On Thu, Mar 14, 2013 at 1:45 PM, Allan, Bruce W <bruce.w.al...@intel.com> wrote:
>> -----Original Message-----
>> From: netdev-ow...@vger.kernel.org [mailto:netdev-
>> ow...@vger.kernel.org] On Behalf Of Andy Lutomirski
>> Sent: Thursday, March 14, 2013 9:54 AM
>> To: Network Development
>> Cc: e1000-devel@lists.sourceforge.net
>> Subject: em1 is visible to iproute2 but not ethtool or tcpdump
>>
>> My e1000e device is em1.  It has this rather strange behavior on
>> Fedora's 3.8.1 (although 3.7.something did the same thing, I believe):
>>
>> $ ip link show dev em1
>> 6: em1: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc
>> pfifo_fast
>> state DOWN qlen 1000
>>     link/ether f0:de:f1:59:f6:c9 brd ff:ff:ff:ff:ff:ff
>> $ strace -e socket,ioctl ethtool em1
>> socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3
>> Settings for em1:
>> ioctl(3, SIOCETHTOOL, 0x7fffa6c554d0)   = -1 ENODEV (No such device)
>> Cannot get device settings: No such device
>> ioctl(3, SIOCETHTOOL, 0x7fffa6c554d0)   = -1 ENODEV (No such device)
>> Cannot get wake-on-lan settings: No such device
>> ioctl(3, SIOCETHTOOL, 0x7fffa6c554d0)   = -1 ENODEV (No such device)
>> Cannot get message level: No such device
>> ioctl(3, SIOCETHTOOL, 0x7fffa6c554d0)   = -1 ENODEV (No such device)
>> Cannot get link status: No such device
>> No data available
>> +++ exited with 75 +++
>> $ sudo tcpdump -i em1
>> tcpdump: em1: SIOETHTOOL(ETHTOOL_GTSO) ioctl failed: No such device
>>
>> A sufficient number of rmmod/modprobe cycles will make it work.
>>
>> This is, perhaps, helpful:
>>
>> $ sudo ip link set dev em1 name foobar
>> RTNETLINK answers: Device or resource busy
>> $ sudo ip link set dev em1 down
>> $ sudo ip link set dev em1 name foobar
>> $ sudo ethtool foobar
>> Settings for foobar:
>>       Supported ports: [ TP ]
>>       Supported link modes:   10baseT/Half 10baseT/Full
>>                               100baseT/Half 100baseT/Full
>>                               1000baseT/Full
>>       Supported pause frame use: No
>>       Supports auto-negotiation: Yes
>>       Advertised link modes:  10baseT/Half 10baseT/Full
>>                               100baseT/Half 100baseT/Full
>>                               1000baseT/Full
>>       Advertised pause frame use: No
>>       Advertised auto-negotiation: Yes
>>       Speed: Unknown!
>>       Duplex: Unknown! (255)
>>       Port: Twisted Pair
>>       PHYAD: 1
>>       Transceiver: internal
>>       Auto-negotiation: on
>>       MDI-X: Unknown
>>       Supports Wake-on: pumbg
>>       Wake-on: g
>>       Current message level: 0x00000007 (7)
>>                              drv probe link
>>       Link detected: no
>> $ sudo ip link set dev foobar name em1
>> $ sudo ethtool em1
>> Settings for em1:
>>       Supported ports: [ TP ]
>>       Supported link modes:   10baseT/Half 10baseT/Full
>>                               100baseT/Half 100baseT/Full
>>                               1000baseT/Full
>>       Supported pause frame use: No
>>       Supports auto-negotiation: Yes
>>       Advertised link modes:  10baseT/Half 10baseT/Full
>>                               100baseT/Half 100baseT/Full
>>                               1000baseT/Full
>>       Advertised pause frame use: No
>>       Advertised auto-negotiation: Yes
>>       Speed: Unknown!
>>       Duplex: Unknown! (255)
>>       Port: Twisted Pair
>>       PHYAD: 1
>>       Transceiver: internal
>>       Auto-negotiation: on
>>       MDI-X: Unknown
>>       Supports Wake-on: pumbg
>>       Wake-on: g
>>       Current message level: 0x00000007 (7)
>>                              drv probe link
>>       Link detected: no
>> $
>>
>> This bug (?) isn't a show-stopper, but it's rather annoying.  I
>> suspect a core networking bug, not an e1000e bug, but I cc'd the e1000
>> people just in case.
>
> This sounds related to some runtime power management (RPM) issues that have
> already been fixed.  First check whether RPM is enabled for your device - the 
> file
> /sys/devices/pciNNNN:NN/NNNN:NN:NN.N/power/control (where NNNN... is
> the PCI domain, bus, slot and function of your device) will contain "auto" 
> when it
> is enabled.  To disable RPM for that device, change the value in the file to 
> "on".

Seems to work.  Oddly, I'm pretty sure I've seen this bug even when
something's plugged in.

Is the fix heading to stable?

--Andy

------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_d2d_mar
_______________________________________________
E1000-devel mailing list
E1000-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel&#174; Ethernet, visit 
http://communities.intel.com/community/wired

Reply via email to