On 26/12/2021 13.28, Axel Braun wrote:
Hello Carlos,
Am Sonntag, 26. Dezember 2021, 13:20:49 CET schrieb Carlos E. R.:
Any idea how to investigate this?
Dez 26 00:16:40 X1E kernel: CPU11 is up Dez 26 00:16:40 X1E kernel: ACPI: PM: Waking up from system sleep state S3 Dez 26 00:16:40 X1E kernel: acpi device:00: Failed to change power state to D0
This is the end of the wake up process.
Correct....
....
The suspend actually happened here:
Dez 25 18:46:10 X1E kscreenlocker_greet[15764]: ...... Dez 26 00:16:40 X1E kernel: bbswitc....
Notice the time jump. Some entries are actually written to the log when the machine wakes up, but belong to the suspend phase.
I'm pretty sure I closed the lid at 18:46 (suspend), and reopened at 00:16 (wakeup)
Yep.
Your problem, on wakeup, is logged here:
Dez 25 18:46:10 X1E kernel: acpi device:00: Failed to change power state to D0 Dez 25 18:46:10 X1E kscreenlocker_greet[15764]: file:///usr/share/plasma/look-and-feel/org.openSUSE.desktop/contents/compon ents/UserList.qml:47:9: Unable to assign [undefine>
That is the suspend.....
"Change power state to D0" is waking up.
Dez 26 00:16:40 X1E kernel: bbswitch: enabling discrete graphics Dez 26 00:16:40 X1E kernel: video LNXVIDEO:00: Cannot transition to power state D0 for parent in (unknown) Dez 26 00:16:40 X1E kernel: pci 0000:01:00.0: can't change power state from D3cold to D0 (config space inaccessible) Dez 26 00:16:40 X1E kernel: video LNXVIDEO:00: Cannot transition to power state D0 for parent in (unknown) Dez 26 00:16:40 X1E kernel: pci 0000:01:00.0: can't change power state from D3cold to D0 (config space inaccessible) Dez 26 00:16:40 X1E kernel: Freezing user space processes ... (elapsed 0.002 seconds) done.
The suspend problem itself is not logged, because it mentions a problem going to D0, not to D3.
...which is the odd thing, as suspend was at 18:46
Yes, but there is a moment during the suspend process when the hard disk is "stopped" and nothing more can be written to it, and after that is when the kernel does the actual suspend work. Any log entries done in that interval are stored in the memory of the syslog process, or the kernel message ring, I don't know for sure, and written when the hard disk is waken up again, so the timestamps for those events are wrong, they write the wake up time.
I have now disabled tlp, and keep monitoring...
-- Cheers / Saludos, Carlos E. R. (from 15.2 x86_64 at Telcontar)