The Hermit Hacker <scrappy@hub.org> writes:
> On Thu, 7 Sep 2000, Tom Lane wrote:
>> I have committed a fix into REL7_0 branch. Although it seems to work,
>> I don't trust it really far because it depends on heap_markpos() and
>> heap_restrpos(), which haven't been used in a long time and are full
>> of alarmed-sounding comments.
> Do you have any thoughts as to what sorts of problems *might*
> arise? Like, are we talking database corruption possibilities, or bad
> results, or ... ? Just want to have an idea of what to try and keep an
> eye out for ...
I may be overstating the cause for worry. All of the "alarmed-sounding
comments" appear to date back to the original Postgres95 sources, and
are probably obsolete. The only thing I really have any concern about
is whether buffer pin/unpin bookkeeping is correct. If it's not,
you'd see an Assert failure from too many unpins (you are running with
--enable-cassert I hope) or "Buffer Leak" notices in the log from too
many pins.
regards, tom lane