summaryrefslogtreecommitdiffstats
path: root/net/sched
diff options
context:
space:
mode:
authorEric Dumazet <edumazet@google.com>2012-05-10 07:51:25 +0000
committerDavid S. Miller <davem@davemloft.net>2012-05-10 23:35:02 -0400
commit76e3cc126bb223013a6b9a0e2a51238d1ef2e409 (patch)
tree37d1c2a3c4f4ebf68e9849262c7d75115652313f /net/sched
parent2dd875ff31ac7ff42d6fc7d7f78ac6c0635439f5 (diff)
downloadblackbird-op-linux-76e3cc126bb223013a6b9a0e2a51238d1ef2e409.tar.gz
blackbird-op-linux-76e3cc126bb223013a6b9a0e2a51238d1ef2e409.zip
codel: Controlled Delay AQM
An implementation of CoDel AQM, from Kathleen Nichols and Van Jacobson. http://queue.acm.org/detail.cfm?id=2209336 This AQM main input is no longer queue size in bytes or packets, but the delay packets stay in (FIFO) queue. As we don't have infinite memory, we still can drop packets in enqueue() in case of massive load, but mean of CoDel is to drop packets in dequeue(), using a control law based on two simple parameters : target : target sojourn time (default 5ms) interval : width of moving time window (default 100ms) Based on initial work from Dave Taht. Refactored to help future codel inclusion as a plugin for other linux qdisc (FQ_CODEL, ...), like RED. include/net/codel.h contains codel algorithm as close as possible than Kathleen reference. net/sched/sch_codel.c contains the linux qdisc specific glue. Separate structures permit a memory efficient implementation of fq_codel (to be sent as a separate work) : Each flow has its own struct codel_vars. timestamps are taken at enqueue() time with 1024 ns precision, allowing a range of 2199 seconds in queue, and 100Gb links support. iproute2 uses usec as base unit. Selected packets are dropped, unless ECN is enabled and packets can get ECN mark instead. Tested from 2Mb to 10Gb speeds with no particular problems, on ixgbe and tg3 drivers (BQL enabled). Usage: tc qdisc ... codel [ limit PACKETS ] [ target TIME ] [ interval TIME ] [ ecn ] qdisc codel 10: parent 1:1 limit 2000p target 3.0ms interval 60.0ms ecn Sent 13347099587 bytes 8815805 pkt (dropped 0, overlimits 0 requeues 0) rate 202365Kbit 16708pps backlog 113550b 75p requeues 0 count 116 lastcount 98 ldelay 4.3ms dropping drop_next 816us maxpacket 1514 ecn_mark 84399 drop_overlimit 0 CoDel must be seen as a base module, and should be used keeping in mind there is still a FIFO queue. So a typical setup will probably need a hierarchy of several qdiscs and packet classifiers to be able to meet whatever constraints a user might have. One possible example would be to use fq_codel, which combines Fair Queueing and CoDel, in replacement of sfq / sfq_red. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Dave Taht <dave.taht@bufferbloat.net> Cc: Kathleen Nichols <nichols@pollere.com> Cc: Van Jacobson <van@pollere.net> Cc: Tom Herbert <therbert@google.com> Cc: Matt Mathis <mattmathis@google.com> Cc: Yuchung Cheng <ycheng@google.com> Cc: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/sched')
-rw-r--r--net/sched/Kconfig11
-rw-r--r--net/sched/Makefile1
-rw-r--r--net/sched/sch_codel.c275
3 files changed, 287 insertions, 0 deletions
diff --git a/net/sched/Kconfig b/net/sched/Kconfig
index 75b58f81d53d..fadd2522053d 100644
--- a/net/sched/Kconfig
+++ b/net/sched/Kconfig
@@ -250,6 +250,17 @@ config NET_SCH_QFQ
If unsure, say N.
+config NET_SCH_CODEL
+ tristate "Controlled Delay AQM (CODEL)"
+ help
+ Say Y here if you want to use the Controlled Delay (CODEL)
+ packet scheduling algorithm.
+
+ To compile this driver as a module, choose M here: the module
+ will be called sch_codel.
+
+ If unsure, say N.
+
config NET_SCH_INGRESS
tristate "Ingress Qdisc"
depends on NET_CLS_ACT
diff --git a/net/sched/Makefile b/net/sched/Makefile
index 8cdf4e2b51d3..30fab03b8516 100644
--- a/net/sched/Makefile
+++ b/net/sched/Makefile
@@ -37,6 +37,7 @@ obj-$(CONFIG_NET_SCH_PLUG) += sch_plug.o
obj-$(CONFIG_NET_SCH_MQPRIO) += sch_mqprio.o
obj-$(CONFIG_NET_SCH_CHOKE) += sch_choke.o
obj-$(CONFIG_NET_SCH_QFQ) += sch_qfq.o
+obj-$(CONFIG_NET_SCH_CODEL) += sch_codel.o
obj-$(CONFIG_NET_CLS_U32) += cls_u32.o
obj-$(CONFIG_NET_CLS_ROUTE4) += cls_route.o
diff --git a/net/sched/sch_codel.c b/net/sched/sch_codel.c
new file mode 100644
index 000000000000..b4a1a81e757e
--- /dev/null
+++ b/net/sched/sch_codel.c
@@ -0,0 +1,275 @@
+/*
+ * Codel - The Controlled-Delay Active Queue Management algorithm
+ *
+ * Copyright (C) 2011-2012 Kathleen Nichols <nichols@pollere.com>
+ * Copyright (C) 2011-2012 Van Jacobson <van@pollere.net>
+ *
+ * Implemented on linux by :
+ * Copyright (C) 2012 Michael D. Taht <dave.taht@bufferbloat.net>
+ * Copyright (C) 2012 Eric Dumazet <edumazet@google.com>
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions, and the following disclaimer,
+ * without modification.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ * 3. The names of the authors may not be used to endorse or promote products
+ * derived from this software without specific prior written permission.
+ *
+ * Alternatively, provided that this notice is retained in full, this
+ * software may be distributed under the terms of the GNU General
+ * Public License ("GPL") version 2, in which case the provisions of the
+ * GPL apply INSTEAD OF those given above.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH
+ * DAMAGE.
+ *
+ */
+
+#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>
+#include <net/codel.h>
+
+
+#define DEFAULT_CODEL_LIMIT 1000
+
+struct codel_sched_data {
+ struct codel_params params;
+ struct codel_vars vars;
+ struct codel_stats stats;
+ u32 drop_overlimit;
+};
+
+/* This is the specific function called from codel_dequeue()
+ * to dequeue a packet from queue. Note: backlog is handled in
+ * codel, we dont need to reduce it here.
+ */
+static struct sk_buff *dequeue(struct codel_vars *vars, struct Qdisc *sch)
+{
+ struct sk_buff *skb = __skb_dequeue(&sch->q);
+
+ prefetch(&skb->end); /* we'll need skb_shinfo() */
+ return skb;
+}
+
+static struct sk_buff *codel_qdisc_dequeue(struct Qdisc *sch)
+{
+ struct codel_sched_data *q = qdisc_priv(sch);
+ struct sk_buff *skb;
+
+ skb = codel_dequeue(sch, &q->params, &q->vars, &q->stats,
+ dequeue, &sch->qstats.backlog);
+ /* We cant call qdisc_tree_decrease_qlen() if our qlen is 0,
+ * or HTB crashes. Defer it for next round.
+ */
+ if (q->stats.drop_count && sch->q.qlen) {
+ qdisc_tree_decrease_qlen(sch, q->stats.drop_count);
+ q->stats.drop_count = 0;
+ }
+ if (skb)
+ qdisc_bstats_update(sch, skb);
+ return skb;
+}
+
+static int codel_qdisc_enqueue(struct sk_buff *skb, struct Qdisc *sch)
+{
+ struct codel_sched_data *q;
+
+ if (likely(qdisc_qlen(sch) < sch->limit)) {
+ codel_set_enqueue_time(skb);
+ return qdisc_enqueue_tail(skb, sch);
+ }
+ q = qdisc_priv(sch);
+ q->drop_overlimit++;
+ return qdisc_drop(skb, sch);
+}
+
+static const struct nla_policy codel_policy[TCA_CODEL_MAX + 1] = {
+ [TCA_CODEL_TARGET] = { .type = NLA_U32 },
+ [TCA_CODEL_LIMIT] = { .type = NLA_U32 },
+ [TCA_CODEL_INTERVAL] = { .type = NLA_U32 },
+ [TCA_CODEL_ECN] = { .type = NLA_U32 },
+};
+
+static int codel_change(struct Qdisc *sch, struct nlattr *opt)
+{
+ struct codel_sched_data *q = qdisc_priv(sch);
+ struct nlattr *tb[TCA_CODEL_MAX + 1];
+ unsigned int qlen;
+ int err;
+
+ if (!opt)
+ return -EINVAL;
+
+ err = nla_parse_nested(tb, TCA_CODEL_MAX, opt, codel_policy);
+ if (err < 0)
+ return err;
+
+ sch_tree_lock(sch);
+
+ if (tb[TCA_CODEL_TARGET]) {
+ u32 target = nla_get_u32(tb[TCA_CODEL_TARGET]);
+
+ q->params.target = ((u64)target * NSEC_PER_USEC) >> CODEL_SHIFT;
+ }
+
+ if (tb[TCA_CODEL_INTERVAL]) {
+ u32 interval = nla_get_u32(tb[TCA_CODEL_INTERVAL]);
+
+ q->params.interval = ((u64)interval * NSEC_PER_USEC) >> CODEL_SHIFT;
+ }
+
+ if (tb[TCA_CODEL_LIMIT])
+ sch->limit = nla_get_u32(tb[TCA_CODEL_LIMIT]);
+
+ if (tb[TCA_CODEL_ECN])
+ q->params.ecn = !!nla_get_u32(tb[TCA_CODEL_ECN]);
+
+ qlen = sch->q.qlen;
+ while (sch->q.qlen > sch->limit) {
+ struct sk_buff *skb = __skb_dequeue(&sch->q);
+
+ sch->qstats.backlog -= qdisc_pkt_len(skb);
+ qdisc_drop(skb, sch);
+ }
+ qdisc_tree_decrease_qlen(sch, qlen - sch->q.qlen);
+
+ sch_tree_unlock(sch);
+ return 0;
+}
+
+static int codel_init(struct Qdisc *sch, struct nlattr *opt)
+{
+ struct codel_sched_data *q = qdisc_priv(sch);
+
+ sch->limit = DEFAULT_CODEL_LIMIT;
+
+ codel_params_init(&q->params);
+ codel_vars_init(&q->vars);
+ codel_stats_init(&q->stats);
+
+ if (opt) {
+ int err = codel_change(sch, opt);
+
+ if (err)
+ return err;
+ }
+
+ if (sch->limit >= 1)
+ sch->flags |= TCQ_F_CAN_BYPASS;
+ else
+ sch->flags &= ~TCQ_F_CAN_BYPASS;
+
+ return 0;
+}
+
+static int codel_dump(struct Qdisc *sch, struct sk_buff *skb)
+{
+ struct codel_sched_data *q = qdisc_priv(sch);
+ struct nlattr *opts;
+
+ opts = nla_nest_start(skb, TCA_OPTIONS);
+ if (opts == NULL)
+ goto nla_put_failure;
+
+ if (nla_put_u32(skb, TCA_CODEL_TARGET,
+ codel_time_to_us(q->params.target)) ||
+ nla_put_u32(skb, TCA_CODEL_LIMIT,
+ sch->limit) ||
+ nla_put_u32(skb, TCA_CODEL_INTERVAL,
+ codel_time_to_us(q->params.interval)) ||
+ nla_put_u32(skb, TCA_CODEL_ECN,
+ q->params.ecn))
+ goto nla_put_failure;
+
+ return nla_nest_end(skb, opts);
+
+nla_put_failure:
+ nla_nest_cancel(skb, opts);
+ return -1;
+}
+
+static int codel_dump_stats(struct Qdisc *sch, struct gnet_dump *d)
+{
+ const struct codel_sched_data *q = qdisc_priv(sch);
+ struct tc_codel_xstats st = {
+ .maxpacket = q->stats.maxpacket,
+ .count = q->vars.count,
+ .lastcount = q->vars.lastcount,
+ .drop_overlimit = q->drop_overlimit,
+ .ldelay = codel_time_to_us(q->vars.ldelay),
+ .dropping = q->vars.dropping,
+ .ecn_mark = q->stats.ecn_mark,
+ };
+
+ if (q->vars.dropping) {
+ codel_tdiff_t delta = q->vars.drop_next - codel_get_time();
+
+ if (delta >= 0)
+ st.drop_next = codel_time_to_us(delta);
+ else
+ st.drop_next = -codel_time_to_us(-delta);
+ }
+
+ return gnet_stats_copy_app(d, &st, sizeof(st));
+}
+
+static void codel_reset(struct Qdisc *sch)
+{
+ struct codel_sched_data *q = qdisc_priv(sch);
+
+ qdisc_reset_queue(sch);
+ codel_vars_init(&q->vars);
+}
+
+static struct Qdisc_ops codel_qdisc_ops __read_mostly = {
+ .id = "codel",
+ .priv_size = sizeof(struct codel_sched_data),
+
+ .enqueue = codel_qdisc_enqueue,
+ .dequeue = codel_qdisc_dequeue,
+ .peek = qdisc_peek_dequeued,
+ .init = codel_init,
+ .reset = codel_reset,
+ .change = codel_change,
+ .dump = codel_dump,
+ .dump_stats = codel_dump_stats,
+ .owner = THIS_MODULE,
+};
+
+static int __init codel_module_init(void)
+{
+ return register_qdisc(&codel_qdisc_ops);
+}
+
+static void __exit codel_module_exit(void)
+{
+ unregister_qdisc(&codel_qdisc_ops);
+}
+
+module_init(codel_module_init)
+module_exit(codel_module_exit)
+
+MODULE_DESCRIPTION("Controlled Delay queue discipline");
+MODULE_AUTHOR("Dave Taht");
+MODULE_AUTHOR("Eric Dumazet");
+MODULE_LICENSE("Dual BSD/GPL");
OpenPOWER on IntegriCloud