Re: PSA: Clock drift and pkgin

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

> How well can we trust your emulator to be fielding the correct number
> of interrupts?

Depending on what you mean by "fielding", that's one of the possible
sources of trouble.

The underlying hardware for that test was my mutant 5.2 on two cores of

cpu0: Intel Core 2 (Merom) (686-class), 2000.64 MHz, id 0x6f2
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 0xe39d<SSE3,DTES64,MONITOR,DS-CPL,EST,TM2,SSSE3,CX16,xTPR,PDCM>
cpu0: features3 0x20100800<SYSCALL/SYSRET,XD,EM64T>
cpu0: "Intel(R) Core(TM)2 CPU          4400  @ 2.00GHz"
cpu0: I-cache 32KB 64B/line 8-way, D-cache 32KB 64B/line 8-way
cpu0: L2 cache 2MB 64B/line 8-way
cpu0: ITLB 128 4KB entries 4-way
cpu0: DTLB 256 4KB entries 4-way, 16 4MB entries 4-way
cpu0: Initial APIC ID 0
cpu0: Cluster/Package ID 0
cpu0: Core ID 0
cpu0: family 06 model 0f extfamily 00 extmodel 00 stepping 02

with very little else going on, certainly not enough to keep more than
part of one core busy.  ICCS interrupts are handled by a child process,
which reads from the timer socket and pokes the main process with a
pipe and a signal; the main loop of the emulator then notices a
volatile variable set by the signal handler and reads from the pipe.

There are lots of places there that ticks could get lost.  That's why I
want to instrument everything.

> 40s in 1 hour is a horrible drift, and for a system not doing
> anything that seems extreme.  I don't have the feeling that a real
> system have that much drift, which makes me wonder about the
> emulator...

It's admittedly been a long time since I ran a real KA630, but what
fuzzy memories I do have indicate that it is not unlike what I saw back
when I did.  I think it's one reason I didn't run the VAX more despite
the emotional attachment I have to VAXen.

At least with the emulator crystal drift is not an issue! :-)  (The
underlying OS has stratum-3 NTP time sync, so unless _its_ kernel has
interrupt handling issues, which I see no reason to think is so, its
100Hz should be very close to true 100Hz.)

/~\ 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