Re: Clock drift and other open issues: Collecting information

simh VAXen running NetBSD have a horrible time keeping for me at least.

   Johnny

On 2023-12-30 22:22, Jan-Benedict Glaw wrote:
> On Sat, 2023-12-30 16:54:53 +0100, Johnny Billquist <[email protected]> wrote:
>> My only clock issue right now seems to be simh related.
> 
> I'm a long-term microvax3900 SIMH user, and tried the 4000/60.
> Installation works, but after reboot:
> 
> KA46-A V1.4-38E-V4.2
> 08-00-2B-A0-0F-0B
> 48MB
> |                                  |
> ####################################
> ?? 001   3        DZ  0032
> ?? 001   4     CACHE  0768
> ?? 001   7        IT  1357
> ?? 001   8       SYS  0128
> ?? 001   9        NI  0024
> 
>>>> sh dev
> 
>    VMS/VMB      ADDR      DEVTYPE    NUMBYTES     RM/FX    WP    DEVNAM      REV
>    -------      ----      -------    --------     -----    --    ------      ---
>    ESA0         08-00-2B-A0-0F-0B
>    DKA0         A/0/0     DISK         2.04GB      RM            RZUSER      0001
>   ..HostID..    A/6       INITR
> 
> 
> 
>>>> b dka0
> 
>   
> -DKA0
>   
> 
> ?43 FILESTRUCT
> HALT instruction, PC: 00000B15 (MOVL (R11),SP)
> Goodbye
> %SIM-INFO: Eth: closed trap_21
> 
> 
> 
> cat run.simh
> SHOW    VERSION
> SET     CPU     48m
> SET     CPU     SIMHALT
> SET     CPU     idle=netbsd
> SET     RZ0     FORMAT=VHD
> SET     RZ0     rzuser
> ATTACH  RZ0     disk.vhd
> SET     RZ1     DISABLE
> SET     RZ2     DISABLE
> SET     RZ3     DISABLE
> SET     RZ4     DISABLE
> SET     RZ5     DISABLE
> SET     RZ6     DISABLE
> SET     RZ7     DISABLE
> SET     NAR     mac=08-00-2b-a0-0f-0b
> ATTACH  XS      tap:trap_21
> ATTACH  nvr     nvram.bin
> DEP     BDR     0
> SHOW    CONFIG
> 
> BOOT    CPU
> EXIT
> 
> 
> (During installation, RZ1 was the cdrom holding the install ISO.)
> That's not a perfect configuration, but ... anybody has a specific
> clue what prevents it from booting?
> 
> MfG, JBG
> NB: A SIMH microvax3900 right now, booted some 20min ago:
> 
> # ntpq -n -p
>       remote           refid      st t when poll reach   delay   offset  jitter
> ==============================================================================
>   2.netbsd.pool.n .POOL.          16 p    -   64    0    0.000   +0.000   7.813
> -88.198.200.96   193.203.3.171    2 u   60   64  377    9.944  -15.422  48.689
> -144.76.66.157   192.53.103.103   2 u   58   64  377    7.813  -22.565   9.526
> +212.112.228.242 237.17.204.95    2 u   54   64  377    7.813  -24.728  12.268
> *130.61.44.4     17.253.14.251    2 u   61   64  377    7.813  -21.598   9.577
> +81.7.16.52      131.188.3.222    2 u   57   64  377    9.850   -0.378  26.232
> 
> # uptime
>   9:19PM  up 23 mins, 1 user, load averages: 0.11, 0.06, 0.03
> 
> # ntptime
> ntp_gettime() returns code 0 (OK)
>    time e93b07ef.0e01eef0  Sat, Dec 30 2023 21:19:43.054, (.054717922),
>    maximum error 124192 us, estimated error 20412 us, TAI offset 0
> ntp_adjtime() returns code 0 (OK)
>    modes 0x0 (),
>    offset -11794.748 us, frequency 72.944 ppm, interval 1 s,
>    maximum error 124192 us, estimated error 20412 us,
>    status 0x2001 (PLL,NANO),
>    time constant 6, precision 0.001 us, tolerance 496 ppm,
> 
> 
> Not too bad. But ISTR that the real /90 and /60 both hat quite stable
> frequencies at about 37ish or so, here it's higher... I'll let it run
> for some time (with NetBSD idle detection) and see what happens.
> (Maybe idle detection plays a role, too. Just sayin'.)
> 

-- 
Johnny Billquist                  || "I'm on a bus
                                   ||  on a psychedelic trip
email: [email protected]             ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol