Earlier I posted problems with the adm8211 driver detecting any wireless
networks and associating with an AP. I switched cards to an ath5k-based
card and some problems improved, but the problem associating is still there.
It seems I am able to associate with an AP (not just my AP) sporadically.
It seems that when it works, it works reliably, but that's only 10% of the
time, if that. I have turned on debugging on ath5 and on mac80211 but the
debug messages don't offer anything useful. The only thing I get is this:
May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
May 2 12:08:28 wlan0 direct probe responded
May 2 12:08:28 wlan0: authenticate with AP 00:30:bd:60:d3:44
May 2 12:08:28 wlan0: authenticate with AP 00:30:bd:60:d3:44
May 2 12:08:28 wlan0: authenticate with AP 00:30:bd:60:d3:44
May 2 12:08:28 wlan0: authentication with AP 00:30:bd:60:d3:44 timed out
I know the WEP key is fine, I know all my hardware is working properly.
Other devices on my wireless network work well and associate without any
troubles, probably because none of them use the ath5k driver. Another
interesting thing is the contents of /sys/class/net/wlan0/wireless:
beacon:0
crypt:0
fragment:0
level:0
link:0
misc:0
noise:0
nwid:0
retries:0
status:0x0
I haven't tried the wireless-testing tree because it causes problems with X
on my system. Even if it were to work, I don't want to go through 200
patches to find the one that works only to find it requires another 20 just
to apply it. :-)
I'm really anxious to get this up and running. I'd be happy to send my
kernel config, the output of lspci or any log in my system. If you send me
a patch, I'll apply it and return any information you want on it.
Google reports several people with these types of problems, but I can't find
any solutions. It would be great if someone with experience with that
driver could take the lead and help me resolve it. Again, ask anything and
I'll do it, even using the wireless-testing tree as long as I can have a
patch for the stable kernel.
Thanks in advance.
--
Alf
@
On Wednesday 06 May 2009 05:54:13 am Nick Kossifidis wrote:
> Can you please post the full dmesg output when ath5k loads ?
This is with debugging turned on.
========== dmesg ==========
15:43:35 ath5k 0000:00:07.0: PCI INT A -> GSI 18 (level, low) -> IRQ 18
15:43:35 ath5k 0000:00:07.0: registered as 'phy0'
15:43:35 phy0: Selected rate control algorithm 'minstrel'
15:43:35 ath5k phy0: Atheros AR2413 chip found (MAC: 0x78, PHY: 0x45)
15:43:39 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:43:39 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
15:43:39 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
15:43:39 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
15:43:50 wlan0: authenticate with AP 00:30:bd:60:d3:44
15:43:50 wlan0: authenticate with AP 00:30:bd:60:d3:44
15:43:50 wlan0: authenticate with AP 00:30:bd:60:d3:44
15:43:51 wlan0: authentication with AP 00:30:bd:60:d3:44 timed out
15:44:01 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:44:01 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
15:44:02 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
15:44:02 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
15:44:12 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:44:12 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
15:44:13 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
15:44:13 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
15:44:29 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:44:29 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:44:30 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
15:44:30 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
15:44:30 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
15:44:47 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:44:47 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:44:47 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
15:44:47 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
15:44:47 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
15:44:55 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:44:55 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
15:44:55 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
15:44:56 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
15:45:05 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:45:05 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:45:05 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:45:05 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
15:45:05 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
15:45:06 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
15:45:16 wlan0: authenticate with AP 00:30:bd:60:d3:44
15:45:16 wlan0: authenticate with AP 00:30:bd:60:d3:44
15:45:16 wlan0: authenticate with AP 00:30:bd:60:d3:44
15:45:17 wlan0: authenticate with AP 00:30:bd:60:d3:44
15:45:17 wlan0: authentication with AP 00:30:bd:60:d3:44 timed out
15:45:34 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:45:34 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:45:35 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
15:45:35 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
15:45:35 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
15:45:46 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:45:46 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
15:45:46 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
15:45:46 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
15:45:46 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
======== end dmesg ========
--
Alf
@
On Saturday 02 May 2009 03:14:54 pm Johannes Berg wrote:
> On Sat, 2009-05-02 at 12:18 -0600, Alf wrote:
> > It seems I am able to associate with an AP (not just my AP)
> > sporadically. It seems that when it works, it works reliably, but
> > that's only 10% of the time, if that. I have turned on debugging on
> > ath5 and on mac80211 but the debug messages don't offer anything
> > useful. The only thing I get is this:
> >
> > May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
> > May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
> > May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
> > May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
> > May 2 12:08:28 wlan0 direct probe responded
>
> What kernel are you using?
I did see your next message, but just to be sure:
2.6.29.1
Also, this is what lspci reports:
00:07.0 Ethernet controller: Atheros Communications Inc. AR2413 802.11bg NIC
(rev 01)
Subsystem: D-Link System Inc D-Link AirPlus G DWL-G510 Wireless PCI
Adapter(rev.B)
Flags: bus master, medium devsel, latency 96, IRQ 18
Memory at e0120000 (32-bit, non-prefetchable) [size=64K]
Capabilities: [44] Power Management version 2
Some Google searches in other lists said this card didn't work in Linux and
that I should use madwifi. I patched my kernel with the latest madwifi
snapshot but that didn't work either. I was able to associate with the the
AP, but I could never get a DHCP response. I also just tried the wireless-
testing tree and it didn't associate.
Does anyone know if this card is supported? Anybody know why this card
works sometimes but most of the time it can't associate with the AP?
Any help would be appreciated.
--
Alf
@
On Sat, 2009-05-02 at 12:18 -0600, Alf wrote:
> It seems I am able to associate with an AP (not just my AP) sporadically.
> It seems that when it works, it works reliably, but that's only 10% of the
> time, if that. I have turned on debugging on ath5 and on mac80211 but the
> debug messages don't offer anything useful. The only thing I get is this:
>
> May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
> May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
> May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
> May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
> May 2 12:08:28 wlan0 direct probe responded
What kernel are you using?
johannes
On Sat, 2009-05-02 at 23:14 +0200, Johannes Berg wrote:
> On Sat, 2009-05-02 at 12:18 -0600, Alf wrote:
>
> > It seems I am able to associate with an AP (not just my AP) sporadically.
> > It seems that when it works, it works reliably, but that's only 10% of the
> > time, if that. I have turned on debugging on ath5 and on mac80211 but the
> > debug messages don't offer anything useful. The only thing I get is this:
> >
> > May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 1
> > May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 2
> > May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 try 3
> > May 2 12:08:20 wlan0: direct probe to AP 00:30:bd:60:d3:44 timed out
> > May 2 12:08:28 wlan0 direct probe responded
>
> What kernel are you using?
Sorry, never mind. Shouldn't reply to email after have a bottle of wine
or so ;)
Sounds like your AP isn't responding -- can you try to move closer for a
start?
johannes
2009/5/6 Alf <[email protected]>:
>
> I tried it a few days ago and it still didn't work. =C2=A0Even the la=
test
> snapshot from madwifi didn't work. =C2=A0They only thing that has wor=
ked is
> ndiswrapper.
Hmm, do you use a driver from your vendor (the inf files etc from CD)
or a generic one from the net ?
It's possible that your vendor uses hardcoded settings on the driver
instead of EEPROM settings so tx power wont work as expected and
that's why you can't connect on the AP. Both ath5k and MadWiFi try to
use EEPROM settings, they don't use any hardcoded settings (not even
default values in case of a corrupted EEPROM -we should fix that in
the future).
--=20
GPG ID: 0xD21DB2DB
As you read this post global entropy rises. Have Fun ;-)
Nick
On Mon, May 4, 2009 at 6:49 PM, Tulio Magno Quites Machado Filho
<[email protected]> wrote:
> On May 02 09 15:53, Alf wrote:
>> Does anyone know if this card is supported? =A0Anybody know why this=
card
>> works sometimes but most of the time it can't associate with the AP?
>
> I think so (http://wireless.kernel.org/en/users/Drivers/ath5k#support=
edchips).
Yes, 2413 should work -- also can you try latest
wireless-testing tree? It has a boatload of updates for
ath5k. If something there works then we can perhaps narrow
it down a bit.
--=20
Bob Copeland %% http://www.bobcopeland.com
On Monday 04 May 2009 11:36:06 pm Bob Copeland wrote:
> Yes, 2413 should work -- also can you try latest
> wireless-testing tree? It has a boatload of updates for
> ath5k. If something there works then we can perhaps narrow
> it down a bit.
I tried it a few days ago and it still didn't work. Even the latest
snapshot from madwifi didn't work. They only thing that has worked is
ndiswrapper.
--
Alf
@
On Tuesday 12 May 2009 07:36:04 am Nick Kossifidis wrote:
> Hmm, do you use a driver from your vendor (the inf files etc from CD)
> or a generic one from the net ?
I downloaded them from DLink's website. They have two versions, I used the
older one. The most recent didn't work.
> It's possible that your vendor uses hardcoded settings on the driver
> instead of EEPROM settings so tx power wont work as expected and
> that's why you can't connect on the AP. Both ath5k and MadWiFi try to
> use EEPROM settings, they don't use any hardcoded settings (not even
> default values in case of a corrupted EEPROM -we should fix that in
> the future).
That's one possible explanation. I can send you the driver I'm using
through ndiswrapper to test your hypothesis. You can also send me a patch
that would set the power level to a hardcoded value instead of the EEPROM to
test it as well. Just let me know how I can be of help in finding and
fixing the problem.
--
Alf
@
2009/5/6 Alf <[email protected]>:
> On Monday 04 May 2009 11:36:06 pm Bob Copeland wrote:
>> Yes, 2413 should work -- also can you try latest
>> wireless-testing tree? =C2=A0It has a boatload of updates for
>> ath5k. =C2=A0If something there works then we can perhaps narrow
>> it down a bit.
>
> I tried it a few days ago and it still didn't work. =C2=A0Even the la=
test
> snapshot from madwifi didn't work. =C2=A0They only thing that has wor=
ked is
> ndiswrapper.
>
Can you please post the full dmesg output when ath5k loads ?
--=20
GPG ID: 0xD21DB2DB
As you read this post global entropy rises. Have Fun ;-)
Nick
On May 02 09 15:53, Alf wrote:
> Also, this is what lspci reports:
> 00:07.0 Ethernet controller: Atheros Communications Inc. AR2413 802.11bg NIC
> (rev 01)
> Subsystem: D-Link System Inc D-Link AirPlus G DWL-G510 Wireless PCI
> Adapter(rev.B)
> Flags: bus master, medium devsel, latency 96, IRQ 18
> Memory at e0120000 (32-bit, non-prefetchable) [size=64K]
> Capabilities: [44] Power Management version 2
>
> Does anyone know if this card is supported? Anybody know why this card
> works sometimes but most of the time it can't associate with the AP?
I think so (http://wireless.kernel.org/en/users/Drivers/ath5k#supportedchips).
Have you tried Johannes sugestion?
--
Tulio Magno