Return-path: Received: from mail-fx0-f213.google.com ([209.85.220.213]:63160 "EHLO mail-fx0-f213.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753690AbZFGJ4e (ORCPT ); Sun, 7 Jun 2009 05:56:34 -0400 Subject: Re: [ANN] Atheros 802.11n USB firmware source code released From: Maxim Levitsky To: Johannes Berg Cc: Maxim Levitsky , "Luis R. Rodriguez" , linux-wireless , linux-kernel@vger.kernel.org, otus-devel@lists.madwifi-project.org, David Woodhouse , "John W. Linville" In-Reply-To: <1244367815.26071.4.camel@johannes.local> References: <43e72e890906051204q22278e4cw6beeae27da50e988@mail.gmail.com> <1244326943.12159.5.camel@localhost> <1244334102.16810.3.camel@localhost> <1244367815.26071.4.camel@johannes.local> Content-Type: text/plain Date: Sun, 07 Jun 2009 12:54:15 +0300 Message-Id: <1244368455.22643.4.camel@localhost> Mime-Version: 1.0 Sender: linux-wireless-owner@vger.kernel.org List-ID: On Sun, 2009-06-07 at 11:43 +0200, Johannes Berg wrote: > On Sun, 2009-06-07 at 03:21 +0300, Maxim Levitsky wrote: > > > > Is this source for same firmware that was closed? > > Yes, but it was compiled differently afaik so you won't arrive at the > same binary. > > > > Will AP mode work in future? > > Well, not TYPE=AP, but we'll integrate whatever is necessary into one > firmware, I expect. It shouldn't be very hard given that AP mode really > works with the TYPE=STA firmware already, almost. > > > What does this mean: > > > > eeprom : ORIGIN = 0x000000, LENGTH = 1024k > > > > > > Does the device have whole 1M of eeprom? > > What is there? > > The eeprom is mapped there, but I don't think it's 1M, it's just the max > size of the mapping to the CPUs address space. Fine, like I expected, but still, I guess there is some 'loader' in eeprom? > > > Also what is this: > > > > sram : ORIGIN = 0x100000, LENGTH = 64k > > This is just memory. It's used for buffering frames and dma descriptors > etc. internally. I want to change the code to do this with the linker > script rather than manually assigning addresses in the source. But if > you grep for this address you'll find how it's used. Great. > > > Description says that pram is 32K, but: > > > > pram : ORIGIN = 0x200000, LENGTH = 16k > > Yeah, we're not exactly sure about this yet, but using 32k definitely > crashes the device. Thats fine too. If there were a documentation too. Isn't it possible to publish it? This firmware contains so many 'magic' registers. Best regards, Maxim Levitsky