Re: [rtl] Release 9F Problems...

[Kurt Rosenfeld <hardware@ees1s0.engr.ccny.cuny.edu>]

Wow.  A picture is worth a thousand words.  I understand the blue trace 
but what is the pink trace in the bit-flipping example?  Are you using the 
make_periodic function or is your task a handler(responding to whtever 
generated the pink trace)?  Can you do an FFT or a histogram on the 
bit-flipping case?  Under different conditions, like running the same 
code under the regular Linux scheduler?  It would be so interesting!  I 
have always looked at the trace of the bit-flipping on my scope but could 
never see it in the frequency/PDF domain.  Thanks for posting your data. 
-kurt 

On Thu, 1 Oct 1998, Donald Gaffney wrote:

> 
> Hello RTLers,
> 
> I've been playing around with release 2.0.35-9F a little bit and my
> feeling is that it is not quite as stable as 2.0.33-0.6 that I did
> earlier development on. I've had my rt-linux apps lock-up and have
> had to resort to a few hard resets. Any 'smbmount'ed filesystems
> usually result in a lockup with a bad freelist error (block at 00000000),
> this is probably ok as I won't have those filesystems mounted in
> production, but I thought I should report it.
> 
> In other news I have a couple of scope traces available at 
> http://www.med.uvm.edu/~gaffney/scope.html
> 
> We just got this scope in and I've been playing with it a bit. The digital
> storage mode shows the jitter around the tasks pretty well, this is the
> same test setup I described earlier.
> 
> It seemed to me the new 9F version is a little more jittery as the blue
> storage traces appear a little thicker. Perhaps this is a poor read.
> It also seems like the system overshoots its 40 us time period fairly
> often, do I need to reset the periodic timer value in the source? 
> 
> Any comments or suggestions for further tests would be appreciated. I
> would like to purchase new hardware shortly and would like to use these
> sorts of data to assess what is required. I'm trying to stuff a lot into a
> 50 us window (2 ADCs, PID calcs, DIO).
> 
> Thanks again for RT-Linux!
> 
> -Don
> 
> 
> Error Message Details:
> 
> On removing an RT-Linux 9F periodic task (with the task running) after
> a front-end lock-up (2.0.33-0.6 never locked the front end, I know
> that 'rmmod'ing a running periodic task is a little silly, but I
> seemed to lose fifo communication with the rt-task and was trying to
> recover):
> 
> (memory addresses, execution stack, etc. - sorry, didn't keep it)
> Aiee, Killing interrupt handler
> 
> Occaisional lockups with 'smbmount'ed filesystems:
> 
> Problem: block on freelist at 00000000 isn't free.
> 
> --- [rtl] ---
> To unsubscribe:
> echo "unsubscribe rtl" | mail majordomo@rtlinux.cs.nmt.edu OR
> echo "unsubscribe rtl <Your_email>" | mail majordomo@rtlinux.cs.nmt.edu
> ----
> For more information on Real-Time Linux see:
> http://www.rtlinux.org/~rtlinux/
> 
--- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail majordomo@rtlinux.cs.nmt.edu OR
echo "unsubscribe rtl <Your_email>" | mail majordomo@rtlinux.cs.nmt.edu
----
For more information on Real-Time Linux see:
http://www.rtlinux.org/~rtlinux/