From mboxrd@z Thu Jan 1 00:00:00 1970 From: Michael Bohan Subject: [PATCH 2/2] hrtimer: Prevent enqueue of hrtimer on dead CPU Date: Wed, 10 Apr 2013 14:07:48 -0700 Message-ID: <1365628068-32738-3-git-send-email-mbohan@codeaurora.org> References: <1365628068-32738-1-git-send-email-mbohan@codeaurora.org> Return-path: In-Reply-To: <1365628068-32738-1-git-send-email-mbohan@codeaurora.org> Sender: linux-kernel-owner@vger.kernel.org To: tglx@linutronix.de Cc: linux-kernel@vger.kernel.org, linux-arm-msm@vger.kernel.org List-Id: linux-arm-msm@vger.kernel.org When switching the hrtimer cpu_base, we briefly allow for preemption to become enabled by unlocking the cpu_base lock. During this time, the CPU corresponding to the new cpu_base that was selected may in fact go offline. In this scenario, the hrtimer is enqueued to a CPU that's not online, and therefore it never fires. As an example, consider this example: CPU #0 CPU #1 ---- ---- ... hrtimer_start() lock_hrtimer_base() switch_hrtimer_base() cpu = hrtimer_get_target() -> 1 spin_unlock(&cpu_base->lock) spin_lock(&new_base->lock) this_cpu = 0 cpu != this_cpu enqueue_hrtimer(cpu_base #1) To prevent this scenario, verify that the CPU corresponding to the new cpu_base is indeed online before selecting it in hrtimer_switch_base(). If it's not online, fallback to using the base of the current CPU. Signed-off-by: Michael Bohan --- kernel/hrtimer.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c index 3f0bce9..54c5d98 100644 --- a/kernel/hrtimer.c +++ b/kernel/hrtimer.c @@ -227,7 +227,8 @@ again: this_cpu = smp_processor_id(); - if (cpu != this_cpu && hrtimer_check_target(timer, new_base)) { + if (cpu != this_cpu && (hrtimer_check_target(timer, new_base) + || !cpu_online(cpu))) { raw_spin_unlock(&new_base->cpu_base->lock); raw_spin_lock(&base->cpu_base->lock); cpu = smp_processor_id(); -- The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum, hosted by The Linux Foundation