Re: SMP lockup

Hans Lermen (
Tue, 4 Mar 1997 02:17:40 +0100 (MET)

On 3 Mar 1997, Linus Torvalds wrote:

> In article <>,
> Richard B. Johnson <> wrote:
> >
> >On my Dual Pentium 166 MHz machine running SMP, I have discovered a
> >repeatable problem.
> > [ ... ]

It is possible to make it just STOP! No errors
> >no obvious problems. After it stops, there is no response to the keyboard,
> >nothing at all. The CPUs __seem__ to be halted with the interrupts off.
> >This is evident by the chip temperature rapidly returning to room temperature.
> >There is no bus activity except for an approximate 26 us CAS for refresh.
> Heh. You have hardware I haven't got access to, but I've seen the same
> thing now that I have access to a SMP box.

Here is one additional report:

I have _exactly_ the same /proc/cpuinfo output as Richard, and I also
can relyable reproduce a SMP bug, but instead of locking it _reboots_
without saying anything, no oops, no info left in the logs :(
It happens when the machine is under havy load ( 2 simultaneous kernel
compile with 'make -j3' do the job).

I had this problem with >2.0.18 as well as with 2.1.27 . The same kernels
compiled for uniprocessor run rock solid (uptime several days).
As I never got the chance to report some valuable information, I'm near to
the point giving up, but if I can help doing some special tests, just
tell me what you want.
( ... frustating, having a SMP box and using only one processor)