Ubuntu 20.04 always restarts after successful wake-up from suspend

Question

Update 1: April 30th, 2020

Found the following logs in the syslog (see Gist link):

Apr 30 15:56:34 aurora-r8-linux kernel: [  607.945292] ACPI: Waking up from system sleep state S3
Apr 30 15:56:34 aurora-r8-linux kernel: [  608.507372] snd_hda_intel 0000:01:00.1: azx_get_response timeout, switching to polling mode: last cmd=0x005f2f04
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.084330] usb usb1: root hub lost power or was reset
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.084335] usb usb2: root hub lost power or was reset
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.088035] ACPI BIOS Error (bug): AE_AML_BUFFER_LIMIT, Field [DRQL] at bit offset/length 136/8 exceeds size of target Buffer (128 bits) (20190816/dsopcode-198)
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.088038] No Local Variables are initialized for Method [DSRS]
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.088039] Initialized Arguments for Method [DSRS]:  (2 arguments defined for method invocation)
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.088039]   Arg0:   00000000d64c2208 <Obj>           Buffer(16) 47 01 F8 03 FF 03 00 08
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.088042]   Arg1:   00000000f4b3cffb <Obj>           Integer 0000000000000000
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.088044] ACPI Error: Aborting method \_SB.PCI0.LPCB.SIO1.DSRS due to previous error (AE_AML_BUFFER_LIMIT) (20190816/psparse-529)
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.088047] ACPI Error: Aborting method \_SB.PCI0.LPCB.UAR1._SRS due to previous error (AE_AML_BUFFER_LIMIT) (20190816/psparse-529)
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.088050] serial 00:02: activation failed
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.088052] PM: dpm_run_callback(): pnp_bus_resume+0x0/0xa0 returns -5
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.088053] PM: Device 00:02 failed to resume: error -5
Apr 30 15:56:34 aurora-r8-linux kernel: [  609.096400] sd 0:0:0:0: [sda] Starting disk

Particularly interesting is:

Apr 30 15:56:34 aurora-r8-linux kernel: [  609.088053] PM: Device 00:02 failed to resume: error -5

Could this be related to the issue? How do I find out what device 00:02 is? What is error -5?

---

TL;DR: Ubuntu restarts about 1 minute after waking up from suspend (able to login successfully into desktop). syslog logs attached.

I'm running into this weird issue where my computer is able to suspend, and also able to successfully wake up from suspend. I'm able to login into desktop without problems. But after a minute or so, my computer would just abruptly reboot, taking me through the BIOS boot up screen and to the default boot manager (Windows Boot Manager for my machine, unless I choose GRUB in BIOS boot menu).

After reboot, my computer would function normally until the next time I try to suspend it.

Here's output of last (looks kinda weird, not sure why there's an entry for 12:02 between 15:46 and 16:02:

➜  ~ last                   
okamayan :1           :1               Thu Apr 30 16:02   still logged in
reboot   system boot  5.4.0-28-generic Thu Apr 30 12:02   still running
okamayan :1           :1               Thu Apr 30 15:46 - crash  (-3:44)
reboot   system boot  5.4.0-28-generic Thu Apr 30 11:46   still running

syslog logs from the relevant time period (I believe the reboot happened at around 15:46-16:02): https://gist.github.com/okamayana/117411aa36263e65d4507a2fcecf6990

Another thing is, in a previous occurence, I was able to find the following log statements in syslog:

Apr 29 09:41:52 aurora-r8-linux boltd[2279]: udev: found 0 domain
Apr 29 09:41:52 aurora-r8-linux boltd[2279]: manager: acquired power guard '1'
Apr 29 09:41:52 aurora-r8-linux boltd[2279]: udev: enumerating devices
Apr 29 09:41:52 aurora-r8-linux boltd[2279]: power: guard '1' for 'boltd' deactivated
Apr 29 09:41:52 aurora-r8-linux boltd[2279]: power: shutdown scheduled (T-20.00s)
Apr 29 09:41:52 aurora-r8-linux boltd[2279]: power: state changed: supported/wait
Apr 29 09:41:52 aurora-r8-linux dbus-daemon[830]: [system] Successfully activated service 'org.freedesktop.bolt'
Apr 29 09:41:52 aurora-r8-linux systemd[1]: Started Thunderbolt system service.
Apr 29 09:41:52 aurora-r8-linux boltd[2279]: power: state changed: supported/on
Apr 29 09:41:52 aurora-r8-linux boltd[2279]: power: guard '2' for 'fwupd' active

I've removed and purged bolt, but the issue still happens! Worth noting the boltd logs don't show up anymore, so I doubt it was the cause. I might just reinstall it back. I don't have any Thunderbolt devices (or even port, not sure), though.

If it helps, my system is an Alienware Aurora R8 desktop, with an i5-9600K and an Nvidia RTX 2070 (I'm using Nvidia closed source drivers).

Any help is much appreciated. I cannot find anything that sticks out in syslog. Should I look at dmesg or any other logs?

I can live without suspend, but I'd like to see if I can get it working. It's especially annoying because unexpected shutdowns like this always disables BIOS overclock settings.

Thanks again in advance.

Answer 1

It helps to know that you are using nvidia. I've had a lot of problems with nvidia, suspend and hibernation. You may get closer to you problem if you circumvent any scripts from your system by opening a root terminal:

#Suspend-to-RAM
echo -n "deep" > /sys/power/mem_sleep 
echo -n "mem" > /sys/power/state

This is the main code the kernel will do on suspend. Wake your device up again. If it does't restart, something in your/systemd's configuration is not correct. You also may want to check out the following description about how to debug it: kernel documentation

Answer 2

Run lsusb to get an idea of what the device is.

Also run: dmidecode | grep "0xa0" and let us know what the output is.