Re: Time passing slowly

[Mouse <mouse%Rodents-Montreal.ORG@localhost>]

>> kern.timecounter.choice = TSC(q=3000, f=11664788060 Hz) clockinterrupt(q=0, f=100 Hz) hpet0(q=2000, f=23999999 Hz) ACPI-Fast(q=1000, f=3579545 Hz) lapic(q=-100, f=24040485 Hz) i8254(q=100, f=1193182 Hz) dummy(q=-1000000, f=1000000 Hz)

>> And, indeed, setting kern.timecounter.hardware=hpet0 makes the clock
>> start behaving itself.

> Which CPU is that?

cpu0 at mainbus0 apid 0: Intel 686-class, 11664MHz, id 0x506e3
cpu1 at mainbus0 apid 2: Intel 686-class, 11664MHz, id 0x506e3
cpu2 at mainbus0 apid 4: Intel 686-class, 11664MHz, id 0x506e3
cpu3 at mainbus0 apid 6: Intel 686-class, 11664MHz, id 0x506e3
cpu4 at mainbus0 apid 1: Intel 686-class, 11664MHz, id 0x506e3
cpu5 at mainbus0 apid 3: Intel 686-class, 11664MHz, id 0x506e3
cpu6 at mainbus0 apid 5: Intel 686-class, 11664MHz, id 0x506e3
cpu7 at mainbus0 apid 7: Intel 686-class, 11664MHz, id 0x506e3

[Pavilion - root] 12> cpuctl identify 0
cpu0: Intel Pentium Pro, II or III (686-class), 11664.83 MHz, id 0x506e3
cpu0: features 0xbfebfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR>
cpu0: features 0xbfebfbff<PGE,MCA,CMOV,PAT,PSE36,CFLUSH,DS,ACPI,MMX>
cpu0: features 0xbfebfbff<FXSR,SSE,SSE2,SS,HTT,TM,SBF>
cpu0: features2 0x77fafbbf<SSE3,DTES64,MONITOR,DS-CPL,VMX,EST,TM2,SSSE3,CX16,xTPR,PDCM,SSE41,SSE42,X2APIC,POPCNT>
cpu0: features3 0x2c100800<SYSCALL/SYSRET,XD,EM64T>
cpu0: "Intel(R) Core(TM) i7-6700HQ CPU @ 2.60GHz"
cpu0: DTLB 64 4KB entries 4-way
cpu0: Initial APIC ID 0
cpu0: Cluster/Package ID 0
cpu0: Core ID 0
cpu0: SMT ID 0
cpu0: family 06 model 0e extfamily 00 extmodel 05 stepping 03

/proc/cpuinfo has basically the same info in a different format.
Here's the stanza for CPU 0

processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 14
model name      : Intel(R) Core(TM) i7-6700HQ CPU @ 2.60GHz
stepping        : 3
cpu MHz         : 11664.83
fdiv_bug        : no
fpu             : yes
fpu_exception   : yes
cpuid level     : 22
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 syscall mmxext fxsr_opt rdtscp lm 3dnow recovery longrun lrti cxmmx cyrix_arr centaur_mcr constant_tsc pni monitor ds_cpi vmx est tm2 cx16 

and the rest are identical except for the processor number (of course)
and...the clock frequency.  Hmm!

cpu MHz         : 11664.83
cpu MHz         : 2607.9
cpu MHz         : 2632.6
cpu MHz         : 3642.10
cpu MHz         : 2853.72
cpu MHz         : 2637.23
cpu MHz         : 2632.60
cpu MHz         : 2844.62

Perhaps this is what's behind the problem.  Interestingly, the observed
clock ratio is pretty close to (probably within measurement error of)
the ratio between the first two clock frequences above.  I'm not sure
why it's claiming 11-plus gigahertz when the text clearly thinks it's a
2.6GHz CPU - actually, the ratio is also pretty close to the ratio
between the claimed 11664MHz and the differently-claimed 2.6GHz.

> TSC should be selected by default only, if the CPU says it is
> invariant, but I don't know from memory if that applies on NetBSD 5
> already.

I'm going to guess that either it doesn't or the invariance detection
is broken on this hardware (perhaps due to my tweaks, perhaps not).

> hpet and ACPI-fast are slower as they require (internal) PCI cycles;
> the LAPIC timer would be a good choice, except noone ever properly
> implemented the cross-CPU phase locking for SMP.

I'd rather have slow(er)-and-working than fast-and-broken, so I think
I'll leave it set on hpet0.  The machine is ludicrously overspecced for
anything I'm likely to try to do on it anyway.

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse%rodents-montreal.org@localhost
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B