https://bugzilla.kernel.org/show_bug.cgi?id=222291

            Bug ID: 222291
           Summary: Lenovo Yoga Slim 7i (Meteor Lake): FocalTech I2C
                    touchpad (FTCS0038) intermittently stalls/jumps; ACPI
                    BIOS Error in EC _Q18 (cannot resolve AWAC.HOR)
           Product: ACPI
           Version: 2.5
          Hardware: All
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P3
         Component: BIOS
          Assignee: [email protected]
          Reporter: [email protected]
        Regression: No

Created attachment 311505
  --> https://bugzilla.kernel.org/attachment.cgi?id=311505&action=edit
dmesg

Hardware: Lenovo 14IMH9, Intel Core Ultra 5 125H (Meteor Lake), BIOS NPCN35WW
(latest available).
OS: Fedora 45 Workstation, kernel 7.3.3-300.fc45.x86_64 (GNOME, Wayland).
Touchpad: FTCS0038:00 2808:0105, i2c-hid over i2c_designware (i2c-0), drivers
i2c_hid_acpi + hid_multitouch.

PROBLEM
The touchpad intermittently stops responding or reports erratic data: the
pointer stays still, clicks are delayed or lost, and a late press is then
interpreted as a click-and-drag (accidental text selection). libinput prints
"kernel bug: Touch jump detected and discarded" for this device. It happens
during normal use, on battery, without suspend and without any screenshot tool.
I first associated it with resume from suspend and screenshot tools, but it
also happens without them.

RECOVERY
Unbinding and rebinding only the device restores normal operation, so the I2C
controller seems fine:
echo i2c-FTCS0038:00 > /sys/bus/i2c/drivers/i2c_hid_acpi/unbind
echo i2c-FTCS0038:00 > /sys/bus/i2c/drivers/i2c_hid_acpi/bind

KERNEL LOG
No I2C timeout or HID error is logged when the failure happens. The only
related messages, repeated at irregular intervals of hours, are:
ACPI BIOS Error (bug): Could not resolve symbol [^^^^AWAC.HOR], AE_NOT_FOUND
(20260408/psargs-365)
ACPI Error: Aborting method \_SB.PC00.LPCB.Q_EC._Q18 due to previous error
(AE_NOT_FOUND) (20260408/psparse-543)
At boot there is also: ACPI Error: Aborting method \_SB.IETM.CHRG.PPSS due to
previous error (AE_NOT_FOUND).
At least once, the _Q18 message was the last line of dmesg right after a
failure, but I have not confirmed that the two are correlated.

ALREADY TRIED
- BIOS is up to date.
- Runtime PM is "unsupported" for the i2c device; setting power/control=on for
the controller (0000:00:15.0) did not prevent the failure.
- Same behavior on Fedora and Ubuntu, with GNOME and KDE. Pop!_OS with COSMIC
looked stable in short testing (not long-term).

QUESTIONS
1. Could the failing EC query _Q18 be related to the touchpad stalls?
2. Is there a known quirk or workaround for this device/firmware?
3. What additional debugging would help (i2c-hid debug, ACPI tracing, etc.)?

I can test patches. Attached: dmesg (Wi-Fi lines removed for privacy) and ACPI
tables (DSDT/SSDTs; license table removed).

-- 
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

Reply via email to