Re: AtAbort_Portsl problem

> Tatsuo Ishii <ishii@postgresql.org> writes:
> > While ago, I reported a problem regarding exec_execute_message crash
> > in transaction abort state if sync message issued right after parse,
> > bind and execute message (which is normal if used with pgpool). After
> > further investigation, I concluded that there's a serious problem with
> > unnamed portal handling.
> 
> I was never able to get your test case to work (or, rather, crash).
> Does this patch address the original issue, or is this something
> different?

The patch addresses the original issue.  The reason why you didn't see
crash was just you were lucky, I believe. I'm sure that your
exec_execute_message looks into already-freed-memory.
--
Tatsuo Ishii
SRA OSS, Inc. Japan