From owner-cvs-lib  Fri Oct 25 21:01:12 1996
Return-Path: owner-cvs-lib
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id VAA21591
          for cvs-lib-outgoing; Fri, 25 Oct 1996 21:01:12 -0700 (PDT)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id VAA21580;
          Fri, 25 Oct 1996 21:01:04 -0700 (PDT)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.7.6/8.6.9) id NAA11678; Sat, 26 Oct 1996 13:56:15 +1000
Date: Sat, 26 Oct 1996 13:56:15 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199610260356.NAA11678@godzilla.zeta.org.au>
To: alex@freefall.freebsd.org, cvs-all@freefall.freebsd.org,
        CVS-committers@freefall.freebsd.org, cvs-lib@freefall.freebsd.org
Subject: Re: cvs commit:  src/lib/libc/stdlib abort.c
Sender: owner-cvs-lib@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

>  Modified:    lib/libc/stdlib  abort.c
>  Log:
>  POSIX requires stdio buffers be flushed on abort.

Actually, it requires abort() to "have the effect of fclose() on
every open stream" if abort() causes process termination, and no
other effect on streams otherwise.

It's not clear that flushing has the same effect in all cases.  I
think it does (a little later when the process exits in the process
termination case).  This depends on application-supplied read and
write functions (if any) not doing anything extra in close() or
write().

Bruce