Re: [BUG] AMD PMC causes system hang on suspend (HP EliteBook X G1a, Strix Point)

[Mario Limonciello <mario.limonciello@amd.com>]

On 12/8/2025 1:22 PM, CX MAIL wrote:
> I have collected more logs with amd-debug-tools and created this issue on drm/freedesktop
> https://gitlab.freedesktop.org/drm/amd/-/issues/4767
> 

Thanks, will follow up with you there.

> I was able to run the script with --force despite the warnings below. I get the same symptom that system hangs before suspend entry.
> 
> Thanks,
> CX
> 
> ```
> Debugging script for s2idle on AMD systems
> 💻 HP HP EliteBook X G1a 14 inch Notebook Next Gen AI PC (103C_5336AN HP EliteBook)
> 🐧 Bazzite
> 🐧 Kernel 6.17.7-ba19.fc43.x86_64
> 🔋 Battery BAT0 (Hewlett-Packard Primary) is operating at 98.37% of design
> Checking prerequisites for s2idle
> ✅ Logs are provided via systemd
> ✅ AMD Ryzen AI 9 HX PRO 375 w/ Radeon 890M (family 1a model 24)
> ✅ ASPM policy set to 'default'
> ✅ SMT enabled
> ✅ LPS0 _DSM enabled
> ✅ ACPI FADT supports Low-power S0 idle
> ✅ PMC driver `amd_pmc` loaded (Program 11 Firmware 93.10.0)
> ✅ USB3 driver `xhci_hcd` bound to 0000:c3:00.4
> ✅ USB3 driver `xhci_hcd` bound to 0000:c5:00.0
> ✅ USB3 driver `xhci_hcd` bound to 0000:c5:00.3
> ✅ USB3 driver `xhci_hcd` bound to 0000:c5:00.4
> ✅ USB4 driver `thunderbolt` bound to 0000:c5:00.5
> ✅ USB4 driver `thunderbolt` bound to 0000:c5:00.6
> ✅ WLAN driver `mt7925e` bound to 0000:c2:00.0
> ✅ GPU driver `amdgpu` bound to 0000:c3:00.0
> ✅ System is configured for s2idle
> ✅ GPIO driver `pinctrl_amd` available
> ✅ PC6 and CC6 states are enabled
> ❌ IOMMU is misconfigured: missing MSFT0201 ACPI device
> ✅ RTC driver `rtc_cmos` configured to use CMOS alarm
> ❌ Kernel is tainted: 64
> Your system does not meet s2idle prerequisites!
> Explanations for your system
> 🚦 Device MSFT0201 missing from ACPI tables
>          The ACPI device {device} is required for suspend to work when the IOMMU is enabled.
>          Please check your BIOS settings and if configured correctly, report a bug to your system vendor.
> 
> For more information on this failure see:
>          https://gitlab.freedesktop.org/drm/amd/-/issues/3738#note_2667140
> 🚦 Kernel is tainted
>          A tainted kernel may exhibit unpredictable bugs that are difficult for this script to characterize.
>          If this is intended behavior run the tool with --force.
> 
> For more information on this failure see:
>          https://gitlab.freedesktop.org/drm/amd/-/issues/3089
> ```
> 
> 
> Sent with Proton Mail secure email.
> 
> On Wednesday, December 3rd, 2025 at 1:56 PM, Mario Limonciello <mario.limonciello@amd.com> wrote:
> 
>> On 12/3/25 11:28 AM, CX MAIL wrote:
>>
>>> Hello,
>>>
>>> I'm reporting a critical suspend issue on HP EliteBook X G1a with AMD Ryzen
>>> AI 9 HX 375 (Strix Point). The system completely hangs when entering s2idle
>>> with the AMD PMC driver loaded.
>>>
>>> Hardware:
>>> - Model: HP EliteBook X G1a 14 inch (Model 8D08)
>>> - CPU: AMD Ryzen AI 9 HX 375 (Strix Point)
>>> - GPU: Radeon 890M (integrated)
>>> - BIOS: X88 Ver. 01.03.11 (2025-09-03)
>>>
>>> Kernel: 6.17.7-ba19.fc43.x86_64 (Fedora-based)
>>> Current cmdline: bluetooth.disable_ertm=1 amdgpu.gpu_recovery=1 pcie_aspm=off
>>>
>>> Problem:
>>> When executing "systemctl suspend", the system enters s2idle but immediately
>>> hangs. The last kernel message is "PM: suspend entry (s2idle)" with no
>>> further output. System becomes completely unresponsive - keyboard, touchpad,
>>> and power button (short press) have no effect. Only hard power-off works.
>>>
>>> BIOS configuration:
>>> - APMF enabled
>>> - Modern Standby (s2idle only, S3 not supported by firmware)
>>> - /sys/power/mem_sleep: [s2idle]
>>>
>>> AMD PMC device present: /sys/bus/platform/devices/AMDI000A:00/
>>>
>>> Workaround:
>>> Adding "initcall_blacklist=amd_pmc_driver_init" to kernel cmdline prevents
>>> the AMD PMC driver from loading. With this workaround, s2idle works perfectly
>>> and the system resumes normally from suspend.
>>>
>>> Verification with workaround:
>>> - lsmod | grep amd_pmc: (no output - driver not loaded)
>>> - Suspend/resume cycle: successful
>>>
>>> Logs attached:
>>> 1. system-info.txt - detailed system information
>>> 2. suspend-fail.log - Journal from failed suspend attempt (hangs after
>>> "PM: suspend entry")
>>> 3. suspend-success.log - Journal from successful suspend/resume with
>>> workaround applied
>>>
>>> The failure log shows the system stops all output immediately after entering
>>> s2idle, suggesting the AMD PMC driver causes a hard hang during early
>>> platform suspend initialization.
>>>
>>> Please let me know if you need any additional debug information or would like
>>> me to test patches.
>>>
>>> Best regards,
>>> CX
>>
>>
>> Hi,
>>
>> Please use amd-s2idle from https://pypi.org/project/amd-debug-tools/ to
>> generate a report. It may root cause it for you. If it doesn't, please
>> attach it to a bugzilla or drm/freedesktop bug report and CC me.
>>
>> Thanks,