Date:   Mon, 10 Feb 2020 15:28:40 -0500
From:   Rich Felker <dalias@libc.org>
To:     "Michael Kerrisk (man-pages)" <mtk.manpages@gmail.com>
Cc:     linux-man <linux-man@vger.kernel.org>
Subject: Remove ESRCH from pthread_* man pages or document it as nonstandard
Message-ID: <20200210202840.GT1663@brightrain.aerifal.cx>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.5.21 (2010-09-15)
Sender: linux-man-owner@vger.kernel.org
Precedence: bulk

Use of a pthread_t after its lifetime has ended is undefined behavior.
Pre-2008 POSIX wrongly had ESRCH errors specified for a number of
interfaces, conflicting with explicitly making such use UB; this was
fixed as a result of Austin Group interpretation 142:

https://collaboration.opengroup.org/austin/interps/documents/14366/AI-142.txt

However, the man pages still have ESRCH listed for most if not all of
the pthread_* interfaces that take a pthread_t argument.

Could this be removed, or annotated to clarify that (1) this is not
standard/portable behavior programmers can rely on (FYI: musl does not
do it), and (2) like all instances of the pattern "use of resource
identifier after the resource it refers to has been released/freed",
passing a pthread_t whose lifetime may have ended is inherently a bug.

This came up again on the musl mailing list today:

https://www.openwall.com/lists/musl/2020/02/10/3

Rich