[Bug 222138] New: hp-wmi: HPWMI_GM queries fail with AE_AML_OPERAND_TYPE on OmniBook 5 16-ag1xxx (8DF4); firmware WHCM never calls GMCF

[bugzilla-daemon@kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=222138

            Bug ID: 222138
           Summary: hp-wmi: HPWMI_GM queries fail with AE_AML_OPERAND_TYPE
                    on OmniBook 5 16-ag1xxx (8DF4); firmware WHCM never
                    calls GMCF
           Product: Drivers
           Version: 2.5
          Hardware: AMD
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P3
         Component: Platform_x86
          Assignee: drivers_platform_x86@kernel-bugs.osdl.org
          Reporter: lorenzocollevecchio17@gmail.com
        Regression: No

Created attachment 311136
  --> https://bugzilla.kernel.org/attachment.cgi?id=311136&action=edit
acpidump

Hardware: HP OmniBook 5 Laptop 16-ag1xxx (16-AG1002SL)
Board: HP 8DF4
BIOS: F.12, 04/10/2026
Kernels tested: 7.2.9-2-cachyos, 7.2.10-1-cachyos; also seen on
Linux Mint 22.3 (6.14.0-37-generic)

Symptoms (every boot):
 - \_SB.WMID.WHCM aborts with AE_AML_OPERAND_TYPE (three times at boot)
 - hp-wmi: "Failed to apply initial fan settings: -22"
 - \_SB.WMID.WQBC / WQBE abort with AE_AML_OPERAND_VALUE (these go
   away with modprobe.blacklist=hp_bioscfg)
 - hp hwmon device exposes only pwm1_enable; no pwm1, no fan inputs
 - acpi_fan works (fan1_input)
 - ACPI thermal: Invalid critical threshold (-274000), no valid trip points

Analysis (Krishna Chomal, on the platform-driver-x86 list): in WHCM,
the case for command 0x00020008 (HPWMI_GM) only does "RTCD = RTCC" and
never calls the dispatcher GMCF. Local2 is therefore never assigned,
and the later DerefOf (Local2 [Zero]) fails on an Integer, producing
AE_AML_OPERAND_TYPE. HPWMI_READ/WRITE commands still work.
This is a firmware bug and cannot be fixed in the driver; a driver
change to hide pwm1_enable/fan inputs when the fan count query fails
has been proposed.

Related: bug 222107 (board 8D9C), bug 221385.
acpidump and dmesg attached.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.