Returns a high-performance monotonically non-decreasing clock for the current
CPU. The clock returned is in nanoseconds.
It provides the following properties:
1) High performance: Many BPF schedulers call bpf_ktime_get_ns() frequently
to account for execution time and track tasks' runtime properties.
Unfortunately, in some hardware platforms, bpf_ktime_get_ns() -- which
eventually reads a hardware timestamp counter -- is neither performant nor
scalable. scx_bpf_now() aims to provide a high-performance clock by
using the rq clock in the scheduler core whenever possible.
2) High enough resolution for the BPF scheduler use cases: In most BPF
scheduler use cases, the required clock resolution is lower than the most
accurate hardware clock (e.g., rdtsc in x86). scx_bpf_now() basically
uses the rq clock in the scheduler core whenever it is valid. It considers
that the rq clock is valid from the time the rq clock is updated
(update_rq_clock) until the rq is unlocked (rq_unpin_lock).
3) Monotonically non-decreasing clock for the same CPU: scx_bpf_now()
guarantees the clock never goes backward when comparing them in the same
CPU. On the other hand, when comparing clocks in different CPUs, there
is no such guarantee -- the clock can go backward. It provides a
monotonically *non-decreasing* clock so that it would provide the same
clock values in two different scx_bpf_now() calls in the same CPU
during the same period of when the rq clock is valid.
An rq clock becomes valid when it is updated using update_rq_clock()
and invalidated when the rq is unlocked using rq_unpin_lock().
Let's suppose the following timeline in the scheduler core:
T1. rq_lock(rq)
T2. update_rq_clock(rq)
T3. a sched_ext BPF operation
T4. rq_unlock(rq)
T5. a sched_ext BPF operation
T6. rq_lock(rq)
T7. update_rq_clock(rq)
For [T2, T4), we consider that rq clock is valid (SCX_RQ_CLK_VALID is
set), so scx_bpf_now() calls during [T2, T4) (including T3) will
return the rq clock updated at T2. For duration [T4, T7), when a BPF
scheduler can still call scx_bpf_now() (T5), we consider the rq clock
is invalid (SCX_RQ_CLK_VALID is unset at T4). So when calling
scx_bpf_now() at T5, we will return a fresh clock value by calling
sched_clock_cpu() internally. Also, to prevent getting outdated rq clocks
from a previous scx scheduler, invalidate all the rq clocks when unloading
a BPF scheduler.
One example of calling scx_bpf_now(), when the rq clock is invalid
(like T5), is in scx_central [1]. The scx_central scheduler uses a BPF
timer for preemptive scheduling. In every msec, the timer callback checks
if the currently running tasks exceed their timeslice. At the beginning of
the BPF timer callback (central_timerfn in scx_central.bpf.c), scx_central
gets the current time. When the BPF timer callback runs, the rq clock could
be invalid, the same as T5. In this case, scx_bpf_now() returns a fresh
clock value rather than returning the old one (T2).
[1] https://github.com/sched-ext/scx/blob/main/scheds/c/scx_central.bpf.c
Signed-off-by: Changwoo Min <changwoo@igalia.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
When a kthread or any other task has an affinity mask that is fully
offline or unallowed, the scheduler reaffines the task to all possible
CPUs as a last resort.
This default decision doesn't mix up very well with nohz_full CPUs that
are part of the possible cpumask but don't want to be disturbed by
unbound kthreads or even detached pinned user tasks.
Make the fallback affinity setting aware of nohz_full.
Suggested-by: Michal Hocko <mhocko@suse.com>
Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Currently we check for bandwidth overflow potentially due to hotplug
operations at the end of sched_cpu_deactivate(), after the cpu going
offline has already been removed from scheduling, active_mask, etc.
This can create issues for DEADLINE tasks, as there is a substantial
race window between the start of sched_cpu_deactivate() and the moment
we possibly decide to roll-back the operation if dl_bw_deactivate()
returns failure in cpuset_cpu_inactive(). An example is a throttled
task that sees its replenishment timer firing while the cpu it was
previously running on is considered offline, but before
dl_bw_deactivate() had a chance to say no and roll-back happened.
Fix this by directly calling dl_bw_deactivate() first thing in
sched_cpu_deactivate() and do the required calculation in the former
function considering the cpu passed as an argument as offline already.
By doing so we also simplify sched_cpu_deactivate(), as there is no need
anymore for any kind of roll-back if we fail early.
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Tested-by: Waiman Long <longman@redhat.com>
Link: https://lore.kernel.org/r/Zzc1DfPhbvqDDIJR@jlelli-thinkpadt14gen4.remote.csb
For hotplug operations, DEADLINE needs to check that there is still enough
bandwidth left after removing the CPU that is going offline. We however
fail to do so currently.
Restore the correct behavior by restructuring dl_bw_manage() a bit, so
that overflow conditions (not enough bandwidth left) are properly
checked. Also account for dl_server bandwidth, i.e. discount such
bandwidth in the calculation since NORMAL tasks will be anyway moved
away from the CPU as a result of the hotplug operation.
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Tested-by: Waiman Long <longman@redhat.com>
Link: https://lore.kernel.org/r/20241114142810.794657-3-juri.lelli@redhat.com
When steal time exceeds the measured delta when updating clock_task, we
currently try to catch up the excess in future updates.
However, this results in inaccurate run times for the future things using
clock_task, in some situations, as they end up getting additional steal
time that did not actually happen.
This is because there is a window between reading the elapsed time in
update_rq_clock() and sampling the steal time in update_rq_clock_task().
If the VCPU gets preempted between those two points, any additional
steal time is accounted to the outgoing task even though the calculated
delta did not actually contain any of that "stolen" time.
When this race happens, we can end up with steal time that exceeds the
calculated delta, and the previous code would try to catch up that excess
steal time in future clock updates, which is given to the next,
incoming task, even though it did not actually have any time stolen.
This behavior is particularly bad when steal time can be very long,
which we've seen when trying to extend steal time to contain the duration
that the host was suspended [0]. When this happens, clock_task stays
frozen, during which the running task stays running for the whole
duration, since its run time doesn't increase.
However the race can happen even under normal operation.
Ideally we would read the elapsed cpu time and the steal time atomically,
to prevent this race from happening in the first place, but doing so
is non-trivial.
Since the time between those two points isn't otherwise accounted anywhere,
neither to the outgoing task nor the incoming task (because the "end of
outgoing task" and "start of incoming task" timestamps are the same),
I would argue that the right thing to do is to simply drop any excess steal
time, in order to prevent these issues.
[0] https://lore.kernel.org/kvm/20240820043543.837914-1-suleiman@google.com/
Signed-off-by: Suleiman Souhlal <suleiman@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20241118043745.1857272-1-suleiman@google.com
Scheduler raises a SCHED_SOFTIRQ to trigger a load balancing event on
from the IPI handler on the idle CPU. If the SMP function is invoked
from an idle CPU via flush_smp_call_function_queue() then the HARD-IRQ
flag is not set and raise_softirq_irqoff() needlessly wakes ksoftirqd
because soft interrupts are handled before ksoftirqd get on the CPU.
Adding a trace_printk() in nohz_csd_func() at the spot of raising
SCHED_SOFTIRQ and enabling trace events for sched_switch, sched_wakeup,
and softirq_entry (for SCHED_SOFTIRQ vector alone) helps observing the
current behavior:
<idle>-0 [000] dN.1.: nohz_csd_func: Raising SCHED_SOFTIRQ from nohz_csd_func
<idle>-0 [000] dN.4.: sched_wakeup: comm=ksoftirqd/0 pid=16 prio=120 target_cpu=000
<idle>-0 [000] .Ns1.: softirq_entry: vec=7 [action=SCHED]
<idle>-0 [000] .Ns1.: softirq_exit: vec=7 [action=SCHED]
<idle>-0 [000] d..2.: sched_switch: prev_comm=swapper/0 prev_pid=0 prev_prio=120 prev_state=R ==> next_comm=ksoftirqd/0 next_pid=16 next_prio=120
ksoftirqd/0-16 [000] d..2.: sched_switch: prev_comm=ksoftirqd/0 prev_pid=16 prev_prio=120 prev_state=S ==> next_comm=swapper/0 next_pid=0 next_prio=120
...
Use __raise_softirq_irqoff() to raise the softirq. The SMP function call
is always invoked on the requested CPU in an interrupt handler. It is
guaranteed that soft interrupts are handled at the end.
Following are the observations with the changes when enabling the same
set of events:
<idle>-0 [000] dN.1.: nohz_csd_func: Raising SCHED_SOFTIRQ for nohz_idle_balance
<idle>-0 [000] dN.1.: softirq_raise: vec=7 [action=SCHED]
<idle>-0 [000] .Ns1.: softirq_entry: vec=7 [action=SCHED]
No unnecessary ksoftirqd wakeups are seen from idle task's context to
service the softirq.
Fixes: b2a02fc43a ("smp: Optimize send_call_function_single_ipi()")
Closes: https://lore.kernel.org/lkml/fcf823f-195e-6c9a-eac3-25f870cb35ac@inria.fr/ [1]
Reported-by: Julia Lawall <julia.lawall@inria.fr>
Suggested-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lore.kernel.org/r/20241119054432.6405-5-kprateek.nayak@amd.com
The need_resched() check currently in nohz_csd_func() can be tracked
to have been added in scheduler_ipi() back in 2011 via commit
ca38062e57 ("sched: Use resched IPI to kick off the nohz idle balance")
Since then, it has travelled quite a bit but it seems like an idle_cpu()
check currently is sufficient to detect the need to bail out from an
idle load balancing. To justify this removal, consider all the following
case where an idle load balancing could race with a task wakeup:
o Since commit f3dd3f6745 ("sched: Remove the limitation of WF_ON_CPU
on wakelist if wakee cpu is idle") a target perceived to be idle
(target_rq->nr_running == 0) will return true for
ttwu_queue_cond(target) which will offload the task wakeup to the idle
target via an IPI.
In all such cases target_rq->ttwu_pending will be set to 1 before
queuing the wake function.
If an idle load balance races here, following scenarios are possible:
- The CPU is not in TIF_POLLING_NRFLAG mode in which case an actual
IPI is sent to the CPU to wake it out of idle. If the
nohz_csd_func() queues before sched_ttwu_pending(), the idle load
balance will bail out since idle_cpu(target) returns 0 since
target_rq->ttwu_pending is 1. If the nohz_csd_func() is queued after
sched_ttwu_pending() it should see rq->nr_running to be non-zero and
bail out of idle load balancing.
- The CPU is in TIF_POLLING_NRFLAG mode and instead of an actual IPI,
the sender will simply set TIF_NEED_RESCHED for the target to put it
out of idle and flush_smp_call_function_queue() in do_idle() will
execute the call function. Depending on the ordering of the queuing
of nohz_csd_func() and sched_ttwu_pending(), the idle_cpu() check in
nohz_csd_func() should either see target_rq->ttwu_pending = 1 or
target_rq->nr_running to be non-zero if there is a genuine task
wakeup racing with the idle load balance kick.
o The waker CPU perceives the target CPU to be busy
(targer_rq->nr_running != 0) but the CPU is in fact going idle and due
to a series of unfortunate events, the system reaches a case where the
waker CPU decides to perform the wakeup by itself in ttwu_queue() on
the target CPU but target is concurrently selected for idle load
balance (XXX: Can this happen? I'm not sure, but we'll consider the
mother of all coincidences to estimate the worst case scenario).
ttwu_do_activate() calls enqueue_task() which would increment
"rq->nr_running" post which it calls wakeup_preempt() which is
responsible for setting TIF_NEED_RESCHED (via a resched IPI or by
setting TIF_NEED_RESCHED on a TIF_POLLING_NRFLAG idle CPU) The key
thing to note in this case is that rq->nr_running is already non-zero
in case of a wakeup before TIF_NEED_RESCHED is set which would
lead to idle_cpu() check returning false.
In all cases, it seems that need_resched() check is unnecessary when
checking for idle_cpu() first since an impending wakeup racing with idle
load balancer will either set the "rq->ttwu_pending" or indicate a newly
woken task via "rq->nr_running".
Chasing the reason why this check might have existed in the first place,
I came across Peter's suggestion on the fist iteration of Suresh's
patch from 2011 [1] where the condition to raise the SCHED_SOFTIRQ was:
sched_ttwu_do_pending(list);
if (unlikely((rq->idle == current) &&
rq->nohz_balance_kick &&
!need_resched()))
raise_softirq_irqoff(SCHED_SOFTIRQ);
Since the condition to raise the SCHED_SOFIRQ was preceded by
sched_ttwu_do_pending() (which is equivalent of sched_ttwu_pending()) in
the current upstream kernel, the need_resched() check was necessary to
catch a newly queued task. Peter suggested modifying it to:
if (idle_cpu() && rq->nohz_balance_kick && !need_resched())
raise_softirq_irqoff(SCHED_SOFTIRQ);
where idle_cpu() seems to have replaced "rq->idle == current" check.
Even back then, the idle_cpu() check would have been sufficient to catch
a new task being enqueued. Since commit b2a02fc43a ("smp: Optimize
send_call_function_single_ipi()") overloads the interpretation of
TIF_NEED_RESCHED for TIF_POLLING_NRFLAG idling, remove the
need_resched() check in nohz_csd_func() to raise SCHED_SOFTIRQ based
on Peter's suggestion.
Fixes: b2a02fc43a ("smp: Optimize send_call_function_single_ipi()")
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20241119054432.6405-3-kprateek.nayak@amd.com
Pull sched_ext fixes from Tejun Heo:
- The fair sched class currently has a bug where its balance() returns
true telling the sched core that it has tasks to run but then NULL
from pick_task(). This makes sched core call sched_ext's pick_task()
without preceding balance() which can lead to stalls in partial mode.
For now, work around by detecting the condition and forcing the CPU
to go through another scheduling cycle.
- Add a missing newline to an error message and fix drgn introspection
tool which went out of sync.
* tag 'sched_ext-for-6.12-rc7-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext:
sched_ext: Handle cases where pick_task_scx() is called without preceding balance_scx()
sched_ext: Update scx_show_state.py to match scx_ops_bypass_depth's new type
sched_ext: Add a missing newline at the end of an error message
sched_ext dispatches tasks from the BPF scheduler from balance_scx() and
thus every pick_task_scx() call must be preceded by balance_scx(). While
this usually holds, due to a bug, there are cases where the fair class's
balance() returns true indicating that it has tasks to run on the CPU and
thus terminating balance() calls but fails to actually find the next task to
run when pick_task() is called. In such cases, pick_task_scx() can be called
without preceding balance_scx().
Detect this condition using SCX_RQ_BAL_PENDING flags. If detected, keep
running the previous task if possible and avoid stalling from entering idle
without balancing.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/Ztj_h5c2LYsdXYbA@slm.duckdns.org
In order to enable PREEMPT_DYNAMIC for PREEMPT_RT, remove PREEMPT_RT
from the 'Preemption Model' choice. Strictly speaking PREEMPT_RT is
not a change in how preemption works, but rather it makes a ton more
code preemptible.
Notably, take away NONE and VOLUNTARY options for PREEMPT_RT, they make
no sense (but are techincally possible).
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lkml.kernel.org/r/20241007075055.441622332@infradead.org
Change fair to use resched_curr_lazy(), which, when the lazy
preemption model is selected, will set TIF_NEED_RESCHED_LAZY.
This LAZY bit will be promoted to the full NEED_RESCHED bit on tick.
As such, the average delay between setting LAZY and actually
rescheduling will be TICK_NSEC/2.
In short, Lazy preemption will delay preemption for fair class but
will function as Full preemption for all the other classes, most
notably the realtime (RR/FIFO/DEADLINE) classes.
The goal is to bridge the performance gap with Voluntary, such that we
might eventually remove that option entirely.
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lkml.kernel.org/r/20241007075055.331243614@infradead.org
Idle tasks are initialized via __sched_fork() twice:
fork_idle()
copy_process()
sched_fork()
__sched_fork()
init_idle()
__sched_fork()
Instead of cleaning this up, sched_ext hacked around it. Even when analyis
and solution were provided in a discussion, nobody cared to clean this up.
init_idle() is also invoked from sched_init() to initialize the boot CPU's
idle task, which requires the __sched_fork() invocation. But this can be
trivially solved by invoking __sched_fork() before init_idle() in
sched_init() and removing the __sched_fork() invocation from init_idle().
Do so and clean up the comments explaining this historical leftover.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20241028103142.359584747@linutronix.de
Commit 98442f0ccd ("sched: Fix delayed_dequeue vs
switched_from_fair()") overlooked that __setscheduler_prio(), now
__setscheduler_class() relies on p->policy for task_should_scx(), and
moved the call before __setscheduler_params() updates it, causing it
to be using the old p->policy value.
Resolve this by changing task_should_scx() to take the policy itself
instead of a task pointer, such that __sched_setscheduler() can pass
in the updated policy.
Fixes: 98442f0ccd ("sched: Fix delayed_dequeue vs switched_from_fair()")
Signed-off-by: Aboorva Devarajan <aboorvad@linux.ibm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Tejun Heo <tj@kernel.org>
What psi needs to do on each enqueue and dequeue has gotten more
subtle, and the generic sched code trying to distill this into a bool
for the callbacks is awkward.
Pass the flags directly and let psi parse them. For that to work, the
#include "stats.h" (which has the psi callback implementations) needs
to be below the flag definitions in "sched.h". Move that section
further down, next to some of the other accounting stuff.
This also puts the ENQUEUE_SAVE/RESTORE branch behind the psi jump
label, slightly reducing overhead when PSI=y but runtime disabled.
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20241014144358.GB1021@cmpxchg.org
Overlapping fixes solving the same bug slightly differently:
7266f0a6d3 fs/bcachefs: Fix __wait_on_freeing_inode() definition of waitqueue entry
3b80552e70 bcachefs: __wait_for_freeing_inode: Switch to wait_bit_queue_entry
Use the upstream version.
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Conflicts:
kernel/sched/ext.c
There's a context conflict between this upstream commit:
3fdb9ebcec sched_ext: Start schedulers with consistent p->scx.slice values
... and this fix in sched/urgent:
98442f0ccd sched: Fix delayed_dequeue vs switched_from_fair()
Resolve it.
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Let's define the "scheduling context" as all the scheduler state
in task_struct for the task chosen to run, which we'll call the
donor task, and the "execution context" as all state required to
actually run the task.
Currently both are intertwined in task_struct. We want to
logically split these such that we can use the scheduling
context of the donor task selected to be scheduled, but use
the execution context of a different task to actually be run.
To this purpose, introduce rq->donor field to point to the
task_struct chosen from the runqueue by the scheduler, and will
be used for scheduler state, and preserve rq->curr to indicate
the execution context of the task that will actually be run.
This patch introduces the donor field as a union with curr, so it
doesn't cause the contexts to be split yet, but adds the logic to
handle everything separately.
[add additional comments and update more sched_class code to use
rq::proxy]
[jstultz: Rebased and resolved minor collisions, reworked to use
accessors, tweaked update_curr_common to use rq_proxy fixing rt
scheduling issues]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Connor O'Brien <connoro@google.com>
Signed-off-by: John Stultz <jstultz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Metin Kaya <metin.kaya@arm.com>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Tested-by: Metin Kaya <metin.kaya@arm.com>
Link: https://lore.kernel.org/r/20241009235352.1614323-8-jstultz@google.com
commit 223baf9d17 ("sched: Fix performance regression introduced by mm_cid")
introduced a per-mm/cpu current concurrency id (mm_cid), which keeps
a reference to the concurrency id allocated for each CPU. This reference
expires shortly after a 100ms delay.
These per-CPU references keep the per-mm-cid data cache-local in
situations where threads are running at least once on each CPU within
each 100ms window, thus keeping the per-cpu reference alive.
However, intermittent workloads behaving in bursts spaced by more than
100ms on each CPU exhibit bad cache locality and degraded performance
compared to purely per-cpu data indexing, because concurrency IDs are
allocated over various CPUs and cores, therefore losing cache locality
of the associated data.
Introduce the following changes to improve per-mm-cid cache locality:
- Add a "recent_cid" field to the per-mm/cpu mm_cid structure to keep
track of which mm_cid value was last used, and use it as a hint to
attempt re-allocating the same concurrency ID the next time this
mm/cpu needs to allocate a concurrency ID,
- Add a per-mm CPUs allowed mask, which keeps track of the union of
CPUs allowed for all threads belonging to this mm. This cpumask is
only set during the lifetime of the mm, never cleared, so it
represents the union of all the CPUs allowed since the beginning of
the mm lifetime (note that the mm_cpumask() is really arch-specific
and tailored to the TLB flush needs, and is thus _not_ a viable
approach for this),
- Add a per-mm nr_cpus_allowed to keep track of the weight of the
per-mm CPUs allowed mask (for fast access),
- Add a per-mm max_nr_cid to keep track of the highest number of
concurrency IDs allocated for the mm. This is used for expanding the
concurrency ID allocation within the upper bound defined by:
min(mm->nr_cpus_allowed, mm->mm_users)
When the next unused CID value reaches this threshold, stop trying
to expand the cid allocation and use the first available cid value
instead.
Spreading allocation to use all the cid values within the range
[ 0, min(mm->nr_cpus_allowed, mm->mm_users) - 1 ]
improves cache locality while preserving mm_cid compactness within the
expected user limits,
- In __mm_cid_try_get, only return cid values within the range
[ 0, mm->nr_cpus_allowed ] rather than [ 0, nr_cpu_ids ]. This
prevents allocating cids above the number of allowed cpus in
rare scenarios where cid allocation races with a concurrent
remote-clear of the per-mm/cpu cid. This improvement is made
possible by the addition of the per-mm CPUs allowed mask,
- In sched_mm_cid_migrate_to, use mm->nr_cpus_allowed rather than
t->nr_cpus_allowed. This criterion was really meant to compare
the number of mm->mm_users to the number of CPUs allowed for the
entire mm. Therefore, the prior comparison worked fine when all
threads shared the same CPUs allowed mask, but not so much in
scenarios where those threads have different masks (e.g. each
thread pinned to a single CPU). This improvement is made
possible by the addition of the per-mm CPUs allowed mask.
* Benchmarks
Each thread increments 16kB worth of 8-bit integers in bursts, with
a configurable delay between each thread's execution. Each thread run
one after the other (no threads run concurrently). The order of
thread execution in the sequence is random. The thread execution
sequence begins again after all threads have executed. The 16kB areas
are allocated with rseq_mempool and indexed by either cpu_id, mm_cid
(not cache-local), or cache-local mm_cid. Each thread is pinned to its
own core.
Testing configurations:
8-core/1-L3: Use 8 cores within a single L3
24-core/24-L3: Use 24 cores, 1 core per L3
192-core/24-L3: Use 192 cores (all cores in the system)
384-thread/24-L3: Use 384 HW threads (all HW threads in the system)
Intermittent workload delays between threads: 200ms, 10ms.
Hardware:
CPU(s): 384
On-line CPU(s) list: 0-383
Vendor ID: AuthenticAMD
Model name: AMD EPYC 9654 96-Core Processor
Thread(s) per core: 2
Core(s) per socket: 96
Socket(s): 2
Caches (sum of all):
L1d: 6 MiB (192 instances)
L1i: 6 MiB (192 instances)
L2: 192 MiB (192 instances)
L3: 768 MiB (24 instances)
Each result is an average of 5 test runs. The cache-local speedup
is calculated as: (cache-local mm_cid) / (mm_cid).
Intermittent workload delay: 200ms
per-cpu mm_cid cache-local mm_cid cache-local speedup
(ns) (ns) (ns)
8-core/1-L3 1374 19289 1336 14.4x
24-core/24-L3 2423 26721 1594 16.7x
192-core/24-L3 2291 15826 2153 7.3x
384-thread/24-L3 1874 13234 1907 6.9x
Intermittent workload delay: 10ms
per-cpu mm_cid cache-local mm_cid cache-local speedup
(ns) (ns) (ns)
8-core/1-L3 662 756 686 1.1x
24-core/24-L3 1378 3648 1035 3.5x
192-core/24-L3 1439 10833 1482 7.3x
384-thread/24-L3 1503 10570 1556 6.8x
[ This deprecates the prior "sched: NUMA-aware per-memory-map concurrency IDs"
patch series with a simpler and more general approach. ]
[ This patch applies on top of v6.12-rc1. ]
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Marco Elver <elver@google.com>
Link: https://lore.kernel.org/lkml/20240823185946.418340-1-mathieu.desnoyers@efficios.com/
Sean noted that ever since commit 152e11f6df ("sched/fair: Implement
delayed dequeue") KVM's preemption notifiers have started
mis-classifying preemption vs blocking.
Notably p->on_rq is no longer sufficient to determine if a task is
runnable or blocked -- the aforementioned commit introduces tasks that
remain on the runqueue even through they will not run again, and
should be considered blocked for many cases.
Add the task_is_runnable() helper to classify things and audit all
external users of the p->on_rq state. Also add a few comments.
Fixes: 152e11f6df ("sched/fair: Implement delayed dequeue")
Reported-by: Sean Christopherson <seanjc@google.com>
Tested-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20241010091843.GK33184@noisy.programming.kicks-ass.net
Since sched_delayed tasks remain queued even after blocking, the load
balancer can migrate them between runqueues while PSI considers them
to be asleep. As a result, it misreads the migration requeue followed
by a wakeup as a double queue:
psi: inconsistent task state! task=... cpu=... psi_flags=4 clear=. set=4
First, call psi_enqueue() after p->sched_class->enqueue_task(). A
wakeup will clear p->se.sched_delayed while a migration will not, so
psi can use that flag to tell them apart.
Then teach psi to migrate any "sleep" state when delayed-dequeue tasks
are being migrated.
Delayed-dequeue tasks can be revived by ttwu_runnable(), which will
call down with a new ENQUEUE_DELAYED. Instead of further complicating
the wakeup conditional in enqueue_task(), identify migration contexts
instead and default to wakeup handling for all other cases.
It's not just the warning in dmesg, the task state corruption causes a
permanent CPU pressure indication, which messes with workload/machine
health monitoring.
Debugged-by-and-original-fix-by: K Prateek Nayak <kprateek.nayak@amd.com>
Fixes: 152e11f6df ("sched/fair: Implement delayed dequeue")
Closes: https://lore.kernel.org/lkml/20240830123458.3557-1-spasswolf@web.de/
Closes: https://lore.kernel.org/all/cd67fbcd-d659-4822-bb90-7e8fbb40a856@molgen.mpg.de/
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Link: https://lkml.kernel.org/r/20241010193712.GC181795@cmpxchg.org
Commit 2e0199df25 ("sched/fair: Prepare exit/cleanup paths for delayed_dequeue")
and its follow up fixes try to deal with a rather unfortunate
situation where is task is enqueued in a new class, even though it
shouldn't have been. Mostly because the existing ->switched_to/from()
hooks are in the wrong place for this case.
This all led to Paul being able to trigger failures at something like
once per 10k CPU hours of RCU torture.
For now, do the ugly thing and move the code to the right place by
ignoring the switch hooks.
Note: Clean up the whole sched_class::switch*_{to,from}() thing.
Fixes: 2e0199df25 ("sched/fair: Prepare exit/cleanup paths for delayed_dequeue")
Reported-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20241003185037.GA5594@noisy.programming.kicks-ass.net
With KASAN and PREEMPT_RT enabled, calling task_work_add() in
task_tick_mm_cid() may cause the following splat.
[ 63.696416] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:48
[ 63.696416] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 610, name: modprobe
[ 63.696416] preempt_count: 10001, expected: 0
[ 63.696416] RCU nest depth: 1, expected: 1
This problem is caused by the following call trace.
sched_tick() [ acquire rq->__lock ]
-> task_tick_mm_cid()
-> task_work_add()
-> __kasan_record_aux_stack()
-> kasan_save_stack()
-> stack_depot_save_flags()
-> alloc_pages_mpol_noprof()
-> __alloc_pages_noprof()
-> get_page_from_freelist()
-> rmqueue()
-> rmqueue_pcplist()
-> __rmqueue_pcplist()
-> rmqueue_bulk()
-> rt_spin_lock()
The rq lock is a raw_spinlock_t. We can't sleep while holding
it. IOW, we can't call alloc_pages() in stack_depot_save_flags().
The task_tick_mm_cid() function with its task_work_add() call was
introduced by commit 223baf9d17 ("sched: Fix performance regression
introduced by mm_cid") in v6.4 kernel.
Fortunately, there is a kasan_record_aux_stack_noalloc() variant that
calls stack_depot_save_flags() while not allowing it to allocate
new pages. To allow task_tick_mm_cid() to use task_work without
page allocation, a new TWAF_NO_ALLOC flag is added to enable calling
kasan_record_aux_stack_noalloc() instead of kasan_record_aux_stack()
if set. The task_tick_mm_cid() function is modified to add this new flag.
The possible downside is the missing stack trace in a KASAN report due
to new page allocation required when task_work_add_noallloc() is called
which should be rare.
Fixes: 223baf9d17 ("sched: Fix performance regression introduced by mm_cid")
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20241010014432.194742-1-longman@redhat.com
During ttwu, ->select_task_rq() can be skipped if only one CPU is allowed or
migration is disabled. sched_ext schedulers may perform operations such as
direct dispatch from ->select_task_rq() path and it is useful for them to
know whether ->select_task_rq() was skipped in the ->enqueue_task() path.
Currently, sched_ext schedulers are using ENQUEUE_WAKEUP for this purpose
and end up assuming incorrectly that ->select_task_rq() was called for tasks
that are bound to a single CPU or migration disabled.
Make select_task_rq() indicate whether ->select_task_rq() was called by
setting WF_RQ_SELECTED in *wake_flags and make ttwu_do_activate() map that
to ENQUEUE_RQ_SELECTED for ->enqueue_task().
This will be used by sched_ext to fix ->select_task_rq() skip detection.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: David Vernet <void@manifault.com>
This will be used to allow select_task_rq() to indicate whether
->select_task_rq() was called by modifying *wake_flags.
This makes try_to_wake_up() call all functions that take wake_flags with
WF_TTWU set. Previously, only select_task_rq() was. Using the same flags is
more consistent, and, as the flag is only tested by ->select_task_rq()
implementations, it doesn't cause any behavior differences.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: David Vernet <void@manifault.com>
Disable the rq empty path when scx is enabled. SCX must consult the BPF
scheduler (via the dispatch path in balance) to determine if rq is empty.
This fixes stalls when scx is enabled.
Signed-off-by: Pat Somaru <patso@likewhatevs.io>
Fixes: 3dcac251b0 ("sched/core: Introduce SM_IDLE and an idle re-entry fast-path in __schedule()")
Signed-off-by: Tejun Heo <tj@kernel.org>
Pull sched_ext support from Tejun Heo:
"This implements a new scheduler class called ‘ext_sched_class’, or
sched_ext, which allows scheduling policies to be implemented as BPF
programs.
The goals of this are:
- Ease of experimentation and exploration: Enabling rapid iteration
of new scheduling policies.
- Customization: Building application-specific schedulers which
implement policies that are not applicable to general-purpose
schedulers.
- Rapid scheduler deployments: Non-disruptive swap outs of scheduling
policies in production environments"
See individual commits for more documentation, but also the cover letter
for the latest series:
Link: https://lore.kernel.org/all/20240618212056.2833381-1-tj@kernel.org/
* tag 'sched_ext-for-6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext: (110 commits)
sched: Move update_other_load_avgs() to kernel/sched/pelt.c
sched_ext: Don't trigger ops.quiescent/runnable() on migrations
sched_ext: Synchronize bypass state changes with rq lock
scx_qmap: Implement highpri boosting
sched_ext: Implement scx_bpf_dispatch[_vtime]_from_dsq()
sched_ext: Compact struct bpf_iter_scx_dsq_kern
sched_ext: Replace consume_local_task() with move_local_task_to_local_dsq()
sched_ext: Move consume_local_task() upward
sched_ext: Move sanity check and dsq_mod_nr() into task_unlink_from_dsq()
sched_ext: Reorder args for consume_local/remote_task()
sched_ext: Restructure dispatch_to_local_dsq()
sched_ext: Fix processs_ddsp_deferred_locals() by unifying DTL_INVALID handling
sched_ext: Make find_dsq_for_dispatch() handle SCX_DSQ_LOCAL_ON
sched_ext: Refactor consume_remote_task()
sched_ext: Rename scx_kfunc_set_sleepable to unlocked and relocate
sched_ext: Add missing static to scx_dump_data
sched_ext: Add missing static to scx_has_op[]
sched_ext: Temporarily work around pick_task_scx() being called without balance_scx()
sched_ext: Add a cgroup scheduler which uses flattened hierarchy
sched_ext: Add cgroup support
...
Pull MM updates from Andrew Morton:
"Along with the usual shower of singleton patches, notable patch series
in this pull request are:
- "Align kvrealloc() with krealloc()" from Danilo Krummrich. Adds
consistency to the APIs and behaviour of these two core allocation
functions. This also simplifies/enables Rustification.
- "Some cleanups for shmem" from Baolin Wang. No functional changes -
mode code reuse, better function naming, logic simplifications.
- "mm: some small page fault cleanups" from Josef Bacik. No
functional changes - code cleanups only.
- "Various memory tiering fixes" from Zi Yan. A small fix and a
little cleanup.
- "mm/swap: remove boilerplate" from Yu Zhao. Code cleanups and
simplifications and .text shrinkage.
- "Kernel stack usage histogram" from Pasha Tatashin and Shakeel
Butt. This is a feature, it adds new feilds to /proc/vmstat such as
$ grep kstack /proc/vmstat
kstack_1k 3
kstack_2k 188
kstack_4k 11391
kstack_8k 243
kstack_16k 0
which tells us that 11391 processes used 4k of stack while none at
all used 16k. Useful for some system tuning things, but
partivularly useful for "the dynamic kernel stack project".
- "kmemleak: support for percpu memory leak detect" from Pavel
Tikhomirov. Teaches kmemleak to detect leaksage of percpu memory.
- "mm: memcg: page counters optimizations" from Roman Gushchin. "3
independent small optimizations of page counters".
- "mm: split PTE/PMD PT table Kconfig cleanups+clarifications" from
David Hildenbrand. Improves PTE/PMD splitlock detection, makes
powerpc/8xx work correctly by design rather than by accident.
- "mm: remove arch_make_page_accessible()" from David Hildenbrand.
Some folio conversions which make arch_make_page_accessible()
unneeded.
- "mm, memcg: cg2 memory{.swap,}.peak write handlers" fro David
Finkel. Cleans up and fixes our handling of the resetting of the
cgroup/process peak-memory-use detector.
- "Make core VMA operations internal and testable" from Lorenzo
Stoakes. Rationalizaion and encapsulation of the VMA manipulation
APIs. With a view to better enable testing of the VMA functions,
even from a userspace-only harness.
- "mm: zswap: fixes for global shrinker" from Takero Funaki. Fix
issues in the zswap global shrinker, resulting in improved
performance.
- "mm: print the promo watermark in zoneinfo" from Kaiyang Zhao. Fill
in some missing info in /proc/zoneinfo.
- "mm: replace follow_page() by folio_walk" from David Hildenbrand.
Code cleanups and rationalizations (conversion to folio_walk())
resulting in the removal of follow_page().
- "improving dynamic zswap shrinker protection scheme" from Nhat
Pham. Some tuning to improve zswap's dynamic shrinker. Significant
reductions in swapin and improvements in performance are shown.
- "mm: Fix several issues with unaccepted memory" from Kirill
Shutemov. Improvements to the new unaccepted memory feature,
- "mm/mprotect: Fix dax puds" from Peter Xu. Implements mprotect on
DAX PUDs. This was missing, although nobody seems to have notied
yet.
- "Introduce a store type enum for the Maple tree" from Sidhartha
Kumar. Cleanups and modest performance improvements for the maple
tree library code.
- "memcg: further decouple v1 code from v2" from Shakeel Butt. Move
more cgroup v1 remnants away from the v2 memcg code.
- "memcg: initiate deprecation of v1 features" from Shakeel Butt.
Adds various warnings telling users that memcg v1 features are
deprecated.
- "mm: swap: mTHP swap allocator base on swap cluster order" from
Chris Li. Greatly improves the success rate of the mTHP swap
allocation.
- "mm: introduce numa_memblks" from Mike Rapoport. Moves various
disparate per-arch implementations of numa_memblk code into generic
code.
- "mm: batch free swaps for zap_pte_range()" from Barry Song. Greatly
improves the performance of munmap() of swap-filled ptes.
- "support large folio swap-out and swap-in for shmem" from Baolin
Wang. With this series we no longer split shmem large folios into
simgle-page folios when swapping out shmem.
- "mm/hugetlb: alloc/free gigantic folios" from Yu Zhao. Nice
performance improvements and code reductions for gigantic folios.
- "support shmem mTHP collapse" from Baolin Wang. Adds support for
khugepaged's collapsing of shmem mTHP folios.
- "mm: Optimize mseal checks" from Pedro Falcato. Fixes an mprotect()
performance regression due to the addition of mseal().
- "Increase the number of bits available in page_type" from Matthew
Wilcox. Increases the number of bits available in page_type!
- "Simplify the page flags a little" from Matthew Wilcox. Many legacy
page flags are now folio flags, so the page-based flags and their
accessors/mutators can be removed.
- "mm: store zero pages to be swapped out in a bitmap" from Usama
Arif. An optimization which permits us to avoid writing/reading
zero-filled zswap pages to backing store.
- "Avoid MAP_FIXED gap exposure" from Liam Howlett. Fixes a race
window which occurs when a MAP_FIXED operqtion is occurring during
an unrelated vma tree walk.
- "mm: remove vma_merge()" from Lorenzo Stoakes. Major rotorooting of
the vma_merge() functionality, making ot cleaner, more testable and
better tested.
- "misc fixups for DAMON {self,kunit} tests" from SeongJae Park.
Minor fixups of DAMON selftests and kunit tests.
- "mm: memory_hotplug: improve do_migrate_range()" from Kefeng Wang.
Code cleanups and folio conversions.
- "Shmem mTHP controls and stats improvements" from Ryan Roberts.
Cleanups for shmem controls and stats.
- "mm: count the number of anonymous THPs per size" from Barry Song.
Expose additional anon THP stats to userspace for improved tuning.
- "mm: finish isolate/putback_lru_page()" from Kefeng Wang: more
folio conversions and removal of now-unused page-based APIs.
- "replace per-quota region priorities histogram buffer with
per-context one" from SeongJae Park. DAMON histogram
rationalization.
- "Docs/damon: update GitHub repo URLs and maintainer-profile" from
SeongJae Park. DAMON documentation updates.
- "mm/vdpa: correct misuse of non-direct-reclaim __GFP_NOFAIL and
improve related doc and warn" from Jason Wang: fixes usage of page
allocator __GFP_NOFAIL and GFP_ATOMIC flags.
- "mm: split underused THPs" from Yu Zhao. Improve THP=always policy.
This was overprovisioning THPs in sparsely accessed memory areas.
- "zram: introduce custom comp backends API" frm Sergey Senozhatsky.
Add support for zram run-time compression algorithm tuning.
- "mm: Care about shadow stack guard gap when getting an unmapped
area" from Mark Brown. Fix up the various arch_get_unmapped_area()
implementations to better respect guard areas.
- "Improve mem_cgroup_iter()" from Kinsey Ho. Improve the reliability
of mem_cgroup_iter() and various code cleanups.
- "mm: Support huge pfnmaps" from Peter Xu. Extends the usage of huge
pfnmap support.
- "resource: Fix region_intersects() vs add_memory_driver_managed()"
from Huang Ying. Fix a bug in region_intersects() for systems with
CXL memory.
- "mm: hwpoison: two more poison recovery" from Kefeng Wang. Teaches
a couple more code paths to correctly recover from the encountering
of poisoned memry.
- "mm: enable large folios swap-in support" from Barry Song. Support
the swapin of mTHP memory into appropriately-sized folios, rather
than into single-page folios"
* tag 'mm-stable-2024-09-20-02-31' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (416 commits)
zram: free secondary algorithms names
uprobes: turn xol_area->pages[2] into xol_area->page
uprobes: introduce the global struct vm_special_mapping xol_mapping
Revert "uprobes: use vm_special_mapping close() functionality"
mm: support large folios swap-in for sync io devices
mm: add nr argument in mem_cgroup_swapin_uncharge_swap() helper to support large folios
mm: fix swap_read_folio_zeromap() for large folios with partial zeromap
mm/debug_vm_pgtable: Use pxdp_get() for accessing page table entries
set_memory: add __must_check to generic stubs
mm/vma: return the exact errno in vms_gather_munmap_vmas()
memcg: cleanup with !CONFIG_MEMCG_V1
mm/show_mem.c: report alloc tags in human readable units
mm: support poison recovery from copy_present_page()
mm: support poison recovery from do_cow_fault()
resource, kunit: add test case for region_intersects()
resource: make alloc_free_mem_region() works for iomem_resource
mm: z3fold: deprecate CONFIG_Z3FOLD
vfio/pci: implement huge_fault support
mm/arm64: support large pfn mappings
mm/x86: support large pfn mappings
...
Pull scheduler updates from Ingo Molnar:
- Implement the SCHED_DEADLINE server infrastructure - Daniel Bristot
de Oliveira's last major contribution to the kernel:
"SCHED_DEADLINE servers can help fixing starvation issues of low
priority tasks (e.g., SCHED_OTHER) when higher priority tasks
monopolize CPU cycles. Today we have RT Throttling; DEADLINE
servers should be able to replace and improve that."
(Daniel Bristot de Oliveira, Peter Zijlstra, Joel Fernandes, Youssef
Esmat, Huang Shijie)
- Preparatory changes for sched_ext integration:
- Use set_next_task(.first) where required
- Fix up set_next_task() implementations
- Clean up DL server vs. core sched
- Split up put_prev_task_balance()
- Rework pick_next_task()
- Combine the last put_prev_task() and the first set_next_task()
- Rework dl_server
- Add put_prev_task(.next)
(Peter Zijlstra, with a fix by Tejun Heo)
- Complete the EEVDF transition and refine EEVDF scheduling:
- Implement delayed dequeue
- Allow shorter slices to wakeup-preempt
- Use sched_attr::sched_runtime to set request/slice suggestion
- Document the new feature flags
- Remove unused and duplicate-functionality fields
- Simplify & unify pick_next_task_fair()
- Misc debuggability enhancements
(Peter Zijlstra, with fixes/cleanups by Dietmar Eggemann, Valentin
Schneider and Chuyi Zhou)
- Initialize the vruntime of a new task when it is first enqueued,
resulting in significant decrease in latency of newly woken tasks
(Zhang Qiao)
- Introduce SM_IDLE and an idle re-entry fast-path in __schedule()
(K Prateek Nayak, Peter Zijlstra)
- Clean up and clarify the usage of Clean up usage of rt_task()
(Qais Yousef)
- Preempt SCHED_IDLE entities in strict cgroup hierarchies
(Tianchen Ding)
- Clarify the documentation of time units for deadline scheduler
parameters (Christian Loehle)
- Remove the HZ_BW chicken-bit feature flag introduced a year ago,
the original change seems to be working fine (Phil Auld)
- Misc fixes and cleanups (Chen Yu, Dan Carpenter, Huang Shijie,
Peilin He, Qais Yousefm and Vincent Guittot)
* tag 'sched-core-2024-09-19' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (64 commits)
sched/cpufreq: Use NSEC_PER_MSEC for deadline task
cpufreq/cppc: Use NSEC_PER_MSEC for deadline task
sched/deadline: Clarify nanoseconds in uapi
sched/deadline: Convert schedtool example to chrt
sched/debug: Fix the runnable tasks output
sched: Fix sched_delayed vs sched_core
kernel/sched: Fix util_est accounting for DELAY_DEQUEUE
kthread: Fix task state in kthread worker if being frozen
sched/pelt: Use rq_clock_task() for hw_pressure
sched/fair: Move effective_cpu_util() and effective_cpu_util() in fair.c
sched/core: Introduce SM_IDLE and an idle re-entry fast-path in __schedule()
sched: Add put_prev_task(.next)
sched: Rework dl_server
sched: Combine the last put_prev_task() and the first set_next_task()
sched: Rework pick_next_task()
sched: Split up put_prev_task_balance()
sched: Clean up DL server vs core sched
sched: Fixup set_next_task() implementations
sched: Use set_next_task(.first) where required
sched/fair: Properly deactivate sched_delayed task upon class change
...
Pull RCU updates from Neeraj Upadhyay:
"Context tracking:
- rename context tracking state related symbols and remove references
to "dynticks" in various context tracking state variables and
related helpers
- force context_tracking_enabled_this_cpu() to be inlined to avoid
leaving a noinstr section
CSD lock:
- enhance CSD-lock diagnostic reports
- add an API to provide an indication of ongoing CSD-lock stall
nocb:
- update and simplify RCU nocb code to handle (de-)offloading of
callbacks only for offline CPUs
- fix RT throttling hrtimer being armed from offline CPU
rcutorture:
- remove redundant rcu_torture_ops get_gp_completed fields
- add SRCU ->same_gp_state and ->get_comp_state functions
- add generic test for NUM_ACTIVE_*RCU_POLL* for testing RCU and SRCU
polled grace periods
- add CFcommon.arch for arch-specific Kconfig options
- print number of update types in rcu_torture_write_types()
- add rcutree.nohz_full_patience_delay testing to the TREE07 scenario
- add a stall_cpu_repeat module parameter to test repeated CPU stalls
- add argument to limit number of CPUs a guest OS can use in
torture.sh
rcustall:
- abbreviate RCU CPU stall warnings during CSD-lock stalls
- Allow dump_cpu_task() to be called without disabling preemption
- defer printing stall-warning backtrace when holding rcu_node lock
srcu:
- make SRCU gp seq wrap-around faster
- add KCSAN checks for concurrent updates to ->srcu_n_exp_nodelay and
->reschedule_count which are used in heuristics governing
auto-expediting of normal SRCU grace periods and
grace-period-state-machine delays
- mark idle SRCU-barrier callbacks to help identify stuck
SRCU-barrier callback
rcu tasks:
- remove RCU Tasks Rude asynchronous APIs as they are no longer used
- stop testing RCU Tasks Rude asynchronous APIs
- fix access to non-existent percpu regions
- check processor-ID assumptions during chosen CPU calculation for
callback enqueuing
- update description of rtp->tasks_gp_seq grace-period sequence
number
- add rcu_barrier_cb_is_done() to identify whether a given
rcu_barrier callback is stuck
- mark idle Tasks-RCU-barrier callbacks
- add *torture_stats_print() functions to print detailed diagnostics
for Tasks-RCU variants
- capture start time of rcu_barrier_tasks*() operation to help
distinguish a hung barrier operation from a long series of barrier
operations
refscale:
- add a TINY scenario to support tests of Tiny RCU and Tiny
SRCU
- optimize process_durations() operation
rcuscale:
- dump stacks of stalled rcu_scale_writer() instances and
grace-period statistics when rcu_scale_writer() stalls
- mark idle RCU-barrier callbacks to identify stuck RCU-barrier
callbacks
- print detailed grace-period and barrier diagnostics on
rcu_scale_writer() hangs for Tasks-RCU variants
- warn if async module parameter is specified for RCU implementations
that do not have async primitives such as RCU Tasks Rude
- make all writer tasks report upon hang
- tolerate repeated GFP_KERNEL failure in rcu_scale_writer()
- use special allocator for rcu_scale_writer()
- NULL out top-level pointers to heap memory to avoid double-free
bugs on modprobe failures
- maintain per-task instead of per-CPU callbacks count to avoid any
issues with migration of either tasks or callbacks
- constify struct ref_scale_ops
Fixes:
- use system_unbound_wq for kfree_rcu work to avoid disturbing
isolated CPUs
Misc:
- warn on unexpected rcu_state.srs_done_tail state
- better define "atomic" for list_replace_rcu() and
hlist_replace_rcu() routines
- annotate struct kvfree_rcu_bulk_data with __counted_by()"
* tag 'rcu.release.v6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/rcu/linux: (90 commits)
rcu: Defer printing stall-warning backtrace when holding rcu_node lock
rcu/nocb: Remove superfluous memory barrier after bypass enqueue
rcu/nocb: Conditionally wake up rcuo if not already waiting on GP
rcu/nocb: Fix RT throttling hrtimer armed from offline CPU
rcu/nocb: Simplify (de-)offloading state machine
context_tracking: Tag context_tracking_enabled_this_cpu() __always_inline
context_tracking, rcu: Rename rcu_dyntick trace event into rcu_watching
rcu: Update stray documentation references to rcu_dynticks_eqs_{enter, exit}()
rcu: Rename rcu_momentary_dyntick_idle() into rcu_momentary_eqs()
rcu: Rename rcu_implicit_dynticks_qs() into rcu_watching_snap_recheck()
rcu: Rename dyntick_save_progress_counter() into rcu_watching_snap_save()
rcu: Rename struct rcu_data .exp_dynticks_snap into .exp_watching_snap
rcu: Rename struct rcu_data .dynticks_snap into .watching_snap
rcu: Rename rcu_dynticks_zero_in_eqs() into rcu_watching_zero_in_eqs()
rcu: Rename rcu_dynticks_in_eqs_since() into rcu_watching_snap_stopped_since()
rcu: Rename rcu_dynticks_in_eqs() into rcu_watching_snap_in_eqs()
rcu: Rename rcu_dynticks_eqs_online() into rcu_watching_online()
context_tracking, rcu: Rename rcu_dynticks_curr_cpu_in_eqs() into rcu_is_watching_curr_cpu()
context_tracking, rcu: Rename rcu_dynticks_task*() into rcu_task*()
refscale: Constify struct ref_scale_ops
...
Pull in tip/sched/core to resolve two merge conflicts:
- 96fd6c65ef ("sched: Factor out update_other_load_avgs() from __update_blocked_others()")
5d871a6399 ("sched/fair: Move effective_cpu_util() and effective_cpu_util() in fair.c")
A simple context conflict. The former added __update_blocked_others() in
the same #ifdef CONFIG_SMP block that effective_cpu_util() and
sched_cpu_util() are in and the latter moved those functions to fair.c.
This makes __update_blocked_others() more out of place. Will follow up
with a patch to relocate.
- 96fd6c65ef ("sched: Factor out update_other_load_avgs() from __update_blocked_others()")
84d265281d ("sched/pelt: Use rq_clock_task() for hw_pressure")
The former factored out the body of __update_blocked_others() into
update_other_load_avgs(). The latter changed how update_hw_load_avg() is
called in the body. Resolved by applying the change to
update_other_load_avgs() instead.
Signed-off-by: Tejun Heo <tj@kernel.org>
Since commit b2a02fc43a ("smp: Optimize
send_call_function_single_ipi()") an idle CPU in TIF_POLLING_NRFLAG mode
can be pulled out of idle by setting TIF_NEED_RESCHED flag to service an
IPI without actually sending an interrupt. Even in cases where the IPI
handler does not queue a task on the idle CPU, do_idle() will call
__schedule() since need_resched() returns true in these cases.
Introduce and use SM_IDLE to identify call to __schedule() from
schedule_idle() and shorten the idle re-entry time by skipping
pick_next_task() when nr_running is 0 and the previous task is the idle
task.
With the SM_IDLE fast-path, the time taken to complete a fixed set of
IPIs using ipistorm improves noticeably. Following are the numbers
from a dual socket Intel Ice Lake Xeon server (2 x 32C/64T) and
3rd Generation AMD EPYC system (2 x 64C/128T) (boost on, C2 disabled)
running ipistorm between CPU8 and CPU16:
cmdline: insmod ipistorm.ko numipi=100000 single=1 offset=8 cpulist=8 wait=1
==================================================================
Test : ipistorm (modified)
Units : Normalized runtime
Interpretation: Lower is better
Statistic : AMean
======================= Intel Ice Lake Xeon ======================
kernel: time [pct imp]
tip:sched/core 1.00 [baseline]
tip:sched/core + SM_IDLE 0.80 [20.51%]
==================== 3rd Generation AMD EPYC =====================
kernel: time [pct imp]
tip:sched/core 1.00 [baseline]
tip:sched/core + SM_IDLE 0.90 [10.17%]
==================================================================
[ kprateek: Commit message, SM_RTLOCK_WAIT fix ]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Not-yet-signed-off-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20240809092240.6921-1-kprateek.nayak@amd.com
Add sched_ext_ops operations to init/exit cgroups, and track task migrations
and config changes. A BPF scheduler may not implement or implement only
subset of cgroup features. The implemented features can be indicated using
%SCX_OPS_HAS_CGOUP_* flags. If cgroup configuration makes use of features
that are not implemented, a warning is triggered.
While a BPF scheduler is being enabled and disabled, relevant cgroup
operations are locked out using scx_cgroup_rwsem. This avoids situations
like task prep taking place while the task is being moved across cgroups,
making things easier for BPF schedulers.
v7: - cgroup interface file visibility toggling is dropped in favor just
warning messages. Dynamically changing interface visiblity caused more
confusion than helping.
v6: - Updated to reflect the removal of SCX_KF_SLEEPABLE.
- Updated to use CONFIG_GROUP_SCHED_WEIGHT and fixes for
!CONFIG_FAIR_GROUP_SCHED && CONFIG_EXT_GROUP_SCHED.
v5: - Flipped the locking order between scx_cgroup_rwsem and
cpus_read_lock() to avoid locking order conflict w/ cpuset. Better
documentation around locking.
- sched_move_task() takes an early exit if the source and destination
are identical. This triggered the warning in scx_cgroup_can_attach()
as it left p->scx.cgrp_moving_from uncleared. Updated the cgroup
migration path so that ops.cgroup_prep_move() is skipped for identity
migrations so that its invocations always match ops.cgroup_move()
one-to-one.
v4: - Example schedulers moved into their own patches.
- Fix build failure when !CONFIG_CGROUP_SCHED, reported by Andrea Righi.
v3: - Make scx_example_pair switch all tasks by default.
- Convert to BPF inline iterators.
- scx_bpf_task_cgroup() is added to determine the current cgroup from
CPU controller's POV. This allows BPF schedulers to accurately track
CPU cgroup membership.
- scx_example_flatcg added. This demonstrates flattened hierarchy
implementation of CPU cgroup control and shows significant performance
improvement when cgroups which are nested multiple levels are under
competition.
v2: - Build fixes for different CONFIG combinations.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: David Vernet <dvernet@meta.com>
Acked-by: Josh Don <joshdon@google.com>
Acked-by: Hao Luo <haoluo@google.com>
Acked-by: Barret Rhoden <brho@google.com>
Reported-by: kernel test robot <lkp@intel.com>
Cc: Andrea Righi <andrea.righi@canonical.com>
sched_ext will soon add cgroup cpu.weigh support. The cgroup interface code
is currently gated behind CONFIG_FAIR_GROUP_SCHED. As the fair class and/or
SCX may implement the feature, put the interface code behind the new
CONFIG_CGROUP_SCHED_WEIGHT which is selected by CONFIG_FAIR_GROUP_SCHED.
This allows either sched class to enable the itnerface code without ading
more complex CONFIG tests.
When !CONFIG_FAIR_GROUP_SCHED, a dummy version of sched_group_set_shares()
is added to support later CONFIG_CGROUP_SCHED_WEIGHT &&
!CONFIG_FAIR_GROUP_SCHED builds.
No functional changes.
Signed-off-by: Tejun Heo <tj@kernel.org>
Move tg_weight() upward and make cpu_shares_read_u64() use it too. This
makes the weight retrieval shared between cgroup v1 and v2 paths and will be
used to implement cgroup support for sched_ext.
No functional changes.
Signed-off-by: Tejun Heo <tj@kernel.org>
A new BPF extensible sched_class will use css_tg() in the init and exit
paths to visit all task_groups by walking cgroups.
v4: __setscheduler_prio() is already exposed. Dropped from this patch.
v3: Dropped SCHED_CHANGE_BLOCK() as upstream is adding more generic cleanup
mechanism.
v2: Expose SCHED_CHANGE_BLOCK() too and update the description.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: David Vernet <dvernet@meta.com>
Acked-by: Josh Don <joshdon@google.com>
Acked-by: Hao Luo <haoluo@google.com>
Acked-by: Barret Rhoden <brho@google.com>
With sched_ext converted to use put_prev_task() for class switch detection,
there's no user of switch_class() left. Drop it.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
- Resolve trivial context conflicts from dl_server clearing being moved
around.
- Add @next to put_prev_task_scx() and @prev to pick_next_task_scx() to
match sched/core.
- Merge sched_class->switch_class() addition from sched_ext with
tip/sched/core changes in __pick_next_task().
- Make pick_next_task_scx() call put_prev_task_scx() to emulate the previous
behavior where sched_class->put_prev_task() was called before
sched_class->pick_next_task().
While this makes sched_ext build and function, the behavior is not in line
with other sched classes. The follow-up patches will address the
discrepancies and remove sched_class->switch_class().
Signed-off-by: Tejun Heo <tj@kernel.org>
