[linux-block.git] / net / netfilter / nfnetlink_acct.c

// SPDX-License-Identifier: GPL-2.0-or-later
/*
 * (C) 2011 Pablo Neira Ayuso <pablo@netfilter.org>
 * (C) 2011 Intra2net AG <https://www.intra2net.com>
 */
#include <linux/init.h>
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/skbuff.h>
#include <linux/atomic.h>
#include <linux/refcount.h>
#include <linux/netlink.h>
#include <linux/rculist.h>
#include <linux/slab.h>
#include <linux/types.h>
#include <linux/errno.h>
#include <net/netlink.h>
#include <net/sock.h>
#include <net/netns/generic.h>

#include <linux/netfilter.h>
#include <linux/netfilter/nfnetlink.h>
#include <linux/netfilter/nfnetlink_acct.h>

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Pablo Neira Ayuso <pablo@netfilter.org>");
MODULE_DESCRIPTION("nfacct: Extended Netfilter accounting infrastructure");

struct nf_acct {
	atomic64_t		pkts;
	atomic64_t		bytes;
	unsigned long		flags;
	struct list_head	head;
	refcount_t		refcnt;
	char			name[NFACCT_NAME_MAX];
	struct rcu_head		rcu_head;
	char			data[];
};

struct nfacct_filter {
	u32 value;
	u32 mask;
};

struct nfnl_acct_net {
	struct list_head        nfnl_acct_list;
};

static unsigned int nfnl_acct_net_id __read_mostly;

static inline struct nfnl_acct_net *nfnl_acct_pernet(struct net *net)
{
	return net_generic(net, nfnl_acct_net_id);
}

#define NFACCT_F_QUOTA (NFACCT_F_QUOTA_PKTS | NFACCT_F_QUOTA_BYTES)
#define NFACCT_OVERQUOTA_BIT	2	/* NFACCT_F_OVERQUOTA */

static int nfnl_acct_new(struct sk_buff *skb, const struct nfnl_info *info,
			 const struct nlattr * const tb[])
{
	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(info->net);
	struct nf_acct *nfacct, *matching = NULL;
	unsigned int size = 0;
	char *acct_name;
	u32 flags = 0;

	if (!tb[NFACCT_NAME])
		return -EINVAL;

	acct_name = nla_data(tb[NFACCT_NAME]);
	if (strlen(acct_name) == 0)
		return -EINVAL;

	list_for_each_entry(nfacct, &nfnl_acct_net->nfnl_acct_list, head) {
		if (strncmp(nfacct->name, acct_name, NFACCT_NAME_MAX) != 0)
			continue;

                if (info->nlh->nlmsg_flags & NLM_F_EXCL)
			return -EEXIST;

		matching = nfacct;
		break;
        }

	if (matching) {
		if (info->nlh->nlmsg_flags & NLM_F_REPLACE) {
			/* reset counters if you request a replacement. */
			atomic64_set(&matching->pkts, 0);
			atomic64_set(&matching->bytes, 0);
			smp_mb__before_atomic();
			/* reset overquota flag if quota is enabled. */
			if ((matching->flags & NFACCT_F_QUOTA))
				clear_bit(NFACCT_OVERQUOTA_BIT,
					  &matching->flags);
			return 0;
		}
		return -EBUSY;
	}

	if (tb[NFACCT_FLAGS]) {
		flags = ntohl(nla_get_be32(tb[NFACCT_FLAGS]));
		if (flags & ~NFACCT_F_QUOTA)
			return -EOPNOTSUPP;
		if ((flags & NFACCT_F_QUOTA) == NFACCT_F_QUOTA)
			return -EINVAL;
		if (flags & NFACCT_F_OVERQUOTA)
			return -EINVAL;
		if ((flags & NFACCT_F_QUOTA) && !tb[NFACCT_QUOTA])
			return -EINVAL;

		size += sizeof(u64);
	}

	nfacct = kzalloc(sizeof(struct nf_acct) + size, GFP_KERNEL);
	if (nfacct == NULL)
		return -ENOMEM;

	if (flags & NFACCT_F_QUOTA) {
		u64 *quota = (u64 *)nfacct->data;

		*quota = be64_to_cpu(nla_get_be64(tb[NFACCT_QUOTA]));
		nfacct->flags = flags;
	}

	nla_strscpy(nfacct->name, tb[NFACCT_NAME], NFACCT_NAME_MAX);

	if (tb[NFACCT_BYTES]) {
		atomic64_set(&nfacct->bytes,
			     be64_to_cpu(nla_get_be64(tb[NFACCT_BYTES])));
	}
	if (tb[NFACCT_PKTS]) {
		atomic64_set(&nfacct->pkts,
			     be64_to_cpu(nla_get_be64(tb[NFACCT_PKTS])));
	}
	refcount_set(&nfacct->refcnt, 1);
	list_add_tail_rcu(&nfacct->head, &nfnl_acct_net->nfnl_acct_list);
	return 0;
}

static int
nfnl_acct_fill_info(struct sk_buff *skb, u32 portid, u32 seq, u32 type,
		   int event, struct nf_acct *acct)
{
	struct nlmsghdr *nlh;
	unsigned int flags = portid ? NLM_F_MULTI : 0;
	u64 pkts, bytes;
	u32 old_flags;

	event = nfnl_msg_type(NFNL_SUBSYS_ACCT, event);
	nlh = nfnl_msg_put(skb, portid, seq, event, flags, AF_UNSPEC,
			   NFNETLINK_V0, 0);
	if (!nlh)
		goto nlmsg_failure;

	if (nla_put_string(skb, NFACCT_NAME, acct->name))
		goto nla_put_failure;

	old_flags = acct->flags;
	if (type == NFNL_MSG_ACCT_GET_CTRZERO) {
		pkts = atomic64_xchg(&acct->pkts, 0);
		bytes = atomic64_xchg(&acct->bytes, 0);
		smp_mb__before_atomic();
		if (acct->flags & NFACCT_F_QUOTA)
			clear_bit(NFACCT_OVERQUOTA_BIT, &acct->flags);
	} else {
		pkts = atomic64_read(&acct->pkts);
		bytes = atomic64_read(&acct->bytes);
	}
	if (nla_put_be64(skb, NFACCT_PKTS, cpu_to_be64(pkts),
			 NFACCT_PAD) ||
	    nla_put_be64(skb, NFACCT_BYTES, cpu_to_be64(bytes),
			 NFACCT_PAD) ||
	    nla_put_be32(skb, NFACCT_USE, htonl(refcount_read(&acct->refcnt))))
		goto nla_put_failure;
	if (acct->flags & NFACCT_F_QUOTA) {
		u64 *quota = (u64 *)acct->data;

		if (nla_put_be32(skb, NFACCT_FLAGS, htonl(old_flags)) ||
		    nla_put_be64(skb, NFACCT_QUOTA, cpu_to_be64(*quota),
				 NFACCT_PAD))
			goto nla_put_failure;
	}
	nlmsg_end(skb, nlh);
	return skb->len;

nlmsg_failure:
nla_put_failure:
	nlmsg_cancel(skb, nlh);
	return -1;
}

static int
nfnl_acct_dump(struct sk_buff *skb, struct netlink_callback *cb)
{
	struct net *net = sock_net(skb->sk);
	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net);
	struct nf_acct *cur, *last;
	const struct nfacct_filter *filter = cb->data;

	if (cb->args[2])
		return 0;

	last = (struct nf_acct *)cb->args[1];
	if (cb->args[1])
		cb->args[1] = 0;

	rcu_read_lock();
	list_for_each_entry_rcu(cur, &nfnl_acct_net->nfnl_acct_list, head) {
		if (last) {
			if (cur != last)
				continue;

			last = NULL;
		}

		if (filter && (cur->flags & filter->mask) != filter->value)
			continue;

		if (nfnl_acct_fill_info(skb, NETLINK_CB(cb->skb).portid,
				       cb->nlh->nlmsg_seq,
				       NFNL_MSG_TYPE(cb->nlh->nlmsg_type),
				       NFNL_MSG_ACCT_NEW, cur) < 0) {
			cb->args[1] = (unsigned long)cur;
			break;
		}
	}
	if (!cb->args[1])
		cb->args[2] = 1;
	rcu_read_unlock();
	return skb->len;
}

static int nfnl_acct_done(struct netlink_callback *cb)
{
	kfree(cb->data);
	return 0;
}

static const struct nla_policy filter_policy[NFACCT_FILTER_MAX + 1] = {
	[NFACCT_FILTER_MASK]	= { .type = NLA_U32 },
	[NFACCT_FILTER_VALUE]	= { .type = NLA_U32 },
};

static int nfnl_acct_start(struct netlink_callback *cb)
{
	const struct nlattr *const attr = cb->data;
	struct nlattr *tb[NFACCT_FILTER_MAX + 1];
	struct nfacct_filter *filter;
	int err;

	if (!attr)
		return 0;

	err = nla_parse_nested_deprecated(tb, NFACCT_FILTER_MAX, attr,
					  filter_policy, NULL);
	if (err < 0)
		return err;

	if (!tb[NFACCT_FILTER_MASK] || !tb[NFACCT_FILTER_VALUE])
		return -EINVAL;

	filter = kzalloc(sizeof(struct nfacct_filter), GFP_KERNEL);
	if (!filter)
		return -ENOMEM;

	filter->mask = ntohl(nla_get_be32(tb[NFACCT_FILTER_MASK]));
	filter->value = ntohl(nla_get_be32(tb[NFACCT_FILTER_VALUE]));
	cb->data = filter;

	return 0;
}

static int nfnl_acct_get(struct sk_buff *skb, const struct nfnl_info *info,
			 const struct nlattr * const tb[])
{
	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(info->net);
	int ret = -ENOENT;
	struct nf_acct *cur;
	char *acct_name;

	if (info->nlh->nlmsg_flags & NLM_F_DUMP) {
		struct netlink_dump_control c = {
			.dump = nfnl_acct_dump,
			.start = nfnl_acct_start,
			.done = nfnl_acct_done,
			.data = (void *)tb[NFACCT_FILTER],
		};

		return netlink_dump_start(info->sk, skb, info->nlh, &c);
	}

	if (!tb[NFACCT_NAME])
		return -EINVAL;
	acct_name = nla_data(tb[NFACCT_NAME]);

	list_for_each_entry(cur, &nfnl_acct_net->nfnl_acct_list, head) {
		struct sk_buff *skb2;

		if (strncmp(cur->name, acct_name, NFACCT_NAME_MAX)!= 0)
			continue;

		skb2 = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
		if (skb2 == NULL) {
			ret = -ENOMEM;
			break;
		}

		ret = nfnl_acct_fill_info(skb2, NETLINK_CB(skb).portid,
					  info->nlh->nlmsg_seq,
					  NFNL_MSG_TYPE(info->nlh->nlmsg_type),
					  NFNL_MSG_ACCT_NEW, cur);
		if (ret <= 0) {
			kfree_skb(skb2);
			break;
		}

		ret = nfnetlink_unicast(skb2, info->net, NETLINK_CB(skb).portid);
		break;
	}

	return ret;
}

/* try to delete object, fail if it is still in use. */
static int nfnl_acct_try_del(struct nf_acct *cur)
{
	int ret = 0;

	/* We want to avoid races with nfnl_acct_put. So only when the current
	 * refcnt is 1, we decrease it to 0.
	 */
	if (refcount_dec_if_one(&cur->refcnt)) {
		/* We are protected by nfnl mutex. */
		list_del_rcu(&cur->head);
		kfree_rcu(cur, rcu_head);
	} else {
		ret = -EBUSY;
	}
	return ret;
}

static int nfnl_acct_del(struct sk_buff *skb, const struct nfnl_info *info,
			 const struct nlattr * const tb[])
{
	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(info->net);
	struct nf_acct *cur, *tmp;
	int ret = -ENOENT;
	char *acct_name;

	if (!tb[NFACCT_NAME]) {
		list_for_each_entry_safe(cur, tmp, &nfnl_acct_net->nfnl_acct_list, head)
			nfnl_acct_try_del(cur);

		return 0;
	}
	acct_name = nla_data(tb[NFACCT_NAME]);

	list_for_each_entry(cur, &nfnl_acct_net->nfnl_acct_list, head) {
		if (strncmp(cur->name, acct_name, NFACCT_NAME_MAX) != 0)
			continue;

		ret = nfnl_acct_try_del(cur);
		if (ret < 0)
			return ret;

		break;
	}
	return ret;
}

static const struct nla_policy nfnl_acct_policy[NFACCT_MAX+1] = {
	[NFACCT_NAME] = { .type = NLA_NUL_STRING, .len = NFACCT_NAME_MAX-1 },
	[NFACCT_BYTES] = { .type = NLA_U64 },
	[NFACCT_PKTS] = { .type = NLA_U64 },
	[NFACCT_FLAGS] = { .type = NLA_U32 },
	[NFACCT_QUOTA] = { .type = NLA_U64 },
	[NFACCT_FILTER] = {.type = NLA_NESTED },
};

static const struct nfnl_callback nfnl_acct_cb[NFNL_MSG_ACCT_MAX] = {
	[NFNL_MSG_ACCT_NEW] = {
		.call		= nfnl_acct_new,
		.type		= NFNL_CB_MUTEX,
		.attr_count	= NFACCT_MAX,
		.policy		= nfnl_acct_policy
	},
	[NFNL_MSG_ACCT_GET] = {
		.call		= nfnl_acct_get,
		.type		= NFNL_CB_MUTEX,
		.attr_count	= NFACCT_MAX,
		.policy		= nfnl_acct_policy
	},
	[NFNL_MSG_ACCT_GET_CTRZERO] = {
		.call		= nfnl_acct_get,
		.type		= NFNL_CB_MUTEX,
		.attr_count	= NFACCT_MAX,
		.policy		= nfnl_acct_policy
	},
	[NFNL_MSG_ACCT_DEL] = {
		.call		= nfnl_acct_del,
		.type		= NFNL_CB_MUTEX,
		.attr_count	= NFACCT_MAX,
		.policy		= nfnl_acct_policy
	},
};

static const struct nfnetlink_subsystem nfnl_acct_subsys = {
	.name				= "acct",
	.subsys_id			= NFNL_SUBSYS_ACCT,
	.cb_count			= NFNL_MSG_ACCT_MAX,
	.cb				= nfnl_acct_cb,
};

MODULE_ALIAS_NFNL_SUBSYS(NFNL_SUBSYS_ACCT);

struct nf_acct *nfnl_acct_find_get(struct net *net, const char *acct_name)
{
	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net);
	struct nf_acct *cur, *acct = NULL;

	rcu_read_lock();
	list_for_each_entry_rcu(cur, &nfnl_acct_net->nfnl_acct_list, head) {
		if (strncmp(cur->name, acct_name, NFACCT_NAME_MAX)!= 0)
			continue;

		if (!try_module_get(THIS_MODULE))
			goto err;

		if (!refcount_inc_not_zero(&cur->refcnt)) {
			module_put(THIS_MODULE);
			goto err;
		}

		acct = cur;
		break;
	}
err:
	rcu_read_unlock();
	return acct;
}
EXPORT_SYMBOL_GPL(nfnl_acct_find_get);

void nfnl_acct_put(struct nf_acct *acct)
{
	if (refcount_dec_and_test(&acct->refcnt))
		kfree_rcu(acct, rcu_head);

	module_put(THIS_MODULE);
}
EXPORT_SYMBOL_GPL(nfnl_acct_put);

void nfnl_acct_update(const struct sk_buff *skb, struct nf_acct *nfacct)
{
	atomic64_inc(&nfacct->pkts);
	atomic64_add(skb->len, &nfacct->bytes);
}
EXPORT_SYMBOL_GPL(nfnl_acct_update);

static void nfnl_overquota_report(struct net *net, struct nf_acct *nfacct)
{
	int ret;
	struct sk_buff *skb;

	skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC);
	if (skb == NULL)
		return;

	ret = nfnl_acct_fill_info(skb, 0, 0, NFNL_MSG_ACCT_OVERQUOTA, 0,
				  nfacct);
	if (ret <= 0) {
		kfree_skb(skb);
		return;
	}
	nfnetlink_broadcast(net, skb, 0, NFNLGRP_ACCT_QUOTA, GFP_ATOMIC);
}

int nfnl_acct_overquota(struct net *net, struct nf_acct *nfacct)
{
	u64 now;
	u64 *quota;
	int ret = NFACCT_UNDERQUOTA;

	/* no place here if we don't have a quota */
	if (!(nfacct->flags & NFACCT_F_QUOTA))
		return NFACCT_NO_QUOTA;

	quota = (u64 *)nfacct->data;
	now = (nfacct->flags & NFACCT_F_QUOTA_PKTS) ?
	       atomic64_read(&nfacct->pkts) : atomic64_read(&nfacct->bytes);

	ret = now > *quota;

	if (now >= *quota &&
	    !test_and_set_bit(NFACCT_OVERQUOTA_BIT, &nfacct->flags)) {
		nfnl_overquota_report(net, nfacct);
	}

	return ret;
}
EXPORT_SYMBOL_GPL(nfnl_acct_overquota);

static int __net_init nfnl_acct_net_init(struct net *net)
{
	INIT_LIST_HEAD(&nfnl_acct_pernet(net)->nfnl_acct_list);

	return 0;
}

static void __net_exit nfnl_acct_net_exit(struct net *net)
{
	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net);
	struct nf_acct *cur, *tmp;

	list_for_each_entry_safe(cur, tmp, &nfnl_acct_net->nfnl_acct_list, head) {
		list_del_rcu(&cur->head);

		if (refcount_dec_and_test(&cur->refcnt))
			kfree_rcu(cur, rcu_head);
	}
}

static struct pernet_operations nfnl_acct_ops = {
        .init   = nfnl_acct_net_init,
        .exit   = nfnl_acct_net_exit,
        .id     = &nfnl_acct_net_id,
        .size   = sizeof(struct nfnl_acct_net),
};

static int __init nfnl_acct_init(void)
{
	int ret;

	ret = register_pernet_subsys(&nfnl_acct_ops);
	if (ret < 0) {
		pr_err("nfnl_acct_init: failed to register pernet ops\n");
		goto err_out;
	}

	ret = nfnetlink_subsys_register(&nfnl_acct_subsys);
	if (ret < 0) {
		pr_err("nfnl_acct_init: cannot register with nfnetlink.\n");
		goto cleanup_pernet;
	}
	return 0;

cleanup_pernet:
	unregister_pernet_subsys(&nfnl_acct_ops);
err_out:
	return ret;
}

static void __exit nfnl_acct_exit(void)
{
	nfnetlink_subsys_unregister(&nfnl_acct_subsys);
	unregister_pernet_subsys(&nfnl_acct_ops);
}

module_init(nfnl_acct_init);
module_exit(nfnl_acct_exit);
Commit	Line	Data
e97150df	1	// SPDX-License-Identifier: GPL-2.0-or-later
94139027 PNA	2	/*
94139027 PNA	3	* (C) 2011 Pablo Neira Ayuso <pablo@netfilter.org>
50935339	4	* (C) 2011 Intra2net AG <https://www.intra2net.com>
94139027 PNA	5	*/
	6	#include <linux/init.h>
	7	#include <linux/module.h>
	8	#include <linux/kernel.h>
	9	#include <linux/skbuff.h>
6523cf9a	10	#include <linux/atomic.h>
b54ab92b	11	#include <linux/refcount.h>
94139027 PNA	12	#include <linux/netlink.h>
	13	#include <linux/rculist.h>
	14	#include <linux/slab.h>
	15	#include <linux/types.h>
	16	#include <linux/errno.h>
	17	#include <net/netlink.h>
	18	#include <net/sock.h>
f7583f02	19	#include <net/netns/generic.h>
94139027 PNA	20
	21	#include <linux/netfilter.h>
	22	#include <linux/netfilter/nfnetlink.h>
	23	#include <linux/netfilter/nfnetlink_acct.h>
	24
	25	MODULE_LICENSE("GPL");
	26	MODULE_AUTHOR("Pablo Neira Ayuso <pablo@netfilter.org>");
	27	MODULE_DESCRIPTION("nfacct: Extended Netfilter accounting infrastructure");
	28
94139027 PNA	29	struct nf_acct {
	30	atomic64_t pkts;
	31	atomic64_t bytes;
683399ed	32	unsigned long flags;
94139027	33	struct list_head head;
b54ab92b	34	refcount_t refcnt;
94139027 PNA	35	char name[NFACCT_NAME_MAX];
94139027 PNA	36	struct rcu_head rcu_head;
6daf1414	37	char data[];
94139027 PNA	38	};
94139027 PNA	39
f111f780 AP	40	struct nfacct_filter {
	41	u32 value;
	42	u32 mask;
	43	};
	44
f7583f02 WS	45	struct nfnl_acct_net {
	46	struct list_head nfnl_acct_list;
	47	};
	48
	49	static unsigned int nfnl_acct_net_id __read_mostly;
	50
	51	static inline struct nfnl_acct_net nfnl_acct_pernet(struct net net)
	52	{
	53	return net_generic(net, nfnl_acct_net_id);
	54	}
	55
683399ed	56	#define NFACCT_F_QUOTA (NFACCT_F_QUOTA_PKTS \| NFACCT_F_QUOTA_BYTES)
b6d04688	57	#define NFACCT_OVERQUOTA_BIT 2 /* NFACCT_F_OVERQUOTA */
683399ed	58
a6555365 PNA	59	static int nfnl_acct_new(struct sk_buff skb, const struct nfnl_info info,
a6555365 PNA	60	const struct nlattr * const tb[])
94139027	61	{
a6555365	62	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(info->net);
94139027	63	struct nf_acct nfacct, matching = NULL;
683399ed	64	unsigned int size = 0;
a6555365	65	char *acct_name;
683399ed	66	u32 flags = 0;
94139027 PNA	67
	68	if (!tb[NFACCT_NAME])
	69	return -EINVAL;
	70
	71	acct_name = nla_data(tb[NFACCT_NAME]);
deadcfc3 PNA	72	if (strlen(acct_name) == 0)
deadcfc3 PNA	73	return -EINVAL;
94139027	74
f7583f02	75	list_for_each_entry(nfacct, &nfnl_acct_net->nfnl_acct_list, head) {
94139027 PNA	76	if (strncmp(nfacct->name, acct_name, NFACCT_NAME_MAX) != 0)
	77	continue;
	78
a6555365	79	if (info->nlh->nlmsg_flags & NLM_F_EXCL)
94139027 PNA	80	return -EEXIST;
	81
	82	matching = nfacct;
	83	break;
	84	}
	85
	86	if (matching) {
a6555365	87	if (info->nlh->nlmsg_flags & NLM_F_REPLACE) {
94139027 PNA	88	/* reset counters if you request a replacement. */
	89	atomic64_set(&matching->pkts, 0);
	90	atomic64_set(&matching->bytes, 0);
f9da455b	91	smp_mb__before_atomic();
683399ed MP	92	/* reset overquota flag if quota is enabled. */
683399ed MP	93	if ((matching->flags & NFACCT_F_QUOTA))
b6d04688 AP	94	clear_bit(NFACCT_OVERQUOTA_BIT,
b6d04688 AP	95	&matching->flags);
94139027 PNA	96	return 0;
	97	}
	98	return -EBUSY;
	99	}
	100
683399ed MP	101	if (tb[NFACCT_FLAGS]) {
	102	flags = ntohl(nla_get_be32(tb[NFACCT_FLAGS]));
	103	if (flags & ~NFACCT_F_QUOTA)
	104	return -EOPNOTSUPP;
	105	if ((flags & NFACCT_F_QUOTA) == NFACCT_F_QUOTA)
	106	return -EINVAL;
	107	if (flags & NFACCT_F_OVERQUOTA)
	108	return -EINVAL;
eda3fc50 PT	109	if ((flags & NFACCT_F_QUOTA) && !tb[NFACCT_QUOTA])
eda3fc50 PT	110	return -EINVAL;
683399ed MP	111
	112	size += sizeof(u64);
	113	}
	114
	115	nfacct = kzalloc(sizeof(struct nf_acct) + size, GFP_KERNEL);
94139027 PNA	116	if (nfacct == NULL)
	117	return -ENOMEM;
	118
683399ed MP	119	if (flags & NFACCT_F_QUOTA) {
	120	u64 quota = (u64 )nfacct->data;
	121
	122	*quota = be64_to_cpu(nla_get_be64(tb[NFACCT_QUOTA]));
	123	nfacct->flags = flags;
	124	}
	125
872f6903	126	nla_strscpy(nfacct->name, tb[NFACCT_NAME], NFACCT_NAME_MAX);
94139027 PNA	127
	128	if (tb[NFACCT_BYTES]) {
	129	atomic64_set(&nfacct->bytes,
fe31d1a8	130	be64_to_cpu(nla_get_be64(tb[NFACCT_BYTES])));
94139027 PNA	131	}
	132	if (tb[NFACCT_PKTS]) {
	133	atomic64_set(&nfacct->pkts,
fe31d1a8	134	be64_to_cpu(nla_get_be64(tb[NFACCT_PKTS])));
94139027	135	}
b54ab92b	136	refcount_set(&nfacct->refcnt, 1);
f7583f02	137	list_add_tail_rcu(&nfacct->head, &nfnl_acct_net->nfnl_acct_list);
94139027 PNA	138	return 0;
	139	}
	140
	141	static int
15e47304	142	nfnl_acct_fill_info(struct sk_buff *skb, u32 portid, u32 seq, u32 type,
94139027 PNA	143	int event, struct nf_acct *acct)
	144	{
	145	struct nlmsghdr *nlh;
15e47304	146	unsigned int flags = portid ? NLM_F_MULTI : 0;
94139027	147	u64 pkts, bytes;
d24675cb	148	u32 old_flags;
94139027	149
dedb67c4	150	event = nfnl_msg_type(NFNL_SUBSYS_ACCT, event);
19c28b13 PNA	151	nlh = nfnl_msg_put(skb, portid, seq, event, flags, AF_UNSPEC,
	152	NFNETLINK_V0, 0);
	153	if (!nlh)
94139027 PNA	154	goto nlmsg_failure;
94139027 PNA	155
7c801189 DM	156	if (nla_put_string(skb, NFACCT_NAME, acct->name))
7c801189 DM	157	goto nla_put_failure;
94139027	158
d24675cb	159	old_flags = acct->flags;
94139027 PNA	160	if (type == NFNL_MSG_ACCT_GET_CTRZERO) {
	161	pkts = atomic64_xchg(&acct->pkts, 0);
	162	bytes = atomic64_xchg(&acct->bytes, 0);
f9da455b	163	smp_mb__before_atomic();
683399ed	164	if (acct->flags & NFACCT_F_QUOTA)
b6d04688	165	clear_bit(NFACCT_OVERQUOTA_BIT, &acct->flags);
94139027 PNA	166	} else {
	167	pkts = atomic64_read(&acct->pkts);
	168	bytes = atomic64_read(&acct->bytes);
	169	}
b46f6ded ND	170	if (nla_put_be64(skb, NFACCT_PKTS, cpu_to_be64(pkts),
	171	NFACCT_PAD) \|\|
	172	nla_put_be64(skb, NFACCT_BYTES, cpu_to_be64(bytes),
	173	NFACCT_PAD) \|\|
b54ab92b	174	nla_put_be32(skb, NFACCT_USE, htonl(refcount_read(&acct->refcnt))))
7c801189	175	goto nla_put_failure;
683399ed MP	176	if (acct->flags & NFACCT_F_QUOTA) {
683399ed MP	177	u64 quota = (u64 )acct->data;
94139027	178
d24675cb	179	if (nla_put_be32(skb, NFACCT_FLAGS, htonl(old_flags)) \|\|
b46f6ded ND	180	nla_put_be64(skb, NFACCT_QUOTA, cpu_to_be64(*quota),
b46f6ded ND	181	NFACCT_PAD))
683399ed MP	182	goto nla_put_failure;
683399ed MP	183	}
94139027 PNA	184	nlmsg_end(skb, nlh);
	185	return skb->len;
	186
	187	nlmsg_failure:
	188	nla_put_failure:
	189	nlmsg_cancel(skb, nlh);
	190	return -1;
	191	}
	192
	193	static int
	194	nfnl_acct_dump(struct sk_buff skb, struct netlink_callback cb)
	195	{
3499abb2	196	struct net *net = sock_net(skb->sk);
f7583f02	197	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net);
94139027	198	struct nf_acct cur, last;
f111f780	199	const struct nfacct_filter *filter = cb->data;
94139027 PNA	200
	201	if (cb->args[2])
	202	return 0;
	203
	204	last = (struct nf_acct *)cb->args[1];
	205	if (cb->args[1])
	206	cb->args[1] = 0;
	207
	208	rcu_read_lock();
f7583f02	209	list_for_each_entry_rcu(cur, &nfnl_acct_net->nfnl_acct_list, head) {
991a6b73 PNA	210	if (last) {
	211	if (cur != last)
	212	continue;
94139027	213
991a6b73 PNA	214	last = NULL;
991a6b73 PNA	215	}
f111f780 AP	216
	217	if (filter && (cur->flags & filter->mask) != filter->value)
	218	continue;
	219
15e47304	220	if (nfnl_acct_fill_info(skb, NETLINK_CB(cb->skb).portid,
94139027 PNA	221	cb->nlh->nlmsg_seq,
	222	NFNL_MSG_TYPE(cb->nlh->nlmsg_type),
	223	NFNL_MSG_ACCT_NEW, cur) < 0) {
	224	cb->args[1] = (unsigned long)cur;
	225	break;
	226	}
	227	}
	228	if (!cb->args[1])
	229	cb->args[2] = 1;
	230	rcu_read_unlock();
	231	return skb->len;
	232	}
	233
f111f780 AP	234	static int nfnl_acct_done(struct netlink_callback *cb)
	235	{
	236	kfree(cb->data);
	237	return 0;
	238	}
	239
	240	static const struct nla_policy filter_policy[NFACCT_FILTER_MAX + 1] = {
	241	[NFACCT_FILTER_MASK] = { .type = NLA_U32 },
	242	[NFACCT_FILTER_VALUE] = { .type = NLA_U32 },
	243	};
	244
3e673b23	245	static int nfnl_acct_start(struct netlink_callback *cb)
f111f780	246	{
3e673b23	247	const struct nlattr *const attr = cb->data;
f111f780	248	struct nlattr *tb[NFACCT_FILTER_MAX + 1];
3e673b23	249	struct nfacct_filter *filter;
f111f780 AP	250	int err;
f111f780 AP	251
3e673b23 FW	252	if (!attr)
	253	return 0;
	254
8cb08174 JB	255	err = nla_parse_nested_deprecated(tb, NFACCT_FILTER_MAX, attr,
8cb08174 JB	256	filter_policy, NULL);
f111f780	257	if (err < 0)
3e673b23	258	return err;
f111f780	259
017b1b6d	260	if (!tb[NFACCT_FILTER_MASK] \|\| !tb[NFACCT_FILTER_VALUE])
3e673b23	261	return -EINVAL;
017b1b6d	262
f111f780 AP	263	filter = kzalloc(sizeof(struct nfacct_filter), GFP_KERNEL);
f111f780 AP	264	if (!filter)
3e673b23	265	return -ENOMEM;
f111f780 AP	266
	267	filter->mask = ntohl(nla_get_be32(tb[NFACCT_FILTER_MASK]));
	268	filter->value = ntohl(nla_get_be32(tb[NFACCT_FILTER_VALUE]));
3e673b23	269	cb->data = filter;
f111f780	270
3e673b23	271	return 0;
f111f780 AP	272	}
f111f780 AP	273
a6555365 PNA	274	static int nfnl_acct_get(struct sk_buff skb, const struct nfnl_info info,
a6555365 PNA	275	const struct nlattr * const tb[])
94139027	276	{
a6555365	277	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(info->net);
3ab0b245	278	int ret = -ENOENT;
94139027 PNA	279	struct nf_acct *cur;
	280	char *acct_name;
	281
a6555365	282	if (info->nlh->nlmsg_flags & NLM_F_DUMP) {
80d326fa PNA	283	struct netlink_dump_control c = {
80d326fa PNA	284	.dump = nfnl_acct_dump,
3e673b23	285	.start = nfnl_acct_start,
f111f780	286	.done = nfnl_acct_done,
3e673b23	287	.data = (void *)tb[NFACCT_FILTER],
80d326fa	288	};
f111f780	289
a6555365	290	return netlink_dump_start(info->sk, skb, info->nlh, &c);
94139027 PNA	291	}
	292
	293	if (!tb[NFACCT_NAME])
	294	return -EINVAL;
	295	acct_name = nla_data(tb[NFACCT_NAME]);
	296
f7583f02	297	list_for_each_entry(cur, &nfnl_acct_net->nfnl_acct_list, head) {
94139027 PNA	298	struct sk_buff *skb2;
	299
	300	if (strncmp(cur->name, acct_name, NFACCT_NAME_MAX)!= 0)
	301	continue;
	302
	303	skb2 = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
3ab0b245 PNA	304	if (skb2 == NULL) {
3ab0b245 PNA	305	ret = -ENOMEM;
94139027	306	break;
3ab0b245	307	}
94139027	308
15e47304	309	ret = nfnl_acct_fill_info(skb2, NETLINK_CB(skb).portid,
a6555365 PNA	310	info->nlh->nlmsg_seq,
	311	NFNL_MSG_TYPE(info->nlh->nlmsg_type),
	312	NFNL_MSG_ACCT_NEW, cur);
3ab0b245	313	if (ret <= 0) {
94139027	314	kfree_skb(skb2);
3ab0b245 PNA	315	break;
3ab0b245 PNA	316	}
94139027	317
e0241ae6 PNA	318	ret = nfnetlink_unicast(skb2, info->net, NETLINK_CB(skb).portid);
e0241ae6 PNA	319	break;
94139027	320	}
e0241ae6	321
94139027 PNA	322	return ret;
	323	}
	324
	325	/* try to delete object, fail if it is still in use. */
	326	static int nfnl_acct_try_del(struct nf_acct *cur)
	327	{
	328	int ret = 0;
	329
12be15dd LZ	330	/* We want to avoid races with nfnl_acct_put. So only when the current
	331	* refcnt is 1, we decrease it to 0.
	332	*/
b54ab92b	333	if (refcount_dec_if_one(&cur->refcnt)) {
94139027 PNA	334	/* We are protected by nfnl mutex. */
	335	list_del_rcu(&cur->head);
	336	kfree_rcu(cur, rcu_head);
	337	} else {
94139027 PNA	338	ret = -EBUSY;
	339	}
	340	return ret;
	341	}
	342
a6555365 PNA	343	static int nfnl_acct_del(struct sk_buff skb, const struct nfnl_info info,
a6555365 PNA	344	const struct nlattr * const tb[])
94139027	345	{
a6555365	346	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(info->net);
93fac10b	347	struct nf_acct cur, tmp;
94139027	348	int ret = -ENOENT;
93fac10b	349	char *acct_name;
94139027 PNA	350
94139027 PNA	351	if (!tb[NFACCT_NAME]) {
f7583f02	352	list_for_each_entry_safe(cur, tmp, &nfnl_acct_net->nfnl_acct_list, head)
94139027 PNA	353	nfnl_acct_try_del(cur);
	354
	355	return 0;
	356	}
	357	acct_name = nla_data(tb[NFACCT_NAME]);
	358
f7583f02	359	list_for_each_entry(cur, &nfnl_acct_net->nfnl_acct_list, head) {
94139027 PNA	360	if (strncmp(cur->name, acct_name, NFACCT_NAME_MAX) != 0)
	361	continue;
	362
	363	ret = nfnl_acct_try_del(cur);
	364	if (ret < 0)
	365	return ret;
	366
	367	break;
	368	}
	369	return ret;
	370	}
	371
	372	static const struct nla_policy nfnl_acct_policy[NFACCT_MAX+1] = {
	373	[NFACCT_NAME] = { .type = NLA_NUL_STRING, .len = NFACCT_NAME_MAX-1 },
	374	[NFACCT_BYTES] = { .type = NLA_U64 },
	375	[NFACCT_PKTS] = { .type = NLA_U64 },
683399ed MP	376	[NFACCT_FLAGS] = { .type = NLA_U32 },
683399ed MP	377	[NFACCT_QUOTA] = { .type = NLA_U64 },
f111f780	378	[NFACCT_FILTER] = {.type = NLA_NESTED },
94139027 PNA	379	};
	380
	381	static const struct nfnl_callback nfnl_acct_cb[NFNL_MSG_ACCT_MAX] = {
50f2db9e PNA	382	[NFNL_MSG_ACCT_NEW] = {
	383	.call = nfnl_acct_new,
	384	.type = NFNL_CB_MUTEX,
	385	.attr_count = NFACCT_MAX,
	386	.policy = nfnl_acct_policy
	387	},
	388	[NFNL_MSG_ACCT_GET] = {
	389	.call = nfnl_acct_get,
	390	.type = NFNL_CB_MUTEX,
	391	.attr_count = NFACCT_MAX,
	392	.policy = nfnl_acct_policy
	393	},
	394	[NFNL_MSG_ACCT_GET_CTRZERO] = {
	395	.call = nfnl_acct_get,
	396	.type = NFNL_CB_MUTEX,
	397	.attr_count = NFACCT_MAX,
	398	.policy = nfnl_acct_policy
	399	},
	400	[NFNL_MSG_ACCT_DEL] = {
	401	.call = nfnl_acct_del,
	402	.type = NFNL_CB_MUTEX,
	403	.attr_count = NFACCT_MAX,
	404	.policy = nfnl_acct_policy
	405	},
94139027 PNA	406	};
	407
	408	static const struct nfnetlink_subsystem nfnl_acct_subsys = {
	409	.name = "acct",
	410	.subsys_id = NFNL_SUBSYS_ACCT,
	411	.cb_count = NFNL_MSG_ACCT_MAX,
	412	.cb = nfnl_acct_cb,
	413	};
	414
	415	MODULE_ALIAS_NFNL_SUBSYS(NFNL_SUBSYS_ACCT);
	416
3499abb2	417	struct nf_acct nfnl_acct_find_get(struct net net, const char *acct_name)
94139027	418	{
f7583f02	419	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net);
94139027 PNA	420	struct nf_acct cur, acct = NULL;
	421
	422	rcu_read_lock();
f7583f02	423	list_for_each_entry_rcu(cur, &nfnl_acct_net->nfnl_acct_list, head) {
94139027 PNA	424	if (strncmp(cur->name, acct_name, NFACCT_NAME_MAX)!= 0)
	425	continue;
	426
	427	if (!try_module_get(THIS_MODULE))
	428	goto err;
	429
b54ab92b	430	if (!refcount_inc_not_zero(&cur->refcnt)) {
94139027 PNA	431	module_put(THIS_MODULE);
	432	goto err;
	433	}
	434
	435	acct = cur;
	436	break;
	437	}
	438	err:
	439	rcu_read_unlock();
	440	return acct;
	441	}
	442	EXPORT_SYMBOL_GPL(nfnl_acct_find_get);
	443
	444	void nfnl_acct_put(struct nf_acct *acct)
	445	{
b54ab92b	446	if (refcount_dec_and_test(&acct->refcnt))
3499abb2 AS	447	kfree_rcu(acct, rcu_head);
3499abb2 AS	448
94139027 PNA	449	module_put(THIS_MODULE);
	450	}
	451	EXPORT_SYMBOL_GPL(nfnl_acct_put);
	452
	453	void nfnl_acct_update(const struct sk_buff skb, struct nf_acct nfacct)
	454	{
	455	atomic64_inc(&nfacct->pkts);
	456	atomic64_add(skb->len, &nfacct->bytes);
	457	}
	458	EXPORT_SYMBOL_GPL(nfnl_acct_update);
	459
aca30018	460	static void nfnl_overquota_report(struct net net, struct nf_acct nfacct)
683399ed MP	461	{
	462	int ret;
	463	struct sk_buff *skb;
	464
	465	skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC);
	466	if (skb == NULL)
	467	return;
	468
	469	ret = nfnl_acct_fill_info(skb, 0, 0, NFNL_MSG_ACCT_OVERQUOTA, 0,
	470	nfacct);
	471	if (ret <= 0) {
	472	kfree_skb(skb);
	473	return;
	474	}
237c609f	475	nfnetlink_broadcast(net, skb, 0, NFNLGRP_ACCT_QUOTA, GFP_ATOMIC);
683399ed MP	476	}
683399ed MP	477
cceae76e	478	int nfnl_acct_overquota(struct net net, struct nf_acct nfacct)
683399ed MP	479	{
	480	u64 now;
	481	u64 *quota;
	482	int ret = NFACCT_UNDERQUOTA;
	483
	484	/* no place here if we don't have a quota */
	485	if (!(nfacct->flags & NFACCT_F_QUOTA))
	486	return NFACCT_NO_QUOTA;
	487
	488	quota = (u64 *)nfacct->data;
	489	now = (nfacct->flags & NFACCT_F_QUOTA_PKTS) ?
	490	atomic64_read(&nfacct->pkts) : atomic64_read(&nfacct->bytes);
	491
	492	ret = now > *quota;
	493
	494	if (now >= *quota &&
b6d04688	495	!test_and_set_bit(NFACCT_OVERQUOTA_BIT, &nfacct->flags)) {
aca30018	496	nfnl_overquota_report(net, nfacct);
683399ed MP	497	}
	498
	499	return ret;
	500	}
	501	EXPORT_SYMBOL_GPL(nfnl_acct_overquota);
	502
3499abb2 AS	503	static int __net_init nfnl_acct_net_init(struct net *net)
3499abb2 AS	504	{
f7583f02	505	INIT_LIST_HEAD(&nfnl_acct_pernet(net)->nfnl_acct_list);
3499abb2 AS	506
	507	return 0;
	508	}
	509
	510	static void __net_exit nfnl_acct_net_exit(struct net *net)
	511	{
f7583f02	512	struct nfnl_acct_net *nfnl_acct_net = nfnl_acct_pernet(net);
3499abb2 AS	513	struct nf_acct cur, tmp;
3499abb2 AS	514
f7583f02	515	list_for_each_entry_safe(cur, tmp, &nfnl_acct_net->nfnl_acct_list, head) {
3499abb2 AS	516	list_del_rcu(&cur->head);
3499abb2 AS	517
b54ab92b	518	if (refcount_dec_and_test(&cur->refcnt))
3499abb2 AS	519	kfree_rcu(cur, rcu_head);
	520	}
	521	}
	522
	523	static struct pernet_operations nfnl_acct_ops = {
	524	.init = nfnl_acct_net_init,
	525	.exit = nfnl_acct_net_exit,
f7583f02 WS	526	.id = &nfnl_acct_net_id,
f7583f02 WS	527	.size = sizeof(struct nfnl_acct_net),
3499abb2 AS	528	};
3499abb2 AS	529
94139027 PNA	530	static int __init nfnl_acct_init(void)
	531	{
	532	int ret;
	533
3499abb2 AS	534	ret = register_pernet_subsys(&nfnl_acct_ops);
	535	if (ret < 0) {
	536	pr_err("nfnl_acct_init: failed to register pernet ops\n");
	537	goto err_out;
	538	}
	539
94139027 PNA	540	ret = nfnetlink_subsys_register(&nfnl_acct_subsys);
	541	if (ret < 0) {
	542	pr_err("nfnl_acct_init: cannot register with nfnetlink.\n");
3499abb2	543	goto cleanup_pernet;
94139027 PNA	544	}
94139027 PNA	545	return 0;
3499abb2 AS	546
	547	cleanup_pernet:
	548	unregister_pernet_subsys(&nfnl_acct_ops);
94139027 PNA	549	err_out:
	550	return ret;
	551	}
	552
	553	static void __exit nfnl_acct_exit(void)
	554	{
94139027	555	nfnetlink_subsys_unregister(&nfnl_acct_subsys);
3499abb2	556	unregister_pernet_subsys(&nfnl_acct_ops);
94139027 PNA	557	}
	558
	559	module_init(nfnl_acct_init);
	560	module_exit(nfnl_acct_exit);