The efficiency of suspend-to-idle depends on being able to keep CPUs
in the deepest available idle states for as much time as possible.
Ideally, they should only be brought out of idle by system wakeup
interrupts.

However, timer interrupts occurring periodically prevent that from
happening and it is not practical to chase all of the "misbehaving"
timers in a whack-a-mole fashion.  A much more effective approach is
to suspend the local ticks for all CPUs and the entire timekeeping
along the lines of what is done during full suspend, which also
helps to keep suspend-to-idle and full suspend reasonably similar.

The idea is to suspend the local tick on each CPU executing
cpuidle_enter_freeze() and to make the last of them suspend the
entire timekeeping.  That should prevent timer interrupts from
triggering until an IO interrupt wakes up one of the CPUs.  It
needs to be done with interrupts disabled on all of the CPUs,
though, because otherwise the suspended clocksource might be
accessed by an interrupt handler which might lead to fatal
consequences.

Unfortunately, the existing ->enter callbacks provided by cpuidle
drivers generally cannot be used for implementing that, because some
of them re-enable interrupts temporarily and some idle entry methods
cause interrupts to be re-enabled automatically on exit.  Also some
of these callbacks manipulate local clock event devices of the CPUs
which really shouldn't be done after suspending their ticks.

To overcome that difficulty, introduce a new cpuidle state callback,
->enter_freeze, that will be guaranteed (1) to keep interrupts
disabled all the time (and return with interrupts disabled) and (2)
not to touch the CPU timer devices.  Modify cpuidle_enter_freeze() to
look for the deepest available idle state with ->enter_freeze present
and to make the CPU execute that callback with suspended tick (and the
last of the online CPUs to execute it with suspended timekeeping).
Suggested-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NPeter Zijlstra (Intel) <peterz@infradead.org>

Theoretically, ktime_get_mono_fast_ns() may be executed after
timekeeping has been suspended (or before it is resumed) which
in turn may lead to undefined behavior, for example, when the
clocksource read from timekeeping_get_ns() called by it is
not accessible at that time.

Prevent that from happening by setting up a dummy readout base for
the fast timekeeper during timekeeping_suspend() such that it will
always return the same number of cycles.

After the last timekeeping_update() in timekeeping_suspend() the
clocksource is read and the result is stored as cycles_at_suspend.
The readout base from the current timekeeper is copied onto the
dummy and the ->read pointer of the dummy is set to a routine
unconditionally returning cycles_at_suspend.  Next, the dummy is
passed to update_fast_timekeeper().

Then, ktime_get_mono_fast_ns() will work until the subsequent
timekeeping_resume() and the proper readout base for the fast
timekeeper will be restored by the timekeeping_update() called
right after clearing timekeeping_suspended.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NJohn Stultz <john.stultz@linaro.org>
Acked-by: NPeter Zijlstra (Intel) <peterz@infradead.org>

Modify update_fast_timekeeper() to take a struct tk_read_base
pointer as its argument (instead of a struct timekeeper pointer)
and update its kerneldoc comment to reflect that.

That will allow a struct tk_read_base that is not part of a
struct timekeeper to be passed to it in the next patch.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: NJohn Stultz <john.stultz@linaro.org>

Adds a timespec64 based getboottime64() implementation
that can be used as we convert internal users of
getboottime away from using timespecs.

Cc: pang.xunlei <pang.xunlei@linaro.org>
Cc: Arnd Bergmann <arnd.bergmann@linaro.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

In commit 6067dc5a ("time: Avoid possible NTP adjustment
mult overflow") a new check was added to watch for adjustments
that could cause a mult overflow.

Unfortunately the check compares a signed with unsigned value
and ignored the case where the adjustment was negative, which
causes spurious warn-ons on some systems (and seems like it
would result in problematic time adjustments there as well, due
to the early return).

Thus this patch adds a check to make sure the adjustment is
positive before we check for an overflow, and resovles the issue
in my testing.
Reported-by: NFengguang Wu <fengguang.wu@intel.com>
Debugged-by: Npang.xunlei <pang.xunlei@linaro.org>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>
Link: http://lkml.kernel.org/r/1416890145-30048-1-git-send-email-john.stultz@linaro.orgSigned-off-by: NIngo Molnar <mingo@kernel.org>

Fix up a few comments that weren't updated when the
functions were converted to use timespec64 structures.
Acked-by: NArnd Bergmann <arnd.bergmann@linaro.org>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Adds a timespec64 based get_monotonic_coarse64() implementation
that can be used as we convert internal users of
get_monotonic_coarse away from using timespecs.
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Adds a timespec64 based getrawmonotonic64() implementation
that can be used as we convert internal users of
getrawmonotonic away from using timespecs.
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

As part of addressing "y2038 problem" for in-kernel uses, this
patch adds timekeeping_inject_sleeptime64() using timespec64.

After this patch, timekeeping_inject_sleeptime() is deprecated
and all its call sites will be fixed using the new interface,
after that it can be removed.

NOTE: timekeeping_inject_sleeptime() is safe actually, but we
want to eliminate timespec eventually, so comes this patch.
Signed-off-by: Npang.xunlei <pang.xunlei@linaro.org>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

The kernel uses 32-bit signed value(time_t) for seconds elapsed
1970-01-01:00:00:00, thus it will overflow at 2038-01-19 03:14:08
on 32-bit systems. This is widely known as the y2038 problem.

As part of addressing "y2038 problem" for in-kernel uses, this patch
adds safe do_settimeofday64() using timespec64.

After this patch, do_settimeofday() is deprecated and all its call
sites will be fixed using do_settimeofday64(), after that it can be
removed.
Signed-off-by: Npang.xunlei <pang.xunlei@linaro.org>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

The clocksource mult-adjustment threshold is [mult-maxadj, mult+maxadj],
timekeeping_adjust() only deals with the upper threshold, but misses the
lower threshold.

This patch adds the lower threshold judging condition.
Signed-off-by: Npang.xunlei <pang.xunlei@linaro.org>
[jstultz: Minor fix for > 80 char line]
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Ideally, __clocksource_updatefreq_scale, selects the largest shift
value possible for a clocksource. This results in the mult memember of
struct clocksource being particularly large, although not so large
that NTP would adjust the clock to cause it to overflow.

That said, nothing actually prohibits an overflow from occuring, its
just that it "shouldn't" occur.

So while very unlikely, and so far never observed, the value of
(cs->mult+cs->maxadj) may have a chance to reach very near 0xFFFFFFFF,
so there is a possibility it may overflow when doing NTP positive
adjustment

See the following detail: When NTP slewes the clock, kernel goes
through update_wall_time()->...->timekeeping_apply_adjustment():
	tk->tkr.mult += mult_adj;

Since there is no guard against it, its possible tk->tkr.mult may
overflow during this operation.

This patch avoids any possible mult overflow by judging the overflow
case before adding mult_adj to mult, also adds the WARNING message
when capturing such case.
Signed-off-by: Npang.xunlei <pang.xunlei@linaro.org>
[jstultz: Reworded commit message]
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

ktime_get_real_seconds() is the replacement function for get_seconds()
returning the seconds portion of CLOCK_REALTIME in a time64_t. For
64bit the function is equivivalent to get_seconds(), but for 32bit it
protects the readout with the timekeeper sequence count. This is
required because 32-bit machines cannot access 64-bit tk->xtime_sec
variable atomically.

[tglx: Massaged changelog and added docbook comment ]
Signed-off-by: NHeena Sirwani <heenasirwani@gmail.com>
Reviewed-by: NArnd Bergman <arnd@arndb.de>
Cc: John Stultz <john.stultz@linaro.org>
Cc: opw-kernel@googlegroups.com
Link: http://lkml.kernel.org/r/7adcfaa8962b8ad58785d9a2456c3f77d93c0ffb.1414578445.git.heenasirwani@gmail.comSigned-off-by: NThomas Gleixner <tglx@linutronix.de>

This is the counterpart to get_seconds() based on CLOCK_MONOTONIC. The
use case for this interface are kernel internal coarse grained
timestamps which do neither require the nanoseconds fraction of
current time nor the CLOCK_REALTIME properties. Such timestamps can
currently only retrieved by calling ktime_get_ts64() and using the
tv_sec field of the returned timespec64. That's inefficient as it
involves the read of the clocksource, math operations and must be
protected by the timekeeper sequence counter.

To avoid the sequence counter protection we restrict the return value
to unsigned 32bit on 32bit machines. This covers ~136 years of uptime
and therefor an overflow is not expected to hit anytime soon.

To avoid math in the function we calculate the current seconds portion
of CLOCK_MONOTONIC when the timekeeper gets updated in
tk_update_ktime_data() similar to the CLOCK_REALTIME counterpart
xtime_sec.

[ tglx: Massaged changelog, simplified and commented the update
  	function, added docbook comment ]
Signed-off-by: NHeena Sirwani <heenasirwani@gmail.com>
Reviewed-by: NArnd Bergman <arnd@arndb.de>
Cc: John Stultz <john.stultz@linaro.org>
Cc: opw-kernel@googlegroups.com
Link: http://lkml.kernel.org/r/da0b63f4bdf3478909f92becb35861197da3a905.1414578445.git.heenasirwani@gmail.comSigned-off-by: NThomas Gleixner <tglx@linutronix.de>

The update_walltime() code works on the shadow timekeeper to make the
seqcount protected region as short as possible. But that update to the
shadow timekeeper does not update all timekeeper fields because it's
sufficient to do that once before it becomes life. One of these fields
is tkr.base_mono. That stays stale in the shadow timekeeper unless an
operation happens which copies the real timekeeper to the shadow.

The update function is called after the update calls to vsyscall and
pvclock. While not correct, it did not cause any problems because none
of the invoked update functions used base_mono.

commit cbcf2dd3 (x86: kvm: Make kvm_get_time_and_clockread()
nanoseconds based) changed that in the kvm pvclock update function, so
the stale mono_base value got used and caused kvm-clock to malfunction.

Put the update where it belongs and fix the issue.
Reported-by: NChris J Arges <chris.j.arges@canonical.com>
Reported-by: NPaolo Bonzini <pbonzini@redhat.com>
Cc: Gleb Natapov <gleb@kernel.org>
Cc: John Stultz <john.stultz@linaro.org>
Link: http://lkml.kernel.org/r/alpine.DEB.2.10.1409050000570.3333@nanosSigned-off-by: NThomas Gleixner <tglx@linutronix.de>

Benjamin Herrenschmidt pointed out that I further missed modifying
update_vsyscall after the wall_to_mono value was changed to a
timespec64.  This causes issues on powerpc32, which expects a 32bit
timespec.

This patch fixes the problem by properly converting from a timespec64 to
a timespec before passing the value on to the arch-specific vsyscall
logic.

[ Thomas is currently on vacation, but reviewed it and wanted me to send
  this fix on to you directly. ]

Cc: LKML <linux-kernel@vger.kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Reported-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
Reviewed-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

By caching the ntp_tick_length() when we correct the frequency error,
and then using that cached value to accumulate error, we avoid large
initial errors when the tick length is changed.

This makes convergence happen much faster in the simulator, since the
initial error doesn't have to be slowly whittled away.

This initially seems like an accounting error, but Miroslav pointed out
that ntp_tick_length() can change mid-tick, so when we apply it in the
error accumulation, we are applying any recent change to the entire tick.

This approach chooses to apply changes in the ntp_tick_length() only to
the next tick, which allows us to calculate the freq correction before
using the new tick length, which avoids accummulating error.

Credit to Miroslav for pointing this out and providing the original patch
this functionality has been pulled out from, along with the rational.

Cc: Miroslav Lichvar <mlichvar@redhat.com>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Prarit Bhargava <prarit@redhat.com>
Reported-by: NMiroslav Lichvar <mlichvar@redhat.com>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

The existing timekeeping_adjust logic has always been complicated
to understand. Further, since it was developed prior to NOHZ becoming
common, its not surprising it performs poorly when NOHZ is enabled.

Since Miroslav pointed out the problematic nature of the existing code
in the NOHZ case, I've tried to refactor the code to perform better.

The problem with the previous approach was that it tried to adjust
for the total cumulative error using a scaled dampening factor. This
resulted in large errors to be corrected slowly, while small errors
were corrected quickly. With NOHZ the timekeeping code doesn't know
how far out the next tick will be, so this results in bad
over-correction to small errors, and insufficient correction to large
errors.

Inspired by Miroslav's patch, I've refactored the code to try to
address the correction in two steps.

1) Check the future freq error for the next tick, and if the frequency
error is large, try to make sure we correct it so it doesn't cause
much accumulated error.

2) Then make a small single unit adjustment to correct any cumulative
error that has collected over time.

This method performs fairly well in the simulator Miroslav created.

Major credit to Miroslav for pointing out the issue, providing the
original patch to resolve this, a simulator for testing, as well as
helping debug and resolve issues in my implementation so that it
performed closer to his original implementation.

Cc: Miroslav Lichvar <mlichvar@redhat.com>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Prarit Bhargava <prarit@redhat.com>
Reported-by: NMiroslav Lichvar <mlichvar@redhat.com>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

In the GENERIC_TIME_VSYSCALL_OLD update_vsyscall implementation,
we take the tk_xtime() value, which returns a timespec64, and
store it in a timespec.

This luckily is ok, since the only architectures that use
GENERIC_TIME_VSYSCALL_OLD are ia64 and ppc64, which are both
64 bit systems where timespec64 is the same as a timespec.

Even so, for cleanliness reasons, use the conversion function
to assign the proper type.
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Tracers want a correlated time between the kernel instrumentation and
user space. We really do not want to export sched_clock() to user
space, so we need to provide something sensible for this.

Using separate data structures with an non blocking sequence count
based update mechanism allows us to do that. The data structure
required for the readout has a sequence counter and two copies of the
timekeeping data.

On the update side:

  smp_wmb();
  tkf->seq++;
  smp_wmb();
  update(tkf->base[0], tk);
  smp_wmb();
  tkf->seq++;
  smp_wmb();
  update(tkf->base[1], tk);

On the reader side:

  do {
     seq = tkf->seq;
     smp_rmb();
     idx = seq & 0x01;
     now = now(tkf->base[idx]);
     smp_rmb();
  } while (seq != tkf->seq)

So if a NMI hits the update of base[0] it will use base[1] which is
still consistent, but this timestamp is not guaranteed to be monotonic
across an update.

The timestamp is calculated by:

	now = base_mono + clock_delta * slope

So if the update lowers the slope, readers who are forced to the
not yet updated second array are still using the old steeper slope.

 tmono
 ^
 |    o  n
 |   o n
 |  u
 | o
 |o
 |12345678---> reader order

 o = old slope
 u = update
 n = new slope

So reader 6 will observe time going backwards versus reader 5.

While other CPUs are likely to be able observe that, the only way
for a CPU local observation is when an NMI hits in the middle of
the update. Timestamps taken from that NMI context might be ahead
of the following timestamps. Callers need to be aware of that and
deal with it.

V2: Got rid of clock monotonic raw and reorganized the data
    structures. Folded in the barrier fix from Mathieu.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

All the function needs is in the tk_read_base struct. No functional
change for the current code, just a preparatory patch for the NMI safe
accessor to clock monotonic which will use struct tk_read_base as well.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

The members of the new struct are the required ones for the new NMI
safe accessor to clcok monotonic. In order to reuse the existing
timekeeping code and to make the update of the fast NMI safe
timekeepers a simple memcpy use the struct for the timekeeper as well
and convert all users.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Access to time requires to touch two cachelines at minimum

   1) The timekeeper data structure

   2) The clocksource data structure

The access to the clocksource data structure can be avoided as almost
all clocksource implementations ignore the argument to the read
callback, which is a pointer to the clocksource.

But the core needs to touch it to access the members @read and @mask.

So we are better off by copying the @read function pointer and the
@mask from the clocksource to the core data structure itself.

For the most used ktime_get() access all required data including the
@read and @mask copies fits together with the sequence counter into a
single 64 byte cacheline.

For the other time access functions we touch in the current code three
cache lines in the worst case. But with the clocksource data copies we
can reduce that to two adjacent cachelines, which is more efficient
than disjunct cache lines.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

cycle_last was added to the clocksource to support the TSC
validation. We moved that to the core code, so we can get rid of the
extra copy.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

We want to move the TSC sanity check into core code to make NMI safe
accessors to clock monotonic[_raw] possible. For this we need to
sanity check the delta calculation. Create a helper function and
convert all sites to use it.

[ Build fix from jstultz ]
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Provide a ktime_t based interface for raw monotonic time.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

timekeeping_clocktai() is not used in fast pathes, so the extra
timespec conversion is not problematic.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

No more users. Remove it
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Subtracting plain nsec values and converting to timespec is simpler
than the whole timespec math. Not really fastpath code, so the
division is not an issue.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

get_monotonic_boottime() is not used in fast pathes, so the extra
timespec conversion is not problematic.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

No more users.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

No more users.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

ktime based conversion function to map a monotonic time stamp to a
different CLOCK.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

No need to juggle with timespecs.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

No need to juggle with timespecs.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Speed up the readout.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Provide a helper function which lets us implement ktime_t based
interfaces for real, boot and tai clocks.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>

Speed up ktime_get() by using ktime_t based data. Text size shrinks by
64 bytes on x8664.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NJohn Stultz <john.stultz@linaro.org>