Dan Ports <drkp(at)csail(dot)mit(dot)edu> wrote:
> Note that this hadn't been a reasonable option until last week
> when we added the check for non-MVCC snapshots, since there are
> lots of things that use heap scans but SeqScan is the only
> (currently-existing) one we want to lock.
That is the sort of thing that I tended to notice going through the
backtraces from heap access I mentioned in another post, and is most
likely the reason the call landed where it did. The MVCC snapshot
tests are then a game-changer. It would be nice to find a way not
to acquire the relation lock if the node is never used, though.
> I am rather uneasy about making changes here unless we can be
> absolutely certain they're right...
Yeah....
-Kevin