Release 4.11 net/sched/cls_api.c

/*
 * net/sched/cls_api.c  Packet classifier API.
 *
 *              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.
 *
 * Authors:     Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
 *
 * Changes:
 *
 * Eduardo J. Blanco <ejbs@netlabs.com.uy> :990222: kmod support
 *
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/errno.h>
#include <linux/err.h>
#include <linux/skbuff.h>
#include <linux/init.h>
#include <linux/kmod.h>
#include <linux/err.h>
#include <linux/slab.h>
#include <net/net_namespace.h>
#include <net/sock.h>
#include <net/netlink.h>
#include <net/pkt_sched.h>
#include <net/pkt_cls.h>

/* The list of all installed classifier types */
static LIST_HEAD(tcf_proto_base);

/* Protects list of registered TC modules. It is pure SMP lock. */
static DEFINE_RWLOCK(cls_mod_lock);

/* Find classifier type by string name */



static const struct tcf_proto_ops *tcf_proto_lookup_ops(const char *kind)
{
	const struct tcf_proto_ops *t, *res = NULL;

	if (kind) {
		read_lock(&cls_mod_lock);
		list_for_each_entry(t, &tcf_proto_base, head) {
			if (strcmp(kind, t->kind) == 0) {
				if (try_module_get(t->owner))
					res = t;
				break;
			}
		}
		read_unlock(&cls_mod_lock);
	}
	return res;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Linus Torvalds (pre-git)
52
61.18%
2
25.00%
Patrick McHardy
13
15.29%
1
12.50%
Américo Wang
8
9.41%
1
12.50%
Eric Dumazet
8
9.41%
2
25.00%
Jiri Pirko
3
3.53%
1
12.50%
Chris Wright
1
1.18%
1
12.50%
Total
85
100.00%
8
100.00%

/* Register(unregister) new classifier type */



int register_tcf_proto_ops(struct tcf_proto_ops *ops)
{
	struct tcf_proto_ops *t;
	int rc = -EEXIST;

	write_lock(&cls_mod_lock);
	list_for_each_entry(t, &tcf_proto_base, head)
		if (!strcmp(ops->kind, t->kind))
			goto out;

	list_add_tail(&ops->head, &tcf_proto_base);
	rc = 0;
out:
	write_unlock(&cls_mod_lock);
	return rc;
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
Linus Torvalds (pre-git)
45
57.69%
2
50.00%
Arnaldo Carvalho de Melo
17
21.79%
1
25.00%
Américo Wang
16
20.51%
1
25.00%
Total
78
100.00%
4
100.00%


EXPORT_SYMBOL(register_tcf_proto_ops);



int unregister_tcf_proto_ops(struct tcf_proto_ops *ops)
{
	struct tcf_proto_ops *t;
	int rc = -ENOENT;

	/* Wait for outstanding call_rcu()s, if any, from a
         * tcf_proto_ops's destroy() handler.
         */
	rcu_barrier();

	write_lock(&cls_mod_lock);
	list_for_each_entry(t, &tcf_proto_base, head) {
		if (t == ops) {
			list_del(&t->head);
			rc = 0;
			break;
		}
	}
	write_unlock(&cls_mod_lock);
	return rc;
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
Linus Torvalds (pre-git)
38
53.52%
2
33.33%
Américo Wang
13
18.31%
1
16.67%
Arnaldo Carvalho de Melo
11
15.49%
1
16.67%
Eric Dumazet
5
7.04%
1
16.67%
Daniel Borkmann
4
5.63%
1
16.67%
Total
71
100.00%
6
100.00%


EXPORT_SYMBOL(unregister_tcf_proto_ops);

static int tfilter_notify(struct net *net, struct sk_buff *oskb,
			  struct nlmsghdr *n, struct tcf_proto *tp,
			  unsigned long fh, int event, bool unicast);



static void tfilter_notify_chain(struct net *net, struct sk_buff *oskb,
				 struct nlmsghdr *n,
				 struct tcf_proto __rcu **chain, int event)
{
	struct tcf_proto __rcu **it_chain;
	struct tcf_proto *tp;

	for (it_chain = chain; (tp = rtnl_dereference(*it_chain)) != NULL;
	     it_chain = &tp->next)
		tfilter_notify(net, oskb, n, tp, 0, event, false);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Daniel Borkmann
82
96.47%
1
33.33%
Eric Dumazet
2
2.35%
1
33.33%
Roman Mashak
1
1.18%
1
33.33%
Total
85
100.00%
3
100.00%

/* Select new prio value from the range, managed by kernel. */



static inline u32 tcf_auto_prio(struct tcf_proto *tp)
{
	u32 first = TC_H_MAKE(0xC0000000U, 0U);

	if (tp)
		first = tp->prio - 1;

	return first;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Linus Torvalds (pre-git)
36
97.30%
1
50.00%
Stephen Hemminger
1
2.70%
1
50.00%
Total
37
100.00%
2
100.00%



static struct tcf_proto *tcf_proto_create(const char *kind, u32 protocol,
					  u32 prio, u32 parent, struct Qdisc *q)
{
	struct tcf_proto *tp;
	int err;

	tp = kzalloc(sizeof(*tp), GFP_KERNEL);
	if (!tp)
		return ERR_PTR(-ENOBUFS);

	err = -ENOENT;
	tp->ops = tcf_proto_lookup_ops(kind);
	if (!tp->ops) {
#ifdef CONFIG_MODULES
		rtnl_unlock();
		request_module("cls_%s", kind);
		rtnl_lock();
		tp->ops = tcf_proto_lookup_ops(kind);
		/* We dropped the RTNL semaphore in order to perform
                 * the module load. So, even if we succeeded in loading
                 * the module we have to replay the request. We indicate
                 * this using -EAGAIN.
                 */
		if (tp->ops) {
			module_put(tp->ops->owner);
			err = -EAGAIN;
		} else {
			err = -ENOENT;
		}
		goto errout;
#endif
	}
	tp->classify = tp->ops->classify;
	tp->protocol = protocol;
	tp->prio = prio;
	tp->classid = parent;
	tp->q = q;

	err = tp->ops->init(tp);
	if (err) {
		module_put(tp->ops->owner);
		goto errout;
	}
	return tp;

errout:
	kfree(tp);
	return ERR_PTR(err);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Jiri Pirko
223
100.00%
1
100.00%
Total
223
100.00%
1
100.00%



static bool tcf_proto_destroy(struct tcf_proto *tp, bool force)
{
	if (tp->ops->destroy(tp, force)) {
		module_put(tp->ops->owner);
		kfree_rcu(tp, rcu);
		return true;
	}
	return false;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Jiri Pirko
51
100.00%
1
100.00%
Total
51
100.00%
1
100.00%



void tcf_destroy_chain(struct tcf_proto __rcu **fl)
{
	struct tcf_proto *tp;

	while ((tp = rtnl_dereference(*fl)) != NULL) {
		RCU_INIT_POINTER(*fl, tp->next);
		tcf_proto_destroy(tp, true);
	}
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
Jiri Pirko
50
100.00%
1
100.00%
Total
50
100.00%
1
100.00%


EXPORT_SYMBOL(tcf_destroy_chain);

/* Add/change/delete/get a filter node */



static int tc_ctl_tfilter(struct sk_buff *skb, struct nlmsghdr *n)
{
	struct net *net = sock_net(skb->sk);
	struct nlattr *tca[TCA_MAX + 1];
	struct tcmsg *t;
	u32 protocol;
	u32 prio;
	u32 nprio;
	u32 parent;
	struct net_device *dev;
	struct Qdisc  *q;
	struct tcf_proto __rcu **back;
	struct tcf_proto __rcu **chain;
	struct tcf_proto *next;
	struct tcf_proto *tp;
	const struct Qdisc_class_ops *cops;
	unsigned long cl;
	unsigned long fh;
	int err;
	int tp_created;

	if ((n->nlmsg_type != RTM_GETTFILTER) &&
	    !netlink_ns_capable(skb, net->user_ns, CAP_NET_ADMIN))
		return -EPERM;

replay:
	tp_created = 0;

	err = nlmsg_parse(n, sizeof(*t), tca, TCA_MAX, NULL);
	if (err < 0)
		return err;

	t = nlmsg_data(n);
	protocol = TC_H_MIN(t->tcm_info);
	prio = TC_H_MAJ(t->tcm_info);
	nprio = prio;
	parent = t->tcm_parent;
	cl = 0;

	if (prio == 0) {
		switch (n->nlmsg_type) {
		case RTM_DELTFILTER:
			if (protocol || t->tcm_handle || tca[TCA_KIND])
				return -ENOENT;
			break;
		case RTM_NEWTFILTER:
			/* If no priority is provided by the user,
                         * we allocate one.
                         */
			if (n->nlmsg_flags & NLM_F_CREATE) {
				prio = TC_H_MAKE(0x80000000U, 0U);
				break;
			}
			/* fall-through */
		default:
			return -ENOENT;
		}
	}

	/* Find head of filter chain. */

	/* Find link */
	dev = __dev_get_by_index(net, t->tcm_ifindex);
	if (dev == NULL)
		return -ENODEV;

	/* Find qdisc */
	if (!parent) {
		q = dev->qdisc;
		parent = q->handle;
	} else {
		q = qdisc_lookup(dev, TC_H_MAJ(t->tcm_parent));
		if (q == NULL)
			return -EINVAL;
	}

	/* Is it classful? */
	cops = q->ops->cl_ops;
	if (!cops)
		return -EINVAL;

	if (cops->tcf_chain == NULL)
		return -EOPNOTSUPP;

	/* Do we search for filter, attached to class? */
	if (TC_H_MIN(parent)) {
		cl = cops->get(q, parent);
		if (cl == 0)
			return -ENOENT;
	}

	/* And the last stroke */
	chain = cops->tcf_chain(q, cl);
	if (chain == NULL) {
		err = -EINVAL;
		goto errout;
	}
	if (n->nlmsg_type == RTM_DELTFILTER && prio == 0) {
		tfilter_notify_chain(net, skb, n, chain, RTM_DELTFILTER);
		tcf_destroy_chain(chain);
		err = 0;
		goto errout;
	}

	/* Check the chain for existence of proto-tcf with this priority */
	for (back = chain;
	     (tp = rtnl_dereference(*back)) != NULL;
	     back = &tp->next) {
		if (tp->prio >= prio) {
			if (tp->prio == prio) {
				if (!nprio ||
				    (tp->protocol != protocol && protocol)) {
					err = -EINVAL;
					goto errout;
				}
			} else {
				tp = NULL;
			}
			break;
		}
	}

	if (tp == NULL) {
		/* Proto-tcf does not exist, create new one */

		if (tca[TCA_KIND] == NULL || !protocol) {
			err = -EINVAL;
			goto errout;
		}

		if (n->nlmsg_type != RTM_NEWTFILTER ||
		    !(n->nlmsg_flags & NLM_F_CREATE)) {
			err = -ENOENT;
			goto errout;
		}

		if (!nprio)
			nprio = TC_H_MAJ(tcf_auto_prio(rtnl_dereference(*back)));

		tp = tcf_proto_create(nla_data(tca[TCA_KIND]),
				      protocol, nprio, parent, q);
		if (IS_ERR(tp)) {
			err = PTR_ERR(tp);
			goto errout;
		}
		tp_created = 1;
	} else if (tca[TCA_KIND] && nla_strcmp(tca[TCA_KIND], tp->ops->kind)) {
		err = -EINVAL;
		goto errout;
	}

	fh = tp->ops->get(tp, t->tcm_handle);

	if (fh == 0) {
		if (n->nlmsg_type == RTM_DELTFILTER && t->tcm_handle == 0) {
			next = rtnl_dereference(tp->next);
			RCU_INIT_POINTER(*back, next);
			tfilter_notify(net, skb, n, tp, fh,
				       RTM_DELTFILTER, false);
			tcf_proto_destroy(tp, true);
			err = 0;
			goto errout;
		}

		if (n->nlmsg_type != RTM_NEWTFILTER ||
		    !(n->nlmsg_flags & NLM_F_CREATE)) {
			err = -ENOENT;
			goto errout;
		}
	} else {
		switch (n->nlmsg_type) {
		case RTM_NEWTFILTER:
			if (n->nlmsg_flags & NLM_F_EXCL) {
				if (tp_created)
					tcf_proto_destroy(tp, true);
				err = -EEXIST;
				goto errout;
			}
			break;
		case RTM_DELTFILTER:
			err = tp->ops->delete(tp, fh);
			if (err)
				goto errout;
			next = rtnl_dereference(tp->next);
			tfilter_notify(net, skb, n, tp, t->tcm_handle,
				       RTM_DELTFILTER, false);
			if (tcf_proto_destroy(tp, false))
				RCU_INIT_POINTER(*back, next);
			goto errout;
		case RTM_GETTFILTER:
			err = tfilter_notify(net, skb, n, tp, fh,
					     RTM_NEWTFILTER, true);
			goto errout;
		default:
			err = -EINVAL;
			goto errout;
		}
	}

	err = tp->ops->change(net, skb, tp, cl, t->tcm_handle, tca, &fh,
			      n->nlmsg_flags & NLM_F_CREATE ? TCA_ACT_NOREPLACE : TCA_ACT_REPLACE);
	if (err == 0) {
		if (tp_created) {
			RCU_INIT_POINTER(tp->next, rtnl_dereference(*back));
			rcu_assign_pointer(*back, tp);
		}
		tfilter_notify(net, skb, n, tp, fh, RTM_NEWTFILTER, false);
	} else {
		if (tp_created)
			tcf_proto_destroy(tp, true);
	}

errout:
	if (cl)
		cops->put(q, cl);
	if (err == -EAGAIN)
		/* Replay the request. */
		goto replay;
	return err;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Linus Torvalds (pre-git)
604
54.86%
6
13.33%
Jiri Pirko
91
8.27%
6
13.33%
Daniel Borkmann
82
7.45%
3
6.67%
Patrick McHardy
75
6.81%
7
15.56%
Minoru Usui
46
4.18%
1
2.22%
Cong Wang
32
2.91%
2
4.44%
Jamal Hadi Salim
31
2.82%
2
4.44%
Hong Zhi Guo
28
2.54%
1
2.22%
John Fastabend
25
2.27%
1
2.22%
Eric W. Biedermann
24
2.18%
4
8.89%
Eric Dumazet
14
1.27%
3
6.67%
Stephen Hemminger
13
1.18%
1
2.22%
Thomas Goff
9
0.82%
1
2.22%
Denis V. Lunev
9
0.82%
1
2.22%
Américo Wang
7
0.64%
1
2.22%
Stéphane Graber
4
0.36%
1
2.22%
Hideaki Yoshifuji / 吉藤英明
3
0.27%
1
2.22%
Benjamin LaHaise
2
0.18%
1
2.22%
David S. Miller
2
0.18%
2
4.44%
Total
1101
100.00%
45
100.00%



static int tcf_fill_node(struct net *net, struct sk_buff *skb,
			 struct tcf_proto *tp, unsigned long fh, u32 portid,
			 u32 seq, u16 flags, int event)
{
	struct tcmsg *tcm;
	struct nlmsghdr  *nlh;
	unsigned char *b = skb_tail_pointer(skb);

	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*tcm), flags);
	if (!nlh)
		goto out_nlmsg_trim;
	tcm = nlmsg_data(nlh);
	tcm->tcm_family = AF_UNSPEC;
	tcm->tcm__pad1 = 0;
	tcm->tcm__pad2 = 0;
	tcm->tcm_ifindex = qdisc_dev(tp->q)->ifindex;
	tcm->tcm_parent = tp->classid;
	tcm->tcm_info = TC_H_MAKE(tp->prio, tp->protocol);
	if (nla_put_string(skb, TCA_KIND, tp->ops->kind))
		goto nla_put_failure;
	tcm->tcm_handle = fh;
	if (RTM_DELTFILTER != event) {
		tcm->tcm_handle = 0;
		if (tp->ops->dump && tp->ops->dump(net, tp, fh, skb, tcm) < 0)
			goto nla_put_failure;
	}
	nlh->nlmsg_len = skb_tail_pointer(skb) - b;
	return skb->len;

out_nlmsg_trim:
nla_put_failure:
	nlmsg_trim(skb, b);
	return -1;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Linus Torvalds (pre-git)
172
69.92%
1
7.69%
Jamal Hadi Salim
23
9.35%
2
15.38%
David S. Miller
21
8.54%
3
23.08%
Patrick McHardy
13
5.28%
2
15.38%
Américo Wang
7
2.85%
1
7.69%
Arnaldo Carvalho de Melo
7
2.85%
2
15.38%
Eric W. Biedermann
2
0.81%
1
7.69%
Jiri Pirko
1
0.41%
1
7.69%
Total
246
100.00%
13
100.00%



static int tfilter_notify(struct net *net, struct sk_buff *oskb,
			  struct nlmsghdr *n, struct tcf_proto *tp,
			  unsigned long fh, int event, bool unicast)
{
	struct sk_buff *skb;
	u32 portid = oskb ? NETLINK_CB(oskb).portid : 0;

	skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
	if (!skb)
		return -ENOBUFS;

	if (tcf_fill_node(net, skb, tp, fh, portid, n->nlmsg_seq,
			  n->nlmsg_flags, event) <= 0) {
		kfree_skb(skb);
		return -EINVAL;
	}

	if (unicast)
		return netlink_unicast(net->rtnl, skb, portid, MSG_DONTWAIT);

	return rtnetlink_send(skb, net, portid, RTNLGRP_TC,
			      n->nlmsg_flags & NLM_F_ECHO);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Linus Torvalds (pre-git)
109
74.15%
2
22.22%
Eric Dumazet
21
14.29%
1
11.11%
Thomas Goff
6
4.08%
1
11.11%
Eric W. Biedermann
4
2.72%
1
11.11%
Roman Mashak
3
2.04%
1
11.11%
Américo Wang
2
1.36%
1
11.11%
Denis V. Lunev
1
0.68%
1
11.11%
Patrick McHardy
1
0.68%
1
11.11%
Total
147
100.00%
9
100.00%


struct tcf_dump_args {
	
struct tcf_walker w;
	
struct sk_buff *skb;
	
struct netlink_callback *cb;
};



static int tcf_node_dump(struct tcf_proto *tp, unsigned long n,
			 struct tcf_walker *arg)
{
	struct tcf_dump_args *a = (void *)arg;
	struct net *net = sock_net(a->skb->sk);

	return tcf_fill_node(net, a->skb, tp, n, NETLINK_CB(a->cb->skb).portid,
			     a->cb->nlh->nlmsg_seq, NLM_F_MULTI,
			     RTM_NEWTFILTER);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Linus Torvalds (pre-git)
65
79.27%
1
33.33%
Américo Wang
16
19.51%
1
33.33%
Eric W. Biedermann
1
1.22%
1
33.33%
Total
82
100.00%
3
100.00%

/* called with RTNL */


static int tc_dump_tfilter(struct sk_buff *skb, struct netlink_callback *cb)
{
	struct net *net = sock_net(skb->sk);
	int t;
	int s_t;
	struct net_device *dev;
	struct Qdisc *q;
	struct tcf_proto *tp, __rcu **chain;
	struct tcmsg *tcm = nlmsg_data(cb->nlh);
	unsigned long cl = 0;
	const struct Qdisc_class_ops *cops;
	struct tcf_dump_args arg;

	if (nlmsg_len(cb->nlh) < sizeof(*tcm))
		return skb->len;
	dev = __dev_get_by_index(net, tcm->tcm_ifindex);
	if (!dev)
		return skb->len;

	if (!tcm->tcm_parent)
		q = dev->qdisc;
	else
		q = qdisc_lookup(dev, TC_H_MAJ(tcm->tcm_parent));
	if (!q)
		goto out;
	cops = q->ops->cl_ops;
	if (!cops)
		goto errout;
	if (cops->tcf_chain == NULL)
		goto errout;
	if (TC_H_MIN(tcm->tcm_parent)) {
		cl = cops->get(q, tcm->tcm_parent);
		if (cl == 0)
			goto errout;
	}
	chain = cops->tcf_chain(q, cl);
	if (chain == NULL)
		goto errout;

	s_t = cb->args[0];

	for (tp = rtnl_dereference(*chain), t = 0;
	     tp; tp = rtnl_dereference(tp->next), t++) {
		if (t < s_t)
			continue;
		if (TC_H_MAJ(tcm->tcm_info) &&
		    TC_H_MAJ(tcm->tcm_info) != tp->prio)
			continue;
		if (TC_H_MIN(tcm->tcm_info) &&
		    TC_H_MIN(tcm->tcm_info) != tp->protocol)
			continue;
		if (t > s_t)
			memset(&cb->args[1], 0,
			       sizeof(cb->args)-sizeof(cb->args[0]));
		if (cb->args[1] == 0) {
			if (tcf_fill_node(net, skb, tp, 0,
					  NETLINK_CB(cb->skb).portid,
					  cb->nlh->nlmsg_seq, NLM_F_MULTI,
					  RTM_NEWTFILTER) <= 0)
				break;

			cb->args[1] = 1;
		}
		if (tp->ops->walk == NULL)
			continue;
		arg.w.fn = tcf_node_dump;
		arg.skb = skb;
		arg.cb = cb;
		arg.w.stop = 0;
		arg.w.skip = cb->args[1] - 1;
		arg.w.count = 0;
		tp->ops->walk(tp, &arg.w);
		cb->args[1] = arg.w.count + 1;
		if (arg.w.stop)
			break;
	}

	cb->args[0] = t;

errout:
	if (cl)
		cops->put(q, cl);
out:
	return skb->len;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Linus Torvalds (pre-git)
492
89.29%
3
16.67%
Patrick McHardy
13
2.36%
2
11.11%
Eric Dumazet
12
2.18%
3
16.67%
Denis V. Lunev
9
1.63%
1
5.56%
John Fastabend
7
1.27%
1
5.56%
Arnaldo Carvalho de Melo
6
1.09%
1
5.56%
Hong Zhi Guo
3
0.54%
1
5.56%
Hideaki Yoshifuji / 吉藤英明
3
0.54%
1
5.56%
Eric W. Biedermann
2
0.36%
2
11.11%
Américo Wang
2
0.36%
1
5.56%
Thomas Goff
1
0.18%
1
5.56%
David S. Miller
1
0.18%
1
5.56%
Total
551
100.00%
18
100.00%



void tcf_exts_destroy(struct tcf_exts *exts)
{
#ifdef CONFIG_NET_CLS_ACT
	LIST_HEAD(actions);

	tcf_exts_to_list(exts, &actions);
	tcf_action_destroy(&actions, TCA_ACT_UNBIND);
	kfree(exts->actions);
	exts->nr_actions = 0;
#endif
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
Américo Wang
25
51.02%
2
50.00%
Thomas Graf
23
46.94%
1
25.00%
Stephen Hemminger
1
2.04%
1
25.00%
Total
49
100.00%
4
100.00%


EXPORT_SYMBOL(tcf_exts_destroy);



int tcf_exts_validate(struct net *net, struct tcf_proto *tp, struct nlattr **tb,
		      struct nlattr *rate_tlv, struct tcf_exts *exts, bool ovr)
{
#ifdef CONFIG_NET_CLS_ACT
	{
		struct tc_action *act;

		if (exts->police && tb[exts->police]) {