X-Git-Url: https://err.no/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=net%2Fipv4%2Fip_fragment.c;h=a2e92f9709db8459bd573e82c6842b6d9ea1d715;hb=7524d7d6de5d5d3f081de8cf5479819fad339661;hp=443b3f89192f5841f78e7783874f9c8b9b479020;hpb=e457f790d8b05977853aa238bbc667b3bb375671;p=linux-2.6

diff --git a/net/ipv4/ip_fragment.c b/net/ipv4/ip_fragment.c
index 443b3f8919..a2e92f9709 100644
--- a/net/ipv4/ip_fragment.c
+++ b/net/ipv4/ip_fragment.c
@@ -50,7 +50,7 @@
  * as well. Or notify me, at least. --ANK
  */
 
-int sysctl_ipfrag_max_dist __read_mostly = 64;
+static int sysctl_ipfrag_max_dist __read_mostly = 64;
 
 struct ipfrag_skb_cb
 {
@@ -74,40 +74,26 @@ struct ipq {
 	struct inet_peer *peer;
 };
 
-struct inet_frags_ctl ip4_frags_ctl __read_mostly = {
-	/*
-	 * Fragment cache limits. We will commit 256K at one time. Should we
-	 * cross that limit we will prune down to 192K. This should cope with
-	 * even the most extreme cases without allowing an attacker to
-	 * measurably harm machine performance.
-	 */
-	.high_thresh	 = 256 * 1024,
-	.low_thresh	 = 192 * 1024,
-
-	/*
-	 * Important NOTE! Fragment queue must be destroyed before MSL expires.
-	 * RFC791 is wrong proposing to prolongate timer each fragment arrival
-	 * by TTL.
-	 */
-	.timeout	 = IP_FRAG_TIME,
-	.secret_interval = 10 * 60 * HZ,
-};
-
 static struct inet_frags ip4_frags;
 
-int ip_frag_nqueues(void)
+int ip_frag_nqueues(struct net *net)
 {
-	return ip4_frags.nqueues;
+	return net->ipv4.frags.nqueues;
 }
 
-int ip_frag_mem(void)
+int ip_frag_mem(struct net *net)
 {
-	return atomic_read(&ip4_frags.mem);
+	return atomic_read(&net->ipv4.frags.mem);
 }
 
 static int ip_frag_reasm(struct ipq *qp, struct sk_buff *prev,
 			 struct net_device *dev);
 
+struct ip4_create_arg {
+	struct iphdr *iph;
+	u32 user;
+};
+
 static unsigned int ipqhashfn(__be16 id, __be32 saddr, __be32 daddr, u8 prot)
 {
 	return jhash_3words((__force u32)id << 16 | prot,
@@ -123,15 +109,43 @@ static unsigned int ip4_hashfn(struct inet_frag_queue *q)
 	return ipqhashfn(ipq->id, ipq->saddr, ipq->daddr, ipq->protocol);
 }
 
+static int ip4_frag_match(struct inet_frag_queue *q, void *a)
+{
+	struct ipq *qp;
+	struct ip4_create_arg *arg = a;
+
+	qp = container_of(q, struct ipq, q);
+	return (qp->id == arg->iph->id &&
+			qp->saddr == arg->iph->saddr &&
+			qp->daddr == arg->iph->daddr &&
+			qp->protocol == arg->iph->protocol &&
+			qp->user == arg->user);
+}
+
 /* Memory Tracking Functions. */
-static __inline__ void frag_kfree_skb(struct sk_buff *skb, int *work)
+static __inline__ void frag_kfree_skb(struct netns_frags *nf,
+		struct sk_buff *skb, int *work)
 {
 	if (work)
 		*work -= skb->truesize;
-	atomic_sub(skb->truesize, &ip4_frags.mem);
+	atomic_sub(skb->truesize, &nf->mem);
 	kfree_skb(skb);
 }
 
+static void ip4_frag_init(struct inet_frag_queue *q, void *a)
+{
+	struct ipq *qp = container_of(q, struct ipq, q);
+	struct ip4_create_arg *arg = a;
+
+	qp->protocol = arg->iph->protocol;
+	qp->id = arg->iph->id;
+	qp->saddr = arg->iph->saddr;
+	qp->daddr = arg->iph->daddr;
+	qp->user = arg->user;
+	qp->peer = sysctl_ipfrag_max_dist ?
+		inet_getpeer(arg->iph->saddr, 1) : NULL;
+}
+
 static __inline__ void ip4_frag_free(struct inet_frag_queue *q)
 {
 	struct ipq *qp;
@@ -139,17 +153,6 @@ static __inline__ void ip4_frag_free(struct inet_frag_queue *q)
 	qp = container_of(q, struct ipq, q);
 	if (qp->peer)
 		inet_putpeer(qp->peer);
-	kfree(qp);
-}
-
-static __inline__ struct ipq *frag_alloc_queue(void)
-{
-	struct ipq *qp = kzalloc(sizeof(struct ipq), GFP_ATOMIC);
-
-	if (!qp)
-		return NULL;
-	atomic_add(sizeof(struct ipq), &ip4_frags.mem);
-	return qp;
 }
 
 
@@ -171,11 +174,11 @@ static void ipq_kill(struct ipq *ipq)
 /* Memory limiting on fragments.  Evictor trashes the oldest
  * fragment queue until we are back under the threshold.
  */
-static void ip_evictor(void)
+static void ip_evictor(struct net *net)
 {
 	int evicted;
 
-	evicted = inet_frag_evictor(&ip4_frags);
+	evicted = inet_frag_evictor(&net->ipv4.frags, &ip4_frags);
 	if (evicted)
 		IP_ADD_STATS_BH(IPSTATS_MIB_REASMFAILS, evicted);
 }
@@ -185,7 +188,9 @@ static void ip_evictor(void)
  */
 static void ip_expire(unsigned long arg)
 {
-	struct ipq *qp = (struct ipq *) arg;
+	struct ipq *qp;
+
+	qp = container_of((struct inet_frag_queue *) arg, struct ipq, q);
 
 	spin_lock(&qp->q.lock);
 
@@ -210,112 +215,30 @@ out:
 	ipq_put(qp);
 }
 
-/* Creation primitives. */
-
-static struct ipq *ip_frag_intern(struct ipq *qp_in)
+/* Find the correct entry in the "incomplete datagrams" queue for
+ * this IP datagram, and create new one, if nothing is found.
+ */
+static inline struct ipq *ip_find(struct net *net, struct iphdr *iph, u32 user)
 {
-	struct ipq *qp;
-#ifdef CONFIG_SMP
-	struct hlist_node *n;
-#endif
+	struct inet_frag_queue *q;
+	struct ip4_create_arg arg;
 	unsigned int hash;
 
-	write_lock(&ip4_frags.lock);
-	hash = ipqhashfn(qp_in->id, qp_in->saddr, qp_in->daddr,
-			 qp_in->protocol);
-#ifdef CONFIG_SMP
-	/* With SMP race we have to recheck hash table, because
-	 * such entry could be created on other cpu, while we
-	 * promoted read lock to write lock.
-	 */
-	hlist_for_each_entry(qp, n, &ip4_frags.hash[hash], q.list) {
-		if (qp->id == qp_in->id		&&
-		    qp->saddr == qp_in->saddr	&&
-		    qp->daddr == qp_in->daddr	&&
-		    qp->protocol == qp_in->protocol &&
-		    qp->user == qp_in->user) {
-			atomic_inc(&qp->q.refcnt);
-			write_unlock(&ip4_frags.lock);
-			qp_in->q.last_in |= COMPLETE;
-			ipq_put(qp_in);
-			return qp;
-		}
-	}
-#endif
-	qp = qp_in;
-
-	if (!mod_timer(&qp->q.timer, jiffies + ip4_frags_ctl.timeout))
-		atomic_inc(&qp->q.refcnt);
+	arg.iph = iph;
+	arg.user = user;
+	hash = ipqhashfn(iph->id, iph->saddr, iph->daddr, iph->protocol);
 
-	atomic_inc(&qp->q.refcnt);
-	hlist_add_head(&qp->q.list, &ip4_frags.hash[hash]);
-	INIT_LIST_HEAD(&qp->q.lru_list);
-	list_add_tail(&qp->q.lru_list, &ip4_frags.lru_list);
-	ip4_frags.nqueues++;
-	write_unlock(&ip4_frags.lock);
-	return qp;
-}
-
-/* Add an entry to the 'ipq' queue for a newly received IP datagram. */
-static struct ipq *ip_frag_create(struct iphdr *iph, u32 user)
-{
-	struct ipq *qp;
-
-	if ((qp = frag_alloc_queue()) == NULL)
+	q = inet_frag_find(&net->ipv4.frags, &ip4_frags, &arg, hash);
+	if (q == NULL)
 		goto out_nomem;
 
-	qp->protocol = iph->protocol;
-	qp->id = iph->id;
-	qp->saddr = iph->saddr;
-	qp->daddr = iph->daddr;
-	qp->user = user;
-	qp->peer = sysctl_ipfrag_max_dist ? inet_getpeer(iph->saddr, 1) : NULL;
-
-	/* Initialize a timer for this entry. */
-	init_timer(&qp->q.timer);
-	qp->q.timer.data = (unsigned long) qp;	/* pointer to queue	*/
-	qp->q.timer.function = ip_expire;		/* expire function	*/
-	spin_lock_init(&qp->q.lock);
-	atomic_set(&qp->q.refcnt, 1);
-
-	return ip_frag_intern(qp);
+	return container_of(q, struct ipq, q);
 
 out_nomem:
 	LIMIT_NETDEBUG(KERN_ERR "ip_frag_create: no memory left !\n");
 	return NULL;
 }
 
-/* Find the correct entry in the "incomplete datagrams" queue for
- * this IP datagram, and create new one, if nothing is found.
- */
-static inline struct ipq *ip_find(struct iphdr *iph, u32 user)
-{
-	__be16 id = iph->id;
-	__be32 saddr = iph->saddr;
-	__be32 daddr = iph->daddr;
-	__u8 protocol = iph->protocol;
-	unsigned int hash;
-	struct ipq *qp;
-	struct hlist_node *n;
-
-	read_lock(&ip4_frags.lock);
-	hash = ipqhashfn(id, saddr, daddr, protocol);
-	hlist_for_each_entry(qp, n, &ip4_frags.hash[hash], q.list) {
-		if (qp->id == id		&&
-		    qp->saddr == saddr	&&
-		    qp->daddr == daddr	&&
-		    qp->protocol == protocol &&
-		    qp->user == user) {
-			atomic_inc(&qp->q.refcnt);
-			read_unlock(&ip4_frags.lock);
-			return qp;
-		}
-	}
-	read_unlock(&ip4_frags.lock);
-
-	return ip_frag_create(iph, user);
-}
-
 /* Is the fragment too far ahead to be part of ipq? */
 static inline int ip_frag_too_far(struct ipq *qp)
 {
@@ -345,7 +268,7 @@ static int ip_frag_reinit(struct ipq *qp)
 {
 	struct sk_buff *fp;
 
-	if (!mod_timer(&qp->q.timer, jiffies + ip4_frags_ctl.timeout)) {
+	if (!mod_timer(&qp->q.timer, jiffies + qp->q.net->timeout)) {
 		atomic_inc(&qp->q.refcnt);
 		return -ETIMEDOUT;
 	}
@@ -353,7 +276,7 @@ static int ip_frag_reinit(struct ipq *qp)
 	fp = qp->q.fragments;
 	do {
 		struct sk_buff *xp = fp->next;
-		frag_kfree_skb(fp, NULL);
+		frag_kfree_skb(qp->q.net, fp, NULL);
 		fp = xp;
 	} while (fp);
 
@@ -490,7 +413,7 @@ static int ip_frag_queue(struct ipq *qp, struct sk_buff *skb)
 				qp->q.fragments = next;
 
 			qp->q.meat -= free_it->len;
-			frag_kfree_skb(free_it, NULL);
+			frag_kfree_skb(qp->q.net, free_it, NULL);
 		}
 	}
 
@@ -510,7 +433,7 @@ static int ip_frag_queue(struct ipq *qp, struct sk_buff *skb)
 	}
 	qp->q.stamp = skb->tstamp;
 	qp->q.meat += skb->len;
-	atomic_add(skb->truesize, &ip4_frags.mem);
+	atomic_add(skb->truesize, &qp->q.net->mem);
 	if (offset == 0)
 		qp->q.last_in |= FIRST_IN;
 
@@ -518,7 +441,7 @@ static int ip_frag_queue(struct ipq *qp, struct sk_buff *skb)
 		return ip_frag_reasm(qp, prev, dev);
 
 	write_lock(&ip4_frags.lock);
-	list_move_tail(&qp->q.lru_list, &ip4_frags.lru_list);
+	list_move_tail(&qp->q.lru_list, &qp->q.net->lru_list);
 	write_unlock(&ip4_frags.lock);
 	return -EINPROGRESS;
 
@@ -545,7 +468,6 @@ static int ip_frag_reasm(struct ipq *qp, struct sk_buff *prev,
 	if (prev) {
 		head = prev->next;
 		fp = skb_clone(head, GFP_ATOMIC);
-
 		if (!fp)
 			goto out_nomem;
 
@@ -571,7 +493,6 @@ static int ip_frag_reasm(struct ipq *qp, struct sk_buff *prev,
 		goto out_oversize;
 
 	/* Head of list must not be cloned. */
-	err = -ENOMEM;
 	if (skb_cloned(head) && pskb_expand_head(head, 0, 0, GFP_ATOMIC))
 		goto out_nomem;
 
@@ -595,12 +516,12 @@ static int ip_frag_reasm(struct ipq *qp, struct sk_buff *prev,
 		head->len -= clone->len;
 		clone->csum = 0;
 		clone->ip_summed = head->ip_summed;
-		atomic_add(clone->truesize, &ip4_frags.mem);
+		atomic_add(clone->truesize, &qp->q.net->mem);
 	}
 
 	skb_shinfo(head)->frag_list = head->next;
 	skb_push(head, head->data - skb_network_header(head));
-	atomic_sub(head->truesize, &ip4_frags.mem);
+	atomic_sub(head->truesize, &qp->q.net->mem);
 
 	for (fp=head->next; fp; fp = fp->next) {
 		head->data_len += fp->len;
@@ -610,7 +531,7 @@ static int ip_frag_reasm(struct ipq *qp, struct sk_buff *prev,
 		else if (head->ip_summed == CHECKSUM_COMPLETE)
 			head->csum = csum_add(head->csum, fp->csum);
 		head->truesize += fp->truesize;
-		atomic_sub(fp->truesize, &ip4_frags.mem);
+		atomic_sub(fp->truesize, &qp->q.net->mem);
 	}
 
 	head->next = NULL;
@@ -627,6 +548,7 @@ static int ip_frag_reasm(struct ipq *qp, struct sk_buff *prev,
 out_nomem:
 	LIMIT_NETDEBUG(KERN_ERR "IP: queue_glue: no memory for gluing "
 			      "queue %p\n", qp);
+	err = -ENOMEM;
 	goto out_fail;
 out_oversize:
 	if (net_ratelimit())
@@ -642,15 +564,17 @@ out_fail:
 int ip_defrag(struct sk_buff *skb, u32 user)
 {
 	struct ipq *qp;
+	struct net *net;
 
 	IP_INC_STATS_BH(IPSTATS_MIB_REASMREQDS);
 
+	net = skb->dev->nd_net;
 	/* Start by cleaning up the memory. */
-	if (atomic_read(&ip4_frags.mem) > ip4_frags_ctl.high_thresh)
-		ip_evictor();
+	if (atomic_read(&net->ipv4.frags.mem) > net->ipv4.frags.high_thresh)
+		ip_evictor(net);
 
 	/* Lookup (or create) queue header */
-	if ((qp = ip_find(ip_hdr(skb), user)) != NULL) {
+	if ((qp = ip_find(net, ip_hdr(skb), user)) != NULL) {
 		int ret;
 
 		spin_lock(&qp->q.lock);
@@ -667,13 +591,150 @@ int ip_defrag(struct sk_buff *skb, u32 user)
 	return -ENOMEM;
 }
 
+#ifdef CONFIG_SYSCTL
+static int zero;
+
+static struct ctl_table ip4_frags_ctl_table[] = {
+	{
+		.ctl_name	= NET_IPV4_IPFRAG_HIGH_THRESH,
+		.procname	= "ipfrag_high_thresh",
+		.data		= &init_net.ipv4.frags.high_thresh,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec
+	},
+	{
+		.ctl_name	= NET_IPV4_IPFRAG_LOW_THRESH,
+		.procname	= "ipfrag_low_thresh",
+		.data		= &init_net.ipv4.frags.low_thresh,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec
+	},
+	{
+		.ctl_name	= NET_IPV4_IPFRAG_TIME,
+		.procname	= "ipfrag_time",
+		.data		= &init_net.ipv4.frags.timeout,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec_jiffies,
+		.strategy	= &sysctl_jiffies
+	},
+	{
+		.ctl_name	= NET_IPV4_IPFRAG_SECRET_INTERVAL,
+		.procname	= "ipfrag_secret_interval",
+		.data		= &ip4_frags.secret_interval,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec_jiffies,
+		.strategy	= &sysctl_jiffies
+	},
+	{
+		.procname	= "ipfrag_max_dist",
+		.data		= &sysctl_ipfrag_max_dist,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec_minmax,
+		.extra1		= &zero
+	},
+	{ }
+};
+
+static int ip4_frags_ctl_register(struct net *net)
+{
+	struct ctl_table *table;
+	struct ctl_table_header *hdr;
+
+	table = ip4_frags_ctl_table;
+	if (net != &init_net) {
+		table = kmemdup(table, sizeof(ip4_frags_ctl_table), GFP_KERNEL);
+		if (table == NULL)
+			goto err_alloc;
+
+		table[0].data = &net->ipv4.frags.high_thresh;
+		table[1].data = &net->ipv4.frags.low_thresh;
+		table[2].data = &net->ipv4.frags.timeout;
+		table[3].mode &= ~0222;
+		table[4].mode &= ~0222;
+	}
+
+	hdr = register_net_sysctl_table(net, net_ipv4_ctl_path, table);
+	if (hdr == NULL)
+		goto err_reg;
+
+	net->ipv4.frags_hdr = hdr;
+	return 0;
+
+err_reg:
+	if (net != &init_net)
+		kfree(table);
+err_alloc:
+	return -ENOMEM;
+}
+
+static void ip4_frags_ctl_unregister(struct net *net)
+{
+	struct ctl_table *table;
+
+	table = net->ipv4.frags_hdr->ctl_table_arg;
+	unregister_net_sysctl_table(net->ipv4.frags_hdr);
+	kfree(table);
+}
+#else
+static inline int ip4_frags_ctl_register(struct net *net)
+{
+	return 0;
+}
+
+static inline void ip4_frags_ctl_unregister(struct net *net)
+{
+}
+#endif
+
+static int ipv4_frags_init_net(struct net *net)
+{
+	/*
+	 * Fragment cache limits. We will commit 256K at one time. Should we
+	 * cross that limit we will prune down to 192K. This should cope with
+	 * even the most extreme cases without allowing an attacker to
+	 * measurably harm machine performance.
+	 */
+	net->ipv4.frags.high_thresh = 256 * 1024;
+	net->ipv4.frags.low_thresh = 192 * 1024;
+	/*
+	 * Important NOTE! Fragment queue must be destroyed before MSL expires.
+	 * RFC791 is wrong proposing to prolongate timer each fragment arrival
+	 * by TTL.
+	 */
+	net->ipv4.frags.timeout = IP_FRAG_TIME;
+
+	inet_frags_init_net(&net->ipv4.frags);
+
+	return ip4_frags_ctl_register(net);
+}
+
+static void ipv4_frags_exit_net(struct net *net)
+{
+	ip4_frags_ctl_unregister(net);
+	inet_frags_exit_net(&net->ipv4.frags, &ip4_frags);
+}
+
+static struct pernet_operations ip4_frags_ops = {
+	.init = ipv4_frags_init_net,
+	.exit = ipv4_frags_exit_net,
+};
+
 void __init ipfrag_init(void)
 {
-	ip4_frags.ctl = &ip4_frags_ctl;
+	register_pernet_subsys(&ip4_frags_ops);
 	ip4_frags.hashfn = ip4_hashfn;
+	ip4_frags.constructor = ip4_frag_init;
 	ip4_frags.destructor = ip4_frag_free;
 	ip4_frags.skb_free = NULL;
 	ip4_frags.qsize = sizeof(struct ipq);
+	ip4_frags.match = ip4_frag_match;
+	ip4_frags.frag_expire = ip_expire;
+	ip4_frags.secret_interval = 10 * 60 * HZ;
 	inet_frags_init(&ip4_frags);
 }