[PATCHv4 01/11] clockevents: Prefer CPU local devices over global devices

[sboyd@codeaurora.org (Stephen Boyd)]

On 04/09/13 03:33, Mark Rutland wrote:
> On Mon, Apr 08, 2013 at 10:27:23PM +0100, Stephen Boyd wrote:
>> On an SMP system with only one global clockevent and a dummy
>> clockevent per CPU we run into problems. We want the dummy
>> clockevents to be registered as the per CPU tick devices, but
>> we can only achieve that if we register the dummy clockevents
>> before the global clockevent or if we artificially inflate the
>> rating of the dummy clockevents to be higher than the rating
>> of the global clockevent. Failure to do so leads to boot
>> hangs when the dummy timers are registered on all other CPUs
>> besides the CPU that accepted the global clockevent as its tick
>> device and there is no broadcast timer to poke the dummy
>> devices.
>>
>> If we're registering multiple clockevents and one clockevent is
>> global and the other is local to a particular CPU we should
>> choose to use the local clockevent regardless of the rating of
>> the device. This way, if the clockevent is a dummy it will take
>> the tick device duty as long as there isn't a higher rated tick
>> device and any global clockevent will be bumped out into
>> broadcast mode, fixing the problem described above.
> It might be worth pointing out that if dummy timers are only registered
> when there's more than one CPU, UP behaviour won't degrade from the
> current state of affairs.

Sure, I'll try to come up with something if I have to repost (looks like
I might have to rework the mct change).

>
>> Reported-by: Mark Rutland <mark.rutland@arm.com>
>> Cc: John Stultz <john.stultz@linaro.org>
>> Cc: Thomas Gleixner <tglx@linutronix.de>
>> Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
>> ---
>>  kernel/time/tick-common.c | 5 +++--
>>  1 file changed, 3 insertions(+), 2 deletions(-)
>>
>> diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c
>> index b1600a6..9ea59b9 100644
>> --- a/kernel/time/tick-common.c
>> +++ b/kernel/time/tick-common.c
>> @@ -251,9 +251,10 @@ static int tick_check_new_device(struct clock_event_device *newdev)
>>  		    !(newdev->features & CLOCK_EVT_FEAT_ONESHOT))
>>  			goto out_bc;
>>  		/*
>> -		 * Check the rating
>> +		 * Check the rating, but prefer CPU local devices
>>  		 */
>> -		if (curdev->rating >= newdev->rating)
>> +		if (curdev->rating >= newdev->rating &&
>> +		    cpumask_equal(curdev->cpumask, newdev->cpumask))
>>  			goto out_bc;
>>  	}
> This looks good to me. I tested this on a TC2 with a v3.9-rc6 kernel
> without architected timer support. The patch restores the ability to use
> the sp804 as a broadcast source.
>
> Tested-by: Mark Rutland <mark.rutland@arm.com>
>
> If possible I think this should sit in -next for a bit to make sure it
> doesn't have any unexpected side effects.

Thanks for testing.

-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
hosted by The Linux Foundation