Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757221AbZD0RCo (ORCPT ); Mon, 27 Apr 2009 13:02:44 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754827AbZD0RCZ (ORCPT ); Mon, 27 Apr 2009 13:02:25 -0400 Received: from hera.kernel.org ([140.211.167.34]:33835 "EHLO hera.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756844AbZD0RCY (ORCPT ); Mon, 27 Apr 2009 13:02:24 -0400 Date: Mon, 27 Apr 2009 17:00:48 GMT From: tip-bot for Yinghai Lu To: linux-tip-commits@vger.kernel.org Cc: linux-kernel@vger.kernel.org, hpa@zytor.com, mingo@redhat.com, kiran@scalex86.org, yinghai@kernel.org, tglx@linutronix.de, mingo@elte.hu Reply-To: mingo@redhat.com, hpa@zytor.com, linux-kernel@vger.kernel.org, yinghai@kernel.org, kiran@scalex86.org, tglx@linutronix.de, mingo@elte.hu In-Reply-To: <49F5532A.5000802@kernel.org> References: <49F5532A.5000802@kernel.org> Subject: [tip:x86/apic] x86: Use dmi check in apic_is_clustered() on 64-bit to mark the TSC unstable Message-ID: Git-Commit-ID: e0e42142bab96404de535cceb85d6533d5ad7942 X-Mailer: tip-git-log-daemon MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Disposition: inline X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.0 (hera.kernel.org [127.0.0.1]); Mon, 27 Apr 2009 17:00:50 +0000 (UTC) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4076 Lines: 146 Commit-ID: e0e42142bab96404de535cceb85d6533d5ad7942 Gitweb: http://git.kernel.org/tip/e0e42142bab96404de535cceb85d6533d5ad7942 Author: Yinghai Lu AuthorDate: Sun, 26 Apr 2009 23:39:38 -0700 Committer: Ingo Molnar CommitDate: Mon, 27 Apr 2009 09:23:52 +0200 x86: Use dmi check in apic_is_clustered() on 64-bit to mark the TSC unstable We will have systems with 2 and more sockets 8cores/2thread, but we treat them as multi chassis - while they could have a stable TSC domain. Use DMI check instead. [ Impact: do not turn possibly stable TSCs off incorrectly ] Signed-off-by: Yinghai Lu Cc: Ravikiran Thirumalai LKML-Reference: <49F5532A.5000802@kernel.org> Signed-off-by: Ingo Molnar --- arch/x86/kernel/apic/apic.c | 86 +++++++++++++++++++++++++++++------------- 1 files changed, 59 insertions(+), 27 deletions(-) diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index 1386dbe..28f747d 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -2138,31 +2138,14 @@ static void apic_pm_activate(void) { } #endif /* CONFIG_PM */ #ifdef CONFIG_X86_64 -/* - * apic_is_clustered_box() -- Check if we can expect good TSC - * - * Thus far, the major user of this is IBM's Summit2 series: - * - * Clustered boxes may have unsynced TSC problems if they are - * multi-chassis. Use available data to take a good guess. - * If in doubt, go HPET. - */ -__cpuinit int apic_is_clustered_box(void) + +static int __cpuinit apic_cluster_num(void) { int i, clusters, zeros; unsigned id; u16 *bios_cpu_apicid; DECLARE_BITMAP(clustermap, NUM_APIC_CLUSTERS); - /* - * there is not this kind of box with AMD CPU yet. - * Some AMD box with quadcore cpu and 8 sockets apicid - * will be [4, 0x23] or [8, 0x27] could be thought to - * vsmp box still need checking... - */ - if ((boot_cpu_data.x86_vendor == X86_VENDOR_AMD) && !is_vsmp_box()) - return 0; - bios_cpu_apicid = early_per_cpu_ptr(x86_bios_cpu_apicid); bitmap_zero(clustermap, NUM_APIC_CLUSTERS); @@ -2198,18 +2181,67 @@ __cpuinit int apic_is_clustered_box(void) ++zeros; } - /* ScaleMP vSMPowered boxes have one cluster per board and TSCs are - * not guaranteed to be synced between boards - */ - if (is_vsmp_box() && clusters > 1) + return clusters; +} + +static int __cpuinitdata multi_checked; +static int __cpuinitdata multi; + +static int __cpuinit set_multi(const struct dmi_system_id *d) +{ + if (multi) + return 0; + printk(KERN_INFO "APIC: %s detected, Multi Chassis\n", d->ident); + multi = 1; + return 0; +} + +static const __cpuinitconst struct dmi_system_id multi_dmi_table[] = { + { + .callback = set_multi, + .ident = "IBM System Summit2", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "IBM"), + DMI_MATCH(DMI_PRODUCT_NAME, "Summit2"), + }, + }, + {} +}; + +static void __cpuinit dmi_check_multi(void) +{ + if (multi_checked) + return; + + dmi_check_system(multi_dmi_table); + multi_checked = 1; +} + +/* + * apic_is_clustered_box() -- Check if we can expect good TSC + * + * Thus far, the major user of this is IBM's Summit2 series: + * Clustered boxes may have unsynced TSC problems if they are + * multi-chassis. + * Use DMI to check them + */ +__cpuinit int apic_is_clustered_box(void) +{ + dmi_check_multi(); + if (multi) return 1; + if (!is_vsmp_box()) + return 0; + /* - * If clusters > 2, then should be multi-chassis. - * May have to revisit this when multi-core + hyperthreaded CPUs come - * out, but AFAIK this will work even for them. + * ScaleMP vSMPowered boxes have one cluster per board and TSCs are + * not guaranteed to be synced between boards */ - return (clusters > 2); + if (apic_cluster_num() > 1) + return 1; + + return 0; } #endif -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/