Re: [PATCH] cfg80211: check for carrier state only when offchanel CAC supported

[Arend van Spriel <arend@broadcom.com>]

On 12/23/14 16:38, Johannes Berg wrote:
> On Tue, 2014-12-23 at 16:34 +0100, Arend van Spriel wrote:
>> On 12/23/14 15:53, Johannes Berg wrote:
>>> On Wed, 2014-12-24 at 01:35 +0530, Avinash Patil wrote:
>>>> Checking for carrier state during start_radar_detection is needed
>>>> only for devices which support offchannel CAC.
>>>> This patch adds one more condition while checking for carrier state
>>>> to see if offchanel CAC is supported, else we need not check carrier
>>>> state.
>>>>
>>>> Signed-off-by: Avinash Patil<patila@marvell.com>
>>>> ---
>>>>    include/net/cfg80211.h | 2 ++
>>>>    net/wireless/nl80211.c | 3 ++-
>>>>    2 files changed, 4 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h
>>>> index 4ebb816..1b80302 100644
>>>> --- a/include/net/cfg80211.h
>>>> +++ b/include/net/cfg80211.h
>>>> @@ -2740,6 +2740,7 @@ struct cfg80211_ops {
>>>>     * @WIPHY_FLAG_SUPPORTS_5_10_MHZ: Device supports 5 MHz and 10 MHz channels.
>>>>     * @WIPHY_FLAG_HAS_CHANNEL_SWITCH: Device supports channel switch in
>>>>     *	beaconing mode (AP, IBSS, Mesh, ...).
>>>> + * @WIPHY_FLAG_OFFCHAN_CAC: Device supports offchannel CAC.
>>>
>>> I think this should be an nl80211 feature flag, so that userspace can
>>> also know whether or not this is supported.
>>>
>>> However, we've run out of nl80211 feature flags. Arend is working on
>>> adding more, see here:
>>> http://mid.gmane.org/1418935344-22159-1-git-send-email-arend@broadcom.com
>>> (ignore the 4-way-HS offload bit he's defining)
>>>
>>> I'm not sure how to proceed here - in a sense your feature flag is far
>>> less controversial and work-intensive than Arend's (who needs a lot more
>>> than just that patch) so I guess you could pick up his patch, but I
>>> expect him to also have fixed already. He's probably also on vacation
>>> for about the same time though...
>>
>> Not on vacation just yet. I indeed reworked the patch, but I was waiting
>> for wpa_supplicant patches so I can verify it works with brcmfmac. I can
>> submit the reworked feature flag patch with empty enum if that would work.
>
> Maybe if you were to do that Avinash could resubmit it again with the
> feature flag they need.

Fine by me. Will come out shortly.

Gr. AvS