> I do not see a defense against "rogue" pointers unless there is a way
> to use the MMU.
VxWorks has MMU support as an option. It may be useful especially during
debugging.
My understanding is that the RTL "solution" to the problem is to keep RT
activities extremely lean and simple and push as much processing as
possible up to the user area, where you have good resource protection
and debugging facilities.
Almost all rogue pointers have to do with dynamically allocated data
structures or hazardous programming practices (like making a variable
point to a local variable). I think your best defense is to avoid doing
those things inside the sensitive kernel.
In our experience a lot can go wrong in a complex real-time system, and
it's a pain to troubleshoot it. But rogue pointers haven't turned out to
be a significant problem. That's why reducing complexity in the
real-time system will eliminate more problems than MMU support. (MMU
support might actually entice you to make your system more complex.)
Marko
--
Marko Rauhamaa [EMAIL PROTECTED] (818) 878-6314
Sr Project Engineer http://www.tekelec.com/ Tekelec Inc
--- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
----
For more information on Real-Time Linux see:
http://www.rtlinux.org/~rtlinux/
