From: Jelle de Jong <jelledejong@powercraft.nl>
To: cpufreq@vger.kernel.org
Subject: Re: can not seem to set cpufreq to max performance with newer kernels for xeons X56xx
Date: Sat, 29 Feb 2020 15:31:39 +0100 [thread overview]
Message-ID: <dd6f0f29-539b-9508-aca8-98c61c6cd7a3@powercraft.nl> (raw)
In-Reply-To: <7c59571c-0693-85e1-6836-64c67399f819@powercraft.nl>
Hello everybody,
I am still having the issue that using the performance driver under
newer kernels does not set the cpu frequency to max. How can I run the
CPU at his max speed on newer kernels with the older Intel X56XX cpu's.
# cpupower frequency-set -g performance
# lscpu | grep "CPU MHz:"
CPU MHz: 1750.195
# uname -a
Linux virt04 4.19.0-8-amd64 #1 SMP Debian 4.19.98-1 (2020-01-26) x86_64
GNU/Linux
# lshw -c processor | grep version
version: Intel(R) Xeon(R) CPU X5675 @ 3.07GHz
version: Intel(R) Xeon(R) CPU X5675 @ 3.07GHz
Kind regards,
Jelle de Jong
On 2020-01-01 07:45, Jelle de Jong wrote:
> Hello everybody,
>
> How do I force the X56xx CPUs to stop scaling and run on maximum speed
> again?
>
> I been experiencing performance issues for on some ceph storage clusters
> with slow requests, and I found at least a partial issue issue in the
> CPU frequency scalling.
>
> I debugged on two different platform to and three different Debian
> versions to make sure it was not a bios or version issue.
>
> The following kernel is working with cpu scalling performance on both
> platforms, the cpu can run on his max without scalling back:
> 4.9.0-11-amd64 #1 SMP Debian 4.9.189-3+deb9u2 (2019-11-11) x86_64 GNU/Linux
>
> # not working performance scalling Intel Xeon X5680 on Intel S5520HC
> 5.4.0-1-amd64 #1 SMP Debian 5.4.6-1 (2019-12-27) x86_64 GNU/Linux
>
> # not working performance scalling Intel Xeon X5670 on HP Z600
> 5.4.0-1-amd64 #1 SMP Debian 5.4.6-1 (2019-12-27) x86_64 GNU/Linux
> 5.3.0-0.bpo.2-amd64 #1 SMP Debian 5.3.9-2~bpo10+1 (2019-11-13) x86_64
> GNU/Linux
> 4.19.0-6-amd64 #1 SMP Debian 4.19.67-2+deb10u2 (2019-11-11) x86_64
> GNU/Linux
>
> I need to use a newer kernel for atlantic.ko 10Gbit driver which is
> buggy on 4.9.0-11.
>
> I tried setting intel_pstate=enable acpi=force and intel_pstate=disable
> acpi=force but both did not work with the 5.x kernels i tried.
>
> BOOT_IMAGE=/vmlinuz-5.4.0-1-amd64 root=/dev/mapper/lvm0--vol-root ro
> rootfstype=ext4 fsck.mode=force console=ttyS0,115200 rootdelay=3
> intel_pstate=enable acpi=force quiet
>
> The CPU just keeps calling down to a lower freq with both the new
> cpupower and older cpufrequtils utils.
>
> cpupower show different current cpu speeds.
>
> root@ceph03:~# cpupower frequency-set --governor performance --min
> 2927000 --max 2927000
> Setting cpu: 0
> Setting cpu: 1
> Setting cpu: 2
> Setting cpu: 3
> Setting cpu: 4
> Setting cpu: 5
> Setting cpu: 6
> Setting cpu: 7
> Setting cpu: 8
> Setting cpu: 9
> Setting cpu: 10
> Setting cpu: 11
> Setting cpu: 12
> Setting cpu: 13
> Setting cpu: 14
> Setting cpu: 15
> Setting cpu: 16
> Setting cpu: 17
> Setting cpu: 18
> Setting cpu: 19
> Setting cpu: 20
> Setting cpu: 21
> Setting cpu: 22
> Setting cpu: 23
>
> # lscpu
> Architecture: x86_64
> CPU op-mode(s): 32-bit, 64-bit
> Byte Order: Little Endian
> CPU(s): 24
> On-line CPU(s) list: 0-23
> Thread(s) per core: 2
> Core(s) per socket: 6
> Socket(s): 2
> NUMA node(s): 2
> Vendor ID: GenuineIntel
> CPU family: 6
> Model: 44
> Model name: Intel(R) Xeon(R) CPU X5670 @ 2.93GHz
> Stepping: 2
> CPU MHz: 1770.170
> CPU max MHz: 2927.0000
> CPU min MHz: 1596.0000
> BogoMIPS: 5852.10
> Virtualization: VT-x
> L1d cache: 32K
> L1i cache: 32K
> L2 cache: 256K
> L3 cache: 12288K
> NUMA node0 CPU(s): 0-5,12-17
> NUMA node1 CPU(s): 6-11,18-23
> Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr
> pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ht tm pbe
> syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good
> nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor
> ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm pcid dca sse4_1 sse4_2
> popcnt aes lahf_lm epb pti ssbd ibrs ibpb stibp tpr_shadow vnmi
> flexpriority ept vpid dtherm ida arat flush_l1d
>
> root@ceph03:~# cpupower frequency-info
> analyzing CPU 0:
> driver: acpi-cpufreq
> CPUs which run at the same hardware frequency: 0
> CPUs which need to have their frequency coordinated by software: 0
> maximum transition latency: 10.0 us
> hardware limits: 1.60 GHz - 2.93 GHz
> available frequency steps: 2.93 GHz, 2.93 GHz, 2.79 GHz, 2.66 GHz,
> 2.53 GHz, 2.39 GHz, 2.26 GHz, 2.13 GHz, 2.00 GHz, 1.86 GHz, 1.73 GHz,
> 1.60 GHz
> available cpufreq governors: powersave userspace conservative
> ondemand performance schedutil
> current policy: frequency should be within 2.93 GHz and 2.93 GHz.
> The governor "performance" may decide which speed to use
> within this range.
> current CPU frequency: 2.93 GHz (asserted by call to hardware)
> boost state support:
> Supported: yes
> Active: yes
> 3200 MHz max turbo 4 active cores
> 3200 MHz max turbo 3 active cores
> 3333 MHz max turbo 2 active cores
> 3333 MHz max turbo 1 active cores
>
> root@ceph03:~# cat /etc/default/cpufrequtils
> GOVERNOR="performance"
> root@ceph03:~# systemctl status cpufrequtils
> ● cpufrequtils.service - LSB: set CPUFreq kernel parameters
> Loaded: loaded (/etc/init.d/cpufrequtils; generated; vendor preset:
> enabled)
> Active: active (exited) since Tue 2019-12-31 14:12:36 CET; 41min ago
> Docs: man:systemd-sysv-generator(8)
> Process: 1734519 ExecStop=/etc/init.d/cpufrequtils stop (code=exited,
> status=0/SUCCESS)
> Process: 1734527 ExecStart=/etc/init.d/cpufrequtils start
> (code=exited, status=0/SUCCESS)
>
> Dec 31 14:12:36 ceph03 systemd[1]: Starting LSB: set CPUFreq kernel
> parameters...
> Dec 31 14:12:36 ceph03 cpufrequtils[1734527]: CPUFreq Utilities: Setting
> performance CPUFreq
> governor...CPU0...CPU1...CPU2...CPU3...CPU4...CPU5...CPU6...CPU7...CPU8...CPU9...CPU10...CPU11...CPU12...CPU13...CPU14...CPU15...CPU16...CPU17...CPU18...CPU19...
>
> Dec 31 14:12:36 ceph03 systemd[1]: Started LSB: set CPUFreq kernel
> parameters.
>
> root@ceph03:~# cpufreq-info --freq
> 1595816
> root@ceph03:~# cpufreq-info --driver
> acpi-cpufreq
> root@ceph03:~# cpufreq-info --policy
> 2927000 2927000 performance
> root@ceph03:~# cpufreq-info --stats
> 2927000:16889032, 2926000:2, 2793000:1, 2660000:3, 2527000:4, 2394000:9,
> 2261000:12, 2128000:12, 1995000:15, 1862000:30, 1729000:85, 1596000:1762
> (167)
>
> Kind regards,
>
> Jelle de Jong
prev parent reply other threads:[~2020-02-29 14:31 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-01-01 6:45 can not seem to set cpufreq to max performance with newer kernels for xeons X56xx Jelle de Jong
2020-02-29 14:31 ` Jelle de Jong [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=dd6f0f29-539b-9508-aca8-98c61c6cd7a3@powercraft.nl \
--to=jelledejong@powercraft.nl \
--cc=cpufreq@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).