Commit | Line | Data |
---|---|---|
1da177e4 LT |
1 | /* |
2 | * net/sched/sch_tbf.c Token Bucket Filter queue. | |
3 | * | |
4 | * This program is free software; you can redistribute it and/or | |
5 | * modify it under the terms of the GNU General Public License | |
6 | * as published by the Free Software Foundation; either version | |
7 | * 2 of the License, or (at your option) any later version. | |
8 | * | |
9 | * Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru> | |
10 | * Dmitry Torokhov <dtor@mail.ru> - allow attaching inner qdiscs - | |
11 | * original idea by Martin Devera | |
12 | * | |
13 | */ | |
14 | ||
1da177e4 | 15 | #include <linux/module.h> |
1da177e4 LT |
16 | #include <linux/types.h> |
17 | #include <linux/kernel.h> | |
1da177e4 | 18 | #include <linux/string.h> |
1da177e4 | 19 | #include <linux/errno.h> |
1da177e4 | 20 | #include <linux/skbuff.h> |
0ba48053 | 21 | #include <net/netlink.h> |
1da177e4 LT |
22 | #include <net/pkt_sched.h> |
23 | ||
24 | ||
25 | /* Simple Token Bucket Filter. | |
26 | ======================================= | |
27 | ||
28 | SOURCE. | |
29 | ------- | |
30 | ||
31 | None. | |
32 | ||
33 | Description. | |
34 | ------------ | |
35 | ||
36 | A data flow obeys TBF with rate R and depth B, if for any | |
37 | time interval t_i...t_f the number of transmitted bits | |
38 | does not exceed B + R*(t_f-t_i). | |
39 | ||
40 | Packetized version of this definition: | |
41 | The sequence of packets of sizes s_i served at moments t_i | |
42 | obeys TBF, if for any i<=k: | |
43 | ||
44 | s_i+....+s_k <= B + R*(t_k - t_i) | |
45 | ||
46 | Algorithm. | |
47 | ---------- | |
48 | ||
49 | Let N(t_i) be B/R initially and N(t) grow continuously with time as: | |
50 | ||
51 | N(t+delta) = min{B/R, N(t) + delta} | |
52 | ||
53 | If the first packet in queue has length S, it may be | |
54 | transmitted only at the time t_* when S/R <= N(t_*), | |
55 | and in this case N(t) jumps: | |
56 | ||
57 | N(t_* + 0) = N(t_* - 0) - S/R. | |
58 | ||
59 | ||
60 | ||
61 | Actually, QoS requires two TBF to be applied to a data stream. | |
62 | One of them controls steady state burst size, another | |
63 | one with rate P (peak rate) and depth M (equal to link MTU) | |
64 | limits bursts at a smaller time scale. | |
65 | ||
66 | It is easy to see that P>R, and B>M. If P is infinity, this double | |
67 | TBF is equivalent to a single one. | |
68 | ||
69 | When TBF works in reshaping mode, latency is estimated as: | |
70 | ||
71 | lat = max ((L-B)/R, (L-M)/P) | |
72 | ||
73 | ||
74 | NOTES. | |
75 | ------ | |
76 | ||
77 | If TBF throttles, it starts a watchdog timer, which will wake it up | |
78 | when it is ready to transmit. | |
79 | Note that the minimal timer resolution is 1/HZ. | |
80 | If no new packets arrive during this period, | |
81 | or if the device is not awaken by EOI for some previous packet, | |
82 | TBF can stop its activity for 1/HZ. | |
83 | ||
84 | ||
85 | This means, that with depth B, the maximal rate is | |
86 | ||
87 | R_crit = B*HZ | |
88 | ||
89 | F.e. for 10Mbit ethernet and HZ=100 the minimal allowed B is ~10Kbytes. | |
90 | ||
91 | Note that the peak rate TBF is much more tough: with MTU 1500 | |
92 | P_crit = 150Kbytes/sec. So, if you need greater peak | |
93 | rates, use alpha with HZ=1000 :-) | |
94 | ||
95 | With classful TBF, limit is just kept for backwards compatibility. | |
96 | It is passed to the default bfifo qdisc - if the inner qdisc is | |
97 | changed the limit is not effective anymore. | |
98 | */ | |
99 | ||
100 | struct tbf_sched_data | |
101 | { | |
102 | /* Parameters */ | |
103 | u32 limit; /* Maximal length of backlog: bytes */ | |
104 | u32 buffer; /* Token bucket depth/rate: MUST BE >= MTU/B */ | |
105 | u32 mtu; | |
106 | u32 max_size; | |
107 | struct qdisc_rate_table *R_tab; | |
108 | struct qdisc_rate_table *P_tab; | |
109 | ||
110 | /* Variables */ | |
111 | long tokens; /* Current number of B tokens */ | |
112 | long ptokens; /* Current number of P tokens */ | |
113 | psched_time_t t_c; /* Time check-point */ | |
1da177e4 | 114 | struct Qdisc *qdisc; /* Inner qdisc, default - bfifo queue */ |
f7f593e3 | 115 | struct qdisc_watchdog watchdog; /* Watchdog timer */ |
1da177e4 LT |
116 | }; |
117 | ||
e9bef55d JDB |
118 | #define L2T(q,L) qdisc_l2t((q)->R_tab,L) |
119 | #define L2T_P(q,L) qdisc_l2t((q)->P_tab,L) | |
1da177e4 LT |
120 | |
121 | static int tbf_enqueue(struct sk_buff *skb, struct Qdisc* sch) | |
122 | { | |
123 | struct tbf_sched_data *q = qdisc_priv(sch); | |
124 | int ret; | |
125 | ||
0abf77e5 | 126 | if (qdisc_pkt_len(skb) > q->max_size) { |
1da177e4 | 127 | sch->qstats.drops++; |
c3bc7cff | 128 | #ifdef CONFIG_NET_CLS_ACT |
1da177e4 LT |
129 | if (sch->reshape_fail == NULL || sch->reshape_fail(skb, sch)) |
130 | #endif | |
131 | kfree_skb(skb); | |
132 | ||
133 | return NET_XMIT_DROP; | |
134 | } | |
135 | ||
5f86173b JK |
136 | ret = qdisc_enqueue(skb, q->qdisc); |
137 | if (ret != 0) { | |
1da177e4 LT |
138 | sch->qstats.drops++; |
139 | return ret; | |
140 | } | |
141 | ||
142 | sch->q.qlen++; | |
0abf77e5 | 143 | sch->bstats.bytes += qdisc_pkt_len(skb); |
1da177e4 LT |
144 | sch->bstats.packets++; |
145 | return 0; | |
146 | } | |
147 | ||
148 | static int tbf_requeue(struct sk_buff *skb, struct Qdisc* sch) | |
149 | { | |
150 | struct tbf_sched_data *q = qdisc_priv(sch); | |
151 | int ret; | |
152 | ||
153 | if ((ret = q->qdisc->ops->requeue(skb, q->qdisc)) == 0) { | |
154 | sch->q.qlen++; | |
155 | sch->qstats.requeues++; | |
156 | } | |
157 | ||
158 | return ret; | |
159 | } | |
160 | ||
161 | static unsigned int tbf_drop(struct Qdisc* sch) | |
162 | { | |
163 | struct tbf_sched_data *q = qdisc_priv(sch); | |
6d037a26 | 164 | unsigned int len = 0; |
1da177e4 | 165 | |
6d037a26 | 166 | if (q->qdisc->ops->drop && (len = q->qdisc->ops->drop(q->qdisc)) != 0) { |
1da177e4 LT |
167 | sch->q.qlen--; |
168 | sch->qstats.drops++; | |
169 | } | |
170 | return len; | |
171 | } | |
172 | ||
1da177e4 LT |
173 | static struct sk_buff *tbf_dequeue(struct Qdisc* sch) |
174 | { | |
175 | struct tbf_sched_data *q = qdisc_priv(sch); | |
176 | struct sk_buff *skb; | |
177 | ||
178 | skb = q->qdisc->dequeue(q->qdisc); | |
179 | ||
180 | if (skb) { | |
181 | psched_time_t now; | |
f7f593e3 | 182 | long toks; |
1da177e4 | 183 | long ptoks = 0; |
0abf77e5 | 184 | unsigned int len = qdisc_pkt_len(skb); |
1da177e4 | 185 | |
3bebcda2 | 186 | now = psched_get_time(); |
03cc45c0 | 187 | toks = psched_tdiff_bounded(now, q->t_c, q->buffer); |
1da177e4 LT |
188 | |
189 | if (q->P_tab) { | |
190 | ptoks = toks + q->ptokens; | |
191 | if (ptoks > (long)q->mtu) | |
192 | ptoks = q->mtu; | |
193 | ptoks -= L2T_P(q, len); | |
194 | } | |
195 | toks += q->tokens; | |
196 | if (toks > (long)q->buffer) | |
197 | toks = q->buffer; | |
198 | toks -= L2T(q, len); | |
199 | ||
200 | if ((toks|ptoks) >= 0) { | |
201 | q->t_c = now; | |
202 | q->tokens = toks; | |
203 | q->ptokens = ptoks; | |
204 | sch->q.qlen--; | |
205 | sch->flags &= ~TCQ_F_THROTTLED; | |
206 | return skb; | |
207 | } | |
208 | ||
f7f593e3 PM |
209 | qdisc_watchdog_schedule(&q->watchdog, |
210 | now + max_t(long, -toks, -ptoks)); | |
1da177e4 LT |
211 | |
212 | /* Maybe we have a shorter packet in the queue, | |
213 | which can be sent now. It sounds cool, | |
214 | but, however, this is wrong in principle. | |
215 | We MUST NOT reorder packets under these circumstances. | |
216 | ||
217 | Really, if we split the flow into independent | |
218 | subflows, it would be a very good solution. | |
219 | This is the main idea of all FQ algorithms | |
220 | (cf. CSZ, HPFQ, HFSC) | |
221 | */ | |
222 | ||
223 | if (q->qdisc->ops->requeue(skb, q->qdisc) != NET_XMIT_SUCCESS) { | |
224 | /* When requeue fails skb is dropped */ | |
e488eafc | 225 | qdisc_tree_decrease_qlen(q->qdisc, 1); |
1da177e4 LT |
226 | sch->qstats.drops++; |
227 | } | |
228 | ||
1da177e4 LT |
229 | sch->qstats.overlimits++; |
230 | } | |
231 | return NULL; | |
232 | } | |
233 | ||
234 | static void tbf_reset(struct Qdisc* sch) | |
235 | { | |
236 | struct tbf_sched_data *q = qdisc_priv(sch); | |
237 | ||
238 | qdisc_reset(q->qdisc); | |
239 | sch->q.qlen = 0; | |
3bebcda2 | 240 | q->t_c = psched_get_time(); |
1da177e4 LT |
241 | q->tokens = q->buffer; |
242 | q->ptokens = q->mtu; | |
f7f593e3 | 243 | qdisc_watchdog_cancel(&q->watchdog); |
1da177e4 LT |
244 | } |
245 | ||
27a3421e PM |
246 | static const struct nla_policy tbf_policy[TCA_TBF_MAX + 1] = { |
247 | [TCA_TBF_PARMS] = { .len = sizeof(struct tc_tbf_qopt) }, | |
248 | [TCA_TBF_RTAB] = { .type = NLA_BINARY, .len = TC_RTAB_SIZE }, | |
249 | [TCA_TBF_PTAB] = { .type = NLA_BINARY, .len = TC_RTAB_SIZE }, | |
250 | }; | |
251 | ||
1e90474c | 252 | static int tbf_change(struct Qdisc* sch, struct nlattr *opt) |
1da177e4 | 253 | { |
cee63723 | 254 | int err; |
1da177e4 | 255 | struct tbf_sched_data *q = qdisc_priv(sch); |
1e90474c | 256 | struct nlattr *tb[TCA_TBF_PTAB + 1]; |
1da177e4 LT |
257 | struct tc_tbf_qopt *qopt; |
258 | struct qdisc_rate_table *rtab = NULL; | |
259 | struct qdisc_rate_table *ptab = NULL; | |
260 | struct Qdisc *child = NULL; | |
261 | int max_size,n; | |
262 | ||
27a3421e | 263 | err = nla_parse_nested(tb, TCA_TBF_PTAB, opt, tbf_policy); |
cee63723 PM |
264 | if (err < 0) |
265 | return err; | |
266 | ||
267 | err = -EINVAL; | |
27a3421e | 268 | if (tb[TCA_TBF_PARMS] == NULL) |
1da177e4 LT |
269 | goto done; |
270 | ||
1e90474c PM |
271 | qopt = nla_data(tb[TCA_TBF_PARMS]); |
272 | rtab = qdisc_get_rtab(&qopt->rate, tb[TCA_TBF_RTAB]); | |
1da177e4 LT |
273 | if (rtab == NULL) |
274 | goto done; | |
275 | ||
276 | if (qopt->peakrate.rate) { | |
277 | if (qopt->peakrate.rate > qopt->rate.rate) | |
1e90474c | 278 | ptab = qdisc_get_rtab(&qopt->peakrate, tb[TCA_TBF_PTAB]); |
1da177e4 LT |
279 | if (ptab == NULL) |
280 | goto done; | |
281 | } | |
282 | ||
283 | for (n = 0; n < 256; n++) | |
284 | if (rtab->data[n] > qopt->buffer) break; | |
285 | max_size = (n << qopt->rate.cell_log)-1; | |
286 | if (ptab) { | |
287 | int size; | |
288 | ||
289 | for (n = 0; n < 256; n++) | |
290 | if (ptab->data[n] > qopt->mtu) break; | |
291 | size = (n << qopt->peakrate.cell_log)-1; | |
292 | if (size < max_size) max_size = size; | |
293 | } | |
294 | if (max_size < 0) | |
295 | goto done; | |
296 | ||
053cfed7 | 297 | if (qopt->limit > 0) { |
fb0305ce PM |
298 | child = fifo_create_dflt(sch, &bfifo_qdisc_ops, qopt->limit); |
299 | if (IS_ERR(child)) { | |
300 | err = PTR_ERR(child); | |
1da177e4 | 301 | goto done; |
fb0305ce | 302 | } |
1da177e4 LT |
303 | } |
304 | ||
305 | sch_tree_lock(sch); | |
5e50da01 PM |
306 | if (child) { |
307 | qdisc_tree_decrease_qlen(q->qdisc, q->qdisc->q.qlen); | |
053cfed7 | 308 | qdisc_destroy(xchg(&q->qdisc, child)); |
5e50da01 | 309 | } |
1da177e4 LT |
310 | q->limit = qopt->limit; |
311 | q->mtu = qopt->mtu; | |
312 | q->max_size = max_size; | |
313 | q->buffer = qopt->buffer; | |
314 | q->tokens = q->buffer; | |
315 | q->ptokens = q->mtu; | |
316 | rtab = xchg(&q->R_tab, rtab); | |
317 | ptab = xchg(&q->P_tab, ptab); | |
318 | sch_tree_unlock(sch); | |
319 | err = 0; | |
320 | done: | |
321 | if (rtab) | |
322 | qdisc_put_rtab(rtab); | |
323 | if (ptab) | |
324 | qdisc_put_rtab(ptab); | |
325 | return err; | |
326 | } | |
327 | ||
1e90474c | 328 | static int tbf_init(struct Qdisc* sch, struct nlattr *opt) |
1da177e4 LT |
329 | { |
330 | struct tbf_sched_data *q = qdisc_priv(sch); | |
331 | ||
332 | if (opt == NULL) | |
333 | return -EINVAL; | |
334 | ||
3bebcda2 | 335 | q->t_c = psched_get_time(); |
f7f593e3 | 336 | qdisc_watchdog_init(&q->watchdog, sch); |
1da177e4 LT |
337 | q->qdisc = &noop_qdisc; |
338 | ||
339 | return tbf_change(sch, opt); | |
340 | } | |
341 | ||
342 | static void tbf_destroy(struct Qdisc *sch) | |
343 | { | |
344 | struct tbf_sched_data *q = qdisc_priv(sch); | |
345 | ||
f7f593e3 | 346 | qdisc_watchdog_cancel(&q->watchdog); |
1da177e4 LT |
347 | |
348 | if (q->P_tab) | |
349 | qdisc_put_rtab(q->P_tab); | |
350 | if (q->R_tab) | |
351 | qdisc_put_rtab(q->R_tab); | |
352 | ||
353 | qdisc_destroy(q->qdisc); | |
354 | } | |
355 | ||
356 | static int tbf_dump(struct Qdisc *sch, struct sk_buff *skb) | |
357 | { | |
358 | struct tbf_sched_data *q = qdisc_priv(sch); | |
4b3550ef | 359 | struct nlattr *nest; |
1da177e4 LT |
360 | struct tc_tbf_qopt opt; |
361 | ||
4b3550ef PM |
362 | nest = nla_nest_start(skb, TCA_OPTIONS); |
363 | if (nest == NULL) | |
364 | goto nla_put_failure; | |
1da177e4 LT |
365 | |
366 | opt.limit = q->limit; | |
367 | opt.rate = q->R_tab->rate; | |
368 | if (q->P_tab) | |
369 | opt.peakrate = q->P_tab->rate; | |
370 | else | |
371 | memset(&opt.peakrate, 0, sizeof(opt.peakrate)); | |
372 | opt.mtu = q->mtu; | |
373 | opt.buffer = q->buffer; | |
1e90474c | 374 | NLA_PUT(skb, TCA_TBF_PARMS, sizeof(opt), &opt); |
1da177e4 | 375 | |
4b3550ef | 376 | nla_nest_end(skb, nest); |
1da177e4 LT |
377 | return skb->len; |
378 | ||
1e90474c | 379 | nla_put_failure: |
4b3550ef | 380 | nla_nest_cancel(skb, nest); |
1da177e4 LT |
381 | return -1; |
382 | } | |
383 | ||
384 | static int tbf_dump_class(struct Qdisc *sch, unsigned long cl, | |
385 | struct sk_buff *skb, struct tcmsg *tcm) | |
386 | { | |
387 | struct tbf_sched_data *q = qdisc_priv(sch); | |
388 | ||
389 | if (cl != 1) /* only one class */ | |
390 | return -ENOENT; | |
391 | ||
392 | tcm->tcm_handle |= TC_H_MIN(1); | |
393 | tcm->tcm_info = q->qdisc->handle; | |
394 | ||
395 | return 0; | |
396 | } | |
397 | ||
398 | static int tbf_graft(struct Qdisc *sch, unsigned long arg, struct Qdisc *new, | |
399 | struct Qdisc **old) | |
400 | { | |
401 | struct tbf_sched_data *q = qdisc_priv(sch); | |
402 | ||
403 | if (new == NULL) | |
404 | new = &noop_qdisc; | |
405 | ||
406 | sch_tree_lock(sch); | |
407 | *old = xchg(&q->qdisc, new); | |
5e50da01 | 408 | qdisc_tree_decrease_qlen(*old, (*old)->q.qlen); |
1da177e4 | 409 | qdisc_reset(*old); |
1da177e4 LT |
410 | sch_tree_unlock(sch); |
411 | ||
412 | return 0; | |
413 | } | |
414 | ||
415 | static struct Qdisc *tbf_leaf(struct Qdisc *sch, unsigned long arg) | |
416 | { | |
417 | struct tbf_sched_data *q = qdisc_priv(sch); | |
418 | return q->qdisc; | |
419 | } | |
420 | ||
421 | static unsigned long tbf_get(struct Qdisc *sch, u32 classid) | |
422 | { | |
423 | return 1; | |
424 | } | |
425 | ||
426 | static void tbf_put(struct Qdisc *sch, unsigned long arg) | |
427 | { | |
428 | } | |
429 | ||
10297b99 | 430 | static int tbf_change_class(struct Qdisc *sch, u32 classid, u32 parentid, |
1e90474c | 431 | struct nlattr **tca, unsigned long *arg) |
1da177e4 LT |
432 | { |
433 | return -ENOSYS; | |
434 | } | |
435 | ||
436 | static int tbf_delete(struct Qdisc *sch, unsigned long arg) | |
437 | { | |
438 | return -ENOSYS; | |
439 | } | |
440 | ||
441 | static void tbf_walk(struct Qdisc *sch, struct qdisc_walker *walker) | |
442 | { | |
443 | if (!walker->stop) { | |
444 | if (walker->count >= walker->skip) | |
445 | if (walker->fn(sch, 1, walker) < 0) { | |
446 | walker->stop = 1; | |
447 | return; | |
448 | } | |
449 | walker->count++; | |
450 | } | |
451 | } | |
452 | ||
453 | static struct tcf_proto **tbf_find_tcf(struct Qdisc *sch, unsigned long cl) | |
454 | { | |
455 | return NULL; | |
456 | } | |
457 | ||
20fea08b | 458 | static const struct Qdisc_class_ops tbf_class_ops = |
1da177e4 LT |
459 | { |
460 | .graft = tbf_graft, | |
461 | .leaf = tbf_leaf, | |
462 | .get = tbf_get, | |
463 | .put = tbf_put, | |
464 | .change = tbf_change_class, | |
465 | .delete = tbf_delete, | |
466 | .walk = tbf_walk, | |
467 | .tcf_chain = tbf_find_tcf, | |
468 | .dump = tbf_dump_class, | |
469 | }; | |
470 | ||
20fea08b | 471 | static struct Qdisc_ops tbf_qdisc_ops __read_mostly = { |
1da177e4 LT |
472 | .next = NULL, |
473 | .cl_ops = &tbf_class_ops, | |
474 | .id = "tbf", | |
475 | .priv_size = sizeof(struct tbf_sched_data), | |
476 | .enqueue = tbf_enqueue, | |
477 | .dequeue = tbf_dequeue, | |
478 | .requeue = tbf_requeue, | |
479 | .drop = tbf_drop, | |
480 | .init = tbf_init, | |
481 | .reset = tbf_reset, | |
482 | .destroy = tbf_destroy, | |
483 | .change = tbf_change, | |
484 | .dump = tbf_dump, | |
485 | .owner = THIS_MODULE, | |
486 | }; | |
487 | ||
488 | static int __init tbf_module_init(void) | |
489 | { | |
490 | return register_qdisc(&tbf_qdisc_ops); | |
491 | } | |
492 | ||
493 | static void __exit tbf_module_exit(void) | |
494 | { | |
495 | unregister_qdisc(&tbf_qdisc_ops); | |
496 | } | |
497 | module_init(tbf_module_init) | |
498 | module_exit(tbf_module_exit) | |
499 | MODULE_LICENSE("GPL"); |