[rt2x00-users] Indirect register access failed

Antonio Quartulli ordex at ritirata.org
Fri Jul 2 06:32:20 AEST 2010




On gio, lug 01, 2010 at 01:15:39 -0700, Blaise Gassend wrote:
> Hi Antonio,
Hi Blaise

> 
> This sounds like a problem I have seen. 
> 
> My current theory is that the signal level reported by the rt2800 driver
> has its sign flipped. Hence, wpa_supplicant tries the AP with the lowest
> signal first. That AP has a really weak signal and is probably far. The
> probe doesn't make it to the AP, and hence no response comes back.
> wpa_supplicant doesn't blacklist the AP when there is a probe failure,
> so it keeps retrying weak APs.
> 
> This problem probably also happens if you are letting the kernel pick an
> AP, for the same reason.
> 
> Which version of wpa_supplicant are you using, or are you letting the
> kernel pick the AP?
> 

Actually I am connecting the device to the AP using the iw command.
So I explicitly targetted the device to a particular AP I chose. The
same AP obviously work with my working version of the driver..

Regards


> You could try explicitly connecting to an AP with a strong signal, the
> connection might work fine in that case.
> 
> Blaise
> 
> You could try my hacked version of wpa_supplicant, which can be found
> here:
> svn co
> https://code.ros.org/svn/ros-pkg/stacks/linux_networking/trunk/wpa_supplicant/wpa_supplicant@30516
> 
> Just run make in the wpa_supplicant directory, and a wpa_supplicant
> binary will show up there.
> 
> On Thu, 2010-07-01 at 19:24 +0200, Antonio Quartulli wrote:
> > On lun, giu 21, 2010 at 11:51:05 +0200, Antonio Quartulli wrote:
> > > > > > > > > The classic time out error:
> > > > > > > > > 
> > > > > > > > > Jun 18 21:54:57 eagle3 kernel: [   96.206662] phy0: device no longeridle - scanning
> > > > > > > > > Jun 18 21:54:58 eagle3 kernel: [   97.161141] wlan0: direct probe to 00:22:3f:03:81:dd (try 1)
> > > > > > > > > Jun 18 21:54:58 eagle3 kernel: [   97.361146] wlan0: direct probe to 00:22:3f:03:81:dd (try 2)
> > > > > > > > > Jun 18 21:54:58 eagle3 kernel: [   97.561100] wlan0: direct probe to 00:22:3f:03:81:dd (try 3)
> > > > > > > > > Jun 18 21:54:59 eagle3 kernel: [   97.761324] wlan0: direct probe to 00:22:3f:03:81:dd timed out
> > > > > > > > > Jun 18 21:54:59 eagle3 kernel: [   97.766249] phy0: device now idle
> > > > > > 
> > > > > > First possibility: the probe requests are not sent out at all
> > > > > > Second possibility: the probe requests are sent out but not received by the AP
> > > > > > 
> > > > > > One way to find that out would be to sniff the wireless traffic with a second
> > > > > > wireless card and check if the rt2800 card sends out probe requests and if
> > > > > > the AP answers them. Do you think you can do that?
> > > > > 
> > Hi all!
> > Sorry for writing only once aweek but in this period I'm a bit in
> > trouble with the university exams :P
> > 
> > About the card, I did some other tests (with the actual version of
> > wireless-testing) and as result I got that: the first time i was not
> > able to associate to the AP and i reported the same timeout as before.
> > Then i turned on wireshark and magically the device has associated, but
> > in wireshark i saw A LOT of retransmission by the AP!
> > I tried to use some application, like dhcp for first and then a simple
> > ping and I got that every packet (dhcp offer, echo reply) is sent many
> > many time..
> > 
> > Here you can find a log of the association scenario:
> > http://www.ritirata.org/ordex/association.cap
> > 
> > Regards,
> > 
> > 
> 
> 
> 
> _______________________________________________
> users mailing list
> users at rt2x00.serialmonkey.com
> http://rt2x00.serialmonkey.com/mailman/listinfo/users_rt2x00.serialmonkey.com

-- 
Antonio Quartulli

..each of us alone is worth nothing..
Ernesto "Che" Guevara




More information about the users mailing list