XL1200 Crashes into FEP

[Kaufman@RIVERSIDE.SCRC.Symbolics.COM (David H. Kaufman)]

    Date: Thu, 3 Jan 1991 11:29 EST
    From: ian@iexist.att.com

    I'm experiencing an XL1200 problem, where (consistently) from some spot
    in some code, the system will "crash"; going all the way through the
    FEP, just like someone cycled the power. No pause in the debugger or
    chance to poke around in the FEP itself -- the machine is gone -- even
    the screen crackles like a power reset. The problem is interesting
    because it becomes very hard to debug. As it so happens, the same code
    will oft times run on a 3600 type machine.

    It often seems to happen in code that may be (erroneously) using up the
    stack, but that sort of thing used to usually land you in the debugger
    for at least a look at where the problem manifested itself. It would
    appear that something is no longer as carefully wrapped as it is on the
    3600 machines.

    I'm curious, if this is a known problem (in which case maybe it can be
    characterized, so that I can better avoid or find it)? Has anyone else
    had this sort of problem on XL1200s (or other Ivory type machines)?

You don't mention what release of Genera you're running, but there was a
bug in 8.0 XL that fits youre description.  The bug could cause an
XL1200 to trap through the power-up firmware (which would cause your
screen crackling) and reload the FEP when Genera tried to use the
cold-load stream.  The bug was fixed in Genera 8.0ECO#1, which should
have been sent to all XL1200 customers with current software maintenance
contracts (among others).  Contact your sales rep if you didn't receive
Genera 8.0ECO#1.

By the way, you ought to be able to tell what Genera was doing just
before the crash by using the Debug command in the FEP.  See the section
Debugging on Ivory-based Machines in the Site Operations book.

    Ian Bruce