Thread (2 messages) flat view 2 messages, 2 authors, 2014-02-21

Re: [PATCH 1/1] rtlwifi: Convert drivers to use new API for ieee80211_is_robust_mgmt_frame()

From: Larry Finger <hidden>
Date: 2014-02-21 15:03:43
Also in: linux-wireless

On 02/21/2014 08:38 AM, Johannes Berg wrote:
On Fri, 2014-02-21 at 08:33 -0600, Larry Finger wrote:
quoted
Drivers rtl8188ee, rtl8192ce, rtl8192se, and rtl8723ae all use this routine.
In addition, all callers of ieee80211_is_robust_mgmt_frame() have access to the
skb. Thus, it is trivial to use the new API rather than the underscored old version.
Unfortunately, I'm pretty sure this is incorrect. I looked at this code
before, and take for example:
quoted
+++ b/drivers/net/wireless/rtlwifi/rtl8723ae/trx.c
@@ -334,8 +334,8 @@ bool rtl8723ae_rx_query_desc(struct ieee80211_hw *hw,
  			/* during testing, hdr could be NULL here */
  			return false;
  		}
-		if ((_ieee80211_is_robust_mgmt_frame(hdr)) &&
-			(ieee80211_has_protected(hdr->frame_control)))
+		if ((ieee80211_is_robust_mgmt_frame(skb)) &&
+		    (ieee80211_has_protected(hdr->frame_control)))
Here, hdr is set like this:

                 hdr = (struct ieee80211_hdr *)(skb->data +
                        status->rx_drvinfo_size + status->rx_bufshift);

so the skb clearly contains something else before the hdr, and the
conversion is probably not right because it assumes hdr == skb->data,
where here it isn't that but instead (skb->data + something).
I thought I had tested the change, but I'm not sure if my AP is using 
ieee802.11w. I need to test this better.

@John: Please drop this patch.

Larry
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help