* from Ren Liu
* - More error checks
*
- *
- *
- * For all the glorious comments look at Alexey's sch_red.c
+ * For all the glorious comments look at include/net/red.h
*/
-#include <linux/config.h>
#include <linux/module.h>
-#include <asm/uaccess.h>
-#include <asm/system.h>
-#include <linux/bitops.h>
#include <linux/types.h>
#include <linux/kernel.h>
-#include <linux/sched.h>
-#include <linux/string.h>
-#include <linux/mm.h>
-#include <linux/socket.h>
-#include <linux/sockios.h>
-#include <linux/in.h>
-#include <linux/errno.h>
-#include <linux/interrupt.h>
-#include <linux/if_ether.h>
-#include <linux/inet.h>
-#include <linux/netdevice.h>
-#include <linux/etherdevice.h>
-#include <linux/notifier.h>
-#include <net/ip.h>
-#include <net/route.h>
#include <linux/skbuff.h>
-#include <net/sock.h>
#include <net/pkt_sched.h>
#include <net/red.h>
-#if 1 /* control */
-#define DPRINTK(format,args...) printk(KERN_DEBUG format,##args)
-#else
-#define DPRINTK(format,args...)
-#endif
-
-#if 0 /* data */
-#define D2PRINTK(format,args...) printk(KERN_DEBUG format,##args)
-#else
-#define D2PRINTK(format,args...)
-#endif
-
#define GRED_DEF_PRIO (MAX_DPs / 2)
#define GRED_VQ_MASK (MAX_DPs - 1)
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 */
+ u8 prio; /* the prio of this vq */
struct red_parms parms;
struct red_stats stats;
{
struct gred_sched_data *tab[MAX_DPs];
unsigned long flags;
- u32 DPs;
- u32 def;
+ u32 red_flags;
+ u32 DPs;
+ u32 def;
struct red_parms wred_set;
};
table->wred_set.qavg = q->parms.qavg;
}
-static int
-gred_enqueue(struct sk_buff *skb, struct Qdisc* sch)
+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;
- int i=0;
u16 dp = tc_index_to_dp(skb);
if (dp >= t->DPs || (q = t->tab[dp]) == NULL) {
* 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)
+ if (skb_queue_len(&sch->q) < qdisc_dev(sch)->tx_queue_len)
return qdisc_enqueue_tail(skb, sch);
else
goto drop;
skb->tc_index = (skb->tc_index & ~GRED_VQ_MASK) | dp;
}
- /* sum up all the qaves of prios <= to ours to get the new qave*/
+ /* sum up all the qaves of prios <= to ours to get the new qave */
if (!gred_wred_mode(t) && gred_rio_mode(t)) {
- for (i=0;i<t->DPs;i++) {
- if ((!t->tab[i]) || (i==q->DP))
- continue;
-
- if (t->tab[i]->prio < q->prio &&
+ 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;
+ q->bytesin += qdisc_pkt_len(skb);
- if (gred_wred_mode(t)) {
- qavg = 0;
+ if (gred_wred_mode(t))
gred_load_wred_set(t, q);
- }
q->parms.qavg = red_calc_qavg(&q->parms, gred_backlog(t, q, sch));
case RED_PROB_MARK:
sch->qstats.overlimits++;
- q->stats.prob_drop++;
- goto congestion_drop;
+ 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++;
- q->stats.forced_drop++;
- goto congestion_drop;
+ 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;
+ if (q->backlog + qdisc_pkt_len(skb) <= q->limit) {
+ q->backlog += qdisc_pkt_len(skb);
return qdisc_enqueue_tail(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)
+static struct sk_buff *gred_dequeue(struct Qdisc* sch)
{
struct sk_buff *skb;
- struct gred_sched_data *q;
- struct gred_sched *t= qdisc_priv(sch);
+ 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) {
"VQ 0x%x after dequeue, screwing up "
"backlog.\n", tc_index_to_dp(skb));
} else {
- q->backlog -= skb->len;
+ q->backlog -= qdisc_pkt_len(skb);
if (!q->backlog && !gred_wred_mode(t))
red_start_of_idle_period(&q->parms);
return skb;
}
- if (gred_wred_mode(t))
+ 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_data *q;
- struct gred_sched *t= qdisc_priv(sch);
+ struct gred_sched *t = qdisc_priv(sch);
skb = qdisc_dequeue_tail(sch);
if (skb) {
- unsigned int len = skb->len;
+ unsigned int len = qdisc_pkt_len(skb);
+ struct gred_sched_data *q;
u16 dp = tc_index_to_dp(skb);
if (dp >= t->DPs || (q = t->tab[dp]) == NULL) {
return len;
}
- if (gred_wred_mode(t))
+ 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_data *q;
- struct gred_sched *t= qdisc_priv(sch);
+ struct gred_sched *t = qdisc_priv(sch);
qdisc_reset_queue(sch);
- for (i=0;i<t->DPs;i++) {
- q= t->tab[i];
- if (!q)
- continue;
+ 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;
}
kfree(q);
}
-static inline int gred_change_table_def(struct Qdisc *sch, struct rtattr *dps)
+static inline int gred_change_table_def(struct Qdisc *sch, struct nlattr *dps)
{
struct gred_sched *table = qdisc_priv(sch);
struct tc_gred_sopt *sopt;
int i;
- if (dps == NULL || RTA_PAYLOAD(dps) < sizeof(*sopt))
+ if (dps == NULL)
return -EINVAL;
- sopt = RTA_DATA(dps);
+ sopt = nla_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
"shadowed VQ 0x%x\n", i);
gred_destroy_vq(table->tab[i]);
table->tab[i] = NULL;
- }
+ }
}
return 0;
struct gred_sched_data *q;
if (table->tab[dp] == NULL) {
- table->tab[dp] = kmalloc(sizeof(*q), GFP_KERNEL);
+ table->tab[dp] = kzalloc(sizeof(*q), GFP_KERNEL);
if (table->tab[dp] == NULL)
return -ENOMEM;
- memset(table->tab[dp], 0, sizeof(*q));
}
q = table->tab[dp];
return 0;
}
-static int gred_change(struct Qdisc *sch, struct rtattr *opt)
+static const struct nla_policy gred_policy[TCA_GRED_MAX + 1] = {
+ [TCA_GRED_PARMS] = { .len = sizeof(struct tc_gred_qopt) },
+ [TCA_GRED_STAB] = { .len = 256 },
+ [TCA_GRED_DPS] = { .len = sizeof(struct tc_gred_sopt) },
+};
+
+static int gred_change(struct Qdisc *sch, struct nlattr *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;
+ struct nlattr *tb[TCA_GRED_MAX + 1];
+ int err, prio = GRED_DEF_PRIO;
u8 *stab;
- if (opt == NULL || rtattr_parse_nested(tb, TCA_GRED_MAX, opt))
+ if (opt == NULL)
return -EINVAL;
- if (tb[TCA_GRED_PARMS-1] == NULL && tb[TCA_GRED_STAB-1] == NULL)
+ err = nla_parse_nested(tb, TCA_GRED_MAX, opt, gred_policy);
+ if (err < 0)
+ return err;
+
+ if (tb[TCA_GRED_PARMS] == NULL && tb[TCA_GRED_STAB] == 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)
+ if (tb[TCA_GRED_PARMS] == NULL ||
+ tb[TCA_GRED_STAB] == NULL)
return -EINVAL;
- ctl = RTA_DATA(tb[TCA_GRED_PARMS-1]);
- stab = RTA_DATA(tb[TCA_GRED_STAB-1]);
+ err = -EINVAL;
+ ctl = nla_data(tb[TCA_GRED_PARMS]);
+ stab = nla_data(tb[TCA_GRED_STAB]);
if (ctl->DP >= table->DPs)
goto errout;
return err;
}
-static int gred_init(struct Qdisc *sch, struct rtattr *opt)
+static int gred_init(struct Qdisc *sch, struct nlattr *opt)
{
- struct rtattr *tb[TCA_GRED_MAX];
+ struct nlattr *tb[TCA_GRED_MAX + 1];
+ int err;
- if (opt == NULL || rtattr_parse_nested(tb, TCA_GRED_MAX, opt))
+ if (opt == NULL)
return -EINVAL;
- if (tb[TCA_GRED_PARMS-1] || tb[TCA_GRED_STAB-1])
+ err = nla_parse_nested(tb, TCA_GRED_MAX, opt, gred_policy);
+ if (err < 0)
+ return err;
+
+ if (tb[TCA_GRED_PARMS] || tb[TCA_GRED_STAB])
return -EINVAL;
- return gred_change_table_def(sch, tb[TCA_GRED_DPS-1]);
+ return gred_change_table_def(sch, tb[TCA_GRED_DPS]);
}
static int gred_dump(struct Qdisc *sch, struct sk_buff *skb)
{
struct gred_sched *table = qdisc_priv(sch);
- struct rtattr *parms, *opts = NULL;
+ struct nlattr *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);
+ opts = nla_nest_start(skb, TCA_OPTIONS);
+ if (opts == NULL)
+ goto nla_put_failure;
+ NLA_PUT(skb, TCA_GRED_DPS, sizeof(sopt), &sopt);
+ 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];
opt.qave = red_calc_qavg(&q->parms, q->parms.qavg);
append_opt:
- RTA_APPEND(skb, sizeof(opt), &opt);
+ if (nla_append(skb, sizeof(opt), &opt) < 0)
+ goto nla_put_failure;
}
- RTA_NEST_END(skb, parms);
+ nla_nest_end(skb, parms);
- return RTA_NEST_END(skb, opts);
+ return nla_nest_end(skb, opts);
-rtattr_failure:
- return RTA_NEST_CANCEL(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++) {
+ for (i = 0; i < table->DPs; i++) {
if (table->tab[i])
gred_destroy_vq(table->tab[i]);
}
}
-static struct Qdisc_ops gred_qdisc_ops = {
- .next = NULL,
- .cl_ops = NULL,
+static struct Qdisc_ops gred_qdisc_ops __read_mostly = {
.id = "gred",
.priv_size = sizeof(struct gred_sched),
.enqueue = gred_enqueue,
.dequeue = gred_dequeue,
- .requeue = gred_requeue,
+ .peek = qdisc_peek_head,
.drop = gred_drop,
.init = gred_init,
.reset = gred_reset,
{
return register_qdisc(&gred_qdisc_ops);
}
-static void __exit gred_module_exit(void)
+
+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");
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff