On 12/19/2022 11:15 AM, Jouni Malinen wrote:
On Mon, Dec 19, 2022 at 10:53:55AM -0800, Aloka Dixit wrote:
On 12/18/2022 7:24 AM, Jouni Malinen wrote:
On Wed, Sep 15, 2021 at 07:54:35PM -0700, Aloka Dixit wrote:
diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c
+static int ieee80211_set_ap_mbssid_options(struct ieee80211_sub_if_data *sdata,
+ struct cfg80211_mbssid_config params)
@@ -1268,10 +1268,17 @@ static int ieee80211_start_ap(struct wiphy *wiphy, struct net_device *dev,
if (sdata->vif.type == NL80211_IFTYPE_AP &&
params->mbssid_config.tx_wdev) {
err = ieee80211_set_ap_mbssid_options(sdata,
- params->mbssid_config,
+ ¶ms->mbssid_config,
link_conf);
if (err)
return err;
+ } else {
+ /* FIX: Is this the correct thing to do here and under which
+ * conditions? At least ema_ap needs to be cleared for AP mode
+ * if mbssid_config.tx_wdev is not set. */
+ link_conf->bssid_index = 0;
+ link_conf->nontransmitted = false;
+ link_conf->ema_ap = false;
Will need to clean MBSSID elements in stored beacon as well.
Would the two HWSIM testcases you mentioned earlier be sufficient for
testing? I've never used P2P related tests, and in fact I only even ran
one testcase at a time so stale configurations didn't show up.
Thanks.
