Thread (2 messages) 2 messages, 2 authors, 2008-06-03

Re: : Emit event stream compat iw_point objects correctly.

From: David Miller <davem@davemloft.net>
Date: 2008-06-03 16:57:53
Also in: netdev

From: Masakazu Mokuno <redacted>
Date: Fri, 11 Jan 2008 13:16:29 +0900
On Thu, 10 Jan 2008 01:16:02 -0800 (PST)
David Miller [off-list ref] wrote:
quoted
diff --git a/include/net/iw_handler.h b/include/net/iw_handler.h
index c99a8ee..d6f0c51 100644
--- a/include/net/iw_handler.h
+++ b/include/net/iw_handler.h
@@ -483,19 +483,26 @@ extern void wireless_spy_update(struct net_device *	dev,
  * Wrapper to add an Wireless Event to a stream of events.
  */
 static inline char *
-iwe_stream_add_event(char *	stream,		/* Stream of events */
-		     char *	ends,		/* End of stream */
-		     struct iw_event *iwe,	/* Payload */
-		     int	event_len)	/* Real size of payload */
+iwe_stream_add_event(struct iw_request_info *info, char *stream, char *ends,
+		     struct iw_event *iwe, int event_len)
 {
+	int lcp_len = IW_EV_LCP_LEN;
+
+#ifdef CONFIG_COMPAT
+	if (info->flags & IW_REQUEST_FLAG_COMPAT) {
+		event_len -= IW_EV_LCP_LEN;
+		event_len += IW_EV_COMPAT_LCP_LEN;
+		lcp_len = IW_EV_COMPAT_LCP_LEN;
+	}
+#endif
 	/* Check if it's possible */
 	if(likely((stream + event_len) < ends)) {
 		iwe->len = event_len;
 		/* Beware of alignement issues on 64 bits */
 		memcpy(stream, (char *) iwe, IW_EV_LCP_PK_LEN);
-		memcpy(stream + IW_EV_LCP_LEN,
-		       ((char *) iwe) + IW_EV_LCP_LEN,
-		       event_len - IW_EV_LCP_LEN);
+		memcpy(stream + lcp_len,
+		       ((char *) iwe) + lcp_len,
The source address does not have to be adjusted.  I think it should be
		       ((char *) iwe) + IW_EV_LCP_LEN,
or just
		       &iwe->u
I think it is more readable what we want to do here.

quoted
+		       event_len - lcp_len);
The length of the real payload does not need to take account of the
destination offset change.  So we can keep the original code like:

		       org_event_len - IW_EV_LCP_LEN);
As long as IW_EV_LCP_LEN and IW_EV_COMPAT_LCP_LEN are different, I am
pretty sure we need to make these adjustments.

There are so many differences in how struct vs. sub-struct alignment will
occur in these cases, that we need to be very careful.

I'll try to look more closely at this and do some testing of my own on
a compat platform.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help