Re: [PATCHv3 RESEND 04/11] cfg80211: allow the user space to change current NAN configuration

[Johannes Berg <johannes@sipsolutions.net>]

On Tue, 2016-03-29 at 12:35 +0300, Emmanuel Grumbach wrote:
> Some NAN configuration paramaters 

parameters

>  /**
> + * enum cfg80211_nan_conf_changes - indicates changed fields in nan
> configuration

again, NAN please

> + *
> + * @CFG80211_NAN_CONF_CHANGED_PREF: master preference
> + * @CFG80211_NAN_CONF_CHANGED_DUAL: dual band operation
> + */
> +enum cfg80211_nan_conf_changes {
> +	CFG80211_NAN_CONF_CHANGED_PREF = BIT(0),
> +	CFG80211_NAN_CONF_CHANGED_DUAL = BIT(1),
> +};

I also think this enum should be declared next to struct
cfg80211_nan_conf, not far later in the file.

> + * @nan_change_conf: changes NAN configuration. The changed
> parameters must
> + *	be specified in @changes. All other parameters must be
> ignored.

This is misleading. "The changed parameters are specified by the bitmap
@changes (using &enum cfg80211_nan_conf_changes); all other parameters
must be ignored" or so. You should always write this for driver
authors.

>   * @NL80211_ATTR_NAN_MASTER_PREF: the master preference to be used
> by
> - *	&NL80211_CMD_START_NAN. Its type is u8 and it can't be 0.
> + *	&NL80211_CMD_START_NAN and optionally with
> + *	&NL80211_CMD_CHANGE_NAN_CONFIG. Its type is u8 and it
> can't be 0.

I didn't notice this before, but generally the & is wrong. That will
try to link to that as a struct/enum/union name, which clearly isn't
the case here. You can use @, perhaps, or just % for a constant.

>   *	Also, values 1 and 255 are reserved for certification
> purposes and
>   *	should not be used during a normal device operation.
>   * @NL80211_ATTR_NAN_DUAL: NAN dual band operation config (see
>   *	&enum nl80211_nan_dual_band_conf). This attribute is used 

Here the & is correct.

> - *	&NL80211_CMD_START_NAN.
> + *	&NL80211_CMD_START_NAN and optionally with
> + *	&NL80211_CMD_CHANGE_NAN_CONFIG.

Here it's wrong again.

johannes