dhcpcd-discuss

Re: Can't ping 169 autoip address with multiple NICs

Dallas Clement

Mon Dec 12 16:42:59 2011

I tried the latest development version and got the same results, i.e.
all of my uncabled interfaces are getting assigned a 169 address.

Here are the printk's that are going to my syslog from the driver:

55821:Dec 13 00:55:14 kernel: [234971.511479] e1000e: eth0 NIC Link is Down
55868:Dec 13 00:55:31 kernel: [234988.523742] e1000e: eth0 NIC Link is
Up 1000 Mbps Full Duplex, Flow Control: Rx/Tx
56047:Dec 13 00:56:18 kernel: [235034.601983] e1000e: eth0 NIC Link is
Up 1000 Mbps Full Duplex, Flow Control: Rx/Tx
57042:Dec 13 01:02:24 kernel: [235400.276320] e1000e: eth0 NIC Link is
Up 1000 Mbps Full Duplex, Flow Control: Rx/Tx
59235:Dec 13 01:10:35 kernel: [   39.630158] e1000e: Intel(R) PRO/1000
Network Driver - 1.6.3-NAPI
59236:Dec 13 01:10:35 kernel: [   39.636270] e1000e: Copyright(c) 1999
- 2011 Intel Corporation.
59237:Dec 13 01:10:35 kernel: [   39.642219] e1000e 0000:04:00.0:
Disabling ASPM L0s
59238:Dec 13 01:10:35 kernel: [   39.647215] e1000e 0000:04:00.0: PCI
INT A -> GSI 16 (level, low) -> IRQ 16
59239:Dec 13 01:10:35 kernel: [   39.868480] e1000e 0000:04:00.0:
eth0: (PCI Express:2.5GT/s:Width x1) cc:52:af:ea:7e:d8
59240:Dec 13 01:10:35 kernel: [   39.876956] e1000e 0000:04:00.0:
eth0: Intel(R) PRO/1000 Network Connection
59241:Dec 13 01:10:35 kernel: [   39.884010] e1000e 0000:04:00.0:
eth0: MAC: 4, PHY: 8, PBA No: FFFFFF-0FF
59242:Dec 13 01:10:35 kernel: [   39.890815] e1000e 0000:05:00.0:
Disabling ASPM L0s
59243:Dec 13 01:10:35 kernel: [   39.895804] e1000e 0000:05:00.0: PCI
INT A -> GSI 17 (level, low) -> IRQ 17
59244:Dec 13 01:10:35 kernel: [   40.079673] e1000e 0000:05:00.0:
eth1: (PCI Express:2.5GT/s:Width x1) cc:52:af:ea:7e:d9
59245:Dec 13 01:10:35 kernel: [   40.087697] e1000e 0000:05:00.0:
eth1: Intel(R) PRO/1000 Network Connection
59246:Dec 13 01:10:35 kernel: [   40.094753] e1000e 0000:05:00.0:
eth1: MAC: 4, PHY: 8, PBA No: FFFFFF-0FF
59247:Dec 13 01:10:35 kernel: [   40.101552] e1000e 0000:06:00.0:
Disabling ASPM L0s
59248:Dec 13 01:10:35 kernel: [   40.106546] e1000e 0000:06:00.0: PCI
INT A -> GSI 18 (level, low) -> IRQ 18
59249:Dec 13 01:10:35 kernel: [   40.279106] e1000e 0000:06:00.0:
eth2: (PCI Express:2.5GT/s:Width x1) cc:52:af:ea:7e:da
59250:Dec 13 01:10:35 kernel: [   40.287128] e1000e 0000:06:00.0:
eth2: Intel(R) PRO/1000 Network Connection
59251:Dec 13 01:10:35 kernel: [   40.294182] e1000e 0000:06:00.0:
eth2: MAC: 4, PHY: 8, PBA No: FFFFFF-0FF
59252:Dec 13 01:10:35 kernel: [   40.300980] e1000e 0000:07:00.0:
Disabling ASPM L0s
59253:Dec 13 01:10:35 kernel: [   40.305975] e1000e 0000:07:00.0: PCI
INT A -> GSI 19 (level, low) -> IRQ 19
59254:Dec 13 01:10:35 kernel: [   40.478535] e1000e 0000:07:00.0:
eth3: (PCI Express:2.5GT/s:Width x1) cc:52:af:ea:7e:db
59255:Dec 13 01:10:35 kernel: [   40.486555] e1000e 0000:07:00.0:
eth3: Intel(R) PRO/1000 Network Connection
59256:Dec 13 01:10:35 kernel: [   40.493610] e1000e 0000:07:00.0:
eth3: MAC: 4, PHY: 8, PBA No: FFFFFF-0FF
59361:Dec 13 01:10:35 kernel: [  121.472787] e1000e: eth0 NIC Link is
Up 1000 Mbps Full Duplex, Flow Control: Rx/Tx

As you can see, eth0 is the only one that reports a link up transition.

Here is what ifconfig is reporting:

root@bubba:# ifconfig
eth0      Link encap:Ethernet  HWaddr CC:52:AF:EA:7E:D8
          inet addr:192.168.1.142  Bcast:192.168.1.255  Mask:255.255.254.0
          UP BROADCAST NOTRAILERS RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:253104 errors:0 dropped:25 overruns:0 frame:0
          TX packets:7686 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:48025107 (45.7 MiB)  TX bytes:1303902 (1.2 MiB)
          Interrupt:16 Memory:fbb00000-fbb20000

eth1      Link encap:Ethernet  HWaddr CC:52:AF:EA:7E:D9
          inet addr:169.254.49.74  Bcast:169.254.255.255  Mask:255.255.0.0
          UP BROADCAST NOTRAILERS MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)
          Interrupt:17 Memory:fba00000-fba20000

eth2      Link encap:Ethernet  HWaddr CC:52:AF:EA:7E:DA
          inet addr:169.254.15.63  Bcast:169.254.255.255  Mask:255.255.0.0
          UP BROADCAST NOTRAILERS MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)
          Interrupt:18 Memory:fb900000-fb920000

eth3      Link encap:Ethernet  HWaddr CC:52:AF:EA:7E:DB
          inet addr:169.254.103.123  Bcast:169.254.255.255  Mask:255.255.0.0
          UP BROADCAST NOTRAILERS MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)
          Interrupt:19 Memory:fb800000-fb820000

And here's what sysfs thinks:

root@bubba:/var/log# cat /sys/class/net/eth0/operstate
up
root@bubba:/var/log# cat /sys/class/net/eth1/operstate
down
root@bubba:/var/log# cat /sys/class/net/eth2/operstate
down
root@bubba:/var/log# cat /sys/class/net/eth3/operstate
down

Does dhcpcd have any debugging that can help us observe why it is
giving a 169 address to a down interface?

On Sat, Dec 10, 2011 at 5:43 AM, Roy Marples <roy@xxxxxxxxxxxx> wrote:
> On Fri, 2011-12-09 at 21:47 -0600, Dallas Clement wrote:
>> Roy,  Thank you for the speedy and most helpful response.  I am indeed
>>  seeing at least one extra spurious link down / up from the driver,
>>  although it's only for the cabled interface.  Ifconfig  shows the
>>  other 3 interfaces as down, as does /sys/class/net/ethX/overstate.  I
>>  am using the very latest e1000e driver.  How is dhcpcd determining
>>  carrier presence?
>
> dhcpcd determines a working carrier from the IFF_RUNNING flag against
> the interface and listens to RTM_NEWLINK and RTM_DELLINK kernel
> messages.
>
> Although this has changed slightly in my latest development version. Why
> not try a trunk copy? You can get the latest code doing this with git:
>
> git clone git://roy.marples.name/dhcpcd.git
>
> Thanks
>
> Roy
>

Follow-Ups:
Re: Can't ping 169 autoip address with multiple NICsDallas Clement
References:
Can't ping 169 autoip address with multiple NICsDallas Clement
Re: Can't ping 169 autoip address with multiple NICsRoy Marples
Re: Can't ping 169 autoip address with multiple NICsDallas Clement
Re: Can't ping 169 autoip address with multiple NICsRoy Marples
Archive administrator: postmaster@marples.name