[Solved]Strange problem with Wake On Lan

barney · 2023-04-08 16:33:05

I have desktop computer which I want to wake up with wol command. I have followed instructions from the Arch Wiki page https://wiki.archlinux.org/title/Wake-on-LAN.
I have used Network Manager to make the changes with ethtool persistent:

nmcli c modify "wired1" 802-3-ethernet.wake-on-lan magic

Rebooted the system and wake on lan is working when I have executed

sudo systemctl poweroff

but it is not working when I am doing suspend

 sudo systemctl suspend

As suggested here https://wiki.archlinux.org/title/Wake-o … leshooting I have executed

nmcli c modify "wired1" 802-3-ethernet.auto-negotiate yes

Didn't do the job.
What is weird is that with shutdown it is working, but with suspend it is not. I suppose that suspend scripts are powering off the network card. I don't know where to start to solve my problem.
Network card is

07:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 06)

I have read the troubleshooting for this card but I don't think that that is the case, because WOL is working with shutdown.

Last edited by barney (2023-04-12 18:07:47)

barney · 2023-04-09 00:33:09

Update. I have tried to blacklist the r8169 module, load the r8168 module, set s5wol=1 module parameter but didn't change anything. I have returned to r8169 module but problem is sill here.

barney · 2023-04-09 03:56:57

I have inspected the journal boot messages and I have found one error message:

kernel: r8168 0000:07:00.0: Unable to load firmware rtl_nic/rtl8168e-4.fw

I have searched Internet for rtl8168e-4.fw file, but I cannot find it.

BluishHumility · 2023-04-09 23:42:58

What suspend state is in use?

Are you using TLP?

barney · 2023-04-10 00:12:56

BluishHumility wrote:

What suspend state is in use?

The output of

 sudo cat /sys/power/mem_sleep

is:

s2idle [deep]

BluishHumility wrote:

Are you using TLP?

No. I don't think that that is the problem. The problem is that r8168 module cannot load the firmware file, because there is no such a file on my system.

BluishHumility · 2023-04-10 01:49:23

barney wrote:

The problem is that r8168 module cannot load the firmware file, because there is no such a file on my system.

I'm not so sure; since WoL is working correctly after a shutdown, that journal error you have found may be a red herring.

I would test another kernel or a few other kernels to see if you get a different result. Try the LTS kernel and the mainline kernel.

barney wrote:

s2idle [deep]

Check if switching to s2idle allows WoL to succeed. https://wiki.archlinux.org/title/Power_ … end_method

seth · 2023-04-10 06:23:47

The problem is that r8168 module cannot load the firmware file, because there is no such a file on my system.

Those "errors" are common, the kernel tries to load the know FWs until one succeeds.

Next to the suggestions in #6, please post your complete system journal for a boot covering a failed WOL:

sudo journalctl -b | curl -F 'file=@-' 0x0.st

for the current one.

Also, ceterum censeo: is there a parallel windows installation?

barney · 2023-04-10 20:09:00

Complete journal file:
http://0x0.st/HXgN.txt

Also, ceterum censeo: is there a parallel windows installation?

Yes, there is. But, fast start and hibernate are disabled.

seth · 2023-04-10 20:18:14

The journal doesn't cover a S3 cycle?

fast start and hibernate are disabled

Known or believed?
Since windows keeps re-enabling this w/ updates, make sure to check.

апр 10 21:56:15 living org_kde_powerdevil[480]: org.kde.powerdevil: org.kde.powerdevil.chargethresholdhelper.getthreshold failed "Charge thresholds are not supported by the kernel for this hardware"
апр 10 21:56:15 living org_kde_powerdevil[480]: org.kde.powerdevil: org.kde.powerdevil.backlighthelper.brightness failed

Try the behavior from the multi-user.target (2nd link below)

barney · 2023-04-10 20:23:17

I have forgot to paste the output of sudo ethtool -i enp7s0:

driver: r8168
version: 8.051.02-NAPI
firmware-version: 
expansion-rom-version: 
bus-info: 0000:07:00.0
supports-statistics: yes
supports-test: no
supports-eeprom-access: no
supports-register-dump: yes
supports-priv-flags: no

which shows that no FW is loaded.

barney · 2023-04-10 21:00:49

The journal doesn't cover a S3 cycle?

What should I do to paste the journal which covers S3 cycle?

Try the behavior from the multi-user.target (2nd link below)

I have tried. It is the same behavior.
I will check about windows.

seth · 2023-04-10 21:13:59

If the driver only wants to load the -4 version of the FW, you could symlink the -3 version there.
Suspend to RAM, try to WOL, resume otherwise and then post the journal.

barney · 2023-04-10 21:58:53

I have made symlink in rtl_nic directory. Now sudo ethtool -i enp7s0 gives:

driver: r8168
version: 8.051.02-NAPI
firmware-version: rtl8168e-3_0.0.4 03/27/12
expansion-rom-version: 
bus-info: 0000:07:00.0
supports-statistics: yes
supports-test: no
supports-eeprom-access: no
supports-register-dump: yes
supports-priv-flags: no

I have suspended the computer, tried WOL withaut success. Here is the jornal boot file:
http://0x0.st/HXg4.txt
I have checked windows, the fast start and hibernate are disabled. I rarely use windows, only when I get hands on some complicated word file that libre office cannot open.

seth · 2023-04-11 06:19:59

Try to add

pcie_aspm=off

to the kernel parameters.
Does a suspended windows respond to WOL?
Despite being a 8168 chip, have you actually tried the r8169 behavior? (you have not "returned" to it and will have to comment the blacklist in /usr/lib/modprobe.d/r8168.conf)

barney · 2023-04-11 08:38:09

seth wrote:

Try to add
pcie_aspm=off
to the kernel parameters.

Did't make any difference.

Does a suspended windows respond to WOL?

Yes. In windows everything is working fine, with shutdown and suspend (sleep).

Despite being a 8168 chip, have you actually tried the r8169 behavior? (you have not "returned" to it and will have to comment the blacklist in /usr/lib/modprobe.d/r8168.conf)

Yes, I have tried. Because you mention it, I have tried second time, without success.

I think that I have a buggy chip that requires 'special' firmware. Is it possible to take the firmware file from windows?

seth · 2023-04-11 12:39:55

I think that I have a buggy chip that requires 'special' firmware. Is it possible to take the firmware file from windows?

Probably not, but you could extract it from the windows driver. Easy, if it comes as separate file, hard, if it's compiled into some dll or exe.

Sanity check:

cat /proc/acpi/wakeup

barney · 2023-04-11 16:12:12

Sanity check:
cat /proc/acpi/wakeup

sudo cat /proc/acpi/wakeup
Device  S-state   Status   Sysfs node
PCI0      S5    *disabled  no-bus:pci0000:00
PEX0      S5    *disabled  pci:0000:00:1c.0
PEX1      S5    *disabled  pci:0000:00:1c.1
PEX2      S5    *disabled
PEX3      S5    *disabled  pci:0000:00:1c.3
PEX4      S5    *disabled
PEX5      S5    *disabled  pci:0000:00:1c.5
PEX6      S5    *disabled
PEX7      S5    *disabled
HUB0      S5    *disabled  pci:0000:00:1e.0
UAR1      S3    *disabled
USBE      S3    *enabled   pci:0000:00:1d.0
USE2      S3    *enabled   pci:0000:00:1a.0
AZAL      S5    *disabled  pci:0000:00:1b.0

seth · 2023-04-11 18:55:39

https://bbs.archlinux.org/viewtopic.php?id=283917 ?

barney · 2023-04-11 23:08:08

seth wrote:

https://bbs.archlinux.org/viewtopic.php?id=283917 ?

That was the problem. I have installed kernel mention there and everything is working fine.
I have stupid question (understand that I am new to Arch Linux and I don't know many things). Now when I install the r8168 package, the module file still goes in modules directory for newer kernel, despite the fact that the running kernel is older. How to resolve this?
When can we expect this bug to be removed from official linux package?
If it is not the secret how did you know that this bug was the problem. The output from cat /proc/acpi/wakeup didn't changed.

Last edited by barney (2023-04-11 23:14:44)

seth · 2023-04-12 06:02:59

https://aur.archlinux.org/packages/r8168-dkms

The pci device simply doesn't show up in the wakeup devices, so that was ruled out as pssible case and the only thing left from all of this (cause I didn't buy the FW thing for a second) was the global regression in 6.2

Please always remember to mark resolved threads by editing your initial posts subject - so others will know that there's no task left, but maybe a solution to find.
Thanks.

Last edited by seth (2023-04-12 06:03:18)

Arch Linux

#1 2023-04-08 16:33:05

[Solved]Strange problem with Wake On Lan

#2 2023-04-09 00:33:09

Re: [Solved]Strange problem with Wake On Lan

#3 2023-04-09 03:56:57

Re: [Solved]Strange problem with Wake On Lan

#4 2023-04-09 23:42:58

Re: [Solved]Strange problem with Wake On Lan

#5 2023-04-10 00:12:56

Re: [Solved]Strange problem with Wake On Lan

#6 2023-04-10 01:49:23

Re: [Solved]Strange problem with Wake On Lan

#7 2023-04-10 06:23:47

Re: [Solved]Strange problem with Wake On Lan

#8 2023-04-10 20:09:00

Re: [Solved]Strange problem with Wake On Lan

#9 2023-04-10 20:18:14

Re: [Solved]Strange problem with Wake On Lan

#10 2023-04-10 20:23:17

Re: [Solved]Strange problem with Wake On Lan

#11 2023-04-10 21:00:49

Re: [Solved]Strange problem with Wake On Lan

#12 2023-04-10 21:13:59

Re: [Solved]Strange problem with Wake On Lan

#13 2023-04-10 21:58:53

Re: [Solved]Strange problem with Wake On Lan

#14 2023-04-11 06:19:59

Re: [Solved]Strange problem with Wake On Lan

#15 2023-04-11 08:38:09

Re: [Solved]Strange problem with Wake On Lan

#16 2023-04-11 12:39:55

Re: [Solved]Strange problem with Wake On Lan

#17 2023-04-11 16:12:12

Re: [Solved]Strange problem with Wake On Lan

#18 2023-04-11 18:55:39

Re: [Solved]Strange problem with Wake On Lan

#19 2023-04-11 23:08:08

Re: [Solved]Strange problem with Wake On Lan

#20 2023-04-12 06:02:59

Re: [Solved]Strange problem with Wake On Lan

Board footer