Re: [Open1x-xsupplicant] Packets sent by xsupplicant not using proper MAC address

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 454-5900

Patrick -

Which driver are you using?  The dmesg output almost looks like the
"linux-wlan.org" driver.  If that is the case, I might suggest you look at
the hostap driver instead.  Few (if any) people have been able to get the
linux-wlan.org driver to work correctly.  I think this may be in part that
the linux-wlan.org driver doesn't use the Linux wireless extensions.

For the hostap driver, check out hostp.epitest.fi.

----------
Chris "Spanky" Hessing                 Networking
				       University of Utah
Chr...@ut...		       Marriott Library

Friends are people that know everything about you, but love you anyway.

On Fri, 26 Mar 2004, Patrick McNamara wrote:

> Here is the relevant dmesg output info (from the point a which I plug the card
> in).  As I noted, this is a prism2 based card, not Atheros. Something you said
> did sound familiar though.  I have been sniffing from a second machine and I
> have not been seeing the response packets on that machine.  I see the AP send
> the request, but never the response.  I'll have to look into that.
> 
> Patrick M
> 
> prism2cs_init: prism2_cs.o: 0.2.1-pre20 Loaded
> prism2cs_init: dev_info is: prism2_cs
> prism2_cs: index 0x01: Vcc 3.3, irq 3, io 0x0100-0x013f
> PDA Read from 0x00390000 in NV space.
> PDA Read from 0x00390000 in NV space.
> ident: nic h/w: id=0x8008 1.0.1
> ident: pri f/w: id=0x15 0.3.0
> ident: sta f/w: id=0x1f 0.8.3
> MFI:SUP:role=0x00:id=0x01:var=0x00:b/t=1/1
> CFI:SUP:role=0x00:id=0x02:var=0x01:b/t=1/1
> PRI:SUP:role=0x00:id=0x03:var=0x01:b/t=1/2
> STA:SUP:role=0x00:id=0x04:var=0x01:b/t=1/6
> PRI-CFI:ACT:role=0x01:id=0x02:var=0x01:b/t=1/1
> STA-CFI:ACT:role=0x01:id=0x02:var=0x01:b/t=1/1
> STA-MFI:ACT:role=0x01:id=0x01:var=0x01:b/t=1/1
> Prism2 card SN: 99SA01000000
> p80211knetdev_hard_start_xmit: Tx attempt prior to association, frame dropped.
> p80211knetdev_hard_start_xmit: Tx attempt prior to association, frame dropped.
> p80211knetdev_hard_start_xmit: Tx attempt prior to association, frame dropped.
> p80211knetdev_hard_start_xmit: Tx attempt prior to association, frame dropped.
> linkstatus=CONNECTED
> linkstatus=AP_CHANGE (unhandled)
> 
> --- Terry Simons <Terry.Simons@m.cc.utah.edu> wrote:
> > Patrick,
> > 
> > This is a problem with the MADWifi driver, but it is cosmetic.
> > 
> > You aren't actually sending out a frame with a bad MAC (at least not 
> > the one you're seeing, if at all).
> > 
> > Sniff the connection from a different computer, instead of the one 
> > containing the wireless card, and you should see that the destination 
> > from your laptop is not actually aa:aa:03:00:00:00.
> > 
> > I reported this bug quite some time ago... are you using an old version 
> > of the driver (I don't know if this actually ever got fixed).
> > 
> > Also, make sure you are using the latest MADWifi CVS to do your testing 
> > (I.E. don't use a gentoo snapshot or anything like that).
> > 
> > Around the time I reported this bug, there was also a problem with the 
> > driver encrypting all of the frames leaving the device, but the 802.1x 
> > frames need to get to the AP unencrypted.
> > 
> > I suspect that if you do a sniff from another laptop, you will see one 
> > of the following:
> > 
> > Your frames are indeed correct OR your frames are seemingly not coming 
> > out of the card at all.
> > 
> > If you send me the driver output from dmesg I can tell you whether or 
> > not you're using a version of the driver with that bug.
> > 
> > You *SHOULD* see:
> > 
> > ath_hal: 0.9.8.6
> > wlan: 0.7.3.1 BETA
> > ath_pci: 0.8.5.4 BETA
> > 
> > If you see anything earlier than that in your dmesg output then you are 
> > definitely using an old driver, and will need to upgrade, as well as 
> > apply the MADWifi patch in the xsupplicant CVS to get the driver 
> > working properly.
> > 
> > Also, the patch in the xsupplicant CVS sort of breaks manual key 
> > sets... in order to set a WEP key you will need to issue the following 
> > commands in the order listed (with whatever key you desire):
> > 
> > iwconfig key 1234567890
> > iwconfig enc on
> > 
> > - Terry
> > 
> > On Mar 26, 2004, at 11:50 AM, Patrick McNamara wrote:
> > 
> > > Nope, haven't been able to authenticate yet.  The access point seems 
> > > to just
> > > ignore the identity response.  It keeps resending a identity request 
> > > packet.
> > > The AP sends the identity request to the correct MAC and I'm guessing 
> > > it
> > > ignores the response as coming from the wrong client because of the 
> > > bad MAC.
> > > I've reproduced this on two different cards, both prism2 based (D-Link 
> > > Air
> > > DWL-650, Airlink+ WLC010).
> > >
> > > Patrick M
> > >
> > > --- Chris Hessing <Chr...@ut...> wrote:
> > >> If I were to guess, I would say that you are using an Atheros based 
> > >> card.
> > >> We have seen this happen with the Atheros based cards, however, the
> > >> packets that get sent are formatted correctly.  I believe this has
> > >> something to do with how the Madwifi driver presents packets to higher
> > >> layers.  Have you been able to authenticate anyway?
> > >>
> > >> ----------
> > >> Chris "Spanky" Hessing                 Networking
> > >> 				       University of Utah
> > >> Chr...@ut...		       Marriott Library
> > >>
> > >> Friends are people that know everything about you, but love you 
> > >> anyway.
> > >>
> > >>
> > >>
> > >>
> > >>
> > >> On Fri, 26 Mar 2004, Patrick McNamara wrote:
> > >>
> > >>> I have recently downloaded an compiled xsupplicant.  Snooping the 
> > >>> traffic
> > >> on
> > >>> wlan0 with ethereal shows that all outbound packets are being sent 
> > >>> with a
> > >> MAC
> > >>> of aa:aa:03:00:00:00 instead of the wireless nic's MAC of
> > >> 00:90:96:24:eb:0b.
> > >>> This behaviour exists both in 0.8b and the latest code from CVS.  In 
> > >>> 0.8b,
> > >> I
> > >>> have verified that the MAC address is correct in the frame sent to 
> > >>> dnet
> > >>> eth_send call.  I have not traced through the latest CVS code yet.  
> > >>> Any
> > >>> thoughts on this behavior?  Has anyone else seen it?
> > >>>
> > >>> Patrick M
> > >>>
> > >>> __________________________________
> > >>> Do you Yahoo!?
> > >>> Yahoo! Finance Tax Center - File online. File on time.
> > >>> http://taxes.yahoo.com/filing.html
> > >>>
> > >>>
> > >>> -------------------------------------------------------
> > >>> This SF.Net email is sponsored by: IBM Linux Tutorials
> > >>> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> > >>> GenToo technologies. Learn everything from fundamentals to system
> > >>> administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> > >>> _______________________________________________
> > >>> Open1x-xsupplicant mailing list
> > >>> Ope...@li...
> > >>> https://lists.sourceforge.net/lists/listinfo/open1x-xsupplicant
> > >>>
> > >>
> > >
> > >
> > > __________________________________
> > > Do you Yahoo!?
> > > Yahoo! Finance Tax Center - File online. File on time.
> > > http://taxes.yahoo.com/filing.html
> > >
> > >
> > > -------------------------------------------------------
> > > This SF.Net email is sponsored by: IBM Linux Tutorials
> > > Free Linux tutorial presented by Daniel Robbins, President and CEO of
> > > GenToo technologies. Learn everything from fundamentals to system
> > > administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
> > > _______________________________________________
> > > Open1x-xsupplicant mailing list
> > > Ope...@li...
> > > https://lists.sourceforge.net/lists/listinfo/open1x-xsupplicant
> > 
> 
> 
> __________________________________
> Do you Yahoo!?
> Yahoo! Finance Tax Center - File online. File on time.
> http://taxes.yahoo.com/filing.html
> 