Re: [patch 5/6] [RFD] timekeeping: Provide optional 128bit math

[John Stultz]

On Thu, Dec 8, 2016 at 8:29 PM, Ingo Molnar <mingo@xxxxxxxxxx> wrote:
>
> * Ingo Molnar <mingo@xxxxxxxxxx> wrote:
>
>>
>> * Thomas Gleixner <tglx@xxxxxxxxxxxxx> wrote:
>>
>> > If the timekeeping CPU is scheduled out long enough by a hypervisor the
>> > clocksource delta multiplication can overflow and as a result time can go
>> > backwards. That's insane to begin with, but people already triggered a
>> > signed multiplication overflow, so a unsigned overflow is not necessarily
>> > impossible.
>> >
>> > Implement optional 128bit math which can be selected by a config option.
>>
>> What's the rough VM interruption time that would trigger an overflow? Given that
>> the clock shift tk_read_base::mult is often 1, isn't it 32-bit nsecs, i.e. 4
>> seconds?
>>
>> That doesn't sound 'insanely long'.
>>
>> Or some other value?
>
> Ok, wasn't fully awake yet: more realistic values of the scaling factor on x86
> would allow cycles input values of up to ~70 billion with 64-bit math, which would
> allow deltas of up to about 1 minute with 64-bit math.

So if I'm remembering properly, we pick mult/shift pairs such that the
mult shouldn't overflow from ~10 minutes worth of cycles.

> I think we should at least detect (and report?) the overflow and sanitize the
> effects to the max offset instead of generating random overflown values.

So with CONFIG_DEBUG_TIMEKEEPING, we do check to see if the cycle
value is larger then the max_cycles and will report a warning. But
this is done at interrupt time and not in the hotpath.

thanks
-john