6 hours ago, veizour said:

Either way, things seem to be working fine now.

Have you isolated cores?

 

CoreFreq is designed to talk directly to the hardware and if you Isolate CPU cores it simply can't and cause issues like that.

😡 Upon running the command my Unraid server immediately crashed. Even IPMI needed a reset. Triggered a parity check (54 TB). Thanks alot.

ts200v-diagnostics-20210913-1547.zip

Edited September 13, 20214 yr by ripeart

11 minutes ago, ripeart said:

😡 Upon running the command my Unraid server immediately crashed.

Command? Do you mean installation?

 

Anyways, thank you for the report and also the Diagnostics, will create a Github issue to let @CyrIng know and also so that he can fix the issue.

 

13 minutes ago, ripeart said:

Triggered a parity check (54 TB). Thanks alot.

Sorry to hear that but if nothing was writing to the Array at the time of installation or better speaking the crash you can abort the Parity check.

For reasons of my own insanity, I want to get temperature readings on me server before and after I make soe risky and likely pointless modifications.

 

I looked and looked for some way to stress test the CPU and get a temperature reading. Eventually realizing I may have that already in this plugin.

 

Is there a way, or does it already, keep ap log of the results? Like running Temperature and frequency?

 

So far I've played with taking screen shots while running the atomic burn tool, but I'm not seeing the numbers I expected. My system is NOT overclocked, and my 10850K is maintaining 4800MHz on all cores at about 72 degrees... Surely I'm doing something wrong for i not to be running hotter.

 

Arbadacarba

7 hours ago, Arbadacarba said:

So far I've played with taking screen shots while running the atomic burn tool, but I'm not seeing the numbers I expected. My system is NOT overclocked, and my 10850K is maintaining 4800MHz on all cores at about 72 degrees... Surely I'm doing something wrong for i not to be running hotter.

Should it run hotter? On my 10600 I can reach temperatures as high as 64C.

 

As far as I know CoreFreq doesn't support logging the frequencies and the temperature, maybe you can create a Github Issue and maybe @CyrIng can look into it if he can add it so that a file is created that logs frequencies and temperatures.

I assumed it would... I've built this system around an air cooled quiet solution.

 

I'm in the industry but I really haven't built a computer other than my servers (built and rebuilt really) since the AMD Athlon. I've been using Laptops for about 20 years.

 

Maybe I just watch too many Linus videos.

 

I'm really just screwing around with the new build on the principle that I have the skills and the equipment so lets lap the IHS and Heatsink to try to get a TINY gain and slow my cooling system down even more.

 

No real reason. Just practicing arcane skills. The whole rig is meticulously cable managed and there is not a window in sight.

 

Thanks

32 minutes ago, Arbadacarba said:

I assumed it would... I've built this system around an air cooled quiet solution.

I think those temps are pretty normal, maybe if you hit it with a AVX load it would ramp up a bit more but I don't think this is a real world scenario where AVX hammers the CPU.

 

Air coolers are not that different to a AiO nowadays and they work pretty well while keeping the CPU cool.

Having Done a little further experimentation... The behavior I'm seeing is EXACTLY right... I was thinking along the frequencies that intel lists for this chip, but not realizing those are SINGLE core boost frequencies.

 

I'm seeing ALL cores hit 4.8Ghz until the package temp hits 80 degrees and then they back off (4.6 or so) until it stays below 80. 

 

Thanks

Ctrl + p

To screenshot current screen. The output file corefreq-xxxxxx can then be read with command less -R

 

Alt + p

To record everything as an ascii-cinema. Video duration can be set in Settings menu.

@Arbadacarba please see the above answer from @CyrIng.

CoreFreq 1.89.1 released

17 minutes ago, CyrIng said:

CoreFreq 1.89.1 released

Already compiled and updated. 🙂

Thanks for the plug in.

 

I'm seeing differences of around 5-6 degrees, between Dynamix System Temp and the temperature reported in CoreFreq.

 

Which is more accurate ?

 

Thanks

3 minutes ago, Myleslewis said:

Which is more accurate ?

I would think that CoreFreq is more accurate but also keep in mind that the WebGUI is maybe a little bit behind because it needs to reflesh and so on...

wanted to use this to see if c-states is/are the reason my server keeps locking up/becoming unresponsive, but not sure where i'm looking. is it able to flip the switch and the server carry on or would a reboot be needed?

50 minutes ago, gellux said:

wanted to use this to see if c-states is/are the reason my server keeps locking up/becoming unresponsive

Don't know if CoreFreq will help you see this...

 

50 minutes ago, gellux said:

is it able to flip the switch and the server carry on or would a reboot be needed?

What do you mean with this?

Is the Supermicro A1SAi with an Atom C2750 supported? I get a blank screen without a prompt in SSH when running corefreq-cli. I get a blinking cursor when I run corefreq-cli -c. When I ctrl-c, I get this:

 

4 hours ago, rbronco21 said:

Is the Supermicro A1SAi with an Atom C2750 supported?

I would recommend that you create an issue on @CyrIng's Github here: Click

 

If your system is not supported and he implements it feel free to contact me and I will create a special build from CoreFreq for you or mention me on Github with @ich777

8 hours ago, ich777 said:

I would recommend that you create an issue on @CyrIng's Github here: Click

 

If your system is not supported and he implements it feel free to contact me and I will create a special build from CoreFreq for you or mention me on Github with @ich777

@ich777

It is warmly recommended to new CoreFreq users to have a look first for processor support (architecture CPUID) within the Wiki - CPU Support:

 

github.com/cyring/CoreFreq/wiki

 

If CPU architecture is not listed, you have to download and try the live ISO image before everything:

 

github.com/cyring/CoreFreq/wiki/Live-CD

 

Fyi, this image is fully automated to build and run CoreFreq with ArchLinux inside. User disks are not mounted to spare them from any crash.

 

If image is starting nicely CoreFreq then you are go for the unRAID plugin.

 

Your screenshots and Cli outputs will be warmly welcomed to complete the CPU Support list.

A fix for the Atom C2750 is awaiting for your testings:

 

github.com/cyring/CoreFreq/issues/319#issuecomment-1024796833

 

@ich777 it might be a difficult for @rbronco21 to boot the ISO image or rebuild from branch. Could you create a dedicated unRAID plugin with this Avoton CoreFreq hot fix ?

 

Thank you both for your returns, and helping me to provide that fix to the mainline code. 

Edited February 5, 20224 yr by CyrIng

I'd love to figure out how to compile it on/for unraid if someone doesn't mind helping me out when I get stuck, even after the fix is included. I've installed devpack and a bunch of packages that got me started with make, including kernel-headers-4.14.37-x86-1.txz. I have the /usr/src/linux-5.10.28-Unraid folder, but it looks pretty sparse. Sounds like there should be headers in there?

 

root@unRAID:~# ls -l /usr/src/linux-5.10.28-Unraid
total 3752
-rw-r--r-- 1 root root     496 Apr  7  2021 COPYING
-rw-r--r-- 1 root root 3785633 Apr  7  2021 System.map
-rw-r--r-- 1 root root    2857 Apr  7  2021 acs_override.patch
drwxr-xr-x 3 root root      60 Apr  7  2021 drivers/
-rw-r--r-- 1 root root    1850 Apr  7  2021 drivers_ata_libata-core.patch
-rw-r--r-- 1 root root     294 Apr  7  2021 drivers_hpsa.patch
-rw-r--r-- 1 root root     529 Apr  7  2021 drivers_md_Kconfig.patch
-rw-r--r-- 1 root root     811 Apr  7  2021 drivers_md_Makefile.patch
-rw-r--r-- 1 root root     401 Apr  7  2021 drivers_mvsas_rr_2782.patch
-rw-r--r-- 1 root root     426 Apr  7  2021 fs_reiserfs_resize.patch
-rw-r--r-- 1 root root     459 Apr  7  2021 fs_reiserfs_super.patch
-rw-r--r-- 1 root root     454 Apr  7  2021 ignore_unhandled_msrs.patch
-rw-r--r-- 1 root root     402 Apr  7  2021 increase_BLK_MAX_REQUEST_COUNT.patch
-rw-r--r-- 1 root root    1348 Apr  7  2021 mozart_395S.patch
-rw-r--r-- 1 root root    2365 Apr  7  2021 raid6_choose_xor.patch

 

edit:

binutils-2.31.1-x86_64-4.txz
gc-8.0.4-x86_64-1.txz
gcc-9.2.0-x86_64-4.txz
glibc-2.30-x86_64-1.txz
guile-2.2.3-x86_64-2.txz
kernel-headers-4.14.37-x86-1.txz
make-4.2.1-x86_64-3.txz

Edited February 5, 20224 yr by rbronco21
installed packages

33 minutes ago, rbronco21 said:

I'd love to figure out how to compile it on/for unraid if someone doesn't mind helping me out when I get stuck, even after the fix is included.

You don't have to do that on your own.

Please tell me your unRAID version and I will compile it for you and I will tell you how to install it.

 

From what I see you are on 6.9.2 is that right?

@ich777 Hello

 

Atom is fixed in 1.89.3

 

Other users are still using plugin with an outdated version of CoreFreq.

 

I would suggest you won't autostart CoreFreq in the plugin, not even the driver, and let Users manage to start manually with a warning "Experimental" and a reference to the GitHub project to read first.

 

What do you think about that ?

35 minutes ago, CyrIng said:

I would suggest you won't autostart CoreFreq in the plugin, not even the driver, and let Users manage to start manually with a warning "Experimental" and a reference to the GitHub project to read first.

Sure thing, will change that in the next couple days, maybe I will build a small plugin page where the users can choose if it started automatically or not, by default auto-start disabled.

Hello there and thanks for the effort.

 

It is the second day in a row the server crashed and became unresponsive.

 

From the logs I managed to grab through my remote log server I see the following (attached at the end). I am not the expert in linux but I suspected corefreq as the cause and uninstalled it before the second crash.

 

Any advice on what is going on and if the issue is indeed related to corefreq?

 

My system has

Gigabyte Technology Co., Ltd. B560M AORUS PRO AX

Intel Core i7-10700

64gb ddr4 HyperX predator ram (4x16 dimms)

 

Before uninstalling corefreq on 2022-02-16

Feb 16 11:18:20 Tower kernel: rcu: INFO: rcu_sched detected stalls on CPUs/tasks:
Feb 16 11:18:20 Tower kernel: rcu: #0116-...0: (0 ticks this GP) idle=772/1/0x4000000000000000 softirq=62494954/62494954 fqs=14977 
Feb 16 11:18:20 Tower kernel: #011(detected by 12, t=60002 jiffies, g=133319997, q=11304)
Feb 16 11:18:20 Tower kernel: Sending NMI from CPU 12 to CPUs 6:
Feb 16 11:18:20 Tower kernel: NMI backtrace for cpu 6
Feb 16 11:18:20 Tower kernel: CPU: 6 PID: 7350 Comm: php Tainted: G        W  O      5.10.28-Unraid #1
Feb 16 11:18:20 Tower kernel: Hardware name: Gigabyte Technology Co., Ltd. B560M AORUS PRO AX/B560M AORUS PRO AX, BIOS F4 03/04/2021
Feb 16 11:18:20 Tower kernel: RIP: 0010:native_queued_spin_lock_slowpath+0x79/0x18a
Feb 16 11:18:20 Tower kernel: Code: c1 e0 08 89 c2 8b 07 30 e4 09 d0 a9 00 01 ff ff 74 0c 0f ba e0 08 72 1a c6 47 01 00 eb 14 85 c0 74 0a 8b 07 84 c0 74 04 f3 90 <eb> f6 66 c7 07 01 00 c3 48 c7 c0 00 30 02 00 65 48 03 05 f0 8e f8
Feb 16 11:18:20 Tower kernel: RSP: 0018:ffffc900002a4e58 EFLAGS: 00000002
Feb 16 11:18:20 Tower kernel: RAX: 0000000000180101 RBX: ffffc900002a4e80 RCX: 0000000000000000
Feb 16 11:18:20 Tower kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8881000b9540
Feb 16 11:18:20 Tower kernel: RBP: ffff88814911a000 R08: ffffffff82013888 R09: ffffffff82013580
Feb 16 11:18:20 Tower kernel: R10: 000a5d1eae38384c R11: 0000000000000000 R12: 0000000000000000
Feb 16 11:18:20 Tower kernel: R13: 0000000000000000 R14: 0000000000000000 R15: 00002c52b99306c0
Feb 16 11:18:20 Tower kernel: FS:  0000148fc63ab780(0000) GS:ffff88907c380000(0000) knlGS:0000000000000000
Feb 16 11:18:20 Tower kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Feb 16 11:18:20 Tower kernel: CR2: 0000148fc96eaf52 CR3: 000000011e53e003 CR4: 00000000007726e0
Feb 16 11:18:20 Tower kernel: PKRU: 55555554
Feb 16 11:18:20 Tower kernel: Call Trace:
Feb 16 11:18:20 Tower kernel: <IRQ>
Feb 16 11:18:20 Tower kernel: queued_spin_lock_slowpath+0x7/0xa
Feb 16 11:18:20 Tower kernel: nr_blockdev_pages+0x13/0x64
Feb 16 11:18:20 Tower kernel: si_meminfo+0x3a/0x57
Feb 16 11:18:20 Tower kernel: Sys_MemInfo+0x20/0x9b [corefreqk]
Feb 16 11:18:20 Tower kernel: ? update_cfs_rq_load_avg+0x14b/0x154
Feb 16 11:18:20 Tower kernel: ? timekeeping_get_ns+0x19/0x2f
Feb 16 11:18:20 Tower kernel: ? Sys_DumpTask+0xe9/0xf1 [corefreqk]
Feb 16 11:18:20 Tower kernel: Cycle_Skylake+0x53a/0x730 [corefreqk]
Feb 16 11:18:20 Tower kernel: __hrtimer_run_queues+0xb7/0x10b
Feb 16 11:18:20 Tower kernel: ? Cycle_Skylake_X+0x711/0x711 [corefreqk]
Feb 16 11:18:20 Tower kernel: hrtimer_interrupt+0x8d/0x15b
Feb 16 11:18:20 Tower kernel: __sysvec_apic_timer_interrupt+0x5d/0x68
Feb 16 11:18:20 Tower kernel: asm_call_irq_on_stack+0xf/0x20
Feb 16 11:18:20 Tower kernel: </IRQ>

 

 

After uninstalling corefreq on 2022-02-17

Feb 17 12:17:56 Tower kernel: rcu: INFO: rcu_sched self-detected stall on CPU
Feb 17 12:17:56 Tower kernel: rcu: #0115-....: (60000 ticks this GP) idle=25a/1/0x4000000000000000 softirq=3595028/3595028 fqs=14991 
Feb 17 12:17:56 Tower kernel: #011(t=60001 jiffies g=8644697 q=12441)
Feb 17 12:17:56 Tower kernel: Sending NMI from CPU 5 to CPUs 2:
Feb 17 12:17:56 Tower kernel: NMI backtrace for cpu 2
Feb 17 12:17:56 Tower kernel: CPU: 2 PID: 31137 Comm: sh Tainted: G        W  O      5.10.28-Unraid #1
Feb 17 12:17:56 Tower kernel: Hardware name: Gigabyte Technology Co., Ltd. B560M AORUS PRO AX/B560M AORUS PRO AX, BIOS F4 03/04/2021
Feb 17 12:17:56 Tower kernel: RIP: 0010:native_queued_spin_lock_slowpath+0x79/0x18a
Feb 17 12:17:56 Tower kernel: Code: c1 e0 08 89 c2 8b 07 30 e4 09 d0 a9 00 01 ff ff 74 0c 0f ba e0 08 72 1a c6 47 01 00 eb 14 85 c0 74 0a 8b 07 84 c0 74 04 f3 90 <eb> f6 66 c7 07 01 00 c3 48 c7 c0 00 30 02 00 65 48 03 05 f0 8e f8
Feb 17 12:17:56 Tower kernel: RSP: 0018:ffffc900001f4e58 EFLAGS: 00000002
Feb 17 12:17:56 Tower kernel: RAX: 0000000000200101 RBX: ffffc900001f4e80 RCX: 0000000000000000
Feb 17 12:17:56 Tower kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8881000b9540
Feb 17 12:17:56 Tower kernel: RBP: ffff888148c5c000 R08: ffffffff82013888 R09: ffffffff82013580
Feb 17 12:17:56 Tower kernel: R10: 0000bbd29ebe3599 R11: 0000000000000000 R12: 0000000000000000
Feb 17 12:17:56 Tower kernel: R13: 0000000000000000 R14: 0000000000000000 R15: 0000040192991d6a
Feb 17 12:17:56 Tower kernel: FS:  0000152b91169740(0000) GS:ffff88907c280000(0000) knlGS:0000000000000000
Feb 17 12:17:56 Tower kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Feb 17 12:17:56 Tower kernel: CR2: 0000152b911ece40 CR3: 0000000165a5a001 CR4: 00000000007726e0
Feb 17 12:17:56 Tower kernel: PKRU: 55555554
Feb 17 12:17:56 Tower kernel: Call Trace:
Feb 17 12:17:56 Tower kernel: <IRQ>
Feb 17 12:17:56 Tower kernel: queued_spin_lock_slowpath+0x7/0xa
Feb 17 12:17:56 Tower kernel: nr_blockdev_pages+0x13/0x64
Feb 17 12:17:56 Tower kernel: si_meminfo+0x3a/0x57
Feb 17 12:17:56 Tower kernel: Sys_MemInfo+0x20/0x9b [corefreqk]
Feb 17 12:17:56 Tower kernel: ? ttwu_do_wakeup.isra.0+0xe/0x8c
Feb 17 12:17:56 Tower kernel: ? _raw_spin_unlock_irqrestore+0xd/0xe
Feb 17 12:17:56 Tower kernel: ? paravirt_sched_clock+0x5/0x8
Feb 17 12:17:56 Tower kernel: ? timekeeping_get_ns+0x19/0x2f
Feb 17 12:17:56 Tower kernel: ? Sys_DumpTask+0xe9/0xf1 [corefreqk]
Feb 17 12:17:56 Tower kernel: Cycle_Skylake+0x53a/0x730 [corefreqk]
Feb 17 12:17:56 Tower kernel: __hrtimer_run_queues+0xb7/0x10b
Feb 17 12:17:56 Tower kernel: ? Cycle_Skylake_X+0x711/0x711 [corefreqk]
Feb 17 12:17:56 Tower kernel: hrtimer_interrupt+0x8d/0x15b
Feb 17 12:17:56 Tower kernel: __sysvec_apic_timer_interrupt+0x5d/0x68
Feb 17 12:17:56 Tower kernel: asm_call_irq_on_stack+0xf/0x20

logs20220217.txt logs20220216.txt