From: Ben Greear <greearb@candelatech.com>
To: netdev@vger.kernel.org
Cc: Ben Greear <greearb@candelatech.com>
Subject: [net-next 1/2] qdisc: Allow qdiscs to provide backpressure up the stack.
Date: Wed, 25 Aug 2010 12:00:50 -0700 [thread overview]
Message-ID: <1282762851-3612-1-git-send-email-greearb@candelatech.com> (raw)
Some qdiscs, in some instances, can reliably detect when they
are about to drop a packet in the dev_queue_xmit path. In
this case, it would be nice to provide backpressure up the
stack, and NOT free the skb in the qdisc logic.
Signed-off-by: Ben Greear <greearb@candelatech.com>
---
:100644 100644 59962db... 20be932... M include/linux/netdevice.h
:100644 100644 3c8728a... 146a97a... M include/net/sch_generic.h
:100644 100644 859e30f... f360a9b... M net/core/dev.c
:100644 100644 2aeb3a4... 0692717... M net/sched/sch_generic.c
include/linux/netdevice.h | 7 +++++++
include/net/sch_generic.h | 19 +++++++++++++++++++
net/core/dev.c | 19 ++++++++++++++-----
net/sched/sch_generic.c | 20 ++++++++++++++++++++
4 files changed, 60 insertions(+), 5 deletions(-)
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 59962db..20be932 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -97,6 +97,7 @@ struct wireless_dev;
#define NET_XMIT_DROP 0x01 /* skb dropped */
#define NET_XMIT_CN 0x02 /* congestion notification */
#define NET_XMIT_POLICED 0x03 /* skb is shot by police */
+#define NET_XMIT_BUSY 0x04 /* congestion, but skb was NOT freed */
#define NET_XMIT_MASK 0x0f /* qdisc flags in net/sch_generic.h */
/* NET_XMIT_CN is special. It does not guarantee that this packet is lost. It
@@ -1296,6 +1297,12 @@ extern int dev_open(struct net_device *dev);
extern int dev_close(struct net_device *dev);
extern void dev_disable_lro(struct net_device *dev);
extern int dev_queue_xmit(struct sk_buff *skb);
+
+/* Similar to dev_queue_xmit, but if try_no_consume != 0,
+ * it may return NET_XMIT_BUSY and NOT free the skb if it detects congestion
+ */
+extern int try_dev_queue_xmit(struct sk_buff *skb, int try_no_consume);
+
extern int register_netdevice(struct net_device *dev);
extern void unregister_netdevice_queue(struct net_device *dev,
struct list_head *head);
diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h
index 3c8728a..146a97a 100644
--- a/include/net/sch_generic.h
+++ b/include/net/sch_generic.h
@@ -43,6 +43,7 @@ struct qdisc_size_table {
struct Qdisc {
int (*enqueue)(struct sk_buff *skb, struct Qdisc *dev);
+ int (*try_enqueue)(struct sk_buff *, struct Qdisc *dev); /* May return NET_XMIT_BUSY and NOT free skb. */
struct sk_buff * (*dequeue)(struct Qdisc *dev);
unsigned flags;
#define TCQ_F_BUILTIN 1
@@ -135,6 +136,7 @@ struct Qdisc_ops {
int priv_size;
int (*enqueue)(struct sk_buff *, struct Qdisc *);
+ int (*try_enqueue)(struct sk_buff *, struct Qdisc *); /* May return NET_XMIT_BUSY and NOT free skb. */
struct sk_buff * (*dequeue)(struct Qdisc *);
struct sk_buff * (*peek)(struct Qdisc *);
unsigned int (*drop)(struct Qdisc *);
@@ -426,6 +428,23 @@ static inline int qdisc_enqueue_root(struct sk_buff *skb, struct Qdisc *sch)
return qdisc_enqueue(skb, sch) & NET_XMIT_MASK;
}
+static inline int try_qdisc_enqueue(struct sk_buff *skb, struct Qdisc *sch)
+{
+#ifdef CONFIG_NET_SCHED
+ if (sch->stab)
+ qdisc_calculate_pkt_len(skb, sch->stab);
+#endif
+ if (sch->try_enqueue)
+ return sch->try_enqueue(skb, sch);
+ return sch->enqueue(skb, sch);
+}
+
+static inline int try_qdisc_enqueue_root(struct sk_buff *skb, struct Qdisc *sch)
+{
+ qdisc_skb_cb(skb)->pkt_len = skb->len;
+ return try_qdisc_enqueue(skb, sch) & NET_XMIT_MASK;
+}
+
static inline void __qdisc_update_bstats(struct Qdisc *sch, unsigned int len)
{
sch->bstats.bytes += len;
diff --git a/net/core/dev.c b/net/core/dev.c
index 859e30f..f360a9b 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -2087,7 +2087,8 @@ static struct netdev_queue *dev_pick_tx(struct net_device *dev,
static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q,
struct net_device *dev,
- struct netdev_queue *txq)
+ struct netdev_queue *txq,
+ bool try_no_consume)
{
spinlock_t *root_lock = qdisc_lock(q);
bool contended = qdisc_is_running(q);
@@ -2128,7 +2129,10 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q,
rc = NET_XMIT_SUCCESS;
} else {
skb_dst_force(skb);
- rc = qdisc_enqueue_root(skb, q);
+ if (try_no_consume)
+ rc = try_qdisc_enqueue_root(skb, q);
+ else
+ rc = qdisc_enqueue_root(skb, q);
if (qdisc_run_begin(q)) {
if (unlikely(contended)) {
spin_unlock(&q->busylock);
@@ -2168,7 +2172,12 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q,
* the BH enable code must have IRQs enabled so that it will not deadlock.
* --BLG
*/
-int dev_queue_xmit(struct sk_buff *skb)
+int dev_queue_xmit(struct sk_buff *skb) {
+ return try_dev_queue_xmit(skb, 0);
+}
+EXPORT_SYMBOL(dev_queue_xmit);
+
+int try_dev_queue_xmit(struct sk_buff *skb, int try_no_consume)
{
struct net_device *dev = skb->dev;
struct netdev_queue *txq;
@@ -2187,7 +2196,7 @@ int dev_queue_xmit(struct sk_buff *skb)
skb->tc_verd = SET_TC_AT(skb->tc_verd, AT_EGRESS);
#endif
if (q->enqueue) {
- rc = __dev_xmit_skb(skb, q, dev, txq);
+ rc = __dev_xmit_skb(skb, q, dev, txq, try_no_consume);
goto out;
}
@@ -2239,7 +2248,7 @@ out:
rcu_read_unlock_bh();
return rc;
}
-EXPORT_SYMBOL(dev_queue_xmit);
+EXPORT_SYMBOL(try_dev_queue_xmit);
/*=======================================================================
diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c
index 2aeb3a4..0692717 100644
--- a/net/sched/sch_generic.c
+++ b/net/sched/sch_generic.c
@@ -460,6 +460,24 @@ static int pfifo_fast_enqueue(struct sk_buff *skb, struct Qdisc* qdisc)
return qdisc_drop(skb, qdisc);
}
+static int pfifo_fast_try_enqueue(struct sk_buff *skb, struct Qdisc* qdisc)
+{
+ if (skb_queue_len(&qdisc->q) < qdisc_dev(qdisc)->tx_queue_len) {
+ int band = prio2band[skb->priority & TC_PRIO_MAX];
+ struct pfifo_fast_priv *priv = qdisc_priv(qdisc);
+ struct sk_buff_head *list = band2list(priv, band);
+
+ priv->bitmap |= (1 << band);
+ qdisc->q.qlen++;
+ return __qdisc_enqueue_tail(skb, qdisc, list);
+ }
+
+ /* no room to enqueue, tell calling code to back off. Do NOT free skb, that is
+ * calling code's to deal with.
+ */
+ return NET_XMIT_BUSY;
+}
+
static struct sk_buff *pfifo_fast_dequeue(struct Qdisc* qdisc)
{
struct pfifo_fast_priv *priv = qdisc_priv(qdisc);
@@ -533,6 +551,7 @@ struct Qdisc_ops pfifo_fast_ops __read_mostly = {
.id = "pfifo_fast",
.priv_size = sizeof(struct pfifo_fast_priv),
.enqueue = pfifo_fast_enqueue,
+ .try_enqueue = pfifo_fast_try_enqueue,
.dequeue = pfifo_fast_dequeue,
.peek = pfifo_fast_peek,
.init = pfifo_fast_init,
@@ -564,6 +583,7 @@ struct Qdisc *qdisc_alloc(struct netdev_queue *dev_queue,
spin_lock_init(&sch->busylock);
sch->ops = ops;
sch->enqueue = ops->enqueue;
+ sch->try_enqueue = ops->try_enqueue;
sch->dequeue = ops->dequeue;
sch->dev_queue = dev_queue;
dev_hold(qdisc_dev(sch));
--
1.6.2.5
next reply other threads:[~2010-08-25 19:00 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-08-25 19:00 Ben Greear [this message]
2010-08-25 19:00 ` [net-next 2/2] macvlan: Enable qdisc backoff logic Ben Greear
2010-08-25 19:24 ` Arnd Bergmann
2010-08-25 19:27 ` Ben Greear
2010-08-25 19:38 ` Hagen Paul Pfeifer
2010-08-25 19:49 ` Ben Greear
2010-08-25 19:59 ` Arnd Bergmann
2010-08-25 20:49 ` Ben Greear
2010-08-26 13:55 ` Arnd Bergmann
2010-08-26 15:33 ` Ben Greear
2010-08-26 17:45 ` Ben Greear
2010-08-27 13:16 ` Arnd Bergmann
2010-08-25 20:44 ` [net-next 1/2] qdisc: Allow qdiscs to provide backpressure up the stack Stephen Hemminger
2010-08-25 20:56 ` Ben Greear
2010-08-26 22:59 ` David Miller
2010-08-27 4:14 ` Ben Greear
2010-08-27 4:34 ` David Miller
2010-08-27 5:22 ` Ben Greear
2010-08-27 5:36 ` David Miller
2010-08-27 5:58 ` Ben Greear
2010-08-27 6:11 ` David Miller
2010-08-27 15:26 ` Ben Greear
2010-08-27 15:59 ` Eric Dumazet
2010-08-27 17:00 ` Ben Greear
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1282762851-3612-1-git-send-email-greearb@candelatech.com \
--to=greearb@candelatech.com \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).