sched_clock_postinit() initializes a generic clock on systems where no
other clock is provided. This function may be called only after
timekeeping_init().

Rename sched_clock_postinit to generic_clock_inti() and call it from
sched_clock_init(). Move the call for sched_clock_init() until after
time_init().
Suggested-by: NPeter Zijlstra <peterz@infradead.org>
Signed-off-by: NPavel Tatashin <pasha.tatashin@oracle.com>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: steven.sistare@oracle.com
Cc: daniel.m.jordan@oracle.com
Cc: linux@armlinux.org.uk
Cc: schwidefsky@de.ibm.com
Cc: heiko.carstens@de.ibm.com
Cc: john.stultz@linaro.org
Cc: sboyd@codeaurora.org
Cc: hpa@zytor.com
Cc: douly.fnst@cn.fujitsu.com
Cc: prarit@redhat.com
Cc: feng.tang@intel.com
Cc: pmladek@suse.com
Cc: gnomes@lxorguk.ukuu.org.uk
Cc: linux-s390@vger.kernel.org
Cc: boris.ostrovsky@oracle.com
Cc: jgross@suse.com
Cc: pbonzini@redhat.com
Link: https://lkml.kernel.org/r/20180719205545.16512-23-pasha.tatashin@oracle.com

read_persistent_wall_and_boot_offset() is called during boot to read
both the persistent clock and also return the offset between the boot time
and the value of persistent clock.

Change the default boot_offset from zero to local_clock() so architectures,
that do not have a dedicated boot_clock but have early sched_clock(), such
as SPARCv9, x86, and possibly more will benefit from this change by getting
a better and more consistent estimate of the boot time without need for an
arch specific implementation.
Signed-off-by: NPavel Tatashin <pasha.tatashin@oracle.com>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: steven.sistare@oracle.com
Cc: daniel.m.jordan@oracle.com
Cc: linux@armlinux.org.uk
Cc: schwidefsky@de.ibm.com
Cc: heiko.carstens@de.ibm.com
Cc: john.stultz@linaro.org
Cc: sboyd@codeaurora.org
Cc: hpa@zytor.com
Cc: douly.fnst@cn.fujitsu.com
Cc: peterz@infradead.org
Cc: prarit@redhat.com
Cc: feng.tang@intel.com
Cc: pmladek@suse.com
Cc: gnomes@lxorguk.ukuu.org.uk
Cc: linux-s390@vger.kernel.org
Cc: boris.ostrovsky@oracle.com
Cc: jgross@suse.com
Cc: pbonzini@redhat.com
Link: https://lkml.kernel.org/r/20180719205545.16512-17-pasha.tatashin@oracle.com

If architecture does not support exact boot time, it is challenging to
estimate boot time without having a reference to the current persistent
clock value. Yet, it cannot read the persistent clock time again, because
this may lead to math discrepancies with the caller of read_boot_clock64()
who have read the persistent clock at a different time.

This is why it is better to provide two values simultaneously: the
persistent clock value, and the boot time.

Replace read_boot_clock64() with:
read_persistent_wall_and_boot_offset(wall_time, boot_offset)

Where wall_time is returned by read_persistent_clock() And boot_offset is
wall_time - boot time, which defaults to 0.
Signed-off-by: NPavel Tatashin <pasha.tatashin@oracle.com>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: steven.sistare@oracle.com
Cc: daniel.m.jordan@oracle.com
Cc: linux@armlinux.org.uk
Cc: schwidefsky@de.ibm.com
Cc: heiko.carstens@de.ibm.com
Cc: john.stultz@linaro.org
Cc: sboyd@codeaurora.org
Cc: hpa@zytor.com
Cc: douly.fnst@cn.fujitsu.com
Cc: peterz@infradead.org
Cc: prarit@redhat.com
Cc: feng.tang@intel.com
Cc: pmladek@suse.com
Cc: gnomes@lxorguk.ukuu.org.uk
Cc: linux-s390@vger.kernel.org
Cc: boris.ostrovsky@oracle.com
Cc: jgross@suse.com
Cc: pbonzini@redhat.com
Link: https://lkml.kernel.org/r/20180719205545.16512-16-pasha.tatashin@oracle.com

This reverts commit 1332a905.

The original issue was not because of incorrect checking of cpumask for
both new and old tick device. It was incorrectly analysed was due to the
misunderstanding of the comment and misinterpretation of the return value
from tick_check_preferred. The main issue is with the clockevent driver
that sets the cpumask to cpu_all_mask instead of cpu_possible_mask.
Signed-off-by: NSudeep Holla <sudeep.holla@arm.com>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Tested-by: NKevin Hilman <khilman@baylibre.com>
Tested-by: NMartin Blumenstingl <martin.blumenstingl@googlemail.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: Marc Zyngier <marc.zyngier@arm.com>
Link: https://lkml.kernel.org/r/1531151136-18297-1-git-send-email-sudeep.holla@arm.com

For the common cases where 1000 is a multiple of HZ, or HZ is a multiple of
1000, jiffies_to_msecs() never returns zero when passed a non-zero time
period.

However, if HZ > 1000 and not an integer multiple of 1000 (e.g. 1024 or
1200, as used on alpha and DECstation), jiffies_to_msecs() may return zero
for small non-zero time periods.  This may break code that relies on
receiving back a non-zero value.

jiffies_to_usecs() does not need such a fix: one jiffy can only be less
than one µs if HZ > 1000000, and such large values of HZ are already
rejected at build time, twice:

  - include/linux/jiffies.h does #error if HZ >= 12288,
  - kernel/time/time.c has BUILD_BUG_ON(HZ > USEC_PER_SEC).

Broken since forever.
Signed-off-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Reviewed-by: NArnd Bergmann <arnd@arndb.de>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: linux-alpha@vger.kernel.org
Cc: linux-mips@linux-mips.org
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20180622143357.7495-1-geert@linux-m68k.org

Commit b5793b0d added support for building the nanosleep compat system
call on 32-bit architectures, but missed one change in nanosleep_copyout(),
which would trigger a BUG() as soon as any architecture is switched over to
use it.

Use the proper config symbol to enable the code path.

Fixes: Commit b5793b0d ("posix-timers: Make compat syscalls depend on CONFIG_COMPAT_32BIT_TIME")
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: y2038@lists.linaro.org
Cc: Anna-Maria Gleixner <anna-maria@linutronix.de>
Cc: Deepa Dinamani <deepa.kernel@gmail.com>
Cc: "Rafael J. Wysocki" <rafael.j.wysocki@intel.com>
Link: https://lkml.kernel.org/r/20180618140811.2998503-1-arnd@arndb.de

The lockdep_assert_irqs_disabled() was a BUG_ON() statement in the
beginning and it was added just before the "spin_lock(siglock)"
statement to ensure this lock was taken with disabled interrupts.

This is no longer the case: the siglock is acquired via
lock_task_sighand() and this function already disables the interrupts.
The lock is also acquired before this "lockdep_assert_irqs_disabled" so
it is best to remove it.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Frederic Weisbecker <frederic@kernel.org>
Link: https://lkml.kernel.org/r20180504152548.7166-1-bigeasy@linutronix.de

timekeeping suspend/resume calls read_persistent_clock() which takes
rtc_lock. That results in might sleep warnings because at that point
we run with interrupts disabled.

We cannot convert rtc_lock to a raw spinlock as that would trigger
other might sleep warnings.

As a workaround we disable the might sleep warnings by setting
system_state to SYSTEM_SUSPEND before calling sysdev_suspend() and
restoring it to SYSTEM_RUNNING afer sysdev_resume(). There is no lock
contention because hibernate / suspend to RAM is single-CPU at this
point.

In s2idle's case the system_state is set to SYSTEM_SUSPEND before
timekeeping_suspend() which is invoked by the last CPU. In the resume
case it set back to SYSTEM_RUNNING after timekeeping_resume() which is
invoked by the first CPU in the resume case. The other CPUs will block
on tick_freeze_lock.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
[bigeasy: cover s2idle in tick_freeze() / tick_unfreeze()]
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

I have run into a couple of drivers using current_kernel_time()
suffering from the y2038 problem, and they could be converted
to using ktime_t, but don't have interfaces that skip the nanosecond
calculation at the moment.

This introduces ktime_get_coarse_with_offset() as a simpler
variant of ktime_get_with_offset(), and adds wrappers for the
three time domains we support with the existing function.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: y2038@lists.linaro.org
Cc: John Stultz <john.stultz@linaro.org>
Link: https://lkml.kernel.org/r/20180427134016.2525989-5-arnd@arndb.de

The current_kernel_time64, get_monotonic_coarse64, getrawmonotonic64,
get_monotonic_boottime64 and timekeeping_clocktai64 interfaces have
rather inconsistent naming, and they differ in the calling conventions
by passing the output either by reference or as a return value.

Rename them to ktime_get_coarse_real_ts64, ktime_get_coarse_ts64,
ktime_get_raw_ts64, ktime_get_boottime_ts64 and ktime_get_clocktai_ts64
respectively, and provide the interfaces with macros or inline
functions as needed.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: y2038@lists.linaro.org
Cc: John Stultz <john.stultz@linaro.org>
Link: https://lkml.kernel.org/r/20180427134016.2525989-4-arnd@arndb.de

In a move to make ktime_get_*() the preferred driver interface into the
timekeeping code, sanitizes ktime_get_real_ts64() to be a proper exported
symbol rather than an alias for getnstimeofday64().

The internal __getnstimeofday64() is no longer used, so remove that
and merge it into ktime_get_real_ts64().
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: y2038@lists.linaro.org
Cc: John Stultz <john.stultz@linaro.org>
Link: https://lkml.kernel.org/r/20180427134016.2525989-3-arnd@arndb.de

At this point, we have converted most of the kernel to use timespec64
consistently in place of timespec, so it seems it's time to make
timespec64 the native structure and define timespec in terms of that
one on 64-bit architectures.

Starting with gcc-5, the compiler can completely optimize away the
timespec_to_timespec64 and timespec64_to_timespec functions on 64-bit
architectures. With older compilers, we introduce a couple of extra
copies of local variables, but those are easily avoided by using
the timespec64 based interfaces consistently, as we do in most of the
important code paths already.

The main upside of removing the hack is that printing the tv_sec
field of a timespec64 structure can now use the %lld format
string on all architectures without a cast to time64_t. Without
this patch, the field is a 'long' type and would have to be printed
using %ld on 64-bit architectures.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: y2038@lists.linaro.org
Cc: John Stultz <john.stultz@linaro.org>
Link: https://lkml.kernel.org/r/20180427134016.2525989-2-arnd@arndb.de

The inline keyword was not at the beginning of the function declarations.
Fix the following warnings triggered when using W=1:

  kernel/time/clocksource.c:456:1: warning: ‘inline’ is not at beginning of declaration [-Wold-style-declaration]
  kernel/time/clocksource.c:457:1: warning: ‘inline’ is not at beginning of declaration [-Wold-style-declaration]
Signed-off-by: NMathieu Malaterre <malat@debian.org>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: John Stultz <john.stultz@linaro.org>
Link: https://lkml.kernel.org/r/20180516195943.31924-1-malat@debian.org

Variant of proc_create_data that directly take a struct seq_operations
argument + a private state size and drastically reduces the boilerplate
code in the callers.

All trivial callers converted over.
Signed-off-by: NChristoph Hellwig <hch@lst.de>

for_each_cpu() unintuitively reports CPU0 as set independent of the actual
cpumask content on UP kernels. This causes an unexpected PIT interrupt
storm on a UP kernel running in an SMP virtual machine on Hyper-V, and as
a result, the virtual machine can suffer from a strange random delay of 1~20
minutes during boot-up, and sometimes it can hang forever.

Protect if by checking whether the cpumask is empty before entering the
for_each_cpu() loop.

[ tglx: Use !IS_ENABLED(CONFIG_SMP) instead of #ifdeffery ]
Signed-off-by: NDexuan Cui <decui@microsoft.com>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Josh Poulson <jopoulso@microsoft.com>
Cc: "Michael Kelley (EOSG)" <Michael.H.Kelley@microsoft.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: stable@vger.kernel.org
Cc: Rakib Mullick <rakib.mullick@gmail.com>
Cc: Jork Loeser <Jork.Loeser@microsoft.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: KY Srinivasan <kys@microsoft.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Link: https://lkml.kernel.org/r/KL1P15301MB000678289FE55BA365B3279ABF990@KL1P15301MB0006.APCP153.PROD.OUTLOOK.COM
Link: https://lkml.kernel.org/r/KL1P15301MB0006FA63BC22BEB64902EAA0BF930@KL1P15301MB0006.APCP153.PROD.OUTLOOK.COM

Remove the 'printf_fn_t' typedef as it is not used.
Signed-off-by: NChen Lin <chen45464546@163.com>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: sboyd@kernel.org
Cc: john.stultz@linaro.org
Link: https://lkml.kernel.org/r/1526053649-24229-1-git-send-email-chen45464546@163.com

Those three warnings can easily solved by using :: to indicate a
code block:

	./kernel/time/timer.c:1259: WARNING: Unexpected indentation.
	./kernel/time/timer.c:1261: WARNING: Unexpected indentation.
	./kernel/time/timer.c:1262: WARNING: Block quote ends without a blank line; unexpected unindent.

While here, align the lines at the block.
Signed-off-by: NMauro Carvalho Chehab <mchehab+samsung@kernel.org>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: Linux Doc Mailing List <linux-doc@vger.kernel.org>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: John Stultz <john.stultz@linaro.org>
Link: https://lkml.kernel.org/r/f02e6a0ce27f3b5e33415d92d07a40598904b3ee.1525684985.git.mchehab%2Bsamsung@kernel.org

Checking the equality of cpumask for both new and old tick device doesn't
ensure that it's CPU local device. This will cause issue if a low rating
clockevent tick device is registered first followed by the registration
of higher rating clockevent tick device.

In such case, clockevents_released list will never get emptied as both
the devices get selected as preferred one and we will loop forever in
clockevents_notify_released.
Signed-off-by: NSudeep Holla <sudeep.holla@arm.com>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Link: https://lkml.kernel.org/r/1525881728-4858-1-git-send-email-sudeep.holla@arm.com

The clocksource watchdog uses a work to spawn a kthread to run the
watchdog. That is about as silly as it sounds, run the watchdog
directly from the work.
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Reviewed-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: diego.viola@gmail.com
Cc: rui.zhang@intel.com
Link: https://lkml.kernel.org/r/20180430100344.713862818@infradead.org

AFAICS the hotplug code no longer uses this function.
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Reviewed-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: diego.viola@gmail.com
Cc: rui.zhang@intel.com
Link: https://lkml.kernel.org/r/20180430100344.656525644@infradead.org

When a registered clocksource gets marked unstable the watchdog_kthread
will de-rate and re-select the clocksource. Ensure it also de-rates
when getting called on an unregistered clocksource.
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Reviewed-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: diego.viola@gmail.com
Cc: rui.zhang@intel.com
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20180430100344.594904898@infradead.org

A number of places relies on list_empty(&cs->wd_list), however the
list_head does not get initialized. Do so upon registration, such that
thereafter it is possible to rely on list_empty() correctly reflecting
the list membership status.
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Tested-by: NDiego Viola <diego.viola@gmail.com>
Reviewed-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: stable@vger.kernel.org
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: rui.zhang@intel.com
Link: https://lkml.kernel.org/r/20180430100344.472662715@infradead.org

Because of how the code flips between tsc-early and tsc clocksources
it might need to mark one or both unstable. The current code in
mark_tsc_unstable() only worked because previously it registered the
tsc clocksource once and then never touched it.

Since it now unregisters the tsc-early clocksource, it needs to know
if a clocksource got unregistered and the current cs->mult test
doesn't work for that. Instead use list_empty(&cs->list) to test for
registration.

Furthermore, since clocksource_mark_unstable() needs to place the cs
on the wd_list, it links the cs->list and cs->wd_list serialization.
It must not see a clocsource registered (!empty cs->list) but already
past dequeue_watchdog(). So place {en,de}queue{,_watchdog}() under the
same lock.

Provided cs->list is initialized to empty, this then allows us to
unconditionally use clocksource_mark_unstable(), regardless of the
registration state.

Fixes: aa83c457 ("x86/tsc: Introduce early tsc clocksource")
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Reviewed-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: NDiego Viola <diego.viola@gmail.com>
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: diego.viola@gmail.com
Cc: rui.zhang@intel.com
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20180502135312.GS12217@hirez.programming.kicks-ass.net

Revert commits

92af4dcb ("tracing: Unify the "boot" and "mono" tracing clocks")
127bfa5f ("hrtimer: Unify MONOTONIC and BOOTTIME clock behavior")
7250a404 ("posix-timers: Unify MONOTONIC and BOOTTIME clock behavior")
d6c7270e ("timekeeping: Remove boot time specific code")
f2d6fdbf ("Input: Evdev - unify MONOTONIC and BOOTTIME clock behavior")
d6ed449a ("timekeeping: Make the MONOTONIC clock behave like the BOOTTIME clock")
72199320 ("timekeeping: Add the new CLOCK_MONOTONIC_ACTIVE clock")

As stated in the pull request for the unification of CLOCK_MONOTONIC and
CLOCK_BOOTTIME, it was clear that we might have to revert the change.

As reported by several folks systemd and other applications rely on the
documented behaviour of CLOCK_MONOTONIC on Linux and break with the above
changes. After resume daemons time out and other timeout related issues are
observed. Rafael compiled this list:

* systemd kills daemons on resume, after >WatchdogSec seconds
  of suspending (Genki Sky).  [Verified that that's because systemd uses
  CLOCK_MONOTONIC and expects it to not include the suspend time.]

* systemd-journald misbehaves after resume:
  systemd-journald[7266]: File /var/log/journal/016627c3c4784cd4812d4b7e96a34226/system.journal
corrupted or uncleanly shut down, renaming and replacing.
  (Mike Galbraith).

* NetworkManager reports "networking disabled" and networking is broken
  after resume 50% of the time (Pavel).  [May be because of systemd.]

* MATE desktop dims the display and starts the screensaver right after
  system resume (Pavel).

* Full system hang during resume (me).  [May be due to systemd or NM or both.]

That happens on debian and open suse systems.

It's sad, that these problems were neither catched in -next nor by those
folks who expressed interest in this change.
Reported-by: NRafael J. Wysocki <rjw@rjwysocki.net>
Reported-by: Genki Sky <sky@genki.is>,
Reported-by: NPavel Machek <pavel@ucw.cz>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Kevin Easton <kevin@guarana.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mark Salyzyn <salyzyn@android.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Petr Mladek <pmladek@suse.com>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>

Kaike reported that in tests rdma hrtimers occasionaly stopped working. He
did great debugging, which provided enough context to decode the problem.

CPU 3			     	      	     CPU 2

idle
start sched_timer expires = 712171000000
 queue->next = sched_timer
					    start rdmavt timer. expires = 712172915662
					    lock(baseof(CPU3))
tick_nohz_stop_tick()
tick = 716767000000			    timerqueue_add(tmr)

hrtimer_set_expires(sched_timer, tick);
  sched_timer->expires = 716767000000  <---- FAIL
					     if (tmr->expires < queue->next->expires)
hrtimer_start(sched_timer)		          queue->next = tmr;
lock(baseof(CPU3))
					     unlock(baseof(CPU3))
timerqueue_remove()
timerqueue_add()

ts->sched_timer is queued and queue->next is pointing to it, but then
ts->sched_timer.expires is modified.

This not only corrupts the ordering of the timerqueue RB tree, it also
makes CPU2 see the new expiry time of timerqueue->next->expires when
checking whether timerqueue->next needs to be updated. So CPU2 sees that
the rdma timer is earlier than timerqueue->next and sets the rdma timer as
new next.

Depending on whether it had also seen the new time at RB tree enqueue, it
might have queued the rdma timer at the wrong place and then after removing
the sched_timer the RB tree is completely hosed.

The problem was introduced with a commit which tried to solve inconsistency
between the hrtimer in the tick_sched data and the underlying hardware
clockevent. It split out hrtimer_set_expires() to store the new tick time
in both the NOHZ and the NOHZ + HIGHRES case, but missed the fact that in
the NOHZ + HIGHRES case the hrtimer might still be queued.

Use hrtimer_start(timer, tick...) for the NOHZ + HIGHRES case which sets
timer->expires after canceling the timer and move the hrtimer_set_expires()
invocation into the NOHZ only code path which is not affected as it merily
uses the hrtimer as next event storage so code pathes can be shared with
the NOHZ + HIGHRES case.

Fixes: d4af6d93 ("nohz: Fix spurious warning when hrtimer and clockevent get out of sync")
Reported-by: N"Wan Kaike" <kaike.wan@intel.com>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Acked-by: NFrederic Weisbecker <frederic@kernel.org>
Cc: "Marciniszyn Mike" <mike.marciniszyn@intel.com>
Cc: Anna-Maria Gleixner <anna-maria@linutronix.de>
Cc: linux-rdma@vger.kernel.org
Cc: "Dalessandro Dennis" <dennis.dalessandro@intel.com>
Cc: "Fleck John" <john.fleck@intel.com>
Cc: stable@vger.kernel.org
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: "Weiny Ira" <ira.weiny@intel.com>
Cc: "linux-rdma@vger.kernel.org"
Link: https://lkml.kernel.org/r/alpine.DEB.2.21.1804241637390.1679@nanos.tec.linutronix.de
Link: https://lkml.kernel.org/r/alpine.DEB.2.21.1804242119210.1597@nanos.tec.linutronix.de

Change over clock_nanosleep syscalls to use y2038 safe
__kernel_timespec times. This will enable changing over
of these syscalls to use new y2038 safe syscalls when
the architectures define the CONFIG_64BIT_TIME.

Note that nanosleep syscall is deprecated and does not have a
plan for making it y2038 safe. But, the syscall should work as
before on 64 bit machines and on 32 bit machines, the syscall
works correctly until y2038 as before using the existing compat
syscall version. There is no new syscall for supporting 64 bit
time_t on 32 bit architectures.

Cc: linux-api@vger.kernel.org
Signed-off-by: NDeepa Dinamani <deepa.kernel@gmail.com>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>

Change over clock_settime, clock_gettime and clock_getres
syscalls to use __kernel_timespec times. This will enable
changing over of these syscalls to use new y2038 safe syscalls
when the architectures define the CONFIG_64BIT_TIME.

Cc: linux-api@vger.kernel.org
Signed-off-by: NDeepa Dinamani <deepa.kernel@gmail.com>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>

get/put_timespec64() interfaces will eventually be used for
conversions between the new y2038 safe struct __kernel_timespec
and struct timespec64.

The new y2038 safe syscalls have a common entry for native
and compat interfaces.
On compat interfaces, the high order bits of nanoseconds
should be zeroed out. This is because the application code
or the libc do not guarantee zeroing of these. If used without
zeroing, kernel might be at risk of using timespec values
incorrectly.

Note that clearing of bits is dependent on CONFIG_64BIT_TIME
for now. This is until COMPAT_USE_64BIT_TIME has been handled
correctly. x86 will be the first architecture that will use the
CONFIG_64BIT_TIME.
Signed-off-by: NDeepa Dinamani <deepa.kernel@gmail.com>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>

clock_gettime, clock_settime, clock_getres and clock_nanosleep
compat syscalls are also repurposed to provide backward compatibility
to support 32 bit time_t on 32 bit systems.

Note that nanosleep compat syscall will also be treated the same way
as the above syscalls as it shares common handler functions with
clock_nanosleep. But, there is no plan to provide y2038 safe solution
for nanosleep.
Signed-off-by: NDeepa Dinamani <deepa.kernel@gmail.com>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>

These functions are used in the repurposed compat syscalls
to provide backward compatibility for using 32 bit time_t
on 32 bit systems.
Signed-off-by: NDeepa Dinamani <deepa.kernel@gmail.com>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>

Commit a9445e47 ("posix-cpu-timers: Make set_process_cpu_timer()
more robust") moved the check into the 'if' statement. Unfortunately,
it did so on the right side of an && which means that it may get short
circuited and never evaluated. This is easily reproduced with:

$ cat loop.c
void main() {
  struct rlimit res;
  /* set the CPU time limit */
  getrlimit(RLIMIT_CPU,&res);
  res.rlim_cur = 2;
  res.rlim_max = 2;
  setrlimit(RLIMIT_CPU,&res);

  while (1);
}

Which will hang forever instead of being killed. Fix this by pulling the
evaluation out of the if statement but checking the return value instead.

Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1568337
Fixes: a9445e47 ("posix-cpu-timers: Make set_process_cpu_timer() more robust")
Signed-off-by: NLaura Abbott <labbott@redhat.com>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Cc: "Max R . P . Grossmann" <m@max.pm>
Cc: John Stultz <john.stultz@linaro.org>
Link: https://lkml.kernel.org/r/20180417215742.2521-1-labbott@redhat.com

The __current_kernel_time() function based on 'struct timespec' is no
longer recommended for new code, and the only user of this function has
been replaced by commit 6909e29f ("kdb: use __ktime_get_real_seconds
instead of __current_kernel_time").

Remove the obsolete interface.
Signed-off-by: NBaolin Wang <baolin.wang@linaro.org>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: arnd@arndb.de
Cc: sboyd@kernel.org
Cc: broonie@kernel.org
Cc: john.stultz@linaro.org
Link: https://lkml.kernel.org/r/1a9dbea7ee2cda7efe9ed330874075cf17fdbff6.1523596316.git.baolin.wang@linaro.org

Convert the clockevents driver from old-style printk() to pr_info() and
pr_cont(), to fix split kernel messages like below:

    Clockevents: could not switch to one-shot mode:
     dummy_timer is not functional.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Link: https://lkml.kernel.org/r/1522942018-14471-1-git-send-email-geert%2Brenesas@glider.be

The use of bitfields seems to confuse gcc, leading to a false-positive
warning in all compiler versions:

kernel/time/tick-sched.c: In function 'tick_nohz_idle_exit':
kernel/time/tick-sched.c:538:2: error: 'now' may be used uninitialized in this function [-Werror=maybe-uninitialized]

This introduces a temporary variable to track the flags so gcc
doesn't have to evaluate twice, eliminating the code path that
leads to the warning.

Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85301
Fixes: 1cae544d42d2 ("nohz: Gather tick_sched booleans under a common flag field")
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Use timerqueue_iterate_next() to get to the next timer in
__hrtimer_next_event_base() without browsing the timerqueue
details diredctly.

No intentional changes in functionality.
Suggested-by: NFrederic Weisbecker <frederic@kernel.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

Move the code setting ts->got_idle_tick into tick_sched_do_timer() to
avoid code duplication.

No intentional changes in functionality.
Suggested-by: NFrederic Weisbecker <frederic@kernel.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: NFrederic Weisbecker <frederic@kernel.org>

Optimize the space and leave plenty of room for further flags.
Signed-off-by: NFrederic Weisbecker <frederic@kernel.org>
[ rjw: Do not use __this_cpu_read() to access tick_stopped and add
       got_idle_tick to avoid overloading inidle ]
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>

If the tick isn't stopped, the target residency of the state selected
by the menu governor may be greater than the actual time to the next
tick and that means lost energy.

To avoid that, make tick_nohz_get_sleep_length() return the current
time to the next event (before stopping the tick) in addition to the
estimated one via an extra pointer argument and make menu_select()
use that value to refine the state selection when necessary.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NPeter Zijlstra (Intel) <peterz@infradead.org>

In order to address the issue with short idle duration predictions
by the idle governor after the scheduler tick has been stopped,
reorder the code in cpuidle_idle_call() so that the governor idle
state selection runs before tick_nohz_idle_go_idle() and use the
"nohz" hint returned by cpuidle_select() to decide whether or not
to stop the tick.

This isn't straightforward, because menu_select() invokes
tick_nohz_get_sleep_length() to get the time to the next timer
event and the number returned by the latter comes from
__tick_nohz_idle_stop_tick().  Fortunately, however, it is possible
to compute that number without actually stopping the tick and with
the help of the existing code.

Namely, tick_nohz_get_sleep_length() can be made call
tick_nohz_next_event(), introduced earlier, to get the time to the
next non-highres timer event.  If that happens, tick_nohz_next_event()
need not be called by __tick_nohz_idle_stop_tick() again.

If it turns out that the scheduler tick cannot be stopped going
forward or the next timer event is too close for the tick to be
stopped, tick_nohz_get_sleep_length() can simply return the time to
the next event currently programmed into the corresponding clock
event device.

In addition to knowing the return value of tick_nohz_next_event(),
however, tick_nohz_get_sleep_length() needs to know the time to the
next highres timer event, but with the scheduler tick timer excluded,
which can be computed with the help of hrtimer_get_next_event().

That minimum of that number and the tick_nohz_next_event() return
value is the total time to the next timer event with the assumption
that the tick will be stopped.  It can be returned to the idle
governor which can use it for predicting idle duration (under the
assumption that the tick will be stopped) and deciding whether or
not it makes sense to stop the tick before putting the CPU into the
selected idle state.

With the above, the sleep_length field in struct tick_sched is not
necessary any more, so drop it.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=199227Reported-by: NDoug Smythies <dsmythies@telus.net>
Reported-by: NThomas Ilsche <thomas.ilsche@tu-dresden.de>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: NFrederic Weisbecker <frederic@kernel.org>

The next set of changes will need to compute the time to the next
hrtimer event over all hrtimers except for the scheduler tick one.

To that end introduce a new helper function,
hrtimer_next_event_without(), for computing the time until the next
hrtimer event over all timers except for one and modify the underlying
code in __hrtimer_next_event_base() to prepare it for being called by
that new function.

No intentional changes in functionality.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: NFrederic Weisbecker <frederic@kernel.org>