a bug in AS scheduler?

a bug in AS scheduler?

[Xiaoning Ding]

Hi,

I am reading the source code AS scheduler in 2.6.18(as-ioscheduler.c).
In function as_close_req, variable delay is in millisecond, while
ad->antic_expire is in jiffies. Doesn't the comparison of delay and
ad->antic_expire make any problem?
The related source code is quoted blow:

	if (ad->antic_status == ANTIC_OFF || !ad->ioc_finished)
		delay = 0;
	else
		delay = ((jiffies - ad->antic_start) * 1000) / HZ;

	if (delay == 0)
		delta = 8192;
	else if (delay <= 20 && delay <= ad->antic_expire)
		delta = 8192 << delay;
	else
		return 1;


Xiaoning

Re: a bug in AS scheduler?

[Jens Axboe]

On Tue, Feb 27 2007, Xiaoning Ding wrote:
> Hi,
> 
> I am reading the source code AS scheduler in 2.6.18(as-ioscheduler.c).
> In function as_close_req, variable delay is in millisecond, while
> ad->antic_expire is in jiffies. Doesn't the comparison of delay and
> ad->antic_expire make any problem?
> The related source code is quoted blow:
> 
> 	if (ad->antic_status == ANTIC_OFF || !ad->ioc_finished)
> 		delay = 0;
> 	else
> 		delay = ((jiffies - ad->antic_start) * 1000) / HZ;

antic_start is in jiffies, the difference is here multiplied by 1000 and
divided by HZ to turn it into msecs. so delay is in msecs.

-- 
Jens Axboe

Re: a bug in AS scheduler?

[Benoit Boissinot]

On 2/28/07, Jens Axboe <jens.axboe@oracle.com> wrote:
> On Tue, Feb 27 2007, Xiaoning Ding wrote:
> > Hi,
> >
> > I am reading the source code AS scheduler in 2.6.18(as-ioscheduler.c).
> > In function as_close_req, variable delay is in millisecond, while
> > ad->antic_expire is in jiffies. Doesn't the comparison of delay and
> > ad->antic_expire make any problem?
> > The related source code is quoted blow:
> >
> >       if (ad->antic_status == ANTIC_OFF || !ad->ioc_finished)
> >               delay = 0;
> >       else
> >               delay = ((jiffies - ad->antic_start) * 1000) / HZ;
>
> antic_start is in jiffies, the difference is here multiplied by 1000 and
> divided by HZ to turn it into msecs. so delay is in msecs.
>
I am pretty sure Xiaoning was talking about antic_expire, particularly
this comparison:

     else if (delay <= 20 && delay <= ad->antic_expire)

regards,

Benoit

Re: a bug in AS scheduler?

[Jens Axboe]

On Wed, Feb 28 2007, Benoit Boissinot wrote:
> On 2/28/07, Jens Axboe <jens.axboe@oracle.com> wrote:
> >On Tue, Feb 27 2007, Xiaoning Ding wrote:
> >> Hi,
> >>
> >> I am reading the source code AS scheduler in 2.6.18(as-ioscheduler.c).
> >> In function as_close_req, variable delay is in millisecond, while
> >> ad->antic_expire is in jiffies. Doesn't the comparison of delay and
> >> ad->antic_expire make any problem?
> >> The related source code is quoted blow:
> >>
> >>       if (ad->antic_status == ANTIC_OFF || !ad->ioc_finished)
> >>               delay = 0;
> >>       else
> >>               delay = ((jiffies - ad->antic_start) * 1000) / HZ;
> >
> >antic_start is in jiffies, the difference is here multiplied by 1000 and
> >divided by HZ to turn it into msecs. so delay is in msecs.
> >
> I am pretty sure Xiaoning was talking about antic_expire, particularly
> this comparison:
> 
>     else if (delay <= 20 && delay <= ad->antic_expire)

Ah, I misread the name. That does look like a bug, antic_expire is in
jiffies.

-- 
Jens Axboe

Re: a bug in AS scheduler?

[Xiaoning Ding]

Jens Axboe wrote:
> On Wed, Feb 28 2007, Benoit Boissinot wrote:
>> On 2/28/07, Jens Axboe <jens.axboe@oracle.com> wrote:
>>> On Tue, Feb 27 2007, Xiaoning Ding wrote:
>>>> Hi,
>>>>
>>>> I am reading the source code AS scheduler in 2.6.18(as-ioscheduler.c).
>>>> In function as_close_req, variable delay is in millisecond, while
>>>> ad->antic_expire is in jiffies. Doesn't the comparison of delay and
>>>> ad->antic_expire make any problem?
>>>> The related source code is quoted blow:
>>>>
>>>>       if (ad->antic_status == ANTIC_OFF || !ad->ioc_finished)
>>>>               delay = 0;
>>>>       else
>>>>               delay = ((jiffies - ad->antic_start) * 1000) / HZ;
>>> antic_start is in jiffies, the difference is here multiplied by 1000 and
>>> divided by HZ to turn it into msecs. so delay is in msecs.
>>>
>> I am pretty sure Xiaoning was talking about antic_expire, particularly
>> this comparison:
>>
>>     else if (delay <= 20 && delay <= ad->antic_expire)
> 
> Ah, I misread the name. That does look like a bug, antic_expire is in
> jiffies.

You got it.

Xiaoning