From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759908Ab3D3JfJ (ORCPT ); Tue, 30 Apr 2013 05:35:09 -0400 Received: from mx1.redhat.com ([209.132.183.28]:53508 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753024Ab3D3JfG (ORCPT ); Tue, 30 Apr 2013 05:35:06 -0400 From: Stanislaw Gruszka To: Ingo Molnar , Peter Zijlstra Cc: Frederic Weisbecker , hpa@zytor.com, rostedt@goodmis.org, akpm@linux-foundation.org, tglx@linutronix.de, Linus Torvalds , linux-kernel@vger.kernel.org, Dave Hansen , Stanislaw Gruszka Subject: [PATCH -tip 2/4] sched: Do not account bogus utime Date: Tue, 30 Apr 2013 11:35:05 +0200 Message-Id: <1367314507-9728-2-git-send-email-sgruszka@redhat.com> In-Reply-To: <1367314507-9728-1-git-send-email-sgruszka@redhat.com> References: <1367314507-9728-1-git-send-email-sgruszka@redhat.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Due to rounding in scale_stime(), for big numbers, scaled stime values will grow in chunks. Since rtime grow in jiffies and we calculate utime like below: prev->stime = max(prev->stime, stime); prev->utime = max(prev->utime, rtime - prev->stime); we could erroneously account stime values as utime. To prevent that only update prev->{u,s}time values when they are smaller than current rtime. Signed-off-by: Stanislaw Gruszka --- kernel/sched/cputime.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c index b3dd984..3f192bf 100644 --- a/kernel/sched/cputime.c +++ b/kernel/sched/cputime.c @@ -581,6 +581,14 @@ static void cputime_adjust(struct task_cputime *curr, */ rtime = nsecs_to_cputime(curr->sum_exec_runtime); + /* + * Update userspace visible utime/stime values only if actual execution + * time is bigger than already exported. Note that can happen, that we + * provided bigger values due to scaling inaccuracy on big numbers. + */ + if (prev->stime + prev->utime >= rtime) + goto out; + if (!rtime) { stime = 0; } else if (!total) { @@ -598,6 +606,7 @@ static void cputime_adjust(struct task_cputime *curr, prev->stime = max(prev->stime, stime); prev->utime = max(prev->utime, rtime - prev->stime); +out: *ut = prev->utime; *st = prev->stime; } -- 1.7.11.7