Re: [OpenWrt-Devel] [PATCHv3] wifi: Introduce 802.11ac support

Re: [OpenWrt-Devel] [PATCHv3] wifi: Introduce 802.11ac support

[Kalle Valo]

(Adding ath10k@lists.infradead.org)

Sven Eckelmann <sven@open-mesh.com> writes:

> On Monday 17 February 2014 05:57:25 Kalle Valo wrote:
>> Sven Eckelmann <sven@open-mesh.com> writes:
>> > The ath10k patch wasn't resent because the driver patches are now
>> > integrated in the OpenWrt mac80211 package. The answer in
>> > 
>> >  http://article.gmane.org/gmane.comp.embedded.openwrt.devel/22422 also
>> >  wasn't> 
>> > quite promising.
>> 
>> What do you mean? What was wrong with my answer?
>
> Nothing is "wrong" but you said that there are two branches. One is "AP mode, 
> nothing else" and the other one is "main branch again has more
> features, like P2P, ...". So I don't see a good reason why a multi-purpose 
> system like openwrt should accept my patch to change it to the "AP-only" 
> ath10k firmware.

AP mode in 10.1 firmware is extensively tested and that's why I strongly
recommend to use 10.1 firmware as the default in openwrt. I don't think
it's wise to knowingle cripple ath10k AP mode in openwrt just because
lack of P2P support. Like Matti also said, AP mode is most important in
openwrt, we should have that properly supported first.

But we should try ways to mitigate the problem of having two firmware
branches. For example, should we have a module parameter in ath10k to
choose which firmware to use? Or is there even a way to do that
automatically based on what interfaces are used?

I'm happy to hear any ideas how to improve this.

-- 
Kalle Valo

_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: [OpenWrt-Devel] [PATCHv3] wifi: Introduce 802.11ac support

[Kalle Valo]

(Adding ath10k@lists.infradead.org)

> I think AP is by far the most common use case for OpenWrt, and DFS support for
> example is only for the AP-branch firmware (as I found out after some intense
> study of the ath10k/cfg80211/mac80211 source

Oh, I didn't know that. But that's not suprising as almost all of AP
mode testing happens with 10.1 firmware.

> -- maybe the firmware features should be better documented...)

Sure, it's just lack of time on my part. Any help with documentation
would be appreciated.

> Does AP firmware support plain STA operation?

Yes, but not that well tested.

> If it does, AP firmware should no doubt be the default in OpenWrt.

I fully agree. This should be the first stage.

> If not, then we should maybe think of a way to support both.

For the second stage I think we should try improve how to make it easier
to switch between firmwares (see my previous email).

-- 
Kalle Valo

_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: [OpenWrt-Devel] [PATCHv3] wifi: Introduce 802.11ac support

[Jonas Gorski]

On Wed, Feb 19, 2014 at 10:41 AM, Kalle Valo <kvalo@qca.qualcomm.com> wrote:
> (...)
> But we should try ways to mitigate the problem of having two firmware
> branches. For example, should we have a module parameter in ath10k to
> choose which firmware to use? Or is there even a way to do that
> automatically based on what interfaces are used?

I don't know if this is a good way to do it, but this is exactly what
mwl8k does. They have dedicated AP and STA firmwares, and reload if
needed when adding the apprpriate interfaces. See
<http://lxr.free-electrons.com/source/drivers/net/wireless/mwl8k.c#L4663>
for the code.


Regards
Jonas

_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

Re: [OpenWrt-Devel] [PATCHv3] wifi: Introduce 802.11ac support

[Ben Greear]

On 02/19/2014 01:41 AM, Kalle Valo wrote:
> (Adding ath10k@lists.infradead.org)
> 
> Sven Eckelmann <sven@open-mesh.com> writes:
> 
>> On Monday 17 February 2014 05:57:25 Kalle Valo wrote:
>>> Sven Eckelmann <sven@open-mesh.com> writes:
>>>> The ath10k patch wasn't resent because the driver patches are now
>>>> integrated in the OpenWrt mac80211 package. The answer in
>>>>
>>>>  http://article.gmane.org/gmane.comp.embedded.openwrt.devel/22422 also
>>>>  wasn't> 
>>>> quite promising.
>>>
>>> What do you mean? What was wrong with my answer?
>>
>> Nothing is "wrong" but you said that there are two branches. One is "AP mode, 
>> nothing else" and the other one is "main branch again has more
>> features, like P2P, ...". So I don't see a good reason why a multi-purpose 
>> system like openwrt should accept my patch to change it to the "AP-only" 
>> ath10k firmware.
> 
> AP mode in 10.1 firmware is extensively tested and that's why I strongly
> recommend to use 10.1 firmware as the default in openwrt. I don't think
> it's wise to knowingle cripple ath10k AP mode in openwrt just because
> lack of P2P support. Like Matti also said, AP mode is most important in
> openwrt, we should have that properly supported first.
> 
> But we should try ways to mitigate the problem of having two firmware
> branches. For example, should we have a module parameter in ath10k to
> choose which firmware to use? Or is there even a way to do that
> automatically based on what interfaces are used?
> 
> I'm happy to hear any ideas how to improve this.

I don't think it can be done automatically, at least not in general,
because users may add/remove interface types after the firmware is
loaded.  Resetting the NIC and reloading firmware is a bit touchy
currently, so I don't think you want to switch to different
firmware dynamically.

A module option is probably the best way to do this in my opinion.


Thanks,
Ben


-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com


_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k