From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from e17.ny.us.ibm.com ([129.33.205.207]:54408 "EHLO e17.ny.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752825AbbFRIGL (ORCPT ); Thu, 18 Jun 2015 04:06:11 -0400 Received: from /spool/local by e17.ny.us.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Thu, 18 Jun 2015 04:06:10 -0400 Message-ID: <55827BEB.4020900@linux.vnet.ibm.com> Date: Thu, 18 Jun 2015 13:36:03 +0530 From: Preeti U Murthy MIME-Version: 1.0 To: stable@vger.kernel.org CC: mingo@kernel.org, Thomas Gleixner , LKML Subject: Re: [PATCH 0/5] Fix race in hrtimer broadcast and take care of dependencies References: <20150513114717.27807.18777.stgit@preeti.in.ibm.com> In-Reply-To: <20150513114717.27807.18777.stgit@preeti.in.ibm.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Sender: stable-owner@vger.kernel.org List-ID: A reminder to pick this series up for stable-4.0 On 05/13/2015 05:19 PM, Preeti U Murthy wrote: > The intention was to backport only PATCH[5/5]: clockevents: Fix > cpu_down() race for hrtimer based broadcasting, > > but this depends on commits upstream which did cleanup and > reorganization of code. However these commits cannot be cherry-picked as > is. There are a few minor changes upstream and hence these commits had > to be backported too. Reasons for backports are given below: > > 1. PATCH[1/5]: The upstream commit introduced a duplicate function which > was removed much later by commit: > > 9eed56e889d: clockevents: Clean up clockchips.h > > We do not need to cherry-pick the entire above commit because the backport > was a simple one. > > 2. PATCH[3/5]: There are newer functions introduced upstream through > > 554ef3876c6: clockevents: Handle tick device's resume separately > > Yet again we did not need to cherry pick the entire above commit, > firstly because it introduces significant new code, secondly because the > backport was a simple one. > > 3. PATCH[4/5]: The function clockevent_set_mode() was renamed to > clockevent_set_state() upstream by commit: > > 77e32c89a7: clockevents: Manage device's state separately for > the core > > We do not need to cherry-pick this commit for the same reason as in (2). > > This series needs to be applied on top of stable-4.0. > --- > > Preeti U Murthy (1): > clockevents: Fix cpu_down() race for hrtimer based broadcasting > > Thomas Gleixner (4): > clockevents: Remove CONFIG_GENERIC_CLOCKEVENTS_BUILD > tick: Move clocksource related stuff to timekeeping.h > tick: Simplify tick-internal.h > tick: Move core only declarations and functions to core > > > include/linux/clockchips.h | 23 ++---- > include/linux/tick.h | 138 ++++++----------------------------- > kernel/cpu.c | 2 + > kernel/time/Kconfig | 6 -- > kernel/time/Makefile | 6 +- > kernel/time/clockevents.c | 3 - > kernel/time/hrtimer.c | 2 - > kernel/time/jiffies.c | 2 - > kernel/time/ntp.c | 1 > kernel/time/tick-broadcast.c | 19 +++-- > kernel/time/tick-internal.h | 166 ++++++++++++++++-------------------------- > kernel/time/tick-sched.c | 7 ++ > kernel/time/tick-sched.h | 64 ++++++++++++++++ > kernel/time/timekeeping.h | 7 ++ > kernel/time/timer_list.c | 2 - > 15 files changed, 190 insertions(+), 258 deletions(-) > create mode 100644 kernel/time/tick-sched.h > > -- >