From: Noah Misch <noah@leadboat.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: pgsql-hackers@postgresql.org
Subject: Re: Obsolete coding in fork_process.c
Date: Fri, 2 May 2014 10:28:49 -0400
Message-ID: <20140502142849.GA1219884@tornado.leadboat.com> (raw)
In-Reply-To: <11846.1399000071@sss.pgh.pa.us>
References: <30040.1398960808@sss.pgh.pa.us>
<20140501183231.GA1190179@tornado.leadboat.com>
<4013.1398971788@sss.pgh.pa.us>
<28165.1398981548@sss.pgh.pa.us>
<20140502000053.GA1191730@tornado.leadboat.com>
<8749.1398991486@sss.pgh.pa.us>
<20140502012456.GB1191730@tornado.leadboat.com>
<11846.1399000071@sss.pgh.pa.us>
List-Unsubscribe: <mailto:majordomo@postgresql.org?body=unsub%20pgsql-hackers>
On Thu, May 01, 2014 at 11:07:51PM -0400, Tom Lane wrote:
> Noah Misch <noah@leadboat.com> writes:
> > On Thu, May 01, 2014 at 08:44:46PM -0400, Tom Lane wrote:
> >> You're only considering one aspect of the problem. Yeah, you might not
> >> get duplicated output unless system() prints something before exec(),
> >> but we would also like to have causality: that is, whatever we sent to
> >> stdout before calling system() should appear there before anything the
> >> child process sends to stdout.
>
> > Good point. I suppose a couple of fflush() calls have negligible cost next to
> > a system() or popen(). Introduce pg_popen()/pg_system(), and adopt a rule
> > that they are [almost] our only callers of raw popen()/system()?
>
> Meh. I'm not usually in favor of adopting nonstandard notation, and
> this doesn't seem like a place to start. In particular, if you don't
> want to use fflush(NULL) in these proposed wrappers, then call sites
> are still going to have an issue with needing to do manual fflushes;
> pg_regress.c's spawn_process is an example:
>
> /*
> * Must flush I/O buffers before fork. Ideally we'd use fflush(NULL) here
> * ... does anyone still care about systems where that doesn't work?
> */
> fflush(stdout);
> fflush(stderr);
> if (logfile)
> fflush(logfile);
>
> pid = fork();
>
> I think that removing the need for fflush(stdout) and fflush(stderr)
> in this context would mostly result in people forgetting to fflush
> other output files. I'd rather have the two lines of boilerplate
> (and a comment about why we're refusing to depend on fflush(NULL))
> than take that risk.
Works for me.
--
Noah Misch
EnterpriseDB http://www.enterprisedb.com
--
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-hackers@postgresql.org
Cc: noah@leadboat.com, tgl@sss.pgh.pa.us
Subject: Re: Obsolete coding in fork_process.c
In-Reply-To: <20140502142849.GA1219884@tornado.leadboat.com>
* 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