Andres Freund <andres@anarazel.de> writes:
> On 2016-10-02 17:59:47 -0400, Tom Lane wrote:
>> So now I'm thinking you're right, it'd be better to have some solution
>> whereby dfmgr.c knows about cross-module dependencies and loads the
>> dependencies first. Not sure how to approach that. The extension
>> "requires" mechanism is tantalizingly close to providing the data
>> we need, but dfmgr.c doesn't know about that, and there's no concept
>> of a reverse mapping from .so names to extensions anyway.
> One, kind of extreme, way to get there would be to resolve the hstore
> symbols hstore_plpython needs with load_external_function, during
> _PG_init(). Most of these files don't actually depend on a large number
> of symbols, so that should actually be doable.
Hm. That would actually not be a bad idea, perhaps, because the method
we're using right now requires that the linker not bitch about unresolved
symbols at build time, which is a really bad thing that I'd prefer to
turn off.
It's still not a very back-patchable answer, but it's something that
we could get to in HEAD without a huge amount of work.
regards, tom lane