From: John Fastabend <john.fastabend@gmail.com>
To: daniel@iogearbox.net, eric.dumazet@gmail.com, jhs@mojatatu.com,
aduyck@mirantis.com, brouer@redhat.com, davem@davemloft.net
Cc: john.r.fastabend@intel.com, netdev@vger.kernel.org,
john.fastabend@gmail.com
Subject: [RFC PATCH 09/12] net: sched: pfifo_fast use alf_queue
Date: Wed, 30 Dec 2015 09:54:20 -0800 [thread overview]
Message-ID: <20151230175420.26257.868.stgit@john-Precision-Tower-5810> (raw)
In-Reply-To: <20151230175000.26257.41532.stgit@john-Precision-Tower-5810>
This converts the pfifo_fast qdisc to use the alf_queue enqueue
and dequeue routines then sets the NOLOCK bit.
This also removes the logic used to pick the next band to dequeue
from and instead just checks each alf_queue for packets from
top priority to lowest. This might need to be a bit more clever
but seems to work for now.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
---
net/sched/sch_generic.c | 120 +++++++++++++++++++++++++----------------------
1 file changed, 65 insertions(+), 55 deletions(-)
diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c
index be5d63a..480cf63 100644
--- a/net/sched/sch_generic.c
+++ b/net/sched/sch_generic.c
@@ -26,6 +26,7 @@
#include <linux/list.h>
#include <linux/slab.h>
#include <linux/if_vlan.h>
+#include <linux/alf_queue.h>
#include <net/sch_generic.h>
#include <net/pkt_sched.h>
#include <net/dst.h>
@@ -467,87 +468,80 @@ static const u8 prio2band[TC_PRIO_MAX + 1] = {
/*
* Private data for a pfifo_fast scheduler containing:
- * - queues for the three band
- * - bitmap indicating which of the bands contain skbs
+ * - rings for the priority bands
*/
struct pfifo_fast_priv {
- u32 bitmap;
- struct sk_buff_head q[PFIFO_FAST_BANDS];
+ struct alf_queue *q[PFIFO_FAST_BANDS];
};
-/*
- * Convert a bitmap to the first band number where an skb is queued, where:
- * bitmap=0 means there are no skbs on any band.
- * bitmap=1 means there is an skb on band 0.
- * bitmap=7 means there are skbs on all 3 bands, etc.
- */
-static const int bitmap2band[] = {-1, 0, 1, 0, 2, 0, 1, 0};
-
-static inline struct sk_buff_head *band2list(struct pfifo_fast_priv *priv,
- int band)
+static inline struct alf_queue *band2list(struct pfifo_fast_priv *priv,
+ int band)
{
- return priv->q + band;
+ return priv->q[band];
}
static int pfifo_fast_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);
- }
-
- return qdisc_drop(skb, qdisc);
-}
-
-static struct sk_buff *pfifo_fast_dequeue(struct Qdisc *qdisc)
-{
+ int band = prio2band[skb->priority & TC_PRIO_MAX];
struct pfifo_fast_priv *priv = qdisc_priv(qdisc);
- int band = bitmap2band[priv->bitmap];
+ struct alf_queue *q = band2list(priv, band);
+ int n;
- if (likely(band >= 0)) {
- struct sk_buff_head *list = band2list(priv, band);
- struct sk_buff *skb = __qdisc_dequeue_head(qdisc, list);
+ if (!q) {
+ WARN_ON(1);
+ return qdisc_drop(skb, qdisc);
+ }
- qdisc->q.qlen--;
- if (skb_queue_empty(list))
- priv->bitmap &= ~(1 << band);
+ n = alf_mp_enqueue(q, &skb, 1);
- return skb;
+ /* If queue is overrun fall through to drop */
+ if (n) {
+ qdisc_qstats_cpu_qlen_inc(qdisc);
+ qdisc_qstats_cpu_backlog_inc(qdisc, skb);
+ return NET_XMIT_SUCCESS;
}
- return NULL;
+ return qdisc_drop_cpu(skb, qdisc);
}
-static struct sk_buff *pfifo_fast_peek(struct Qdisc *qdisc)
+static struct sk_buff *pfifo_fast_dequeue(struct Qdisc *qdisc)
{
struct pfifo_fast_priv *priv = qdisc_priv(qdisc);
- int band = bitmap2band[priv->bitmap];
+ struct sk_buff *skb = NULL;
+ int band;
+
+ for (band = 0; band < PFIFO_FAST_BANDS && !skb; band++) {
+ struct alf_queue *q = band2list(priv, band);
- if (band >= 0) {
- struct sk_buff_head *list = band2list(priv, band);
+ if (alf_queue_empty(q))
+ continue;
- return skb_peek(list);
+ alf_mc_dequeue(q, &skb, 1);
}
- return NULL;
+ if (likely(skb)) {
+ qdisc_qstats_cpu_backlog_dec(qdisc, skb);
+ qdisc_bstats_cpu_update(qdisc, skb);
+ qdisc_qstats_cpu_qlen_dec(qdisc);
+ }
+
+ return skb;
}
static void pfifo_fast_reset(struct Qdisc *qdisc)
{
- int prio;
+ int i, band;
struct pfifo_fast_priv *priv = qdisc_priv(qdisc);
- for (prio = 0; prio < PFIFO_FAST_BANDS; prio++)
- __qdisc_reset_queue(qdisc, band2list(priv, prio));
+ for (band = 0; band < PFIFO_FAST_BANDS; band++)
+ alf_queue_flush(band2list(priv, band));
- priv->bitmap = 0;
- qdisc->qstats.backlog = 0;
- qdisc->q.qlen = 0;
+ for_each_possible_cpu(i) {
+ struct gnet_stats_queue *q = per_cpu_ptr(qdisc->cpu_qstats, i);
+
+ q->backlog = 0;
+ q->qlen = 0;
+ }
}
static int pfifo_fast_dump(struct Qdisc *qdisc, struct sk_buff *skb)
@@ -565,14 +559,30 @@ nla_put_failure:
static int pfifo_fast_init(struct Qdisc *qdisc, struct nlattr *opt)
{
- int prio;
+ unsigned int qlen = qdisc_dev(qdisc)->tx_queue_len;
struct pfifo_fast_priv *priv = qdisc_priv(qdisc);
+ int prio;
+
+ /* guard against zero length rings */
+ if (!qlen)
+ return -EINVAL;
+
+ for (prio = 0; prio < PFIFO_FAST_BANDS; prio++) {
+ unsigned int size = roundup_pow_of_two(qlen);
+ struct alf_queue *q;
- for (prio = 0; prio < PFIFO_FAST_BANDS; prio++)
- __skb_queue_head_init(band2list(priv, prio));
+ q = alf_queue_alloc(size, GFP_KERNEL);
+ if (IS_ERR_OR_NULL(q))
+ return -ENOMEM;
+
+ priv->q[prio] = q;
+ }
/* Can by-pass the queue discipline */
qdisc->flags |= TCQ_F_CAN_BYPASS;
+ qdisc->flags |= TCQ_F_NOLOCK;
+ qdisc->flags |= TCQ_F_CPUSTATS;
+
return 0;
}
@@ -581,7 +591,7 @@ struct Qdisc_ops pfifo_fast_ops __read_mostly = {
.priv_size = sizeof(struct pfifo_fast_priv),
.enqueue = pfifo_fast_enqueue,
.dequeue = pfifo_fast_dequeue,
- .peek = pfifo_fast_peek,
+ .peek = qdisc_peek_dequeued,
.init = pfifo_fast_init,
.reset = pfifo_fast_reset,
.dump = pfifo_fast_dump,
next prev parent reply other threads:[~2015-12-30 17:54 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-12-30 17:50 [RFC PATCH 00/12] drop the qdisc lock for pfifo_fast/mq John Fastabend
2015-12-30 17:51 ` [RFC PATCH 01/12] lib: array based lock free queue John Fastabend
2016-01-13 19:28 ` Jesper Dangaard Brouer
2015-12-30 17:51 ` [RFC PATCH 02/12] net: sched: free per cpu bstats John Fastabend
2016-01-04 15:21 ` Daniel Borkmann
2016-01-04 17:32 ` Eric Dumazet
2016-01-04 18:08 ` John Fastabend
2015-12-30 17:51 ` [RFC PATCH 03/12] net: sched: allow qdiscs to handle locking John Fastabend
2015-12-30 17:52 ` [RFC PATCH 04/12] net: sched: provide per cpu qstat helpers John Fastabend
2015-12-30 17:52 ` [RFC PATCH 05/12] net: sched: per cpu gso handlers John Fastabend
2015-12-30 20:26 ` Jesper Dangaard Brouer
2015-12-30 20:42 ` John Fastabend
2015-12-30 17:53 ` [RFC PATCH 06/12] net: sched: support qdisc_reset on NOLOCK qdisc John Fastabend
2016-01-01 2:30 ` Alexei Starovoitov
2016-01-03 19:37 ` John Fastabend
2016-01-13 16:20 ` David Miller
2016-01-13 18:03 ` John Fastabend
2016-01-15 19:44 ` David Miller
2015-12-30 17:53 ` [RFC PATCH 07/12] net: sched: qdisc_qlen for per cpu logic John Fastabend
2015-12-30 17:53 ` [RFC PATCH 08/12] net: sched: a dflt qdisc may be used with per cpu stats John Fastabend
2015-12-30 17:54 ` John Fastabend [this message]
2016-01-13 16:24 ` [RFC PATCH 09/12] net: sched: pfifo_fast use alf_queue David Miller
2016-01-13 18:18 ` John Fastabend
2015-12-30 17:54 ` [RFC PATCH 10/12] net: sched: helper to sum qlen John Fastabend
2015-12-30 17:55 ` [RFC PATCH 11/12] net: sched: add support for TCQ_F_NOLOCK subqueues to sch_mq John Fastabend
2015-12-30 17:55 ` [RFC PATCH 12/12] net: sched: pfifo_fast new option to deque multiple pkts John Fastabend
2015-12-30 18:13 ` John Fastabend
2016-01-06 13:14 ` [RFC PATCH 00/12] drop the qdisc lock for pfifo_fast/mq Jamal Hadi Salim
2016-01-07 23:30 ` 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=20151230175420.26257.868.stgit@john-Precision-Tower-5810 \
--to=john.fastabend@gmail.com \
--cc=aduyck@mirantis.com \
--cc=brouer@redhat.com \
--cc=daniel@iogearbox.net \
--cc=davem@davemloft.net \
--cc=eric.dumazet@gmail.com \
--cc=jhs@mojatatu.com \
--cc=john.r.fastabend@intel.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