Re: [PATCH 06/12] irq: implement irq_schedule_poll()

[Tejun Heo <tj@kernel.org>]

On 06/15/2010 07:40 PM, Jonathan Corbet wrote:
> On Sun, 13 Jun 2010 17:31:32 +0200
> Tejun Heo <tj@kernel.org> wrote:
> 
>> Implement and use irq_schedule_poll() to schedule desc->poll_timer
>> instead of calling mod_timer directly.  irq_schedule_poll() is called
>> with desc->lock held and schedules the timer iff necessary - ie. if
>> the timer is offline or scheduled to expire later than requested.
>> This will be used to share desc->poll_timer.
> 
> One quick question...
> 
>> diff --git a/kernel/irq/spurious.c b/kernel/irq/spurious.c
>> index 0bce0e3..0b8fd0a 100644
>> --- a/kernel/irq/spurious.c
>> +++ b/kernel/irq/spurious.c
>> @@ -23,6 +23,8 @@ enum {
>>  
>>  	/* IRQ polling common parameters */
>>  	IRQ_POLL_INTV			= HZ / 100,	/* from the good ol' 100HZ tick */
>> +
>> +	IRQ_POLL_SLACK			= HZ / 1000,	/* 10% slack */
>>  };
> 
> Why would IRQ_POLL_SLACK be anything other than zero for HZ < 1000?
> It seems like you'd get no slack at all on most systems.  Am I missing
> something?

Jiri questioned the same thing.  I'll bump it to HZ / 250 so that it
at least has one tick slack on 250HZ configuration which is pretty
common these days.

Thanks.

-- 
tejun