From: Simon Wunderlich <[email protected]>
Responding to probe requests for scanning clients will often create
excessive rtries, as it happens quite often that the scanning client
already left the channel. Therefore do it like hostapd and send probe
responses for broadcast SSID only once by using the noack flag.
Signed-off-by: Simon Wunderlich <[email protected]>
---
net/mac80211/ibss.c | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c
index ed7eec3..e22eb95 100644
--- a/net/mac80211/ibss.c
+++ b/net/mac80211/ibss.c
@@ -1465,6 +1465,11 @@ static void ieee80211_rx_mgmt_probe_req(struct ieee80211_sub_if_data *sdata,
memcpy(((struct ieee80211_mgmt *) skb->data)->da, mgmt->sa, ETH_ALEN);
ibss_dbg(sdata, "Sending ProbeResp to %pM\n", mgmt->sa);
IEEE80211_SKB_CB(skb)->flags |= IEEE80211_TX_INTFL_DONT_ENCRYPT;
+
+ /* avoid excessive retries for probe request to wildcard SSIDs */
+ if (pos[1] == 0)
+ IEEE80211_SKB_CB(skb)->flags |= IEEE80211_TX_CTL_NO_ACK;
+
ieee80211_tx_skb(sdata, skb);
}
--
1.7.10.4
On Fri, 2014-01-24 at 23:48 +0100, Simon Wunderlich wrote:
> From: Simon Wunderlich <[email protected]>
>
> Responding to probe requests for scanning clients will often create
> excessive rtries, as it happens quite often that the scanning client
> already left the channel. Therefore do it like hostapd and send probe
> responses for broadcast SSID only once by using the noack flag.
Applied. Should have said "wildcard" though, technically :)
johannes