In article <200512280830.jBS8UCF0014226@vtn1.victoria.tc.ca>,
John Nemeth <jnemeth@victoria.tc.ca> wrote:
>On Apr 14,  6:26pm, Christos Zoulas wrote:
>} In article <200512280242.jBS2geUq020037@vtn1.victoria.tc.ca>,
>} John Nemeth <jnemeth@victoria.tc.ca> wrote:
>} >     Are there known cardbus/pcmcia regressions in 3.0?  I recently
>} >upgraded my laptop and I just went through and tested all my cards.  The
>} >Cisco Aironet 350, 3Com 3C589D, and IBM Token Ring (which is documented
>} >in GENERIC as not working with cardbus bridges) adapters all worked.
>} >However, the US Robotics XJ/CC1336 modem and Adaptec 1460D SCSI
>} >adapter don't.  In the case of the modem, when I cu to it, I don't
>} >get any response to commands.  In the case of the SCSI adapter, it prints
>} >"aic0: unable to detect chip!"
>} 
>} Yes, there is a problem and there is a patch to fix it. The patch works
>} around the problem, but it is not entirely correct. Setting cfe->iomask = 0;
>} does the trick, but the problem is really in the way we allocate regions.
>
>     And, where does this setting go?

Add it to aic_pcmcia_validate_config() before you return 0;

christos