sch_tbf.c 11.5 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
/*
 * net/sched/sch_tbf.c	Token Bucket Filter queue.
 *
 *		This program is free software; you can redistribute it and/or
 *		modify it under the terms of the GNU General Public License
 *		as published by the Free Software Foundation; either version
 *		2 of the License, or (at your option) any later version.
 *
 * Authors:	Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
 *		Dmitry Torokhov <dtor@mail.ru> - allow attaching inner qdiscs -
 *						 original idea by Martin Devera
 *
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/errno.h>
#include <linux/skbuff.h>
21
#include <net/netlink.h>
L
Linus Torvalds 已提交
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
#include <net/pkt_sched.h>


/*	Simple Token Bucket Filter.
	=======================================

	SOURCE.
	-------

	None.

	Description.
	------------

	A data flow obeys TBF with rate R and depth B, if for any
	time interval t_i...t_f the number of transmitted bits
	does not exceed B + R*(t_f-t_i).

	Packetized version of this definition:
	The sequence of packets of sizes s_i served at moments t_i
	obeys TBF, if for any i<=k:

	s_i+....+s_k <= B + R*(t_k - t_i)

	Algorithm.
	----------

	Let N(t_i) be B/R initially and N(t) grow continuously with time as:

	N(t+delta) = min{B/R, N(t) + delta}

	If the first packet in queue has length S, it may be
	transmitted only at the time t_* when S/R <= N(t_*),
	and in this case N(t) jumps:

	N(t_* + 0) = N(t_* - 0) - S/R.



	Actually, QoS requires two TBF to be applied to a data stream.
	One of them controls steady state burst size, another
	one with rate P (peak rate) and depth M (equal to link MTU)
	limits bursts at a smaller time scale.

	It is easy to see that P>R, and B>M. If P is infinity, this double
	TBF is equivalent to a single one.

	When TBF works in reshaping mode, latency is estimated as:

	lat = max ((L-B)/R, (L-M)/P)


	NOTES.
	------

	If TBF throttles, it starts a watchdog timer, which will wake it up
	when it is ready to transmit.
	Note that the minimal timer resolution is 1/HZ.
	If no new packets arrive during this period,
	or if the device is not awaken by EOI for some previous packet,
	TBF can stop its activity for 1/HZ.


	This means, that with depth B, the maximal rate is

	R_crit = B*HZ

	F.e. for 10Mbit ethernet and HZ=100 the minimal allowed B is ~10Kbytes.

	Note that the peak rate TBF is much more tough: with MTU 1500
	P_crit = 150Kbytes/sec. So, if you need greater peak
	rates, use alpha with HZ=1000 :-)

	With classful TBF, limit is just kept for backwards compatibility.
	It is passed to the default bfifo qdisc - if the inner qdisc is
	changed the limit is not effective anymore.
*/

struct tbf_sched_data
{
/* Parameters */
	u32		limit;		/* Maximal length of backlog: bytes */
	u32		buffer;		/* Token bucket depth/rate: MUST BE >= MTU/B */
	u32		mtu;
	u32		max_size;
	struct qdisc_rate_table	*R_tab;
	struct qdisc_rate_table	*P_tab;

/* Variables */
	long	tokens;			/* Current number of B tokens */
	long	ptokens;		/* Current number of P tokens */
	psched_time_t	t_c;		/* Time check-point */
	struct Qdisc	*qdisc;		/* Inner qdisc, default - bfifo queue */
115
	struct qdisc_watchdog watchdog;	/* Watchdog timer */
L
Linus Torvalds 已提交
116 117
};

118 119
#define L2T(q,L)   qdisc_l2t((q)->R_tab,L)
#define L2T_P(q,L) qdisc_l2t((q)->P_tab,L)
L
Linus Torvalds 已提交
120 121 122 123 124 125 126 127

static int tbf_enqueue(struct sk_buff *skb, struct Qdisc* sch)
{
	struct tbf_sched_data *q = qdisc_priv(sch);
	int ret;

	if (skb->len > q->max_size) {
		sch->qstats.drops++;
128
#ifdef CONFIG_NET_CLS_ACT
L
Linus Torvalds 已提交
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
		if (sch->reshape_fail == NULL || sch->reshape_fail(skb, sch))
#endif
			kfree_skb(skb);

		return NET_XMIT_DROP;
	}

	if ((ret = q->qdisc->enqueue(skb, q->qdisc)) != 0) {
		sch->qstats.drops++;
		return ret;
	}

	sch->q.qlen++;
	sch->bstats.bytes += skb->len;
	sch->bstats.packets++;
	return 0;
}

static int tbf_requeue(struct sk_buff *skb, struct Qdisc* sch)
{
	struct tbf_sched_data *q = qdisc_priv(sch);
	int ret;

	if ((ret = q->qdisc->ops->requeue(skb, q->qdisc)) == 0) {
		sch->q.qlen++;
		sch->qstats.requeues++;
	}

	return ret;
}

static unsigned int tbf_drop(struct Qdisc* sch)
{
	struct tbf_sched_data *q = qdisc_priv(sch);
163
	unsigned int len = 0;
L
Linus Torvalds 已提交
164

165
	if (q->qdisc->ops->drop && (len = q->qdisc->ops->drop(q->qdisc)) != 0) {
L
Linus Torvalds 已提交
166 167 168 169 170 171 172 173 174 175 176 177 178 179 180
		sch->q.qlen--;
		sch->qstats.drops++;
	}
	return len;
}

static struct sk_buff *tbf_dequeue(struct Qdisc* sch)
{
	struct tbf_sched_data *q = qdisc_priv(sch);
	struct sk_buff *skb;

	skb = q->qdisc->dequeue(q->qdisc);

	if (skb) {
		psched_time_t now;
181
		long toks;
L
Linus Torvalds 已提交
182 183 184
		long ptoks = 0;
		unsigned int len = skb->len;

185
		now = psched_get_time();
186
		toks = psched_tdiff_bounded(now, q->t_c, q->buffer);
L
Linus Torvalds 已提交
187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207

		if (q->P_tab) {
			ptoks = toks + q->ptokens;
			if (ptoks > (long)q->mtu)
				ptoks = q->mtu;
			ptoks -= L2T_P(q, len);
		}
		toks += q->tokens;
		if (toks > (long)q->buffer)
			toks = q->buffer;
		toks -= L2T(q, len);

		if ((toks|ptoks) >= 0) {
			q->t_c = now;
			q->tokens = toks;
			q->ptokens = ptoks;
			sch->q.qlen--;
			sch->flags &= ~TCQ_F_THROTTLED;
			return skb;
		}

208 209
		qdisc_watchdog_schedule(&q->watchdog,
					now + max_t(long, -toks, -ptoks));
L
Linus Torvalds 已提交
210 211 212 213 214 215 216 217 218 219 220 221 222 223

		/* Maybe we have a shorter packet in the queue,
		   which can be sent now. It sounds cool,
		   but, however, this is wrong in principle.
		   We MUST NOT reorder packets under these circumstances.

		   Really, if we split the flow into independent
		   subflows, it would be a very good solution.
		   This is the main idea of all FQ algorithms
		   (cf. CSZ, HPFQ, HFSC)
		 */

		if (q->qdisc->ops->requeue(skb, q->qdisc) != NET_XMIT_SUCCESS) {
			/* When requeue fails skb is dropped */
224
			qdisc_tree_decrease_qlen(q->qdisc, 1);
L
Linus Torvalds 已提交
225 226 227 228 229 230 231 232 233 234 235 236 237 238
			sch->qstats.drops++;
		}

		sch->qstats.overlimits++;
	}
	return NULL;
}

static void tbf_reset(struct Qdisc* sch)
{
	struct tbf_sched_data *q = qdisc_priv(sch);

	qdisc_reset(q->qdisc);
	sch->q.qlen = 0;
239
	q->t_c = psched_get_time();
L
Linus Torvalds 已提交
240 241
	q->tokens = q->buffer;
	q->ptokens = q->mtu;
242
	qdisc_watchdog_cancel(&q->watchdog);
L
Linus Torvalds 已提交
243 244
}

245
static struct Qdisc *tbf_create_dflt_qdisc(struct Qdisc *sch, u32 limit)
L
Linus Torvalds 已提交
246
{
247
	struct Qdisc *q;
248
	struct nlattr *nla;
L
Linus Torvalds 已提交
249 250
	int ret;

251 252
	q = qdisc_create_dflt(sch->dev, &bfifo_qdisc_ops,
			      TC_H_MAKE(sch->handle, 1));
L
Linus Torvalds 已提交
253
	if (q) {
254 255 256 257 258 259
		nla = kmalloc(nla_attr_size(sizeof(struct tc_fifo_qopt)),
			      GFP_KERNEL);
		if (nla) {
			nla->nla_type = RTM_NEWQDISC;
			nla->nla_len = nla_attr_size(sizeof(struct tc_fifo_qopt));
			((struct tc_fifo_qopt *)nla_data(nla))->limit = limit;
L
Linus Torvalds 已提交
260

261 262
			ret = q->ops->change(q, nla);
			kfree(nla);
L
Linus Torvalds 已提交
263 264 265 266 267 268 269 270 271 272

			if (ret == 0)
				return q;
		}
		qdisc_destroy(q);
	}

	return NULL;
}

273
static int tbf_change(struct Qdisc* sch, struct nlattr *opt)
L
Linus Torvalds 已提交
274 275 276
{
	int err = -EINVAL;
	struct tbf_sched_data *q = qdisc_priv(sch);
277
	struct nlattr *tb[TCA_TBF_PTAB + 1];
L
Linus Torvalds 已提交
278 279 280 281 282 283
	struct tc_tbf_qopt *qopt;
	struct qdisc_rate_table *rtab = NULL;
	struct qdisc_rate_table *ptab = NULL;
	struct Qdisc *child = NULL;
	int max_size,n;

284 285 286
	if (nla_parse_nested(tb, TCA_TBF_PTAB, opt, NULL) ||
	    tb[TCA_TBF_PARMS] == NULL ||
	    nla_len(tb[TCA_TBF_PARMS]) < sizeof(*qopt))
L
Linus Torvalds 已提交
287 288
		goto done;

289 290
	qopt = nla_data(tb[TCA_TBF_PARMS]);
	rtab = qdisc_get_rtab(&qopt->rate, tb[TCA_TBF_RTAB]);
L
Linus Torvalds 已提交
291 292 293 294 295
	if (rtab == NULL)
		goto done;

	if (qopt->peakrate.rate) {
		if (qopt->peakrate.rate > qopt->rate.rate)
296
			ptab = qdisc_get_rtab(&qopt->peakrate, tb[TCA_TBF_PTAB]);
L
Linus Torvalds 已提交
297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314
		if (ptab == NULL)
			goto done;
	}

	for (n = 0; n < 256; n++)
		if (rtab->data[n] > qopt->buffer) break;
	max_size = (n << qopt->rate.cell_log)-1;
	if (ptab) {
		int size;

		for (n = 0; n < 256; n++)
			if (ptab->data[n] > qopt->mtu) break;
		size = (n << qopt->peakrate.cell_log)-1;
		if (size < max_size) max_size = size;
	}
	if (max_size < 0)
		goto done;

315
	if (qopt->limit > 0) {
316
		if ((child = tbf_create_dflt_qdisc(sch, qopt->limit)) == NULL)
L
Linus Torvalds 已提交
317 318 319 320
			goto done;
	}

	sch_tree_lock(sch);
321 322
	if (child) {
		qdisc_tree_decrease_qlen(q->qdisc, q->qdisc->q.qlen);
323
		qdisc_destroy(xchg(&q->qdisc, child));
324
	}
L
Linus Torvalds 已提交
325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342
	q->limit = qopt->limit;
	q->mtu = qopt->mtu;
	q->max_size = max_size;
	q->buffer = qopt->buffer;
	q->tokens = q->buffer;
	q->ptokens = q->mtu;
	rtab = xchg(&q->R_tab, rtab);
	ptab = xchg(&q->P_tab, ptab);
	sch_tree_unlock(sch);
	err = 0;
done:
	if (rtab)
		qdisc_put_rtab(rtab);
	if (ptab)
		qdisc_put_rtab(ptab);
	return err;
}

343
static int tbf_init(struct Qdisc* sch, struct nlattr *opt)
L
Linus Torvalds 已提交
344 345 346 347 348 349
{
	struct tbf_sched_data *q = qdisc_priv(sch);

	if (opt == NULL)
		return -EINVAL;

350
	q->t_c = psched_get_time();
351
	qdisc_watchdog_init(&q->watchdog, sch);
L
Linus Torvalds 已提交
352 353 354 355 356 357 358 359 360
	q->qdisc = &noop_qdisc;

	return tbf_change(sch, opt);
}

static void tbf_destroy(struct Qdisc *sch)
{
	struct tbf_sched_data *q = qdisc_priv(sch);

361
	qdisc_watchdog_cancel(&q->watchdog);
L
Linus Torvalds 已提交
362 363 364 365 366 367 368 369 370 371 372 373

	if (q->P_tab)
		qdisc_put_rtab(q->P_tab);
	if (q->R_tab)
		qdisc_put_rtab(q->R_tab);

	qdisc_destroy(q->qdisc);
}

static int tbf_dump(struct Qdisc *sch, struct sk_buff *skb)
{
	struct tbf_sched_data *q = qdisc_priv(sch);
374
	unsigned char *b = skb_tail_pointer(skb);
375
	struct nlattr *nla;
L
Linus Torvalds 已提交
376 377
	struct tc_tbf_qopt opt;

378 379
	nla = (struct nlattr*)b;
	NLA_PUT(skb, TCA_OPTIONS, 0, NULL);
L
Linus Torvalds 已提交
380 381 382 383 384 385 386 387 388

	opt.limit = q->limit;
	opt.rate = q->R_tab->rate;
	if (q->P_tab)
		opt.peakrate = q->P_tab->rate;
	else
		memset(&opt.peakrate, 0, sizeof(opt.peakrate));
	opt.mtu = q->mtu;
	opt.buffer = q->buffer;
389 390
	NLA_PUT(skb, TCA_TBF_PARMS, sizeof(opt), &opt);
	nla->nla_len = skb_tail_pointer(skb) - b;
L
Linus Torvalds 已提交
391 392 393

	return skb->len;

394
nla_put_failure:
395
	nlmsg_trim(skb, b);
L
Linus Torvalds 已提交
396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422
	return -1;
}

static int tbf_dump_class(struct Qdisc *sch, unsigned long cl,
			  struct sk_buff *skb, struct tcmsg *tcm)
{
	struct tbf_sched_data *q = qdisc_priv(sch);

	if (cl != 1) 	/* only one class */
		return -ENOENT;

	tcm->tcm_handle |= TC_H_MIN(1);
	tcm->tcm_info = q->qdisc->handle;

	return 0;
}

static int tbf_graft(struct Qdisc *sch, unsigned long arg, struct Qdisc *new,
		     struct Qdisc **old)
{
	struct tbf_sched_data *q = qdisc_priv(sch);

	if (new == NULL)
		new = &noop_qdisc;

	sch_tree_lock(sch);
	*old = xchg(&q->qdisc, new);
423
	qdisc_tree_decrease_qlen(*old, (*old)->q.qlen);
L
Linus Torvalds 已提交
424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444
	qdisc_reset(*old);
	sch_tree_unlock(sch);

	return 0;
}

static struct Qdisc *tbf_leaf(struct Qdisc *sch, unsigned long arg)
{
	struct tbf_sched_data *q = qdisc_priv(sch);
	return q->qdisc;
}

static unsigned long tbf_get(struct Qdisc *sch, u32 classid)
{
	return 1;
}

static void tbf_put(struct Qdisc *sch, unsigned long arg)
{
}

445
static int tbf_change_class(struct Qdisc *sch, u32 classid, u32 parentid,
446
			    struct nlattr **tca, unsigned long *arg)
L
Linus Torvalds 已提交
447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472
{
	return -ENOSYS;
}

static int tbf_delete(struct Qdisc *sch, unsigned long arg)
{
	return -ENOSYS;
}

static void tbf_walk(struct Qdisc *sch, struct qdisc_walker *walker)
{
	if (!walker->stop) {
		if (walker->count >= walker->skip)
			if (walker->fn(sch, 1, walker) < 0) {
				walker->stop = 1;
				return;
			}
		walker->count++;
	}
}

static struct tcf_proto **tbf_find_tcf(struct Qdisc *sch, unsigned long cl)
{
	return NULL;
}

473
static const struct Qdisc_class_ops tbf_class_ops =
L
Linus Torvalds 已提交
474 475 476 477 478 479 480 481 482 483 484 485
{
	.graft		=	tbf_graft,
	.leaf		=	tbf_leaf,
	.get		=	tbf_get,
	.put		=	tbf_put,
	.change		=	tbf_change_class,
	.delete		=	tbf_delete,
	.walk		=	tbf_walk,
	.tcf_chain	=	tbf_find_tcf,
	.dump		=	tbf_dump_class,
};

486
static struct Qdisc_ops tbf_qdisc_ops __read_mostly = {
L
Linus Torvalds 已提交
487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514
	.next		=	NULL,
	.cl_ops		=	&tbf_class_ops,
	.id		=	"tbf",
	.priv_size	=	sizeof(struct tbf_sched_data),
	.enqueue	=	tbf_enqueue,
	.dequeue	=	tbf_dequeue,
	.requeue	=	tbf_requeue,
	.drop		=	tbf_drop,
	.init		=	tbf_init,
	.reset		=	tbf_reset,
	.destroy	=	tbf_destroy,
	.change		=	tbf_change,
	.dump		=	tbf_dump,
	.owner		=	THIS_MODULE,
};

static int __init tbf_module_init(void)
{
	return register_qdisc(&tbf_qdisc_ops);
}

static void __exit tbf_module_exit(void)
{
	unregister_qdisc(&tbf_qdisc_ops);
}
module_init(tbf_module_init)
module_exit(tbf_module_exit)
MODULE_LICENSE("GPL");