Priority Inheritance per se?

Priority Inheritance per se?

[Martin Däumler]

  Hello,

our team develops a Linux-based embedded system to
realise a programmable logic controller. So we investigate
several possibilities to augment Linux real-time capabilities.
Naturally, the PREEMPT_RT-patch is interesting. I searched
the wiki, the mailing list archive, listened to Mr. Assmann's
talk at the Chemnitz Linux Days 2010 and read the corresponding
chapter in "Building Embedded Linux Systems" by Karim Yaghmour.
Nevertheless, I have a question to makes thinks clear once and
for all:

Does the priority inheritance mechanism work without using
special (Pthread-) mutexes in userspace? (As far as I understood,
it does not.)

Imagine a hypothetical case: a high RT-priority (60) task triggers
a synchronous/blocking system call which triggers a device driver
whose ISR is a kernel thread with RT-priority 49. So, a medium
RT-priority (55) task may block the high RT-priority task, doesn't
it?


With kind regards,
Martin Daeumler

Re: Priority Inheritance per se?

[Sergio Ruocco]

Martin Däumler wrote:
> [snip]

Hi Martin,

before really deciding for the use of PI, you may want to read what
Victor Yodaiken had to say:

http://www.yodaiken.com/papers-and-talks/
Against priority inheritance, Victor Yodaiken. September 23, 2004
http://www.yodaiken.com/papers/inherit.pdf

Here is a general Linux RT workshop report, with some comments on PI:
http://lwn.net/Articles/354690/

	Sergio

-- 

Dr. Sergio Ruocco   Research Fellow    http://www.disco.unimib.it/ruocco
mailto:ruocco@disco.unimib.it / sergio.ruocco@gmail.com      NOMADIS Lab
phone: +39-02-6448-7914               Mobile, embedded real-time systems
skype: 'sergioruocco'  Dip. di Informatica, Sistemistica e COmunicazione
Building U14, room 1003  Università degli Studi di Milano-Bicocca, Italy

W La ricerca http://video.google.com/videoplay?docid=-6518779854944519929

Your manuscript is both good and original; but the part that is good is
not original, and the part that is original is not good.
Samuel Johnson


--
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

Re: Priority Inheritance per se?

[Martin Däumler]

Sergio Ruocco wrote:

> before really deciding for the use of PI, you may want to read what
> Victor Yodaiken had to say:
>
> http://www.yodaiken.com/papers-and-talks/
> Against priority inheritance, Victor Yodaiken. September 23, 2004
> http://www.yodaiken.com/papers/inherit.pdf
>
> Here is a general Linux RT workshop report, with some comments on PI:
> http://lwn.net/Articles/354690/


Hi Sergio,

thanks for that reading. I (hope to) know which drawbacks and
shortcomings priority inheritance has. Before I can make a
decision about using an RT_PREEMPT patched Linux, I want to
know if the RT_PREEMPT-patch introduces priority inheritance
automatically, that is, without using any locks?

For example (from my opening post):

Imagine a hypothetical case: a high RT-priority (60) task
triggers a blocking system call which triggers a device driver
whose ISR is a kernel thread with RT-priority 49. So, a medium
RT-priority (55) task is able block the high RT-priority task,
isn't it?


With kind regards,
Martin Däumler
--
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