Experimental Physics and
Industrial Control System

Maren Purves <[email protected]> · Wed, 07 Feb 2001 16:33:36 -1000

This rings a bell. 

I made some stupid programming mistake that under certain circumstances
caused the CPU to be 100% busy. The console serial line was the only 
way in to find out (dm windows stayed blank/white because there was 
no CPU time left. It appeared dead).
As far as I remember this was on vxWorks 5.3.1 and some earlier
EPICS 3.13 though.

Maren

Jeff Hill wrote:
> 
> Don't assume that the IOC is 100% dead before you connect
> a terminal to the serial port and try to get console
> access w/o depending on the network. Sometimes the IP kernel
> will be dead but vxWorks is still running. Console serial
> access via a terminal can prove to be invaluable when
> debugging vxWorks problems.
> 
> > -----Original Message-----
> > From: [email protected]
> >
> > We are having run time problems running EPICS on the new Motorola MV162
> > P242E cards. I am running VxWorks 5.2, and have tried EPICS R3.12.2 and
> > R3.13.3. In all cases the CPU freezes or locks up, usually within 24
> > hours. We think it crashes more quickly if sequencers are running.
> >
> > Has anyone else had problems running EPICS on these cards?
> >
> > I'm in the process of seeing if we require a BSP upgrade from Wind River
> > (we may have to upgrade to tornado in order to get the latest BSPs)

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System