DomainKey-Signature: a=rsa-sha1; c=nofws; d=sourceware.org; h=list-id
	:list-unsubscribe:list-subscribe:list-archive:list-post
	:list-help:sender:message-id:date:from:mime-version:to:subject
	:references:in-reply-to:content-type:content-transfer-encoding;
	 q=dns; s=default; b=yFWgEgDOrEtBocNpvHLSvfFngSRfIHJ3r2Td7Up5ypJ
	QK0II5Eermlh3XjuVfAmi3mWTazqVTosO6jQPx+2TRY6EH7ExbN6stdwerfrYeAT
	BH2DgXmfhTXxaUdTz3xu/2s/p1/otDba0xxsc/7jACUINoK1KkIwLp8hDx1SdMV4
	=
Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
Message-ID: <5166BCA1.5040204@net-b.de>
Date: Thu, 11 Apr 2013 15:37:37 +0200
From: Tobias Burnus <burnus AT net-b DOT de>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130329 Thunderbird/17.0.5
MIME-Version: 1.0
To: cygwin AT cygwin DOT com, Corinna Vinschen <corinna-cygwin AT cygwin DOT com>,
        Christopher Faylor <cgf AT gcc DOT gnu DOT org>
Subject: Cygwin with clock_gettime and CLOCK_MONOTONIC - gives always 0
References: <5166A0B3 DOT 70801 AT net-b DOT de>
In-Reply-To: <5166A0B3.70801@net-b.de>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Please CC your reply to fortran AT gcc DOT gnu DOT org

Dear all,

using clock_gettime with CLOCK_MONOTONIC fails on Cygwin; it always 
gives 0. That breaks code compiled with GCC's gfortran which uses 
system_clock in libgfortran.

libgfortran uses CLOCK_MONOTONIC if defined and otherwise it falls back 
to CLOCK_REALTIME (which works); however, Cygwin defines both. Maybe a 
better check would be _POSIX_MONOTONIC_CLOCK [2], however, that is also 
defined with Cygwin [1].


First, is a known problem that clock_gettime gives the wrong result?

Secondly, do you have an idea how to test for it? One could use an exec 
test in configure, but that does not nicely work with cross 
compilations, which are commonly done for GCC.


See http://gcc.gnu.org/ml/fortran/2013-04/msg00085.html for the thread 
at GCC, where the issue showed up.

(The issue is new since GCC 4.6 (2011-01-31) as that's the first GCC 
version where libgfortran uses clock_gettime (when available); before 
gettimeofday was used. See also GCC bugreport 56919. And CLOCK_MONOTONIC 
appeared in Cygwin around the same time [Aug 2010], cf. [1])


Tobias


[1] Cf. Cygwin-1.7.6 announcement at 
http://cygwin.com/ml/cygwin-announce/2010-08/msg00019.html and the 
"#define _POSIX_MONOTONIC_CLOCK 200112L" patch at 
http://sourceware.org/ml/newlib/2010/msg00454.html Note that Linux's 
glibc has in /usr/include/bits/posix_opt.h a more careful value: "/* The 
monotonic clock might be available.  */", "#define 
_POSIX_MONOTONIC_CLOCK    0"
[2] http://pubs.opengroup.org/onlinepubs/009695399/basedefs/xbd_chap02.html

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple