From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754655AbbFKMzu (ORCPT ); Thu, 11 Jun 2015 08:55:50 -0400 Received: from casper.infradead.org ([85.118.1.10]:39694 "EHLO casper.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752881AbbFKMxw (ORCPT ); Thu, 11 Jun 2015 08:53:52 -0400 Message-Id: <20150611124743.072387650@infradead.org> User-Agent: quilt/0.61-1 Date: Thu, 11 Jun 2015 14:46:44 +0200 From: Peter Zijlstra To: umgwanakikbuti@gmail.com, mingo@elte.hu Cc: ktkhai@parallels.com, rostedt@goodmis.org, tglx@linutronix.de, juri.lelli@gmail.com, pang.xunlei@linaro.org, oleg@redhat.com, wanpeng.li@linux.intel.com, linux-kernel@vger.kernel.org, peterz@infradead.org Subject: [PATCH 08/18] hrtimer: Remove HRTIMER_STATE_MIGRATE References: <20150611124636.448700267@infradead.org> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Disposition: inline; filename=oleg_nesterov-hrtimer-allow_hrtimer__function_to_free_the_timer.patch Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org I do not understand HRTIMER_STATE_MIGRATE. Unless I am totally confused it looks buggy and simply unneeded. migrate_hrtimer_list() sets it to keep hrtimer_active() == T, but this is not enough: this can fool, say, hrtimer_is_queued() in dequeue_signal(). Can't migrate_hrtimer_list() simply use HRTIMER_STATE_ENQUEUED? This fixes the race and we can kill STATE_MIGRATE. Signed-off-by: Oleg Nesterov Signed-off-by: Peter Zijlstra (Intel) --- include/linux/hrtimer.h | 6 +----- kernel/time/hrtimer.c | 7 ++----- 2 files changed, 3 insertions(+), 10 deletions(-) --- a/include/linux/hrtimer.h +++ b/include/linux/hrtimer.h @@ -70,17 +70,13 @@ enum hrtimer_restart { * the handling of the timer. * * The HRTIMER_STATE_ENQUEUED bit is always or'ed to the current state - * to preserve the HRTIMER_STATE_CALLBACK in the above scenario. This - * also affects HRTIMER_STATE_MIGRATE where the preservation is not - * necessary. HRTIMER_STATE_MIGRATE is cleared after the timer is - * enqueued on the new cpu. + * to preserve the HRTIMER_STATE_CALLBACK in the above scenario. * * All state transitions are protected by cpu_base->lock. */ #define HRTIMER_STATE_INACTIVE 0x00 #define HRTIMER_STATE_ENQUEUED 0x01 #define HRTIMER_STATE_CALLBACK 0x02 -#define HRTIMER_STATE_MIGRATE 0x04 /** * struct hrtimer - the basic hrtimer structure --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -1510,11 +1510,11 @@ static void migrate_hrtimer_list(struct debug_deactivate(timer); /* - * Mark it as STATE_MIGRATE not INACTIVE otherwise the + * Mark it as ENQUEUED not INACTIVE otherwise the * timer could be seen as !active and just vanish away * under us on another CPU */ - __remove_hrtimer(timer, old_base, HRTIMER_STATE_MIGRATE, 0); + __remove_hrtimer(timer, old_base, HRTIMER_STATE_ENQUEUED, 0); timer->base = new_base; /* * Enqueue the timers on the new cpu. This does not @@ -1525,9 +1525,6 @@ static void migrate_hrtimer_list(struct * event device. */ enqueue_hrtimer(timer, new_base); - - /* Clear the migration state bit */ - timer->state &= ~HRTIMER_STATE_MIGRATE; } }