[MAC80211]: don't send invalid QoS frames
authorJohannes Berg <johannes@sipsolutions.net>
Fri, 14 Sep 2007 15:10:24 +0000 (11:10 -0400)
committerDavid S. Miller <davem@sunset.davemloft.net>
Wed, 10 Oct 2007 23:49:26 +0000 (16:49 -0700)
Kalle Valo noticed that QoS frames are sent with an invalid QoS control
field; this is because we increase the header length but neither
initialise the space nor actually have enough space in the header
structure for the QoS control field.

This patch fixes it by treating the QoS field specially and appending it
explicitly, initialising it to zero.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Acked-by: Michael Wu <flamingice@sourmilk.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/mac80211/tx.c

index 9e952e37b7dfa12d554d2d2226740089494f2c6e..0820f127da2b34dfe8d5698b4ace996b143c7675 100644 (file)
@@ -1487,7 +1487,20 @@ int ieee80211_subif_start_xmit(struct sk_buff *skb,
                nh_pos += encaps_len;
                h_pos += encaps_len;
        }
-       memcpy(skb_push(skb, hdrlen), &hdr, hdrlen);
+
+       if (fc & IEEE80211_STYPE_QOS_DATA) {
+               __le16 *qos_control;
+
+               qos_control = (__le16*) skb_push(skb, 2);
+               memcpy(skb_push(skb, hdrlen - 2), &hdr, hdrlen - 2);
+               /*
+                * Maybe we could actually set some fields here, for now just
+                * initialise to zero to indicate no special operation.
+                */
+               *qos_control = 0;
+       } else
+               memcpy(skb_push(skb, hdrlen), &hdr, hdrlen);
+
        nh_pos += hdrlen;
        h_pos += hdrlen;