Multithreading

Multithreading

Post by Douglas Beattie Jr » Mon, 01 Apr 1996 04:00:00



Does the use of multithreading conflict with the methodology
of an RTOS?  How so?

--
Douglas Beattie Jr.      http://www.whidbey.net/~beattidp

 
 
 

Multithreading

Post by Harry Franke » Mon, 01 Apr 1996 04:00:00



> Does the use of multithreading conflict with the methodology
> of an RTOS?  How so?

> --
> Douglas Beattie Jr.      http://www.whidbey.net/~beattidp

I don't understand you're question, please explain

Harry

 
 
 

Multithreading

Post by Peter Herma » Tue, 02 Apr 1996 04:00:00



: Does the use of multithreading conflict with the methodology
: of an RTOS?  How so?

A good Real Time Operating System must be eager to use
multithreading and multiple CPUs (central processing units)
for truly concurrent processing.
In advanced systems the applications are supported by
the programming language (e.g. Ada) to easily program
concurrency and the resulting conflict resolution
when accessing shared memory.
The programming language works like the hand in the glove
of the OS. These much more comfortable concepts of the
future are realized today with a free compiler from GNU,
the GNAT (GNU Ada Translator) from ftp cs.nyu.edu .
I easily ran multiprocessing+multiaccessing on both
SGI Iris and Sun Sparc, both with 2 processors.
More than 2 processors are tested by colleagues of mine.

BTW, WindowsNT on DecAlphas or on multiprocessor-PCs
are readily willing to work this way. It's just great.

: --
: Douglas Beattie Jr.      http://www.whidbey.net/~beattidp

--

Pfaffenwaldring 27, 70569 Stuttgart Uni Computeranwendungen
Team Ada: "C'mon people let the world begin" (Paul McCartney)

 
 
 

Multithreading

Post by Alex Colv » Tue, 02 Apr 1996 04:00:00


Quote:>Does the use of multithreading conflict with the methodology
>of an RTOS?  How so?

no.
--
        Alex Colvin

 
 
 

Multithreading

Post by Christian Jakobsso » Wed, 03 Apr 1996 04:00:00


Win NT supports at most 4 processors, any good RTOS should be able to
handle any number of processors, from 1 to (**insert largest imaginable
number here**), not just 4.
A system should have no other limits than it's hardware.

--
In RL: Christian Jakobsson
Snail: Tybblegatana 12, S-702 19  OREBRO
Voice: +46 (0)19 33 54 83


W++>+++$

t+>+$


 
 
 

Multithreading

Post by David A. Pau » Wed, 03 Apr 1996 04:00:00



> Does the use of multithreading conflict with the methodology
> of an RTOS?  How so?

> --
> Douglas Beattie Jr.      http://www.whidbey.net/~beattidp

If anything, multithreading brings protected mode OS's closer to
becoming a RTOS.. but only if the OS can support priority based
scheduling. I'm not saying that a full blown Unixware system will
perform like a VRTX-based embedded control, but threads offer the
designer the capability to build a responsive system using a protected
mode operating system. Check out Lynx OS which supports threads.

dap

------------------------------------------------------------------------------

David A. Paul                                   phone:    (309)578-8633
Research Engineer                               fax:      (309)578-4572/3605

Caterpillar Inc. USA

------------------------------------------------------------------------------

 
 
 

Multithreading

Post by DGrana » Sat, 06 Apr 1996 04:00:00


I've built space vehicle software for systems with three main processors
and
up to a dozen subsystem processors.  All of them have been by convetional
definition multithreaded. The trick is not are they multithreaded but how
many
threads do you actually need (smaller and simpler the better) and how do
they
interact with one another.

Dave Granath