From: Tom Lane <tgl@sss.pgh.pa.us>
To: Peter Eisentraut <peter_e@gmx.net>
Cc: pgsql-hackers@postgreSQL.org
Subject: Re: [COMMITTERS] pgsql: Fix bool/int type confusion
Date: Thu, 21 Sep 2017 11:46:40 -0400
Message-ID: <20053.1506008800@sss.pgh.pa.us> (raw)
In-Reply-To: <E1dsWhf-0000pT-F9@gemulon.postgresql.org>
References: <E1dsWhf-0000pT-F9@gemulon.postgresql.org>
List-Unsubscribe: <mailto:majordomo@postgresql.org?body=unsub%20pgsql-hackers>
Peter Eisentraut <peter_e@gmx.net> writes:
> Fix bool/int type confusion
> Using ++ on a bool variable doesn't work well when stdbool.h is in use.
> The original BSD code appears to use int here, so use that instead.
I'm fairly unhappy with this approach to fixing this problem, because
localtime.c is not Postgres-maintained code; we try to keep it in sync
with the upstream copy maintained by the IANA timezone crew. Patching it
like this will interfere with the next sync attempt, and patching it only
in master will cause back-patching of that sync patch to fail.
I checked the tz list archives and discovered that this problem was
already reported to them back in May:
http://mm.icann.org/pipermail/tz/2017-May/024995.html
Although indeed their previous coding had had the variable as "int",
their response was not to change it back to int, but to get rid
of the loop incrementing it, because that was intended to support
the possibility of 2 leap seconds on the same day, which according
to them is a widespread misunderstanding of the applicable standard.
So the code in their git repo still has the variable as bool, but
there's no ++ operator on it anymore.
This also means that the portability problem is purely hypothetical;
given valid tz reference data the code wouldn't ever try to increment
"hit" to 2 anyway. It's even more hypothetical for us, because we don't
use leap-second-aware zones.
They've not made a new tzcode release since May, due to lack of political
activity in this sphere this year, although I gather that one is likely
by November or so. However, we have precedent for sometimes syncing with
their git tip to absorb code bug fixes, cf commit 7afafe8af.
Therefore, what I would like to do is revert this commit (0ec2e908b),
and then either leave the problem to be fixed by our next regular sync
with a released tzcode version, or else force a sync with their git tip
to absorb their fix now. In either case we should keep all our live
branches in sync. I'm willing to do the code-sync work either way.
Comments?
regards, tom lane
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Reply to all the recipients using the --to and --cc options:
reply via email
To: pgsql-committers@postgresql.org
Cc: tgl@sss.pgh.pa.us, peter_e@gmx.net, pgsql-hackers@postgreSQL.org
Subject: Re: [COMMITTERS] pgsql: Fix bool/int type confusion
In-Reply-To: <20053.1506008800@sss.pgh.pa.us>
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox