On Fri, Oct 01, 2021 at 01:02:23AM +0200, Thomas Gleixner wrote:
> Marcelo,
>
> On Thu, Sep 30 2021 at 16:21, Marcelo Tosatti wrote:
> > On Wed, Sep 29, 2021 at 03:56:29PM -0300, Marcelo Tosatti wrote:
> >> On Thu, Sep 16, 2021 at 06:15:35PM +0000, Oliver Upton wrote:
> >>
> >> Thomas, CC'ed, has deeper understanding of problems with
> >> forward time jumps than I do. Thomas, any comments?
> >
> > Based on the earlier discussion about the problems of synchronizing
> > the guests clock via a notification to the NTP/Chrony daemon
> > (where there is a window where applications can read the stale
> > value of the clock), a possible solution would be triggering
> > an NMI on the destination (so that it runs ASAP, with higher
> > priority than application/kernel).
> >
> > What would this NMI do, exactly?
>
> Nothing. You cannot do anything time related in an NMI.
>
> You might queue irq work which handles that, but that would still not
> prevent user space or kernel space from observing the stale time stamp
> depending on the execution state from where it resumes.
Yes.
> >> As a note: this makes it not OK to use KVM_CLOCK_REALTIME flag
> >> for either vm pause / vm resume (well, if paused for long periods of time)
> >> or savevm / restorevm.
> >
> > Maybe with the NMI above, it would be possible to use
> > the realtime clock as a way to know time elapsed between
> > events and advance guest clock without the current
> > problematic window.
>
> As much duct tape you throw at the problem, it cannot be solved ever
> because it's fundamentally broken. All you can do is to make the
> observation windows smaller, but that's just curing the symptom.
Yes.
> The problem is that the guest is paused/resumed without getting any
> information about that and the execution of the guest is stopped at an
> arbitrary instruction boundary from which it resumes after migration or
> restore. So there is no way to guarantee that after resume all vCPUs are
> executing in a state which can handle that.
>
> But even if that would be the case, then what prevents the stale time
> stamps to be visible? Nothing:
>
> T0: t = now();
> -> pause
> -> resume
> -> magic "fixup"
> T1: dostuff(t);
Yes.
BTW, you could have a userspace notification (then applications
could handle this if desired).
> But that's not a fundamental problem because every preemptible or
> interruptible code has the same issue:
>
> T0: t = now();
> -> preemption or interrupt
> T1: dostuff(t);
>
> Which is usually not a problem, but It becomes a problem when T1 - T0 is
> greater than the usual expectations which can obviously be trivially
> achieved by guest migration or a savevm, restorevm cycle.
>
> But let's go a step back and look at the clocks and their expectations:
>
> CLOCK_MONOTONIC:
>
> Monotonically increasing clock which counts unless the system
> is in suspend. On resume it continues counting without jumping
> forward.
>
> That's the reference clock for everything else and therefore it
> is important that it does _not_ jump around.
>
> The reasons why CLOCK_MONOTONIC stops during suspend is
> historical and any attempt to change that breaks the world and
> some more because making it jump forward will trigger all sorts
> of timeouts, watchdogs and whatever. The last attempt to make
> CLOCK_MONOTONIC behave like CLOCK_BOOTTIME was reverted within 3
> weeks. It's not going to be attempted again. See a3ed0e4393d6
> ("Revert: Unify CLOCK_MONOTONIC and CLOCK_BOOTTIME") for
> details.
>
> Now the proposed change is creating exactly the same problem:
>
> >> > + if (data.flags & KVM_CLOCK_REALTIME) {
> >> > + u64 now_real_ns = ktime_get_real_ns();
> >> > +
> >> > + /*
> >> > + * Avoid stepping the kvmclock backwards.
> >> > + */
> >> > + if (now_real_ns > data.realtime)
> >> > + data.clock += now_real_ns - data.realtime;
> >> > + }
>
> IOW, it takes the time between pause and resume into account and
> forwards the underlying base clock which makes CLOCK_MONOTONIC
> jump forward by exactly that amount of time.
Well, it is assuming that the
T0: t = now();
T1: pause vm()
T2: finish vm migration()
T3: dostuff(t);
Interval between T1 and T2 is small (and that the guest
clocks are synchronized up to a given boundary).
But i suppose adding a limit to the forward clock advance
(in the migration case) is useful:
1) If migration (well actually, only the final steps
to finish migration, the time between when guest is paused
on source and is resumed on destination) takes too long,
then too bad: fix it to be shorter if you want the clocks
to have close to zero change to realtime on migration.
2) Avoid the other bugs in case of large forward advance.
Maybe having it configurable, with a say, 1 minute maximum by default
is a good choice?
An alternative would be to advance only the guests REALTIME clock, from
data about how long steps T1-T2 took.
> So depending on the size of the delta you are running into exactly the
> same problem as the final failed attempt to unify CLOCK_MONOTONIC and
> CLOCK_BOOTTIME which btw. would have been a magic cure for virt.
>
> Too bad, not going to happen ever unless you fix all affected user
> space and kernel side issues.
>
>
> CLOCK_BOOTTIME:
>
> CLOCK_MONOTONIC + time spent in suspend
>
>
> CLOCK_REALTIME/TAI:
>
> CLOCK_MONOTONIC + offset
>
> The offset is established by reading RTC at boot time and can be
> changed by clock_settime(2) and adjtimex(2). The latter is used
> by NTP/PTP.
>
> Any user of CLOCK_REALTIME has to be prepared for it to jump in
> both directions, but of course NTP/PTP daemons have expectations
> vs. such time jumps.
>
> They rightfully assume on a properly configured and administrated
> system that there are only two things which can make CLOCK_REALTIME
> jump:
>
> 1) NTP/PTP daemon controlled
> 2) Suspend/resume related updates by the kernel
>
>
> Just for the record, these assumptions predate virtualization.
>
> So now virt came along and created a hard to solve circular dependency
> problem:
>
> - If CLOCK_MONOTONIC stops for too long then NTP/PTP gets out of
> sync, but everything else is happy.
>
> - If CLOCK_MONOTONIC jumps too far forward, then all hell breaks
> lose, but NTP/PTP is happy.
One must handle the
T0: t = now();
-> pause
-> resume
-> magic "fixup"
T1: dostuff(t);
fact if one is going to use savevm/restorevm anyway, so...
(it is kind of unfixable, unless you modify your application
to accept notifications to redo any computation based on t, isnt it?).
> IOW, you are up a creek without a paddle and you have to chose one evil.
>
> That's simply a design fail because there has been no design for this
> from day one. But I'm not surprised at all by that simply because
> virtualization followed the hardware design fails vs. time and
> timekeeping which keep us entertained for the past 20 years on various
> architectures but most prominently on X86 which is the uncontended
> master of disaster in that regard.
>
> Of course virt follows the same approach of hardware by ignoring the
> problem and coming up with more duct tape and the assumption that lack
> of design can be "fixed in software". Just the timeframe is slightly
> different: We're discussing this only for about 10 years now.
>
> Seriously? All you folks can come up with in 10 years is piling duct
> tape on duct tape instead of sitting down and fixing the underlying root
> cause once and forever?
Been fixing bugs that are reported over 10+ years, yeah.
Hopefully this thread is the "sitting down and fixing the underyling root
cause" :-)
> I'm aware that especially chrony has tried to deal with this nonsense
> more gracefully, but that still does not make it great and it still gets
> upset.
>
> The reason why suspend/resume works perfectly fine is that it's fully
> coordinated and NTP state is cleared on resume which makes it easy for
> the deamon to accomodate.
This is what is in place now (which is executed on the destination):
/* Now, if user has passed a time to set and the system time is set, we
* just need to synchronize the hardware clock. However, if no time was
* passed, user is requesting the opposite: set the system time from the
* hardware clock (RTC). */
pid = fork();
if (pid == 0) {
setsid();
reopen_fd_to_null(0);
reopen_fd_to_null(1);
reopen_fd_to_null(2);
/* Use '/sbin/hwclock -w' to set RTC from the system time,
* or '/sbin/hwclock -s' to set the system time from RTC. */
execle(hwclock_path, "hwclock", has_time ? "-w" : "-s",
NULL, environ);
_exit(EXIT_FAILURE);
} else if (pid < 0) {
error_setg_errno(errp, errno, "failed to create child process");
return;
}
>
> So again and I'm telling this for a decade now:
>
> 1) Stop pretending that you can fix the lack of design with duct tape
> engineering
>
> 2) Accept the fundamental properties of Linux time keeping as they are
> not going to change as explained above
>
> 3) Either accept that CLOCK_REALTIME is off and jumping around which
> confuses NTP/PTP or get your act together and design and implement a
> proper synchronization mechanism for this:
>
> - Notify the guest about the intended "pause" or "savevm" event
>
> - Let the guest go into a lightweight freeze similar to S2IDLE
>
> - Save the VM for later resume or migrate the saved state
>
> - Watch everything working as expected on resume
>
> - Have the benefit that pause/resume and savevm/restorevm have
> exactly the same behaviour
OK!
> That won't solve the problem for frankenkernels and !paravirt setups,
> but that's unsolvable and you can keep the pieces by chosing one of two
> evils. While I do not care at all, I still recommend to chose
> CLOCK_MONOTONIC correctness for obvious reasons.
>
> The frankenkernel/legacy problem aside, I know you are going to tell me
> that this is too much overhead and has VM customer visible impact.
No, i think it boils down to someone implementing it.
> It's
> your choice, really:
>
> Either you chose correctness or you decide to ignore correctness for
> whatever reason.
>
> There is no middle ground simply because you _cannot_ guarantee that
> your migration time is within the acceptable limits of the
> CLOCK_MONOTONIC or the CLOCK_REALTIME expectations.
>
> You can limit the damage somehow by making some arbitrary cutoff of
> how much you forward CLOCK_MONOTONIC, but don't ask me about the right
> value.
> If you decide that correctness is overrated, then please document it
> clearly instead of trying to pretend being correct.
Based on the above, advancing only CLOCK_REALTIME (and not CLOCK_MONOTONIC)
would be correct, right? And its probably not very hard to do.
> I'm curious whether the hardware people or the virt folks come to senses
> first, but honestly I'm not expecting this to happen before I retire.
>
> Thanks,
>
> tglx
Thanks very much for the detailed information! Its a good basis
for the document you ask.
