networking fails to activate - even more info
Pieter
phenning at vastech.co.za
Fri Oct 17 02:02:51 EST 2008
>>> Hi
>>>
>>> We have a powerpc (8548) based board. running eldk4.1. the board has
>>> three ethernet ports using Gainfair ethernet controller v1.2.
>>>
>>> If the board powers up with any network cable plugged in, that specific
>>> network is accessible.
>>> if the board powers up without network cables connected and a cable is
>>> plugged in, the network does not become active.
>>> The kernal makes a log entry stating that the phy is up when the cable
>>> is plugged in - but the interface is not pingable
>>> Running ifdown and ifup manually does not result in a pingable interface.
>>> Neither does running /etc/init.d/network restart
>>>
>>> can anyone help me pls.
>>>
>>>
>> Check the statistics counters to see if you are getting any bad
>> packets, or to see if you are getting interrupts. Several people have
>> reported that their PHYs have powerdown issues, though I wouldn't
>> expect the PHY to report the link as up, then. Make sure that your
>> dts specifies the correct PHY address for the connected PHY.
>>
>> What happens if you start up with the PHY connected, bring up the
>> interface, and *then* disconnect and reconnect the cable (after a
>> delay)?
>>
> Hi
>
> I have checked the counters in both ifconfig and with ethtool. both show
> TX counts but no RX count and no errors
> ethtool reports that the link is detected.
> The phy addresses are correct (ethtool and kernel log shows the
> exspected addresses)
> I have found that the networking scripts are not called (ifup or down)
> when a cable is pluged or unplugged.
> I also found that udev is not in my current linux platform - can this
> contribute to the problem?
>
> To rule out the phys. I powered on the system with all network cables
> plugged in. ping all three to confirm that they work. i then issue a
> reset signal (hardware routed) to the phys. when the phys reappear im
> still able to ping all three interfaces.
>
> The log file entries for unplugging and connecting the cable are :
> (logged by /driver/phy/phy.c )
> Oct 15 14:36:56 equus user.info kernel: PHY: 0:00 - Link is
> Down
> Oct 15 14:36:59 equus user.info kernel: PHY: 0:00 - Link is Up - 100/Full
> what is the sequence of events that should follow?
>
> Another test i tried, powered the system up with no cables plugged in
> and "ONBOOT=no" in ifcgf-eth sctipts. I then manually run the ifup
> scripts one by one. then plug in the three network cables. when i ping
> the three interfaces and ping targets on the three networks i find that
> one interface is pingable, one is not and one returns "ping: sendto:
> Network is unreachable" - these results were reproducible about 25% of
> attempts but varied according to order in with cables were plugged in.
>
> I would appreciate any help.
> thanks pieter
>
hi
I did some more digging and testing today. To recap the problem: When i
power on the system with network cables plugged in, the interface is
pingable. If the system powers up without network cables plugged in,
ifconfig reports tat all interfaces are up and running. but when
connecting the network cables none of the interfaces are pingable (even
after running ifdown and ifup scripts for the interfaces. )
I think the problem lies some ware in that both ifconfig and ethtool
reports the link as being up even though the system was powered up
without network cables plugged in.
Ifconfigs output:
SDH-25108>
ifconfig
eth0 Link encap:Ethernet HWaddr
00:50:C2:52:51:08
inet addr:10.0.0.200 Bcast:10.0.0.255
Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500
Metric:1
RX packets:0 errors:0 dropped:0 overruns:0
frame:0
TX packets:394 errors:0 dropped:0 overruns:0
carrier:0
collisions:0
txqueuelen:1000
RX bytes:0 (0.0 b) TX bytes:16548 (16.1
KiB)
Base
address:0xe000
eth1 Link encap:Ethernet HWaddr
00:50:C2:52:51:09
inet addr:172.16.63.237 Bcast:172.16.63.255
Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500
Metric:1
RX packets:8676 errors:0 dropped:0 overruns:0
frame:0
TX packets:149 errors:0 dropped:0 overruns:0
carrier:0
collisions:0
txqueuelen:1000
RX bytes:942270 (920.1 KiB) TX bytes:10268 (10.0
KiB)
eth2 Link encap:Ethernet HWaddr
00:50:C2:52:51:0A
inet addr:20.0.0.200 Bcast:20.0.0.255
Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500
Metric:1
RX packets:0 errors:0 dropped:0 overruns:0
frame:0
TX packets:4 errors:0 dropped:0 overruns:0
carrier:0
collisions:0
txqueuelen:1000
RX bytes:0 (0.0 b) TX bytes:168 (168.0
b)
Base
address:0x2000
lo Link encap:Local
Loopback
inet addr:127.0.0.1
Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:16436
Metric:1
RX packets:3733 errors:0 dropped:0 overruns:0
frame:0
TX packets:3733 errors:0 dropped:0 overruns:0
carrier:0
collisions:0
txqueuelen:0
RX bytes:674220 (658.4 KiB) TX bytes:674220 (658.4 KiB)
ethtools output:
SDH-25108> ethtool
eth0
Settings for
eth0:
Supported ports: [ TP AUI BNC MII FIBRE
]
Supported link modes: 10baseT/Half
10baseT/Full
100baseT/Half
100baseT/Full
1000baseT/Half
1000baseT/Full
Supports auto-negotiation:
Yes
Advertised link modes: 10baseT/Half
10baseT/Full
100baseT/Half
100baseT/Full
1000baseT/Half
1000baseT/Full
Advertised auto-negotiation:
Yes
Speed:
1000Mb/s
Duplex:
Half
Port:
MII
PHYAD:
0
Transceiver:
external
Auto-negotiation:
off
Current message level: 0x0000003f
(63)
Link detected: yes
If I then plug in the cables and unplug them again the link status goes
down. When plugging in the cables again link status is detected again,
but none of teh interfaces are pingable - even after trying ifdown and up.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 2722 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://ozlabs.org/pipermail/linuxppc-embedded/attachments/20081016/710c1e6a/attachment-0001.bin>
More information about the Linuxppc-embedded
mailing list