[driverloader] Now for the real problems...

Al Niessner Al.Niessner at gmx.net
Thu Feb 12 04:05:05 EST 2004

The license problem was just two transposed values in the MAC address.
My oopsy. It is now valid through March 12. Lets see if that is enough

Now the real problems start. I have an eMachine (Mandrake 9.2) which has
a built in BroadCom 94306 (AirForce) chip set -- or so lspci says. I
extracted the drivers from the "repair" disk and had 3 files of
interest: bcmwl5.inf, bcmwl5a.inf, and bcmwl.sys. I chose
bcmwl5.{inf,sys} as my driver pair -- I also downloaded the ones off the
net but they locked my machine up hard -- tried the 94306 not the 43xx.

A list of things that I consider significant effects from using the
LinuxAnt drivers:

1) the driver reported on the LinuxAnt configuration page is the bcm43xx
-- in conflict with lspci.

2) when the driver is loaded and the wireless is powered, the kernel CPU
usage is between 70 and 100 percent.

3) a spurious 8295A interrupt is detected on IRQ 7 (reported in

4) the NetGear access point never receives a broadcast from MAC address
-- can view this via the access points configuration page.

Ignoring all those weird items, I try and start the wireless anyway and
get the following message:

[root at charon]# ifup eth1
Determining IP information for eth1... failed; no link present. Check

Not too surprising because of the list above. In fact, given the list I
think I can firmly say that the wireless card is never transmitting at
the 802.11b frequency (2.422 GHz) -- the access point is an 802.11b.

Here is my ifcfg-eth1:

KEY="open 798220"

There is no specific question other than does some else see what is
wrong or know why this is not functioning?

Any explanation as to why I should not find 1-4 in the above list as
being weird would also be helpful.

Thanks in advance.

Al Niessner <Al.Niessner at gmx.net>

More information about the driverloader mailing list