Re: pgsql: Add support for basic NUMA awareness

Hi,

Walther, adding you because of the failure on buildfarm animal dogfish. The
main issue is something unrelated to your animal (except it was the only one
to catch it), but there's one oddity:

On 2025-04-08 10:36:32 -0400, Tom Lane wrote:
> Andres Freund <andres(at)anarazel(dot)de> writes:
> > On April 8, 2025 10:05:48 AM EDT, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> >> I noticed this a little further up in dogfish's log:
> >>
> >> ninja: bad depfile: expected ':', saw 's'
>
> > I noticed that too - but it happened even before the hard failures. Will look at it in a bit.
>
> After catching up on the hackers thread, I wondered if this isn't
> actually part of the identical issue, that is ninja is noticing
> that there's not a dependency leading to that include file.

No, it seems to be something else - that "error" is happening even in the
backbranches, which have been building without an issue for 167 days.

But I'm rather bewildered - I can't actually build with the professed version
of ninja, it errors out:

ninja-1.9.0 src/bin/psql/sql_help.c
ninja: error: build.ninja:8462: multiple outputs aren't (yet?) supported by depslog; bring this up on the mailing list if it affects you

So this must be some version of ninja between 1.9 and 1.10 (where the multiple
outputs support was added).

What's particularly weird about that is that the rest of the distro seems much
newer. gcc 14.2.1, meson 1.6.1, LLVM 18. What's a < 2020 ninja doing with
those surroundings?

> > I think part of the file would make sense to eventually use in FE code, we should move the SQL function and the page size function out.
>
> Works for me. But a file in src/port/ should really use only c.h
> if at all possible. Otherwise (as Andrew said) it has to choose
> one of postgres.h and postgres_fe.h depending on FRONTEND.

Yea. I think after moving the rest c.h should suffice.

Greetings,

Andres Freund