On 2023-12-17 20:58, Johnny Billquist wrote:
On 2023-12-17 16:45, Paul Koning wrote:On Dec 16, 2023, at 9:16 PM, Johnny Billquist <bqt%softjar.se@localhost> wrote:For what it's worth, I think these lines also suggest something is wrong with the clock (this run under simh, simulating a VAX8650):[ 91.2854336] WARNING: lwp 854 (sh): negative runtime: (-1 + 0xfee8c21d5277ae5a/2^64) sec [ 91.2854336] WARNING: pid 854 (sh): negative runtime; monotonic clock has gone backwardsThat would be a different issue, though. If interrupts are lost or clock tick interrupt rates are not accurate, the system time will be wrong. But that would not cause the "monotonic" clock to go backwards. A supposed monotonic value going backwards is just a bug, no matter what the clock source is doing.For sure. But at least in my mind, this is what seems to be a problem more in general. The comments about the time moving all over the place, ntp having issues syncinc, and eventually just loosing sync altogether. That is more than lost interrupts. Time for processes reporting 106% is not a lost interrupt thing either. Something is seriously broken with time on VAX in general, I suspect. But as the above was observed running under simh, at this time I should not exclude bugs in simh as well.
By the way - another comment/reflection on this. I don't really understand how a monotonic clock could ever go backwards. That in itself seems to be something that would be a real bug no matter anything else. Independent of wether I would run it on simh or a real machine. Such a clock should only ever be possibly moving forward. But I won't exclude that maybe some bug in simh could cause this effect. It certainly needs more looking into. But right now I'm busy trying to just do a build if/when that succeeds, I plan to start looking at this some more if noone beats me to it.
Johnny
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt%softjar.se@localhost || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol