Re: index on function confuses drop table cascade on child

"Kevin Grittner" <Kevin.Grittner@wicourts.gov> writes:
> create index t_sname on t (sname(t));

Huh, interesting.  The reason the DROP misbehaves is that the index
doesn't have any dependency at all on table "t".  Which appears to
be exposing the folly of this bit in find_expr_references_walker:

        /*
         * A whole-row Var references no specific columns, so adds no new
         * dependency.
         */
        if (var->varattno == InvalidAttrNumber)
            return false;

This is broken at least as far back as 8.1.  Surprising no one's noticed
before.

            regards, tom lane