From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S268166AbUH2BnK (ORCPT ); Sat, 28 Aug 2004 21:43:10 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S268158AbUH2BnK (ORCPT ); Sat, 28 Aug 2004 21:43:10 -0400 Received: from dci.doncaster.on.ca ([66.11.168.194]:43662 "EHLO smtp.istop.com") by vger.kernel.org with ESMTP id S268166AbUH2BnC (ORCPT ); Sat, 28 Aug 2004 21:43:02 -0400 From: Andrew Miklas Reply-To: public@mikl.as To: Gianni Tedesco Subject: Re: Linux Incompatibility List Date: Sat, 28 Aug 2004 21:42:43 -0400 User-Agent: KMail/1.7 Cc: linux-kernel References: <87r7q0th2n.fsf@dedasys.com> <200408250159.20606.public@mikl.as> <1093418493.18600.10.camel@sherbert> In-Reply-To: <1093418493.18600.10.camel@sherbert> MIME-Version: 1.0 Content-Type: multipart/signed; boundary="nextPart156251825.WiuWTpF0tG"; protocol="application/pgp-signature"; micalg=pgp-sha1 Content-Transfer-Encoding: 7bit Message-Id: <200408282143.05304.public@mikl.as> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org --nextPart156251825.WiuWTpF0tG Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Hi, On August 25, 2004 03:21 am, Gianni Tedesco wrote: > Also I think that the pciproxy[0] technique currently offers a simpler > solution. Analyzing the data produced is more akin to reverse > engineering a network protocol than machine code. Reverse engineering > protocols is much less 'copyright sensitive' than decompiling machine > code and, I think, more easily shared. (Sorry for the long delay...) Yeah, we tried that at first too (actually, we were using Frank Cornelis' = =20 patches to Bochs). The problem with dumping all the PCI activity (even whi= le=20 the interface isn't sending/receiving) is that there is a huge amount of da= ta=20 to sift through. Even capturing for just a for a few seconds generates=20 megabytes of data. You also have to deal with various events (like the=20 watchdog timer) going off at random times and getting mixed in with the=20 send/receive data. We also found DMA a little tricky too (ie. you need to= =20 dump out any data that the chip will bus-master to itself to see how it is= =20 structured). Also, we figured it would cause problems for supporting the whole range of= =20 devices that can be handled by the wl.o drivers. For example, from looking= =20 at the module, we can see that the driver will have somewhat different=20 behaviour according to exactly what MAC and radio chips are present, the=20 interface being used (ie. PCI, cardbus, etc.), the vendor, model number, an= d=20 revision of the board itself, the contents of the ROM, etc. We decided tha= t=20 there was simply too many combinations to make the data capture approach=20 useful over the entire range of Broadcom hardware, unless you repeat the=20 process on every variation. > Anyway, perhaps once I've had some time to make a little more progress > we would be able to compare some notes? Sure, let me know when you're ready. =2D- Andrew --nextPart156251825.WiuWTpF0tG Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Comment: Key ID: EC3F6CCD (www.keyserver.net) iD8DBQBBMTSpTHKGaOw/bM0RAvvlAJ9Gfw9BInED8HixHAR1E9hjvKsiEQCgmhNK jKLbfJLQYEpRYSmklYQuqGM= =jsIn -----END PGP SIGNATURE----- --nextPart156251825.WiuWTpF0tG--