From: Morten Rasmussen <morten.rasmussen@arm.com>
To: peterz@infradead.org, mingo@redhat.com
Cc: vincent.guittot@linaro.org, dietmar.eggemann@arm.com,
yuyang.du@intel.com, preeti@linux.vnet.ibm.com,
mturquette@linaro.org, nico@linaro.org, rjw@rjwysocki.net,
juri.lelli@arm.com, linux-kernel@vger.kernel.org,
Russell King <linux@arm.linux.org.uk>
Subject: [RFCv3 PATCH 16/48] arm: Cpu invariant scheduler load-tracking support
Date: Wed, 4 Feb 2015 18:30:53 +0000 [thread overview]
Message-ID: <1423074685-6336-17-git-send-email-morten.rasmussen@arm.com> (raw)
In-Reply-To: <1423074685-6336-1-git-send-email-morten.rasmussen@arm.com>
From: Dietmar Eggemann <dietmar.eggemann@arm.com>
Reuses the existing infrastructure for cpu_scale to provide the scheduler
with a cpu scaling correction factor for more accurate load-tracking.
This factor comprises a micro-architectural part, which is based on the
cpu efficiency value of a cpu as well as a platform-wide max frequency
part, which relates to the dtb property clock-frequency of a cpu node.
The calculation of cpu_scale, return value of arch_scale_cpu_capacity,
changes from:
capacity / middle_capacity
with capacity = (clock_frequency >> 20) * cpu_efficiency
to:
SCHED_CAPACITY_SCALE * cpu_perf / max_cpu_perf
The range of the cpu_scale value changes from
[0..3*SCHED_CAPACITY_SCALE/2] to [0..SCHED_CAPACITY_SCALE].
The functionality to calculate the middle_capacity which corresponds to an
'average' cpu has been taken out since the scaling is now done
differently.
In the case that either the cpu efficiency or the clock-frequency value
for a cpu is missing, no cpu scaling is done for any cpu.
The platform-wide max frequency part of the factor should not be confused
with the frequency invariant scheduler load-tracking support which deals
with frequency related scaling due to DFVS functionality on a cpu.
Cc: Russell King <linux@arm.linux.org.uk>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
---
arch/arm/kernel/topology.c | 64 +++++++++++++++++-----------------------------
1 file changed, 23 insertions(+), 41 deletions(-)
diff --git a/arch/arm/kernel/topology.c b/arch/arm/kernel/topology.c
index a1274e6..34ecbdc 100644
--- a/arch/arm/kernel/topology.c
+++ b/arch/arm/kernel/topology.c
@@ -62,9 +62,7 @@ struct cpu_efficiency {
* Table of relative efficiency of each processors
* The efficiency value must fit in 20bit and the final
* cpu_scale value must be in the range
- * 0 < cpu_scale < 3*SCHED_CAPACITY_SCALE/2
- * in order to return at most 1 when DIV_ROUND_CLOSEST
- * is used to compute the capacity of a CPU.
+ * 0 < cpu_scale < SCHED_CAPACITY_SCALE.
* Processors that are not defined in the table,
* use the default SCHED_CAPACITY_SCALE value for cpu_scale.
*/
@@ -77,24 +75,18 @@ static const struct cpu_efficiency table_efficiency[] = {
static unsigned long *__cpu_capacity;
#define cpu_capacity(cpu) __cpu_capacity[cpu]
-static unsigned long middle_capacity = 1;
+static unsigned long max_cpu_perf;
/*
* Iterate all CPUs' descriptor in DT and compute the efficiency
- * (as per table_efficiency). Also calculate a middle efficiency
- * as close as possible to (max{eff_i} - min{eff_i}) / 2
- * This is later used to scale the cpu_capacity field such that an
- * 'average' CPU is of middle capacity. Also see the comments near
- * table_efficiency[] and update_cpu_capacity().
+ * (as per table_efficiency). Calculate the max cpu performance too.
*/
+
static void __init parse_dt_topology(void)
{
const struct cpu_efficiency *cpu_eff;
struct device_node *cn = NULL;
- unsigned long min_capacity = ULONG_MAX;
- unsigned long max_capacity = 0;
- unsigned long capacity = 0;
- int cpu = 0;
+ int cpu = 0, i = 0;
__cpu_capacity = kcalloc(nr_cpu_ids, sizeof(*__cpu_capacity),
GFP_NOWAIT);
@@ -102,6 +94,7 @@ static void __init parse_dt_topology(void)
for_each_possible_cpu(cpu) {
const u32 *rate;
int len;
+ unsigned long cpu_perf;
/* too early to use cpu->of_node */
cn = of_get_cpu_node(cpu, NULL);
@@ -124,46 +117,35 @@ static void __init parse_dt_topology(void)
continue;
}
- capacity = ((be32_to_cpup(rate)) >> 20) * cpu_eff->efficiency;
-
- /* Save min capacity of the system */
- if (capacity < min_capacity)
- min_capacity = capacity;
-
- /* Save max capacity of the system */
- if (capacity > max_capacity)
- max_capacity = capacity;
-
- cpu_capacity(cpu) = capacity;
+ cpu_perf = ((be32_to_cpup(rate)) >> 20) * cpu_eff->efficiency;
+ cpu_capacity(cpu) = cpu_perf;
+ max_cpu_perf = max(max_cpu_perf, cpu_perf);
+ i++;
}
- /* If min and max capacities are equals, we bypass the update of the
- * cpu_scale because all CPUs have the same capacity. Otherwise, we
- * compute a middle_capacity factor that will ensure that the capacity
- * of an 'average' CPU of the system will be as close as possible to
- * SCHED_CAPACITY_SCALE, which is the default value, but with the
- * constraint explained near table_efficiency[].
- */
- if (4*max_capacity < (3*(max_capacity + min_capacity)))
- middle_capacity = (min_capacity + max_capacity)
- >> (SCHED_CAPACITY_SHIFT+1);
- else
- middle_capacity = ((max_capacity / 3)
- >> (SCHED_CAPACITY_SHIFT-1)) + 1;
-
+ if (i < num_possible_cpus())
+ max_cpu_perf = 0;
}
/*
* Look for a customed capacity of a CPU in the cpu_capacity table during the
* boot. The update of all CPUs is in O(n^2) for heteregeneous system but the
- * function returns directly for SMP system.
+ * function returns directly for SMP systems or if there is no complete set
+ * of cpu efficiency, clock frequency data for each cpu.
*/
static void update_cpu_capacity(unsigned int cpu)
{
- if (!cpu_capacity(cpu))
+ unsigned long capacity = cpu_capacity(cpu);
+
+ if (!capacity || !max_cpu_perf) {
+ cpu_capacity(cpu) = 0;
return;
+ }
+
+ capacity *= SCHED_CAPACITY_SCALE;
+ capacity /= max_cpu_perf;
- set_capacity_scale(cpu, cpu_capacity(cpu) / middle_capacity);
+ set_capacity_scale(cpu, capacity);
pr_info("CPU%u: update cpu_capacity %lu\n",
cpu, arch_scale_cpu_capacity(NULL, cpu));
--
1.9.1
next prev parent reply other threads:[~2015-02-04 18:40 UTC|newest]
Thread overview: 124+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-02-04 18:30 [RFCv3 PATCH 00/48] sched: Energy cost model for energy-aware scheduling Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 01/48] sched: add utilization_avg_contrib Morten Rasmussen
2015-02-11 8:50 ` Preeti U Murthy
2015-02-12 1:07 ` Vincent Guittot
2015-02-04 18:30 ` [RFCv3 PATCH 02/48] sched: Track group sched_entity usage contributions Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 03/48] sched: remove frequency scaling from cpu_capacity Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 04/48] sched: Make sched entity usage tracking frequency-invariant Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 05/48] sched: make scale_rt invariant with frequency Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 06/48] sched: add per rq cpu_capacity_orig Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 07/48] sched: get CPU's usage statistic Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 08/48] sched: replace capacity_factor by usage Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 09/48] sched: add SD_PREFER_SIBLING for SMT level Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 10/48] sched: move cfs task on a CPU with higher capacity Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 11/48] sched: Make load tracking frequency scale-invariant Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 12/48] sched: Make usage tracking cpu scale-invariant Morten Rasmussen
2015-03-23 14:46 ` Peter Zijlstra
2015-03-23 19:19 ` Dietmar Eggemann
[not found] ` <OF8A3E3617.0D4400A5-ON48257E3A.001B38D9-48257E3A.002379A4@zte.com.cn>
2015-05-06 9:49 ` Dietmar Eggemann
2015-02-04 18:30 ` [RFCv3 PATCH 13/48] cpufreq: Architecture specific callback for frequency changes Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 14/48] arm: Frequency invariant scheduler load-tracking support Morten Rasmussen
2015-03-23 13:39 ` Peter Zijlstra
2015-03-24 9:41 ` Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 15/48] arm: vexpress: Add CPU clock-frequencies to TC2 device-tree Morten Rasmussen
2015-02-04 18:30 ` Morten Rasmussen [this message]
2015-02-04 18:30 ` [RFCv3 PATCH 17/48] sched: Get rid of scaling usage by cpu_capacity_orig Morten Rasmussen
[not found] ` <OFFC493540.15A92099-ON48257E35.0026F60C-48257E35.0027A5FB@zte.com.cn>
2015-04-28 16:54 ` Dietmar Eggemann
2015-02-04 18:30 ` [RFCv3 PATCH 18/48] sched: Track blocked utilization contributions Morten Rasmussen
2015-03-23 14:08 ` Peter Zijlstra
2015-03-24 9:43 ` Morten Rasmussen
2015-03-24 16:07 ` Peter Zijlstra
2015-03-24 17:44 ` Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 19/48] sched: Include blocked utilization in usage tracking Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 20/48] sched: Documentation for scheduler energy cost model Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 21/48] sched: Make energy awareness a sched feature Morten Rasmussen
2015-02-04 18:30 ` [RFCv3 PATCH 22/48] sched: Introduce energy data structures Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 23/48] sched: Allocate and initialize " Morten Rasmussen
[not found] ` <OF29F384AC.37929D8E-ON48257E35.002FCB0C-48257E35.003156FE@zte.com.cn>
2015-04-29 15:43 ` Dietmar Eggemann
2015-02-04 18:31 ` [RFCv3 PATCH 24/48] sched: Introduce SD_SHARE_CAP_STATES sched_domain flag Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 25/48] arm: topology: Define TC2 energy and provide it to the scheduler Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 26/48] sched: Compute cpu capacity available at current frequency Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 27/48] sched: Relocated get_cpu_usage() Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 28/48] sched: Use capacity_curr to cap utilization in get_cpu_usage() Morten Rasmussen
2015-03-23 16:14 ` Peter Zijlstra
2015-03-24 11:36 ` Morten Rasmussen
2015-03-24 12:59 ` Peter Zijlstra
2015-02-04 18:31 ` [RFCv3 PATCH 29/48] sched: Highest energy aware balancing sched_domain level pointer Morten Rasmussen
2015-03-23 16:16 ` Peter Zijlstra
2015-03-24 10:52 ` Morten Rasmussen
[not found] ` <OF5977496A.A21A7B96-ON48257E35.002EC23C-48257E35.00324DAD@zte.com.cn>
2015-04-29 15:54 ` Dietmar Eggemann
2015-02-04 18:31 ` [RFCv3 PATCH 30/48] sched: Calculate energy consumption of sched_group Morten Rasmussen
2015-03-13 22:54 ` Sai Gurrappadi
2015-03-16 14:15 ` Morten Rasmussen
2015-03-23 16:47 ` Peter Zijlstra
2015-03-23 20:21 ` Dietmar Eggemann
2015-03-24 10:44 ` Morten Rasmussen
2015-03-24 16:10 ` Peter Zijlstra
2015-03-24 17:39 ` Morten Rasmussen
2015-03-26 15:23 ` Dietmar Eggemann
2015-03-20 18:40 ` Sai Gurrappadi
2015-03-27 15:58 ` Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 31/48] sched: Extend sched_group_energy to test load-balancing decisions Morten Rasmussen
[not found] ` <OF081FBA75.F80B8844-ON48257E37.00261E89-48257E37.00267F24@zte.com.cn>
2015-04-30 20:26 ` Dietmar Eggemann
2015-02-04 18:31 ` [RFCv3 PATCH 32/48] sched: Estimate energy impact of scheduling decisions Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 33/48] sched: Energy-aware wake-up task placement Morten Rasmussen
2015-03-13 22:47 ` Sai Gurrappadi
2015-03-16 14:47 ` Morten Rasmussen
2015-03-18 20:15 ` Sai Gurrappadi
2015-03-27 16:37 ` Morten Rasmussen
2015-03-24 13:00 ` Peter Zijlstra
2015-03-24 15:24 ` Morten Rasmussen
2015-03-24 13:00 ` Peter Zijlstra
2015-03-24 15:42 ` Morten Rasmussen
2015-03-24 15:53 ` Peter Zijlstra
2015-03-24 17:47 ` Morten Rasmussen
2015-03-24 16:35 ` Peter Zijlstra
2015-03-25 18:01 ` Juri Lelli
2015-03-25 18:14 ` Peter Zijlstra
2015-03-26 10:21 ` Juri Lelli
2015-03-26 10:41 ` Peter Zijlstra
2015-04-27 16:01 ` Michael Turquette
2015-04-28 13:06 ` Peter Zijlstra
2015-02-04 18:31 ` [RFCv3 PATCH 34/48] sched: Bias new task wakeups towards higher capacity cpus Morten Rasmussen
2015-03-24 13:33 ` Peter Zijlstra
2015-03-25 18:18 ` Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 35/48] sched, cpuidle: Track cpuidle state index in the scheduler Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 36/48] sched: Count number of shallower idle-states in struct sched_group_energy Morten Rasmussen
2015-03-24 13:14 ` Peter Zijlstra
2015-03-24 17:13 ` Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 37/48] sched: Determine the current sched_group idle-state Morten Rasmussen
[not found] ` <OF1FDC99CD.22435E74-ON48257E37.001BA739-48257E37.001CA5ED@zte.com.cn>
2015-04-30 20:17 ` Dietmar Eggemann
[not found] ` <OF2F4202E4.8A4AF229-ON48257E38.00312CD4-48257E38.0036ADB6@zte.com.cn>
2015-05-01 15:09 ` Dietmar Eggemann
2015-02-04 18:31 ` [RFCv3 PATCH 38/48] sched: Infrastructure to query if load balancing is energy-aware Morten Rasmussen
2015-03-24 13:41 ` Peter Zijlstra
2015-03-24 16:17 ` Dietmar Eggemann
2015-03-24 13:56 ` Peter Zijlstra
2015-03-24 16:22 ` Dietmar Eggemann
2015-02-04 18:31 ` [RFCv3 PATCH 39/48] sched: Introduce energy awareness into update_sg_lb_stats Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 40/48] sched: Introduce energy awareness into update_sd_lb_stats Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 41/48] sched: Introduce energy awareness into find_busiest_group Morten Rasmussen
2015-02-04 18:31 ` [RFCv3 PATCH 42/48] sched: Introduce energy awareness into find_busiest_queue Morten Rasmussen
2015-03-24 15:21 ` Peter Zijlstra
2015-03-24 18:04 ` Dietmar Eggemann
2015-02-04 18:31 ` [RFCv3 PATCH 43/48] sched: Introduce energy awareness into detach_tasks Morten Rasmussen
2015-03-24 15:25 ` Peter Zijlstra
2015-03-25 23:50 ` Sai Gurrappadi
2015-03-27 15:03 ` Dietmar Eggemann
[not found] ` <OFDCE15EEF.2F536D7F-ON48257E37.002565ED-48257E37.0027A8B9@zte.com.cn>
2015-04-30 20:35 ` Dietmar Eggemann
2015-02-04 18:31 ` [RFCv3 PATCH 44/48] sched: Tipping point from energy-aware to conventional load balancing Morten Rasmussen
2015-03-24 15:26 ` Peter Zijlstra
2015-03-24 18:47 ` Dietmar Eggemann
2015-02-04 18:31 ` [RFCv3 PATCH 45/48] sched: Skip cpu as lb src which has one task and capacity gte the dst cpu Morten Rasmussen
2015-03-24 15:27 ` Peter Zijlstra
2015-03-25 18:44 ` Dietmar Eggemann
[not found] ` <OF9320540C.255228F9-ON48257E37.002A02D1-48257E37.002AB5EE@zte.com.cn>
2015-05-05 10:01 ` Dietmar Eggemann
2015-02-04 18:31 ` [RFCv3 PATCH 46/48] sched: Turn off fast idling of cpus on a partially loaded system Morten Rasmussen
2015-03-24 16:01 ` Peter Zijlstra
2015-02-04 18:31 ` [RFCv3 PATCH 47/48] sched: Enable active migration for cpus of lower capacity Morten Rasmussen
2015-03-24 16:02 ` Peter Zijlstra
2015-02-04 18:31 ` [RFCv3 PATCH 48/48] sched: Disable energy-unfriendly nohz kicks Morten Rasmussen
2015-02-20 19:26 ` Dietmar Eggemann
2015-04-02 12:43 ` [RFCv3 PATCH 00/48] sched: Energy cost model for energy-aware scheduling Vincent Guittot
2015-04-08 13:33 ` Morten Rasmussen
2015-04-09 7:41 ` Vincent Guittot
2015-04-10 14:46 ` Morten Rasmussen
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=1423074685-6336-17-git-send-email-morten.rasmussen@arm.com \
--to=morten.rasmussen@arm.com \
--cc=dietmar.eggemann@arm.com \
--cc=juri.lelli@arm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@arm.linux.org.uk \
--cc=mingo@redhat.com \
--cc=mturquette@linaro.org \
--cc=nico@linaro.org \
--cc=peterz@infradead.org \
--cc=preeti@linux.vnet.ibm.com \
--cc=rjw@rjwysocki.net \
--cc=vincent.guittot@linaro.org \
--cc=yuyang.du@intel.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
all inboxes | Powered by JetHome®