On Sat, Jan 8, 2022 at 1:00 PM Alexander Lakhin <exclusion@gmail.com> wrote:
> 07.01.2022 09:15, Etsuro Fujita wrote:
> > On Mon, Jan 3, 2022 at 4:00 AM Alexander Lakhin <exclusion@gmail.com> wrote:
> >> propose a simple test case for the issue. Maybe you will find it useful.
> > Thanks, but in my environment, the test case doesn’t cause any
> > failure.
> It's interesting. I perform on REL_14_STABLE (61c8da50) the following:
> wget
> https://www.postgresql.org/message-id/attachment/129455/posgres_fdw.sql.patch
> -O - | git apply
> ./configure --enable-debug --enable-cassert >/dev/null && make -j8
> >/dev/null && make -j8 -C contrib >/dev/null && make check -C
> contrib/postgres_fdw/
> and get:
> test postgres_fdw ... FAILED (test process exited with
> exit code 2) 1116 ms
Reproduced. Sorry for that. Maybe, I tested this on a
non-assertion-enabled build by mistake. :-(
> > How about something like the attached, which is made by
> > modifying the original test case to avoid the connection-limit-error
> > issue. In the attached, I also added a comment to a function to match
> > other places.
> Your modified test fails with unpatched postgres_fdw.c too (as
> expected). I would prefer your test as more elaborated.
Ok, but I modified the added test case a bit further so that foreign
tables are well-defined like in your second test case, which I think
is good. Attached is an updated version of the patch.
Thanks!
Best regards,
Etsuro Fujita