Created attachment 301691 [details] journalctl -b 0 after booting, closing and reopening laptop lid (hence triggering sleep) I have an HP Spectre x360 Convertible Laptop PC 13t-aw200 (Intel® Core™ i7-1165G7 Processor, Tiger Lake) that does not entirely reach sleep state. My primary evidence is that the battery discharges too quickly and still gives off a minor amount of heat even though it appears to be asleep. I tried to do as much debugging as I could given my limited experience with power management. This led me to this page on linux S0ix troubleshooting: https://01.org/blogs/qwang59/2020/linux-s0ix-troubleshooting which recommended building with "CONFIG_ACPI_DEBUG=y" and booting with the "initcall_debug" option. I did this line in the journalctl: "ACPI FADT declares the system doesn't support PCIe ASPM, so disable it" (on line 619) and hence also included info about my pcie tree but again, I don't know enough to say if this is the issue. Thanks in advance for your help.
Created attachment 301692 [details] dmesg from boot
Created attachment 301693 [details] lspci
Created attachment 301694 [details] lspci -tvvv
Created attachment 301695 [details] turbostat --show Pk%pc10 rtcwake -m freeze -s 15 Turbotstat command recommended for debugging from 01.org
Created attachment 301696 [details] turbostat --show SYS%LPI rtcwake -m freeze -s 15 also as recommended for debugging by 01.org
is this a new install, or did earlier kernel run fine and the problem only started to happen after updating to 5.19? If the latter: is the latest version (5.19.6) is still affected?
Thanks for helping to track this down. This is not a regression (i.e. it happened on older versions as well). I just built/installed 5.19.6 and this line: "ACPI FADT declares the system doesn't support PCIe ASPM, so disable it" was still in the journalctl. I did not yet see if the battery drained at the same rate. I'll do that and publish my findings later.
Th(In reply to Michael Shantzis from comment #7) > > This is not a regression Then that's nothing for my todo list. I assume some developer will look into this report, as according to the maintainers file (https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/plain/MAINTAINERS ) bugzilla.kernel.org is the official place to report ACPI subsystem bugs. If you don't get any help here you might want to report the problem by mail to the maintainers with the lists in CC, as explained here: https://docs.kernel.org/admin-guide/reporting-issues.html HTH!
Hi, If it's possible to disable RST in BIOS (and if you're not dual booting with Windows) you can try that. Note that changing from RST to AHCI mode may prevent a Windows partition from booting. Also, - What Linux distribution and version are you using? - Please provide the output from sudo lspci -vvv. Adding -t will actually cause the verbose output not to display. - Please run and attach the log from turbostat rtcwake -m freeze -s 300 > ts.log. Thanks.
Created attachment 303447 [details] dmesg from boot 2 Requested by rafael@kernel.org for further debugging
Just for clarity, the original version on which I reported this bug (i.e. the one for which I built it with CONFIG_ACPI_DEBUG=y) was 5.19.0-rc1.
Created attachment 303453 [details] lspci -vvv
Created attachment 303454 [details] turbostat rtcwake -m freeze -s 300
Uploaded new logs. Also tried running it on 6.0.12-200 and the problem is still there. Unfortunately, the BIOS on this machine is anemic, with very few options. I did some searches to see if there is an advanced "back door" but there isn't. Rafael and David - I just did some searches and noticed that there was this similar bug reported last year that you commented on: https://bugzilla.kernel.org/show_bug.cgi?id=213717 possibly even a dupe? And it looks like that one wasn't resolved either. Just thought I'd point that out in case it's helpful.