Thanks for the quick reply.  So are you saying that it is impossible
that the b43 driver could have somehow made my wireless card unable to
detect any networks after a reboot (in either Windows or Linux)?

On Wed, Mar 24, 2010 at 2:46 AM, Larry Finger <[email protected]> wrote:
> On 03/23/2010 02:30 PM, Chris Lopes wrote:
>> Hi,
>>
>> I had a perfectly working Dell system running Windows Vista and using
>> version 5.60.188.1 of the Broadcom windows driver (distributed by
>> Dell) for what Windows calls the "Dell Wireless 1397 WLAN Mini-Card",
>> which is actually 14e4:4315.
>>
>> I booted Parted Magic 4.9 from a USB flash drive, which uses the
>> 2.6.32 linux kernel.
>> I started up networking in Parted Magic and asked it to configure the
>> wireless driver.
>>
>> At first it found my normal SSID and I asked it to try to connect, but
>> it failed.  I thought this was just a transient issue specific to
>> Parted Magic and/or linux.
>>
>> So I booted bank into Vista, only to see that although Vista reports
>> everything as being fine with the card, no wireless networks are ever
>> detected.  I have verified that the wireless networks that I expect
>> indeed exist and are in range (the machine has not moved).
>>
>> Is it possible that the b43 driver somehow corrupted the firmware/ROM
>> of my card such that it does not function correctly?  Can I fix this?
>> There is a DMA error (also shown below in context): b43-phy0 ERROR:
>> Fatal DMA error: 0x00000400, 0x00000000, 0x00000000, 0x00000000,
>> 0x00000000, 0x00000000
>
> No, b43 did not corrupt your firmware. The firmware is not contained in
> non-volatile memory. With a power-down, it goes away. There is an EEPROM
> containing some initial values, but changing that requires specific actions
> through the sysfs, and does not happen accidentally. The wl and Windows 
> drivers
> have the firmware embedded in the driver. For copyright reasons, we cannot do
> that, which is why there is external firmware.
>
> Your DMA error is a known problem with 14e4:4315 devices with Netbooks and 
> some
> other systems with Intel processors and a Phoenix BIOS. Again, we have no fix.
> With a 2.6.33 kernel, the system will seamlessly switch to PIO mode when the
> fatal DMA error occurs.
>
> Larry
>
>
_______________________________________________
Bcm43xx-dev mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/bcm43xx-dev

Reply via email to