Marti Raudsepp <marti@juffo.org> writes:
> On Sun, Jun 8, 2014 at 3:36 PM, David Rowley <dgrowleyml@gmail.com> wrote:
>> Currently pull_up_sublinks_qual_recurse only changes the plan for NOT EXISTS
>> queries and leaves NOT IN alone. The reason for this is because the values
>> returned by a subquery in the IN clause could have NULLs.
> I believe the reason why this hasn't been done yet, is that the plan
> becomes invalid when another backend modifies the nullability of the
> column. To get it to replan, you'd have to introduce a dependency on
> the "NOT NULL" constraint, but it's impossible for now because there's
> no pg_constraint entry for NOT NULLs.
I don't believe this is an issue, because we are only talking about a
*plan* depending on the NOT NULL condition. ALTER TABLE DROP NOT NULL
would result in a relcache inval event against the table, which would
result in invalidating all cached plans mentioning the table.
I forget exactly what context we were discussing needing a NOT NULL
constraint's OID for, but it would have to be something where the
dependency was longer-lived than a plan; perhaps semantics of a view?
The existing comparable case is that a view containing ungrouped
variable references is allowed if the GROUP BY includes a primary key,
which means the semantic validity of the view depends on the pkey.
regards, tom lane
