rtc: cmos: rtcwake broken on NUC8i7HVK

REGRESSION v6.1-rc1 Kalle Valo 3 participants 5 messages Oct 17, 2022 – Oct 18, 2022

Subsystems (2)

Real time clock (RTC) Regressions

lore

Share & tools

lore

Kalle Valo

author

2022-10-17 16:58 UTC #0 lore raw

Hi Rafael and Alexandre,

I upgraded my NUC8i7HVK test box to v6.1-rc1 and noticed that my suspend tests with ath11k were broken, the box never woke up from suspend automatically and I had to manually push the power button to trigger resume. This is the command I used:

sudo rtcwake -m mem -s 30

v6.0 works without problems and a bisect found this commit:

commit 4919d3eb2ec0ee364f7e3cf2d99646c1b224fae8

Author:     Rafael J. Wysocki <[email protected]>
AuthorDate: Wed Oct 12 20:07:01 2022 +0200
Commit:     Alexandre Belloni <[email protected]>
CommitDate: Thu Oct 13 23:27:52 2022 +0200

rtc: cmos: Fix event handler registration ordering issue

And indeed reverting that commit fixed the issue and suspend works as before. Here's some info about my hardware:

$ dmesg | grep rtc
[   10.884460] rtc_cmos 00:03: RTC can wake from S4
[   10.888439] rtc_cmos 00:03: registered as rtc0
[   10.888930] rtc_cmos 00:03: alarms up to one month, y3k, 242 bytes nvram

$ lshw
nuc2
    description: Desktop Computer
    product: NUC8i7HVK
    vendor: Intel(R) Client Systems
    version: J71485-504
[...]
     *-cpu
          description: CPU
          product: Intel(R) Core(TM) i7-8809G CPU @ 3.10GHz
[...]
     *-pnp00:03
          product: PnP device PNP0b00
          physical id: 4
          capabilities: pnp
          configuration: driver=rtc_cmos

Do let me know if more info is needed, I can also test patches. But next week I'm travelling and do not have access to the box.

Kalle

Thorsten Leemhuis

2022-10-18 06:33 UTC #1 lore raw

Hi, this is your Linux kernel regression tracker.

Quoted 10 lines · Kalle Valo

I upgraded my NUC8i7HVK test box to v6.1-rc1 and noticed that my suspend tests with ath11k were broken, the box never woke up from suspend automatically and I had to manually push the power button to trigger resume. This is the command I used:
sudo rtcwake -m mem -s 30
v6.0 works without problems and a bisect found this commit:

Quoted 7 lines

commit 4919d3eb2ec0ee364f7e3cf2d99646c1b224fae8
Author:     Rafael J. Wysocki <[email protected]>
AuthorDate: Wed Oct 12 20:07:01 2022 +0200
Commit:     Alexandre Belloni <[email protected]>
CommitDate: Thu Oct 13 23:27:52 2022 +0200
rtc: cmos: Fix event handler registration ordering issue

FWIW, Todd also reported this yesterday:
https://lore.kernel.org/all/[email protected]/
https://bugzilla.kernel.org/show_bug.cgi?id=216600

And Zhang Rui provided a patch in bugzilla that should fix this.

HTH, Ciao, Thorsten

Quoted 31 lines

And indeed reverting that commit fixed the issue and suspend works as before. Here's some info about my hardware:

$ dmesg | grep rtc
[   10.884460] rtc_cmos 00:03: RTC can wake from S4
[   10.888439] rtc_cmos 00:03: registered as rtc0
[   10.888930] rtc_cmos 00:03: alarms up to one month, y3k, 242 bytes nvram

$ lshw
nuc2
    description: Desktop Computer
    product: NUC8i7HVK
    vendor: Intel(R) Client Systems
    version: J71485-504
[...]
     *-cpu
          description: CPU
          product: Intel(R) Core(TM) i7-8809G CPU @ 3.10GHz
[...]
     *-pnp00:03
          product: PnP device PNP0b00
          physical id: 4
          capabilities: pnp
          configuration: driver=rtc_cmos

Do let me know if more info is needed, I can also test patches. But next week I'm travelling and do not have access to the box.

Kalle

Kalle Valo

author

2022-10-18 07:09 UTC #2 lore raw

Quoted 17 lines · Thorsten Leemhuis

Quoted 16 lines · Kalle Valo
I upgraded my NUC8i7HVK test box to v6.1-rc1 and noticed that my suspend tests with ath11k were broken, the box never woke up from suspend automatically and I had to manually push the power button to trigger resume. This is the command I used:
sudo rtcwake -m mem -s 30
v6.0 works without problems and a bisect found this commit:
commit 4919d3eb2ec0ee364f7e3cf2d99646c1b224fae8
Author:     Rafael J. Wysocki <[email protected]>
AuthorDate: Wed Oct 12 20:07:01 2022 +0200
Commit:     Alexandre Belloni <[email protected]>
CommitDate: Thu Oct 13 23:27:52 2022 +0200

Quoted 7 lines

Quoted 1 line
rtc: cmos: Fix event handler registration ordering issue
FWIW, Todd also reported this yesterday:
https://lore.kernel.org/all/[email protected]/
https://bugzilla.kernel.org/show_bug.cgi?id=216600
And Zhang Rui provided a patch in bugzilla that should fix this.

Thanks, this is good info. Strangely enough Zhang's workaround in [1] did not fix the issue for me, rtcwake is still broken. Unfortunately no time right now to investigate further.

[1] https://bugzilla.kernel.org/show_bug.cgi?id=216600#c3

Signature · 3 lines

https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

Rafael J. Wysocki

2022-10-18 16:36 UTC #3 lore raw

Quoted 23 lines · Kalle Valo

Quoted 21 lines · Thorsten Leemhuis
Quoted 9 lines · Kalle Valo
I upgraded my NUC8i7HVK test box to v6.1-rc1 and noticed that my suspend tests with ath11k were broken, the box never woke up from suspend automatically and I had to manually push the power button to trigger resume. This is the command I used:
sudo rtcwake -m mem -s 30
v6.0 works without problems and a bisect found this commit:
Quoted 7 lines
commit 4919d3eb2ec0ee364f7e3cf2d99646c1b224fae8
Author:     Rafael J. Wysocki <[email protected]>
AuthorDate: Wed Oct 12 20:07:01 2022 +0200
Commit:     Alexandre Belloni <[email protected]>
CommitDate: Thu Oct 13 23:27:52 2022 +0200
rtc: cmos: Fix event handler registration ordering issue
FWIW, Todd also reported this yesterday:
https://lore.kernel.org/all/[email protected]/
https://bugzilla.kernel.org/show_bug.cgi?id=216600

Quoted 7 lines

Quoted 1 line
And Zhang Rui provided a patch in bugzilla that should fix this.
Thanks, this is good info. Strangely enough Zhang's workaround in [1] did not fix the issue for me, rtcwake is still broken. Unfortunately no time right now to investigate further.
[1] https://bugzilla.kernel.org/show_bug.cgi?id=216600#c3

Please try https://lore.kernel.org/linux-acpi/5887691.lOV4Wx5bFT@kreacher/

Kalle Valo

author Tested-by

2022-10-18 17:40 UTC #4 lore raw

Quoted 25 lines · Rafael J. Wysocki

Quoted 23 lines · Kalle Valo
Quoted 21 lines · Thorsten Leemhuis
Quoted 9 lines · Kalle Valo
I upgraded my NUC8i7HVK test box to v6.1-rc1 and noticed that my suspend tests with ath11k were broken, the box never woke up from suspend automatically and I had to manually push the power button to trigger resume. This is the command I used:
sudo rtcwake -m mem -s 30
v6.0 works without problems and a bisect found this commit:
Quoted 7 lines
commit 4919d3eb2ec0ee364f7e3cf2d99646c1b224fae8
Author:     Rafael J. Wysocki <[email protected]>
AuthorDate: Wed Oct 12 20:07:01 2022 +0200
Commit:     Alexandre Belloni <[email protected]>
CommitDate: Thu Oct 13 23:27:52 2022 +0200
rtc: cmos: Fix event handler registration ordering issue
FWIW, Todd also reported this yesterday:
https://lore.kernel.org/all/[email protected]/
https://bugzilla.kernel.org/show_bug.cgi?id=216600

Quoted 8 lines

Quoted 6 lines
Quoted 1 line
And Zhang Rui provided a patch in bugzilla that should fix this.
Thanks, this is good info. Strangely enough Zhang's workaround in [1] did not fix the issue for me, rtcwake is still broken. Unfortunately no time right now to investigate further.
[1] https://bugzilla.kernel.org/show_bug.cgi?id=216600#c3
Please try https://lore.kernel.org/linux-acpi/5887691.lOV4Wx5bFT@kreacher/

This works, thanks!

Tested-by Kalle Valo <[email protected]>

Signature · 3 lines

https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches