From: Ricardo Neri <ricardo.neri-calderon@linux.intel.com>
To: "Peter Zijlstra (Intel)" <peterz@infradead.org>,
Juri Lelli <juri.lelli@redhat.com>,
Vincent Guittot <vincent.guittot@linaro.org>
Cc: Ricardo Neri <ricardo.neri@intel.com>,
"Ravi V. Shankar" <ravi.v.shankar@intel.com>,
Ben Segall <bsegall@google.com>,
Daniel Bristot de Oliveira <bristot@redhat.com>,
Dietmar Eggemann <dietmar.eggemann@arm.com>,
Len Brown <len.brown@intel.com>, Mel Gorman <mgorman@suse.de>,
"Rafael J. Wysocki" <rafael.j.wysocki@intel.com>,
Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>,
Steven Rostedt <rostedt@goodmis.org>,
Tim Chen <tim.c.chen@linux.intel.com>,
Valentin Schneider <vschneid@redhat.com>,
x86@kernel.org, linux-kernel@vger.kernel.org,
Ricardo Neri <ricardo.neri-calderon@linux.intel.com>,
"Tim C . Chen" <tim.c.chen@intel.com>
Subject: [RFC PATCH 09/23] sched/fair: Use task-class performance score to pick the busiest group
Date: Fri, 9 Sep 2022 16:11:51 -0700 [thread overview]
Message-ID: <20220909231205.14009-10-ricardo.neri-calderon@linux.intel.com> (raw)
In-Reply-To: <20220909231205.14009-1-ricardo.neri-calderon@linux.intel.com>
update_sd_pick_busiest() keeps on selecting as the busiest group scheduling
groups of identical priority. Since both groups have the same priority,
either group is a good choice. The classes of tasks in the scheduling
groups can break this tie.
Pick as busiest the scheduling group that yields a higher task-class
performance score after load balancing.
Cc: Ben Segall <bsegall@google.com>
Cc: Daniel Bristot de Oliveira <bristot@redhat.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Tim C. Chen <tim.c.chen@intel.com>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: x86@kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ricardo Neri <ricardo.neri-calderon@linux.intel.com>
---
kernel/sched/fair.c | 68 +++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 68 insertions(+)
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 97731f81b570..7368a0b453ee 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -8777,6 +8777,60 @@ static void compute_ilb_sg_task_class_scores(struct sg_lb_task_class_stats *clas
sgs->task_class_score_before = group_score;
}
+/**
+ * sched_asym_class_prefer - Select a sched group based on its classes of tasks
+ * @a: Load balancing statistics of @sg_a
+ * @b: Load balancing statistics of @sg_b
+ *
+ * Returns: true if preferring @a yields a higher overall throughput after
+ * balancing load. Returns false otherwise.
+ */
+static bool sched_asym_class_prefer(struct sg_lb_stats *a,
+ struct sg_lb_stats *b)
+{
+ if (!sched_task_classes_enabled())
+ return false;
+
+ /* @a increases overall throughput after load balance. */
+ if (a->task_class_score_after > b->task_class_score_after)
+ return true;
+
+ /*
+ * If @a and @b yield the same overall throughput, pick @a if
+ * its current throughput is lower than that of @b.
+ */
+ if (a->task_class_score_after == b->task_class_score_after)
+ return a->task_class_score_before < b->task_class_score_before;
+
+ return false;
+}
+
+/**
+ * sched_asym_class_pick - Select a sched group based on classes of tasks
+ * @a: A scheduling group
+ * @b: A second scheduling group
+ * @a_stats: Load balancing statistics of @a
+ * @b_stats: Load balancing statistics of @b
+ *
+ * Returns: true if @a has the same priority and @a has classes of tasks that
+ * yield higher overall throughput after load balance. Returns false otherwise.
+ */
+static bool sched_asym_class_pick(struct sched_group *a,
+ struct sched_group *b,
+ struct sg_lb_stats *a_stats,
+ struct sg_lb_stats *b_stats)
+{
+ /*
+ * Only use the class-specific preference selection if both sched
+ * groups have the same priority.
+ */
+ if (arch_asym_cpu_priority(a->asym_prefer_cpu) !=
+ arch_asym_cpu_priority(b->asym_prefer_cpu))
+ return false;
+
+ return sched_asym_class_prefer(a_stats, b_stats);
+}
+
#else /* CONFIG_SCHED_TASK_CLASSES */
static void update_rq_task_classes_stats(struct sg_lb_task_class_stats *class_sgs,
struct rq *rq)
@@ -8793,6 +8847,14 @@ static void compute_ilb_sg_task_class_scores(struct sg_lb_task_class_stats *clas
{
}
+static bool sched_asym_class_pick(struct sched_group *a,
+ struct sched_group *b,
+ struct sg_lb_stats *a_stats,
+ struct sg_lb_stats *b_stats)
+{
+ return false;
+}
+
#endif /* CONFIG_SCHED_TASK_CLASSES */
/**
@@ -9049,6 +9111,12 @@ static bool update_sd_pick_busiest(struct lb_env *env,
/* Prefer to move from lowest priority CPU's work */
if (sched_asym_prefer(sg->asym_prefer_cpu, sds->busiest->asym_prefer_cpu))
return false;
+
+ /* @sg and @sds::busiest have the same priority. */
+ if (sched_asym_class_pick(sds->busiest, sg, &sds->busiest_stat, sgs))
+ return false;
+
+ /* @sg has lower priority than @sds::busiest. */
break;
case group_misfit_task:
--
2.25.1
next prev parent reply other threads:[~2022-09-09 23:08 UTC|newest]
Thread overview: 78+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-09 23:11 [RFC PATCH 00/23] sched: Introduce classes of tasks for load balance Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 01/23] sched/task_struct: Introduce classes of tasks Ricardo Neri
2022-09-14 13:46 ` Peter Zijlstra
2022-09-16 14:41 ` Ricardo Neri
2022-09-27 13:01 ` Peter Zijlstra
2022-10-02 22:32 ` Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 02/23] sched: Add interfaces for " Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 03/23] sched/core: Initialize the class of a new task Ricardo Neri
2022-09-26 14:57 ` Joel Fernandes
2022-09-26 21:53 ` Ricardo Neri
2022-09-27 13:04 ` Peter Zijlstra
2022-09-27 15:48 ` Joel Fernandes
2022-10-01 20:32 ` Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 04/23] sched/core: Add user_tick as argument to scheduler_tick() Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 05/23] sched/core: Move is_core_idle() out of fair.c Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 06/23] sched/core: Update the classification of the current task Ricardo Neri
2022-09-14 13:44 ` Peter Zijlstra
2022-09-16 14:42 ` Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 07/23] sched/fair: Collect load-balancing stats for task classes Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 08/23] sched/fair: Compute task-class performance scores for load balancing Ricardo Neri
2022-09-27 9:15 ` Peter Zijlstra
2022-10-26 3:57 ` Ricardo Neri
2022-10-26 8:55 ` Peter Zijlstra
2022-10-27 3:30 ` Ricardo Neri
2022-09-09 23:11 ` Ricardo Neri [this message]
2022-09-27 11:01 ` [RFC PATCH 09/23] sched/fair: Use task-class performance score to pick the busiest group Peter Zijlstra
2022-10-05 23:38 ` Ricardo Neri
2022-10-06 8:37 ` Peter Zijlstra
2022-10-06 19:07 ` Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 10/23] sched/fair: Use classes of tasks when selecting a busiest runqueue Ricardo Neri
2022-09-27 11:25 ` Peter Zijlstra
2022-10-07 23:36 ` Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 11/23] thermal: intel: hfi: Introduce Hardware Feedback Interface classes Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 12/23] thermal: intel: hfi: Convert table_lock to use flags-handling variants Ricardo Neri
2022-09-27 11:34 ` Peter Zijlstra
2022-09-27 11:36 ` Peter Zijlstra
2022-10-26 3:59 ` Ricardo Neri
2022-10-26 3:58 ` Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 13/23] x86/cpufeatures: Add the Intel Thread Director feature definitions Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 14/23] thermal: intel: hfi: Update the class of the current task Ricardo Neri
2022-09-27 11:46 ` Peter Zijlstra
2022-10-07 20:34 ` Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 15/23] thermal: intel: hfi: Report per-cpu class-specific performance scores Ricardo Neri
2022-09-27 11:59 ` Peter Zijlstra
2022-10-05 23:59 ` Ricardo Neri
2022-10-06 8:52 ` Peter Zijlstra
2022-10-06 9:14 ` Peter Zijlstra
2022-10-06 15:05 ` Brown, Len
2022-10-06 16:14 ` Peter Zijlstra
2022-10-07 11:20 ` Len Brown
2022-09-09 23:11 ` [RFC PATCH 16/23] thermal: intel: hfi: Define a default classification for unclassified tasks Ricardo Neri
2022-09-09 23:11 ` [RFC PATCH 17/23] thermal: intel: hfi: Enable the Intel Thread Director Ricardo Neri
2022-09-27 12:00 ` Peter Zijlstra
2022-10-06 1:50 ` Ricardo Neri
2022-09-09 23:12 ` [RFC PATCH 18/23] sched/task_struct: Add helpers for task classification Ricardo Neri
2022-09-27 11:52 ` Peter Zijlstra
2022-10-08 0:38 ` Ricardo Neri
2022-09-09 23:12 ` [RFC PATCH 19/23] sched/core: Initialize helpers of " Ricardo Neri
2022-09-09 23:12 ` [RFC PATCH 20/23] thermal: intel: hfi: Implement model-specific checks for " Ricardo Neri
2022-09-09 23:12 ` [RFC PATCH 21/23] x86/cpufeatures: Add feature bit for HRESET Ricardo Neri
2022-09-09 23:12 ` [RFC PATCH 22/23] x86/hreset: Configure history reset Ricardo Neri
2022-09-27 12:03 ` Peter Zijlstra
2022-10-02 22:34 ` Ricardo Neri
2022-09-09 23:12 ` [RFC PATCH 23/23] x86/process: Reset hardware history in context switch Ricardo Neri
2022-09-27 12:52 ` Peter Zijlstra
2022-10-03 23:07 ` Ricardo Neri
2022-10-06 8:35 ` Peter Zijlstra
2022-10-06 22:55 ` Ricardo Neri
2022-09-27 12:53 ` Peter Zijlstra
2022-10-02 22:02 ` Ricardo Neri
2022-09-27 13:15 ` Borislav Petkov
2022-10-02 22:12 ` Ricardo Neri
2022-10-02 22:15 ` Borislav Petkov
2022-10-03 19:49 ` Ricardo Neri
2022-10-03 19:55 ` Borislav Petkov
[not found] ` <20220910072120.2651-1-hdanton@sina.com>
2022-09-16 14:51 ` [RFC PATCH 06/23] sched/core: Update the classification of the current task Ricardo Neri
2022-10-11 19:12 ` Trying to apply patch set Carlos Bilbao
2022-10-18 2:31 ` Ricardo Neri
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=20220909231205.14009-10-ricardo.neri-calderon@linux.intel.com \
--to=ricardo.neri-calderon@linux.intel.com \
--cc=bristot@redhat.com \
--cc=bsegall@google.com \
--cc=dietmar.eggemann@arm.com \
--cc=juri.lelli@redhat.com \
--cc=len.brown@intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mgorman@suse.de \
--cc=peterz@infradead.org \
--cc=rafael.j.wysocki@intel.com \
--cc=ravi.v.shankar@intel.com \
--cc=ricardo.neri@intel.com \
--cc=rostedt@goodmis.org \
--cc=srinivas.pandruvada@linux.intel.com \
--cc=tim.c.chen@intel.com \
--cc=tim.c.chen@linux.intel.com \
--cc=vincent.guittot@linaro.org \
--cc=vschneid@redhat.com \
--cc=x86@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;
as well as URLs for NNTP newsgroup(s).