real-time priority

real-time priority

[Rodrigo Amestica]

Hi,

general questions from somebody who's just starting to look into the 
preempt patch.

the article linked below makes direct reference to the behavior of the 
preempt rt patch. The article was written in 2007. How much of that 
still holds true in the current version of the patch?

I'm basically wondering about the first sentences in the abstract of 
that article. The big picture seems to be rather simple: calls to 
specific linux services could break the performance of a real time task 
by depending on lower priority tasks to complete that service.

Does this mean that one must always be alert to which system services 
are invoked during the execution of a real time task? Does the vanilla 
kernel or the preempt-rt patch tries to somehow detect this condition 
and alert somehow?

thanks,
 Rodrigo

Re: real-time priority

[Rodrigo Amestica]

yes, i forgot to include the link. now it goes.

http://www.ddj.com/linux-open-source/200900827

Rodrigo Amestica wrote:
> Hi,
>
> general questions from somebody who's just starting to look into the 
> preempt patch.
>
> the article linked below makes direct reference to the behavior of the 
> preempt rt patch. The article was written in 2007. How much of that 
> still holds true in the current version of the patch?
>
> I'm basically wondering about the first sentences in the abstract of 
> that article. The big picture seems to be rather simple: calls to 
> specific linux services could break the performance of a real time 
> task by depending on lower priority tasks to complete that service.
>
> Does this mean that one must always be alert to which system services 
> are invoked during the execution of a real time task? Does the vanilla 
> kernel or the preempt-rt patch tries to somehow detect this condition 
> and alert somehow?
>
> thanks,
> Rodrigo
>
> -- 
> To unsubscribe from this list: send the line "unsubscribe 
> linux-rt-users" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
> ____________________________________________________________________________________ 
>
> Click here to find the perfect banking opportunity!
> http://ads.lavabit.com/fc/BLSrjwrz8JR8QtO2thrVR65MkKPO7Tm4zsO0sj7xJYVXikFfIocXosXcLFO/ 
>
> ____________________________________________________________________________________ 
>

Re: real-time priority

[Frank Rowand]

Rodrigo Amestica wrote:
> yes, i forgot to include the link. now it goes.
> 
> http://www.ddj.com/linux-open-source/200900827
> 
> Rodrigo Amestica wrote:
>> Hi,
>>
>> general questions from somebody who's just starting to look into the
>> preempt patch.
>>
>> the article linked below makes direct reference to the behavior of the
>> preempt rt patch. The article was written in 2007. How much of that
>> still holds true in the current version of the patch?
>>
>> I'm basically wondering about the first sentences in the abstract of
>> that article. The big picture seems to be rather simple: calls to
>> specific linux services could break the performance of a real time
>> task by depending on lower priority tasks to complete that service.
>>
>> Does this mean that one must always be alert to which system services
>> are invoked during the execution of a real time task?

Yes.

>> Does the vanilla
>> kernel or the preempt-rt patch tries to somehow detect this condition
>> and alert somehow?

I am not aware of any such test, but it sounds like an interesting idea.

>>
>> thanks,
>> Rodrigo
>>

Re: real-time priority

[Frank Rowand]

Rodrigo Amestica wrote:
> Hi,
> 
> general questions from somebody who's just starting to look into the
> preempt patch.
> 
> the article linked below makes direct reference to the behavior of the

The link is missing.


> preempt rt patch. The article was written in 2007. How much of that
> still holds true in the current version of the patch?
> 
> I'm basically wondering about the first sentences in the abstract of
> that article. The big picture seems to be rather simple: calls to
> specific linux services could break the performance of a real time task
> by depending on lower priority tasks to complete that service.
> 
> Does this mean that one must always be alert to which system services
> are invoked during the execution of a real time task? Does the vanilla
> kernel or the preempt-rt patch tries to somehow detect this condition
> and alert somehow?
> 
> thanks,
> Rodrigo
> 
> -- 
> To unsubscribe from this list: send the line "unsubscribe
> linux-rt-users" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
>