ip6_flowlabel.c

linux 内核源代码

/*
 *	ip6_flowlabel.c		IPv6 flowlabel manager.
 *
 *	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, <[e-mail]>
 */

#include <linux/capability.h>
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/net.h>
#include <linux/netdevice.h>
#include <linux/if_arp.h>
#include <linux/in6.h>
#include <linux/route.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>

#include <net/net_namespace.h>
#include <net/sock.h>

#include <net/ipv6.h>
#include <net/ndisc.h>
#include <net/protocol.h>
#include <net/ip6_route.h>
#include <net/addrconf.h>
#include <net/rawv6.h>
#include <net/icmp.h>
#include <net/transp_v6.h>

#include <asm/uaccess.h>

#define FL_MIN_LINGER	6	/* Minimal linger. It is set to 6sec specified
				   in old IPv6 RFC. Well, it was reasonable value.
				 */
#define FL_MAX_LINGER	60	/* Maximal linger timeout */

/* FL hash table */

#define FL_MAX_PER_SOCK	32
#define FL_MAX_SIZE	4096
#define FL_HASH_MASK	255
#define FL_HASH(l)	(ntohl(l)&FL_HASH_MASK)

static atomic_t fl_size = ATOMIC_INIT(0);
static struct ip6_flowlabel *fl_ht[FL_HASH_MASK+1];

static void ip6_fl_gc(unsigned long dummy);
static DEFINE_TIMER(ip6_fl_gc_timer, ip6_fl_gc, 0, 0);

/* FL hash table lock: it protects only of GC */

static DEFINE_RWLOCK(ip6_fl_lock);

/* Big socket sock */

static DEFINE_RWLOCK(ip6_sk_fl_lock);


static __inline__ struct ip6_flowlabel * __fl_lookup(__be32 label)
{
	struct ip6_flowlabel *fl;

	for (fl=fl_ht[FL_HASH(label)]; fl; fl = fl->next) {
		if (fl->label == label)
			return fl;
	}
	return NULL;
}

static struct ip6_flowlabel * fl_lookup(__be32 label)
{
	struct ip6_flowlabel *fl;

	read_lock_bh(&ip6_fl_lock);
	fl = __fl_lookup(label);
	if (fl)
		atomic_inc(&fl->users);
	read_unlock_bh(&ip6_fl_lock);
	return fl;
}


static void fl_free(struct ip6_flowlabel *fl)
{
	if (fl)
		kfree(fl->opt);
	kfree(fl);
}

static void fl_release(struct ip6_flowlabel *fl)
{
	write_lock_bh(&ip6_fl_lock);

	fl->lastuse = jiffies;
	if (atomic_dec_and_test(&fl->users)) {
		unsigned long ttd = fl->lastuse + fl->linger;
		if (time_after(ttd, fl->expires))
			fl->expires = ttd;
		ttd = fl->expires;
		if (fl->opt && fl->share == IPV6_FL_S_EXCL) {
			struct ipv6_txoptions *opt = fl->opt;
			fl->opt = NULL;
			kfree(opt);
		}
		if (!timer_pending(&ip6_fl_gc_timer) ||
		    time_after(ip6_fl_gc_timer.expires, ttd))
			mod_timer(&ip6_fl_gc_timer, ttd);
	}

	write_unlock_bh(&ip6_fl_lock);
}

static void ip6_fl_gc(unsigned long dummy)
{
	int i;
	unsigned long now = jiffies;
	unsigned long sched = 0;

	write_lock(&ip6_fl_lock);

	for (i=0; i<=FL_HASH_MASK; i++) {
		struct ip6_flowlabel *fl, **flp;
		flp = &fl_ht[i];
		while ((fl=*flp) != NULL) {
			if (atomic_read(&fl->users) == 0) {
				unsigned long ttd = fl->lastuse + fl->linger;
				if (time_after(ttd, fl->expires))
					fl->expires = ttd;
				ttd = fl->expires;
				if (time_after_eq(now, ttd)) {
					*flp = fl->next;
					fl_free(fl);
					atomic_dec(&fl_size);
					continue;
				}
				if (!sched || time_before(ttd, sched))
					sched = ttd;
			}
			flp = &fl->next;
		}
	}
	if (!sched && atomic_read(&fl_size))
		sched = now + FL_MAX_LINGER;
	if (sched) {
		ip6_fl_gc_timer.expires = sched;
		add_timer(&ip6_fl_gc_timer);
	}
	write_unlock(&ip6_fl_lock);
}

static struct ip6_flowlabel *fl_intern(struct ip6_flowlabel *fl, __be32 label)
{
	struct ip6_flowlabel *lfl;

	fl->label = label & IPV6_FLOWLABEL_MASK;

	write_lock_bh(&ip6_fl_lock);
	if (label == 0) {
		for (;;) {
			fl->label = htonl(net_random())&IPV6_FLOWLABEL_MASK;
			if (fl->label) {
				lfl = __fl_lookup(fl->label);
				if (lfl == NULL)
					break;
			}
		}
	} else {
		/*
		 * we dropper the ip6_fl_lock, so this entry could reappear
		 * and we need to recheck with it.
		 *
		 * OTOH no need to search the active socket first, like it is
		 * done in ipv6_flowlabel_opt - sock is locked, so new entry
		 * with the same label can only appear on another sock
		 */
		lfl = __fl_lookup(fl->label);
		if (lfl != NULL) {
			atomic_inc(&lfl->users);
			write_unlock_bh(&ip6_fl_lock);
			return lfl;
		}
	}

	fl->lastuse = jiffies;
	fl->next = fl_ht[FL_HASH(fl->label)];
	fl_ht[FL_HASH(fl->label)] = fl;
	atomic_inc(&fl_size);
	write_unlock_bh(&ip6_fl_lock);
	return NULL;
}



/* Socket flowlabel lists */

struct ip6_flowlabel * fl6_sock_lookup(struct sock *sk, __be32 label)
{
	struct ipv6_fl_socklist *sfl;
	struct ipv6_pinfo *np = inet6_sk(sk);

	label &= IPV6_FLOWLABEL_MASK;

	read_lock_bh(&ip6_sk_fl_lock);
	for (sfl=np->ipv6_fl_list; sfl; sfl = sfl->next) {
		struct ip6_flowlabel *fl = sfl->fl;
		if (fl->label == label) {
			fl->lastuse = jiffies;
			atomic_inc(&fl->users);
			read_unlock_bh(&ip6_sk_fl_lock);
			return fl;
		}
	}
	read_unlock_bh(&ip6_sk_fl_lock);
	return NULL;
}

EXPORT_SYMBOL_GPL(fl6_sock_lookup);

void fl6_free_socklist(struct sock *sk)
{
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct ipv6_fl_socklist *sfl;

	while ((sfl = np->ipv6_fl_list) != NULL) {
		np->ipv6_fl_list = sfl->next;
		fl_release(sfl->fl);
		kfree(sfl);
	}
}

/* Service routines */


/*
   It is the only difficult place. flowlabel enforces equal headers
   before and including routing header, however user may supply options
   following rthdr.
 */

struct ipv6_txoptions *fl6_merge_options(struct ipv6_txoptions * opt_space,
					 struct ip6_flowlabel * fl,
					 struct ipv6_txoptions * fopt)
{
	struct ipv6_txoptions * fl_opt = fl->opt;

	if (fopt == NULL || fopt->opt_flen == 0)
		return fl_opt;

	if (fl_opt != NULL) {
		opt_space->hopopt = fl_opt->hopopt;
		opt_space->dst0opt = fl_opt->dst0opt;
		opt_space->srcrt = fl_opt->srcrt;
		opt_space->opt_nflen = fl_opt->opt_nflen;
	} else {
		if (fopt->opt_nflen == 0)
			return fopt;
		opt_space->hopopt = NULL;
		opt_space->dst0opt = NULL;
		opt_space->srcrt = NULL;
		opt_space->opt_nflen = 0;
	}
	opt_space->dst1opt = fopt->dst1opt;
	opt_space->opt_flen = fopt->opt_flen;
	return opt_space;
}

static unsigned long check_linger(unsigned long ttl)
{
	if (ttl < FL_MIN_LINGER)
		return FL_MIN_LINGER*HZ;
	if (ttl > FL_MAX_LINGER && !capable(CAP_NET_ADMIN))
		return 0;
	return ttl*HZ;
}

static int fl6_renew(struct ip6_flowlabel *fl, unsigned long linger, unsigned long expires)
{
	linger = check_linger(linger);
	if (!linger)
		return -EPERM;
	expires = check_linger(expires);
	if (!expires)
		return -EPERM;
	fl->lastuse = jiffies;
	if (time_before(fl->linger, linger))
		fl->linger = linger;
	if (time_before(expires, fl->linger))
		expires = fl->linger;
	if (time_before(fl->expires, fl->lastuse + expires))
		fl->expires = fl->lastuse + expires;
	return 0;
}

static struct ip6_flowlabel *
fl_create(struct in6_flowlabel_req *freq, char __user *optval, int optlen, int *err_p)
{
	struct ip6_flowlabel *fl;
	int olen;
	int addr_type;
	int err;

	err = -ENOMEM;
	fl = kzalloc(sizeof(*fl), GFP_KERNEL);
	if (fl == NULL)
		goto done;

	olen = optlen - CMSG_ALIGN(sizeof(*freq));
	if (olen > 0) {
		struct msghdr msg;
		struct flowi flowi;
		int junk;

		err = -ENOMEM;
		fl->opt = kmalloc(sizeof(*fl->opt) + olen, GFP_KERNEL);
		if (fl->opt == NULL)
			goto done;

		memset(fl->opt, 0, sizeof(*fl->opt));
		fl->opt->tot_len = sizeof(*fl->opt) + olen;
		err = -EFAULT;
		if (copy_from_user(fl->opt+1, optval+CMSG_ALIGN(sizeof(*freq)), olen))
			goto done;

		msg.msg_controllen = olen;
		msg.msg_control = (void*)(fl->opt+1);
		flowi.oif = 0;

		err = datagram_send_ctl(&msg, &flowi, fl->opt, &junk, &junk);
		if (err)
			goto done;
		err = -EINVAL;
		if (fl->opt->opt_flen)
			goto done;
		if (fl->opt->opt_nflen == 0) {
			kfree(fl->opt);
			fl->opt = NULL;
		}
	}

	fl->expires = jiffies;
	err = fl6_renew(fl, freq->flr_linger, freq->flr_expires);
	if (err)
		goto done;
	fl->share = freq->flr_share;
	addr_type = ipv6_addr_type(&freq->flr_dst);
	if ((addr_type&IPV6_ADDR_MAPPED)
	    || addr_type == IPV6_ADDR_ANY) {
		err = -EINVAL;
		goto done;
	}