sch_prio.c 8.3 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9
/*
 * net/sched/sch_prio.c	Simple 3-band priority "scheduler".
 *
 *		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>
10
 * Fixes:       19990609: J Hadi Salim <hadi@nortelnetworks.com>:
L
Linus Torvalds 已提交
11 12 13 14
 *              Init --  EINVAL when opt undefined
 */

#include <linux/module.h>
15
#include <linux/slab.h>
L
Linus Torvalds 已提交
16 17 18 19 20
#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
#include <net/pkt_sched.h>


struct prio_sched_data
{
	int bands;
	struct tcf_proto *filter_list;
	u8  prio2band[TC_PRIO_MAX+1];
	struct Qdisc *queues[TCQ_PRIO_BANDS];
};


static struct Qdisc *
prio_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr)
{
	struct prio_sched_data *q = qdisc_priv(sch);
	u32 band = skb->priority;
	struct tcf_result res;
40
	int err;
L
Linus Torvalds 已提交
41

42
	*qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
L
Linus Torvalds 已提交
43
	if (TC_H_MAJ(skb->priority) != sch->handle) {
44
		err = tc_classify(skb, q->filter_list, &res);
L
Linus Torvalds 已提交
45
#ifdef CONFIG_NET_CLS_ACT
46
		switch (err) {
L
Linus Torvalds 已提交
47 48
		case TC_ACT_STOLEN:
		case TC_ACT_QUEUED:
49
			*qerr = NET_XMIT_SUCCESS | __NET_XMIT_STOLEN;
L
Linus Torvalds 已提交
50 51
		case TC_ACT_SHOT:
			return NULL;
52
		}
L
Linus Torvalds 已提交
53
#endif
54
		if (!q->filter_list || err < 0) {
L
Linus Torvalds 已提交
55 56
			if (TC_H_MAJ(band))
				band = 0;
57
			return q->queues[q->prio2band[band&TC_PRIO_MAX]];
L
Linus Torvalds 已提交
58 59 60 61
		}
		band = res.classid;
	}
	band = TC_H_MIN(band) - 1;
62
	if (band >= q->bands)
63 64
		return q->queues[q->prio2band[0]];

L
Linus Torvalds 已提交
65 66 67 68 69 70 71 72 73 74 75 76
	return q->queues[band];
}

static int
prio_enqueue(struct sk_buff *skb, struct Qdisc *sch)
{
	struct Qdisc *qdisc;
	int ret;

	qdisc = prio_classify(skb, sch, &ret);
#ifdef CONFIG_NET_CLS_ACT
	if (qdisc == NULL) {
77

78
		if (ret & __NET_XMIT_BYPASS)
L
Linus Torvalds 已提交
79 80 81 82 83 84
			sch->qstats.drops++;
		kfree_skb(skb);
		return ret;
	}
#endif

85 86
	ret = qdisc_enqueue(skb, qdisc);
	if (ret == NET_XMIT_SUCCESS) {
87
		qdisc_bstats_update(sch, skb);
L
Linus Torvalds 已提交
88 89 90
		sch->q.qlen++;
		return NET_XMIT_SUCCESS;
	}
91 92
	if (net_xmit_drop_count(ret))
		sch->qstats.drops++;
93
	return ret;
L
Linus Torvalds 已提交
94 95
}

96 97 98 99 100 101 102 103 104 105 106 107 108
static struct sk_buff *prio_peek(struct Qdisc *sch)
{
	struct prio_sched_data *q = qdisc_priv(sch);
	int prio;

	for (prio = 0; prio < q->bands; prio++) {
		struct Qdisc *qdisc = q->queues[prio];
		struct sk_buff *skb = qdisc->ops->peek(qdisc);
		if (skb)
			return skb;
	}
	return NULL;
}
L
Linus Torvalds 已提交
109

110
static struct sk_buff *prio_dequeue(struct Qdisc* sch)
L
Linus Torvalds 已提交
111 112 113 114 115
{
	struct prio_sched_data *q = qdisc_priv(sch);
	int prio;

	for (prio = 0; prio < q->bands; prio++) {
116 117 118 119 120
		struct Qdisc *qdisc = q->queues[prio];
		struct sk_buff *skb = qdisc->dequeue(qdisc);
		if (skb) {
			sch->q.qlen--;
			return skb;
L
Linus Torvalds 已提交
121 122 123 124 125 126 127 128 129 130 131 132 133 134 135
		}
	}
	return NULL;

}

static unsigned int prio_drop(struct Qdisc* sch)
{
	struct prio_sched_data *q = qdisc_priv(sch);
	int prio;
	unsigned int len;
	struct Qdisc *qdisc;

	for (prio = q->bands-1; prio >= 0; prio--) {
		qdisc = q->queues[prio];
136
		if (qdisc->ops->drop && (len = qdisc->ops->drop(qdisc)) != 0) {
L
Linus Torvalds 已提交
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
			sch->q.qlen--;
			return len;
		}
	}
	return 0;
}


static void
prio_reset(struct Qdisc* sch)
{
	int prio;
	struct prio_sched_data *q = qdisc_priv(sch);

	for (prio=0; prio<q->bands; prio++)
		qdisc_reset(q->queues[prio]);
	sch->q.qlen = 0;
}

static void
prio_destroy(struct Qdisc* sch)
{
	int prio;
	struct prio_sched_data *q = qdisc_priv(sch);

162
	tcf_destroy_chain(&q->filter_list);
L
Linus Torvalds 已提交
163 164 165 166
	for (prio=0; prio<q->bands; prio++)
		qdisc_destroy(q->queues[prio]);
}

167
static int prio_tune(struct Qdisc *sch, struct nlattr *opt)
L
Linus Torvalds 已提交
168 169
{
	struct prio_sched_data *q = qdisc_priv(sch);
170
	struct tc_prio_qopt *qopt;
L
Linus Torvalds 已提交
171 172
	int i;

173 174 175
	if (nla_len(opt) < sizeof(*qopt))
		return -EINVAL;
	qopt = nla_data(opt);
176

177
	if (qopt->bands > TCQ_PRIO_BANDS || qopt->bands < 2)
L
Linus Torvalds 已提交
178 179 180
		return -EINVAL;

	for (i=0; i<=TC_PRIO_MAX; i++) {
181
		if (qopt->priomap[i] >= qopt->bands)
L
Linus Torvalds 已提交
182 183 184 185
			return -EINVAL;
	}

	sch_tree_lock(sch);
186
	q->bands = qopt->bands;
L
Linus Torvalds 已提交
187 188 189
	memcpy(q->prio2band, qopt->priomap, TC_PRIO_MAX+1);

	for (i=q->bands; i<TCQ_PRIO_BANDS; i++) {
190 191
		struct Qdisc *child = q->queues[i];
		q->queues[i] = &noop_qdisc;
192 193
		if (child != &noop_qdisc) {
			qdisc_tree_decrease_qlen(child, child->q.qlen);
L
Linus Torvalds 已提交
194
			qdisc_destroy(child);
195
		}
L
Linus Torvalds 已提交
196 197 198
	}
	sch_tree_unlock(sch);

199 200
	for (i=0; i<q->bands; i++) {
		if (q->queues[i] == &noop_qdisc) {
201
			struct Qdisc *child, *old;
202
			child = qdisc_create_dflt(sch->dev_queue,
203
						  &pfifo_qdisc_ops,
204
						  TC_H_MAKE(sch->handle, i + 1));
L
Linus Torvalds 已提交
205 206
			if (child) {
				sch_tree_lock(sch);
207 208
				old = q->queues[i];
				q->queues[i] = child;
L
Linus Torvalds 已提交
209

210 211 212 213
				if (old != &noop_qdisc) {
					qdisc_tree_decrease_qlen(old,
								 old->q.qlen);
					qdisc_destroy(old);
214
				}
L
Linus Torvalds 已提交
215 216 217 218 219 220 221
				sch_tree_unlock(sch);
			}
		}
	}
	return 0;
}

222
static int prio_init(struct Qdisc *sch, struct nlattr *opt)
L
Linus Torvalds 已提交
223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243
{
	struct prio_sched_data *q = qdisc_priv(sch);
	int i;

	for (i=0; i<TCQ_PRIO_BANDS; i++)
		q->queues[i] = &noop_qdisc;

	if (opt == NULL) {
		return -EINVAL;
	} else {
		int err;

		if ((err= prio_tune(sch, opt)) != 0)
			return err;
	}
	return 0;
}

static int prio_dump(struct Qdisc *sch, struct sk_buff *skb)
{
	struct prio_sched_data *q = qdisc_priv(sch);
244
	unsigned char *b = skb_tail_pointer(skb);
L
Linus Torvalds 已提交
245 246 247 248
	struct tc_prio_qopt opt;

	opt.bands = q->bands;
	memcpy(&opt.priomap, q->prio2band, TC_PRIO_MAX+1);
249

250
	NLA_PUT(skb, TCA_OPTIONS, sizeof(opt), &opt);
251

L
Linus Torvalds 已提交
252 253
	return skb->len;

254
nla_put_failure:
255
	nlmsg_trim(skb, b);
L
Linus Torvalds 已提交
256 257 258 259 260 261 262 263 264 265 266 267 268 269 270
	return -1;
}

static int prio_graft(struct Qdisc *sch, unsigned long arg, struct Qdisc *new,
		      struct Qdisc **old)
{
	struct prio_sched_data *q = qdisc_priv(sch);
	unsigned long band = arg - 1;

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

	sch_tree_lock(sch);
	*old = q->queues[band];
	q->queues[band] = new;
271
	qdisc_tree_decrease_qlen(*old, (*old)->q.qlen);
L
Linus Torvalds 已提交
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
	qdisc_reset(*old);
	sch_tree_unlock(sch);

	return 0;
}

static struct Qdisc *
prio_leaf(struct Qdisc *sch, unsigned long arg)
{
	struct prio_sched_data *q = qdisc_priv(sch);
	unsigned long band = arg - 1;

	return q->queues[band];
}

static unsigned long prio_get(struct Qdisc *sch, u32 classid)
{
	struct prio_sched_data *q = qdisc_priv(sch);
	unsigned long band = TC_H_MIN(classid);

	if (band - 1 >= q->bands)
		return 0;
	return band;
}

static unsigned long prio_bind(struct Qdisc *sch, unsigned long parent, u32 classid)
{
	return prio_get(sch, classid);
}


static void prio_put(struct Qdisc *q, unsigned long cl)
{
}

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

	tcm->tcm_handle |= TC_H_MIN(cl);
313
	tcm->tcm_info = q->queues[cl-1]->handle;
L
Linus Torvalds 已提交
314 315 316
	return 0;
}

317 318 319 320 321 322 323
static int prio_dump_class_stats(struct Qdisc *sch, unsigned long cl,
				 struct gnet_dump *d)
{
	struct prio_sched_data *q = qdisc_priv(sch);
	struct Qdisc *cl_q;

	cl_q = q->queues[cl - 1];
324
	cl_q->qstats.qlen = cl_q->q.qlen;
325 326 327 328 329 330 331
	if (gnet_stats_copy_basic(d, &cl_q->bstats) < 0 ||
	    gnet_stats_copy_queue(d, &cl_q->qstats) < 0)
		return -1;

	return 0;
}

L
Linus Torvalds 已提交
332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361
static void prio_walk(struct Qdisc *sch, struct qdisc_walker *arg)
{
	struct prio_sched_data *q = qdisc_priv(sch);
	int prio;

	if (arg->stop)
		return;

	for (prio = 0; prio < q->bands; prio++) {
		if (arg->count < arg->skip) {
			arg->count++;
			continue;
		}
		if (arg->fn(sch, prio+1, arg) < 0) {
			arg->stop = 1;
			break;
		}
		arg->count++;
	}
}

static struct tcf_proto ** prio_find_tcf(struct Qdisc *sch, unsigned long cl)
{
	struct prio_sched_data *q = qdisc_priv(sch);

	if (cl)
		return NULL;
	return &q->filter_list;
}

362
static const struct Qdisc_class_ops prio_class_ops = {
L
Linus Torvalds 已提交
363 364 365 366 367 368 369 370 371
	.graft		=	prio_graft,
	.leaf		=	prio_leaf,
	.get		=	prio_get,
	.put		=	prio_put,
	.walk		=	prio_walk,
	.tcf_chain	=	prio_find_tcf,
	.bind_tcf	=	prio_bind,
	.unbind_tcf	=	prio_put,
	.dump		=	prio_dump_class,
372
	.dump_stats	=	prio_dump_class_stats,
L
Linus Torvalds 已提交
373 374
};

375
static struct Qdisc_ops prio_qdisc_ops __read_mostly = {
L
Linus Torvalds 已提交
376 377 378 379 380 381
	.next		=	NULL,
	.cl_ops		=	&prio_class_ops,
	.id		=	"prio",
	.priv_size	=	sizeof(struct prio_sched_data),
	.enqueue	=	prio_enqueue,
	.dequeue	=	prio_dequeue,
382
	.peek		=	prio_peek,
L
Linus Torvalds 已提交
383 384 385 386 387 388 389 390 391 392 393
	.drop		=	prio_drop,
	.init		=	prio_init,
	.reset		=	prio_reset,
	.destroy	=	prio_destroy,
	.change		=	prio_tune,
	.dump		=	prio_dump,
	.owner		=	THIS_MODULE,
};

static int __init prio_module_init(void)
{
394
	return register_qdisc(&prio_qdisc_ops);
L
Linus Torvalds 已提交
395 396
}

397
static void __exit prio_module_exit(void)
L
Linus Torvalds 已提交
398 399 400 401 402 403 404 405
{
	unregister_qdisc(&prio_qdisc_ops);
}

module_init(prio_module_init)
module_exit(prio_module_exit)

MODULE_LICENSE("GPL");