From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <pthreads-win32-return-692-listarch-pthreads-win32=sources.redhat.com@sources.redhat.com>
Received: (qmail 13258 invoked by alias); 23 Sep 2003 12:06:49 -0000
Mailing-List: contact pthreads-win32-help@sources.redhat.com; run by ezmlm
Precedence: bulk
List-Subscribe: <mailto:pthreads-win32-subscribe@sources.redhat.com>
List-Archive: <http://sources.redhat.com/ml/pthreads-win32/>
List-Post: <mailto:pthreads-win32@sources.redhat.com>
List-Help: <mailto:pthreads-win32-help@sources.redhat.com>, <http://sources.redhat.com/ml/#faqs>
Sender: pthreads-win32-owner@sources.redhat.com
Received: (qmail 13225 invoked from network); 23 Sep 2003 12:06:47 -0000
Received: from unknown (HELO relay.materna.de) (193.96.115.65)
  by sources.redhat.com with SMTP; 23 Sep 2003 12:06:47 -0000
Received: from ganymed.materna.de (ganymed [139.2.34.141])
	by relay.materna.de (Postfix) with ESMTP id 4D41F67F3
	for <pthreads-win32@sources.redhat.com>; Tue, 23 Sep 2003 14:06:43 +0200 (MEST)
Received: from ganymed.materna.de (localhost [127.0.0.1])
	by mail.materna.de (Postfix) with ESMTP id 12420D901
	for <pthreads-win32@sources.redhat.com>; Tue, 23 Sep 2003 14:06:46 +0200 (MET DST)
Received: from hb-castor.hb.buc.materna.com (hb-castor.hb.materna.de [139.2.228.234])
	by mail.materna.de (Postfix) with ESMTP id E3805D900
	for <pthreads-win32@sources.redhat.com>; Tue, 23 Sep 2003 14:06:45 +0200 (MET DST)
Received: by hb-castor.hb.buc.materna.com with Internet Mail Service (5.5.2653.19)
	id <TK13M110>; Tue, 23 Sep 2003 14:06:45 +0200
Message-ID: <5BD61062E524E244ADD87F7988480EC3C82369@hb-castor.hb.buc.materna.com>
From: "Brozinski, Stefan" <stefan.brozinski@materna.de>
To: "'pthreads-win32@sources.redhat.com'" <pthreads-win32@sources.redhat.com>
Subject: Dangerous: localtime_r, gmtime_r
Date: Tue, 23 Sep 2003 12:06:00 -0000
MIME-Version: 1.0
Content-Type: text/plain
X-SW-Source: 2003/txt/msg00092.txt.bz2

Hi all,

there is a problem with the current implementation of the thread safe
version of "localtime" and "gmtime" currently implemented as a macro in
<pthread.h>.

The current definition for localtime_r is:

#define localtime_r( _clock, _result ) \
	( *(_result) = *localtime( (_clock) ), \
	  (_result) )

The problem is that localtime() may return NULL. At least Microsoft's
implementation in MSVCRT[D].DLL does in fact return NULL if certain
conditions are met. Thus, a page fault may occur when using this
implementation.

I suggest to change the implementation from the current #define to a 'real'
function like:

struct tm *localtime_r(time_t *_clock, struct tm *_result)
{
  struct tm *p = localtime(_clock);

  if (p)
    *(_result) = *p;

  return p;
}

Regards
Stefan