Re: b43: firmware loading problem and sleeping BUG

[Michael Buesch <mb@bu3sch.de>]

On Friday 06 November 2009 17:19:51 Martin Fuzzey wrote:
> I posted a couple of days ago to the bcm43xx_dev list about problems
> getting b43 to run on a
> arm (MX21 SoC). That list appears to have died recently (at least
> nothing added to the archives
> since last month)

Well, it never really worked. It's hosted on berlios. That's why it's crap. ;)

> However I now get:
> pcmcia_socket pcmcia_socket0: pccard: PCMCIA card inserted into slot 0
> pcmcia 0.0: pcmcia: registering new device pcmcia0.0
> ssb: Core 0 found: ChipCommon (cc 0x800, rev 0x0D, vendor 0x4243)
> ssb: Core 1 found: IEEE 802.11 (cc 0x812, rev 0x09, vendor 0x4243)
> ssb: Core 2 found: PCI (cc 0x804, rev 0x0C, vendor 0x4243)
> ssb: Core 3 found: PCMCIA (cc 0x80D, rev 0x07, vendor 0x4243)
> BUG: sleeping function called from invalid context at mm/vmalloc.c:1367
> in_atomic(): 1, irqs_disabled(): 128, pid: 1361, name: modprobe
> 2 locks held by modprobe/1361:
>  #0:  (buses_mutex){+.+.+.}, at: [<bf03761c>] ssb_bus_register+0x48/0x1a4 [ssb]
>  #1:  (&bus->bar_lock){......}, at: [<bf038e38>]
> ssb_pcmcia_read32+0x24/0x74 [ssb]
> irq event stamp: 105917
> hardirqs last  enabled at (105916): [<c027cf34>]
> __mutex_unlock_slowpath+0x120/0x150
> hardirqs last disabled at (105917): [<c027eee0>] _spin_lock_irqsave+0x20/0x60
> softirqs last  enabled at (105853): [<c0042f60>] irq_exit+0x50/0x64
> softirqs last disabled at (105832): [<c0042f60>] irq_exit+0x50/0x64
> ------------[ cut here ]------------
> WARNING: at kernel/lockdep.c:2465 lockdep_trace_alloc+0xac/0xec()
> Modules linked in: b43(+) ssb mac80211 mxc_pcmcia
> ---[ end trace 7a542bbcadb0bb88 ]---

These logs look weird. Is that the full log from "dmesg" command?
I guess the pcmcia_access_configuration_register() is doing nonatomic vmalloc stuff.

I think this is easy to fix, because we can replace the spinlock by a mutex, as
the b43 driver (which is the only user of the code) always allows sleeping now.
I'll send a patch for testing soon.

> ssb: Sonics Silicon Backplane found on PCMCIA device pcmcia0.0
> b43-phy0: Broadcom 4318 WLAN found (core revision 9)
> b43-phy0 debug: Found PHY: Analog 3, Type 2, Revision 7
> b43-phy0 debug: Found Radio: Manuf 0x17F, Version 0x2050, Revision 8
> phy0: Selected rate control algorithm 'minstrel'
> Broadcom 43xx driver loaded [ Features: M, Firmware-ID: FW13 ]
> 
> And then when I do ifconfig wlan0 up:
> 
> b43 ssb0:0: firmware: requesting b43/ucode5.fw
> b43 ssb0:0: firmware: requesting b43/pcm5.fw
> b43 ssb0:0: firmware: requesting b43/b0g0initvals5.fw
> b43 ssb0:0: firmware: requesting b43/b0g0bsinitvals5.fw
> b43-phy0: Loading firmware version 410.2160 (2007-05-26 15:32:10)
> b43-phy0 ERROR: Initial Values Firmware file-format error.
> b43-phy0 ERROR: You must go to
> http://wireless.kernel.org/en/users/Drivers/b43#devicefirmware
> and download the correct firmware for this driver version. Please
> carefully read all instructions on this website.
> 
> I am using firmware broadcom-wl-4.150.10.5 extracted with
> b43-fwcutter-012 as described
> http://wireless.kernel.org/en/users/Drivers/b43#devicefirmware
> 
> 
> Any ideas why the versions don't match?

These are two different version numbers. One for the broadcom driver
and one for the firmware.

> The firmware files are arch independant right? (since I run fwcutter
> on x86 and run the kernel on ARM)

Yes.

> Kernel version is 2.6.32-rc5
> The PCMCIA controller driver is new (I recently posted it to
> linux-pcmcia and linux-arm)

Thanks for testing. The ssb-pcmcia code is not tested a lot, so
I'm not surprised that there are bugs.

-- 
Greetings, Michael.