Commit f38a9eb1 authored by Thomas Graf's avatar Thomas Graf Committed by David S. Miller

dst: Metadata destinations

Introduces a new dst_metadata which enables to carry per packet metadata
between forwarding and processing elements via the skb->dst pointer.

The structure is set up to be a union. Thus, each separate type of
metadata requires its own dst instance. If demand arises to carry
multiple types of metadata concurrently, metadata dst entries can be
made stackable.

The metadata dst entry is refcnt'ed as expected for now but a non
reference counted use is possible if the reference is forced before
queueing the skb.

In order to allow allocating dsts with variable length, the existing
dst_alloc() is split into a dst_alloc() and dst_init() function. The
existing dst_init() function to initialize the subsystem is being
renamed to dst_subsys_init() to make it clear what is what.

The check before ip_route_input() is changed to ignore metadata dsts
and drop the dst inside the routing function thus allowing to interpret
metadata in a later commit.
Signed-off-by: default avatarThomas Graf <tgraf@suug.ch>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent 773a69d6
...@@ -57,6 +57,7 @@ struct dst_entry { ...@@ -57,6 +57,7 @@ struct dst_entry {
#define DST_FAKE_RTABLE 0x0040 #define DST_FAKE_RTABLE 0x0040
#define DST_XFRM_TUNNEL 0x0080 #define DST_XFRM_TUNNEL 0x0080
#define DST_XFRM_QUEUE 0x0100 #define DST_XFRM_QUEUE 0x0100
#define DST_METADATA 0x0200
unsigned short pending_confirm; unsigned short pending_confirm;
...@@ -356,6 +357,9 @@ static inline int dst_discard(struct sk_buff *skb) ...@@ -356,6 +357,9 @@ static inline int dst_discard(struct sk_buff *skb)
} }
void *dst_alloc(struct dst_ops *ops, struct net_device *dev, int initial_ref, void *dst_alloc(struct dst_ops *ops, struct net_device *dev, int initial_ref,
int initial_obsolete, unsigned short flags); int initial_obsolete, unsigned short flags);
void dst_init(struct dst_entry *dst, struct dst_ops *ops,
struct net_device *dev, int initial_ref, int initial_obsolete,
unsigned short flags);
void __dst_free(struct dst_entry *dst); void __dst_free(struct dst_entry *dst);
struct dst_entry *dst_destroy(struct dst_entry *dst); struct dst_entry *dst_destroy(struct dst_entry *dst);
...@@ -457,7 +461,7 @@ static inline struct dst_entry *dst_check(struct dst_entry *dst, u32 cookie) ...@@ -457,7 +461,7 @@ static inline struct dst_entry *dst_check(struct dst_entry *dst, u32 cookie)
return dst; return dst;
} }
void dst_init(void); void dst_subsys_init(void);
/* Flags for xfrm_lookup flags argument. */ /* Flags for xfrm_lookup flags argument. */
enum { enum {
......
#ifndef __NET_DST_METADATA_H
#define __NET_DST_METADATA_H 1
#include <linux/skbuff.h>
#include <net/ip_tunnels.h>
#include <net/dst.h>
struct metadata_dst {
struct dst_entry dst;
size_t opts_len;
};
static inline struct metadata_dst *skb_metadata_dst(struct sk_buff *skb)
{
struct metadata_dst *md_dst = (struct metadata_dst *) skb_dst(skb);
if (md_dst && md_dst->dst.flags & DST_METADATA)
return md_dst;
return NULL;
}
static inline bool skb_valid_dst(const struct sk_buff *skb)
{
struct dst_entry *dst = skb_dst(skb);
return dst && !(dst->flags & DST_METADATA);
}
struct metadata_dst *metadata_dst_alloc(u8 optslen, gfp_t flags);
#endif /* __NET_DST_METADATA_H */
...@@ -7669,7 +7669,7 @@ static int __init net_dev_init(void) ...@@ -7669,7 +7669,7 @@ static int __init net_dev_init(void)
open_softirq(NET_RX_SOFTIRQ, net_rx_action); open_softirq(NET_RX_SOFTIRQ, net_rx_action);
hotcpu_notifier(dev_cpu_callback, 0); hotcpu_notifier(dev_cpu_callback, 0);
dst_init(); dst_subsys_init();
rc = 0; rc = 0;
out: out:
return rc; return rc;
......
...@@ -22,6 +22,7 @@ ...@@ -22,6 +22,7 @@
#include <linux/prefetch.h> #include <linux/prefetch.h>
#include <net/dst.h> #include <net/dst.h>
#include <net/dst_metadata.h>
/* /*
* Theory of operations: * Theory of operations:
...@@ -158,19 +159,10 @@ const u32 dst_default_metrics[RTAX_MAX + 1] = { ...@@ -158,19 +159,10 @@ const u32 dst_default_metrics[RTAX_MAX + 1] = {
[RTAX_MAX] = 0xdeadbeef, [RTAX_MAX] = 0xdeadbeef,
}; };
void dst_init(struct dst_entry *dst, struct dst_ops *ops,
void *dst_alloc(struct dst_ops *ops, struct net_device *dev, struct net_device *dev, int initial_ref, int initial_obsolete,
int initial_ref, int initial_obsolete, unsigned short flags) unsigned short flags)
{ {
struct dst_entry *dst;
if (ops->gc && dst_entries_get_fast(ops) > ops->gc_thresh) {
if (ops->gc(ops))
return NULL;
}
dst = kmem_cache_alloc(ops->kmem_cachep, GFP_ATOMIC);
if (!dst)
return NULL;
dst->child = NULL; dst->child = NULL;
dst->dev = dev; dst->dev = dev;
if (dev) if (dev)
...@@ -200,6 +192,25 @@ void *dst_alloc(struct dst_ops *ops, struct net_device *dev, ...@@ -200,6 +192,25 @@ void *dst_alloc(struct dst_ops *ops, struct net_device *dev,
dst->next = NULL; dst->next = NULL;
if (!(flags & DST_NOCOUNT)) if (!(flags & DST_NOCOUNT))
dst_entries_add(ops, 1); 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 && dst_entries_get_fast(ops) > ops->gc_thresh) {
if (ops->gc(ops))
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; return dst;
} }
EXPORT_SYMBOL(dst_alloc); EXPORT_SYMBOL(dst_alloc);
...@@ -248,6 +259,10 @@ struct dst_entry *dst_destroy(struct dst_entry * dst) ...@@ -248,6 +259,10 @@ struct dst_entry *dst_destroy(struct dst_entry * dst)
dst->ops->destroy(dst); dst->ops->destroy(dst);
if (dst->dev) if (dst->dev)
dev_put(dst->dev); dev_put(dst->dev);
if (dst->flags & DST_METADATA)
kfree(dst);
else
kmem_cache_free(dst->ops->kmem_cachep, dst); kmem_cache_free(dst->ops->kmem_cachep, dst);
dst = child; dst = child;
...@@ -327,6 +342,47 @@ void __dst_destroy_metrics_generic(struct dst_entry *dst, unsigned long old) ...@@ -327,6 +342,47 @@ void __dst_destroy_metrics_generic(struct dst_entry *dst, unsigned long old)
} }
EXPORT_SYMBOL(__dst_destroy_metrics_generic); EXPORT_SYMBOL(__dst_destroy_metrics_generic);
static struct dst_ops md_dst_ops = {
.family = AF_UNSPEC,
};
static int dst_md_discard_sk(struct sock *sk, struct sk_buff *skb)
{
WARN_ONCE(1, "Attempting to call output on metadata dst\n");
kfree_skb(skb);
return 0;
}
static int dst_md_discard(struct sk_buff *skb)
{
WARN_ONCE(1, "Attempting to call input on metadata dst\n");
kfree_skb(skb);
return 0;
}
struct metadata_dst *metadata_dst_alloc(u8 optslen, gfp_t flags)
{
struct metadata_dst *md_dst;
struct dst_entry *dst;
md_dst = kmalloc(sizeof(*md_dst) + optslen, flags);
if (!md_dst)
return ERR_PTR(-ENOMEM);
dst = &md_dst->dst;
dst_init(dst, &md_dst_ops, NULL, 1, DST_OBSOLETE_NONE,
DST_METADATA | DST_NOCACHE | DST_NOCOUNT);
dst->input = dst_md_discard;
dst->output = dst_md_discard_sk;
memset(dst + 1, 0, sizeof(*md_dst) + optslen - sizeof(*dst));
md_dst->opts_len = optslen;
return md_dst;
}
EXPORT_SYMBOL_GPL(metadata_dst_alloc);
/* Dirty hack. We did it in 2.2 (in __dst_free), /* Dirty hack. We did it in 2.2 (in __dst_free),
* we have _very_ good reasons not to repeat * we have _very_ good reasons not to repeat
* this mistake in 2.3, but we have no choice * this mistake in 2.3, but we have no choice
...@@ -391,7 +447,7 @@ static struct notifier_block dst_dev_notifier = { ...@@ -391,7 +447,7 @@ static struct notifier_block dst_dev_notifier = {
.priority = -10, /* must be called after other network notifiers */ .priority = -10, /* must be called after other network notifiers */
}; };
void __init dst_init(void) void __init dst_subsys_init(void)
{ {
register_netdevice_notifier(&dst_dev_notifier); register_netdevice_notifier(&dst_dev_notifier);
} }
...@@ -146,6 +146,7 @@ ...@@ -146,6 +146,7 @@
#include <net/xfrm.h> #include <net/xfrm.h>
#include <linux/mroute.h> #include <linux/mroute.h>
#include <linux/netlink.h> #include <linux/netlink.h>
#include <net/dst_metadata.h>
/* /*
* Process Router Attention IP option (RFC 2113) * Process Router Attention IP option (RFC 2113)
...@@ -331,7 +332,7 @@ static int ip_rcv_finish(struct sock *sk, struct sk_buff *skb) ...@@ -331,7 +332,7 @@ static int ip_rcv_finish(struct sock *sk, struct sk_buff *skb)
* Initialise the virtual path cache for the packet. It describes * Initialise the virtual path cache for the packet. It describes
* how the packet travels inside Linux networking. * how the packet travels inside Linux networking.
*/ */
if (!skb_dst(skb)) { if (!skb_valid_dst(skb)) {
int err = ip_route_input_noref(skb, iph->daddr, iph->saddr, int err = ip_route_input_noref(skb, iph->daddr, iph->saddr,
iph->tos, skb->dev); iph->tos, skb->dev);
if (unlikely(err)) { if (unlikely(err)) {
......
...@@ -1690,6 +1690,8 @@ static int ip_route_input_slow(struct sk_buff *skb, __be32 daddr, __be32 saddr, ...@@ -1690,6 +1690,8 @@ static int ip_route_input_slow(struct sk_buff *skb, __be32 daddr, __be32 saddr,
by fib_lookup. by fib_lookup.
*/ */
skb_dst_drop(skb);
if (ipv4_is_multicast(saddr) || ipv4_is_lbcast(saddr)) if (ipv4_is_multicast(saddr) || ipv4_is_lbcast(saddr))
goto martian_source; goto martian_source;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment