| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Thanks to kernel cmdline being available early, before any
cgroup hierarchy exists, we can achieve the RT_GROUP_SCHED boottime
disabling goal by simply skipping any creation (and destruction) of
RT_GROUP data and its exposure via RT attributes.
We can do this thanks to previously placed runtime guards that would
redirect all operations to root_task_group's data when RT_GROUP_SCHED
disabled.
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250310170442.504716-8-mkoutny@suse.com
|
|
The rcu_seq_done_exact() function checks if a grace period has completed by
comparing sequence numbers. It includes a guard band to handle sequence number
wraparound, which was previously expressed using the magic number calculation
'3 * RCU_SEQ_STATE_MASK + 1'.
This magic number is not immediately obvious in terms of what it represents.
Instead, the reason we need this tiny guardband is because of the lag between
the setting of rcu_state.gp_seq_polled and root rnp's gp_seq in rcu_gp_init().
This guardband needs to be at least 2 GPs worth of counts, to avoid recognizing
the newly started GP as completed immediately, due to the following sequence
which arises due to the delay between update of rcu_state.gp_seq_polled and
root rnp's gp_seq:
rnp->gp_seq = rcu_state.gp_seq = 0
CPU 0 CPU 1
----- -----
// rcu_state.gp_seq = 1
rcu_seq_start(&rcu_state.gp_seq)
// snap = 8
snap = rcu_seq_snap(&rcu_state.gp_seq)
// Two full GP differences
rcu_seq_done_exact(&rnp->gp_seq, snap)
// rnp->gp_seq = 1
WRITE_ONCE(rnp->gp_seq, rcu_state.gp_seq);
This can happen due to get_state_synchronize_rcu_full() sampling
rcu_state.gp_seq_polled, however the poll_state_synchronize_rcu_full()
sampling the root rnp's gp_seq. The delay between the update of the 2
counters occurs in rcu_gp_init() during which the counters briefly go
out of sync.
Make the guardband explictly 2 GPs. This improves code readability and
maintainability by making the intent clearer as well.
Suggested-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
|
|
When RT_GROUPs are compiled but not exposed, their bandwidth cannot
be configured (and it is not initialized for non-root task_groups neither).
Therefore bypass any checks of task vs task_group bandwidth.
This will achieve behavior very similar to setups that have
!CONFIG_RT_GROUP_SCHED and attach cpu controller to cgroup v2 hierarchy.
(On a related note, this may allow having RT tasks with
CONFIG_RT_GROUP_SCHED and cgroup v2 hierarchy.)
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250310170442.504716-7-mkoutny@suse.com
|
|
First, we want to prevent placement of RT tasks on non-root rt_rqs which
we achieve in the task migration code that'd fall back to
root_task_group's rt_rq.
Second, we want to work with only root_task_group's rt_rq when iterating
all "real" rt_rqs when RT_GROUP is disabled. To achieve this we keep
root_task_group as the first one on the task_groups and break out
quickly.
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250310170442.504716-6-mkoutny@suse.com
|
|
Only simple implementation with a static key wrapper, it will be wired
in later.
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250310170442.504716-5-mkoutny@suse.com
|
|
rt_rq->tg may be NULL which denotes the root task_group.
Store the pointer to root_task_group directly so that callers may use
rt_rq->tg homogenously.
root_task_group exists always with CONFIG_CGROUPS_SCHED,
CONFIG_RT_GROUP_SCHED depends on that.
This changes root level rt_rq's default limit from infinity to the
value of (originally) global RT throttling.
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250310170442.504716-4-mkoutny@suse.com
|
|
rt_entity_is_task has split definitions based on CONFIG_RT_GROUP_SCHED,
therefore we can use it always. No functional change intended.
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250310170442.504716-3-mkoutny@suse.com
|
|
Convert the blocks guarded by macros to regular code so that the RT
group code gets more compile validation. Reasoning is in
Documentation/process/coding-style.rst 21) Conditional Compilation.
With that, no functional change is expected.
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250310170442.504716-2-mkoutny@suse.com
|
|
commit 10a35e6812aa ("sched/pelt: Skip updating util_est when
utilization is higher than CPU's capacity")
prevents util_est from being updated if util_avg is higher than the
underlying CPU capacity to avoid overestimating the task when the CPU
is capped (due to thermal issue for instance). In this scenario, the
task will miss its deadlines and start overlapping its wake-up events
for instance. The task will appear as always running when the CPU is
just not powerful enough to allow having a good estimation of the
task.
commit b8c96361402a ("sched/fair/util_est: Implement faster ramp-up
EWMA on utilization increases")
sets ewma to util_avg when ewma > util_avg, allowing ewma to quickly
grow instead of slowly converge to the new util_avg value when a task
profile changes from small to big.
However, the 2 conditions:
- Check util_avg against max CPU capacity
- Check whether util_est > util_avg
are placed in an order such as it is possible to set util_est to a
value higher than the CPU capacity if util_est > util_avg, but
util_est is prevented to decay as long as:
CPU capacity < util_avg < util_est.
Just remove the check as either:
1.
There is idle time on the CPU. In that case the util_avg value of the
task is actually correct. It is possible that the task missed a
deadline and appears bigger, but this is also the case when the
util_avg of the task is lower than the maximum CPU capacity.
2.
There is no idle time. In that case, the util_avg value might aswell
be an under estimation of the size of the task.
It is possible that undesired frequency spikes will appear when the
task is later enqueued with an inflated util_est value, but the
frequency spike might aswell be deserved. The absence of idle time
prevents from drawing any conclusion.
Signed-off-by: Pierre Gondois <pierre.gondois@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.rog>
Link: https://lore.kernel.org/r/20250325150542.1077344-1-pierre.gondois@arm.com
|
|
Simplify the topology_span_sane code further, removing the need to
allocate an array and gotos used to make sure the array gets freed.
This version is in a separate commit because it could return a
different sanity result than the previous code, but only in odd
circumstances that are not expected to actually occur; for example,
when a CPU is not listed in its own mask.
Signed-off-by: Steve Wahl <steve.wahl@hpe.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <vschneid@redhat.com>
Reviewed-by: Madadi Vineeth Reddy <vineethr@linux.ibm.com>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Tested-by: Valentin Schneider <vschneid@redhat.com>
Tested-by: Madadi Vineeth Reddy <vineethr@linux.ibm.com>
Link: https://lore.kernel.org/r/20250304160844.75373-3-steve.wahl@hpe.com
|
|
Use a different approach to topology_span_sane(), that checks for the
same constraint of no partial overlaps for any two CPU sets for
non-NUMA topology levels, but does so in a way that is O(N) rather
than O(N^2).
Instead of comparing with all other masks to detect collisions, keep
one mask that includes all CPUs seen so far and detect collisions with
a single cpumask_intersects test.
If the current mask has no collisions with previously seen masks, it
should be a new mask, which can be uniquely identified by the lowest
bit set in this mask. Keep a pointer to this mask for future
reference (in an array indexed by the lowest bit set), and add the
CPUs in this mask to the list of those seen.
If the current mask does collide with previously seen masks, it should
be exactly equal to a mask seen before, looked up in the same array
indexed by the lowest bit set in the mask, a single comparison.
Move the topology_span_sane() check out of the existing topology level
loop, let it use its own loop so that the array allocation can be done
only once, shared across levels.
On a system with 1920 processors (16 sockets, 60 cores, 2 threads),
the average time to take one processor offline is reduced from 2.18
seconds to 1.01 seconds. (Off-lining 959 of 1920 processors took
34m49.765s without this change, 16m10.038s with this change in place.)
Signed-off-by: Steve Wahl <steve.wahl@hpe.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <vschneid@redhat.com>
Reviewed-by: Madadi Vineeth Reddy <vineethr@linux.ibm.com>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Tested-by: Valentin Schneider <vschneid@redhat.com>
Tested-by: Madadi Vineeth Reddy <vineethr@linux.ibm.com>
Link: https://lore.kernel.org/r/20250304160844.75373-2-steve.wahl@hpe.com
|
|
Gabriele noted that in case of signal_pending_state(), the tracepoint
sees a stale task-state.
Fixes: fa2c3254d7cf ("sched/tracing: Don't re-read p->state when emitting sched_switch event")
Reported-by: Gabriele Monaco <gmonaco@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Valentin Schneider <vschneid@redhat.com>
|
|
Previously it was only safe to call perf_pmu_unregister() if there
were no active events of that pmu around -- which was impossible to
guarantee since it races all sorts against perf_init_event().
Rework the whole thing by:
- keeping track of all events for a given pmu
- 'hiding' the pmu from perf_init_event()
- waiting for the appropriate (s)rcu grace periods such that all
prior references to the PMU will be completed
- detaching all still existing events of that pmu (see first point)
and moving them to a new REVOKED state.
- actually freeing the pmu data.
Where notably the new REVOKED state must inhibit all event actions
from reaching code that wants to use event->pmu.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Ravi Bangoria <ravi.bangoria@amd.com>
Link: https://lkml.kernel.org/r/20250307193723.525402029@infradead.org
|
|
The task passed to perf_event_exit_task() is not a child, it is
current. Fix this confusing naming, since much of the rest of the code
also relies on it being current.
Specifically, both exec() and exit() callers use it with current as
the argument.
Notably, task_ctx_sched_out() doesn't make much sense outside of
current.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Ravi Bangoria <ravi.bangoria@amd.com>
Link: https://lkml.kernel.org/r/20250307193305.486326750@infradead.org
|
|
Both perf_event_free_task() and perf_event_exit_task_context() are
very similar, except perf_event_exit_task_context() is a little more
generic / makes less assumptions.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Ravi Bangoria <ravi.bangoria@amd.com>
Link: https://lkml.kernel.org/r/20250307193723.274039710@infradead.org
|
|
There is no good reason to have the free list anymore. It is possible
to call free_event() after the locks have been dropped in the main
loop.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Ravi Bangoria <ravi.bangoria@amd.com>
Link: https://lkml.kernel.org/r/20250307193723.151721102@infradead.org
|
|
Simplify the code by moving the duplicated wakeup condition into
put_ctx().
Notably, wait_var_event() is in perf_event_free_task() and will have
set ctx->task = TASK_TOMBSTONE.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Ravi Bangoria <ravi.bangoria@amd.com>
Link: https://lkml.kernel.org/r/20250307193723.044499344@infradead.org
|
|
Currently perf_event_release_kernel() will iterate the child events and attempt
tear-down. However, it removes them from the child_list using list_move(),
notably skipping the state management done by perf_child_detach().
Crucially, it fails to clear PERF_ATTACH_CHILD, which opens the door for a
concurrent perf_remove_from_context() to race.
This way child_list management stays fully serialized using child_mutex.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Ravi Bangoria <ravi.bangoria@amd.com>
Link: https://lkml.kernel.org/r/20250307193305.486326750@infradead.org
|
|
Ravi reported that the bpf_perf_link_attach() usage of
perf_event_set_bpf_prog() is not serialized by ctx->mutex, unlike the
PERF_EVENT_IOC_SET_BPF case.
Reported-by: Ravi Bangoria <ravi.bangoria@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Ravi Bangoria <ravi.bangoria@amd.com>
Link: https://lkml.kernel.org/r/20250307193305.486326750@infradead.org
|
|
Perf can hang while freeing a sigtrap event if a related deferred
signal hadn't managed to be sent before the file got closed:
perf_event_overflow()
task_work_add(perf_pending_task)
fput()
task_work_add(____fput())
task_work_run()
____fput()
perf_release()
perf_event_release_kernel()
_free_event()
perf_pending_task_sync()
task_work_cancel() -> FAILED
rcuwait_wait_event()
Once task_work_run() is running, the list of pending callbacks is
removed from the task_struct and from this point on task_work_cancel()
can't remove any pending and not yet started work items, hence the
task_work_cancel() failure and the hang on rcuwait_wait_event().
Task work could be changed to remove one work at a time, so a work
running on the current task can always cancel a pending one, however
the wait / wake design is still subject to inverted dependencies when
remote targets are involved, as pictured by Oleg:
T1 T2
fd = perf_event_open(pid => T2->pid); fd = perf_event_open(pid => T1->pid);
close(fd) close(fd)
<IRQ> <IRQ>
perf_event_overflow() perf_event_overflow()
task_work_add(perf_pending_task) task_work_add(perf_pending_task)
</IRQ> </IRQ>
fput() fput()
task_work_add(____fput()) task_work_add(____fput())
task_work_run() task_work_run()
____fput() ____fput()
perf_release() perf_release()
perf_event_release_kernel() perf_event_release_kernel()
_free_event() _free_event()
perf_pending_task_sync() perf_pending_task_sync()
rcuwait_wait_event() rcuwait_wait_event()
Therefore the only option left is to acquire the event reference count
upon queueing the perf task work and release it from the task work, just
like it was done before 3a5465418f5f ("perf: Fix event leak upon exec and file release")
but without the leaks it fixed.
Some adjustments are necessary to make it work:
* A child event might dereference its parent upon freeing. Care must be
taken to release the parent last.
* Some places assuming the event doesn't have any reference held and
therefore can be freed right away must instead put the reference and
let the reference counting to its job.
Reported-by: "Yi Lai" <yi1.lai@linux.intel.com>
Closes: https://lore.kernel.org/all/Zx9Losv4YcJowaP%2F@ly-workstation/
Reported-by: syzbot+3c4321e10eea460eb606@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/all/673adf75.050a0220.87769.0024.GAE@google.com/
Fixes: 3a5465418f5f ("perf: Fix event leak upon exec and file release")
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250304135446.18905-1-frederic@kernel.org
|
|
The rcu_torture_one_read() function is designed for RCU readers that are
confined to a task, such that a single thread of control extends from the
beginning of a given RCU read-side critical section to its end. This does
not suffice for things like srcu_down_read() and srcu_up_read(), where
the critical section might start at task level and end in a timer handler.
This commit therefore creates separate init_rcu_torture_one_read_state(),
rcu_torture_one_read_start(), and rcu_torture_one_read_end() functions,
along with a rcu_torture_one_read_state structure to coordinate their
actions. These will be used to create tests for srcu_down_read()
and friends.
One caution: The caller to rcu_torture_one_read_start() must enter the
initial read-side critical section prior to the call. This enables use
of non-standard primitives such as srcu_down_read() while still using
the same validation code.
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
|
|
SCX_OPS_HAS_CGROUP_WEIGHT was only used to suppress the missing cgroup
weight support warnings. Now that the warnings are removed, the flag doesn't
do anything. Mark it for deprecation and remove its usage from scx_flatcg.
v2: Actually include the scx_flatcg update.
Signed-off-by: Tejun Heo <tj@kernel.org>
Suggested-and-reviewed-by: Andrea Righi <arighi@nvidia.com>
|
|
sched_ext generates warnings when cpu.weight / cpu.idle are set to
non-default values if the BPF scheduler doesn't implement weight support.
These warnings don't provide much value while adding constant annoyance. A
BPF scheduler may not implement any particular behavior and there's nothing
particularly special about missing cgroup weight support. Drop the warnings.
Signed-off-by: Tejun Heo <tj@kernel.org>
|
|
Replace kzalloc with kvzalloc for the exit_dump buffer allocation, which
can require large contiguous memory depending on the implementation.
This change prevents allocation failures by allowing the system to fall
back to vmalloc when contiguous memory allocation fails.
Since this buffer is only used for debugging purposes, physical memory
contiguity is not required, making vmalloc a suitable alternative.
Cc: stable@vger.kernel.org
Fixes: 07814a9439a3b0 ("sched_ext: Print debug dump after an error exit")
Suggested-by: Rik van Riel <riel@surriel.com>
Signed-off-by: Breno Leitao <leitao@debian.org>
Acked-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
|
|
The lookup_one_len family of functions is (now) only used internally by
a filesystem on itself either
- in a context where permission checking is irrelevant such as by a
virtual filesystem populating itself, or xfs accessing its ORPHANAGE
or dquota accessing the quota file; or
- in a context where a permission check (MAY_EXEC on the parent) has just
been performed such as a network filesystem finding in "silly-rename"
file in the same directory. This is also the context after the
_parentat() functions where currently lookup_one_qstr_excl() is used.
So the permission check is pointless.
The name "one_len" is unhelpful in understanding the purpose of these
functions and should be changed. Most of the callers pass the len as
"strlen()" so using a qstr and QSTR() can simplify the code.
This patch renames these functions (include lookup_positive_unlocked()
which is part of the family despite the name) to have a name based on
"lookup_noperm". They are changed to receive a 'struct qstr' instead
of separate name and len. In a few cases the use of QSTR() results in a
new call to strlen().
try_lookup_noperm() takes a pointer to a qstr instead of the whole
qstr. This is consistent with d_hash_and_lookup() (which is nearly
identical) and useful for lookup_noperm_unlocked().
The new lookup_noperm_common() doesn't take a qstr yet. That will be
tidied up in a subsequent patch.
Signed-off-by: NeilBrown <neil@brown.name>
Link: https://lore.kernel.org/r/20250319031545.2999807-5-neil@brown.name
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
Cleanup fprobe address hash table on module unloading because the
target symbols will be disappeared when unloading module and not
sure the same symbol is mapped on the same address.
Note that this is at least disables the fprobes if a part of target
symbols on the unloaded modules. Unlike kprobes, fprobe does not
re-enable the probe point by itself. To do that, the caller should
take care register/unregister fprobe when loading/unloading modules.
This simplifies the fprobe state managememt related to the module
loading/unloading.
Link: https://lore.kernel.org/all/174343534473.843280.13988101014957210732.stgit@devnote2/
Fixes: 4346ba160409 ("fprobe: Rewrite fprobe on function-graph tracer")
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
|
|
Add WARN_ON_ONCE() statements whenever new exclusive CPUs are being
added to a partition root to catch inconsistency in the way exclusive
CPUs are being handled in the cpuset code.
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
|
|
Move the partition disablement comment from cpuset_css_offline()
to cpuset_css_killed() and reword it to remove the remnant of
sched.partition.
Suggested-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
|
|
The current cpuset code uses both cpu_active_mask and cpu_online_mask
and it can be confusing which one should be used if we need to update
the code.
The top_cpuset is always synchronized to cpu_active_mask and we should
avoid using cpu_online_mask as much as possible. An active CPU is always
an online CPU, but not vice versa. cpu_active_mask and cpu_online_mask
can differ during hotplug operations.
A CPU is marked active at the last stage of CPU bringup (CPUHP_AP_ACTIVE).
It is also the stage where cpuset hotplug code will be called to update
the sched domains so that the scheduler can move a normal task to a
newly active CPU or remove tasks away from a newly inactivated CPU. The
online bit is set much earlier in the CPU bringup process and cleared
much later in CPU teardown.
If cpu_online_mask is used while a hotunplug operation is happening in
parallel, we may leave an offline CPU in cpu_allowed or have a higher
chance of leaving an offline CPU in some other masks. Avoid this
problem by always using cpu_active_mask in the cpuset code and leave
a comment as to why the use of cpu_online_mask is discouraged.
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
|
|
Avoid a false-positive lockdep warning in the CONFIG_PREEMPT_RT=y
configuration when using write_seqcount_begin() in the uprobe timer
callback by using raw_write_* APIs.
Uprobe's use of timer callback is guaranteed to not race with itself
for a given uprobe_task, and as such seqcount's insistence on having
preemption disabled on the writer side is irrelevant. So switch to
raw_ variants of seqcount API instead of disabling preemption unnecessarily.
Also, point out in the comments more explicitly why we use seqcount
despite our reader side being rather simple and never retrying. We favor
well-maintained kernel primitive in favor of open-coding our own memory
barriers.
Fixes: 8622e45b5da1 ("uprobes: Reuse return_instances between multiple uretprobes within task")
Reported-by: Alexei Starovoitov <ast@kernel.org>
Suggested-by: Sebastian Siewior <bigeasy@linutronix.de>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: stable@kernel.org
Link: https://lore.kernel.org/r/20250404194848.2109539-1-andrii@kernel.org
|
|
Provide a new kfunc, scx_bpf_select_cpu_and(), that can be used to apply
the built-in idle CPU selection policy to a subset of allowed CPU.
This new helper is basically an extension of scx_bpf_select_cpu_dfl().
However, when an idle CPU can't be found, it returns a negative value
instead of @prev_cpu, aligning its behavior more closely with
scx_bpf_pick_idle_cpu().
It also accepts %SCX_PICK_IDLE_* flags, which can be used to enforce
strict selection to @prev_cpu's node (%SCX_PICK_IDLE_IN_NODE), or to
request only a full-idle SMT core (%SCX_PICK_IDLE_CORE), while applying
the built-in selection logic.
With this helper, BPF schedulers can apply the built-in idle CPU
selection policy restricted to any arbitrary subset of CPUs.
Example usage
=============
Possible usage in ops.select_cpu():
s32 BPF_STRUCT_OPS(foo_select_cpu, struct task_struct *p,
s32 prev_cpu, u64 wake_flags)
{
const struct cpumask *cpus = task_allowed_cpus(p) ?: p->cpus_ptr;
s32 cpu;
cpu = scx_bpf_select_cpu_and(p, prev_cpu, wake_flags, cpus, 0);
if (cpu >= 0) {
scx_bpf_dsq_insert(p, SCX_DSQ_LOCAL, SCX_SLICE_DFL, 0);
return cpu;
}
return prev_cpu;
}
Results
=======
Load distribution on a 4 sockets, 4 cores per socket system, simulated
using virtme-ng, running a modified version of scx_bpfland that uses
scx_bpf_select_cpu_and() with 0xff00 as the allowed subset of CPUs:
$ vng --cpu 16,sockets=4,cores=4,threads=1
...
$ stress-ng -c 16
...
$ htop
...
0[ 0.0%] 8[||||||||||||||||||||||||100.0%]
1[ 0.0%] 9[||||||||||||||||||||||||100.0%]
2[ 0.0%] 10[||||||||||||||||||||||||100.0%]
3[ 0.0%] 11[||||||||||||||||||||||||100.0%]
4[ 0.0%] 12[||||||||||||||||||||||||100.0%]
5[ 0.0%] 13[||||||||||||||||||||||||100.0%]
6[ 0.0%] 14[||||||||||||||||||||||||100.0%]
7[ 0.0%] 15[||||||||||||||||||||||||100.0%]
With scx_bpf_select_cpu_dfl() tasks would be distributed evenly across
all the available CPUs.
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
|
|
Many scx schedulers implement their own hard or soft-affinity rules
to support topology characteristics, such as heterogeneous architectures
(e.g., big.LITTLE, P-cores/E-cores), or to categorize tasks based on
specific properties (e.g., running certain tasks only in a subset of
CPUs).
Currently, there is no mechanism that allows to use the built-in idle
CPU selection policy to an arbitrary subset of CPUs. As a result,
schedulers often implement their own idle CPU selection policies, which
are typically similar to one another, leading to a lot of code
duplication.
To address this, modify scx_select_cpu_dfl() to accept an arbitrary
cpumask, that can be used by the BPF schedulers to apply the existent
built-in idle CPU selection policy to a subset of allowed CPUs.
With this concept the idle CPU selection policy becomes the following:
- always prioritize CPUs from fully idle SMT cores (if SMT is enabled),
- select the same CPU if it's idle and in the allowed CPUs,
- select an idle CPU within the same LLC, if the LLC cpumask is a
subset of the allowed CPUs,
- select an idle CPU within the same node, if the node cpumask is a
subset of the allowed CPUs,
- select an idle CPU within the allowed CPUs.
This functionality will be exposed through a dedicated kfunc in a
separate patch.
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
|
|
Modify scx_select_cpu_dfl() to take the allowed cpumask as an explicit
argument, instead of implicitly using @p->cpus_ptr.
This prepares for future changes where arbitrary cpumasks may be passed
to the built-in idle CPU selection policy.
This is a pure refactoring with no functional changes.
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
|
|
The built-in idle selection policy, scx_select_cpu_dfl(), always
prioritizes picking idle CPUs within the same LLC or NUMA node, but
these optimizations are currently applied only when a task has no CPU
affinity constraints.
This is done primarily for efficiency, as it avoids the overhead of
updating a cpumask every time we need to select an idle CPU (which can
be costly in large SMP systems).
However, this approach limits the effectiveness of the built-in idle
policy and results in inconsistent behavior, as affinity-restricted
tasks don't benefit from topology-aware optimizations.
To address this, modify the policy to apply LLC and NUMA-aware
optimizations even when a task is constrained to a subset of CPUs.
We can still avoid updating the cpumasks by checking if the subset of
LLC and node CPUs are contained in the subset of allowed CPUs usable by
the task (which is true in most of the cases - for tasks that don't have
affinity constratints).
Moreover, use temporary local per-CPU cpumasks to determine the LLC and
node subsets, minimizing potential overhead even on large SMP systems.
Signed-off-by: Andrea Righi <arighi@nvidia.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
|
|
The function get_vm_area() is not defined for non-MMU builds and causes a
build error if it is used. Hide the map_pages() function around a:
#ifdef CONFIG_MMU
to keep it from being compiled when CONFIG_MMU is not set.
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Link: https://lore.kernel.org/20250407120111.2ccc9319@gandalf.local.home
Reported-by: Guenter Roeck <linux@roeck-us.net>
Tested-by: Guenter Roeck <linux@roeck-us.net>
Closes: https://lore.kernel.org/all/4f8ece8b-8862-4f7c-8ede-febd28f8a9fe@roeck-us.net/
Fixes: 394f3f02de531 ("tracing: Use vmap_page_range() to map memmap ring buffer")
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
|
|
Replace the irq_gc_lock/unlock() pairs with guards. There is no point to
implement a guard wrapper for them as they just wrap around raw_spin_lock*().
Switch the other lock instances in the core code to guards as well.
Conversion was done with Coccinelle plus manual fixups.
No functional change.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/all/20250313142524.073826193@linutronix.de
|
|
We currently report EINVAL whenever a struct pid has no tasked attached
anymore thereby conflating two concepts:
(1) The task has already been reaped.
(2) The caller requested a pidfd for a thread-group leader but the pid
actually references a struct pid that isn't used as a thread-group
leader.
This is causing issues for non-threaded workloads as in [1].
This patch tries to allow userspace to distinguish between (1) and (2).
This is racy of course but that shouldn't matter.
Link: https://github.com/systemd/systemd/pull/36982 [1]
Link: https://lore.kernel.org/r/20250403-work-pidfd-fixes-v1-3-a123b6ed6716@kernel.org
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
None of the caller actually pass a NULL pid in there.
Link: https://lore.kernel.org/r/20250403-work-pidfd-fixes-v1-2-a123b6ed6716@kernel.org
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
Percpu-rwsems are used for superblock locking. However, we know the
read percpu-rwsem we take for sb_start_write() on a frozen filesystem
needs not to inhibit system from suspending or hibernating. That
means it needs to wait with TASK_UNINTERRUPTIBLE | TASK_FREEZABLE.
Introduce a new percpu_down_read_freezable() that allows us to control
whether TASK_FREEZABLE is added to the wait flags.
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Link: https://lore.kernel.org/r/20250327140613.25178-2-James.Bottomley@HansenPartnership.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
Add new function *_twothreecell() to extend support to parse three-cell
interrupts which encoded as <instance hwirq irqflag>, the translate
function will retrieve irq number and flag from last two cells.
This API will be used in gpio irq driver which need to work with
two or three cells cases.
Signed-off-by: Yixun Lan <dlan@gentoo.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20250326-04-gpio-irq-threecell-v3-1-aab006ab0e00@gentoo.org
|
|
Move the get_ctx(child_ctx) call and the child_event->ctx assignment to
occur immediately after the child event is allocated. Ensure that
child_event->ctx is non-NULL before any subsequent error path within
inherit_event calls free_event(), satisfying the assumptions of the
cleanup code.
Details:
There's no clear Fixes tag, because this bug is a side-effect of
multiple interacting commits over time (up to 15 years old), not
a single regression.
The code initially incremented refcount then assigned context
immediately after the child_event was created. Later, an early
validity check for child_event was added before the
refcount/assignment. Even later, a WARN_ON_ONCE() cleanup check was
added, assuming event->ctx is valid if the pmu_ctx is valid.
The problem is that the WARN_ON_ONCE() could trigger after the initial
check passed but before child_event->ctx was assigned, violating its
precondition. The solution is to assign child_event->ctx right after
its initial validation. This ensures the context exists for any
subsequent checks or cleanup routines, resolving the WARN_ON_ONCE().
To resolve it, defer the refcount update and child_event->ctx assignment
directly after child_event->pmu_ctx is set but before checking if the
parent event is orphaned. The cleanup routine depends on
event->pmu_ctx being non-NULL before it verifies event->ctx is
non-NULL. This also maintains the author's original intent of passing
in child_ctx to find_get_pmu_context before its refcount/assignment.
[ mingo: Expanded the changelog from another email by Gabriel Shahrouzi. ]
Reported-by: syzbot+ff3aa851d46ab82953a3@syzkaller.appspotmail.com
Signed-off-by: Gabriel Shahrouzi <gshahrouzi@gmail.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ravi Bangoria <ravi.bangoria@amd.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Link: https://lore.kernel.org/r/20250405203036.582721-1-gshahrouzi@gmail.com
Closes: https://syzkaller.appspot.com/bug?extid=ff3aa851d46ab82953a3
|
|
Pull perf event fix from Ingo Molnar:
"Fix a perf events time accounting bug"
* tag 'perf-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/core: Fix child_total_time_enabled accounting bug at task exit
|
|
Pull scheduler fixes from Ingo Molnar:
- Fix a nonsensical Kconfig combination
- Remove an unnecessary rseq-notification
* tag 'sched-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
rseq: Eliminate useless task_work on execve
sched/isolation: Make CONFIG_CPU_ISOLATION depend on CONFIG_SMP
|
|
Pull timer cleanups from Thomas Gleixner:
"A set of final cleanups for the timer subsystem:
- Convert all del_timer[_sync]() instances over to the new
timer_delete[_sync]() API and remove the legacy wrappers.
Conversion was done with coccinelle plus some manual fixups as
coccinelle chokes on scoped_guard().
- The final cleanup of the hrtimer_init() to hrtimer_setup()
conversion.
This has been delayed to the end of the merge window, so that all
patches which have been merged through other trees are in mainline
and all new users are catched.
Doing this right before rc1 ensures that new code which is merged post
rc1 is not introducing new instances of the original functionality"
* tag 'timers-cleanups-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
tracing/timers: Rename the hrtimer_init event to hrtimer_setup
hrtimers: Rename debug_init_on_stack() to debug_setup_on_stack()
hrtimers: Rename debug_init() to debug_setup()
hrtimers: Rename __hrtimer_init_sleeper() to __hrtimer_setup_sleeper()
hrtimers: Remove unnecessary NULL check in hrtimer_start_range_ns()
hrtimers: Make callback function pointer private
hrtimers: Merge __hrtimer_init() into __hrtimer_setup()
hrtimers: Switch to use __htimer_setup()
hrtimers: Delete hrtimer_init()
treewide: Convert new and leftover hrtimer_init() users
treewide: Switch/rename to timer_delete[_sync]()
|
|
Pull more irq updates from Thomas Gleixner:
"A set of updates for the interrupt subsystem:
- A treewide cleanup for the irq_domain code, which makes the naming
consistent and gets rid of the original oddity of naming domains
'host'.
This is a trivial mechanical change and is done late to ensure that
all instances have been catched and new code merged post rc1 wont
reintroduce new instances.
- A trivial consistency fix in the migration code
The recent introduction of irq_force_complete_move() in the core
code, causes a problem for the nostalgia crowd who maintains ia64
out of tree.
The code assumes that hierarchical interrupt domains are enabled
and dereferences irq_data::parent_data unconditionally. That works
in mainline because both architectures which enable that code have
hierarchical domains enabled. Though it breaks the ia64 build,
which enables the functionality, but does not have hierarchical
domains.
While it's not really a problem for mainline today, this
unconditional dereference is inconsistent and trivially fixable by
using the existing helper function irqd_get_parent_data(), which
has the appropriate #ifdeffery in place"
* tag 'irq-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
genirq/migration: Use irqd_get_parent_data() in irq_force_complete_move()
irqdomain: Stop using 'host' for domain
irqdomain: Rename irq_get_default_host() to irq_get_default_domain()
irqdomain: Rename irq_set_default_host() to irq_set_default_domain()
|
|
Pull timer fix from Thomas Gleixner:
"A revert to fix a adjtimex() regression:
The recent change to prevent that time goes backwards for the coarse
time getters due to immediate multiplier adjustments via adjtimex(),
changed the way how the timekeeping core treats that.
That change result in a regression on the adjtimex() side, which is
user space visible:
1) The forwarding of the base time moves the update out of the
original period and establishes a new one. That's changing the
behaviour of the [PF]LL control, which user space expects to be
applied periodically.
2) The clearing of the accumulated NTP error due to #1, changes the
behaviour as well.
An attempt to delay the multiplier/frequency update to the next tick
did not solve the problem as userspace expects that the multiplier or
frequency updates are in effect, when the syscall returns.
There is a different solution for the coarse time problem available,
so revert the offending commit to restore the existing adjtimex()
behaviour"
* tag 'timers-urgent-2025-04-06' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
Revert "timekeeping: Fix possible inconsistencies in _COARSE clockids"
|
|
Pull Kbuild updates from Masahiro Yamada:
- Improve performance in gendwarfksyms
- Remove deprecated EXTRA_*FLAGS and KBUILD_ENABLE_EXTRA_GCC_CHECKS
- Support CONFIG_HEADERS_INSTALL for ARCH=um
- Use more relative paths to sources files for better reproducibility
- Support the loong64 Debian architecture
- Add Kbuild bash completion
- Introduce intermediate vmlinux.unstripped for architectures that need
static relocations to be stripped from the final vmlinux
- Fix versioning in Debian packages for -rc releases
- Treat missing MODULE_DESCRIPTION() as an error
- Convert Nios2 Makefiles to use the generic rule for built-in DTB
- Add debuginfo support to the RPM package
* tag 'kbuild-v6.15' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild: (40 commits)
kbuild: rpm-pkg: build a debuginfo RPM
kconfig: merge_config: use an empty file as initfile
nios2: migrate to the generic rule for built-in DTB
rust: kbuild: skip `--remap-path-prefix` for `rustdoc`
kbuild: pacman-pkg: hardcode module installation path
kbuild: deb-pkg: don't set KBUILD_BUILD_VERSION unconditionally
modpost: require a MODULE_DESCRIPTION()
kbuild: make all file references relative to source root
x86: drop unnecessary prefix map configuration
kbuild: deb-pkg: add comment about future removal of KDEB_COMPRESS
kbuild: Add a help message for "headers"
kbuild: deb-pkg: remove "version" variable in mkdebian
kbuild: deb-pkg: fix versioning for -rc releases
Documentation/kbuild: Fix indentation in modules.rst example
x86: Get rid of Makefile.postlink
kbuild: Create intermediate vmlinux build with relocations preserved
kbuild: Introduce Kconfig symbol for linking vmlinux with relocations
kbuild: link-vmlinux.sh: Make output file name configurable
kbuild: do not generate .tmp_vmlinux*.map when CONFIG_VMLINUX_MAP=y
Revert "kheaders: Ignore silly-rename files"
...
|
|
The function hrtimer_init() doesn't exist anymore. It was replaced by
hrtimer_setup().
Thus, rename the hrtimer_init trace event to hrtimer_setup to keep it
consistent.
Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/cba84c3d853c5258aa3a262363a6eac08e2c7afc.1738746927.git.namcao@linutronix.de
|
|
All the hrtimer_init*() functions have been renamed to hrtimer_setup*().
Rename debug_init_on_stack() to debug_setup_on_stack() as well, to keep the
names consistent.
Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/073cf6162779a2f5b12624677d4c49ee7eccc1ed.1738746927.git.namcao@linutronix.de
|
|
All the hrtimer_init*() functions have been renamed to hrtimer_setup*().
Rename debug_init() to debug_setup() as well, to keep the names consistent.
Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/all/4b730c1f79648b16a1c5413f928fdc2e138dfc43.1738746927.git.namcao@linutronix.de
|
Michal Koutný
Joel Fernandes
Pierre Gondois
Steve Wahl
Peter Zijlstra
Frederic Weisbecker
Paul E. McKenney
Tejun Heo
Breno Leitao
NeilBrown
Masami Hiramatsu (Google)
Waiman Long
Andrii Nakryiko
Andrea Righi
Steven Rostedt
Thomas Gleixner
Christian Brauner
James Bottomley
Yixun Lan
Gabriel Shahrouzi
Linus Torvalds
Nam Cao