Robert Haas <robertmhaas@gmail.com> writes:
> I did not back-patch, because the code is in a different file in v11,
> none of the hunks of the patch apply on v11, and v11 is not failing on
> hyrax.
Hmm, I wonder why not. I suppose the answer is that
the leak is worse in HEAD than before, but how come?
I followed your reference to 898e5e329, and I've got to say that
the hunk it adds in relcache.c looks fishy as can be. The argument
that the rd_pdcxt "will get cleaned up eventually" reads to me like
"this leaks memory like a sieve", especially in the repeated-rebuild
scenario which is exactly what CLOBBER_CACHE_ALWAYS would provoke.
Probably the only thing that keeps it from being effectively a
session-lifespan leak is that CCA will generally result in relcache
entries being flushed entirely as soon as their refcount goes to 0.
Still, because of that, I wouldn't think it'd move the needle very
much on a CCA animal; so my guess is that there's something else.
regards, tom lane