both;
 			}
 		} endfor_ifa(in_dev);
-		read_unlock(&in_dev->lock);
 	}
 out_unlock_both:
-	read_unlock(&inetdev_lock);
 	read_unlock(&dev_base_lock);
+	rcu_read_unlock();
 out:
 	return addr;
-out_unlock_inetdev:
-	read_unlock(&inetdev_lock);
-	goto out;
 }
 
 static u32 confirm_addr_indev(struct in_device *in_dev, u32 dst,
@@ -874,29 +865,24 @@ u32 inet_confirm_addr(const struct net_device *dev, u32 dst, u32 local, int scop
 	struct in_device *in_dev;
 
 	if (dev) {
-		read_lock(&inetdev_lock);
-		if ((in_dev = __in_dev_get(dev))) {
-			read_lock(&in_dev->lock);
+		rcu_read_lock();
+		if ((in_dev = __in_dev_get(dev)))
 			addr = confirm_addr_indev(in_dev, dst, local, scope);
-			read_unlock(&in_dev->lock);
-		}
-		read_unlock(&inetdev_lock);
+		rcu_read_unlock();
 
 		return addr;
 	}
 
 	read_lock(&dev_base_lock);
-	read_lock(&inetdev_lock);
+	rcu_read_lock();
 	for (dev = dev_base; dev; dev = dev->next) {
 		if ((in_dev = __in_dev_get(dev))) {
-			read_lock(&in_dev->lock);
 			addr = confirm_addr_indev(in_dev, dst, local, scope);
-			read_unlock(&in_dev->lock);
 			if (addr)
 				break;
 		}
 	}
-	read_unlock(&inetdev_lock);
+	rcu_read_unlock();
 	read_unlock(&dev_base_lock);
 
 	return addr;
@@ -1065,12 +1051,12 @@ static int inet_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
 			continue;
 		if (idx > s_idx)
 			s_ip_idx = 0;
-		read_lock(&inetdev_lock);
+		rcu_read_lock();
 		if ((in_dev = __in_dev_get(dev)) == NULL) {
-			read_unlock(&inetdev_lock);
+			rcu_read_unlock();
 			continue;
 		}
-		read_lock(&in_dev->lock);
+
 		for (ifa = in_dev->ifa_list, ip_idx = 0; ifa;
 		     ifa = ifa->ifa_next, ip_idx++) {
 			if (ip_idx < s_ip_idx)
@@ -1078,13 +1064,11 @@ static int inet_dump_ifaddr(struct sk_buff *skb, struct netlink_callback *cb)
 			if (inet_fill_ifaddr(skb, ifa, NETLINK_CB(cb->skb).pid,
 					     cb->nlh->nlmsg_seq,
 					     RTM_NEWADDR) <= 0) {
-				read_unlock(&in_dev->lock);
-				read_unlock(&inetdev_lock);
+				rcu_read_unlock();
 				goto done;
 			}
 		}
-		read_unlock(&in_dev->lock);
-		read_unlock(&inetdev_lock);
+		rcu_read_unlock();
 	}
 
 done:
@@ -1138,11 +1122,11 @@ void inet_forward_change(void)
 	read_lock(&dev_base_lock);
 	for (dev = dev_base; dev; dev = dev->next) {
 		struct in_device *in_dev;
-		read_lock(&inetdev_lock);
+		rcu_read_lock();
 		in_dev = __in_dev_get(dev);
 		if (in_dev)
 			in_dev->cnf.forwarding = on;
-		read_unlock(&inetdev_lock);
+		rcu_read_unlock();
 	}
 	read_unlock(&dev_base_lock);
 
@@ -1508,6 +1492,5 @@ EXPORT_SYMBOL(devinet_ioctl);
 EXPORT_SYMBOL(in_dev_finish_destroy);
 EXPORT_SYMBOL(inet_select_addr);
 EXPORT_SYMBOL(inetdev_by_index);
-EXPORT_SYMBOL(inetdev_lock);
 EXPORT_SYMBOL(register_inetaddr_notifier);
 EXPORT_SYMBOL(unregister_inetaddr_notifier);

net/ipv4/fib_frontend.c

@@ -172,13 +172,13 @@ int fib_validate_source(u32 src, u32 dst, u8 tos, int oif,
 	int ret;
 
 	no_addr = rpf = 0;
-	read_lock(&inetdev_lock);
+	rcu_read_lock();
 	in_dev = __in_dev_get(dev);
 	if (in_dev) {
 		no_addr = in_dev->ifa_list == NULL;
 		rpf = IN_DEV_RPFILTER(in_dev);
 	}
-	read_unlock(&inetdev_lock);
+	rcu_read_unlock();
 
 	if (in_dev == NULL)
 		goto e_inval;

net/ipv4/fib_rules.c

@@ -176,7 +176,7 @@ int inet_rtm_newrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 	table_id = rtm->rtm_table;
 	if (table_id == RT_TABLE_UNSPEC) {
 		struct fib_table *table;
-		if (rtm->rtm_type == RTN_UNICAST || rtm->rtm_type == RTN_NAT) {
+		if (rtm->rtm_type == RTN_UNICAST) {
 			if ((table = fib_empty_table()) == NULL)
 				return -ENOBUFS;
 			table_id = table->tb_id;
@@ -251,26 +251,6 @@ u32 fib_rules_map_destination(u32 daddr, struct fib_result *res)
 	return (daddr&~mask)|res->fi->fib_nh->nh_gw;
 }
 
-u32 fib_rules_policy(u32 saddr, struct fib_result *res, unsigned *flags)
-{
-	struct fib_rule *r = res->r;
-
-	if (r->r_action == RTN_NAT) {
-		int addrtype = inet_addr_type(r->r_srcmap);
-
-		if (addrtype == RTN_NAT) {
-			/* Packet is from  translated source; remember it */
-			saddr = (saddr&~r->r_srcmask)|r->r_srcmap;
-			*flags |= RTCF_SNAT;
-		} else if (addrtype == RTN_LOCAL || r->r_srcmap == 0) {
-			/* Packet is from masqueraded source; remember it */
-			saddr = r->r_srcmap;
-			*flags |= RTCF_MASQ;
-		}
-	}
-	return saddr;
-}
-
 #ifdef CONFIG_NET_CLS_ROUTE
 u32 fib_rules_tclass(struct fib_result *res)
 {
@@ -334,7 +314,6 @@ FRprintk("Lookup: %u.%u.%u.%u <- %u.%u.%u.%u ",
 FRprintk("tb %d r %d ", r->r_table, r->r_action);
 		switch (r->r_action) {
 		case RTN_UNICAST:
-		case RTN_NAT:
 			policy = r;
 			break;
 		case RTN_UNREACHABLE:

net/ipv4/fib_semantics.c

@@ -124,17 +124,10 @@ static struct
 		.error	= -EAGAIN,
 		.scope	= RT_SCOPE_UNIVERSE,
 	},	/* RTN_THROW */
-#ifdef CONFIG_IP_ROUTE_NAT
-	{
-		.error	= 0,
-		.scope	= RT_SCOPE_HOST,
-	},	/* RTN_NAT */
-#else
 	{
 		.error	= -EINVAL,
 		.scope	= RT_SCOPE_NOWHERE,
 	},	/* RTN_NAT */
-#endif
 	{
 		.error	= -EINVAL,
 		.scope	= RT_SCOPE_NOWHERE,
@@ -543,15 +536,6 @@ fib_create_info(const struct rtmsg *r, struct kern_rta *rta,
 #endif
 	}
 
-#ifdef CONFIG_IP_ROUTE_NAT
-	if (r->rtm_type == RTN_NAT) {
-		if (rta->rta_gw == NULL || nhs != 1 || rta->rta_oif)
-			goto err_inval;
-		memcpy(&fi->fib_nh->nh_gw, rta->rta_gw, 4);
-		goto link_it;
-	}
-#endif
-
 	if (fib_props[r->rtm_type].error) {
 		if (rta->rta_gw || rta->rta_oif || rta->rta_mp)
 			goto err_inval;
@@ -629,12 +613,6 @@ fib_semantic_match(int type, struct fib_info *fi, const struct flowi *flp, struc
 		res->fi = fi;
 
 		switch (type) {
-#ifdef CONFIG_IP_ROUTE_NAT
-		case RTN_NAT:
-			FIB_RES_RESET(*res);
-			atomic_inc(&fi->fib_clntref);
-			return 0;
-#endif
 		case RTN_UNICAST:
 		case RTN_LOCAL:
 		case RTN_BROADCAST:

net/ipv4/icmp.c

@@ -508,16 +508,6 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, u32 info)
 	 *	Construct source address and options.
 	 */
 
-#ifdef CONFIG_IP_ROUTE_NAT
-	/*
-	 *	Restore original addresses if packet has been translated.
-	 */
-	if (rt->rt_flags & RTCF_NAT && IPCB(skb_in)->flags & IPSKB_TRANSLATED) {
-		iph->daddr = rt->fl.fl4_dst;
-		iph->saddr = rt->fl.fl4_src;
-	}
-#endif
-
 	saddr = iph->daddr;
 	if (!(rt->rt_flags & RTCF_LOCAL))
 		saddr = 0;
@@ -892,7 +882,7 @@ static void icmp_address_reply(struct sk_buff *skb)
 	in_dev = in_dev_get(dev);
 	if (!in_dev)
 		goto out;
-	read_lock(&in_dev->lock);
+	rcu_read_lock();
 	if (in_dev->ifa_list &&
 	    IN_DEV_LOG_MARTIANS(in_dev) &&
 	    IN_DEV_FORWARD(in_dev)) {
@@ -912,7 +902,7 @@ static void icmp_address_reply(struct sk_buff *skb)
 			       NIPQUAD(*mp), dev->name, NIPQUAD(rt->rt_src));
 		}
 	}
-	read_unlock(&in_dev->lock);
+	rcu_read_unlock();
 	in_dev_put(in_dev);
 out:;
 }

net/ipv4/igmp.c

@@ -487,7 +487,7 @@ static int igmpv3_send_report(struct in_device *in_dev, struct ip_mc_list *pmc)
 	int type;
 
 	if (!pmc) {
-		read_lock(&in_dev->lock);
+		read_lock(&in_dev->mc_list_lock);
 		for (pmc=in_dev->mc_list; pmc; pmc=pmc->next) {
 			if (pmc->multiaddr == IGMP_ALL_HOSTS)
 				continue;
@@ -499,7 +499,7 @@ static int igmpv3_send_report(struct in_device *in_dev, struct ip_mc_list *pmc)
 			skb = add_grec(skb, pmc, type, 0, 0);
 			spin_unlock_bh(&pmc->lock);
 		}
-		read_unlock(&in_dev->lock);
+		read_unlock(&in_dev->mc_list_lock);
 	} else {
 		spin_lock_bh(&pmc->lock);
 		if (pmc->sfcount[MCAST_EXCLUDE])
@@ -541,8 +541,8 @@ static void igmpv3_send_cr(struct in_device *in_dev)
 	struct sk_buff *skb = NULL;
 	int type, dtype;
 
-	read_lock(&in_dev->lock);
-	write_lock_bh(&in_dev->mc_lock);
+	read_lock(&in_dev->mc_list_lock);
+	spin_lock_bh(&in_dev->mc_tomb_lock);
 
 	/* deleted MCA's */
 	pmc_prev = NULL;
@@ -575,7 +575,7 @@ static void igmpv3_send_cr(struct in_device *in_dev)
 		} else
 			pmc_prev = pmc;
 	}
-	write_unlock_bh(&in_dev->mc_lock);
+	spin_unlock_bh(&in_dev->mc_tomb_lock);
 
 	/* change recs */
 	for (pmc=in_dev->mc_list; pmc; pmc=pmc->next) {
@@ -601,7 +601,8 @@ static void igmpv3_send_cr(struct in_device *in_dev)
 		}
 		spin_unlock_bh(&pmc->lock);
 	}
-	read_unlock(&in_dev->lock);
+	read_unlock(&in_dev->mc_list_lock);
+
 	if (!skb)
 		return;
 	(void) igmpv3_sendpack(skb);
@@ -759,14 +760,14 @@ static void igmp_heard_report(struct in_device *in_dev, u32 group)
 	if (group == IGMP_ALL_HOSTS)
 		return;
 
-	read_lock(&in_dev->lock);
+	read_lock(&in_dev->mc_list_lock);
 	for (im=in_dev->mc_list; im!=NULL; im=im->next) {
 		if (im->multiaddr == group) {
 			igmp_stop_timer(im);
 			break;
 		}
 	}
-	read_unlock(&in_dev->lock);
+	read_unlock(&in_dev->mc_list_lock);
 }
 
 static void igmp_heard_query(struct in_device *in_dev, struct sk_buff *skb,
@@ -840,7 +841,7 @@ static void igmp_heard_query(struct in_device *in_dev, struct sk_buff *skb,
 	 * - Use the igmp->igmp_code field as the maximum
 	 *   delay possible
 	 */
-	read_lock(&in_dev->lock);
+	read_lock(&in_dev->mc_list_lock);
 	for (im=in_dev->mc_list; im!=NULL; im=im->next) {
 		if (group && group != im->multiaddr)
 			continue;
@@ -856,7 +857,7 @@ static void igmp_heard_query(struct in_device *in_dev, struct sk_buff *skb,
 		spin_unlock_bh(&im->lock);
 		igmp_mod_timer(im, max_delay);
 	}
-	read_unlock(&in_dev->lock);
+	read_unlock(&in_dev->mc_list_lock);
 }
 
 int igmp_rcv(struct sk_buff *skb)
@@ -982,10 +983,10 @@ static void igmpv3_add_delrec(struct in_device *in_dev, struct ip_mc_list *im)
 	}
 	spin_unlock_bh(&im->lock);
 
-	write_lock_bh(&in_dev->mc_lock);
+	spin_lock_bh(&in_dev->mc_tomb_lock);
 	pmc->next = in_dev->mc_tomb;
 	in_dev->mc_tomb = pmc;
-	write_unlock_bh(&in_dev->mc_lock);
+	spin_unlock_bh(&in_dev->mc_tomb_lock);
 }
 
 static void igmpv3_del_delrec(struct in_device *in_dev, __u32 multiaddr)
@@ -993,7 +994,7 @@ static void igmpv3_del_delrec(struct in_device *in_dev, __u32 multiaddr)
 	struct ip_mc_list *pmc, *pmc_prev;
 	struct ip_sf_list *psf, *psf_next;
 
-	write_lock_bh(&in_dev->mc_lock);
+	spin_lock_bh(&in_dev->mc_tomb_lock);
 	pmc_prev = NULL;
 	for (pmc=in_dev->mc_tomb; pmc; pmc=pmc->next) {
 		if (pmc->multiaddr == multiaddr)
@@ -1006,7 +1007,7 @@ static void igmpv3_del_delrec(struct in_device *in_dev, __u32 multiaddr)
 		else
 			in_dev->mc_tomb = pmc->next;
 	}
-	write_unlock_bh(&in_dev->mc_lock);
+	spin_unlock_bh(&in_dev->mc_tomb_lock);
 	if (pmc) {
 		for (psf=pmc->tomb; psf; psf=psf_next) {
 			psf_next = psf->sf_next;
@@ -1021,10 +1022,10 @@ static void igmpv3_clear_delrec(struct in_device *in_dev)
 {
 	struct ip_mc_list *pmc, *nextpmc;
 
-	write_lock_bh(&in_dev->mc_lock);
+	spin_lock_bh(&in_dev->mc_tomb_lock);
 	pmc = in_dev->mc_tomb;
 	in_dev->mc_tomb = NULL;
-	write_unlock_bh(&in_dev->mc_lock);
+	spin_unlock_bh(&in_dev->mc_tomb_lock);
 
 	for (; pmc; pmc = nextpmc) {
 		nextpmc = pmc->next;
@@ -1033,7 +1034,7 @@ static void igmpv3_clear_delrec(struct in_device *in_dev)
 		kfree(pmc);
 	}
 	/* clear dead sources, too */
-	read_lock(&in_dev->lock);
+	read_lock(&in_dev->mc_list_lock);
 	for (pmc=in_dev->mc_list; pmc; pmc=pmc->next) {
 		struct ip_sf_list *psf, *psf_next;
 
@@ -1046,7 +1047,7 @@ static void igmpv3_clear_delrec(struct in_device *in_dev)
 			kfree(psf);
 		}
 	}
-	read_unlock(&in_dev->lock);
+	read_unlock(&in_dev->mc_list_lock);
 }
 #endif
 
@@ -1167,10 +1168,10 @@ void ip_mc_inc_group(struct in_device *in_dev, u32 addr)
 	im->gsquery = 0;
 #endif
 	im->loaded = 0;
-	write_lock_bh(&in_dev->lock);
+	write_lock_bh(&in_dev->mc_list_lock);
 	im->next=in_dev->mc_list;
 	in_dev->mc_list=im;
-	write_unlock_bh(&in_dev->lock);
+	write_unlock_bh(&in_dev->mc_list_lock);
 #ifdef CONFIG_IP_MULTICAST
 	igmpv3_del_delrec(in_dev, im->multiaddr);
 #endif
@@ -1194,9 +1195,9 @@ void ip_mc_dec_group(struct in_device *in_dev, u32 addr)
 	for (ip=&in_dev->mc_list; (i=*ip)!=NULL; ip=&i->next) {
 		if (i->multiaddr==addr) {
 			if (--i->users == 0) {
-				write_lock_bh(&in_dev->lock);
+				write_lock_bh(&in_dev->mc_list_lock);
 				*ip = i->next;
-				write_unlock_bh(&in_dev->lock);
+				write_unlock_bh(&in_dev->mc_list_lock);
 				igmp_group_dropped(i);
 
 				if (!in_dev->dead)
@@ -1251,7 +1252,8 @@ void ip_mc_init_dev(struct in_device *in_dev)
 	in_dev->mr_qrv = IGMP_Unsolicited_Report_Count;
 #endif
 
-	in_dev->mc_lock = RW_LOCK_UNLOCKED;
+	in_dev->mc_list_lock = RW_LOCK_UNLOCKED;
+	in_dev->mc_tomb_lock = SPIN_LOCK_UNLOCKED;
 }
 
 /* Device going up */
@@ -1281,17 +1283,17 @@ void ip_mc_destroy_dev(struct in_device *in_dev)
 	/* Deactivate timers */
 	ip_mc_down(in_dev);
 
-	write_lock_bh(&in_dev->lock);
+	write_lock_bh(&in_dev->mc_list_lock);
 	while ((i = in_dev->mc_list) != NULL) {
 		in_dev->mc_list = i->next;
-		write_unlock_bh(&in_dev->lock);
+		write_unlock_bh(&in_dev->mc_list_lock);
 
 		igmp_group_dropped(i);
 		ip_ma_put(i);
 
-		write_lock_bh(&in_dev->lock);
+		write_lock_bh(&in_dev->mc_list_lock);
 	}
-	write_unlock_bh(&in_dev->lock);
+	write_unlock_bh(&in_dev->mc_list_lock);
 }
 
 static struct in_device * ip_mc_find_dev(struct ip_mreqn *imr)
@@ -1391,18 +1393,18 @@ int ip_mc_del_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
 
 	if (!in_dev)
 		return -ENODEV;
-	read_lock(&in_dev->lock);
+	read_lock(&in_dev->mc_list_lock);
 	for (pmc=in_dev->mc_list; pmc; pmc=pmc->next) {
 		if (*pmca == pmc->multiaddr)
 			break;
 	}
 	if (!pmc) {
 		/* MCA not found?? bug */
-		read_unlock(&in_dev->lock);
+		read_unlock(&in_dev->mc_list_lock);
 		return -ESRCH;
 	}
 	spin_lock_bh(&pmc->lock);
-	read_unlock(&in_dev->lock);
+	read_unlock(&in_dev->mc_list_lock);
 #ifdef CONFIG_IP_MULTICAST
 	sf_markstate(pmc);
 #endif
@@ -1527,18 +1529,18 @@ int ip_mc_add_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
 
 	if (!in_dev)
 		return -ENODEV;
-	read_lock(&in_dev->lock);
+	read_lock(&in_dev->mc_list_lock);
 	for (pmc=in_dev->mc_list; pmc; pmc=pmc->next) {
 		if (*pmca == pmc->multiaddr)
 			break;
 	}
 	if (!pmc) {
 		/* MCA not found?? bug */
-		read_unlock(&in_dev->lock);
+		read_unlock(&in_dev->mc_list_lock);
 		return -ESRCH;
 	}
 	spin_lock_bh(&pmc->lock);
-	read_unlock(&in_dev->lock);
+	read_unlock(&in_dev->mc_list_lock);
 
 #ifdef CONFIG_IP_MULTICAST
 	sf_markstate(pmc);
@@ -2095,7 +2097,7 @@ int ip_check_mc(struct in_device *in_dev, u32 mc_addr, u32 src_addr, u16 proto)
 	struct ip_sf_list *psf;
 	int rv = 0;
 
-	read_lock(&in_dev->lock);
+	read_lock(&in_dev->mc_list_lock);
 	for (im=in_dev->mc_list; im; im=im->next) {
 		if (im->multiaddr == mc_addr)
 			break;
@@ -2117,7 +2119,7 @@ int ip_check_mc(struct in_device *in_dev, u32 mc_addr, u32 src_addr, u16 proto)
 		} else
 			rv = 1; /* unspecified source; tentatively allow */
 	}
-	read_unlock(&in_dev->lock);
+	read_unlock(&in_dev->mc_list_lock);
 	return rv;
 }
 
@@ -2141,13 +2143,13 @@ static inline struct ip_mc_list *igmp_mc_get_first(struct seq_file *seq)
 		in_dev = in_dev_get(state->dev);
 		if (!in_dev)
 			continue;
-		read_lock(&in_dev->lock);
+		read_lock(&in_dev->mc_list_lock);
 		im = in_dev->mc_list;
 		if (im) {
 			state->in_dev = in_dev;
 			break;
 		}
-		read_unlock(&in_dev->lock);
+		read_unlock(&in_dev->mc_list_lock);
 		in_dev_put(in_dev);
 	}
 	return im;
@@ -2159,7 +2161,7 @@ static struct ip_mc_list *igmp_mc_get_next(struct seq_file *seq, struct ip_mc_li
 	im = im->next;
 	while (!im) {
 		if (likely(state->in_dev != NULL)) {
-			read_unlock(&state->in_dev->lock);
+			read_unlock(&state->in_dev->mc_list_lock);
 			in_dev_put(state->in_dev);
 		}
 		state->dev = state->dev->next;
@@ -2170,7 +2172,7 @@ static struct ip_mc_list *igmp_mc_get_next(struct seq_file *seq, struct ip_mc_li
 		state->in_dev = in_dev_get(state->dev);
 		if (!state->in_dev)
 			continue;
-		read_lock(&state->in_dev->lock);
+		read_lock(&state->in_dev->mc_list_lock);
 		im = state->in_dev->mc_list;
 	}
 	return im;
@@ -2206,7 +2208,7 @@ static void igmp_mc_seq_stop(struct seq_file *seq, void *v)
 {
 	struct igmp_mc_iter_state *state = igmp_mc_seq_private(seq);
 	if (likely(state->in_dev != NULL)) {
-		read_unlock(&state->in_dev->lock);
+		read_unlock(&state->in_dev->mc_list_lock);
 		in_dev_put(state->in_dev);
 		state->in_dev = NULL;
 	}
@@ -2304,7 +2306,7 @@ static inline struct ip_sf_list *igmp_mcf_get_first(struct seq_file *seq)
 		idev = in_dev_get(state->dev);
 		if (unlikely(idev == NULL))
 			continue;
-		read_lock_bh(&idev->lock);
+		read_lock(&idev->mc_list_lock);
 		im = idev->mc_list;
 		if (likely(im != NULL)) {
 			spin_lock_bh(&im->lock);
@@ -2316,7 +2318,7 @@ static inline struct ip_sf_list *igmp_mcf_get_first(struct seq_file *seq)
 			}
 			spin_unlock_bh(&im->lock);
 		}
-		read_unlock_bh(&idev->lock);
+		read_unlock(&idev->mc_list_lock);
 		in_dev_put(idev);
 	}
 	return psf;
@@ -2332,7 +2334,7 @@ static struct ip_sf_list *igmp_mcf_get_next(struct seq_file *seq, struct ip_sf_l
 		state->im = state->im->next;
 		while (!state->im) {
 			if (likely(state->idev != NULL)) {
-				read_unlock_bh(&state->idev->lock);
+				read_unlock(&state->idev->mc_list_lock);
 				in_dev_put(state->idev);
 			}
 			state->dev = state->dev->next;
@@ -2343,7 +2345,7 @@ static struct ip_sf_list *igmp_mcf_get_next(struct seq_file *seq, struct ip_sf_l
 			state->idev = in_dev_get(state->dev);
 			if (!state->idev)
 				continue;
-			read_lock_bh(&state->idev->lock);
+			read_lock(&state->idev->mc_list_lock);
 			state->im = state->idev->mc_list;
 		}
 		if (!state->im)
@@ -2389,7 +2391,7 @@ static void igmp_mcf_seq_stop(struct seq_file *seq, void *v)
 		state->im = NULL;
 	}
 	if (likely(state->idev != NULL)) {
-		read_unlock_bh(&state->idev->lock);
+		read_unlock(&state->idev->mc_list_lock);
 		in_dev_put(state->idev);
 		state->idev = NULL;
 	}

net/ipv4/ip_nat_dumb.c

-/*
- * INET		An implementation of the TCP/IP protocol suite for the LINUX
- *		operating system.  INET is implemented using the  BSD Socket
- *		interface as the means of communication with the user level.
- *
- *		Dumb Network Address Translation.
- *
- * Version:	$Id: ip_nat_dumb.c,v 1.11 2000/12/13 18:31:48 davem Exp $
- *
- * 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.
- *
- * Fixes:
- *		Rani Assaf	:	A zero checksum is a special case
- *					only in UDP
- * 		Rani Assaf	:	Added ICMP messages rewriting
- * 		Rani Assaf	:	Repaired wrong changes, made by ANK.
- *
- *
- * NOTE:	It is just working model of real NAT.
- */
-
-#include <linux/config.h>
-#include <linux/types.h>
-#include <linux/mm.h>
-#include <linux/sched.h>
-#include <linux/skbuff.h>
-#include <linux/ip.h>
-#include <linux/icmp.h>
-#include <linux/netdevice.h>
-#include <net/sock.h>
-#include <net/ip.h>
-#include <net/icmp.h>
-#include <linux/tcp.h>
-#include <linux/udp.h>
-#include <net/checksum.h>
-#include <linux/route.h>
-#include <net/route.h>
-#include <net/ip_fib.h>
-
-
-int
-ip_do_nat(struct sk_buff *skb)
-{
-	struct rtable *rt = (struct rtable*)skb->dst;
-	struct iphdr *iph = skb->nh.iph;
-	u32 odaddr = iph->daddr;
-	u32 osaddr = iph->saddr;
-	u16	check;
-
-	IPCB(skb)->flags |= IPSKB_TRANSLATED;
-
-	/* Rewrite IP header */
-	iph->daddr = rt->rt_dst_map;
-	iph->saddr = rt->rt_src_map;
-	iph->check = 0;
-	iph->check = ip_fast_csum((unsigned char *)iph, iph->ihl);
-
-	/* If it is the first fragment, rewrite protocol headers */
-
-	if (!(iph->frag_off & htons(IP_OFFSET))) {
-		u16	*cksum;
-
-		switch(iph->protocol) {
-		case IPPROTO_TCP:
-			cksum  = (u16*)&((struct tcphdr*)(((char*)iph) + (iph->ihl<<2)))->check;
-			if ((u8*)(cksum+1) > skb->tail)
-				goto truncated;
-			check = *cksum;
-			if (skb->ip_summed != CHECKSUM_HW)
-				check = ~check;
-			check = csum_tcpudp_magic(iph->saddr, iph->daddr, 0, 0, check);
-			check = csum_tcpudp_magic(~osaddr, ~odaddr, 0, 0, ~check);
-			if (skb->ip_summed == CHECKSUM_HW)
-				check = ~check;
-			*cksum = check;
-			break;
-		case IPPROTO_UDP:
-			cksum  = (u16*)&((struct udphdr*)(((char*)iph) + (iph->ihl<<2)))->check;
-			if ((u8*)(cksum+1) > skb->tail)
-				goto truncated;
-			if ((check = *cksum) != 0) {
-				check = csum_tcpudp_magic(iph->saddr, iph->daddr, 0, 0, ~check);
-				check = csum_tcpudp_magic(~osaddr, ~odaddr, 0, 0, ~check);
-				*cksum = check ? : 0xFFFF;
-			}
-			break;
-		case IPPROTO_ICMP:
-		{
-			struct icmphdr *icmph = (struct icmphdr*)((char*)iph + (iph->ihl<<2));
-			struct   iphdr *ciph;
-			u32 idaddr, isaddr;
-			int updated;
-
-			if ((icmph->type != ICMP_DEST_UNREACH) &&
-			    (icmph->type != ICMP_TIME_EXCEEDED) &&
-			    (icmph->type != ICMP_PARAMETERPROB))
-				break;
-
-			ciph = (struct iphdr *) (icmph + 1);
-
-			if ((u8*)(ciph+1) > skb->tail)
-				goto truncated;
-
-			isaddr = ciph->saddr;
-			idaddr = ciph->daddr;
-			updated = 0;
-
-			if (rt->rt_flags&RTCF_DNAT && ciph->saddr == odaddr) {
-				ciph->saddr = iph->daddr;
-				updated = 1;
-			}
-			if (rt->rt_flags&RTCF_SNAT) {
-				if (ciph->daddr != osaddr) {
-					struct   fib_result res;
-					unsigned flags = 0;
-					struct flowi fl = {
-						.iif = skb->dev->ifindex,
-						.nl_u =
-						{ .ip4_u =
-						  { .daddr = ciph->saddr,
-						    .saddr = ciph->daddr,
-#ifdef CONFIG_IP_ROUTE_TOS
-						    .tos = RT_TOS(ciph->tos)
-#endif
-						  } },
-						.proto = ciph->protocol };
-
-					/* Use fib_lookup() until we get our own
-					 * hash table of NATed hosts -- Rani
-				 	 */
-					if (fib_lookup(&fl, &res) == 0) {
-						if (res.r) {
-							ciph->daddr = fib_rules_policy(ciph->daddr, &res, &flags);
-							if (ciph->daddr != idaddr)
-								updated = 1;
-						}
-						fib_res_put(&res);
-					}
-				} else {
-					ciph->daddr = iph->saddr;
-					updated = 1;
-				}
-			}
-			if (updated) {
-				cksum  = &icmph->checksum;
-				/* Using tcpudp primitive. Why not? */
-				check  = csum_tcpudp_magic(ciph->saddr, ciph->daddr, 0, 0, ~(*cksum));
-				*cksum = csum_tcpudp_magic(~isaddr, ~idaddr, 0, 0, ~check);
-			}
-			break;
-		}
-		default:
-			break;
-		}
-	}
-	return NET_RX_SUCCESS;
-
-truncated:
-	/* should be return NET_RX_BAD; */
-	return -EINVAL;
-}

net/ipv4/route.c

@@ -1390,13 +1390,7 @@ void ip_rt_get_source(u8 *addr, struct rtable *rt)
 	if (rt->fl.iif == 0)
 		src = rt->rt_src;
 	else if (fib_lookup(&rt->fl, &res) == 0) {
-#ifdef CONFIG_IP_ROUTE_NAT
-		if (res.type == RTN_NAT)
-			src = inet_select_addr(rt->u.dst.dev, rt->rt_gateway,
-						RT_SCOPE_UNIVERSE);
-		else
-#endif
-			src = FIB_RES_PREFSRC(res);
+		src = FIB_RES_PREFSRC(res);
 		fib_res_put(&res);
 	} else
 		src = inet_select_addr(rt->u.dst.dev, rt->rt_gateway,
@@ -1500,10 +1494,6 @@ static int ip_route_input_mc(struct sk_buff *skb, u32 daddr, u32 saddr,
 #endif
 	rth->fl.fl4_src	= saddr;
 	rth->rt_src	= saddr;
-#ifdef CONFIG_IP_ROUTE_NAT
-	rth->rt_dst_map	= daddr;
-	rth->rt_src_map	= saddr;
-#endif
 #ifdef CONFIG_NET_CLS_ROUTE
 	rth->u.dst.tclassid = itag;
 #endif
@@ -1613,31 +1603,6 @@ static int ip_route_input_slow(struct sk_buff *skb, u32 daddr, u32 saddr,
 
 	RT_CACHE_STAT_INC(in_slow_tot);
 
-#ifdef CONFIG_IP_ROUTE_NAT
-	/* Policy is applied before mapping destination,
-	   but rerouting after map should be made with old source.
-	 */
-
-	if (1) {
-		u32 src_map = saddr;
-		if (res.r)
-			src_map = fib_rules_policy(saddr, &res, &flags);
-
-		if (res.type == RTN_NAT) {
-			fl.fl4_dst = fib_rules_map_destination(daddr, &res);
-			fib_res_put(&res);
-			free_res = 0;
-			if (fib_lookup(&fl, &res))
-				goto e_inval;
-			free_res = 1;
-			if (res.type != RTN_UNICAST)
-				goto e_inval;
-			flags |= RTCF_DNAT;
-		}
-		fl.fl4_src = src_map;
-	}
-#endif
-
 	if (res.type == RTN_BROADCAST)
 		goto brd_input;
 
@@ -1711,12 +1676,6 @@ static int ip_route_input_slow(struct sk_buff *skb, u32 daddr, u32 saddr,
 	rth->fl.fl4_src	= saddr;
 	rth->rt_src	= saddr;
 	rth->rt_gateway	= daddr;
-#ifdef CONFIG_IP_ROUTE_NAT
-	rth->rt_src_map	= fl.fl4_src;
-	rth->rt_dst_map	= fl.fl4_dst;
-	if (flags&RTCF_DNAT)
-		rth->rt_gateway	= fl.fl4_dst;
-#endif
 	rth->rt_iif 	=
 	rth->fl.iif	= dev->ifindex;
 	rth->u.dst.dev	= out_dev->dev;
@@ -1779,10 +1738,6 @@ out:	return err;
 #endif
 	rth->fl.fl4_src	= saddr;
 	rth->rt_src	= saddr;
-#ifdef CONFIG_IP_ROUTE_NAT
-	rth->rt_dst_map	= fl.fl4_dst;
-	rth->rt_src_map	= fl.fl4_src;
-#endif
 #ifdef CONFIG_NET_CLS_ROUTE
 	rth->u.dst.tclassid = itag;
 #endif
@@ -1903,7 +1858,7 @@ int ip_route_input(struct sk_buff *skb, u32 daddr, u32 saddr,
 	if (MULTICAST(daddr)) {
 		struct in_device *in_dev;
 
-		read_lock(&inetdev_lock);
+		rcu_read_lock();
 		if ((in_dev = __in_dev_get(dev)) != NULL) {
 			int our = ip_check_mc(in_dev, daddr, saddr,
 				skb->nh.iph->protocol);
@@ -1912,12 +1867,12 @@ int ip_route_input(struct sk_buff *skb, u32 daddr, u32 saddr,
 			    || (!LOCAL_MCAST(daddr) && IN_DEV_MFORWARD(in_dev))
 #endif
 			    ) {
-				read_unlock(&inetdev_lock);
+				rcu_read_unlock();
 				return ip_route_input_mc(skb, daddr, saddr,
 							 tos, dev, our);
 			}
 		}
-		read_unlock(&inetdev_lock);
+		rcu_read_unlock();
 		return -EINVAL;
 	}
 	return ip_route_input_slow(skb, daddr, saddr, tos, dev);
@@ -2075,9 +2030,6 @@ static int ip_route_output_slow(struct rtable **rp, const struct flowi *oldflp)
 	}
 	free_res = 1;
 
-	if (res.type == RTN_NAT)
-		goto e_inval;
-
 	if (res.type == RTN_LOCAL) {
 		if (!fl.fl4_src)
 			fl.fl4_src = fl.fl4_dst;
@@ -2167,10 +2119,6 @@ static int ip_route_output_slow(struct rtable **rp, const struct flowi *oldflp)
 #endif
 	rth->rt_dst	= fl.fl4_dst;
 	rth->rt_src	= fl.fl4_src;
-#ifdef CONFIG_IP_ROUTE_NAT
-	rth->rt_dst_map	= fl.fl4_dst;
-	rth->rt_src_map	= fl.fl4_src;
-#endif
 	rth->rt_iif	= oldflp->oif ? : dev_out->ifindex;
 	rth->u.dst.dev	= dev_out;
 	dev_hold(dev_out);

net/irda/irlan/irlan_eth.c

@@ -303,10 +303,10 @@ void irlan_eth_send_gratuitous_arp(struct net_device *dev)
 	 */
 #ifdef CONFIG_INET
 	IRDA_DEBUG(4, "IrLAN: Sending gratuitous ARP\n");
-	in_dev = in_dev_get(dev);
+	rcu_read_lock();
+	in_dev = __in_dev_get(dev);
 	if (in_dev == NULL)
-		return;
-	read_lock(&in_dev->lock);
+		goto out;
 	if (in_dev->ifa_list)
 		
 	arp_send(ARPOP_REQUEST, ETH_P_ARP, 
@@ -314,8 +314,8 @@ void irlan_eth_send_gratuitous_arp(struct net_device *dev)
 		 dev, 
 		 in_dev->ifa_list->ifa_address,
 		 NULL, dev->dev_addr, NULL);
-	read_unlock(&in_dev->lock);
-	in_dev_put(in_dev);
+out:
+	rcu_read_unlock();
 #endif /* CONFIG_INET */
 }
 

net/sctp/protocol.c

@@ -148,13 +148,12 @@ static void sctp_v4_copy_addrlist(struct list_head *addrlist,
 	struct in_ifaddr *ifa;
 	struct sctp_sockaddr_entry *addr;
 
-	read_lock(&inetdev_lock);
+	rcu_read_lock();
 	if ((in_dev = __in_dev_get(dev)) == NULL) {
-		read_unlock(&inetdev_lock);
+		rcu_read_unlock();
 		return;
 	}
 
-	read_lock(&in_dev->lock);
 	for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next) {
 		/* Add the address to the local list.  */
 		addr = t_new(struct sctp_sockaddr_entry, GFP_ATOMIC);
@@ -166,8 +165,7 @@ static void sctp_v4_copy_addrlist(struct list_head *addrlist,
 		}
 	}
 
-	read_unlock(&in_dev->lock);
-	read_unlock(&inetdev_lock);
+	rcu_read_unlock();
 }
 
 /* Extract our IP addresses from the system and stash them in the