intel_pstate not working. CPU frequency fixed at 800MHz. No cpu frequency scaling.

Bug #1838141 reported by Srijan Kumar Sharma
22
This bug affects 4 people
Affects Status Importance Assigned to Milestone
linux (Ubuntu)
Fix Released
Undecided
Unassigned
Bionic
Confirmed
Undecided
Unassigned

Bug Description

intel_pstate not working. CPU frequency fixed at 800MHz.

I've tried to modify frequency using "cpufreq-set" and also by directly writing to files in "/sys/devices/system/cpu/cpu0/cpufreq/*" to no success.

I tried different versions of kernel, it started working on 4.17.0 but stopped working soon. These are the versions I tested.

5.2.3 - Not Working
5.1.0 - Not Working
4.19.60 - Not Working
4.18.20 - Not Working
4.17.19 - Not Working
4.17.10 - Not Working
4.17.5 - Not Working
4.17.3 - Not Working
4.17.2 - Working
4.17.1 - Working
4.17.0 - Working
4.16.18 - Not Working
4.15.0 - Not Working

Please let me know if anything else is needed.
thanks

ProblemType: Bug
DistroRelease: Ubuntu 18.04
Package: linux-image-4.15.0-55-generic 4.15.0-55.60
ProcVersionSignature: Ubuntu 4.15.0-55.60-generic 4.15.18
Uname: Linux 4.15.0-55-generic x86_64
ApportVersion: 2.20.9-0ubuntu7.7
Architecture: amd64
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: srijan 2219 F.... pulseaudio
CurrentDesktop: KDE
Date: Sat Jul 27 11:34:10 2019
InstallationDate: Installed on 2018-11-25 (243 days ago)
InstallationMedia: Ubuntu 18.04.1 LTS "Bionic Beaver" - Release amd64 (20180725)
Lsusb:
 Bus 002 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
 Bus 001 Device 003: ID 0bda:5621 Realtek Semiconductor Corp.
 Bus 001 Device 002: ID 046d:c534 Logitech, Inc. Unifying Receiver
 Bus 001 Device 004: ID 8087:0aaa Intel Corp.
 Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
MachineType: Acer Predator PH315-51
ProcFB: 0 inteldrmfb
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-4.15.0-55-generic root=UUID=a76bd009-8058-4e96-a8b5-a82b7fbe1ede ro acpi_backlight=vendor i915.modeset=1 i915.enable_guc=3 i915.enable_dc=2 i915.enable_fbc=1 nvme_core.default_ps_max_latency_us=18000000 pcie_aspm=force i915.enable_psr=2 i915.disable_power_well=1 nmi_watchdog=0
RelatedPackageVersions:
 linux-restricted-modules-4.15.0-55-generic N/A
 linux-backports-modules-4.15.0-55-generic N/A
 linux-firmware 1.173.10
SourcePackage: linux
UpgradeStatus: No upgrade log present (probably fresh install)
WifiSyslog:

dmi.bios.date: 12/05/2018
dmi.bios.vendor: Insyde Corp.
dmi.bios.version: V1.24
dmi.board.asset.tag: Type2 - Board Serial Number
dmi.board.name: Sienta_CFS
dmi.board.vendor: CFL
dmi.board.version: V1.24
dmi.chassis.type: 10
dmi.chassis.vendor: Acer
dmi.chassis.version: V1.24
dmi.modalias: dmi:bvnInsydeCorp.:bvrV1.24:bd12/05/2018:svnAcer:pnPredatorPH315-51:pvrV1.24:rvnCFL:rnSienta_CFS:rvrV1.24:cvnAcer:ct10:cvrV1.24:
dmi.product.family: Predator Helios 300
dmi.product.name: Predator PH315-51
dmi.product.version: V1.24
dmi.sys.vendor: Acer

Revision history for this message
Srijan Kumar Sharma (raja-srijan) wrote :
summary: - intel_pstate not working. CPU frequency fixed at 800MHz
+ intel_pstate not working. CPU frequency fixed at 800MHz. No cpu
+ frequency scaling.
Revision history for this message
Ubuntu Kernel Bot (ubuntu-kernel-bot) wrote : Status changed to Confirmed

This change was made by a bot.

Changed in linux (Ubuntu):
status: New → Confirmed
Revision history for this message
Anthony Wong (anthonywong) wrote :

Setting specific CPU frequencies via cpufreq-set is incompatible with the intel_pstate driver.
Is your problem about unable to fix the CPU to specific frequencies, or the CPU frequencies do not increase even you are under heavy workload?

Revision history for this message
Srijan Kumar Sharma (raja-srijan) wrote :

CPU frequency does not increase under heavy load.

Revision history for this message
Kai-Heng Feng (kaihengfeng) wrote :
Revision history for this message
Anthony Wong (anthonywong) wrote :

Does the problem go away if "intel_pstate=disable" is provided in kernel cmdline?
There were many accounts of CPU being stuck at 800MHz, and they were not caused by OS at all:
 https://forums.bit-tech.net/index.php?threads/solved-laptop-stuck-at-800mhz.338423/
 https://linustechtips.com/main/topic/531278-skylake-system-stuck-at-800mhz/
 https://www.dell.com/community/Laptops-General-Read-Only/Dell-XPS-9550-CPU-Multiplier-Stuck-at-8/td-p/4740654

Revision history for this message
Srijan Kumar Sharma (raja-srijan) wrote :

@kaihengfeng I'll test and post the results soon.

@anthonywong If I disable pstate with "intel_pstate=disable", frequency goes up to 1000Mhz max, also laptop runs very hot and battery lasts 30-60mins without pstate.

On Windows 10 CPU goes all the way to 4.1GHz so it's not a hardware issue. Linux kernel 4.17.3, 4.17.2, 4.17.1 are also working as expected (4.1GHz) without overheating and >3hrs battery. I would try 1st and 2nd link you posted but I don't have problem on win10. 3rd link is dell model specific but I've already tried that.

Revision history for this message
Srijan Kumar Sharma (raja-srijan) wrote :

@kaihengfeng I'm unable to boot with v5.3-rc2

description: updated
Revision history for this message
Kai-Heng Feng (kaihengfeng) wrote :

If v4.17 works and (presumably) v4.18-rc1 doesn't work, perform a bisection between theses two version to find the regression point.

Revision history for this message
Srijan Kumar Sharma (raja-srijan) wrote :

I upgraded to Ubuntu 19.04, and the issue didn't reproduce with default installed kernel 5.0.0-27-generic. Issue still reproduces with Ubuntu 18.04.

Should I close this issue?

Changed in linux (Ubuntu):
status: Confirmed → Fix Released
Revision history for this message
vinibali (vinibali) wrote :

Hello there,

I'd like to just note, that I had the same problem with 18.04 LTS HWE (5.0.0-23-generic) and an i5-8350U, but mine was clocking around 3.6GHz on each core.
Upgrade to 5.0.0-31-generic has resolved this issue with intel_pstate driver, but using the kernel parameter intel_pstate=disable was a usable workaround as well.

Regards

Revision history for this message
semidark (nthomaier) wrote :

I had the same Problem with the CPU in my Lenovo T470s. Intel(R) Core(TM) i5-6300U CPU @ 2.40GHz

i tried intel_pstate=disable which did fix the stuck @ 800 MHz Problem. But Managing CPU govenors ans frequencies was still kind of broken with Kernel 5.3.0-40.

So I deleted the intel_pstate kernel Parameter and checked the BIOS / EFI for the CPU management settings. After I disabled Intel Speed Stepping in the Bios, everything was working as expected in Linux. intel_pstate is still availible for the Kernel and the Management of Govenors and Frequencies is just working fine.

After a bit of reading up un intel_pstate i think I found another possible Solution or the core reason for This Problem.

I did not try it since my problem was solved via the Bios Setting, but it reads promissing:

https://www.kernel.org/doc/html/v5.3/admin-guide/pm/intel_pstate.html

--- snip
If the processor supports the HWP feature, it will be enabled during the processor initialization and cannot be disabled after that. It is possible to avoid enabling it by passing the intel_pstate=no_hwp argument to the kernel in the command line.

If the HWP feature has been enabled, intel_pstate relies on the processor to select P-states by itself, but still it can give hints to the processor's internal P-state selection logic. What those hints are depends on which P-state selection algorithm has been applied to the given policy (or to the CPU it corresponds to).

---snap

Revision history for this message
Launchpad Janitor (janitor) wrote :

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in linux (Ubuntu Bionic):
status: New → Confirmed
Revision history for this message
semidark (nthomaier) wrote :

After disabling the Speed Step in BIOS everythig seems to work fine... Until i Suspended my Notebook. After That the Frequency was stuck again @800MHz.

I'm now testing the Kernel Parameter intel_pstate=no_hwp

Revision history for this message
semidark (nthomaier) wrote :

After a few days of Testing two Ubuntu Installations on the same laptop, one with the intel_pstate=no_hwp and one without, this seems like a valid solution for me. Until now the Frequency never got stuck again @800MHz with the kernel parameter intel_pstate=no_hwp set.

Revision history for this message
semidark (nthomaier) wrote :

and again. it got stuck. even with the intel_pstate=no_hwp kernel parameter.Damn!

Revision history for this message
Ryan Gubele (5pau) wrote :

Thinkpad T460 with Intel(R) Core(TM) i7-6600U CPU @ 2.60GHz

Recently went Ubuntu 16.04 -> 18.04.

Using intel_pstate in default configuration.

HWE low latency kernel:

Linux machine 5.3.0-62-lowlatency #56~18.04.1-Ubuntu SMP PREEMPT Wed Jun 24 16:57:19 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux

For context, this laptop mostly moves from docking station to docking station. It lives most of it's life in "clamshell" mode and is only occasionally actually opened.

CPU gets "stuck" at 1.6GHz after clamshell->clamshell suspend/resume cycle.

Opening the lid (!!) resolves the issue. Doesn't come back when I close it again. :wtf:

I am at a loss for words. Definitely did not have this issue in 16.04.

I guess my hinge will be getting a workout until this is resolved.

Revision history for this message
Srijan Kumar Sharma (raja-srijan) wrote :

Personally I've disabled pstate driver from kernel command line. Events like suspend, lid close and restart completely changes the behavior.

While working on battery I sometimes will close the lid and when I open back CPU is at 4Ghz and battery gets drained in 30mins. Not to mention the heating.

*It simply doesn't work.*

The acpi driver is much better and battery life is great too. No lid/suspend issue so far.

I'm on Ubuntu 20.04 stock kernel:

Linux Predator-300 5.4.0-40-lowlatency #44-Ubuntu SMP PREEMPT Tue Jun 23 01:32:24 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux

Revision history for this message
Julian Andres Klode (juliank) wrote :

XPS 13 (Core i5-10210U) with performance governor idles (0.2% usage) at 4 GHz instead of 900 MHz.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.