Andres Freund <andres(at)anarazel(dot)de> writes:
> We push down expressions below a sort node, even though they could be
> evaluated above.
Yeah. That's a hangover from an ancient decision that sort/limit
would always be applied at the top of the plan tree. I'm too
lazy to check the details right now, but I think we already relaxed
that in some cases (or maybe it was about evaluating stuff
before/after GROUP BY?).
> That can very substantially increase the space needed for the
> sort.
Could decrease it too, eg if what you are outputting is
md5(some-wide-column). Not sure we are smart enough to tell
which way is better.
regards, tom lane