On Fri, 15 Aug 2025 at 17:46, Hayato Kuroda (Fujitsu)
<kuroda.hayato@fujitsu.com> wrote:
> Just in case: actually, even if the HASH_DEBUG part is fixed on PG17, it could
> not pass some tests. One example is initdb test [1].
> ISTM, command_like() assumed that there are no outputs in stderr but this option
> does. This meant no BF animals cannot set this option as-is.
> After I changed them to elog(DEBUG1) (and set debug1 as default) I ran tests
> under src/test, and they could pass.
Yeah, I noticed the tests wouldn't pass with it enabled. I think now
that the format bug is fixed in the back branches, we can decide what
we're going to do in master.
I think if we're going to keep HASH_DEBUG, then we should try and do
something to help ensure it does not get broken again for an excessive
period of time. I'm not against switching to elog(DEBUG), but then I
don't have any idea why it was outputting to stderr. If someone wants
to switch their BF animal to define HASH_DEBUG then maybe we can
commit the elog(DEBUG) change to master only and see if anyone
complains. (I suspect nobody will).
I think the votes to keep it should outweigh the votes to get rid of
it, as someone voting to keep it probably has some idea that it'll be
useful to them, and keeping it shouldn't really hinder the people who
don't want it. I think Michael has voted to keep it.
Michael, any thoughts on switching from stderr to elog(DEBUG)? And
also about defining HASH_DEBUG on one of your BF members?
David
