1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
|
/*
* 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/netdevice.h>
#include <linux/skbuff.h>
#include <linux/string.h>
#include <linux/types.h>
#include <net/net_namespace.h>
#include <net/dst.h>
/*
* Theory of operations:
* 1) We use a list, protected by a spinlock, to add
* new entries from both BH and non-BH context.
* 2) In order to keep spinlock held for a small delay,
* we use a second list where are stored long lived
* entries, that are handled by the garbage collect thread
* fired by a workqueue.
* 3) This list is guarded by a mutex,
* so that the gc_task and dst_dev_event() can be synchronized.
*/
#if RT_CACHE_DEBUG >= 2
static atomic_t dst_total = ATOMIC_INIT(0);
#endif
/*
* We want to keep lock & list close together
* to dirty as few cache lines as possible in __dst_free().
* As this is not a very strong hint, we dont force an alignment on SMP.
*/
static struct {
spinlock_t lock;
struct dst_entry *list;
unsigned long timer_inc;
unsigned long timer_expires;
} dst_garbage = {
.lock = __SPIN_LOCK_UNLOCKED(dst_garbage.lock),
.timer_inc = DST_GC_MAX,
};
static void dst_gc_task(struct work_struct *work);
static void ___dst_free(struct dst_entry * dst);
static DECLARE_DELAYED_WORK(dst_gc_work, dst_gc_task);
static DEFINE_MUTEX(dst_gc_mutex);
/*
* long lived entries are maintained in this list, guarded by dst_gc_mutex
*/
static struct dst_entry *dst_busy_list;
static void dst_gc_task(struct work_struct *work)
{
int delayed = 0;
int work_performed = 0;
unsigned long expires = ~0L;
struct dst_entry *dst, *next, head;
struct dst_entry *last = &head;
#if RT_CACHE_DEBUG >= 2
ktime_t time_start = ktime_get();
struct timespec elapsed;
#endif
mutex_lock(&dst_gc_mutex);
next = dst_busy_list;
loop:
while ((dst = next) != NULL) {
next = dst->next;
prefetch(&next->next);
if (likely(atomic_read(&dst->__refcnt))) {
last->next = dst;
last = dst;
delayed++;
continue;
}
work_performed++;
dst = dst_destroy(dst);
if (dst) {
/* NOHASH and still referenced. Unless it is already
* on gc list, invalidate it and add to gc list.
*
* Note: this is temporary. Actually, NOHASH dst's
* must be obsoleted when parent is obsoleted.
* But we do not have state "obsoleted, but
* referenced by parent", so it is right.
*/
if (dst->obsolete > 1)
continue;
___dst_free(dst);
dst->next = next;
next = dst;
}
}
spin_lock_bh(&dst_garbage.lock);
next = dst_garbage.list;
if (next) {
dst_garbage.list = NULL;
spin_unlock_bh(&dst_garbage.lock);
goto loop;
}
last->next = NULL;
dst_busy_list = head.next;
if (!dst_busy_list)
dst_garbage.timer_inc = DST_GC_MAX;
else {
/*
* if we freed less than 1/10 of delayed entries,
* we can sleep longer.
*/
if (work_performed <= delayed/10) {
dst_garbage.timer_expires += dst_garbage.timer_inc;
if (dst_garbage.timer_expires > DST_GC_MAX)
dst_garbage.timer_expires = DST_GC_MAX;
dst_garbage.timer_inc += DST_GC_INC;
} else {
dst_garbage.timer_inc = DST_GC_INC;
dst_garbage.timer_expires = DST_GC_MIN;
}
expires = dst_garbage.timer_expires;
/*
* if the next desired timer is more than 4 seconds in the future
* then round the timer to whole seconds
*/
if (expires > 4*HZ)
expires = round_jiffies_relative(expires);
schedule_delayed_work(&dst_gc_work, expires);
}
spin_unlock_bh(&dst_garbage.lock);
mutex_unlock(&dst_gc_mutex);
#if RT_CACHE_DEBUG >= 2
elapsed = ktime_to_timespec(ktime_sub(ktime_get(), time_start));
printk(KERN_DEBUG "dst_total: %d delayed: %d work_perf: %d"
" expires: %lu elapsed: %lu us\n",
atomic_read(&dst_total), delayed, work_performed,
expires,
elapsed.tv_sec * USEC_PER_SEC + elapsed.tv_nsec / NSEC_PER_USEC);
#endif
}
static int dst_discard(struct sk_buff *skb)
{
kfree_skb(skb);
return 0;
}
void * dst_alloc(struct dst_ops * ops)
{
struct dst_entry * dst;
if (ops->gc && atomic_read(&ops->entries) > ops->gc_thresh) {
if (ops->gc())
return NULL;
}
dst = kmem_cache_zalloc(ops->kmem_cachep, GFP_ATOMIC);
if (!dst)
return NULL;
atomic_set(&dst->__refcnt, 0);
dst->ops = ops;
dst->lastuse = jiffies;
dst->path = dst;
dst->input = dst->output = dst_discard;
#if RT_CACHE_DEBUG >= 2
atomic_inc(&dst_total);
#endif
atomic_inc(&ops->entries);
return dst;
}
static void ___dst_free(struct dst_entry * dst)
{
/* The first case (dev==NULL) is required, when
protocol module is unloaded.
*/
if (dst->dev == NULL || !(dst->dev->flags&IFF_UP)) {
dst->input = dst->output = dst_discard;
}
dst->obsolete = 2;
}
void __dst_free(struct dst_entry * dst)
{
spin_lock_bh(&dst_garbage.lock);
___dst_free(dst);
dst->next = dst_garbage.list;
dst_garbage.list = dst;
if (dst_garbage.timer_inc > DST_GC_INC) {
dst_garbage.timer_inc = DST_GC_INC;
dst_garbage.timer_expires = DST_GC_MIN;
schedule_delayed_work(&dst_gc_work, dst_garbage.timer_expires);
}
spin_unlock_bh(&dst_garbage.lock);
}
struct dst_entry *dst_destroy(struct dst_entry * dst)
{
struct dst_entry *child;
struct neighbour *neigh;
struct hh_cache *hh;
smp_rmb();
again:
neigh = dst->neighbour;
hh = dst->hh;
child = dst->child;
dst->hh = NULL;
if (hh && atomic_dec_and_test(&hh->hh_refcnt))
kfree(hh);
if (neigh) {
dst->neighbour = NULL;
neigh_release(neigh);
}
atomic_dec(&dst->ops->entries);
if (dst->ops->destroy)
dst->ops->destroy(dst);
if (dst->dev)
dev_put(dst->dev);
#if RT_CACHE_DEBUG >= 2
atomic_dec(&dst_total);
#endif
kmem_cache_free(dst->ops->kmem_cachep, dst);
dst = child;
if (dst) {
int nohash = dst->flags & DST_NOHASH;
if (atomic_dec_and_test(&dst->__refcnt)) {
/* We were real parent of this dst, so kill child. */
if (nohash)
goto again;
} else {
/* Child is still referenced, return it for freeing. */
if (nohash)
return dst;
/* Child is still in his hash table */
}
}
return NULL;
}
/* Dirty hack. We did it in 2.2 (in __dst_free),
* we have _very_ good reasons not to repeat
* this mistake in 2.3, but we have no choice
* now. _It_ _is_ _explicit_ _deliberate_
* _race_ _condition_.
*
* Commented and originally written by Alexey.
*/
static inline void dst_ifdown(struct dst_entry *dst, struct net_device *dev,
int unregister)
{
if (dst->ops->ifdown)
dst->ops->ifdown(dst, dev, unregister);
if (dev != dst->dev)
return;
if (!unregister) {
dst->input = dst->output = dst_discard;
} else {
dst->dev = init_net.loopback_dev;
dev_hold(dst->dev);
dev_put(dev);
if (dst->neighbour && dst->neighbour->dev == dev) {
dst->neighbour->dev = init_net.loopback_dev;
dev_put(dev);
dev_hold(dst->neighbour->dev);
}
}
}
static int dst_dev_event(struct notifier_block *this, unsigned long event, void *ptr)
{
struct net_device *dev = ptr;
struct dst_entry *dst, *last = NULL;
if (dev->nd_net != &init_net)
return NOTIFY_DONE;
switch<
|