Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 13 Mar 2023 16:22:57 +0000
From:      void <void@f-m.fm>
To:        freebsd-hackers@freebsd.org
Subject:   CPU supports Enhanced Speedstep, but is not recognized.
Message-ID:  <ZA9N4SFhoILqCeIr@int21h>

next in thread | raw e-mail | index | archive | help
Hello hackers@

On a 13/stable system built today with generic kernel, there are 
these lines for each cpu in the dmesg:-

###
est0: <Enhanced SpeedStep Frequency Control> numa-domain 0 on cpu0
est: CPU supports Enhanced Speedstep, but is not recognized.
est: cpu_vendor GenuineIntel, msr 231100002100
device_attach: est0 attach returned 6
###

and then, later

###
interface hwpstate_intel.1 already present in the KLD 'kernel'!
linker_load_file: /boot/kernel/cpufreq.ko - unsupported file type
###

The system is a HP ProLiant dual Xeon E5-2690 v2, so there are 40 "cpus"
ie 20 if hyperthreading is turned off, in total.

I was kind of surprised this didn't work out-of-the-box considering
the age of the cpu, its use-context and how widely it was/is used
in that context. Clearly I'm doing something wrong or missing something.

Should it work?
Do I need to build a custom kernel? How can I debug this?
I tried using powerd but I understand from later reading that powerd
is incompatible with hwpstate_intel, so it's deactivated at the moment.

-- 



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?ZA9N4SFhoILqCeIr>