2005-04-17 05:20:36 +07:00
|
|
|
/*
|
|
|
|
* net/dst.h Protocol independent destination cache definitions.
|
|
|
|
*
|
|
|
|
* Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifndef _NET_DST_H
|
|
|
|
#define _NET_DST_H
|
|
|
|
|
2009-08-29 08:34:49 +07:00
|
|
|
#include <net/dst_ops.h>
|
2005-12-27 11:43:12 +07:00
|
|
|
#include <linux/netdevice.h>
|
2005-04-17 05:20:36 +07:00
|
|
|
#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;
|
|
|
|
|
2009-11-03 10:26:03 +07:00
|
|
|
struct dst_entry {
|
2007-02-10 07:26:55 +07:00
|
|
|
struct rcu_head rcu_head;
|
2005-04-17 05:20:36 +07:00
|
|
|
struct dst_entry *child;
|
|
|
|
struct net_device *dev;
|
2005-04-20 10:46:37 +07:00
|
|
|
short error;
|
|
|
|
short obsolete;
|
2005-04-17 05:20:36 +07:00
|
|
|
int flags;
|
net: introduce DST_NOCACHE flag
While doing stress tests with IP route cache disabled, and multi queue
devices, I noticed a very high contention on one rwlock used in
neighbour code.
When many cpus are trying to send frames (possibly using a high
performance multiqueue device) to the same neighbour, they fight for the
neigh->lock rwlock in order to call neigh_hh_init(), and fight on
hh->hh_refcnt (a pair of atomic_inc/atomic_dec_and_test())
But we dont need to call neigh_hh_init() for dst that are used only
once. It costs four atomic operations at least, on two contended cache
lines, plus the high contention on neigh->lock rwlock.
Introduce a new dst flag, DST_NOCACHE, that is set when dst was not
inserted in route cache.
With the stress test bench, sending 160000000 frames on one neighbour,
results are :
Before patch:
real 2m28.406s
user 0m11.781s
sys 36m17.964s
After patch:
real 1m26.532s
user 0m12.185s
sys 20m3.903s
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-10-04 12:17:54 +07:00
|
|
|
#define DST_HOST 0x0001
|
|
|
|
#define DST_NOXFRM 0x0002
|
|
|
|
#define DST_NOPOLICY 0x0004
|
|
|
|
#define DST_NOHASH 0x0008
|
|
|
|
#define DST_NOCACHE 0x0010
|
2005-04-17 05:20:36 +07:00
|
|
|
unsigned long expires;
|
|
|
|
|
|
|
|
unsigned short header_len; /* more space at head required */
|
|
|
|
unsigned short trailer_len; /* space to reserve at tail */
|
|
|
|
|
2008-01-22 21:18:34 +07:00
|
|
|
unsigned int rate_tokens;
|
[NET]: Fix tbench regression in 2.6.25-rc1
Comparing with kernel 2.6.24, tbench result has regression with
2.6.25-rc1.
1) On 2 quad-core processor stoakley: 4%.
2) On 4 quad-core processor tigerton: more than 30%.
bisect located below patch.
b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b is first bad commit
commit b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b
Author: Herbert Xu <herbert@gondor.apana.org.au>
Date: Tue Nov 13 21:33:32 2007 -0800
[IPV6]: Move nfheader_len into rt6_info
The dst member nfheader_len is only used by IPv6. It's also currently
creating a rather ugly alignment hole in struct dst. Therefore this patch
moves it from there into struct rt6_info.
Above patch changes the cache line alignment, especially member
__refcnt. I did a testing by adding 2 unsigned long pading before
lastuse, so the 3 members, lastuse/__refcnt/__use, are moved to next
cache line. The performance is recovered.
I created a patch to rearrange the members in struct dst_entry.
With Eric and Valdis Kletnieks's suggestion, I made finer arrangement.
1) Move tclassid under ops in case CONFIG_NET_CLS_ROUTE=y. So
sizeof(dst_entry)=200 no matter if CONFIG_NET_CLS_ROUTE=y/n. I
tested many patches on my 16-core tigerton by moving tclassid to
different place. It looks like tclassid could also have impact on
performance. If moving tclassid before metrics, or just don't move
tclassid, the performance isn't good. So I move it behind metrics.
2) Add comments before __refcnt.
On 16-core tigerton:
If CONFIG_NET_CLS_ROUTE=y, the result with below patch is about 18%
better than the one without the patch;
If CONFIG_NET_CLS_ROUTE=n, the result with below patch is about 30%
better than the one without the patch.
With 32bit 2.6.25-rc1 on 8-core stoakley, the new patch doesn't
introduce regression.
Thank Eric, Valdis, and David!
Signed-off-by: Zhang Yanmin <yanmin.zhang@intel.com>
Acked-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-13 12:52:37 +07:00
|
|
|
unsigned long rate_last; /* rate limiting for ICMP */
|
2008-01-22 21:18:34 +07:00
|
|
|
|
[NET]: Fix tbench regression in 2.6.25-rc1
Comparing with kernel 2.6.24, tbench result has regression with
2.6.25-rc1.
1) On 2 quad-core processor stoakley: 4%.
2) On 4 quad-core processor tigerton: more than 30%.
bisect located below patch.
b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b is first bad commit
commit b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b
Author: Herbert Xu <herbert@gondor.apana.org.au>
Date: Tue Nov 13 21:33:32 2007 -0800
[IPV6]: Move nfheader_len into rt6_info
The dst member nfheader_len is only used by IPv6. It's also currently
creating a rather ugly alignment hole in struct dst. Therefore this patch
moves it from there into struct rt6_info.
Above patch changes the cache line alignment, especially member
__refcnt. I did a testing by adding 2 unsigned long pading before
lastuse, so the 3 members, lastuse/__refcnt/__use, are moved to next
cache line. The performance is recovered.
I created a patch to rearrange the members in struct dst_entry.
With Eric and Valdis Kletnieks's suggestion, I made finer arrangement.
1) Move tclassid under ops in case CONFIG_NET_CLS_ROUTE=y. So
sizeof(dst_entry)=200 no matter if CONFIG_NET_CLS_ROUTE=y/n. I
tested many patches on my 16-core tigerton by moving tclassid to
different place. It looks like tclassid could also have impact on
performance. If moving tclassid before metrics, or just don't move
tclassid, the performance isn't good. So I move it behind metrics.
2) Add comments before __refcnt.
On 16-core tigerton:
If CONFIG_NET_CLS_ROUTE=y, the result with below patch is about 18%
better than the one without the patch;
If CONFIG_NET_CLS_ROUTE=n, the result with below patch is about 30%
better than the one without the patch.
With 32bit 2.6.25-rc1 on 8-core stoakley, the new patch doesn't
introduce regression.
Thank Eric, Valdis, and David!
Signed-off-by: Zhang Yanmin <yanmin.zhang@intel.com>
Acked-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-13 12:52:37 +07:00
|
|
|
struct dst_entry *path;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
struct neighbour *neighbour;
|
|
|
|
struct hh_cache *hh;
|
2008-10-29 03:24:06 +07:00
|
|
|
#ifdef CONFIG_XFRM
|
2005-04-17 05:20:36 +07:00
|
|
|
struct xfrm_state *xfrm;
|
2008-11-17 10:46:36 +07:00
|
|
|
#else
|
|
|
|
void *__pad1;
|
2008-10-29 03:24:06 +07:00
|
|
|
#endif
|
2005-04-17 05:20:36 +07:00
|
|
|
int (*input)(struct sk_buff*);
|
|
|
|
int (*output)(struct sk_buff*);
|
|
|
|
|
|
|
|
struct dst_ops *ops;
|
[NET]: Fix tbench regression in 2.6.25-rc1
Comparing with kernel 2.6.24, tbench result has regression with
2.6.25-rc1.
1) On 2 quad-core processor stoakley: 4%.
2) On 4 quad-core processor tigerton: more than 30%.
bisect located below patch.
b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b is first bad commit
commit b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b
Author: Herbert Xu <herbert@gondor.apana.org.au>
Date: Tue Nov 13 21:33:32 2007 -0800
[IPV6]: Move nfheader_len into rt6_info
The dst member nfheader_len is only used by IPv6. It's also currently
creating a rather ugly alignment hole in struct dst. Therefore this patch
moves it from there into struct rt6_info.
Above patch changes the cache line alignment, especially member
__refcnt. I did a testing by adding 2 unsigned long pading before
lastuse, so the 3 members, lastuse/__refcnt/__use, are moved to next
cache line. The performance is recovered.
I created a patch to rearrange the members in struct dst_entry.
With Eric and Valdis Kletnieks's suggestion, I made finer arrangement.
1) Move tclassid under ops in case CONFIG_NET_CLS_ROUTE=y. So
sizeof(dst_entry)=200 no matter if CONFIG_NET_CLS_ROUTE=y/n. I
tested many patches on my 16-core tigerton by moving tclassid to
different place. It looks like tclassid could also have impact on
performance. If moving tclassid before metrics, or just don't move
tclassid, the performance isn't good. So I move it behind metrics.
2) Add comments before __refcnt.
On 16-core tigerton:
If CONFIG_NET_CLS_ROUTE=y, the result with below patch is about 18%
better than the one without the patch;
If CONFIG_NET_CLS_ROUTE=n, the result with below patch is about 30%
better than the one without the patch.
With 32bit 2.6.25-rc1 on 8-core stoakley, the new patch doesn't
introduce regression.
Thank Eric, Valdis, and David!
Signed-off-by: Zhang Yanmin <yanmin.zhang@intel.com>
Acked-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-13 12:52:37 +07:00
|
|
|
|
|
|
|
u32 metrics[RTAX_MAX];
|
|
|
|
|
|
|
|
#ifdef CONFIG_NET_CLS_ROUTE
|
|
|
|
__u32 tclassid;
|
2008-11-17 10:46:36 +07:00
|
|
|
#else
|
|
|
|
__u32 __pad2;
|
[NET]: Fix tbench regression in 2.6.25-rc1
Comparing with kernel 2.6.24, tbench result has regression with
2.6.25-rc1.
1) On 2 quad-core processor stoakley: 4%.
2) On 4 quad-core processor tigerton: more than 30%.
bisect located below patch.
b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b is first bad commit
commit b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b
Author: Herbert Xu <herbert@gondor.apana.org.au>
Date: Tue Nov 13 21:33:32 2007 -0800
[IPV6]: Move nfheader_len into rt6_info
The dst member nfheader_len is only used by IPv6. It's also currently
creating a rather ugly alignment hole in struct dst. Therefore this patch
moves it from there into struct rt6_info.
Above patch changes the cache line alignment, especially member
__refcnt. I did a testing by adding 2 unsigned long pading before
lastuse, so the 3 members, lastuse/__refcnt/__use, are moved to next
cache line. The performance is recovered.
I created a patch to rearrange the members in struct dst_entry.
With Eric and Valdis Kletnieks's suggestion, I made finer arrangement.
1) Move tclassid under ops in case CONFIG_NET_CLS_ROUTE=y. So
sizeof(dst_entry)=200 no matter if CONFIG_NET_CLS_ROUTE=y/n. I
tested many patches on my 16-core tigerton by moving tclassid to
different place. It looks like tclassid could also have impact on
performance. If moving tclassid before metrics, or just don't move
tclassid, the performance isn't good. So I move it behind metrics.
2) Add comments before __refcnt.
On 16-core tigerton:
If CONFIG_NET_CLS_ROUTE=y, the result with below patch is about 18%
better than the one without the patch;
If CONFIG_NET_CLS_ROUTE=n, the result with below patch is about 30%
better than the one without the patch.
With 32bit 2.6.25-rc1 on 8-core stoakley, the new patch doesn't
introduce regression.
Thank Eric, Valdis, and David!
Signed-off-by: Zhang Yanmin <yanmin.zhang@intel.com>
Acked-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-13 12:52:37 +07:00
|
|
|
#endif
|
|
|
|
|
2008-11-17 10:46:36 +07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Align __refcnt to a 64 bytes alignment
|
|
|
|
* (L1_CACHE_SIZE would be too much)
|
|
|
|
*/
|
|
|
|
#ifdef CONFIG_64BIT
|
|
|
|
long __pad_to_align_refcnt[1];
|
|
|
|
#endif
|
[NET]: Fix tbench regression in 2.6.25-rc1
Comparing with kernel 2.6.24, tbench result has regression with
2.6.25-rc1.
1) On 2 quad-core processor stoakley: 4%.
2) On 4 quad-core processor tigerton: more than 30%.
bisect located below patch.
b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b is first bad commit
commit b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b
Author: Herbert Xu <herbert@gondor.apana.org.au>
Date: Tue Nov 13 21:33:32 2007 -0800
[IPV6]: Move nfheader_len into rt6_info
The dst member nfheader_len is only used by IPv6. It's also currently
creating a rather ugly alignment hole in struct dst. Therefore this patch
moves it from there into struct rt6_info.
Above patch changes the cache line alignment, especially member
__refcnt. I did a testing by adding 2 unsigned long pading before
lastuse, so the 3 members, lastuse/__refcnt/__use, are moved to next
cache line. The performance is recovered.
I created a patch to rearrange the members in struct dst_entry.
With Eric and Valdis Kletnieks's suggestion, I made finer arrangement.
1) Move tclassid under ops in case CONFIG_NET_CLS_ROUTE=y. So
sizeof(dst_entry)=200 no matter if CONFIG_NET_CLS_ROUTE=y/n. I
tested many patches on my 16-core tigerton by moving tclassid to
different place. It looks like tclassid could also have impact on
performance. If moving tclassid before metrics, or just don't move
tclassid, the performance isn't good. So I move it behind metrics.
2) Add comments before __refcnt.
On 16-core tigerton:
If CONFIG_NET_CLS_ROUTE=y, the result with below patch is about 18%
better than the one without the patch;
If CONFIG_NET_CLS_ROUTE=n, the result with below patch is about 30%
better than the one without the patch.
With 32bit 2.6.25-rc1 on 8-core stoakley, the new patch doesn't
introduce regression.
Thank Eric, Valdis, and David!
Signed-off-by: Zhang Yanmin <yanmin.zhang@intel.com>
Acked-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-13 12:52:37 +07:00
|
|
|
/*
|
|
|
|
* __refcnt wants to be on a different cache line from
|
|
|
|
* input/output/ops or performance tanks badly
|
|
|
|
*/
|
2007-02-10 07:26:55 +07:00
|
|
|
atomic_t __refcnt; /* client references */
|
|
|
|
int __use;
|
[NET]: Fix tbench regression in 2.6.25-rc1
Comparing with kernel 2.6.24, tbench result has regression with
2.6.25-rc1.
1) On 2 quad-core processor stoakley: 4%.
2) On 4 quad-core processor tigerton: more than 30%.
bisect located below patch.
b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b is first bad commit
commit b4ce92775c2e7ff9cf79cca4e0a19c8c5fd6287b
Author: Herbert Xu <herbert@gondor.apana.org.au>
Date: Tue Nov 13 21:33:32 2007 -0800
[IPV6]: Move nfheader_len into rt6_info
The dst member nfheader_len is only used by IPv6. It's also currently
creating a rather ugly alignment hole in struct dst. Therefore this patch
moves it from there into struct rt6_info.
Above patch changes the cache line alignment, especially member
__refcnt. I did a testing by adding 2 unsigned long pading before
lastuse, so the 3 members, lastuse/__refcnt/__use, are moved to next
cache line. The performance is recovered.
I created a patch to rearrange the members in struct dst_entry.
With Eric and Valdis Kletnieks's suggestion, I made finer arrangement.
1) Move tclassid under ops in case CONFIG_NET_CLS_ROUTE=y. So
sizeof(dst_entry)=200 no matter if CONFIG_NET_CLS_ROUTE=y/n. I
tested many patches on my 16-core tigerton by moving tclassid to
different place. It looks like tclassid could also have impact on
performance. If moving tclassid before metrics, or just don't move
tclassid, the performance isn't good. So I move it behind metrics.
2) Add comments before __refcnt.
On 16-core tigerton:
If CONFIG_NET_CLS_ROUTE=y, the result with below patch is about 18%
better than the one without the patch;
If CONFIG_NET_CLS_ROUTE=n, the result with below patch is about 30%
better than the one without the patch.
With 32bit 2.6.25-rc1 on 8-core stoakley, the new patch doesn't
introduce regression.
Thank Eric, Valdis, and David!
Signed-off-by: Zhang Yanmin <yanmin.zhang@intel.com>
Acked-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-13 12:52:37 +07:00
|
|
|
unsigned long lastuse;
|
2007-02-10 07:26:55 +07:00
|
|
|
union {
|
2010-10-29 10:09:24 +07:00
|
|
|
struct dst_entry *next;
|
|
|
|
struct rtable __rcu *rt_next;
|
|
|
|
struct rt6_info *rt6_next;
|
|
|
|
struct dn_route __rcu *dn_next;
|
2007-02-10 07:26:55 +07:00
|
|
|
};
|
2005-04-17 05:20:36 +07:00
|
|
|
};
|
|
|
|
|
|
|
|
#ifdef __KERNEL__
|
|
|
|
|
|
|
|
static inline u32
|
|
|
|
dst_metric(const struct dst_entry *dst, int metric)
|
|
|
|
{
|
|
|
|
return dst->metrics[metric-1];
|
|
|
|
}
|
|
|
|
|
2009-10-28 11:15:23 +07:00
|
|
|
static inline u32
|
|
|
|
dst_feature(const struct dst_entry *dst, u32 feature)
|
|
|
|
{
|
tcp: Revert per-route SACK/DSACK/TIMESTAMP changes.
It creates a regression, triggering badness for SYN_RECV
sockets, for example:
[19148.022102] Badness at net/ipv4/inet_connection_sock.c:293
[19148.022570] NIP: c02a0914 LR: c02a0904 CTR: 00000000
[19148.023035] REGS: eeecbd30 TRAP: 0700 Not tainted (2.6.32)
[19148.023496] MSR: 00029032 <EE,ME,CE,IR,DR> CR: 24002442 XER: 00000000
[19148.024012] TASK = eee9a820[1756] 'privoxy' THREAD: eeeca000
This is likely caused by the change in the 'estab' parameter
passed to tcp_parse_options() when invoked by the functions
in net/ipv4/tcp_minisocks.c
But even if that is fixed, the ->conn_request() changes made in
this patch series is fundamentally wrong. They try to use the
listening socket's 'dst' to probe the route settings. The
listening socket doesn't even have a route, and you can't
get the right route (the child request one) until much later
after we setup all of the state, and it must be done by hand.
This stuff really isn't ready, so the best thing to do is a
full revert. This reverts the following commits:
f55017a93f1a74d50244b1254b9a2bd7ac9bbf7d
022c3f7d82f0f1c68018696f2f027b87b9bb45c2
1aba721eba1d84a2defce45b950272cee1e6c72a
cda42ebd67ee5fdf09d7057b5a4584d36fe8a335
345cda2fd695534be5a4494f1b59da9daed33663
dc343475ed062e13fc260acccaab91d7d80fd5b2
05eaade2782fb0c90d3034fd7a7d5a16266182bb
6a2a2d6bf8581216e08be15fcb563cfd6c430e1e
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-16 11:56:42 +07:00
|
|
|
return dst_metric(dst, RTAX_FEATURES) & feature;
|
2009-10-28 11:15:23 +07:00
|
|
|
}
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
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;
|
|
|
|
}
|
|
|
|
|
2008-07-19 13:02:15 +07:00
|
|
|
/* 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);
|
|
|
|
}
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
static inline u32
|
|
|
|
dst_allfrag(const struct dst_entry *dst)
|
|
|
|
{
|
2009-10-28 11:15:23 +07:00
|
|
|
int ret = dst_feature(dst, RTAX_FEATURE_ALLFRAG);
|
2005-04-17 05:20:36 +07:00
|
|
|
/* 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)
|
|
|
|
{
|
2008-11-17 10:46:36 +07:00
|
|
|
/*
|
|
|
|
* 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);
|
2005-04-17 05:20:36 +07:00
|
|
|
atomic_inc(&dst->__refcnt);
|
|
|
|
}
|
|
|
|
|
2007-11-11 12:28:34 +07:00
|
|
|
static inline void dst_use(struct dst_entry *dst, unsigned long time)
|
|
|
|
{
|
|
|
|
dst_hold(dst);
|
|
|
|
dst->__use++;
|
|
|
|
dst->lastuse = time;
|
|
|
|
}
|
|
|
|
|
2010-05-12 06:19:48 +07:00
|
|
|
static inline void dst_use_noref(struct dst_entry *dst, unsigned long time)
|
|
|
|
{
|
|
|
|
dst->__use++;
|
|
|
|
dst->lastuse = time;
|
|
|
|
}
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
static inline
|
|
|
|
struct dst_entry * dst_clone(struct dst_entry * dst)
|
|
|
|
{
|
|
|
|
if (dst)
|
|
|
|
atomic_inc(&dst->__refcnt);
|
|
|
|
return dst;
|
|
|
|
}
|
|
|
|
|
2008-03-28 07:53:31 +07:00
|
|
|
extern void dst_release(struct dst_entry *dst);
|
2010-05-12 06:19:48 +07:00
|
|
|
|
|
|
|
static inline void refdst_drop(unsigned long refdst)
|
|
|
|
{
|
|
|
|
if (!(refdst & SKB_DST_NOREF))
|
|
|
|
dst_release((struct dst_entry *)(refdst & SKB_DST_PTRMASK));
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* skb_dst_drop - drops skb dst
|
|
|
|
* @skb: buffer
|
|
|
|
*
|
|
|
|
* Drops dst reference count if a reference was taken.
|
|
|
|
*/
|
2009-06-02 12:19:30 +07:00
|
|
|
static inline void skb_dst_drop(struct sk_buff *skb)
|
|
|
|
{
|
2010-05-12 06:19:48 +07:00
|
|
|
if (skb->_skb_refdst) {
|
|
|
|
refdst_drop(skb->_skb_refdst);
|
|
|
|
skb->_skb_refdst = 0UL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void skb_dst_copy(struct sk_buff *nskb, const struct sk_buff *oskb)
|
|
|
|
{
|
|
|
|
nskb->_skb_refdst = oskb->_skb_refdst;
|
|
|
|
if (!(nskb->_skb_refdst & SKB_DST_NOREF))
|
|
|
|
dst_clone(skb_dst(nskb));
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* skb_dst_force - makes sure skb dst is refcounted
|
|
|
|
* @skb: buffer
|
|
|
|
*
|
|
|
|
* If dst is not yet refcounted, let's do it
|
|
|
|
*/
|
|
|
|
static inline void skb_dst_force(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
if (skb_dst_is_noref(skb)) {
|
|
|
|
WARN_ON(!rcu_read_lock_held());
|
|
|
|
skb->_skb_refdst &= ~SKB_DST_NOREF;
|
|
|
|
dst_clone(skb_dst(skb));
|
|
|
|
}
|
2009-06-02 12:19:30 +07:00
|
|
|
}
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2010-05-18 12:36:55 +07:00
|
|
|
|
2010-09-27 07:33:35 +07:00
|
|
|
/**
|
|
|
|
* __skb_tunnel_rx - prepare skb for rx reinsert
|
|
|
|
* @skb: buffer
|
|
|
|
* @dev: tunnel device
|
|
|
|
*
|
|
|
|
* After decapsulation, packet is going to re-enter (netif_rx()) our stack,
|
|
|
|
* so make some cleanups. (no accounting done)
|
|
|
|
*/
|
|
|
|
static inline void __skb_tunnel_rx(struct sk_buff *skb, struct net_device *dev)
|
|
|
|
{
|
|
|
|
skb->dev = dev;
|
|
|
|
skb->rxhash = 0;
|
|
|
|
skb_set_queue_mapping(skb, 0);
|
|
|
|
skb_dst_drop(skb);
|
|
|
|
nf_reset(skb);
|
|
|
|
}
|
|
|
|
|
2010-05-18 12:36:55 +07:00
|
|
|
/**
|
|
|
|
* skb_tunnel_rx - prepare skb for rx reinsert
|
|
|
|
* @skb: buffer
|
|
|
|
* @dev: tunnel device
|
|
|
|
*
|
|
|
|
* After decapsulation, packet is going to re-enter (netif_rx()) our stack,
|
|
|
|
* so make some cleanups, and perform accounting.
|
2010-09-27 07:33:35 +07:00
|
|
|
* Note: this accounting is not SMP safe.
|
2010-05-18 12:36:55 +07:00
|
|
|
*/
|
|
|
|
static inline void skb_tunnel_rx(struct sk_buff *skb, struct net_device *dev)
|
|
|
|
{
|
|
|
|
/* TODO : stats should be SMP safe */
|
|
|
|
dev->stats.rx_packets++;
|
|
|
|
dev->stats.rx_bytes += skb->len;
|
2010-09-27 07:33:35 +07:00
|
|
|
__skb_tunnel_rx(skb, dev);
|
2010-05-18 12:36:55 +07:00
|
|
|
}
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
/* Children define the path of the packet through the
|
|
|
|
* Linux networking. Thus, destinations are stackable.
|
|
|
|
*/
|
|
|
|
|
2010-06-04 08:57:38 +07:00
|
|
|
static inline struct dst_entry *skb_dst_pop(struct sk_buff *skb)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
2010-06-04 08:57:38 +07:00
|
|
|
struct dst_entry *child = skb_dst(skb)->child;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2010-06-04 08:57:38 +07:00
|
|
|
skb_dst_drop(skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
return child;
|
|
|
|
}
|
|
|
|
|
2007-11-14 12:34:06 +07:00
|
|
|
extern int dst_discard(struct sk_buff *skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
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_link_failure(struct sk_buff *skb)
|
|
|
|
{
|
2009-06-02 12:19:30 +07:00
|
|
|
struct dst_entry *dst = skb_dst(skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
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)
|
|
|
|
{
|
2009-06-02 12:19:30 +07:00
|
|
|
return skb_dst(skb)->output(skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Input packet from network to transport. */
|
|
|
|
static inline int dst_input(struct sk_buff *skb)
|
|
|
|
{
|
2009-06-02 12:19:30 +07:00
|
|
|
return skb_dst(skb)->input(skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
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);
|
|
|
|
|
2007-12-13 01:36:59 +07:00
|
|
|
/* Flags for xfrm_lookup flags argument. */
|
|
|
|
enum {
|
|
|
|
XFRM_LOOKUP_WAIT = 1 << 0,
|
2007-12-13 01:44:43 +07:00
|
|
|
XFRM_LOOKUP_ICMP = 1 << 1,
|
2007-12-13 01:36:59 +07:00
|
|
|
};
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
struct flowi;
|
|
|
|
#ifndef CONFIG_XFRM
|
2008-11-26 08:35:18 +07:00
|
|
|
static inline int xfrm_lookup(struct net *net, struct dst_entry **dst_p,
|
|
|
|
struct flowi *fl, struct sock *sk, int flags)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
2008-11-26 08:35:18 +07:00
|
|
|
static inline int __xfrm_lookup(struct net *net, struct dst_entry **dst_p,
|
|
|
|
struct flowi *fl, struct sock *sk, int flags)
|
2007-05-25 08:17:54 +07:00
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
2005-04-17 05:20:36 +07:00
|
|
|
#else
|
2008-11-26 08:35:18 +07:00
|
|
|
extern int xfrm_lookup(struct net *net, struct dst_entry **dst_p,
|
|
|
|
struct flowi *fl, struct sock *sk, int flags);
|
|
|
|
extern int __xfrm_lookup(struct net *net, struct dst_entry **dst_p,
|
|
|
|
struct flowi *fl, struct sock *sk, int flags);
|
2005-04-17 05:20:36 +07:00
|
|
|
#endif
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#endif /* _NET_DST_H */
|