Re: COPY FREEZE and setting PD_ALL_VISIBLE/visibility map bits

On 1/23/21 1:10 AM, Tom Lane wrote:
> Pavan Deolasee <pavan(dot)deolasee(at)gmail(dot)com> writes:
>> On Mon, Jan 18, 2021 at 3:02 AM Tomas Vondra <tomas(dot)vondra(at)enterprisedb(dot)com>
>> wrote:
>>> Pushed. Thanks everyone for the effort put into this patch. The first
>>> version was sent in 2015, so it took quite a bit of time.
>
>> Thanks Tomas, Anastasia and everyone else who worked on the patch and
>> ensured that it gets into the tree.
>
> Buildfarm results suggest that the test case is unstable under
> CLOBBER_CACHE_ALWAYS:
>
> https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=hyrax&dt=2021-01-19%2020%3A27%3A46
>
> This might mean an actual bug, or just that the test isn't robust.
>

Yeah :-( It seems I've committed the v5 patch, not the v7 addressing
exactly this issue (which I've actually pointed out and asked to be
fixed). Oh well ... I'll get this fixed tomorrow - I have the fix, and I
have verified that it passes with CLOBBER_CACHE_ALWAYS, but pushing it
at 5AM does not seem like a good idea.

regards

--
Tomas Vondra
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company