af_key.c

linux 内核源代码

/*
 * net/key/af_key.c	An implementation of PF_KEYv2 sockets.
 *
 *		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:	Maxim Giryaev	<[e-mail]>
 *		David S. Miller	<[e-mail]>
 *		Alexey Kuznetsov <[e-mail]>
 *		Kunihiro Ishiguro <[e-mail]>
 *		Kazunori MIYAZAWA / USAGI Project <[e-mail]>
 *		Derek Atkins <[e-mail]>
 */

#include <linux/capability.h>
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/socket.h>
#include <linux/pfkeyv2.h>
#include <linux/ipsec.h>
#include <linux/skbuff.h>
#include <linux/rtnetlink.h>
#include <linux/in.h>
#include <linux/in6.h>
#include <linux/proc_fs.h>
#include <linux/init.h>
#include <net/net_namespace.h>
#include <net/xfrm.h>

#include <net/sock.h>

#define _X2KEY(x) ((x) == XFRM_INF ? 0 : (x))
#define _KEY2X(x) ((x) == 0 ? XFRM_INF : (x))


/* List of all pfkey sockets. */
static HLIST_HEAD(pfkey_table);
static DECLARE_WAIT_QUEUE_HEAD(pfkey_table_wait);
static DEFINE_RWLOCK(pfkey_table_lock);
static atomic_t pfkey_table_users = ATOMIC_INIT(0);

static atomic_t pfkey_socks_nr = ATOMIC_INIT(0);

struct pfkey_sock {
	/* struct sock must be the first member of struct pfkey_sock */
	struct sock	sk;
	int		registered;
	int		promisc;
};

static inline struct pfkey_sock *pfkey_sk(struct sock *sk)
{
	return (struct pfkey_sock *)sk;
}

static void pfkey_sock_destruct(struct sock *sk)
{
	skb_queue_purge(&sk->sk_receive_queue);

	if (!sock_flag(sk, SOCK_DEAD)) {
		printk("Attempt to release alive pfkey socket: %p\n", sk);
		return;
	}

	BUG_TRAP(!atomic_read(&sk->sk_rmem_alloc));
	BUG_TRAP(!atomic_read(&sk->sk_wmem_alloc));

	atomic_dec(&pfkey_socks_nr);
}

static void pfkey_table_grab(void)
{
	write_lock_bh(&pfkey_table_lock);

	if (atomic_read(&pfkey_table_users)) {
		DECLARE_WAITQUEUE(wait, current);

		add_wait_queue_exclusive(&pfkey_table_wait, &wait);
		for(;;) {
			set_current_state(TASK_UNINTERRUPTIBLE);
			if (atomic_read(&pfkey_table_users) == 0)
				break;
			write_unlock_bh(&pfkey_table_lock);
			schedule();
			write_lock_bh(&pfkey_table_lock);
		}

		__set_current_state(TASK_RUNNING);
		remove_wait_queue(&pfkey_table_wait, &wait);
	}
}

static __inline__ void pfkey_table_ungrab(void)
{
	write_unlock_bh(&pfkey_table_lock);
	wake_up(&pfkey_table_wait);
}

static __inline__ void pfkey_lock_table(void)
{
	/* read_lock() synchronizes us to pfkey_table_grab */

	read_lock(&pfkey_table_lock);
	atomic_inc(&pfkey_table_users);
	read_unlock(&pfkey_table_lock);
}

static __inline__ void pfkey_unlock_table(void)
{
	if (atomic_dec_and_test(&pfkey_table_users))
		wake_up(&pfkey_table_wait);
}


static const struct proto_ops pfkey_ops;

static void pfkey_insert(struct sock *sk)
{
	pfkey_table_grab();
	sk_add_node(sk, &pfkey_table);
	pfkey_table_ungrab();
}

static void pfkey_remove(struct sock *sk)
{
	pfkey_table_grab();
	sk_del_node_init(sk);
	pfkey_table_ungrab();
}

static struct proto key_proto = {
	.name	  = "KEY",
	.owner	  = THIS_MODULE,
	.obj_size = sizeof(struct pfkey_sock),
};

static int pfkey_create(struct net *net, struct socket *sock, int protocol)
{
	struct sock *sk;
	int err;

	if (net != &init_net)
		return -EAFNOSUPPORT;

	if (!capable(CAP_NET_ADMIN))
		return -EPERM;
	if (sock->type != SOCK_RAW)
		return -ESOCKTNOSUPPORT;
	if (protocol != PF_KEY_V2)
		return -EPROTONOSUPPORT;

	err = -ENOMEM;
	sk = sk_alloc(net, PF_KEY, GFP_KERNEL, &key_proto);
	if (sk == NULL)
		goto out;

	sock->ops = &pfkey_ops;
	sock_init_data(sock, sk);

	sk->sk_family = PF_KEY;
	sk->sk_destruct = pfkey_sock_destruct;

	atomic_inc(&pfkey_socks_nr);

	pfkey_insert(sk);

	return 0;
out:
	return err;
}

static int pfkey_release(struct socket *sock)
{
	struct sock *sk = sock->sk;

	if (!sk)
		return 0;

	pfkey_remove(sk);

	sock_orphan(sk);
	sock->sk = NULL;
	skb_queue_purge(&sk->sk_write_queue);
	sock_put(sk);

	return 0;
}

static int pfkey_broadcast_one(struct sk_buff *skb, struct sk_buff **skb2,
			       gfp_t allocation, struct sock *sk)
{
	int err = -ENOBUFS;

	sock_hold(sk);
	if (*skb2 == NULL) {
		if (atomic_read(&skb->users) != 1) {
			*skb2 = skb_clone(skb, allocation);
		} else {
			*skb2 = skb;
			atomic_inc(&skb->users);
		}
	}
	if (*skb2 != NULL) {
		if (atomic_read(&sk->sk_rmem_alloc) <= sk->sk_rcvbuf) {
			skb_orphan(*skb2);
			skb_set_owner_r(*skb2, sk);
			skb_queue_tail(&sk->sk_receive_queue, *skb2);
			sk->sk_data_ready(sk, (*skb2)->len);
			*skb2 = NULL;
			err = 0;
		}
	}
	sock_put(sk);
	return err;
}

/* Send SKB to all pfkey sockets matching selected criteria.  */
#define BROADCAST_ALL		0
#define BROADCAST_ONE		1
#define BROADCAST_REGISTERED	2
#define BROADCAST_PROMISC_ONLY	4
static int pfkey_broadcast(struct sk_buff *skb, gfp_t allocation,
			   int broadcast_flags, struct sock *one_sk)
{
	struct sock *sk;
	struct hlist_node *node;
	struct sk_buff *skb2 = NULL;
	int err = -ESRCH;

	/* XXX Do we need something like netlink_overrun?  I think
	 * XXX PF_KEY socket apps will not mind current behavior.
	 */
	if (!skb)
		return -ENOMEM;

	pfkey_lock_table();
	sk_for_each(sk, node, &pfkey_table) {
		struct pfkey_sock *pfk = pfkey_sk(sk);
		int err2;

		/* Yes, it means that if you are meant to receive this
		 * pfkey message you receive it twice as promiscuous
		 * socket.
		 */
		if (pfk->promisc)
			pfkey_broadcast_one(skb, &skb2, allocation, sk);

		/* the exact target will be processed later */
		if (sk == one_sk)
			continue;
		if (broadcast_flags != BROADCAST_ALL) {
			if (broadcast_flags & BROADCAST_PROMISC_ONLY)
				continue;
			if ((broadcast_flags & BROADCAST_REGISTERED) &&
			    !pfk->registered)
				continue;
			if (broadcast_flags & BROADCAST_ONE)
				continue;
		}

		err2 = pfkey_broadcast_one(skb, &skb2, allocation, sk);

		/* Error is cleare after succecful sending to at least one
		 * registered KM */
		if ((broadcast_flags & BROADCAST_REGISTERED) && err)
			err = err2;
	}
	pfkey_unlock_table();

	if (one_sk != NULL)
		err = pfkey_broadcast_one(skb, &skb2, allocation, one_sk);

	if (skb2)
		kfree_skb(skb2);
	kfree_skb(skb);
	return err;
}

static inline void pfkey_hdr_dup(struct sadb_msg *new, struct sadb_msg *orig)
{
	*new = *orig;
}

static int pfkey_error(struct sadb_msg *orig, int err, struct sock *sk)
{
	struct sk_buff *skb = alloc_skb(sizeof(struct sadb_msg) + 16, GFP_KERNEL);
	struct sadb_msg *hdr;

	if (!skb)
		return -ENOBUFS;

	/* Woe be to the platform trying to support PFKEY yet
	 * having normal errnos outside the 1-255 range, inclusive.
	 */
	err = -err;
	if (err == ERESTARTSYS ||
	    err == ERESTARTNOHAND ||
	    err == ERESTARTNOINTR)
		err = EINTR;
	if (err >= 512)
		err = EINVAL;
	BUG_ON(err <= 0 || err >= 256);

	hdr = (struct sadb_msg *) skb_put(skb, sizeof(struct sadb_msg));
	pfkey_hdr_dup(hdr, orig);
	hdr->sadb_msg_errno = (uint8_t) err;
	hdr->sadb_msg_len = (sizeof(struct sadb_msg) /
			     sizeof(uint64_t));

	pfkey_broadcast(skb, GFP_KERNEL, BROADCAST_ONE, sk);

	return 0;
}

static u8 sadb_ext_min_len[] = {
	[SADB_EXT_RESERVED]		= (u8) 0,
	[SADB_EXT_SA]			= (u8) sizeof(struct sadb_sa),
	[SADB_EXT_LIFETIME_CURRENT]	= (u8) sizeof(struct sadb_lifetime),
	[SADB_EXT_LIFETIME_HARD]	= (u8) sizeof(struct sadb_lifetime),
	[SADB_EXT_LIFETIME_SOFT]	= (u8) sizeof(struct sadb_lifetime),
	[SADB_EXT_ADDRESS_SRC]		= (u8) sizeof(struct sadb_address),
	[SADB_EXT_ADDRESS_DST]		= (u8) sizeof(struct sadb_address),
	[SADB_EXT_ADDRESS_PROXY]	= (u8) sizeof(struct sadb_address),
	[SADB_EXT_KEY_AUTH]		= (u8) sizeof(struct sadb_key),
	[SADB_EXT_KEY_ENCRYPT]		= (u8) sizeof(struct sadb_key),
	[SADB_EXT_IDENTITY_SRC]		= (u8) sizeof(struct sadb_ident),
	[SADB_EXT_IDENTITY_DST]		= (u8) sizeof(struct sadb_ident),
	[SADB_EXT_SENSITIVITY]		= (u8) sizeof(struct sadb_sens),
	[SADB_EXT_PROPOSAL]		= (u8) sizeof(struct sadb_prop),
	[SADB_EXT_SUPPORTED_AUTH]	= (u8) sizeof(struct sadb_supported),
	[SADB_EXT_SUPPORTED_ENCRYPT]	= (u8) sizeof(struct sadb_supported),
	[SADB_EXT_SPIRANGE]		= (u8) sizeof(struct sadb_spirange),
	[SADB_X_EXT_KMPRIVATE]		= (u8) sizeof(struct sadb_x_kmprivate),
	[SADB_X_EXT_POLICY]		= (u8) sizeof(struct sadb_x_policy),
	[SADB_X_EXT_SA2]		= (u8) sizeof(struct sadb_x_sa2),
	[SADB_X_EXT_NAT_T_TYPE]		= (u8) sizeof(struct sadb_x_nat_t_type),
	[SADB_X_EXT_NAT_T_SPORT]	= (u8) sizeof(struct sadb_x_nat_t_port),
	[SADB_X_EXT_NAT_T_DPORT]	= (u8) sizeof(struct sadb_x_nat_t_port),
	[SADB_X_EXT_NAT_T_OA]		= (u8) sizeof(struct sadb_address),
	[SADB_X_EXT_SEC_CTX]		= (u8) sizeof(struct sadb_x_sec_ctx),
};

/* Verify sadb_address_{len,prefixlen} against sa_family.  */
static int verify_address_len(void *p)
{
	struct sadb_address *sp = p;
	struct sockaddr *addr = (struct sockaddr *)(sp + 1);
	struct sockaddr_in *sin;
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
	struct sockaddr_in6 *sin6;
#endif
	int len;

	switch (addr->sa_family) {
	case AF_INET:
		len = DIV_ROUND_UP(sizeof(*sp) + sizeof(*sin), sizeof(uint64_t));
		if (sp->sadb_address_len != len ||
		    sp->sadb_address_prefixlen > 32)
			return -EINVAL;
		break;
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
	case AF_INET6:
		len = DIV_ROUND_UP(sizeof(*sp) + sizeof(*sin6), sizeof(uint64_t));
		if (sp->sadb_address_len != len ||
		    sp->sadb_address_prefixlen > 128)
			return -EINVAL;
		break;
#endif
	default:
		/* It is user using kernel to keep track of security
		 * associations for another protocol, such as
		 * OSPF/RSVP/RIPV2/MIP.  It is user's job to verify
		 * lengths.
		 *
		 * XXX Actually, association/policy database is not yet
		 * XXX able to cope with arbitrary sockaddr families.
		 * XXX When it can, remove this -EINVAL.  -DaveM
		 */
		return -EINVAL;
		break;
	}

	return 0;
}

static inline int pfkey_sec_ctx_len(struct sadb_x_sec_ctx *sec_ctx)
{
	return DIV_ROUND_UP(sizeof(struct sadb_x_sec_ctx) +
			    sec_ctx->sadb_x_ctx_len,
			    sizeof(uint64_t));
}

static inline int verify_sec_ctx_len(void *p)
{
	struct sadb_x_sec_ctx *sec_ctx = (struct sadb_x_sec_ctx *)p;
	int len = sec_ctx->sadb_x_ctx_len;

	if (len > PAGE_SIZE)
		return -EINVAL;

	len = pfkey_sec_ctx_len(sec_ctx);

	if (sec_ctx->sadb_x_sec_len != len)
		return -EINVAL;

	return 0;
}

static inline struct xfrm_user_sec_ctx *pfkey_sadb2xfrm_user_sec_ctx(struct sadb_x_sec_ctx *sec_ctx)
{
	struct xfrm_user_sec_ctx *uctx = NULL;
	int ctx_size = sec_ctx->sadb_x_ctx_len;

	uctx = kmalloc((sizeof(*uctx)+ctx_size), GFP_KERNEL);

	if (!uctx)
		return NULL;

	uctx->len = pfkey_sec_ctx_len(sec_ctx);
	uctx->exttype = sec_ctx->sadb_x_sec_exttype;
	uctx->ctx_doi = sec_ctx->sadb_x_ctx_doi;
	uctx->ctx_alg = sec_ctx->sadb_x_ctx_alg;
	uctx->ctx_len = sec_ctx->sadb_x_ctx_len;
	memcpy(uctx + 1, sec_ctx + 1,
	       uctx->ctx_len);

	return uctx;
}

static int present_and_same_family(struct sadb_address *src,
				   struct sadb_address *dst)
{
	struct sockaddr *s_addr, *d_addr;

	if (!src || !dst)
		return 0;

	s_addr = (struct sockaddr *)(src + 1);
	d_addr = (struct sockaddr *)(dst + 1);
	if (s_addr->sa_family != d_addr->sa_family)
		return 0;
	if (s_addr->sa_family != AF_INET
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
	    && s_addr->sa_family != AF_INET6
#endif
		)
		return 0;

	return 1;
}

static int parse_exthdrs(struct sk_buff *skb, struct sadb_msg *hdr, void **ext_hdrs)
{
	char *p = (char *) hdr;
	int len = skb->len;

	len -= sizeof(*hdr);
	p += sizeof(*hdr);
	while (len > 0) {
		struct sadb_ext *ehdr = (struct sadb_ext *) p;
		uint16_t ext_type;
		int ext_len;

		ext_len  = ehdr->sadb_ext_len;
		ext_len *= sizeof(uint64_t);
		ext_type = ehdr->sadb_ext_type;
		if (ext_len < sizeof(uint64_t) ||
		    ext_len > len ||
		    ext_type == SADB_EXT_RESERVED)
			return -EINVAL;

		if (ext_type <= SADB_EXT_MAX) {
			int min = (int) sadb_ext_min_len[ext_type];
			if (ext_len < min)
				return -EINVAL;
			if (ext_hdrs[ext_type-1] != NULL)
				return -EINVAL;