Re: 2.6.32: Attempted to kill the idle task?

[Christian Kujau]

On Wed, 1 Jun 2011 at 15:06, Connor H wrote:
> > we have a Debian/lenny system (Dell R410) with its "latest"
> > 2.6.32+27~bpo50+1 installed. From time to time the system tends to oops.
> > Can someone have a quick look at the following, if this makes sense at
> > all? I suspect the "real" oops message has scrolled away, but this is
> > all we got so far:
> >
> > ------------------------------------
> > [4330423.071859] RSP <ffff88012fffdcb8>
> > [4330423.071926] ---[ end trace b471cad78827bf8a ]--â
> > [4330423.071994] Kernel panic - not syncing: Attempted to kill the idle
> > task? [4330423.072068] rid: 0, comm: swapper Tainted: G D 2.6.32-5-amd64
> > 81 [4330423.072139] Call Trace:
> > [4330423.072206] [afffffff81216ee5>] ? panic+0x86/0x141
> > [4330423.072277] [afffffff812f6fee>1 ? printk+0x4e/0x58
> > [4330423.072349] [afffffff81041c07>] ? do_exit+0x72/0x6b5
> > [4330423.072419] [afffffff8104d011>] ? release_console_sem+0xl7e/0xlaf
> > [4330423.072493] [afffffff812f9c4d>1 ? oops_end+0xaf/0xb4
> > [4330423.072563] [afffffff81012734>] ? do_divide_error+0x85/0x81
> Here is the division fail
> kernel/sched_fair.c #3164
> 
> in find_busiest_group

Thanks for looking into it!

> sds.avg_load = (SCHED_POWER_SCALE * sds.total_load) / sds.total_pwr;
> 
> kernel/sched_fair.c #2853
> 	sds set from this call
> 	update_sd_lb_stats(sd, this_cpu, idle, cpus, balance, &sds);
> 
> 	which calculates total_pwr by looping all sched_group 's  in
> 	sds->total_pwr += sg->cpu_power;
> 	
> so hopefully Ingo or Peter can shed some light,
> 
> Connor

Christian.
-- 
BOFH excuse #365:

parallel processors running perpendicular today
--
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/