Re: pgsql: Move SQL-callable code related to multixacts into its own file

Hi,

On Tue, Aug 19, 2025 at 11:27:25AM -0400, Tom Lane wrote:
> =?utf-8?Q?=C3=81lvaro?= Herrera <alvherre@kurilemu.de> writes:
> > Yeah, I'm not on board with making changes to the makefiles, because
> > AFAIR the current arrangement is purposefully what is is.  I did
> > discover that my build script does "make install ; make -C contrib
> > install", which means src/test/modules is not built.  But that's a local
> > fix for me.
> 
> I think "make world-bin"

We have:

$(call recurse,world-bin,src config contrib,all)

means src/Makefile is reached but will not reach src/test/modules/Makefile
because test/modules is not part of src/Makefile's SUBDIRS.

> or "make install-world" are the accepted targets for building everything.

We have:

$(call recurse,install-world,doc src config contrib,install)

so, same as above (src/test/modules/Makefile will not be reached).

From what I can see, only:

check-world:

$(call recurse,check-world,src/test src/pl src/interfaces contrib src/bin src/tools/pg_bsd_indent,check)

checkprep:

$(call recurse,checkprep,  src/test src/pl src/interfaces contrib src/bin)

installcheck-world:

$(call recurse,installcheck-world,src/test src/pl src/interfaces contrib src/bin,installcheck)

build src/test/modules.

I just found surprising that a "default" make (no target specified) does not
build src/test/modules while a "default" meson/ninja does.

Regards,

-- 
Bertrand Drouvot
PostgreSQL Contributors Team
RDS Open Source Databases
Amazon Web Services: https://aws.amazon.com