Messages in this thread | | | From | "Jason A. Donenfeld" <> | Date | Wed, 12 Jun 2019 14:58:21 +0200 | Subject | Re: infinite loop in read_hpet from ktime_get_boot_fast_ns |
| |
Hi Peter,
Thanks for the explanation.
On Wed, Jun 12, 2019 at 2:29 PM Peter Zijlstra <peterz@infradead.org> wrote: > Either local_clock() or cpu_clock(cpu). The sleep hooks are not > something the consumer has to worry about.
Alright. Just so long as it *is* tracking sleep, then that's fine. If it isn't some important aspects of the protocol will be violated.
> If an architecture doesn't provide a sched_clock(), you're on a > seriously handicapped arch. It wraps in ~500 days, and aside from > changing jiffies_lock to a latch, I don't think we can do much about it.
Are you sure? The base definition I'm looking at uses jiffies:
unsigned long long __weak sched_clock(void) { return (unsigned long long)(jiffies - INITIAL_JIFFIES) * (NSEC_PER_SEC / HZ); }
On a CONFIG_HZ_1000 machine, jiffies wraps in ~49.7 days: >>> ((1<<32)-1)/1000/(60*60*24) 49.710269618055555
Why not just use get_jiffies_64()? The lock is too costly on 32bit?
> (the scheduler too expects sched_clock() to not wrap short of the u64 > and so having those machines online for 500 days will get you 'funny' > results)
Ahh. So if, on the other hand, the whole machine explodes at the wrap mark, I guess my silly protocol is the least of concerns, and so this shouldn't matter?
Jason
| |