Bug 902371

Description of problem:


Version-Release number of selected component (if applicable):


How reproducible:
always

Steps to Reproduce:
0. stop all network-related services

1.
/etc/sysconfig/network-scripts/ifcfg-br0:
DEVICE=br0
TYPE=Bridge
BOOTPROTO=dhcp
ONBOOT=yes
DELAY=0
STP=off
NM_CONTROLLED=yes
HWADDR=f0:de:f1:04:c0:fa  # MAC address of eth0

/etc/sysconfig/network-scripts/ifcfg-eth0
DEVICE=eth0
BRIDGE=br0
ONBOOT=yes
BOOTPROTO=none
NM_CONTROLLED=no   # or yes, it doesn't affect reproducer yet

user-defined connection using eth0 device:
$ gconftool-2 -R /system/networking/connections/1
 /system/networking/connections/1/ipv4:
  addresses = []
  name = ipv4
  dns = []
  routes = []
  method = auto
 /system/networking/connections/1/ipv6:
  addresses = []
  name = ipv6
  dns = []
  routes = []
  method = auto
  may-fail = true
 /system/networking/connections/1/connection:
  uuid = 9b887145-365a-447a-900e-4ad3a80bf157
  name = connection
  id = Auto Ethernet
  type = 802-3-ethernet
  timestamp = 1358772786
 /system/networking/connections/1/802-3-ethernet:
  name = 802-3-ethernet
  duplex = full

2. log in to gnome session as a user to whom "Auto Ethernet" connection belongs and make sure that nm-applet is running
3. service network start
4. service NetworkManager start
  
Actual results:
3. br0 is enabled and configured successfully by network-scripts (good)
4. nm-applet senses that it "can" reconnect "Auto Ethernet" connection
5a bridge is not recognized (br0 device is "unavailable" according to nmcli)
5b NM daemon happily configures "Auto Ethernet" connection on eth0 despite:
  * the device already being part of the bridge
  * NM_CONTROLLED=no in ifcfg-eth0 (that's for separate bug)

Expected results:
4 NM daemon should see that eth0 is a part of a bridge and it should not "offer" it to the clients at all
5. even if the eth0 is visible to clients such as nm-applet who ask to enable the device, NM should refuse

Additional info: