https://bugzilla.kernel.org/show_bug.cgi?id=218750
Bug ID: 218750
Summary: Regression: S0ix sleep broken on 13th Gen Lenovo
laptops
Product: ACPI
Version: 2.5
Hardware: Intel
OS: Linux
Status: NEW
Severity: blocking
Priority: P3
Component: Power-Sleep-Wake
Assignee: [email protected]
Reporter: [email protected]
Regression: No
Hi,
S0ix sleep is broken on X1 Carbon G11 and X1 Yoga G8 platforms (Raptorlake
CPU).
Still checking if other platforms from that generation are impacted.
Bisect is pointing at:
[banther@x1c11 linux]$ git bisect bad
073237281a508ac80ec025872ad7de50cfb5a28a is the first bad commit
commit 073237281a508ac80ec025872ad7de50cfb5a28a
Author: Rafael J. Wysocki <[email protected]>
Date: Tue Feb 6 20:33:45 2024 +0100
ACPI: PM: s2idle: Enable Low-Power S0 Idle MSFT UUID for non-AMD systems
Systems based on Intel platforms that use the MSFT UUID for Low-Power S0
Idle (LPS0) have started to ship, so allow the kernel to use the MSFT
UUID in the non-AMD case too, but in that case make it avoid evaluating
the same _DSM function for two different UUIDs and prioritize the MSFT
one.
While at it, combine two MSFT _DSM function mask checks in
acpi_s2idle_restore_early() so as to make it reflect the
acpi_s2idle_prepare_late() flow more closely and adjust the
Modern Standby entry and exit comments slightly.
Non-AMD systems that do not support MSFT UUID for Low-power S0 Idle are
not expected to be affected by this change in any way.
Signed-off-by: Rafael J. Wysocki <[email protected]>
Reviewed-by: Mario Limonciello <[email protected]>
drivers/acpi/x86/s2idle.c | 37 +++++++++++++++++++++++++++----------
1 file changed, 27 insertions(+), 10 deletions(-)
I did some sanity checking and confirmed that if I force
lps0_dsm_func_mask_microsoft = -EINVAL that S0ix works again.
Can this change be reverted while we figure out how to fix this please?
I'm happy to help with any testing, and can pull in Lenovo FW teams if FW level
fixes are needed (with a note that those will take time, and I think this has
to be reverted from 6.9 until the issue is root caused and understood).
I'll note that I'm not seeing any problems on the Meteorlake system I have - so
it may be platform or CPU version specific. I've not had time to confirm
Thanks to Hans de Goede for notifying me of the issue in the 6.9-rc builds.
Mark
--
You may reply to this email to add a comment.
You are receiving this mail because:
You are watching the assignee of the bug.
_______________________________________________
acpi-bugzilla mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/acpi-bugzilla
