Re: NM not connecting on resume, ipw2200

[Dan Williams <dcbw redhat com>]

On Thu, 2006-07-13 at 09:40 -0400, Derek Atkins wrote:
> Bill Moss <bmoss CLEMSON EDU> writes:
> 
> > Some recent patches to NM and ipw2200 should take care of the failure to 
> > connect on resume from suspend.
> >
> > 1. NM failed to set the essid to ANY but instead left the essid 
> > unchanged when deactivating a wireless network. Consequently, when NM 
> > resumed the essid still had its previous value. Dan applied patches at 
> > HEAD 1.934 to fix this.
> 
> What file was changed?  Is/has this patch been backported to 0.6?
> And would this affect madwifi as well?

src/nm-device-802-11-wireless.c, revision 1.69 -> 1.70

It doesn't appear to be in STABLE_06, patch against 0.6 attached.
Please test this out and see if there are problems.  I think I
originally didn't backport it to 0.6 because stuff was working fine, and
if not broken, why fix it.  So if this patch fixes any of your problems,
and doesn't present new ones, please let me know.

Dan

? nm-clear-settings.patch
? po/stamp-it
? vpn-daemons/vpnc/NetworkManager-vpnc-0.5.0
? vpn-daemons/vpnc/NetworkManager-vpnc-0.6.0
Index: src/nm-device-802-11-wireless.c
===================================================================
RCS file: /cvs/gnome/NetworkManager/src/nm-device-802-11-wireless.c,v
retrieving revision 1.60.2.15
diff -u -r1.60.2.15 nm-device-802-11-wireless.c
--- src/nm-device-802-11-wireless.c	20 Jun 2006 02:24:57 -0000	1.60.2.15
+++ src/nm-device-802-11-wireless.c	13 Jul 2006 16:13:46 -0000
@@ -131,9 +131,7 @@
 
 static void		remove_link_timeout (NMDevice80211Wireless *self);
 
-static void		nm_device_802_11_wireless_set_wep_enc_key (NMDevice80211Wireless *self,
-                                           const char *key,
-                                           int auth_method);
+static void		nm_device_802_11_wireless_disable_encryption (NMDevice80211Wireless *self);
 
 
 /*
@@ -522,6 +520,10 @@
 
 	supplicant_cleanup (self);
 	remove_link_timeout (self);
+
+	/* Clean up stuff, don't leave the card associated */
+	nm_device_802_11_wireless_set_essid (self, "");
+	nm_device_802_11_wireless_disable_encryption (self);
 }
 
 
@@ -534,9 +536,6 @@
 	app_data = nm_device_get_app_data (dev);
 	g_assert (app_data);
 
-	/* Clean up stuff, don't leave the card associated */
-	nm_device_802_11_wireless_set_essid (self, "");
-	nm_device_802_11_wireless_set_wep_enc_key (self, NULL, 0);
 	nm_device_802_11_wireless_set_mode (self, IW_MODE_INFRA);
 	nm_device_802_11_wireless_set_scan_interval (app_data, self, NM_WIRELESS_SCAN_INTERVAL_ACTIVE);
 }
@@ -1596,91 +1595,36 @@
 
 
 /*
- * nm_device_set_wep_enc_key
+ * nm_device_802_11_wireless_disable_encryption
  *
- * If a device is wireless, set the encryption key that it should use.
+ * Clear any encryption keys the device may have set.
  *
- * key:	encryption key to use, or NULL or "" to disable encryption.
  */
 static void
-nm_device_802_11_wireless_set_wep_enc_key (NMDevice80211Wireless *self,
-                                           const char *key,
-                                           int auth_method)
+nm_device_802_11_wireless_disable_encryption (NMDevice80211Wireless *self)
 {
+	const char * iface = nm_device_get_iface (NM_DEVICE (self));
 	NMSock *		sk;
-	struct iwreq	wreq;
-	int			keylen;
-	unsigned char	safe_key[IW_ENCODING_TOKEN_MAX + 1];
-	gboolean		set_key = FALSE;
-	const char *	iface;
 
 	g_return_if_fail (self != NULL);
 
-	/* Make sure the essid we get passed is a valid size */
-	if (!key)
-		safe_key[0] = '\0';
-	else
-	{
-		strncpy ((char *) safe_key, key, IW_ENCODING_TOKEN_MAX);
-		safe_key[IW_ENCODING_TOKEN_MAX] = '\0';
-	}
-
-	iface = nm_device_get_iface (NM_DEVICE (self));
 	if ((sk = nm_dev_sock_open (NM_DEVICE (self), DEV_WIRELESS, __FUNCTION__, NULL)))
 	{
-		wreq.u.data.pointer = (caddr_t) NULL;
-		wreq.u.data.length = 0;
-		wreq.u.data.flags = IW_ENCODE_ENABLED;
-
-		/* Unfortunately, some drivers (Cisco) don't make a distinction between
-		 * Open System authentication mode and whether or not to use WEP.  You
-		 * DON'T have to use WEP when using Open System, but these cards force
-		 * it.  Therefore, we have to set Open System mode when using WEP.
-		 */
+		struct iwreq	wreq = {
+			.u.data.pointer = (caddr_t) NULL,
+			.u.data.length = 0,
+			.u.data.flags = IW_ENCODE_DISABLED
+		};
 
-		if (strlen ((char *) safe_key) == 0)
-		{
-			wreq.u.data.flags |= IW_ENCODE_DISABLED | IW_ENCODE_NOKEY;
-			set_key = TRUE;
-		}
-		else
-		{
-			unsigned char		parsed_key[IW_ENCODING_TOKEN_MAX + 1];
-
-			keylen = iw_in_key_full (nm_dev_sock_get_fd (sk), iface,
-						(char *) safe_key, &parsed_key[0], &wreq.u.data.flags);
-			if (keylen > 0)
-			{
-				switch (auth_method)
-				{
-					case IW_AUTH_ALG_OPEN_SYSTEM:
-						wreq.u.data.flags |= IW_ENCODE_OPEN;
-						break;
-					case IW_AUTH_ALG_SHARED_KEY:
-						wreq.u.data.flags |= IW_ENCODE_RESTRICTED;
-						break;
-					default:
-						wreq.u.data.flags |= IW_ENCODE_RESTRICTED;
-						break;
-				}
-				wreq.u.data.pointer	=  (caddr_t) &parsed_key;
-				wreq.u.data.length	=  keylen;
-				set_key = TRUE;
-			}
-		}
-
-		if (set_key)
-		{
 #ifdef IOCTL_DEBUG
-			nm_info ("%s: About to SET IWENCODE.", iface);
+		nm_info ("%s: About to SET IWENCODE.", iface);
 #endif
-			if (iw_set_ext (nm_dev_sock_get_fd (sk), iface, SIOCSIWENCODE, &wreq) == -1)
+		if (iw_set_ext (nm_dev_sock_get_fd (sk), iface, SIOCSIWENCODE, &wreq) == -1)
+		{
+			if (errno != ENODEV)
 			{
-				if (errno != ENODEV)
-				{
-					nm_warning ("error setting key for device %s: %s",
-							iface, strerror (errno));
-				}
+				nm_warning ("error setting key for device %s: %s",
+						iface, strerror (errno));
 			}
 		}