[deliverable/linux.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;
	int err;

	*qerr = NET_XMIT_BYPASS;
	if (TC_H_MAJ(skb->priority) != sch->handle) {
		err = tc_classify(skb, q->filter_list, &res);
#ifdef CONFIG_NET_CLS_ACT
		switch (err) {
		case TC_ACT_STOLEN:
		case TC_ACT_QUEUED:
			*qerr = NET_XMIT_SUCCESS;
		case TC_ACT_SHOT:
			return NULL;
		}
#endif
		if (!q->filter_list || err < 0) {
			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 nlattr *opt)
{
	struct prio_sched_data *q = qdisc_priv(sch);
	struct tc_prio_qopt *qopt;
	struct nlattr *tb[TCA_PRIO_MAX + 1];
	int i;

	if (nla_parse_nested_compat(tb, TCA_PRIO_MAX, opt, NULL, 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.  Also, the root qdisc must be the
	 * only one that is enabled for multiqueue, since it's the only one
	 * that interacts with the underlying device.
	 */
	q->mq = nla_get_flag(tb[TCA_PRIO_MQ]);
	if (q->mq) {
		if (sch->parent != 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 nlattr *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 nlattr *nest;
	struct tc_prio_qopt opt;

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

	nest = nla_nest_compat_start(skb, TCA_OPTIONS, sizeof(opt), &opt);
	if (nest == NULL)
		goto nla_put_failure;
	if (q->mq) {
		if (nla_put_flag(skb, TCA_PRIO_MQ) < 0)
			goto nla_put_failure;
	}
	nla_nest_compat_end(skb, nest);

	return skb->len;

nla_put_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 nlattr **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 const 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 __read_mostly = {
	.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 __read_mostly = {
	.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;
bdba91ec	41	int err;
1da177e4	42
29f1df6c	43	*qerr = NET_XMIT_BYPASS;
1da177e4	44	if (TC_H_MAJ(skb->priority) != sch->handle) {
bdba91ec	45	err = tc_classify(skb, q->filter_list, &res);
1da177e4	46	#ifdef CONFIG_NET_CLS_ACT
dbaaa07a	47	switch (err) {
1da177e4 LT	48	case TC_ACT_STOLEN:
	49	case TC_ACT_QUEUED:
	50	*qerr = NET_XMIT_SUCCESS;
	51	case TC_ACT_SHOT:
	52	return NULL;
3ff50b79	53	}
1da177e4	54	#endif
bdba91ec	55	if (!q->filter_list \|\| err < 0) {
1da177e4 LT	56	if (TC_H_MAJ(band))
1da177e4 LT	57	band = 0;
d62733c8 PWJ	58	band = q->prio2band[band&TC_PRIO_MAX];
d62733c8 PWJ	59	goto out;
1da177e4 LT	60	}
	61	band = res.classid;
	62	}
	63	band = TC_H_MIN(band) - 1;
3e5c2d3b	64	if (band >= q->bands)
d62733c8 PWJ	65	band = q->prio2band[0];
	66	out:
	67	if (q->mq)
	68	skb_set_queue_mapping(skb, band);
1da177e4 LT	69	return q->queues[band];
	70	}
	71
	72	static int
	73	prio_enqueue(struct sk_buff skb, struct Qdisc sch)
	74	{
	75	struct Qdisc *qdisc;
	76	int ret;
	77
	78	qdisc = prio_classify(skb, sch, &ret);
	79	#ifdef CONFIG_NET_CLS_ACT
	80	if (qdisc == NULL) {
29f1df6c JHS	81
29f1df6c JHS	82	if (ret == NET_XMIT_BYPASS)
1da177e4 LT	83	sch->qstats.drops++;
	84	kfree_skb(skb);
	85	return ret;
	86	}
	87	#endif
	88
	89	if ((ret = qdisc->enqueue(skb, qdisc)) == NET_XMIT_SUCCESS) {
	90	sch->bstats.bytes += skb->len;
	91	sch->bstats.packets++;
	92	sch->q.qlen++;
	93	return NET_XMIT_SUCCESS;
	94	}
	95	sch->qstats.drops++;
10297b99	96	return ret;
1da177e4 LT	97	}
	98
	99
	100	static int
	101	prio_requeue(struct sk_buff skb, struct Qdisc sch)
	102	{
	103	struct Qdisc *qdisc;
	104	int ret;
	105
	106	qdisc = prio_classify(skb, sch, &ret);
	107	#ifdef CONFIG_NET_CLS_ACT
	108	if (qdisc == NULL) {
29f1df6c	109	if (ret == NET_XMIT_BYPASS)
1da177e4 LT	110	sch->qstats.drops++;
	111	kfree_skb(skb);
	112	return ret;
	113	}
	114	#endif
	115
	116	if ((ret = qdisc->ops->requeue(skb, qdisc)) == NET_XMIT_SUCCESS) {
	117	sch->q.qlen++;
	118	sch->qstats.requeues++;
	119	return 0;
	120	}
	121	sch->qstats.drops++;
	122	return NET_XMIT_DROP;
	123	}
	124
	125
	126	static struct sk_buff *
	127	prio_dequeue(struct Qdisc* sch)
	128	{
	129	struct sk_buff *skb;
	130	struct prio_sched_data *q = qdisc_priv(sch);
	131	int prio;
	132	struct Qdisc *qdisc;
	133
	134	for (prio = 0; prio < q->bands; prio++) {
d62733c8 PWJ	135	/* Check if the target subqueue is available before
	136	* pulling an skb. This way we avoid excessive requeues
	137	* for slower queues.
	138	*/
00346226	139	if (!__netif_subqueue_stopped(sch->dev, (q->mq ? prio : 0))) {
d62733c8 PWJ	140	qdisc = q->queues[prio];
	141	skb = qdisc->dequeue(qdisc);
	142	if (skb) {
	143	sch->q.qlen--;
	144	return skb;
	145	}
1da177e4 LT	146	}
	147	}
	148	return NULL;
	149
	150	}
	151
d62733c8 PWJ	152	static struct sk_buff rr_dequeue(struct Qdisc sch)
	153	{
	154	struct sk_buff *skb;
	155	struct prio_sched_data *q = qdisc_priv(sch);
	156	struct Qdisc *qdisc;
	157	int bandcount;
	158
	159	/* Only take one pass through the queues. If nothing is available,
	160	* return nothing.
	161	*/
	162	for (bandcount = 0; bandcount < q->bands; bandcount++) {
	163	/* Check if the target subqueue is available before
	164	* pulling an skb. This way we avoid excessive requeues
	165	* for slower queues. If the queue is stopped, try the
	166	* next queue.
	167	*/
00346226	168	if (!__netif_subqueue_stopped(sch->dev,
d62733c8 PWJ	169	(q->mq ? q->curband : 0))) {
	170	qdisc = q->queues[q->curband];
	171	skb = qdisc->dequeue(qdisc);
	172	if (skb) {
	173	sch->q.qlen--;
	174	q->curband++;
	175	if (q->curband >= q->bands)
	176	q->curband = 0;
	177	return skb;
	178	}
	179	}
	180	q->curband++;
	181	if (q->curband >= q->bands)
	182	q->curband = 0;
	183	}
	184	return NULL;
	185	}
	186
1da177e4 LT	187	static unsigned int prio_drop(struct Qdisc* sch)
	188	{
	189	struct prio_sched_data *q = qdisc_priv(sch);
	190	int prio;
	191	unsigned int len;
	192	struct Qdisc *qdisc;
	193
	194	for (prio = q->bands-1; prio >= 0; prio--) {
	195	qdisc = q->queues[prio];
6d037a26	196	if (qdisc->ops->drop && (len = qdisc->ops->drop(qdisc)) != 0) {
1da177e4 LT	197	sch->q.qlen--;
	198	return len;
	199	}
	200	}
	201	return 0;
	202	}
	203
	204
	205	static void
	206	prio_reset(struct Qdisc* sch)
	207	{
	208	int prio;
	209	struct prio_sched_data *q = qdisc_priv(sch);
	210
	211	for (prio=0; prio<q->bands; prio++)
	212	qdisc_reset(q->queues[prio]);
	213	sch->q.qlen = 0;
	214	}
	215
	216	static void
	217	prio_destroy(struct Qdisc* sch)
	218	{
	219	int prio;
	220	struct prio_sched_data *q = qdisc_priv(sch);
1da177e4	221
a48b5a61	222	tcf_destroy_chain(q->filter_list);
1da177e4 LT	223	for (prio=0; prio<q->bands; prio++)
	224	qdisc_destroy(q->queues[prio]);
	225	}
	226
1e90474c	227	static int prio_tune(struct Qdisc sch, struct nlattr opt)
1da177e4 LT	228	{
1da177e4 LT	229	struct prio_sched_data *q = qdisc_priv(sch);
d62733c8	230	struct tc_prio_qopt *qopt;
1e90474c	231	struct nlattr *tb[TCA_PRIO_MAX + 1];
1da177e4 LT	232	int i;
1da177e4 LT	233
1e90474c PM	234	if (nla_parse_nested_compat(tb, TCA_PRIO_MAX, opt, NULL, qopt,
1e90474c PM	235	sizeof(*qopt)))
1da177e4	236	return -EINVAL;
d62733c8 PWJ	237	q->bands = qopt->bands;
	238	/* If we're multiqueue, make sure the number of incoming bands
	239	* matches the number of queues on the device we're associating with.
	240	* If the number of bands requested is zero, then set q->bands to
0773192b PWJ	241	* dev->egress_subqueue_count. Also, the root qdisc must be the
	242	* only one that is enabled for multiqueue, since it's the only one
	243	* that interacts with the underlying device.
d62733c8	244	*/
1e90474c	245	q->mq = nla_get_flag(tb[TCA_PRIO_MQ]);
d62733c8	246	if (q->mq) {
0773192b	247	if (sch->parent != TC_H_ROOT)
d62733c8 PWJ	248	return -EINVAL;
	249	if (netif_is_multiqueue(sch->dev)) {
	250	if (q->bands == 0)
	251	q->bands = sch->dev->egress_subqueue_count;
	252	else if (q->bands != sch->dev->egress_subqueue_count)
	253	return -EINVAL;
	254	} else
	255	return -EOPNOTSUPP;
	256	}
	257
	258	if (q->bands > TCQ_PRIO_BANDS \|\| q->bands < 2)
1da177e4 LT	259	return -EINVAL;
	260
	261	for (i=0; i<=TC_PRIO_MAX; i++) {
d62733c8	262	if (qopt->priomap[i] >= q->bands)
1da177e4 LT	263	return -EINVAL;
	264	}
	265
	266	sch_tree_lock(sch);
1da177e4 LT	267	memcpy(q->prio2band, qopt->priomap, TC_PRIO_MAX+1);
	268
	269	for (i=q->bands; i<TCQ_PRIO_BANDS; i++) {
	270	struct Qdisc *child = xchg(&q->queues[i], &noop_qdisc);
5e50da01 PM	271	if (child != &noop_qdisc) {
5e50da01 PM	272	qdisc_tree_decrease_qlen(child, child->q.qlen);
1da177e4	273	qdisc_destroy(child);
5e50da01	274	}
1da177e4 LT	275	}
	276	sch_tree_unlock(sch);
	277
dd914b40 AA	278	for (i=0; i<q->bands; i++) {
dd914b40 AA	279	if (q->queues[i] == &noop_qdisc) {
1da177e4	280	struct Qdisc *child;
9f9afec4 PM	281	child = qdisc_create_dflt(sch->dev, &pfifo_qdisc_ops,
9f9afec4 PM	282	TC_H_MAKE(sch->handle, i + 1));
1da177e4 LT	283	if (child) {
1da177e4 LT	284	sch_tree_lock(sch);
dd914b40	285	child = xchg(&q->queues[i], child);
1da177e4	286
5e50da01 PM	287	if (child != &noop_qdisc) {
	288	qdisc_tree_decrease_qlen(child,
	289	child->q.qlen);
1da177e4	290	qdisc_destroy(child);
5e50da01	291	}
1da177e4 LT	292	sch_tree_unlock(sch);
	293	}
	294	}
	295	}
	296	return 0;
	297	}
	298
1e90474c	299	static int prio_init(struct Qdisc sch, struct nlattr opt)
1da177e4 LT	300	{
	301	struct prio_sched_data *q = qdisc_priv(sch);
	302	int i;
	303
	304	for (i=0; i<TCQ_PRIO_BANDS; i++)
	305	q->queues[i] = &noop_qdisc;
	306
	307	if (opt == NULL) {
	308	return -EINVAL;
	309	} else {
	310	int err;
	311
	312	if ((err= prio_tune(sch, opt)) != 0)
	313	return err;
	314	}
	315	return 0;
	316	}
	317
	318	static int prio_dump(struct Qdisc sch, struct sk_buff skb)
	319	{
	320	struct prio_sched_data *q = qdisc_priv(sch);
27a884dc	321	unsigned char *b = skb_tail_pointer(skb);
1e90474c	322	struct nlattr *nest;
1da177e4 LT	323	struct tc_prio_qopt opt;
	324
	325	opt.bands = q->bands;
	326	memcpy(&opt.priomap, q->prio2band, TC_PRIO_MAX+1);
d62733c8	327
1e90474c PM	328	nest = nla_nest_compat_start(skb, TCA_OPTIONS, sizeof(opt), &opt);
	329	if (nest == NULL)
	330	goto nla_put_failure;
	331	if (q->mq) {
	332	if (nla_put_flag(skb, TCA_PRIO_MQ) < 0)
	333	goto nla_put_failure;
	334	}
	335	nla_nest_compat_end(skb, nest);
d62733c8	336
1da177e4 LT	337	return skb->len;
1da177e4 LT	338
1e90474c	339	nla_put_failure:
dc5fc579	340	nlmsg_trim(skb, b);
1da177e4 LT	341	return -1;
	342	}
	343
	344	static int prio_graft(struct Qdisc sch, unsigned long arg, struct Qdisc new,
	345	struct Qdisc **old)
	346	{
	347	struct prio_sched_data *q = qdisc_priv(sch);
	348	unsigned long band = arg - 1;
	349
	350	if (band >= q->bands)
	351	return -EINVAL;
	352
	353	if (new == NULL)
	354	new = &noop_qdisc;
	355
	356	sch_tree_lock(sch);
	357	*old = q->queues[band];
	358	q->queues[band] = new;
5e50da01	359	qdisc_tree_decrease_qlen(old, (old)->q.qlen);
1da177e4 LT	360	qdisc_reset(*old);
	361	sch_tree_unlock(sch);
	362
	363	return 0;
	364	}
	365
	366	static struct Qdisc *
	367	prio_leaf(struct Qdisc *sch, unsigned long arg)
	368	{
	369	struct prio_sched_data *q = qdisc_priv(sch);
	370	unsigned long band = arg - 1;
	371
	372	if (band >= q->bands)
	373	return NULL;
	374
	375	return q->queues[band];
	376	}
	377
	378	static unsigned long prio_get(struct Qdisc *sch, u32 classid)
	379	{
	380	struct prio_sched_data *q = qdisc_priv(sch);
	381	unsigned long band = TC_H_MIN(classid);
	382
	383	if (band - 1 >= q->bands)
	384	return 0;
	385	return band;
	386	}
	387
	388	static unsigned long prio_bind(struct Qdisc *sch, unsigned long parent, u32 classid)
	389	{
	390	return prio_get(sch, classid);
	391	}
	392
	393
	394	static void prio_put(struct Qdisc *q, unsigned long cl)
	395	{
	396	return;
	397	}
	398
1e90474c	399	static int prio_change(struct Qdisc sch, u32 handle, u32 parent, struct nlattr tca, unsigned long arg)
1da177e4 LT	400	{
	401	unsigned long cl = *arg;
	402	struct prio_sched_data *q = qdisc_priv(sch);
	403
	404	if (cl - 1 > q->bands)
	405	return -ENOENT;
	406	return 0;
	407	}
	408
	409	static int prio_delete(struct Qdisc *sch, unsigned long cl)
	410	{
	411	struct prio_sched_data *q = qdisc_priv(sch);
	412	if (cl - 1 > q->bands)
	413	return -ENOENT;
	414	return 0;
	415	}
	416
	417
	418	static int prio_dump_class(struct Qdisc sch, unsigned long cl, struct sk_buff skb,
	419	struct tcmsg *tcm)
	420	{
	421	struct prio_sched_data *q = qdisc_priv(sch);
	422
	423	if (cl - 1 > q->bands)
	424	return -ENOENT;
	425	tcm->tcm_handle \|= TC_H_MIN(cl);
	426	if (q->queues[cl-1])
	427	tcm->tcm_info = q->queues[cl-1]->handle;
	428	return 0;
	429	}
	430
2cf6c36c JP	431	static int prio_dump_class_stats(struct Qdisc *sch, unsigned long cl,
	432	struct gnet_dump *d)
	433	{
	434	struct prio_sched_data *q = qdisc_priv(sch);
	435	struct Qdisc *cl_q;
	436
	437	cl_q = q->queues[cl - 1];
	438	if (gnet_stats_copy_basic(d, &cl_q->bstats) < 0 \|\|
	439	gnet_stats_copy_queue(d, &cl_q->qstats) < 0)
	440	return -1;
	441
	442	return 0;
	443	}
	444
1da177e4 LT	445	static void prio_walk(struct Qdisc sch, struct qdisc_walker arg)
	446	{
	447	struct prio_sched_data *q = qdisc_priv(sch);
	448	int prio;
	449
	450	if (arg->stop)
	451	return;
	452
	453	for (prio = 0; prio < q->bands; prio++) {
	454	if (arg->count < arg->skip) {
	455	arg->count++;
	456	continue;
	457	}
	458	if (arg->fn(sch, prio+1, arg) < 0) {
	459	arg->stop = 1;
	460	break;
	461	}
	462	arg->count++;
	463	}
	464	}
	465
	466	static struct tcf_proto ** prio_find_tcf(struct Qdisc *sch, unsigned long cl)
	467	{
	468	struct prio_sched_data *q = qdisc_priv(sch);
	469
	470	if (cl)
	471	return NULL;
	472	return &q->filter_list;
	473	}
	474
20fea08b	475	static const struct Qdisc_class_ops prio_class_ops = {
1da177e4 LT	476	.graft = prio_graft,
	477	.leaf = prio_leaf,
	478	.get = prio_get,
	479	.put = prio_put,
	480	.change = prio_change,
	481	.delete = prio_delete,
	482	.walk = prio_walk,
	483	.tcf_chain = prio_find_tcf,
	484	.bind_tcf = prio_bind,
	485	.unbind_tcf = prio_put,
	486	.dump = prio_dump_class,
2cf6c36c	487	.dump_stats = prio_dump_class_stats,
1da177e4 LT	488	};
1da177e4 LT	489
20fea08b	490	static struct Qdisc_ops prio_qdisc_ops __read_mostly = {
1da177e4 LT	491	.next = NULL,
	492	.cl_ops = &prio_class_ops,
	493	.id = "prio",
	494	.priv_size = sizeof(struct prio_sched_data),
	495	.enqueue = prio_enqueue,
	496	.dequeue = prio_dequeue,
	497	.requeue = prio_requeue,
	498	.drop = prio_drop,
	499	.init = prio_init,
	500	.reset = prio_reset,
	501	.destroy = prio_destroy,
	502	.change = prio_tune,
	503	.dump = prio_dump,
	504	.owner = THIS_MODULE,
	505	};
	506
20fea08b	507	static struct Qdisc_ops rr_qdisc_ops __read_mostly = {
d62733c8 PWJ	508	.next = NULL,
	509	.cl_ops = &prio_class_ops,
	510	.id = "rr",
	511	.priv_size = sizeof(struct prio_sched_data),
	512	.enqueue = prio_enqueue,
	513	.dequeue = rr_dequeue,
	514	.requeue = prio_requeue,
	515	.drop = prio_drop,
	516	.init = prio_init,
	517	.reset = prio_reset,
	518	.destroy = prio_destroy,
	519	.change = prio_tune,
	520	.dump = prio_dump,
	521	.owner = THIS_MODULE,
	522	};
	523
1da177e4 LT	524	static int __init prio_module_init(void)
1da177e4 LT	525	{
d62733c8 PWJ	526	int err;
	527
	528	err = register_qdisc(&prio_qdisc_ops);
	529	if (err < 0)
	530	return err;
	531	err = register_qdisc(&rr_qdisc_ops);
	532	if (err < 0)
	533	unregister_qdisc(&prio_qdisc_ops);
	534	return err;
1da177e4 LT	535	}
1da177e4 LT	536
10297b99	537	static void __exit prio_module_exit(void)
1da177e4 LT	538	{
1da177e4 LT	539	unregister_qdisc(&prio_qdisc_ops);
d62733c8	540	unregister_qdisc(&rr_qdisc_ops);
1da177e4 LT	541	}
	542
	543	module_init(prio_module_init)
	544	module_exit(prio_module_exit)
	545
	546	MODULE_LICENSE("GPL");
d62733c8	547	MODULE_ALIAS("sch_rr");