I have got the Broadcom WLAN Chip BCM943111MCG in my PC.
"b43-fwcutter" installed some firmware.
But it works in Debian SID with ndiswrapper only so far.
But, you know, ndiswrapper imports parts of WINDOWS into Debian. I don't
want that.

Searching the web have I found, that there are two new Linux drivers:

* b43
* b43legacy

I didn't find these in a Debian kernel source.
So I have built my new kernel from 2.6.24
<http://kernel.org/pub/linux/kernel/v2.6/linux-2.6.24.tar.bz2> from
kernel.org

Now it seems, that the b43 kernel module does not serve the hardware,
but ssb does:
###$ lshw -C net description: Network controller
product: BCM94311MCG wlan mini-PCI
vendor: Broadcom Corporation
physical id: 0
bus info: pci@0000:03:00.0
version: 02
width: 64 bits
clock: 33MHz
capabilities: bus_master cap_list
configuration: driver=b43-pci-bridge latency=0 module=ssb

^^^^^^^^
If I understand things correctly, then should the module be: module=b43
rather.

^^^^^^^^
Googling tells me, that this problem is caused by "udev" and could be
corrected by an entry in the file

* z75_persistent-net.rules, which I suppose, is in Debian
/etc/udev/rules.d/z25_persistent-net-rules, which looks in my
system after trials of modifications like this:

-------------------------------------------------------
SUBSYSTEM=="net", DRIVERS=="?*", ATTR{address}=="00:1b:24:aa:41:9d",
NAME="eth0"

SUBSYSTEM=="net", DRIVERS=="?*", ATTR{address}=="00:00:00:1a:73:a8",
ATTRS{type}=="1", NAME="wlan0"

SUBSYSTEM=="net", DRIVERS=="?*", ATTR{address}=="00:1a:73:a8:0f:1f",
ATTRS{type}=="1", NAME="eth1"
---------------------------------------------------------

When googling deeper, I run into discrepancies:
Is the adding of ATTRS{type}=="1", correct
or should it be ATTR{type}=="1" without "S"?

And the names of the device seem to vary from distribution to distribution:
from "wlan0" ... "eth2".
What is correct in the Debian environment?

Something is still wrong with my system.
Does anybody have some clue, how to get this Wireless LAN up and running
in Debian?

Hugo


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

On Tue, Jan 29, 2008 at 11:11:19 -0700, wauhugo AT yahoo DOT com wrote:
> I have got the Broadcom WLAN Chip BCM943111MCG in my PC.
> "b43-fwcutter" installed some firmware.
> But it works in Debian SID with ndiswrapper only so far.
> But, you know, ndiswrapper imports parts of WINDOWS into Debian. I don't
> want that.
>
> Searching the web have I found, that there are two new Linux drivers:
>
> * b43
> * b43legacy
>
> I didn't find these in a Debian kernel source.

$ ls -ld /usr/src/linux-source-2.6.24/drivers/net/wireless/b43*
drwxrwxr-x 2 root src 4096 2008-01-24 23:58 /usr/src/linux-source-2.6.24/drivers/net/wireless/b43
drwxrwxr-x 2 root src 4096 2008-01-24 23:58 /usr/src/linux-source-2.6.24/drivers/net/wireless/b43legacy

I have the Debian linux sources version 2.6.24-1 installed.

> So I have built my new kernel from 2.6.24
> <http://kernel.org/pub/linux/kernel/v2.6/linux-2.6.24.tar.bz2> from
> kernel.org

I think neither the b43 nor the b43legacy is the correct choice; they
both seem to be intended for vendor-ID 0x4243 and device-ID 0x812. (See
below for how to find out these identifiers for your chip.)

> Now it seems, that the b43 kernel module does not serve the hardware,
> but ssb does:
> ###$ lshw -C net description: Network controller
> product: BCM94311MCG wlan mini-PCI
> vendor: Broadcom Corporation
> physical id: 0
> bus info: pci@0000:03:00.0
> version: 02
> width: 64 bits
> clock: 33MHz
> capabilities: bus_master cap_list
> configuration: driver=b43-pci-bridge latency=0 module=ssb
>
> If I understand things correctly, then should the module be: module=b43
> rather.

According to /usr/share/misc/pci.ids your chip should have the vendor-ID
0x14e4 and the prodcut-ID 0x4311. Check with "lspci -nn" if this is
correct; you should see these numbers listed as "[14e4:4311]" right
after the name of the card/chip. This would mean that you need the
bcm43xx module if I read the kernel sources correctly. (Note: I don't
have any direct experience with this particular chip/card.)

I am not sure what to make of this ssb module, though. Part of its
description reads: "Support for the Sonics Silicon Backplane bus.
You only need to enable this option, if you are configuring a kernel for
an embedded system with this bus. It will be auto-selected if needed in
other environments." For now I would assume that the kernel knows what
it is doing and leave this module loaded.

So, make sure that you have the bcm43xx module compiled and try to load
it. If there are any problems use "modprobe -v bcm43xx" to get more
details about what is going on.

> Googling tells me, that this problem is caused by "udev" and could be
> corrected by an entry in the file
>
> * z75_persistent-net.rules, which I suppose, is in Debian
> /etc/udev/rules.d/z25_persistent-net-rules, which looks in my
> system after trials of modifications like this:

Leave udev and this file alone for now. You have to find the right
module for the broadcom chip and make it load correctly. Once that is
settled you can fine-tune the naming your various network devices.

--
Regards, | http://users.icfo.es/Florian.Kulzer
Florian |


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Il giorno Tue, 29 Jan 2008 20:46:56 +0100
Florian Kulzer <florian.kulzer+debian@icfo.es> ha scritto:

> According to /usr/share/misc/pci.ids your chip should have the vendor-ID
> 0x14e4 and the prodcut-ID 0x4311. Check with "lspci -nn" if this is
> correct; you should see these numbers listed as "[14e4:4311]" right
> after the name of the card/chip. This would mean that you need the
> bcm43xx module if I read the kernel sources correctly. (Note: I don't
> have any direct experience with this particular chip/card.)

I've got 14e4:4312, and bcm43xx doesn't work (I'm actually using ndiswrapper).
I know I need to get the firmware:

apt-get install bcm43xx-fwcutter

it asks if I want to fetch and extract the firmware: YES! Ok, everything's
in /lib/firmware now. Well:

$ sudo modprobe bcm43xx
$ ifconfig eth1
eth1 Link encap:Ethernet HWaddr 00:00:00:1a:73:a5
inet addr:192.168.1.41 Bcast:192.168.1.255 Mask:255.255.255.0
UP BROADCAST MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:13 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:0 (0.0 B) TX bytes:546 (546.0 B)
Interrupt:17
$ iwconfig eth1
eth1 IEEE 802.11b/g ESSIDff/any Nickname:"Broadcom 4311"
Mode:Managed Frequency=2.472 GHz Access Point: Invalid
Bit Rate=1 Mb/s Tx-Power=18 dBm
RTS thrff Fragment thrff
Link Quality=0/100 Signal level=0 dBm Noise level=0 dBm
Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0
$ sudo iwlist eth1 scan
eth1 No scan results
$ sudo rmmod bcm43xx
$ sudo modprobe ndiswrapper
$ sudo iwlist wlan0 scan
wlan0 Scan completed :
Cell 01 - Address: *
ESSID:"*"
...
Cell 02 - Address: *
ESSID:"*"
...
Cell 03 - Address: *
ESSID:"*"
...
$

(yes, ndiswrapper sees it as wlan0, bcm43xx as eth1)

I believe that the reason is that bcm43xx doesn't correctly recognize the MAC
address:

bcm43xx -> 00:00:00:AA:BB:CC
ndiswrapper -> 00:AA:BB:CC:*:*

(masking the address )

It seems like bcm43xx is "shifting" the MAC address of two bytes :s. Any idea?

Kindly,
David

--
. ''`. Debian maintainer | http://wiki.debian.org/DavidPaleino
: :' : Linuxer #334216 --|-- http://www.hanskalabs.net/
`. `'` GPG: 1392B174 ----|---- http://snipr.com/qa_page
`- 2BAB C625 4E66 E7B8 450A C3E1 E6AA 9017 1392 B174

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFHn4qJ5qqQFxOSsXQRAgN2AJ9o9JKw4qPymr5jH2bIJv 247Hs3MwCgma+2
fMGKuLHrAMWPIOiPYsAS2Yg=
=v0Ix
-----END PGP SIGNATURE-----

On Tue, Jan 29, 2008 at 21:20:25 +0100, David Paleino wrote:
> Il giorno Tue, 29 Jan 2008 20:46:56 +0100 Florian Kulzer ha scritto:
>
> > According to /usr/share/misc/pci.ids your chip should have the vendor-ID
> > 0x14e4 and the prodcut-ID 0x4311. Check with "lspci -nn" if this is
> > correct; you should see these numbers listed as "[14e4:4311]" right
> > after the name of the card/chip. This would mean that you need the
> > bcm43xx module if I read the kernel sources correctly. (Note: I don't
> > have any direct experience with this particular chip/card.)
>
> I've got 14e4:4312, and bcm43xx doesn't work (I'm actually using ndiswrapper).
> I know I need to get the firmware:
>
> apt-get install bcm43xx-fwcutter
>
> it asks if I want to fetch and extract the firmware: YES! Ok, everything's
> in /lib/firmware now. Well:
>
> $ sudo modprobe bcm43xx
> $ ifconfig eth1
> eth1 Link encap:Ethernet HWaddr 00:00:00:1a:73:a5
> inet addr:192.168.1.41 Bcast:192.168.1.255 Mask:255.255.255.0

[...]

> $ iwconfig eth1
> eth1 IEEE 802.11b/g ESSIDff/any Nickname:"Broadcom 4311"
> Mode:Managed Frequency=2.472 GHz Access Point: Invalid
^^^^^^^
I think that this might already be a sign of trouble.

[...]

> $ sudo iwlist eth1 scan
> eth1 No scan results
> $ sudo rmmod bcm43xx
> $ sudo modprobe ndiswrapper
> $ sudo iwlist wlan0 scan
> wlan0 Scan completed :
> Cell 01 - Address: *
> ESSID:"*"
> ...
> Cell 02 - Address: *
> ESSID:"*"
> ...
> Cell 03 - Address: *
> ESSID:"*"
> ...
> $
>
> (yes, ndiswrapper sees it as wlan0, bcm43xx as eth1)
>
> I believe that the reason is that bcm43xx doesn't correctly recognize the MAC
> address:
>
> bcm43xx -> 00:00:00:AA:BB:CC
> ndiswrapper -> 00:AA:BB:CC:*:*
>
> (masking the address )
>
> It seems like bcm43xx is "shifting" the MAC address of two bytes :s. Any idea?

OK, it turns out that I had overlooked a few things when I wrote in my
previous mail. For a start, this is in the changelog for
2.6.24~rc8-1~experimental.1:

* Disable bcm43xx, deprecated by upstream. Enable B43 (needs v4
firmware) and B43LEGACY (needs v3 firmware).

So it seems that with 2.6.24 kernel you should indeed try b43 or
b43legacy, as the OP had stated. (When looking at the sources in a
hurry, I mistook the SSB_* identifiers for PCI_* ones - sorry!)
Unfortunately I have no idea how to debug this SBB bus. In any case,
most of my previous email can safely be ignored...

The two b43* modules depend on the mac80211 stack instead of the
ieee80211 one; with a self-compiled kernel and an older "inherited"
configuration file you might first have to activate the new stack in
"Networking > Wireless" to be able to select the b43* modules in "Device
driver > Network device support > Wireless LAN".

--
Regards, | http://users.icfo.es/Florian.Kulzer
Florian |


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

On Tue, 29 Jan 2008 22:40:16 +0100
Florian Kulzer <florian.kulzer+debian@icfo.es> wrote:

> On Tue, Jan 29, 2008 at 21:20:25 +0100, David Paleino wrote:
> > Il giorno Tue, 29 Jan 2008 20:46:56 +0100 Florian Kulzer ha scritto:
> >
> > > According to /usr/share/misc/pci.ids your chip should have the vendor-ID
> > > 0x14e4 and the prodcut-ID 0x4311. Check with "lspci -nn" if this is
> > > correct; you should see these numbers listed as "[14e4:4311]" right
> > > after the name of the card/chip. This would mean that you need the
> > > bcm43xx module if I read the kernel sources correctly. (Note: I don't
> > > have any direct experience with this particular chip/card.)
> >
> > I've got 14e4:4312, and bcm43xx doesn't work (I'm actually using ndiswrapper).
> > I know I need to get the firmware:

Don't use bcm43xx; see below.

> > apt-get install bcm43xx-fwcutter
> >
> > it asks if I want to fetch and extract the firmware: YES! Ok, everything's
> > in /lib/firmware now. Well:
> >
> > $ sudo modprobe bcm43xx
> > $ ifconfig eth1
> > eth1 Link encap:Ethernet HWaddr 00:00:00:1a:73:a5
> > inet addr:192.168.1.41 Bcast:192.168.1.255 Mask:255.255.255.0
>
> [...]
>
> > $ iwconfig eth1
> > eth1 IEEE 802.11b/g ESSIDff/any Nickname:"Broadcom 4311"
> > Mode:Managed Frequency=2.472 GHz Access Point: Invalid
> ^^^^^^^
> I think that this might already be a sign of trouble.

With the bcm43xx or b43 drivers, that generally means that the card is
not associated.


[snip]

> OK, it turns out that I had overlooked a few things when I wrote
in my
> previous mail. For a start, this is in the changelog for
> 2.6.24~rc8-1~experimental.1:
>
> * Disable bcm43xx, deprecated by upstream. Enable B43 (needs v4
> firmware) and B43LEGACY (needs v3 firmware).
>
> So it seems that with 2.6.24 kernel you should indeed try b43 or
> b43legacy, as the OP had stated. (When looking at the sources in a
> hurry, I mistook the SSB_* identifiers for PCI_* ones - sorry!)
> Unfortunately I have no idea how to debug this SBB bus. In any case,
> most of my previous email can safely be ignored...

b43[legacy] is the completely rewritten driver for Broadcom chips. The
devs have been literally deprecating bcm43xx for quite a while, stating
that it is no longer developed or supported, and that everyone should
use b43[legacy] instead.

> The two b43* modules depend on the mac80211 stack instead of the
> ieee80211 one; with a self-compiled kernel and an older "inherited"
> configuration file you might first have to activate the new stack in
> "Networking > Wireless" to be able to select the b43* modules in "Device
> driver > Network device support > Wireless LAN".

The b43 devs hang out on the bcm43xx (it's still called that) ML at:
http://lists.berlios.de/mailman/listinfo/bcm43xx-dev

Both dev and support work occurs there. The devs will , but make
sure you RTFM, cooperate with their suggestions and stay polite.

> Florian |

Celejar
--
mailmin.sourceforge.net - remote access via secure (OpenPGP) email
ssuds.sourceforge.net - A Simple Sudoku Solver and Generator


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

On Tue, 29 Jan 2008 11:11:19 -0700
"wauhugo@yahoo.com" <wauhugo@yahoo.com> wrote:

> I have got the Broadcom WLAN Chip BCM943111MCG in my PC.
> "b43-fwcutter" installed some firmware.
> But it works in Debian SID with ndiswrapper only so far.
> But, you know, ndiswrapper imports parts of WINDOWS into Debian. I don't
> want that.
>
> Searching the web have I found, that there are two new Linux drivers:
>
> * b43
> * b43legacy
>
> I didn't find these in a Debian kernel source.
> So I have built my new kernel from 2.6.24
> <http://kernel.org/pub/linux/kernel/v2.6/linux-2.6.24.tar.bz2> from
> kernel.org
>
> Now it seems, that the b43 kernel module does not serve the hardware,
> but ssb does:

b43 depends on ssb

> ###$ lshw -C net description: Network controller
> product: BCM94311MCG wlan mini-PCI
> vendor: Broadcom Corporation
> physical id: 0
> bus info: pci@0000:03:00.0
> version: 02
> width: 64 bits
> clock: 33MHz
> capabilities: bus_master cap_list
> configuration: driver=b43-pci-bridge latency=0 module=ssb
>
> ^^^^^^^^
> If I understand things correctly, then should the module be: module=b43
> rather.
>
> ^^^^^^^^
> Googling tells me, that this problem is caused by "udev" and could be
> corrected by an entry in the file
>
> * z75_persistent-net.rules, which I suppose, is in Debian
> /etc/udev/rules.d/z25_persistent-net-rules, which looks in my
> system after trials of modifications like this:
>
> -------------------------------------------------------
> SUBSYSTEM=="net", DRIVERS=="?*", ATTR{address}=="00:1b:24:aa:41:9d",
> NAME="eth0"
>
> SUBSYSTEM=="net", DRIVERS=="?*", ATTR{address}=="00:00:00:1a:73:a8",
> ATTRS{type}=="1", NAME="wlan0"
>
> SUBSYSTEM=="net", DRIVERS=="?*", ATTR{address}=="00:1a:73:a8:0f:1f",
> ATTRS{type}=="1", NAME="eth1"
> ---------------------------------------------------------
>
> When googling deeper, I run into discrepancies:
> Is the adding of ATTRS{type}=="1", correct
> or should it be ATTR{type}=="1" without "S"?

I use ATTRS; it works.

> And the names of the device seem to vary from distribution to distribution:
> from "wlan0" ... "eth2".
> What is correct in the Debian environment?

I believe it's your choice; you can tell udev to name an interface
anything you want.

> Something is still wrong with my system.
> Does anybody have some clue, how to get this Wireless LAN up and running
> in Debian?

See my other mail in this thread about the ML; and also take a look at
this page:

http://linuxwireless.org/en/users/Drivers/b43

> Hugo

Celejar
--
mailmin.sourceforge.net - remote access via secure (OpenPGP) email
ssuds.sourceforge.net - A Simple Sudoku Solver and Generator


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Florian Kulzer wrote:
> The two b43* modules depend on the mac80211 stack instead of the
> ieee80211 one; with a self-compiled kernel and an older "inherited"
> configuration file you might first have to activate the new stack in
> "Networking > Wireless" to be able to select the b43* modules in "Device
> driver > Network device support > Wireless LAN".
>
>
Before posting on this list had I read through some posts, which
describe the same effect, which I see. And finally the "ssb" versus
"b43" issue had there been blamed on "udev" rules:
http://bugzilla.kernel.org/show_bug.cgi?id=9402#c14
http://bugzilla.kernel.org/show_bug.cgi?id=9402#c12

But the posts, which I had read were not for original Debian.

I want a solution for Debian but I don't understand "udev", nor many
other things in depth.

Hugo


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

--- "wauhugo@yahoo.com" <wauhugo@yahoo.com> wrote:

> I have got the Broadcom WLAN Chip BCM943111MCG in my
> PC.
> "b43-fwcutter" installed some firmware.
> But it works in Debian SID with ndiswrapper only so
> far.
> But, you know, ndiswrapper imports parts of WINDOWS
> into Debian. I don't
> want that.
Ndiswrapper doesn't really import part of windows into
Debian,
rather it just uses the software driver which is
provided for ms,
but not for us Linux users.
Until hardware manufacturers get their heads out of
Bill Gates butt, we'll have to do the best we can.

It is getting better.

Regards.

-- Angus

"All churches, whether Jewish, Christian, or Muslim, appear
to me no other than human inventions, setup to terrify and
enslave mankind - and to monopolize power and profit."
-- Thomas Paine (1737-1809)

######Linux Laptop powered by Debian Linux######
###########Reg. Linux User #278931###########


__________________________________________________ __________________________________
Looking for last minute shopping deals?
Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsea...egory=shopping


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org