You are not logged in.
Hi all,
Since the update to Linux 4.10.2 my Ethernet link does not work anymore. I cannot communicate with my router, the link status is UP,LOWER_UP, but the link is 10Mbps only. It looks as if the PHY is not able to get a proper link.
[ 5.216787] e1000e: Intel(R) PRO/1000 Network Driver - 3.2.6-k
[ 5.216788] e1000e: Copyright(c) 1999 - 2015 Intel Corporation.
[ 5.216963] e1000e 0000:00:19.0: Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
[ 5.588335] e1000e 0000:00:19.0 0000:00:19.0 (uninitialized): registered PHC clock
[ 5.686950] e1000e 0000:00:19.0 eth0: (PCI Express:2.5GT/s:Width x1) 28:d2:44:39:a8:e4
[ 5.686955] e1000e 0000:00:19.0 eth0: Intel(R) PRO/1000 Network Connection
[ 5.686988] e1000e 0000:00:19.0 eth0: MAC: 11, PHY: 12, PBA No: 1000FF-0FF
[ 5.688779] e1000e 0000:00:19.0 enp0s25: renamed from eth0
[ 10.130497] e1000e: enp0s25 NIC Link is Up 10 Mbps Full Duplex, Flow Control: None
[ 187.529015] e1000e: enp0s25 NIC Link is Down
[ 190.599893] e1000e 0000:00:19.0 enp0s25: PHY Wakeup cause - Link Status Change
[ 193.703506] e1000e: enp0s25 NIC Link is Up 10 Mbps Full Duplex, Flow Control: None
3: enp0s25: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
link/ether 28:d2:44:39:a8:e4 brd ff:ff:ff:ff:ff:ff
Anybody else seen this?
--
Stefan
Offline
what is the output of ethtool enp0s25 ?
Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.
(A works at time B) && (time C > time B ) ≠ (A works at time C)
Offline
This is more complex than I thought it was: I downgraded the kernel and still observed the issue. I downgraded all the way to some previous linux-lts 4.4 package I still had around, but still could observe the issue (10MBit link, but no real communication)!?
But then I realized I only do reboots... How about a power cycle? And it helped! Back to 4.10.2, and... Network still worked with Gigabit just fine. I often use suspend, so I also tried a suspend/resume cycle and ethernet continued to work just fine.
Today, I see the issue again, and I have a dmesg history back to when it worked:
$ dmesg | grep e1000
[ 5.331915] e1000e: Intel(R) PRO/1000 Network Driver - 3.2.6-k
[ 5.331916] e1000e: Copyright(c) 1999 - 2015 Intel Corporation.
[ 5.333352] e1000e 0000:00:19.0: Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
[ 5.447461] e1000e 0000:00:19.0 0000:00:19.0 (uninitialized): registered PHC clock
[ 5.529917] e1000e 0000:00:19.0 eth0: (PCI Express:2.5GT/s:Width x1) 28:d2:44:39:b7:e4
[ 5.529924] e1000e 0000:00:19.0 eth0: Intel(R) PRO/1000 Network Connection
[ 5.529957] e1000e 0000:00:19.0 eth0: MAC: 11, PHY: 12, PBA No: 1000FF-0FF
[ 5.551742] e1000e 0000:00:19.0 enp0s25: renamed from eth0
[ 9.763802] e1000e: enp0s25 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
[ 8643.979305] e1000e: EEE TX LPI TIMER: 00000011
[ 8644.756208] e1000e 0000:00:19.0: System wakeup enabled by ACPI
[ 8644.924340] e1000e 0000:00:19.0: System wakeup disabled by ACPI
[ 8648.196404] e1000e: enp0s25 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
[ 8648.592917] e1000e: enp0s25 NIC Link is Down
[ 8652.382781] e1000e: enp0s25 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
[10215.316904] e1000e: enp0s25 NIC Link is Down
[12056.659755] e1000e: enp0s25 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
[19235.693891] e1000e: EEE TX LPI TIMER: 00000011
[19236.434903] e1000e 0000:00:19.0: System wakeup enabled by ACPI
[19236.615550] e1000e 0000:00:19.0: System wakeup disabled by ACPI
[19240.406182] e1000e: enp0s25 NIC Link is Down
[19243.649587] e1000e: enp0s25 NIC Link is Up 10 Mbps Full Duplex, Flow Control: None
[45885.757309] e1000e: EEE TX LPI TIMER: 00000011
[45886.527502] e1000e 0000:00:19.0: System wakeup enabled by ACPI
[45886.706931] e1000e 0000:00:19.0: System wakeup disabled by ACPI
[45889.919162] e1000e: enp0s25 NIC Link is Down
[45893.030882] e1000e: enp0s25 NIC Link is Up 10 Mbps Full Duplex, Flow Control: None
Output of ethtool:
$ ethtool enp0s25
Settings for enp0s25:
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: 10Mb/s
Duplex: Full
Port: Twisted Pair
PHYAD: 1
Transceiver: internal
Auto-negotiation: on
MDI-X: off (auto)
Cannot get wake-on-lan settings: Operation not permitted
Current message level: 0x00000007 (7)
drv probe link
Link detected: yes
Offline
Hmm, ethtool output looks fine (apart from the 10 Mb/s) .
check /usr/lib/modules/4.10.4-1-ARCH/build/Documentation/networking/e1000e.txt .
Does enabling/disabling SmartPowerDownEnable make a difference ?
Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.
(A works at time B) && (time C > time B ) ≠ (A works at time C)
Offline
Networking is no longer within my realm of even semi-expertise, but, is this the old Intel Eagle card we used way-back-when (circa 2001)? Isn't it hardware-limited and incapable of modern, higher speeds?
That being queried, I'll bow out and let you younger, network wogs continue on...
Regards
UNIX was not designed to stop you from doing stupid things, because that would also stop you from doing clever things. -- Doug Gwyn
Offline
The SmartPowerDownEnable parameter really seems to make a difference. This is after I had the issue, I enabled the parameter, rebooted and put the system into sleep once:
[ 5.646824] e1000e: Intel(R) PRO/1000 Network Driver - 3.2.6-k
[ 5.646826] e1000e: Copyright(c) 1999 - 2015 Intel Corporation.
[ 5.647007] e1000e 0000:00:19.0: Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
[ 5.647010] e1000e 0000:00:19.0: PHY Smart Power Down Enabled
[ 6.019089] e1000e 0000:00:19.0 0000:00:19.0 (uninitialized): registered PHC clock
[ 6.117435] e1000e 0000:00:19.0 eth0: (PCI Express:2.5GT/s:Width x1) 28:d2:44:39:b7:e4
[ 6.117440] e1000e 0000:00:19.0 eth0: Intel(R) PRO/1000 Network Connection
[ 6.117474] e1000e 0000:00:19.0 eth0: MAC: 11, PHY: 12, PBA No: 1000FF-0FF
[ 6.119357] e1000e 0000:00:19.0 enp0s25: renamed from eth0
[ 10.967845] e1000e: enp0s25 NIC Link is Up 10 Mbps Full Duplex, Flow Control: None
[ 107.278475] e1000e: EEE TX LPI TIMER: 00000011
[ 108.098781] e1000e 0000:00:19.0: System wakeup enabled by ACPI
[ 108.279125] e1000e 0000:00:19.0: System wakeup disabled by ACPI
[ 111.457363] e1000e: enp0s25 NIC Link is Down
[ 114.892992] e1000e: enp0s25 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
After the reboot, it came up with 10Mbps only, but after entering sleep mode once, the PHY properly detected the link again. So it seems that the parameter really "restarts" the PHY.. However, given that it worked before without the parameter, it seems somewhat a workaround to me.
Offline