Re: Unix runs faster, maybe

In article <Ag30HWowCH6L@xxxxxxxxxxxxxxxxxxxxxxxx>,
koehler@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx (Bob Koehler) writes:
In article <4e8b6hF1ahl9iU3@xxxxxxxxxxxxxx>, bill@xxxxxxxxxxx (Bill Gunshannon) writes:

Boy would I like to see a list of those needs!! :-)

A kernel which can do hard real time is at the top of the list.

OK, I learned something else. I didn't think VMS was hard realtime.


Linus originally wanted the kernel (which he controls) to remain

Well, he controls the msainstream Linux kernel. That doesn't stop the
realtime guys from doing their own.

simple (the same design criteria was selected for the original UNIX
kernel).

True, I know AT&T did a realtime kernel (used to run the phone system
at one time, that's why they created the WE32000 family and the real
3B line, not to be confused witht he 3B1) but I don't know that they
ever let it out of house.

You need a fairly complex kernel to do both timesharing and
real time effectively.

Linus felt that the issues brought up by real time should be handled
by the hardware.

Yeah, proving once again that Andy was right.

Now he seems to realise that there are processors
in that "hardware" and software in those processors. Real time Linux
vendors have contributed to the Linux kernel and Linus has accepted
those changes.

I don't ever expect real time Linux to keep up with pure real time OS
like VxWorks, but if a real effort is made it could keep up with the
VMS kernel. That, however, would take a bigger change than I think
most vendors are willing to donate.

I don't htink Linux will ever catch up with or keep up with any serious
kernel. Too much baggage.


Meanwhile VxWorks is becoming complex enough to offer some of the
features not traditionally found in pure real time OS (like separate
memory maps for every task). Those of us who use it are not running
away from it.

But, can I take it you are also using VMS for realtime tasks? Another
facet of VMS I was unfamiliar with. Where would I read about doing
realtime programming on VMS? We have a professor who teaches a course
in realtime here and he might be convince to at least touch on VMS
realtime.

bill

--
Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves
bill@xxxxxxxxxxxxxxx | and a sheep voting on what's for dinner.
University of Scranton |
Scranton, Pennsylvania | #include <std.disclaimer.h>
.

Relevant Pages

  • Re: RT patch acceptance
    ... > Bill, you can belittle me to your heart's content. ... you mean pervasively throughout the kernel for every system, ... should run within an interrupt plus a thread to enable if assuming ... practicality we're very close to hard real time. ...
    (Linux-Kernel)
  • Re: Newbie Question About Device Drivers
    ... I guess we are trying to achieve a more "Real Time" operating system. ... in the kernel you have access to the ... >if you can achieve what you want in user space, it is probably better to do ...
    (microsoft.public.development.device.drivers)
  • JAD and AppArmor problem
    ... during boot up that I can not modify, add or delete any AppArmor profiles. ... Has anybody booted to the rt kernel and got AppArmor ... jackd runs with real time capabilities enabled. ...
    (alt.os.linux.suse)
  • Re: Unix runs faster, maybe
    ... A kernel which can do hard real time is at the top of the list. ... Linus felt that the issues brought up by real time should be handled ... I don't ever expect real time Linux to keep up with pure real time OS ...
    (comp.os.vms)
  • Re: Unix runs faster, maybe
    ... Well it's my understanding that one of them is near real time data ... times in the past that Unix isn't up to his needs. ... Well, I'm probably wrong again, but I wouldn't think that VMS was hard ... been modified to do realtime, but I don;t think they are hard realtime ...
    (comp.os.vms)