Weird CPU speed
-
@stephenw10 But why is the damn thing running at so high frequency when it's not doing anything?
-
Speedshift responds orders of magnitude faster than speedstep. When I was first testing it I had problems seeing what it's doing because simply running the command to report it bumps the CPU speed.
However I would not expect that if it's set to maximum efficiency.
-
@stephenw10 Hm, I have disabled Speed Shift, and now pfSense doesn't even display any frequency at all. That's not right I guess?
-
Mmm, perhaps SpeedStep (EIST) is not enabled in the BIOS then.
-
@stephenw10 Weirdly, the frequency info appeared after a few seconds. It also disappears after page refresh.
Speedstep is enabled. -
@stephenw10 Btw does the Intel Core somerthing thermal sensor option apply to any Intel CPU? I'm not sure what to choose: this or the generic ACPI value? Both seem to work.
-
The coretemp driver should work there.
Yes, with speedstep it generally only shows frequency info when the CPU is running at a reduced frequency.
Try running
sysctl -a | grep freq
again to see what levels are now shown. -
@stephenw10 said in Weird CPU speed:
sysctl -a | grep freq
Timecounter "HPET" frequency 19200000 Hz quality 950
Event timer "HPET" frequency 19200000 Hz quality 550
Event timer "HPET1" frequency 19200000 Hz quality 440
Event timer "HPET2" frequency 19200000 Hz quality 440
Event timer "HPET3" frequency 19200000 Hz quality 440
Event timer "HPET4" frequency 19200000 Hz quality 440
Event timer "RTC" frequency 32768 Hz quality 0
Timecounter "i8254" frequency 1193182 Hz quality 0
Event timer "i8254" frequency 1193182 Hz quality 100
Timecounter "ACPI-fast" frequency 3579545 Hz quality 900
Timecounter "TSC" frequency 806401669 Hz quality 1000
Timecounter "HPET" frequency 19200000 Hz quality 950
Event timer "HPET" frequency 19200000 Hz quality 550
Event timer "HPET1" frequency 19200000 Hz quality 440
Event timer "HPET2" frequency 19200000 Hz quality 440
Event timer "HPET3" frequency 19200000 Hz quality 440
Event timer "HPET4" frequency 19200000 Hz quality 440
Event timer "RTC" frequency 32768 Hz quality 0
Timecounter "i8254" frequency 1193182 Hz quality 0
Event timer "i8254" frequency 1193182 Hz quality 100
Timecounter "ACPI-fast" frequency 3579545 Hz quality 900
Timecounter "TSC" frequency 806402450 Hz quality 1000
Timecounter "HPET" frequency 19200000 Hz quality 950
Event timer "HPET" frequency 19200000 Hz quality 550
Event timer "HPET1" frequency 19200000 Hz quality 440
Event timer "HPET2" frequency 19200000 Hz quality 440
Event timer "HPET3" frequency 19200000 Hz quality 440
Event timer "HPET4" frequency 19200000 Hz quality 440
Event timer "RTC" frequency 32768 Hz quality 0
Timecounter "i8254" frequency 1193182 Hz quality 0
Event timer "i8254" frequency 1193182 Hz quality 100
Timecounter "ACPI-fast" frequency 3579545 Hz quality 900
Timecounter "TSC" frequency 806401594 Hz quality 1000
Timecounter "HPET" frequency 19200000 Hz quality 950
Event timer "HPET" frequency 19200000 Hz quality 550
Event timer "HPET1" frequency 19200000 Hz quality 440
Event timer "HPET2" frequency 19200000 Hz quality 440
Event timer "HPET3" frequency 19200000 Hz quality 440
Event timer "HPET4" frequency 19200000 Hz quality 440
Event timer "RTC" frequency 32768 Hz quality 0
Timecounter "i8254" frequency 1193182 Hz quality 0
Event timer "i8254" frequency 1193182 Hz quality 100
Timecounter "ACPI-fast" frequency 3579545 Hz quality 900
Timecounter "TSC" frequency 806402560 Hz quality 1000
kern.timecounter.tc.TSC.frequency: 806402560
kern.timecounter.tc.ACPI-fast.frequency: 3579545
kern.timecounter.tc.i8254.frequency: 1193182
kern.timecounter.tc.HPET.frequency: 19200000
kern.ntp_pll.time_freq: 5467471872000
kern.ntp_pll.pps_freq: 5467471872000
device cpufreq
kern.eventtimer.et.i8254.frequency: 1193182
kern.eventtimer.et.RTC.frequency: 32768
kern.eventtimer.et.HPET4.frequency: 19200000
kern.eventtimer.et.HPET3.frequency: 19200000
kern.eventtimer.et.HPET2.frequency: 19200000
kern.eventtimer.et.HPET1.frequency: 19200000
kern.eventtimer.et.HPET.frequency: 19200000
kern.eventtimer.et.LAPIC.frequency: 806402560
kern.acct_chkfreq: 15
debug.cpufreq.verbose: 0
debug.cpufreq.lowest: 0
debug.uart_poll_freq: 50
machdep.tsc_freq: 806402560
machdep.i8254_freq: 1193182
machdep.acpi_timer_freq: 3579545
dev.cpufreq.3.freq_driver: est3
dev.cpufreq.3.%parent: cpu3
dev.cpufreq.3.%pnpinfo:
dev.cpufreq.3.%location:
dev.cpufreq.3.%driver: cpufreq
dev.cpufreq.3.%desc:
dev.cpufreq.2.freq_driver: est2
dev.cpufreq.2.%parent: cpu2
dev.cpufreq.2.%pnpinfo:
dev.cpufreq.2.%location:
dev.cpufreq.2.%driver: cpufreq
dev.cpufreq.2.%desc:
dev.cpufreq.1.freq_driver: est1
dev.cpufreq.1.%parent: cpu1
dev.cpufreq.1.%pnpinfo:
dev.cpufreq.1.%location:
dev.cpufreq.1.%driver: cpufreq
dev.cpufreq.1.%desc:
dev.cpufreq.0.freq_driver: est0
dev.cpufreq.0.%parent: cpu0
dev.cpufreq.0.%pnpinfo:
dev.cpufreq.0.%location:
dev.cpufreq.0.%driver: cpufreq
dev.cpufreq.0.%desc:
dev.cpufreq.%parent:
dev.est.3.freq_settings: 801/6000 800/6000 700/5160
dev.est.2.freq_settings: 801/6000 800/6000 700/5160
dev.est.1.freq_settings: 801/6000 800/6000 700/5160
dev.est.0.freq_settings: 801/6000 800/6000 700/5160
dev.cpu.3.freq_levels: 801/6000 800/6000 700/5160
dev.cpu.3.freq: 801
dev.cpu.2.freq_levels: 801/6000 800/6000 700/5160
dev.cpu.2.freq: 801
dev.cpu.1.freq_levels: 801/6000 800/6000 700/5160
dev.cpu.1.freq: 801
dev.cpu.0.freq_levels: 801/6000 800/6000 700/5160No idea what part of it is relevant.
PowerD is disabled if it matters. -
It must have been enabled at some point because the est(4) driver, speedstep, is loaded. The levels it's reporting, which are passed by the BIOS, are not useful though.
-
@stephenw10 Yes it was. I previously flipped several settings around.
What baffles me is the reported speed. And there's no way to tell whether it's real or not. -
Which one?
The speedstep levels there could be correct just not helpful. 700MHz and 800MHz are just that. 801MHz implies 800+turbo which could be anything, speedstep doesn't see the actual turbo speeds.
-
@stephenw10 The 1700-1900MHz one. That's with Speed Shift.
-
Speedshift does report the actual speed so I'd expect that to be real.
-
@stephenw10 Hmm, ok. But then something is not right anyway, because 1700MHz when there's less than 1% load.... that's just weird.
-
That's unexpected if it set to maximum efficiency for sure. At higher performance levels it might be since as I said just opening the dashboard would ramp it up significantly.
-
@stephenw10 I tried both extremes, and at max performance it went even higher, but never lower than 1700. But again, I don't know if it's the real speed, because the BIOS might be sketchy.
-
Mmm, I guess I'd be reading through the epic length threads there are for those boxes on the serverthehome forums. Seems like people have done a lot of investigation there:
https://forums.servethehome.com/index.php?threads/cwwk-topton-nxxx-quad-nic-router.39685/ -
@Octopuss I believe that the machine that I have is identical, and I had similar problems. I was ultimately able to resolve them by loading a modded BIOS that exposed more settings than the one it came loaded with. I believe it was the PL1 setting that made the difference, but I'm not entirely sure. But I would direct you to this thread that I started:
https://forum.netgate.com/topic/186104/topton-n100-reporting-402-mhz -
@TheNarc Thanks, I'll take a look.
But modding a BIOS... I'm not sure about that. But do you think the BIOS isn't bugged then, technically speaking? -
@Octopuss The way I look at, it's no more or less suspect than the BIOSes that come with these boxes. The way I think about it is that large companies (and even small companies) are not buying and using these boxes for business applications, so they do not present as an attractive target in that respect. So I feel that pretty much leaves "botnet nodes" as the most likely rationale for someone to compromise them. But there are far easier ways to amass botnet nodes than to make nefarious modifications to the BIOSes of niche devices used by some pfSense home users, then post them in a random forum thread and hope that a few dozen people use them. In my mind, the cost/benefit analysis just doesn't make sense for someone looking to attack/compromise vulnerable devices. But of course you have to come to your own decision.