Re: Frustrating Random Reboots, seeking suggestions

[Hui Zhou <hzhou@hzsolution.net>]

On Sat, Jun 10, 2006 at 10:52:03AM +0200, Ingo Oeser wrote:
> Hi Hui Zhou,
>
> On Saturday, 10. June 2006 04:37, Hui Zhou wrote:
> > Thanks. memtest86 passes 6 times without errors. Serial console didn't 
> > show up anything (it just reboots). 
> >
> > Anyway, I finally suspect the debian libmpeg binary is at fault. I 
> > manually build it from src and statically linked to the `bkmark' 
> > program. It seems cured the random reboots problem. It runs 
> > successfully for 4 times. However, the fifth time it ended up in a `D' 
> > state. The only system call it uses is libc file IO and some signal 
> > passing. Any comment on the cause?
>
> Do you also see the problem if you decode from file to memory only.
> without any display?
>
> NO: You have some problem with your peripherals.

There is no display. The program just marks the blank scene or scene 
changes and dumps the results to a text file for another program to 
analyze.
> YES: Check for heat and power problems.
>
> 	If you are brave you could try some cpuburn variant to put the heat 
> 	to the maximum.
>
> 	WARNING: This could kill your CPU and might void your warranty, 
> 		since this is not "normal use" of your CPU :-)

No, I am not that brave. :) However, I am now faily certain it is not 
heat problem. After relinked with a new libmpeg binary, it hasn't 
rebooted yet (8+hours). Any possibility that some binary code can 
randomly trigger reboots on certain CPUs? (Sounds absurd, but only you 
kernel guys have better answers.)

Now I am concerned with the `D' state. Is that some problem from the 
kernel?

Thanks.

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