From patchwork Tue Apr 21 05:19:19 2009 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Len Brown X-Patchwork-Id: 19119 Received: from vger.kernel.org (vger.kernel.org [209.132.176.167]) by demeter.kernel.org (8.14.2/8.14.2) with ESMTP id n3L5JZK5019416 for ; Tue, 21 Apr 2009 05:19:35 GMT Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1750807AbZDUFTe (ORCPT ); Tue, 21 Apr 2009 01:19:34 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1750811AbZDUFTe (ORCPT ); Tue, 21 Apr 2009 01:19:34 -0400 Received: from vms173005pub.verizon.net ([206.46.173.5]:14379 "EHLO vms173005pub.verizon.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750807AbZDUFTd (ORCPT ); Tue, 21 Apr 2009 01:19:33 -0400 Received: from localhost.localdomain ([96.237.168.40]) by vms173005.mailsrvcs.net (Sun Java(tm) System Messaging Server 6.3-7.04 (built Sep 26 2008; 32bit)) with ESMTPA id <0KIF003L2QS9TY0S@vms173005.mailsrvcs.net> for linux-acpi@vger.kernel.org; Tue, 21 Apr 2009 00:19:27 -0500 (CDT) Received: from localhost.localdomain (d975xbx2 [127.0.0.1]) by localhost.localdomain (8.14.2/8.14.2) with ESMTP id n3L5JK55018414; Tue, 21 Apr 2009 01:19:21 -0400 Received: from localhost (lenb@localhost) by localhost.localdomain (8.14.2/8.14.2/Submit) with ESMTP id n3L5JJRM018409; Tue, 21 Apr 2009 01:19:20 -0400 X-Authentication-warning: localhost.localdomain: lenb owned process doing -bs Date: Tue, 21 Apr 2009 01:19:19 -0400 (EDT) From: Len Brown X-X-Sender: lenb@localhost.localdomain To: linux-acpi@vger.kernel.org Cc: Thomas Gleixner Subject: [PATCH] ACPI: idle: mark_tsc_unstable() at init-time, not run-time Message-id: User-Agent: Alpine 2.00 (LFD 1167 2008-08-23) MIME-version: 1.0 Content-type: TEXT/PLAIN; charset=US-ASCII Sender: linux-acpi-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-acpi@vger.kernel.org From: Len Brown The c2 and c3 idle handlers check tsc_halts_in_c() after every time they return from idle. Um, when?:-) Move this check to init-time to remove the unnecessary run-time overhead, and also to have the check complete before the first entry into the idle handler. ff69f2bba67bd45514923aaedbf40fe351787c59 (acpi: fix of pmtimer overflow that make Cx states time incorrect) replaced the hard-coded use of the PM-timer inside idle, with ktime_get_readl(), which possibly uses the TSC -- so it is now especially prudent to detect a broken TSC before entering idle. http://bugzilla.kernel.org/show_bug.cgi?id=13087 Signed-off-by: Len Brown Acked-by: Thomas Gleixner --- drivers/acpi/processor_idle.c | 15 +++++---------- 1 files changed, 5 insertions(+), 10 deletions(-) diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index 6fe1214..9d1f01e 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -581,6 +581,11 @@ static int acpi_processor_power_verify(struct acpi_processor *pr) for (i = 1; i < ACPI_PROCESSOR_MAX_POWER; i++) { struct acpi_processor_cx *cx = &pr->power.states[i]; +#if defined (CONFIG_GENERIC_TIME) && defined (CONFIG_X86) + /* TSC could halt in idle, so notify users */ + if (tsc_halts_in_c(cx->type)) + mark_tsc_unstable("TSC halts in idle");; +#endif switch (cx->type) { case ACPI_STATE_C1: cx->valid = 1; @@ -871,11 +876,6 @@ static int acpi_idle_enter_simple(struct cpuidle_device *dev, kt2 = ktime_get_real(); idle_time = ktime_to_us(ktime_sub(kt2, kt1)); -#if defined (CONFIG_GENERIC_TIME) && defined (CONFIG_X86) - /* TSC could halt in idle, so notify users */ - if (tsc_halts_in_c(cx->type)) - mark_tsc_unstable("TSC halts in idle");; -#endif sleep_ticks = us_to_pm_timer_ticks(idle_time); /* Tell the scheduler how much we idled: */ @@ -989,11 +989,6 @@ static int acpi_idle_enter_bm(struct cpuidle_device *dev, spin_unlock(&c3_lock); } -#if defined (CONFIG_GENERIC_TIME) && defined (CONFIG_X86) - /* TSC could halt in idle, so notify users */ - if (tsc_halts_in_c(ACPI_STATE_C3)) - mark_tsc_unstable("TSC halts in idle"); -#endif sleep_ticks = us_to_pm_timer_ticks(idle_time); /* Tell the scheduler how much we idled: */ sched_clock_idle_wakeup_event(sleep_ticks*PM_TIMER_TICK_NS);