From patchwork Thu Jul 26 13:08:54 2018 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Prarit Bhargava X-Patchwork-Id: 10545881 X-Patchwork-Delegate: lenb@kernel.org Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id E843114E2 for ; Thu, 26 Jul 2018 13:09:03 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id D6A172B1FE for ; Thu, 26 Jul 2018 13:09:03 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id CAEBD2B201; Thu, 26 Jul 2018 13:09:03 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-7.9 required=2.0 tests=BAYES_00,MAILING_LIST_MULTI, RCVD_IN_DNSWL_HI autolearn=ham version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 32FAD2B1FE for ; Thu, 26 Jul 2018 13:09:03 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1730173AbeGZOZt (ORCPT ); Thu, 26 Jul 2018 10:25:49 -0400 Received: from mx3-rdu2.redhat.com ([66.187.233.73]:52292 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1729873AbeGZOZs (ORCPT ); Thu, 26 Jul 2018 10:25:48 -0400 Received: from smtp.corp.redhat.com (int-mx03.intmail.prod.int.rdu2.redhat.com [10.11.54.3]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 50E37818B107; Thu, 26 Jul 2018 13:09:00 +0000 (UTC) Received: from prarit.bos.redhat.com (prarit-guest.khw.lab.eng.bos.redhat.com [10.16.186.145]) by smtp.corp.redhat.com (Postfix) with ESMTP id 165E7111AF3C; Thu, 26 Jul 2018 13:09:00 +0000 (UTC) From: Prarit Bhargava To: linux-pm@vger.kernel.org Cc: Prarit Bhargava , lenb@kernel.org, artem.bityutskiy@intel.com Subject: [PATCH] tools/power/turbostat: Fix logical node enumeration to allow for non-sequential physical nodes Date: Thu, 26 Jul 2018 09:08:54 -0400 Message-Id: <20180726130854.5139-1-prarit@redhat.com> X-Scanned-By: MIMEDefang 2.78 on 10.11.54.3 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.11.55.8]); Thu, 26 Jul 2018 13:09:00 +0000 (UTC) X-Greylist: inspected by milter-greylist-4.5.16 (mx1.redhat.com [10.11.55.8]); Thu, 26 Jul 2018 13:09:00 +0000 (UTC) for IP:'10.11.54.3' DOMAIN:'int-mx03.intmail.prod.int.rdu2.redhat.com' HELO:'smtp.corp.redhat.com' FROM:'prarit@redhat.com' RCPT:'' Sender: linux-pm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-pm@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP turbostat fails on some multi-package topologies because the logical node enumeration assumes that the nodes are sequentially numbered which causes the logical numa nodes to not be enumerated, or enumerated incorrectly. Use a more robust enumeration algorithm which allows for non-seqential physical nodes. Signed-off-by: Prarit Bhargava Cc: lenb@kernel.org Cc: artem.bityutskiy@intel.com --- tools/power/x86/turbostat/turbostat.c | 106 ++++++++++++++++------------------ 1 file changed, 50 insertions(+), 56 deletions(-) diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index 02e71accad16..2b0135599f37 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -2471,55 +2471,43 @@ int get_core_id(int cpu) void set_node_data(void) { - char path[80]; - FILE *filep; - int pkg, node, cpu; - - struct pkg_node_info { - int count; - int min; - } *pni; - - pni = calloc(topo.num_packages, sizeof(struct pkg_node_info)); - if (!pni) - err(1, "calloc pkg_node_count"); - - for (pkg = 0; pkg < topo.num_packages; pkg++) - pni[pkg].min = topo.num_cpus; - - for (node = 0; node <= topo.max_node_num; node++) { - /* find the "first" cpu in the node */ - sprintf(path, "/sys/bus/node/devices/node%d/cpulist", node); - filep = fopen(path, "r"); - if (!filep) - continue; - fscanf(filep, "%d", &cpu); - fclose(filep); - - pkg = cpus[cpu].physical_package_id; - pni[pkg].count++; - - if (node < pni[pkg].min) - pni[pkg].min = node; - } - - for (pkg = 0; pkg < topo.num_packages; pkg++) - if (pni[pkg].count > topo.nodes_per_pkg) - topo.nodes_per_pkg = pni[0].count; - - /* Fake 1 node per pkg for machines that don't - * expose nodes and thus avoid -nan results - */ - if (topo.nodes_per_pkg == 0) - topo.nodes_per_pkg = 1; - - for (cpu = 0; cpu < topo.num_cpus; cpu++) { - pkg = cpus[cpu].physical_package_id; - node = cpus[cpu].physical_node_id; - cpus[cpu].logical_node_id = node - pni[pkg].min; + int pkg, node, lnode, cpu, cpux; + int cpu_count; + + /* initialize logical_node_id */ + for (cpu = 0; cpu <= topo.max_cpu_num; ++cpu) + cpus[cpu].logical_node_id = -1; + + cpu_count = 0; + for (pkg = 0; pkg < topo.num_packages; pkg++) { + lnode = 0; + for (cpu = 0; cpu <= topo.max_cpu_num; ++cpu) { + if (cpus[cpu].physical_package_id != pkg) + continue; + /* find a cpu with an unset logical_node_id */ + if (cpus[cpu].logical_node_id != -1) + continue; + cpus[cpu].logical_node_id = lnode; + node = cpus[cpu].physical_node_id; + cpu_count++; + /* + * find all matching cpus on this pkg and set + * the logical_node_id + */ + for (cpux = cpu; cpux <= topo.max_cpu_num; cpux++) { + if ((cpus[cpux].physical_package_id == pkg) && + (cpus[cpux].physical_node_id == node)) { + cpus[cpux].logical_node_id = lnode; + cpu_count++; + } + } + lnode++; + if (lnode > topo.nodes_per_pkg) + topo.nodes_per_pkg = lnode; + } + if (cpu_count >= topo.max_cpu_num) + break; } - free(pni); - } int get_physical_node_id(struct cpu_topology *thiscpu) @@ -4840,14 +4828,6 @@ void topology_probe() max_siblings = siblings; if (cpus[i].thread_id == 0) topo.num_cores++; - - if (debug > 1) - fprintf(outf, - "cpu %d pkg %d node %d core %d thread %d\n", - i, cpus[i].physical_package_id, - cpus[i].physical_node_id, - cpus[i].physical_core_id, - cpus[i].thread_id); } topo.cores_per_node = max_core_id + 1; @@ -4873,6 +4853,20 @@ void topology_probe() topo.threads_per_core = max_siblings; if (debug > 1) fprintf(outf, "max_siblings %d\n", max_siblings); + + if (debug < 1) + return; + + for (i = 0; i <= topo.max_cpu_num; ++i) { + fprintf(outf, + "cpu %d pkg %d node %d lnode %d core %d thread %d\n", + i, cpus[i].physical_package_id, + cpus[i].physical_node_id, + cpus[i].logical_node_id, + cpus[i].physical_core_id, + cpus[i].thread_id); + } + } void