From patchwork Tue Feb 7 05:10:49 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ricardo Neri X-Patchwork-Id: 13131028 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 797D2C636CD for ; Tue, 7 Feb 2023 05:02:27 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230250AbjBGFCZ (ORCPT ); Tue, 7 Feb 2023 00:02:25 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59782 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229743AbjBGFB4 (ORCPT ); Tue, 7 Feb 2023 00:01:56 -0500 Received: from mga05.intel.com (mga05.intel.com [192.55.52.43]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id D302A901E; Mon, 6 Feb 2023 21:01:53 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1675746113; x=1707282113; h=from:to:cc:subject:date:message-id:in-reply-to: references; bh=aOX48G3IsFHDTBbZEmk9IpTBJOBpnREOfkswBIUTlho=; b=RLeOi2k2COE/zVoVK2CzAPc1wrUepDlkJUycQaQHSSR1n/8qq/0sJlNZ GPYShstWgk8b5Hy9AmRz+aBmQh7NA+pz4GGd8QkkbSKGmfwN+UCzrk526 i/gGkySJivGiS2ha1Gc6hmdZHYqWfhv3/XF9u3+gTwQBtxXxxo0tI2Yff 6JZB2SHgXRlGAzLRUwGAGtLZIaD+iJG8uaQeCc+d5Z+Lrm2hE//h9iFuO NSr7jbawJR/1fYl3d8+eUnHSncWS8BbbHppBztLlLekXu+zeHxCVq/wlX X8NQ33ekK3pVM5e5nGNyIPufIMnbY078qGGFWDIVXVqrur40i825mAhCv A==; X-IronPort-AV: E=McAfee;i="6500,9779,10613"; a="415625820" X-IronPort-AV: E=Sophos;i="5.97,278,1669104000"; d="scan'208";a="415625820" Received: from fmsmga007.fm.intel.com ([10.253.24.52]) by fmsmga105.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 06 Feb 2023 21:01:42 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6500,9779,10613"; a="668657720" X-IronPort-AV: E=Sophos;i="5.97,278,1669104000"; d="scan'208";a="668657720" Received: from ranerica-svr.sc.intel.com ([172.25.110.23]) by fmsmga007.fm.intel.com with ESMTP; 06 Feb 2023 21:01:42 -0800 From: Ricardo Neri To: "Peter Zijlstra (Intel)" , Juri Lelli , Vincent Guittot Cc: Ricardo Neri , "Ravi V. Shankar" , Ben Segall , Daniel Bristot de Oliveira , Dietmar Eggemann , Len Brown , Mel Gorman , "Rafael J. Wysocki" , Srinivas Pandruvada , Steven Rostedt , Tim Chen , Valentin Schneider , Lukasz Luba , Ionela Voinescu , x86@kernel.org, "Joel Fernandes (Google)" , linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org, Ricardo Neri , "Tim C . Chen" Subject: [PATCH v3 08/24] sched/fair: Use IPCC stats to break ties between asym_packing sched groups Date: Mon, 6 Feb 2023 21:10:49 -0800 Message-Id: <20230207051105.11575-9-ricardo.neri-calderon@linux.intel.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20230207051105.11575-1-ricardo.neri-calderon@linux.intel.com> References: <20230207051105.11575-1-ricardo.neri-calderon@linux.intel.com> Precedence: bulk List-ID: X-Mailing-List: linux-pm@vger.kernel.org As it iterates, update_sd_pick_busiest() keeps on selecting as busiest sched groups of identical priority. Since both groups have the same priority, either group is a good choice. The IPCC statistics provide a measure of the throughput before and after load balance. Use them to pick a busiest scheduling group from otherwise identical asym_packing scheduling groups. Pick as busiest the scheduling group that yields a higher IPCC score after load balancing. Cc: Ben Segall Cc: Daniel Bristot de Oliveira Cc: Dietmar Eggemann Cc: Ionela Voinescu Cc: Joel Fernandes (Google) Cc: Len Brown Cc: Lukasz Luba Cc: Mel Gorman Cc: Rafael J. Wysocki Cc: Srinivas Pandruvada Cc: Steven Rostedt Cc: Tim C. Chen Cc: Valentin Schneider Cc: x86@kernel.org Cc: linux-pm@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Ricardo Neri --- Changes since v2: * None Changes since v1: * Added a comment to clarify why sched_asym_prefer() needs a tie breaker only in update_sd_pick_busiest(). (PeterZ) * Renamed functions for accuracy: sched_asym_class_prefer() >> sched_asym_ipcc_prefer() sched_asym_class_pick() >> sched_asym_ipcc_pick() * Reworded commit message for clarity. --- kernel/sched/fair.c | 72 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 72 insertions(+) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index b6165aa8a376..841927b9b192 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -9345,6 +9345,60 @@ static void update_sg_lb_stats_scores(struct sg_lb_stats *sgs, sgs->ipcc_score_before = before; } +/** + * sched_asym_ipcc_prefer - Select a sched group based on its IPCC score + * @a: Load balancing statistics of a sched group + * @b: Load balancing statistics of a second sched group + * + * Returns: true if @a has a higher IPCC score than @b after load balance. + * False otherwise. + */ +static bool sched_asym_ipcc_prefer(struct sg_lb_stats *a, + struct sg_lb_stats *b) +{ + if (!sched_ipcc_enabled()) + return false; + + /* @a increases overall throughput after load balance. */ + if (a->ipcc_score_after > b->ipcc_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->ipcc_score_after == b->ipcc_score_after) + return a->ipcc_score_before < b->ipcc_score_before; + + return false; +} + +/** + * sched_asym_ipcc_pick - Select a sched group based on its IPCC score + * @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 tasks with IPC classes + * that yield higher overall throughput after load balance. False otherwise. + */ +static bool sched_asym_ipcc_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_ipcc_prefer(a_stats, b_stats); +} + #else /* CONFIG_IPC_CLASSES */ static void update_sg_lb_ipcc_stats(int dst_cpu, struct sg_lb_stats *sgs, struct rq *rq) @@ -9361,6 +9415,14 @@ static void update_sg_lb_stats_scores(struct sg_lb_stats *sgs, { } +static bool sched_asym_ipcc_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_IPC_CLASSES */ /** @@ -9596,6 +9658,16 @@ 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; + + /* + * Unlike other callers of sched_asym_prefer(), here both @sg + * and @sds::busiest have tasks running. When they have equal + * priority, their IPC class scores can be used to select a + * better busiest. + */ + if (sched_asym_ipcc_pick(sds->busiest, sg, &sds->busiest_stat, sgs)) + return false; + break; case group_misfit_task: