linux/net/sched/sch_fifo.c
Eric Dumazet 520ac30f45 net_sched: drop packets after root qdisc lock is released
Qdisc performance suffers when packets are dropped at enqueue()
time because drops (kfree_skb()) are done while qdisc lock is held,
delaying a dequeue() draining the queue.

Nominal throughput can be reduced by 50 % when this happens,
at a time we would like the dequeue() to proceed as fast as possible.

Even FQ is vulnerable to this problem, while one of FQ goals was
to provide some flow isolation.

This patch adds a 'struct sk_buff **to_free' parameter to all
qdisc->enqueue(), and in qdisc_drop() helper.

I measured a performance increase of up to 12 %, but this patch
is a prereq so that future batches in enqueue() can fly.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-06-25 12:19:35 -04:00

181 lines
4.2 KiB
C

/*
* net/sched/sch_fifo.c The simplest FIFO 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: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
*/
#include <linux/module.h>
#include <linux/slab.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/skbuff.h>
#include <net/pkt_sched.h>
/* 1 band FIFO pseudo-"scheduler" */
static int bfifo_enqueue(struct sk_buff *skb, struct Qdisc *sch,
struct sk_buff **to_free)
{
if (likely(sch->qstats.backlog + qdisc_pkt_len(skb) <= sch->limit))
return qdisc_enqueue_tail(skb, sch);
return qdisc_drop(skb, sch, to_free);
}
static int pfifo_enqueue(struct sk_buff *skb, struct Qdisc *sch,
struct sk_buff **to_free)
{
if (likely(skb_queue_len(&sch->q) < sch->limit))
return qdisc_enqueue_tail(skb, sch);
return qdisc_drop(skb, sch, to_free);
}
static int pfifo_tail_enqueue(struct sk_buff *skb, struct Qdisc *sch,
struct sk_buff **to_free)
{
if (likely(skb_queue_len(&sch->q) < sch->limit))
return qdisc_enqueue_tail(skb, sch);
/* queue full, remove one skb to fulfill the limit */
__qdisc_queue_drop_head(sch, &sch->q, to_free);
qdisc_qstats_drop(sch);
qdisc_enqueue_tail(skb, sch);
return NET_XMIT_CN;
}
static int fifo_init(struct Qdisc *sch, struct nlattr *opt)
{
bool bypass;
bool is_bfifo = sch->ops == &bfifo_qdisc_ops;
if (opt == NULL) {
u32 limit = qdisc_dev(sch)->tx_queue_len;
if (is_bfifo)
limit *= psched_mtu(qdisc_dev(sch));
sch->limit = limit;
} else {
struct tc_fifo_qopt *ctl = nla_data(opt);
if (nla_len(opt) < sizeof(*ctl))
return -EINVAL;
sch->limit = ctl->limit;
}
if (is_bfifo)
bypass = sch->limit >= psched_mtu(qdisc_dev(sch));
else
bypass = sch->limit >= 1;
if (bypass)
sch->flags |= TCQ_F_CAN_BYPASS;
else
sch->flags &= ~TCQ_F_CAN_BYPASS;
return 0;
}
static int fifo_dump(struct Qdisc *sch, struct sk_buff *skb)
{
struct tc_fifo_qopt opt = { .limit = sch->limit };
if (nla_put(skb, TCA_OPTIONS, sizeof(opt), &opt))
goto nla_put_failure;
return skb->len;
nla_put_failure:
return -1;
}
struct Qdisc_ops pfifo_qdisc_ops __read_mostly = {
.id = "pfifo",
.priv_size = 0,
.enqueue = pfifo_enqueue,
.dequeue = qdisc_dequeue_head,
.peek = qdisc_peek_head,
.init = fifo_init,
.reset = qdisc_reset_queue,
.change = fifo_init,
.dump = fifo_dump,
.owner = THIS_MODULE,
};
EXPORT_SYMBOL(pfifo_qdisc_ops);
struct Qdisc_ops bfifo_qdisc_ops __read_mostly = {
.id = "bfifo",
.priv_size = 0,
.enqueue = bfifo_enqueue,
.dequeue = qdisc_dequeue_head,
.peek = qdisc_peek_head,
.init = fifo_init,
.reset = qdisc_reset_queue,
.change = fifo_init,
.dump = fifo_dump,
.owner = THIS_MODULE,
};
EXPORT_SYMBOL(bfifo_qdisc_ops);
struct Qdisc_ops pfifo_head_drop_qdisc_ops __read_mostly = {
.id = "pfifo_head_drop",
.priv_size = 0,
.enqueue = pfifo_tail_enqueue,
.dequeue = qdisc_dequeue_head,
.peek = qdisc_peek_head,
.init = fifo_init,
.reset = qdisc_reset_queue,
.change = fifo_init,
.dump = fifo_dump,
.owner = THIS_MODULE,
};
/* Pass size change message down to embedded FIFO */
int fifo_set_limit(struct Qdisc *q, unsigned int limit)
{
struct nlattr *nla;
int ret = -ENOMEM;
/* Hack to avoid sending change message to non-FIFO */
if (strncmp(q->ops->id + 1, "fifo", 4) != 0)
return 0;
nla = kmalloc(nla_attr_size(sizeof(struct tc_fifo_qopt)), GFP_KERNEL);
if (nla) {
nla->nla_type = RTM_NEWQDISC;
nla->nla_len = nla_attr_size(sizeof(struct tc_fifo_qopt));
((struct tc_fifo_qopt *)nla_data(nla))->limit = limit;
ret = q->ops->change(q, nla);
kfree(nla);
}
return ret;
}
EXPORT_SYMBOL(fifo_set_limit);
struct Qdisc *fifo_create_dflt(struct Qdisc *sch, struct Qdisc_ops *ops,
unsigned int limit)
{
struct Qdisc *q;
int err = -ENOMEM;
q = qdisc_create_dflt(sch->dev_queue, ops, TC_H_MAKE(sch->handle, 1));
if (q) {
err = fifo_set_limit(q, limit);
if (err < 0) {
qdisc_destroy(q);
q = NULL;
}
}
return q ? : ERR_PTR(err);
}
EXPORT_SYMBOL(fifo_create_dflt);