[linux-block.git] / net / core / dst.c

// SPDX-License-Identifier: GPL-2.0-only
/*
 * net/core/dst.c	Protocol independent destination cache.
 *
 * Authors:		Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
 *
 */

#include <linux/bitops.h>
#include <linux/errno.h>
#include <linux/init.h>
#include <linux/kernel.h>
#include <linux/workqueue.h>
#include <linux/mm.h>
#include <linux/module.h>
#include <linux/slab.h>
#include <linux/netdevice.h>
#include <linux/skbuff.h>
#include <linux/string.h>
#include <linux/types.h>
#include <net/net_namespace.h>
#include <linux/sched.h>
#include <linux/prefetch.h>
#include <net/lwtunnel.h>
#include <net/xfrm.h>

#include <net/dst.h>
#include <net/dst_metadata.h>

int dst_discard_out(struct net *net, struct sock *sk, struct sk_buff *skb)
{
	kfree_skb(skb);
	return 0;
}
EXPORT_SYMBOL(dst_discard_out);

const struct dst_metrics dst_default_metrics = {
	/* This initializer is needed to force linker to place this variable
	 * into const section. Otherwise it might end into bss section.
	 * We really want to avoid false sharing on this variable, and catch
	 * any writes on it.
	 */
	.refcnt = REFCOUNT_INIT(1),
};
EXPORT_SYMBOL(dst_default_metrics);

void dst_init(struct dst_entry *dst, struct dst_ops *ops,
	      struct net_device *dev, int initial_ref, int initial_obsolete,
	      unsigned short flags)
{
	dst->dev = dev;
	netdev_hold(dev, &dst->dev_tracker, GFP_ATOMIC);
	dst->ops = ops;
	dst_init_metrics(dst, dst_default_metrics.metrics, true);
	dst->expires = 0UL;
#ifdef CONFIG_XFRM
	dst->xfrm = NULL;
#endif
	dst->input = dst_discard;
	dst->output = dst_discard_out;
	dst->error = 0;
	dst->obsolete = initial_obsolete;
	dst->header_len = 0;
	dst->trailer_len = 0;
#ifdef CONFIG_IP_ROUTE_CLASSID
	dst->tclassid = 0;
#endif
	dst->lwtstate = NULL;
	atomic_set(&dst->__refcnt, initial_ref);
	dst->__use = 0;
	dst->lastuse = jiffies;
	dst->flags = flags;
	if (!(flags & DST_NOCOUNT))
		dst_entries_add(ops, 1);
}
EXPORT_SYMBOL(dst_init);

void *dst_alloc(struct dst_ops *ops, struct net_device *dev,
		int initial_ref, int initial_obsolete, unsigned short flags)
{
	struct dst_entry *dst;

	if (ops->gc &&
	    !(flags & DST_NOCOUNT) &&
	    dst_entries_get_fast(ops) > ops->gc_thresh) {
		if (ops->gc(ops)) {
			pr_notice_ratelimited("Route cache is full: consider increasing sysctl net.ipv6.route.max_size.\n");
			return NULL;
		}
	}

	dst = kmem_cache_alloc(ops->kmem_cachep, GFP_ATOMIC);
	if (!dst)
		return NULL;

	dst_init(dst, ops, dev, initial_ref, initial_obsolete, flags);

	return dst;
}
EXPORT_SYMBOL(dst_alloc);

struct dst_entry *dst_destroy(struct dst_entry * dst)
{
	struct dst_entry *child = NULL;

	smp_rmb();

#ifdef CONFIG_XFRM
	if (dst->xfrm) {
		struct xfrm_dst *xdst = (struct xfrm_dst *) dst;

		child = xdst->child;
	}
#endif
	if (!(dst->flags & DST_NOCOUNT))
		dst_entries_add(dst->ops, -1);

	if (dst->ops->destroy)
		dst->ops->destroy(dst);
	netdev_put(dst->dev, &dst->dev_tracker);

	lwtstate_put(dst->lwtstate);

	if (dst->flags & DST_METADATA)
		metadata_dst_free((struct metadata_dst *)dst);
	else
		kmem_cache_free(dst->ops->kmem_cachep, dst);

	dst = child;
	if (dst)
		dst_release_immediate(dst);
	return NULL;
}
EXPORT_SYMBOL(dst_destroy);

static void dst_destroy_rcu(struct rcu_head *head)
{
	struct dst_entry *dst = container_of(head, struct dst_entry, rcu_head);

	dst = dst_destroy(dst);
}

/* Operations to mark dst as DEAD and clean up the net device referenced
 * by dst:
 * 1. put the dst under blackhole interface and discard all tx/rx packets
 *    on this route.
 * 2. release the net_device
 * This function should be called when removing routes from the fib tree
 * in preparation for a NETDEV_DOWN/NETDEV_UNREGISTER event and also to
 * make the next dst_ops->check() fail.
 */
void dst_dev_put(struct dst_entry *dst)
{
	struct net_device *dev = dst->dev;

	dst->obsolete = DST_OBSOLETE_DEAD;
	if (dst->ops->ifdown)
		dst->ops->ifdown(dst, dev, true);
	dst->input = dst_discard;
	dst->output = dst_discard_out;
	dst->dev = blackhole_netdev;
	netdev_ref_replace(dev, blackhole_netdev, &dst->dev_tracker,
			   GFP_ATOMIC);
}
EXPORT_SYMBOL(dst_dev_put);

void dst_release(struct dst_entry *dst)
{
	if (dst) {
		int newrefcnt;

		newrefcnt = atomic_dec_return(&dst->__refcnt);
		if (WARN_ONCE(newrefcnt < 0, "dst_release underflow"))
			net_warn_ratelimited("%s: dst:%p refcnt:%d\n",
					     __func__, dst, newrefcnt);
		if (!newrefcnt)
			call_rcu(&dst->rcu_head, dst_destroy_rcu);
	}
}
EXPORT_SYMBOL(dst_release);

void dst_release_immediate(struct dst_entry *dst)
{
	if (dst) {
		int newrefcnt;

		newrefcnt = atomic_dec_return(&dst->__refcnt);
		if (WARN_ONCE(newrefcnt < 0, "dst_release_immediate underflow"))
			net_warn_ratelimited("%s: dst:%p refcnt:%d\n",
					     __func__, dst, newrefcnt);
		if (!newrefcnt)
			dst_destroy(dst);
	}
}
EXPORT_SYMBOL(dst_release_immediate);

u32 *dst_cow_metrics_generic(struct dst_entry *dst, unsigned long old)
{
	struct dst_metrics *p = kmalloc(sizeof(*p), GFP_ATOMIC);

	if (p) {
		struct dst_metrics *old_p = (struct dst_metrics *)__DST_METRICS_PTR(old);
		unsigned long prev, new;

		refcount_set(&p->refcnt, 1);
		memcpy(p->metrics, old_p->metrics, sizeof(p->metrics));

		new = (unsigned long) p;
		prev = cmpxchg(&dst->_metrics, old, new);

		if (prev != old) {
			kfree(p);
			p = (struct dst_metrics *)__DST_METRICS_PTR(prev);
			if (prev & DST_METRICS_READ_ONLY)
				p = NULL;
		} else if (prev & DST_METRICS_REFCOUNTED) {
			if (refcount_dec_and_test(&old_p->refcnt))
				kfree(old_p);
		}
	}
	BUILD_BUG_ON(offsetof(struct dst_metrics, metrics) != 0);
	return (u32 *)p;
}
EXPORT_SYMBOL(dst_cow_metrics_generic);

/* Caller asserts that dst_metrics_read_only(dst) is false.  */
void __dst_destroy_metrics_generic(struct dst_entry *dst, unsigned long old)
{
	unsigned long prev, new;

	new = ((unsigned long) &dst_default_metrics) | DST_METRICS_READ_ONLY;
	prev = cmpxchg(&dst->_metrics, old, new);
	if (prev == old)
		kfree(__DST_METRICS_PTR(old));
}
EXPORT_SYMBOL(__dst_destroy_metrics_generic);

struct dst_entry *dst_blackhole_check(struct dst_entry *dst, u32 cookie)
{
	return NULL;
}

u32 *dst_blackhole_cow_metrics(struct dst_entry *dst, unsigned long old)
{
	return NULL;
}

struct neighbour *dst_blackhole_neigh_lookup(const struct dst_entry *dst,
					     struct sk_buff *skb,
					     const void *daddr)
{
	return NULL;
}

void dst_blackhole_update_pmtu(struct dst_entry *dst, struct sock *sk,
			       struct sk_buff *skb, u32 mtu,
			       bool confirm_neigh)
{
}
EXPORT_SYMBOL_GPL(dst_blackhole_update_pmtu);

void dst_blackhole_redirect(struct dst_entry *dst, struct sock *sk,
			    struct sk_buff *skb)
{
}
EXPORT_SYMBOL_GPL(dst_blackhole_redirect);

unsigned int dst_blackhole_mtu(const struct dst_entry *dst)
{
	unsigned int mtu = dst_metric_raw(dst, RTAX_MTU);

	return mtu ? : dst->dev->mtu;
}
EXPORT_SYMBOL_GPL(dst_blackhole_mtu);

static struct dst_ops dst_blackhole_ops = {
	.family		= AF_UNSPEC,
	.neigh_lookup	= dst_blackhole_neigh_lookup,
	.check		= dst_blackhole_check,
	.cow_metrics	= dst_blackhole_cow_metrics,
	.update_pmtu	= dst_blackhole_update_pmtu,
	.redirect	= dst_blackhole_redirect,
	.mtu		= dst_blackhole_mtu,
};

static void __metadata_dst_init(struct metadata_dst *md_dst,
				enum metadata_type type, u8 optslen)
{
	struct dst_entry *dst;

	dst = &md_dst->dst;
	dst_init(dst, &dst_blackhole_ops, NULL, 1, DST_OBSOLETE_NONE,
		 DST_METADATA | DST_NOCOUNT);
	memset(dst + 1, 0, sizeof(*md_dst) + optslen - sizeof(*dst));
	md_dst->type = type;
}

struct metadata_dst *metadata_dst_alloc(u8 optslen, enum metadata_type type,
					gfp_t flags)
{
	struct metadata_dst *md_dst;

	md_dst = kmalloc(sizeof(*md_dst) + optslen, flags);
	if (!md_dst)
		return NULL;

	__metadata_dst_init(md_dst, type, optslen);

	return md_dst;
}
EXPORT_SYMBOL_GPL(metadata_dst_alloc);

void metadata_dst_free(struct metadata_dst *md_dst)
{
#ifdef CONFIG_DST_CACHE
	if (md_dst->type == METADATA_IP_TUNNEL)
		dst_cache_destroy(&md_dst->u.tun_info.dst_cache);
#endif
	kfree(md_dst);
}
EXPORT_SYMBOL_GPL(metadata_dst_free);

struct metadata_dst __percpu *
metadata_dst_alloc_percpu(u8 optslen, enum metadata_type type, gfp_t flags)
{
	int cpu;
	struct metadata_dst __percpu *md_dst;

	md_dst = __alloc_percpu_gfp(sizeof(struct metadata_dst) + optslen,
				    __alignof__(struct metadata_dst), flags);
	if (!md_dst)
		return NULL;

	for_each_possible_cpu(cpu)
		__metadata_dst_init(per_cpu_ptr(md_dst, cpu), type, optslen);

	return md_dst;
}
EXPORT_SYMBOL_GPL(metadata_dst_alloc_percpu);

void metadata_dst_free_percpu(struct metadata_dst __percpu *md_dst)
{
#ifdef CONFIG_DST_CACHE
	int cpu;

	for_each_possible_cpu(cpu) {
		struct metadata_dst *one_md_dst = per_cpu_ptr(md_dst, cpu);

		if (one_md_dst->type == METADATA_IP_TUNNEL)
			dst_cache_destroy(&one_md_dst->u.tun_info.dst_cache);
	}
#endif
	free_percpu(md_dst);
}
EXPORT_SYMBOL_GPL(metadata_dst_free_percpu);
Commit	Line	Data
457c8996	1	// SPDX-License-Identifier: GPL-2.0-only
1da177e4 LT	2	/*
	3	* net/core/dst.c Protocol independent destination cache.
	4	*
	5	* Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
	6	*
	7	*/
	8
	9	#include <linux/bitops.h>
	10	#include <linux/errno.h>
	11	#include <linux/init.h>
	12	#include <linux/kernel.h>
86bba269	13	#include <linux/workqueue.h>
1da177e4 LT	14	#include <linux/mm.h>
1da177e4 LT	15	#include <linux/module.h>
5a0e3ad6	16	#include <linux/slab.h>
1da177e4	17	#include <linux/netdevice.h>
1da177e4 LT	18	#include <linux/skbuff.h>
	19	#include <linux/string.h>
	20	#include <linux/types.h>
e9dc8653	21	#include <net/net_namespace.h>
2fc1b5dd	22	#include <linux/sched.h>
268bb0ce	23	#include <linux/prefetch.h>
61adedf3	24	#include <net/lwtunnel.h>
b6ca8bd5	25	#include <net/xfrm.h>
1da177e4 LT	26
1da177e4 LT	27	#include <net/dst.h>
f38a9eb1	28	#include <net/dst_metadata.h>
1da177e4	29
ede2059d	30	int dst_discard_out(struct net net, struct sock sk, struct sk_buff *skb)
1da177e4 LT	31	{
	32	kfree_skb(skb);
	33	return 0;
	34	}
ede2059d	35	EXPORT_SYMBOL(dst_discard_out);
1da177e4	36
3fb07daf	37	const struct dst_metrics dst_default_metrics = {
a37e6e34 ED	38	/* This initializer is needed to force linker to place this variable
	39	* into const section. Otherwise it might end into bss section.
	40	* We really want to avoid false sharing on this variable, and catch
	41	* any writes on it.
	42	*/
9620fef2	43	.refcnt = REFCOUNT_INIT(1),
a37e6e34	44	};
d4ead6b3	45	EXPORT_SYMBOL(dst_default_metrics);
a37e6e34	46
f38a9eb1 TG	47	void dst_init(struct dst_entry dst, struct dst_ops ops,
	48	struct net_device *dev, int initial_ref, int initial_obsolete,
	49	unsigned short flags)
1da177e4	50	{
5c1e6aa3	51	dst->dev = dev;
d62607c3	52	netdev_hold(dev, &dst->dev_tracker, GFP_ATOMIC);
1da177e4	53	dst->ops = ops;
3fb07daf	54	dst_init_metrics(dst, dst_default_metrics.metrics, true);
cf911662	55	dst->expires = 0UL;
cf911662 DM	56	#ifdef CONFIG_XFRM
	57	dst->xfrm = NULL;
	58	#endif
5c1e6aa3	59	dst->input = dst_discard;
ede2059d	60	dst->output = dst_discard_out;
cf911662	61	dst->error = 0;
5c1e6aa3	62	dst->obsolete = initial_obsolete;
cf911662 DM	63	dst->header_len = 0;
	64	dst->trailer_len = 0;
	65	#ifdef CONFIG_IP_ROUTE_CLASSID
	66	dst->tclassid = 0;
1da177e4	67	#endif
61adedf3	68	dst->lwtstate = NULL;
5c1e6aa3	69	atomic_set(&dst->__refcnt, initial_ref);
cf911662	70	dst->__use = 0;
5c1e6aa3 DM	71	dst->lastuse = jiffies;
5c1e6aa3 DM	72	dst->flags = flags;
957c665f DM	73	if (!(flags & DST_NOCOUNT))
957c665f DM	74	dst_entries_add(ops, 1);
f38a9eb1 TG	75	}
	76	EXPORT_SYMBOL(dst_init);
	77
	78	void dst_alloc(struct dst_ops ops, struct net_device *dev,
	79	int initial_ref, int initial_obsolete, unsigned short flags)
	80	{
	81	struct dst_entry *dst;
	82
cf86a086 ED	83	if (ops->gc &&
	84	!(flags & DST_NOCOUNT) &&
	85	dst_entries_get_fast(ops) > ops->gc_thresh) {
22c2ad61	86	if (ops->gc(ops)) {
cf86a086	87	pr_notice_ratelimited("Route cache is full: consider increasing sysctl net.ipv6.route.max_size.\n");
f38a9eb1	88	return NULL;
22c2ad61	89	}
f38a9eb1 TG	90	}
	91
	92	dst = kmem_cache_alloc(ops->kmem_cachep, GFP_ATOMIC);
	93	if (!dst)
	94	return NULL;
	95
	96	dst_init(dst, ops, dev, initial_ref, initial_obsolete, flags);
	97
1da177e4 LT	98	return dst;
1da177e4 LT	99	}
598ed936	100	EXPORT_SYMBOL(dst_alloc);
1da177e4	101
1da177e4 LT	102	struct dst_entry dst_destroy(struct dst_entry dst)
1da177e4 LT	103	{
b92cf4aa	104	struct dst_entry *child = NULL;
1da177e4 LT	105
	106	smp_rmb();
	107
b92cf4aa	108	#ifdef CONFIG_XFRM
b6ca8bd5 DM	109	if (dst->xfrm) {
	110	struct xfrm_dst xdst = (struct xfrm_dst ) dst;
	111
	112	child = xdst->child;
	113	}
b92cf4aa	114	#endif
957c665f DM	115	if (!(dst->flags & DST_NOCOUNT))
957c665f DM	116	dst_entries_add(dst->ops, -1);
1da177e4 LT	117
	118	if (dst->ops->destroy)
	119	dst->ops->destroy(dst);
d62607c3	120	netdev_put(dst->dev, &dst->dev_tracker);
f38a9eb1	121
e252b3d1 WC	122	lwtstate_put(dst->lwtstate);
e252b3d1 WC	123
f38a9eb1	124	if (dst->flags & DST_METADATA)
d71785ff	125	metadata_dst_free((struct metadata_dst *)dst);
f38a9eb1 TG	126	else
f38a9eb1 TG	127	kmem_cache_free(dst->ops->kmem_cachep, dst);
1da177e4 LT	128
1da177e4 LT	129	dst = child;
52df157f WW	130	if (dst)
52df157f WW	131	dst_release_immediate(dst);
1da177e4 LT	132	return NULL;
1da177e4 LT	133	}
598ed936	134	EXPORT_SYMBOL(dst_destroy);
1da177e4	135
f8864972 ED	136	static void dst_destroy_rcu(struct rcu_head *head)
	137	{
	138	struct dst_entry *dst = container_of(head, struct dst_entry, rcu_head);
	139
	140	dst = dst_destroy(dst);
f8864972 ED	141	}
f8864972 ED	142
4a6ce2b6 WW	143	/* Operations to mark dst as DEAD and clean up the net device referenced
4a6ce2b6 WW	144	* by dst:
1be107de	145	* 1. put the dst under blackhole interface and discard all tx/rx packets
4a6ce2b6 WW	146	* on this route.
	147	* 2. release the net_device
	148	* This function should be called when removing routes from the fib tree
	149	* in preparation for a NETDEV_DOWN/NETDEV_UNREGISTER event and also to
	150	* make the next dst_ops->check() fail.
	151	*/
	152	void dst_dev_put(struct dst_entry *dst)
	153	{
	154	struct net_device *dev = dst->dev;
	155
	156	dst->obsolete = DST_OBSOLETE_DEAD;
	157	if (dst->ops->ifdown)
	158	dst->ops->ifdown(dst, dev, true);
	159	dst->input = dst_discard;
	160	dst->output = dst_discard_out;
8d7017fd	161	dst->dev = blackhole_netdev;
d62607c3 JK	162	netdev_ref_replace(dev, blackhole_netdev, &dst->dev_tracker,
d62607c3 JK	163	GFP_ATOMIC);
4a6ce2b6 WW	164	}
	165	EXPORT_SYMBOL(dst_dev_put);
	166
8d330868 IJ	167	void dst_release(struct dst_entry *dst)
	168	{
	169	if (dst) {
598ed936	170	int newrefcnt;
ef711cf1	171
598ed936	172	newrefcnt = atomic_dec_return(&dst->__refcnt);
adecda5b	173	if (WARN_ONCE(newrefcnt < 0, "dst_release underflow"))
8bf4ada2 KK	174	net_warn_ratelimited("%s: dst:%p refcnt:%d\n",
8bf4ada2 KK	175	__func__, dst, newrefcnt);
b2a9c0ed	176	if (!newrefcnt)
f8864972	177	call_rcu(&dst->rcu_head, dst_destroy_rcu);
8d330868 IJ	178	}
	179	}
	180	EXPORT_SYMBOL(dst_release);
	181
5f56f409 WW	182	void dst_release_immediate(struct dst_entry *dst)
	183	{
	184	if (dst) {
	185	int newrefcnt;
	186
	187	newrefcnt = atomic_dec_return(&dst->__refcnt);
adecda5b	188	if (WARN_ONCE(newrefcnt < 0, "dst_release_immediate underflow"))
5f56f409 WW	189	net_warn_ratelimited("%s: dst:%p refcnt:%d\n",
	190	__func__, dst, newrefcnt);
	191	if (!newrefcnt)
	192	dst_destroy(dst);
	193	}
	194	}
	195	EXPORT_SYMBOL(dst_release_immediate);
	196
62fa8a84 DM	197	u32 dst_cow_metrics_generic(struct dst_entry dst, unsigned long old)
62fa8a84 DM	198	{
3fb07daf	199	struct dst_metrics p = kmalloc(sizeof(p), GFP_ATOMIC);
62fa8a84 DM	200
62fa8a84 DM	201	if (p) {
3fb07daf	202	struct dst_metrics old_p = (struct dst_metrics )__DST_METRICS_PTR(old);
62fa8a84 DM	203	unsigned long prev, new;
62fa8a84 DM	204
9620fef2	205	refcount_set(&p->refcnt, 1);
3fb07daf	206	memcpy(p->metrics, old_p->metrics, sizeof(p->metrics));
62fa8a84 DM	207
	208	new = (unsigned long) p;
	209	prev = cmpxchg(&dst->_metrics, old, new);
	210
	211	if (prev != old) {
	212	kfree(p);
3fb07daf	213	p = (struct dst_metrics *)__DST_METRICS_PTR(prev);
62fa8a84 DM	214	if (prev & DST_METRICS_READ_ONLY)
62fa8a84 DM	215	p = NULL;
3fb07daf	216	} else if (prev & DST_METRICS_REFCOUNTED) {
9620fef2	217	if (refcount_dec_and_test(&old_p->refcnt))
3fb07daf	218	kfree(old_p);
62fa8a84 DM	219	}
62fa8a84 DM	220	}
3fb07daf ED	221	BUILD_BUG_ON(offsetof(struct dst_metrics, metrics) != 0);
3fb07daf ED	222	return (u32 *)p;
62fa8a84 DM	223	}
	224	EXPORT_SYMBOL(dst_cow_metrics_generic);
	225
	226	/* Caller asserts that dst_metrics_read_only(dst) is false. */
	227	void __dst_destroy_metrics_generic(struct dst_entry *dst, unsigned long old)
	228	{
	229	unsigned long prev, new;
	230
3fb07daf	231	new = ((unsigned long) &dst_default_metrics) \| DST_METRICS_READ_ONLY;
62fa8a84 DM	232	prev = cmpxchg(&dst->_metrics, old, new);
	233	if (prev == old)
	234	kfree(__DST_METRICS_PTR(old));
	235	}
	236	EXPORT_SYMBOL(__dst_destroy_metrics_generic);
	237
c4c877b2 DB	238	struct dst_entry dst_blackhole_check(struct dst_entry dst, u32 cookie)
	239	{
	240	return NULL;
	241	}
	242
	243	u32 dst_blackhole_cow_metrics(struct dst_entry dst, unsigned long old)
	244	{
	245	return NULL;
	246	}
	247
	248	struct neighbour dst_blackhole_neigh_lookup(const struct dst_entry dst,
	249	struct sk_buff *skb,
	250	const void *daddr)
	251	{
	252	return NULL;
	253	}
	254
	255	void dst_blackhole_update_pmtu(struct dst_entry dst, struct sock sk,
	256	struct sk_buff *skb, u32 mtu,
	257	bool confirm_neigh)
	258	{
	259	}
	260	EXPORT_SYMBOL_GPL(dst_blackhole_update_pmtu);
	261
	262	void dst_blackhole_redirect(struct dst_entry dst, struct sock sk,
	263	struct sk_buff *skb)
	264	{
	265	}
	266	EXPORT_SYMBOL_GPL(dst_blackhole_redirect);
	267
	268	unsigned int dst_blackhole_mtu(const struct dst_entry *dst)
	269	{
	270	unsigned int mtu = dst_metric_raw(dst, RTAX_MTU);
	271
	272	return mtu ? : dst->dev->mtu;
	273	}
	274	EXPORT_SYMBOL_GPL(dst_blackhole_mtu);
	275
a188bb56 DB	276	static struct dst_ops dst_blackhole_ops = {
	277	.family = AF_UNSPEC,
	278	.neigh_lookup = dst_blackhole_neigh_lookup,
	279	.check = dst_blackhole_check,
	280	.cow_metrics = dst_blackhole_cow_metrics,
	281	.update_pmtu = dst_blackhole_update_pmtu,
	282	.redirect = dst_blackhole_redirect,
	283	.mtu = dst_blackhole_mtu,
f38a9eb1 TG	284	};
f38a9eb1 TG	285
3fcece12 JK	286	static void __metadata_dst_init(struct metadata_dst *md_dst,
3fcece12 JK	287	enum metadata_type type, u8 optslen)
f38a9eb1	288	{
f38a9eb1 TG	289	struct dst_entry *dst;
f38a9eb1 TG	290
f38a9eb1	291	dst = &md_dst->dst;
a188bb56	292	dst_init(dst, &dst_blackhole_ops, NULL, 1, DST_OBSOLETE_NONE,
a4c2fd7f	293	DST_METADATA \| DST_NOCOUNT);
f38a9eb1	294	memset(dst + 1, 0, sizeof(md_dst) + optslen - sizeof(dst));
3fcece12	295	md_dst->type = type;
d3aa45ce AS	296	}
d3aa45ce AS	297
3fcece12 JK	298	struct metadata_dst *metadata_dst_alloc(u8 optslen, enum metadata_type type,
3fcece12 JK	299	gfp_t flags)
d3aa45ce AS	300	{
	301	struct metadata_dst *md_dst;
	302
	303	md_dst = kmalloc(sizeof(*md_dst) + optslen, flags);
	304	if (!md_dst)
	305	return NULL;
	306
3fcece12	307	__metadata_dst_init(md_dst, type, optslen);
f38a9eb1 TG	308
	309	return md_dst;
	310	}
	311	EXPORT_SYMBOL_GPL(metadata_dst_alloc);
	312
d71785ff PA	313	void metadata_dst_free(struct metadata_dst *md_dst)
	314	{
	315	#ifdef CONFIG_DST_CACHE
e65a4955 DL	316	if (md_dst->type == METADATA_IP_TUNNEL)
e65a4955 DL	317	dst_cache_destroy(&md_dst->u.tun_info.dst_cache);
d71785ff PA	318	#endif
	319	kfree(md_dst);
	320	}
af308b94	321	EXPORT_SYMBOL_GPL(metadata_dst_free);
d71785ff	322
3fcece12 JK	323	struct metadata_dst __percpu *
3fcece12 JK	324	metadata_dst_alloc_percpu(u8 optslen, enum metadata_type type, gfp_t flags)
d3aa45ce AS	325	{
	326	int cpu;
	327	struct metadata_dst __percpu *md_dst;
	328
	329	md_dst = __alloc_percpu_gfp(sizeof(struct metadata_dst) + optslen,
	330	__alignof__(struct metadata_dst), flags);
	331	if (!md_dst)
	332	return NULL;
	333
	334	for_each_possible_cpu(cpu)
3fcece12	335	__metadata_dst_init(per_cpu_ptr(md_dst, cpu), type, optslen);
d3aa45ce AS	336
	337	return md_dst;
	338	}
	339	EXPORT_SYMBOL_GPL(metadata_dst_alloc_percpu);
d66f2b91 JK	340
	341	void metadata_dst_free_percpu(struct metadata_dst __percpu *md_dst)
	342	{
833e0e2f	343	#ifdef CONFIG_DST_CACHE
d66f2b91 JK	344	int cpu;
d66f2b91 JK	345
d66f2b91 JK	346	for_each_possible_cpu(cpu) {
	347	struct metadata_dst *one_md_dst = per_cpu_ptr(md_dst, cpu);
	348
	349	if (one_md_dst->type == METADATA_IP_TUNNEL)
	350	dst_cache_destroy(&one_md_dst->u.tun_info.dst_cache);
	351	}
	352	#endif
	353	free_percpu(md_dst);
	354	}
	355	EXPORT_SYMBOL_GPL(metadata_dst_free_percpu);