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

// SPDX-License-Identifier: GPL-2.0

/* net/sched/sch_etf.c  Earliest TxTime First queueing discipline.
 *
 * Authors:	Jesus Sanchez-Palencia <jesus.sanchez-palencia@intel.com>
 *		Vinicius Costa Gomes <vinicius.gomes@intel.com>
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/errno.h>
#include <linux/errqueue.h>
#include <linux/rbtree.h>
#include <linux/skbuff.h>
#include <linux/posix-timers.h>
#include <net/netlink.h>
#include <net/sch_generic.h>
#include <net/pkt_sched.h>
#include <net/sock.h>

#define DEADLINE_MODE_IS_ON(x) ((x)->flags & TC_ETF_DEADLINE_MODE_ON)
#define OFFLOAD_IS_ON(x) ((x)->flags & TC_ETF_OFFLOAD_ON)

struct etf_sched_data {
	bool offload;
	bool deadline_mode;
	int clockid;
	int queue;
	s32 delta; /* in ns */
	ktime_t last; /* The txtime of the last skb sent to the netdevice. */
	struct rb_root_cached head;
	struct qdisc_watchdog watchdog;
	ktime_t (*get_time)(void);
};

static const struct nla_policy etf_policy[TCA_ETF_MAX + 1] = {
	[TCA_ETF_PARMS]	= { .len = sizeof(struct tc_etf_qopt) },
};

static inline int validate_input_params(struct tc_etf_qopt *qopt,
					struct netlink_ext_ack *extack)
{
	/* Check if params comply to the following rules:
	 *	* Clockid and delta must be valid.
	 *
	 *	* Dynamic clockids are not supported.
	 *
	 *	* Delta must be a positive integer.
	 *
	 * Also note that for the HW offload case, we must
	 * expect that system clocks have been synchronized to PHC.
	 */
	if (qopt->clockid < 0) {
		NL_SET_ERR_MSG(extack, "Dynamic clockids are not supported");
		return -ENOTSUPP;
	}

	if (qopt->clockid != CLOCK_TAI) {
		NL_SET_ERR_MSG(extack, "Invalid clockid. CLOCK_TAI must be used");
		return -EINVAL;
	}

	if (qopt->delta < 0) {
		NL_SET_ERR_MSG(extack, "Delta must be positive");
		return -EINVAL;
	}

	return 0;
}

static bool is_packet_valid(struct Qdisc *sch, struct sk_buff *nskb)
{
	struct etf_sched_data *q = qdisc_priv(sch);
	ktime_t txtime = nskb->tstamp;
	struct sock *sk = nskb->sk;
	ktime_t now;

	if (!sk)
		return false;

	if (!sock_flag(sk, SOCK_TXTIME))
		return false;

	/* We don't perform crosstimestamping.
	 * Drop if packet's clockid differs from qdisc's.
	 */
	if (sk->sk_clockid != q->clockid)
		return false;

	if (sk->sk_txtime_deadline_mode != q->deadline_mode)
		return false;

	now = q->get_time();
	if (ktime_before(txtime, now) || ktime_before(txtime, q->last))
		return false;

	return true;
}

static struct sk_buff *etf_peek_timesortedlist(struct Qdisc *sch)
{
	struct etf_sched_data *q = qdisc_priv(sch);
	struct rb_node *p;

	p = rb_first_cached(&q->head);
	if (!p)
		return NULL;

	return rb_to_skb(p);
}

static void reset_watchdog(struct Qdisc *sch)
{
	struct etf_sched_data *q = qdisc_priv(sch);
	struct sk_buff *skb = etf_peek_timesortedlist(sch);
	ktime_t next;

	if (!skb) {
		qdisc_watchdog_cancel(&q->watchdog);
		return;
	}

	next = ktime_sub_ns(skb->tstamp, q->delta);
	qdisc_watchdog_schedule_ns(&q->watchdog, ktime_to_ns(next));
}

static void report_sock_error(struct sk_buff *skb, u32 err, u8 code)
{
	struct sock_exterr_skb *serr;
	struct sk_buff *clone;
	ktime_t txtime = skb->tstamp;

	if (!skb->sk || !(skb->sk->sk_txtime_report_errors))
		return;

	clone = skb_clone(skb, GFP_ATOMIC);
	if (!clone)
		return;

	serr = SKB_EXT_ERR(clone);
	serr->ee.ee_errno = err;
	serr->ee.ee_origin = SO_EE_ORIGIN_TXTIME;
	serr->ee.ee_type = 0;
	serr->ee.ee_code = code;
	serr->ee.ee_pad = 0;
	serr->ee.ee_data = (txtime >> 32); /* high part of tstamp */
	serr->ee.ee_info = txtime; /* low part of tstamp */

	if (sock_queue_err_skb(skb->sk, clone))
		kfree_skb(clone);
}

static int etf_enqueue_timesortedlist(struct sk_buff *nskb, struct Qdisc *sch,
				      struct sk_buff **to_free)
{
	struct etf_sched_data *q = qdisc_priv(sch);
	struct rb_node **p = &q->head.rb_root.rb_node, *parent = NULL;
	ktime_t txtime = nskb->tstamp;
	bool leftmost = true;

	if (!is_packet_valid(sch, nskb)) {
		report_sock_error(nskb, EINVAL,
				  SO_EE_CODE_TXTIME_INVALID_PARAM);
		return qdisc_drop(nskb, sch, to_free);
	}

	while (*p) {
		struct sk_buff *skb;

		parent = *p;
		skb = rb_to_skb(parent);
		if (ktime_after(txtime, skb->tstamp)) {
			p = &parent->rb_right;
			leftmost = false;
		} else {
			p = &parent->rb_left;
		}
	}
	rb_link_node(&nskb->rbnode, parent, p);
	rb_insert_color_cached(&nskb->rbnode, &q->head, leftmost);

	qdisc_qstats_backlog_inc(sch, nskb);
	sch->q.qlen++;

	/* Now we may need to re-arm the qdisc watchdog for the next packet. */
	reset_watchdog(sch);

	return NET_XMIT_SUCCESS;
}

static void timesortedlist_drop(struct Qdisc *sch, struct sk_buff *skb,
				ktime_t now)
{
	struct etf_sched_data *q = qdisc_priv(sch);
	struct sk_buff *to_free = NULL;
	struct sk_buff *tmp = NULL;

	skb_rbtree_walk_from_safe(skb, tmp) {
		if (ktime_after(skb->tstamp, now))
			break;

		rb_erase_cached(&skb->rbnode, &q->head);

		/* The rbnode field in the skb re-uses these fields, now that
		 * we are done with the rbnode, reset them.
		 */
		skb->next = NULL;
		skb->prev = NULL;
		skb->dev = qdisc_dev(sch);

		report_sock_error(skb, ECANCELED, SO_EE_CODE_TXTIME_MISSED);

		qdisc_qstats_backlog_dec(sch, skb);
		qdisc_drop(skb, sch, &to_free);
		qdisc_qstats_overlimit(sch);
		sch->q.qlen--;
	}

	kfree_skb_list(to_free);
}

static void timesortedlist_remove(struct Qdisc *sch, struct sk_buff *skb)
{
	struct etf_sched_data *q = qdisc_priv(sch);

	rb_erase_cached(&skb->rbnode, &q->head);

	/* The rbnode field in the skb re-uses these fields, now that
	 * we are done with the rbnode, reset them.
	 */
	skb->next = NULL;
	skb->prev = NULL;
	skb->dev = qdisc_dev(sch);

	qdisc_qstats_backlog_dec(sch, skb);

	qdisc_bstats_update(sch, skb);

	q->last = skb->tstamp;

	sch->q.qlen--;
}

static struct sk_buff *etf_dequeue_timesortedlist(struct Qdisc *sch)
{
	struct etf_sched_data *q = qdisc_priv(sch);
	struct sk_buff *skb;
	ktime_t now, next;

	skb = etf_peek_timesortedlist(sch);
	if (!skb)
		return NULL;

	now = q->get_time();

	/* Drop if packet has expired while in queue. */
	if (ktime_before(skb->tstamp, now)) {
		timesortedlist_drop(sch, skb, now);
		skb = NULL;
		goto out;
	}

	/* When in deadline mode, dequeue as soon as possible and change the
	 * txtime from deadline to (now + delta).
	 */
	if (q->deadline_mode) {
		timesortedlist_remove(sch, skb);
		skb->tstamp = now;
		goto out;
	}

	next = ktime_sub_ns(skb->tstamp, q->delta);

	/* Dequeue only if now is within the [txtime - delta, txtime] range. */
	if (ktime_after(now, next))
		timesortedlist_remove(sch, skb);
	else
		skb = NULL;

out:
	/* Now we may need to re-arm the qdisc watchdog for the next packet. */
	reset_watchdog(sch);

	return skb;
}

static void etf_disable_offload(struct net_device *dev,
				struct etf_sched_data *q)
{
	struct tc_etf_qopt_offload etf = { };
	const struct net_device_ops *ops;
	int err;

	if (!q->offload)
		return;

	ops = dev->netdev_ops;
	if (!ops->ndo_setup_tc)
		return;

	etf.queue = q->queue;
	etf.enable = 0;

	err = ops->ndo_setup_tc(dev, TC_SETUP_QDISC_ETF, &etf);
	if (err < 0)
		pr_warn("Couldn't disable ETF offload for queue %d\n",
			etf.queue);
}

static int etf_enable_offload(struct net_device *dev, struct etf_sched_data *q,
			      struct netlink_ext_ack *extack)
{
	const struct net_device_ops *ops = dev->netdev_ops;
	struct tc_etf_qopt_offload etf = { };
	int err;

	if (q->offload)
		return 0;

	if (!ops->ndo_setup_tc) {
		NL_SET_ERR_MSG(extack, "Specified device does not support ETF offload");
		return -EOPNOTSUPP;
	}

	etf.queue = q->queue;
	etf.enable = 1;

	err = ops->ndo_setup_tc(dev, TC_SETUP_QDISC_ETF, &etf);
	if (err < 0) {
		NL_SET_ERR_MSG(extack, "Specified device failed to setup ETF hardware offload");
		return err;
	}

	return 0;
}

static int etf_init(struct Qdisc *sch, struct nlattr *opt,
		    struct netlink_ext_ack *extack)
{
	struct etf_sched_data *q = qdisc_priv(sch);
	struct net_device *dev = qdisc_dev(sch);
	struct nlattr *tb[TCA_ETF_MAX + 1];
	struct tc_etf_qopt *qopt;
	int err;

	if (!opt) {
		NL_SET_ERR_MSG(extack,
			       "Missing ETF qdisc options which are mandatory");
		return -EINVAL;
	}

	err = nla_parse_nested(tb, TCA_ETF_MAX, opt, etf_policy, extack);
	if (err < 0)
		return err;

	if (!tb[TCA_ETF_PARMS]) {
		NL_SET_ERR_MSG(extack, "Missing mandatory ETF parameters");
		return -EINVAL;
	}

	qopt = nla_data(tb[TCA_ETF_PARMS]);

	pr_debug("delta %d clockid %d offload %s deadline %s\n",
		 qopt->delta, qopt->clockid,
		 OFFLOAD_IS_ON(qopt) ? "on" : "off",
		 DEADLINE_MODE_IS_ON(qopt) ? "on" : "off");

	err = validate_input_params(qopt, extack);
	if (err < 0)
		return err;

	q->queue = sch->dev_queue - netdev_get_tx_queue(dev, 0);

	if (OFFLOAD_IS_ON(qopt)) {
		err = etf_enable_offload(dev, q, extack);
		if (err < 0)
			return err;
	}

	/* Everything went OK, save the parameters used. */
	q->delta = qopt->delta;
	q->clockid = qopt->clockid;
	q->offload = OFFLOAD_IS_ON(qopt);
	q->deadline_mode = DEADLINE_MODE_IS_ON(qopt);

	switch (q->clockid) {
	case CLOCK_REALTIME:
		q->get_time = ktime_get_real;
		break;
	case CLOCK_MONOTONIC:
		q->get_time = ktime_get;
		break;
	case CLOCK_BOOTTIME:
		q->get_time = ktime_get_boottime;
		break;
	case CLOCK_TAI:
		q->get_time = ktime_get_clocktai;
		break;
	default:
		NL_SET_ERR_MSG(extack, "Clockid is not supported");
		return -ENOTSUPP;
	}

	qdisc_watchdog_init_clockid(&q->watchdog, sch, q->clockid);

	return 0;
}

static void timesortedlist_clear(struct Qdisc *sch)
{
	struct etf_sched_data *q = qdisc_priv(sch);
	struct rb_node *p = rb_first_cached(&q->head);

	while (p) {
		struct sk_buff *skb = rb_to_skb(p);

		p = rb_next(p);

		rb_erase_cached(&skb->rbnode, &q->head);
		rtnl_kfree_skbs(skb, skb);
		sch->q.qlen--;
	}
}

static void etf_reset(struct Qdisc *sch)
{
	struct etf_sched_data *q = qdisc_priv(sch);

	/* Only cancel watchdog if it's been initialized. */
	if (q->watchdog.qdisc == sch)
		qdisc_watchdog_cancel(&q->watchdog);

	/* No matter which mode we are on, it's safe to clear both lists. */
	timesortedlist_clear(sch);
	__qdisc_reset_queue(&sch->q);

	sch->qstats.backlog = 0;
	sch->q.qlen = 0;

	q->last = 0;
}

static void etf_destroy(struct Qdisc *sch)
{
	struct etf_sched_data *q = qdisc_priv(sch);
	struct net_device *dev = qdisc_dev(sch);

	/* Only cancel watchdog if it's been initialized. */
	if (q->watchdog.qdisc == sch)
		qdisc_watchdog_cancel(&q->watchdog);

	etf_disable_offload(dev, q);
}

static int etf_dump(struct Qdisc *sch, struct sk_buff *skb)
{
	struct etf_sched_data *q = qdisc_priv(sch);
	struct tc_etf_qopt opt = { };
	struct nlattr *nest;

	nest = nla_nest_start(skb, TCA_OPTIONS);
	if (!nest)
		goto nla_put_failure;

	opt.delta = q->delta;
	opt.clockid = q->clockid;
	if (q->offload)
		opt.flags |= TC_ETF_OFFLOAD_ON;

	if (q->deadline_mode)
		opt.flags |= TC_ETF_DEADLINE_MODE_ON;

	if (nla_put(skb, TCA_ETF_PARMS, sizeof(opt), &opt))
		goto nla_put_failure;

	return nla_nest_end(skb, nest);

nla_put_failure:
	nla_nest_cancel(skb, nest);
	return -1;
}

static struct Qdisc_ops etf_qdisc_ops __read_mostly = {
	.id		=	"etf",
	.priv_size	=	sizeof(struct etf_sched_data),
	.enqueue	=	etf_enqueue_timesortedlist,
	.dequeue	=	etf_dequeue_timesortedlist,
	.peek		=	etf_peek_timesortedlist,
	.init		=	etf_init,
	.reset		=	etf_reset,
	.destroy	=	etf_destroy,
	.dump		=	etf_dump,
	.owner		=	THIS_MODULE,
};

static int __init etf_module_init(void)
{
	return register_qdisc(&etf_qdisc_ops);
}

static void __exit etf_module_exit(void)
{
	unregister_qdisc(&etf_qdisc_ops);
}
module_init(etf_module_init)
module_exit(etf_module_exit)
MODULE_LICENSE("GPL");
Commit	Line	Data
25db26a9 VCG	1	// SPDX-License-Identifier: GPL-2.0
	2
	3	/* net/sched/sch_etf.c Earliest TxTime First queueing discipline.
	4	*
	5	* Authors: Jesus Sanchez-Palencia <jesus.sanchez-palencia@intel.com>
	6	* Vinicius Costa Gomes <vinicius.gomes@intel.com>
	7	*/
	8
	9	#include <linux/module.h>
	10	#include <linux/types.h>
	11	#include <linux/kernel.h>
	12	#include <linux/string.h>
	13	#include <linux/errno.h>
4b15c707	14	#include <linux/errqueue.h>
25db26a9 VCG	15	#include <linux/rbtree.h>
	16	#include <linux/skbuff.h>
	17	#include <linux/posix-timers.h>
	18	#include <net/netlink.h>
	19	#include <net/sch_generic.h>
	20	#include <net/pkt_sched.h>
	21	#include <net/sock.h>
	22
	23	#define DEADLINE_MODE_IS_ON(x) ((x)->flags & TC_ETF_DEADLINE_MODE_ON)
88cab771	24	#define OFFLOAD_IS_ON(x) ((x)->flags & TC_ETF_OFFLOAD_ON)
25db26a9 VCG	25
25db26a9 VCG	26	struct etf_sched_data {
88cab771	27	bool offload;
25db26a9 VCG	28	bool deadline_mode;
	29	int clockid;
	30	int queue;
	31	s32 delta; /* in ns */
	32	ktime_t last; /* The txtime of the last skb sent to the netdevice. */
09fd4860	33	struct rb_root_cached head;
25db26a9 VCG	34	struct qdisc_watchdog watchdog;
	35	ktime_t (*get_time)(void);
	36	};
	37
	38	static const struct nla_policy etf_policy[TCA_ETF_MAX + 1] = {
	39	[TCA_ETF_PARMS] = { .len = sizeof(struct tc_etf_qopt) },
	40	};
	41
	42	static inline int validate_input_params(struct tc_etf_qopt *qopt,
	43	struct netlink_ext_ack *extack)
	44	{
	45	/* Check if params comply to the following rules:
	46	* * Clockid and delta must be valid.
	47	*
	48	* * Dynamic clockids are not supported.
	49	*
	50	* * Delta must be a positive integer.
88cab771 JSP	51	*
	52	* Also note that for the HW offload case, we must
	53	* expect that system clocks have been synchronized to PHC.
25db26a9 VCG	54	*/
	55	if (qopt->clockid < 0) {
	56	NL_SET_ERR_MSG(extack, "Dynamic clockids are not supported");
	57	return -ENOTSUPP;
	58	}
	59
	60	if (qopt->clockid != CLOCK_TAI) {
	61	NL_SET_ERR_MSG(extack, "Invalid clockid. CLOCK_TAI must be used");
	62	return -EINVAL;
	63	}
	64
	65	if (qopt->delta < 0) {
	66	NL_SET_ERR_MSG(extack, "Delta must be positive");
	67	return -EINVAL;
	68	}
	69
	70	return 0;
	71	}
	72
	73	static bool is_packet_valid(struct Qdisc sch, struct sk_buff nskb)
	74	{
	75	struct etf_sched_data *q = qdisc_priv(sch);
	76	ktime_t txtime = nskb->tstamp;
	77	struct sock *sk = nskb->sk;
	78	ktime_t now;
	79
	80	if (!sk)
	81	return false;
	82
	83	if (!sock_flag(sk, SOCK_TXTIME))
	84	return false;
	85
	86	/* We don't perform crosstimestamping.
	87	* Drop if packet's clockid differs from qdisc's.
	88	*/
	89	if (sk->sk_clockid != q->clockid)
	90	return false;
	91
	92	if (sk->sk_txtime_deadline_mode != q->deadline_mode)
	93	return false;
	94
	95	now = q->get_time();
	96	if (ktime_before(txtime, now) \|\| ktime_before(txtime, q->last))
	97	return false;
	98
	99	return true;
	100	}
	101
	102	static struct sk_buff etf_peek_timesortedlist(struct Qdisc sch)
	103	{
	104	struct etf_sched_data *q = qdisc_priv(sch);
	105	struct rb_node *p;
	106
09fd4860	107	p = rb_first_cached(&q->head);
25db26a9 VCG	108	if (!p)
	109	return NULL;
	110
	111	return rb_to_skb(p);
	112	}
	113
	114	static void reset_watchdog(struct Qdisc *sch)
	115	{
	116	struct etf_sched_data *q = qdisc_priv(sch);
	117	struct sk_buff *skb = etf_peek_timesortedlist(sch);
	118	ktime_t next;
	119
3fcbdaee JSP	120	if (!skb) {
3fcbdaee JSP	121	qdisc_watchdog_cancel(&q->watchdog);
25db26a9	122	return;
3fcbdaee	123	}
25db26a9 VCG	124
	125	next = ktime_sub_ns(skb->tstamp, q->delta);
	126	qdisc_watchdog_schedule_ns(&q->watchdog, ktime_to_ns(next));
	127	}
	128
4b15c707 JSP	129	static void report_sock_error(struct sk_buff *skb, u32 err, u8 code)
	130	{
	131	struct sock_exterr_skb *serr;
	132	struct sk_buff *clone;
	133	ktime_t txtime = skb->tstamp;
	134
	135	if (!skb->sk \|\| !(skb->sk->sk_txtime_report_errors))
	136	return;
	137
	138	clone = skb_clone(skb, GFP_ATOMIC);
	139	if (!clone)
	140	return;
	141
	142	serr = SKB_EXT_ERR(clone);
	143	serr->ee.ee_errno = err;
	144	serr->ee.ee_origin = SO_EE_ORIGIN_TXTIME;
	145	serr->ee.ee_type = 0;
	146	serr->ee.ee_code = code;
	147	serr->ee.ee_pad = 0;
	148	serr->ee.ee_data = (txtime >> 32); /* high part of tstamp */
	149	serr->ee.ee_info = txtime; /* low part of tstamp */
	150
	151	if (sock_queue_err_skb(skb->sk, clone))
	152	kfree_skb(clone);
	153	}
	154
25db26a9 VCG	155	static int etf_enqueue_timesortedlist(struct sk_buff nskb, struct Qdisc sch,
	156	struct sk_buff **to_free)
	157	{
	158	struct etf_sched_data *q = qdisc_priv(sch);
09fd4860	159	struct rb_node *p = &q->head.rb_root.rb_node, parent = NULL;
25db26a9	160	ktime_t txtime = nskb->tstamp;
09fd4860	161	bool leftmost = true;
25db26a9	162
4b15c707 JSP	163	if (!is_packet_valid(sch, nskb)) {
	164	report_sock_error(nskb, EINVAL,
	165	SO_EE_CODE_TXTIME_INVALID_PARAM);
25db26a9	166	return qdisc_drop(nskb, sch, to_free);
4b15c707	167	}
25db26a9 VCG	168
	169	while (*p) {
	170	struct sk_buff *skb;
	171
	172	parent = *p;
	173	skb = rb_to_skb(parent);
09fd4860	174	if (ktime_after(txtime, skb->tstamp)) {
25db26a9	175	p = &parent->rb_right;
09fd4860 JSP	176	leftmost = false;
09fd4860 JSP	177	} else {
25db26a9	178	p = &parent->rb_left;
09fd4860	179	}
25db26a9 VCG	180	}
25db26a9 VCG	181	rb_link_node(&nskb->rbnode, parent, p);
09fd4860	182	rb_insert_color_cached(&nskb->rbnode, &q->head, leftmost);
25db26a9 VCG	183
	184	qdisc_qstats_backlog_inc(sch, nskb);
	185	sch->q.qlen++;
	186
	187	/* Now we may need to re-arm the qdisc watchdog for the next packet. */
	188	reset_watchdog(sch);
	189
	190	return NET_XMIT_SUCCESS;
	191	}
	192
37342bda JSP	193	static void timesortedlist_drop(struct Qdisc sch, struct sk_buff skb,
37342bda JSP	194	ktime_t now)
25db26a9 VCG	195	{
25db26a9 VCG	196	struct etf_sched_data *q = qdisc_priv(sch);
cbeeb8ef	197	struct sk_buff *to_free = NULL;
37342bda	198	struct sk_buff *tmp = NULL;
25db26a9	199
37342bda JSP	200	skb_rbtree_walk_from_safe(skb, tmp) {
	201	if (ktime_after(skb->tstamp, now))
	202	break;
25db26a9	203
37342bda	204	rb_erase_cached(&skb->rbnode, &q->head);
25db26a9	205
37342bda JSP	206	/* The rbnode field in the skb re-uses these fields, now that
	207	* we are done with the rbnode, reset them.
	208	*/
	209	skb->next = NULL;
	210	skb->prev = NULL;
	211	skb->dev = qdisc_dev(sch);
25db26a9	212
37342bda	213	report_sock_error(skb, ECANCELED, SO_EE_CODE_TXTIME_MISSED);
25db26a9	214
37342bda JSP	215	qdisc_qstats_backlog_dec(sch, skb);
	216	qdisc_drop(skb, sch, &to_free);
	217	qdisc_qstats_overlimit(sch);
	218	sch->q.qlen--;
	219	}
4b15c707	220
37342bda	221	kfree_skb_list(to_free);
cbeeb8ef	222	}
25db26a9	223
cbeeb8ef JSP	224	static void timesortedlist_remove(struct Qdisc sch, struct sk_buff skb)
	225	{
	226	struct etf_sched_data *q = qdisc_priv(sch);
	227
	228	rb_erase_cached(&skb->rbnode, &q->head);
	229
	230	/* The rbnode field in the skb re-uses these fields, now that
	231	* we are done with the rbnode, reset them.
	232	*/
	233	skb->next = NULL;
	234	skb->prev = NULL;
	235	skb->dev = qdisc_dev(sch);
	236
	237	qdisc_qstats_backlog_dec(sch, skb);
	238
	239	qdisc_bstats_update(sch, skb);
	240
	241	q->last = skb->tstamp;
25db26a9 VCG	242
	243	sch->q.qlen--;
	244	}
	245
	246	static struct sk_buff etf_dequeue_timesortedlist(struct Qdisc sch)
	247	{
	248	struct etf_sched_data *q = qdisc_priv(sch);
	249	struct sk_buff *skb;
	250	ktime_t now, next;
	251
	252	skb = etf_peek_timesortedlist(sch);
	253	if (!skb)
	254	return NULL;
	255
	256	now = q->get_time();
	257
	258	/* Drop if packet has expired while in queue. */
25db26a9	259	if (ktime_before(skb->tstamp, now)) {
37342bda	260	timesortedlist_drop(sch, skb, now);
25db26a9 VCG	261	skb = NULL;
	262	goto out;
	263	}
	264
	265	/* When in deadline mode, dequeue as soon as possible and change the
	266	* txtime from deadline to (now + delta).
	267	*/
	268	if (q->deadline_mode) {
cbeeb8ef	269	timesortedlist_remove(sch, skb);
25db26a9 VCG	270	skb->tstamp = now;
	271	goto out;
	272	}
	273
	274	next = ktime_sub_ns(skb->tstamp, q->delta);
	275
	276	/* Dequeue only if now is within the [txtime - delta, txtime] range. */
	277	if (ktime_after(now, next))
cbeeb8ef	278	timesortedlist_remove(sch, skb);
25db26a9 VCG	279	else
	280	skb = NULL;
	281
	282	out:
	283	/* Now we may need to re-arm the qdisc watchdog for the next packet. */
	284	reset_watchdog(sch);
	285
	286	return skb;
	287	}
	288
88cab771 JSP	289	static void etf_disable_offload(struct net_device *dev,
	290	struct etf_sched_data *q)
	291	{
	292	struct tc_etf_qopt_offload etf = { };
	293	const struct net_device_ops *ops;
	294	int err;
	295
	296	if (!q->offload)
	297	return;
	298
	299	ops = dev->netdev_ops;
	300	if (!ops->ndo_setup_tc)
	301	return;
	302
	303	etf.queue = q->queue;
	304	etf.enable = 0;
	305
	306	err = ops->ndo_setup_tc(dev, TC_SETUP_QDISC_ETF, &etf);
	307	if (err < 0)
	308	pr_warn("Couldn't disable ETF offload for queue %d\n",
	309	etf.queue);
	310	}
	311
	312	static int etf_enable_offload(struct net_device dev, struct etf_sched_data q,
	313	struct netlink_ext_ack *extack)
	314	{
	315	const struct net_device_ops *ops = dev->netdev_ops;
	316	struct tc_etf_qopt_offload etf = { };
	317	int err;
	318
	319	if (q->offload)
	320	return 0;
	321
	322	if (!ops->ndo_setup_tc) {
	323	NL_SET_ERR_MSG(extack, "Specified device does not support ETF offload");
	324	return -EOPNOTSUPP;
	325	}
	326
	327	etf.queue = q->queue;
	328	etf.enable = 1;
	329
	330	err = ops->ndo_setup_tc(dev, TC_SETUP_QDISC_ETF, &etf);
	331	if (err < 0) {
	332	NL_SET_ERR_MSG(extack, "Specified device failed to setup ETF hardware offload");
	333	return err;
	334	}
	335
	336	return 0;
	337	}
	338
25db26a9 VCG	339	static int etf_init(struct Qdisc sch, struct nlattr opt,
	340	struct netlink_ext_ack *extack)
	341	{
	342	struct etf_sched_data *q = qdisc_priv(sch);
	343	struct net_device *dev = qdisc_dev(sch);
	344	struct nlattr *tb[TCA_ETF_MAX + 1];
	345	struct tc_etf_qopt *qopt;
	346	int err;
	347
	348	if (!opt) {
	349	NL_SET_ERR_MSG(extack,
	350	"Missing ETF qdisc options which are mandatory");
	351	return -EINVAL;
	352	}
	353
	354	err = nla_parse_nested(tb, TCA_ETF_MAX, opt, etf_policy, extack);
	355	if (err < 0)
	356	return err;
	357
	358	if (!tb[TCA_ETF_PARMS]) {
	359	NL_SET_ERR_MSG(extack, "Missing mandatory ETF parameters");
	360	return -EINVAL;
	361	}
	362
	363	qopt = nla_data(tb[TCA_ETF_PARMS]);
	364
88cab771	365	pr_debug("delta %d clockid %d offload %s deadline %s\n",
25db26a9	366	qopt->delta, qopt->clockid,
88cab771	367	OFFLOAD_IS_ON(qopt) ? "on" : "off",
25db26a9 VCG	368	DEADLINE_MODE_IS_ON(qopt) ? "on" : "off");
	369
	370	err = validate_input_params(qopt, extack);
	371	if (err < 0)
	372	return err;
	373
	374	q->queue = sch->dev_queue - netdev_get_tx_queue(dev, 0);
	375
88cab771 JSP	376	if (OFFLOAD_IS_ON(qopt)) {
	377	err = etf_enable_offload(dev, q, extack);
	378	if (err < 0)
	379	return err;
	380	}
	381
25db26a9 VCG	382	/* Everything went OK, save the parameters used. */
	383	q->delta = qopt->delta;
	384	q->clockid = qopt->clockid;
88cab771	385	q->offload = OFFLOAD_IS_ON(qopt);
25db26a9 VCG	386	q->deadline_mode = DEADLINE_MODE_IS_ON(qopt);
	387
	388	switch (q->clockid) {
	389	case CLOCK_REALTIME:
	390	q->get_time = ktime_get_real;
	391	break;
	392	case CLOCK_MONOTONIC:
	393	q->get_time = ktime_get;
	394	break;
	395	case CLOCK_BOOTTIME:
	396	q->get_time = ktime_get_boottime;
	397	break;
	398	case CLOCK_TAI:
	399	q->get_time = ktime_get_clocktai;
	400	break;
	401	default:
	402	NL_SET_ERR_MSG(extack, "Clockid is not supported");
	403	return -ENOTSUPP;
	404	}
	405
	406	qdisc_watchdog_init_clockid(&q->watchdog, sch, q->clockid);
	407
	408	return 0;
	409	}
	410
	411	static void timesortedlist_clear(struct Qdisc *sch)
	412	{
	413	struct etf_sched_data *q = qdisc_priv(sch);
09fd4860	414	struct rb_node *p = rb_first_cached(&q->head);
25db26a9 VCG	415
	416	while (p) {
	417	struct sk_buff *skb = rb_to_skb(p);
	418
	419	p = rb_next(p);
	420
09fd4860	421	rb_erase_cached(&skb->rbnode, &q->head);
25db26a9 VCG	422	rtnl_kfree_skbs(skb, skb);
	423	sch->q.qlen--;
	424	}
	425	}
	426
	427	static void etf_reset(struct Qdisc *sch)
	428	{
	429	struct etf_sched_data *q = qdisc_priv(sch);
	430
	431	/* Only cancel watchdog if it's been initialized. */
	432	if (q->watchdog.qdisc == sch)
	433	qdisc_watchdog_cancel(&q->watchdog);
	434
	435	/* No matter which mode we are on, it's safe to clear both lists. */
	436	timesortedlist_clear(sch);
	437	__qdisc_reset_queue(&sch->q);
	438
	439	sch->qstats.backlog = 0;
	440	sch->q.qlen = 0;
	441
	442	q->last = 0;
	443	}
	444
	445	static void etf_destroy(struct Qdisc *sch)
	446	{
	447	struct etf_sched_data *q = qdisc_priv(sch);
88cab771	448	struct net_device *dev = qdisc_dev(sch);
25db26a9 VCG	449
	450	/* Only cancel watchdog if it's been initialized. */
	451	if (q->watchdog.qdisc == sch)
	452	qdisc_watchdog_cancel(&q->watchdog);
88cab771 JSP	453
88cab771 JSP	454	etf_disable_offload(dev, q);
25db26a9 VCG	455	}
	456
	457	static int etf_dump(struct Qdisc sch, struct sk_buff skb)
	458	{
	459	struct etf_sched_data *q = qdisc_priv(sch);
	460	struct tc_etf_qopt opt = { };
	461	struct nlattr *nest;
	462
	463	nest = nla_nest_start(skb, TCA_OPTIONS);
	464	if (!nest)
	465	goto nla_put_failure;
	466
	467	opt.delta = q->delta;
	468	opt.clockid = q->clockid;
88cab771 JSP	469	if (q->offload)
	470	opt.flags \|= TC_ETF_OFFLOAD_ON;
	471
25db26a9 VCG	472	if (q->deadline_mode)
	473	opt.flags \|= TC_ETF_DEADLINE_MODE_ON;
	474
	475	if (nla_put(skb, TCA_ETF_PARMS, sizeof(opt), &opt))
	476	goto nla_put_failure;
	477
	478	return nla_nest_end(skb, nest);
	479
	480	nla_put_failure:
	481	nla_nest_cancel(skb, nest);
	482	return -1;
	483	}
	484
	485	static struct Qdisc_ops etf_qdisc_ops __read_mostly = {
	486	.id = "etf",
	487	.priv_size = sizeof(struct etf_sched_data),
	488	.enqueue = etf_enqueue_timesortedlist,
	489	.dequeue = etf_dequeue_timesortedlist,
	490	.peek = etf_peek_timesortedlist,
	491	.init = etf_init,
	492	.reset = etf_reset,
	493	.destroy = etf_destroy,
	494	.dump = etf_dump,
	495	.owner = THIS_MODULE,
	496	};
	497
	498	static int __init etf_module_init(void)
	499	{
	500	return register_qdisc(&etf_qdisc_ops);
	501	}
	502
	503	static void __exit etf_module_exit(void)
	504	{
	505	unregister_qdisc(&etf_qdisc_ops);
	506	}
	507	module_init(etf_module_init)
	508	module_exit(etf_module_exit)
	509	MODULE_LICENSE("GPL");