Re: IWN driver (5000 & 5100 series)

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Wed, Mar 25, 2009 at 03:21:54PM -0500, Jeremy C. Reed wrote:
> I tried your patch. I am on amd64 and had to patch a little due to:
> 
> ../../../../dev/pci/if_iwn.c: In function 'iwn_set_timing':
> ../../../../dev/pci/if_iwn.c:3367: warning: format '%llu' expects type 
> 'long long unsigned int', but argument 3 has type '__uint64_t'
> ../../../../dev/pci/if_iwn.c: In function 'iwn_read_firmware':
> ../../../../dev/pci/if_iwn.c:5178: warning: format '%d' expects type 
> 'int', but argument 3 has type 'size_t'
> 
> I used PRIu64

Yes, I ran into this after sending the patch

> 
>         DPRINTF(("timing bintval=%u, tstamp=%" PRIu64", init=%u\n",
>             ni->ni_intval, le64toh(cmd.tstamp), (uint32_t)(val - mod)));
>  
> and %zd
> 
>                 aprint_error_dev(sc->sc_dev,
>                     "firmware file too short: %zd bytes\n", size);
> 
> hopefully that is correct.
> 
> I rebooted with new 5.0_RC3 but didn't see any change.
> 
>       vendor 0x8086 product 0x4232 (miscellaneous network) at pci2 dev 0 
>       function 0 not configured
> 
> I rebooted again (with the switch on front of laptop turned to on and that 
> made a light come on for the wireless) but still no change.
> 
> No extra device showed with "ifconfig -l".
> 
> Why is the pci_matchid list not used with "#if 0"? Why isn't the 
> PCI_PRODUCT_INTEL_PRO_WL_5100AGN_1 et cetera in the pcidevices? Is that 
> not needed?

it's needed of course. I just added the adapter I was trying to get
working as a quick&dirty hack because I didn't have all the PCI IDs in
hands. See iwn_match() ...

-- 
Manuel Bouyer <bouyer%antioche.eu.org@localhost>
     NetBSD: 26 ans d'experience feront toujours la difference
--