[linux.git] / net / sched / sch_mqprio.c

/*
 * net/sched/sch_mqprio.c
 *
 * Copyright (c) 2010 John Fastabend <[email protected]>
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * version 2 as published by the Free Software Foundation.
 */

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

struct mqprio_sched {
	struct Qdisc		**qdiscs;
	int hw_owned;
};

static void mqprio_destroy(struct Qdisc *sch)
{
	struct net_device *dev = qdisc_dev(sch);
	struct mqprio_sched *priv = qdisc_priv(sch);
	unsigned int ntx;

	if (priv->qdiscs) {
		for (ntx = 0;
		     ntx < dev->num_tx_queues && priv->qdiscs[ntx];
		     ntx++)
			qdisc_destroy(priv->qdiscs[ntx]);
		kfree(priv->qdiscs);
	}

	if (priv->hw_owned && dev->netdev_ops->ndo_setup_tc)
		dev->netdev_ops->ndo_setup_tc(dev, 0);
	else
		netdev_set_num_tc(dev, 0);
}

static int mqprio_parse_opt(struct net_device *dev, struct tc_mqprio_qopt *qopt)
{
	int i, j;

	/* Verify num_tc is not out of max range */
	if (qopt->num_tc > TC_MAX_QUEUE)
		return -EINVAL;

	/* Verify priority mapping uses valid tcs */
	for (i = 0; i < TC_BITMASK + 1; i++) {
		if (qopt->prio_tc_map[i] >= qopt->num_tc)
			return -EINVAL;
	}

	/* net_device does not support requested operation */
	if (qopt->hw && !dev->netdev_ops->ndo_setup_tc)
		return -EINVAL;

	/* if hw owned qcount and qoffset are taken from LLD so
	 * no reason to verify them here
	 */
	if (qopt->hw)
		return 0;

	for (i = 0; i < qopt->num_tc; i++) {
		unsigned int last = qopt->offset[i] + qopt->count[i];

		/* Verify the queue count is in tx range being equal to the
		 * real_num_tx_queues indicates the last queue is in use.
		 */
		if (qopt->offset[i] >= dev->real_num_tx_queues ||
		    !qopt->count[i] ||
		    last > dev->real_num_tx_queues)
			return -EINVAL;

		/* Verify that the offset and counts do not overlap */
		for (j = i + 1; j < qopt->num_tc; j++) {
			if (last > qopt->offset[j])
				return -EINVAL;
		}
	}

	return 0;
}

static int mqprio_init(struct Qdisc *sch, struct nlattr *opt)
{
	struct net_device *dev = qdisc_dev(sch);
	struct mqprio_sched *priv = qdisc_priv(sch);
	struct netdev_queue *dev_queue;
	struct Qdisc *qdisc;
	int i, err = -EOPNOTSUPP;
	struct tc_mqprio_qopt *qopt = NULL;

	BUILD_BUG_ON(TC_MAX_QUEUE != TC_QOPT_MAX_QUEUE);
	BUILD_BUG_ON(TC_BITMASK != TC_QOPT_BITMASK);

	if (sch->parent != TC_H_ROOT)
		return -EOPNOTSUPP;

	if (!netif_is_multiqueue(dev))
		return -EOPNOTSUPP;

	if (!opt || nla_len(opt) < sizeof(*qopt))
		return -EINVAL;

	qopt = nla_data(opt);
	if (mqprio_parse_opt(dev, qopt))
		return -EINVAL;

	/* pre-allocate qdisc, attachment can't fail */
	priv->qdiscs = kcalloc(dev->num_tx_queues, sizeof(priv->qdiscs[0]),
			       GFP_KERNEL);
	if (priv->qdiscs == NULL) {
		err = -ENOMEM;
		goto err;
	}

	for (i = 0; i < dev->num_tx_queues; i++) {
		dev_queue = netdev_get_tx_queue(dev, i);
		qdisc = qdisc_create_dflt(dev_queue, &pfifo_fast_ops,
					  TC_H_MAKE(TC_H_MAJ(sch->handle),
						    TC_H_MIN(i + 1)));
		if (qdisc == NULL) {
			err = -ENOMEM;
			goto err;
		}
		priv->qdiscs[i] = qdisc;
	}

	/* If the mqprio options indicate that hardware should own
	 * the queue mapping then run ndo_setup_tc otherwise use the
	 * supplied and verified mapping
	 */
	if (qopt->hw) {
		priv->hw_owned = 1;
		err = dev->netdev_ops->ndo_setup_tc(dev, qopt->num_tc);
		if (err)
			goto err;
	} else {
		netdev_set_num_tc(dev, qopt->num_tc);
		for (i = 0; i < qopt->num_tc; i++)
			netdev_set_tc_queue(dev, i,
					    qopt->count[i], qopt->offset[i]);
	}

	/* Always use supplied priority mappings */
	for (i = 0; i < TC_BITMASK + 1; i++)
		netdev_set_prio_tc_map(dev, i, qopt->prio_tc_map[i]);

	sch->flags |= TCQ_F_MQROOT;
	return 0;

err:
	mqprio_destroy(sch);
	return err;
}

static void mqprio_attach(struct Qdisc *sch)
{
	struct net_device *dev = qdisc_dev(sch);
	struct mqprio_sched *priv = qdisc_priv(sch);
	struct Qdisc *qdisc;
	unsigned int ntx;

	/* Attach underlying qdisc */
	for (ntx = 0; ntx < dev->num_tx_queues; ntx++) {
		qdisc = priv->qdiscs[ntx];
		qdisc = dev_graft_qdisc(qdisc->dev_queue, qdisc);
		if (qdisc)
			qdisc_destroy(qdisc);
	}
	kfree(priv->qdiscs);
	priv->qdiscs = NULL;
}

static struct netdev_queue *mqprio_queue_get(struct Qdisc *sch,
					     unsigned long cl)
{
	struct net_device *dev = qdisc_dev(sch);
	unsigned long ntx = cl - 1 - netdev_get_num_tc(dev);

	if (ntx >= dev->num_tx_queues)
		return NULL;
	return netdev_get_tx_queue(dev, ntx);
}

static int mqprio_graft(struct Qdisc *sch, unsigned long cl, struct Qdisc *new,
		    struct Qdisc **old)
{
	struct net_device *dev = qdisc_dev(sch);
	struct netdev_queue *dev_queue = mqprio_queue_get(sch, cl);

	if (!dev_queue)
		return -EINVAL;

	if (dev->flags & IFF_UP)
		dev_deactivate(dev);

	*old = dev_graft_qdisc(dev_queue, new);

	if (dev->flags & IFF_UP)
		dev_activate(dev);

	return 0;
}

static int mqprio_dump(struct Qdisc *sch, struct sk_buff *skb)
{
	struct net_device *dev = qdisc_dev(sch);
	struct mqprio_sched *priv = qdisc_priv(sch);
	unsigned char *b = skb_tail_pointer(skb);
	struct tc_mqprio_qopt opt = { 0 };
	struct Qdisc *qdisc;
	unsigned int i;

	sch->q.qlen = 0;
	memset(&sch->bstats, 0, sizeof(sch->bstats));
	memset(&sch->qstats, 0, sizeof(sch->qstats));

	for (i = 0; i < dev->num_tx_queues; i++) {
		qdisc = netdev_get_tx_queue(dev, i)->qdisc;
		spin_lock_bh(qdisc_lock(qdisc));
		sch->q.qlen		+= qdisc->q.qlen;
		sch->bstats.bytes	+= qdisc->bstats.bytes;
		sch->bstats.packets	+= qdisc->bstats.packets;
		sch->qstats.qlen	+= qdisc->qstats.qlen;
		sch->qstats.backlog	+= qdisc->qstats.backlog;
		sch->qstats.drops	+= qdisc->qstats.drops;
		sch->qstats.requeues	+= qdisc->qstats.requeues;
		sch->qstats.overlimits	+= qdisc->qstats.overlimits;
		spin_unlock_bh(qdisc_lock(qdisc));
	}

	opt.num_tc = netdev_get_num_tc(dev);
	memcpy(opt.prio_tc_map, dev->prio_tc_map, sizeof(opt.prio_tc_map));
	opt.hw = priv->hw_owned;

	for (i = 0; i < netdev_get_num_tc(dev); i++) {
		opt.count[i] = dev->tc_to_txq[i].count;
		opt.offset[i] = dev->tc_to_txq[i].offset;
	}

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

	return skb->len;
nla_put_failure:
	nlmsg_trim(skb, b);
	return -1;
}

static struct Qdisc *mqprio_leaf(struct Qdisc *sch, unsigned long cl)
{
	struct netdev_queue *dev_queue = mqprio_queue_get(sch, cl);

	if (!dev_queue)
		return NULL;

	return dev_queue->qdisc_sleeping;
}

static unsigned long mqprio_get(struct Qdisc *sch, u32 classid)
{
	struct net_device *dev = qdisc_dev(sch);
	unsigned int ntx = TC_H_MIN(classid);

	if (ntx > dev->num_tx_queues + netdev_get_num_tc(dev))
		return 0;
	return ntx;
}

static void mqprio_put(struct Qdisc *sch, unsigned long cl)
{
}

static int mqprio_dump_class(struct Qdisc *sch, unsigned long cl,
			 struct sk_buff *skb, struct tcmsg *tcm)
{
	struct net_device *dev = qdisc_dev(sch);

	if (cl <= netdev_get_num_tc(dev)) {
		tcm->tcm_parent = TC_H_ROOT;
		tcm->tcm_info = 0;
	} else {
		int i;
		struct netdev_queue *dev_queue;

		dev_queue = mqprio_queue_get(sch, cl);
		tcm->tcm_parent = 0;
		for (i = 0; i < netdev_get_num_tc(dev); i++) {
			struct netdev_tc_txq tc = dev->tc_to_txq[i];
			int q_idx = cl - netdev_get_num_tc(dev);

			if (q_idx > tc.offset &&
			    q_idx <= tc.offset + tc.count) {
				tcm->tcm_parent =
					TC_H_MAKE(TC_H_MAJ(sch->handle),
						  TC_H_MIN(i + 1));
				break;
			}
		}
		tcm->tcm_info = dev_queue->qdisc_sleeping->handle;
	}
	tcm->tcm_handle |= TC_H_MIN(cl);
	return 0;
}

static int mqprio_dump_class_stats(struct Qdisc *sch, unsigned long cl,
				   struct gnet_dump *d)
	__releases(d->lock)
	__acquires(d->lock)
{
	struct net_device *dev = qdisc_dev(sch);

	if (cl <= netdev_get_num_tc(dev)) {
		int i;
		struct Qdisc *qdisc;
		struct gnet_stats_queue qstats = {0};
		struct gnet_stats_basic_packed bstats = {0};
		struct netdev_tc_txq tc = dev->tc_to_txq[cl - 1];

		/* Drop lock here it will be reclaimed before touching
		 * statistics this is required because the d->lock we
		 * hold here is the look on dev_queue->qdisc_sleeping
		 * also acquired below.
		 */
		spin_unlock_bh(d->lock);

		for (i = tc.offset; i < tc.offset + tc.count; i++) {
			qdisc = netdev_get_tx_queue(dev, i)->qdisc;
			spin_lock_bh(qdisc_lock(qdisc));
			bstats.bytes      += qdisc->bstats.bytes;
			bstats.packets    += qdisc->bstats.packets;
			qstats.qlen       += qdisc->qstats.qlen;
			qstats.backlog    += qdisc->qstats.backlog;
			qstats.drops      += qdisc->qstats.drops;
			qstats.requeues   += qdisc->qstats.requeues;
			qstats.overlimits += qdisc->qstats.overlimits;
			spin_unlock_bh(qdisc_lock(qdisc));
		}
		/* Reclaim root sleeping lock before completing stats */
		spin_lock_bh(d->lock);
		if (gnet_stats_copy_basic(d, &bstats) < 0 ||
		    gnet_stats_copy_queue(d, &qstats) < 0)
			return -1;
	} else {
		struct netdev_queue *dev_queue = mqprio_queue_get(sch, cl);

		sch = dev_queue->qdisc_sleeping;
		sch->qstats.qlen = sch->q.qlen;
		if (gnet_stats_copy_basic(d, &sch->bstats) < 0 ||
		    gnet_stats_copy_queue(d, &sch->qstats) < 0)
			return -1;
	}
	return 0;
}

static void mqprio_walk(struct Qdisc *sch, struct qdisc_walker *arg)
{
	struct net_device *dev = qdisc_dev(sch);
	unsigned long ntx;

	if (arg->stop)
		return;

	/* Walk hierarchy with a virtual class per tc */
	arg->count = arg->skip;
	for (ntx = arg->skip;
	     ntx < dev->num_tx_queues + netdev_get_num_tc(dev);
	     ntx++) {
		if (arg->fn(sch, ntx + 1, arg) < 0) {
			arg->stop = 1;
			break;
		}
		arg->count++;
	}
}

static const struct Qdisc_class_ops mqprio_class_ops = {
	.graft		= mqprio_graft,
	.leaf		= mqprio_leaf,
	.get		= mqprio_get,
	.put		= mqprio_put,
	.walk		= mqprio_walk,
	.dump		= mqprio_dump_class,
	.dump_stats	= mqprio_dump_class_stats,
};

static struct Qdisc_ops mqprio_qdisc_ops __read_mostly = {
	.cl_ops		= &mqprio_class_ops,
	.id		= "mqprio",
	.priv_size	= sizeof(struct mqprio_sched),
	.init		= mqprio_init,
	.destroy	= mqprio_destroy,
	.attach		= mqprio_attach,
	.dump		= mqprio_dump,
	.owner		= THIS_MODULE,
};

static int __init mqprio_module_init(void)
{
	return register_qdisc(&mqprio_qdisc_ops);
}

static void __exit mqprio_module_exit(void)
{
	unregister_qdisc(&mqprio_qdisc_ops);
}

module_init(mqprio_module_init);
module_exit(mqprio_module_exit);

MODULE_LICENSE("GPL");
Commit	Line	Data
b8970f0b JF	1	/*
	2	* net/sched/sch_mqprio.c
	3	*
	4	* Copyright (c) 2010 John Fastabend <[email protected]>
	5	*
	6	* This program is free software; you can redistribute it and/or
	7	* modify it under the terms of the GNU General Public License
	8	* version 2 as published by the Free Software Foundation.
	9	*/
	10
	11	#include <linux/types.h>
	12	#include <linux/slab.h>
	13	#include <linux/kernel.h>
	14	#include <linux/string.h>
	15	#include <linux/errno.h>
	16	#include <linux/skbuff.h>
3a9a231d	17	#include <linux/module.h>
b8970f0b JF	18	#include <net/netlink.h>
	19	#include <net/pkt_sched.h>
	20	#include <net/sch_generic.h>
	21
	22	struct mqprio_sched {
	23	struct Qdisc **qdiscs;
	24	int hw_owned;
	25	};
	26
	27	static void mqprio_destroy(struct Qdisc *sch)
	28	{
	29	struct net_device *dev = qdisc_dev(sch);
	30	struct mqprio_sched *priv = qdisc_priv(sch);
	31	unsigned int ntx;
	32
ac7100ba BH	33	if (priv->qdiscs) {
	34	for (ntx = 0;
	35	ntx < dev->num_tx_queues && priv->qdiscs[ntx];
	36	ntx++)
	37	qdisc_destroy(priv->qdiscs[ntx]);
	38	kfree(priv->qdiscs);
	39	}
b8970f0b JF	40
	41	if (priv->hw_owned && dev->netdev_ops->ndo_setup_tc)
	42	dev->netdev_ops->ndo_setup_tc(dev, 0);
	43	else
	44	netdev_set_num_tc(dev, 0);
b8970f0b JF	45	}
	46
	47	static int mqprio_parse_opt(struct net_device dev, struct tc_mqprio_qopt qopt)
	48	{
	49	int i, j;
	50
	51	/* Verify num_tc is not out of max range */
	52	if (qopt->num_tc > TC_MAX_QUEUE)
	53	return -EINVAL;
	54
	55	/* Verify priority mapping uses valid tcs */
	56	for (i = 0; i < TC_BITMASK + 1; i++) {
	57	if (qopt->prio_tc_map[i] >= qopt->num_tc)
	58	return -EINVAL;
	59	}
	60
	61	/* net_device does not support requested operation */
	62	if (qopt->hw && !dev->netdev_ops->ndo_setup_tc)
	63	return -EINVAL;
	64
	65	/* if hw owned qcount and qoffset are taken from LLD so
	66	* no reason to verify them here
	67	*/
	68	if (qopt->hw)
	69	return 0;
	70
	71	for (i = 0; i < qopt->num_tc; i++) {
	72	unsigned int last = qopt->offset[i] + qopt->count[i];
	73
	74	/* Verify the queue count is in tx range being equal to the
	75	* real_num_tx_queues indicates the last queue is in use.
	76	*/
	77	if (qopt->offset[i] >= dev->real_num_tx_queues \|\|
	78	!qopt->count[i] \|\|
	79	last > dev->real_num_tx_queues)
	80	return -EINVAL;
	81
	82	/* Verify that the offset and counts do not overlap */
	83	for (j = i + 1; j < qopt->num_tc; j++) {
	84	if (last > qopt->offset[j])
	85	return -EINVAL;
	86	}
	87	}
	88
	89	return 0;
	90	}
	91
	92	static int mqprio_init(struct Qdisc sch, struct nlattr opt)
	93	{
	94	struct net_device *dev = qdisc_dev(sch);
	95	struct mqprio_sched *priv = qdisc_priv(sch);
	96	struct netdev_queue *dev_queue;
	97	struct Qdisc *qdisc;
	98	int i, err = -EOPNOTSUPP;
	99	struct tc_mqprio_qopt *qopt = NULL;
	100
	101	BUILD_BUG_ON(TC_MAX_QUEUE != TC_QOPT_MAX_QUEUE);
	102	BUILD_BUG_ON(TC_BITMASK != TC_QOPT_BITMASK);
	103
	104	if (sch->parent != TC_H_ROOT)
	105	return -EOPNOTSUPP;
	106
	107	if (!netif_is_multiqueue(dev))
	108	return -EOPNOTSUPP;
109
7838f2ce	110	if (!opt \|\| nla_len(opt) < sizeof(*qopt))
b8970f0b JF	111	return -EINVAL;
	112
	113	qopt = nla_data(opt);
	114	if (mqprio_parse_opt(dev, qopt))
	115	return -EINVAL;
	116
	117	/* pre-allocate qdisc, attachment can't fail */
	118	priv->qdiscs = kcalloc(dev->num_tx_queues, sizeof(priv->qdiscs[0]),
	119	GFP_KERNEL);
	120	if (priv->qdiscs == NULL) {
	121	err = -ENOMEM;
	122	goto err;
	123	}
	124
	125	for (i = 0; i < dev->num_tx_queues; i++) {
	126	dev_queue = netdev_get_tx_queue(dev, i);
	127	qdisc = qdisc_create_dflt(dev_queue, &pfifo_fast_ops,
	128	TC_H_MAKE(TC_H_MAJ(sch->handle),
	129	TC_H_MIN(i + 1)));
	130	if (qdisc == NULL) {
	131	err = -ENOMEM;
	132	goto err;
	133	}
b8970f0b JF	134	priv->qdiscs[i] = qdisc;
	135	}
	136
	137	/* If the mqprio options indicate that hardware should own
	138	* the queue mapping then run ndo_setup_tc otherwise use the
	139	* supplied and verified mapping
	140	*/
	141	if (qopt->hw) {
	142	priv->hw_owned = 1;
	143	err = dev->netdev_ops->ndo_setup_tc(dev, qopt->num_tc);
	144	if (err)
	145	goto err;
	146	} else {
	147	netdev_set_num_tc(dev, qopt->num_tc);
	148	for (i = 0; i < qopt->num_tc; i++)
	149	netdev_set_tc_queue(dev, i,
	150	qopt->count[i], qopt->offset[i]);
	151	}
	152
	153	/* Always use supplied priority mappings */
	154	for (i = 0; i < TC_BITMASK + 1; i++)
	155	netdev_set_prio_tc_map(dev, i, qopt->prio_tc_map[i]);
	156
	157	sch->flags \|= TCQ_F_MQROOT;
	158	return 0;
	159
	160	err:
	161	mqprio_destroy(sch);
	162	return err;
	163	}
	164
	165	static void mqprio_attach(struct Qdisc *sch)
	166	{
	167	struct net_device *dev = qdisc_dev(sch);
	168	struct mqprio_sched *priv = qdisc_priv(sch);
	169	struct Qdisc *qdisc;
	170	unsigned int ntx;
	171
	172	/* Attach underlying qdisc */
	173	for (ntx = 0; ntx < dev->num_tx_queues; ntx++) {
	174	qdisc = priv->qdiscs[ntx];
	175	qdisc = dev_graft_qdisc(qdisc->dev_queue, qdisc);
	176	if (qdisc)
	177	qdisc_destroy(qdisc);
	178	}
	179	kfree(priv->qdiscs);
	180	priv->qdiscs = NULL;
	181	}
	182
	183	static struct netdev_queue mqprio_queue_get(struct Qdisc sch,
	184	unsigned long cl)
	185	{
	186	struct net_device *dev = qdisc_dev(sch);
	187	unsigned long ntx = cl - 1 - netdev_get_num_tc(dev);
	188
	189	if (ntx >= dev->num_tx_queues)
	190	return NULL;
	191	return netdev_get_tx_queue(dev, ntx);
	192	}
	193
	194	static int mqprio_graft(struct Qdisc sch, unsigned long cl, struct Qdisc new,
	195	struct Qdisc **old)
	196	{
	197	struct net_device *dev = qdisc_dev(sch);
198	struct netdev_queue *dev_queue = mqprio_queue_get(sch, cl);
199
200	if (!dev_queue)
201	return -EINVAL;
202
203	if (dev->flags & IFF_UP)
204	dev_deactivate(dev);
205
206	*old = dev_graft_qdisc(dev_queue, new);
207
208	if (dev->flags & IFF_UP)
209	dev_activate(dev);
210
211	return 0;
212	}
213
214	static int mqprio_dump(struct Qdisc sch, struct sk_buff skb)
215	{
216	struct net_device *dev = qdisc_dev(sch);
217	struct mqprio_sched *priv = qdisc_priv(sch);
218	unsigned char *b = skb_tail_pointer(skb);
144ce879	219	struct tc_mqprio_qopt opt = { 0 };
b8970f0b JF	220	struct Qdisc *qdisc;
	221	unsigned int i;
	222
	223	sch->q.qlen = 0;
	224	memset(&sch->bstats, 0, sizeof(sch->bstats));
	225	memset(&sch->qstats, 0, sizeof(sch->qstats));
	226
	227	for (i = 0; i < dev->num_tx_queues; i++) {
	228	qdisc = netdev_get_tx_queue(dev, i)->qdisc;
	229	spin_lock_bh(qdisc_lock(qdisc));
	230	sch->q.qlen += qdisc->q.qlen;
	231	sch->bstats.bytes += qdisc->bstats.bytes;
	232	sch->bstats.packets += qdisc->bstats.packets;
	233	sch->qstats.qlen += qdisc->qstats.qlen;
	234	sch->qstats.backlog += qdisc->qstats.backlog;
	235	sch->qstats.drops += qdisc->qstats.drops;
	236	sch->qstats.requeues += qdisc->qstats.requeues;
	237	sch->qstats.overlimits += qdisc->qstats.overlimits;
	238	spin_unlock_bh(qdisc_lock(qdisc));
	239	}
	240
	241	opt.num_tc = netdev_get_num_tc(dev);
	242	memcpy(opt.prio_tc_map, dev->prio_tc_map, sizeof(opt.prio_tc_map));
	243	opt.hw = priv->hw_owned;
	244
	245	for (i = 0; i < netdev_get_num_tc(dev); i++) {
	246	opt.count[i] = dev->tc_to_txq[i].count;
	247	opt.offset[i] = dev->tc_to_txq[i].offset;
	248	}
	249
	250	NLA_PUT(skb, TCA_OPTIONS, sizeof(opt), &opt);
	251
	252	return skb->len;
	253	nla_put_failure:
	254	nlmsg_trim(skb, b);
	255	return -1;
	256	}
	257
	258	static struct Qdisc mqprio_leaf(struct Qdisc sch, unsigned long cl)
	259	{
	260	struct netdev_queue *dev_queue = mqprio_queue_get(sch, cl);
	261
	262	if (!dev_queue)
	263	return NULL;
	264
	265	return dev_queue->qdisc_sleeping;
	266	}
	267
	268	static unsigned long mqprio_get(struct Qdisc *sch, u32 classid)
	269	{
	270	struct net_device *dev = qdisc_dev(sch);
	271	unsigned int ntx = TC_H_MIN(classid);
	272
	273	if (ntx > dev->num_tx_queues + netdev_get_num_tc(dev))
	274	return 0;
	275	return ntx;
	276	}
	277
	278	static void mqprio_put(struct Qdisc *sch, unsigned long cl)
	279	{
	280	}
	281
	282	static int mqprio_dump_class(struct Qdisc *sch, unsigned long cl,
	283	struct sk_buff skb, struct tcmsg tcm)
284	{
285	struct net_device *dev = qdisc_dev(sch);
286
287	if (cl <= netdev_get_num_tc(dev)) {
288	tcm->tcm_parent = TC_H_ROOT;
289	tcm->tcm_info = 0;
290	} else {
291	int i;
292	struct netdev_queue *dev_queue;
293
294	dev_queue = mqprio_queue_get(sch, cl);
295	tcm->tcm_parent = 0;
296	for (i = 0; i < netdev_get_num_tc(dev); i++) {
297	struct netdev_tc_txq tc = dev->tc_to_txq[i];
298	int q_idx = cl - netdev_get_num_tc(dev);
299
300	if (q_idx > tc.offset &&
301	q_idx <= tc.offset + tc.count) {
302	tcm->tcm_parent =
303	TC_H_MAKE(TC_H_MAJ(sch->handle),
304	TC_H_MIN(i + 1));
305	break;
306	}
307	}
308	tcm->tcm_info = dev_queue->qdisc_sleeping->handle;
309	}
310	tcm->tcm_handle \|= TC_H_MIN(cl);
311	return 0;
312	}
313
314	static int mqprio_dump_class_stats(struct Qdisc *sch, unsigned long cl,
ea18fd95	315	struct gnet_dump *d)
	316	__releases(d->lock)
	317	__acquires(d->lock)
b8970f0b JF	318	{
	319	struct net_device *dev = qdisc_dev(sch);
	320
	321	if (cl <= netdev_get_num_tc(dev)) {
	322	int i;
	323	struct Qdisc *qdisc;
	324	struct gnet_stats_queue qstats = {0};
	325	struct gnet_stats_basic_packed bstats = {0};
	326	struct netdev_tc_txq tc = dev->tc_to_txq[cl - 1];
	327
	328	/* Drop lock here it will be reclaimed before touching
	329	* statistics this is required because the d->lock we
	330	* hold here is the look on dev_queue->qdisc_sleeping
	331	* also acquired below.
	332	*/
	333	spin_unlock_bh(d->lock);
	334
	335	for (i = tc.offset; i < tc.offset + tc.count; i++) {
	336	qdisc = netdev_get_tx_queue(dev, i)->qdisc;
	337	spin_lock_bh(qdisc_lock(qdisc));
	338	bstats.bytes += qdisc->bstats.bytes;
	339	bstats.packets += qdisc->bstats.packets;
	340	qstats.qlen += qdisc->qstats.qlen;
	341	qstats.backlog += qdisc->qstats.backlog;
	342	qstats.drops += qdisc->qstats.drops;
	343	qstats.requeues += qdisc->qstats.requeues;
	344	qstats.overlimits += qdisc->qstats.overlimits;
	345	spin_unlock_bh(qdisc_lock(qdisc));
	346	}
	347	/* Reclaim root sleeping lock before completing stats */
	348	spin_lock_bh(d->lock);
	349	if (gnet_stats_copy_basic(d, &bstats) < 0 \|\|
	350	gnet_stats_copy_queue(d, &qstats) < 0)
	351	return -1;
	352	} else {
	353	struct netdev_queue *dev_queue = mqprio_queue_get(sch, cl);
	354
	355	sch = dev_queue->qdisc_sleeping;
	356	sch->qstats.qlen = sch->q.qlen;
	357	if (gnet_stats_copy_basic(d, &sch->bstats) < 0 \|\|
	358	gnet_stats_copy_queue(d, &sch->qstats) < 0)
	359	return -1;
	360	}
	361	return 0;
	362	}
	363
	364	static void mqprio_walk(struct Qdisc sch, struct qdisc_walker arg)
	365	{
	366	struct net_device *dev = qdisc_dev(sch);
	367	unsigned long ntx;
	368
	369	if (arg->stop)
	370	return;
	371
	372	/* Walk hierarchy with a virtual class per tc */
	373	arg->count = arg->skip;
	374	for (ntx = arg->skip;
	375	ntx < dev->num_tx_queues + netdev_get_num_tc(dev);
	376	ntx++) {
	377	if (arg->fn(sch, ntx + 1, arg) < 0) {
	378	arg->stop = 1;
	379	break;
	380	}
	381	arg->count++;
382	}
383	}
384
385	static const struct Qdisc_class_ops mqprio_class_ops = {
386	.graft = mqprio_graft,
387	.leaf = mqprio_leaf,
388	.get = mqprio_get,
389	.put = mqprio_put,
390	.walk = mqprio_walk,
391	.dump = mqprio_dump_class,
392	.dump_stats = mqprio_dump_class_stats,
393	};
394
ea18fd95	395	static struct Qdisc_ops mqprio_qdisc_ops __read_mostly = {
b8970f0b JF	396	.cl_ops = &mqprio_class_ops,
	397	.id = "mqprio",
	398	.priv_size = sizeof(struct mqprio_sched),
	399	.init = mqprio_init,
	400	.destroy = mqprio_destroy,
	401	.attach = mqprio_attach,
	402	.dump = mqprio_dump,
	403	.owner = THIS_MODULE,
	404	};
	405
	406	static int __init mqprio_module_init(void)
	407	{
	408	return register_qdisc(&mqprio_qdisc_ops);
	409	}
	410
	411	static void __exit mqprio_module_exit(void)
	412	{
	413	unregister_qdisc(&mqprio_qdisc_ops);
	414	}
	415
	416	module_init(mqprio_module_init);
	417	module_exit(mqprio_module_exit);
	418
	419	MODULE_LICENSE("GPL");