Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!mips!bridge2!jarthur!elroy.jpl.nasa.gov!decwrl!shelby!neon!Gang-of-Four!weening
From: weening@Gang-of-Four.Stanford.EDU (Joe Weening)
Newsgroups: comp.sys.alliant
Subject: Re: non-zero ipl
Message-ID: <1990Mar7.013905.29368@Neon.Stanford.EDU>
Date: 7 Mar 90 01:39:05 GMT
References: <1990Mar6.222236.7098@uokmax.uucp>
Sender: root@Neon.Stanford.EDU (System PRIVILEGED Account)
Organization: Computer Science Department, Stanford University
Lines: 25
In-Reply-To: jay@alliant.backbone.uoknor.edu's message of 6 Mar 90 22:22:36 GMT

Here's my understanding based on conversations with Alliant C.S.
people: "ipl" refers to the interrupt level at which a processor is
running.  At various points, the kernel assumes it is at level 0; if
it checks and finds that it isn't, you get a panic message.

Note that there are several of these: "kernel non-zero ipl" and just
plain "non-zero ipl"; also "fbuserr" and "cbuserr" and maybe others.
Sometimes, it is OK to proceed through these problems.  To do this,
type <esc>P at the console.  However, this doesn't always work and you
may get strange events after doing so.

After we brought up 5.5, we got one of the non-zero ipl errors so
often that Alliant installed a patch to automatically proceed and not
type an error message.  This kept us running for a while, but then we
found that an application run by one of our users consistently caused
a different bug, that we could not proceed through.

Alliant told us that Concentrix 5.5.1 would soon be sent to people
having problems, so I debated whether to wait for it, but finally
decided to roll back to 5.0.  This was at the end of January, and we
haven't yet received 5.5.1, so I'm glad I decided to roll back.  We
don't really need any of the features that 5.5 supports.
--
Joe Weening                                Computer Science Dept.
weening@Gang-of-Four.Stanford.EDU          Stanford University