On Wed, Dec 26, 2012 at 6:33 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Yeah, that destroys my theory that there's something broken about index
> management specifically. Now we're looking for something that can
> affect any buffer's refcount, which more than likely means it has
> nothing to do with the buffer's contents ...
Hardware problem is still a plausible candidate. Have you run any
memory checker software or seen anything else crash? Classically gcc
is as good at detecting memory problems as memory checking software.
Or a bad cpu can also sometimes cause problems like this. Have you
been able to reproduce on any other machines?
Did you ever say what kind of hardware it was? This is the local
reference count so I can't see how it could be a race condition or
anything like that but it sure smells a bit like one.
--
greg