dn_route.c

Linux Kernel 2.6.9 for OMAP1710

				     { .daddr = cb->dst,
				       .saddr = cb->src,
				       .scope = RT_SCOPE_UNIVERSE,
#ifdef CONFIG_DECNET_ROUTE_FWMARK
				       .fwmark = skb->nfmark
#endif
				    } },
			    .iif = skb->dev->ifindex };
	struct dn_fib_res res = { .fi = NULL, .type = RTN_UNREACHABLE };
	int err = -EINVAL;
	int free_res = 0;

	dev_hold(in_dev);

	if ((dn_db = in_dev->dn_ptr) == NULL)
		goto out;

	/* Zero source addresses are not allowed */
	if (fl.fld_src == 0)
		goto out;

	/*
	 * In this case we've just received a packet from a source
	 * outside ourselves pretending to come from us. We don't
	 * allow it any further to prevent routing loops, spoofing and
	 * other nasties. Loopback packets already have the dst attached
	 * so this only affects packets which have originated elsewhere.
	 */
	err  = -ENOTUNIQ;
	if (dn_dev_islocal(in_dev, cb->src))
		goto out;

	err = dn_fib_lookup(&fl, &res);
	if (err) {
		if (err != -ESRCH)
			goto out;
		/*
		 * Is the destination us ?
		 */
		if (!dn_dev_islocal(in_dev, cb->dst))
			goto e_inval;

		res.type = RTN_LOCAL;
		flags |= RTCF_DIRECTSRC;
	} else {
		__u16 src_map = fl.fld_src;
		free_res = 1;

		out_dev = DN_FIB_RES_DEV(res);
		if (out_dev == NULL) {
			if (net_ratelimit())
				printk(KERN_CRIT "Bug in dn_route_input_slow() "
						 "No output device\n");
			goto e_inval;
		}
		dev_hold(out_dev);

		if (res.r)
			src_map = dn_fib_rules_policy(fl.fld_src, &res, &flags);

		gateway = DN_FIB_RES_GW(res);
		if (res.type == RTN_NAT) {
			fl.fld_dst = dn_fib_rules_map_destination(fl.fld_dst, &res);
			dn_fib_res_put(&res);
			free_res = 0;
			if (dn_fib_lookup(&fl, &res))
				goto e_inval;
			free_res = 1;
			if (res.type != RTN_UNICAST)
				goto e_inval;
			flags |= RTCF_DNAT;
			gateway = fl.fld_dst;
		}
		fl.fld_src = src_map;
	}

	switch(res.type) {
	case RTN_UNICAST:
		/*
		 * Forwarding check here, we only check for forwarding
		 * being turned off, if you want to only forward intra
		 * area, its up to you to set the routing tables up
		 * correctly.
		 */
		if (dn_db->parms.forwarding == 0)
			goto e_inval;

		if (res.fi->fib_nhs > 1 && fl.oif == 0)
			dn_fib_select_multipath(&fl, &res);

		/* 
		 * Check for out_dev == in_dev. We use the RTCF_DOREDIRECT
		 * flag as a hint to set the intra-ethernet bit when
		 * forwarding. If we've got NAT in operation, we don't do
		 * this optimisation.
		 */
		if (out_dev == in_dev && !(flags & RTCF_NAT))
			flags |= RTCF_DOREDIRECT;

		local_src = DN_FIB_RES_PREFSRC(res);

	case RTN_BLACKHOLE:
	case RTN_UNREACHABLE:
		break;
	case RTN_LOCAL:
		flags |= RTCF_LOCAL;
		fl.fld_src = cb->dst;
		fl.fld_dst = cb->src;

		/* Routing tables gave us a gateway */
		if (gateway)
			goto make_route;

		/* Packet was intra-ethernet, so we know its on-link */
		if (cb->rt_flags | DN_RT_F_IE) {
			gateway = cb->src;
			flags |= RTCF_DIRECTSRC;
			goto make_route;
		}

		/* Use the default router if there is one */
		neigh = neigh_clone(dn_db->router);
		if (neigh) {
			gateway = ((struct dn_neigh *)neigh)->addr;
			goto make_route;
		}

		/* Close eyes and pray */
		gateway = cb->src;
		flags |= RTCF_DIRECTSRC;
		goto make_route;
	default:
		goto e_inval;
	}

make_route:
	rt = dst_alloc(&dn_dst_ops);
	if (rt == NULL)
		goto e_nobufs;

	rt->rt_saddr      = fl.fld_src;
	rt->rt_daddr      = fl.fld_dst;
	rt->rt_gateway    = fl.fld_dst;
	if (gateway)
		rt->rt_gateway = gateway;
	rt->rt_local_src  = local_src ? local_src : rt->rt_saddr;

	rt->rt_dst_map    = fl.fld_dst;
	rt->rt_src_map    = fl.fld_src;

	rt->fl.fld_src    = cb->src;
	rt->fl.fld_dst    = cb->dst;
	rt->fl.oif        = 0;
	rt->fl.iif        = in_dev->ifindex;
	rt->fl.fld_fwmark = fl.fld_fwmark;

	rt->u.dst.flags = DST_HOST;
	rt->u.dst.neighbour = neigh;
	rt->u.dst.dev = out_dev;
	rt->u.dst.lastuse = jiffies;
	rt->u.dst.output = dn_rt_bug_out;
	switch(res.type) {
		case RTN_UNICAST:
			rt->u.dst.input = dn_forward;
			break;
		case RTN_LOCAL:
			rt->u.dst.output = dn_output;
			rt->u.dst.input = dn_nsp_rx;
			rt->u.dst.dev = in_dev;
			flags |= RTCF_LOCAL;
			break;
		default:
		case RTN_UNREACHABLE:
		case RTN_BLACKHOLE:
			rt->u.dst.input = dn_blackhole;
	}
	rt->rt_flags = flags;
	if (rt->u.dst.dev)
		dev_hold(rt->u.dst.dev);

	err = dn_rt_set_next_hop(rt, &res);
	if (err)
		goto e_neighbour;

	hash = dn_hash(rt->fl.fld_src, rt->fl.fld_dst);
	dn_insert_route(rt, hash, (struct dn_route **)&skb->dst);

done:
	if (neigh)
		neigh_release(neigh);
	if (free_res)
		dn_fib_res_put(&res);
	dev_put(in_dev);
	if (out_dev)
		dev_put(out_dev);
out:
	return err;

e_inval:
	err = -EINVAL;
	goto done;

e_nobufs:
	err = -ENOBUFS;
	goto done;

e_neighbour:
	dst_free(&rt->u.dst);
	goto done;
}

int dn_route_input(struct sk_buff *skb)
{
	struct dn_route *rt;
	struct dn_skb_cb *cb = DN_SKB_CB(skb);
	unsigned hash = dn_hash(cb->src, cb->dst);

	if (skb->dst)
		return 0;

	rcu_read_lock();
	for(rt = rcu_dereference(dn_rt_hash_table[hash].chain); rt != NULL;
	    rt = rcu_dereference(rt->u.rt_next)) {
		if ((rt->fl.fld_src == cb->src) &&
	 	    (rt->fl.fld_dst == cb->dst) &&
		    (rt->fl.oif == 0) &&
#ifdef CONFIG_DECNET_ROUTE_FWMARK
		    (rt->fl.fld_fwmark == skb->nfmark) &&
#endif
		    (rt->fl.iif == cb->iif)) {
			rt->u.dst.lastuse = jiffies;
			dst_hold(&rt->u.dst);
			rt->u.dst.__use++;
			rcu_read_unlock();
			skb->dst = (struct dst_entry *)rt;
			return 0;
		}
	}
	rcu_read_unlock();

	return dn_route_input_slow(skb);
}

static int dn_rt_fill_info(struct sk_buff *skb, u32 pid, u32 seq, int event, int nowait)
{
	struct dn_route *rt = (struct dn_route *)skb->dst;
	struct rtmsg *r;
	struct nlmsghdr *nlh;
	unsigned char *b = skb->tail;
	struct rta_cacheinfo ci;

	nlh = NLMSG_PUT(skb, pid, seq, event, sizeof(*r));
	r = NLMSG_DATA(nlh);
	nlh->nlmsg_flags = (nowait && pid) ? NLM_F_MULTI : 0;
	r->rtm_family = AF_DECnet;
	r->rtm_dst_len = 16;
	r->rtm_src_len = 0;
	r->rtm_tos = 0;
	r->rtm_table = RT_TABLE_MAIN;
	r->rtm_type = rt->rt_type;
	r->rtm_flags = (rt->rt_flags & ~0xFFFF) | RTM_F_CLONED;
	r->rtm_scope = RT_SCOPE_UNIVERSE;
	r->rtm_protocol = RTPROT_UNSPEC;
	if (rt->rt_flags & RTCF_NOTIFY)
		r->rtm_flags |= RTM_F_NOTIFY;
	RTA_PUT(skb, RTA_DST, 2, &rt->rt_daddr);
	if (rt->fl.fld_src) {
		r->rtm_src_len = 16;
		RTA_PUT(skb, RTA_SRC, 2, &rt->fl.fld_src);
	}
	if (rt->u.dst.dev)
		RTA_PUT(skb, RTA_OIF, sizeof(int), &rt->u.dst.dev->ifindex);
	/*
	 * Note to self - change this if input routes reverse direction when
	 * they deal only with inputs and not with replies like they do
	 * currently.
	 */
	RTA_PUT(skb, RTA_PREFSRC, 2, &rt->rt_local_src);
	if (rt->rt_daddr != rt->rt_gateway)
		RTA_PUT(skb, RTA_GATEWAY, 2, &rt->rt_gateway);
	if (rtnetlink_put_metrics(skb, rt->u.dst.metrics) < 0)
		goto rtattr_failure;
	ci.rta_lastuse = jiffies_to_clock_t(jiffies - rt->u.dst.lastuse);
	ci.rta_used     = rt->u.dst.__use;
	ci.rta_clntref  = atomic_read(&rt->u.dst.__refcnt);
	if (rt->u.dst.expires)
		ci.rta_expires = jiffies_to_clock_t(rt->u.dst.expires - jiffies);
	else
		ci.rta_expires = 0;
	ci.rta_error    = rt->u.dst.error;
	ci.rta_id       = ci.rta_ts = ci.rta_tsage = 0;
	RTA_PUT(skb, RTA_CACHEINFO, sizeof(ci), &ci);
	if (rt->fl.iif)
		RTA_PUT(skb, RTA_IIF, sizeof(int), &rt->fl.iif);

	nlh->nlmsg_len = skb->tail - b;
	return skb->len;

nlmsg_failure:
rtattr_failure:
        skb_trim(skb, b - skb->data);
        return -1;
}

/*
 * This is called by both endnodes and routers now.
 */
int dn_cache_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh, void *arg)
{
	struct rtattr **rta = arg;
	struct rtmsg *rtm = NLMSG_DATA(nlh);
	struct dn_route *rt = NULL;
	struct dn_skb_cb *cb;
	int err;
	struct sk_buff *skb;
	struct flowi fl;

	memset(&fl, 0, sizeof(fl));
	fl.proto = DNPROTO_NSP;

	skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
	if (skb == NULL)
		return -ENOBUFS;
	skb->mac.raw = skb->data;
	cb = DN_SKB_CB(skb);

	if (rta[RTA_SRC-1])
		memcpy(&fl.fld_src, RTA_DATA(rta[RTA_SRC-1]), 2);
	if (rta[RTA_DST-1])
		memcpy(&fl.fld_dst, RTA_DATA(rta[RTA_DST-1]), 2);
	if (rta[RTA_IIF-1])
		memcpy(&fl.iif, RTA_DATA(rta[RTA_IIF-1]), sizeof(int));

	if (fl.iif) {
		struct net_device *dev;
		if ((dev = dev_get_by_index(fl.iif)) == NULL) {
			kfree_skb(skb);
			return -ENODEV;
		}
		if (!dev->dn_ptr) {
			dev_put(dev);
			kfree_skb(skb);
			return -ENODEV;
		}
		skb->protocol = __constant_htons(ETH_P_DNA_RT);
		skb->dev = dev;
		cb->src = fl.fld_src;
		cb->dst = fl.fld_dst;
		local_bh_disable();
		err = dn_route_input(skb);
		local_bh_enable();
		memset(cb, 0, sizeof(struct dn_skb_cb));
		rt = (struct dn_route *)skb->dst;
		if (!err && -rt->u.dst.error)
			err = rt->u.dst.error;
	} else {
		int oif = 0;
		if (rta[RTA_OIF - 1])
			memcpy(&oif, RTA_DATA(rta[RTA_OIF - 1]), sizeof(int));
		fl.oif = oif;
		err = dn_route_output_key((struct dst_entry **)&rt, &fl, 0);
	}

	if (skb->dev)
		dev_put(skb->dev);
	skb->dev = NULL;
	if (err)
		goto out_free;
	skb->dst = &rt->u.dst;
	if (rtm->rtm_flags & RTM_F_NOTIFY)
		rt->rt_flags |= RTCF_NOTIFY;

	NETLINK_CB(skb).dst_pid = NETLINK_CB(in_skb).pid;

	err = dn_rt_fill_info(skb, NETLINK_CB(in_skb).pid, nlh->nlmsg_seq, RTM_NEWROUTE, 0);

	if (err == 0)
		goto out_free;
	if (err < 0) {
		err = -EMSGSIZE;
		goto out_free;
	}

	err = netlink_unicast(rtnl, skb, NETLINK_CB(in_skb).pid, MSG_DONTWAIT);

	return err;

out_free:
	kfree_skb(skb);
	return err;
}

/*
 * For routers, this is called from dn_fib_dump, but for endnodes its
 * called directly from the rtnetlink dispatch table.
 */
int dn_cache_dump(struct sk_buff *skb, struct netlink_callback *cb)
{
	struct dn_route *rt;
	int h, s_h;
	int idx, s_idx;

	if (NLMSG_PAYLOAD(cb->nlh, 0) < sizeof(struct rtmsg))
		return -EINVAL;
	if (!(((struct rtmsg *)NLMSG_DATA(cb->nlh))->rtm_flags&RTM_F_CLONED))
		return 0;

	s_h = cb->args[0];
	s_idx = idx = cb->args[1];
	for(h = 0; h <= dn_rt_hash_mask; h++) {
		if (h < s_h)
			continue;
		if (h > s_h)
			s_idx = 0;
		rcu_read_lock_bh();
		for(rt = rcu_dereference(dn_rt_hash_table[h].chain), idx = 0;
			rt;
			rt = rcu_dereference(rt->u.rt_next), idx++) {
			if (idx < s_idx)
				continue;
			skb->dst = dst_clone(&rt->u.dst);
			if (dn_rt_fill_info(skb, NETLINK_CB(cb->skb).pid,
					cb->nlh->nlmsg_seq, RTM_NEWROUTE, 1) <= 0) {
				dst_release(xchg(&skb->dst, NULL));
				rcu_read_unlock_bh();
				goto done;
			}
			dst_release(xchg(&skb->dst, NULL));
		}
		rcu_read_unlock_bh();
	}

done:
	cb->args[0] = h;
	cb->args[1] = idx;
	return skb->len;
}

#ifdef CONFIG_PROC_FS
struct dn_rt_cache_iter_state {
	int bucket;
};

static struct dn_route *dn_rt_cache_get_first(struct seq_file *seq)
{
	struct dn_route *rt = NULL;
	struct dn_rt_cache_iter_state *s = seq->private;

	for(s->bucket = dn_rt_hash_mask; s->bucket >= 0; --s->bucket) {
		rcu_read_lock_bh();
		rt = dn_rt_hash_table[s->bucket].chain;
		if (rt)
			break;
		rcu_read_unlock();
	}
	return rt;
}

static struct dn_route *dn_rt_cache_get_next(struct seq_file *seq, struct dn_route *rt)
{
	struct dn_rt_cache_iter_state *s = rcu_dereference(seq->private);

	rt = rt->u.rt_next;
	while(!rt) {
		rcu_read_unlock_bh();
		if (--s->bucket < 0)
			break;
		rcu_read_lock_bh();
		rt = dn_rt_hash_table[s->bucket].chain;
	}
	return rt;
}

static void *dn_rt_cache_seq_start(struct seq_file *seq, loff_t *pos)
{
	struct dn_route *rt = dn_rt_cache_get_first(seq);

	if (rt) {
		while(*pos && (rt = dn_rt_cache_get_next(seq, rt)))
			--*pos;
	}
	return *pos ? NULL : rt;
}

static void *dn_rt_cache_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
	struct dn_route *rt = dn_rt_cache_get_next(seq, v);
	++*pos;
	return rt;
}

static void dn_rt_cache_seq_stop(struct seq_file *seq, void *v)
{
	if (v)
		rcu_read_unlock_bh();
}

static int dn_rt_cache_seq_show(struct seq_file *seq, void *v)
{
	struct dn_route *rt = v;
	char buf1[DN_ASCBUF_LEN], buf2[DN_ASCBUF_LEN];

	seq_printf(seq, "%-8s %-7s %-7s %04d %04d %04d\n",
			rt->u.dst.dev ? rt->u.dst.dev->name : "*",
			dn_addr2asc(dn_ntohs(rt->rt_daddr), buf1),
			dn_addr2asc(dn_ntohs(rt->rt_saddr), buf2),
			atomic_read(&rt->u.dst.__refcnt),
			rt->u.dst.__use,
			(int) dst_metric(&rt->u.dst, RTAX_RTT));
	return 0;
} 

static struct seq_operations dn_rt_cache_seq_ops = {
	.start	= dn_rt_cache_seq_start,
	.next	= dn_rt_cache_seq_next,
	.stop	= dn_rt_cache_seq_stop,
	.show	= dn_rt_cache_seq_show,
};

static int dn_rt_cache_seq_open(struct inode *inode, struct file *file)
{
	struct seq_file *seq;
	int rc = -ENOMEM;
	struct dn_rt_cache_iter_state *s = kmalloc(sizeof(*s), GFP_KERNEL);

	if (!s)
		goto out;
	rc = seq_open(file, &dn_rt_cache_seq_ops);
	if (rc)
		goto out_kfree;
	seq		= file->private_data;
	seq->private	= s;
	memset(s, 0, sizeof(*s));
out:
	return rc;
out_kfree:
	kfree(s);
	goto out;
}

static struct file_operations dn_rt_cache_seq_fops = {
	.owner	 = THIS_MODULE,
	.open	 = dn_rt_cache_seq_open,
	.read	 = seq_read,
	.llseek	 = seq_lseek,
	.release = seq_release_private,
};

#endif /* CONFIG_PROC_FS */

void __init dn_route_init(void)
{
	int i, goal, order;

	dn_dst_ops.kmem_cachep = kmem_cache_create("dn_dst_cache",
						   sizeof(struct dn_route),
						   0, SLAB_HWCACHE_ALIGN,
						   NULL, NULL);

	if (!dn_dst_ops.kmem_cachep)
		panic("DECnet: Failed to allocate dn_dst_cache\n");

	init_timer(&dn_route_timer);
	dn_route_timer.function = dn_dst_check_expire;
	dn_route_timer.expires = jiffies + decnet_dst_gc_interval * HZ;
	add_timer(&dn_route_timer);

	goal = num_physpages >> (26 - PAGE_SHIFT);

	for(order = 0; (1UL << order) < goal; order++)
		/* NOTHING */;

        /*
         * Only want 1024 entries max, since the table is very, very unlikely
         * to be larger than that.
         */
        while(order && ((((1UL << order) * PAGE_SIZE) / 
                                sizeof(struct dn_rt_hash_bucket)) >= 2048))
                order--;

        do {
                dn_rt_hash_mask = (1UL << order) * PAGE_SIZE /
                        sizeof(struct dn_rt_hash_bucket);
                while(dn_rt_hash_mask & (dn_rt_hash_mask - 1))
                        dn_rt_hash_mask--;
                dn_rt_hash_table = (struct dn_rt_hash_bucket *)
                        __get_free_pages(GFP_ATOMIC, order);
        } while (dn_rt_hash_table == NULL && --order > 0);

	if (!dn_rt_hash_table)
                panic("Failed to allocate DECnet route cache hash table\n");

	printk(KERN_INFO 
		"DECnet: Routing cache hash table of %u buckets, %ldKbytes\n", 
		dn_rt_hash_mask, 
		(long)(dn_rt_hash_mask*sizeof(struct dn_rt_hash_bucket))/1024);

	dn_rt_hash_mask--;
        for(i = 0; i <= dn_rt_hash_mask; i++) {
                dn_rt_hash_table[i].lock = SPIN_LOCK_UNLOCKED;
                dn_rt_hash_table[i].chain = NULL;
        }

        dn_dst_ops.gc_thresh = (dn_rt_hash_mask + 1);

	proc_net_fops_create("decnet_cache", S_IRUGO, &dn_rt_cache_seq_fops);
}

void __exit dn_route_cleanup(void)
{
	del_timer(&dn_route_timer);
	dn_run_flush(0);

	proc_net_remove("decnet_cache");
}