kern/132722: [ath] Wifi ath0 associates fine with AP, but DHCP or IP does not work

[Bruce Simpson]

Hi,

One crucial piece of info missing is exactly which Atheros part you are 
using, can you please attach
this to the PR? you can  get this from pciconf -lv and look for the 
ath0, also please post your boot time
dmesg / or when you insert the part. thanks.


Matthias Apitz wrote:
> The following reply was made to PR kern/132722; it has been noted by GNATS.
>
> From: Matthias Apitz <guru at unixarea.de>
> To: bug-followup at FreeBSD.org, guru at unixarea.de
> Cc:  
> Subject: Re: kern/132722: [ath] Wifi ath0 associates fine with AP, but DHCP or IP does not work
> Date: Tue, 17 Mar 2009 09:42:43 +0100
>
>  I forgot to mention that I've applied this patch from Sam to my RELENG_7
>  kernel:
>  
>  http://article.gmane.org/gmane.os.freebsd.stable/60383/match=cft+ath+hal+src
>  
>  I think, updating to 
>  
>  Date: Thu Mar 12 03:09:11 2009
>  New Revision: 189720
>  URL: http://svn.freebsd.org/changeset/base/189720
>  
>  as suggested will not change anything, or I'm wrong?
>   

I didn't make many changes apart from fixups in a few files for the diff,
so if you're running with the updated HAL, then in theory an svn 
update/cvsup should make no change.
Of course the only way to be sure is to look under the hood.

I didn't see any issues with the backported HAL, and I don't really know 
enough about ath(4) to
suggest any workarounds at this point other than reversing your 
application of the patch, and cvsupping
to before the date the patch was MFCed, can we be absolutely sure that 
the regression you observe
happens with the new HAL?

Sam would know more, but of course one of the reasons I've taken this on 
is because a project
I'm involved with needs working ath(4) on pci-e in 7.x, I don't know 
about chip specific problems
at this point in time, so I'm just as in the dark as you are!

thanks,
BMS