Horton

Re: Mark all GUC variable as PGDLLIMPORT

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Andres Freund <andres@anarazel.de>, Robert Haas <robertmhaas@gmail.com>, John Naylor <john.naylor@enterprisedb.com>, Andrew Dunstan <andrew@dunslane.net>, Chapman Flack <chap@anastigmatix.net>, Julien Rouhaud <rjuju123@gmail.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2022-05-13T00:05:44Z
Lists: pgsql-hackers 
On Thu, May 12, 2022 at 11:59:49AM -0400, Tom Lane wrote:
>> It seems pretty nonsensical to add PGDLLIMPORT to frontend only headers /
>> variables. What is that supposed to mean?
> 
> Yeah, that's why I removed it in 9a374b77.

Perhaps we should try to remove it from the header itself in the long
run, even if that's used in a couple of macros?  pgbench relies on it
to avoid building a debug string for a meta-command, and logging.h has
it in those compat macros..  I won't fight on that, though.

Anyway, I'll go remove the marking.  My apologies for the noise.
--
Michael

Commits

  1. Remove PGDLLIMPORT marker from __pg_log_level

  2. Mark a few 'bbsink' related functions / variables static.

  3. Add some missing PGDLLIMPORT markings

  4. Apply PGDLLIMPORT markings broadly.

  5. Helper script to apply PGDLLIMPORT markings.

  6. Simplify declaring variables exported from libpgcommon and libpgport.