Changing Wake-from-Suspend Sensitivity

BelowZero · 2018-05-15 00:33:19

I now have my system suspending well, but I cannot figure out how to adjust the wake-from-suspend sensitivity. I come back to my computer and find that it has woken up, and then suspended automatically several times during the day. Is there something I can adjust to make it less sensitive?

Last edited by BelowZero (2018-06-25 22:39:51)

Trilby · 2018-05-15 00:37:39

Sensitive to *what*. There is not some general sensitivity threshold live a volume level, the system just wakes due to specific events. Figure out what is waking it. Do you have wake-on-lan or similar network triggers that are waking it? Do you have cat-on-keyboard or similar fur-bearing triggers that are waking it?

BelowZero · 2018-05-15 00:54:43

No cat. I assumed it was random air movements or a ghost.

Is there a log someplace that I can check to see what caused it to wake? If so, I'll monitor that.

I'm also not sure how to tell if I have wake-on-lan enabled. It does not appear in my settings, and I'm still new to Arch.

seth · 2018-05-15 05:25:00

[cat]
cat /proc/acpi/wakeup
lsusb
[/cat]
It's probably soem USB device, so try disabing USB wakeup (initially maybe all, powerbutton and/or lid should still work for wakeup)

BelowZero · 2018-05-15 21:53:09

Everything is disabled except these items. I'm not sure what they are:

XHC	  S0	*enabled   pci:0000:00:14.0
LID0	  S3	*enabled   platform:PNP0C0D:00
PBTN	  S3	*enabled   platform:PNP0C0C:00

progandy · 2018-05-15 21:58:40

LID0 is a common name for the lid switch of a laptop, PBTN stands most likely for power button.

Edit: XHC should be the USB3 controller. This is the one you can try to disable.

echo XHC > /proc/acpi/wakeup

Last edited by progandy (2018-05-15 22:02:06)

seth · 2018-05-15 22:10:25

It's only listed with S0, though.
My money would then be on the lid switch (auto-suspend sounds like the lid isn't closed and it might generate spurious events)

BelowZero · 2018-05-15 22:46:42

My lid is closed the entire day. Today is woke up a three random times, and I wasn't even home.

Disabling the USB3 controller sounds like it will have other consequences...

loqs · 2018-05-15 23:50:44

BelowZero wrote:

Disabling the USB3 controller sounds like it will have other consequences...

It would disable it from waking the system from S0 it should not have any other impact on use of the controller. The setting would not persist beyond reboot.

Trilby · 2018-05-16 01:36:45

To build on that, you could disable *all* of them (perhaps one by one) to test. Worst case you cannot wake from suspend at all, so you'd have to power off and restart (save any work before trying this, obviously). But if you disable all the triggers and it still wakes up, then we know to look elsewhere.

seth · 2018-05-16 05:18:57

Oh, and one more thing about the weird wakeups: do you also have a windows installation?

BelowZero · 2018-05-16 21:28:54

seth wrote:

Oh, and one more thing about the weird wakeups: do you also have a windows installation?

Yes, I set this machine up to dual boot with Windows 10, although the Windows partition is currently not booting correctly. I plan to troubleshoot that when I have some additional time. Why did you ask?

loqs · 2018-05-17 00:00:34

Windows fast boot shutdown leaves devices in an undefined state for a boot for another OS. Assuming you have shutdown and started the system multiple times since the last use of Windows I doubt that is connected.

seth · 2018-05-17 05:45:16

The problem is that if windows was never properly shut down, it might still have ACPI timers active to wake the system for maintainance tasks (downloading updates, disk defragmentation, stuff...)

Check the logs about how random those wake ups really are (amount and time) - if it always wakes up about the same times of the day, that's probably not some random trigger ...

BelowZero · 2018-06-05 21:07:10

My Windows installation is now running in VirtualBox, which is not running, so I'm not sure it could be waking my machine.

I tried this command:

echo XHC > /proc/acpi/wakeup

It has reduced the waking to once per day, but now I'm having trouble waking the machine by using my mouse (which is plugged in via USB, and I expect, one of the intended consequences). Any other ideas?

seth · 2018-06-06 05:46:56

If disabling USB wakeups has significant impact on the self-wakeup, the natural assumption is that one of the USB devices causes spurious input.
I will say that mice are incredibly prone to this (the sensor cannot decide between two positions, this could even depend on the ambient light) … so try suspending w/ XHC enabled, but your mouse detached.
Also check what other devices are attached via USB (lsusb) - the next natural contenders would be the keyboard or the touchpad.
I'm afraid unbinding the culprit (if you can isolate it) on the kernel layer won't do, since the wake is triggered by the BIOS.

That said: one random creepy self-wakeup per day is still one incident too creepy. DId you also try to disable ONE of the other two (eg. the lid)?

BelowZero · 2018-06-16 16:45:55

I've isolated the problem to my mouse, which is a Logitech mouse connected via a USB dongle. I'm going to pair it with Bluetooth instead and see if that doesn't resolve the problem. For now, I'll mark this as solved.

BelowZero · 2018-06-25 22:45:40

Okay, I take back what I said. It's not only my mouse. Switching my mouse to Bluetooth seemed to solve most of the issues, but I'm still experiencing random wake-ups when nothing is connected except an external monitor and the laptop lid is left open (but the machine is suspended). What else could it be?

If it helps, it's a Dell XPS 9370.

Last edited by BelowZero (2018-06-25 22:46:15)

seth · 2018-06-26 06:23:15

We actually arrived at this in post #15 (even ruling out the entire usb stack as remaining offender)
Unless the wakeup is periodic (in which case it's safe to say something™ setup an ACPI timer in the BIOS/UEFI) i'd bet on the lid switch.

BelowZero · 2018-06-26 13:07:25

The wakeup does not appear periodic, and how would the lid switch cause it while the lid is open?

The Wiki entry for the 9370 indicates that the machine may not be entering deep sleep properly. See https://wiki.archlinux.org/index.php/Dell_XPS_13_(9370). Now I just need to figure out how to edit the kernel parameters.

Last edited by BelowZero (2018-06-26 13:09:33)

seth · 2018-06-26 15:18:04

https://wiki.archlinux.org/index.php/Kernel_parameters - but that's not "random wakeup"…
The idea about the lid would be some false positive trigger (it's often a magnetic switch) - since you don't close it anyway, just disable it for wakeup.

BelowZero · 2018-06-26 23:28:36

Well, I've disabled absolutely everything, and it still wakes from suspend by itself.

DasRohr · 2018-06-27 00:10:42

did you check the BIOS? Some have settings there to enable/disable devices to wake the system from certain states. Like PCIe devices, Network, PS/2, USB, etc.
Some also have the options to modify the S_ state the system goes to when sleeping.

Had something similar with my Desktop.
I allowed the BIOS to boot the system when receiving magic packages on the network-Port. Mainly of course to use WOL, but I also enabled a setting on my FritzBox, that if a certain Port is connected to from the Internet (had a port forwarding enabled) it fires up my PC.
For some reason my PC got booted automatically without me initialising a connection on that certain Port from the outside. Never found out why though ... just disabled the option on my FB.

BelowZero · 2018-06-27 16:37:55

I thought the BIOS idea was a good one, but the only "wake" feature in the BIOS is related to using USB devices through a Dell dock. I disabled it anyway to see if it would help, and it did not. My device woke up again for no reason this morning.

seth · 2018-06-27 17:09:24

Did you try to enforce the deep sleep?

Let's check on the alarms…

cat /proc/driver/rtc

Arch Linux

#1 2018-05-15 00:33:19

Changing Wake-from-Suspend Sensitivity

#2 2018-05-15 00:37:39

Re: Changing Wake-from-Suspend Sensitivity

#3 2018-05-15 00:54:43

Re: Changing Wake-from-Suspend Sensitivity

#4 2018-05-15 05:25:00

Re: Changing Wake-from-Suspend Sensitivity

#5 2018-05-15 21:53:09

Re: Changing Wake-from-Suspend Sensitivity

#6 2018-05-15 21:58:40

Re: Changing Wake-from-Suspend Sensitivity

#7 2018-05-15 22:10:25

Re: Changing Wake-from-Suspend Sensitivity

#8 2018-05-15 22:46:42

Re: Changing Wake-from-Suspend Sensitivity

#9 2018-05-15 23:50:44

Re: Changing Wake-from-Suspend Sensitivity

#10 2018-05-16 01:36:45

Re: Changing Wake-from-Suspend Sensitivity

#11 2018-05-16 05:18:57

Re: Changing Wake-from-Suspend Sensitivity

#12 2018-05-16 21:28:54

Re: Changing Wake-from-Suspend Sensitivity

#13 2018-05-17 00:00:34

Re: Changing Wake-from-Suspend Sensitivity

#14 2018-05-17 05:45:16

Re: Changing Wake-from-Suspend Sensitivity

#15 2018-06-05 21:07:10

Re: Changing Wake-from-Suspend Sensitivity

#16 2018-06-06 05:46:56

Re: Changing Wake-from-Suspend Sensitivity

#17 2018-06-16 16:45:55

Re: Changing Wake-from-Suspend Sensitivity

#18 2018-06-25 22:45:40

Re: Changing Wake-from-Suspend Sensitivity

#19 2018-06-26 06:23:15

Re: Changing Wake-from-Suspend Sensitivity

#20 2018-06-26 13:07:25

Re: Changing Wake-from-Suspend Sensitivity

#21 2018-06-26 15:18:04

Re: Changing Wake-from-Suspend Sensitivity

#22 2018-06-26 23:28:36

Re: Changing Wake-from-Suspend Sensitivity

#23 2018-06-27 00:10:42

Re: Changing Wake-from-Suspend Sensitivity

#24 2018-06-27 16:37:55

Re: Changing Wake-from-Suspend Sensitivity

#25 2018-06-27 17:09:24

Re: Changing Wake-from-Suspend Sensitivity

Board footer