[Bug 220117] New: cpufreq performance governor doesn't correctly apply in combination with isolcpus

[Bug 220117] New: cpufreq performance governor doesn't correctly apply in combination with isolcpus

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220117

            Bug ID: 220117
           Summary: cpufreq performance governor doesn't correctly apply
                    in combination with isolcpus
           Product: Power Management
           Version: 2.5
          Hardware: Intel
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P3
         Component: cpufreq
          Assignee: linux-pm@vger.kernel.org
          Reporter: robbert.proost@vitestro.com
        Regression: No

During testing on my setup with kernel 6.6.69 I have found that even though the
cpufreq uses the performance governor, it doesn't seem to correctly apply on
CPUs that are isolated using `isolcpus`.

In my setup if have isolated core 2 and 3 set
CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE in my kernel configuration to 'Y'.
When verifying the used governor with 'cat
/sys/devices/system/cpu/cpu2/cpufreq/scaling_governor' it correctly outputs
'performance'.

However, when running 'cat /proc/cpuinfo' the clock frequency of the isolated
CPUs is 800 MHz instead of the expected 3000 MHz. For the not isolated CPUs it
does correctly show 3000 MHz. Changing
'/sys/devices/system/cpu/cpuX/cpufreq/scaling_min_freq' to 3000 MHz does change
the clock frequency to 3000 MHz, but that shouldn't be needed.

Kernel bootline: console=tty console=ttyS0,115200 acpi_enforce_resources=lax
panic=1 earlyprintk isolcpus=2,3 rootwait

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

You are receiving this mail because:
You are the assignee for the bug.

[Bug 220117] cpufreq performance governor doesn't correctly apply in combination with isolcpus

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220117

Doug Smythies (dsmythies@telus.net) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |dsmythies@telus.net

--- Comment #1 from Doug Smythies (dsmythies@telus.net) ---
Depending on the CPU frequency scaling driver, yes listed stale frequencies
might just be the current value for scaling_min_freq. If the processor is idle
and has been for awhile the CPU frequency is considered to be "stale" and a
default value is listed.

Here is an example for my test computer with an Intel i5-10600K processor, with
CPUs 2 and 3 isolated (which isn't actually relevant):

doug@s19:~$ grep . /sys/devices/system/cpu/cpu*/cpufreq/scaling_driver
/sys/devices/system/cpu/cpu0/cpufreq/scaling_driver:intel_pstate
/sys/devices/system/cpu/cpu10/cpufreq/scaling_driver:intel_pstate
/sys/devices/system/cpu/cpu11/cpufreq/scaling_driver:intel_pstate
/sys/devices/system/cpu/cpu1/cpufreq/scaling_driver:intel_pstate
/sys/devices/system/cpu/cpu2/cpufreq/scaling_driver:intel_pstate
/sys/devices/system/cpu/cpu3/cpufreq/scaling_driver:intel_pstate
/sys/devices/system/cpu/cpu4/cpufreq/scaling_driver:intel_pstate
/sys/devices/system/cpu/cpu5/cpufreq/scaling_driver:intel_pstate
/sys/devices/system/cpu/cpu6/cpufreq/scaling_driver:intel_pstate
/sys/devices/system/cpu/cpu7/cpufreq/scaling_driver:intel_pstate
/sys/devices/system/cpu/cpu8/cpufreq/scaling_driver:intel_pstate
/sys/devices/system/cpu/cpu9/cpufreq/scaling_driver:intel_pstate
doug@s19:~$ grep . /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
/sys/devices/system/cpu/cpu0/cpufreq/scaling_governor:performance
/sys/devices/system/cpu/cpu10/cpufreq/scaling_governor:performance
/sys/devices/system/cpu/cpu11/cpufreq/scaling_governor:performance
/sys/devices/system/cpu/cpu1/cpufreq/scaling_governor:performance
/sys/devices/system/cpu/cpu2/cpufreq/scaling_governor:performance
/sys/devices/system/cpu/cpu3/cpufreq/scaling_governor:performance
/sys/devices/system/cpu/cpu4/cpufreq/scaling_governor:performance
/sys/devices/system/cpu/cpu5/cpufreq/scaling_governor:performance
/sys/devices/system/cpu/cpu6/cpufreq/scaling_governor:performance
/sys/devices/system/cpu/cpu7/cpufreq/scaling_governor:performance
/sys/devices/system/cpu/cpu8/cpufreq/scaling_governor:performance
/sys/devices/system/cpu/cpu9/cpufreq/scaling_governor:performance
doug@s19:~$ grep MHz /proc/cpuinfo
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000

My test computer is extremely idle and all cpu frequencies were stale.
Now, adding a load to isolated CPU 2 gives:

doug@s19:~$ grep MHz /proc/cpuinfo
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 4799.997
cpu MHz         : 800.000
cpu MHz         : 4800.906
cpu MHz         : 4800.362
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 800.000
cpu MHz         : 4800.017
cpu MHz         : 800.000

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

You are receiving this mail because:
You are the assignee for the bug.

[Bug 220117] cpufreq performance governor doesn't correctly apply in combination with isolcpus

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220117

Rolf Leggewie (bugzilla.kernel.org@21.leggewie.org) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |bugzilla.kernel.org@21.legg
                   |                            |ewie.org

--- Comment #2 from Rolf Leggewie (bugzilla.kernel.org@21.leggewie.org) ---
so, this is not a bug and can be closed?

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

You are receiving this mail because:
You are the assignee for the bug.

[Bug 220117] cpufreq performance governor doesn't correctly apply in combination with isolcpus

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220117

--- Comment #3 from Doug Smythies (dsmythies@telus.net) ---
Yes, this is not a bug and can be closed.

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

You are receiving this mail because:
You are the assignee for the bug.