On Sun, Apr 27, 2008 at 1:33 PM, Cheney Tang <[email protected]> wrote:
> On Fri, Apr 25, 2008 at 3:50 PM, Wessam Baghdadi <[email protected]> wrote:
> > hi,
> >
> > I've been looking into the out of tree driver for Airgo MIMO wireless
> > - the current status is detaied in a post on the linux wireless
> > mailing list http://article.gmane.org/gmane.linux.kernel.wireless.general/13784
> I am reversing engineer the airgo driver on Windows. There are some
> difficulty, especially
> no one disasseembly tool can deal with it fully. IDA can disassemble
> only some parts and others is not available. I try another method to
> reverse it. From the current partial result obtained, it is very
> different between Window driver and Jeff'
> Specifications(http://airgo.wdwconsulting.net/mymoin).
>
> >
> > Airgo is now part of Qualcomm, but surprisingly enough there is a
> > claim that one of the airgo engineers had developed linux drivers that
> > were held back.
> Yanbo had download a Airgo driver from internet and share it with me.
> The driver is for MIPS on Linux 2.4. It composes two files: one is a simple
> module and another is low level driver. IDA can disassemble the former
> and do the
> latter partly. I had try some objdump for MIPS to disassemble it, all failed.
> I guess the driver is for specific chipset.
> >
You can download the whole toolchains from the link
http://www.linksys.com/servlet/Satellite?childpagename=US%2FLayout&packedargs=c%3DL_Content_C1%26cid%3D1115416836002&pagename=Linksys%2FCommon%2FVisitorWrapper
IIRIC the type should be WRT54GX
Yanbo
>(http://www.linuxelectrons.com/news/mobile/airgo-and-asus-team-launch-faster-wired-laptop).
> >
> > He still works at Qualcomm and has a linkedin profile here
> > (http://www.linkedin.com/in/cfliu)
> >
> > Is it possible that someone could get in touch with him (or Qualcomm) to see:
> >
> > * if these mythical drivers were indeed developed, the status of the
> > drivers and the possibility of Qualcomm releasing the source?
> >
> > * the possibility of releasing the complete specifications of the chipset?
> >
> > thanks,
> >
> > wessam
> > _______________________________________________
> > prjmgr mailing list
> > [email protected]
> > http://driverdev.linuxdriverproject.org/mailman/listinfo/prjmgr
> >
>
________________________________________
From: YanBo [[email protected]]
Sent: Sunday, April 27, 2008 1:59 AM
To: Cheney Tang
Cc: Wessam Baghdadi; [email protected]; Williams, Jeffrey D.; wireless
Subject: Re: airgo wireless drivers
On Sun, Apr 27, 2008 at 1:33 PM, Cheney Tang <[email protected]> wrote:
> On Fri, Apr 25, 2008 at 3:50 PM, Wessam Baghdadi <[email protected]> wrote:
> > hi,
> >
> > I've been looking into the out of tree driver for Airgo MIMO wireless
> > - the current status is detaied in a post on the linux wireless
> > mailing list http://article.gmane.org/gmane.linux.kernel.wireless.general/13784
> I am reversing engineer the airgo driver on Windows. There are some
> difficulty, especially
> no one disasseembly tool can deal with it fully. IDA can disassemble
> only some parts and others is not available. I try another method to
> reverse it. From the current partial result obtained, it is very
> different between Window driver and Jeff'
> Specifications(http://airgo.wdwconsulting.net/mymoin).
>
> >
> > Airgo is now part of Qualcomm, but surprisingly enough there is a
> > claim that one of the airgo engineers had developed linux drivers that
> > were held back.
> Yanbo had download a Airgo driver from internet and share it with me.
> The driver is for MIPS on Linux 2.4. It composes two files: one is a simple
> module and another is low level driver. IDA can disassemble the former
> and do the
> latter partly. I had try some objdump for MIPS to disassemble it, all failed.
> I guess the driver is for specific chipset.
> >
I've already got a lot of it reverse engineered. I'd share the sources with you that I do have, but they will be inaccessable to me for at least 4-6 weeks. All I can offer you is to get in touch with Felix Fietkau (nbd) of the openwrt project, he has some source code for the agn300 that you can look at. You'll also need a cross-compiled toolchain for the mips chipset, not the standard objdump. Use crossdev for gentoo, or ubuntu has them in its repos.
Jeff
On Mon, Apr 28, 2008 at 12:22 AM, Williams, Jeffrey D. <[email protected]> wrote:
>
>
> ________________________________________
> From: YanBo [[email protected]]
> Sent: Sunday, April 27, 2008 1:59 AM
> To: Cheney Tang
> Cc: Wessam Baghdadi; [email protected]; Williams, Jeffrey D.; wireless
> Subject: Re: airgo wireless drivers
>
>
> On Sun, Apr 27, 2008 at 1:33 PM, Cheney Tang <[email protected]> wrote:
> > On Fri, Apr 25, 2008 at 3:50 PM, Wessam Baghdadi <[email protected]> wrote:
> > > hi,
> > >
> > > I've been looking into the out of tree driver for Airgo MIMO wireless
> > > - the current status is detaied in a post on the linux wireless
> > > mailing list http://article.gmane.org/gmane.linux.kernel.wireless.general/13784
> > I am reversing engineer the airgo driver on Windows. There are some
> > difficulty, especially
> > no one disasseembly tool can deal with it fully. IDA can disassemble
> > only some parts and others is not available. I try another method to
> > reverse it. From the current partial result obtained, it is very
> > different between Window driver and Jeff'
> > Specifications(http://airgo.wdwconsulting.net/mymoin).
> >
> > >
> > > Airgo is now part of Qualcomm, but surprisingly enough there is a
> > > claim that one of the airgo engineers had developed linux drivers that
> > > were held back.
> > Yanbo had download a Airgo driver from internet and share it with me.
> > The driver is for MIPS on Linux 2.4. It composes two files: one is a simple
> > module and another is low level driver. IDA can disassemble the former
> > and do the
> > latter partly. I had try some objdump for MIPS to disassemble it, all failed.
> > I guess the driver is for specific chipset.
> > >
>
> I've already got a lot of it reverse engineered. I'd share the sources with you that I do have, but they will be inaccessable to me for at least 4-6 weeks. All I can offer you is to get in touch with Felix Fietkau (nbd) of the openwrt project, he has some source code for the agn300 that you can look at. You'll also need a cross-compiled toolchain for the mips chipset, not the standard objdump. Use crossdev for gentoo, or ubuntu has them in its repos.
I had got the cross compiled toolchain from Yanbo. Does crossdev work on Fedora?
Is the MIPS hardware required? I have only X86 hardware.
I had achieve greatly on reversing the driver on Windows, though there
is no good tool to disassemble it fully yet. I think I can overcome
it. The driver on Windows is for F5D8010 chipset. Some of its action
are identical with the driver on MIPS, otheres not so.
I guess they are not for same chipset and they are the same family chipset.
>
> Jeff
>
On Sun, May 4, 2008 at 10:32 PM, Williams, Jeffrey D. <[email protected]> wrote:
>
>
> > I had got the cross compiled toolchain from Yanbo. Does crossdev work on Fedora?
> > Is the MIPS hardware required? I have only X86 hardware.
> > I had achieve greatly on reversing the driver on Windows, though there
> > is no good tool to disassemble it fully yet. I think I can overcome
> > it. The driver on Windows is for F5D8010 chipset. Some of its action
> > are identical with the driver on MIPS, otheres not so.
> > I guess they are not for same chipset and they are the same family chipset.
>
> No, MIPS hardware isn't required, I did all of the reverse engineering on X86 machines. And from what I can tell, all of the drivers for airgo chipsets, whether on linux-based routers or windows are generated from the more or less the same sources. In fact, the AGN300 drivers use most of the original code written for the agn100.
Did any one run the driver? Or is it possible to run it on x86 hardware?
>
> Jeff
>
> I had got the cross compiled toolchain from Yanbo. Does crossdev work on Fedora?
> Is the MIPS hardware required? I have only X86 hardware.
> I had achieve greatly on reversing the driver on Windows, though there
> is no good tool to disassemble it fully yet. I think I can overcome
> it. The driver on Windows is for F5D8010 chipset. Some of its action
> are identical with the driver on MIPS, otheres not so.
> I guess they are not for same chipset and they are the same family chipset.
No, MIPS hardware isn't required, I did all of the reverse engineering on X86 machines. And from what I can tell, all of the drivers for airgo chipsets, whether on linux-based routers or windows are generated from the more or less the same sources. In fact, the AGN300 drivers use most of the original code written for the agn100.
Jeff