From: vincent.guittot@linaro.org (Vincent Guittot)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH 06/14] sched: add a knob to choose the packing level
Date: Thu, 25 Apr 2013 19:23:22 +0200 [thread overview]
Message-ID: <1366910611-20048-7-git-send-email-vincent.guittot@linaro.org> (raw)
In-Reply-To: <1366910611-20048-1-git-send-email-vincent.guittot@linaro.org>
There are 3 packing levels:
- the default one only packs the small tasks when the system is not busy
- the none level doesn't pack anything
- the full level uses as few as possible number of CPUs based on the current
activity of the system
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
---
include/linux/sched/sysctl.h | 8 ++++++++
kernel/sched/fair.c | 12 ++++++++++++
kernel/sysctl.c | 13 +++++++++++++
3 files changed, 33 insertions(+)
diff --git a/include/linux/sched/sysctl.h b/include/linux/sched/sysctl.h
index bf8086b..b72a8b8 100644
--- a/include/linux/sched/sysctl.h
+++ b/include/linux/sched/sysctl.h
@@ -44,6 +44,14 @@ enum sched_tunable_scaling {
};
extern enum sched_tunable_scaling sysctl_sched_tunable_scaling;
+enum sched_tunable_packing {
+ SCHED_PACKING_NONE,
+ SCHED_PACKING_DEFAULT,
+ SCHED_PACKING_FULL,
+};
+
+extern enum sched_tunable_packing __read_mostly sysctl_sched_packing_mode;
+
extern unsigned int sysctl_numa_balancing_scan_delay;
extern unsigned int sysctl_numa_balancing_scan_period_min;
extern unsigned int sysctl_numa_balancing_scan_period_max;
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index a985c98..98166aa 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -113,6 +113,18 @@ unsigned int __read_mostly sysctl_sched_shares_window = 10000000UL;
unsigned int sysctl_sched_cfs_bandwidth_slice = 5000UL;
#endif
+#ifdef CONFIG_SMP
+/*
+ * The packing policy of the scheduler
+ *
+ * Options are:
+ * SCHED_PACKING_NONE - No buddy is used for packing some tasks
+ * SCHED_PACKING_DEFAULT - The small tasks are packed on a not busy CPUs
+ * SCHED_PACKING_FULL - All Tasks are packed in a minimum number of CPUs
+ */
+enum sched_tunable_packing sysctl_sched_packing_mode = SCHED_PACKING_DEFAULT;
+
+#endif
/*
* Increase the granularity value when there are more CPUs,
* because with more CPUs the 'effective latency' as visible
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index afc1dc6..ca22f59 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -265,6 +265,8 @@ static int max_wakeup_granularity_ns = NSEC_PER_SEC; /* 1 second */
#ifdef CONFIG_SMP
static int min_sched_tunable_scaling = SCHED_TUNABLESCALING_NONE;
static int max_sched_tunable_scaling = SCHED_TUNABLESCALING_END-1;
+static int min_sched_packing_mode = SCHED_PACKING_NONE;
+static int max_sched_packing_mode = SCHED_PACKING_FULL-1;
#endif /* CONFIG_SMP */
#endif /* CONFIG_SCHED_DEBUG */
@@ -281,6 +283,17 @@ static struct ctl_table kern_table[] = {
.mode = 0644,
.proc_handler = proc_dointvec,
},
+#ifdef CONFIG_SMP
+ {
+ .procname = "sched_packing_mode",
+ .data = &sysctl_sched_packing_mode,
+ .maxlen = sizeof(enum sched_tunable_packing),
+ .mode = 0644,
+ .proc_handler = proc_dointvec,
+ .extra1 = &min_sched_packing_mode,
+ .extra2 = &max_sched_packing_mode,
+ },
+#endif
#ifdef CONFIG_SCHED_DEBUG
{
.procname = "sched_min_granularity_ns",
--
1.7.9.5
next prev parent reply other threads:[~2013-04-25 17:23 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-04-25 17:23 [RFC PATCH v4 00/14] sched: packing small tasks Vincent Guittot
2013-04-25 17:23 ` [PATCH 01/14] Revert "sched: Introduce temporary FAIR_GROUP_SCHED dependency for load-tracking" Vincent Guittot
2013-04-25 17:23 ` [PATCH 02/14] sched: add a new SD_SHARE_POWERDOMAIN flag for sched_domain Vincent Guittot
2013-04-25 17:23 ` [PATCH 03/14] sched: pack small tasks Vincent Guittot
2013-04-26 12:30 ` Peter Zijlstra
2013-04-26 13:16 ` Vincent Guittot
2013-04-26 12:38 ` Peter Zijlstra
2013-04-26 13:38 ` Vincent Guittot
2013-04-25 17:23 ` [PATCH 04/14] sched: pack the idle load balance Vincent Guittot
2013-04-26 12:49 ` Peter Zijlstra
2013-04-26 13:47 ` Vincent Guittot
2013-04-25 17:23 ` [PATCH 05/14] ARM: sched: clear SD_SHARE_POWERDOMAIN Vincent Guittot
2013-04-25 17:23 ` Vincent Guittot [this message]
2013-04-25 17:23 ` [PATCH 07/14] sched: agressively pack at wake/fork/exec Vincent Guittot
2013-04-26 13:08 ` Peter Zijlstra
2013-04-26 14:23 ` Vincent Guittot
2013-04-25 17:23 ` [PATCH 08/14] sched: trig ILB on an idle buddy Vincent Guittot
2013-04-26 13:15 ` Peter Zijlstra
2013-04-26 14:52 ` Vincent Guittot
2013-04-25 17:23 ` [PATCH 09/14] sched: evaluate the activity level of the system Vincent Guittot
2013-05-22 16:50 ` Morten Rasmussen
2013-05-23 8:11 ` Vincent Guittot
2013-04-25 17:23 ` [PATCH 10/14] sched: update the buddy CPU Vincent Guittot
2013-04-28 8:20 ` Francesco Lavra
2013-04-29 7:32 ` Vincent Guittot
2013-04-25 17:23 ` [PATCH 11/14] sched: filter task pull request Vincent Guittot
2013-04-26 10:00 ` Vincent Guittot
2013-05-22 15:56 ` Morten Rasmussen
2013-05-22 16:03 ` Vincent Guittot
2013-04-25 17:23 ` [PATCH 12/14] sched: create a new field with available capacity Vincent Guittot
2013-04-25 17:23 ` [PATCH 13/14] sched: update the cpu_power Vincent Guittot
2013-05-22 15:46 ` Morten Rasmussen
2013-05-22 15:58 ` Vincent Guittot
2013-04-25 17:23 ` [PATCH 14/14] sched: force migration on buddy CPU Vincent Guittot
2013-04-26 12:08 ` [RFC PATCH v4 00/14] sched: packing small tasks Vincent Guittot
2013-04-26 15:00 ` Arjan van de Ven
2013-04-26 15:40 ` Vincent Guittot
2013-04-26 15:46 ` Arjan van de Ven
2013-04-26 15:56 ` Vincent Guittot
2013-05-02 9:12 ` Vincent Guittot
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=1366910611-20048-7-git-send-email-vincent.guittot@linaro.org \
--to=vincent.guittot@linaro.org \
--cc=linux-arm-kernel@lists.infradead.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).