Re: [PERFORM] Foreign key performance

Tom Lane wrote:
> Stephan Szabo <sszabo(at)megazone23(dot)bigpanda(dot)com> writes:
> > Okay, this changes the second paragraph of that comment. I left in the
> > comment that's really similar next to where I actually do the selection of
> > which start point to use.
>
> This had a bit of a problem yet: the loop in deferredTriggerInvokeEvents
> expects 'prev_event' to point to the list entry just before 'event'.
> A nice byproduct of fixing that is we don't uselessly rescan the last list
> entry. I also tried to improve the comments a little. You can see what
> I actually applied at
>
> http://developer.postgresql.org/cvsweb.cgi/pgsql-server/src/backend/commands/trigger.c

Any chance of backporting these changes to 7_3_STABLE (when you're
satisfied they don't break anything)? Just looking at the CVS log for
trigger.c, it appears there have been enough changes since then that
it might not be easy to do (and since it's not necessarily a "bug fix"
as such, it might not qualify for backporting to a stable version).

Even if it's not something that can be put into another release of
7.3, it would certainly be useful to me. It might be useful to enough
people to justify releasing it as a patch on -patches, if nothing
else.

I'd do it myself but I don't understand the trigger code at all (and
if there's any documentation you can point me to that describes the
various functions and supporting data structures in trigger.c, that
would help a lot), and I'd rather not touch something like that until
I understand it thoroughly.

--
Kevin Brown kevin(at)sysexperts(dot)com