System ------ - Dell XPS 15 9500 - 6.8.1 kernel (Arch) Problem description ------------------- Whenever I try to hibernate the system, hibernation fails with the message "PM: hibernation: Wakeup event detected during hibernation, rolling back." (and, as a side effect, the NVIDIA GPU becomes unresponsive). This is likely a regression bug since, with the LTS kernel 6.6.22, everything works as expected. Here's the link to the Arch forum where we were discussing the issue: https://bbs.archlinux.org/viewtopic.php?id=294111
Please attach full dmesg. Make sure you're running the latest BIOS version.
Created attachment 306032 [details] dmesg of the working setup (kernel 6.6.22)
Created attachment 306033 [details] dmesg of the faulty setup (kernel 6.8.1)
(In reply to Artem S. Tashkinov from comment #1) > Please attach full dmesg. > > Make sure you're running the latest BIOS version. I checked with fwupd and confirm I'm on the latest BIOS version (1.27.0). I've attached the dmesg of both kernel versions.
Created attachment 306034 [details] dmesg of the ThinkPad (kernel 6.8.1) I also confirm that the issue is hardware-specific. I've just tested another laptop (a ThinkPad T14) running the kernel 6.8.1, and hibernation works.
Same problem on Dell XPS 9520, also with latest BIOS, kernel Linux XPS9520 6.8.1-arch1-1 Mär 26 19:35:36 XPS9520 kernel: PM: hibernation: Wakeup event detected during hibernation, rolling back. Mär 26 19:35:36 XPS9520 kernel: printk: Suspending console(s) (use no_console_suspend to debug) Mär 26 19:35:36 XPS9520 kernel: PM: S| Mär 26 19:35:36 XPS9520 kernel: PM: hibernation: Wrote 9406880 kbytes in 8.08 seconds (1164.21 MB/s) Mär 26 19:35:36 XPS9520 kernel: PM: Image saving done Mär 26 19:35:36 XPS9520 kernel: PM: Image saving progress: 90% Mär 26 19:35:36 XPS9520 kernel: PM: Image saving progress: 80% Mär 26 19:35:36 XPS9520 kernel: PM: Image saving progress: 70% Mär 26 19:35:36 XPS9520 kernel: PM: Image saving progress: 60% Mär 26 19:35:36 XPS9520 kernel: PM: Image saving progress: 50% Mär 26 19:35:36 XPS9520 kernel: PM: Image saving progress: 40% Mär 26 19:35:36 XPS9520 kernel: PM: Image saving progress: 30% Mär 26 19:35:36 XPS9520 kernel: PM: Image saving progress: 20% Mär 26 19:35:36 XPS9520 kernel: PM: Image saving progress: 10% Mär 26 19:35:36 XPS9520 kernel: PM: Image saving progress: 0% Mär 26 19:35:36 XPS9520 kernel: PM: Compressing and saving image data (2351720 pages)... Mär 26 19:35:36 XPS9520 kernel: PM: Using 3 thread(s) for compression
(In reply to Mattia Orlandi from comment #5) > Created attachment 306034 [details] > dmesg of the ThinkPad (kernel 6.8.1) > > I also confirm that the issue is hardware-specific. I've just tested another > laptop (a ThinkPad T14) running the kernel 6.8.1, and hibernation works. Strange. I can confirm that the problem is not hardware-specific. On my desktop, the system freezes completely just before shutting down. Sleep works on both. I have rolled back to kernel 6.7.9 and everything works perfectly again.
I'm getting the same problem on a Dell Latitude, on 6.8.1-arch1-1. Rolling back to 6.7.9 fixes it. Suspend works, hibernate doesn't. When I set HibernateMode to shutdown in systemd-sleep, the system powered off anyway after the hibernate failed. When I powered it back on, it seemingly resumed successfully (the tty had a blinking cursor), but no input devices or network devices worked. I had to do a hard shutdown.
I'm having the exact same problem with a Dell Inspiron 16 (7640) BIOS 1.3.0. Staying with 6.7.9 for now.
Same error on Dell XPS 9510 on Kernel 6.8.2.arch1-1. Switching to the latest lts release (6.6.23-1-lts) fixed this issue.
I did some debugging of the issue based on the arch linux 6.8.2 kernel. The first observation is that the changes introduced by commit 0c4cae1bc00d31c78858c184ede351baea232bdb, which now check the return value of hibernation_platform_enter(), lead to the cancellation of the hibernation request. But these changes are not the root cause, they make sense. Closer to the root cause seems to be a kind of spurious 0xCF (power button up) event from the Intel 5 button device that seems to be generated / handled / whatever when the actual hibernation state is entered through platform / ACPI methods. This is nothing new AFAICT and may have been there for a long time. It's not easily visible in the logs when hibernation works, as this happens just after the hibernation has been generated. (At least this is my limited understanding of the hibernation process.) If I change the code in drivers/platform/x86/intel/hid.c:notify_handler() to ignore this "power button up", hibernation works again. (see attached patch 0001-WIP-Ignore-power-button-up-while-entering-sleep-stat.patch) However, there might be a much better way to fix the issue by e.g. not having the "platform" generate this event, at all.
Created attachment 306071 [details] Ignore power-button-up while entering sleep state
FYI, same error on a Dell Inspiron 5406 2n1 running kernel 6.8.2-arch2-1
The same problem occurs on my laptop: OS: openSUSE Tumbleweed x86_64 Host: Latitude 3420 Kernel: 6.8.1-1-default
There is already a fix in the queue: https://lore.kernel.org/platform-driver-x86/878r1tpd6u.fsf_-_@gmail.com/ You might want helping to test it. It also applies to 6.8.2.
Hi, I have installed kernel 6.8.6 that includes the patch but i still can't hibernate with my laptop: OS: Archlinux x86_64 Host: Dell XPS 9380 Kernel: 6.8.6-arch1-1 (Core-Testing) https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/diff/drivers/platform/x86/intel/hid.c?id=v6.8.6&id2=v6.8.5 Apr 14 14:02:05 ACMEPC systemd[1]: Reached target Sleep. Apr 14 14:02:05 ACMEPC systemd[1]: Starting System Hibernate... Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Successfully made thread 1825 of process 1795 owned by '1000' high priority at nice level 0. Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Supervising 7 threads of 5 processes of 1 users. Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Supervising 6 threads of 4 processes of 1 users. Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Supervising 6 threads of 4 processes of 1 users. Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Successfully made thread 1825 of process 1795 owned by '1000' RT at priority 20. Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Supervising 7 threads of 5 processes of 1 users. Apr 14 14:02:05 ACMEPC systemd-sleep[5369]: Performing sleep operation 'hibernate'... Apr 14 14:02:05 ACMEPC kernel: PM: hibernation: hibernation entry Apr 14 14:02:13 ACMEPC kernel: Filesystems sync: 0.002 seconds Apr 14 14:02:13 ACMEPC kernel: Freezing user space processes Apr 14 14:02:13 ACMEPC kernel: Freezing user space processes completed (elapsed 0.001 seconds) Apr 14 14:02:13 ACMEPC kernel: OOM killer disabled. Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem 0x00000000-0x00000fff] Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem 0x0009e000-0x0009efff] Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem 0x000a0000-0x000fffff] Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem 0x2b871000-0x2b872fff] Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem 0x35941000-0x359b4fff] Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem 0x3a551000-0x3fffefff] Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem 0x40000000-0x47ffffff] Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem 0x48e00000-0xffffffff] Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Basic memory bitmaps created Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Preallocating image memory Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Allocated 749300 pages for snapshot Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Allocated 2997200 kbytes in 3.77 seconds (795.01 MB/s) Apr 14 14:02:13 ACMEPC kernel: Freezing remaining freezable tasks Apr 14 14:02:13 ACMEPC kernel: Freezing remaining freezable tasks completed (elapsed 0.778 seconds) Apr 14 14:02:13 ACMEPC kernel: printk: Suspending console(s) (use no_console_suspend to debug) Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: interrupt blocked Apr 14 14:02:13 ACMEPC kernel: ACPI: PM: Preparing to enter system sleep state S4 Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: event blocked Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: EC stopped Apr 14 14:02:13 ACMEPC kernel: ACPI: PM: Saving platform NVS memory Apr 14 14:02:13 ACMEPC kernel: Disabling non-boot CPUs ... Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 1 is now offline Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 2 is now offline Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 3 is now offline Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 4 is now offline Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 5 is now offline Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 6 is now offline Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 7 is now offline Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Creating image: Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Need to copy 717514 pages Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Normal pages needed: 717514 + 1024, available pages: 1287420 Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Image created (650205 pages copied, 67309 zero pages) Apr 14 14:02:13 ACMEPC kernel: ACPI: PM: Restoring platform NVS memory Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: EC started Apr 14 14:02:13 ACMEPC kernel: Enabling non-boot CPUs ... Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 1 APIC 0x2 Apr 14 14:02:13 ACMEPC kernel: CPU1 is up Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 2 APIC 0x4 Apr 14 14:02:13 ACMEPC kernel: CPU2 is up Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 3 APIC 0x6 Apr 14 14:02:13 ACMEPC kernel: CPU3 is up Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 4 APIC 0x1 Apr 14 14:02:13 ACMEPC kernel: CPU4 is up Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 5 APIC 0x3 Apr 14 14:02:13 ACMEPC kernel: CPU5 is up Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 6 APIC 0x5 Apr 14 14:02:13 ACMEPC kernel: CPU6 is up Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 7 APIC 0x7 Apr 14 14:02:13 ACMEPC kernel: CPU7 is up Apr 14 14:02:13 ACMEPC kernel: ACPI: PM: Waking up from system sleep state S4 Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: interrupt unblocked Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: event unblocked Apr 14 14:02:13 ACMEPC kernel: nvme nvme0: 8/0/0 default/read/poll queues Apr 14 14:02:13 ACMEPC kernel: PM: Using 3 thread(s) for compression Apr 14 14:02:13 ACMEPC kernel: PM: Compressing and saving image data (651607 pages)... Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 0% Apr 14 14:02:13 ACMEPC kernel: mei_hdcp 0000:00:16.0-b638ab7e-94e2-4ea2-a552-d1c54b627f04: bound 0000:00:02.0 (ops i915_hdcp_ops [i915]) Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 10% Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 20% Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 30% Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 40% Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 50% Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 60% Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 70% Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 80% Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 90% Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 100% Apr 14 14:02:13 ACMEPC kernel: PM: Image saving done Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Wrote 2606428 kbytes in 2.12 seconds (1229.44 MB/s) Apr 14 14:02:13 ACMEPC kernel: PM: S| Apr 14 14:02:13 ACMEPC kernel: printk: Suspending console(s) (use no_console_suspend to debug) Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Wakeup event detected during hibernation, rolling back. Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Basic memory bitmaps freed Apr 14 14:02:13 ACMEPC kernel: pci_bus 0000:05: Allocating resources Apr 14 14:02:13 ACMEPC kernel: pci_bus 0000:39: Allocating resources Apr 14 14:02:13 ACMEPC kernel: OOM killer enabled. Apr 14 14:02:13 ACMEPC kernel: Restarting tasks ... done. Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: The canary thread is apparently starving. Taking action. Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Demoting known real-time threads. Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 1825 of process 1795. Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 3048 of process 3030. Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 2337 of process 2327. Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 2327 of process 2327. Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 2325 of process 2325. Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 2329 of process 2312. Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 2312 of process 2312. Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Demoted 7 threads. Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: hibernation exit Apr 14 14:02:13 ACMEPC systemd-sleep[5369]: System returned from sleep operation 'hibernate'. Apr 14 14:02:13 ACMEPC systemd[1]: systemd-hibernate.service: Deactivated successfully. Apr 14 14:02:13 ACMEPC systemd[1]: Finished System Hibernate. Apr 14 14:02:13 ACMEPC systemd[1]: systemd-hibernate.service: Consumed 4.762s CPU time. Apr 14 14:02:13 ACMEPC systemd[1]: Reached target System Hibernation. Apr 14 14:02:13 ACMEPC systemd[1]: Stopped target Sleep. Apr 14 14:02:13 ACMEPC systemd-logind[589]: Operation 'hibernate' finished. Apr 14 14:02:13 ACMEPC systemd[1]: Stopped target System Hibernation. Best regards,
Using Kernel: 6.8.6-arch1-1 (Core-Testing) makes hibernation on the Dell Inspiron 16 (7640) work again.
Ok, 6.8.6 doesn't work fully for me, either. systemctl hibernate: OK power button (via systemd): OK Gnome menu entries: NOT OK Will try to investigate as soon as time permits.
(In reply to Enrik Berkhan from comment #18) > Ok, 6.8.6 doesn't work fully for me, either. > > systemctl hibernate: OK > power button (via systemd): OK > Gnome menu entries: NOT OK > > Will try to investigate as soon as time permits. I have performed the test using all the methods without success: systemctl hibernate: NOT OK power button (via systemd): NOT OK Gnome menu entries: NOT OK I will be testing with kernel v6.7.9 to validate that my system is not the cause. Thank you for your help!
Created attachment 306150 [details] xps 9380 systemd hibernate kernel v6.7.9
Tested with kernel v6.7.9 with good results. Linux ACMEPC 6.7.9-arch1-1 #1 SMP PREEMPT_DYNAMIC Fri, 08 Mar 2024 01:59:01 +0000 x86_64 GNU/Linux systemctl hibernate: OK power button (via systemd): OK Gnome menu entries: OK
(In reply to Jonathan Colon from comment #16) > Hi, > > I have installed kernel 6.8.6 that includes the patch but i still can't > hibernate with my laptop: > > OS: Archlinux x86_64 > Host: Dell XPS 9380 > Kernel: 6.8.6-arch1-1 (Core-Testing) > > https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/diff/ > drivers/platform/x86/intel/hid.c?id=v6.8.6&id2=v6.8.5 > > Apr 14 14:02:05 ACMEPC systemd[1]: Reached target Sleep. > Apr 14 14:02:05 ACMEPC systemd[1]: Starting System Hibernate... > Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Successfully made thread 1825 of > process 1795 owned by '1000' high priority at nice level 0. > Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Supervising 7 threads of 5 > processes of 1 users. > Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Supervising 6 threads of 4 > processes of 1 users. > Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Supervising 6 threads of 4 > processes of 1 users. > Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Successfully made thread 1825 of > process 1795 owned by '1000' RT at priority 20. > Apr 14 14:02:05 ACMEPC rtkit-daemon[859]: Supervising 7 threads of 5 > processes of 1 users. > Apr 14 14:02:05 ACMEPC systemd-sleep[5369]: Performing sleep operation > 'hibernate'... > Apr 14 14:02:05 ACMEPC kernel: PM: hibernation: hibernation entry > Apr 14 14:02:13 ACMEPC kernel: Filesystems sync: 0.002 seconds > Apr 14 14:02:13 ACMEPC kernel: Freezing user space processes > Apr 14 14:02:13 ACMEPC kernel: Freezing user space processes completed > (elapsed 0.001 seconds) > Apr 14 14:02:13 ACMEPC kernel: OOM killer disabled. > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem > 0x00000000-0x00000fff] > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem > 0x0009e000-0x0009efff] > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem > 0x000a0000-0x000fffff] > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem > 0x2b871000-0x2b872fff] > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem > 0x35941000-0x359b4fff] > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem > 0x3a551000-0x3fffefff] > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem > 0x40000000-0x47ffffff] > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Marking nosave pages: [mem > 0x48e00000-0xffffffff] > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Basic memory bitmaps created > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Preallocating image memory > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Allocated 749300 pages for > snapshot > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Allocated 2997200 kbytes in > 3.77 seconds (795.01 MB/s) > Apr 14 14:02:13 ACMEPC kernel: Freezing remaining freezable tasks > Apr 14 14:02:13 ACMEPC kernel: Freezing remaining freezable tasks completed > (elapsed 0.778 seconds) > Apr 14 14:02:13 ACMEPC kernel: printk: Suspending console(s) (use > no_console_suspend to debug) > Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: interrupt blocked > Apr 14 14:02:13 ACMEPC kernel: ACPI: PM: Preparing to enter system sleep > state S4 > Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: event blocked > Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: EC stopped > Apr 14 14:02:13 ACMEPC kernel: ACPI: PM: Saving platform NVS memory > Apr 14 14:02:13 ACMEPC kernel: Disabling non-boot CPUs ... > Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 1 is now offline > Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 2 is now offline > Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 3 is now offline > Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 4 is now offline > Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 5 is now offline > Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 6 is now offline > Apr 14 14:02:13 ACMEPC kernel: smpboot: CPU 7 is now offline > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Creating image: > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Need to copy 717514 pages > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Normal pages needed: 717514 > + 1024, available pages: 1287420 > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Image created (650205 pages > copied, 67309 zero pages) > Apr 14 14:02:13 ACMEPC kernel: ACPI: PM: Restoring platform NVS memory > Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: EC started > Apr 14 14:02:13 ACMEPC kernel: Enabling non-boot CPUs ... > Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 1 APIC 0x2 > Apr 14 14:02:13 ACMEPC kernel: CPU1 is up > Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 2 APIC 0x4 > Apr 14 14:02:13 ACMEPC kernel: CPU2 is up > Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 3 APIC 0x6 > Apr 14 14:02:13 ACMEPC kernel: CPU3 is up > Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 4 APIC 0x1 > Apr 14 14:02:13 ACMEPC kernel: CPU4 is up > Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 5 APIC 0x3 > Apr 14 14:02:13 ACMEPC kernel: CPU5 is up > Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 6 APIC 0x5 > Apr 14 14:02:13 ACMEPC kernel: CPU6 is up > Apr 14 14:02:13 ACMEPC kernel: smpboot: Booting Node 0 Processor 7 APIC 0x7 > Apr 14 14:02:13 ACMEPC kernel: CPU7 is up > Apr 14 14:02:13 ACMEPC kernel: ACPI: PM: Waking up from system sleep state S4 > Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: interrupt unblocked > Apr 14 14:02:13 ACMEPC kernel: ACPI: EC: event unblocked > Apr 14 14:02:13 ACMEPC kernel: nvme nvme0: 8/0/0 default/read/poll queues > Apr 14 14:02:13 ACMEPC kernel: PM: Using 3 thread(s) for compression > Apr 14 14:02:13 ACMEPC kernel: PM: Compressing and saving image data (651607 > pages)... > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 0% > Apr 14 14:02:13 ACMEPC kernel: mei_hdcp > 0000:00:16.0-b638ab7e-94e2-4ea2-a552-d1c54b627f04: bound 0000:00:02.0 (ops > i915_hdcp_ops [i915]) > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 10% > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 20% > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 30% > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 40% > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 50% > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 60% > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 70% > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 80% > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 90% > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving progress: 100% > Apr 14 14:02:13 ACMEPC kernel: PM: Image saving done > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Wrote 2606428 kbytes in 2.12 > seconds (1229.44 MB/s) > Apr 14 14:02:13 ACMEPC kernel: PM: S| > Apr 14 14:02:13 ACMEPC kernel: printk: Suspending console(s) (use > no_console_suspend to debug) > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Wakeup event detected during > hibernation, rolling back. > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: Basic memory bitmaps freed > Apr 14 14:02:13 ACMEPC kernel: pci_bus 0000:05: Allocating resources > Apr 14 14:02:13 ACMEPC kernel: pci_bus 0000:39: Allocating resources > Apr 14 14:02:13 ACMEPC kernel: OOM killer enabled. > Apr 14 14:02:13 ACMEPC kernel: Restarting tasks ... done. > Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: The canary thread is apparently > starving. Taking action. > Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Demoting known real-time threads. > Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 1825 > of process 1795. > Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 3048 > of process 3030. > Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 2337 > of process 2327. > Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 2327 > of process 2327. > Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 2325 > of process 2325. > Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 2329 > of process 2312. > Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Successfully demoted thread 2312 > of process 2312. > Apr 14 14:02:13 ACMEPC rtkit-daemon[859]: Demoted 7 threads. > Apr 14 14:02:13 ACMEPC kernel: PM: hibernation: hibernation exit > Apr 14 14:02:13 ACMEPC systemd-sleep[5369]: System returned from sleep > operation 'hibernate'. > Apr 14 14:02:13 ACMEPC systemd[1]: systemd-hibernate.service: Deactivated > successfully. > Apr 14 14:02:13 ACMEPC systemd[1]: Finished System Hibernate. > Apr 14 14:02:13 ACMEPC systemd[1]: systemd-hibernate.service: Consumed > 4.762s CPU time. > Apr 14 14:02:13 ACMEPC systemd[1]: Reached target System Hibernation. > Apr 14 14:02:13 ACMEPC systemd[1]: Stopped target Sleep. > Apr 14 14:02:13 ACMEPC systemd-logind[589]: Operation 'hibernate' finished. > Apr 14 14:02:13 ACMEPC systemd[1]: Stopped target System Hibernation. > > Best regards, I compiled intel-hid.ko for both 6.8.4 and 6.8.5 with this patch applied on OpenSuSE Tumbleweed and it seems to work. I can hibernate.
(In reply to Enrik Berkhan from comment #15) > There is already a fix in the queue: > > https://lore.kernel.org/platform-driver-x86/878r1tpd6u.fsf_-_@gmail.com/ > > You might want helping to test it. It also applies to 6.8.2. Sorry, replying to the wrong comment. So: I compiled intel-hid.ko for both 6.8.4 and 6.8.5 with this patch applied on OpenSuSE Tumbleweed and it seems to work. I can hibernate.
(In reply to Tamás Németh from comment #23) > (In reply to Enrik Berkhan from comment #15) > > There is already a fix in the queue: > > > > https://lore.kernel.org/platform-driver-x86/878r1tpd6u.fsf_-_@gmail.com/ > > > > You might want helping to test it. It also applies to 6.8.2. > > Sorry, replying to the wrong comment. So: > > I compiled intel-hid.ko for both 6.8.4 and 6.8.5 with this patch applied on > OpenSuSE Tumbleweed and it seems to work. I can hibernate. ...but it doesn't work always. Sometimes(!) my laptop bounces back from hibernation just like before the patch :-(
I compiled a vanilla kernel 6.8.5 only with the patch but I still can't hibernate with my laptop. Additionally I performed the "Per sub-system hibernate testing" procedures but in every result I never receive the message "Wakeup event detected during hibernation, rolling back" as if everything were working fine. https://wiki.ubuntu.com/DebuggingKernelHibernate
After adding some kernel logs telling me when there are events from the "five button device", I maybe got some insight, which might well be specific for my machine. If and only if AC is plugged and I enter hibernate via software (like using the Gnome menu or 'systemctl hibernate'), I see 0xCE (Power button down) events while the driver is in the wakeup_mode == 1 state. These events are not (yet) suppressed by the existing patch, which made it into 6.8.6 BTW. These events seem to be the root cause for cancelling hibernation in my case. When on battery, I only see 0xCF (Power button up) events. These are now filtered having the fix applied. Maybe Dell uses the different events depending on AC connected or not to do some magic? No idea ... Anyone else seeing different behavior depending on AC connected?
Did the test on battery with same results. systemctl hibernate: NOT OK power button (via systemd): NOT OK Gnome menu entries: NOT OK How can i get the events generate by the kernel? Thanks!
I added this line dev_info(&device->dev, "event 0x%x wakeup_mode %d", event, priv->wakeup_mode); in drivers/platform/x86/intel/hid.c at the top of notify_handler() to get dmesg output for the "Intel HID" events. However, I'm following a new strategy right now: blacklisting the intel_hid module from the kernel command line. In this case (unloading the module after it had been loaded once does not work!), the ACPI BIOS seems to send "classic" power button events that can be used e.g. to enter hibernate mode and there will be no more "five button" (ACPI BIOS simulated?) events that puzzle the code. YMMV, e.g. for hardware that actually has fancy extra buttons ...
(In reply to Enrik Berkhan from comment #28) > I added this line > > dev_info(&device->dev, "event 0x%x wakeup_mode %d", event, > priv->wakeup_mode); > > in drivers/platform/x86/intel/hid.c at the top of notify_handler() to get > dmesg output for the "Intel HID" events. Another option not needing a kernel rebuild is to enable some runtime ACPI debugging, e.g. to see event notifications: echo 4 > /sys/module/acpi/parameters/debug_layer You should also verify that at least ACPI_LV_INFO is enabled in /sys/module/acpi/parameters/debug_level.
Enabled the debugging on ACPI. [root@ACMEPC rebelinux]# cat /sys/module/acpi/parameters/debug_layer | grep [*] ACPI_EVENTS 0x00000004 [*] debug_layer = 0x00000004 ( * = enabled) [root@ACMEPC rebelinux]# cat /sys/module/acpi/parameters/debug_level | grep [*] ACPI_LV_INFO 0x00000004 [*] debug_level = 0x00000004 (* = enabled) [root@ACMEPC rebelinux]#
Created attachment 306174 [details] xps 9380 dmesg after ACPI debugging
(In reply to Jonathan Colon from comment #31) > Created attachment 306174 [details] > xps 9380 dmesg after ACPI debugging I think these are the interesting lines, logged when the system comes back from the ephemeral S4 state that had been entered for image creation, before writing the hibernation image out to disk: [ 2934.829876] ACPI: PM: Waking up from system sleep state S4 [ 2934.830050] evmisc-0132 ev_queue_notify_reques: Dispatching Notify on [PBTN] (Device) Value 0x02 (Device Wake) Node 00000000b9046256 The firmware of your notebook sends a (spurious?) Power Button Wakeup. This wakeup is very likely the root cause for cancelling the hibernation process. This also explains why the patch does not work for you: the patch only addresses similar behavior when it comes from the "5 button device" power button. (The "5 button device", when present, takes over the power button functionality and disables the "classic" power button when its driver is loaded.) Maybe a more general solution for this issue is to ignore all wakeup events that arrive between the "waking up from S4 to save the data to disk" and the actual power down. I guess this needs to be discussed on the x86 platform kernel mailing list.
Adding to the conversation if I put the module intel_hid on the blacklist I can do the hibernate without problem. module_blacklist=intel_hid thanks for your time and help!
(In reply to Enrik Berkhan from comment #26) > Maybe Dell uses the different events depending on AC connected or not to do > some magic? No idea ... > > Anyone else seeing different behavior depending on AC connected? Maybe... I couldn't find the exact causality. Sometimes (probably if I load the newly compiled intel_hid while AC is plugged in?) hibernation still bounces back on my Dell Latitude 3420 laptop. But it mostly works.
Created attachment 306188 [details] dmesg of the XPS 9500 on kernel 6.8.7 (working) I've just tested hibernation on the XPS 15 9500 and hibernation seems to work again with kernel 6.8.7.