LP PHY Dell device not working properly with b43 even in PIO mode

[Anisse Astier <anisse@astier.eu>]

On Thu, 10 Feb 2011 11:41:57 -0600, Larry Finger <Larry.Finger@lwfinger.net> wrote :

> On 02/10/2011 11:16 AM, Anisse Astier wrote:
> > [  103.627184] wlan0: associated
> > [  103.677583] b43-phy0 debug: Using hardware based encryption for keyidx: 0, mac: 68:7f:74:16:57:c9
> > [  148.586756] b43-phy0 ERROR: Fatal DMA error: 0x00000800, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000
> > [  148.586788] b43-phy0 ERROR: This device does not support DMA on your system. It will now be switched to PIO.
> > [  148.586791] b43-phy0: Controller RESET (DMA error) ...
> > [  148.586794] b43-phy0 ERROR: Fatal DMA error: 0x00000800, 0x00000000, 0x00000000, 0x00000000, 0x00000000, 0x00000000
> > [  148.586797] b43-phy0 ERROR: This device does not support DMA on your system. It will now be switched to PIO.
> > [  148.586800] b43-phy0: Controller RESET (DMA error) ...
> > [  148.586882] b43-phy0 debug: Wireless interface stopped
> 
> On some boxes, 14e4:4315 devices fail in b43 with with this DMA error. We don't
> know why, nor do we know how to fix it. Unfortunately, none of the developers
> have such a system, which makes the debugging harder.
> 
> I don't think the problem is with the Dell 1397, but with the motherboard on the
> laptop. In fact, most of the problems have been with netbook computers with a
> Phoenix BIOS. Does your system fit that description?
It's indeed a Phoenix Bios. Machine model is Dell Vostro 1520, but
definitely not a netbook.

> 
> > [  297.703041] b43-phy0 ERROR: MAC suspend failed
> > [  297.854043] b43-phy0 ERROR: MAC suspend failed
> > [  298.003042] b43-phy0 ERROR: MAC suspend failed
> 
> Are the MAC suspend failures in PIO mode a regression that recently appeared? I
> cannot think of any changes that should have affected the LP PHY devices, but
> anything is possible.
> 
> The only workaround I can suggest is to use wl. I hate to recommend
> closed-source software, but there may be no alternative.
Yep, I already packaged that driver, and planned to use it. I was in the
mode give-every-time-I-have-left to b43 developers to test patches and
try to find a solution.


Anisse