netfilter: fixup the quota2, and enable.

The xt_quota2 came from
  http://sourceforge.net/projects/xtables-addons/develop

It needed tweaking for it to compile within the kernel tree.
Fixed kmalloc() and create_proc_entry() invocations within
 a non-interruptible context.
Removed useless copying of current quota back to the iptable's
struct matchinfo:
  - those are per CPU: they will change randomly based on which
    cpu gets to update the value.
  - they prevent matching a rule: e.g.
      -A chain -m quota2 --name q1 --quota 123
     can't be followed by
      -D chain -m quota2 --name q1 --quota 123
    as the 123 will be compared to the struct matchinfo's quota member.
Use the NETLINK NETLINK_NFLOG family to log a single message
when the quota limit is reached.
It uses the same packet type as ipt_ULOG, but
 - never copies skb data,
 - uses 112 as the event number (ULOG's +1)
It doesn't log if the module param "event_num" is 0.

Change-Id: I021d3b743db3b22158cc49acb5c94d905b501492
Signed-off-by: JP Abgrall <jpa@google.com>
diff --git a/net/netfilter/Kconfig b/net/netfilter/Kconfig
index 3681229..532e33d 100644
--- a/net/netfilter/Kconfig
+++ b/net/netfilter/Kconfig
@@ -1359,6 +1359,30 @@
 	  If you want to compile it as a module, say M here and read
 	  <file:Documentation/kbuild/modules.txt>.  If unsure, say `N'.
 
+config NETFILTER_XT_MATCH_QUOTA2
+	tristate '"quota2" match support'
+	depends on NETFILTER_ADVANCED
+	help
+	  This option adds a `quota2' match, which allows to match on a
+	  byte counter correctly and not per CPU.
+	  It allows naming the quotas.
+	  This is based on http://xtables-addons.git.sourceforge.net
+
+	  If you want to compile it as a module, say M here and read
+	  <file:Documentation/kbuild/modules.txt>.  If unsure, say `N'.
+
+config NETFILTER_XT_MATCH_QUOTA2_LOG
+	bool '"quota2" Netfilter LOG support'
+	depends on NETFILTER_XT_MATCH_QUOTA2
+	depends on IP_NF_TARGET_ULOG=n    # not yes, not module, just no
+	default n
+	help
+	  This option allows `quota2' to log ONCE when a quota limit
+	  is passed. It logs via NETLINK using the NETLINK_NFLOG family.
+	  It logs similarly to how ipt_ULOG would without data.
+
+	  If unsure, say `N'.
+
 config NETFILTER_XT_MATCH_RATEEST
 	tristate '"rateest" match support'
 	depends on NETFILTER_ADVANCED
diff --git a/net/netfilter/Makefile b/net/netfilter/Makefile
index 2c8455a..631f75c 100644
--- a/net/netfilter/Makefile
+++ b/net/netfilter/Makefile
@@ -165,6 +165,7 @@
 obj-$(CONFIG_NETFILTER_XT_MATCH_POLICY) += xt_policy.o
 obj-$(CONFIG_NETFILTER_XT_MATCH_QTAGUID) += xt_qtaguid_print.o xt_qtaguid.o
 obj-$(CONFIG_NETFILTER_XT_MATCH_QUOTA) += xt_quota.o
+obj-$(CONFIG_NETFILTER_XT_MATCH_QUOTA2) += xt_quota2.o
 obj-$(CONFIG_NETFILTER_XT_MATCH_RATEEST) += xt_rateest.o
 obj-$(CONFIG_NETFILTER_XT_MATCH_REALM) += xt_realm.o
 obj-$(CONFIG_NETFILTER_XT_MATCH_RECENT) += xt_recent.o
diff --git a/net/netfilter/xt_quota2.c b/net/netfilter/xt_quota2.c
index 4857008..aace729 100644
--- a/net/netfilter/xt_quota2.c
+++ b/net/netfilter/xt_quota2.c
@@ -12,14 +12,18 @@
  *	version 2 of the License, as published by the Free Software Foundation.
  */
 #include <linux/list.h>
+#include <linux/module.h>
 #include <linux/proc_fs.h>
 #include <linux/skbuff.h>
 #include <linux/spinlock.h>
 #include <asm/atomic.h>
+#include <net/netlink.h>
 
 #include <linux/netfilter/x_tables.h>
-#include "xt_quota2.h"
-#include "compat_xtables.h"
+#include <linux/netfilter/xt_quota2.h>
+#ifdef CONFIG_NETFILTER_XT_MATCH_QUOTA2_LOG
+#include <linux/netfilter_ipv4/ipt_ULOG.h>
+#endif
 
 /**
  * @lock:	lock to protect quota writers from each other
@@ -33,6 +37,16 @@
 	struct proc_dir_entry *procfs_entry;
 };
 
+#ifdef CONFIG_NETFILTER_XT_MATCH_QUOTA2_LOG
+/* Harald's favorite number +1 :D From ipt_ULOG.C */
+static int qlog_nl_event = 112;
+module_param_named(event_num, qlog_nl_event, uint, S_IRUGO | S_IWUSR);
+MODULE_PARM_DESC(event_num,
+		 "Event number for NETLINK_NFLOG message. 0 disables log."
+		 "111 is what ipt_ULOG uses.");
+static struct sock *nflognl;
+#endif
+
 static LIST_HEAD(counter_list);
 static DEFINE_SPINLOCK(counter_list_lock);
 
@@ -44,6 +58,70 @@
 module_param_named(uid, quota_list_uid, uint, S_IRUGO | S_IWUSR);
 module_param_named(gid, quota_list_gid, uint, S_IRUGO | S_IWUSR);
 
+
+#ifdef CONFIG_NETFILTER_XT_MATCH_QUOTA2_LOG
+static void quota2_log(unsigned int hooknum,
+		       const struct sk_buff *skb,
+		       const struct net_device *in,
+		       const struct net_device *out,
+		       const char *prefix)
+{
+	ulog_packet_msg_t *pm;
+	struct sk_buff *log_skb;
+	size_t size;
+	struct nlmsghdr *nlh;
+
+	if (!qlog_nl_event)
+		return;
+
+	size = NLMSG_SPACE(sizeof(*pm));
+	size = max(size, (size_t)NLMSG_GOODSIZE);
+	log_skb = alloc_skb(size, GFP_ATOMIC);
+	if (!log_skb) {
+		pr_err("xt_quota2: cannot alloc skb for logging\n");
+		return;
+	}
+
+	nlh = nlmsg_put(log_skb, /*pid*/0, /*seq*/0, qlog_nl_event,
+			sizeof(*pm), 0);
+	if (!nlh) {
+		pr_err("xt_quota2: nlmsg_put failed\n");
+		kfree_skb(log_skb);
+		return;
+	}
+	pm = nlmsg_data(nlh);
+	if (skb->tstamp.tv64 == 0)
+		__net_timestamp((struct sk_buff *)skb);
+	pm->data_len = 0;
+	pm->hook = hooknum;
+	if (prefix != NULL)
+		strlcpy(pm->prefix, prefix, sizeof(pm->prefix));
+	else
+		*(pm->prefix) = '\0';
+	if (in)
+		strlcpy(pm->indev_name, in->name, sizeof(pm->indev_name));
+	else
+		pm->indev_name[0] = '\0';
+
+	if (out)
+		strlcpy(pm->outdev_name, out->name, sizeof(pm->outdev_name));
+	else
+		pm->outdev_name[0] = '\0';
+
+	NETLINK_CB(log_skb).dst_group = 1;
+	pr_debug("throwing 1 packets to netlink group 1\n");
+	netlink_broadcast(nflognl, log_skb, 0, 1, GFP_ATOMIC);
+}
+#else
+static void quota2_log(unsigned int hooknum,
+		       const struct sk_buff *skb,
+		       const struct net_device *in,
+		       const struct net_device *out,
+		       const char *prefix)
+{
+}
+#endif  /* if+else CONFIG_NETFILTER_XT_MATCH_QUOTA2_LOG */
+
 static int quota_proc_read(char *page, char **start, off_t offset,
                            int count, int *eof, void *data)
 {
@@ -91,7 +169,7 @@
 	if (!anon) {
 		INIT_LIST_HEAD(&e->list);
 		atomic_set(&e->ref, 1);
-		strncpy(e->name, q->name, sizeof(e->name));
+		strlcpy(e->name, q->name, sizeof(e->name));
 	}
 	return e;
 }
@@ -104,42 +182,56 @@
 q2_get_counter(const struct xt_quota_mtinfo2 *q)
 {
 	struct proc_dir_entry *p;
-	struct xt_quota_counter *e;
+	struct xt_quota_counter *e = NULL;
+	struct xt_quota_counter *new_e;
 
 	if (*q->name == '\0')
 		return q2_new_counter(q, true);
 
+	/* No need to hold a lock while getting a new counter */
+	new_e = q2_new_counter(q, false);
+	if (new_e == NULL)
+		goto out;
+
 	spin_lock_bh(&counter_list_lock);
 	list_for_each_entry(e, &counter_list, list)
 		if (strcmp(e->name, q->name) == 0) {
 			atomic_inc(&e->ref);
 			spin_unlock_bh(&counter_list_lock);
+			kfree(new_e);
+			pr_debug("xt_quota2: old counter name=%s", e->name);
 			return e;
 		}
+	e = new_e;
+	pr_debug("xt_quota2: new_counter name=%s", e->name);
+	list_add_tail(&e->list, &counter_list);
+	/* The entry having a refcount of 1 is not directly destructible.
+	 * This func has not yet returned the new entry, thus iptables
+	 * has not references for destroying this entry.
+	 * For another rule to try to destroy it, it would 1st need for this
+	 * func* to be re-invoked, acquire a new ref for the same named quota.
+	 * Nobody will access the e->procfs_entry either.
+	 * So release the lock. */
+	spin_unlock_bh(&counter_list_lock);
 
-	e = q2_new_counter(q, false);
-	if (e == NULL)
-		goto out;
-
+	/* create_proc_entry() is not spin_lock happy */
 	p = e->procfs_entry = create_proc_entry(e->name, quota_list_perms,
 	                      proc_xt_quota);
-	if (p == NULL || IS_ERR(p))
-		goto out;
 
-#if LINUX_VERSION_CODE <= KERNEL_VERSION(2, 6, 29)
-	p->owner        = THIS_MODULE;
-#endif
+	if (IS_ERR_OR_NULL(p)) {
+		spin_lock_bh(&counter_list_lock);
+		list_del(&e->list);
+		spin_unlock_bh(&counter_list_lock);
+		goto out;
+	}
 	p->data         = e;
 	p->read_proc    = quota_proc_read;
 	p->write_proc   = quota_proc_write;
 	p->uid          = quota_list_uid;
 	p->gid          = quota_list_gid;
-	list_add_tail(&e->list, &counter_list);
-	spin_unlock_bh(&counter_list_lock);
 	return e;
 
  out:
-	spin_unlock_bh(&counter_list_lock);
 	kfree(e);
 	return NULL;
 }
@@ -148,6 +240,8 @@
 {
 	struct xt_quota_mtinfo2 *q = par->matchinfo;
 
+	pr_debug("xt_quota2: check() flags=0x%04x", q->flags);
+
 	if (q->flags & ~XT_QUOTA_MASK)
 		return -EINVAL;
 
@@ -203,7 +297,6 @@
 		 */
 		if (!(q->flags & XT_QUOTA_NO_CHANGE)) {
 			e->quota += (q->flags & XT_QUOTA_PACKET) ? 1 : skb->len;
-			q->quota = e->quota;
 		}
 		ret = true;
 	} else {
@@ -212,10 +305,17 @@
 				e->quota -= (q->flags & XT_QUOTA_PACKET) ? 1 : skb->len;
 			ret = !ret;
 		} else {
+			/* We are transitioning, log that fact. */
+			if (e->quota) {
+				quota2_log(par->hooknum,
+					   skb,
+					   par->in,
+					   par->out,
+					   q->name);
+			}
 			/* we do not allow even small packets from now on */
 			e->quota = 0;
 		}
-		q->quota = e->quota;
 	}
 	spin_unlock_bh(&e->lock);
 	return ret;
@@ -228,7 +328,7 @@
 		.family     = NFPROTO_IPV4,
 		.checkentry = quota_mt2_check,
 		.match      = quota_mt2,
-		.destroy    = quota_mt2_destroy,  
+		.destroy    = quota_mt2_destroy,
 		.matchsize  = sizeof(struct xt_quota_mtinfo2),
 		.me         = THIS_MODULE,
 	},
@@ -238,7 +338,7 @@
 		.family     = NFPROTO_IPV6,
 		.checkentry = quota_mt2_check,
 		.match      = quota_mt2,
-		.destroy    = quota_mt2_destroy,  
+		.destroy    = quota_mt2_destroy,
 		.matchsize  = sizeof(struct xt_quota_mtinfo2),
 		.me         = THIS_MODULE,
 	},
@@ -247,21 +347,29 @@
 static int __init quota_mt2_init(void)
 {
 	int ret;
+	pr_debug("xt_quota2: init()");
 
-	proc_xt_quota = proc_mkdir("xt_quota", init_net__proc_net);
+#ifdef CONFIG_NETFILTER_XT_MATCH_QUOTA2_LOG
+	nflognl = netlink_kernel_create(&init_net, NETLINK_NFLOG, NULL);
+	if (!nflognl)
+		return -ENOMEM;
+#endif
+
+	proc_xt_quota = proc_mkdir("xt_quota", init_net.proc_net);
 	if (proc_xt_quota == NULL)
 		return -EACCES;
 
 	ret = xt_register_matches(quota_mt2_reg, ARRAY_SIZE(quota_mt2_reg));
 	if (ret < 0)
-		remove_proc_entry("xt_quota", init_net__proc_net);
+		remove_proc_entry("xt_quota", init_net.proc_net);
+	pr_debug("xt_quota2: init() %d", ret);
 	return ret;
 }
 
 static void __exit quota_mt2_exit(void)
 {
 	xt_unregister_matches(quota_mt2_reg, ARRAY_SIZE(quota_mt2_reg));
-	remove_proc_entry("xt_quota", init_net__proc_net);
+	remove_proc_entry("xt_quota", init_net.proc_net);
 }
 
 module_init(quota_mt2_init);