Please note: This project is no longer active. The website is kept online for historic purposes only.
If you´re looking for a Linux driver for your Atheros WLAN device, you should continue here .

Ticket #847 (new defect)

Opened 15 years ago

Last modified 15 years ago

Madwifi on Ubuntu going terribly wrong

Reported by: chase.voisin+madwifi@gmail.com Assigned to:
Priority: major Milestone:
Component: madwifi: driver Version: trunk
Keywords: driver unloaded Cc:
Patch is attached: 0 Pending:

Description (Last modified by mrenzmann)

I've got a 3com 3CRPAG175 with an Atheros chipset on it that is shown to work with Madwifi. I've checked the documentation and tried it out with the default Ubuntu Dapper drivers and was htexperiencing a few scanning problems. I decided to upgrade to the newest drivers.

I pulled the most recent copy with subversion (sudo svn checkout svn.madwifi.org/trunk madwifi-ng)

make
sudo make install
depmod -a
modprobe ath_pci

It goes just fine but if I pull and ifconfig I have no wifi0 or ath0 dev devices. I check the dmesg and here is the output.

[17180544.092000] ath_hal: 0.9.17.2 (AR5210, AR5211, AR5212, RF5111, RF5112, RF2413, RF5413)
[17180544.092000] ath_rate_sample: 1.2 (svn r1705)
[17180544.100000] ath_pci: 0.9.4.5 (svn r1705)
[17180544.100000] ACPI: PCI Interrupt 0000:04:00.0[A] -> GSI 19 (level, low) -> IRQ 193
[17180544.404000] wifi0: 11a rates: 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
[17180544.404000] wifi0: 11b rates: 1Mbps 2Mbps 5.5Mbps 11Mbps
[17180544.404000] wifi0: 11g rates: 1Mbps 2Mbps 5.5Mbps 11Mbps 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
[17180544.404000] wifi0: turboA rates: 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
[17180544.404000] wifi0: H/W encryption support: WEP AES AES_CCM TKIP
[17180544.404000] wifi0: mac 5.6 phy 4.1 5ghz radio 1.7 2ghz radio 2.3
[17180544.404000] wifi0: Use hw queue 1 for WME_AC_BE traffic
[17180544.404000] wifi0: Use hw queue 0 for WME_AC_BK traffic
[17180544.404000] wifi0: Use hw queue 2 for WME_AC_VI traffic
[17180544.404000] wifi0: Use hw queue 3 for WME_AC_VO traffic
[17180544.404000] wifi0: Use hw queue 8 for CAB traffic
[17180544.404000] wifi0: Use hw queue 9 for beacons
[17180544.408000] wifi0: Atheros 5212: mem=0x8a000000, irq=193
[17180544.700000] ACPI: PCI interrupt for device 0000:04:00.0 disabled
[17180544.700000] ath_pci: driver unloaded
[17180544.704000] ath_rate_sample: unloaded
[17180544.704000] ath_hal: driver unloaded

I've been fighing with this for about 3 hours now, do you have any idea of what I can do to fix this?

Attachments

modprobe (1.4 kB) - added by anonymous on 08/29/06 20:53:34.
better looking version of the dmesg output since the formatting came in poorly

Change History

08/29/06 20:53:34 changed by anonymous

  • attachment modprobe added.

better looking version of the dmesg output since the formatting came in poorly

(follow-up: ↓ 2 ) 08/30/06 09:08:29 changed by mrenzmann

  • description changed.

I'm quite sure that this is a problem caused by Ubuntu (some of their scripts or whatever), not by the driver. The dmesg dump above indicates that the driver gets loaded and successfully initializes your card, but then immediately get unloaded again (last 4 lines). The driver cannot unload itself, so the culprit must be something that sits "behind the scenes". And without the driver being loaded at the time you run ifconfig/iwconfig, you of course don't see something like wifi0 or ath0.

(in reply to: ↑ 1 ; follow-up: ↓ 3 ) 03/17/07 05:51:58 changed by mattf

Replying to mrenzmann:

I have the same problem, also on Ubuntu. I can't find anything funky in /etc/modprobe.d or /etc/defaults/ or /etc/modules... any clues as to what might be happening, or where to begin looking?

This worked fine 'till a couple of weeks ago, and I just re-compiled and re-installed the latest .9.2.1. I suppose Ubuntu must have upgraded the kernel for Dapper and that messed me up.

(in reply to: ↑ 2 ) 03/17/07 07:19:04 changed by anonymous

Replying to mrenzmann:

I have what might be a useful clue. When trying to insert the module, my dmesg says this: [17180574.944000] wifi0: Atheros 5212: mem=0xe8900000, irq=193 [17180575.340000] ACPI: PCI interrupt for device 0000:00:0b.0 disabled

but inside of /proc/interrupts, I see this: 193: 2421 IO-APIC-level eth1

So, it seems that the irq 193 is already snagged by a network card. I'm not sure if this matters, or not, but perhaps it might be a clue to smarter people than me. ;-)

03/21/07 05:56:37 changed by proski

PCI device drivers should be able to share the irq. MadWifi certainly can do it. But even if there is in irq conflict, I don't see how it would lead to driver unloading. Whatever happens in the driver, unloading can only be only requested by the userspace. I don't think it's a MadWifi bug.