[Fwd: Re: [Fwd: Linksys and Dlink wireless access points]]

Marcelo Tosatti marcelo at kvack.org
Thu Feb 1 13:35:13 EST 2007


Lilian,

Two things:

1) Setting MAC address with CMD_802_11_MAC_ADDRESS

I was unable to write to the MAC register using the CMD_MAC_REG_ACCESS
command:

http://lists.infradead.org/pipermail/libertas-dev/2006-November/000153.h
tml

Ronak, in the thread above there was no sufficient information back
from you about the MAC register access.

Can you please provide us with consistent information on that?

2) Speculation that RA=0 is the reason for the association failures 
with certain AP's.

Can you sniff the air and verify that a common wireless device sets RA
properly, thus associating successfully?

I don't have hardware for that at present moment.

On Tue, Jan 30, 2007 at 11:40:35AM -0500, Dan Williams wrote:
> Over to marcelo...

Content-Description: Forwarded message - Re: [Fwd: Linksys and Dlink wireless access points]
> From: "Lilian Walter" <lwalter42 at gmail.com>
> Date: Mon, 29 Jan 2007 20:11:11 -0800
> To: "Dan Williams" <dcbw at redhat.com>
> In-Reply-To: <1169812592.2936.1.camel at localhost.localdomain>
> Subject: Re: [Fwd: Linksys and Dlink wireless access points]
> DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta;
> 	h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references;
> 	b=rppavnpDojN5si6vLL4GN85c0DgPXSEvtZH2HsNRxuyULd11+2Y233BradAWKeWHjg6igT2Qzdsohdfv+Xi5BjiYQbQoOCbHhnD+hOqExSs57c2kANyPHasCrk1F66VzyxAYR3Kl+7r7iNoTW8w5882O3Bi2f2nfIpN/+tIOXr4=
> 
> Hi, Dan,
> 
> I'm speculating that the dlink and linksys do not like the fact that RA is
> 00:00:00:00:00:00.  The dump of the CMD_802_11_ASSOCIATE shows:
> 
> 08:57:57.029372 BSSID:00:19:5b:2c:16:4d (oui Unknown) DA:00:19:5b:2c:16:4d
> (oui Unknown) SA:00:00:00:00:00:00 (oui Ethernet) Authentication (Open
> System)-1: Succesful
>        0x0000:  b000 3a01 0019 5b2c 164d 0000 0000 0000  ..:...[,.M......
>        0x0010:  0019 5b2c 164d 2001 0000 0100 0000       ..[,.M........
> 08:57:57.029685 RA:00:00:00:00:00:00 (oui Ethernet) Acknowledgment
>        0x0000:  d400 0000 0000 0000 0000                 ..........
> 
> It never got to Authentication-2.
> 
> When I send a CMD_802_11_MAC_ADDRESS get to the Marvel firmware, it does
> return reasonable looking MAC address.  But the CMD_802_11_GET_HW_SPEC
> returns MAC address 0.
> 
> And then when I try to set the MAC address with the CMD_802_11_MAC_ADDRESS,
> the next command to read the MAC address returns 0.
> 
> So, how do I set the MAC address?
> 
> Lilian
> 
> On 1/26/07, Dan Williams <dcbw at redhat.com> wrote:
> >
> >On Thu, 2007-01-25 at 15:48 -0800, Lilian Walter wrote:
> >> I activated the wireless interface on my laptop and it is now
> >> associated with the dlink AP.  I can ping via the IP address assigned
> >> to it but tcpdump showed nothing while the ping was commencing.  And
> >> iwconfig scan found only the dlink.
> >
> >Scans won't always give you a complete picture of the environment,
> >because the card has to jump channels and dwell on a channel for a
> >period of time, and if the beacon from the AP doesn't get broadcast
> >during that dwell time, or collides with another frame and gets
> >scrambled, or whatever, it won't show up in your results.  To get a more
> >complete picture, you have to do a few scans over a short period of time
> >(a minute or so) and combine them.
> >
> >Dan
> >
> >> Maybe I need an external standalone wireless analyzer.
> >>
> >> Lilian
> >>
> >> On 1/25/07, Dan Williams <dcbw at redhat.com> wrote:
> >>         On Thu, 2007-01-25 at 14:15 -0800, Lilian Walter wrote:
> >>         > Thanks, Dan,
> >>         >
> >>         > That's great insight.  I'll see what I can do about timing
> >>         and
> >>         > authentication.
> >>
> >>         Can you get some 802.11 frame dumps with airsnort or some sort
> >>         of 802.11
> >>         frame grabber?  That would go a long way towards figuring out
> >>         why the
> >>         Marvell firmware and the APs can't talk to each other.
> >>
> >>         Dan
> >>
> >>         > Lilian
> >>         >
> >>         > On 1/25/07, Dan Williams <dcbw at redhat.com> wrote:
> >>         >         On Thu, 2007-01-25 at 11:22 -1000, Mitch Bradley
> >>         wrote:
> >>         >         > Dan -
> >>         >         >
> >>         >         > Any thoughts?
> >>         >
> >>         >         It looks like 2 is an "internal timer expired" on
> >>         association,
> >>         >         and 4 is
> >>         >         a few things.
> >>         >
> >>         >         >From Ronak:
> >>         >
> >>         >         The failure codes (3 and 4) that you are seeing are,
> >>         in fact,
> >>         >         associated
> >>         >         to one of the IEEE 802.11 failure codes that you
> >>         listed in
> >>         >         your email
> >>         >         below. The Marvell firmware actually maps them to
> >>         numbers 1 -
> >>         >         4
> >>         >         according to the table below before sending the
> >>         Assoc Command
> >>         >         Response
> >>         >         to the host.
> >>         >
> >>         >         Status Code Mapping for Association Response
> >>         currently
> >>         >         implemented in
> >>         >         FW:
> >>         >
> >>         >         IEEE Status Code                Marvell Status Code
> >>         >         0                       ->      0x0000
> >>         ASSOC_RESULT_SUCCESS
> >>         >         13                      ->      0x0004
> >>         >         ASSOC_RESULT_AUTH_REFUSED
> >>         >         14                      ->      0x0004
> >>         >         ASSOC_RESULT_AUTH_REFUSED
> >>         >         15                      ->      0x0004
> >>         >         ASSOC_RESULT_AUTH_REFUSED
> >>         >         16                      ->      0x0004
> >>         >         ASSOC_RESULT_AUTH_REFUSED
> >>         >         others          ->      0x0003 ASSOC_RESULT_REFUSED
> >>         >
> >>         ------------------------------------------------------------
> >>         >
> >>         >         0x0001 ASSOC_RESULT_INVALID_PARAMETERS
> >>         >         - Currently not used
> >>         >
> >>         >         0x0002 ASSOC_RESULT_TIMEOUT
> >>         >         - Internal Timer expired
> >>         >         for waiting Assoc Rsp from AP
> >>         >
> >>         >         This has been a very old code and is definitely not
> >>         the right
> >>         >         thing to
> >>         >         do. But we have already correctly this (i.e. the
> >>         host gets to
> >>         >         know the
> >>         >         actual status code that we get from the Association
> >>         response
> >>         >         frame from
> >>         >         the AP) in our later versions of the firmware for
> >>         other
> >>         >         products. I am
> >>         >         trying to correct this in the firmware scheduled to
> >>         be sent to
> >>         >         OLPC
> >>         >         later this month.
> >>         >
> >>         >         > email message attachment (Linksys and Dlink
> >>         wireless access
> >>         >         points)
> >>         >         > > -------- Forwarded Message --------
> >>         >         > > From: Lilian Walter < lwalter42 at gmail.com>
> >>         >         > > To: Mitch Bradley <wmb at firmworks.com>
> >>         >         > > Subject: Linksys and Dlink wireless access
> >>         points
> >>         >         > > Date: Thu, 25 Jan 2007 13:21:58 -0800
> >>         >         > >
> >>         >         > > Hi, Mitch,
> >>         >         > >
> >>         >         > > I have two APs and I cannot associate with
> >>         them.  However,
> >>         >         I
> >>         >         > > successfully associated with my neighbor's
> >>         Netgear and
> >>         >         watched
> >>         >         > > broadcast packets from his net.  None of them
> >>         have WEP or
> >>         >         other
> >>         >         > > security enabled.
> >>         >         > >
> >>         >         > > Do you have any thoughts why that maybe?  The
> >>         Linksys
> >>         >         returns the
> >>         >         > > status code of 2 (reserved) and the Dlink
> >>         returns 4
> >>         >         (reserved also).
> >>         >         > >
> >>         >         > > Thanks,
> >>         >         > > Lilian
> >>         >
> >>         >
> >>
> >>
> >
> >




More information about the libertas-dev mailing list