ECC errors causing panic on Sun Fire V210

[Marius Strobl]

On Sat, Jul 03, 2010 at 12:12:29PM -0400, John Floren wrote:
> Hi everyone.
> 
> I recently installed FreeBSD on my newly-acquired Sun Fire V210. In
> the three days I've had it up, it has spontaneously rebooted 4 times.
> The first 3 times I didn't catch any error messages on the serial
> console (I had disconnected my laptop), but this morning it did it
> again after I left my laptop hooked up. Here's what I saw:
> 
> panic: trap: corrected ecc error
> cpuid = 0
> [forgot to write this part, it was something like not being able to
> find the dump device]
> [starts rebooting in OBP]
> 
> It seems to me that the kernel should not be panicking when an ecc
> error is *corrected*. Have any of you come across this before? I
> attempted to search the archives but got an internal server error.
> 

CURRENT as well as 7-STABLE and 8-STABLE (8.1-RELEASE will be the
first release to include this) actually no longer should panic in
case of ECC errors. The problem was to find a machine exhibiting
ECC errors frequently enough to ensure the kernel does the right
thing in order to recover gracefully from them. Nevertheless
they are a sign of hardware problems.

Marius