Commit 6b5d,.. breaks ath5k in WPA-enabled networks

Commit 6b5d,.. breaks ath5k in WPA-enabled networks

[Peter Chubb]

With Linus's current head-of-tree my ath5k cannot associate with
WPA-enabled access points.  I bisected back to this patch:

	    Since commit 6b5d117eddc09cd976ad8030d715f4350f598a22 
	    ath5k: clean up queue manipulation

I suspect at least one of the two wake-up calls that were removed is
actually necessary.

Reverting the patch restores correct behaviour.
--
Dr Peter Chubb                                  peter DOT chubb AT nicta.com.au
http://www.ertos.nicta.com.au               ERTOS within National ICT Australia
All things shall perish from under the sky/Music alone shall live, never to die

Re: Commit 6b5d,.. breaks ath5k in WPA-enabled networks

[Bruno Randolf]

On Monday 31 May 2010 18:23:43 Peter Chubb wrote:
> With Linus's current head-of-tree my ath5k cannot associate with
> WPA-enabled access points.  I bisected back to this patch:
> 
> 	    Since commit 6b5d117eddc09cd976ad8030d715f4350f598a22
> 	    ath5k: clean up queue manipulation
> 
> I suspect at least one of the two wake-up calls that were removed is
> actually necessary.
> 
> Reverting the patch restores correct behaviour.

try this, please:

https://patchwork.kernel.org/patch/100681/

i have sent the patch a long time ago but it didnt get merged yet.

bruno

> --
> Dr Peter Chubb                                  peter DOT chubb AT
> nicta.com.au http://www.ertos.nicta.com.au               ERTOS within
> National ICT Australia All things shall perish from under the sky/Music
> alone shall live, never to die --
> To unsubscribe from this list: send the line "unsubscribe linux-wireless"
> in the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Commit 6b5d,.. breaks ath5k in WPA-enabled networks

[Peter Chubb]

>>>>> "Bruno" == Bruno Randolf <br1@einfach.org> writes:

Bruno> On Monday 31 May 2010 18:23:43 Peter Chubb wrote:
>> With Linus's current head-of-tree my ath5k cannot associate with
>> WPA-enabled access points.  I bisected back to this patch:
>> 
>> Since commit 6b5d117eddc09cd976ad8030d715f4350f598a22 ath5k: clean
>> up queue manipulation
>> 
>> I suspect at least one of the two wake-up calls that were removed
>> is actually necessary.
>> 
>> Reverting the patch restores correct behaviour.

Bruno> try this, please:

Bruno> https://patchwork.kernel.org/patch/100681/

This works, thanks.

--
Dr Peter Chubb                                  peter DOT chubb AT nicta.com.au
http://www.ertos.nicta.com.au               ERTOS within National ICT Australia
All things shall perish from under the sky/Music alone shall live, never to die

Re: Commit 6b5d,.. breaks ath5k in WPA-enabled networks

[me]

On Mon, May 31, 2010 at 09:25:55PM +1000, Peter Chubb wrote:
> Bruno> On Monday 31 May 2010 18:23:43 Peter Chubb wrote:
> >> With Linus's current head-of-tree my ath5k cannot associate with
> >> WPA-enabled access points.  I bisected back to this patch:
> >> 
> >> Since commit 6b5d117eddc09cd976ad8030d715f4350f598a22 ath5k: clean
> >> up queue manipulation
> 
> Bruno> try this, please:
> Bruno> https://patchwork.kernel.org/patch/100681/
> 
> This works, thanks.

Yeah, big oops. :\

So what do we do now?  I guess we should just revert it for now rather
than taking the half that works plus Bruno's fix?

-- 
Bob Copeland %% www.bobcopeland.com

Re: Commit 6b5d,.. breaks ath5k in WPA-enabled networks

[Jaroslav Fojtik]

> With Linus's current head-of-tree my ath5k cannot associate with
> WPA-enabled access points.  I bisected back to this patch:

I hope that not only WPA association is corrupted. I have already 
reported association failures for plain connections.

> 	    Since commit 6b5d117eddc09cd976ad8030d715f4350f598a22 
> 	    ath5k: clean up queue manipulation
I have tested to revert this patch, but I did not observe any effect.

There must be something wrong somewhere else.

regards
   Jara


> I suspect at least one of the two wake-up calls that were removed is
> actually necessary.
> 
> Reverting the patch restores correct behaviour.
> --
> Dr Peter Chubb                                  peter DOT chubb AT nicta.com.au
> http://www.ertos.nicta.com.au               ERTOS within National ICT Australia
> All things shall perish from under the sky/Music alone shall live, never to die
> --
> To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Commit 6b5d,.. breaks ath5k in WPA-enabled networks

[Bruno Randolf]

On Monday 31 May 2010 21:56:40 you wrote:
> On Mon, May 31, 2010 at 09:25:55PM +1000, Peter Chubb wrote:
> > Bruno> On Monday 31 May 2010 18:23:43 Peter Chubb wrote:
> > >> With Linus's current head-of-tree my ath5k cannot associate with
> > >> WPA-enabled access points.  I bisected back to this patch:
> > >> 
> > >> Since commit 6b5d117eddc09cd976ad8030d715f4350f598a22 ath5k: clean
> > >> up queue manipulation
> > 
> > Bruno> try this, please:
> > Bruno> https://patchwork.kernel.org/patch/100681/
> > 
> > This works, thanks.
> 
> Yeah, big oops. :\
> 
> So what do we do now?  I guess we should just revert it for now rather
> than taking the half that works plus Bruno's fix?

cc: stable for my patch?
john?

bruno

Re: Commit 6b5d,.. breaks ath5k in WPA-enabled networks

[Peter Chubb]

>>>>> "Bruno" == Bruno Randolf <br1@einfach.org> writes:

>> 
>> Yeah, big oops. :\
>> 
>> So what do we do now?  I guess we should just revert it for now
>> rather than taking the half that works plus Bruno's fix?

Bruno> cc: stable for my patch?  john?

Bruno's patch on top of current head-of-tree works fine.

Peter C



--
Dr Peter Chubb                                  peter DOT chubb AT nicta.com.au
http://www.ertos.nicta.com.au               ERTOS within National ICT Australia
All things shall perish from under the sky/Music alone shall live, never to die