The timekeeping duty is handed over from the outgoing CPU on stop
machine, then the oneshot tick is stopped right after. Therefore it's
guaranteed that the current CPU isn't the timekeeper upon its last call
to idle.
Besides, calling tick_nohz_idle_stop_tick() while the dying CPU goes
into idle suggests that the tick is going to be stopped while it is
actually stopped already from the appropriate CPU hotplug state.
Remove the confusing call and the obsolete case handling and convert it
to a sanity check that verifies the above assumption.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-16-frederic@kernel.org
The timekeeping duty is handed over from the outgoing CPU within stop
machine. This works well if CONFIG_NO_HZ_COMMON=n or the tick is in
high-res mode. However in low-res dynticks mode, the tick isn't
cancelled until the clockevent is shut down, which can happen later. The
tick may therefore fire again once IRQs are re-enabled on stop machine
and until IRQs are disabled for good upon the last call to idle.
That's so many opportunities for a timekeeper to go idle and the
outgoing CPU to take over that duty. This is why
tick_nohz_idle_stop_tick() is called one last time on idle if the CPU
is seen offline: so that the timekeeping duty is handed over again in
case the CPU has re-taken the duty.
This means there are two timekeeping handovers on CPU down hotplug with
different undocumented constraints and purposes:
1) A handover on stop machine for !dynticks || highres. All online CPUs
are guaranteed to be non-idle and the timekeeping duty can be safely
handed-over. The hrtimer tick is cancelled so it is guaranteed that in
dynticks mode the outgoing CPU won't take again the duty.
2) A handover on last idle call for dynticks && lowres. Setting the
duty to TICK_DO_TIMER_NONE makes sure that a CPU will take over the
timekeeping.
Prepare for consolidating the handover to a single place (the first one)
with shutting down the low-res tick as well from
tick_cancel_sched_timer() as well. This will simplify the handover and
unify the tick cancellation between high-res and low-res.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-15-frederic@kernel.org
The nohz mode field tells about low resolution nohz mode or high
resolution nohz mode but it doesn't tell about high resolution non-nohz
mode.
In order to retrieve the latter state, tick_cancel_sched_timer() must
fiddle with struct hrtimer's internals to guess if the tick has been
initialized in high resolution.
Move instead the nohz mode field information into the tick flags and
provide two new bits: one to know if the tick is in nohz mode and
another one to know if the tick is in high resolution. The combination
of those two flags provides all the needed informations to determine
which of the three tick modes is running.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-14-frederic@kernel.org
The individual bitfields of struct tick_sched must be modified from
IRQs disabled places, otherwise local modifications can race due to them
sharing the same memory storage.
The recent move of the "got_idle_tick" bitfield to its own storage shows
that the use of these bitfields, as pretty as they look, can be as much
error prone.
In order to avoid future issues of the like and make sure that those
bitfields are safely accessed, move those flags to an explicit mask
along with a mutator function performing the basic IRQs disabled sanity
check.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-13-frederic@kernel.org
tick_nohz_idle_got_tick() is called by cpuidle_reflect() within the idle
loop with interrupts enabled. This function modifies the struct
tick_sched's bitfield "got_idle_tick". However this bitfield is stored
within the same mask as other bitfields that can be modified from
interrupts.
Fortunately so far it looks like the only race that can happen is while
writing ->got_idle_tick to 0, an interrupt fires and writes the
->idle_active field to 0. It's then possible that the interrupted write
to ->got_idle_tick writes back the old value of ->idle_active back to 1.
However if that happens, the worst possible outcome is that the time
spent between that interrupt and the upcoming call to
tick_nohz_idle_exit() is accounted as idle, which is negligible quantity.
Still all the bitfield writes within this struct tick_sched's shadow
mask should be IRQ-safe. Therefore move this bitfield out to its own
storage to avoid further suprises.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-12-frederic@kernel.org
The full-nohz update function checks if the nohz mode is active before
proceeding. It considers one exception though: if the tick is already
stopped even though the nohz mode is inactive, it still moves on in
order to update/restart the tick if needed.
However in order for the tick to be stopped, the nohz_mode has to be
either NOHZ_MODE_LOWRES or NOHZ_MODE_HIGHRES. Therefore it doesn't make
sense to test if the tick is stopped before verifying NOHZ_MODE_INACTIVE
mode.
Remove the needless related condition.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-11-frederic@kernel.org
The broadcast shutdown code is executed through a random explicit call
within stop machine from the outgoing CPU.
However the tick broadcast is a midware between the tick callback and
the clocksource, therefore it makes more sense to shut it down after the
tick callback and before the clocksource drivers.
Move it instead to the common tick shutdown CPU hotplug state where
related operations can be ordered from highest to lowest level.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-10-frederic@kernel.org
The tick hrtimer is cancelled right before hrtimers are migrated. This
is done from the hrtimer subsystem even though it shouldn't know about
its actual users.
Move instead the tick hrtimer cancellation to the relevant CPU hotplug
state that aims at centralizing high level tick shutdown operations so
that the related flow is easy to follow.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-9-frederic@kernel.org
During the CPU offlining process, the various timer tick features are
shut down from scattered places, sometimes from teardown callbacks on
stop machine, sometimes through explicit calls, sometimes from the
control CPU after the CPU died. The reason why these shutdown operations
are spread around is not always clear and it makes the tick lifecycle
hard to follow.
The tick should be shut down in order from highest to lowest level:
On stop machine from the dying CPU (high-level):
1) Hand-over the timekeeping duty (tick_handover_do_timer())
2) Cancel the tick implementation called by the clockevent callback
(tick_cancel_sched_timer())
3) Shutdown broadcasting (tick_offline_cpu() / tick_broadcast_offline())
On stop machine from the dying CPU (low-level):
4) Shutdown clockevents drivers (CPUHP_AP_*_TIMER_STARTING states)
From the control CPU after the CPU died (low-level):
5) Shutdown/unregister/cleanup clockevents for the dead CPU
(tick_cleanup_dead_cpu())
Instead the current order is 2, 4 (both from CPU hotplug states), then
1 and 3 through direct calls. This layout and order don't make much
sense. The operations 1, 2, 3 should be gathered together and in order.
Sort this situation with creating a new TICK shut-down CPU hotplug state
and start with introducing the timekeeping duty hand-over there. The
state must precede hrtimers migration because the tick hrtimer will be
stopped from it in a further patch.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-8-frederic@kernel.org
The tick sched structure is already cleared from tick_cancel_sched_timer(),
so there is no need to clear that field again.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-7-frederic@kernel.org
tick_nohz_stop_sched_tick() is only about NOHZ_full and not about
dynticks-idle. Reflect that in the function name to avoid confusion.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-6-frederic@kernel.org
Avoid ifdeferry if it can be converted to IS_ENABLED() whenever possible
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-5-frederic@kernel.org
tick-sched.c is only built when CONFIG_TICK_ONESHOT=y, which is selected
only if CONFIG_NO_HZ_COMMON=y or CONFIG_HIGH_RES_TIMERS=y. Therefore
the related ifdeferry in this file is needless and can be removed.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-4-frederic@kernel.org
tick_nohz_lowres_handler() does the same work as
tick_nohz_highres_handler() plus the clockevent device reprogramming, so
make the former reuse the latter and rename it accordingly.
Signed-off-by: Peng Liu <liupeng17@lenovo.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-3-frederic@kernel.org
The ts->sched_timer initialization work of tick_nohz_switch_to_nohz()
is almost the same as that of tick_setup_sched_timer(), so adjust the
latter to get it reused by tick_nohz_switch_to_nohz().
This also makes the low resolution mode sched_timer benefit from the tick
skew boot option.
Signed-off-by: Peng Liu <liupeng17@lenovo.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240225225508.11587-2-frederic@kernel.org
During CPU-down hotplug, hrtimers may migrate to isolated CPUs,
compromising CPU isolation.
Address this issue by masking valid CPUs for hrtimers using
housekeeping_cpumask(HK_TYPE_TIMER).
Suggested-by: Waiman Long <longman@redhat.com>
Signed-off-by: Costa Shulyupin <costa.shul@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Waiman Long <longman@redhat.com>
Link: https://lore.kernel.org/r/20240222200856.569036-1-costa.shul@redhat.com
The timer pull model is in place so we can remove the heuristics which try
to guess the best target CPU at enqueue/modification time.
All non pinned timers are queued on the local CPU in the separate storage
and eventually pulled at expiry time to a remote CPU.
Originally-by: Richard Cochran (linutronix GmbH) <richardcochran@gmail.com>
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-21-anna-maria@linutronix.de
The timer pull logic needs proper debugging aids. Add tracepoints so the
hierarchical idle machinery can be diagnosed.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240222103403.31923-1-anna-maria@linutronix.de
Placing timers at enqueue time on a target CPU based on dubious heuristics
does not make any sense:
1) Most timer wheel timers are canceled or rearmed before they expire.
2) The heuristics to predict which CPU will be busy when the timer expires
are wrong by definition.
So placing the timers at enqueue wastes precious cycles.
The proper solution to this problem is to always queue the timers on the
local CPU and allow the non pinned timers to be pulled onto a busy CPU at
expiry time.
Therefore split the timer storage into local pinned and global timers:
Local pinned timers are always expired on the CPU on which they have been
queued. Global timers can be expired on any CPU.
As long as a CPU is busy it expires both local and global timers. When a
CPU goes idle it arms for the first expiring local timer. If the first
expiring pinned (local) timer is before the first expiring movable timer,
then no action is required because the CPU will wake up before the first
movable timer expires. If the first expiring movable timer is before the
first expiring pinned (local) timer, then this timer is queued into an idle
timerqueue and eventually expired by another active CPU.
To avoid global locking the timerqueues are implemented as a hierarchy. The
lowest level of the hierarchy holds the CPUs. The CPUs are associated to
groups of 8, which are separated per node. If more than one CPU group
exist, then a second level in the hierarchy collects the groups. Depending
on the size of the system more than 2 levels are required. Each group has a
"migrator" which checks the timerqueue during the tick for remote expirable
timers.
If the last CPU in a group goes idle it reports the first expiring event in
the group up to the next group(s) in the hierarchy. If the last CPU goes
idle it arms its timer for the first system wide expiring timer to ensure
that no timer event is missed.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240222103710.32582-1-anna-maria@linutronix.de
To prepare for the conversion of the NOHZ timer placement to a pull at
expiry time model it's required to have a function that returns the value
of the is_idle flag of the timer base to keep the hierarchy states during
online in sync with timer base state.
No functional change.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-18-anna-maria@linutronix.de
The logic to get the time of the last jiffies update will be needed by
the timer pull model as well.
Move the code into a global function in anticipation of the new caller.
No functional change.
Signed-off-by: Richard Cochran (linutronix GmbH) <richardcochran@gmail.com>
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-17-anna-maria@linutronix.de
Due to the conversion of the NOHZ timer placement to a pull at expiry
time model, the per CPU timer bases with non pinned timers are no
longer handled only by the local CPU. In case a remote CPU already
expires the non pinned timers base of the local CPU, nothing more
needs to be done by the local CPU. A check at the begin of the expire
timers routine is required, because timer base lock is dropped before
executing the timer callback function.
This is a preparatory work, but has no functional impact right now.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-16-anna-maria@linutronix.de
Move the locking out from __run_timers() to the call sites, so the
protected section can be extended at the call site. Preparatory work for
changing the NOHZ timer placement to a pull at expiry time model.
No functional change.
Signed-off-by: Richard Cochran (linutronix GmbH) <richardcochran@gmail.com>
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-15-anna-maria@linutronix.de
To prepare for the conversion of the NOHZ timer placement to a pull at
expiry time model it's required to have functionality available getting the
next timer interrupt on a remote CPU.
Locking of the timer bases and getting the information for the next timer
interrupt functionality is split into separate functions. This is required
to be compliant with lock ordering when the new model is in place.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-14-anna-maria@linutronix.de
The functionality for getting the next timer interrupt in
get_next_timer_interrupt() is split into a separate function
fetch_next_timer_interrupt() to be usable by other call sites.
This is preparatory work for the conversion of the NOHZ timer
placement to a pull at expiry time model. No functional change.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-13-anna-maria@linutronix.de
For the conversion of the NOHZ timer placement to a pull at expiry time
model it's required to have separate expiry times for the pinned and the
non-pinned (movable) timers. Therefore struct timer_events is introduced.
No functional change
Originally-by: Richard Cochran (linutronix GmbH) <richardcochran@gmail.com>
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-12-anna-maria@linutronix.de
Separate the storage space for pinned timers. Deferrable timers (doesn't
matter if pinned or non pinned) are still enqueued into their own base.
This is preparatory work for changing the NOHZ timer placement from a push
at enqueue time to a pull at expiry time model.
Originally-by: Richard Cochran (linutronix GmbH) <richardcochran@gmail.com>
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-11-anna-maria@linutronix.de
Split the logic for getting next timer interrupt (no matter of recalculated
or already stored in base->next_expiry) into a separate function named
next_timer_interrupt(). Make it available to local call sites only.
No functional change.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-10-anna-maria@linutronix.de
The logic for raising a softirq the way it is implemented right now, is
readable for two timer bases. When increasing the number of timer bases,
code gets harder to read. With the introduction of the timer migration
hierarchy, there will be three timer bases.
Therefore restructure the code to use a loop. No functional change.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-9-anna-maria@linutronix.de
When adding a timer to the timer wheel using add_timer_on(), it is an
implicitly pinned timer. With the timer pull at expiry time model in place,
the TIMER_PINNED flag is required to make sure timers end up in proper
base.
Set the TIMER_PINNED flag unconditionally when add_timer_on() is executed.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-8-anna-maria@linutronix.de
A timer might be used as a pinned timer (using add_timer_on()) and later on
as non-pinned timer using add_timer(). When the "NOHZ timer pull at expiry
model" is in place, the TIMER_PINNED flag is required to be used whenever a
timer needs to expire on a dedicated CPU. Otherwise the flag must not be
set if expiration on a dedicated CPU is not required.
add_timer_on()'s behavior will be changed during the preparation patches
for the "NOHZ timer pull at expiry model" to unconditionally set the
TIMER_PINNED flag. To be able to clear/ set the flag when queueing a
timer, two variants of add_timer() are introduced.
This is a preparatory step and has no functional change.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-6-anna-maria@linutronix.de
When tick is stopped also the timer base is_idle flag is set. When
reentering timer_base_try_to_set_idle() with the tick stopped, there is no
need to check whether the timer base needs to be set idle again. When a
timer was enqueued in the meantime, this is already handled by the
tick_nohz_next_event() call which was executed before
tick_nohz_stop_tick().
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-5-anna-maria@linutronix.de
The timer base is marked idle when get_next_timer_interrupt() is
executed. But the decision whether the tick will be stopped and whether the
system is able to go idle is done later. When the timer bases is marked
idle and a new first timer is enqueued remote an IPI is raised. Even if it
is not required because the tick is not stopped and the timer base is
evaluated again at the next tick.
To prevent this, the timer base is marked idle in tick_nohz_stop_tick() and
get_next_timer_interrupt() is streamlined by only looking for the next timer
interrupt. All other work is postponed to timer_base_try_to_set_idle() which is
called by tick_nohz_stop_tick(). timer_base_try_to_set_idle() never resets
timer_base::is_idle state. This is done when the tick is restarted via
tick_nohz_restart_sched_tick().
With this, tick_sched::tick_stopped and timer_base::is_idle are always in
sync. So there is no longer the need to execute timer_clear_idle() in
tick_nohz_idle_retain_tick(). This was required before, as
tick_nohz_next_event() set timer_base::is_idle even if the tick would not be
stopped. So timer_clear_idle() is only executed, when timer base is idle. So the
check whether timer base is idle, is now no longer required as well.
While at it fix some nearby whitespace damage as well.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-4-anna-maria@linutronix.de
Split out get_next_timer_interrupt() to be able to extend it and make it
reusable for other call sites.
No functional change.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-3-anna-maria@linutronix.de
get_next_timer_interrupt() contains two parts for the next timer interrupt
calculation. Those two parts are separated by forwarding the base
clock. But the second part does not depend on the forwarded base
clock.
Therefore restructure get_next_timer_interrupt() to keep things together
which belong together.
No functional change.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20240221090548.36600-2-anna-maria@linutronix.de
On a 8-socket server the TSC is wrongly marked as 'unstable' and disabled
during boot time on about one out of 120 boot attempts:
clocksource: timekeeping watchdog on CPU227: wd-tsc-wd excessive read-back delay of 153560ns vs. limit of 125000ns,
wd-wd read-back delay only 11440ns, attempt 3, marking tsc unstable
tsc: Marking TSC unstable due to clocksource watchdog
TSC found unstable after boot, most likely due to broken BIOS. Use 'tsc=unstable'.
sched_clock: Marking unstable (119294969739, 159204297)<-(125446229205, -5992055152)
clocksource: Checking clocksource tsc synchronization from CPU 319 to CPUs 0,99,136,180,210,542,601,896.
clocksource: Switched to clocksource hpet
The reason is that for platform with a large number of CPUs, there are
sporadic big or huge read latencies while reading the watchog/clocksource
during boot or when system is under stress work load, and the frequency and
maximum value of the latency goes up with the number of online CPUs.
The cCurrent code already has logic to detect and filter such high latency
case by reading the watchdog twice and checking the two deltas. Due to the
randomness of the latency, there is a low probabilty that the first delta
(latency) is big, but the second delta is small and looks valid. The
watchdog code retries the readouts by default twice, which is not
necessarily sufficient for systems with a large number of CPUs.
There is a command line parameter 'max_cswd_read_retries' which allows to
increase the number of retries, but that's not user friendly as it needs to
be tweaked per system. As the number of required retries is proportional to
the number of online CPUs, this parameter can be calculated at runtime.
Scale and enlarge the number of retries according to the number of online
CPUs and remove the command line parameter completely.
[ tglx: Massaged change log and comments ]
Signed-off-by: Feng Tang <feng.tang@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Jin Wang <jin1.wang@intel.com>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Waiman Long <longman@redhat.com>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Link: https://lore.kernel.org/r/20240221060859.1027450-1-feng.tang@intel.com
'days' is a s64 (from div_s64), and so should use a %lld specifier.
This was found by extending KUnit's assertion macros to use gcc's
__printf attribute.
Fixes: 2760105516 ("time: Improve performance of time64_to_tm()")
Signed-off-by: David Gow <davidgow@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240221092728.1281499-5-davidgow@google.com
There's a conflict between this recent upstream fix:
dad6a09f31 ("hrtimer: Report offline hrtimer enqueue")
and a pending commit in the timers tree:
1a4729ecaf ("hrtimers: Move hrtimer base related definitions into hrtimer_defs.h")
Resolve it by applying the upstream fix to the new <linux/hrtimer_defs.h> header.
Conflict:
include/linux/hrtimer.h
Semantic conflict:
include/linux/hrtimer_defs.h
Signed-off-by: Ingo Molnar <mingo@kernel.org>
So far, get_device_system_crosststamp() unconditionally passes
system_counterval.cycles to timekeeping_cycles_to_ns(). But when
interpolating system time (do_interp == true), system_counterval.cycles is
before tkr_mono.cycle_last, contrary to the timekeeping_cycles_to_ns()
expectations.
On x86, CONFIG_CLOCKSOURCE_VALIDATE_LAST_CYCLE will mitigate on
interpolating, setting delta to 0. With delta == 0, xtstamp->sys_monoraw
and xtstamp->sys_realtime are then set to the last update time, as
implicitly expected by adjust_historical_crosststamp(). On other
architectures, the resulting nonsense xtstamp->sys_monoraw and
xtstamp->sys_realtime corrupt the xtstamp (ts) adjustment in
adjust_historical_crosststamp().
Fix this by deriving xtstamp->sys_monoraw and xtstamp->sys_realtime from
the last update time when interpolating, by using the local variable
"cycles". The local variable already has the right value when
interpolating, unlike system_counterval.cycles.
Fixes: 2c756feb18 ("time: Add history to cross timestamp interface supporting slower devices")
Signed-off-by: Peter Hilber <peter.hilber@opensynergy.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: John Stultz <jstultz@google.com>
Link: https://lore.kernel.org/r/20231218073849.35294-4-peter.hilber@opensynergy.com
The cycle_between() helper checks if parameter test is in the open interval
(before, after). Colloquially speaking, this also applies to the counter
wrap-around special case before > after. get_device_system_crosststamp()
currently uses cycle_between() at the first call site to decide whether to
interpolate for older counter readings.
get_device_system_crosststamp() has the following problem with
cycle_between() testing against an open interval: Assume that, by chance,
cycles == tk->tkr_mono.cycle_last (in the following, "cycle_last" for
brevity). Then, cycle_between() at the first call site, with effective
argument values cycle_between(cycle_last, cycles, now), returns false,
enabling interpolation. During interpolation,
get_device_system_crosststamp() will then call cycle_between() at the
second call site (if a history_begin was supplied). The effective argument
values are cycle_between(history_begin->cycles, cycles, cycles), since
system_counterval.cycles == interval_start == cycles, per the assumption.
Due to the test against the open interval, cycle_between() returns false
again. This causes get_device_system_crosststamp() to return -EINVAL.
This failure should be avoided, since get_device_system_crosststamp() works
both when cycles follows cycle_last (no interpolation), and when cycles
precedes cycle_last (interpolation). For the case cycles == cycle_last,
interpolation is actually unneeded.
Fix this by changing cycle_between() into timestamp_in_interval(), which
now checks against the closed interval, rather than the open interval.
This changes the get_device_system_crosststamp() behavior for three corner
cases:
1. Bypass interpolation in the case cycles == tk->tkr_mono.cycle_last,
fixing the problem described above.
2. At the first timestamp_in_interval() call site, cycles == now no longer
causes failure.
3. At the second timestamp_in_interval() call site, history_begin->cycles
== system_counterval.cycles no longer causes failure.
adjust_historical_crosststamp() also works for this corner case,
where partial_history_cycles == total_history_cycles.
These behavioral changes should not cause any problems.
Fixes: 2c756feb18 ("time: Add history to cross timestamp interface supporting slower devices")
Signed-off-by: Peter Hilber <peter.hilber@opensynergy.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20231218073849.35294-3-peter.hilber@opensynergy.com
cycle_between() decides whether get_device_system_crosststamp() will
interpolate for older counter readings.
cycle_between() yields wrong results for a counter wrap-around where after
< before < test, and for the case after < test < before.
Fix the comparison logic.
Fixes: 2c756feb18 ("time: Add history to cross timestamp interface supporting slower devices")
Signed-off-by: Peter Hilber <peter.hilber@opensynergy.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: John Stultz <jstultz@google.com>
Link: https://lore.kernel.org/r/20231218073849.35294-2-peter.hilber@opensynergy.com
timer_base struct lacks description of struct members. Important struct
member information is sprinkled in comments or in code all over the place.
Collect information and write struct description to keep track of most
important information in a single place.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240123164702.55612-5-anna-maria@linutronix.de
The return value of tick_nohz_next_event() is not obvious at the first
glance. Add a kernel-doc compatible function description which also covers
return values.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240123164702.55612-4-anna-maria@linutronix.de
Documentation of functions lacks the annotations which are used by
kernel-doc and *.rst to make appearance in rendered documents more
user-friendly.
Use those annotations to improve user-friendliness. While at it prevent
duplication of comments and use a reference instead.
Signed-off-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20240123164702.55612-3-anna-maria@linutronix.de
Now that the driver core can properly handle constant struct bus_type,
move the clockevents_subsys variable to be a constant structure as well,
placing it into read-only memory which can not be modified at runtime.
Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ricardo B. Marliere <ricardo@marliere.net>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://lore.kernel.org/r/20240204-bus_cleanup-time-v1-2-207ec18e24b8@marliere.net
Now that the driver core can properly handle constant struct bus_type,
move the clocksource_subsys variable to be a constant structure as well,
placing it into read-only memory which can not be modified at runtime.
Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ricardo B. Marliere <ricardo@marliere.net>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: John Stultz <jstultz@google.com>
Link: https://lore.kernel.org/r/20240204-bus_cleanup-time-v1-1-207ec18e24b8@marliere.net
The hrtimers migration on CPU-down hotplug process has been moved
earlier, before the CPU actually goes to die. This leaves a small window
of opportunity to queue an hrtimer in a blind spot, leaving it ignored.
For example a practical case has been reported with RCU waking up a
SCHED_FIFO task right before the CPUHP_AP_IDLE_DEAD stage, queuing that
way a sched/rt timer to the local offline CPU.
Make sure such situations never go unnoticed and warn when that happens.
Fixes: 5c0930ccaa ("hrtimers: Push pending hrtimers away from outgoing CPU earlier")
Reported-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20240129235646.3171983-4-boqun.feng@gmail.com
Commit 71fee48f ("tick-sched: Fix idle and iowait sleeptime accounting vs
CPU hotplug") preserved total idle sleep time and iowait sleeptime across
CPU hotplug events.
Similar reasoning applies to the number of idle calls and idle sleeps to
get the proper average of sleep time per idle invocation.
Preserve those fields too.
Fixes: 71fee48f ("tick-sched: Fix idle and iowait sleeptime accounting vs CPU hotplug")
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20240122233534.3094238-1-tim.c.chen@linux.intel.com
There have been reports of the watchdog marking clocksources unstable on
machines with 8 NUMA nodes:
clocksource: timekeeping watchdog on CPU373:
Marking clocksource 'tsc' as unstable because the skew is too large:
clocksource: 'hpet' wd_nsec: 14523447520
clocksource: 'tsc' cs_nsec: 14524115132
The measured clocksource skew - the absolute difference between cs_nsec
and wd_nsec - was 668 microseconds:
cs_nsec - wd_nsec = 14524115132 - 14523447520 = 667612
The kernel used 200 microseconds for the uncertainty_margin of both the
clocksource and watchdog, resulting in a threshold of 400 microseconds (the
md variable). Both the cs_nsec and the wd_nsec value indicate that the
readout interval was circa 14.5 seconds. The observed behaviour is that
watchdog checks failed for large readout intervals on 8 NUMA node
machines. This indicates that the size of the skew was directly proportinal
to the length of the readout interval on those machines. The measured
clocksource skew, 668 microseconds, was evaluated against a threshold (the
md variable) that is suited for readout intervals of roughly
WATCHDOG_INTERVAL, i.e. HZ >> 1, which is 0.5 second.
The intention of 2e27e793e2 ("clocksource: Reduce clocksource-skew
threshold") was to tighten the threshold for evaluating skew and set the
lower bound for the uncertainty_margin of clocksources to twice
WATCHDOG_MAX_SKEW. Later in c37e85c135 ("clocksource: Loosen clocksource
watchdog constraints"), the WATCHDOG_MAX_SKEW constant was increased to
125 microseconds to fit the limit of NTP, which is able to use a
clocksource that suffers from up to 500 microseconds of skew per second.
Both the TSC and the HPET use default uncertainty_margin. When the
readout interval gets stretched the default uncertainty_margin is no
longer a suitable lower bound for evaluating skew - it imposes a limit
that is far stricter than the skew with which NTP can deal.
The root causes of the skew being directly proportinal to the length of
the readout interval are:
* the inaccuracy of the shift/mult pairs of clocksources and the watchdog
* the conversion to nanoseconds is imprecise for large readout intervals
Prevent this by skipping the current watchdog check if the readout
interval exceeds 2 * WATCHDOG_INTERVAL. Considering the maximum readout
interval of 2 * WATCHDOG_INTERVAL, the current default uncertainty margin
(of the TSC and HPET) corresponds to a limit on clocksource skew of 250
ppm (microseconds of skew per second). To keep the limit imposed by NTP
(500 microseconds of skew per second) for all possible readout intervals,
the margins would have to be scaled so that the threshold value is
proportional to the length of the actual readout interval.
As for why the readout interval may get stretched: Since the watchdog is
executed in softirq context the expiration of the watchdog timer can get
severely delayed on account of a ksoftirqd thread not getting to run in a
timely manner. Surely, a system with such belated softirq execution is not
working well and the scheduling issue should be looked into but the
clocksource watchdog should be able to deal with it accordingly.
Fixes: 2e27e793e2 ("clocksource: Reduce clocksource-skew threshold")
Suggested-by: Feng Tang <feng.tang@intel.com>
Signed-off-by: Jiri Wiesner <jwiesner@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Feng Tang <feng.tang@intel.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20240122172350.GA740@incl
- A fix for the idle and iowait time accounting vs. CPU hotplug.
The time is reset on CPU hotplug which makes the accumulated
systemwide time jump backwards.
- Assorted fixes and improvements for clocksource/event drivers
-----BEGIN PGP SIGNATURE-----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=RD6E
-----END PGP SIGNATURE-----
Merge tag 'timers-core-2024-01-21' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer updates from Thomas Gleixner:
"Updates for time and clocksources:
- A fix for the idle and iowait time accounting vs CPU hotplug.
The time is reset on CPU hotplug which makes the accumulated
systemwide time jump backwards.
- Assorted fixes and improvements for clocksource/event drivers"
* tag 'timers-core-2024-01-21' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
tick-sched: Fix idle and iowait sleeptime accounting vs CPU hotplug
clocksource/drivers/ep93xx: Fix error handling during probe
clocksource/drivers/cadence-ttc: Fix some kernel-doc warnings
clocksource/drivers/timer-ti-dm: Fix make W=n kerneldoc warnings
clocksource/timer-riscv: Add riscv_clock_shutdown callback
dt-bindings: timer: Add StarFive JH8100 clint
dt-bindings: timer: thead,c900-aclint-mtimer: separate mtime and mtimecmp regs
