Re: pg_trgm comparison bug on cross-architecture replication due to different char implementation

On Tue, Sep 10, 2024 at 3:05 PM Noah Misch <noah(at)leadboat(dot)com> wrote:
>
> On Tue, Sep 10, 2024 at 05:56:47PM -0400, Tom Lane wrote:
> > Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com> writes:
> > > On Tue, Sep 10, 2024 at 11:57 AM Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> > >> Yeah, that seems like it could work. But are we sure that replicas
> > >> get a copy of the primary's control file rather than creating their
> > >> own?
> >
> > > Yes, I think so. Since at least the system identifiers of primary and
> > > replicas must be identical for physical replication, if replicas use
> > > their own control files then they cannot start the replication.
> >
> > Got it. So now I'm wondering if we need all the complexity of storing
> > stuff in the GIN metapages. Could we simply read the (primary's)
> > signedness out of pg_control and use that?
>
> Yes.

Indeed.

I've attached a PoC patch for this idea. We write the default char
signedness to the control file at initdb time. Then when comparing two
trgms, pg_trgm opclasses use a comparison function based on the char
signedness of the cluster. I've confirmed that the patch fixes the
reported case at least.

Regards,

--
Masahiko Sawada
Amazon Web Services: https://aws.amazon.com