Faulty 14e4:4315 (BCM4312)?

[Rafał Miłecki]

W dniu 8 czerwca 2011 23:08 użytkownik Gábor Stefanik
<netrolller.3d at gmail.com> napisał:
> 2011/6/8 Rafał Miłecki <zajec5 at gmail.com>:
>> W dniu 8 czerwca 2011 22:37 użytkownik Gábor Stefanik
>> <netrolller.3d at gmail.com> napisał:
>>> 2011/6/8 Rafał Miłecki <zajec5 at gmail.com>:
>>>> Hey,
>>>>
>>>> I've just discovered I can not connect with my 14e4:4315 to my home network.
>>>>
>>>> dmesg:
>>>> [  119.744050] b43-phy0: Loading firmware version 478.104 (2008-07-01 00:50:23)
>>>> [  119.746744] b43-phy0 debug: b2062: Using crystal tab entry 19200 kHz.
>>>> [  121.130067] b43-phy0 debug: Chip initialized
>>>> [  121.130348] b43-phy0 debug: 64-bit DMA initialized
>>>> [  121.130416] b43-phy0 debug: QoS enabled
>>>> [  121.139857] b43-phy0 debug: Wireless interface started
>>>> [  121.145095] b43-phy0 debug: Adding Interface type 2
>>>> [  121.145605] ADDRCONF(NETDEV_UP): wlan0: link is not ready
>>>> [  128.295371] wlan0: authenticate with 00:23:69:c2:65:06 (try 1)
>>>> [  128.300285] wlan0: authenticated
>>>> [  128.300319] wlan0: associate with 00:23:69:c2:65:06 (try 1)
>>>> [  128.303709] wlan0: RX AssocResp from 00:23:69:c2:65:06 (capab=0x411
>>>> status=0 aid=3)
>>>> [  128.303715] wlan0: associated
>>>> [  128.304711] ADDRCONF(NETDEV_CHANGE): wlan0: link becomes ready
>>>> [  128.429557] b43-phy0 debug: Using hardware based encryption for
>>>> keyidx: 0, mac: 00:23:69:c2:65:06
>>>> [  128.429611] b43-phy0 debug: Using hardware based encryption for
>>>> keyidx: 2, mac: ff:ff:ff:ff:ff:ff
>>>> [  132.208036] ieee80211 phy0: wlan0: No probe response from AP
>>>> 00:23:69:c2:65:06 after 500ms, disconnecting.
>>>>
>>>> wpa_supplicant fills my console with:
>>>>> ioctl[SIOCSIWENCODEEXT]: Invalid argument
>>>>> ioctl[SIOCSIWENCODEEXT]: Invalid argument
>>>>> Trying to associate with 00:23:69:c2:65:06 (SSID='zajec_poznan_160' freq=2427 MHz)
>>>>> Associated with 00:23:69:c2:65:06
>>>>> WPA: Key negotiation completed with 00:23:69:c2:65:06 [PTK=CCMP GTK=CCMP]
>>>>> CTRL-EVENT-CONNECTED - Connection to 00:23:69:c2:65:06 completed (auth) [id=0 id_str=]
>>>>> CTRL-EVENT-DISCONNECTED bssid=00:23:69:c2:65:06 reason=0
>>>>
>>>> It seems it was already reported, for example in:
>>>> "No probe response from AP <address> after 500ms, disconnecting." by
>>>> Miklos Vajna on Thu, 14 Jan 2010 17:29:20 -0800.
>>>> http://www.mail-archive.com/bcm43xx-dev@lists.berlios.de/msg09481.html
>>>>
>>>> What's really surprising I can not connect to this network also when
>>>> using wl 5.100.82.38!
>>>>
>>>> Did anyone of you already got some faulty BCM4312 hardware? Did it
>>>> behave like this? I'm surprised than scanning works, but connection is
>>>> quickly dropped. It does not really sound like a faulty hardware.
>>>
>>> Chances are, your card can happily communicate with the AP using CCK,
>>> but once it tries to do OFDM, it fails.
>>> Is the AP close enough for OFDM communication?
>>
>> My AP is about 2 meters from the card, antennas are attached OFC.
>>
>> --
>> Rafał
>>
>
> You are too close to the AP then - the AP's signal is jamming the
> card's receiver (or vice versa). Again, CCK is more resilient in this
> aspect than OFDM, hence no problems until association.

I believe that even if one is jamming the other, connection should
still work. I can accept lower performance, packet loss, but not
disconnection at all.

Actually after receiving tip about OFDM from you, I went back to the
machine and checked for the news. It was running wpa_supplicant for
all this time. I'm glad to say dmesg said the same thing you did. TX
transmission errors :) The old known friend from N-PHY cards.

Forcing rate to 1M workarounded this issue.

-- 
Rafał