[comp.realtime] WANTED: paper on priority inheritance

srm@Matrix.COM (Steve Morris) (11/09/90)

I'm looking for a paper (Tech. Rept. CMU-CS-87-181) called

   Priority Inheritance Protocols: An Approach to Real-Time Synchronization

by L. Sha, R. Rajkumar, and J.P. Lehoczky. Is it obtainable via FTP?

Thank you.

-- 
 _____________________________________________________________
| Steve Morris                         srm@matrix.com         |
| Matrix Corporation                   mcnc!matrx!srm         |
| 1203 New Hope Road                   (919)231-8000 Telephone|

walden@sparky.eecs.umich.edu (Eugene Marvin Walden) (11/13/90)

In article <1990Nov9.132404.5094@Matrix.COM> srm@matrx.matrix.com (Steve Morris) writes:
>I'm looking for a paper (Tech. Rept. CMU-CS-87-181) called
>   Priority Inheritance Protocols: An Approach to Real-Time Synchronization
>by L. Sha, R. Rajkumar, and J.P. Lehoczky. Is it obtainable via FTP?
>Thank you.

>| Steve Morris                         srm@matrix.com         |

   I don't know about that particular technical report, but the authors 
describe the approach in:

   Lehoczky, J. P., Sha, L., and Strosnider, J. K., "Enhanced Aperiodic 
Responsiveness in Hard Real-Time Environments," IEEE Real-Time Systems
Symposium, 1987.

   You can get your hands on copies of the real-time systems symposia by
contacting the IEEE Computer Society Press. 

   I don't know if you have seen reference to their idea before, but in case
you haven't, here's the basic idea. You have a real-time system, consisting 
primarily of hard, real-time, periodic tasks. However, you also have aperiodic
jobs that need to be serviced occasionally. 
   In order to fit aperiodicity into the rate-monotonic model, which assumes
that all tasks are periodic, create a periodic task called the Aperiodic 
Server. This server is just treated as a normal periodic task, that is assigned
a priority based upon the rate-monotonic property. When the Aperiodic Server
is invoked for its period, it will give its periodic time slice to one of the
waiting aperiodic tasks. Thus, periodically, aperiodic tasks get service. 
   Anyway, that's the basic idea, and it's kind of neat. It allows you to fit
aperiodic service into the commonly desirable model of an entirely periodic
task set.

   - Eugene Walden (walden@dip.eecs.umich.edu)

parks@bennett.berkeley.edu (Tom Parks) (11/13/90)

In article <1990Nov9.132404.5094@Matrix.COM>, srm@Matrix.COM (Steve
Morris) writes:
> I'm looking for a paper (Tech. Rept. CMU-CS-87-181) called
> 
>    Priority Inheritance Protocols: An Approach to Real-Time Synchronization
> 
> by L. Sha, R. Rajkumar, and J.P. Lehoczky. Is it obtainable via FTP?

Why not try this paper instead:

Priority Inheritance Protocols:  An Approach to Real-Time Synchronization
L. Sha, R. Rajkumar, and J. P. Lehoczky
IEEE Transactions on Computers, Vol. 39, No. 9, September 1990, pp. 1175-1185

> 
> Thank you.
> 

You're welcome.

Tom Parks
Electronics Research Laboratory
University of California, Berkeley