[SOLVED] Suspend broken after recent update

branch · 2022-02-27 23:56:02

This is the offending commit:

commit 4287509b4d21e34dc49266c6239920811e189e40
Author: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Date:   Fri Feb 4 18:31:02 2022 +0100

    ACPI: PM: s2idle: Cancel wakeup before dispatching EC GPE

I have a custom linux-5.16.10.arch1-1 build with just this commit reverted, and it is behaving normally.

JoeyCorleone · 2022-02-28 11:38:03

Good job identifying the offending commit.

Have you already considered reporting it to upstream? A quick search for "s2idle" leads to this similar(?) ticket. Look for other similar tickets and make sure to put Rafael J. Wysocki and Sasha Levin on CC when reporting.

branch · 2022-02-28 12:51:24

Yes, but I should do due diligence first, so I'm currently testing with 5.17-rc6. It will take more time to be convinced, but initial results are encouraging.

willy9 · 2022-02-28 21:21:06

I also have the issue but reverting to upower 0.99.15-1 fixes it in my case. upower 0.99.16-2 does not fix it for me. My kernel is 5.16.11-arch1-1. Could it be cause by some kernel + upower combination ?

branch · 2022-03-01 01:55:38

I guess it's possible, I never tried linux-5.16.11 with reverted upower, just 5.16.10.

Anyway, linux 5.17-rc6 is behaving well for me, so it may be fixed upstream already. I'll probably just stick with linux 5.17-rc6 or linux 5.16.9 until the official 5.17 is released.

j-james · 2022-03-02 10:57:40

@branch I think you're right - after downgrading to linux-lts 5.15.22 the issue appears to be gone.

flan_suse · 2022-03-04 00:56:30

willy9 wrote:

I also have the issue but reverting to upower 0.99.15-1 fixes it in my case. upower 0.99.16-2 does not fix it for me. My kernel is 5.16.11-arch1-1. Could it be cause by some kernel + upower combination ?

This is what I'm thinking too.

Even with 5.15.22, the issue persists, so it may be upower and/or a combination with upower + recent kernel update.

V1del · 2022-03-04 01:56:07

The upower bug only affected/affects the trigger for the lid close event, it should not have an effect on actual suspension via other means e.g. systemctl suspend. Upower got a update that fixed this (again this should've only affected the trigger for the lid event, not the suspension in and of itself). If you have actual issues suspending then you likely have a kernel regression without much relation to upower

Last edited by V1del (2022-03-04 01:57:29)

flan_suse · 2022-03-04 02:37:21

adamjbc wrote:

V1del wrote:
If systemctl suspend works but closing the lid doesn't then this is possibly this bug in the current upower package: https://bugs.archlinux.org/task/73942
Thanks. Unfortunately I don't use upower, so I don't think this can be related to the issue I am having. For me, it appears that the laptop screen is turning off correctly. And according to dmesg it is reaching s2idle. But it carries on doing something because it starts to get warm. When I open the lid, it does not resume (or rather the screen does not come on) until I start hitting buttons.
The problem happens regardless of whether I shut the lid or use systemctl suspend.

Then it must be the regression in the kernel, since the above precisely describes the same issue.

I can try with the 5.17-rc kernel, but I'm not a fan (no pun intended) of using a release candidate kernel. (Quick test is fine though.)

flan_suse · 2022-03-07 13:46:07

branch wrote:

Anyway, linux 5.17-rc6 is behaving well for me, so it may be fixed upstream already.

I wonder if this will be backported for kernel 5.15?

dcy3rka · 2022-03-08 23:54:24

Sounds like the same bug: https://forums.lenovo.com/t5/Other-Linu … -p/5132432

flan_suse · 2022-03-09 00:29:32

dcy3rka wrote:

Sounds like the same bug: https://forums.lenovo.com/t5/Other-Linu … -p/5132432

I'm perplexed that a bug/commit this serious was backported into an "LTS" kernel, and the effects are more than "inconvenient": It's literally cooking unsuspecting users' laptops and stressing their batteries. The whole reason for sticking with an LTS kernel (such as 5.15), is to be spared of these types of issues that might affect future kernels. You trade new features for more consistent stability. (At least that's what I thought.)

I'm using 5.17-rc7 for now, but it also introduced new issues with CIFS connections. Hopefully 5.15 can be a safe-haven until the next LTS kernel after this is resolved. Thankfully, with 5.17-rc7, I can safely suspend my laptop and not worry about setting my backpack or house on fire.

TreeForDev · 2022-03-10 18:26:41

upower updated to 0.99.17 https://gitlab.freedesktop.org/upower/u … s/v0.99.17
Looks like it solves the issue

Last edited by TreeForDev (2022-03-10 19:27:09)

Cakegg · 2022-03-11 00:46:13

EDIT: I do not have upower installed.

My laptop did not suspend with linux-lts-5.15.22. With 5.17-rc7, it seems fine (I am not sure as I have only tested it for two days).

Last edited by Cakegg (2022-03-11 00:48:32)

dcy3rka · 2022-03-20 10:37:47

Since a few day it seems fine. Not sure if it is from the upower update (0.99.17-1) or a kernel update (5.16.15-arch1-1) or something else.

flan_suse · 2022-03-20 12:21:32

dcy3rka wrote:

Since a few day it seems fine. Not sure if it is from the upower update (0.99.17-1) or a kernel update (5.16.15-arch1-1) or something else.

Fixed in kernel. Can confirm it was fixed (backported) in LTS 5.15 as well.

Last edited by flan_suse (2022-03-20 12:21:46)

ifaigios · 2022-03-30 15:35:47

While Linux 5.16.16 seems to be working perfectly for me, Linux 5.17 again breaks suspend unfortunately
My X1 Carbon Gen 9 wakes up immediately after sleeping. My system is fully upgraded and standby mode is set in BIOS to S0ix "Windows and Linux".

codl · 2022-03-31 05:33:48

ifaigios wrote:

While Linux 5.16.16 seems to be working perfectly for me, Linux 5.17 again breaks suspend unfortunately
My X1 Carbon Gen 9 wakes up immediately after sleeping. My system is fully upgraded and standby mode is set in BIOS to S0ix "Windows and Linux".

Same here on a Dell Latitude 5310. linux 5.16.16.arch1-1 suspends fine, linux 5.17.arch1-1 or 5.17.1.arch1-1 wakes up on its own a few seconds after suspending.

ndhakara · 2022-03-31 08:40:37

I have the same issue with new 5.17.1 kernel and now switched to the LTS kernel. there is no issue with LTS kernel.

ifaigios · 2022-03-31 22:21:15

Found a workaround: Disabling "Always-on USB" from BIOS settings (Config > USB) stops making the laptop wake up by itself. This is still a clear kernel regression so I think a kernel bug should be opened.

EDIT: Scratch that, problem is back again.

Last edited by ifaigios (2022-04-04 23:59:25)

plumlis · 2022-04-05 00:58:52

Disable bluetooth can make laptop suspend work again.

ivanbogomolov · 2022-04-05 05:39:23

In my case suspend works but it takes long time to wake up. Laptop woke up quickly before update, but now it's slow. It takes about 1 minute to wake up.
I have dell g15 with intel i7 and what i see in dmesg after slow wake up.

[Вт апр 5 14:57:34 2022] systemd[1]: systemd-suspend.service: Consumed 8.346s CPU time.
[Вт апр 5 14:57:34 2022] systemd[1]: Stopped target Sleep.
[Вт апр 5 14:57:34 2022] systemd[1]: Reached target Suspend.
[Вт апр 5 14:57:34 2022] systemd[1]: Stopped target Suspend.
[Вт апр 5 14:57:34 2022] Generic FE-GE Realtek PHY r8169-0-300:00: attached PHY driver (mii_bus:phy_addr=r8169-0-300:00, irq=MAC)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: Process 350 (systemd-journal) of user 0 dumped core.
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: Coredump diverted to /var/lib/systemd/coredump/core.systemd-journal.0.9fe6502531ee4906a880fe4c77a21258.350.1649136648000000.zst
...
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: Module linux-vdso.so.1 with build-id ae8518a0710c2679844504efb751b01abf13cef5
...
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: Stack trace of thread 350:
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #0 0x00007fd4da6ff15a __futex_abstimed_wait_common (libc.so.6 + 0x8a15a)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #1 0x00007fd4da7040e4 __pthread_clockjoin_ex (libc.so.6 + 0x8f0e4)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #2 0x00007fd4daadc56f journal_file_set_offline_thread_join (libsystemd-shared-250.so + 0x21656f)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #3 0x00007fd4daadc668 n/a (libsystemd-shared-250.so + 0x216668)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #4 0x00007fd4daadd948 journal_file_append_object (libsystemd-shared-250.so + 0x217948)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #5 0x00007fd4daade8cb n/a (libsystemd-shared-250.so + 0x2188cb)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #6 0x00007fd4daadf1d1 journal_file_append_entry (libsystemd-shared-250.so + 0x2191d1)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #7 0x0000557f3f04f479 n/a (systemd-journald + 0xf479)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #8 0x0000557f3f05462b n/a (systemd-journald + 0x1462b)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #9 0x00007fd4dab03d9e n/a (libsystemd-shared-250.so + 0x23dd9e)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #10 0x00007fd4dab03f8e sd_event_dispatch (libsystemd-shared-250.so + 0x23df8e)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #11 0x00007fd4dab05d49 sd_event_run (libsystemd-shared-250.so + 0x23fd49)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #12 0x0000557f3f047b6e n/a (systemd-journald + 0x7b6e)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #13 0x00007fd4da6a2310 __libc_start_call_main (libc.so.6 + 0x2d310)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #14 0x00007fd4da6a23c1 __libc_start_main@@GLIBC_2.34 (libc.so.6 + 0x2d3c1)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #15 0x0000557f3f0482a5 n/a (systemd-journald + 0x82a5)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: Stack trace of thread 89574:
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #0 0x00007fd4da77ddbb fsync (libc.so.6 + 0x108dbb)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #1 0x0000557f3f05759a n/a (systemd-journald + 0x1759a)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #2 0x0000557f3f057886 n/a (systemd-journald + 0x17886)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #3 0x00007fd4da7025c2 start_thread (libc.so.6 + 0x8d5c2)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: #4 0x00007fd4da787584 __clone (libc.so.6 + 0x112584)
[Вт апр 5 14:57:34 2022] systemd-coredump[89615]: ELF object binary architecture: AMD x86-64

----
uname -a
Linux 5.17.1-arch1-1 #1 SMP PREEMPT Mon, 28 Mar 2022 20:55:33 +0000 x86_64 GNU/Linux

Deneb · 2022-04-06 00:36:49

My ThinkPad T480 has also been having issues with somewhat consistently waking from sleep mode seconds after closing the lid. Does not happen when suspending manually. Issue is present in both linux and linux-zen 5.17.1-arch1-1, not present in 5.16.9-arch1-1 as reported earlier. Seems to be Bluetooth-related in my case as the issue only occurs when it's enabled.

Last edited by Deneb (2022-04-06 00:49:16)

ifaigios · 2022-04-07 12:30:30

Opened bug:

https://bugs.archlinux.org/task/74387

plumlis · 2022-05-04 21:24:59

Some progress on this bug.

https://bugzilla.kernel.org/show_bug.cgi?id=215768

It related to this commit

80740ebb7e1ad15ab9c11425dcd26e073f86d74b
Bluetooth: hci_sync: Fix hci_update_accept_list_sync

Arch Linux

#26 2022-02-27 23:56:02

Re: [SOLVED] Suspend broken after recent update

#27 2022-02-28 11:38:03

Re: [SOLVED] Suspend broken after recent update

#28 2022-02-28 12:51:24

Re: [SOLVED] Suspend broken after recent update

#29 2022-02-28 21:21:06

Re: [SOLVED] Suspend broken after recent update

#30 2022-03-01 01:55:38

Re: [SOLVED] Suspend broken after recent update

#31 2022-03-02 10:57:40

Re: [SOLVED] Suspend broken after recent update

#32 2022-03-04 00:56:30

Re: [SOLVED] Suspend broken after recent update

#33 2022-03-04 01:56:07

Re: [SOLVED] Suspend broken after recent update

#34 2022-03-04 02:37:21

Re: [SOLVED] Suspend broken after recent update

#35 2022-03-07 13:46:07

Re: [SOLVED] Suspend broken after recent update

#36 2022-03-08 23:54:24

Re: [SOLVED] Suspend broken after recent update

#37 2022-03-09 00:29:32

Re: [SOLVED] Suspend broken after recent update

#38 2022-03-10 18:26:41

Re: [SOLVED] Suspend broken after recent update

#39 2022-03-11 00:46:13

Re: [SOLVED] Suspend broken after recent update

#40 2022-03-20 10:37:47

Re: [SOLVED] Suspend broken after recent update

#41 2022-03-20 12:21:32

Re: [SOLVED] Suspend broken after recent update

#42 2022-03-30 15:35:47

Re: [SOLVED] Suspend broken after recent update

#43 2022-03-31 05:33:48

Re: [SOLVED] Suspend broken after recent update

#44 2022-03-31 08:40:37

Re: [SOLVED] Suspend broken after recent update

#45 2022-03-31 22:21:15

Re: [SOLVED] Suspend broken after recent update

#46 2022-04-05 00:58:52

Re: [SOLVED] Suspend broken after recent update

#47 2022-04-05 05:39:23

Re: [SOLVED] Suspend broken after recent update

#48 2022-04-06 00:36:49

Re: [SOLVED] Suspend broken after recent update

#49 2022-04-07 12:30:30

Re: [SOLVED] Suspend broken after recent update

#50 2022-05-04 21:24:59

Re: [SOLVED] Suspend broken after recent update

Board footer