From: Tom Lane <tgl@sss.pgh.pa.us>
To: Tobias Bussmann <t.bussmann@gmx.net>
cc: John Naylor <johncnaylorls@gmail.com>, Lukas Fittl <lukas@fittl.com>,
        Jakob Egger <jakob@eggerapps.at>,
        pgsql-hackers <pgsql-hackers@postgresql.org>,
        Andres Freund <andres@anarazel.de>,
        Sandeep Thakkar <sandeep.thakkar@enterprisedb.com>
Subject: Re: Broken build on macOS (Universal / Intel): cpuid instruction not
 available
In-reply-to: <15574903-87C9-478A-B2D7-CC8F4C275DBB@gmx.net>
References: <223EA201-A0E8-4A13-B220-EB903E8DF817@eggerapps.at>
 <871806.1778168884@sss.pgh.pa.us> <873909.1778170924@sss.pgh.pa.us>
 <CAP53PkzLmeNUiNtcOyw62KkD7OzthdCbkwGTyAk+Lu3LKsmOcg@mail.gmail.com>
 <CANWCAZa5TEmjLy0xtBjP9NyuVCJpCOpVkBSzw-fqmH8D9tFzDg@mail.gmail.com>
 <471E4CB3-1690-4168-9A99-5F83D97C12AF@gmx.net>
 <CANWCAZZNu5nLzeeNs8TyC6Xs4HKzHni=yeTqCH7_QNpktR2nMQ@mail.gmail.com>
 <15574903-87C9-478A-B2D7-CC8F4C275DBB@gmx.net>
Comments: In-reply-to Tobias Bussmann <t.bussmann@gmx.net>
	message dated "Tue, 02 Jun 2026 14:49:54 +0200"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <2925607.1780411691.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Tue, 02 Jun 2026 10:48:11 -0400
Message-ID: <2925608.1780411691@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/2925608.1780411691%40sss.pgh.pa.us>
Precedence: bulk

Tobias Bussmann <t.bussmann@gmx.net> writes:
> Indeed. The universal builds seem to disable optimisations that are supp=
orted =

> natively.

Yes, we realized that upthread.  It's been like that since day 1,
not something that's new in v19.  I don't see that we're likely
to do anything about it: it'd require a fairly fundamental
re-architecting of our configure tests, and the end result would
only be to improve support for machines that are going to be dead
to macOS within a year.

However, it definitely is a regression that the build fails
altogether.  Too bad nobody tried the x86 -> ARM case earlier.

			regards, tom lane