Skip to contentSkip to conversation

New: hibernation seems to be broken after commit bb31fef0d03ed17d587b40e3458786be408fb9df on v6.19.0-rc4

Bug 220946 [email protected] 1 participant 9 messages Jan 5 – Jan 7

lore
Share & tools
lore

[email protected]

author
#0 lore raw

https://bugzilla.kernel.org/show_bug.cgi?id=220946

            Bug ID: 220946
           Summary: hibernation seems to be broken after commit
                    bb31fef0d03ed17d587b40e3458786be408fb9df on
                    v6.19.0-rc4
           Product: Power Management
           Version: 2.5
          Hardware: AMD
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P3
         Component: cpufreq
          Assignee: [email protected]
          Reporter: [email protected]
        Regression: No

Created attachment 309127
--> https://bugzilla.kernel.org/attachment.cgi?id=309127&action=edit
dmesg (successful hibernation after reverting commit, the cpufreq errors still appear)

Overview:
Hibernation appears to be broken after commit bb31fef0d03ed17d587b40e3458786be408fb9df - I built v6.19.0-rc4 (build info provided). My goal was to originally figure out amd_pstate errors in dmesg, while reverting the commit seemed to fix hibernation, it did not fix the error. Regardless, I think there is some relation, so I put it under cpufreq component.

Steps to Reproduce:
- Build kernel from v6.19.0-rc4 release tag. I did this at around 00:00 UTC on 05 January. Config attached. Install modules and all.
- Boot, and try to suspend. Will not work (displays were off, but my CPU fan and keyboard/mouse lights stayed on, meaning suspend did not complete.
- Force reboot. You can test on a backup kernel version (I used 6.17.0-8) and I did not experience issues.
- Revert the aforementioned commit.
- Rebuild, reinstall, reboot.
- Hibernation works as expected.

Hardware:
Motherboard: ASRock X570M Pro4
CPU: Ryzen 5 3600
GPU: Radeon RX 6700 XT

Signature · 4 lines
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.
author
#1 lore raw

[Bug 220946] hibernation seems to be broken after commit bb31fef0d03ed17d587b40e3458786be408fb9df on v6.19.0-rc4

https://bugzilla.kernel.org/show_bug.cgi?id=220946

--- Comment #1 from Joshua Peisach ([email protected]) --- Created attachment 309128
--> https://bugzilla.kernel.org/attachment.cgi?id=309128&action=edit
config file used to build

Signature · 4 lines
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.
author
#2 lore raw

[Bug 220946] hibernation seems to be broken after commit bb31fef0d03ed17d587b40e3458786be408fb9df on v6.19.0-rc4

In reply to [email protected]

https://bugzilla.kernel.org/show_bug.cgi?id=220946

--- Comment #2 from Mario Limonciello (AMD) ([email protected]) --- Two things that can help figure out what's going on.

1) Can you please boot the kernel with dynamic debugging enabled for amd-pstate and share the logs?
2) Can you please share an acpidump for your system

Also; I notice a really really old BIOS:

[    0.000000] DMI: To Be Filled By O.E.M. To Be Filled By O.E.M./X570M Pro4,

BIOS P3.10 10/26/2020

By chance anything newer you can apply?

Signature · 4 lines
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.
author
#3 lore raw

[Bug 220946] hibernation seems to be broken after commit bb31fef0d03ed17d587b40e3458786be408fb9df on v6.19.0-rc4

In reply to [email protected]

https://bugzilla.kernel.org/show_bug.cgi?id=220946

--- Comment #3 from Joshua Peisach ([email protected]) --- (In reply to Mario Limonciello (AMD) from comment #2)

Quoted 5 lines

Two things that can help figure out what's going on.

1) Can you please boot the kernel with dynamic debugging enabled for amd-pstate and share the logs?
2) Can you please share an acpidump for your system

Sure - when I get home this evening I'll do so. It will be at least 12 hours though.

Quoted 6 lines

Also; I notice a really really old BIOS:

[    0.000000] DMI: To Be Filled By O.E.M. To Be Filled By O.E.M./X570M

Pro4, BIOS P3.10 10/26/2020

By chance anything newer you can apply?

Yes, there are newer BIOS versions I can upgrade to.

Signature · 4 lines
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.
author
#4 lore raw

[Bug 220946] hibernation seems to be broken after commit bb31fef0d03ed17d587b40e3458786be408fb9df on v6.19.0-rc4

In reply to [email protected]

https://bugzilla.kernel.org/show_bug.cgi?id=220946

Joshua Peisach ([email protected]) changed:

What |Removed |Added
----------------------------------------------------------------------------

             Status|NEW                         |RESOLVED
         Resolution|---                         |UNREPRODUCIBLE

--- Comment #4 from Joshua Peisach ([email protected]) --- Annoyingly, the issues are now unreproducible. I swear I'm not crazy, but here are the new dmesg and acpidump. I did not upgrade BIOS yet, but it's working now even with the patch which I thought caused a regression.

Signature · 4 lines
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.
author
#5 lore raw

[Bug 220946] hibernation seems to be broken after commit bb31fef0d03ed17d587b40e3458786be408fb9df on v6.19.0-rc4

In reply to [email protected]

https://bugzilla.kernel.org/show_bug.cgi?id=220946

--- Comment #5 from Joshua Peisach ([email protected]) --- Created attachment 309134
--> https://bugzilla.kernel.org/attachment.cgi?id=309134&action=edit
acpidump

Signature · 4 lines
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.
author
#6 lore raw

[Bug 220946] hibernation seems to be broken after commit bb31fef0d03ed17d587b40e3458786be408fb9df on v6.19.0-rc4

In reply to [email protected]

https://bugzilla.kernel.org/show_bug.cgi?id=220946

--- Comment #6 from Joshua Peisach ([email protected]) --- Created attachment 309135
--> https://bugzilla.kernel.org/attachment.cgi?id=309135&action=edit
dmesg (successful hibernation WITH commit - proving bug is resolved/unreproducible)

Sorry about the trouble.

Signature · 4 lines
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.
author
#7 lore raw

[Bug 220946] hibernation seems to be broken after commit bb31fef0d03ed17d587b40e3458786be408fb9df on v6.19.0-rc4

In reply to [email protected]

https://bugzilla.kernel.org/show_bug.cgi?id=220946

--- Comment #7 from Mario Limonciello (AMD) ([email protected]) --- OK thanks for updating.

I suggest updating BIOS, hopefully amd-pstate will actually be compatible with your system rather than the failure to probe from invalid values from the BIOS:

Quoted 4 lines
[    0.515949] amd_pstate: min_freq(0) or max_freq(0) or nominal_freq(0)
value is incorrect
[    0.515950] amd_pstate: Failed to initialize CPU 11: -22
[    0.515965] amd_pstate: failed to register with return -19
Signature · 4 lines
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.
author
#8 lore raw

[Bug 220946] hibernation seems to be broken after commit bb31fef0d03ed17d587b40e3458786be408fb9df on v6.19.0-rc4

In reply to [email protected]

https://bugzilla.kernel.org/show_bug.cgi?id=220946

--- Comment #8 from Joshua Peisach ([email protected]) --- BIOS update fixed that.

Sorry for the false report. But it's good to have a fully up to date system before getting into kernel hacking...

Thanks again.

Signature · 4 lines
You may reply to this email to add a comment.

You are receiving this mail because:
You are the assignee for the bug.