On Thu, May 31, 2012 at 3:17 PM, Arend van Spriel <arend@xxxxxxxxxxxx> wrote:
> This documentation is provided for the driver. mac80211 already allows
> this callback to sleep. So rephrase 'must be allowed to sleep' to 'is
> allowed to sleep' or 'can sleep'.
>
I will change it, thanks.
>> @@ -517,7 +520,7 @@ static void ieee80211_get_et_stats(struct wiphy *wiphy,
>> * network device.
>> */
>>
>> - rcu_read_lock();
>> + mutex_lock(&local->sta_mtx);
>>
>> if (sdata->vif.type == NL80211_IFTYPE_STATION) {
>> sta = sta_info_get_bss(sdata, sdata->u.mgd.bssid);
>> @@ -546,7 +549,7 @@ static void ieee80211_get_et_stats(struct wiphy *wiphy,
>> data[i] = (u8)sinfo.signal_avg;
>> i++;
>> } else {
>> - list_for_each_entry_rcu(sta, &local->sta_list, list) {
>> + list_for_each_entry(sta, &local->sta_list, list) {
>> /* Make sure this station belongs to the proper dev */
>> if (sta->sdata->dev != dev)
>> continue;
>
> Just for my curiosity/learning: How are these related to the new
> callback? Is it because you are now doing the callback in sta_set_sinfo()?
>
Yes, since sta_set_sinfo now might sleep we must replace rcu locks
around it with sta_mtx.
Thanks for the review,
Victor.
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
