"Eric B. Ridge" <ebr@tcdi.com> writes:
> Wow, thanks for spending the time on this. What about for gettuple?
> Do calls to it take advantage of the cache? If not, this likely
> explains some of my custom am's performance troubles.
gettuple is looked up once at the start of a scan, so there's no
per-tuple overhead involved there. As I said before, we're usually
pretty good about avoiding per-tuple overheads --- the cost you
identified here is a per-query overhead.
> If there's anything I can do to help, let me know. I'll be happy to
> test any patches you might come up with too.
I have committed a patch into CVS HEAD --- give it a try.
regards, tom lane