Andrew Dunstan <andrew(at)dunslane(dot)net> writes:
> We might be able to solve that for plpgsql, but in general we can't,
> ISTM. What if I write a plperl function that loops forever? We have no
> chance there to call CHECK_FOR_INTERRUPTS.
Yeah, I was thinking about that too. Still, we can/should/already did
fix it in plpgsql.
regards, tom lane