SUMMARY: console says "no timer interrupts on CPU 0"

From: Peter Marquardt <wwwutz_at_MPIMG-Berlin-Dahlem.MPG.DE>
Date: Wed, 18 Mar 1998 09:16:32 +0100 (MET)

Concerning the console-message "*** no timer interrupts on CPU 0 ***"
I got two replies. The short summary would be:

        - "It _may_ be a problem."
        - "Replace cache memory."

The practial solution was:

        - "Replace the motherboard."

This is what our Digital engineers did. The message disappeared,
neither the system crashed the last 3 days nor the harddisk
created cam-errors so I would call the system "fixed". The motherboard
is now on its way to Digital engineering to find out more.

It might have been enough to just replace the cache-memory, as Hartmut Duecker
replied, but we did it the hard way [and expensive one, so Digital should
really think about this console message 8-)].

As usual Dr. Tom Blinn was very helpful, and I wanted to thank him
for his efforts to help us all. I think he should at least get
a grammy or oscar for his work 8-) (the pope should canonize him). ;-)


    =============== original question ===================

Hi,

does anyone know what this message on console boot means:

*** no timer interrupts on CPU 0 ***

This message appeared on several 433au, 600au, pciaxp33 we have here.
Sometimes the message just disappears after a while, but we also had
whole motherboards replaced by digital-service to fix this.

Since it's always a hassle to make up a call and to have a motherboard replaced
I would like to know what this message means and how I could get around with
it. We have had systems which had this message and which suddenly crashed
without leaving a dump anywhere, so I don't think it's ignorable.

        =============== answers ===================

Date: Wed, 04 Mar 98 10:34:43 -0500
From: "Dr. Tom Blinn, 603-884-0646" <tpb_at_zk3.dec.com>
To: Peter Marquardt <wwwutz_at_mpimg-berlin-dahlem.mpg.de>
Subject: Origin of "no timer interrupts on CPU 0" messages

Peter,

I checked with a colleague who works with things like the SROM, PALcode, and
the SRM console firmware on a number of our Alpha systems. He says this
about the question you asked:


------- Forwarded Message

This message is generated by a console routine that checks to
see that timer interrupts are happening at (more or less)
regular intervals. This error message is only printed once,
and only during a system power up.

The message may be indicative of a real problem with clock
interrupts on the system, or just a transient occurrence
during the power up sequence. I guess the real test would
be to boot the OS and see if it complains about the clock.

------- End of Forwarded Message

So, it's coming from console firmware, and it's more a sanity check than a
definitive "there is something really wrong with your system". But if it's
happening a lot, and the system software (operating system) is failing with
no trace of a cause (just stops cold, no messages coming out even if you've
got a serial console hooked up and producing a reliable log), then there may
be a real hardware problem that needs to be repaired. But that message, in
and of itself, isn't proof of broken hardware, just that a "sanity check" in
the firmware seemed to fail.

=======================================================================
Date: Mon, 9 Mar 1998 10:34:45 +0100
From: Hartmut Duecker <i3040104_at_thcpch.nat.tu-bs.de>
To: marquardt_p_at_mpimg-berlin-dahlem.mpg.de
Subject: Re: no timer interrupts on CPU 0

Hallo Herr Marquardt,

dasselbe Problem beobachte ich auf einem neu beschafften DEC Workstation
Cluster
mit einigen 500au- und Alpha-Server-500/500-Rechnern. Eine Workstation stuerzte
in der von Ihnen beschriebenen Form ab, was erst dadurch behoben wurde, dass
ein Digital-Techniker den Cache ausbaute bzw. ersetzte. Die Fehlermeldung
tritt sporadisch beim Booten oder Systemshutdown auf, ohne dass ich wuesste,
was sie zu besagen haette.

=======================================================================

        Peter
--
##===========================##============================================##
Dipl.Ing. Peter Marquardt/// B.O.F.H. Max-Planck-Institute Molecular Genetics
|| +49-30-8413-1430     ///  ||   --==> Earth-Germany-Berlin-Dahlem <==--  ||
::        fax -1773 \\\///   :: marquardt_p_at_mpimg-berlin-dahlem.mpg.de     ::
..================== *The Turtle Moves* ===================================..
Received on Wed Mar 18 1998 - 12:25:17 NZST

This archive was generated by hypermail 2.4.0 : Wed Nov 08 2023 - 11:53:37 NZDT