The problem is context switch time. With Hard real time systems it is <10
microseconds. I am trying to find out context switch times in Linux also.
There will be several processors and DSP's many pegs many holes.
From: owner-ucsimm@uClinux.com [mailto:owner-ucsimm@uClinux.com]On
Behalf Of Tom Walsh
Sent: Wednesday, February 23, 2000 2:58 PM
Subject: Re: [uCsimm] Hard Real time Linux
Jeff Bauknecht wrote:
My company has been looking at different operating systems for our
product. We are probably going to have a PowerPC processor (860).
product is going to be a PBX. The real time extensions to Linux are
better. We have one problem that is keeping Linux from being our
That is a conference bridge. A conference bridge takes several
sources sums them together, averages them and sends them back out.
processor intensive. It has to be hard real time. LynxOS has put
Bluecat, but it is only soft real time. This might be to encourage
to use their LynxOS for the hard real time applications. Is anyone
of a solution for this? I.E Hard real time extensions for Linux?
I have been pushing Linux here at work and need help.
Not all solutions have to be a 'single chip' one, nor can they be,
it sounds to me that your solution is that of a multiple processor one: a
processor that passes data around on the IP network and the other processor
to do the heavy work of summing the voice data.
Round pegs + square holes type of problem, like trying to do DSP
with a PIC controller.
This message resent by the email@example.com list server
-- Tom Walsh - WN3L - Embedded Systems Consultant - tom over_at mytoys(.)com 'www.mytoys.com', 'www.openhardware.net', 'www.cyberiansoftware.com' "Windows? No thanks, I have work to do..."
This message resent by the firstname.lastname@example.org list server http://www.uClinux.com/
This archive was generated by hypermail 2b30 : Sun Apr 07 2002 - 00:01:34 EST