[LITMUS^RT] rtspin running 100% or 90% of wcet? -- Sisu

[Sisu Xi]

Hi, Jonathan:

Could you clarify why without budget enforcement, it will be off?

I understand that with budget enforcement, the job will be preempted when
they run out of budget. When configured as NO_ENFORCEMENT, I assume the job
would run to completion, as long as it has the earliest deadline. Right?

Also, I don't see any code to off the jobs. In the loop_for() function,
while (now + last_loop < start + exec_time) {
loop_start = now;
tmp += loop_once();
now = cputime();
last_loop = now - loop_start;

Looks to me that there is no off at all.

Thanks for your help!

Sisu




On Wed, Apr 17, 2013 at 5:17 PM, Jonathan Herman <hermanjl at cs.unc.edu>wrote:

> Disclaimer: the <2% is a guess, not a fact. I haven't measured the exact
> difference between rtspins WCET and actual execution time with a 1.0 scale.
>
>
> On Wed, Apr 17, 2013 at 6:16 PM, Jonathan Herman <hermanjl at cs.unc.edu>wrote:
>
>> If you use budget enforcement, then yes, that is a reasonable number for
>> preventing WCET overruns. .98 would probably work as well. Without budget
>> enforcement, I wouldn't worry about specifying a scale. Leave it at 1.0 as
>> your jobs will be off by <2%.
>>
>>
>> On Wed, Apr 17, 2013 at 5:07 PM, Sisu Xi <xisisu at gmail.com> wrote:
>>
>>> HI, Glenn:
>>>
>>> Thanks for your reply.
>>>
>>> If the scale is set to 1.0, which means all the overhead is *not*
>>> counted.
>>>
>>> Usually what's the value for the scale parameter as a reasonable setting
>>> on a standard intel 64bit platform? 0.95?
>>>
>>> Thanks.
>>>
>>> Sisu
>>>
>>>
>>> On Wed, Apr 17, 2013 at 4:04 PM, Glenn Elliott <gelliott at cs.unc.edu>wrote:
>>>
>>>> Yes.  The 90% thing is out of date.  Use the scale parameter.
>>>>
>>>> -Glenn
>>>>
>>>> On Apr 17, 2013, at 4:58 PM, Sisu Xi <xisisu at gmail.com> wrote:
>>>>
>>>> Hi, all:
>>>>
>>>> I was reading the mailing list and found this discussion:
>>>>
>>>> https://lists.litmus-rt.org/pipermail/litmus-dev/2012/000095.html
>>>>
>>>> It says rtspin cuts the wcet to 90% of its actual specified wcet.
>>>>
>>>> Is it the same for the current release?
>>>>
>>>> I check the code and cannot find where does this 90% coming from.
>>>>
>>>> It has the "scale" variable, but is set to 1.0
>>>>
>>>> Thanks very much!
>>>>
>>>> Sisu
>>>>
>>>>
>>>> --
>>>> Sisu Xi, PhD Candidate
>>>>
>>>> http://www.cse.wustl.edu/~xis/
>>>> Department of Computer Science and Engineering
>>>> Campus Box 1045
>>>> Washington University in St. Louis
>>>> One Brookings Drive
>>>> St. Louis, MO 63130
>>>>
>>>> _______________________________________________
>>>> litmus-dev mailing list
>>>> litmus-dev at lists.litmus-rt.org
>>>> https://lists.litmus-rt.org/listinfo/litmus-dev
>>>>
>>>>
>>>> _______________________________________________
>>>> litmus-dev mailing list
>>>> litmus-dev at lists.litmus-rt.org
>>>> https://lists.litmus-rt.org/listinfo/litmus-dev
>>>>
>>>>
>>>
>>>
>>> --
>>> Sisu Xi, PhD Candidate
>>>
>>> http://www.cse.wustl.edu/~xis/
>>> Department of Computer Science and Engineering
>>> Campus Box 1045
>>> Washington University in St. Louis
>>> One Brookings Drive
>>> St. Louis, MO 63130
>>>
>>> _______________________________________________
>>> litmus-dev mailing list
>>> litmus-dev at lists.litmus-rt.org
>>> https://lists.litmus-rt.org/listinfo/litmus-dev
>>>
>>>
>>
>>
>> --
>> Jonathan Herman
>> Department of Computer Science at UNC Chapel Hill
>>
>
>
>
> --
> Jonathan Herman
> Department of Computer Science at UNC Chapel Hill
>
> _______________________________________________
> litmus-dev mailing list
> litmus-dev at lists.litmus-rt.org
> https://lists.litmus-rt.org/listinfo/litmus-dev
>
>


-- 
Sisu Xi, PhD Candidate

http://www.cse.wustl.edu/~xis/
Department of Computer Science and Engineering
Campus Box 1045
Washington University in St. Louis
One Brookings Drive
St. Louis, MO 63130
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.litmus-rt.org/pipermail/litmus-dev/attachments/20130417/b19acac2/attachment.html>