From: Johannes Berg johannes.berg@intel.com
[ Upstream commit 67dfa589aa8806c7959cbca2f4613b8d41c75a06 ]
When probing a client, first check if we have it, and then check for the channel context, otherwise you can trigger the warning there easily by probing when the AP isn't even started yet. Since a client existing means the AP is also operating, we can then keep the warning.
Also simplify the moved code a bit.
Reported-by: syzbot+999fac712d84878a7379@syzkaller.appspotmail.com Signed-off-by: Johannes Berg johannes.berg@intel.com Signed-off-by: Alexander Ofitserov oficerovas@altlinux.org Cc: stable@vger.kernel.org --- net/mac80211/cfg.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-)
diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index 45bb6f27559877..8aef97b92c5213 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -3642,19 +3642,20 @@ static int ieee80211_probe_client(struct wiphy *wiphy, struct net_device *dev, mutex_lock(&local->mtx);
rcu_read_lock(); + sta = sta_info_get_bss(sdata, peer); + if (!sta) { + ret = -ENOLINK; + goto unlock; + } + + qos = sta->sta.wme; + chanctx_conf = rcu_dereference(sdata->vif.chanctx_conf); if (WARN_ON(!chanctx_conf)) { ret = -EINVAL; goto unlock; } band = chanctx_conf->def.chan->band; - sta = sta_info_get_bss(sdata, peer); - if (sta) { - qos = sta->sta.wme; - } else { - ret = -ENOLINK; - goto unlock; - }
if (qos) { fc = cpu_to_le16(IEEE80211_FTYPE_DATA |