From: Tom Lane <tgl@sss.pgh.pa.us>
To: Jesse Zhang <sbjesse@gmail.com>
cc: PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Subject: Re: Residual cpluspluscheck issues
In-reply-to: 
 <CAGf+fX5W2JwS3c0DW+c=tnmoCGEs5tvKNYMQZ+DV2XaLWAJ81g@mail.gmail.com>
References: <20518.1559494394@sss.pgh.pa.us>
 <CAGf+fX5W2JwS3c0DW+c=tnmoCGEs5tvKNYMQZ+DV2XaLWAJ81g@mail.gmail.com>
Comments: In-reply-to Jesse Zhang <sbjesse@gmail.com>
	message dated "Wed, 30 Sep 2020 08:20:45 -0700"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <844759.1601480838.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Wed, 30 Sep 2020 11:47:18 -0400
Message-ID: <844760.1601480838@sss.pgh.pa.us>
Precedence: bulk

Jesse Zhang <sbjesse@gmail.com> writes:
> So it's been 17 months since you sent this email, so I'm not sure that
> nothing has happened (off list or in the code base), but...

Well, we fixed the case that was discussed at the time [1].

I'm not exactly convinced about removing the register keywords in
s_lock.h.  Those are all associated with asm blocks, which are already
extremely C/GCC specific; complaining that the register declarations
aren't portable seems to be missing the forest for the trees.

BTW, grepping my local tree says that plperl/ppport.h also has some
register variables, which is something we have no control over.

			regards, tom lane

[1] https://git.postgresql.org/gitweb/?p=3Dpostgresql.git&a=3Dcommitdiff&h=
=3D232720be9