On Feb 19, 2004, at 6:18 AM, Greg Troxel wrote:

>   ath0 at pci0 dev 10 function 0
>   ath0: interrupting at irq 12
>   ath0: unable to attach hardware; HAL status 3
>
> I haven't tried code in the last few months, but I got a number of
> atheros cards (4 I think) to work.  I got messages that look like this
> when my cardbus memory space mappings were messed up (IBM TP600E,
> needs RBUS_MIN_START bashing).  So I would first check that you can
> use other cardbus cards.

This error is likely related to h/w setup:

typedef enum {
         HAL_OK          = 0,    /* No error */
         HAL_ENXIO       = 1,    /* No hardware present */
         HAL_ENOMEM      = 2,    /* Memory allocation failed */
         HAL_EIO         = 3,    /* Hardware didn't respond as expected 
*/
         HAL_EEMAGIC     = 4,    /* EEPROM magic number invalid */
         HAL_EEVERSION   = 5,    /* EEPROM version invalid */
         HAL_EELOCKED    = 6,    /* EEPROM unreadable */
         HAL_EEBADSUM    = 7,    /* EEPROM checksum invalid */
         HAL_EEREAD      = 8,    /* EEPROM read problem */
         HAL_EEBADMAC    = 9,    /* EEPROM mac address invalid */
         HAL_EESIZE      = 10,   /* EEPROM size not supported */
         HAL_EEWRITE     = 11,   /* Attempt to change write-locked 
EEPROM */
         HAL_EINVAL      = 12,   /* Invalid parameter to function */
         HAL_ENOTSUPP    = 13,   /* Hardware revision not supported */
         HAL_ESELFTEST   = 14,   /* Hardware self-test failed */
         HAL_EINPROGRESS = 15,   /* Operation incomplete */
} HAL_STATUS;

(i.e. 3 == HAL_EIO which means basically the HAL couldn't talk to the 
hardware).

>
> Also, I think there is only HAL support for i386 at the moment, but
> presumably that would have been a compile failure if you are on
> another arch.

There's actually support for many non-x86 architectures; look on 
sourceforge (the madwifi project).

	Sam