The WakeOvfIsDeferred code path in __call_rcu_nocb_wake() attempts to
wake rcuog when the callback count exceeds qhimark and callbacks aren't
done with their GP (newly queued or awaiting GP). However, a lot of
testing proves this wake is always redundant or useless.
In the flooding case, rcuog is always waiting for a GP to finish. So
waking up the rcuog thread is pointless. The timer wakeup adds overhead,
rcuog simply wakes up and goes back to sleep achieving nothing.
This path also adds a full memory barrier, and additional timer expiry
modifications unnecessarily.
The root cause is that WakeOvfIsDeferred fires when
!rcu_segcblist_ready_cbs() (GP not complete), but waking rcuog cannot
accelerate GP completion.
This commit therefore removes this path.
Tested with rcutorture scenarios: TREE01, TREE05, TREE08 (all NOCB
configurations) - all pass. Also stress tested using a kernel module
that floods call_rcu() to trigger the overload conditions and made the
observations confirming the findings.
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
RCU re-initializes the deferred QS irq work everytime before attempting
to queue it. However there are situations where the irq work is
attempted to be queued even though it is already queued. In that case
re-initializing messes-up with the irq work queue that is about to be
handled.
The chances for that to happen are higher when the architecture doesn't
support self-IPIs and irq work are then all lazy, such as with the
following sequence:
1) rcu_read_unlock() is called when IRQs are disabled and there is a
grace period involving blocked tasks on the node. The irq work
is then initialized and queued.
2) The related tasks are unblocked and the CPU quiescent state
is reported. rdp->defer_qs_iw_pending is reset to DEFER_QS_IDLE,
allowing the irq work to be requeued in the future (note the previous
one hasn't fired yet).
3) A new grace period starts and the node has blocked tasks.
4) rcu_read_unlock() is called when IRQs are disabled again. The irq work
is re-initialized (but it's queued! and its node is cleared) and
requeued. Which means it's requeued to itself.
5) The irq work finally fires with the tick. But since it was requeued
to itself, it loops and hangs.
Fix this with initializing the irq work only once before the CPU boots.
Fixes: b41642c877 ("rcu: Fix rcu_read_unlock() deadloop due to IRQ work")
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202508071303.c1134cce-lkp@intel.com
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Joel Fernandes <joelagnelf@nvidia.com>
Signed-off-by: Neeraj Upadhyay (AMD) <neeraj.upadhyay@kernel.org>
During rcu_read_unlock_special(), if this happens during irq_exit(), we
can lockup if an IPI is issued. This is because the IPI itself triggers
the irq_exit() path causing a recursive lock up.
This is precisely what Xiongfeng found when invoking a BPF program on
the trace_tick_stop() tracepoint As shown in the trace below. Fix by
managing the irq_work state correctly.
irq_exit()
__irq_exit_rcu()
/* in_hardirq() returns false after this */
preempt_count_sub(HARDIRQ_OFFSET)
tick_irq_exit()
tick_nohz_irq_exit()
tick_nohz_stop_sched_tick()
trace_tick_stop() /* a bpf prog is hooked on this trace point */
__bpf_trace_tick_stop()
bpf_trace_run2()
rcu_read_unlock_special()
/* will send a IPI to itself */
irq_work_queue_on(&rdp->defer_qs_iw, rdp->cpu);
A simple reproducer can also be obtained by doing the following in
tick_irq_exit(). It will hang on boot without the patch:
static inline void tick_irq_exit(void)
{
+ rcu_read_lock();
+ WRITE_ONCE(current->rcu_read_unlock_special.b.need_qs, true);
+ rcu_read_unlock();
+
Reported-by: Xiongfeng Wang <wangxiongfeng2@huawei.com>
Closes: https://lore.kernel.org/all/9acd5f9f-6732-7701-6880-4b51190aa070@huawei.com/
Tested-by: Qi Xi <xiqi2@huawei.com>
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
Reviewed-by: "Paul E. McKenney" <paulmck@kernel.org>
Reported-by: Linux Kernel Functional Testing <lkft@linaro.org>
[neeraj: Apply Frederic's suggested fix for PREEMPT_RT]
Signed-off-by: Neeraj Upadhyay (AMD) <neeraj.upadhyay@kernel.org>
Currently, the ->gpwrap is not tested (at all per my testing) due to the
requirement of a large delta between a CPU's rdp->gp_seq and its node's
rnp->gpseq.
This results in no testing of ->gpwrap being set. This patch by default
adds 5 minutes of testing with ->gpwrap forced by lowering the delta
between rdp->gp_seq and rnp->gp_seq to just 8 GPs. All of this is
configurable, including the active time for the setting and a full
testing cycle.
By default, the first 25 minutes of a test will have the _default_
behavior there is right now (ULONG_MAX / 4) delta. Then for 5 minutes,
we switch to a smaller delta causing 1-2 wraps in 5 minutes. I believe
this is reasonable since we at least add a little bit of testing for
usecases where ->gpwrap is set.
[ Apply fix for Dan Carpenter's bug report on init path cleanup. ]
[ Apply kernel doc warning fix from Akira Yokosawa. ]
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
The context_tracking.state RCU_DYNTICKS subvariable has been renamed to
RCU_WATCHING, and the snapshot helpers are now prefix by
"rcu_watching". Reflect that change into the storage variables for these
snapshots.
Signed-off-by: Valentin Schneider <vschneid@redhat.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
The context_tracking.state RCU_DYNTICKS subvariable has been renamed to
RCU_WATCHING, and the snapshot helpers are now prefix by
"rcu_watching". Reflect that change into the storage variables for these
snapshots.
Signed-off-by: Valentin Schneider <vschneid@redhat.com>
Reviewed-by: Frederic Weisbecker <frederic@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>
nocb_is_setup is a rarely used field, mostly on boot and CPU hotplug.
It shouldn't occupy the middle of the rcu state hot fields cacheline.
Move it to the end and build it conditionally while at it. More cold
NOCB fields are to come.
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>
In the synchronize_rcu() common case, we will have less than
SR_MAX_USERS_WAKE_FROM_GP number of users per GP. Waking up the kworker
is pointless just to free the last injected wait head since at that point,
all the users have already been awakened.
Introduce a new counter to track this and prevent the wakeup in the
common case.
[ paulmck: Remove atomic_dec_return_release in cannot-happen state. ]
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Reviewed-by: Uladzislau Rezki (Sony) <urezki@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@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>
This patch introduces a small enhancement which allows to do a
direct wake-up of synchronize_rcu() callers. It occurs after a
completion of grace period, thus by the gp-kthread.
Number of clients is limited by the hard-coded maximum allowed
threshold. The remaining part, if still exists is deferred to
a main worker.
Link: https://lore.kernel.org/lkml/Zd0ZtNu+Rt0qXkfS@lothringen/
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Uladzislau Rezki (Sony) <urezki@gmail.com>
Currently, there are rcu_data structure fields named ->rcu_onl_gp_seq
and ->rcu_ofl_gp_seq that track the rcu_state.gp_flags field at the
time of the corresponding CPU's last online or offline operation,
respectively. However, this information is not particularly useful.
It would be better to instead track the grace period state kept
in rcu_state.gp_state. This would also be consistent with the
initialization in rcu_boot_init_percpu_data(), which is to RCU_GP_CLEANED
(an rcu_state.gp_state value), and also with the diagnostics in
rcu_implicit_dynticks_qs(), whose format is consistent with an integer,
not a bitmask.
This commit therefore makes this change and changes the names to
->rcu_onl_gp_flags and ->rcu_ofl_gp_flags, respectively.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Uladzislau Rezki (Sony) <urezki@gmail.com>
The synchronize_rcu() call is going to be reworked, thus
this patch adds dedicated fields into the rcu_state structure.
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Uladzislau Rezki (Sony) <urezki@gmail.com>
TREE04 running on short iterations can produce writer stalls of the
following kind:
??? Writer stall state RTWS_EXP_SYNC(4) g3968 f0x0 ->state 0x2 cpu 0
task:rcu_torture_wri state:D stack:14568 pid:83 ppid:2 flags:0x00004000
Call Trace:
<TASK>
__schedule+0x2de/0x850
? trace_event_raw_event_rcu_exp_funnel_lock+0x6d/0xb0
schedule+0x4f/0x90
synchronize_rcu_expedited+0x430/0x670
? __pfx_autoremove_wake_function+0x10/0x10
? __pfx_synchronize_rcu_expedited+0x10/0x10
do_rtws_sync.constprop.0+0xde/0x230
rcu_torture_writer+0x4b4/0xcd0
? __pfx_rcu_torture_writer+0x10/0x10
kthread+0xc7/0xf0
? __pfx_kthread+0x10/0x10
ret_from_fork+0x2f/0x50
? __pfx_kthread+0x10/0x10
ret_from_fork_asm+0x1b/0x30
</TASK>
Waiting for an expedited grace period and polling for an expedited
grace period both are operations that internally rely on the same
workqueue performing necessary asynchronous work.
However, a dependency chain is involved between those two operations,
as depicted below:
====== CPU 0 ======= ====== CPU 1 =======
synchronize_rcu_expedited()
exp_funnel_lock()
mutex_lock(&rcu_state.exp_mutex);
start_poll_synchronize_rcu_expedited
queue_work(rcu_gp_wq, &rnp->exp_poll_wq);
synchronize_rcu_expedited_queue_work()
queue_work(rcu_gp_wq, &rew->rew_work);
wait_event() // A, wait for &rew->rew_work completion
mutex_unlock() // B
//======> switch to kworker
sync_rcu_do_polled_gp() {
synchronize_rcu_expedited()
exp_funnel_lock()
mutex_lock(&rcu_state.exp_mutex); // C, wait B
....
} // D
Since workqueues are usually implemented on top of several kworkers
handling the queue concurrently, the above situation wouldn't deadlock
most of the time because A then doesn't depend on D. But in case of
memory stress, a single kworker may end up handling alone all the works
in a serialized way. In that case the above layout becomes a problem
because A then waits for D, closing a circular dependency:
A -> D -> C -> B -> A
This however only happens when CONFIG_RCU_EXP_KTHREAD=n. Indeed
synchronize_rcu_expedited() is otherwise implemented on top of a kthread
worker while polling still relies on rcu_gp_wq workqueue, breaking the
above circular dependency chain.
Fix this with making expedited grace period to always rely on kthread
worker. The workqueue based implementation is essentially a duplicate
anyway now that the per-node initialization is performed by per-node
kthread workers.
Meanwhile the CONFIG_RCU_EXP_KTHREAD switch is still kept around to
manage the scheduler policy of these kthread workers.
Reported-by: Anna-Maria Behnsen <anna-maria@linutronix.de>
Reported-by: Thomas Gleixner <tglx@linutronix.de>
Suggested-by: Joel Fernandes <joel@joelfernandes.org>
Suggested-by: Paul E. McKenney <paulmck@kernel.org>
Suggested-by: Neeraj upadhyay <Neeraj.Upadhyay@amd.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>
When CONFIG_RCU_EXP_KTHREAD=n, the expedited grace period per node
initialization is performed in parallel via workqueues (one work per
node).
However in CONFIG_RCU_EXP_KTHREAD=y, this per node initialization is
performed by a single kworker serializing each node initialization (one
work for all nodes).
The second part is certainly less scalable and efficient beyond a single
leaf node.
To improve this, expand this single kworker into per-node kworkers. This
new layout is eventually intended to remove the workqueues based
implementation since it will essentially now become duplicate code.
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>
This mutex is currently protecting per node boost kthreads creation and
affinity setting across CPU hotplug operations.
Since the expedited kworkers will soon be split per node as well, they
will be subject to the same concurrency constraints against hotplug.
Therefore their creation and affinity tuning operations will be grouped
with those of boost kthreads and then rely on the same mutex.
To prepare for that, generalize its name.
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 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>
