Port-amd64 archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: spurious reboots



On Wed, Aug 26, 2009 at 08:34:37PM +0200, Bernd Ernesti wrote:
> On Wed, Aug 26, 2009 at 08:26:30PM +0200, Manuel Bouyer wrote:
> > On Wed, Aug 26, 2009 at 09:57:06AM +0000, Emmanuel Dreyfus wrote:
> > > On Wed, Aug 26, 2009 at 10:33:20AM +0300, AR wrote:
> > > > Maybe you could try swapping the RAM chips in their sockets or try a  
> > > > memtest. It is likely that your problem is hardware related. If not, 
> > > > try 
> > > > a live cd with say, Free/OpenBSD and compile stuff there, see if your  
> > > > problems persist.
> > > 
> > > It wokred fine with NetBSD 4, si I have trouble to buy the hardware
> > > issue.
> > 
> > It could be, because NetBSD 5 stresses hardware in a much different way.
> > It makes heavy use of locked instructions for example. I know some AMD
> > processors have bugs in this area.
> 
> Do you have more information about that?
> 
> I may be affected by it too, because the system panics and memtest showed
> no problem. Which means I would have to do a downgrade to get it in a
> working state again.

With a current kernel, when the CPU attaches you should see a note about
CPU erratas presents (maybe needs boot -v). I think netbsd-5 also has
this code but I'm not sure.
There are workaround for some of them in the NetBSD kernel, but I
don't remember if they're only in HEAD or in netbsd-5 too.

-- 
Manuel Bouyer <bouyer%antioche.eu.org@localhost>
     NetBSD: 26 ans d'experience feront toujours la difference
--


Home | Main Index | Thread Index | Old Index