> At 08:03 AM 10-11-99 -0700, [EMAIL PROTECTED] wrote:
> >
> >This would reduce jitter at the expense of making timings always worst
> >case. I think it would be most useful for testing worst case timing.
> >
> >
>
> What if RTLinux was cachable and everything else uncachable (user space and
> maybe the Linux kernel)?
>
I think that turning it arround will not improve jitter because you will not
get the kernel and your app+shm into the cache (assuming you don't have 4MB
cache...) and so you still would have jitter due to cache-misses .
a solution maby would be to setup a shm section that is EXACTLY the size of
your cache and then only enable cache for that memory area , that would though
slow down the kernel substatialy .So I don't know if you could realy winn with
this strategy .
nmg
--- [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/
