Bug 52021 - e1000e unsupported speed/duplex
Summary: e1000e unsupported speed/duplex
Alias: None
Product: Drivers
Classification: Unclassified
Component: Network (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: drivers_network@kernel-bugs.osdl.org
Depends on:
Reported: 2012-12-26 18:42 UTC by onorua
Modified: 2021-08-18 08:34 UTC (History)
5 users (show)

See Also:
Kernel Version: 3.8-rc1
Regression: Yes
Bisected commit-id:

lspci output (10.38 KB, application/octet-stream)
2013-01-04 06:14 UTC, onorua
Dmesg with 3.8-rc2 (20.76 KB, text/plain)
2013-01-04 11:11 UTC, onorua

Description onorua 2012-12-26 18:42:31 UTC
After upgrade to 3.8-rc1, and suspend/resume/reboot I got error message:
e100e unsupported speed/duplex 

Looks similar to Bug 36132. Ethernet doesn't work:
# ethtool eth0
Settings for eth0:
Cannot get device settings: No such device
Cannot get wake-on-lan settings: No such device
Cannot get message level: No such device
Cannot get link status: No such device
No data available

with 3.5.5 kernel works as expected. 
Please let me know if you need more info.
Comment 1 Tushar 2013-01-02 21:53:35 UTC
I will check if this is a known bug.
Does this bug happen to occur with all kernels > 3.5.5?
Comment 2 onorua 2013-01-03 01:20:29 UTC
I was observing it only on 3.8-rc1, but to be honest, I was not extensively checking it on other kernels due to bug 48721 which occur starting from 3.5.5.
Comment 3 Tushar 2013-01-03 20:30:02 UTC
I installed 3.8-rc1 but don't see the issue.
What is the network device are you using (lspci -vvv)
What is the output of ifconfig eth0 on your failing system.

Comment 4 onorua 2013-01-04 06:14:11 UTC
Created attachment 90381 [details]
lspci output

I have ThinkPad x220
eth0      Link encap:Ethernet  HWaddr f0:de:f1:5d:33:37  
          UP BROADCAST MULTICAST  MTU:1500  Metric:1
          RX packets:609374 errors:0 dropped:0 overruns:0 frame:0
          TX packets:692882 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:477726923 (455.5 MiB)  TX bytes:497892520 (474.8 MiB)
          Interrupt:20 Memory:f2500000-f2520000 

Probbaly this is not problem with driver itself, but with ACPI+driver. Because this happen time to time, due to hibernate/suspend/reboot or due to electricity lost.
Comment 5 onorua 2013-01-04 11:11:10 UTC
Created attachment 90391 [details]
Dmesg with 3.8-rc2

I've updated to 3.8-rc2, jsut in case something had been fixed meanwhile. As soon as I startup - my ethernet stop to work. 
3.5.5 -> working correctly. I decided that dmesg would be useful to you. If not - just ignore it.
Comment 6 Tushar 2013-01-04 19:03:09 UTC
There is definitely something changed between 3.5.5 and 3.8-rc1 that is causing this error on you system. And I don't think this is a driver bug. Have you tried using .config of your 3.5.5 kernel with 3.8-rc1?
Comment 7 onorua 2013-01-05 08:11:27 UTC
I've recompiled the 3.8-rc2 kernel without any extra options, just copied .config from 3.5.5 to 3.8-rc2, and enabled my webcam (since 3.6.x config had been changed and webcam enabled in 3.5.5 doesn't work). 
So, basically - no changes at all. Reboot -> ethernet doesn't work. 

Is there any way I can identify what exactly the problem? What ever, I'll be glad to help to identify the problem.
Comment 8 Tushar 2013-01-07 21:39:39 UTC
What is the ethtool vesrion? Is it different or same?
Comment 9 Stefan Huber 2013-01-29 18:54:07 UTC
I have the same problem on a X220 and the bug is already discussed at https://lkml.org/lkml/2013/1/5/100 along with an underflow issue of /sys/bus/pci/devices/0000\:00\:19.0/enable

In the meantime, as a "workaround", I enable the ethernet controller with
echo on > /sys/bus/pci/devices/0000\:00\:19.0/power/control
Comment 10 Stefan Huber 2013-01-29 19:09:12 UTC
Ah, and patches appear to be on the way: https://lkml.org/lkml/2013/1/18/147
Comment 11 onorua 2013-01-30 05:56:01 UTC
Thank you Stefan, the workaround does work!
Comment 12 Stefan Huber 2013-01-30 07:25:51 UTC
I have applied the patches proposed by Konstantin Khlebnikov on top of v3.8-rc4 and they solved the problem for me.

Note You need to log in before you can comment on or make changes to this bug.