[linux-2.6-block.git] / net / sched / sch_prio.c

/*
 * 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>
 * Fixes:       19990609: J Hadi Salim <hadi@nortelnetworks.com>:
 *              Init --  EINVAL when opt undefined
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/errno.h>
#include <linux/skbuff.h>
#include <net/netlink.h>
#include <net/pkt_sched.h>


struct prio_sched_data
{
	int bands;
	int curband; /* for round-robin */
	struct tcf_proto *filter_list;
	u8  prio2band[TC_PRIO_MAX+1];
	struct Qdisc *queues[TCQ_PRIO_BANDS];
	int mq;
};


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;

	*qerr = NET_XMIT_BYPASS;
	if (TC_H_MAJ(skb->priority) != sch->handle) {
#ifdef CONFIG_NET_CLS_ACT
		switch (tc_classify(skb, q->filter_list, &res)) {
		case TC_ACT_STOLEN:
		case TC_ACT_QUEUED:
			*qerr = NET_XMIT_SUCCESS;
		case TC_ACT_SHOT:
			return NULL;
		}

		if (!q->filter_list ) {
#else
		if (!q->filter_list || tc_classify(skb, q->filter_list, &res)) {
#endif
			if (TC_H_MAJ(band))
				band = 0;
			band = q->prio2band[band&TC_PRIO_MAX];
			goto out;
		}
		band = res.classid;
	}
	band = TC_H_MIN(band) - 1;
	if (band >= q->bands)
		band = q->prio2band[0];
out:
	if (q->mq)
		skb_set_queue_mapping(skb, band);
	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) {

		if (ret == NET_XMIT_BYPASS)
			sch->qstats.drops++;
		kfree_skb(skb);
		return ret;
	}
#endif

	if ((ret = qdisc->enqueue(skb, qdisc)) == NET_XMIT_SUCCESS) {
		sch->bstats.bytes += skb->len;
		sch->bstats.packets++;
		sch->q.qlen++;
		return NET_XMIT_SUCCESS;
	}
	sch->qstats.drops++;
	return ret;
}


static int
prio_requeue(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) {
		if (ret == NET_XMIT_BYPASS)
			sch->qstats.drops++;
		kfree_skb(skb);
		return ret;
	}
#endif

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


static struct sk_buff *
prio_dequeue(struct Qdisc* sch)
{
	struct sk_buff *skb;
	struct prio_sched_data *q = qdisc_priv(sch);
	int prio;
	struct Qdisc *qdisc;

	for (prio = 0; prio < q->bands; prio++) {
		/* Check if the target subqueue is available before
		 * pulling an skb.  This way we avoid excessive requeues
		 * for slower queues.
		 */
		if (!netif_subqueue_stopped(sch->dev, (q->mq ? prio : 0))) {
			qdisc = q->queues[prio];
			skb = qdisc->dequeue(qdisc);
			if (skb) {
				sch->q.qlen--;
				return skb;
			}
		}
	}
	return NULL;

}

static struct sk_buff *rr_dequeue(struct Qdisc* sch)
{
	struct sk_buff *skb;
	struct prio_sched_data *q = qdisc_priv(sch);
	struct Qdisc *qdisc;
	int bandcount;

	/* Only take one pass through the queues.  If nothing is available,
	 * return nothing.
	 */
	for (bandcount = 0; bandcount < q->bands; bandcount++) {
		/* Check if the target subqueue is available before
		 * pulling an skb.  This way we avoid excessive requeues
		 * for slower queues.  If the queue is stopped, try the
		 * next queue.
		 */
		if (!netif_subqueue_stopped(sch->dev,
					    (q->mq ? q->curband : 0))) {
			qdisc = q->queues[q->curband];
			skb = qdisc->dequeue(qdisc);
			if (skb) {
				sch->q.qlen--;
				q->curband++;
				if (q->curband >= q->bands)
					q->curband = 0;
				return skb;
			}
		}
		q->curband++;
		if (q->curband >= q->bands)
			q->curband = 0;
	}
	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];
		if (qdisc->ops->drop && (len = qdisc->ops->drop(qdisc)) != 0) {
			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);

	tcf_destroy_chain(q->filter_list);
	for (prio=0; prio<q->bands; prio++)
		qdisc_destroy(q->queues[prio]);
}

static int prio_tune(struct Qdisc *sch, struct rtattr *opt)
{
	struct prio_sched_data *q = qdisc_priv(sch);
	struct tc_prio_qopt *qopt;
	struct rtattr *tb[TCA_PRIO_MAX];
	int i;

	if (rtattr_parse_nested_compat(tb, TCA_PRIO_MAX, opt, qopt,
				       sizeof(*qopt)))
		return -EINVAL;
	q->bands = qopt->bands;
	/* If we're multiqueue, make sure the number of incoming bands
	 * matches the number of queues on the device we're associating with.
	 * If the number of bands requested is zero, then set q->bands to
	 * dev->egress_subqueue_count.
	 */
	q->mq = RTA_GET_FLAG(tb[TCA_PRIO_MQ - 1]);
	if (q->mq) {
		if (sch->handle != TC_H_ROOT)
			return -EINVAL;
		if (netif_is_multiqueue(sch->dev)) {
			if (q->bands == 0)
				q->bands = sch->dev->egress_subqueue_count;
			else if (q->bands != sch->dev->egress_subqueue_count)
				return -EINVAL;
		} else
			return -EOPNOTSUPP;
	}

	if (q->bands > TCQ_PRIO_BANDS || q->bands < 2)
		return -EINVAL;

	for (i=0; i<=TC_PRIO_MAX; i++) {
		if (qopt->priomap[i] >= q->bands)
			return -EINVAL;
	}

	sch_tree_lock(sch);
	memcpy(q->prio2band, qopt->priomap, TC_PRIO_MAX+1);

	for (i=q->bands; i<TCQ_PRIO_BANDS; i++) {
		struct Qdisc *child = xchg(&q->queues[i], &noop_qdisc);
		if (child != &noop_qdisc) {
			qdisc_tree_decrease_qlen(child, child->q.qlen);
			qdisc_destroy(child);
		}
	}
	sch_tree_unlock(sch);

	for (i=0; i<q->bands; i++) {
		if (q->queues[i] == &noop_qdisc) {
			struct Qdisc *child;
			child = qdisc_create_dflt(sch->dev, &pfifo_qdisc_ops,
						  TC_H_MAKE(sch->handle, i + 1));
			if (child) {
				sch_tree_lock(sch);
				child = xchg(&q->queues[i], child);

				if (child != &noop_qdisc) {
					qdisc_tree_decrease_qlen(child,
								 child->q.qlen);
					qdisc_destroy(child);
				}
				sch_tree_unlock(sch);
			}
		}
	}
	return 0;
}

static int prio_init(struct Qdisc *sch, struct rtattr *opt)
{
	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);
	unsigned char *b = skb_tail_pointer(skb);
	struct rtattr *nest;
	struct tc_prio_qopt opt;

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

	nest = RTA_NEST_COMPAT(skb, TCA_OPTIONS, sizeof(opt), &opt);
	if (q->mq)
		RTA_PUT_FLAG(skb, TCA_PRIO_MQ);
	RTA_NEST_COMPAT_END(skb, nest);

	return skb->len;

rtattr_failure:
	nlmsg_trim(skb, b);
	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 (band >= q->bands)
		return -EINVAL;

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

	sch_tree_lock(sch);
	*old = q->queues[band];
	q->queues[band] = new;
	qdisc_tree_decrease_qlen(*old, (*old)->q.qlen);
	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;

	if (band >= q->bands)
		return NULL;

	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)
{
	return;
}

static int prio_change(struct Qdisc *sch, u32 handle, u32 parent, struct rtattr **tca, unsigned long *arg)
{
	unsigned long cl = *arg;
	struct prio_sched_data *q = qdisc_priv(sch);

	if (cl - 1 > q->bands)
		return -ENOENT;
	return 0;
}

static int prio_delete(struct Qdisc *sch, unsigned long cl)
{
	struct prio_sched_data *q = qdisc_priv(sch);
	if (cl - 1 > q->bands)
		return -ENOENT;
	return 0;
}


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);

	if (cl - 1 > q->bands)
		return -ENOENT;
	tcm->tcm_handle |= TC_H_MIN(cl);
	if (q->queues[cl-1])
		tcm->tcm_info = q->queues[cl-1]->handle;
	return 0;
}

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];
	if (gnet_stats_copy_basic(d, &cl_q->bstats) < 0 ||
	    gnet_stats_copy_queue(d, &cl_q->qstats) < 0)
		return -1;

	return 0;
}

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;
}

static struct Qdisc_class_ops prio_class_ops = {
	.graft		=	prio_graft,
	.leaf		=	prio_leaf,
	.get		=	prio_get,
	.put		=	prio_put,
	.change		=	prio_change,
	.delete		=	prio_delete,
	.walk		=	prio_walk,
	.tcf_chain	=	prio_find_tcf,
	.bind_tcf	=	prio_bind,
	.unbind_tcf	=	prio_put,
	.dump		=	prio_dump_class,
	.dump_stats	=	prio_dump_class_stats,
};

static struct Qdisc_ops prio_qdisc_ops = {
	.next		=	NULL,
	.cl_ops		=	&prio_class_ops,
	.id		=	"prio",
	.priv_size	=	sizeof(struct prio_sched_data),
	.enqueue	=	prio_enqueue,
	.dequeue	=	prio_dequeue,
	.requeue	=	prio_requeue,
	.drop		=	prio_drop,
	.init		=	prio_init,
	.reset		=	prio_reset,
	.destroy	=	prio_destroy,
	.change		=	prio_tune,
	.dump		=	prio_dump,
	.owner		=	THIS_MODULE,
};

static struct Qdisc_ops rr_qdisc_ops = {
	.next		=	NULL,
	.cl_ops		=	&prio_class_ops,
	.id		=	"rr",
	.priv_size	=	sizeof(struct prio_sched_data),
	.enqueue	=	prio_enqueue,
	.dequeue	=	rr_dequeue,
	.requeue	=	prio_requeue,
	.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)
{
	int err;

	err = register_qdisc(&prio_qdisc_ops);
	if (err < 0)
		return err;
	err = register_qdisc(&rr_qdisc_ops);
	if (err < 0)
		unregister_qdisc(&prio_qdisc_ops);
	return err;
}

static void __exit prio_module_exit(void)
{
	unregister_qdisc(&prio_qdisc_ops);
	unregister_qdisc(&rr_qdisc_ops);
}

module_init(prio_module_init)
module_exit(prio_module_exit)

MODULE_LICENSE("GPL");
MODULE_ALIAS("sch_rr");
Commit	Line	Data
1da177e4 LT	1	/*
	2	* net/sched/sch_prio.c Simple 3-band priority "scheduler".
	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>
10297b99	10	* Fixes: 19990609: J Hadi Salim <hadi@nortelnetworks.com>:
1da177e4 LT	11	* Init -- EINVAL when opt undefined
	12	*/
	13
1da177e4	14	#include <linux/module.h>
1da177e4 LT	15	#include <linux/types.h>
1da177e4 LT	16	#include <linux/kernel.h>
1da177e4	17	#include <linux/string.h>
1da177e4	18	#include <linux/errno.h>
1da177e4	19	#include <linux/skbuff.h>
dc5fc579	20	#include <net/netlink.h>
1da177e4 LT	21	#include <net/pkt_sched.h>
	22
	23
	24	struct prio_sched_data
	25	{
	26	int bands;
d62733c8	27	int curband; /* for round-robin */
1da177e4 LT	28	struct tcf_proto *filter_list;
	29	u8 prio2band[TC_PRIO_MAX+1];
	30	struct Qdisc *queues[TCQ_PRIO_BANDS];
d62733c8	31	int mq;
1da177e4 LT	32	};
	33
	34
	35	static struct Qdisc *
	36	prio_classify(struct sk_buff skb, struct Qdisc sch, int *qerr)
	37	{
	38	struct prio_sched_data *q = qdisc_priv(sch);
	39	u32 band = skb->priority;
	40	struct tcf_result res;
	41
29f1df6c	42	*qerr = NET_XMIT_BYPASS;
1da177e4 LT	43	if (TC_H_MAJ(skb->priority) != sch->handle) {
	44	#ifdef CONFIG_NET_CLS_ACT
	45	switch (tc_classify(skb, q->filter_list, &res)) {
	46	case TC_ACT_STOLEN:
	47	case TC_ACT_QUEUED:
	48	*qerr = NET_XMIT_SUCCESS;
	49	case TC_ACT_SHOT:
	50	return NULL;
3ff50b79	51	}
1da177e4 LT	52
	53	if (!q->filter_list ) {
	54	#else
	55	if (!q->filter_list \|\| tc_classify(skb, q->filter_list, &res)) {
	56	#endif
	57	if (TC_H_MAJ(band))
	58	band = 0;
d62733c8 PWJ	59	band = q->prio2band[band&TC_PRIO_MAX];
d62733c8 PWJ	60	goto out;
1da177e4 LT	61	}
	62	band = res.classid;
	63	}
	64	band = TC_H_MIN(band) - 1;
3e5c2d3b	65	if (band >= q->bands)
d62733c8 PWJ	66	band = q->prio2band[0];
	67	out:
	68	if (q->mq)
	69	skb_set_queue_mapping(skb, band);
1da177e4 LT	70	return q->queues[band];
	71	}
	72
	73	static int
	74	prio_enqueue(struct sk_buff skb, struct Qdisc sch)
	75	{
	76	struct Qdisc *qdisc;
	77	int ret;
	78
	79	qdisc = prio_classify(skb, sch, &ret);
	80	#ifdef CONFIG_NET_CLS_ACT
	81	if (qdisc == NULL) {
29f1df6c JHS	82
29f1df6c JHS	83	if (ret == NET_XMIT_BYPASS)
1da177e4 LT	84	sch->qstats.drops++;
	85	kfree_skb(skb);
	86	return ret;
	87	}
	88	#endif
	89
	90	if ((ret = qdisc->enqueue(skb, qdisc)) == NET_XMIT_SUCCESS) {
	91	sch->bstats.bytes += skb->len;
	92	sch->bstats.packets++;
	93	sch->q.qlen++;
	94	return NET_XMIT_SUCCESS;
	95	}
	96	sch->qstats.drops++;
10297b99	97	return ret;
1da177e4 LT	98	}
	99
	100
	101	static int
	102	prio_requeue(struct sk_buff skb, struct Qdisc sch)
	103	{
	104	struct Qdisc *qdisc;
	105	int ret;
	106
	107	qdisc = prio_classify(skb, sch, &ret);
	108	#ifdef CONFIG_NET_CLS_ACT
	109	if (qdisc == NULL) {
29f1df6c	110	if (ret == NET_XMIT_BYPASS)
1da177e4 LT	111	sch->qstats.drops++;
	112	kfree_skb(skb);
	113	return ret;
	114	}
	115	#endif
	116
	117	if ((ret = qdisc->ops->requeue(skb, qdisc)) == NET_XMIT_SUCCESS) {
	118	sch->q.qlen++;
	119	sch->qstats.requeues++;
	120	return 0;
	121	}
	122	sch->qstats.drops++;
	123	return NET_XMIT_DROP;
	124	}
	125
	126
	127	static struct sk_buff *
	128	prio_dequeue(struct Qdisc* sch)
	129	{
	130	struct sk_buff *skb;
	131	struct prio_sched_data *q = qdisc_priv(sch);
	132	int prio;
	133	struct Qdisc *qdisc;
	134
	135	for (prio = 0; prio < q->bands; prio++) {
d62733c8 PWJ	136	/* Check if the target subqueue is available before
	137	* pulling an skb. This way we avoid excessive requeues
	138	* for slower queues.
	139	*/
	140	if (!netif_subqueue_stopped(sch->dev, (q->mq ? prio : 0))) {
	141	qdisc = q->queues[prio];
	142	skb = qdisc->dequeue(qdisc);
	143	if (skb) {
	144	sch->q.qlen--;
	145	return skb;
	146	}
1da177e4 LT	147	}
	148	}
	149	return NULL;
	150
	151	}
	152
d62733c8 PWJ	153	static struct sk_buff rr_dequeue(struct Qdisc sch)
	154	{
	155	struct sk_buff *skb;
	156	struct prio_sched_data *q = qdisc_priv(sch);
	157	struct Qdisc *qdisc;
	158	int bandcount;
	159
	160	/* Only take one pass through the queues. If nothing is available,
	161	* return nothing.
	162	*/
	163	for (bandcount = 0; bandcount < q->bands; bandcount++) {
	164	/* Check if the target subqueue is available before
	165	* pulling an skb. This way we avoid excessive requeues
	166	* for slower queues. If the queue is stopped, try the
	167	* next queue.
	168	*/
	169	if (!netif_subqueue_stopped(sch->dev,
	170	(q->mq ? q->curband : 0))) {
	171	qdisc = q->queues[q->curband];
	172	skb = qdisc->dequeue(qdisc);
	173	if (skb) {
	174	sch->q.qlen--;
	175	q->curband++;
	176	if (q->curband >= q->bands)
	177	q->curband = 0;
	178	return skb;
	179	}
	180	}
	181	q->curband++;
	182	if (q->curband >= q->bands)
	183	q->curband = 0;
	184	}
	185	return NULL;
	186	}
	187
1da177e4 LT	188	static unsigned int prio_drop(struct Qdisc* sch)
	189	{
	190	struct prio_sched_data *q = qdisc_priv(sch);
	191	int prio;
	192	unsigned int len;
	193	struct Qdisc *qdisc;
	194
	195	for (prio = q->bands-1; prio >= 0; prio--) {
	196	qdisc = q->queues[prio];
6d037a26	197	if (qdisc->ops->drop && (len = qdisc->ops->drop(qdisc)) != 0) {
1da177e4 LT	198	sch->q.qlen--;
	199	return len;
	200	}
	201	}
	202	return 0;
	203	}
	204
	205
	206	static void
	207	prio_reset(struct Qdisc* sch)
	208	{
	209	int prio;
	210	struct prio_sched_data *q = qdisc_priv(sch);
	211
	212	for (prio=0; prio<q->bands; prio++)
	213	qdisc_reset(q->queues[prio]);
	214	sch->q.qlen = 0;
	215	}
	216
	217	static void
	218	prio_destroy(struct Qdisc* sch)
	219	{
	220	int prio;
	221	struct prio_sched_data *q = qdisc_priv(sch);
1da177e4	222
a48b5a61	223	tcf_destroy_chain(q->filter_list);
1da177e4 LT	224	for (prio=0; prio<q->bands; prio++)
	225	qdisc_destroy(q->queues[prio]);
	226	}
	227
	228	static int prio_tune(struct Qdisc sch, struct rtattr opt)
	229	{
	230	struct prio_sched_data *q = qdisc_priv(sch);
d62733c8 PWJ	231	struct tc_prio_qopt *qopt;
d62733c8 PWJ	232	struct rtattr *tb[TCA_PRIO_MAX];
1da177e4 LT	233	int i;
1da177e4 LT	234
d62733c8 PWJ	235	if (rtattr_parse_nested_compat(tb, TCA_PRIO_MAX, opt, qopt,
d62733c8 PWJ	236	sizeof(*qopt)))
1da177e4	237	return -EINVAL;
d62733c8 PWJ	238	q->bands = qopt->bands;
	239	/* If we're multiqueue, make sure the number of incoming bands
	240	* matches the number of queues on the device we're associating with.
	241	* If the number of bands requested is zero, then set q->bands to
	242	* dev->egress_subqueue_count.
	243	*/
	244	q->mq = RTA_GET_FLAG(tb[TCA_PRIO_MQ - 1]);
	245	if (q->mq) {
	246	if (sch->handle != TC_H_ROOT)
	247	return -EINVAL;
	248	if (netif_is_multiqueue(sch->dev)) {
	249	if (q->bands == 0)
	250	q->bands = sch->dev->egress_subqueue_count;
	251	else if (q->bands != sch->dev->egress_subqueue_count)
	252	return -EINVAL;
	253	} else
	254	return -EOPNOTSUPP;
	255	}
	256
	257	if (q->bands > TCQ_PRIO_BANDS \|\| q->bands < 2)
1da177e4 LT	258	return -EINVAL;
	259
	260	for (i=0; i<=TC_PRIO_MAX; i++) {
d62733c8	261	if (qopt->priomap[i] >= q->bands)
1da177e4 LT	262	return -EINVAL;
	263	}
	264
	265	sch_tree_lock(sch);
1da177e4 LT	266	memcpy(q->prio2band, qopt->priomap, TC_PRIO_MAX+1);
	267
	268	for (i=q->bands; i<TCQ_PRIO_BANDS; i++) {
	269	struct Qdisc *child = xchg(&q->queues[i], &noop_qdisc);
5e50da01 PM	270	if (child != &noop_qdisc) {
5e50da01 PM	271	qdisc_tree_decrease_qlen(child, child->q.qlen);
1da177e4	272	qdisc_destroy(child);
5e50da01	273	}
1da177e4 LT	274	}
	275	sch_tree_unlock(sch);
	276
dd914b40 AA	277	for (i=0; i<q->bands; i++) {
dd914b40 AA	278	if (q->queues[i] == &noop_qdisc) {
1da177e4	279	struct Qdisc *child;
9f9afec4 PM	280	child = qdisc_create_dflt(sch->dev, &pfifo_qdisc_ops,
9f9afec4 PM	281	TC_H_MAKE(sch->handle, i + 1));
1da177e4 LT	282	if (child) {
1da177e4 LT	283	sch_tree_lock(sch);
dd914b40	284	child = xchg(&q->queues[i], child);
1da177e4	285
5e50da01 PM	286	if (child != &noop_qdisc) {
	287	qdisc_tree_decrease_qlen(child,
	288	child->q.qlen);
1da177e4	289	qdisc_destroy(child);
5e50da01	290	}
1da177e4 LT	291	sch_tree_unlock(sch);
	292	}
	293	}
	294	}
	295	return 0;
	296	}
	297
	298	static int prio_init(struct Qdisc sch, struct rtattr opt)
	299	{
	300	struct prio_sched_data *q = qdisc_priv(sch);
	301	int i;
	302
	303	for (i=0; i<TCQ_PRIO_BANDS; i++)
	304	q->queues[i] = &noop_qdisc;
	305
	306	if (opt == NULL) {
	307	return -EINVAL;
	308	} else {
	309	int err;
	310
	311	if ((err= prio_tune(sch, opt)) != 0)
	312	return err;
	313	}
	314	return 0;
	315	}
	316
	317	static int prio_dump(struct Qdisc sch, struct sk_buff skb)
	318	{
	319	struct prio_sched_data *q = qdisc_priv(sch);
27a884dc	320	unsigned char *b = skb_tail_pointer(skb);
d62733c8	321	struct rtattr *nest;
1da177e4 LT	322	struct tc_prio_qopt opt;
	323
	324	opt.bands = q->bands;
	325	memcpy(&opt.priomap, q->prio2band, TC_PRIO_MAX+1);
d62733c8 PWJ	326
	327	nest = RTA_NEST_COMPAT(skb, TCA_OPTIONS, sizeof(opt), &opt);
	328	if (q->mq)
	329	RTA_PUT_FLAG(skb, TCA_PRIO_MQ);
	330	RTA_NEST_COMPAT_END(skb, nest);
	331
1da177e4 LT	332	return skb->len;
	333
	334	rtattr_failure:
dc5fc579	335	nlmsg_trim(skb, b);
1da177e4 LT	336	return -1;
	337	}
	338
	339	static int prio_graft(struct Qdisc sch, unsigned long arg, struct Qdisc new,
	340	struct Qdisc **old)
	341	{
	342	struct prio_sched_data *q = qdisc_priv(sch);
	343	unsigned long band = arg - 1;
	344
	345	if (band >= q->bands)
	346	return -EINVAL;
	347
	348	if (new == NULL)
	349	new = &noop_qdisc;
	350
	351	sch_tree_lock(sch);
	352	*old = q->queues[band];
	353	q->queues[band] = new;
5e50da01	354	qdisc_tree_decrease_qlen(old, (old)->q.qlen);
1da177e4 LT	355	qdisc_reset(*old);
	356	sch_tree_unlock(sch);
	357
	358	return 0;
	359	}
	360
	361	static struct Qdisc *
	362	prio_leaf(struct Qdisc *sch, unsigned long arg)
	363	{
	364	struct prio_sched_data *q = qdisc_priv(sch);
	365	unsigned long band = arg - 1;
	366
	367	if (band >= q->bands)
	368	return NULL;
	369
	370	return q->queues[band];
	371	}
	372
	373	static unsigned long prio_get(struct Qdisc *sch, u32 classid)
	374	{
	375	struct prio_sched_data *q = qdisc_priv(sch);
	376	unsigned long band = TC_H_MIN(classid);
	377
	378	if (band - 1 >= q->bands)
	379	return 0;
	380	return band;
	381	}
	382
	383	static unsigned long prio_bind(struct Qdisc *sch, unsigned long parent, u32 classid)
	384	{
	385	return prio_get(sch, classid);
	386	}
	387
	388
	389	static void prio_put(struct Qdisc *q, unsigned long cl)
	390	{
	391	return;
	392	}
	393
	394	static int prio_change(struct Qdisc sch, u32 handle, u32 parent, struct rtattr tca, unsigned long arg)
	395	{
	396	unsigned long cl = *arg;
	397	struct prio_sched_data *q = qdisc_priv(sch);
	398
	399	if (cl - 1 > q->bands)
	400	return -ENOENT;
	401	return 0;
	402	}
	403
	404	static int prio_delete(struct Qdisc *sch, unsigned long cl)
	405	{
	406	struct prio_sched_data *q = qdisc_priv(sch);
	407	if (cl - 1 > q->bands)
	408	return -ENOENT;
	409	return 0;
	410	}
	411
	412
	413	static int prio_dump_class(struct Qdisc sch, unsigned long cl, struct sk_buff skb,
	414	struct tcmsg *tcm)
	415	{
	416	struct prio_sched_data *q = qdisc_priv(sch);
	417
	418	if (cl - 1 > q->bands)
419	return -ENOENT;
420	tcm->tcm_handle \|= TC_H_MIN(cl);
421	if (q->queues[cl-1])
422	tcm->tcm_info = q->queues[cl-1]->handle;
423	return 0;
424	}
425
2cf6c36c JP	426	static int prio_dump_class_stats(struct Qdisc *sch, unsigned long cl,
	427	struct gnet_dump *d)
	428	{
	429	struct prio_sched_data *q = qdisc_priv(sch);
	430	struct Qdisc *cl_q;
	431
	432	cl_q = q->queues[cl - 1];
	433	if (gnet_stats_copy_basic(d, &cl_q->bstats) < 0 \|\|
	434	gnet_stats_copy_queue(d, &cl_q->qstats) < 0)
	435	return -1;
	436
	437	return 0;
	438	}
	439
1da177e4 LT	440	static void prio_walk(struct Qdisc sch, struct qdisc_walker arg)
	441	{
	442	struct prio_sched_data *q = qdisc_priv(sch);
	443	int prio;
	444
	445	if (arg->stop)
	446	return;
	447
	448	for (prio = 0; prio < q->bands; prio++) {
	449	if (arg->count < arg->skip) {
	450	arg->count++;
	451	continue;
	452	}
	453	if (arg->fn(sch, prio+1, arg) < 0) {
	454	arg->stop = 1;
	455	break;
	456	}
	457	arg->count++;
	458	}
	459	}
	460
	461	static struct tcf_proto ** prio_find_tcf(struct Qdisc *sch, unsigned long cl)
	462	{
	463	struct prio_sched_data *q = qdisc_priv(sch);
	464
	465	if (cl)
	466	return NULL;
	467	return &q->filter_list;
	468	}
	469
	470	static struct Qdisc_class_ops prio_class_ops = {
	471	.graft = prio_graft,
	472	.leaf = prio_leaf,
	473	.get = prio_get,
	474	.put = prio_put,
	475	.change = prio_change,
	476	.delete = prio_delete,
	477	.walk = prio_walk,
	478	.tcf_chain = prio_find_tcf,
	479	.bind_tcf = prio_bind,
	480	.unbind_tcf = prio_put,
	481	.dump = prio_dump_class,
2cf6c36c	482	.dump_stats = prio_dump_class_stats,
1da177e4 LT	483	};
	484
	485	static struct Qdisc_ops prio_qdisc_ops = {
	486	.next = NULL,
	487	.cl_ops = &prio_class_ops,
	488	.id = "prio",
	489	.priv_size = sizeof(struct prio_sched_data),
	490	.enqueue = prio_enqueue,
	491	.dequeue = prio_dequeue,
	492	.requeue = prio_requeue,
	493	.drop = prio_drop,
	494	.init = prio_init,
	495	.reset = prio_reset,
	496	.destroy = prio_destroy,
	497	.change = prio_tune,
	498	.dump = prio_dump,
	499	.owner = THIS_MODULE,
	500	};
	501
d62733c8 PWJ	502	static struct Qdisc_ops rr_qdisc_ops = {
	503	.next = NULL,
	504	.cl_ops = &prio_class_ops,
	505	.id = "rr",
	506	.priv_size = sizeof(struct prio_sched_data),
	507	.enqueue = prio_enqueue,
	508	.dequeue = rr_dequeue,
	509	.requeue = prio_requeue,
	510	.drop = prio_drop,
	511	.init = prio_init,
	512	.reset = prio_reset,
	513	.destroy = prio_destroy,
	514	.change = prio_tune,
	515	.dump = prio_dump,
	516	.owner = THIS_MODULE,
	517	};
	518
1da177e4 LT	519	static int __init prio_module_init(void)
1da177e4 LT	520	{
d62733c8 PWJ	521	int err;
	522
	523	err = register_qdisc(&prio_qdisc_ops);
	524	if (err < 0)
	525	return err;
	526	err = register_qdisc(&rr_qdisc_ops);
	527	if (err < 0)
	528	unregister_qdisc(&prio_qdisc_ops);
	529	return err;
1da177e4 LT	530	}
1da177e4 LT	531
10297b99	532	static void __exit prio_module_exit(void)
1da177e4 LT	533	{
1da177e4 LT	534	unregister_qdisc(&prio_qdisc_ops);
d62733c8	535	unregister_qdisc(&rr_qdisc_ops);
1da177e4 LT	536	}
	537
	538	module_init(prio_module_init)
	539	module_exit(prio_module_exit)
	540
	541	MODULE_LICENSE("GPL");
d62733c8	542	MODULE_ALIAS("sch_rr");