Created attachment 277047 [details] dmesg including the ACPI errors, among other errors I get this on each boot in dmesg on an Atom x7-Z8750 SoC: [ 33.601564] ACPI Error: AE_BAD_PARAMETER, Returned by Handler for [UserDefinedRegion] (20180531/evregion-266) [ 33.601573] ACPI Error: Method parse/execution failed \_SB.STR1.PAT0, AE_BAD_PARAMETER (20180531/psparse-516) [ 33.601717] ACPI Error: AE_BAD_PARAMETER, Returned by Handler for [UserDefinedRegion] (20180531/evregion-266) [ 33.601723] ACPI Error: Method parse/execution failed \_SB.STR1.PAT1, AE_BAD_PARAMETER (20180531/psparse-516) I'll add a dmesg from the current boot. I am not sure if the ACPI errors have any consequence, but since I have issues with certain hardware coming back into function after a resume, it is a guess that they might. There is a failure to initialize the USB C charging in this boot just before the ACPI errors, but those have been present also without that before. Lots of issues still with this Cherry Trail system. I hope these ACPI errors give an easy start to fix them.
Rebooting this particular instance hangs, just after emitting some further ACPI errors: Jun 29 11:47:36 kernel: [drm:intel_pipe_update_end [i915]] *ERROR* Atomic update failure on pipe B (start=140096 end=140097) time 423 us, min 1908, max 1919, scanline start 1874, end 1924 Jun 29 12:58:31 kernel: xfce4-power-man[2816]: segfault at 0 ip 0000000000000000 sp 00007ffe67739a68 error 14 in xfce4-power-manager[55849ba98000+34000] Jun 29 12:58:31 kernel: Code: Bad RIP value. Jun 29 12:58:33 kernel: ACPI Error: AE_BAD_PARAMETER, Returned by Handler for [UserDefinedRegion] (20180531/evregion-266) Jun 29 12:58:33 kernel: ACPI Error: Method parse/execution failed \_SB.STR2.PAT0, AE_BAD_PARAMETER (20180531/psparse-516) Jun 29 12:58:33 kernel: ACPI Error: AE_BAD_PARAMETER, Returned by Handler for [UserDefinedRegion] (20180531/evregion-266) Jun 29 12:58:33 kernel: ACPI Error: Method parse/execution failed \_SB.STR2.PAT1, AE_BAD_PARAMETER (20180531/psparse-516) Jun 29 12:58:33 kernel: ACPI Error: AE_BAD_PARAMETER, Returned by Handler for [UserDefinedRegion] (20180531/evregion-266) Jun 29 12:58:33 kernel: ACPI Error: Method parse/execution failed \_SB.STR0.PAT0, AE_BAD_PARAMETER (20180531/psparse-516) Jun 29 12:58:33 kernel: ACPI Error: AE_BAD_PARAMETER, Returned by Handler for [UserDefinedRegion] (20180531/evregion-266) Jun 29 12:58:33 kernel: ACPI Error: Method parse/execution failed \_SB.STR0.PAT1, AE_BAD_PARAMETER (20180531/psparse-516) Jun 29 12:58:35 kernel: systemd-shutdow: 36 output lines suppressed due to ratelimiting Jun 29 12:58:35 systemd-shutdown[1]: Syncing filesystems and block devices. Jun 29 12:58:35 systemd-shutdown[1]: Sending SIGTERM to remaining processes... Not sure what role the xfce power manager segfault plays.
please attach the acpidump output.
Created attachment 278207 [details] acpidump output Sure, here is the output of acpidump, without any switches. I hope it contains clues …
please make sure your kernel is built with CONFIG_PINCTRL_CHERRYVIEW. It seems that there are a lot of gaps in linux kernel for the vendor specific operation region support on this machine. So to me, the " AE_BAD_PARAMETER, Returned by Handler for .." messages are not kernel bug, but instead, they are Linux gaps to support some vendor specific devices on this platform, and I don't expect we can get this resolved in this bug report.
I have the same machine and I'm not seeing any of these errors. Have you perhaps changed some BIOS settings from the defaults?
I am not sure if I will be able to contribute more to this. My main GPD Pocket died (not as resilient to environment influences as it needs to be in my possession, apparently), and the other is in pieces of varying defectiveness. Yes, I actually spent money for two of those and even tried to use it as main working device for some months. It mostly worked, but I did not try suspend to disk in a long time. I wish everyone luck in getting the Linux support completed.