From nobody Mon Dec 15 11:25:25 2025 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by smtp.subspace.kernel.org (Postfix) with ESMTP id 9509C3B7A8 for ; Fri, 14 Mar 2025 15:13:57 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=217.140.110.172 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741965240; cv=none; b=MAq9fzvFU71ZzP+rw6OfVumaXQstLMrxqNJjJmqdq80uhyx0ExgQ8oETLYwpoUWFLrFvHEziP0Gk3TEhMBZbJoVq5GYIhzBQgbctCUcKHrmTbtQuff3ICzhOMEnw1GYyAGnfqiT+HXlgu3IXwyXOdgVg4NYnonTlvVtSgpjxg0g= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1741965240; c=relaxed/simple; bh=RiXXWX3kH29UPxw8aYNz01QrT3fmUFBQQ/Ye3YBfEjE=; h=From:To:Cc:Subject:Date:Message-Id:MIME-Version; b=jey31DFy0xJEylippqg3J8NsrMyD+ytvX+Sja5Y9II5tEYIO3NWQ6s7f5Y52atYPAnUf3Bcjqo7INqAhbp/oeS3szWlWCldWE0U9sr+lUpLQWMkfJVJyqoi6Wi5Wn6hSsNsQ9tutHmoS2Kb6Bk/oMMjT/Ao3hOT7umeKXQue7uU= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=arm.com; spf=pass smtp.mailfrom=arm.com; arc=none smtp.client-ip=217.140.110.172 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=arm.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=arm.com Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id A40DA1424; Fri, 14 Mar 2025 08:14:06 -0700 (PDT) Received: from e125579.fritz.box (usa-sjc-mx-foss1.foss.arm.com [172.31.20.19]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id F20453F673; Fri, 14 Mar 2025 08:13:54 -0700 (PDT) From: Dietmar Eggemann To: Ingo Molnar , Peter Zijlstra , Juri Lelli , Vincent Guittot Cc: Steven Rostedt , Ben Segall , Mel Gorman , Valentin Schneider , Hagar Hemdan , Hazem Mohamed Abuelfotoh , linux-kernel@vger.kernel.org Subject: [PATCH] Revert "sched/core: Reduce cost of sched_move_task when config autogroup" Date: Fri, 14 Mar 2025 16:13:45 +0100 Message-Id: <20250314151345.275739-1-dietmar.eggemann@arm.com> X-Mailer: git-send-email 2.34.1 Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" This reverts commit eff6c8ce8d4d7faef75f66614dd20bb50595d261. Hazem reported a 30% drop in UnixBench spawn test with commit eff6c8ce8d4d ("sched/core: Reduce cost of sched_move_task when config autogroup") on a m6g.xlarge AWS EC2 instance with 4 vCPUs and 16 GiB RAM (aarch64) (single level MC sched domain) [1]. There is an early bail from sched_move_task() if p->sched_task_group is equal to p's 'cpu cgroup' (sched_get_task_group()). E.g. both are pointing to taskgroup '/user.slice/user-1000.slice/session-1.scope' (Ubuntu '22.04.5 LTS'). So in: do_exit() sched_autogroup_exit_task() sched_move_task() if sched_get_task_group(p) =3D=3D p->sched_task_group return /* p is enqueued */ dequeue_task() \ sched_change_group() | task_change_group_fair() | detach_task_cfs_rq() | (1) set_task_rq() | attach_task_cfs_rq() | enqueue_task() / (1) isn't called for p anymore. Turns out that the regression is related to sgs->group_util in group_is_overloaded() and group_has_capacity(). If (1) isn't called for all the 'spawn' tasks then sgs->group_util is ~900 and sgs->group_capacity =3D 1024 (single CPU sched domain) and this leads to group_is_overloaded() returning true (2) and group_has_capacity() false (3) much more often compared to the case when (1) is called. I.e. there are much more cases of 'group_is_overloaded' and 'group_fully_busy' in WF_FORK wakeup sched_balance_find_dst_cpu() which then returns much more often a CPU !=3D smp_processor_id() (5). This isn't good for these extremely short running tasks (FORK + EXIT) and also involves calling sched_balance_find_dst_group_cpu() unnecessary (single CPU sched domain). Instead if (1) is called for 'p->flags & PF_EXITING' then the path (4),(6) is taken much more often. select_task_rq_fair(..., wake_flags =3D WF_FORK) cpu =3D smp_processor_id() new_cpu =3D sched_balance_find_dst_cpu(..., cpu, ...) group =3D sched_balance_find_dst_group(..., cpu) do { update_sg_wakeup_stats() sgs->group_type =3D group_classify() if group_is_overloaded() (2) return group_overloaded if !group_has_capacity() (3) return group_fully_busy return group_has_spare (4) } while group if local_sgs.group_type > idlest_sgs.group_type return idlest (5) case group_has_spare: if local_sgs.idle_cpus >=3D idlest_sgs.idle_cpus return NULL (6) Unixbench Tests './Run -c 4 spawn' on: (a) VM AWS instance (m7gd.16xlarge) with v6.13 ('maxcpus=3D4 nr_cpus=3D4') and Ubuntu 22.04.5 LTS (aarch64). Shell & test run in '/user.slice/user-1000.slice/session-1.scope'. w/o patch w/ patch 21005 27120 (b) i7-13700K with tip/sched/core ('nosmt maxcpus=3D8 nr_cpus=3D8') and Ubuntu 22.04.5 LTS (x86_64). Shell & test run in '/A'. w/o patch w/ patch 67675 88806 CONFIG_SCHED_AUTOGROUP=3Dy & /sys/proc/kernel/sched_autogroup_enabled equal 0 or 1. [1] https://lkml.kernel.org/r/20250205151026.13061-1-hagarhem@amazon.com Reported-by: Hazem Mohamed Abuelfotoh Tested-by: Hagar Hemdan Signed-off-by: Dietmar Eggemann Reviewed-by: Vincent Guittot --- Previous discussion under: https://lkml.kernel.org/r/20250306162635.2614376-1-dietmar.eggemann@arm.com kernel/sched/core.c | 21 +++------------------ 1 file changed, 3 insertions(+), 18 deletions(-) diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 621cfc731c5b..532119a3f72b 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -9015,7 +9015,7 @@ void sched_release_group(struct task_group *tg) spin_unlock_irqrestore(&task_group_lock, flags); } =20 -static struct task_group *sched_get_task_group(struct task_struct *tsk) +static void sched_change_group(struct task_struct *tsk) { struct task_group *tg; =20 @@ -9027,13 +9027,7 @@ static struct task_group *sched_get_task_group(struc= t task_struct *tsk) tg =3D container_of(task_css_check(tsk, cpu_cgrp_id, true), struct task_group, css); tg =3D autogroup_task_group(tsk, tg); - - return tg; -} - -static void sched_change_group(struct task_struct *tsk, struct task_group = *group) -{ - tsk->sched_task_group =3D group; + tsk->sched_task_group =3D tg; =20 #ifdef CONFIG_FAIR_GROUP_SCHED if (tsk->sched_class->task_change_group) @@ -9054,20 +9048,11 @@ void sched_move_task(struct task_struct *tsk, bool = for_autogroup) { int queued, running, queue_flags =3D DEQUEUE_SAVE | DEQUEUE_MOVE | DEQUEUE_NOCLOCK; - struct task_group *group; struct rq *rq; =20 CLASS(task_rq_lock, rq_guard)(tsk); rq =3D rq_guard.rq; =20 - /* - * Esp. with SCHED_AUTOGROUP enabled it is possible to get superfluous - * group changes. - */ - group =3D sched_get_task_group(tsk); - if (group =3D=3D tsk->sched_task_group) - return; - update_rq_clock(rq); =20 running =3D task_current_donor(rq, tsk); @@ -9078,7 +9063,7 @@ void sched_move_task(struct task_struct *tsk, bool fo= r_autogroup) if (running) put_prev_task(rq, tsk); =20 - sched_change_group(tsk, group); + sched_change_group(tsk); if (!for_autogroup) scx_cgroup_move_task(tsk); =20 --=20 2.34.1