From owner-freebsd-wireless@FreeBSD.ORG  Thu Dec  1 00:39:39 2011
Return-Path: <owner-freebsd-wireless@FreeBSD.ORG>
Delivered-To: freebsd-wireless@freebsd.org
Received: by hub.freebsd.org (Postfix, from userid 1233)
	id 59E751065670; Thu,  1 Dec 2011 00:39:38 +0000 (UTC)
Date: Thu, 1 Dec 2011 00:39:38 +0000
From: Alexander Best <arundel@freebsd.org>
To: Igor Mozolevsky <mozolevsky@gmail.com>
Message-ID: <20111201003938.GA52503@freebsd.org>
References: <20111130224422.GA36424@freebsd.org>
	<CADWvR2herXkcMaZXFKm_ALBXnVFDhRCNZ-9+-MwHVx5XJPG4AA@mail.gmail.com>
	<20111201000243.GB44877@freebsd.org>
	<CADWvR2irEU0ANHNG4mvC1voVP6Y3YMTQxs5d35o5SbBTJ8JyfQ@mail.gmail.com>
	<20111201001158.GA48051@freebsd.org>
	<CADWvR2iUB4mDFuwbWzqe6bnjShTZEQzQ-ubAx8h7yqGwjcUHKA@mail.gmail.com>
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="8t9RHnE3ZwKMSgU+"
Content-Disposition: inline
In-Reply-To: <CADWvR2iUB4mDFuwbWzqe6bnjShTZEQzQ-ubAx8h7yqGwjcUHKA@mail.gmail.com>
Cc: freebsd-wireless@freebsd.org
Subject: Re: comparing floating points via "==" or "!="
X-BeenThere: freebsd-wireless@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: "Discussions of 802.11 stack,
	tools device driver development." <freebsd-wireless.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-wireless>, 
	<mailto:freebsd-wireless-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-wireless>
List-Post: <mailto:freebsd-wireless@freebsd.org>
List-Help: <mailto:freebsd-wireless-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-wireless>, 
	<mailto:freebsd-wireless-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Thu, 01 Dec 2011 00:39:39 -0000


--8t9RHnE3ZwKMSgU+
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

On Thu Dec  1 11, Igor Mozolevsky wrote:
> On 1 December 2011 00:11, Alexander Best <arundel@freebsd.org> wrote:
> 
> > i sorry. since sys/dev/ath/ath_rate/sample/tx_schedules.h also handles a lot
> > of integer values, i thought you referred to abs(3), because you saw also
> > some issues with the integer computations in that file. so i thought you were
> > referring to yet another issue in that file.
> 
> Oh, I see! No, I was referring to a generic |a-b| expression, not
> abs(3) specifically. Retrospectively, using |a-b| would've been better
> :-)

so wrapping fabs(3) inside a macro, like in the following example might work.

cheers.
alex

ps: just a 60 second hack. ;)

> 
> Cheers,
> 
> --
> Igor

--8t9RHnE3ZwKMSgU+
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="tx_schedules.h.diff"

Index: sys/dev/ath/ath_rate/sample/tx_schedules.h
===================================================================
--- sys/dev/ath/ath_rate/sample/tx_schedules.h	(revision 228164)
+++ sys/dev/ath/ath_rate/sample/tx_schedules.h	(working copy)
@@ -40,6 +40,11 @@
 #include <sys/cdefs.h>
 __FBSDID("$FreeBSD$");
 
+#define TOLERANCE 0.00001
+
+#define F(_r1, _r2)\
+	((fabs(_r1-_r2) < TOLLERANCE) ? 1 : 0)
+
 #define A(_r) \
     (((_r) == 6)   ? 0 : (((_r) == 9)   ? 1 : (((_r) == 12)  ? 2 : \
     (((_r) == 18)  ? 3 : (((_r) == 24)  ? 4 : (((_r) == 36)  ? 5 : \
@@ -56,7 +61,7 @@
 };
 
 #define NA1(_r) \
-	(((_r) == 6.5)  ? 8 : (((_r) == 13)  ?  9 : (((_r) == 19.5)? 10 : \
+	((F(_r, 6.5)   ? 8 : (((_r) == 13)  ?  9 : (((_r) == 19.5)? 10 : \
 	(((_r) == 26)  ? 11 : (((_r) == 39)  ? 12 : (((_r) == 52)  ? 13 : \
 	(((_r) == 58.5)? 14 : (((_r) == 65)  ? 15 : 0))))))))
 #define NA2(_r) \

--8t9RHnE3ZwKMSgU+--