Greyed out modes on unspecified shift_mode

[miguelcr1982]

Hello,

First, thank you for all the work on mControlCenter—it’s really helpful!

I’m running the latest msi-ec driver on my MSI GF63. On a fresh boot, /sys/devices/platform/msi-ec/shift_mode shows unspecified, which causes mControlCenter to grey out all the shift modes. Manually setting a mode like this:

echo comfort | sudo tee /sys/devices/platform/msi-ec/shift_mode

makes all modes visible and fully functional within the app.

One additional observation: while the driver correctly exposes the EC modes (eco, comfort, sport, turbo), switching to ECO does not reduce the CPU frequency. My system stays around ~2.7 GHz, likely because there’s no automatic bridge between the EC modes and kernel power management (intel_pstate/cpufreq). I’m currently using TLP, which may also interfere with automatic power profile adjustments (e.g., Plasma’s power-profiles-daemon).

Could you recommend the best approach here? Should I adjust TLP, or is this an expected limitation on Linux until further integration is implemented?

Thank you again for your support!

[mutchiko]

what is your distro?

also, please uninstall tlp, use only PPD or tuneD (not both at the same time)

when you change shift mode with msi-ec, you talk directly to the motherboard and tell it to switch some power parameters (vrms, total power, thermal limiting, fan curve), and this is not related to what PPD does (changing cpu/gpu frequency profiles)

what you want to do is to manually match msi-ec shift mode (in MCC for example) with the setting in PPD to get the desired results.

also please don't directly compare windows to linux, cpu scaling isn't exactly the same, and linux reports average frequencies differently .

once we get this clear, we can start working on your issue.

[miguelcr1982]

Hi @mutchiko,

Thank you for your detailed response and guidance.

I’m using Kubuntu 25.04 with kernel 6.14.0-28-generic. I have already removed TLP and reverted to using only power-profiles-daemon (PPD).

I ran some stress tests while setting msi-ec shift mode to ECO and the corresponding PPD profile to power-saver. I observed that the CPU frequency still spikes to ~4 GHz under load and then stabilizes around 2.7 GHz, so the ECO mode does not effectively limit CPU frequencies as one might expect.

I realize now that my previous comparison with Windows was not accurate, sorry for that. With TLP active in battery mode, it did limit the CPU to ~2.5 GHz, which is why I initially assumed the ECO mode on Linux should behave similarly.

Please let me know if there’s any other information I should provide or specific tests I can run to help diagnose the interaction between msi-ec and PPD.

UPDATE
I disabled intel_pstate via the GRUB kernel parameter and now the CPU frequencies behave much better with PPD profiles. ECO, balanced, and performance profiles reflect the expected frequency limits.

Regarding MSI-EC, I don’t see any unusual behavior at this point, but it doesn’t directly affect CPU frequency scaling.

Thank you again for your support!

[mutchiko]

I observed that the CPU frequency still spikes to ~4 GHz

test with the charger disconnected and intel-pstate activated, PPD is less configurable but kind of smart about small details.

Please let me know if there’s any other information I should provide

there isn't much you can provide, we already know how things work, just keep in mind that your laptop EC firmware is kind of old, so the real effects of shift mode might be different than today's msi laptops.

i will try to make PPD automatically change msi-ec shift mode in the future, either through MCC (dbus) or by exposing correct sysfs files in acpi power profiles.

as for cpu scaling, i highly recommend you use the pstate driver (for both amd and intel cpus), it gives the system many more freq steps rather than the built in kernel driver.

4Ghz means the cpu is boosting, i recommend looking at the arch wiki, since you have a modern intel cpu

https://wiki.archlinux.org/title/CPU_frequency_scaling

also look try cpupower frequency-info it has some interesting info to look at:

  driver: amd-pstate-epp
  CPUs which run at the same hardware frequency: 8
  CPUs which need to have their frequency coordinated by software: 8
  energy performance preference: power
  hardware limits: 403 MHz - 3.20 GHz
  available cpufreq governors: performance powersave
  current policy: frequency should be within 403 MHz and 3.20 GHz.
                  The governor "powersave" may decide which speed to use
                  within this range.
  current CPU frequency: 1.92 GHz (asserted by call to kernel)
  boost state support:
    Supported: yes
    Active: no
  amd-pstate limits:
    Highest Performance: 166. Maximum Frequency: 4.47 GHz.
    Nominal Performance: 119. Nominal Frequency: 3.20 GHz.
    Lowest Non-linear Performance: 41. Lowest Non-linear Frequency: 1.10 GHz.
    Lowest Performance: 15. Lowest Frequency: 400 MHz.
    Preferred Core Support: 1. Preferred Core Ranking: 196.

[miguelcr1982]

Hi @mutchiko,

I just wanted to share an update: I created a systemd service to set the MSI Shift mode to Comfort at startup, which solved the issue of MCC not recognizing the modes.

sudo nano /etc/systemd/system/msi-shift.service

[Unit]
Description=Set MSI Shift Mode to eco
After=multi-user.target

[Service]
Type=oneshot
ExecStartPre=/bin/bash -c 'until [ -e /sys/devices/platform/msi-ec/shift_mode ]; do sleep 1; done'
ExecStart=/usr/bin/bash -c 'echo comfort > /sys/devices/platform/msi-ec/shift_mode'
user=root

[Install]
WantedBy=multi-user.target

I also tried running the laptop without intel_pstate, using only PPD profiles. Interestingly, CPU scaling now behaves well on battery, here’s the current output from cpupower frequency-info:

analyzing CPU 10:
driver: acpi-cpufreq
CPUs which run at the same hardware frequency: 10
CPUs which need to have their frequency coordinated by software: 10
maximum transition latency: 10.0 us
hardware limits: 800 MHz - 2.20 GHz
available frequency steps: 2.20 GHz, 2.20 GHz, 2.10 GHz, 2.00 GHz, 1.90 GHz, 1.80 GHz, 1.70 GHz, 1.60 GHz, 1.50 GHz, 1.40 GHz, 1.30 GHz, 1.20 GHz, 1.10 GHz, 1000 MHz, 900 MHz, 800 MHz
available cpufreq governors: conservative ondemand userspace powersave performance schedutil
current policy: frequency should be within 800 MHz and 2.20 GHz.
The governor "ondemand" may decide which speed to use
within this range.
current CPU frequency: 2.20 GHz (asserted by call to kernel)
boost state support:
Supported: yes
Active: yes

Battery life seems to have improved, roughly an extra hour compared to before. It’s a bit surprising that PPD performs better without intel_pstate.

Thanks again for your guidance and help!

[mutchiko]

good to hear, pstate drivers also rely on a good implementation in UEFI, sometimes its better to have it off if its not done well, which is the case in older laptops unfortunately.

now lets go back to the main issue, MCC shouldn't show you gray performance modes IF you have the msi-ec installed, so you shouldn't even need that systemd service in the first. i will provide test builds for you to try

[miguelcr1982]

Hi @mutchiko,

Thanks for the clarification! Makes sense regarding the P-State driver, it's curious that my i7-8750H doesn’t behave well with intel_pstate, considering it’s not that old.

Looking forward to testing the builds you’ll provide for MCC.

Best regards,