From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754130AbdBHKZD (ORCPT ); Wed, 8 Feb 2017 05:25:03 -0500 Received: from Galois.linutronix.de ([146.0.238.70]:53404 "EHLO Galois.linutronix.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753737AbdBHKYn (ORCPT ); Wed, 8 Feb 2017 05:24:43 -0500 Date: Wed, 8 Feb 2017 11:24:01 +0100 (CET) From: Thomas Gleixner To: John Stultz cc: Kees Cook , Xing Gao , Jessica Frazelle , "Eric W. Biederman" , Jonathan Corbet , Tejun Heo , Lai Jiangshan , Petr Mladek , Andrew Morton , Oleg Nesterov , Nicolas Iooss , Nicolas Pitre , Richard Cochran , "Paul E. McKenney" , Michal Marek , Josh Poimboeuf , Dmitry Vyukov , Olof Johansson , Shuah Khan , linux-doc@vger.kernel.org, lkml , kernel-hardening@lists.openwall.com Subject: Re: [PATCH] time: Remove CONFIG_TIMER_STATS In-Reply-To: Message-ID: References: <20170207234031.GA27933@beast> User-Agent: Alpine 2.20 (DEB 67 2015-01-07) MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, 7 Feb 2017, John Stultz wrote: > On Tue, Feb 7, 2017 at 3:40 PM, Kees Cook wrote: > > Currently CONFIG_TIMER_STATS exposes process information across namespaces: > > > > kernel/time/timer_list.c print_timer(): > > > > SEQ_printf(m, ", %s/%d", tmp, timer->start_pid); > > > > /proc/timer_list: > > > > #11: <0000000000000000>, hrtimer_wakeup, S:01, do_nanosleep, cron/2570 > > > > Given that the tracer can give the same information, this patch entirely > > removes CONFIG_TIMER_STATS. > > > > Suggested-by: Thomas Gleixner > > Signed-off-by: Kees Cook > > I don't have an issue with this, but I worry this would break some > tooling out there. Should it be marked as deprecated first? > > Or maybe just pulling the band-aid off is the best way? I think we should just kill it. No tools can really rely on the behaviour of that file because it depends on CONFIG_TIMER_STATS and the information available there is just a random snapshot of queued timers at a given point of time, which is in no way usefull. Thanks, tglx From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Wed, 8 Feb 2017 11:24:01 +0100 (CET) From: Thomas Gleixner In-Reply-To: Message-ID: References: <20170207234031.GA27933@beast> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Subject: [kernel-hardening] Re: [PATCH] time: Remove CONFIG_TIMER_STATS To: John Stultz Cc: Kees Cook , Xing Gao , Jessica Frazelle , "Eric W. Biederman" , Jonathan Corbet , Tejun Heo , Lai Jiangshan , Petr Mladek , Andrew Morton , Oleg Nesterov , Nicolas Iooss , Nicolas Pitre , Richard Cochran , "Paul E. McKenney" , Michal Marek , Josh Poimboeuf , Dmitry Vyukov , Olof Johansson , Shuah Khan , linux-doc@vger.kernel.org, lkml , kernel-hardening@lists.openwall.com List-ID: On Tue, 7 Feb 2017, John Stultz wrote: > On Tue, Feb 7, 2017 at 3:40 PM, Kees Cook wrote: > > Currently CONFIG_TIMER_STATS exposes process information across namespaces: > > > > kernel/time/timer_list.c print_timer(): > > > > SEQ_printf(m, ", %s/%d", tmp, timer->start_pid); > > > > /proc/timer_list: > > > > #11: <0000000000000000>, hrtimer_wakeup, S:01, do_nanosleep, cron/2570 > > > > Given that the tracer can give the same information, this patch entirely > > removes CONFIG_TIMER_STATS. > > > > Suggested-by: Thomas Gleixner > > Signed-off-by: Kees Cook > > I don't have an issue with this, but I worry this would break some > tooling out there. Should it be marked as deprecated first? > > Or maybe just pulling the band-aid off is the best way? I think we should just kill it. No tools can really rely on the behaviour of that file because it depends on CONFIG_TIMER_STATS and the information available there is just a random snapshot of queued timers at a given point of time, which is in no way usefull. Thanks, tglx