Loading b43legacy freezes computer

Bug #192720 reported by Ralf Hildebrandt
86
Affects Status Importance Assigned to Milestone
linux (Ubuntu)
Fix Released
High
Tim Gardner
Hardy
Fix Released
High
Tim Gardner
Intrepid
Fix Released
High
Tim Gardner

Bug Description

Binary package hint: linux-image-2.6.24-8-generic

# lspci -vvnn -s 02:02.0
02:02.0 Network controller [0280]: Broadcom Corporation BCM4303 802.11b Wireless LAN Controller [14e4:4301] (rev 02)
        Subsystem: Hewlett-Packard Company Unknown device [103c:12f3]
        Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
        Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
        Latency: 64
        Interrupt: pin A routed to IRQ 19
        Region 0: Memory at e8100000 (32-bit, non-prefetchable) [size=8K]

# uname -a
Linux pavilion 2.6.24-8-generic #1 SMP Thu Feb 14 20:40:45 UTC 2008 i686 GNU/Linux

# cat /etc/issue
Ubuntu hardy (development branch) \n \l

Loading the b43legacy driver causes a hard lockup. Magic-sysreq-key doesn't work, I have to power cycle.
I already tried removing the firmware to see if the lockup is related to the firmware, but loading the driver still causes a crash, even when there's no firmware to load :)

Loading the b43 (no legacy driver) has no effect at all (as expected).

According to http://linuxwireless.org/en/users/Drivers/b43
b43legacy is the right driver to use.

Revision history for this message
Basilio Kublik (sourcercito) wrote :

Hi Ralf
I can confirm this behavior, I've the same card and my machine locks up at boot time, this is specially delicate since this hardware is installed inside the laptop, so there's no easy workaround other than open the machine and remove the wireless card or rename/delete/blacklist the kernel modules so won't load up at boot time, anyhow this is only feasible by booting an installed old kernel (<= 2.6.24-3-generic, IIRC) or using a livecd, in my opinion this should be take in consideration as soon as possible due the closeness to the distribution release date.

lspci -vvnn -s 02:02.0
02:02.0 Network controller [0280]: Broadcom Corporation BCM4303 802.11b Wireless LAN Controller [14e4:4301] (rev 02)
        Subsystem: Hewlett-Packard Company Unknown device [103c:12f3]
        Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
        Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
        Latency: 64
        Interrupt: pin A routed to IRQ 18
        Region 0: Memory at d0204000 (32-bit, non-prefetchable) [size=8K]

I think is is enough information to start working on this, if not, just ask.

Thanks

Changed in linux:
importance: Undecided → Medium
status: New → Triaged
Revision history for this message
Michael Mittler (launchpad-michael-mittler) wrote :

I can confirm this problem too.

# lspci -vvnn -s 04:00.0
04:00.0 Network controller [0280]: Broadcom Corporation BCM94311MCG wlan mini-PCI [14e4:4311] (rev 01)
 Subsystem: AMBIT Microsystem Corp. Unknown device [1468:0422]
 Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
 Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
 Latency: 0, Cache Line Size: 64 bytes
 Interrupt: pin A routed to IRQ 17
 Region 0: Memory at f8000000 (32-bit, non-prefetchable) [size=16K]
 Capabilities: <access denied>

# uname -a
Linux ALTAIR 2.6.24-8-generic #1 SMP Thu Feb 14 20:40:45 UTC 2008 i686 GNU/Linux

Revision history for this message
moa3333 (moa3333) wrote :

Look at this one using b43 (and not the legacy version):

https://bugs.launchpad.net/ubuntu/+source/linux-meta/+bug/194523

Revision history for this message
Darek L. (luohan) wrote :

Unfortunately I have the same problem. I can't boot my machine without blacklisting b43legacy in /etc/modprobe.d/blacklist. In consequence, I'm not able to use current CD's shipped with (K)Ubuntu with alpha versions of Hardy Heron. Are there any kernel boot options to blacklist b43legacy? I've tried the following options without any success:

b43legacy.blacklist=yes
b43legacy=off
blacklist=b43legacy

Revision history for this message
Ralf Hildebrandt (ralf-hildebrandt) wrote :

More interesting would be how to debug this...
Which (modprobe) options does the b43legacy driver have?

Revision history for this message
Darek L. (luohan) wrote : Re: [Bug 192720] Re: Loading b43legacy freezes computer

> Which (modprobe) options does the b43legacy driver have?
How can I check this? I can't boot my machine (even with LiveCD!) without blacklisting b43legacy and b43xx.

Revision history for this message
Ralf Hildebrandt (ralf-hildebrandt) wrote :

* Darek L. <email address hidden>:
> > Which (modprobe) options does the b43legacy driver have?
> How can I check this? I can't boot my machine (even with LiveCD!) without blacklisting b43legacy and b43xx.

Boot it (with b43legacy blacklisted) and then

% modprobe b43legacy option

"option" being the options to the driver,m which I happen not to know, that's why I asked)

Revision history for this message
Rondeth (roneisele) wrote :

If it's of any help at all, I was bitten by this problem as well with Fedora 8...the following link MAY shed some more light, as my laptop (HP Pavilion ZV5000) has the same problem trying to boot the lastest Hardy releases (hard lock-up). If it's not of any assistance, my apologies.

https://bugzilla.redhat.com/show_bug.cgi?id=384981

Revision history for this message
Ralf Hildebrandt (ralf-hildebrandt) wrote :

Well, the https://bugzilla.redhat.com/show_bug.cgi?id=384981 page gives me a few things to try.

Revision history for this message
Ralf Hildebrandt (ralf-hildebrandt) wrote :

Alas, it seems to be exactly the same problem. With modprobe -v also get:
# sync && modprobe -v b43legacy
insmod /lib/modules/2.6.24-11-generic/kernel/drivers/input/input-polldev.ko
insmod /lib/modules/2.6.24-11-generic/kernel/drivers/leds/led-class.ko
insmod /lib/modules/2.6.24-11-generic/kernel/net/wireless/cfg80211.ko
insmod /lib/modules/2.6.24-11-generic/kernel/net/mac80211/mac80211.ko
insmod /lib/modules/2.6.24-11-generic/kernel/net/rfkill/rfkill.ko
insmod /lib/modules/2.6.24-11-generic/kernel/drivers/ssb/ssb.ko
... hard lockup ...

Revision history for this message
David Fokkema (dfokkema) wrote :

Is there any progress with this on the Ubuntu side? With only 25 days to go, I'd guess the dev people will be very busy, but this is biting me painfully, :-/

Revision history for this message
rodri.cilla (rodri-cilla) wrote :

Hi,
I'm suffering the same problem. But in my laptop (HP Zv5000) the problem also arises using ndiswrapper. If I load ndiswrapper after removing the ssb kernel module, the computer hangs up. I'm using ndiswrapper driver named netbc564.
I hope that this information will be of aid. If anybody needs aditional information, please ask.
Rodrigo

Revision history for this message
toratora (frenchtoratora) wrote :

I confirm this problem with same card and b43legacy. ndiswrapper also freeze my computer with kernel 2.6.24-15.
ndiswrapper works well with kernel 2.6.22-14.

Revision history for this message
toratora (frenchtoratora) wrote :

No more problem with kernel 2.6.24-16.
Wifi works fine with my broadcom bcm4303 rev 02

Thx

Revision history for this message
rodri.cilla (rodri-cilla) wrote :

Really? My laptop is still crashing loading both modules with kernel 2.6.24-16.

Some info about my device:
:~$ lspci -vvnn -s 02:02.0
02:02.0 Network controller [0280]: Broadcom Corporation BCM4303
802.11b Wireless LAN Controller [14e4:4301] (rev 02)
 Subsystem: Hewlett-Packard Company Unknown device [103c:12f3]
 Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR+ FastB2B-
 Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort- >SERR- <PERR-
 Latency: 64
 Interrupt: pin A routed to IRQ 17
 Region 0: Memory at e0104000 (32-bit, non-prefetchable) [size=8K]

:~$ uname -a
Linux mojo 2.6.24-16-generic #1 SMP Thu Apr 10 12:47:45 UTC 2008
x86_64 GNU/Linux

Rodrigo

2008/4/14 toratora <email address hidden>:
> No more problem with kernel 2.6.24-16.
> Wifi works fine with my broadcom bcm4303 rev 02
>
> Thx
>
>
>
> --
> Loading b43legacy freezes computer
> https://bugs.launchpad.net/bugs/192720
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in Source Package "linux" in Ubuntu: Triaged
>
> Bug description:
> Binary package hint: linux-image-2.6.24-8-generic
>
> # lspci -vvnn -s 02:02.0
> 02:02.0 Network controller [0280]: Broadcom Corporation BCM4303 802.11b Wireless LAN Controller [14e4:4301] (rev 02)
> Subsystem: Hewlett-Packard Company Unknown device [103c:12f3]
> Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
> Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
> Latency: 64
> Interrupt: pin A routed to IRQ 19
> Region 0: Memory at e8100000 (32-bit, non-prefetchable) [size=8K]
>
> # uname -a
> Linux pavilion 2.6.24-8-generic #1 SMP Thu Feb 14 20:40:45 UTC 2008 i686 GNU/Linux
>
> # cat /etc/issue
> Ubuntu hardy (development branch) \n \l
>
> Loading the b43legacy driver causes a hard lockup. Magic-sysreq-key doesn't work, I have to power cycle.
> I already tried removing the firmware to see if the lockup is related to the firmware, but loading the driver still causes a crash, even when there's no firmware to load :)
>
> Loading the b43 (no legacy driver) has no effect at all (as expected).
>
> According to http://linuxwireless.org/en/users/Drivers/b43
> b43legacy is the right driver to use.
>

Revision history for this message
toratora (frenchtoratora) wrote :

Because I unblacklisted the b43legacy module in /etc/modprobe.d/blacklist, I was thinking it works fine with b43legacy.
But looking at messages in dmesg it seems to be working with the ndiswrapper and bcmwl5a driver.

Hardware seems to be the same, try with ndiswrapper.

My hardware :

lspci -vvnn -s 02:02
02:02.0 Network controller [0280]: Broadcom Corporation BCM4303 802.11b Wireless LAN Controller [14e4:4301] (rev 02)
 Subsystem: Hewlett-Packard Company Unknown device [103c:12f3]
 Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
 Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
 Latency: 64
 Interrupt: pin A routed to IRQ 21
 Region 0: Memory at e0104000 (32-bit, non-prefetchable) [size=8K]

uname -a
Linux ordenadora 2.6.24-16-generic #1 SMP Thu Apr 10 13:23:42 UTC 2008 i686 GNU/Linux

Revision history for this message
rodri.cilla (rodri-cilla) wrote :

I didn't realize. I'm using an x84_64 kernel, whereas you are using an
i686 one, so maybe the problem is only on 64 bits platforms.

2008/4/15 toratora <email address hidden>:
> Because I unblacklisted the b43legacy module in /etc/modprobe.d/blacklist, I was thinking it works fine with b43legacy.
> But looking at messages in dmesg it seems to be working with the ndiswrapper and bcmwl5a driver.
>
> Hardware seems to be the same, try with ndiswrapper.
>
> My hardware :
>
>
> lspci -vvnn -s 02:02
>
> 02:02.0 Network controller [0280]: Broadcom Corporation BCM4303 802.11b Wireless LAN Controller [14e4:4301] (rev 02)
> Subsystem: Hewlett-Packard Company Unknown device [103c:12f3]
> Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
> Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
> Latency: 64
> Interrupt: pin A routed to IRQ 21
>
> Region 0: Memory at e0104000 (32-bit, non-prefetchable) [size=8K]
>
>
> uname -a
> Linux ordenadora 2.6.24-16-generic #1 SMP Thu Apr 10 13:23:42 UTC 2008 i686 GNU/Linux
>
> --
>
>
> Loading b43legacy freezes computer
> https://bugs.launchpad.net/bugs/192720
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in Source Package "linux" in Ubuntu: Triaged
>
> Bug description:
> Binary package hint: linux-image-2.6.24-8-generic
>
> # lspci -vvnn -s 02:02.0
> 02:02.0 Network controller [0280]: Broadcom Corporation BCM4303 802.11b Wireless LAN Controller [14e4:4301] (rev 02)
> Subsystem: Hewlett-Packard Company Unknown device [103c:12f3]
> Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
> Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
> Latency: 64
> Interrupt: pin A routed to IRQ 19
> Region 0: Memory at e8100000 (32-bit, non-prefetchable) [size=8K]
>
> # uname -a
> Linux pavilion 2.6.24-8-generic #1 SMP Thu Feb 14 20:40:45 UTC 2008 i686 GNU/Linux
>
> # cat /etc/issue
> Ubuntu hardy (development branch) \n \l
>
> Loading the b43legacy driver causes a hard lockup. Magic-sysreq-key doesn't work, I have to power cycle.
> I already tried removing the firmware to see if the lockup is related to the firmware, but loading the driver still causes a crash, even when there's no firmware to load :)
>
> Loading the b43 (no legacy driver) has no effect at all (as expected).
>
> According to http://linuxwireless.org/en/users/Drivers/b43
> b43legacy is the right driver to use.
>

Revision history for this message
David Fokkema (dfokkema) wrote :

This bug has a medium importance. Well, maybe...

I really like to ask for at least this before hardy: add a note in the release notes! I'd like to know how I can boot hardy on my laptop. I consider it at least important that we don't have systems failing to boot the desktop cd because of a known bug without mentioning any of this in the caveats. It would be very nice to have a boot option available next to acpi=off and the like, but since time is short, I understand that this is probably not possible.

But please, document this bug in the release notes and mention the right boot option to manually disable this driver.

Thanks!

(Usual praise and kudos still apply for hardy, of course)

Revision history for this message
darkmatter14B (darkmatter14b) wrote :

Installing on a compaq presario r3000 laptop.

Man, I can't boot at all. Wasn't able to install using the normal install cd because of the boot freeze, but did install 8.04 rc using the alternate cd.
After trying to boot in rescue mode, boot seems to hang at the line "b43legacy-phy0: Broadcom 4301 WLAN found"
Is there anyway to get around this? Tried noacpi to no avail.

Thanks,

Revision history for this message
JoiHap (astronomy) wrote :

Same problem with an ASUS Laptop (L5000GA),
lspci -vvnn -s 02:02.0
02:02.0 Network controller [0280]: Broadcom Corporation BCM4306 802.11b/g Wireless LAN Controller [14e4:4320] (rev 02)
 Subsystem: ASUSTeK Computer Inc. Unknown device [1043:120f]
 Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
 Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
 Latency: 64
 Interrupt: pin A routed to IRQ 5
 Region 0: Memory at feaf8000 (32-bit, non-prefetchable) [size=8K]
 Capabilities: [40] Power Management version 2
  Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=375mA PME(D0+,D1+,D2+,D3hot+,D3cold+)
  Status: D0 PME-Enable- DSel=0 DScale=2 PME-

Hardy doesn't boot at all after a fresh installation. Only blacklisting b43legacy in /etc/modprobe.d/blacklist I was able to boot.

Revision history for this message
JoiHap (astronomy) wrote :

The way I got around this problem was: boot from CD and select Rescue Broken System; open a shell in the root system; use "vi" to edit /etc/modprobe.d/blacklist; blacklist bcm43xx and b43 legacy, as follows:

# replaced by b43 and ssb.
blacklist bcm43xx
blacklist b43legacy

The system then boots. To make wireless work I then followed these instructions: http://linuxwireless.org/en/users/Drivers/b43, specifically "You are using the b43-legacy driver". Of course, I didn't compile b43-fwcutter, but used the one from the repositories. The directory to put the firmware is "/lib/firmware/2.6.24-16-generic".

Now, to use the wireless I just type "sudo modprobe b43legacy", and it works (odd). If I don't "sudo modprobe -r b43legacy" before shutting down, the system hangs and does not power down.

Revision history for this message
Ralf Hildebrandt (ralf-hildebrandt) wrote :

b43legacy crashes even WITH firmware installed...

Revision history for this message
JoiHap (astronomy) wrote :

Since b43legacy is causing hard locks on hardy installations, making computers unusable, unbootable, shouldn't the bug have an importance of "high", according to this scheme: https://wiki.ubuntu.com/Bugs/Importance?

Revision history for this message
chongma (matt-bodyaspect) wrote :

Ubuntu 8.04 cannot install from CD

HP Pavilion zv5000
Broadcom 4301
AMD Athlon 64

b43legacy-phy0:Broadcom 4301 WLAN found
clocksource tsc unstable

bug #218256 looks like a duplicate

Revision history for this message
chongma (matt-bodyaspect) wrote :

Bug #220384 could be duplicate

Revision history for this message
chongma (matt-bodyaspect) wrote :
Revision history for this message
Saivann Carignan (oxmosys) wrote :

According to Kernel bug policies, I assign this bug to the ubuntu kernel team. Since this bug can prevent ubuntu from booting and that it affects essential hardware, I set the Importance to High.

Changed in linux:
assignee: nobody → ubuntu-kernel-team
importance: Medium → High
Revision history for this message
chongma (matt-bodyaspect) wrote :

This bug may also be a duplicate of bug #182716

Revision history for this message
harley_prime (harley-prime) wrote :

8.04 live cd will not boot.

8.04 alt cd will as well as 7.10 live cd.

7.10 install boots, wifi will not work.

8.04 install from alt cd will not boot, hangs here in recovery mode:
b43 legacy-phy0:Broadcom 4301 WLAN found

Athlon XP 3200+
Asus a7n8x-x
Linksys WMP11 v2.7 pci wireless

sounds like the same issue

Revision history for this message
David Fokkema (dfokkema) wrote :

I can confirm the following, asus l5800c (l5c) with broadcom 4303:

Install 7.10 (live cd). Install ndiswrapper and drivers. Confirm wifi works.

Upgrade to 8.04, blacklist b43legacy and reboot.

Ndiswrapper is loaded, but the wifi is off (led's off, and every iwlist or iwconfig returns an error (no wireless extensions)).

Trying to reload ndiswrapper and unload various other modules often results in hard lockups.

Revision history for this message
pjina3 (pjina3) wrote :

darkmatter14B : I have the same machine with you and I have solved the problem by edit /etc/modprobe.d/blacklist; blacklist bcm43xx and b43 legacy, as follows:

# replaced by b43 and ssb.
blacklist bcm43xx
blacklist b43legacy

if you can't boot the system to edit it, you could use a knoppix's liveCD to boot and then remount the hda1 (in mode rw) then edit /media/hda1/etc/modprobe.d/blacklist

the solution can solve the problem but you can't use the b43legacy driver, it isn't problem for me because I use a D-Link usb wifi which use rt73 driver

Revision history for this message
Stefan (stefan13) wrote :

Same problem here with HP/Compaq nx9110. Same error message about bcm 4301, although I am using bcm 4303. By blacklisting b43legacy and bcm43xx I can boot normally, but of course have no wireless.
Thanks for fixing this, if you need any more info, please let me know.

Revision history for this message
Larry Finger (larry-finger) wrote :

A patch for this bug has just been submitted to the Ubuntu Kernel Team. The problem was a kernel NULL pointer and affects BCM4301 and BCM4303 cards, which use the b43legacy driver. After patching, the Ubuntu kernel works with these devices.

Revision history for this message
Tim Gardner (timg-tpi) wrote :
Revision history for this message
Tim Gardner (timg-tpi) wrote :
Revision history for this message
Tim Gardner (timg-tpi) wrote :
Revision history for this message
Tim Gardner (timg-tpi) wrote :
Revision history for this message
Tim Gardner (timg-tpi) wrote :

There is now a test kernel with these 3 patches at http://ppa.launchpad.net/timg-tpi/ubuntu. Please give it a try and report the results.

Changed in linux:
assignee: ubuntu-kernel-team → timg-tpi
milestone: none → ubuntu-8.04.1
status: Triaged → In Progress
Revision history for this message
Larry Finger (larry-finger) wrote :

Works for me.

Revision history for this message
Ralf Hildebrandt (ralf-hildebrandt) wrote :

Works for me as well:
[ 792.255845] b43legacy-phy0: Broadcom 4301 WLAN found
[ 792.279238] b43legacy-phy0 debug: Found PHY: Analog 0, Type 1, Revision 4
[ 792.279269] b43legacy-phy0 debug: Found Radio: Manuf 0x17F, Version 0x2050, Revision 2
[ 792.303586] b43legacy-phy0 debug: Radio initialized
[ 792.308860] phy0: Selected rate control algorithm 'simple'
[ 792.482070] udev: renamed network interface wlan0 to eth1
[ 792.618958] b43legacy-phy0 debug: Loading firmware version 0x127, patch level 14 (2005-04-18 02:36:27)
[ 792.719375] b43legacy-phy0 debug: Chip initialized
[ 792.720321] b43legacy-phy0 debug: 30-bit DMA initialized
[ 792.723641] b43legacy-phy0 debug: Wireless interface started
[ 792.731132] b43legacy-phy0: Radio hardware status changed to disabled
[ 792.731810] b43legacy-phy0 debug: Adding Interface type 2
[ 792.732755] ADDRCONF(NETDEV_UP): eth1: link is not ready

Revision history for this message
Tim Gardner (timg-tpi) wrote :

SRU Justification:

Impact: b43kegacy can crash or lock during initialization.

Fix Description: One oops fix plus 2 important upstream maintenance patches suggested by Larry Finger (the b43legacy maintainer).

1) Upon driver load when a BCM4304 rev 2 card is present, the kernel undergoes a hard crash with power off the only recovery. This crash is due to a missing check for NULL pointers, which only occur for a limited set os devices.

2) In the current code, the presence of improper LED encoding in the SPROM generates a kernel bug message. As a number of older cards have faulty encoding, these conditions are now silently accepted.

3) Initialize 802.11b correctly.

Patch: See previous 3 attachments

TEST CASE: b43legacy locks on init, or cannot connect using 802.11b.

Revision history for this message
Tim Gardner (timg-tpi) wrote :
Changed in linux:
status: In Progress → Fix Committed
Revision history for this message
David Fokkema (dfokkema) wrote :

I confirm that it works, but not entirely without problems.

00:0c.0 Network controller: Broadcom Corporation BCM4303 802.11b Wireless LAN Controller (rev 02)
00:0c.0 0280: 14e4:4301 (rev 02)

With firmware (b43-fwcutter, yes for download) it works flawlessly. But before installing b43-fwcutter, so without firmware, I get the following in my logs:

ssb: FATAL ERROR: Bus powered down while accessing PCI MMIO space

A new error message is added every second. When I unload the module with `sudo modprobe -r b43legacy' I get a kernel oops. It might be important to add the following: I blacklisted b43legacy so it's not loaded on boot. However, the b44 module _is_ loaded which also uses the ssb module. I don't know if this is relevant, but the b44/b43legacy without firmware combination might screw up ssb.

To summarize: it works, but not on fresh installations without firmware, they produce ominous error messages every second.

To conclude: thanks for the fix! I only have wifi access, not ethernet, so the fix is much appreciated!

Revision history for this message
Larry Finger (larry-finger) wrote :

In many cases, the first error is the only one that has any significance. I'm confident that your log has a message that the firmware could not be loaded. I know where periodic messages arise, and I'll look into stopping them. I don't think it has anything to do with b44. It does not need any firmware to function and should have been fully operational when these messages were being printed.

Thanks for the report.

Revision history for this message
David Fokkema (dfokkema) wrote :

Of course, the log has the missing firmware error message. No problem with that. But the ssb error seemed severe, especially since unloading b43legacy causes a kernel oops. But maybe that's not directly related to the ssb error message. I think that it should be possible to unload b43legacy without a crash. Anything I can do to pin down the problem?

Revision history for this message
jimmy muldoon (jimmymuldoon) wrote :

Confirm this error on an intel-based compaq presario r3000 laptop; didn't boot, hung on boot screen with progress bar 1/5 way across; in second boot option (verbose mode or whatever it's called) card reported as 4301 and boot hung then.
Blacklisting as per pjina3's advice works (thanks pjina3!); I blacklisted both bcm43xx and b43legacy and the machine boots. Un-blacklisted bcm43xx and is didn't boot. Which is where I'm at now, actually.

Can anyone difinitively advise best course of action from here on in to enable wifi; should I wait a bit until patches are applied to the kernel, or use ndiswrapper, or what is the recommended course of action?

Thanks to those who have provided advice on this thread by the way..

Revision history for this message
nondas (nondasg) wrote :

Hi to all, i am a new linux user. Can anyone provide an installation description for beginners?
I just installed Hardy on a Fujitsu-Siemens D1840 without the BCM4301. Everything works great apart from the card-reader ( i dont care ) and ...wifi.
What i must do next?
Obviously i cant understand how to install the test kernel mentioned above or the patching thing...

Revision history for this message
David Fokkema (dfokkema) wrote :

@nondas:

There are two ways to go about this:

1. Wait it out. There is a fix committed. This means that in due time (when?) the fix is included in an automatic update. First as an proposed update (you might want to enable these System->Administration->Software Sources->Updates enable Pre-released updates) and eventually as a recommended update.

2. Learn about compiling your own kernel and enabling it with grub. It is fun. It is a lot of work. You'll learn a lot from it. But it might not pay off the time invested. If computing is a hobby for you, go ahead. If it is not, don't even try.

In the mean time, you could buy a supported usb wireless device or just use an ethernet connection.

Good luck!

Revision history for this message
Larry Finger (larry-finger) wrote :

You don't need to buy a USB device. From a terminal, do the following commands:

sudo su -
echo "deb http://ppa.launchpad.net/timg-tpi/ubuntu hardy main" > \
/etc/apt/sources.list.d/kernel-ppa.list
apt-get update
apt-get -u upgrade

After these commands are executed, a new kernel with patches for the BCM4301 problems will become a candidate for installation. Install it and the necessary firmware and your wifi will work.

Larry

Revision history for this message
nondas (nondasg) wrote :

Thanks for the fast responce! Everything is clear now and I 'll give it a try.
As for compiling the kernel i believe sooner or later i must learn it...

Revision history for this message
Brian Piltz, MD (brianpiltz) wrote :

I am a newb. I am trying to install Ubuntu HH on my Dell desktop which has the Linksys WMP11 ver2.7. The install fails at the attempt to load the Broadcom controller. Am I going to have to wait until the next version of Ubuntu comes out in 6 months for the updates to part of the install CD? I don't see how I can download the automatic update patches when I can't get the OS installed on my machine, in the first place. Any help would be greatly appreciated.

Revision history for this message
nondas (nondasg) wrote :

TheLarry's workaround did not work for me...
At the pre-released updates nothing new.
The only thing i could do was blacklisting the b43legacy to continue working.
As for you Brian i suggest to install without the wifi card ( take it out of your box ), then try the solutions described above.
Don't forget to sign the petition :
http://www.petitiononline.com/BCM4301/petition.html

Revision history for this message
jimmy muldoon (jimmymuldoon) wrote :

Brian, you might not have to remove the wifi card to install the OS, you should hopefully be able to install from the 'Alternate Install CD' which uses a text-based installer instead of actually booting up a full ubuntu (which hangs as it reaches that driver). This is what I did, and it installed ubuntu fine.

Just be aware that even when you have it installed and try to boot then, it might hang again as it tries to load that driver. (again, this is what I experienced).

The way to make your system boot is by blacklisting the guilty driver as described previously in this thread by pjina3, by using either the ubuntu cd in repair mode (not ridiculously difficult but might be a bit hard if you're new at this stuff) or by using a knoppix cd as mentioned, to edit /etc/modprobe.d/blacklist as described. I was actually able to boot an ubuntu 7.10 livecd on the same laptop to edit the blacklist file, you might want to give that a shot.

Once you have that driver blacklisted (I eventually blacklisted both the bcm43xx and b43legacy drivers) and reboot without the cd inserted you should hopefully have a working install (apart from wireless); then if you have a wired ethernet port you can download updates using that.

As the guys say, a fix to the wifi driver problem has been submitted so hopefully when that gets pushed down via updates you should be able to get it over your wired connection. Failing that if you need the wifi card working in a hurry you can try the fixes suggested above by David Fokkema or Larry Finger.

Best of luck.

Revision history for this message
Brian Piltz, MD (brianpiltz) wrote :

Thanks for the explanation. I decided to go the simple route and buy a compatible WLAN card.

Revision history for this message
Larry Finger (larry-finger) wrote :

That decision is yours, of course; however, you will still have to remove your existing card before you will be able to boot.

Larry

Revision history for this message
chongma (matt-bodyaspect) wrote :

BCM4303
HP zv5000

Installed the kernel as described by Larry, and the b43legacy module now loads fine. The card lights up and it finds access points but cannot seem to connect using the password. This may be an unrelated problem.

Also there seems to be no restricted driver module present so it cannot load the nvidia card. Maybe this will come with the next kernel release.

Revision history for this message
JoiHap (astronomy) wrote :

Testing Tim's latest kernel (-18-generic) this is what happens in a fresh hardy installation:

1) CPU Goes 100% maybe because

2) "ssb: FATAL ERROR: Bus powered down while accessing PCI MMIO space" is always being printed.

3) attached follows /var/log/messages. Is it normal?

The led indicating that the wireless is turned on, turns on for a few moments during boot time and then turns off again.

My hardware is listed above.

Revision history for this message
scott healey (smhealey) wrote :

no solution yet? the fix larry posted above seems to be no longer available.

Revision history for this message
JoiHap (astronomy) wrote :

The CPU at 100% was not related to this issue. So, remains 2) and 3)... Was the wireless supposed to work out-of-the-box with Tim's new kernel? or you still need to download firmware as a warning during boot-time seems to suggest?

Revision history for this message
Furlot (furlot) wrote :

same than chongma : load is ok, scan is ok but I can't connect
Presario R3215 : Broadcom Corporation BCM4303 (rev 02)

here is the dmesg modprobe b43legacy and then try to connect :
[ 9770.120955] ssb: Core 0 found: IEEE 802.11 (cc 0x812, rev 0x02, vendor 0x4243)
[ 9770.120965] ssb: Core 1 found: PCMCIA (cc 0x80D, rev 0x00, vendor 0x4243)
[ 9770.120974] ssb: Core 2 found: Fast Ethernet (cc 0x806, rev 0x02, vendor 0x4243)
[ 9770.120982] ssb: Core 3 found: V90 (cc 0x807, rev 0x01, vendor 0x4243)
[ 9770.120991] ssb: Core 4 found: PCI (cc 0x804, rev 0x03, vendor 0x4243)
[ 9770.124362] ssb: Sonics Silicon Backplane found on PCI device 0000:02:02.0
[ 4881.698311] b44.c:v2.0
[ 4881.698331] b44: Invalid MAC address found in EEPROM
[ 4881.698335] b44 ssb0:1: Problem fetching invariants of chip, aborting.
[ 4881.699227] b44: probe of ssb0:1 failed with error -22
[ 4881.717462] b43legacy-phy0: Broadcom 4301 WLAN found
[ 4881.744374] b43legacy-phy0 debug: Found PHY: Analog 0, Type 1, Revision 4
[ 4881.744400] b43legacy-phy0 debug: Found Radio: Manuf 0x17F, Version 0x2050, Revision 2
[ 4881.767887] b43legacy-phy0 debug: Radio initialized
[ 4881.769041] phy0: Selected rate control algorithm 'simple'
[ 4882.158614] udev: renamed network interface wlan0 to wlan1
[ 4883.159840] b43legacy-phy0 debug: Loading firmware version 0x127, patch level 14 (2005-04-18 02:36:27)
[ 4883.258541] b43legacy-phy0 debug: Chip initialized
[ 4883.259086] b43legacy-phy0 debug: 30-bit DMA initialized
[ 4883.260902] b43legacy-phy0 debug: Wireless interface started
[ 4883.266295] b43legacy-phy0: Radio hardware status changed to disabled
[ 4883.266679] b43legacy-phy0 debug: Adding Interface type 2
[ 4883.267251] ADDRCONF(NETDEV_UP): wlan1: link is not ready
[ 9837.600296] wlan1: Initial auth_alg=0
[ 9837.600308] wlan1: authenticate with AP xx:xx:xx:xx:xx:xx
[ 9837.800098] wlan1: authenticate with AP xx:xx:xx:xx:xx:xx
[ 9838.000036] wlan1: authenticate with AP xx:xx:xx:xx:xx:xx
[ 9838.199962] wlan1: authentication with AP xx:xx:xx:xx:xx:xx timed out
[ 9852.141114] wlan1: Initial auth_alg=0
[ 9852.141128] wlan1: authenticate with AP xx:xx:xx:xx:xx:xx
[ 9852.339822] wlan1: authenticate with AP xx:xx:xx:xx:xx:xx
[ 9852.539781] wlan1: authenticate with AP xx:xx:xx:xx:xx:xx
[ 9852.739704] wlan1: authentication with AP xx:xx:xx:xx:xx:xx timed out

Steve Langasek (vorlon)
Changed in linux:
assignee: nobody → timg-tpi
importance: Undecided → High
milestone: none → ubuntu-8.04.1
status: New → Triaged
milestone: ubuntu-8.04.1 → none
Revision history for this message
Léo Studer (leo-studer) wrote :

Same than Furlot: loading and scanning are going on properly, but no connection is possible.

Hardware: HP Pavillon zv5000, chipset broadcom BCM4303 802.11b (rev 02)

The dmesg output is the same too

Revision history for this message
Martin Pitt (pitti) wrote :

Accepted into -proposed, please test and give feedback here

Changed in linux:
milestone: ubuntu-8.04.1 → none
status: Triaged → Fix Committed
Steve Langasek (vorlon)
Changed in linux:
milestone: none → ubuntu-8.04.1
Revision history for this message
Furlot (furlot) wrote :

test with new kernel 2.6.24-19-generic
Presario R3215 : Broadcom Corporation BCM4303 (rev 02) (14e4:4301)

module load ok but no scan, here is the dmesg :

[ 2243.055144] ssb: Core 0 found: IEEE 802.11 (cc 0x812, rev 0x02, vendor 0x4243)
[ 2243.055156] ssb: Core 1 found: PCMCIA (cc 0x80D, rev 0x00, vendor 0x4243)
[ 2243.055165] ssb: Core 2 found: Fast Ethernet (cc 0x806, rev 0x02, vendor 0x4243)
[ 2243.055174] ssb: Core 3 found: V90 (cc 0x807, rev 0x01, vendor 0x4243)
[ 2243.055182] ssb: Core 4 found: PCI (cc 0x804, rev 0x03, vendor 0x4243)
[ 2243.059856] ssb: Sonics Silicon Backplane found on PCI device 0000:02:02.0
[ 2243.103691] b43legacy-phy0: Broadcom 4301 WLAN found
[ 2243.127062] b43legacy-phy0 debug: Found PHY: Analog 0, Type 1, Revision 4
[ 2243.127094] b43legacy-phy0 debug: Found Radio: Manuf 0x17F, Version 0x2050, Revision 2
[ 2243.151065] b43legacy-phy0 debug: Radio initialized
[ 2243.160493] phy0: Selected rate control algorithm 'simple'
[ 1121.017175] b44.c:v2.0
[ 1121.017195] b44: Invalid MAC address found in EEPROM
[ 1121.017199] b44 ssb0:1: Problem fetching invariants of chip, aborting.
[ 1121.017207] b44: probe of ssb0:1 failed with error -22
[ 1121.062362] udev: renamed network interface wlan0 to wlan1
[ 1121.208450] b43legacy-phy0 debug: Loading firmware version 0x127, patch level 14 (2005-04-18 02:36:27)
[ 1121.306953] b43legacy-phy0 debug: Chip initialized
[ 1121.307465] b43legacy-phy0 debug: 30-bit DMA initialized
[ 1121.309277] b43legacy-phy0 debug: Wireless interface started
[ 1121.314704] b43legacy-phy0: Radio hardware status changed to disabled
[ 1121.315094] b43legacy-phy0 debug: Adding Interface type 2
[ 1121.315567] ADDRCONF(NETDEV_UP): wlan1: link is not ready

Revision history for this message
Steve Langasek (vorlon) wrote :

Furlot, please file a separate bug report about the failure to run; this bug report is about hard lock-ups when loading the driver, which is separate from your issue.

Is anyone who experienced the hard lockups before able to verify that the new kernel corrects this problem?

Revision history for this message
JoiHap (astronomy) wrote :

It does not hard lock anymore. The message "ssb: FATAL ERROR: Bus powered down while accessing PCI MMIO space" still gets printed out a lot.

Revision history for this message
Larry Finger (larry-finger) wrote :

Are any other messages logged around these "ssb: FATAL ERROR ..." messages?

Revision history for this message
Damien Dixon (damienxdixon) wrote :
  • unnamed Edit (236 bytes, text/html; charset=ISO-8859-1)

I am following this bug because it has rendered Hardy un-installable on my
laptop. I was just wondering if someone could clear up what the end result
will be - how will I be able to get it install on my laptop? Is it fixed
already?

Revision history for this message
Larry Finger (larry-finger) wrote :

The bug is fixed in the updated kernels; however, the CD has not been updated. To install from the current CD, you need to do the following:

1. Boot the CD, choose your language, and select "install". Next press F6 and insert "break=top" into the options line. I'm not sure if it was necessary, but I put it ahead of the two minus signs at the end.

2. At the (initramfs) prompt, type the following lines:

       echo "blacklist ssb" >> /etc/modprobe.d/blacklist
       exit

3. Installation will proceed. Once it completes, reboot the computer with the CD in the drive. I know this is against the on-screen instructions, but we need to change a file on the new installation. This time, use the "Try Ubuntu with ..." option.

4. Once the system is up and running, open a console, and enter the following commands:

    sudo -s
    mount -t ext3 /dev/sda1 /mnt
    echo "blacklist ssb" >> /mnt/etc/modprobe.d/blacklist
    unmount /mnt
    echo

    Note: The /sda1 in the second line above should be replaced with the correct device and partition for / on your system. If you get it wrong, either the mount or the echo commands will fail.

6. You will now be able to reboot normally without the system hanging. After you have updated the kernel (2.6.24-17 or later), you should install the b43legacy firmware and edit /etc/modprobe.d/blacklist to remove the line that blacklists ssb. Once these two steps are completed, your BCM4303 should work with the new driver.

Revision history for this message
ianO (iogilvy) wrote :

We have the same problem with Hardy on a laptop- however this laptop was upgraded to Hardy and is currently able to boot when selecting an older kernel (2.6.22-14) from the grub menu.

However despite all updates being applied, selecting the newest kernel (2.6.24-18) still presents the problem. We wish to get this laptop running properly on Hardy as a key step in migrating a set of laptops to ubuntu.
Is there as step missing? What is the "b43legacy firmware" step? Beyond simply having the latest updates from the standard repositories, what else is required?

Any assistance greatly appreciated

Revision history for this message
Steve Langasek (vorlon) wrote :

ian0,

The kernel package that includes this fix is the -19 package in hardy-proposed, not the -18 package in hardy-updates. You will need to enable the "proposed updates" software source on your system, or wait until -19 passes all its QA tests and is copied to hardy-updates.

Revision history for this message
Stefan (stefan13) wrote :

On my Compaq nx 9110, I can now boot without a problem or blacklisting drivers, however the wireless still does not detect any networks, or shows any sign of life.

Revision history for this message
Damien Dixon (damienxdixon) wrote :
  • unnamed Edit (349 bytes, text/html; charset=ISO-8859-1)

I have the same issue. I installed Gutsy then upgraded it using Update
Manager to Hardy. No freezing; acknowledgment that there is a wireless card,
but it's not finding any networks.

I am going to hard wire it and do a full update, including proposed updates
and see if that helps.

Chip is a BCM4301 on an Advent 7027 laptop.

Revision history for this message
Larry Finger (larry-finger) wrote :

For both Stefan and Damien: What, if anything, does dmesg have to say regarding b43legacy or ssb?

Revision history for this message
Steve Langasek (vorlon) wrote :

At the risk of repeating myself, this is a bug report about hard lockups when loading the driver. If the issue you're seeing is that the driver loads but doesn't work, that needs to be tracked as a distinct bug report.

Revision history for this message
ianO (iogilvy) wrote :

To Steve Langasek.
Thank you- clear and informative answer. Perfect!

Revision history for this message
Stefan (stefan13) wrote :
Download full text (9.3 KiB)

Here is my relevant part from the dmesg output - any help is much appreciated, thanks:

 27.508169] ssb: Core 0 found: IEEE 802.11 (cc 0x812, rev 0x02, vendor 0x4243)
[ 27.508175] ssb: Core 1 found: PCMCIA (cc 0x80D, rev 0x00, vendor 0x4243)
[ 27.508180] ssb: Core 2 found: Fast Ethernet (cc 0x806, rev 0x02, vendor 0x4243)
[ 27.508185] ssb: Core 3 found: V90 (cc 0x807, rev 0x01, vendor 0x4243)
[ 27.508189] ssb: Core 4 found: PCI (cc 0x804, rev 0x03, vendor 0x4243)
[ 27.511512] ssb: Sonics Silicon Backplane found on PCI device 0000:02:02.0
[ 27.512219] ACPI: PCI Interrupt 0000:02:00.0[A] -> GSI 16 (level, low) -> IRQ 17
[ 27.561861] ohci1394: fw-host0: OHCI-1394 1.1 (PCI): IRQ=[17] MMIO=[d0208000-d02087ff] Max Packet=[2048] IR/IT contexts=[4/8]
[ 27.566369] ACPI: PCI Interrupt 0000:02:03.0[A] -> GSI 19 (level, low) -> IRQ 18
[ 27.567005] eth0: RealTek RTL8139 at 0xa000, 00:0f:b0:01:b2:1d, IRQ 18
[ 27.567007] eth0: Identified 8139 chip type 'RTL-8101'
[ 27.569203] 8139cp: 10/100 PCI Ethernet driver v1.3 (Mar 22, 2004)
[ 27.569357] ACPI: PCI Interrupt 0000:02:07.2[C] -> GSI 19 (level, low) -> IRQ 18
[ 27.569370] ehci_hcd 0000:02:07.2: EHCI Host Controller
[ 27.569404] ehci_hcd 0000:02:07.2: new USB bus registered, assigned bus number 5
[ 27.569448] ehci_hcd 0000:02:07.2: irq 18, io mem 0xd0208c00
[ 27.579463] ehci_hcd 0000:02:07.2: USB 2.0 started, EHCI 1.00, driver 10 Dec 2004
[ 27.579605] usb usb5: configuration #1 chosen from 1 choice
[ 27.579629] hub 5-0:1.0: USB hub found
[ 27.579636] hub 5-0:1.0: 5 ports detected
[ 27.687453] b44.c:v2.0
[ 27.687474] b44: Invalid MAC address found in EEPROM
[ 27.687478] b44 ssb0:1: Problem fetching invariants of chip, aborting.
[ 27.687484] b44: probe of ssb0:1 failed with error -22
[ 27.698443] hda: max request size: 128KiB
[ 27.703581] hda: 117210240 sectors (60011 MB) w/8192KiB Cache, CHS=65535/16/63
[ 27.706001] hda: cache flushes supported
[ 27.706043] hda: hda1 hda2
[ 27.768700] hdc: ATAPI 24X DVD-ROM CD-R/RW drive, 2048kB Cache
[ 27.768709] Uniform CD-ROM driver Revision: 3.20
[ 27.935603] Attempting manual resume
[ 27.935608] swsusp: Resume From Partition 3:2
[ 27.935609] PM: Checking swsusp image.
[ 27.935805] PM: Resume from disk failed.
[ 27.973066] kjournald starting. Commit interval 5 seconds
[ 27.973083] EXT3-fs: mounted filesystem with ordered data mode.
[ 28.831672] ieee1394: Host added: ID:BUS[0-00:1023] GUID[463f0200145b403d]
[ 39.397894] Linux agpgart interface v0.102
[ 39.497340] Real Time Clock Driver v1.12ac
[ 39.507379] agpgart: Detected Ati IGP9100/M chipset
[ 39.510528] agpgart: AGP aperture is 32M @ 0xd2000000
[ 39.539183] input: PC Speaker as /devices/platform/pcspkr/input/input2
[ 39.541385] pci_hotplug: PCI Hot Plug PCI Core version: 0.5
[ 39.543282] shpchp: Standard Hot Plug PCI Controller Driver version: 0.4
[ 39.552266] piix4_smbus 0000:00:14.0: Found 0000:00:14.0 device
[ 39.841724] ACPI: PCI Interrupt 0000:00:14.5[B] -> GSI 5 (level, low) -> IRQ 5
[ 39.866860] input: Power Button (FF) as /devices/virtual/input/input3
[ 39.879092] ACPI: Power Button (FF) [PWRF]
[ 39.879179] ...

Read more...

Revision history for this message
Furlot (furlot) wrote :

for issue with module loading but no wifi I create this report : #238896

Revision history for this message
Martin Pitt (pitti) wrote :

Copied to hardy-updates. The new kernel was tested extensively by many people, who reported back in other bug reports. Due to lack of feedback, this particular bug was not confirmed to be tested, though. Please report back here if the bug still occurs for you with the new kernel packages, then we will reopen this bug.

Changed in linux:
status: Fix Committed → Fix Released
Revision history for this message
Martin Wickman (martin-wickman) wrote :

I tried Larry Finger's advice above (with 'break=top') and then blacklisting ssb whilst trying to install ubuntu. It did not work. It still hangs during boot with

b43legacy-phy0: Broadcom 4303 WLAN found

I also tried blacklisting b43legacy, b43, b44 and any other combination I could think of. Still no success. I get the feeling that blacklisting does not work?

$ sudo lspci -nn|grep Broad
00:0c.0 Network controller [0280]: Broadcom Corporation BCM4303 802.11b Wireless LAN Controller [14e4:4301] (rev 02)

$ sudo ndiswrapper -l
bcmwl5 : driver installed
        device (14E4:4301) present (alternate driver: bcm43xx)

It works fine in gentoo using ndiswrapper.

Any ideas what to do next?

Revision history for this message
Martin Wickman (martin-wickman) wrote :

I managed to get through the installation by deleting the **** modules. Also needed is a knoppix cd (or other working linux live cd)

1. Append 'break=bottom' to the boot options (F6)
2. Start the ubuntu installation
3. Delete ('rm') all modules called b43 and ssb from /lib/modules and /root/lib/modules
4. 'exit'
5. Complete the installation as usual and reboot
6. Insert knoppix cd
7. Mount your / partition and 'rm' the modules as in step 3.

Reboot. Installation should be complete and ubuntu working.

After that use ndiswrapper to get the broadcom card working.

Revision history for this message
nightwish (rssfeeds) wrote :

I must confirm the problem described by chongma above (2008-05-17) for Hardy Heron with kernel (2.6.24-19) on a HP Pavilion zv5000 (internal Broadcom bcm4303 wireless card):

When using the b43legacy with ssb and the firmware extracted by b43-fwcutter, the laptop does not freeze, but shows the behaviour following behaviour: The b43legacy module loads fine. The card lights up and it finds access points but cannot seem to connect. This happens both to encrypted aswell to unencrypted networks.

The laptop freezes, on the other hand, when using ndiswrapper. I manually deactivated the b43, b43legacy and ssb modules. The freeze occours right when modprobing ndiswrapper.

As for now I have reverted back to the 2.5.22-14 which works fine with ndiswrapper.

Revision history for this message
David Fokkema (dfokkema) wrote :

@Martin:

Please don't use this. While it does work, it is not a clean way of
doing things. Deleting things is ugly, especially since the next kernel
update will render your system unusable again (since it reinstalls the
deleted files). Also, there is a clean(er) way, which probably should be
documented instead of the 'deleting things way'.

1. Download the ubuntu alternate install cd (mind the 'alternate').
2. Install your system using this (yes, it uses debian-installer instead
of a live cd approach). Write down the chosen root partition
(/dev/sda1 ?)
3. Reboot again from CD and choose: 'Rescue a broken system'
4. Say yes to a couple of things and then choose the correct root
partition (you wrote it down, didn't you?)
5. Execute a shell in the root partition:

# cd /etc/modprobe.d
# vim blacklist (or use your favorite command-line editor)
add the line: 'blacklist b43legacy' (without the quotes)
save and exit
# exit

6. Reboot

Then, hook up to a cable or use a supported usb wifi card, enable
updates, install kernel updates, remove the blacklist line and reboot.
It should work now. If not, re-add the blacklist line and come back.

Good luck!

David

Revision history for this message
Dilton McGowan II (diltonm) wrote :

Will the CD on the main site for 8.04 ever work as did 7.04 and 7.10 on my system? Maybe we should wait for 8.10?

Revision history for this message
Steve Langasek (vorlon) wrote :

Tim, is this fix included in the version of linux 2.6.26 present in intrepid?

Revision history for this message
Damien Dixon (damienxdixon) wrote :

I'm running intrepid on a b43 imbedded chipset with no problems.

Revision history for this message
Tim Gardner (timg-tpi) wrote :

The wireless stack and device drivers are hugely different from Hardy to Intrepid, so I would guess that this particular freeze was fixed.

Revision history for this message
Larry Finger (larry-finger) wrote :

I can confirm the b43legacy does not crash Intrepid.

Revision history for this message
Steve Langasek (vorlon) wrote :

Thanks for the confirmation that this is fixed in intrepid; closing the report.

Changed in linux:
status: Fix Committed → Fix Released
Revision history for this message
zeojex (zeojex) wrote :

Hi all,

Same problem for me : "clocksource tsc unstable" message booting on the Ubuntu Studio 9.04 Alternate AMD64.
I am on a Acer Laptop Aspire 6530G (amd turion X2).

I have tried the two solutions below founded on the Internet :
- From a ubuntu live CD ==> mount the "/" partition and Blacklist bcm43xx and b43legacy in /etc/modprobe.d/blacklist (save modifications...)
=> Not fix the problem, same message : "clocksource tsc unstable"
- From a ubuntu live CD ==> mount the "/boot" partition and add the line "clocksource=hpet" in /boot (grub/menu.lst) (save modifications...)
=> Not fix the problem, same message : "clocksource tsc unstable"

Can you help me please ?

Revision history for this message
Larry Finger (larry-finger-lwfinger) wrote :

The "clocksource tsc unstable" message has nothing to do with b43, b43legacy, or any other wireless driver. It comes from your CPU. Many, if not all, AMD 64-bit processors will show this message. Mine does, but it does not affect any operations. Linux uses other clocks that are stable, and is just informing you of the situation. Please IGNORE this message.

Revision history for this message
zeojex (zeojex) wrote :

Yes infact I was suprised reading a Wireless Driver could cause this message which concern clock...thx.
But my system doesn't boot and is blocked precisely at this message display.
I don't know what is my boot problem.

Revision history for this message
Larry Finger (larry-finger-lwfinger) wrote :

I downloaded the "final" version of the Ubuntu Studio 9.04 Alternate AMD64 DVD and installed it on my Turion AMD64 X2-based computer. It works well except that I could not find a way to get the Gnome applet to accept a WPA2 secret. It only offered me WEP password or WEP key.

What are the 10 or so messages that print before the clocksource one.

Revision history for this message
Damien Dixon (damienxdixon) wrote : Opa
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.