From nobody Mon Feb 9 23:04:17 2026 Received: from mail-dl1-f99.google.com (mail-dl1-f99.google.com [74.125.82.99]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id D96192F6167 for ; Wed, 3 Dec 2025 11:40:23 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=74.125.82.99 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764762025; cv=none; b=dB+MUuiuJDjUdy7vicdzCOIeVy324upVijlfVCuFsZp0wkinVdBDq2sLJuHzZ50NW9CYrPqm/Q8tQa4t8LzaEq8gs/CZxteXEdrFYc1uFmMUiRjhFYVPYtNl2LZvLazyyy5pAbucjqsHVlTUleiG8q/PubxqfihTDc7RfThXDkI= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764762025; c=relaxed/simple; bh=02Yp53TuowJehhoJcF+Iw5981qDC8pXTcvwqF/knsI4=; h=From:To:Cc:Subject:Date:Message-Id:In-Reply-To:References: MIME-Version; b=kTSgVjCxaTGs5D535ejNzSFGiEXGKA0wBf2uYKpz3u/k2KEAWrzwuf2D182x5+qSGMhDwNX4RcWvP+vlDBwVJQKf1Amw0olclLHld3I9QPGOHw+P7ZWiINP631K8sRJud3WkX3MHMrjYULe1JqPKkF7VwTXHU9xe1zJD7KOS06M= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=broadcom.com; spf=fail smtp.mailfrom=broadcom.com; dkim=pass (1024-bit key) header.d=broadcom.com header.i=@broadcom.com header.b=PQfQtqB5; arc=none smtp.client-ip=74.125.82.99 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=broadcom.com Authentication-Results: smtp.subspace.kernel.org; spf=fail smtp.mailfrom=broadcom.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=broadcom.com header.i=@broadcom.com header.b="PQfQtqB5" Received: by mail-dl1-f99.google.com with SMTP id a92af1059eb24-11beb0a7bd6so956630c88.1 for ; Wed, 03 Dec 2025 03:40:23 -0800 (PST) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1764762023; x=1765366823; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:dkim-signature:x-gm-gg :x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=b/SKWEA8ISmyxBr9KByGjrpmJhdyEvH8fSbpB6G/k0w=; b=EkGKT6qzSAlEWvWzXOzYmXTGOky0Qd5LaOrIAXW98GaORjRMuSpq0Pa3eTxMrQh2iC lzHbaXmzc70gpUzCCGZEbMi+gDdDGW/cD1+4VH+NaCJWWk+MSimliC+yzmZmLYCXjs/s 5rRUhWWhPbSZxocFQnnxI+k77cErnI1K/c2lZQUeWsvvWfuA9bZlfaibfpDY6W66eXlv sPxSbe4sknx7t+yq4+X6++vYbQfr/UlI74ey/iRuKvnVhWlkCwK/mNupZ2XzZw8JwiDM YQ+/MbrLZWQC+ZWM6LMENVFcdpnf4DYP8OCq3K9X+vGIlqeyZJsDfSyMHRviWM4SsK6w TxWA== X-Forwarded-Encrypted: i=1; AJvYcCWS/h0rNhOvdsNZJm7vupypuf0kHt0vUcuKYhKiPKXTBG5XfKZiaYCidRJwFnfTSC+eaatkFkA9CBsFmdM=@vger.kernel.org X-Gm-Message-State: AOJu0YwGei0LtZ8A/US61Js09YjO92wewYavmD5wrIiPqDnp2+EnpppZ 1qlpUd3O48ntl5BmW8jmrhEYUVMb1hwtcUm8InCU/bkT4+UMHza8SlG8opgQyCJ+MvMhmCzcxRf rOQhSj55z16mQOZEnMfDzcLSj94pJxFd3l9NKEkZRX1spMBTh9/JSYnvcvaZe6FuVwu/ORzMd1r VKJBURXF+ThwbyDvUTNvRuILZMWpN32oxvfV8ucW12LNXzwUk0fsCFIszRK8sGZJ4vZf1s+37m0 ubKuVjPN++g2bCKyBI= X-Gm-Gg: ASbGncuXkYnN48U3sO/f0MM/Euk64LRv16kGqRF8Vxu3uWa6zhiO7vUoo/A5fXIzlah Y7gYAri6W+Xp2cdVPdZffjQu679O4vFz/9cO5HdAWuRtWa3xiCMLrMDOp++DsqA0pu+96NNrDIx u7YXegT2x3zrGXHjQHAeTyjVRNiZ61ep4xB8WiwK0EoCZ/MpNkYSK0nSocY8y1NW7lrG8RjqrgN 9meOApP7C+QwSrbhSaOF8UVZOlI287w6rYT9DRzoo+gzhZFT3nXHq/TaAQ4lB8wAto8A0gd8TbP P8tDgcAS6w7TpvQYMV4UDl0VtDLrmvlNbHmpN0lAJhMLfl2d9NXUWlrElxWreI3JuzMLUcndBpP aBpWh/i8gquvQC2zNs6eafksxyPXPrU+G21fKOX8DGQJuk7iUa+8ouPmZs7k9Tr8+WQY+LumUQ/ o0fT28/RhMpZ4OMiksL4BLP05VdFGUFRhJcBUMIEiMag== X-Google-Smtp-Source: AGHT+IEP1eI+IcGvKs97ffOFBLQKH4WwJ5jzCoD5EdEmN6l3o8e77MbWhORSr+htWX/EV3vQgfASLo3g7FNp X-Received: by 2002:a05:7022:2604:b0:119:e569:f85b with SMTP id a92af1059eb24-11de948a95amr4638640c88.18.1764762022713; Wed, 03 Dec 2025 03:40:22 -0800 (PST) Received: from smtp-us-east1-p01-i01-si01.dlp.protect.broadcom.com (address-144-49-247-11.dlp.protect.broadcom.com. [144.49.247.11]) by smtp-relay.gmail.com with ESMTPS id a92af1059eb24-11dcb03f38fsm2356024c88.6.2025.12.03.03.40.22 for (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Wed, 03 Dec 2025 03:40:22 -0800 (PST) X-Relaying-Domain: broadcom.com X-CFilter-Loop: Reflected Received: by mail-dy1-f200.google.com with SMTP id 5a478bee46e88-2a467c4e74bso1067671eec.1 for ; Wed, 03 Dec 2025 03:40:21 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=broadcom.com; s=google; t=1764762021; x=1765366821; darn=vger.kernel.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=b/SKWEA8ISmyxBr9KByGjrpmJhdyEvH8fSbpB6G/k0w=; b=PQfQtqB5M3kK8u318NO1Q8A2ZhMCieTbmATYvgwbDKFtzolSkhl9h+9QaEa3RYbcAv nOYuvybn1LfYVQ2m/8Fr6Mk8KdgtYUyf0Nit0ZWveoNO8rV33Sju7Em9x1fsA51WBHxq UNJ/t4WzBb9TYCGiOoA7WCqCvW/IHoxYvTqYY= X-Forwarded-Encrypted: i=1; AJvYcCWP6VRPMmaCC6JJ2jkxx9bQ9mswTNYVL9prr3g+5PoB2Jd0Bi6UZbl9x0+Lkp61U9eFd+YxKun2WK19gSY=@vger.kernel.org X-Received: by 2002:a05:693c:2b10:b0:2ab:9990:79a9 with SMTP id 5a478bee46e88-2ab99908053mr504400eec.17.1764762020794; Wed, 03 Dec 2025 03:40:20 -0800 (PST) X-Received: by 2002:a05:693c:2b10:b0:2ab:9990:79a9 with SMTP id 5a478bee46e88-2ab99908053mr504383eec.17.1764762020180; Wed, 03 Dec 2025 03:40:20 -0800 (PST) Received: from photon-dev-haas.. ([192.19.161.250]) by smtp.gmail.com with ESMTPSA id 5a478bee46e88-2a965b1ceeesm63324781eec.5.2025.12.03.03.40.19 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 03 Dec 2025 03:40:19 -0800 (PST) From: Ajay Kaher To: stable@vger.kernel.org, gregkh@linuxfoundation.org Cc: mingo@redhat.com, peterz@infradead.org, juri.lelli@redhat.com, vincent.guittot@linaro.org, dietmar.eggemann@arm.com, rostedt@goodmis.org, bsegall@google.com, mgorman@suse.de, vschneid@redhat.com, linux-kernel@vger.kernel.org, ajay.kaher@broadcom.com, alexey.makhalov@broadcom.com, yin.ding@broadcom.com, tapas.kundu@broadcom.com, Chris Mason Subject: [PATCH v6.6 4/4] sched/fair: Proportional newidle balance Date: Wed, 3 Dec 2025 11:22:55 +0000 Message-Id: <20251203112255.1738272-5-ajay.kaher@broadcom.com> X-Mailer: git-send-email 2.40.4 In-Reply-To: <20251203112255.1738272-1-ajay.kaher@broadcom.com> References: <20251203112255.1738272-1-ajay.kaher@broadcom.com> 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 X-DetectorID-Processed: b00c1d49-9d2e-4205-b15f-d015386d3d5e Content-Type: text/plain; charset="utf-8" From: Peter Zijlstra (Intel) commit 33cf66d88306663d16e4759e9d24766b0aaa2e17 upstream. Add a randomized algorithm that runs newidle balancing proportional to its success rate. This improves schbench significantly: 6.18-rc4: 2.22 Mrps/s 6.18-rc4+revert: 2.04 Mrps/s 6.18-rc4+revert+random: 2.18 Mrps/S Conversely, per Adam Li this affects SpecJBB slightly, reducing it by 1%: 6.17: -6% 6.17+revert: 0% 6.17+revert+random: -1% Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Dietmar Eggemann Tested-by: Dietmar Eggemann Tested-by: Chris Mason Link: https://lkml.kernel.org/r/6825c50d-7fa7-45d8-9b81-c6e7e25738e2@meta.c= om Link: https://patch.msgid.link/20251107161739.770122091@infradead.org [ Ajay: Modified to apply on v6.6 ] Signed-off-by: Ajay Kaher --- include/linux/sched/topology.h | 3 +++ kernel/sched/core.c | 3 +++ kernel/sched/fair.c | 44 ++++++++++++++++++++++++++++++---- kernel/sched/features.h | 5 ++++ kernel/sched/sched.h | 7 ++++++ kernel/sched/topology.c | 6 +++++ 6 files changed, 64 insertions(+), 4 deletions(-) diff --git a/include/linux/sched/topology.h b/include/linux/sched/topology.h index 9671b7234..197039bab 100644 --- a/include/linux/sched/topology.h +++ b/include/linux/sched/topology.h @@ -106,6 +106,9 @@ struct sched_domain { unsigned int nr_balance_failed; /* initialise to 0 */ =20 /* idle_balance() stats */ + unsigned int newidle_call; + unsigned int newidle_success; + unsigned int newidle_ratio; u64 max_newidle_lb_cost; unsigned long last_decay_max_lb_cost; =20 diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 1b5e4389f..c4a9797e9 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -116,6 +116,7 @@ EXPORT_TRACEPOINT_SYMBOL_GPL(sched_util_est_se_tp); EXPORT_TRACEPOINT_SYMBOL_GPL(sched_update_nr_running_tp); =20 DEFINE_PER_CPU_SHARED_ALIGNED(struct rq, runqueues); +DEFINE_PER_CPU(struct rnd_state, sched_rnd_state); =20 #ifdef CONFIG_SCHED_DEBUG /* @@ -9872,6 +9873,8 @@ void __init sched_init_smp(void) { sched_init_numa(NUMA_NO_NODE); =20 + prandom_init_once(&sched_rnd_state); + /* * There's no userspace yet to cause hotplug operations; hence all the * CPU masks are stable and all blatant races in the below code cannot diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index f93a6a12e..a10df85e4 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -11704,11 +11704,27 @@ void update_max_interval(void) max_load_balance_interval =3D HZ*num_online_cpus()/10; } =20 -static inline bool update_newidle_cost(struct sched_domain *sd, u64 cost) +static inline void update_newidle_stats(struct sched_domain *sd, unsigned = int success) +{ + sd->newidle_call++; + sd->newidle_success +=3D success; + + if (sd->newidle_call >=3D 1024) { + sd->newidle_ratio =3D sd->newidle_success; + sd->newidle_call /=3D 2; + sd->newidle_success /=3D 2; + } +} + +static inline bool +update_newidle_cost(struct sched_domain *sd, u64 cost, unsigned int succes= s) { unsigned long next_decay =3D sd->last_decay_max_lb_cost + HZ; unsigned long now =3D jiffies; =20 + if (cost) + update_newidle_stats(sd, success); + if (cost > sd->max_newidle_lb_cost) { /* * Track max cost of a domain to make sure to not delay the @@ -11756,7 +11772,7 @@ static void rebalance_domains(struct rq *rq, enum c= pu_idle_type idle) * Decay the newidle max times here because this is a regular * visit to all the domains. */ - need_decay =3D update_newidle_cost(sd, 0); + need_decay =3D update_newidle_cost(sd, 0, 0); max_cost +=3D sd->max_newidle_lb_cost; =20 /* @@ -12394,6 +12410,22 @@ static int sched_balance_newidle(struct rq *this_r= q, struct rq_flags *rf) break; =20 if (sd->flags & SD_BALANCE_NEWIDLE) { + unsigned int weight =3D 1; + + if (sched_feat(NI_RANDOM)) { + /* + * Throw a 1k sided dice; and only run + * newidle_balance according to the success + * rate. + */ + u32 d1k =3D sched_rng() % 1024; + weight =3D 1 + sd->newidle_ratio; + if (d1k > weight) { + update_newidle_stats(sd, 0); + continue; + } + weight =3D (1024 + weight/2) / weight; + } =20 pulled_task =3D load_balance(this_cpu, this_rq, sd, CPU_NEWLY_IDLE, @@ -12401,10 +12433,14 @@ static int sched_balance_newidle(struct rq *this_= rq, struct rq_flags *rf) =20 t1 =3D sched_clock_cpu(this_cpu); domain_cost =3D t1 - t0; - update_newidle_cost(sd, domain_cost); - curr_cost +=3D domain_cost; t0 =3D t1; + + /* + * Track max cost of a domain to make sure to not delay the + * next wakeup on the CPU. + */ + update_newidle_cost(sd, domain_cost, weight * !!pulled_task); } =20 /* diff --git a/kernel/sched/features.h b/kernel/sched/features.h index f77016823..48b104ab5 100644 --- a/kernel/sched/features.h +++ b/kernel/sched/features.h @@ -88,4 +88,9 @@ SCHED_FEAT(UTIL_EST_FASTUP, true) =20 SCHED_FEAT(LATENCY_WARN, false) =20 +/* + * Do newidle balancing proportional to its success rate using randomizati= on. + */ +SCHED_FEAT(NI_RANDOM, true) + SCHED_FEAT(HZ_BW, true) diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 64634314a..e1913e253 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -5,6 +5,7 @@ #ifndef _KERNEL_SCHED_SCHED_H #define _KERNEL_SCHED_SCHED_H =20 +#include #include #include #include @@ -1205,6 +1206,12 @@ static inline bool is_migration_disabled(struct task= _struct *p) } =20 DECLARE_PER_CPU_SHARED_ALIGNED(struct rq, runqueues); +DECLARE_PER_CPU(struct rnd_state, sched_rnd_state); + +static inline u32 sched_rng(void) +{ + return prandom_u32_state(this_cpu_ptr(&sched_rnd_state)); +} =20 #define cpu_rq(cpu) (&per_cpu(runqueues, (cpu))) #define this_rq() this_cpu_ptr(&runqueues) diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c index b87426b74..9fa77e7a6 100644 --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -1600,6 +1600,12 @@ sd_init(struct sched_domain_topology_level *tl, =20 .last_balance =3D jiffies, .balance_interval =3D sd_weight, + + /* 50% success rate */ + .newidle_call =3D 512, + .newidle_success =3D 256, + .newidle_ratio =3D 512, + .max_newidle_lb_cost =3D 0, .last_decay_max_lb_cost =3D jiffies, .child =3D child, --=20 2.40.4