Re: Dumb question: BKL on reboot ?

[Andrew Morton]

Hannes Reinecke <Hannes.Reinecke@xxxxxxx> wrote:
>
> Hiya,
> 
> I've got a dumb question: Why is the BKL held on entering sys_reboot() 
> in kernel/sys.c:405 ?

Probably for no good reason.

> It is getting especially interesting on SMP, when one cpu is entering 
> sys_reboot, acquires the BKL and then waits (via machine_restart) for 
> all other cpus to shut down. If any of the other cpus is executing a 
> task which also needs the BKL, we have a nice deadlock.
> We've seen this here on 2-way s390, where the other cpu tried to execute 
> kupdated() (what did it try that for? Anyway...), which of course 
> resulted in a deadlock.

I guess that dropping the BKL around the machine_restart() call would be an
appropriate 2.4 fix.

Where exactly does the rebooting CPU get stuck in machine_restart()?  If
someone has done lock_kernel() with local interrupts disabled then yes,
it'll deadlock.  But that's unlikely?  Confused.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/