Re: Cannot find a working 64-bit integer type on Illumos

Tom Lane <tgl@sss.pgh.pa.us>

From: Tom Lane <tgl@sss.pgh.pa.us>
To: Thomas Munro <thomas.munro@gmail.com>
Cc: Tomas Vondra <tomas@vondra.me>, Peter Eisentraut <peter@eisentraut.org>, Heikki Linnakangas <hlinnaka@iki.fi>, Japin Li <japinli@hotmail.com>, Andres Freund <andres@anarazel.de>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2024-12-05T03:29:38Z
Lists: pgsql-hackers
Thomas Munro <thomas.munro@gmail.com> writes:
> On Thu, Dec 5, 2024 at 1:59 PM Thomas Munro <thomas.munro@gmail.com> wrote:
>> I was trying to figure out how I missed this, and I think it might be
>> that the meson build scripts didn't port AC_SYS_LARGEFILES.  So if you
>> build on a 32 bit Linux system with meson (like one of CI's tasks, and
>> also build farm animal adder) then I think you finish up with 32 bit
>> off_t and no SIZEOF_OFF_T, because we don't do AC_SYS_LARGEFILES'
>> dance to figure out if this system needs -D_FILE_OFFSET_BITS=64 (or
>> other similar macros for AIX, Solaris etc).  I will look into that.

> Ahh, correction, it does define it (or else perl would have
> complained), but it seems that meson magically puts it into the
> compiler command line without being asked.  So it is defined without
> pg_config.h being involved, and thus earlier.  Huh.

That does not seem great.  Compile an extension without the same
CPPFLAGS, you silently get an ABI-incompatible module.  We really
ought to be putting these ABI-critical flags into pg_config.h.
It's especially bad that this works differently between autoconf
and meson builds.

			regards, tom lane



Commits

  1. Move pg_int64 back to postgres_ext.h

  2. pgbench: Make set_random_seed() 64-bit everywhere.

  3. Use PRI?64 instead of "ll?" in format strings (continued).

  4. Fix order of -I switches for building pg_regress.o.

  5. libpq: Deprecate pg_int64.

  6. Use PRI*64 instead of "ll*" in format strings (minimal trial)

  7. Fix header inclusion order in c.h.

  8. Use <stdint.h> and <inttypes.h> for c.h integers.

  9. Remove traces of BeOS.

  10. More correct way to check for existence of types, which allows to specify