Re: Issue with the PRNG used by Postgres

On Wed, Apr 10, 2024 at 4:40 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
> I'm not worried about it being slower, but about whether it could
> report "stuck spinlock" in cases where the existing code succeeds.
> While that seems at least theoretically possible, it seems like
> if you hit it you have got problems that need to be fixed anyway.
> Nonetheless, I'm kind of leaning to not back-patching.  I do agree
> on getting it into HEAD sooner not later though.

I just want to mention that I have heard of "stuck spinlock" happening
in production just because the server was busy. And I think that's not
intended. The timeout is supposed to be high enough that you only hit
it if there's a bug in the code. At least AIUI. But it isn't.

I know that's a separate issue, but I think it's an important one. It
shouldn't happen that a system which was installed to defend against
bugs in the code causes more problems than the bugs themselves.

--
Robert Haas
EDB: http://www.enterprisedb.com