Re: SMC2662W V.4

Hi Alex,

On Mon, 28 Jun 2004, Alex Schultz wrote:

> Hi,
>
> My friend finally got a chance to test it. At first we had some keyboard
> lockup problems when one plugged the device in, but that didn't seem to
> be related to this driver, more the usb system in the kernel. We worked
> around that by having the SMC2662W v4 plugged in at startup and it
> doesn't give us the keyboard lock then (that's a good enough work around
> for us). When we load the driver though, it still doesn't work. The
> dmesg results that apply are:
> /root/newianwireless/at76c503a/usbdfu.c: USB Device Firmware Upgrade
> (DFU) handler v0.12beta14 loading
> /root/newianwireless/at76c503a/at76c503.c: Generic Atmel
> at76c503/at76c505 routines v0.12beta14
> /root/newianwireless/at76c503a/at76c503-fw_skel.c: Atmel at76c505a
> (RFMD2958) Wireless LAN Driver v0.12beta14 loading
> at76c505a-rfmd2958 1-2:1.0: usb_probe_interface
> at76c505a-rfmd2958 1-2:1.0: usb_probe_interface - got id
> /root/newianwireless/at76c503a/at76c503-fw_skel.c: using compiled-in
> firmware
> /root/newianwireless/at76c503a/at76c503.c: opmode 4
> /root/newianwireless/at76c503a/at76c503.c: wlan%d: state 9 -> 0 (INIT)
> /root/newianwireless/at76c503a/at76c503.c: USB interface: 2 endpoints
> /root/newianwireless/at76c503a/at76c503.c: $Id: at76c503.c,v 1.60
> 2004/06/15 18:20:37 jal2 Exp $ compiled Jun 26 2004 16:07:53
> /root/newianwireless/at76c503a/at76c503.c: firmware version 1.102.0 #113
> (fcs_len 4)
> /root/newianwireless/at76c503a/at76c503.c: Bad board type set (7).
> Unable to get hardware config.
> /root/newianwireless/at76c503a/at76c503.c: Get HW Config failed (-22)
> /root/newianwireless/at76c503a/at76c503.c: could not get MAC address
> usb 1-2: hcd_unlink_urb cf1a1d40 fail -22
> drivers/usb/core/usb.c: registered new driver at76c505a-rfmd2958
>
>
>  From what I can tell, the "Bad board type set" error occurs when a
> board type is returned that is not any listed in this switch and case
> block, but BOARDTYPE_505_RFMD_2958 is listed there and defined as 7 in
> at76c503.h. So anyways this is a weird looking problem. Does anyone know
> what's going on?

No, this is in at76c503.h:

#define BOARDTYPE_505_RFMD_2958     6
#define BOARDTYPE_505A_RFMD_2958    7

the 505A_* is the new one, so just add a
	case BOARDTYPE_505A_RFMD_2958:
after the
	case BOARDTYPE_505_RFMD_2958:
in at76c503.c around line 972 and the above error should go away.

/Jörg