Currently, running rcutorture test with torture_type=rcu fwd_progress=8
n_barrier_cbs=8 nocbs_nthreads=8 nocbs_toggle=100 onoff_interval=60
test_boost=2, will trigger the following warning:
WARNING: CPU: 19 PID: 100 at kernel/rcu/tree_nocb.h:1061 rcu_nocb_rdp_deoffload+0x292/0x2a0
RIP: 0010:rcu_nocb_rdp_deoffload+0x292/0x2a0
Call Trace:
<TASK>
? __warn+0x7e/0x120
? rcu_nocb_rdp_deoffload+0x292/0x2a0
? report_bug+0x18e/0x1a0
? handle_bug+0x3d/0x70
? exc_invalid_op+0x18/0x70
? asm_exc_invalid_op+0x1a/0x20
? rcu_nocb_rdp_deoffload+0x292/0x2a0
rcu_nocb_cpu_deoffload+0x70/0xa0
rcu_nocb_toggle+0x136/0x1c0
? __pfx_rcu_nocb_toggle+0x10/0x10
kthread+0xd1/0x100
? __pfx_kthread+0x10/0x10
ret_from_fork+0x2f/0x50
? __pfx_kthread+0x10/0x10
ret_from_fork_asm+0x1a/0x30
</TASK>
CPU0 CPU2 CPU3
//rcu_nocb_toggle //nocb_cb_wait //rcutorture
// deoffload CPU1 // process CPU1's rdp
rcu_barrier()
rcu_segcblist_entrain()
rcu_segcblist_add_len(1);
// len == 2
// enqueue barrier
// callback to CPU1's
// rdp->cblist
rcu_do_batch()
// invoke CPU1's rdp->cblist
// callback
rcu_barrier_callback()
rcu_barrier()
mutex_lock(&rcu_state.barrier_mutex);
// still see len == 2
// enqueue barrier callback
// to CPU1's rdp->cblist
rcu_segcblist_entrain()
rcu_segcblist_add_len(1);
// len == 3
// decrement len
rcu_segcblist_add_len(-2);
kthread_parkme()
// CPU1's rdp->cblist len == 1
// Warn because there is
// still a pending barrier
// trigger warning
WARN_ON_ONCE(rcu_segcblist_n_cbs(&rdp->cblist));
cpus_read_unlock();
// wait CPU1 to comes online and
// invoke barrier callback on
// CPU1 rdp's->cblist
wait_for_completion(&rcu_state.barrier_completion);
// deoffload CPU4
cpus_read_lock()
rcu_barrier()
mutex_lock(&rcu_state.barrier_mutex);
// block on barrier_mutex
// wait rcu_barrier() on
// CPU3 to unlock barrier_mutex
// but CPU3 unlock barrier_mutex
// need to wait CPU1 comes online
// when CPU1 going online will block on cpus_write_lock
The above scenario will not only trigger a WARN_ON_ONCE(), but also
trigger a deadlock.
Thanks to nocb locking, a second racing rcu_barrier() on an offline CPU
will either observe the decremented callback counter down to 0 and spare
the callback enqueue, or rcuo will observe the new callback and keep
rdp->nocb_cb_sleep to false.
Therefore check rdp->nocb_cb_sleep before parking to make sure no
further rcu_barrier() is waiting on the rdp.
Fixes: 1fcb932c8b5c ("rcu/nocb: Simplify (de-)offloading state machine")
Suggested-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Zqiang <qiang.zhang1211@gmail.com>
Reviewed-by: Neeraj Upadhyay <Neeraj.Upadhyay@amd.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
After a CPU has set itself offline and before it eventually calls
rcutree_report_cpu_dead(), there are still opportunities for callbacks
to be enqueued, for example from a softirq. When that happens on NOCB,
the rcuog wake-up is deferred through an IPI to an online CPU in order
not to call into the scheduler and risk arming the RT-bandwidth after
hrtimers have been migrated out and disabled.
But performing a synchronized IPI from a softirq is buggy as reported in
the following scenario:
WARNING: CPU: 1 PID: 26 at kernel/smp.c:633 smp_call_function_single
Modules linked in: rcutorture torture
CPU: 1 UID: 0 PID: 26 Comm: migration/1 Not tainted 6.11.0-rc1-00012-g9139f93209d1 #1
Stopper: multi_cpu_stop+0x0/0x320 <- __stop_cpus+0xd0/0x120
RIP: 0010:smp_call_function_single
<IRQ>
swake_up_one_online
__call_rcu_nocb_wake
__call_rcu_common
? rcu_torture_one_read
call_timer_fn
__run_timers
run_timer_softirq
handle_softirqs
irq_exit_rcu
? tick_handle_periodic
sysvec_apic_timer_interrupt
</IRQ>
Fix this with forcing deferred rcuog wake up through the NOCB timer when
the CPU is offline. The actual wake up will happen from
rcutree_report_cpu_dead().
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202409231644.4c55582d-lkp@intel.com
Fixes: 9139f93209d1 ("rcu/nocb: Fix RT throttling hrtimer armed from offline CPU")
Reviewed-by: "Joel Fernandes (Google)" <joel@joelfernandes.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
Pre-GP accesses performed by the update side must be ordered against
post-GP accesses performed by the readers. This is ensured by the
bypass or nocb locking on enqueue time, followed by the fully ordered
rnp locking initiated while callbacks are accelerated, and then
propagated throughout the whole GP lifecyle associated with the
callbacks.
Therefore the explicit barrier advertizing ordering between bypass
enqueue and rcuo wakeup is superfluous. If anything, it would even only
order the first bypass callback enqueue against the rcuo wakeup and
ignore all the subsequent ones.
Remove the needless barrier.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
A callback enqueuer currently wakes up the rcuo kthread if it is adding
the first non-done callback of a CPU, whether the kthread is waiting on
a grace period or not (unless the CPU is offline).
This looks like a desired behaviour because then the rcuo kthread
doesn't wait for the end of the current grace period to handle the
callback. It is accelerated right away and assigned to the next grace
period. The GP kthread is notified about that fact and iterates with
the upcoming GP without sleeping in-between.
However this best-case scenario is contradicted by a few details,
depending on the situation:
1) If the callback is a non-bypass one queued with IRQs enabled, the
wake up only occurs if no other pending callbacks are on the list.
Therefore the theoretical "optimization" actually applies on rare
occasions.
2) If the callback is a non-bypass one queued with IRQs disabled, the
situation is similar with even more uncertainty due to the deferred
wake up.
3) If the callback is lazy, a few jiffies don't make any difference.
4) If the callback is bypass, the wake up timer is programmed 2 jiffies
ahead by rcuo in case the regular pending queue has been handled
in the meantime. The rare storm of callbacks can otherwise wait for
the currently elapsing grace period to be flushed and handled.
For all those reasons, the optimization is only theoretical and
occasional. Therefore it is reasonable that callbacks enqueuers only
wake up the rcuo kthread when it is not already waiting on a grace
period to complete.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
After a CPU is marked offline and until it reaches its final trip to
idle, rcuo has several opportunities to be woken up, either because
a callback has been queued in the meantime or because
rcutree_report_cpu_dead() has issued the final deferred NOCB wake up.
If RCU-boosting is enabled, RCU kthreads are set to SCHED_FIFO policy.
And if RT-bandwidth is enabled, the related hrtimer might be armed.
However this then happens after hrtimers have been migrated at the
CPUHP_AP_HRTIMERS_DYING stage, which is broken as reported by the
following warning:
Call trace:
enqueue_hrtimer+0x7c/0xf8
hrtimer_start_range_ns+0x2b8/0x300
enqueue_task_rt+0x298/0x3f0
enqueue_task+0x94/0x188
ttwu_do_activate+0xb4/0x27c
try_to_wake_up+0x2d8/0x79c
wake_up_process+0x18/0x28
__wake_nocb_gp+0x80/0x1a0
do_nocb_deferred_wakeup_common+0x3c/0xcc
rcu_report_dead+0x68/0x1ac
cpuhp_report_idle_dead+0x48/0x9c
do_idle+0x288/0x294
cpu_startup_entry+0x34/0x3c
secondary_start_kernel+0x138/0x158
Fix this with waking up rcuo using an IPI if necessary. Since the
existing API to deal with this situation only handles swait queue, rcuo
is only woken up from offline CPUs if it's not already waiting on a
grace period. In the worst case some callbacks will just wait for a
grace period to complete before being assigned to a subsequent one.
Reported-by: "Cheng-Jui Wang (王正睿)" <Cheng-Jui.Wang@mediatek.com>
Fixes: 5c0930ccaad5 ("hrtimers: Push pending hrtimers away from outgoing CPU earlier")
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
Now that the (de-)offloading process can only apply to offline CPUs,
there is no more concurrency between rcu_core and nocb kthreads. Also
the mutation now happens on empty queues.
Therefore the state machine can be reduced to a single bit called
SEGCBLIST_OFFLOADED. Simplify the transition as follows:
* Upon offloading: queue the rdp to be added to the rcuog list and
wait for the rcuog kthread to set the SEGCBLIST_OFFLOADED bit. Unpark
rcuo kthread.
* Upon de-offloading: Park rcuo kthread. Queue the rdp to be removed
from the rcuog list and wait for the rcuog kthread to clear the
SEGCBLIST_OFFLOADED bit.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
The context_tracking.state RCU_DYNTICKS subvariable has been renamed to
RCU_WATCHING, replace "dyntick_idle" into "eqs" to drop the dyntick
reference.
Signed-off-by: Valentin Schneider <vschneid@redhat.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
RCU core can't be running anymore while in the middle of (de-)offloading
since this sort of transition now only applies to offline CPUs.
The SEGCBLIST_RCU_CORE state can therefore be removed.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
Bypass enqueue can't happen anymore in the middle of (de-)offloading
since this sort of transition now only applies to offline CPUs.
The related safety check can therefore be removed.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
Currently callbacks can be (de-)offloaded only on online CPUs. This
involves an overly elaborated state machine in order to make sure that
callbacks are always handled during the process while ensuring
synchronization between rcu_core and NOCB kthreads.
The only potential user of NOCB (de-)offloading appears to be a
nohz_full toggling interface through cpusets. And the general agreement
is now to work toward toggling the nohz_full state on offline CPUs to
simplify the whole picture.
Therefore, convert the (de-)offloading to only support offline CPUs.
This involves the following changes:
* Call rcu_barrier() before deoffloading. An offline offloaded CPU may
still carry callbacks in its queue ignored by
rcutree_migrate_callbacks(). Those callbacks must all be flushed
before switching to a regular queue because no more kthreads will
handle those before the CPU ever gets re-onlined.
This means that further calls to rcu_barrier() will find an empty
queue until the CPU goes through rcutree_report_cpu_starting(). As a
result it is guaranteed that further rcu_barrier() won't try to lock
the nocb_lock for that target and thus won't risk an imbalance.
Therefore barrier_mutex doesn't need to be locked anymore upon
deoffloading.
* Assume the queue is empty before offloading, as
rcutree_migrate_callbacks() took care of everything.
This means that further calls to rcu_barrier() will find an empty
queue until the CPU goes through rcutree_report_cpu_starting(). As a
result it is guaranteed that further rcu_barrier() won't risk a
nocb_lock imbalance.
Therefore barrier_mutex doesn't need to be locked anymore upon
offloading.
* No need to flush bypass anymore.
Further simplifications will follow in upcoming patches.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
The barrier_mutex is used currently to protect (de-)offloading
operations and prevent from nocb_lock locking imbalance in rcu_barrier()
and shrinker, and also from misordered RCU barrier invocation.
Now since RCU (de-)offloading is going to happen on offline CPUs, an RCU
barrier will have to be executed while transitionning from offloaded to
de-offloaded state. And this can't happen while holding the
barrier_mutex.
Introduce a NOCB mutex to protect (de-)offloading transitions. The
barrier_mutex is still held for now when necessary to avoid barrier
callbacks reordering and nocb_lock imbalance.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
When a NOCB CPU fails to create a nocb kthread on bringup, the CPU is
then deoffloaded. The barrier mutex is locked at this stage. It is
typically used to protect against concurrent (de-)offloading and/or
concurrent rcu_barrier() that would otherwise risk a nocb locking
imbalance. However:
* rcu_barrier() can't run concurrently if it's the boot CPU on early
boot-up.
* rcu_barrier() can run concurrently if it's a secondary CPU but it is
expected to see 0 callbacks on this target because it's the first
time it boots.
* (de-)offloading can't happen concurrently with smp_init(), as
rcutorture is initialized later, at least not before device_initcall(),
and userspace isn't available yet.
* (de-)offloading can't happen concurrently with cpu_up(), courtesy of
cpu_hotplug_lock.
But:
* The lazy shrinker might run concurrently with cpu_up(). It shouldn't
try to grab the nocb_lock and risk an imbalance due to lazy_len
supposed to be 0 but be extra cautious.
* Also be cautious against resume from hibernation potential subtleties.
So keep the locking and add some assertions and comments.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
Checking for races against concurrent (de-)offloading implies the
creation of !CONFIG_RCU_NOCB_CPU stubs to check if each relevant lock
is held. For now this only implies the nocb_lock but more are to be
expected.
Create instead a NOCB specific version of RCU_LOCKDEP_WARN() to avoid
the proliferation of stubs.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
The bypass lock contention mitigation assumes there can be at most
2 contenders on the bypass lock, following this scheme:
1) One kthread takes the bypass lock
2) Another one spins on it and increment the contended counter
3) A third one (a bypass enqueuer) sees the contended counter on and
busy loops waiting on it to decrement.
However this assumption is wrong. There can be only one CPU to find the
lock contended because call_rcu() (the bypass enqueuer) is the only
bypass lock acquire site that may not already hold the NOCB lock
beforehand, all the other sites must first contend on the NOCB lock.
Therefore step 2) is impossible.
The other problem is that the mitigation assumes that contenders all
belong to the same rdp CPU, which is also impossible for a raw spinlock.
In theory the warning could trigger if the enqueuer holds the bypass
lock and another CPU flushes the bypass queue concurrently but this is
prevented from all flush users:
1) NOCB kthreads only flush if they successfully _tried_ to lock the
bypass lock. So no contention management here.
2) Flush on callbacks migration happen remotely when the CPU is offline.
No concurrency against bypass enqueue.
3) Flush on deoffloading happen either locally with IRQs disabled or
remotely when the CPU is not yet online. No concurrency against
bypass enqueue.
4) Flush on barrier entrain happen either locally with IRQs disabled or
remotely when the CPU is offline. No concurrency against
bypass enqueue.
For those reasons, the bypass lock contention mitigation isn't needed
and is even wrong. Remove it but keep the warning reporting a contended
bypass lock on a remote CPU, to keep unexpected contention awareness.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Upon NOCB deoffloading, the rcuo kthread must be forced to sleep
until the corresponding rdp is ever offloaded again. The deoffloader
clears the SEGCBLIST_OFFLOADED flag, wakes up the rcuo kthread which
then notices that change and clears in turn its SEGCBLIST_KTHREAD_CB
flag before going to sleep, until it ever sees the SEGCBLIST_OFFLOADED
flag again, should a re-offloading happen.
Upon NOCB offloading, the rcuo kthread must be forced to wake up and
handle callbacks until the corresponding rdp is ever deoffloaded again.
The offloader sets the SEGCBLIST_OFFLOADED flag, wakes up the rcuo
kthread which then notices that change and sets in turn its
SEGCBLIST_KTHREAD_CB flag before going to check callbacks, until it
ever sees the SEGCBLIST_OFFLOADED flag cleared again, should a
de-offloading happen again.
This is all a crude ad-hoc and error-prone kthread (un-)parking
re-implementation.
Consolidate the behaviour with the appropriate API instead.
[ paulmck: Apply Qiang Zhang feedback provided in Link: below. ]
Link: https://lore.kernel.org/all/20240509074046.15629-1-qiang.zhang1211@gmail.com/
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The variable name jiffies_till_flush is too generic and therefore:
* It may shadow a global variable
* It doesn't tell on what it operates
Make the name more precise, along with the related APIs.
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Currently, only rdp_gp->nocb_timer is used, for nocb_timer of
no-rdp_gp structure, the timer_pending() is always return false,
this commit therefore need to check rdp_gp->nocb_timer in
__call_rcu_nocb_wake().
Signed-off-by: Zqiang <qiang.zhang1211@gmail.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
For the kernels built with CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y and
CONFIG_RCU_LAZY=y, the following scenarios will trigger WARN_ON_ONCE()
in the rcu_nocb_bypass_lock() and rcu_nocb_wait_contended() functions:
CPU2 CPU11
kthread
rcu_nocb_cb_kthread ksys_write
rcu_do_batch vfs_write
rcu_torture_timer_cb proc_sys_write
__kmem_cache_free proc_sys_call_handler
kmemleak_free drop_caches_sysctl_handler
delete_object_full drop_slab
__delete_object shrink_slab
put_object lazy_rcu_shrink_scan
call_rcu rcu_nocb_flush_bypass
__call_rcu_commn rcu_nocb_bypass_lock
raw_spin_trylock(&rdp->nocb_bypass_lock) fail
atomic_inc(&rdp->nocb_lock_contended);
rcu_nocb_wait_contended WARN_ON_ONCE(smp_processor_id() != rdp->cpu);
WARN_ON_ONCE(atomic_read(&rdp->nocb_lock_contended)) |
|_ _ _ _ _ _ _ _ _ _same rdp and rdp->cpu != 11_ _ _ _ _ _ _ _ _ __|
Reproduce this bug with "echo 3 > /proc/sys/vm/drop_caches".
This commit therefore uses rcu_nocb_try_flush_bypass() instead of
rcu_nocb_flush_bypass() in lazy_rcu_shrink_scan(). If the nocb_bypass
queue is being flushed, then rcu_nocb_try_flush_bypass will return
directly.
Signed-off-by: Zqiang <qiang.zhang1211@gmail.com>
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Currently the call_rcu() function interleaves NOCB and !NOCB enqueue
code in a complicated way such that:
* The bypass enqueue code may or may not have enqueued and may or may
not have locked the ->nocb_lock. Everything that follows is in a
Schrödinger locking state for the unwary reviewer's eyes.
* The was_alldone is always set but only used in NOCB related code.
* The NOCB wake up is distantly related to the locking hopefully
performed by the bypass enqueue code that did not enqueue on the
bypass list.
Unconfuse the whole and gather NOCB and !NOCB specific enqueue code to
their own functions.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Currently IRQs are disabled on call_rcu() and then depending on the
context:
* If the CPU is in nocb mode:
- If the callback is enqueued in the bypass list, IRQs are re-enabled
implictly by rcu_nocb_try_bypass()
- If the callback is enqueued in the normal list, IRQs are re-enabled
implicitly by __call_rcu_nocb_wake()
* If the CPU is NOT in nocb mode, IRQs are reenabled explicitly from call_rcu()
This makes the code a bit hard to follow, especially as it interleaves
with nocb locking.
To make the IRQ flags coverage clearer and also in order to prepare for
moving all the nocb enqueue code to its own function, always re-enable
the IRQ flags explicitly from call_rcu().
Reviewed-by: Neeraj Upadhyay (AMD) <neeraj.iitr10@gmail.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
A full barrier is issued from nocb_gp_wait() upon callbacks advancing
to order grace period completion with callbacks execution.
However these two events are already ordered by the
smp_mb__after_unlock_lock() barrier within the call to
raw_spin_lock_rcu_node() that is necessary for callbacks advancing to
happen.
The following litmus test shows the kind of guarantee that this barrier
provides:
C smp_mb__after_unlock_lock
{}
// rcu_gp_cleanup()
P0(spinlock_t *rnp_lock, int *gpnum)
{
// Grace period cleanup increase gp sequence number
spin_lock(rnp_lock);
WRITE_ONCE(*gpnum, 1);
spin_unlock(rnp_lock);
}
// nocb_gp_wait()
P1(spinlock_t *rnp_lock, spinlock_t *nocb_lock, int *gpnum, int *cb_ready)
{
int r1;
// Call rcu_advance_cbs() from nocb_gp_wait()
spin_lock(nocb_lock);
spin_lock(rnp_lock);
smp_mb__after_unlock_lock();
r1 = READ_ONCE(*gpnum);
WRITE_ONCE(*cb_ready, 1);
spin_unlock(rnp_lock);
spin_unlock(nocb_lock);
}
// nocb_cb_wait()
P2(spinlock_t *nocb_lock, int *cb_ready, int *cb_executed)
{
int r2;
// rcu_do_batch() -> rcu_segcblist_extract_done_cbs()
spin_lock(nocb_lock);
r2 = READ_ONCE(*cb_ready);
spin_unlock(nocb_lock);
// Actual callback execution
WRITE_ONCE(*cb_executed, 1);
}
P3(int *cb_executed, int *gpnum)
{
int r3;
WRITE_ONCE(*cb_executed, 2);
smp_mb();
r3 = READ_ONCE(*gpnum);
}
exists (1:r1=1 /\ 2:r2=1 /\ cb_executed=2 /\ 3:r3=0) (* Bad outcome. *)
Here the bad outcome only occurs if the smp_mb__after_unlock_lock() is
removed. This barrier orders the grace period completion against
callbacks advancing and even later callbacks invocation, thanks to the
opportunistic propagation via the ->nocb_lock to nocb_cb_wait().
Therefore the smp_mb() placed after callbacks advancing can be safely
removed.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
The LOAD-ACQUIRE access performed on rdp->nocb_cb_sleep advertizes
ordering callback execution against grace period completion. However
this is contradicted by the following:
* This LOAD-ACQUIRE doesn't pair with anything. The only counterpart
barrier that can be found is the smp_mb() placed after callbacks
advancing in nocb_gp_wait(). However the barrier is placed _after_
->nocb_cb_sleep write.
* Callbacks can be concurrently advanced between the LOAD-ACQUIRE on
->nocb_cb_sleep and the call to rcu_segcblist_extract_done_cbs() in
rcu_do_batch(), making any ordering based on ->nocb_cb_sleep broken.
* Both rcu_segcblist_extract_done_cbs() and rcu_advance_cbs() are called
under the nocb_lock, the latter hereby providing already the desired
ACQUIRE semantics.
Therefore it is safe to access ->nocb_cb_sleep with a simple compiler
barrier.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
The rcu_nocb_poll kernel boot parameter is defined via early_param(),
whose parsing functions are invoked from parse_early_param() which
is in turn invoked by setup_arch(), which is very early indeed. It
is invoked so early that the console output timestamps read 0.000000,
in other words, before time begins.
This use of early_param() means that the rcu_nocb_poll kernel boot
parameter cannot usefully be embedded into the kernel image. Yes, you
can embed it, but setup_boot_config() is invoked from start_kernel()
too late for it to be parsed.
But it makes no sense to parse this parameter so early. After all,
it cannot do anything until the rcuog kthreads are created, which is
long after rcu_init() time, let alone setup_boot_config() time.
This commit therefore switches the rcu_nocb_poll kernel boot parameter
from early_param() to __setup(), which allows boot-config parsing of
this parameter, in turn allowing it to be embedded into the kernel image.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Callbacks can only be queued as lazy on NOCB CPUs, therefore iterating
over the NOCB mask is enough for both counting and scanning. Just lock
the mostly uncontended barrier mutex on counting as well in order to
keep rcu_nocb_mask stable.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The ->lazy_len is only checked locklessly. Recheck again under the
->nocb_lock to avoid spending more time on flushing/waking if not
necessary. The ->lazy_len can still increment concurrently (from 1 to
infinity) but under the ->nocb_lock we at least know for sure if there
are lazy callbacks at all (->lazy_len > 0).
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The shrinker resets the lazy callbacks counter in order to trigger the
pending lazy queue flush though the rcuog kthread. The counter reset is
protected by the ->nocb_lock against concurrent accesses...except
for one of them. Here is a list of existing synchronized readers/writer:
1) The first lazy enqueuer (incrementing ->lazy_len to 1) does so under
->nocb_lock and ->nocb_bypass_lock.
2) The further lazy enqueuers (incrementing ->lazy_len above 1) do so
under ->nocb_bypass_lock _only_.
3) The lazy flush checks and resets to 0 under ->nocb_lock and
->nocb_bypass_lock.
The shrinker protects its ->lazy_len reset against cases 1) and 3) but
not against 2). As such, setting ->lazy_len to 0 under the ->nocb_lock
may be cancelled right away by an overwrite from an enqueuer, leading
rcuog to ignore the flush.
To avoid that, use the proper bypass flush API which takes care of all
those details.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The shrinker may run concurrently with callbacks (de-)offloading. As
such, calling rcu_nocb_lock() is very dangerous because it does a
conditional locking. The worst outcome is that rcu_nocb_lock() doesn't
lock but rcu_nocb_unlock() eventually unlocks, or the reverse, creating
an imbalance.
Fix this with protecting against (de-)offloading using the barrier mutex.
Although if the barrier mutex is contended, which should be rare, then
step aside so as not to trigger a mutex VS allocation
dependency chain.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The lazy_rcu_shrink_count() shrinker function is registered even in
kernels built with CONFIG_RCU_LAZY=n, in which case this function
uselessly consumes cycles learning that no CPU has any lazy callbacks
queued.
This commit therefore registers this shrinker function only in the kernels
built with CONFIG_RCU_LAZY=y, where it might actually do something useful.
Signed-off-by: Zqiang <qiang1.zhang@intel.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
The shrinker is used to speed up the free'ing of memory potentially held
by RCU lazy callbacks. RCU kernel module test cases show this to be
effective. Test is introduced in a later patch.
Signed-off-by: Vineeth Pillai <vineeth@bitbyteword.org>
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This consolidates the code a bit and makes it cleaner. Functionally it
is the same.
Reported-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Implement timer-based RCU callback batching (also known as lazy
callbacks). With this we save about 5-10% of power consumed due
to RCU requests that happen when system is lightly loaded or idle.
By default, all async callbacks (queued via call_rcu) are marked
lazy. An alternate API call_rcu_hurry() is provided for the few users,
for example synchronize_rcu(), that need the old behavior.
The batch is flushed whenever a certain amount of time has passed, or
the batch on a particular CPU grows too big. Also memory pressure will
flush it in a future patch.
To handle several corner cases automagically (such as rcu_barrier() and
hotplug), we re-use bypass lists which were originally introduced to
address lock contention, to handle lazy CBs as well. The bypass list
length has the lazy CB length included in it. A separate lazy CB length
counter is also introduced to keep track of the number of lazy CBs.
[ paulmck: Fix formatting of inline call_rcu_lazy() definition. ]
[ paulmck: Apply Zqiang feedback. ]
[ paulmck: Apply s/call_rcu_flush/call_rcu_hurry/ feedback from Tejun Heo. ]
Suggested-by: Paul McKenney <paulmck@kernel.org>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
In preparation for RCU lazy changes, wake up the RCU nocb gp thread if
needed after an entrain. This change prevents the RCU barrier callback
from waiting in the queue for several seconds before the lazy callbacks
in front of it are serviced.
Reported-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
When the bypass cblist gets too big or its timeout has occurred, it is
flushed into the main cblist. However, the bypass timer is still running
and the behavior is that it would eventually expire and wake the GP
thread.
Since we are going to use the bypass cblist for lazy CBs, do the wakeup
soon as the flush for "too big or too long" bypass list happens.
Otherwise, long delays can happen for callbacks which get promoted from
lazy to non-lazy.
This is a good thing to do anyway (regardless of future lazy patches),
since it makes the behavior consistent with behavior of other code paths
where flushing into the ->cblist makes the GP kthread into a
non-sleeping state quickly.
[ Frederic Weisbecker: Changes to avoid unnecessary GP-thread wakeups plus
comment changes. ]
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
In kernels built with either CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y or
CONFIG_NO_HZ_FULL=y, additional CPUs must be added to rcu_nocb_mask.
Except that kernels booted without the rcu_nocbs= will not have
allocated rcu_nocb_mask. And the current rcu_init_nohz() function uses
its need_rcu_nocb_mask and offload_all local variables to track the
rcu_nocb and nohz_full state.
But there is a much simpler approach, namely creating a cpumask pointer
to track the default and then using cpumask_available() to check the
rcu_nocb_mask state. This commit takes this approach, thereby simplifying
and shortening the rcu_init_nohz() function.
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Offline CPUs cannot be offloaded or deoffloaded. Any attempt to offload
or deoffload an offline CPU causes a message to be printed on the console,
which is good, but this message does not contain the CPU number, which
is bad. Such a CPU number can be helpful when debugging, as it gives a
clear indication that the CPU in question is in fact offline. This commit
therefore adds the CPU number to the CPU-{,de}offload failure messages.
Cc: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The show_rcu_nocb_gp_state() function is supposed to dump out the rcuog
kthread and the show_rcu_nocb_state() function is supposed to dump out
the rcuo[ps] kthread. Currently, both do a mixture, which is not optimal
for debugging, even though it does not affect functionality.
This commit therefore adjusts these two functions to focus on their
respective kthreads.
Signed-off-by: Zqiang <qiang1.zhang@intel.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Currently, if the 'rcu_nocb_poll' kernel boot parameter is enabled, all
rcuog kthreads enter polling mode. However, if all of a given group
of rcuo kthreads correspond to CPUs that have been de-offloaded, the
corresponding rcuog kthread will nonetheless still wake up periodically,
unnecessarily consuming power and perturbing workloads. Fortunately,
this situation is easily detected by the fact that the rcuog kthread's
CPU's rcu_data structure's ->nocb_head_rdp list is empty.
This commit saves power and avoids unnecessarily perturbing workloads
by putting an rcuog kthread to sleep during any time period when all of
its rcuo kthreads' CPUs are de-offloaded.
Co-developed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Zqiang <qiang1.zhang@intel.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Neeraj Upadhyay <quic_neeraju@quicinc.com>
This commit introduces a RCU_NOCB_CPU_CB_BOOST Kconfig option that
prevents rcuo kthreads from running at real-time priority, even in
kernels built with RCU_BOOST. This capability is important to devices
needing low-latency (as in a few milliseconds) response from expedited
RCU grace periods, but which are not running a classic real-time workload.
On such devices, permitting the rcuo kthreads to run at real-time priority
results in unacceptable latencies imposed on the application tasks,
which run as SCHED_OTHER.
See for example the following trace output:
<snip>
<...>-60 [006] d..1 2979.028717: rcu_batch_start: rcu_preempt CBs=34619 bl=270
<snip>
If that rcuop kthread were permitted to run at real-time SCHED_FIFO
priority, it would monopolize its CPU for hundreds of milliseconds
while invoking those 34619 RCU callback functions, which would cause an
unacceptably long latency spike for many application stacks on Android
platforms.
However, some existing real-time workloads require that callback
invocation run at SCHED_FIFO priority, for example, those running on
systems with heavy SCHED_OTHER background loads. (It is the real-time
system's administrator's responsibility to make sure that important
real-time tasks run at a higher priority than do RCU's kthreads.)
Therefore, this new RCU_NOCB_CPU_CB_BOOST Kconfig option defaults to
"y" on kernels built with PREEMPT_RT and defaults to "n" otherwise.
The effect is to preserve current behavior for real-time systems, but for
other systems to allow expedited RCU grace periods to run with real-time
priority while continuing to invoke RCU callbacks as SCHED_OTHER.
As you would expect, this RCU_NOCB_CPU_CB_BOOST Kconfig option has no
effect except on CPUs with offloaded RCU callbacks.
Signed-off-by: Uladzislau Rezki (Sony) <urezki@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Acked-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Reviewed-by: Neeraj Upadhyay <quic_neeraju@quicinc.com>
Systems built with CONFIG_RCU_NOCB_CPU=y but booted without either
the rcu_nocbs= or rcu_nohz_full= kernel-boot parameters will not have
callback offloading on any of the CPUs, nor can any of the CPUs be
switched to enable callback offloading at runtime. Although this is
intentional, it would be nice to have a way to offload all the CPUs
without having to make random bootloaders specify either the rcu_nocbs=
or the rcu_nohz_full= kernel-boot parameters.
This commit therefore provides a new CONFIG_RCU_NOCB_CPU_DEFAULT_ALL
Kconfig option that switches the default so as to offload callback
processing on all of the CPUs. This default can still be overridden
using the rcu_nocbs= and rcu_nohz_full= kernel-boot parameters.
Reviewed-by: Kalesh Singh <kaleshsingh@google.com>
Reviewed-by: Uladzislau Rezki <urezki@gmail.com>
(In v4.1, fixed issues with CONFIG maze reported by kernel test robot).
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Joel Fernandes <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Neeraj Upadhyay <quic_neeraju@quicinc.com>
If the rcuog/o[p] kthreads spawn failed, the offloaded rdp needs to
be explicitly deoffloaded, otherwise the target rdp is still considered
offloaded even though nothing actually handles the callbacks.
Signed-off-by: Zqiang <qiang1.zhang@intel.com>
Cc: Neeraj Upadhyay <quic_neeraju@quicinc.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Uladzislau Rezki <uladzislau.rezki@sony.com>
Cc: Joel Fernandes <joel@joelfernandes.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Neeraj Upadhyay <quic_neeraju@quicinc.com>
In case of failure to spawn either rcuog or rcuo[p] kthreads for a given
rdp, rcu_nocb_rdp_deoffload() needs to be called with the hotplug
lock and the barrier_mutex held. However cpus write lock is already held
while calling rcutree_prepare_cpu(). It's not possible to call
rcu_nocb_rdp_deoffload() from there with just locking the barrier_mutex
or this would result in a locking inversion against
rcu_nocb_cpu_deoffload() which holds both locks in the reverse order.
Simply solve this with inverting the locking order inside
rcu_nocb_cpu_[de]offload(). This will be a pre-requisite to toggle NOCB
states toward cpusets anyway.
Signed-off-by: Zqiang <qiang1.zhang@intel.com>
Cc: Neeraj Upadhyay <quic_neeraju@quicinc.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Uladzislau Rezki <uladzislau.rezki@sony.com>
Cc: Joel Fernandes <joel@joelfernandes.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Neeraj Upadhyay <quic_neeraju@quicinc.com>
NOCB rdp's are part of a group whose list is iterated by the
corresponding rdp leader.
This list is RCU traversed because an rdp can be either added or
deleted concurrently. Upon addition, a new iteration to the list after
a synchronization point (a pair of LOCK/UNLOCK ->nocb_gp_lock) is forced
to make sure:
1) we didn't miss a new element added in the middle of an iteration
2) we didn't ignore a whole subset of the list due to an element being
quickly deleted and then re-added.
3) we prevent from probably other surprises...
Although this layout is expected to be safe, it doesn't help anybody
to sleep well.
Simplify instead the nocb state toggling with moving the list
modification from the nocb (de-)offloading workqueue to the rcuog
kthreads instead.
Whenever the rdp leader is expected to (re-)set the SEGCBLIST_KTHREAD_GP
flag of a target rdp, the latter is queued so that the leader handles
the flag flip along with adding or deleting the target rdp to the list
to iterate. This way the list modification and iteration happen from the
same kthread and those operations can't race altogether.
As a bonus, the flags for each rdp don't need to be checked locklessly
before each iteration, which is one less opportunity to produce
nightmares.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Cc: Neeraj Upadhyay <quic_neeraju@quicinc.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Uladzislau Rezki <uladzislau.rezki@sony.com>
Cc: Joel Fernandes <joel@joelfernandes.org>
Cc: Zqiang <qiang1.zhang@intel.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Neeraj Upadhyay <quic_neeraju@quicinc.com>
The rcu_spawn_gp_kthread() function is called as an early initcall, which
means that SMP initialization hasn't happened yet and only the boot CPU is
online. Therefore, create only the NOCB kthreads related to the boot CPU.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Cc: Neeraj Upadhyay <quic_neeraju@quicinc.com>
Cc: Uladzislau Rezki <uladzislau.rezki@sony.com>
Cc: Joel Fernandes <joel@joelfernandes.org>
Cc: Boqun Feng <boqun.feng@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
This commit moves the RCU nocb initialization witness within rcu_state
to consolidate RCU's global state.
Reported-by: Paul E. McKenney <paulmck@kernel.org>
Cc: Neeraj Upadhyay <quic_neeraju@quicinc.com>
Cc: Uladzislau Rezki <uladzislau.rezki@sony.com>
Cc: Joel Fernandes <joel@joelfernandes.org>
Cc: Boqun Feng <boqun.feng@gmail.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The rcu_is_nocb_cpu() function is no longer used, so this commmit
removes it.
Reported-by: Neeraj Upadhyay <quic_neeraju@quicinc.com>
Cc: Uladzislau Rezki <uladzislau.rezki@sony.com>
Cc: Joel Fernandes <joel@joelfernandes.org>
Cc: Boqun Feng <boqun.feng@gmail.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
In some places, RCU code calls cpumask_weight() to check if any bit of a
given cpumask is set. We can do it more efficiently with cpumask_empty()
because cpumask_empty() stops traversing the cpumask as soon as it finds
first set bit, while cpumask_weight() counts all bits unconditionally.
Signed-off-by: Yury Norov <yury.norov@gmail.com>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
