aha/include/net/dst.h
Eric Dumazet 5635c10d97 net: make sure struct dst_entry refcount is aligned on 64 bytes
As found in the past (commit f1dd9c379c
[NET]: Fix tbench regression in 2.6.25-rc1), it is really
important that struct dst_entry refcount is aligned on a cache line.

We cannot use __atribute((aligned)), so manually pad the structure
for 32 and 64 bit arches.

for 32bit : offsetof(truct dst_entry, __refcnt) is 0x80
for 64bit : offsetof(truct dst_entry, __refcnt) is 0xc0

As it is not possible to guess at compile time cache line size,
we use a generic value of 64 bytes, that satisfies many current arches.
(Using 128 bytes alignment on 64bit arches would waste 64 bytes)

Add a BUILD_BUG_ON to catch future updates to "struct dst_entry" dont
break this alignment.

"tbench 8" is 4.4 % faster on a dual quad core (HP BL460c G1), Intel E5450 @3.00GHz
(2350 MB/s instead of 2250 MB/s)

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-16 19:46:36 -08:00

312 lines
6.9 KiB
C

/*
* net/dst.h Protocol independent destination cache definitions.
*
* Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
*
*/
#ifndef _NET_DST_H
#define _NET_DST_H
#include <linux/netdevice.h>
#include <linux/rtnetlink.h>
#include <linux/rcupdate.h>
#include <linux/jiffies.h>
#include <net/neighbour.h>
#include <asm/processor.h>
/*
* 0 - no debugging messages
* 1 - rare events and bugs (default)
* 2 - trace mode.
*/
#define RT_CACHE_DEBUG 0
#define DST_GC_MIN (HZ/10)
#define DST_GC_INC (HZ/2)
#define DST_GC_MAX (120*HZ)
/* Each dst_entry has reference count and sits in some parent list(s).
* When it is removed from parent list, it is "freed" (dst_free).
* After this it enters dead state (dst->obsolete > 0) and if its refcnt
* is zero, it can be destroyed immediately, otherwise it is added
* to gc list and garbage collector periodically checks the refcnt.
*/
struct sk_buff;
struct dst_entry
{
struct rcu_head rcu_head;
struct dst_entry *child;
struct net_device *dev;
short error;
short obsolete;
int flags;
#define DST_HOST 1
#define DST_NOXFRM 2
#define DST_NOPOLICY 4
#define DST_NOHASH 8
unsigned long expires;
unsigned short header_len; /* more space at head required */
unsigned short trailer_len; /* space to reserve at tail */
unsigned int rate_tokens;
unsigned long rate_last; /* rate limiting for ICMP */
struct dst_entry *path;
struct neighbour *neighbour;
struct hh_cache *hh;
#ifdef CONFIG_XFRM
struct xfrm_state *xfrm;
#else
void *__pad1;
#endif
int (*input)(struct sk_buff*);
int (*output)(struct sk_buff*);
struct dst_ops *ops;
u32 metrics[RTAX_MAX];
#ifdef CONFIG_NET_CLS_ROUTE
__u32 tclassid;
#else
__u32 __pad2;
#endif
/*
* Align __refcnt to a 64 bytes alignment
* (L1_CACHE_SIZE would be too much)
*/
#ifdef CONFIG_64BIT
long __pad_to_align_refcnt[2];
#else
long __pad_to_align_refcnt[1];
#endif
/*
* __refcnt wants to be on a different cache line from
* input/output/ops or performance tanks badly
*/
atomic_t __refcnt; /* client references */
int __use;
unsigned long lastuse;
union {
struct dst_entry *next;
struct rtable *rt_next;
struct rt6_info *rt6_next;
struct dn_route *dn_next;
};
};
struct dst_ops
{
unsigned short family;
__be16 protocol;
unsigned gc_thresh;
int (*gc)(struct dst_ops *ops);
struct dst_entry * (*check)(struct dst_entry *, __u32 cookie);
void (*destroy)(struct dst_entry *);
void (*ifdown)(struct dst_entry *,
struct net_device *dev, int how);
struct dst_entry * (*negative_advice)(struct dst_entry *);
void (*link_failure)(struct sk_buff *);
void (*update_pmtu)(struct dst_entry *dst, u32 mtu);
int (*local_out)(struct sk_buff *skb);
atomic_t entries;
struct kmem_cache *kmem_cachep;
struct net *dst_net;
};
#ifdef __KERNEL__
static inline u32
dst_metric(const struct dst_entry *dst, int metric)
{
return dst->metrics[metric-1];
}
static inline u32 dst_mtu(const struct dst_entry *dst)
{
u32 mtu = dst_metric(dst, RTAX_MTU);
/*
* Alexey put it here, so ask him about it :)
*/
barrier();
return mtu;
}
/* RTT metrics are stored in milliseconds for user ABI, but used as jiffies */
static inline unsigned long dst_metric_rtt(const struct dst_entry *dst, int metric)
{
return msecs_to_jiffies(dst_metric(dst, metric));
}
static inline void set_dst_metric_rtt(struct dst_entry *dst, int metric,
unsigned long rtt)
{
dst->metrics[metric-1] = jiffies_to_msecs(rtt);
}
static inline u32
dst_allfrag(const struct dst_entry *dst)
{
int ret = dst_metric(dst, RTAX_FEATURES) & RTAX_FEATURE_ALLFRAG;
/* Yes, _exactly_. This is paranoia. */
barrier();
return ret;
}
static inline int
dst_metric_locked(struct dst_entry *dst, int metric)
{
return dst_metric(dst, RTAX_LOCK) & (1<<metric);
}
static inline void dst_hold(struct dst_entry * dst)
{
/*
* If your kernel compilation stops here, please check
* __pad_to_align_refcnt declaration in struct dst_entry
*/
BUILD_BUG_ON(offsetof(struct dst_entry, __refcnt) & 63);
atomic_inc(&dst->__refcnt);
}
static inline void dst_use(struct dst_entry *dst, unsigned long time)
{
dst_hold(dst);
dst->__use++;
dst->lastuse = time;
}
static inline
struct dst_entry * dst_clone(struct dst_entry * dst)
{
if (dst)
atomic_inc(&dst->__refcnt);
return dst;
}
extern void dst_release(struct dst_entry *dst);
/* Children define the path of the packet through the
* Linux networking. Thus, destinations are stackable.
*/
static inline struct dst_entry *dst_pop(struct dst_entry *dst)
{
struct dst_entry *child = dst_clone(dst->child);
dst_release(dst);
return child;
}
extern int dst_discard(struct sk_buff *skb);
extern void * dst_alloc(struct dst_ops * ops);
extern void __dst_free(struct dst_entry * dst);
extern struct dst_entry *dst_destroy(struct dst_entry * dst);
static inline void dst_free(struct dst_entry * dst)
{
if (dst->obsolete > 1)
return;
if (!atomic_read(&dst->__refcnt)) {
dst = dst_destroy(dst);
if (!dst)
return;
}
__dst_free(dst);
}
static inline void dst_rcu_free(struct rcu_head *head)
{
struct dst_entry *dst = container_of(head, struct dst_entry, rcu_head);
dst_free(dst);
}
static inline void dst_confirm(struct dst_entry *dst)
{
if (dst)
neigh_confirm(dst->neighbour);
}
static inline void dst_negative_advice(struct dst_entry **dst_p)
{
struct dst_entry * dst = *dst_p;
if (dst && dst->ops->negative_advice)
*dst_p = dst->ops->negative_advice(dst);
}
static inline void dst_link_failure(struct sk_buff *skb)
{
struct dst_entry * dst = skb->dst;
if (dst && dst->ops && dst->ops->link_failure)
dst->ops->link_failure(skb);
}
static inline void dst_set_expires(struct dst_entry *dst, int timeout)
{
unsigned long expires = jiffies + timeout;
if (expires == 0)
expires = 1;
if (dst->expires == 0 || time_before(expires, dst->expires))
dst->expires = expires;
}
/* Output packet to network from transport. */
static inline int dst_output(struct sk_buff *skb)
{
return skb->dst->output(skb);
}
/* Input packet from network to transport. */
static inline int dst_input(struct sk_buff *skb)
{
return skb->dst->input(skb);
}
static inline struct dst_entry *dst_check(struct dst_entry *dst, u32 cookie)
{
if (dst->obsolete)
dst = dst->ops->check(dst, cookie);
return dst;
}
extern void dst_init(void);
/* Flags for xfrm_lookup flags argument. */
enum {
XFRM_LOOKUP_WAIT = 1 << 0,
XFRM_LOOKUP_ICMP = 1 << 1,
};
struct flowi;
#ifndef CONFIG_XFRM
static inline int xfrm_lookup(struct dst_entry **dst_p, struct flowi *fl,
struct sock *sk, int flags)
{
return 0;
}
static inline int __xfrm_lookup(struct dst_entry **dst_p, struct flowi *fl,
struct sock *sk, int flags)
{
return 0;
}
#else
extern int xfrm_lookup(struct dst_entry **dst_p, struct flowi *fl,
struct sock *sk, int flags);
extern int __xfrm_lookup(struct dst_entry **dst_p, struct flowi *fl,
struct sock *sk, int flags);
#endif
#endif
#endif /* _NET_DST_H */