kpriv/linux-hardened
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

sched: manipulate __QDISC_STATE_RUNNING in qdisc_run_* helpers

Browse source 
Currently NOLOCK qdiscs pay a measurable overhead to atomically
manipulate the __QDISC_STATE_RUNNING. Such bit is flipped twice per
packet in the uncontended scenario with packet rate below the
line rate: on packed dequeue and on the next, failing dequeue attempt.

This changeset moves the bit manipulation into the qdisc_run_{begin,end}
helpers, so that the bit is now flipped only once per packet, with
measurable performance improvement in the uncontended scenario.

This also allows simplifying the qdisc teardown code path - since
qdisc_is_running() is now effective for each qdisc type - and avoid a
possible race between qdisc_run() and dev_deactivate_many(), as now
the some_qdisc_is_busy() can properly detect NOLOCK qdiscs being busy
dequeuing packets.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
Paolo Abeni 2018-05-15 10:50:31 +02:00 committed by David S. Miller
parent 5a63f77af7
commit 32f7b44d0f
3 changed files with 19 additions and 24 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

10
include/net/sch_generic.h
View file

@ -113,13 +113,19 @@ static inline void qdisc_refcount_inc(struct Qdisc *qdisc)
static inline bool qdisc_is_running(const struct Qdisc *qdisc)
{
if (qdisc->flags & TCQ_F_NOLOCK)
return test_bit(__QDISC_STATE_RUNNING, &qdisc->state);
return (raw_read_seqcount(&qdisc->running) & 1) ? true : false;
}
static inline bool qdisc_run_begin(struct Qdisc *qdisc)
{
if (qdisc_is_running(qdisc))
if (qdisc->flags & TCQ_F_NOLOCK) {
if (test_and_set_bit(__QDISC_STATE_RUNNING, &qdisc->state))
return false;
} else if (qdisc_is_running(qdisc)) {
return false;
}
/* Variant of write_seqcount_begin() telling lockdep a trylock
* was attempted.
*/
@ -131,6 +137,8 @@ static inline bool qdisc_run_begin(struct Qdisc *qdisc)
static inline void qdisc_run_end(struct Qdisc *qdisc)
{
write_seqcount_end(&qdisc->running);
if (qdisc->flags & TCQ_F_NOLOCK)
clear_bit(__QDISC_STATE_RUNNING, &qdisc->state);
}
static inline bool qdisc_may_bulk(const struct Qdisc *qdisc)

2
net/core/dev.c
View file

@ -3244,7 +3244,7 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q,
rc = NET_XMIT_DROP;
} else {
rc = q->enqueue(skb, q, &to_free) & NET_XMIT_MASK;
__qdisc_run(q);
qdisc_run(q);
}
if (unlikely(to_free))

21
net/sched/sch_generic.c
View file

@ -373,33 +373,24 @@ bool sch_direct_xmit(struct sk_buff *skb, struct Qdisc *q,
*/
static inline bool qdisc_restart(struct Qdisc *q, int *packets)
{
bool more, validate, nolock = q->flags & TCQ_F_NOLOCK;
spinlock_t *root_lock = NULL;
struct netdev_queue *txq;
struct net_device *dev;
struct sk_buff *skb;
bool validate;
/* Dequeue packet */
if (nolock && test_and_set_bit(__QDISC_STATE_RUNNING, &q->state))
return false;
skb = dequeue_skb(q, &validate, packets);
if (unlikely(!skb)) {
if (nolock)
clear_bit(__QDISC_STATE_RUNNING, &q->state);
if (unlikely(!skb))
return false;
}
if (!nolock)
if (!(q->flags & TCQ_F_NOLOCK))
root_lock = qdisc_lock(q);
dev = qdisc_dev(q);
txq = skb_get_tx_queue(dev, skb);
more = sch_direct_xmit(skb, q, dev, txq, root_lock, validate);
if (nolock)
clear_bit(__QDISC_STATE_RUNNING, &q->state);
return more;
return sch_direct_xmit(skb, q, dev, txq, root_lock, validate);
}
void __qdisc_run(struct Qdisc *q)
@ -1131,9 +1122,6 @@ static bool some_qdisc_is_busy(struct net_device *dev)
dev_queue = netdev_get_tx_queue(dev, i);
q = dev_queue->qdisc_sleeping;
if (q->flags & TCQ_F_NOLOCK) {
val = test_bit(__QDISC_STATE_SCHED, &q->state);
} else {
root_lock = qdisc_lock(q);
spin_lock_bh(root_lock);
@ -1141,7 +1129,6 @@ static bool some_qdisc_is_busy(struct net_device *dev)
test_bit(__QDISC_STATE_SCHED, &q->state));
spin_unlock_bh(root_lock);
}
if (val)
return true;