Re: [PATCH] Fix memory corruption in pg_shdepend.c

Michael Paquier <michael@paquier.xyz> writes:
> I was thinking on this one over the last couple of days, and doing a
> copy of shared deps from a template within createdb still feels
> natural, as of this patch:
> https://www.postgresql.org/message-id/YXDTl+PfSnqmbbkE@paquier.xyz
> Would somebody object to the addition of this test?  Or perhaps
> somebody has a better idea?

I agree that we're not testing that area well enough.  Proposed
patch seems basically OK, but I think the test needs to be stricter
about what the expected output looks like --- for instance, it
wouldn't complain if tab_foobar were described as something other
than a table.

            regards, tom lane