PANIC during VACUUM

Hi,

I am testing version 9.1.9 before putting it in production. One of my tests
involved deleting a the contents of a big table ( ~ 13 GB size) and then
VACUUMing it. During VACUUM PANICS. Here is the message:

PANIC: corrupted item pointer: offset = 8128, size = 80

I found the error a couple of times, allways during VACUUM after deleting
the context of the same big table (after re-polpulating it of course).

The error message is always *exactly* the same i.e. the same offset and
size.

When this happens the backend gets restarted and if I issue the same VACUUM
command, I get the same error.

I also tried triggering the backup server (hot-standby with streaming
replication, and trying the VACUUM there (to see if it may be a hardware
problem in the primary) and got the same issue.

What might be causing this? Should I reported as a bug? Thanks!

--
Germán