netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: John Fastabend <john.fastabend@gmail.com>
To: xiyou.wangcong@gmail.com, davem@davemloft.net,
	eric.dumazet@gmail.com, jhs@mojatatu.com
Cc: netdev@vger.kernel.org, paulmck@linux.vnet.ibm.com, brouer@redhat.com
Subject: [net-next PATCH v5 16/16] net: sched: drop ingress qdisc lock
Date: Fri, 12 Sep 2014 09:35:20 -0700	[thread overview]
Message-ID: <20140912163519.19588.54966.stgit@nitbit.x32> (raw)
In-Reply-To: <20140912162748.19588.39677.stgit@nitbit.x32>

After the previous patches to make the filters RCU safe and
support per cpu counters we can drop the qdisc lock around
the ingress qdisc hook.

This is possible because the ingress qdisc is a very basic
qdisc and only updates stats and runs tc_classify. Its the
simplest qdiscs we have.

In order for the per-cpu counters to get invoked the
ingress qdisc must set the LLQDISC flag. We could use per-cpu
counters on all counters but it is only necessary when the
qdisc lock is not held.

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
---
 net/core/dev.c          |    2 --
 net/sched/sch_ingress.c |    6 ++++++
 2 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/net/core/dev.c b/net/core/dev.c
index b3d6dbc..e400df9 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3481,10 +3481,8 @@ static int ing_filter(struct sk_buff *skb, struct netdev_queue *rxq)
 
 	q = rcu_dereference(rxq->qdisc);
 	if (q != &noop_qdisc) {
-		spin_lock(qdisc_lock(q));
 		if (likely(!test_bit(__QDISC_STATE_DEACTIVATED, &q->state)))
 			result = qdisc_enqueue_root(skb, q);
-		spin_unlock(qdisc_lock(q));
 	}
 
 	return result;
diff --git a/net/sched/sch_ingress.c b/net/sched/sch_ingress.c
index 88a6289..57ea680 100644
--- a/net/sched/sch_ingress.c
+++ b/net/sched/sch_ingress.c
@@ -89,6 +89,11 @@ static int ingress_enqueue(struct sk_buff *skb, struct Qdisc *sch)
 }
 
 /* ------------------------------------------------------------- */
+static int ingress_init(struct Qdisc *sch, struct nlattr *opt)
+{
+	sch->flags |= TCQ_F_LLQDISC;
+	return 0;
+}
 
 static void ingress_destroy(struct Qdisc *sch)
 {
@@ -122,6 +127,7 @@ static const struct Qdisc_class_ops ingress_class_ops = {
 };
 
 static struct Qdisc_ops ingress_qdisc_ops __read_mostly = {
+	.init		=	ingress_init,
 	.cl_ops		=	&ingress_class_ops,
 	.id		=	"ingress",
 	.priv_size	=	sizeof(struct ingress_qdisc_data),

  parent reply	other threads:[~2014-09-12 16:35 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-09-12 16:28 [net-next PATCH v5 00/16] net/sched rcu filters John Fastabend
2014-09-12 16:28 ` [net-next PATCH v5 01/16] net: qdisc: use rcu prefix and silence sparse warnings John Fastabend
2014-09-12 16:36   ` Eric Dumazet
2014-09-12 16:29 ` [net-next PATCH v5 02/16] net: rcu-ify tcf_proto John Fastabend
2014-09-12 16:40   ` Eric Dumazet
2014-09-12 16:29 ` [net-next PATCH v5 03/16] net: sched: cls_basic use RCU John Fastabend
2014-09-12 16:30 ` [net-next PATCH v5 04/16] net: sched: cls_cgroup " John Fastabend
2014-09-12 16:30 ` [net-next PATCH v5 05/16] net: sched: cls_flow " John Fastabend
2014-09-12 16:31 ` [net-next PATCH v5 06/16] net: sched: fw " John Fastabend
2014-09-12 16:31 ` [net-next PATCH v5 07/16] net: sched: RCU cls_route John Fastabend
2014-09-12 16:31 ` [net-next PATCH v5 08/16] net: sched: RCU cls_tcindex John Fastabend
2014-09-12 16:32 ` [net-next PATCH v5 09/16] net: sched: make cls_u32 per cpu John Fastabend
2014-09-12 16:32 ` [net-next PATCH v5 10/16] net: sched: make cls_u32 lockless John Fastabend
2014-09-12 16:33 ` [net-next PATCH v5 11/16] net: sched: rcu'ify cls_rsvp John Fastabend
2014-09-12 18:21   ` Eric Dumazet
2014-09-12 16:33 ` [net-next PATCH v5 12/16] net: sched: rcu'ify cls_bpf John Fastabend
2014-09-12 18:22   ` Eric Dumazet
2014-09-12 16:34 ` [net-next PATCH v5 13/16] net: sched: make tc_action safe to walk under RCU John Fastabend
2014-09-12 20:51   ` Eric Dumazet
2014-09-13  0:05     ` John Fastabend
2014-09-12 16:34 ` [net-next PATCH v5 14/16] net: sched: make bstats per cpu and estimator RCU safe John Fastabend
2014-09-12 23:49   ` Eric Dumazet
2014-09-12 16:34 ` [net-next PATCH v5 15/16] net: sched: make qstats per cpu John Fastabend
2014-09-12 23:55   ` Eric Dumazet
2014-09-12 16:35 ` John Fastabend [this message]
2014-09-12 23:56   ` [net-next PATCH v5 16/16] net: sched: drop ingress qdisc lock Eric Dumazet
2014-09-13  1:33     ` Eric Dumazet
2014-09-13  2:36 ` [net-next PATCH v5 00/16] net/sched rcu filters David Miller
2014-09-13  2:49   ` Eric Dumazet
2014-09-13  3:12     ` John Fastabend

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=20140912163519.19588.54966.stgit@nitbit.x32 \
    --to=john.fastabend@gmail.com \
    --cc=brouer@redhat.com \
    --cc=davem@davemloft.net \
    --cc=eric.dumazet@gmail.com \
    --cc=jhs@mojatatu.com \
    --cc=netdev@vger.kernel.org \
    --cc=paulmck@linux.vnet.ibm.com \
    --cc=xiyou.wangcong@gmail.com \
    /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).