[ABANDONED] Can't wake up from sleep

light9876 · 2023-03-21 15:01:13

Hello, I use the following command to sleep my laptop (Lenovo T14):

systemctl suspend

Since last week, whenever I sent it to sleep, it won't wake up. Not by any keyboard or trackpad buttons, not even by pressing the power button!

Any ideas please?

Thanks.

Last edited by light9876 (2023-12-09 23:18:34)

seth · 2023-03-21 21:39:48

https://01.org/blogs/rzhang/2015/best-p … ate-issues

light9876 · 2023-03-22 12:49:44

Thank you, but this article contains information about updating a system file:

 echo mem> /sys/power/state

which requires sudo permissions.. I don't want to use sudo permissions...

seth · 2023-03-22 13:06:10

Well, and the gods don't want your system to wake from sleep.
I guess there's nothing that can be done about that.

jchau_xyz · 2023-03-31 04:35:09

Hi, I actually have been having this problem for a couple weeks now.

I used to be able to close/open my lid, which would suspend/wake up my laptop without any issues.

Now, whenever I do that, run `systemctl suspend`, or `echo mem > /sys/power/state`, I get a black screen and can't interact with anything. I need to hard reboot.

The only notable thing I could think about that could have changed during that time span is going from `mesa` to `mesa-git`, but I'm not sure whether this is a graphic issue (more on this later).

So I tried to figure things out through the link you provided:

I tried using the lts kernel, but the issue seems to persist there as well, so I don't think it's a regression? EDIT: I am currently on 6.2.8-arch1-1 if that matters.
journalctl (with the boot options) only reports that I've entered sleep state 'suspend' and "PM: suspend entry (s2idle)".
I tried to do 4.7 since it seems like it describes my issue, but my knowledge is sorta limited and I don't know where I would be looking for these enable_irq_wake(). Maybe mesa drivers?
I can't ssh into my laptop after "waking up" my laptop, so maybe not related to `mesa-git`?

I also tried using hibernate, but the outcome is similar.

Here's my `cat /proc/acpi/wakeup`:

Device	S-state	 Status   Sysfs node
GPP0	 S4	*disabled
GPP1	 S4	*enabled   pci:0000:00:01.2
GPP2	 S0	*disabled
L860	 S0	*disabled
GPP5	 S4	*disabled
GPP7	 S4	*disabled
GP11	 S4	*enabled   pci:0000:00:03.1
SWUS	 S4	*disabled
GP12	 S4	*enabled   pci:0000:00:04.1
SWUS	 S4	*disabled
GP17	 S4	*enabled   pci:0000:00:08.1
XHC0	 S4	*enabled   pci:0000:63:00.3
XHC1	 S4	*enabled   pci:0000:63:00.4
XHC2	 S4	*enabled   pci:0000:64:00.0
NHI0	 S3	*enabled   pci:0000:64:00.5
XHC3	 S3	*enabled   pci:0000:64:00.3
NHI1	 S3	*enabled   pci:0000:64:00.6
XHC4	 S3	*enabled   pci:0000:64:00.4
LID	 S4	*enabled   platform:PNP0C0D:00
SLPB	 S3	*enabled   platform:PNP0C0E:00

Any idea where I should start looking?

Last edited by jchau_xyz (2023-03-31 04:37:59)

seth · 2023-03-31 06:24:36

Maybe mesa drivers?

This is rather about kernel modules.

PM: suspend entry (s2idle)

cat /sys/power/mem_sleep

Please post your complete system journal for the boot:

sudo journalctl -b | curl -F 'file=@-' 0x0.st

And for convenience, "lspci -v" and "lsusb -tv"

Ceterum censeo: if there's a parallel windows installation, see the 3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.

Also look out for BIOS updates.

jchau_xyz · 2023-04-01 09:16:57

cat /sys/power/mem_sleep

[s2idle]

Please post your complete system journal for the boot:

http://0x0.st/HH9M.txt

And for convenience, "lspci -v" and "lsusb -tv"

Here's lspci -v:
http://0x0.st/HH9S.txt

Here's lsusb -tv:
http://0x0.st/HH9Q.txt

Ceterum censeo: if there's a parallel windows installation, see the 3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.

I do not have Windows installed at all on this laptop, so I guess I don't have to worry about this

Also look out for BIOS updates.

Seems like I am up-to-date according to fwupdmgr.

Last edited by jchau_xyz (2023-04-01 09:20:35)

seth · 2023-04-01 12:49:48

[s2idle]

https://wiki.archlinux.org/title/Lenovo … Management
https://wiki.archlinux.org/title/Power_ … end_method

The journal looks cut? The kernel commandline isn't part of it.

Try to rfkill the wifi and disable APST for the nvme, https://wiki.archlinux.org/title/Solid_ … leshooting
Also "iommu=soft" (see the following paragraph)

jchau_xyz · 2023-04-02 07:45:02

The journal looks cut? The kernel commandline isn't part of it.

Sorry, I'm not too sure what this means.
But here's what's in my /etc/default/grub file:

GRUB_CMDLINE_LINUX_DEFAULT="pcie_aspm=force amdgpu.ppfeaturemask=0xfffd7fff initcall_debug no_console_suspend loglevel=3 quiet"

https://wiki.archlinux.org/title/Lenovo … Management
https://wiki.archlinux.org/title/Power_ … end_method

I am not too sure which adjustments are not available for the Z13, I'm not sure I see them specified.
Are these saying that my laptop only supports s2idle as highlighted by `cat /sys/power/mem_sleep`?

Also, related to power management, I also tried disabling TLP, and that did not seem to address the issue.

Try to rfkill the wifi and disable APST for the nvme, https://wiki.archlinux.org/title/Solid_ … leshooting
Also "iommu=soft" (see the following paragraph)

I've tried all of these with little success.

I added `iommu=soft/off/fullflush` and `nvme_core.default_ps_max_latency_us=0` in the grub cmdline above.

seth · 2023-04-02 08:12:26

The head of the journal is missing.

cat /proc/cmdline

Are these saying that my laptop only supports s2idle as highlighted by `cat /sys/power/mem_sleep`?

the wiki wrote:

If your hardware does not advertise the deep sleep status, check first if your UEFI advertise some settings for it, generally under Power or Sleep state or similar wordings, with options named Windows 10, Windows and Linux or S3/Modern standby support for S0ix, and Legacy, Linux, Linux S3 or S3 enabled for S3 sleep.

jchau_xyz · 2023-04-02 21:50:06

cat /proc/cmdline

BOOT_IMAGE=/boot/vmlinuz-linux root=UUID=0833bbaa-5d9d-4bee-99c4-0f45e4701857 rw pcie_aspm=force amdgpu.ppfeaturemask=0xfffd7fff initcall_debug no_console_suspend loglevel=3 quiet

the wiki wrote:
If your hardware does not advertise the deep sleep status, check first if your UEFI advertise some settings for it, generally under Power or Sleep state or similar wordings, with options named Windows 10, Windows and Linux or S3/Modern standby support for S0ix, and Legacy, Linux, Linux S3 or S3 enabled for S3 sleep.

I have gone through the BIOS a couple times now and I have never found any options related to sleep settings. I'll take another look tonight and do some research to see if they are hidden somewhere.

EDIT: I found this forum post that may be related to my issue:
https://forums.lenovo.com/t5/Debian/Lon … =1#5940511

It mentions that 1.55 BIOS update "did address a sleep-related issue, but seems like it may have also added a regression under Linux", although that it also says that I shouldn't have this issue since I am still on 1.54.

Last edited by jchau_xyz (2023-04-02 22:43:07)

seth · 2023-04-03 07:39:03

pcie_aspm=force amdgpu.ppfeaturemask=0xfffd7fff … no_console_suspend

Why are those there?
Did you try w/o and if they're relevant try to isolate the culprit.

jchau_xyz · 2023-04-03 13:52:35

pcie_aspm=force was to allow TLP to change to ASPM, I've had it on for a while so I didn't think of changing it. I could try later and see if that fixes it.

I had some issues where my screen would be completely glitched out or white while watching videos (audio would continue) and the ppfeaturemask line seems to have helped out with that situation.

seth · 2023-04-03 14:20:30

The kernel documentation wrote:

WARNING: Forcing ASPM on may cause system lockups.

Is "no_console_suspend" to analyze the present issue?

"amdgpu.ppfeaturemask=0xfffd7fff" *allows* you to under- or overvolt the GPU, but (afaiu) doesn't do anything by itself?

jchau_xyz · 2023-04-03 22:43:02

Is "no_console_suspend" to analyze the present issue?

Yeah, it was part of the "Best practice to debug suspend issues" link.

"amdgpu.ppfeaturemask=0xfffd7fff" *allows* you to under- or overvolt the GPU, but (afaiu) doesn't do anything by itself?

I'm going to be honest, I have no clue, but it is affecting something as removing it fixes the suspend issue
However, I'm going to keep testing to see if the graphics issue comes back before declaring this as a full fix.

jchau_xyz · 2023-04-06 04:46:35

Unfortunately, after removing the ppfeaturemask line,
my system crashed.

Here are the logs of when it happened:
http://0x0.st/HXKN.txt

seth · 2023-04-06 06:41:16

Apr 06 00:42:08 jchau kernel: amdgpu 0000:63:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:40 vmid:4 pasid:32779, for process RDD Process pid 31802 thread firefox-bi:cs0 pid 31888)
Apr 06 00:42:08 jchau kernel: amdgpu 0000:63:00.0: amdgpu:   in page starting at address 0x0000800103280000 from client 0x1b (UTCL2)
Apr 06 00:42:08 jchau kernel: amdgpu 0000:63:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00441051
Apr 06 00:42:08 jchau kernel: amdgpu 0000:63:00.0: amdgpu:          Faulty UTCL2 client ID: TCP (0x8)
Apr 06 00:42:08 jchau kernel: amdgpu 0000:63:00.0: amdgpu:          MORE_FAULTS: 0x1
Apr 06 00:42:08 jchau kernel: amdgpu 0000:63:00.0: amdgpu:          WALKER_ERROR: 0x0
Apr 06 00:42:08 jchau kernel: amdgpu 0000:63:00.0: amdgpu:          PERMISSION_FAULTS: 0x5
Apr 06 00:42:08 jchau kernel: amdgpu 0000:63:00.0: amdgpu:          MAPPING_ERROR: 0x0
Apr 06 00:42:08 jchau kernel: amdgpu 0000:63:00.0: amdgpu:          RW: 0x1

This increasingly shows up - is it (for you) deterministically bound to the ppfeaturemask?
Did you try the isolated removal of pcie_aspm=force wrt the S3 situation?

https://bbs.archlinux.org/viewtopic.php?id=282273
https://bbs.archlinux.org/viewtopic.php?id=282511
https://bbs.archlinux.org/viewtopic.php?id=284076
https://bbs.archlinux.org/viewtopic.php?id=284895

jchau_xyz · 2023-04-06 07:19:29

This increasingly shows up - is it (for you) deterministically bound to the ppfeaturemask?

I am not sure, but I can say that after setting it up, I had 2-3 weeks without that issue, and within a couple hours of using the system without it, the crash happened.

I am not testing with these following parameters:

amdgpu.ppfeaturemask=0xffff7fff
amdgpu.noretry=0
amdgpu.gpu_recovery=1

Did you try the isolated removal of pcie_aspm=force wrt the S3 situation?

Yeah, that's the first thing I tried, and that made no difference.

Only after the removal of the initial ppfeaturemask fixed suspend independently from pcie_aspm=force. However, I decided to remove both of them because in the case that they were causing other instabilities.

Interestingly, the new ppfeaturemask (the one from the snippet right above) does not seem to affect suspend.

I'll update if these changes fix or improve the situation

jchau_xyz · 2023-04-07 09:07:50

I don't want to jinx it, but I've been using the system for about 15-17h normally (streaming videos, browsing, coding, ...) without any issues with the parameters listed above.

Suspend also seems to be working as expected.

I will edit this reply if ever I encounter any other issues in the following weeks.

EDIT:
still true as of today, have not crashed due to something related to these I believe.
I do experience a crash maybe once every 2 weeks, but it seems to be unrelated this and something I'll have to check out in the near future

Last edited by jchau_xyz (2023-04-22 02:59:54)

Arch Linux

#1 2023-03-21 15:01:13

[ABANDONED] Can't wake up from sleep

#2 2023-03-21 21:39:48

Re: [ABANDONED] Can't wake up from sleep

#3 2023-03-22 12:49:44

Re: [ABANDONED] Can't wake up from sleep

#4 2023-03-22 13:06:10

Re: [ABANDONED] Can't wake up from sleep

#5 2023-03-31 04:35:09

Re: [ABANDONED] Can't wake up from sleep

#6 2023-03-31 06:24:36

Re: [ABANDONED] Can't wake up from sleep

#7 2023-04-01 09:16:57

Re: [ABANDONED] Can't wake up from sleep

#8 2023-04-01 12:49:48

Re: [ABANDONED] Can't wake up from sleep

#9 2023-04-02 07:45:02

Re: [ABANDONED] Can't wake up from sleep

#10 2023-04-02 08:12:26

Re: [ABANDONED] Can't wake up from sleep

#11 2023-04-02 21:50:06

Re: [ABANDONED] Can't wake up from sleep

#12 2023-04-03 07:39:03

Re: [ABANDONED] Can't wake up from sleep

#13 2023-04-03 13:52:35

Re: [ABANDONED] Can't wake up from sleep

#14 2023-04-03 14:20:30

Re: [ABANDONED] Can't wake up from sleep

#15 2023-04-03 22:43:02

Re: [ABANDONED] Can't wake up from sleep

#16 2023-04-06 04:46:35

Re: [ABANDONED] Can't wake up from sleep

#17 2023-04-06 06:41:16

Re: [ABANDONED] Can't wake up from sleep

#18 2023-04-06 07:19:29

Re: [ABANDONED] Can't wake up from sleep

#19 2023-04-07 09:07:50

Re: [ABANDONED] Can't wake up from sleep

Board footer