Table corrupt?

Hey, guys...
   I suppose this probably belongs in questions, but I thought I might
get a quicker answer here.  I have a table in a customer's database
that has gotten quite large via lack of vacuuming (57MB).  Queries
involving this table started failing in the form of the backend just
sitting there chugging away on the CPU (but not disk so much) for
hours on end.
   This began about 24 hours ago, and as of about 12 hours ago, no
queries on this table work.  I started a vacuum about 3 hours ago, and
it has had upper-90s percent CPU usage the whole time, and still
hasn't completed.
   Any ideas on what might be going on here?  And, if postgres won't
be able to access the table, is there any hope of extracting rows from
the raw database file, such that I could reconstruct the table?
   Please cc responses to me directly, as I'm only on the digest list,
and thanks in advance for any advice/help!

-Brandon :)