Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner AT cygwin DOT com
Delivered-To: mailing list cygwin AT cygwin DOT com
From: "Ralf Habacker" <Ralf DOT Habacker AT freenet DOT de>
To: "Cygwin" <cygwin AT sources DOT redhat DOT com>
Subject: RE: gettimeofday() does not returns usec resolution
Date: Thu, 24 Jan 2002 21:05:54 +0100
Message-ID: <002e01c1a512$877216c0$0fa807d5@BRAMSCHE>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook 8.5, Build 4.71.2173.0
Importance: Normal
In-reply-to: <3C4F64B4.380E84AF@lapo.it>
X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400

>
>
> > Hmm.  I do remember this patch.  Can you submit this to cygwin-patches
> > in the correct format: http://cygwin.com/contrib.html ?
>
> Given enough time, yes (never hacked that one, but it seems that only
> winsup/cygwin/times.cc would need to be modified).. anyway of course I'm one of
> the few that actually would like to have this feature so I guess there's no big
> hurry.
>
> Of course the number reported by PerformanceTimer has an unknown offset with
> system time (it appears to start at 0 at program start, but that's not clearly
> specified) so I would have calculate an initial "offset" in the "old" way... this
> would give 3-4 microsecond precision on "deltas" but an initial uncertainty of
             ^^^^^
> the usual 15 milliseconds or so.
>
This is a output of the mentioned profiler. It say about a minimal time of about 1-2 us. May the difference be
caused by the cygwin.dll overhead ? (We're using the native functions, not gettimeofday()).

[1828] EmptyTest   count:    100  time: (min)          1 [us] (max)          2 [us] (sum)        169 [us]

> I guess that it's better that having that uncertainty between ANY two
> measurations... but if someone has some other idea.. please suggest =)
>
> > Do you have an assignment on file with Red Hat?
>
> Sorry but english knowledge doesn't let me understand "assignment on file", could
> you explain it to me with other words?
>
> > >-------
> > >int gettimeofday(struct timeval *tv, struct timezone *tz)
> > >{....
> > >}
> > >------------
>
> A note for everyone: that implementation was taken from tim Prince's message on
> lmbench and of course lacked any syncronization with real time as was used only
> for deltas.
>
> --
> Lapo 'Raist' Luchini
> lapo AT lapo DOT it (PGP & X.509 keys available)
> http://www.lapo.it (ICQ UIN: 529796)
>
>
> --
> Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
> Bug reporting:         http://cygwin.com/bugs.html
> Documentation:         http://cygwin.com/docs.html
> FAQ:                   http://cygwin.com/faq/
>
>


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Bug reporting:         http://cygwin.com/bugs.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/