> ...I'd be fairly sure that I shouldn't see this either ;
> 
> wlan0: authenticate with AP 00:1e:2a:61:7d:2c
> phy3: TX to low-level driver (len=30) FC=0x00b0 DUR=0x013a 
> A1=00:1e:2a:61:7d:2c A2=00:1b:fc:61:15:e1 A3=00:1e:2a:61:7d:2c
> wlan0: authenticate with AP 00:1e:2a:61:7d:2c
> phy3: TX to low-level driver (len=30) FC=0x00b0 DUR=0x013a 
> A1=00:1e:2a:61:7d:2c A2=00:1b:fc:61:15:e1 A3=00:1e:2a:61:7d:2c
> wlan0: authentication with AP 00:1e:2a:61:7d:2c timed out

For one you should turn off low-level debugging ;)

But the timeouts are another matter.

> ....on the other hand, if RX packets = 0 on the interface and that's
> true, then I probably would expect to see something like that...

Not entirely sure those stats are kept properly, but I think they are.

> Strangely, turning encryption off altogether in the AP itself and the
> linux setup here, made absolutely no difference apart from the fact
> that you then don't get debug messages about authentication attempts
> and failures.....and this make me think it's not
> authenticaion/encryption, seeing as the use (or not) of such security
> measures makes absolutly no difference to the tcpip layer - it seems
> dead. AFAICT, the radio (wireless) layer is doing all the right
> things, and the tcpip networking & routing>should< work, but and alas
> nothing happens...in fact, I doubt I've see ANY count on RX packets on
> wlan0 ever....

No, it's more likely a transmission bug. Can you enable b43 debugfs,
mount debugfs, and check the output of
cat /debug/b43/phy*/txstat

> One other curious thing -- if this machine is at 172.16.1.234 and the
> wireless AP is at 172.16.1.1 and I try to ping that end address across
> wlan0, the ping command itself performs very slowly, and I'll have to
> send it several rapidfire ^C breaks to get it to return. It might take
> it up to 10second to release and give me my prompt back, but in doing
> so it reports that it only got around to sending 3 packets in
> 10-15seconds worth of realtime - this is fairly bizarre behaviour, and
> no..there's no firewall in anyone's way.....

That's pretty normal if it cannot reverse-resolve the name.

johannes

Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
Bcm43xx-dev mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev

Reply via email to