em_meta.c

linux 内核源代码

/*
 * net/sched/em_meta.c	Metadata ematch
 *
 *		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:	Thomas Graf <[e-mail]>
 *
 * ==========================================================================
 *
 * 	The metadata ematch compares two meta objects where each object
 * 	represents either a meta value stored in the kernel or a static
 * 	value provided by userspace. The objects are not provided by
 * 	userspace itself but rather a definition providing the information
 * 	to build them. Every object is of a certain type which must be
 * 	equal to the object it is being compared to.
 *
 * 	The definition of a objects conists of the type (meta type), a
 * 	identifier (meta id) and additional type specific information.
 * 	The meta id is either TCF_META_TYPE_VALUE for values provided by
 * 	userspace or a index to the meta operations table consisting of
 * 	function pointers to type specific meta data collectors returning
 * 	the value of the requested meta value.
 *
 * 	         lvalue                                   rvalue
 * 	      +-----------+                           +-----------+
 * 	      | type: INT |                           | type: INT |
 * 	 def  | id: DEV   |                           | id: VALUE |
 * 	      | data:     |                           | data: 3   |
 * 	      +-----------+                           +-----------+
 * 	            |                                       |
 * 	            ---> meta_ops[INT][DEV](...)            |
 *	                      |                             |
 * 	            -----------                             |
 * 	            V                                       V
 * 	      +-----------+                           +-----------+
 * 	      | type: INT |                           | type: INT |
 * 	 obj  | id: DEV |                             | id: VALUE |
 * 	      | data: 2   |<--data got filled out     | data: 3   |
 * 	      +-----------+                           +-----------+
 * 	            |                                         |
 * 	            --------------> 2  equals 3 <--------------
 *
 * 	This is a simplified schema, the complexity varies depending
 * 	on the meta type. Obviously, the length of the data must also
 * 	be provided for non-numeric types.
 *
 * 	Additionaly, type dependant modifiers such as shift operators
 * 	or mask may be applied to extend the functionaliy. As of now,
 * 	the variable length type supports shifting the byte string to
 * 	the right, eating up any number of octets and thus supporting
 * 	wildcard interface name comparisons such as "ppp%" matching
 * 	ppp0..9.
 *
 * 	NOTE: Certain meta values depend on other subsystems and are
 * 	      only available if that subsystem is enabled in the kernel.
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/sched.h>
#include <linux/string.h>
#include <linux/skbuff.h>
#include <linux/random.h>
#include <linux/tc_ematch/tc_em_meta.h>
#include <net/dst.h>
#include <net/route.h>
#include <net/pkt_cls.h>
#include <net/sock.h>

struct meta_obj
{
	unsigned long		value;
	unsigned int		len;
};

struct meta_value
{
	struct tcf_meta_val	hdr;
	unsigned long		val;
	unsigned int		len;
};

struct meta_match
{
	struct meta_value	lvalue;
	struct meta_value	rvalue;
};

static inline int meta_id(struct meta_value *v)
{
	return TCF_META_ID(v->hdr.kind);
}

static inline int meta_type(struct meta_value *v)
{
	return TCF_META_TYPE(v->hdr.kind);
}

#define META_COLLECTOR(FUNC) static void meta_##FUNC(struct sk_buff *skb, \
	struct tcf_pkt_info *info, struct meta_value *v, \
	struct meta_obj *dst, int *err)

/**************************************************************************
 * System status & misc
 **************************************************************************/

META_COLLECTOR(int_random)
{
	get_random_bytes(&dst->value, sizeof(dst->value));
}

static inline unsigned long fixed_loadavg(int load)
{
	int rnd_load = load + (FIXED_1/200);
	int rnd_frac = ((rnd_load & (FIXED_1-1)) * 100) >> FSHIFT;

	return ((rnd_load >> FSHIFT) * 100) + rnd_frac;
}

META_COLLECTOR(int_loadavg_0)
{
	dst->value = fixed_loadavg(avenrun[0]);
}

META_COLLECTOR(int_loadavg_1)
{
	dst->value = fixed_loadavg(avenrun[1]);
}

META_COLLECTOR(int_loadavg_2)
{
	dst->value = fixed_loadavg(avenrun[2]);
}

/**************************************************************************
 * Device names & indices
 **************************************************************************/

static inline int int_dev(struct net_device *dev, struct meta_obj *dst)
{
	if (unlikely(dev == NULL))
		return -1;

	dst->value = dev->ifindex;
	return 0;
}

static inline int var_dev(struct net_device *dev, struct meta_obj *dst)
{
	if (unlikely(dev == NULL))
		return -1;

	dst->value = (unsigned long) dev->name;
	dst->len = strlen(dev->name);
	return 0;
}

META_COLLECTOR(int_dev)
{
	*err = int_dev(skb->dev, dst);
}

META_COLLECTOR(var_dev)
{
	*err = var_dev(skb->dev, dst);
}

/**************************************************************************
 * skb attributes
 **************************************************************************/

META_COLLECTOR(int_priority)
{
	dst->value = skb->priority;
}

META_COLLECTOR(int_protocol)
{
	/* Let userspace take care of the byte ordering */
	dst->value = skb->protocol;
}

META_COLLECTOR(int_pkttype)
{
	dst->value = skb->pkt_type;
}

META_COLLECTOR(int_pktlen)
{
	dst->value = skb->len;
}

META_COLLECTOR(int_datalen)
{
	dst->value = skb->data_len;
}

META_COLLECTOR(int_maclen)
{
	dst->value = skb->mac_len;
}

/**************************************************************************
 * Netfilter
 **************************************************************************/

META_COLLECTOR(int_mark)
{
	dst->value = skb->mark;
}

/**************************************************************************
 * Traffic Control
 **************************************************************************/

META_COLLECTOR(int_tcindex)
{
	dst->value = skb->tc_index;
}

/**************************************************************************
 * Routing
 **************************************************************************/

META_COLLECTOR(int_rtclassid)
{
	if (unlikely(skb->dst == NULL))
		*err = -1;
	else
#ifdef CONFIG_NET_CLS_ROUTE
		dst->value = skb->dst->tclassid;
#else
		dst->value = 0;
#endif
}

META_COLLECTOR(int_rtiif)
{
	if (unlikely(skb->dst == NULL))
		*err = -1;
	else
		dst->value = ((struct rtable*) skb->dst)->fl.iif;
}

/**************************************************************************
 * Socket Attributes
 **************************************************************************/

#define SKIP_NONLOCAL(skb)			\
	if (unlikely(skb->sk == NULL)) {	\
		*err = -1;			\
		return;				\
	}

META_COLLECTOR(int_sk_family)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_family;
}

META_COLLECTOR(int_sk_state)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_state;
}

META_COLLECTOR(int_sk_reuse)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_reuse;
}

META_COLLECTOR(int_sk_bound_if)
{
	SKIP_NONLOCAL(skb);
	/* No error if bound_dev_if is 0, legal userspace check */
	dst->value = skb->sk->sk_bound_dev_if;
}

META_COLLECTOR(var_sk_bound_if)
{
	SKIP_NONLOCAL(skb);

	 if (skb->sk->sk_bound_dev_if == 0) {
		dst->value = (unsigned long) "any";
		dst->len = 3;
	 } else  {
		struct net_device *dev;

		dev = dev_get_by_index(&init_net, skb->sk->sk_bound_dev_if);
		*err = var_dev(dev, dst);
		if (dev)
			dev_put(dev);
	 }
}

META_COLLECTOR(int_sk_refcnt)
{
	SKIP_NONLOCAL(skb);
	dst->value = atomic_read(&skb->sk->sk_refcnt);
}

META_COLLECTOR(int_sk_rcvbuf)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_rcvbuf;
}

META_COLLECTOR(int_sk_shutdown)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_shutdown;
}

META_COLLECTOR(int_sk_proto)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_protocol;
}

META_COLLECTOR(int_sk_type)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_type;
}

META_COLLECTOR(int_sk_rmem_alloc)
{
	SKIP_NONLOCAL(skb);
	dst->value = atomic_read(&skb->sk->sk_rmem_alloc);
}

META_COLLECTOR(int_sk_wmem_alloc)
{
	SKIP_NONLOCAL(skb);
	dst->value = atomic_read(&skb->sk->sk_wmem_alloc);
}

META_COLLECTOR(int_sk_omem_alloc)
{
	SKIP_NONLOCAL(skb);
	dst->value = atomic_read(&skb->sk->sk_omem_alloc);
}

META_COLLECTOR(int_sk_rcv_qlen)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_receive_queue.qlen;
}

META_COLLECTOR(int_sk_snd_qlen)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_write_queue.qlen;
}

META_COLLECTOR(int_sk_wmem_queued)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_wmem_queued;
}

META_COLLECTOR(int_sk_fwd_alloc)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_forward_alloc;
}

META_COLLECTOR(int_sk_sndbuf)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_sndbuf;
}

META_COLLECTOR(int_sk_alloc)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_allocation;
}

META_COLLECTOR(int_sk_route_caps)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_route_caps;
}

META_COLLECTOR(int_sk_hash)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_hash;
}

META_COLLECTOR(int_sk_lingertime)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_lingertime / HZ;
}

META_COLLECTOR(int_sk_err_qlen)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_error_queue.qlen;
}

META_COLLECTOR(int_sk_ack_bl)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_ack_backlog;
}

META_COLLECTOR(int_sk_max_ack_bl)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_max_ack_backlog;
}

META_COLLECTOR(int_sk_prio)
{
	SKIP_NONLOCAL(skb);
	dst->value = skb->sk->sk_priority;
}

META_COLLECTOR(int_sk_rcvlowat)