LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Valentin Schneider <valentin.schneider@arm.com>
To: linux-kernel@vger.kernel.org
Cc: mingo@kernel.org, peterz@infradead.org,
	vincent.guittot@linaro.org, dietmar.eggemann@arm.com
Subject: [PATCH v2 9/9] sched/topology: Define and use shortcut pointers for wakeup sd_flag scan
Date: Wed, 11 Mar 2020 18:16:01 +0000	[thread overview]
Message-ID: <20200311181601.18314-10-valentin.schneider@arm.com> (raw)
In-Reply-To: <20200311181601.18314-1-valentin.schneider@arm.com>

Reworking select_task_rq_fair()'s domain walk exposed that !want_affine
wakeups only look for highest sched_domain with the required sd_flag
set. This is something we can cache at sched domain build time to slightly
optimize select_task_rq_fair(). Note that this isn't a "free" optimization:
it costs us 3 pointers per CPU.

Add cached per-CPU pointers for the highest domains with SD_BALANCE_WAKE,
SD_BALANCE_EXEC and SD_BALANCE_FORK. Use them in select_task_rq_fair().

Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
---
 kernel/sched/fair.c     | 25 +++++++++++++------------
 kernel/sched/sched.h    |  3 +++
 kernel/sched/topology.c | 12 ++++++++++++
 3 files changed, 28 insertions(+), 12 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index a6fca6817e92..40fb97062157 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -6595,17 +6595,6 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int wake_flags)
 	int want_affine = 0;
 	int sd_flag;
 
-	switch (wake_flags & (WF_TTWU | WF_FORK | WF_EXEC)) {
-	case WF_TTWU:
-		sd_flag = SD_BALANCE_WAKE;
-		break;
-	case WF_FORK:
-		sd_flag = SD_BALANCE_FORK;
-		break;
-	default:
-		sd_flag = SD_BALANCE_EXEC;
-	}
-
 	if (wake_flags & WF_TTWU) {
 		record_wakee(p);
 
@@ -6621,7 +6610,19 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int wake_flags)
 
 	rcu_read_lock();
 
-	sd = highest_flag_domain(cpu, sd_flag);
+	switch (wake_flags & (WF_TTWU | WF_FORK | WF_EXEC)) {
+	case WF_TTWU:
+		sd_flag = SD_BALANCE_WAKE;
+		sd = rcu_dereference(per_cpu(sd_balance_wake, cpu));
+		break;
+	case WF_FORK:
+		sd_flag = SD_BALANCE_FORK;
+		sd = rcu_dereference(per_cpu(sd_balance_fork, cpu));
+		break;
+	default:
+		sd_flag = SD_BALANCE_EXEC;
+		sd = rcu_dereference(per_cpu(sd_balance_exec, cpu));
+	}
 
 	/*
 	 * If !want_affine, we just look for the highest domain where
diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index ad2106245e12..3a0e38f2f713 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -1393,6 +1393,9 @@ DECLARE_PER_CPU(int, sd_llc_size);
 DECLARE_PER_CPU(int, sd_llc_id);
 DECLARE_PER_CPU(struct sched_domain_shared __rcu *, sd_llc_shared);
 DECLARE_PER_CPU(struct sched_domain __rcu *, sd_numa);
+DECLARE_PER_CPU(struct sched_domain __rcu *, sd_balance_wake);
+DECLARE_PER_CPU(struct sched_domain __rcu *, sd_balance_fork);
+DECLARE_PER_CPU(struct sched_domain __rcu *, sd_balance_exec);
 DECLARE_PER_CPU(struct sched_domain __rcu *, sd_asym_packing);
 DECLARE_PER_CPU(struct sched_domain __rcu *, sd_asym_cpucapacity);
 extern struct static_key_false sched_asym_cpucapacity;
diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c
index 6077b23f9723..0270252a964b 100644
--- a/kernel/sched/topology.c
+++ b/kernel/sched/topology.c
@@ -610,6 +610,9 @@ DEFINE_PER_CPU(int, sd_llc_size);
 DEFINE_PER_CPU(int, sd_llc_id);
 DEFINE_PER_CPU(struct sched_domain_shared __rcu *, sd_llc_shared);
 DEFINE_PER_CPU(struct sched_domain __rcu *, sd_numa);
+DEFINE_PER_CPU(struct sched_domain __rcu *, sd_balance_wake);
+DEFINE_PER_CPU(struct sched_domain __rcu *, sd_balance_fork);
+DEFINE_PER_CPU(struct sched_domain __rcu *, sd_balance_exec);
 DEFINE_PER_CPU(struct sched_domain __rcu *, sd_asym_packing);
 DEFINE_PER_CPU(struct sched_domain __rcu *, sd_asym_cpucapacity);
 DEFINE_STATIC_KEY_FALSE(sched_asym_cpucapacity);
@@ -636,6 +639,15 @@ static void update_top_cache_domain(int cpu)
 	sd = lowest_flag_domain(cpu, SD_NUMA);
 	rcu_assign_pointer(per_cpu(sd_numa, cpu), sd);
 
+	sd = highest_flag_domain(cpu, SD_BALANCE_WAKE);
+	rcu_assign_pointer(per_cpu(sd_balance_wake, cpu), sd);
+
+	sd = highest_flag_domain(cpu, SD_BALANCE_FORK);
+	rcu_assign_pointer(per_cpu(sd_balance_fork, cpu), sd);
+
+	sd = highest_flag_domain(cpu, SD_BALANCE_EXEC);
+	rcu_assign_pointer(per_cpu(sd_balance_exec, cpu), sd);
+
 	sd = highest_flag_domain(cpu, SD_ASYM_PACKING);
 	rcu_assign_pointer(per_cpu(sd_asym_packing, cpu), sd);
 
-- 
2.24.0


  parent reply	other threads:[~2020-03-11 18:17 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-03-11 18:15 [PATCH v2 0/9] sched: Streamline select_task_rq() & select_task_rq_fair() Valentin Schneider
2020-03-11 18:15 ` [PATCH v2 1/9] sched/fair: find_idlest_group(): Remove unused sd_flag parameter Valentin Schneider
2020-03-19  9:05   ` Dietmar Eggemann
2020-03-11 18:15 ` [PATCH v2 2/9] sched/debug: Make sd->flags sysctl read-only Valentin Schneider
2020-03-19  9:07   ` Dietmar Eggemann
2020-03-19 12:04     ` Valentin Schneider
2020-03-11 18:15 ` [PATCH v2 3/9] sched: Remove checks against SD_LOAD_BALANCE Valentin Schneider
2020-03-19 10:28   ` Dietmar Eggemann
2020-03-19 12:05     ` Valentin Schneider
2020-03-23 14:26       ` Dietmar Eggemann
2020-03-23 17:17         ` Valentin Schneider
2020-03-11 18:15 ` [PATCH v2 4/9] sched/topology: Kill SD_LOAD_BALANCE Valentin Schneider
2020-03-19 10:29   ` Dietmar Eggemann
2020-03-19 12:06     ` Valentin Schneider
2020-03-11 18:15 ` [PATCH v2 5/9] sched: Add WF_TTWU, WF_EXEC wakeup flags Valentin Schneider
2020-03-11 18:15 ` [PATCH v2 6/9] sched: Kill select_task_rq()'s sd_flag parameter Valentin Schneider
2020-03-11 18:15 ` [PATCH v2 7/9] sched/fair: Dissociate wakeup decisions from SD flag value Valentin Schneider
2020-03-11 18:16 ` [PATCH v2 8/9] sched/fair: Split select_task_rq_fair want_affine logic Valentin Schneider
2020-03-19 10:30   ` Dietmar Eggemann
2020-03-19 12:06     ` Valentin Schneider
2020-03-11 18:16 ` Valentin Schneider [this message]
2020-03-19 10:46   ` [PATCH v2 9/9] sched/topology: Define and use shortcut pointers for wakeup sd_flag scan Dietmar Eggemann
2020-03-19 12:22     ` Valentin Schneider

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=20200311181601.18314-10-valentin.schneider@arm.com \
    --to=valentin.schneider@arm.com \
    --cc=dietmar.eggemann@arm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@kernel.org \
    --cc=peterz@infradead.org \
    --cc=vincent.guittot@linaro.org \
    --subject='Re: [PATCH v2 9/9] sched/topology: Define and use shortcut pointers for wakeup sd_flag scan' \
    /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

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).