Subversion Repositories or1k_soc_on_altera_embedded_dev_kit

/*
 * net/sched/sch_gred.c	Generic Random Early Detection queue.
 *
 *
 *              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:    J Hadi Salim (hadi@cyberus.ca) 1998-2002
 *
 *             991129: -  Bug fix with grio mode
 *		       - a better sing. AvgQ mode with Grio(WRED)
 *		       - A finer grained VQ dequeue based on sugestion
 *		         from Ren Liu
 *		       - More error checks
 *
 *  For all the glorious comments look at include/net/red.h
 */
 
#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/skbuff.h>
#include <net/pkt_sched.h>
#include <net/red.h>
 
#define GRED_DEF_PRIO (MAX_DPs / 2)
#define GRED_VQ_MASK (MAX_DPs - 1)
 
struct gred_sched_data;
struct gred_sched;
 
struct gred_sched_data
{
	u32		limit;		/* HARD maximal queue length	*/
	u32      	DP;		/* the drop pramaters */
	u32		bytesin;	/* bytes seen on virtualQ so far*/
	u32		packetsin;	/* packets seen on virtualQ so far*/
	u32		backlog;	/* bytes on the virtualQ */
	u8		prio;		/* the prio of this vq */
 
	struct red_parms parms;
	struct red_stats stats;
};
 
enum {
	GRED_WRED_MODE = 1,
	GRED_RIO_MODE,
};
 
struct gred_sched
{
	struct gred_sched_data *tab[MAX_DPs];
	unsigned long	flags;
	u32		red_flags;
	u32 		DPs;
	u32 		def;
	struct red_parms wred_set;
};
 
static inline int gred_wred_mode(struct gred_sched *table)
{
	return test_bit(GRED_WRED_MODE, &table->flags);
}
 
static inline void gred_enable_wred_mode(struct gred_sched *table)
{
	__set_bit(GRED_WRED_MODE, &table->flags);
}
 
static inline void gred_disable_wred_mode(struct gred_sched *table)
{
	__clear_bit(GRED_WRED_MODE, &table->flags);
}
 
static inline int gred_rio_mode(struct gred_sched *table)
{
	return test_bit(GRED_RIO_MODE, &table->flags);
}
 
static inline void gred_enable_rio_mode(struct gred_sched *table)
{
	__set_bit(GRED_RIO_MODE, &table->flags);
}
 
static inline void gred_disable_rio_mode(struct gred_sched *table)
{
	__clear_bit(GRED_RIO_MODE, &table->flags);
}
 
static inline int gred_wred_mode_check(struct Qdisc *sch)
{
	struct gred_sched *table = qdisc_priv(sch);
	int i;
 
	/* Really ugly O(n^2) but shouldn't be necessary too frequent. */
	for (i = 0; i < table->DPs; i++) {
		struct gred_sched_data *q = table->tab[i];
		int n;
 
		if (q == NULL)
			continue;
 
		for (n = 0; n < table->DPs; n++)
			if (table->tab[n] && table->tab[n] != q &&
			    table->tab[n]->prio == q->prio)
				return 1;
	}
 
	return 0;
}
 
static inline unsigned int gred_backlog(struct gred_sched *table,
					struct gred_sched_data *q,
					struct Qdisc *sch)
{
	if (gred_wred_mode(table))
		return sch->qstats.backlog;
	else
		return q->backlog;
}
 
static inline u16 tc_index_to_dp(struct sk_buff *skb)
{
	return skb->tc_index & GRED_VQ_MASK;
}
 
static inline void gred_load_wred_set(struct gred_sched *table,
				      struct gred_sched_data *q)
{
	q->parms.qavg = table->wred_set.qavg;
	q->parms.qidlestart = table->wred_set.qidlestart;
}
 
static inline void gred_store_wred_set(struct gred_sched *table,
				       struct gred_sched_data *q)
{
	table->wred_set.qavg = q->parms.qavg;
}
 
static inline int gred_use_ecn(struct gred_sched *t)
{
	return t->red_flags & TC_RED_ECN;
}
 
static inline int gred_use_harddrop(struct gred_sched *t)
{
	return t->red_flags & TC_RED_HARDDROP;
}
 
static int gred_enqueue(struct sk_buff *skb, struct Qdisc* sch)
{
	struct gred_sched_data *q=NULL;
	struct gred_sched *t= qdisc_priv(sch);
	unsigned long qavg = 0;
	u16 dp = tc_index_to_dp(skb);
 
	if (dp >= t->DPs  || (q = t->tab[dp]) == NULL) {
		dp = t->def;
 
		if ((q = t->tab[dp]) == NULL) {
			/* Pass through packets not assigned to a DP
			 * if no default DP has been configured. This
			 * allows for DP flows to be left untouched.
			 */
			if (skb_queue_len(&sch->q) < sch->dev->tx_queue_len)
				return qdisc_enqueue_tail(skb, sch);
			else
				goto drop;
		}
 
		/* fix tc_index? --could be controvesial but needed for
		   requeueing */
		skb->tc_index = (skb->tc_index & ~GRED_VQ_MASK) | dp;
	}
 
	/* sum up all the qaves of prios <= to ours to get the new qave */
	if (!gred_wred_mode(t) && gred_rio_mode(t)) {
		int i;
 
		for (i = 0; i < t->DPs; i++) {
			if (t->tab[i] && t->tab[i]->prio < q->prio &&
			    !red_is_idling(&t->tab[i]->parms))
				qavg +=t->tab[i]->parms.qavg;
		}
 
	}
 
	q->packetsin++;
	q->bytesin += skb->len;
 
	if (gred_wred_mode(t))
		gred_load_wred_set(t, q);
 
	q->parms.qavg = red_calc_qavg(&q->parms, gred_backlog(t, q, sch));
 
	if (red_is_idling(&q->parms))
		red_end_of_idle_period(&q->parms);
 
	if (gred_wred_mode(t))
		gred_store_wred_set(t, q);
 
	switch (red_action(&q->parms, q->parms.qavg + qavg)) {
		case RED_DONT_MARK:
			break;
 
		case RED_PROB_MARK:
			sch->qstats.overlimits++;
			if (!gred_use_ecn(t) || !INET_ECN_set_ce(skb)) {
				q->stats.prob_drop++;
				goto congestion_drop;
			}
 
			q->stats.prob_mark++;
			break;
 
		case RED_HARD_MARK:
			sch->qstats.overlimits++;
			if (gred_use_harddrop(t) || !gred_use_ecn(t) ||
			    !INET_ECN_set_ce(skb)) {
				q->stats.forced_drop++;
				goto congestion_drop;
			}
			q->stats.forced_mark++;
			break;
	}
 
	if (q->backlog + skb->len <= q->limit) {
		q->backlog += skb->len;
		return qdisc_enqueue_tail(skb, sch);
	}
 
	q->stats.pdrop++;
drop:
	return qdisc_drop(skb, sch);
 
congestion_drop:
	qdisc_drop(skb, sch);
	return NET_XMIT_CN;
}
 
static int gred_requeue(struct sk_buff *skb, struct Qdisc* sch)
{
	struct gred_sched *t = qdisc_priv(sch);
	struct gred_sched_data *q;
	u16 dp = tc_index_to_dp(skb);
 
	if (dp >= t->DPs || (q = t->tab[dp]) == NULL) {
		if (net_ratelimit())
			printk(KERN_WARNING "GRED: Unable to relocate VQ 0x%x "
			       "for requeue, screwing up backlog.\n",
			       tc_index_to_dp(skb));
	} else {
		if (red_is_idling(&q->parms))
			red_end_of_idle_period(&q->parms);
		q->backlog += skb->len;
	}
 
	return qdisc_requeue(skb, sch);
}
 
static struct sk_buff *gred_dequeue(struct Qdisc* sch)
{
	struct sk_buff *skb;
	struct gred_sched *t = qdisc_priv(sch);
 
	skb = qdisc_dequeue_head(sch);
 
	if (skb) {
		struct gred_sched_data *q;
		u16 dp = tc_index_to_dp(skb);
 
		if (dp >= t->DPs || (q = t->tab[dp]) == NULL) {
			if (net_ratelimit())
				printk(KERN_WARNING "GRED: Unable to relocate "
				       "VQ 0x%x after dequeue, screwing up "
				       "backlog.\n", tc_index_to_dp(skb));
		} else {
			q->backlog -= skb->len;
 
			if (!q->backlog && !gred_wred_mode(t))
				red_start_of_idle_period(&q->parms);
		}
 
		return skb;
	}
 
	if (gred_wred_mode(t) && !red_is_idling(&t->wred_set))
		red_start_of_idle_period(&t->wred_set);
 
	return NULL;
}
 
static unsigned int gred_drop(struct Qdisc* sch)
{
	struct sk_buff *skb;
	struct gred_sched *t = qdisc_priv(sch);
 
	skb = qdisc_dequeue_tail(sch);
	if (skb) {
		unsigned int len = skb->len;
		struct gred_sched_data *q;
		u16 dp = tc_index_to_dp(skb);
 
		if (dp >= t->DPs || (q = t->tab[dp]) == NULL) {
			if (net_ratelimit())
				printk(KERN_WARNING "GRED: Unable to relocate "
				       "VQ 0x%x while dropping, screwing up "
				       "backlog.\n", tc_index_to_dp(skb));
		} else {
			q->backlog -= len;
			q->stats.other++;
 
			if (!q->backlog && !gred_wred_mode(t))
				red_start_of_idle_period(&q->parms);
		}
 
		qdisc_drop(skb, sch);
		return len;
	}
 
	if (gred_wred_mode(t) && !red_is_idling(&t->wred_set))
		red_start_of_idle_period(&t->wred_set);
 
	return 0;
 
}
 
static void gred_reset(struct Qdisc* sch)
{
	int i;
	struct gred_sched *t = qdisc_priv(sch);
 
	qdisc_reset_queue(sch);
 
	for (i = 0; i < t->DPs; i++) {
		struct gred_sched_data *q = t->tab[i];
 
		if (!q)
			continue;
 
		red_restart(&q->parms);
		q->backlog = 0;
	}
}
 
static inline void gred_destroy_vq(struct gred_sched_data *q)
{
	kfree(q);
}
 
static inline int gred_change_table_def(struct Qdisc *sch, struct rtattr *dps)
{
	struct gred_sched *table = qdisc_priv(sch);
	struct tc_gred_sopt *sopt;
	int i;
 
	if (dps == NULL || RTA_PAYLOAD(dps) < sizeof(*sopt))
		return -EINVAL;
 
	sopt = RTA_DATA(dps);
 
	if (sopt->DPs > MAX_DPs || sopt->DPs == 0 || sopt->def_DP >= sopt->DPs)
		return -EINVAL;
 
	sch_tree_lock(sch);
	table->DPs = sopt->DPs;
	table->def = sopt->def_DP;
	table->red_flags = sopt->flags;
 
	/*
	 * Every entry point to GRED is synchronized with the above code
	 * and the DP is checked against DPs, i.e. shadowed VQs can no
	 * longer be found so we can unlock right here.
	 */
	sch_tree_unlock(sch);
 
	if (sopt->grio) {
		gred_enable_rio_mode(table);
		gred_disable_wred_mode(table);
		if (gred_wred_mode_check(sch))
			gred_enable_wred_mode(table);
	} else {
		gred_disable_rio_mode(table);
		gred_disable_wred_mode(table);
	}
 
	for (i = table->DPs; i < MAX_DPs; i++) {
		if (table->tab[i]) {
			printk(KERN_WARNING "GRED: Warning: Destroying "
			       "shadowed VQ 0x%x\n", i);
			gred_destroy_vq(table->tab[i]);
			table->tab[i] = NULL;
		}
	}
 
	return 0;
}
 
static inline int gred_change_vq(struct Qdisc *sch, int dp,
				 struct tc_gred_qopt *ctl, int prio, u8 *stab)
{
	struct gred_sched *table = qdisc_priv(sch);
	struct gred_sched_data *q;
 
	if (table->tab[dp] == NULL) {
		table->tab[dp] = kzalloc(sizeof(*q), GFP_KERNEL);
		if (table->tab[dp] == NULL)
			return -ENOMEM;
	}
 
	q = table->tab[dp];
	q->DP = dp;
	q->prio = prio;
	q->limit = ctl->limit;
 
	if (q->backlog == 0)
		red_end_of_idle_period(&q->parms);
 
	red_set_parms(&q->parms,
		      ctl->qth_min, ctl->qth_max, ctl->Wlog, ctl->Plog,
		      ctl->Scell_log, stab);
 
	return 0;
}
 
static int gred_change(struct Qdisc *sch, struct rtattr *opt)
{
	struct gred_sched *table = qdisc_priv(sch);
	struct tc_gred_qopt *ctl;
	struct rtattr *tb[TCA_GRED_MAX];
	int err = -EINVAL, prio = GRED_DEF_PRIO;
	u8 *stab;
 
	if (opt == NULL || rtattr_parse_nested(tb, TCA_GRED_MAX, opt))
		return -EINVAL;
 
	if (tb[TCA_GRED_PARMS-1] == NULL && tb[TCA_GRED_STAB-1] == NULL)
		return gred_change_table_def(sch, opt);
 
	if (tb[TCA_GRED_PARMS-1] == NULL ||
	    RTA_PAYLOAD(tb[TCA_GRED_PARMS-1]) < sizeof(*ctl) ||
	    tb[TCA_GRED_STAB-1] == NULL ||
	    RTA_PAYLOAD(tb[TCA_GRED_STAB-1]) < 256)
		return -EINVAL;
 
	ctl = RTA_DATA(tb[TCA_GRED_PARMS-1]);
	stab = RTA_DATA(tb[TCA_GRED_STAB-1]);
 
	if (ctl->DP >= table->DPs)
		goto errout;
 
	if (gred_rio_mode(table)) {
		if (ctl->prio == 0) {
			int def_prio = GRED_DEF_PRIO;
 
			if (table->tab[table->def])
				def_prio = table->tab[table->def]->prio;
 
			printk(KERN_DEBUG "GRED: DP %u does not have a prio "
			       "setting default to %d\n", ctl->DP, def_prio);
 
			prio = def_prio;
		} else
			prio = ctl->prio;
	}
 
	sch_tree_lock(sch);
 
	err = gred_change_vq(sch, ctl->DP, ctl, prio, stab);
	if (err < 0)
		goto errout_locked;
 
	if (gred_rio_mode(table)) {
		gred_disable_wred_mode(table);
		if (gred_wred_mode_check(sch))
			gred_enable_wred_mode(table);
	}
 
	err = 0;
 
errout_locked:
	sch_tree_unlock(sch);
errout:
	return err;
}
 
static int gred_init(struct Qdisc *sch, struct rtattr *opt)
{
	struct rtattr *tb[TCA_GRED_MAX];
 
	if (opt == NULL || rtattr_parse_nested(tb, TCA_GRED_MAX, opt))
		return -EINVAL;
 
	if (tb[TCA_GRED_PARMS-1] || tb[TCA_GRED_STAB-1])
		return -EINVAL;
 
	return gred_change_table_def(sch, tb[TCA_GRED_DPS-1]);
}
 
static int gred_dump(struct Qdisc *sch, struct sk_buff *skb)
{
	struct gred_sched *table = qdisc_priv(sch);
	struct rtattr *parms, *opts = NULL;
	int i;
	struct tc_gred_sopt sopt = {
		.DPs	= table->DPs,
		.def_DP	= table->def,
		.grio	= gred_rio_mode(table),
		.flags	= table->red_flags,
	};
 
	opts = RTA_NEST(skb, TCA_OPTIONS);
	RTA_PUT(skb, TCA_GRED_DPS, sizeof(sopt), &sopt);
	parms = RTA_NEST(skb, TCA_GRED_PARMS);
 
	for (i = 0; i < MAX_DPs; i++) {
		struct gred_sched_data *q = table->tab[i];
		struct tc_gred_qopt opt;
 
		memset(&opt, 0, sizeof(opt));
 
		if (!q) {
			/* hack -- fix at some point with proper message
			   This is how we indicate to tc that there is no VQ
			   at this DP */
 
			opt.DP = MAX_DPs + i;
			goto append_opt;
		}
 
		opt.limit	= q->limit;
		opt.DP		= q->DP;
		opt.backlog	= q->backlog;
		opt.prio	= q->prio;
		opt.qth_min	= q->parms.qth_min >> q->parms.Wlog;
		opt.qth_max	= q->parms.qth_max >> q->parms.Wlog;
		opt.Wlog	= q->parms.Wlog;
		opt.Plog	= q->parms.Plog;
		opt.Scell_log	= q->parms.Scell_log;
		opt.other	= q->stats.other;
		opt.early	= q->stats.prob_drop;
		opt.forced	= q->stats.forced_drop;
		opt.pdrop	= q->stats.pdrop;
		opt.packets	= q->packetsin;
		opt.bytesin	= q->bytesin;
 
		if (gred_wred_mode(table)) {
			q->parms.qidlestart =
				table->tab[table->def]->parms.qidlestart;
			q->parms.qavg = table->tab[table->def]->parms.qavg;
		}
 
		opt.qave = red_calc_qavg(&q->parms, q->parms.qavg);
 
append_opt:
		RTA_APPEND(skb, sizeof(opt), &opt);
	}
 
	RTA_NEST_END(skb, parms);
 
	return RTA_NEST_END(skb, opts);
 
rtattr_failure:
	return RTA_NEST_CANCEL(skb, opts);
}
 
static void gred_destroy(struct Qdisc *sch)
{
	struct gred_sched *table = qdisc_priv(sch);
	int i;
 
	for (i = 0; i < table->DPs; i++) {
		if (table->tab[i])
			gred_destroy_vq(table->tab[i]);
	}
}
 
static struct Qdisc_ops gred_qdisc_ops = {
	.id		=	"gred",
	.priv_size	=	sizeof(struct gred_sched),
	.enqueue	=	gred_enqueue,
	.dequeue	=	gred_dequeue,
	.requeue	=	gred_requeue,
	.drop		=	gred_drop,
	.init		=	gred_init,
	.reset		=	gred_reset,
	.destroy	=	gred_destroy,
	.change		=	gred_change,
	.dump		=	gred_dump,
	.owner		=	THIS_MODULE,
};
 
static int __init gred_module_init(void)
{
	return register_qdisc(&gred_qdisc_ops);
}
 
static void __exit gred_module_exit(void)
{
	unregister_qdisc(&gred_qdisc_ops);
}
 
module_init(gred_module_init)
module_exit(gred_module_exit)
 
MODULE_LICENSE("GPL");