Hi Tomasz,
On 07/10/2014 10:33 AM, Tomasz Bursztyka wrote:
>>> +++ b/gsupplicant/gsupplicant.h
>>> @@ -78,6 +78,8 @@ extern "C" {
>>> #define G_SUPPLICANT_WPS_PIN (1 << 2)
>>> #define G_SUPPLICANT_WPS_REGISTRAR (1 << 3)
>>> +#define G_SUPPLICANT_WPS_CONFIG_PBC 0x0080
>>> +
>> Is there a special reason it is not like the defines above (1 << x)?
>
> This value is the one given by wpa_supplicant into wps_capabilities
> (yes, internal values exposed...).
> So, it's use to parse wps_capabilities.
It stands out, that is way I am asking. The rest of the defines use the
(1 << x) version.
>>> +bool g_supplicant_peer_is_wps_pin(GSupplicantPeer *peer)
>>> +{
>>> + if (!peer)
>>> + return false;
>>> +
>>> + return ((peer->wps_capabilities & G_SUPPLICANT_WPS_PIN) ==
>>> + G_SUPPLICANT_WPS_PIN);
>>> +}
>> In the linse above this change the pattern used is:
>>
>> dbus_bool_t g_supplicant_network_is_wps_advertizing(GSupplicantNetwork
>> *network)
>> {
>> if (!network)
>> return FALSE;
>>
>> if (network->wps_capabilities & G_SUPPLICANT_WPS_REGISTRAR)
>> return TRUE;
>>
>> return FALSE;
>> }
>>
>> I would suggest to stick to the pattern. That brings the next question
>> if the return type 'bool' should not be dbus_bool_t to be consistent.
>
> Actually, I would rather change this
> g_supplicant_network_is_wps_advertizing()
> returning a bool rather than a dbus_bool_t.
Yes, that is also good idea. I don't mind just be be careful where the
allee uses the type for D-Bus messages (1 byte vs 4 byte issue). :D
cheers,
daniel
>
> Tomasz
_______________________________________________
connman mailing list
[email protected]
https://lists.connman.net/mailman/listinfo/connman
