From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1IoOSN-0006wW-CL for qemu-devel@nongnu.org; Sat, 03 Nov 2007 15:14:55 -0400 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1IoOSM-0006wA-R4 for qemu-devel@nongnu.org; Sat, 03 Nov 2007 15:14:54 -0400 Received: from [199.232.76.173] (helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1IoOSM-0006w4-KZ for qemu-devel@nongnu.org; Sat, 03 Nov 2007 15:14:54 -0400 Received: from hall.aurel32.net ([88.191.38.19]) by monty-python.gnu.org with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.60) (envelope-from ) id 1IoOSL-0001j5-VV for qemu-devel@nongnu.org; Sat, 03 Nov 2007 15:14:54 -0400 Received: from volta.aurel32.net ([2001:618:400:fc13:216:d3ff:fe17:fd00]) by hall.aurel32.net with esmtpsa (TLS-1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.63) (envelope-from ) id 1IoOSK-0006UI-F0 for qemu-devel@nongnu.org; Sat, 03 Nov 2007 20:14:52 +0100 Received: from localhost.aurel32.net ([127.0.0.1] ident=aurel32) by volta.aurel32.net with esmtp (Exim 4.68) (envelope-from ) id 1IoOSM-0003Bf-3M for qemu-devel@nongnu.org; Sat, 03 Nov 2007 20:14:54 +0100 Message-ID: <472CC8AD.1000904@aurel32.net> Date: Sat, 03 Nov 2007 20:14:53 +0100 From: Aurelien Jarno MIME-Version: 1.0 Subject: Re: [Qemu-devel] [PATCH] Fix NaN handling in softfloat References: <20071103173548.GA16847@hall.aurel32.net> <20071103180604.GA14403@caradoc.them.org> In-Reply-To: <20071103180604.GA14403@caradoc.them.org> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Reply-To: qemu-devel@nongnu.org List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Daniel Jacobowitz a écrit : > On Sat, Nov 03, 2007 at 06:35:48PM +0100, Aurelien Jarno wrote: >> Hi all, >> >> The current softfloat implementation changes qNaN into sNaN when >> converting between formats, for no reason. The attached patch fixes >> that. It also fixes an off-by-one in the extended double precision >> format (aka floatx80), the mantissa is 64-bit long and not 63-bit >> long. >> >> With this patch applied all the glibc 2.7 floating point tests >> are successfull on MIPS and MIPSEL. > > FYI, I posted a similar patch and haven't had time to get back to it. > Andreas reminded me that we need to make sure at least one mantissa > bit is set. If we're confident that the common NaN format will > already have some bit other than the qnan/snan bit set, this is fine; > otherwise, we might want to forcibly set some other mantissa bit. > Good point. I will try to improve my patch. -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' aurel32@debian.org | aurelien@aurel32.net `- people.debian.org/~aurel32 | www.aurel32.net