Re: [PATCH v2] cpufreq: intel_pstate: Add Icelake servers support in no-HWP mode

[Doug Smythies]

On Fri, May 14, 2021 at 1:33 PM Giovanni Gherdovich <ggherdovich@xxxxxxx> wrote:
> On Fri, 2021-05-14 at 08:31 -0700, Doug Smythies wrote:
>
> > Can I on-board to this patch or do you want me to submit another?
> > I want to add COMETLAKE (tested), as below:
> >
> > ... Doug
>
> Hello Doug!

Hi Giovanni,
Thank you for your reply.
>
> Wait, why you don't want to use HWP? It's such a fantastic technology!
>
> :) I'm just teasing you.
>
> More seriously:
>
> when COMETLAKE is not in that list, can you confirm that if you go into the
> BIOS config at boot, and disable HWP from there, then intel_pstate does *not* load?

Yes, already tested before my original reply.

> Does it say "intel_pstate: CPU model not supported" in the dmesg log?

That I did not check, but if I boot now with an unmodified kernel
5.13-rc1 (i.e. without this patch):

[    0.369323] intel_pstate: CPU model not supported

> The control may be somewhere around "power mangement" in the BIOS config, and
> may be called "Enable/disable Intel Speed Shift".

Yes.

> I'm asking because I've just checked on two Dell laptops, one Skylake and the
> other Kabylake, and the menu is there in the BIOS config to disable HWP,
> but if I disable it... nothing happens. "lscpu" shows all the hwp flags as usual:

Motherboard here is ASUS PRIME Z490-A.
CPU: Intel(R) Core(TM) i5-10600K CPU @ 4.10GHz

>     # lscpu | grep Flags | tr ' ' '\n' | grep hwp
>     hwp
>     hwp_notify
>     hwp_act_window
>     hwp_epp

Here, for some reason I have to do it this way (sudo) or your command
doesn't work properly. Results herein confirmed by looking at the
"Flags" output manually without filtering:

intel_speed_shift = Disabled in BIOS:

doug@s19:~$ sudo lscpu | tr ' ' '\n' | grep hwp
doug@s19:~$

intel_speed_shift = Auto in BIOS

$ sudo lscpu | tr ' ' '\n' | grep hwp
hwp
hwp_notify
hwp_act_window
hwp_epp

> and turbostat gives me:
>
>     # turbostat -Summary -i 1 : 2>&1 | grep MSR_PM_ENABLE
>     cpu0: MSR_PM_ENABLE: 0x00000001 (HWP)

Here:

intel_speed_shift = Disabled in BIOS:

root@s19:/home/doug#
/home/doug/temp-k-git/linux/tools/power/x86/turbostat/turbostat
-Summary -i 1 : 2>&1 | grep MSR_PM_ENABLE
root@s19:/home/doug#

intel_speed_shift = Auto in BIOS (the default setting)

root@s19:/home/doug#
/home/doug/temp-k-git/linux/tools/power/x86/turbostat/turbostat
-Summary -i 1 : 2>&1 | grep MSR_PM_ENABLE
cpu0: MSR_PM_ENABLE: 0x00000001 (HWP)

or with "intel_pstate=no_hwp"

root@s19:/home/doug#
/home/doug/temp-k-git/linux/tools/power/x86/turbostat/turbostat
-Summary -i 1 : 2>&1 | grep MSR_PM_ENABLE
cpu0: MSR_PM_ENABLE: 0x00000000 (No-HWP)

> Which is to say, on the Intel client machines I have, the firmware doesn't
> seem to be able to hide HWP from the OS. Buggy BIOS? Maybe, the fact of the
> matter is, I wouldn't need to add, say, KABYLAKE to that list, based on my
> experience.

My experience (hardware) differs from yours with respect to this.

> The other side of the issue is that, from my understanding, the
> preferred/supported way to disable HWP is to boot with intel_pstate=no_hwp,
> and that list is a sort of "known exceptions" that people really can't live
> without (it's mostly server CPUs, and mostly because of unfortunate firmware
> defaults). Otherwise you'd see the entire intel-family.h file in there.

I'm not sure how to respond here. Yes, I'd expect to see a big list
here, and in the recently added TCC Offset thermal stuff and in the
recently added turbostat patches to deal with a TCC offset. I do not
understand doing things partially only. But that is a bigger/broader
subject than herein.

That said, yes, "intel_pstate=no_hwp" is what I normally do. And my
BIOS normally has "Intel Speed Shift = AUTO", which is the default.

... deleted the rest ...

... Doug