I'm not so worried about this given so few people use 32-bit for production
these days and the error is not a crasher just gives wrong results. Even
then I can only exercise the issue using my 32-bit compiled mingw64
PostgreSQL (when running against the VC++ 32-bit build, the issue seems to
be non-existent which is what most people will be using that use 32-bit on
Björn also noted he is seeing a regress issue with Brin under 32-bit
Debian. I'm not seeing that at all in windows 32-bit. Given brin hasn't
changed in 2.4, I'm not so concerned about this one either.
P.S. Björn, you going to ticket that one?
No, I don't feel I have enough info to ticket it. Ignore until it becomes an issue for anyone else.
Aside from that we do have a bunch of tickets open for 2.2.6 and 2.3.4 that
I presume are also issues on 2.4.0. So would be nice to close as many of
these as possible.