sch_gred.c revision a73ed26bbae7327370c5bd298f07de78df9e3466 
1/*
2 * net/sched/sch_gred.c	Generic Random Early Detection queue.
3 *
4 *
5 *              This program is free software; you can redistribute it and/or
6 *              modify it under the terms of the GNU General Public License
7 *              as published by the Free Software Foundation; either version
8 *              2 of the License, or (at your option) any later version.
9 *
10 * Authors:    J Hadi Salim (hadi@cyberus.ca) 1998-2002
11 *
12 *             991129: -  Bug fix with grio mode
13 *		       - a better sing. AvgQ mode with Grio(WRED)
14 *		       - A finer grained VQ dequeue based on sugestion
15 *		         from Ren Liu
16 *		       - More error checks
17 *
18 *  For all the glorious comments look at include/net/red.h
19 */
20
21#include <linux/slab.h>
22#include <linux/module.h>
23#include <linux/types.h>
24#include <linux/kernel.h>
25#include <linux/skbuff.h>
26#include <net/pkt_sched.h>
27#include <net/red.h>
28
29#define GRED_DEF_PRIO (MAX_DPs / 2)
30#define GRED_VQ_MASK (MAX_DPs - 1)
31
32struct gred_sched_data;
33struct gred_sched;
34
35struct gred_sched_data {
36	u32		limit;		/* HARD maximal queue length	*/
37	u32		DP;		/* the drop parameters */
38	u32		bytesin;	/* bytes seen on virtualQ so far*/
39	u32		packetsin;	/* packets seen on virtualQ so far*/
40	u32		backlog;	/* bytes on the virtualQ */
41	u8		prio;		/* the prio of this vq */
42
43	struct red_parms parms;
44	struct red_stats stats;
45};
46
47enum {
48	GRED_WRED_MODE = 1,
49	GRED_RIO_MODE,
50};
51
52struct gred_sched {
53	struct gred_sched_data *tab[MAX_DPs];
54	unsigned long	flags;
55	u32		red_flags;
56	u32 		DPs;
57	u32 		def;
58	struct red_parms wred_set;
59};
60
61static inline int gred_wred_mode(struct gred_sched *table)
62{
63	return test_bit(GRED_WRED_MODE, &table->flags);
64}
65
66static inline void gred_enable_wred_mode(struct gred_sched *table)
67{
68	__set_bit(GRED_WRED_MODE, &table->flags);
69}
70
71static inline void gred_disable_wred_mode(struct gred_sched *table)
72{
73	__clear_bit(GRED_WRED_MODE, &table->flags);
74}
75
76static inline int gred_rio_mode(struct gred_sched *table)
77{
78	return test_bit(GRED_RIO_MODE, &table->flags);
79}
80
81static inline void gred_enable_rio_mode(struct gred_sched *table)
82{
83	__set_bit(GRED_RIO_MODE, &table->flags);
84}
85
86static inline void gred_disable_rio_mode(struct gred_sched *table)
87{
88	__clear_bit(GRED_RIO_MODE, &table->flags);
89}
90
91static inline int gred_wred_mode_check(struct Qdisc *sch)
92{
93	struct gred_sched *table = qdisc_priv(sch);
94	int i;
95
96	/* Really ugly O(n^2) but shouldn't be necessary too frequent. */
97	for (i = 0; i < table->DPs; i++) {
98		struct gred_sched_data *q = table->tab[i];
99		int n;
100
101		if (q == NULL)
102			continue;
103
104		for (n = 0; n < table->DPs; n++)
105			if (table->tab[n] && table->tab[n] != q &&
106			    table->tab[n]->prio == q->prio)
107				return 1;
108	}
109
110	return 0;
111}
112
113static inline unsigned int gred_backlog(struct gred_sched *table,
114					struct gred_sched_data *q,
115					struct Qdisc *sch)
116{
117	if (gred_wred_mode(table))
118		return sch->qstats.backlog;
119	else
120		return q->backlog;
121}
122
123static inline u16 tc_index_to_dp(struct sk_buff *skb)
124{
125	return skb->tc_index & GRED_VQ_MASK;
126}
127
128static inline void gred_load_wred_set(struct gred_sched *table,
129				      struct gred_sched_data *q)
130{
131	q->parms.qavg = table->wred_set.qavg;
132	q->parms.qidlestart = table->wred_set.qidlestart;
133}
134
135static inline void gred_store_wred_set(struct gred_sched *table,
136				       struct gred_sched_data *q)
137{
138	table->wred_set.qavg = q->parms.qavg;
139}
140
141static inline int gred_use_ecn(struct gred_sched *t)
142{
143	return t->red_flags & TC_RED_ECN;
144}
145
146static inline int gred_use_harddrop(struct gred_sched *t)
147{
148	return t->red_flags & TC_RED_HARDDROP;
149}
150
151static int gred_enqueue(struct sk_buff *skb, struct Qdisc *sch)
152{
153	struct gred_sched_data *q = NULL;
154	struct gred_sched *t = qdisc_priv(sch);
155	unsigned long qavg = 0;
156	u16 dp = tc_index_to_dp(skb);
157
158	if (dp >= t->DPs || (q = t->tab[dp]) == NULL) {
159		dp = t->def;
160
161		q = t->tab[dp];
162		if (!q) {
163			/* Pass through packets not assigned to a DP
164			 * if no default DP has been configured. This
165			 * allows for DP flows to be left untouched.
166			 */
167			if (skb_queue_len(&sch->q) < qdisc_dev(sch)->tx_queue_len)
168				return qdisc_enqueue_tail(skb, sch);
169			else
170				goto drop;
171		}
172
173		/* fix tc_index? --could be controvesial but needed for
174		   requeueing */
175		skb->tc_index = (skb->tc_index & ~GRED_VQ_MASK) | dp;
176	}
177
178	/* sum up all the qaves of prios <= to ours to get the new qave */
179	if (!gred_wred_mode(t) && gred_rio_mode(t)) {
180		int i;
181
182		for (i = 0; i < t->DPs; i++) {
183			if (t->tab[i] && t->tab[i]->prio < q->prio &&
184			    !red_is_idling(&t->tab[i]->parms))
185				qavg += t->tab[i]->parms.qavg;
186		}
187
188	}
189
190	q->packetsin++;
191	q->bytesin += qdisc_pkt_len(skb);
192
193	if (gred_wred_mode(t))
194		gred_load_wred_set(t, q);
195
196	q->parms.qavg = red_calc_qavg(&q->parms, gred_backlog(t, q, sch));
197
198	if (red_is_idling(&q->parms))
199		red_end_of_idle_period(&q->parms);
200
201	if (gred_wred_mode(t))
202		gred_store_wred_set(t, q);
203
204	switch (red_action(&q->parms, q->parms.qavg + qavg)) {
205	case RED_DONT_MARK:
206		break;
207
208	case RED_PROB_MARK:
209		sch->qstats.overlimits++;
210		if (!gred_use_ecn(t) || !INET_ECN_set_ce(skb)) {
211			q->stats.prob_drop++;
212			goto congestion_drop;
213		}
214
215		q->stats.prob_mark++;
216		break;
217
218	case RED_HARD_MARK:
219		sch->qstats.overlimits++;
220		if (gred_use_harddrop(t) || !gred_use_ecn(t) ||
221		    !INET_ECN_set_ce(skb)) {
222			q->stats.forced_drop++;
223			goto congestion_drop;
224		}
225		q->stats.forced_mark++;
226		break;
227	}
228
229	if (q->backlog + qdisc_pkt_len(skb) <= q->limit) {
230		q->backlog += qdisc_pkt_len(skb);
231		return qdisc_enqueue_tail(skb, sch);
232	}
233
234	q->stats.pdrop++;
235drop:
236	return qdisc_drop(skb, sch);
237
238congestion_drop:
239	qdisc_drop(skb, sch);
240	return NET_XMIT_CN;
241}
242
243static struct sk_buff *gred_dequeue(struct Qdisc *sch)
244{
245	struct sk_buff *skb;
246	struct gred_sched *t = qdisc_priv(sch);
247
248	skb = qdisc_dequeue_head(sch);
249
250	if (skb) {
251		struct gred_sched_data *q;
252		u16 dp = tc_index_to_dp(skb);
253
254		if (dp >= t->DPs || (q = t->tab[dp]) == NULL) {
255			if (net_ratelimit())
256				pr_warning("GRED: Unable to relocate VQ 0x%x "
257					   "after dequeue, screwing up "
258					   "backlog.\n", tc_index_to_dp(skb));
259		} else {
260			q->backlog -= qdisc_pkt_len(skb);
261
262			if (!q->backlog && !gred_wred_mode(t))
263				red_start_of_idle_period(&q->parms);
264		}
265
266		return skb;
267	}
268
269	if (gred_wred_mode(t) && !red_is_idling(&t->wred_set))
270		red_start_of_idle_period(&t->wred_set);
271
272	return NULL;
273}
274
275static unsigned int gred_drop(struct Qdisc *sch)
276{
277	struct sk_buff *skb;
278	struct gred_sched *t = qdisc_priv(sch);
279
280	skb = qdisc_dequeue_tail(sch);
281	if (skb) {
282		unsigned int len = qdisc_pkt_len(skb);
283		struct gred_sched_data *q;
284		u16 dp = tc_index_to_dp(skb);
285
286		if (dp >= t->DPs || (q = t->tab[dp]) == NULL) {
287			if (net_ratelimit())
288				pr_warning("GRED: Unable to relocate VQ 0x%x "
289					   "while dropping, screwing up "
290					   "backlog.\n", tc_index_to_dp(skb));
291		} else {
292			q->backlog -= len;
293			q->stats.other++;
294
295			if (!q->backlog && !gred_wred_mode(t))
296				red_start_of_idle_period(&q->parms);
297		}
298
299		qdisc_drop(skb, sch);
300		return len;
301	}
302
303	if (gred_wred_mode(t) && !red_is_idling(&t->wred_set))
304		red_start_of_idle_period(&t->wred_set);
305
306	return 0;
307
308}
309
310static void gred_reset(struct Qdisc *sch)
311{
312	int i;
313	struct gred_sched *t = qdisc_priv(sch);
314
315	qdisc_reset_queue(sch);
316
317	for (i = 0; i < t->DPs; i++) {
318		struct gred_sched_data *q = t->tab[i];
319
320		if (!q)
321			continue;
322
323		red_restart(&q->parms);
324		q->backlog = 0;
325	}
326}
327
328static inline void gred_destroy_vq(struct gred_sched_data *q)
329{
330	kfree(q);
331}
332
333static inline int gred_change_table_def(struct Qdisc *sch, struct nlattr *dps)
334{
335	struct gred_sched *table = qdisc_priv(sch);
336	struct tc_gred_sopt *sopt;
337	int i;
338
339	if (dps == NULL)
340		return -EINVAL;
341
342	sopt = nla_data(dps);
343
344	if (sopt->DPs > MAX_DPs || sopt->DPs == 0 || sopt->def_DP >= sopt->DPs)
345		return -EINVAL;
346
347	sch_tree_lock(sch);
348	table->DPs = sopt->DPs;
349	table->def = sopt->def_DP;
350	table->red_flags = sopt->flags;
351
352	/*
353	 * Every entry point to GRED is synchronized with the above code
354	 * and the DP is checked against DPs, i.e. shadowed VQs can no
355	 * longer be found so we can unlock right here.
356	 */
357	sch_tree_unlock(sch);
358
359	if (sopt->grio) {
360		gred_enable_rio_mode(table);
361		gred_disable_wred_mode(table);
362		if (gred_wred_mode_check(sch))
363			gred_enable_wred_mode(table);
364	} else {
365		gred_disable_rio_mode(table);
366		gred_disable_wred_mode(table);
367	}
368
369	for (i = table->DPs; i < MAX_DPs; i++) {
370		if (table->tab[i]) {
371			pr_warning("GRED: Warning: Destroying "
372				   "shadowed VQ 0x%x\n", i);
373			gred_destroy_vq(table->tab[i]);
374			table->tab[i] = NULL;
375		}
376	}
377
378	return 0;
379}
380
381static inline int gred_change_vq(struct Qdisc *sch, int dp,
382				 struct tc_gred_qopt *ctl, int prio,
383				 u8 *stab, u32 max_P)
384{
385	struct gred_sched *table = qdisc_priv(sch);
386	struct gred_sched_data *q;
387
388	if (table->tab[dp] == NULL) {
389		table->tab[dp] = kzalloc(sizeof(*q), GFP_KERNEL);
390		if (table->tab[dp] == NULL)
391			return -ENOMEM;
392	}
393
394	q = table->tab[dp];
395	q->DP = dp;
396	q->prio = prio;
397	q->limit = ctl->limit;
398
399	if (q->backlog == 0)
400		red_end_of_idle_period(&q->parms);
401
402	red_set_parms(&q->parms,
403		      ctl->qth_min, ctl->qth_max, ctl->Wlog, ctl->Plog,
404		      ctl->Scell_log, stab, max_P);
405
406	return 0;
407}
408
409static const struct nla_policy gred_policy[TCA_GRED_MAX + 1] = {
410	[TCA_GRED_PARMS]	= { .len = sizeof(struct tc_gred_qopt) },
411	[TCA_GRED_STAB]		= { .len = 256 },
412	[TCA_GRED_DPS]		= { .len = sizeof(struct tc_gred_sopt) },
413	[TCA_GRED_MAX_P]	= { .type = NLA_U32 },
414};
415
416static int gred_change(struct Qdisc *sch, struct nlattr *opt)
417{
418	struct gred_sched *table = qdisc_priv(sch);
419	struct tc_gred_qopt *ctl;
420	struct nlattr *tb[TCA_GRED_MAX + 1];
421	int err, prio = GRED_DEF_PRIO;
422	u8 *stab;
423	u32 max_P;
424
425	if (opt == NULL)
426		return -EINVAL;
427
428	err = nla_parse_nested(tb, TCA_GRED_MAX, opt, gred_policy);
429	if (err < 0)
430		return err;
431
432	if (tb[TCA_GRED_PARMS] == NULL && tb[TCA_GRED_STAB] == NULL)
433		return gred_change_table_def(sch, opt);
434
435	if (tb[TCA_GRED_PARMS] == NULL ||
436	    tb[TCA_GRED_STAB] == NULL)
437		return -EINVAL;
438
439	max_P = tb[TCA_GRED_MAX_P] ? nla_get_u32(tb[TCA_GRED_MAX_P]) : 0;
440
441	err = -EINVAL;
442	ctl = nla_data(tb[TCA_GRED_PARMS]);
443	stab = nla_data(tb[TCA_GRED_STAB]);
444
445	if (ctl->DP >= table->DPs)
446		goto errout;
447
448	if (gred_rio_mode(table)) {
449		if (ctl->prio == 0) {
450			int def_prio = GRED_DEF_PRIO;
451
452			if (table->tab[table->def])
453				def_prio = table->tab[table->def]->prio;
454
455			printk(KERN_DEBUG "GRED: DP %u does not have a prio "
456			       "setting default to %d\n", ctl->DP, def_prio);
457
458			prio = def_prio;
459		} else
460			prio = ctl->prio;
461	}
462
463	sch_tree_lock(sch);
464
465	err = gred_change_vq(sch, ctl->DP, ctl, prio, stab, max_P);
466	if (err < 0)
467		goto errout_locked;
468
469	if (gred_rio_mode(table)) {
470		gred_disable_wred_mode(table);
471		if (gred_wred_mode_check(sch))
472			gred_enable_wred_mode(table);
473	}
474
475	err = 0;
476
477errout_locked:
478	sch_tree_unlock(sch);
479errout:
480	return err;
481}
482
483static int gred_init(struct Qdisc *sch, struct nlattr *opt)
484{
485	struct nlattr *tb[TCA_GRED_MAX + 1];
486	int err;
487
488	if (opt == NULL)
489		return -EINVAL;
490
491	err = nla_parse_nested(tb, TCA_GRED_MAX, opt, gred_policy);
492	if (err < 0)
493		return err;
494
495	if (tb[TCA_GRED_PARMS] || tb[TCA_GRED_STAB])
496		return -EINVAL;
497
498	return gred_change_table_def(sch, tb[TCA_GRED_DPS]);
499}
500
501static int gred_dump(struct Qdisc *sch, struct sk_buff *skb)
502{
503	struct gred_sched *table = qdisc_priv(sch);
504	struct nlattr *parms, *opts = NULL;
505	int i;
506	u32 max_p[MAX_DPs];
507	struct tc_gred_sopt sopt = {
508		.DPs	= table->DPs,
509		.def_DP	= table->def,
510		.grio	= gred_rio_mode(table),
511		.flags	= table->red_flags,
512	};
513
514	opts = nla_nest_start(skb, TCA_OPTIONS);
515	if (opts == NULL)
516		goto nla_put_failure;
517	NLA_PUT(skb, TCA_GRED_DPS, sizeof(sopt), &sopt);
518
519	for (i = 0; i < MAX_DPs; i++) {
520		struct gred_sched_data *q = table->tab[i];
521
522		max_p[i] = q ? q->parms.max_P : 0;
523	}
524	NLA_PUT(skb, TCA_GRED_MAX_P, sizeof(max_p), max_p);
525
526	parms = nla_nest_start(skb, TCA_GRED_PARMS);
527	if (parms == NULL)
528		goto nla_put_failure;
529
530	for (i = 0; i < MAX_DPs; i++) {
531		struct gred_sched_data *q = table->tab[i];
532		struct tc_gred_qopt opt;
533
534		memset(&opt, 0, sizeof(opt));
535
536		if (!q) {
537			/* hack -- fix at some point with proper message
538			   This is how we indicate to tc that there is no VQ
539			   at this DP */
540
541			opt.DP = MAX_DPs + i;
542			goto append_opt;
543		}
544
545		opt.limit	= q->limit;
546		opt.DP		= q->DP;
547		opt.backlog	= q->backlog;
548		opt.prio	= q->prio;
549		opt.qth_min	= q->parms.qth_min >> q->parms.Wlog;
550		opt.qth_max	= q->parms.qth_max >> q->parms.Wlog;
551		opt.Wlog	= q->parms.Wlog;
552		opt.Plog	= q->parms.Plog;
553		opt.Scell_log	= q->parms.Scell_log;
554		opt.other	= q->stats.other;
555		opt.early	= q->stats.prob_drop;
556		opt.forced	= q->stats.forced_drop;
557		opt.pdrop	= q->stats.pdrop;
558		opt.packets	= q->packetsin;
559		opt.bytesin	= q->bytesin;
560
561		if (gred_wred_mode(table)) {
562			q->parms.qidlestart =
563				table->tab[table->def]->parms.qidlestart;
564			q->parms.qavg = table->tab[table->def]->parms.qavg;
565		}
566
567		opt.qave = red_calc_qavg(&q->parms, q->parms.qavg);
568
569append_opt:
570		if (nla_append(skb, sizeof(opt), &opt) < 0)
571			goto nla_put_failure;
572	}
573
574	nla_nest_end(skb, parms);
575
576	return nla_nest_end(skb, opts);
577
578nla_put_failure:
579	nla_nest_cancel(skb, opts);
580	return -EMSGSIZE;
581}
582
583static void gred_destroy(struct Qdisc *sch)
584{
585	struct gred_sched *table = qdisc_priv(sch);
586	int i;
587
588	for (i = 0; i < table->DPs; i++) {
589		if (table->tab[i])
590			gred_destroy_vq(table->tab[i]);
591	}
592}
593
594static struct Qdisc_ops gred_qdisc_ops __read_mostly = {
595	.id		=	"gred",
596	.priv_size	=	sizeof(struct gred_sched),
597	.enqueue	=	gred_enqueue,
598	.dequeue	=	gred_dequeue,
599	.peek		=	qdisc_peek_head,
600	.drop		=	gred_drop,
601	.init		=	gred_init,
602	.reset		=	gred_reset,
603	.destroy	=	gred_destroy,
604	.change		=	gred_change,
605	.dump		=	gred_dump,
606	.owner		=	THIS_MODULE,
607};
608
609static int __init gred_module_init(void)
610{
611	return register_qdisc(&gred_qdisc_ops);
612}
613
614static void __exit gred_module_exit(void)
615{
616	unregister_qdisc(&gred_qdisc_ops);
617}
618
619module_init(gred_module_init)
620module_exit(gred_module_exit)
621
622MODULE_LICENSE("GPL");
623