[ Upstream commit b7082cdfc464bf9231300605d03eebf943dda307 ]
Bugs have been reported on 8 sockets x86 machines in which the TSC was
wrongly disabled when the system is under heavy workload.
[ 818.380354] clocksource: timekeeping watchdog on CPU336: hpet wd-wd read-back delay of 1203520ns
[ 818.436160] clocksource: wd-tsc-wd read-back delay of 181880ns, clock-skew test skipped!
[ 819.402962] clocksource: timekeeping watchdog on CPU338: hpet wd-wd read-back delay of 324000ns
[ 819.448036] clocksource: wd-tsc-wd read-back delay of 337240ns, clock-skew test skipped!
[ 819.880863] clocksource: timekeeping watchdog on CPU339: hpet read-back delay of 150280ns, attempt 3, marking unstable
[ 819.936243] tsc: Marking TSC unstable due to clocksource watchdog
[ 820.068173] TSC found unstable after boot, most likely due to broken BIOS. Use 'tsc=unstable'.
[ 820.092382] sched_clock: Marking unstable (818769414384, 1195404998)
[ 820.643627] clocksource: Checking clocksource tsc synchronization from CPU 267 to CPUs 0,4,25,70,126,430,557,564.
[ 821.067990] clocksource: Switched to clocksource hpet
This can be reproduced by running memory intensive 'stream' tests,
or some of the stress-ng subcases such as 'ioport'.
The reason for these issues is the when system is under heavy load, the
read latency of the clocksources can be very high. Even lightweight TSC
reads can show high latencies, and latencies are much worse for external
clocksources such as HPET or the APIC PM timer. These latencies can
result in false-positive clocksource-unstable determinations.
These issues were initially reported by a customer running on a production
system, and this problem was reproduced on several generations of Xeon
servers, especially when running the stress-ng test. These Xeon servers
were not production systems, but they did have the latest steppings
and firmware.
Given that the clocksource watchdog is a continual diagnostic check with
frequency of twice a second, there is no need to rush it when the system
is under heavy load. Therefore, when high clocksource read latencies
are detected, suspend the watchdog timer for 5 minutes.
Signed-off-by: Feng Tang <feng.tang@intel.com>
Acked-by: Waiman Long <longman@redhat.com>
Cc: John Stultz <jstultz@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: Feng Tang <feng.tang@intel.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9f76d59173d9d146e96c66886b671c1915a5c5e5 ]
The nanosleep syscalls use the restart_block mechanism, with a quirk:
The `type` and `rmtp`/`compat_rmtp` fields are set up unconditionally on
syscall entry, while the rest of the restart_block is only set up in the
unlikely case that the syscall is actually interrupted by a signal (or
pseudo-signal) that doesn't have a signal handler.
If the restart_block was set up by a previous syscall (futex(...,
FUTEX_WAIT, ...) or poll()) and hasn't been invalidated somehow since then,
this will clobber some of the union fields used by futex_wait_restart() and
do_restart_poll().
If userspace afterwards wrongly calls the restart_syscall syscall,
futex_wait_restart()/do_restart_poll() will read struct fields that have
been clobbered.
This doesn't actually lead to anything particularly interesting because
none of the union fields contain trusted kernel data, and
futex(..., FUTEX_WAIT, ...) and poll() aren't syscalls where it makes much
sense to apply seccomp filters to their arguments.
So the current consequences are just of the "if userspace does bad stuff,
it can damage itself, and that's not a problem" flavor.
But still, it seems like a hazard for future developers, so invalidate the
restart_block when partly setting it up in the nanosleep syscalls.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20230105134403.754986-1-jannh@google.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit d125d1349abeb46945dc5e98f7824bf688266f13 upstream.
syzbot reported a RCU stall which is caused by setting up an alarmtimer
with a very small interval and ignoring the signal. The reproducer arms the
alarm timer with a relative expiry of 8ns and an interval of 9ns. Not a
problem per se, but that's an issue when the signal is ignored because then
the timer is immediately rearmed because there is no way to delay that
rearming to the signal delivery path. See posix_timer_fn() and commit
58229a1899 ("posix-timers: Prevent softirq starvation by small intervals
and SIG_IGN") for details.
The reproducer does not set SIG_IGN explicitely, but it sets up the timers
signal with SIGCONT. That has the same effect as explicitely setting
SIG_IGN for a signal as SIGCONT is ignored if there is no handler set and
the task is not ptraced.
The log clearly shows that:
[pid 5102] --- SIGCONT {si_signo=SIGCONT, si_code=SI_TIMER, si_timerid=0, si_overrun=316014, si_int=0, si_ptr=NULL} ---
It works because the tasks are traced and therefore the signal is queued so
the tracer can see it, which delays the restart of the timer to the signal
delivery path. But then the tracer is killed:
[pid 5087] kill(-5102, SIGKILL <unfinished ...>
...
./strace-static-x86_64: Process 5107 detached
and after it's gone the stall can be observed:
syzkaller login: [ 79.439102][ C0] hrtimer: interrupt took 68471 ns
[ 184.460538][ C1] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
...
[ 184.658237][ C1] rcu: Stack dump where RCU GP kthread last ran:
[ 184.664574][ C1] Sending NMI from CPU 1 to CPUs 0:
[ 184.669821][ C0] NMI backtrace for cpu 0
[ 184.669831][ C0] CPU: 0 PID: 5108 Comm: syz-executor192 Not tainted 6.2.0-rc6-next-20230203-syzkaller #0
...
[ 184.670036][ C0] Call Trace:
[ 184.670041][ C0] <IRQ>
[ 184.670045][ C0] alarmtimer_fired+0x327/0x670
posix_timer_fn() prevents that by checking whether the interval for
timers which have the signal ignored is smaller than a jiffie and
artifically delay it by shifting the next expiry out by a jiffie. That's
accurate vs. the overrun accounting, but slightly inaccurate
vs. timer_gettimer(2).
The comment in that function says what needs to be done and there was a fix
available for the regular userspace induced SIG_IGN mechanism, but that did
not work due to the implicit ignore for SIGCONT and similar signals. This
needs to be worked on, but for now the only available workaround is to do
exactly what posix_timer_fn() does:
Increase the interval of self-rearming timers, which have their signal
ignored, to at least a jiffie.
Interestingly this has been fixed before via commit ff86bf0c65
("alarmtimer: Rate limit periodic intervals") already, but that fix got
lost in a later rework.
Reported-by: syzbot+b9564ba6e8e00694511b@syzkaller.appspotmail.com
Fixes: f2c45807d3 ("alarmtimer: Switch over to generic set/get/rearm routine")
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: John Stultz <jstultz@google.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/87k00q1no2.ffs@tglx
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit b8ac29b401 ]
The rng's random_init() function contributes the real time to the rng at
boot time, so that events can at least start in relation to something
particular in the real world. But this clock might not yet be set that
point in boot, so nothing is contributed. In addition, the relation
between minor clock changes from, say, NTP, and the cycle counter is
potentially useful entropic data.
This commit addresses this by mixing in a time stamp on calls to
settimeofday and adjtimex. No entropy is credited in doing so, so it
doesn't make initialization faster, but it is still useful input to
have.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 151c8e499f ]
Using msleep() is problematic because it's compared against
ratelimiter.c's ktime_get_coarse_boottime_ns(), which means on systems
with slow jiffies (such as UML's forced HZ=100), the result is
inaccurate. So switch to using schedule_hrtimeout().
However, hrtimer gives us access only to the traditional posix timers,
and none of the _COARSE variants. So now, rather than being too
imprecise like jiffies, it's too precise.
One solution would be to give it a large "range" value, but this will
still fire early on a loaded system. A better solution is to align the
timeout to the actual coarse timer, and then round up to the nearest
tick, plus change.
So add the timeout to the current coarse time, and then
schedule_hrtimer() until the absolute computed time.
This should hopefully reduce flakes in CI as well. Note that we keep the
retry loop in case the entire function is running behind, because the
test could still be scheduled out, by either the kernel or by the
hypervisor's kernel, in which case restarting the test and hoping to not
be scheduled out still helps.
Fixes: e7096c131e ("net: WireGuard secure network tunnel")
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 2390095113 upstream.
EXPORT_SYMBOL and __init is a bad combination because the .init.text
section is freed up after the initialization. Hence, modules cannot
use symbols annotated __init. The access to a freed symbol may end up
with kernel panic.
modpost used to detect it, but it had been broken for a decade.
Commit 28438794ab ("modpost: fix section mismatch check for exported
init/exit sections") fixed it so modpost started to warn it again, then
this showed up:
MODPOST vmlinux.symvers
WARNING: modpost: vmlinux.o(___ksymtab_gpl+tick_nohz_full_setup+0x0): Section mismatch in reference from the variable __ksymtab_tick_nohz_full_setup to the function .init.text:tick_nohz_full_setup()
The symbol tick_nohz_full_setup is exported and annotated __init
Fix this by removing the __init annotation of tick_nohz_full_setup or drop the export.
Drop the export because tick_nohz_full_setup() is only called from the
built-in code in kernel/sched/isolation.c.
Fixes: ae9e557b5b ("time: Export tick start/stop functions for rcutorture")
Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Backlund <tmb@tmb.nu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 1366992e16 upstream.
The addition of random_get_entropy_fallback() provides access to
whichever time source has the highest frequency, which is useful for
gathering entropy on platforms without available cycle counters. It's
not necessarily as good as being able to quickly access a cycle counter
that the CPU has, but it's still something, even when it falls back to
being jiffies-based.
In the event that a given arch does not define get_cycles(), falling
back to the get_cycles() default implementation that returns 0 is really
not the best we can do. Instead, at least calling
random_get_entropy_fallback() would be preferable, because that always
needs to return _something_, even falling back to jiffies eventually.
It's not as though random_get_entropy_fallback() is super high precision
or guaranteed to be entropic, but basically anything that's not zero all
the time is better than returning zero all the time.
Finally, since random_get_entropy_fallback() is used during extremely
early boot when randomizing freelists in mm_init(), it can be called
before timekeeping has been initialized. In that case there really is
nothing we can do; jiffies hasn't even started ticking yet. So just give
up and return 0.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c54bc0fc84 upstream.
When the timer base is empty, base::next_expiry is set to base::clk +
NEXT_TIMER_MAX_DELTA and base::next_expiry_recalc is false. When no timer
is queued until jiffies reaches base::next_expiry value, the warning for
not finding any expired timer and base::next_expiry_recalc is false in
__run_timers() triggers.
To prevent triggering the warning in this valid scenario
base::timers_pending needs to be added to the warning condition.
Fixes: 31cd0e119d ("timers: Recalculate next timer interrupt only when necessary")
Reported-by: Johannes Berg <johannes@sipsolutions.net>
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/20220405191732.7438-3-anna-maria@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 40e97e4296 upstream.
While running some testing on code that happened to allow the variable
tick_nohz_full_running to get set but with no "possible" NOHZ cores to
back up that setting, this warning triggered:
if (unlikely(tick_do_timer_cpu == TICK_DO_TIMER_NONE))
WARN_ON(tick_nohz_full_running);
The console was overwhemled with an endless stream of one WARN per tick
per core and there was no way to even see what was going on w/o using a
serial console to capture it and then trace it back to this.
Change it to WARN_ON_ONCE().
Fixes: 08ae95f4fd ("nohz_full: Allow the boot CPU to be nohz_full")
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20211206145950.10927-3-paul.gortmaker@windriver.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit c86ff8c55b ]
Since commit db3a34e174 ("clocksource: Retry clock read if long delays
detected") and commit 2e27e793e2 ("clocksource: Reduce clocksource-skew
threshold"), it is found that tsc clocksource fallback to hpet can
sometimes happen on both Intel and AMD systems especially when they are
running stressful benchmarking workloads. Of the 23 systems tested with
a v5.14 kernel, 10 of them have switched to hpet clock source during
the test run.
The result of falling back to hpet is a drastic reduction of performance
when running benchmarks. For example, the fio performance tests can
drop up to 70% whereas the iperf3 performance can drop up to 80%.
4 hpet fallbacks happened during bootup. They were:
[ 8.749399] clocksource: timekeeping watchdog on CPU13: hpet read-back delay of 263750ns, attempt 4, marking unstable
[ 12.044610] clocksource: timekeeping watchdog on CPU19: hpet read-back delay of 186166ns, attempt 4, marking unstable
[ 17.336941] clocksource: timekeeping watchdog on CPU28: hpet read-back delay of 182291ns, attempt 4, marking unstable
[ 17.518565] clocksource: timekeeping watchdog on CPU34: hpet read-back delay of 252196ns, attempt 4, marking unstable
Other fallbacks happen when the systems were running stressful
benchmarks. For example:
[ 2685.867873] clocksource: timekeeping watchdog on CPU117: hpet read-back delay of 57269ns, attempt 4, marking unstable
[46215.471228] clocksource: timekeeping watchdog on CPU8: hpet read-back delay of 61460ns, attempt 4, marking unstable
Commit 2e27e793e2 ("clocksource: Reduce clocksource-skew threshold"),
changed the skew margin from 100us to 50us. I think this is too small
and can easily be exceeded when running some stressful workloads on a
thermally stressed system. So it is switched back to 100us.
Even a maximum skew margin of 100us may be too small in for some systems
when booting up especially if those systems are under thermal stress. To
eliminate the case that the large skew is due to the system being too
busy slowing down the reading of both the watchdog and the clocksource,
an extra consecutive read of watchdog clock is being done to check this.
The consecutive watchdog read delay is compared against
WATCHDOG_MAX_SKEW/2. If the delay exceeds the limit, we assume that
the system is just too busy. A warning will be printed to the console
and the clock skew check is skipped for this round.
Fixes: db3a34e174 ("clocksource: Retry clock read if long delays detected")
Fixes: 2e27e793e2 ("clocksource: Reduce clocksource-skew threshold")
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2e27e793e2 ]
Currently, WATCHDOG_THRESHOLD is set to detect a 62.5-millisecond skew in
a 500-millisecond WATCHDOG_INTERVAL. This requires that clocks be skewed
by more than 12.5% in order to be marked unstable. Except that a clock
that is skewed by that much is probably destroying unsuspecting software
right and left. And given that there are now checks for false-positive
skews due to delays between reading the two clocks, it should be possible
to greatly decrease WATCHDOG_THRESHOLD, at least for fine-grained clocks
such as TSC.
Therefore, add a new uncertainty_margin field to the clocksource structure
that contains the maximum uncertainty in nanoseconds for the corresponding
clock. This field may be initialized manually, as it is for
clocksource_tsc_early and clocksource_jiffies, which is copied to
refined_jiffies. If the field is not initialized manually, it will be
computed at clock-registry time as the period of the clock in question
based on the scale and freq parameters to __clocksource_update_freq_scale()
function. If either of those two parameters are zero, the
tens-of-milliseconds WATCHDOG_THRESHOLD is used as a cowardly alternative
to dividing by zero. No matter how the uncertainty_margin field is
calculated, it is bounded below by twice WATCHDOG_MAX_SKEW, that is, by 100
microseconds.
Note that manually initialized uncertainty_margin fields are not adjusted,
but there is a WARN_ON_ONCE() that triggers if any such field is less than
twice WATCHDOG_MAX_SKEW. This WARN_ON_ONCE() is intended to discourage
production use of the one-nanosecond uncertainty_margin values that are
used to test the clock-skew code itself.
The actual clock-skew check uses the sum of the uncertainty_margin fields
of the two clocksource structures being compared. Integer overflow is
avoided because the largest computed value of the uncertainty_margin
fields is one billion (10^9), and double that value fits into an
unsigned int. However, if someone manually specifies (say) UINT_MAX,
they will get what they deserve.
Note that the refined_jiffies uncertainty_margin field is initialized to
TICK_NSEC, which means that skew checks involving this clocksource will
be sufficently forgiving. In a similar vein, the clocksource_tsc_early
uncertainty_margin field is initialized to 32*NSEC_PER_MSEC, which
replicates the current behavior and allows custom setting if needed
in order to address the rare skews detected for this clocksource in
current mainline.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Feng Tang <feng.tang@intel.com>
Link: https://lore.kernel.org/r/20210527190124.440372-4-paulmck@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 4e8c11b6b3 upstream.
Even after commit e1d7ba8735 ("time: Always make sure wall_to_monotonic
isn't positive") it is still possible to make wall_to_monotonic positive
by running the following code:
int main(void)
{
struct timespec time;
clock_gettime(CLOCK_MONOTONIC, &time);
time.tv_nsec = 0;
clock_settime(CLOCK_REALTIME, &time);
return 0;
}
The reason is that the second parameter of timespec64_compare(), ts_delta,
may be unnormalized because the delta is calculated with an open coded
substraction which causes the comparison of tv_sec to yield the wrong
result:
wall_to_monotonic = { .tv_sec = -10, .tv_nsec = 900000000 }
ts_delta = { .tv_sec = -9, .tv_nsec = -900000000 }
That makes timespec64_compare() claim that wall_to_monotonic < ts_delta,
but actually the result should be wall_to_monotonic > ts_delta.
After normalization, the result of timespec64_compare() is correct because
the tv_sec comparison is not longer misleading:
wall_to_monotonic = { .tv_sec = -10, .tv_nsec = 900000000 }
ts_delta = { .tv_sec = -10, .tv_nsec = 100000000 }
Use timespec64_sub() to ensure that ts_delta is normalized, which fixes the
issue.
Fixes: e1d7ba8735 ("time: Always make sure wall_to_monotonic isn't positive")
Signed-off-by: Yu Liao <liaoyu15@huawei.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20211213135727.1656662-1-liaoyu15@huawei.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ca7752caea upstream.
copy_process currently copies task_struct.posix_cputimers_work as-is. If a
timer interrupt arrives while handling clone and before dup_task_struct
completes then the child task will have:
1. posix_cputimers_work.scheduled = true
2. posix_cputimers_work.work queued.
copy_process clears task_struct.task_works, so (2) will have no effect and
posix_cpu_timers_work will never run (not to mention it doesn't make sense
for two tasks to share a common linked list).
Since posix_cpu_timers_work never runs, posix_cputimers_work.scheduled is
never cleared. Since scheduled is set, future timer interrupts will skip
scheduling work, with the ultimate result that the task will never receive
timer expirations.
Together, the complete flow is:
1. Task 1 calls clone(), enters kernel.
2. Timer interrupt fires, schedules task work on Task 1.
2a. task_struct.posix_cputimers_work.scheduled = true
2b. task_struct.posix_cputimers_work.work added to
task_struct.task_works.
3. dup_task_struct() copies Task 1 to Task 2.
4. copy_process() clears task_struct.task_works for Task 2.
5. Future timer interrupts on Task 2 see
task_struct.posix_cputimers_work.scheduled = true and skip scheduling
work.
Fix this by explicitly clearing contents of task_struct.posix_cputimers_work
in copy_process(). This was never meant to be shared or inherited across
tasks in the first place.
Fixes: 1fb497dd00 ("posix-cpu-timers: Provide mechanisms to defer timer handling to task_work")
Reported-by: Rhys Hiltner <rhys@justin.tv>
Signed-off-by: Michael Pratt <mpratt@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20211101210615.716522-1-mpratt@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 8c3b5e6ec0 ]
If high resolution timers are disabled the timerfd notification about a
clock was set event is not happening for all cases which use
clock_was_set_delayed() because that's a NOP for HIGHRES=n, which is wrong.
Make clock_was_set_delayed() unconditially available to fix that.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20210713135158.196661266@linutronix.de
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 627ef5ae2d ]
If __hrtimer_start_range_ns() is invoked with an already armed hrtimer then
the timer has to be canceled first and then added back. If the timer is the
first expiring timer then on removal the clockevent device is reprogrammed
to the next expiring timer to avoid that the pending expiry fires needlessly.
If the new expiry time ends up to be the first expiry again then the clock
event device has to reprogrammed again.
Avoid this by checking whether the timer is the first to expire and in that
case, keep the timer on the current CPU and delay the reprogramming up to
the point where the timer has been enqueued again.
Reported-by: Lorenzo Colitti <lorenzo@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20210713135157.873137732@linutronix.de
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 406dd42bd1 ]
When an itimer deactivates a previously armed expiration, it simply doesn't
do anything. As a result the process wide cputime counter keeps running and
the tick dependency stays set until it reaches the old ghost expiration
value.
This can be reproduced with the following snippet:
void trigger_process_counter(void)
{
struct itimerval n = {};
n.it_value.tv_sec = 100;
setitimer(ITIMER_VIRTUAL, &n, NULL);
n.it_value.tv_sec = 0;
setitimer(ITIMER_VIRTUAL, &n, NULL);
}
Fix this with resetting the relevant base expiration. This is similar to
disarming a timer.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20210726125513.271824-4-frederic@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit bb7262b295 upstream.
syzbot reported KCSAN data races vs. timer_base::timer_running being set to
NULL without holding base::lock in expire_timers().
This looks innocent and most reads are clearly not problematic, but
Frederic identified an issue which is:
int data = 0;
void timer_func(struct timer_list *t)
{
data = 1;
}
CPU 0 CPU 1
------------------------------ --------------------------
base = lock_timer_base(timer, &flags); raw_spin_unlock(&base->lock);
if (base->running_timer != timer) call_timer_fn(timer, fn, baseclk);
ret = detach_if_pending(timer, base, true); base->running_timer = NULL;
raw_spin_unlock_irqrestore(&base->lock, flags); raw_spin_lock(&base->lock);
x = data;
If the timer has previously executed on CPU 1 and then CPU 0 can observe
base->running_timer == NULL and returns, assuming the timer has completed,
but it's not guaranteed on all architectures. The comment for
del_timer_sync() makes that guarantee. Moving the assignment under
base->lock prevents this.
For non-RT kernel it's performance wise completely irrelevant whether the
store happens before or after taking the lock. For an RT kernel moving the
store under the lock requires an extra unlock/lock pair in the case that
there is a waiter for the timer, but that's not the end of the world.
Reported-by: syzbot+aa7c2385d46c5eba0b89@syzkaller.appspotmail.com
Reported-by: syzbot+abea4558531bae1ba9fe@syzkaller.appspotmail.com
Fixes: 030dcdd197 ("timers: Prepare support for PREEMPT_RT")
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lore.kernel.org/r/87lfea7gw8.fsf@nanos.tec.linutronix.de
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 1a3402d93c upstream.
Since the process wide cputime counter is started locklessly from
posix_cpu_timer_rearm(), it can be concurrently stopped by operations
on other timers from the same thread group, such as in the following
unlucky scenario:
CPU 0 CPU 1
----- -----
timer_settime(TIMER B)
posix_cpu_timer_rearm(TIMER A)
cpu_clock_sample_group()
(pct->timers_active already true)
handle_posix_cpu_timers()
check_process_timers()
stop_process_timers()
pct->timers_active = false
arm_timer(TIMER A)
tick -> run_posix_cpu_timers()
// sees !pct->timers_active, ignore
// our TIMER A
Fix this with simply locking process wide cputime counting start and
timer arm in the same block.
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Fixes: 60f2ceaa81 ("posix-cpu-timers: Remove unnecessary locking around cpu_clock_sample_group")
Cc: stable@vger.kernel.org
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit aebacb7f6c ]
31cd0e119d ("timers: Recalculate next timer interrupt only when
necessary") subtly altered get_next_timer_interrupt()'s behaviour. The
function no longer consistently returns KTIME_MAX with no timers
pending.
In order to decide if there are any timers pending we check whether the
next expiry will happen NEXT_TIMER_MAX_DELTA jiffies from now.
Unfortunately, the next expiry time and the timer base clock are no
longer updated in unison. The former changes upon certain timer
operations (enqueue, expire, detach), whereas the latter keeps track of
jiffies as they move forward. Ultimately breaking the logic above.
A simplified example:
- Upon entering get_next_timer_interrupt() with:
jiffies = 1
base->clk = 0;
base->next_expiry = NEXT_TIMER_MAX_DELTA;
'base->next_expiry == base->clk + NEXT_TIMER_MAX_DELTA', the function
returns KTIME_MAX.
- 'base->clk' is updated to the jiffies value.
- The next time we enter get_next_timer_interrupt(), taking into account
no timer operations happened:
base->clk = 1;
base->next_expiry = NEXT_TIMER_MAX_DELTA;
'base->next_expiry != base->clk + NEXT_TIMER_MAX_DELTA', the function
returns a valid expire time, which is incorrect.
This ultimately might unnecessarily rearm sched's timer on nohz_full
setups, and add latency to the system[1].
So, introduce 'base->timers_pending'[2], update it every time
'base->next_expiry' changes, and use it in get_next_timer_interrupt().
[1] See tick_nohz_stop_tick().
[2] A quick pahole check on x86_64 and arm64 shows it doesn't make
'struct timer_base' any bigger.
Fixes: 31cd0e119d ("timers: Recalculate next timer interrupt only when necessary")
Signed-off-by: Nicolas Saenz Julienne <nsaenzju@redhat.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7560c02bdf ]
Some sorts of per-CPU clock sources have a history of going out of
synchronization with each other. However, this problem has purportedy been
solved in the past ten years. Except that it is all too possible that the
problem has instead simply been made less likely, which might mean that
some of the occasional "Marking clocksource 'tsc' as unstable" messages
might be due to desynchronization. How would anyone know?
Therefore apply CPU-to-CPU synchronization checking to newly unstable
clocksource that are marked with the new CLOCK_SOURCE_VERIFY_PERCPU flag.
Lists of desynchronized CPUs are printed, with the caveat that if it
is the reporting CPU that is itself desynchronized, it will appear that
all the other clocks are wrong. Just like in real life.
Reported-by: Chris Mason <clm@fb.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Feng Tang <feng.tang@intel.com>
Link: https://lore.kernel.org/r/20210527190124.440372-2-paulmck@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit db3a34e174 ]
When the clocksource watchdog marks a clock as unstable, this might be due
to that clock being unstable or it might be due to delays that happen to
occur between the reads of the two clocks. Yes, interrupts are disabled
across those two reads, but there are no shortage of things that can delay
interrupts-disabled regions of code ranging from SMI handlers to vCPU
preemption. It would be good to have some indication as to why the clock
was marked unstable.
Therefore, re-read the watchdog clock on either side of the read from the
clock under test. If the watchdog clock shows an excessive time delta
between its pair of reads, the reads are retried.
The maximum number of retries is specified by a new kernel boot parameter
clocksource.max_cswd_read_retries, which defaults to three, that is, up to
four reads, one initial and up to three retries. If more than one retry
was required, a message is printed on the console (the occasional single
retry is expected behavior, especially in guest OSes). If the maximum
number of retries is exceeded, the clock under test will be marked
unstable. However, the probability of this happening due to various sorts
of delays is quite small. In addition, the reason (clock-read delays) for
the unstable marking will be apparent.
Reported-by: Chris Mason <clm@fb.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Feng Tang <feng.tang@intel.com>
Link: https://lore.kernel.org/r/20210527190124.440372-1-paulmck@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
