Page: 1, 2  Next

[RESOLVED] intel i7 12700H CPU running at lowest frequency
Eric97
Status: New User - Welcome
Joined: 18 Jun 2022
Posts: 4
Reply Quote
Hello,

My linux kernel version on my MSI GE76 laptop as below:

Linux GE76 5.18.0-5.1-liquorix-amd64 #1 ZEN SMP PREEMPT_DYNAMIC liquorix 5.18-1ubuntu1~focal (2022-06 x86_64 x86_64 x86_64 GNU/Linux
linux-firmware at 1.187.31

The CPU is always running at lowest frequency at 400.000 MHz from the /proc/cpuinfo. I have also disabled the thermald service. still the same problem.

:: Quote ::

processor : 19
vendor_id : GenuineIntel
cpu family : 6
model : 154
model name : 12th Gen Intel(R) Core(TM) i7-12700H
stepping : 3
microcode : 0x413
cpu MHz : 400.000
cache size : 24576 KB
physical id : 0
siblings : 20
core id : 31
cpu cores : 14
apicid : 62
initial apicid : 62
fpu : yes
fpu_exception : yes
cpuid level : 32
wp : yes
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 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l2 invpcid_single cdp_l2 ssbd ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid rdt_a rdseed adx smap clflushopt clwb intel_pt sha_ni xsaveopt xsavec xgetbv1 xsaves split_lock_detect avx_vnni dtherm ida arat pln pts hwp hwp_notify hwp_act_window hwp_epp hwp_pkg_req hfi umip pku ospke waitpkg gfni vaes vpclmulqdq rdpid movdiri movdir64b fsrm md_clear serialize arch_lbr ibt flush_l1d arch_capabilities
vmx flags : vnmi preemption_timer posted_intr invvpid ept_x_only ept_ad ept_1gb flexpriority apicv tsc_offset vtpr mtf vapic ept vpid unrestricted_guest vapic_reg vid ple shadow_vmcs ept_mode_based_exec tsc_scaling usr_wait_pause
bugs : spectre_v1 spectre_v2 spec_store_bypass swapgs
bogomips : 5376.00
clflush size : 64
cache_alignment : 64
address sizes : 39 bits physical, 48 bits virtual
power management:



Any suggestions to fix this issue? Thanks
Back to top
techAdmin
Status: Site Admin
Joined: 26 Sep 2003
Posts: 4127
Location: East Coast, West Coast? I know it's one of them.
Reply Quote
That's alder lake, 12th gen Intel. I don't believe Linux support is fully there for Alder Lake.

inxi -Ca --zv would be useful to confirm some things.

But you appear to be unclear about what Alder lake is, it's got a set of performance cores, and a set of efficiency cores, I do not know if Linux fully supports those things yet or not. It didn't as of a few months back, don't know the full support status yet. In terms of the real performance you get, I believe on Linux you're betting off with a Ryzen, though that may change, hard to say. Maybe Damentz can fill in on the current Alder lake status.

inxi -Ca --zv will show a more clear picture of what is going on there across your cores.
Back to top
Eric97
Status: New User - Welcome
Joined: 18 Jun 2022
Posts: 4
Reply Quote
Thanks for the heads up.

The output of inxi:
:: Quote ::

CPU:
Topology: 10-Core model: 12th Gen Intel Core i7-12700H bits: 64
type: MT MCP arch: N/A family: 6 model-id: 9A (154) stepping: 3
microcode: 413 L2 cache: 24.0 MiB
flags: avx avx2 lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx
bogomips: 107520
Speed: 457 MHz min/max: 400/2701 MHz boost: enabled Core speeds (MHz):
1: 456 2: 401 3: 400 4: 400 5: 399 6: 400 7: 400 8: 400 9: 400 10: 400
11: 400 12: 400 13: 400 14: 494 15: 751 16: 586 17: 1119 18: 465 19: 400
20: 400
Vulnerabilities: Type: itlb_multihit status: Not affected
Type: l1tf status: Not affected
Type: mds status: Not affected
Type: meltdown status: Not affected
Type: mmio_stale_data status: Not affected
Type: spec_store_bypass
mitigation: Speculative Store Bypass disabled via prctl
Type: spectre_v1
mitigation: usercopy/swapgs barriers and __user pointer sanitization
Type: spectre_v2 status: Vulnerable: eIBRS with unprivileged eBPF
Type: srbds status: Not affected
Type: tsx_async_abort status: Not affected


Back to top
damentz
Status: Assistant
Joined: 09 Sep 2008
Posts: 1122
Reply Quote
Try adding intel_pstate=enable to your kernel boot parameters. There may be a bug with Alder Lake CPUs where the more robust acpi-cpufreq doesn't work properly.

The Intel P-state governor is disabled in Liquorix by default since its behavior is not desirable for interactive systems. You can read more about it in the commit that introduced the "enable" parameter: github.com/zen-kernel/zen-kernel/commit/38340ce3e7354f61213653ffd78fc7ab3f924f9d
Back to top
Eric97
Status: New User - Welcome
Joined: 18 Jun 2022
Posts: 4
Reply Quote
Thanks.
Adding in as GRUB_CMDLINE_LINUX_DEFAULT="quiet splash intel_pstate=enable". but seems nothing changed.
Back to top
techAdmin
Status: Site Admin
Joined: 26 Sep 2003
Posts: 4127
Location: East Coast, West Coast? I know it's one of them.
Reply Quote
Note you are using a legacy inxi, the current version does a far better job with complicated cpus like alder lake, including a far more accurate report on per cpu core type speeds, min max, and a much advanced topology report.

It was designed specifically due to the issues with getting alder lake data to show more accurately.

In /etc/inxi.conf change to: B_ALLOW_UPDATES=true
then do the self updater: sudo inxi -U
then again show: inxi -Ca --zv

and you'll be showing a far more useful report of what is going on there.

I can't quite tell the core sets there in the legacy inxi.

But alder lake, you should have waited probably 1 year to buy that if you run linux (and maybe never if you run a bsd), that was a huge change that it is taking the kernel a while to catch up with. Would have been better off imo with a ryzen 3, where you'd get all the performance and features out of the box. Even microsoft was unable to support it for older windows initially, it was a big change.
Back to top
damentz
Status: Assistant
Joined: 09 Sep 2008
Posts: 1122
Reply Quote
It's also important to note, is that your CPUs are scaling. Per your own post, some cores are scaling over 400mhz.

Quoting the exact section:
:: Code ::
Speed: 457 MHz min/max: 400/2701 MHz boost: enabled Core speeds (MHz):
1: 456 2: 401 3: 400 4: 400 5: 399 6: 400 7: 400 8: 400 9: 400 10: 400
11: 400 12: 400 13: 400 14: 494 15: 751 16: 586 17: 1119 18: 465 19: 400
20: 400


So with this in mind, is there a problem here? With inxi updated as techAdmin posted, what's the output of: inxi -Cabz?

Also, what's the output of:

:: Code ::
$ cat /sys/devices/system/cpu/cpufreq/boost
$ cat /sys/devices/system/cpu/cpufreq/policy0/scaling_driver
$ cat /sys/devices/system/cpu/cpufreq/policy0/scaling_governor


If the values are anything besides below, something on your system is altering Liquorix defaults and giving you a horrible experience:
:: Code ::
1
acpi-cpufreq
performance


If that's the case, you'll need to chase down whatever is changing it or wholesale change your distribution.

And last thing, I recommend disabling your efficiency cores through the BIOS. The scheduler used by Liquorix, PDS, has no understanding of what they are and you'll randomly get poor performance with them turned on. Overall, Liquorix is not compatible with big/little processors like Alder Lake and you need to make adjustments to have all dissimilar cores be disabled.
Back to top
Eric97
Status: New User - Welcome
Joined: 18 Jun 2022
Posts: 4
Reply Quote
After update inxi, the output of inxi -Ca --zv:
:: Quote ::

CPU:
Info: model: 12th Gen Intel Core i7-12700H bits: 64 type: MST AMCP
arch: Alder Lake gen: core 12 built: 2021 process: Intel 7 (10nm ESF)
family: 6 model-id: 0x9A (154) stepping: 3 microcode: 0x413
Topology: cpus: 1x cores: 14 mt: 6 tpc: 2 st: 8 threads: 20 smt: enabled
cache: L1: 1.2 MiB desc: d-8x32 KiB, 6x48 KiB; i-6x32 KiB, 8x64 KiB
L2: 11.5 MiB desc: 6x1.2 MiB, 2x2 MiB L3: 24 MiB desc: 1x24 MiB
Speed (MHz): avg: 420 high: 495 min/max: 400/4679:4700:3500 scaling:
driver: intel_pstate governor: powersave cores: 1: 420 2: 429 3: 423 4: 495
5: 477 6: 430 7: 424 8: 418 9: 421 10: 419 11: 414 12: 432 13: 400
14: 400 15: 400 16: 400 17: 400 18: 400 19: 400 20: 400 bogomips: 107520
Flags: avx avx2 ht lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx
Vulnerabilities: <filter>


All the cat outputs are not showing anything that alters the liquorix defaults, so should be fine.

Seems have to wait a few more months to have Linux full support.. or maybe need to buy another laptop with either Ryzen or Intel 11th gen CPU.
Back to top
damentz
Status: Assistant
Joined: 09 Sep 2008
Posts: 1122
Reply Quote
I missed this originally, but you're running an MSI laptop. MSI is well known for making gaming laptops for gamers, but I don't know their history for Linux workstations.

There's a pretty high chance that the BIOS is reading you're booting with Linux and not exposing the right power management features. Even Lenovo has messed this up with Thinkpads and had to release an update on a laptop I own so that the correct TDP levels are applied on boot. Without the update, the laptop would use power settings more similar to running on battery.

I recommend you update the bios. If you had posted the output of inxi -b, we could see if it was out of date and exactly which model you have (there's 3 models of the MSI GE76), but it's highly likely the BIOS update won't fix anything. Outside of the major laptop makers that support enterprise, the smaller ones like MSI don't typically fix bugs in the BIOS/EFI that are reported as affecting Linux. Still, it's worth trying.

:: Quote ::
Seems have to wait a few more months to have Linux full support.. or maybe need to buy another laptop with either Ryzen or Intel 11th gen CPU.


Also, considering you run Linux, try sticking with with Dell XPS/Precision, Lenovo Thinkpad, System76, or even the new HP Dev One that runs a Ryzen. Outside of these knowns, your mileage may vary.
Back to top
techAdmin
Status: Site Admin
Joined: 26 Sep 2003
Posts: 4127
Location: East Coast, West Coast? I know it's one of them.
Reply Quote
As noted, inxi -bza would have shown all the bits.

There's a few tests you want to run, since all your cores are running at 400+ Mhz, first, you want to see what happens when you start a bunch of high cpu processes:

open a terminal, create 8 tabs, in each tab, do this:

:: Code ::
cat /dev/urandom > /dev/null


You will use ctrl + c to terminate the process when you are done with this test.

After they are all running for maybe 15 seconds, do:

inxi -C

This will show the list of speeds. The above command tells one thread to do this task as many times as it can per second, the goal here is you want to load the cpu threads as much as possible.

If you aren't seeing 8 threads at very high speed/cpu utilization, then I think damentz has probably nailed the cause.

There's a very common misconception that laptops are like desktops, they 'just work', but that's not the case, there are several laptop brands you want to in general avoid like the plague, toshiba is one, due to poor free software support, and others. mainly the ones damentz listed, that you want to buy if you care about everything working, and, important, continuing to work, in Linux kernel.

I'm hoping in a few years that the used market shapes up enough to let me buy a ryzen laptop from dell or lenovo.

Alder lake was intel's first attempt to _catch up_ with AMD. I credit them for trying, but based on power consumption numbers I see for alder lake for similar performance loads, we're probably going to have to wait a few steppings, or maybe even gen 13, before they catch zen/ryzen cpus, and zen is about to go to 3n process node, it's at 5n now, at zen 3/3+. note that the process numbers don't really reflect anything real beyond each generation getting x percent higher transistor density, wire widths, and x percent greater efficiency, so the best way to compare nodes now is really to do a comparable cpu load test and check the heat generated and power consumption required to achieve that load. Alder lake is behind in this metric, and is HOT, very hot, incredibly high power consumption.

Given intel's return to engineering from the black hole of marketing they had fallen into, and their renewed focus on scaling out chip fabs globally, I think they will catch up, roughly, by about 2024, maybe, unless amd maintains their lead, but it's getting a lot harder to hit the next nodes now, taking longer and longer.

But I don't see intel shipping a competitive product until 2024ish.
Back to top
Display posts from previous:   
Page: 1, 2  Next
All times are GMT - 8 Hours