Email List: Xaustin-group-lX
[All Lists]

Re: [Fwd: Re: whether ext() and atexit() are safe w.r.t each other

To: Anoop <acv@xxxxxxxxxxxxxxxxxx>
Subject: Re: [Fwd: Re: whether ext() and atexit() are safe w.r.t each other
From: Ulrich Drepper <drepper@xxxxxxxxxx>
Date: Tue, 10 Jun 2008 07:10:20 -0700
Cc: David Butenhof <david.butenhof@xxxxxx>, Geoff Clare <gwc@xxxxxxxxxxxxx>, "austin-group-l@xxxxxxxxxxxxx" <austin-group-l@xxxxxxxxxxxxx>
Organization: Red Hat, Inc.
References: <1212759114.12557.52.camel@localhost.localdomain> <20080606154514.GB39160@finch-staff-1.thus.net> <48496905.2050606@hp.com> <20080609075139.GC74480@finch-staff-1.thus.net> <20080609085814.GA23187@squonk.masqnet> <484CFFA2.6000701@hp.com> <484E2D05.9020008@linux.vnet.ibm.com>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Anoop wrote:
> All successfully registered handlers are invoked before exiting, or in
> other words,
> Fail all the registrations that cannot be handled. Is that correct?

No.  It is perfectly reasonable to have an implementation which takes a
copy of the atexit handlers when the exit() function is entered and then
work on that copy.  New additions won't be seen.  There is no reason to
give programmers with questionable (at best) code reason to demand that
implementations are more complex than they are already.

- --
➧ Ulrich Drepper ➧ Red Hat, Inc. ➧ 444 Castro St ➧ Mountain View, CA ❖
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAkhOi0wACgkQ2ijCOnn/RHS/mACfe7HJjI1dNxPkcpdlOmIMiIW/
HVMAoJ4btdwC5QEB8LI+76hqdfgYeYsZ
=jHxW
-----END PGP SIGNATURE-----

<Prev in Thread] Current Thread [Next in Thread>