Re: Rethinking opclass member checks and dependency strength

Hamid Akhtar <hamid(dot)akhtar(at)gmail(dot)com> writes:
> I've gone through the patch and applied on the master branch, other than a few hunks, and whether as suggested upthread, the default case for "switch (op->number)" should throw an error or not, I found that bloom regression is crashing.
> -------------
> test bloom ... FAILED (test process exited with exit code 2) 20 ms

Hmm ... I think you must have done something wrong. For me,
am-check-members-callback-5.patch still applies cleanly (just a few
small offsets), and it passes that test as well as the rest of
check-world. The cfbot agrees [1].

Maybe you didn't "make clean" before rebuilding?

regards, tom lane

[1] https://travis-ci.org/github/postgresql-cfbot/postgresql/builds/712599990