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

[Eduardo Abinader]

I got a situation with both cases.

To support the decision for 1st case, I would say that today hostapd
also allows such "invalid" configurations and fallsback to ht20.

On 1 April 2017 at 13:39, Jouni Malinen <j at w1.fi> wrote:
> On Sat, Apr 01, 2017 at 01:02:11PM +0200, Eduardo Abinader wrote:
>> As hostapd currently allows some flexibility (fallback to ht20) on
>> setting channel to channel 20MHz, although ht40+/- has been initially
>> configured.
>> This patch just extends such flexibility. That said, ch165 fcc would
>> be then supported.
>> Doesn´t that make sense to allow such flexibility?
>
> That depends.. If the configuration file is explicitly specifying a
> HT40+ channel with a given primary and secondary channels and either of
> those channels is not allowed, that configuration file should be
> rejected. On the other hand, if the configuration file is not specifying
> a channel (i.e., using ACS) and indicates HT40+ can be enabled, then it
> would make sense to allow fallback. I'm not sure which of those cases
> this patch is trying to address. I know it does address the first one
> and that's the one I don't think should be applied. If it is also needed
> for the second case, it would make sense to add it, but I'd like to
> understand what configuration parameters can hit this case.
>
> --
> Jouni Malinen                                            PGP id EFC895FA