[Click] Feedback from madwifi in Click

[Javier Sánchez]

Hello,

I see now what is your problem now. You use stripped drivers and feedback
loop.

I suppose that feedback loop were generated artificially by madwifi stripped
driver (but certainly i am not completally sure) to pass stadistics to click
in form of ExtraEncap packet. Obviously this feature may not be implemented
on regular drivers, they are not focus on click and only pass this
information to kernel in other ways than ExtraEncap packet.

Salutations
Javier Sánchez


On Wed, Feb 27, 2008 at 4:13 PM, Peter De Cleyn <peter.decleyn at gmail.com>
wrote:

> Hi David,
>
> Thanks for trying! I'll try to see where this support was dropped and will
> address it to the madwifi mailinglist too. I already contacted John
> Bicket,
> the original author of the madwifi-stripped driver and who ported the
> functionality in to the regular driver.
>
> I'll keep you posted
>
> Peter
>
> On Wed, Feb 27, 2008 at 4:04 PM, David Fuste <dfuste at ac.upc.edu> wrote:
>
> >  Hi Peter,
> > I understand...but unfortunately, I have checked your problem and I get
> > the same results as you. No packets are received via output port 1 in
> the
> > FilterTX element, neither in monitor nor in adhoc mode. In fact, I never
> > realized of that problem before because I only need feedback from
> received
> > packets and not from transmitted packets. It seems a problem of the new
> > madwifi drivers, which do not send back the packets transmitted by
> click.
> > Please, let me know if you get some solution...
> >
> > Good luck!
> >
> > david
> >
> > Peter De Cleyn wrote:
> >
> > Hi David,
> > I know I can find it in these headers, actually  all I need for now is
> to
> > know if a packet was successfully transmitted (using the FilterFailures
> > etc). I have running Click scripts for this, but I used to use them in
> > combination with the older madwifi drivers, so I had to set up a ath0raw
> > interface at that time.
> >
> > Now in the current driver, this functionality should be available
> through
> > the monitor VAP, but I do not receive any packets back from the the
> system.
> > I hope I make myself clear. Normally this is what happens:
> > a packet to be transmitted via the wireless interface is encapsulated
> > (Prism2Encap/radiotapencap/..) and delivered to the ToDevice element.
> > Lateron, the same packet will be received back from the FromDevice
> element,
> > with the Prism2/ratiota.... headers filled with the transmission
> > information. It this last packet I miss. I don't receive any feedback
> > packets.
> >
> > If you have a setup running which does gets these feedback packets,
> could
> > you describe me what you did to get it running? I'm using a CVS click,
> an
> > unpatched 2.6.22 kernel, madwifi 0.9.3.1.99-2.
> >
> > On Wed, Feb 27, 2008 at 10:19 AM, David Fuste <dfuste at ac.upc.edu> <
> dfuste at ac.upc.edu> wrote:
> >
> >
> >
> >  Hi Peter,
> > what do you mean exactly for the wifi feedback mechanism?
> > If you mean SSI signal, SSI noise, data rate, etc...you can obtain them
> > in the radiotap, prism or atheros headers, added by madwifi to the top
> > of the packet.
> >
> > david
> >
> > Peter De Cleyn wrote:
> >
> >
> >  Has anyone recently used the Wifi Feedback mechanism in Click in
> >
> >
> >  combination
> >
> >
> >  with the recent MadWifi drivers?
> >
> > Following the information from the MadWifi site I am able to contruct a
> > monitor VAP which allows me to send and receive raw 802.11 frames.
> >
> >
> >  However I
> >
> >
> >  don't seem to get any feedback information anymore from the wireless
> >
> >
> >  driver.
> >
> >
> >  Any hints on how to (re)activate this behaviour are welcome. In the
> past
> >
> >
> >  I
> >
> >
> >  used the raw device with previous version of the Madwifi driver, but I
> >
> >
> >  seem
> >
> >
> >  not succeed in compiling this for the more recent kernels.
> >
> > regards
> >
> > Peter
> > _______________________________________________
> > click mailing listclick at amsterdam.lcs.mit
> .eduhttps://amsterdam.lcs.mit.edu/mailman/listinfo/click
> >
> >        --
> > ========================================================================
> >  0 0 0  David Fusté i Vilella       |    Dept. of Comp. Architecture
> >  0 0 0  Ph.D. Student               |    UPC-Campus Nord, D6-001
> >  0 0 0  email: dfuste at ac.upc.edu    |    C/ Jordi Girona, 1-3
> >  U P C  phone: 0034 93 401 67 83    |    08034 Barcelona - SPAIN
> >                www: http://personals.ac.upc.edu/dfuste/
> > ========================================================================
> >
> >
> >
> >
> >  _______________________________________________
> > click mailing listclick at amsterdam.lcs.mit
> .eduhttps://amsterdam.lcs.mit.edu/mailman/listinfo/click
> >
> >
> > --
> > ========================================================================
> >   0 0 0  David Fusté i Vilella       |    Dept. of Comp. Architecture
> >   0 0 0  Ph.D. Student               |    UPC-Campus Nord, D6-001
> >   0 0 0  email: dfuste at ac.upc.edu    |    C/ Jordi Girona, 1-3
> >   U P C  phone: 0034 93 401 67 83    |    08034 Barcelona - SPAIN
> >                 www: http://personals.ac.upc.edu/dfuste/
> > ========================================================================
> >
> >
> _______________________________________________
> click mailing list
> click at amsterdam.lcs.mit.edu
> https://amsterdam.lcs.mit.edu/mailman/listinfo/click
>