[PATCH] hw_features: increase fallback to HT20 by defering channel pair check

[Eduardo Abinader]

Thanks, Jouni.

I will respin the patch with you suggestion.

On 29 March 2017 at 13:37, Eduardo Abinader <eduardoabinader at gmail.com> wrote:
> Thanks, Jouni.
>
> I will respin the patch with you suggestion.
>
> On 29 March 2017 at 13:20, Jouni Malinen <j at w1.fi> wrote:
>>
>> On Tue, Mar 21, 2017 at 12:24:54PM +0100, Eduardo Abinader wrote:
>> > Defer decision, so we could possibly benefit from a fallback.
>> > ---
>>
>> This is missing the Signed-off-by: line (see the top level CONTRIBUTIONS
>> file for details).
>>
>> > diff --git a/src/ap/hw_features.c b/src/ap/hw_features.c
>> > index 93d923a..fd93a4c 100644
>> > --- a/src/ap/hw_features.c
>> > +++ b/src/ap/hw_features.c
>> > @@ -725,11 +725,7 @@ static int hostapd_is_usable_chans(struct
>> > hostapd_iface *iface)
>> >       if (!hostapd_is_usable_chan(iface, iface->conf->channel, 1))
>> >               return 0;
>> >
>> > -     if (!iface->conf->secondary_channel)
>> > -             return 1;
>> > -
>> > -     return hostapd_is_usable_chan(iface, iface->conf->channel +
>> > -                                   iface->conf->secondary_channel * 4,
>> > 0);
>> > +     return 1;
>> >  }
>>
>>
>> There is not much remaining in hostapd_is_usable_chans() after this.. In
>> practice, it would be simply return !hostapd_is_usable_chan(iface,
>> iface->conf->channel, 1). It would likely be simpler to just get rid of
>> the function completely and handle this within the only caller in
>> hostapd_check_chans().
>>
>> --
>> Jouni Malinen                                            PGP id EFC895FA
>
>