Neil Conway <neilc@samurai.com> writes:
> Someone contacted me in IRC about a bug in PL/PgSQL. I've confirmed that
> the example SQL they sent me causes a segfault on my machine (CVS HEAD),
> but I've so far not had a lot of success tracking down the exact cause
> of the problem.
The core dump appears to occur when the SELECT INTO fails to retrieve
a row, leaving "finalrec" undefined. Clearly, RETURN NEXT with an
undefined record variable shouldn't dump core, but what should it do?
Raise an error, or perhaps be a no-op?
regards, tom lane