On Wed, Nov 11, 2026 at 6:34 AM Angela Arch Linux <[email protected]>
wrote:

> Hi Timothy,
>
> I appreciate your enthusiasm, and I'm glad to see you are leveraging AI to
> find a solution. However, I must kindly point out a few critical
> inaccuracies in the AI's response.
>
> 1. The AI claims you are running "Linux Kernel 7.3.3 (Debian 14
> Trixie/Testing)". This is factually incorrect. You are on Debian 13 Stable,
> running Kernel 6.12.116. The AI is hallucinating your environment.
>

I upgraded to sid to try the 7.3.3 kernel. I was on Debian Forky Testing
before.

I also installed sudo apt install linux-cpupower. I tested it and it
provides automatic speed stepping for acpi-cpufreq. Like with old Intel it
just works.

sudo cpupower frequency-info

boost state support:
 Supported: yes
 Active: yes
 Boost States: 0
 Total States: 3
 Pstate-P0:  2000MHz
 Pstate-P1:  1700MHz
 Pstate-P2:  1400MHz

2. More importantly, the AI's theoretical explanation completely ignores
> your own empirical evidence: `modinfo: ERROR: Module amd_pstate not found.`
>

You are correct that the module still does not show up. I emailed the
kernel distribution list about it. I solved the hardware issues now I just
need kernel modules. I may compile from source just to see if it works.

Modules
amd-pstate-epp (if you chose active)
amd-pstate (if you chose passive or guided)

Thanks again for all of your help.

Tim


> You cannot load a module that doesn't exist. Adding
> `amd_pstate.shared_mem=1` to your GRUB will not magically compile the
> missing module. It will simply fail silently and leave you stuck on
> `acpi-cpufreq`.
>
> If you wish to test those parameters, you must first install a newer
> kernel (e.g., from Backports) that actually includes the module. Otherwise,
> I must kindly suggest that you not waste your time rebooting.
>
> AI is a great assistant, but in Linux troubleshooting, your own terminal
> output is the ultimate truth.
>
> Best regards,
> Angela Arch Linux
>
>
> ---- 回复的原邮件 ----
> 发件人 Timothy M Butterworth<[email protected]>
> <[email protected]>
> 日期 2026年11月11日 19:10
> 收件人 Angela Arch Linux<[email protected]> <[email protected]>
> 抄送至 heck-schneider<[email protected]>
> <[email protected]>、debian-user<[email protected]>
> <[email protected]>
> 主题 Re: Re: Re: Re: Re: Power Profiles not supported
>
>
> On Wed, Nov 11, 2026 at 1:03 AM Angela Arch Linux <[email protected]>
> wrote:
>
>> Hi Timothy,
>>
>> Thanks for the precise output. Now we have the definitive answer.
>>
>> Since your CPU has the `cppc` flag, the hardware is perfectly capable.
>> However, the `modinfo` output reveals the real issue: `Module amd_pstate
>> not found`.
>>
>> This means the Debian kernel you are running was compiled WITHOUT the
>> `amd_pstate` module. No matter what you put in GRUB, the kernel simply
>> doesn't have the code to load it. This also explains why Fedora works—their
>> kernel has it built-in.
>>
>> There are only a few options left:
>> 1. Use the `linux-image-amd64` from backports (if available), which might
>> have a newer kernel config.
>> 2. Or, accept that Debian's default kernel won't support it, and
>> `acpi-cpufreq` is the only driver you can use. In this case, the
>> "Performance" profile simply won't appear in `power-profiles-daemon`.
>>
>> You can safely remove `amd_pstate=active` from your GRUB to avoid any
>> unnecessary conflicts, and just enjoy the two profiles you have.
>>
>> It's a kernel packaging limitation, not your fault.
>>
>> Hope this helps.
>>
>
> I found the problem. Thank you AI!
> For your AMD Ryzen 7 4700U (a Zen 2 Renoir laptop architecture) running on
> Linux Kernel 7.3.3 (Debian 14 Trixie/Testing), you must include an explicit
> memory management flag to bypass the legacy acpi-cpufreq driver.
> Because Zen 2 mobile processors do not have dedicated hardware MSR
> registers for frequency scaling, they rely entirely on an ACPI Shared
> Memory interface. By default, the Linux kernel disables P-State for these
> chips unless you manually authorize shared memory access. On platforms like
> the  forums, users have noted that mobile processors like the Ryzen 7 4700U
> will completely fail to load amd-pstate without specific arguments.
> amd_pstate=active amd_pstate.shared_mem=1
>
>
>
>>
>>
>> ---- 回复的原邮件 ----
>> 发件人 Timothy M Butterworth<[email protected]>
>> <[email protected]>
>> 日期 2026年11月11日 13:59
>> 收件人 Angela Arch Linux<[email protected]> <[email protected]>
>> 抄送至 heck-schneider<[email protected]>
>> <[email protected]>、debian-user<[email protected]>
>> <[email protected]>
>> 主题 Re: Re: Re: Re: Power Profiles not supported
>>
>>
>> On Wed, Nov 11, 2026 at 12:53 AM Angela Arch Linux <[email protected]>
>> wrote:
>>
>>> Hi Timothy,
>>>
>>> First, don't worry about the `bash: acpi-cpufreq: command not found`
>>> error—you accidentally typed the *output* of `cat` as a command. The actual
>>> output was `acpi-cpufreq`, which confirms the driver is still the legacy
>>> one.
>>>
>>> Now, about why `amd_pstate=active` didn't take effect. This is likely a
>>> deeper hardware/firmware issue:
>>>
>>> 1. Check if your CPU actually supports `amd_pstate`. Run:
>>>   `grep -i cppc /proc/cpuinfo`
>>>   If there's no `cppc` flag, your CPU or BIOS doesn't expose the
>>> required ACPI CPPC interface, and `amd_pstate` will refuse to load.
>>>
>>
>> cppc is present.
>>
>>
>>> 2. Check your BIOS settings. Look for options like "CPPC", "ACPI CPPC",
>>> or "Global C-state Control" and make sure they are enabled.
>>>
>>> 3. Check if the `amd_pstate` module is even available:
>>>   `modinfo amd_pstate`
>>>   If it says "not found", your Debian kernel might not have it compiled
>>> in for your CPU family.
>>>
>>
>> sudo modinfo amd_pstate
>> modinfo: ERROR: Module amd_pstate not found.
>>
>>
>>>
>>> If none of these work, your HP Pavilion might be too old for
>>> `amd_pstate`, and `acpi-cpufreq` is the only driver it can use. In that
>>> case, the "Performance" profile might simply not be supported by your
>>> hardware on Debian.
>>>
>>> Let me know what `grep -i cppc /proc/cpuinfo` shows.
>>>
>>> Hope this helps.
>>>
>>>
>>> ---- 回复的原邮件 ----
>>> 发件人 Timothy M Butterworth<[email protected]>
>>> <[email protected]>
>>> 日期 2026年11月11日 13:50
>>> 收件人 Angela Arch Linux<[email protected]> <[email protected]>
>>> 抄送至 heck-schneider<[email protected]>
>>> <[email protected]>、debian-user<[email protected]>
>>> <[email protected]>
>>> 主题 Re: Re: Re: Power Profiles not supported
>>>
>>>
>>> On Wed, Nov 11, 2026 at 12:24 AM Angela Arch Linux <[email protected]>
>>> wrote:
>>>
>>>> To add the parameter, just run this command in your terminal (no need
>>>> to use Vim):
>>>>
>>>> sudo sed -i
>>>> 's/GRUB_CMDLINE_LINUX_DEFAULT="/GRUB_CMDLINE_LINUX_DEFAULT="amd_pstate=active
>>>> /' /etc/default/grub
>>>> sudo update-grub
>>>> sudo reboot
>>>>
>>>
>>> I modified the GRUB command line as shown:
>>>
>>> GRUB_CMDLINE_LINUX_DEFAULT="amd_pstate=active quiet splash
>>>
>>> I updated GRUB and I still load acpi-cpufreq
>>>
>>> `cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver`
>>> bash: acpi-cpufreq: command not found
>>>
>>>
>>>
>>>
>>>>
>>>> ---- 回复的原邮件 ----
>>>> 发件人 Timothy M Butterworth<[email protected]>
>>>> <[email protected]>
>>>> 日期 2026年11月11日 13:15
>>>> 收件人 Angela Arch Linux<[email protected]> <[email protected]>
>>>> 抄送至 heck-schneider<[email protected]>
>>>> <[email protected]>、debian-user<[email protected]>
>>>> <[email protected]>
>>>> 主题 Re: Re: Power Profiles not supported
>>>>
>>>>
>>>> On Tue, Nov 10, 2026 at 10:49 PM Angela Arch Linux <[email protected]>
>>>> wrote:
>>>>
>>>>> Hi,
>>>>>
>>>>> Since you mentioned the same kernel works on Fedora, it's likely a
>>>>> missing
>>>>> power management daemon or firmware dependency in Debian.
>>>>>
>>>>> Try installing "thermald" (Intel) or "amd64-microcode" (AMD).
>>>>> Also, check if your CPU requires the specific P-State driver:
>>>>>
>>>>
>>>> I have:
>>>> amd64-microcode/testing,now 3.20260923.1 amd64 [installed]
>>>>  Platform firmware and microcode for AMD CPUs and SoCs
>>>>
>>>> and
>>>> intel-microcode/testing,now 3.20260925.1 amd64 [installed]
>>>>  Processor microcode firmware for Intel CPUs
>>>>
>>>> My processor is actually AMD.
>>>>
>>>> For PState I have:
>>>> librte-power-amd-pstate26/testing,now 25.11.4-1 amd64 [installed]
>>>>  Data Plane Development Kit (librte-power-amd-pstate runtime library)
>>>>
>>>> sudo apt install thermald intel-microcode
>>>>> sudo systemctl enable --now thermald
>>>>>
>>>>
>>>> I installed thermald but it is not supported.
>>>> Nov 11 00:04:54 hp-pavilion.home.arpa systemd[1]: Starting
>>>> thermald.service - Thermal Daemon Service...
>>>> Nov 11 00:04:54 hp-pavilion.home.arpa thermald[33197]: Unsupported cpu
>>>> model or platform
>>>>
>>>> After rebooting, check if "Performance" appears in "powerprofilesctl
>>>>> list".
>>>>>
>>>>> Another possibility: Fedora might be loading a specific firmware that
>>>>> Debian
>>>>> keeps in "non-free-firmware". Check "dmesg | grep -i power" to see if
>>>>> the CPU is complaining about missing ACPI or thermal tables.
>>>>>
>>>>
>>>> sudo dmesg | grep -i power
>>>> [    0.246664] thermal_sys: Registered thermal governor
>>>> 'power_allocator'
>>>> [    0.418642] ACPI: \_SB_.PCI0.GPP2.DEV0.PWRE: New power resource
>>>> [    0.419128] ACPI: \_SB_.PCI0.GP17.XHC0.P0U0: New power resource
>>>> [    0.419172] ACPI: \_SB_.PCI0.GP17.XHC0.P3U0: New power resource
>>>> [    0.419434] ACPI: \_SB_.PCI0.GP17.XHC1.P0U1: New power resource
>>>> [    0.419477] ACPI: \_SB_.PCI0.GP17.XHC1.P3U1: New power resource
>>>> [    0.419741] ACPI: \_SB_.P0S0: New power resource
>>>> [    0.419757] ACPI: \_SB_.P3S0: New power resource
>>>> [    0.419802] ACPI: \_SB_.P0S1: New power resource
>>>> [    0.419817] ACPI: \_SB_.P3S1: New power resource
>>>> [    0.422872] ACPI: \_SB_.PRWL: New power resource
>>>> [    0.429670] Low-power S0 idle used by default for system suspend
>>>> [    0.448425] pci 0000:04:00.1: D0 power state depends on 0000:04:00.0
>>>> [    0.448434] pci 0000:04:00.3: extending delay after power-on from
>>>> D3hot to 20 msec
>>>> [    0.448579] pci 0000:04:00.4: extending delay after power-on from
>>>> D3hot to 20 msec
>>>> [    0.958457] PM: genpd: Disabling unused power domains
>>>> [    1.122915] input: Power Button as
>>>> /devices/platform/PNP0C0C:00/input/input1
>>>> [    1.124077] ACPI: button: Power Button [PWRB]
>>>> [    6.458155] sr 1:0:0:0: Power-on or device reset occurred
>>>> [   24.492652] wlo1: Limiting TX power to 23 (23 - 0) dBm as advertised
>>>> by 18:a5:ff:1c:da:27
>>>>
>>>> The only thing that stood out was:
>>>> [    0.958457] PM: genpd: Disabling unused power domains
>>>>
>>>> Tim
>>>>
>>>>
>>>>> Hope this helps.
>>>>>
>>>>>
>>>>> ---- 回复的原邮件 ----
>>>>> 发件人 Timothy M Butterworth<[email protected]>
>>>>> <[email protected]>
>>>>> 日期 2026年11月11日 08:50
>>>>> 收件人 derlyne<[email protected]> <[email protected]>
>>>>> 抄送至 Debian User<[email protected]>
>>>>> <[email protected]>
>>>>> 主题 Re: Power Profiles not supported
>>>>>
>>>>>
>>>>> On Tue, Nov 10, 2026 at 7:26 AM derlyne <[email protected]>
>>>>> wrote:
>>>>>
>>>>>> The package to install to manage power profiles is
>>>>>> `power-profiles-daemon`. It provides the system service and the
>>>>>> `powerprofilesctl` command; the required dependencies are installed
>>>>>> automatically.
>>>>>>
>>>>>> sudo apt install power-profiles-daemon
>>>>>> sudo systemctl enable --now power-profiles-daemon
>>>>>>
>>>>>
>>>>> I already have this installed and running.
>>>>>
>>>>> To display and change profiles graphically, the interface depends on
>>>>>> your desktop environment: GNOME and KDE Plasma include this functionality
>>>>>> in their settings or power-management tools. You generally don’t need to
>>>>>> install a separate package if those components are already installed.
>>>>>>
>>>>>
>>>>> I have KDE.
>>>>>
>>>>>
>>>>>>
>>>>>> Check the detected profiles with:
>>>>>>
>>>>>> powerprofilesctl list
>>>>>>
>>>>>>
>>>>>> “Power Saver” and “Balanced” are normally available; “Performance”
>>>>>> depends on hardware support.
>>>>>>
>>>>>
>>>>>  Performance shows up if I boot Fedora and I am able to switch the
>>>>> profiles from the system tray battery widget. Both Fedora and Forky have
>>>>> the 7.2 Linux kernel so it does not appear to be a kernel support issue.
>>>>>
>>>>>
>>>>>> (... gmail...)
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> ⢀⣴⠾⠻⢶⣦⠀
>>>>> ⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
>>>>> ⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org/
>>>>> ⠈⠳⣄⠀⠀
>>>>>
>>>>
>>>>
>>>> --
>>>> ⢀⣴⠾⠻⢶⣦⠀
>>>> ⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
>>>> ⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org/
>>>> ⠈⠳⣄⠀⠀
>>>>
>>>
>>>
>>> --
>>> ⢀⣴⠾⠻⢶⣦⠀
>>> ⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
>>> ⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org/
>>> ⠈⠳⣄⠀⠀
>>>
>>
>>
>> --
>> ⢀⣴⠾⠻⢶⣦⠀
>> ⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
>> ⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org/
>> ⠈⠳⣄⠀⠀
>>
>
>
> --
> ⢀⣴⠾⠻⢶⣦⠀
> ⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
> ⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org/
> ⠈⠳⣄⠀⠀
>


-- 
⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org/
⠈⠳⣄⠀⠀

Reply via email to