| Message ID | 20191209102808.12014-5-nicolas.cavallari@green-communications.fr (mailing list archive) |
|---|---|
| State | RFC |
| Delegated to: | Johannes Berg |
| Headers | show |
| Series | 802.11n IBSS: wlan0 stops receiving packets due to aggregation after sender reboot | expand |
On Mon, 2019-12-09 at 11:28 +0100, Nicolas Cavallari wrote: > > diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c > index da5262b2298b..82046c990a2a 100644 > --- a/net/wireless/nl80211.c > +++ b/net/wireless/nl80211.c > @@ -6073,7 +6073,8 @@ static int nl80211_del_station(struct sk_buff *skb, struct genl_info *info) > if (dev->ieee80211_ptr->iftype != NL80211_IFTYPE_AP && > dev->ieee80211_ptr->iftype != NL80211_IFTYPE_AP_VLAN && > dev->ieee80211_ptr->iftype != NL80211_IFTYPE_MESH_POINT && > - dev->ieee80211_ptr->iftype != NL80211_IFTYPE_P2P_GO) > + dev->ieee80211_ptr->iftype != NL80211_IFTYPE_P2P_GO && > + dev->ieee80211_ptr->iftype != NL80211_IFTYPE_ADHOC) > return -EINVAL; If we go this route (and I'm not sure, shouldn't be _that_ hard to do some kind of auth/deauth thing?) then you probably should make this depend on an nl80211 extended feature bit. That way, not only do you get visibility in userspace whether it's supported, but also avoid the need to change those non-mac80211 drivers (by having only mac80211 set the extended feature) johannes
On 11/12/2019 22:32, Johannes Berg wrote: > and I'm not sure, shouldn't be _that_ hard to do some kind of auth/deauth > thing? wpa_supplicant already does this for IBSS RSN. And it does not reset the station for each received auth frame, the logic is more advanced than that. (not to mention that it should switch to SAE at some point, but there is this long standing issue of wpa_supplicant not being able to get the rsn ie of the peer's probe response) But for the unencrypted case, i guess simply re-reverting 52874a5e (Revert "mac80211: in IBSS use the Auth frame to trigger STA reinsertion") and making it conditional on !privacy should be enough ? > That way, not only do you get visibility in userspace whether it's > supported, but also avoid the need to change those non-mac80211 drivers > (by having only mac80211 set the extended feature) Ok.
diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index da5262b2298b..82046c990a2a 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -6073,7 +6073,8 @@ static int nl80211_del_station(struct sk_buff *skb, struct genl_info *info) if (dev->ieee80211_ptr->iftype != NL80211_IFTYPE_AP && dev->ieee80211_ptr->iftype != NL80211_IFTYPE_AP_VLAN && dev->ieee80211_ptr->iftype != NL80211_IFTYPE_MESH_POINT && - dev->ieee80211_ptr->iftype != NL80211_IFTYPE_P2P_GO) + dev->ieee80211_ptr->iftype != NL80211_IFTYPE_P2P_GO && + dev->ieee80211_ptr->iftype != NL80211_IFTYPE_ADHOC) return -EINVAL; if (!rdev->ops->del_station)
Sometimes, userspace is able to detect that a peer silently lost its state (like, if the peer reboots). wpa_supplicant does this for IBSS-RSN and currently only resets the key of the peer so that it can attempt another handshake. However, the kernel also hold state about the station, such as BA sessions, probe response parameters and the like. They also need to be resetted correctly. This patch allow userspace to use NL80211_CMD_DEL_STATION in IBSS mode, which should send a deauth and reset the state of the station, just like in mesh point mode. This has successfully been tested with mac80211/ath9k. Drivers that do not support it should just return an error, so that userspace can fall back to the old methods. Signed-off-by: Nicolas Cavallari <nicolas.cavallari@green-communications.fr> --- net/wireless/nl80211.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)