static int red_enqueue(struct sk_buff *skb, struct Qdisc* sch)
{
struct red_sched_data *q = qdisc_priv(sch);
q->parms.qavg = red_calc_qavg(&q->parms, sch->qstats.backlog);
if (red_is_idling(&q->parms))
red_end_of_idle_period(&q->parms);
switch (red_action(&q->parms, q->parms.qavg)) {
case RED_DONT_MARK:
break;
case RED_PROB_MARK:
sch->qstats.overlimits++;
if (!red_use_ecn(q) || !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 (red_use_harddrop(q) || !red_use_ecn(q) ||
!INET_ECN_set_ce(skb)) {
q->stats.forced_drop++;
goto congestion_drop;
}
q->stats.forced_mark++;
break;
}
if (sch->qstats.backlog + skb->len <= q->limit)
return qdisc_enqueue_tail(skb, sch);
q->stats.pdrop++;
return qdisc_drop(skb, sch);
congestion_drop:
qdisc_drop(skb, sch);
return NET_XMIT_CN;
}
/*
* When q->perturbation is changed, we rehash all queued skbs
* to avoid OOO (Out Of Order) effects.
* We dont use sfq_dequeue()/sfq_enqueue() because we dont want to change
* counters.
*/
static void sfq_rehash(struct Qdisc *sch)
{
struct sfq_sched_data *q = qdisc_priv(sch);
struct sk_buff *skb;
int i;
struct sfq_slot *slot;
struct sk_buff_head list;
int dropped = 0;
__skb_queue_head_init(&list);
for (i = 0; i < q->maxflows; i++) {
slot = &q->slots[i];
if (!slot->qlen)
continue;
while (slot->qlen) {
skb = slot_dequeue_head(slot);
sfq_dec(q, i);
__skb_queue_tail(&list, skb);
}
slot->backlog = 0;
red_set_vars(&slot->vars);
q->ht[slot->hash] = SFQ_EMPTY_SLOT;
}
q->tail = NULL;
while ((skb = __skb_dequeue(&list)) != NULL) {
unsigned int hash = sfq_hash(q, skb);
sfq_index x = q->ht[hash];
slot = &q->slots[x];
if (x == SFQ_EMPTY_SLOT) {
x = q->dep[0].next; /* get a free slot */
if (x >= SFQ_MAX_FLOWS) {
drop: sch->qstats.backlog -= qdisc_pkt_len(skb);
kfree_skb(skb);
dropped++;
continue;
}
q->ht[hash] = x;
slot = &q->slots[x];
slot->hash = hash;
}
if (slot->qlen >= q->maxdepth)
goto drop;
slot_queue_add(slot, skb);
if (q->red_parms)
slot->vars.qavg = red_calc_qavg(q->red_parms,
&slot->vars,
slot->backlog);
slot->backlog += qdisc_pkt_len(skb);
sfq_inc(q, x);
if (slot->qlen == 1) { /* The flow is new */
if (q->tail == NULL) { /* It is the first flow */
slot->next = x;
} else {
slot->next = q->tail->next;
q->tail->next = x;
}
q->tail = slot;
slot->allot = q->scaled_quantum;
}
}
sch->q.qlen -= dropped;
qdisc_tree_decrease_qlen(sch, dropped);
}
static int choke_enqueue(struct sk_buff *skb, struct Qdisc *sch)
{
struct choke_sched_data *q = qdisc_priv(sch);
const struct red_parms *p = &q->parms;
int ret = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
if (q->filter_list) {
/* If using external classifiers, get result and record it. */
if (!choke_classify(skb, sch, &ret))
goto other_drop; /* Packet was eaten by filter */
}
choke_skb_cb(skb)->keys_valid = 0;
/* Compute average queue usage (see RED) */
q->vars.qavg = red_calc_qavg(p, &q->vars, sch->q.qlen);
if (red_is_idling(&q->vars))
red_end_of_idle_period(&q->vars);
/* Is queue small? */
if (q->vars.qavg <= p->qth_min)
q->vars.qcount = -1;
else {
unsigned int idx;
/* Draw a packet at random from queue and compare flow */
if (choke_match_random(q, skb, &idx)) {
q->stats.matched++;
choke_drop_by_idx(sch, idx);
goto congestion_drop;
}
/* Queue is large, always mark/drop */
if (q->vars.qavg > p->qth_max) {
q->vars.qcount = -1;
sch->qstats.overlimits++;
if (use_harddrop(q) || !use_ecn(q) ||
!INET_ECN_set_ce(skb)) {
q->stats.forced_drop++;
goto congestion_drop;
}
q->stats.forced_mark++;
} else if (++q->vars.qcount) {
if (red_mark_probability(p, &q->vars, q->vars.qavg)) {
q->vars.qcount = 0;
q->vars.qR = red_random(p);
sch->qstats.overlimits++;
if (!use_ecn(q) || !INET_ECN_set_ce(skb)) {
q->stats.prob_drop++;
goto congestion_drop;
}
q->stats.prob_mark++;
}
} else
q->vars.qR = red_random(p);
}
/* Admit new packet */
if (sch->q.qlen < q->limit) {
q->tab[q->tail] = skb;
q->tail = (q->tail + 1) & q->tab_mask;
++sch->q.qlen;
sch->qstats.backlog += qdisc_pkt_len(skb);
return NET_XMIT_SUCCESS;
}
q->stats.pdrop++;
sch->qstats.drops++;
kfree_skb(skb);
return NET_XMIT_DROP;
congestion_drop:
qdisc_drop(skb, sch);
return NET_XMIT_CN;
other_drop:
if (ret & __NET_XMIT_BYPASS)
sch->qstats.drops++;
kfree_skb(skb);
return ret;
}
static int choke_enqueue(struct sk_buff *skb, struct Qdisc *sch,
struct sk_buff **to_free)
{
struct choke_sched_data *q = qdisc_priv(sch);
const struct red_parms *p = &q->parms;
choke_skb_cb(skb)->keys_valid = 0;
/* Compute average queue usage (see RED) */
q->vars.qavg = red_calc_qavg(p, &q->vars, sch->q.qlen);
if (red_is_idling(&q->vars))
red_end_of_idle_period(&q->vars);
/* Is queue small? */
if (q->vars.qavg <= p->qth_min)
q->vars.qcount = -1;
else {
unsigned int idx;
/* Draw a packet at random from queue and compare flow */
if (choke_match_random(q, skb, &idx)) {
q->stats.matched++;
choke_drop_by_idx(sch, idx, to_free);
goto congestion_drop;
}
/* Queue is large, always mark/drop */
if (q->vars.qavg > p->qth_max) {
q->vars.qcount = -1;
qdisc_qstats_overlimit(sch);
if (use_harddrop(q) || !use_ecn(q) ||
!INET_ECN_set_ce(skb)) {
q->stats.forced_drop++;
goto congestion_drop;
}
q->stats.forced_mark++;
} else if (++q->vars.qcount) {
if (red_mark_probability(p, &q->vars, q->vars.qavg)) {
q->vars.qcount = 0;
q->vars.qR = red_random(p);
qdisc_qstats_overlimit(sch);
if (!use_ecn(q) || !INET_ECN_set_ce(skb)) {
q->stats.prob_drop++;
goto congestion_drop;
}
q->stats.prob_mark++;
}
} else
q->vars.qR = red_random(p);
}
/* Admit new packet */
if (sch->q.qlen < q->limit) {
q->tab[q->tail] = skb;
q->tail = (q->tail + 1) & q->tab_mask;
++sch->q.qlen;
qdisc_qstats_backlog_inc(sch, skb);
return NET_XMIT_SUCCESS;
}
q->stats.pdrop++;
return qdisc_drop(skb, sch, to_free);
congestion_drop:
qdisc_drop(skb, sch, to_free);
return NET_XMIT_CN;
}
static int gred_enqueue(struct sk_buff *skb, struct Qdisc *sch,
struct sk_buff **to_free)
{
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;
q = t->tab[dp];
if (!q) {
/* 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 (likely(sch->qstats.backlog + qdisc_pkt_len(skb) <=
sch->limit))
return qdisc_enqueue_tail(skb, sch);
else
goto drop;
}
/* fix tc_index? --could be controversial but needed for
requeueing */
skb->tc_index = (skb->tc_index & ~GRED_VQ_MASK) | dp;
}
/* sum up all the qaves of prios < 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]->vars))
qavg += t->tab[i]->vars.qavg;
}
}
q->packetsin++;
q->bytesin += qdisc_pkt_len(skb);
if (gred_wred_mode(t))
gred_load_wred_set(t, q);
q->vars.qavg = red_calc_qavg(&q->parms,
&q->vars,
gred_backlog(t, q, sch));
if (red_is_idling(&q->vars))
red_end_of_idle_period(&q->vars);
if (gred_wred_mode(t))
gred_store_wred_set(t, q);
switch (red_action(&q->parms, &q->vars, q->vars.qavg + qavg)) {
case RED_DONT_MARK:
break;
case RED_PROB_MARK:
qdisc_qstats_overlimit(sch);
if (!gred_use_ecn(q) || !INET_ECN_set_ce(skb)) {
q->stats.prob_drop++;
goto congestion_drop;
}
q->stats.prob_mark++;
break;
case RED_HARD_MARK:
qdisc_qstats_overlimit(sch);
if (gred_use_harddrop(q) || !gred_use_ecn(q) ||
!INET_ECN_set_ce(skb)) {
q->stats.forced_drop++;
goto congestion_drop;
}
q->stats.forced_mark++;
break;
}
if (gred_backlog(t, q, sch) + qdisc_pkt_len(skb) <= q->limit) {
q->backlog += qdisc_pkt_len(skb);
return qdisc_enqueue_tail(skb, sch);
}
q->stats.pdrop++;
drop:
return qdisc_drop(skb, sch, to_free);
congestion_drop:
qdisc_drop(skb, sch, to_free);
return NET_XMIT_CN;
}
static int gred_dump(struct Qdisc *sch, struct sk_buff *skb)
{
struct gred_sched *table = qdisc_priv(sch);
struct nlattr *parms, *opts = NULL;
int i;
u32 max_p[MAX_DPs];
struct tc_gred_sopt sopt = {
.DPs = table->DPs,
.def_DP = table->def,
.grio = gred_rio_mode(table),
.flags = table->red_flags,
};
opts = nla_nest_start(skb, TCA_OPTIONS);
if (opts == NULL)
goto nla_put_failure;
nla_put(skb, TCA_GRED_DPS, sizeof(sopt), &sopt);
for (i = 0; i < MAX_DPs; i++) {
struct gred_sched_data *q = table->tab[i];
max_p[i] = q ? q->parms.max_P : 0;
}
nla_put(skb, TCA_GRED_MAX_P, sizeof(max_p), max_p);
parms = nla_nest_start(skb, TCA_GRED_PARMS);
if (parms == NULL)
goto nla_put_failure;
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))
gred_load_wred_set(table, q);
opt.qave = red_calc_qavg(&q->parms, &q->vars, q->vars.qavg);
append_opt:
if (nla_append(skb, sizeof(opt), &opt) < 0)
goto nla_put_failure;
}
nla_nest_end(skb, parms);
return nla_nest_end(skb, opts);
nla_put_failure:
nla_nest_cancel(skb, opts);
return -EMSGSIZE;
}
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 __read_mostly = {
.id = "gred",
.priv_size = sizeof(struct gred_sched),
.enqueue = gred_enqueue,
.dequeue = gred_dequeue,
.peek = qdisc_peek_head,
.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");
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");
