* Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
*
* Changes:
- * J Hadi Salim <hadi@nortel.com> 980914: computation fixes
+ * J Hadi Salim 980914: computation fixes
* Alexey Makarenko <makar@phoenix.kharkov.ua> 990814: qave on idle link was calculated incorrectly.
- * J Hadi Salim <hadi@nortelnetworks.com> 980816: ECN support
+ * J Hadi Salim 980816: ECN support
*/
-#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/inet_ecn.h>
-#include <net/dsfield.h>
#include <net/red.h>
unsigned char flags;
struct red_parms parms;
struct red_stats stats;
+ struct Qdisc *qdisc;
};
static inline int red_use_ecn(struct red_sched_data *q)
return q->flags & TC_RED_ECN;
}
-static int
-red_enqueue(struct sk_buff *skb, struct Qdisc* sch)
+static inline int red_use_harddrop(struct red_sched_data *q)
+{
+ return q->flags & TC_RED_HARDDROP;
+}
+
+static int red_enqueue(struct sk_buff *skb, struct Qdisc* sch)
{
struct red_sched_data *q = qdisc_priv(sch);
+ struct Qdisc *child = q->qdisc;
+ int ret;
- q->parms.qavg = red_calc_qavg(&q->parms, sch->qstats.backlog);
+ q->parms.qavg = red_calc_qavg(&q->parms, child->qstats.backlog);
if (red_is_idling(&q->parms))
red_end_of_idle_period(&q->parms);
case RED_HARD_MARK:
sch->qstats.overlimits++;
- if (!red_use_ecn(q) || !INET_ECN_set_ce(skb)) {
+ if (red_use_harddrop(q) || !red_use_ecn(q) ||
+ !INET_ECN_set_ce(skb)) {
q->stats.forced_drop++;
goto congestion_drop;
}
break;
}
- if (sch->qstats.backlog + skb->len <= q->limit) {
- __skb_queue_tail(&sch->q, skb);
- sch->qstats.backlog += skb->len;
+ ret = child->enqueue(skb, child);
+ if (likely(ret == NET_XMIT_SUCCESS)) {
sch->bstats.bytes += skb->len;
sch->bstats.packets++;
- return NET_XMIT_SUCCESS;
+ sch->q.qlen++;
+ } else {
+ q->stats.pdrop++;
+ sch->qstats.drops++;
}
-
- q->stats.pdrop++;
- kfree_skb(skb);
- sch->qstats.drops++;
- return NET_XMIT_DROP;
+ return ret;
congestion_drop:
- kfree_skb(skb);
- sch->qstats.drops++;
+ qdisc_drop(skb, sch);
return NET_XMIT_CN;
}
-static int
-red_requeue(struct sk_buff *skb, struct Qdisc* sch)
+static int red_requeue(struct sk_buff *skb, struct Qdisc* sch)
{
struct red_sched_data *q = qdisc_priv(sch);
+ struct Qdisc *child = q->qdisc;
+ int ret;
if (red_is_idling(&q->parms))
red_end_of_idle_period(&q->parms);
- __skb_queue_head(&sch->q, skb);
- sch->qstats.backlog += skb->len;
- sch->qstats.requeues++;
- return 0;
+ ret = child->ops->requeue(skb, child);
+ if (likely(ret == NET_XMIT_SUCCESS)) {
+ sch->qstats.requeues++;
+ sch->q.qlen++;
+ }
+ return ret;
}
-static struct sk_buff *
-red_dequeue(struct Qdisc* sch)
+static struct sk_buff * red_dequeue(struct Qdisc* sch)
{
struct sk_buff *skb;
struct red_sched_data *q = qdisc_priv(sch);
+ struct Qdisc *child = q->qdisc;
- skb = __skb_dequeue(&sch->q);
- if (skb) {
- sch->qstats.backlog -= skb->len;
- return skb;
- }
+ skb = child->dequeue(child);
+ if (skb)
+ sch->q.qlen--;
+ else if (!red_is_idling(&q->parms))
+ red_start_of_idle_period(&q->parms);
- red_start_of_idle_period(&q->parms);
- return NULL;
+ return skb;
}
static unsigned int red_drop(struct Qdisc* sch)
{
- struct sk_buff *skb;
struct red_sched_data *q = qdisc_priv(sch);
+ struct Qdisc *child = q->qdisc;
+ unsigned int len;
- skb = __skb_dequeue_tail(&sch->q);
- if (skb) {
- unsigned int len = skb->len;
- sch->qstats.backlog -= len;
- sch->qstats.drops++;
+ if (child->ops->drop && (len = child->ops->drop(child)) > 0) {
q->stats.other++;
- kfree_skb(skb);
+ sch->qstats.drops++;
+ sch->q.qlen--;
return len;
}
- red_start_of_idle_period(&q->parms);
+ if (!red_is_idling(&q->parms))
+ red_start_of_idle_period(&q->parms);
+
return 0;
}
{
struct red_sched_data *q = qdisc_priv(sch);
- __skb_queue_purge(&sch->q);
- sch->qstats.backlog = 0;
+ qdisc_reset(q->qdisc);
+ sch->q.qlen = 0;
red_restart(&q->parms);
}
+static void red_destroy(struct Qdisc *sch)
+{
+ struct red_sched_data *q = qdisc_priv(sch);
+ qdisc_destroy(q->qdisc);
+}
+
+static struct Qdisc *red_create_dflt(struct Qdisc *sch, u32 limit)
+{
+ struct Qdisc *q;
+ struct rtattr *rta;
+ int ret;
+
+ q = qdisc_create_dflt(sch->dev, &bfifo_qdisc_ops,
+ TC_H_MAKE(sch->handle, 1));
+ if (q) {
+ rta = kmalloc(RTA_LENGTH(sizeof(struct tc_fifo_qopt)),
+ GFP_KERNEL);
+ if (rta) {
+ rta->rta_type = RTM_NEWQDISC;
+ rta->rta_len = RTA_LENGTH(sizeof(struct tc_fifo_qopt));
+ ((struct tc_fifo_qopt *)RTA_DATA(rta))->limit = limit;
+
+ ret = q->ops->change(q, rta);
+ kfree(rta);
+
+ if (ret == 0)
+ return q;
+ }
+ qdisc_destroy(q);
+ }
+ return NULL;
+}
+
static int red_change(struct Qdisc *sch, struct rtattr *opt)
{
struct red_sched_data *q = qdisc_priv(sch);
- struct rtattr *tb[TCA_RED_STAB];
+ struct rtattr *tb[TCA_RED_MAX];
struct tc_red_qopt *ctl;
+ struct Qdisc *child = NULL;
+
+ if (opt == NULL || rtattr_parse_nested(tb, TCA_RED_MAX, opt))
+ return -EINVAL;
- if (opt == NULL ||
- rtattr_parse_nested(tb, TCA_RED_STAB, opt) ||
- tb[TCA_RED_PARMS-1] == 0 || tb[TCA_RED_STAB-1] == 0 ||
+ if (tb[TCA_RED_PARMS-1] == NULL ||
RTA_PAYLOAD(tb[TCA_RED_PARMS-1]) < sizeof(*ctl) ||
- RTA_PAYLOAD(tb[TCA_RED_STAB-1]) < 256)
+ tb[TCA_RED_STAB-1] == NULL ||
+ RTA_PAYLOAD(tb[TCA_RED_STAB-1]) < RED_STAB_SIZE)
return -EINVAL;
ctl = RTA_DATA(tb[TCA_RED_PARMS-1]);
+ if (ctl->limit > 0) {
+ child = red_create_dflt(sch, ctl->limit);
+ if (child == NULL)
+ return -ENOMEM;
+ }
+
sch_tree_lock(sch);
q->flags = ctl->flags;
q->limit = ctl->limit;
+ if (child) {
+ qdisc_tree_decrease_qlen(q->qdisc, q->qdisc->q.qlen);
+ qdisc_destroy(xchg(&q->qdisc, child));
+ }
red_set_parms(&q->parms, ctl->qth_min, ctl->qth_max, ctl->Wlog,
ctl->Plog, ctl->Scell_log,
if (skb_queue_empty(&sch->q))
red_end_of_idle_period(&q->parms);
+
sch_tree_unlock(sch);
return 0;
}
static int red_init(struct Qdisc* sch, struct rtattr *opt)
{
+ struct red_sched_data *q = qdisc_priv(sch);
+
+ q->qdisc = &noop_qdisc;
return red_change(sch, opt);
}
static int red_dump(struct Qdisc *sch, struct sk_buff *skb)
{
struct red_sched_data *q = qdisc_priv(sch);
- unsigned char *b = skb->tail;
- struct rtattr *rta;
+ struct rtattr *opts = NULL;
struct tc_red_qopt opt = {
.limit = q->limit,
.flags = q->flags,
.Scell_log = q->parms.Scell_log,
};
- rta = (struct rtattr*)b;
- RTA_PUT(skb, TCA_OPTIONS, 0, NULL);
+ opts = RTA_NEST(skb, TCA_OPTIONS);
RTA_PUT(skb, TCA_RED_PARMS, sizeof(opt), &opt);
- rta->rta_len = skb->tail - b;
-
- return skb->len;
+ return RTA_NEST_END(skb, opts);
rtattr_failure:
- skb_trim(skb, b - skb->data);
- return -1;
+ return RTA_NEST_CANCEL(skb, opts);
}
static int red_dump_stats(struct Qdisc *sch, struct gnet_dump *d)
return gnet_stats_copy_app(d, &st, sizeof(st));
}
+static int red_dump_class(struct Qdisc *sch, unsigned long cl,
+ struct sk_buff *skb, struct tcmsg *tcm)
+{
+ struct red_sched_data *q = qdisc_priv(sch);
+
+ if (cl != 1)
+ return -ENOENT;
+ tcm->tcm_handle |= TC_H_MIN(1);
+ tcm->tcm_info = q->qdisc->handle;
+ return 0;
+}
+
+static int red_graft(struct Qdisc *sch, unsigned long arg, struct Qdisc *new,
+ struct Qdisc **old)
+{
+ struct red_sched_data *q = qdisc_priv(sch);
+
+ if (new == NULL)
+ new = &noop_qdisc;
+
+ sch_tree_lock(sch);
+ *old = xchg(&q->qdisc, new);
+ qdisc_tree_decrease_qlen(*old, (*old)->q.qlen);
+ qdisc_reset(*old);
+ sch_tree_unlock(sch);
+ return 0;
+}
+
+static struct Qdisc *red_leaf(struct Qdisc *sch, unsigned long arg)
+{
+ struct red_sched_data *q = qdisc_priv(sch);
+ return q->qdisc;
+}
+
+static unsigned long red_get(struct Qdisc *sch, u32 classid)
+{
+ return 1;
+}
+
+static void red_put(struct Qdisc *sch, unsigned long arg)
+{
+ return;
+}
+
+static int red_change_class(struct Qdisc *sch, u32 classid, u32 parentid,
+ struct rtattr **tca, unsigned long *arg)
+{
+ return -ENOSYS;
+}
+
+static int red_delete(struct Qdisc *sch, unsigned long cl)
+{
+ return -ENOSYS;
+}
+
+static void red_walk(struct Qdisc *sch, struct qdisc_walker *walker)
+{
+ if (!walker->stop) {
+ if (walker->count >= walker->skip)
+ if (walker->fn(sch, 1, walker) < 0) {
+ walker->stop = 1;
+ return;
+ }
+ walker->count++;
+ }
+}
+
+static struct tcf_proto **red_find_tcf(struct Qdisc *sch, unsigned long cl)
+{
+ return NULL;
+}
+
+static struct Qdisc_class_ops red_class_ops = {
+ .graft = red_graft,
+ .leaf = red_leaf,
+ .get = red_get,
+ .put = red_put,
+ .change = red_change_class,
+ .delete = red_delete,
+ .walk = red_walk,
+ .tcf_chain = red_find_tcf,
+ .dump = red_dump_class,
+};
+
static struct Qdisc_ops red_qdisc_ops = {
- .next = NULL,
- .cl_ops = NULL,
.id = "red",
.priv_size = sizeof(struct red_sched_data),
+ .cl_ops = &red_class_ops,
.enqueue = red_enqueue,
.dequeue = red_dequeue,
.requeue = red_requeue,
.drop = red_drop,
.init = red_init,
.reset = red_reset,
+ .destroy = red_destroy,
.change = red_change,
.dump = red_dump,
.dump_stats = red_dump_stats,
{
return register_qdisc(&red_qdisc_ops);
}
-static void __exit red_module_exit(void)
+
+static void __exit red_module_exit(void)
{
unregister_qdisc(&red_qdisc_ops);
}
+
module_init(red_module_init)
module_exit(red_module_exit)
+
MODULE_LICENSE("GPL");
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff