Bug 218936 - e1000e won't enter suspend without a network connection
Summary: e1000e won't enter suspend without a network connection
Status: NEW
Alias: None
Product: Drivers
Classification: Unclassified
Component: Network (show other bugs)
Hardware: Intel Linux
: P3 high
Assignee: drivers_network@kernel-bugs.osdl.org
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2024-06-05 05:07 UTC by Dieter Mummenschanz
Modified: 2025-02-14 09:12 UTC (History)
10 users (show)

See Also:
Kernel Version: 6.10-rc2
Subsystem:
Regression: Yes
Bisected commit-id: bfd546a552e140b0a4c8a21527c39d6d21addb28


Attachments
full dmesg (101.51 KB, text/plain)
2024-06-05 05:07 UTC, Dieter Mummenschanz
Details

Description Dieter Mummenschanz 2024-06-05 05:07:04 UTC
Created attachment 306412 [details]
full dmesg

Hallo,

starting with 6.10-rc1 I'm unable to suspend my laptop when the LAN cable is not connected. dmesg shows:

[42394.581194] e1000e 0000:00:1f.6: PM: pci_pm_suspend(): e1000e_pm_suspend [e1000e] returns -2
[42394.581205] e1000e 0000:00:1f.6: PM: dpm_run_callback(): pci_pm_suspend returns -2
[42394.581209] e1000e 0000:00:1f.6: PM: failed to suspend async: error -2
[42394.704458] PM: Some devices failed to suspend, or early wake event detected
[42401.167550] e1000e 0000:00:1f.6 eth0: Hardware Error

This doesn't happen when the lan cable is plugged in.

Full dmesg is attached with several scsessful suspend/resumes when connected and the last failed suspend when disconnected.

Can anyone confirm this?
Comment 1 Dieter Mummenschanz 2024-06-05 19:50:59 UTC
lspci -nn -s 00:1f.6
00:1f.6 Ethernet controller [0200]: Intel Corporation Ethernet Connection (7) I219-V [8086:15bc] (rev 10)
Comment 2 Todd Brandt 2024-06-05 22:12:03 UTC
I believe I'm seeing something similar that is the result of a commit added to 6.10.0-rc2. I've bisected it here. Perhaps it's the issue for you as well? https://bugzilla.kernel.org/show_bug.cgi?id=218940
Comment 3 Todd Brandt 2024-06-05 22:23:36 UTC
oh, I didn't notice you have a bisect commit as well, ok clearly not the same issue. Never mind.
Comment 4 Dieter Mummenschanz 2024-06-06 05:35:12 UTC
No you were right. They are related as a matter of fact. I've added the wrong commit id here.
Comment 5 Randy Barlow 2024-08-19 23:50:39 UTC
I see this too with this NIC:

❯ lspci | grep Ethernet
00:1f.6 Ethernet controller: Intel Corporation Device 550b (rev 20)

A workaround is to rmmod e1000e (obviously only when you aren't using the NIC). I find that it suspends with the module unloaded.
Comment 6 Yury Vidineev 2024-11-29 07:15:50 UTC
Hi. Same issue here on Dell Precision 3590

❯ lspci | grep Ethernet
00:1f.6 Ethernet controller: Intel Corporation Device 550a (rev 20)

Same workaround, to remove e1000e. Sometimes disabling WoL helps as well (sudo ethtool -s enp0s31f6  wol d) but this solution is less reliable to me.
Comment 7 Randy Barlow 2024-11-29 13:57:46 UTC
I've noticed that recent kernels I've been running from Gentoo do not have this issue. I am unsure if it's because Gentoo has applied a patch, or if perhaps it got fixed in upstream. I am currently running 6.11.8.
Comment 10 Yury Vidineev 2025-02-14 09:12:36 UTC
I still have this issue on 6.11 (haven't tested anything newer yet)

$ lspci | grep Ethernet
00:1f.6 Ethernet controller: Intel Corporation Device 550a (rev 20)

$ uname -a
Linux adept-framework 6.11.0-1011-oem #11-Ubuntu SMP PREEMPT_DYNAMIC Thu Dec 19 11:35:20 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux

$ journalctl -k --since "30 minutes ago" | egrep 'e1000e|PM'


Feb 14 01:02:52 adept-framework kernel: PM: suspend entry (s2idle)
Feb 14 01:02:54 adept-framework kernel: e1000e: EEE TX LPI TIMER: 00000011
Feb 14 01:02:54 adept-framework kernel: e1000e 0000:00:1f.6: PM: pci_pm_suspend(): e1000e_pm_suspend [e1000e] returns -2
Feb 14 01:02:54 adept-framework kernel: e1000e 0000:00:1f.6: PM: dpm_run_callback(): pci_pm_suspend returns -2
Feb 14 01:02:54 adept-framework kernel: e1000e 0000:00:1f.6: PM: failed to suspend async: error -2
Feb 14 01:02:54 adept-framework kernel: workqueue: e1000_watchdog_task [e1000e] hogged CPU for >10000us 4 times, consider switching to WQ_UNBOUND
Feb 14 01:02:54 adept-framework kernel: PM: Some devices failed to suspend, or early wake event detected
Feb 14 01:02:54 adept-framework kernel: spd5118 2-0050: PM: dpm_run_callback(): spd5118_resume [spd5118] returns -6
Feb 14 01:02:54 adept-framework kernel: spd5118 2-0050: PM: failed to resume async: error -6
Feb 14 01:02:54 adept-framework kernel: PM: suspend exit
Feb 14 01:02:57 adept-framework kernel: e1000e 0000:00:1f.6 enp0s31f6: Hardware Error
Feb 14 01:02:58 adept-framework kernel: e1000e 0000:00:1f.6 enp0s31f6: NIC Link is Down
Feb 14 01:03:00 adept-framework kernel: e1000e 0000:00:1f.6 enp0s31f6: Hardware Error

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