1 files changed, 183 insertions, 62 deletions

diff --git a/net/netfilter/xt_hashlimit.c b/net/netfilter/xt_hashlimit.c
index 9228ee0dc11..a3910fc2122 100644
--- a/net/netfilter/xt_hashlimit.c
+++ b/net/netfilter/xt_hashlimit.c
@@ -3,6 +3,7 @@
  *	separately for each hashbucket (sourceip/sourceport/dstip/dstport)
  *
  *	(C) 2003-2004 by Harald Welte <laforge@netfilter.org>
+ *	(C) 2006-2012 Patrick McHardy <kaber@trash.net>
  *	Copyright © CC Computer Consultants GmbH, 2007 - 2008
  *
  * Development of this code was funded by Astaro AG, http://www.astaro.com/
@@ -21,7 +22,7 @@
 #include <linux/mm.h>
 #include <linux/in.h>
 #include <linux/ip.h>
-#if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE)
+#if IS_ENABLED(CONFIG_IP6_NF_IPTABLES)
 #include <linux/ipv6.h>
 #include <net/ipv6.h>
 #endif
@@ -64,7 +65,7 @@ struct dsthash_dst {
 			__be32 src;
 			__be32 dst;
 		} ip;
-#if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE)
+#if IS_ENABLED(CONFIG_IP6_NF_IPTABLES)
 		struct {
 			__be32 src[4];
 			__be32 dst[4];
@@ -107,6 +108,7 @@ struct xt_hashlimit_htable {
 
 	/* seq_file stuff */
 	struct proc_dir_entry *pde;
+	const char *name;
 	struct net *net;
 
 	struct hlist_head hash[0];	/* hashtable itself */
@@ -141,11 +143,10 @@ dsthash_find(const struct xt_hashlimit_htable *ht,
 	     const struct dsthash_dst *dst)
 {
 	struct dsthash_ent *ent;
-	struct hlist_node *pos;
 	u_int32_t hash = hash_dst(ht, dst);
 
 	if (!hlist_empty(&ht->hash[hash])) {
-		hlist_for_each_entry_rcu(ent, pos, &ht->hash[hash], node)
+		hlist_for_each_entry_rcu(ent, &ht->hash[hash], node)
 			if (dst_cmp(ent, dst)) {
 				spin_lock(&ent->lock);
 				return ent;
@@ -157,11 +158,22 @@ dsthash_find(const struct xt_hashlimit_htable *ht,
 /* allocate dsthash_ent, initialize dst, put in htable and lock it */
 static struct dsthash_ent *
 dsthash_alloc_init(struct xt_hashlimit_htable *ht,
-		   const struct dsthash_dst *dst)
+		   const struct dsthash_dst *dst, bool *race)
 {
 	struct dsthash_ent *ent;
 
 	spin_lock(&ht->lock);
+
+	/* Two or more packets may race to create the same entry in the
+	 * hashtable, double check if this packet lost race.
+	 */
+	ent = dsthash_find(ht, dst);
+	if (ent != NULL) {
+		spin_unlock(&ht->lock);
+		*race = true;
+		return ent;
+	}
+
 	/* initialize hash with random val at the time we allocate
 	 * the first hashtable entry */
 	if (unlikely(!ht->rnd_initialized)) {
@@ -171,15 +183,11 @@ dsthash_alloc_init(struct xt_hashlimit_htable *ht,
 
 	if (ht->cfg.max && ht->count >= ht->cfg.max) {
 		/* FIXME: do something. question is what.. */
-		if (net_ratelimit())
-			pr_err("max count of %u reached\n", ht->cfg.max);
+		net_err_ratelimited("max count of %u reached\n", ht->cfg.max);
 		ent = NULL;
 	} else
 		ent = kmem_cache_alloc(hashlimit_cachep, GFP_ATOMIC);
-	if (!ent) {
-		if (net_ratelimit())
-			pr_err("cannot allocate dsthash_ent\n");
-	} else {
+	if (ent) {
 		memcpy(&ent->dst, dst, sizeof(ent->dst));
 		spin_lock_init(&ent->lock);
 
@@ -247,6 +255,11 @@ static int htable_create(struct net *net, struct xt_hashlimit_mtinfo1 *minfo,
 	hinfo->count = 0;
 	hinfo->family = family;
 	hinfo->rnd_initialized = false;
+	hinfo->name = kstrdup(minfo->name, GFP_KERNEL);
+	if (!hinfo->name) {
+		vfree(hinfo);
+		return -ENOMEM;
+	}
 	spin_lock_init(&hinfo->lock);
 
 	hinfo->pde = proc_create_data(minfo->name, 0,
@@ -254,6 +267,7 @@ static int htable_create(struct net *net, struct xt_hashlimit_mtinfo1 *minfo,
 		hashlimit_net->ipt_hashlimit : hashlimit_net->ip6t_hashlimit,
 		&dl_file_ops, hinfo);
 	if (hinfo->pde == NULL) {
+		kfree(hinfo->name);
 		vfree(hinfo);
 		return -ENOMEM;
 	}
@@ -290,8 +304,8 @@ static void htable_selective_cleanup(struct xt_hashlimit_htable *ht,
 	spin_lock_bh(&ht->lock);
 	for (i = 0; i < ht->cfg.size; i++) {
 		struct dsthash_ent *dh;
-		struct hlist_node *pos, *n;
-		hlist_for_each_entry_safe(dh, pos, n, &ht->hash[i], node) {
+		struct hlist_node *n;
+		hlist_for_each_entry_safe(dh, n, &ht->hash[i], node) {
 			if ((*select)(ht, dh))
 				dsthash_free(ht, dh);
 		}
@@ -311,19 +325,26 @@ static void htable_gc(unsigned long htlong)
 	add_timer(&ht->timer);
 }
 
-static void htable_destroy(struct xt_hashlimit_htable *hinfo)
+static void htable_remove_proc_entry(struct xt_hashlimit_htable *hinfo)
 {
 	struct hashlimit_net *hashlimit_net = hashlimit_pernet(hinfo->net);
 	struct proc_dir_entry *parent;
 
-	del_timer_sync(&hinfo->timer);
-
 	if (hinfo->family == NFPROTO_IPV4)
 		parent = hashlimit_net->ipt_hashlimit;
 	else
 		parent = hashlimit_net->ip6t_hashlimit;
-	remove_proc_entry(hinfo->pde->name, parent);
+
+	if (parent != NULL)
+		remove_proc_entry(hinfo->name, parent);
+}
+
+static void htable_destroy(struct xt_hashlimit_htable *hinfo)
+{
+	del_timer_sync(&hinfo->timer);
+	htable_remove_proc_entry(hinfo);
 	htable_selective_cleanup(hinfo, select_all);
+	kfree(hinfo->name);
 	vfree(hinfo);
 }
 
@@ -333,10 +354,9 @@ static struct xt_hashlimit_htable *htable_find_get(struct net *net,
 {
 	struct hashlimit_net *hashlimit_net = hashlimit_pernet(net);
 	struct xt_hashlimit_htable *hinfo;
-	struct hlist_node *pos;
 
-	hlist_for_each_entry(hinfo, pos, &hashlimit_net->htables, node) {
-		if (!strcmp(name, hinfo->pde->name) &&
+	hlist_for_each_entry(hinfo, &hashlimit_net->htables, node) {
+		if (!strcmp(name, hinfo->name) &&
 		    hinfo->family == family) {
 			hinfo->use++;
 			return hinfo;
@@ -391,9 +411,20 @@ static void htable_put(struct xt_hashlimit_htable *hinfo)
 
 #define CREDITS_PER_JIFFY POW2_BELOW32(MAX_CPJ)
 
+/* in byte mode, the lowest possible rate is one packet/second.
+ * credit_cap is used as a counter that tells us how many times we can
+ * refill the "credits available" counter when it becomes empty.
+ */
+#define MAX_CPJ_BYTES (0xFFFFFFFF / HZ)
+#define CREDITS_PER_JIFFY_BYTES POW2_BELOW32(MAX_CPJ_BYTES)
+
+static u32 xt_hashlimit_len_to_chunks(u32 len)
+{
+	return (len >> XT_HASHLIMIT_BYTE_SHIFT) + 1;
+}
+
 /* Precision saver. */
-static inline u_int32_t
-user2credits(u_int32_t user)
+static u32 user2credits(u32 user)
 {
 	/* If multiplying would overflow... */
 	if (user > 0xFFFFFFFF / (HZ*CREDITS_PER_JIFFY))
@@ -403,12 +434,53 @@ user2credits(u_int32_t user)
 	return (user * HZ * CREDITS_PER_JIFFY) / XT_HASHLIMIT_SCALE;
 }
 
-static inline void rateinfo_recalc(struct dsthash_ent *dh, unsigned long now)
+static u32 user2credits_byte(u32 user)
 {
-	dh->rateinfo.credit += (now - dh->rateinfo.prev) * CREDITS_PER_JIFFY;
-	if (dh->rateinfo.credit > dh->rateinfo.credit_cap)
-		dh->rateinfo.credit = dh->rateinfo.credit_cap;
+	u64 us = user;
+	us *= HZ * CREDITS_PER_JIFFY_BYTES;
+	return (u32) (us >> 32);
+}
+
+static void rateinfo_recalc(struct dsthash_ent *dh, unsigned long now, u32 mode)
+{
+	unsigned long delta = now - dh->rateinfo.prev;
+	u32 cap;
+
+	if (delta == 0)
+		return;
+
 	dh->rateinfo.prev = now;
+
+	if (mode & XT_HASHLIMIT_BYTES) {
+		u32 tmp = dh->rateinfo.credit;
+		dh->rateinfo.credit += CREDITS_PER_JIFFY_BYTES * delta;
+		cap = CREDITS_PER_JIFFY_BYTES * HZ;
+		if (tmp >= dh->rateinfo.credit) {/* overflow */
+			dh->rateinfo.credit = cap;
+			return;
+		}
+	} else {
+		dh->rateinfo.credit += delta * CREDITS_PER_JIFFY;
+		cap = dh->rateinfo.credit_cap;
+	}
+	if (dh->rateinfo.credit > cap)
+		dh->rateinfo.credit = cap;
+}
+
+static void rateinfo_init(struct dsthash_ent *dh,
+			  struct xt_hashlimit_htable *hinfo)
+{
+	dh->rateinfo.prev = jiffies;
+	if (hinfo->cfg.mode & XT_HASHLIMIT_BYTES) {
+		dh->rateinfo.credit = CREDITS_PER_JIFFY_BYTES * HZ;
+		dh->rateinfo.cost = user2credits_byte(hinfo->cfg.avg);
+		dh->rateinfo.credit_cap = hinfo->cfg.burst;
+	} else {
+		dh->rateinfo.credit = user2credits(hinfo->cfg.avg *
+						   hinfo->cfg.burst);
+		dh->rateinfo.cost = user2credits(hinfo->cfg.avg);
+		dh->rateinfo.credit_cap = dh->rateinfo.credit;
+	}
 }
 
 static inline __be32 maskl(__be32 a, unsigned int l)
@@ -416,7 +488,7 @@ static inline __be32 maskl(__be32 a, unsigned int l)
 	return l ? htonl(ntohl(a) & ~0 << (32 - l)) : 0;
 }
 
-#if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE)
+#if IS_ENABLED(CONFIG_IP6_NF_IPTABLES)
 static void hashlimit_ipv6_mask(__be32 *i, unsigned int p)
 {
 	switch (p) {
@@ -466,8 +538,11 @@ hashlimit_init_dst(const struct xt_hashlimit_htable *hinfo,
 			return 0;
 		nexthdr = ip_hdr(skb)->protocol;
 		break;
-#if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE)
+#if IS_ENABLED(CONFIG_IP6_NF_IPTABLES)
 	case NFPROTO_IPV6:
+	{
+		__be16 frag_off;
+
 		if (hinfo->cfg.mode & XT_HASHLIMIT_HASH_DIP) {
 			memcpy(&dst->ip6.dst, &ipv6_hdr(skb)->daddr,
 			       sizeof(dst->ip6.dst));
@@ -483,10 +558,11 @@ hashlimit_init_dst(const struct xt_hashlimit_htable *hinfo,
 		      (XT_HASHLIMIT_HASH_DPT | XT_HASHLIMIT_HASH_SPT)))
 			return 0;
 		nexthdr = ipv6_hdr(skb)->nexthdr;
-		protoff = ipv6_skip_exthdr(skb, sizeof(struct ipv6hdr), &nexthdr);
+		protoff = ipv6_skip_exthdr(skb, sizeof(struct ipv6hdr), &nexthdr, &frag_off);
 		if ((int)protoff < 0)
 			return -1;
 		break;
+	}
 #endif
 	default:
 		BUG();
@@ -510,6 +586,21 @@ hashlimit_init_dst(const struct xt_hashlimit_htable *hinfo,
 	return 0;
 }
 
+static u32 hashlimit_byte_cost(unsigned int len, struct dsthash_ent *dh)
+{
+	u64 tmp = xt_hashlimit_len_to_chunks(len);
+	tmp = tmp * dh->rateinfo.cost;
+
+	if (unlikely(tmp > CREDITS_PER_JIFFY_BYTES * HZ))
+		tmp = CREDITS_PER_JIFFY_BYTES * HZ;
+
+	if (dh->rateinfo.credit < tmp && dh->rateinfo.credit_cap) {
+		dh->rateinfo.credit_cap--;
+		dh->rateinfo.credit = CREDITS_PER_JIFFY_BYTES * HZ;
+	}
+	return (u32) tmp;
+}
+
 static bool
 hashlimit_mt(const struct sk_buff *skb, struct xt_action_param *par)
 {
@@ -518,6 +609,8 @@ hashlimit_mt(const struct sk_buff *skb, struct xt_action_param *par)
 	unsigned long now = jiffies;
 	struct dsthash_ent *dh;
 	struct dsthash_dst dst;
+	bool race = false;
+	u32 cost;
 
 	if (hashlimit_init_dst(hinfo, &dst, skb, par->thoff) < 0)
 		goto hotdrop;
@@ -525,27 +618,32 @@ hashlimit_mt(const struct sk_buff *skb, struct xt_action_param *par)
 	rcu_read_lock_bh();
 	dh = dsthash_find(hinfo, &dst);
 	if (dh == NULL) {
-		dh = dsthash_alloc_init(hinfo, &dst);
+		dh = dsthash_alloc_init(hinfo, &dst, &race);
 		if (dh == NULL) {
 			rcu_read_unlock_bh();
 			goto hotdrop;
+		} else if (race) {
+			/* Already got an entry, update expiration timeout */
+			dh->expires = now + msecs_to_jiffies(hinfo->cfg.expire);
+			rateinfo_recalc(dh, now, hinfo->cfg.mode);
+		} else {
+			dh->expires = jiffies + msecs_to_jiffies(hinfo->cfg.expire);
+			rateinfo_init(dh, hinfo);
 		}
-		dh->expires = jiffies + msecs_to_jiffies(hinfo->cfg.expire);
-		dh->rateinfo.prev = jiffies;
-		dh->rateinfo.credit = user2credits(hinfo->cfg.avg *
-		                      hinfo->cfg.burst);
-		dh->rateinfo.credit_cap = user2credits(hinfo->cfg.avg *
-		                          hinfo->cfg.burst);
-		dh->rateinfo.cost = user2credits(hinfo->cfg.avg);
 	} else {
 		/* update expiration timeout */
 		dh->expires = now + msecs_to_jiffies(hinfo->cfg.expire);
-		rateinfo_recalc(dh, now);
+		rateinfo_recalc(dh, now, hinfo->cfg.mode);
 	}
 
-	if (dh->rateinfo.credit >= dh->rateinfo.cost) {
+	if (info->cfg.mode & XT_HASHLIMIT_BYTES)
+		cost = hashlimit_byte_cost(skb->len, dh);
+	else
+		cost = dh->rateinfo.cost;
+
+	if (dh->rateinfo.credit >= cost) {
 		/* below the limit */
-		dh->rateinfo.credit -= dh->rateinfo.cost;
+		dh->rateinfo.credit -= cost;
 		spin_unlock(&dh->lock);
 		rcu_read_unlock_bh();
 		return !(info->cfg.mode & XT_HASHLIMIT_INVERT);
@@ -567,14 +665,6 @@ static int hashlimit_mt_check(const struct xt_mtchk_param *par)
 	struct xt_hashlimit_mtinfo1 *info = par->matchinfo;
 	int ret;
 
-	/* Check for overflow. */
-	if (info->cfg.burst == 0 ||
-	    user2credits(info->cfg.avg * info->cfg.burst) <
-	    user2credits(info->cfg.avg)) {
-		pr_info("overflow, try lower: %u/%u\n",
-			info->cfg.avg, info->cfg.burst);
-		return -ERANGE;
-	}
 	if (info->cfg.gc_interval == 0 || info->cfg.expire == 0)
 		return -EINVAL;
 	if (info->name[sizeof(info->name)-1] != '\0')
@@ -587,6 +677,26 @@ static int hashlimit_mt_check(const struct xt_mtchk_param *par)
 			return -EINVAL;
 	}
 
+	if (info->cfg.mode & ~XT_HASHLIMIT_ALL) {
+		pr_info("Unknown mode mask %X, kernel too old?\n",
+						info->cfg.mode);
+		return -EINVAL;
+	}
+
+	/* Check for overflow. */
+	if (info->cfg.mode & XT_HASHLIMIT_BYTES) {
+		if (user2credits_byte(info->cfg.avg) == 0) {
+			pr_info("overflow, rate too high: %u\n", info->cfg.avg);
+			return -EINVAL;
+		}
+	} else if (info->cfg.burst == 0 ||
+		    user2credits(info->cfg.avg * info->cfg.burst) <
+		    user2credits(info->cfg.avg)) {
+			pr_info("overflow, try lower: %u/%u\n",
+				info->cfg.avg, info->cfg.burst);
+			return -ERANGE;
+	}
+
 	mutex_lock(&hashlimit_mutex);
 	info->hinfo = htable_find_get(net, info->name, par->family);
 	if (info->hinfo == NULL) {
@@ -618,7 +728,7 @@ static struct xt_match hashlimit_mt_reg[] __read_mostly = {
 		.destroy        = hashlimit_mt_destroy,
 		.me             = THIS_MODULE,
 	},
-#if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE)
+#if IS_ENABLED(CONFIG_IP6_NF_IPTABLES)
 	{
 		.name           = "hashlimit",
 		.revision       = 1,
@@ -679,10 +789,11 @@ static int dl_seq_real_show(struct dsthash_ent *ent, u_int8_t family,
 				   struct seq_file *s)
 {
 	int res;
+	const struct xt_hashlimit_htable *ht = s->private;
 
 	spin_lock(&ent->lock);
 	/* recalculate to show accurate numbers */
-	rateinfo_recalc(ent, jiffies);
+	rateinfo_recalc(ent, jiffies, ht->cfg.mode);
 
 	switch (family) {
 	case NFPROTO_IPV4:
@@ -695,7 +806,7 @@ static int dl_seq_real_show(struct dsthash_ent *ent, u_int8_t family,
 				 ent->rateinfo.credit, ent->rateinfo.credit_cap,
 				 ent->rateinfo.cost);
 		break;
-#if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE)
+#if IS_ENABLED(CONFIG_IP6_NF_IPTABLES)
 	case NFPROTO_IPV6:
 		res = seq_printf(s, "%ld %pI6:%u->%pI6:%u %u %u %u\n",
 				 (long)(ent->expires - jiffies)/HZ,
@@ -720,10 +831,9 @@ static int dl_seq_show(struct seq_file *s, void *v)
 	struct xt_hashlimit_htable *htable = s->private;
 	unsigned int *bucket = (unsigned int *)v;
 	struct dsthash_ent *ent;
-	struct hlist_node *pos;
 
 	if (!hlist_empty(&htable->hash[*bucket])) {
-		hlist_for_each_entry(ent, pos, &htable->hash[*bucket], node)
+		hlist_for_each_entry(ent, &htable->hash[*bucket], node)
 			if (dl_seq_real_show(ent, htable->family, s))
 				return -1;
 	}
@@ -743,7 +853,7 @@ static int dl_proc_open(struct inode *inode, struct file *file)
 
 	if (!ret) {
 		struct seq_file *sf = file->private_data;
-		sf->private = PDE(inode)->data;
+		sf->private = PDE_DATA(inode);
 	}
 	return ret;
 }
@@ -763,10 +873,10 @@ static int __net_init hashlimit_proc_net_init(struct net *net)
 	hashlimit_net->ipt_hashlimit = proc_mkdir("ipt_hashlimit", net->proc_net);
 	if (!hashlimit_net->ipt_hashlimit)
 		return -ENOMEM;
-#if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE)
+#if IS_ENABLED(CONFIG_IP6_NF_IPTABLES)
 	hashlimit_net->ip6t_hashlimit = proc_mkdir("ip6t_hashlimit", net->proc_net);
 	if (!hashlimit_net->ip6t_hashlimit) {
-		proc_net_remove(net, "ipt_hashlimit");
+		remove_proc_entry("ipt_hashlimit", net->proc_net);
 		return -ENOMEM;
 	}
 #endif
@@ -775,9 +885,23 @@ static int __net_init hashlimit_proc_net_init(struct net *net)
 
 static void __net_exit hashlimit_proc_net_exit(struct net *net)
 {
-	proc_net_remove(net, "ipt_hashlimit");
-#if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE)
-	proc_net_remove(net, "ip6t_hashlimit");
+	struct xt_hashlimit_htable *hinfo;
+	struct hashlimit_net *hashlimit_net = hashlimit_pernet(net);
+
+	/* hashlimit_net_exit() is called before hashlimit_mt_destroy().
+	 * Make sure that the parent ipt_hashlimit and ip6t_hashlimit proc
+	 * entries is empty before trying to remove it.
+	 */
+	mutex_lock(&hashlimit_mutex);
+	hlist_for_each_entry(hinfo, &hashlimit_net->htables, node)
+		htable_remove_proc_entry(hinfo);
+	hashlimit_net->ipt_hashlimit = NULL;
+	hashlimit_net->ip6t_hashlimit = NULL;
+	mutex_unlock(&hashlimit_mutex);
+
+	remove_proc_entry("ipt_hashlimit", net->proc_net);
+#if IS_ENABLED(CONFIG_IP6_NF_IPTABLES)
+	remove_proc_entry("ip6t_hashlimit", net->proc_net);
 #endif
 }
 
@@ -791,9 +915,6 @@ static int __net_init hashlimit_net_init(struct net *net)
 
 static void __net_exit hashlimit_net_exit(struct net *net)
 {
-	struct hashlimit_net *hashlimit_net = hashlimit_pernet(net);
-
-	BUG_ON(!hlist_empty(&hashlimit_net->htables));
 	hashlimit_proc_net_exit(net);
 }