sched/topology: Remove sched_domain_topology_level::flags

Support for overlapping domains added in commit e3589f6c81e4 ("sched:
Allow for overlapping sched_domain spans") also allowed forcefully
setting SD_OVERLAP for !NUMA domains via FORCE_SD_OVERLAP sched_feat().

Since NUMA domains had to be presumed overlapping to ensure correct
behavior, "sched_domain_topology_level::flags" was introduced. NUMA
domains added the SDTL_OVERLAP flag would ensure SD_OVERLAP was always
added during build_sched_domains() for these domains, even when
FORCE_SD_OVERLAP was off.

Condition for adding the SD_OVERLAP flag at the aforementioned commit
was as follows:

    if (tl->flags & SDTL_OVERLAP || sched_feat(FORCE_SD_OVERLAP))
            sd->flags |= SD_OVERLAP;

The FORCE_SD_OVERLAP debug feature was removed in commit af85596c74de
("sched/topology: Remove FORCE_SD_OVERLAP") which left the NUMA domains
as the exclusive users of SDTL_OVERLAP, SD_OVERLAP, and SD_NUMA flags.

Get rid of SDTL_OVERLAP and SD_OVERLAP as they have become redundant
and instead rely on SD_NUMA to detect the only overlapping domain
currently supported. Since SDTL_OVERLAP was the only user of
"tl->flags", get rid of "sched_domain_topology_level::flags" too.

Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/ba4dbdf8-bc37-493d-b2e0-2efb00ea3e19@amd.com

x86/smpboot: avoid SMT domain attach/destroy if SMT is not enabled

Currently, the SMT domain is added into sched_domain_topology by default.

If cpu_attach_domain() finds that the CPU SMT domain’s cpumask_weight
is just 1, it will destroy it.

On a large machine, such as one with 512 cores, this results in
512 redundant domain attach/destroy operations.

Avoid these unnecessary operations by simply checking
cpu_smt_num_threads and skip SMT domain if the SMT domain is not
enabled.

Suggested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Li Chen <chenl311@chinatelecom.cn>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: K Prateek Nayak <kprateek.nayak@amd.com>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Link: https://lore.kernel.org/r/20250710105715.66594-5-me@linux.beauty

x86/smpboot: moves x86_topology to static initialize and truncate

The #ifdeffery and the initializers in build_sched_topology() are just
disgusting.

Statically initialize the domain levels in the topology array and let
build_sched_topology() invalidate the package domain level when NUMA in
package is available.

Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Li Chen <chenl311@chinatelecom.cn>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: K Prateek Nayak <kprateek.nayak@amd.com>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Link: https://lore.kernel.org/r/20250710105715.66594-4-me@linux.beauty

x86/smpboot: remove redundant CONFIG_SCHED_SMT

On x86 CONFIG_SCHED_SMT is default y if SMP is enabled, so let's
simply drop CONFIG_SCHED_SMT.

Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Li Chen <chenl311@chinatelecom.cn>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: K Prateek Nayak <kprateek.nayak@amd.com>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Link: https://lore.kernel.org/r/20250710105715.66594-3-me@linux.beauty

smpboot: introduce SDTL_INIT() helper to tidy sched topology setup

Define a small SDTL_INIT(maskfn, flagsfn, name) macro and use it to build the
sched_domain_topology_level array. Purely a cleanup; behaviour is unchanged.

Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Li Chen <chenl311@chinatelecom.cn>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: K Prateek Nayak <kprateek.nayak@amd.com>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Link: https://lore.kernel.org/r/20250710105715.66594-2-me@linux.beauty

tools/sched: Add dl_bw_dump.py for printing bandwidth accounting info

dl_rq bandwidth accounting information is crucial for the correct
functioning of SCHED_DEADLINE.

Add a drgn script for accessing that information at runtime, so that
it's easier to check and debug issues related to it.

Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Marcel Ziswiler <marcel.ziswiler@codethink.co.uk> # nuc & rock5b
Link: https://lore.kernel.org/r/20250627115118.438797-6-juri.lelli@redhat.com

tools/sched: Add root_domains_dump.py which dumps root domains info

Root domains information is somewhat hard to access at runtime. Even
with sched_debug and sched_verbose, such information is only printed
on kernel console when domains are modified.

Add a simple drgn script to more easily retrieve root domains
information at runtime.

Since tools/sched is a new directory, add it to MAINTAINERS as well.

Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Marcel Ziswiler <marcel.ziswiler@codethink.co.uk> # nuc & rock5b
Link: https://lore.kernel.org/r/20250627115118.438797-5-juri.lelli@redhat.com

sched/deadline: Fix accounting after global limits change

A global limits change (sched_rt_handler() logic) currently leaves stale
and/or incorrect values in variables related to accounting (e.g.
extra_bw).

Properly clean up per runqueue variables before implementing the change
and rebuild scheduling domains (so that accounting is also properly
restored) after such a change is complete.

Reported-by: Marcel Ziswiler <marcel.ziswiler@codethink.co.uk>
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Marcel Ziswiler <marcel.ziswiler@codethink.co.uk> # nuc & rock5b
Link: https://lore.kernel.org/r/20250627115118.438797-4-juri.lelli@redhat.com

sched/deadline: Reset extra_bw to max_bw when clearing root domains

dl_clear_root_domain() doesn't take into account the fact that per-rq
extra_bw variables retain values computed before root domain changes,
resulting in broken accounting.

Fix it by resetting extra_bw to max_bw before restoring back dl-servers
contributions.

Fixes: 2ff899e351643 ("sched/deadline: Rebuild root domain accounting after every update")
Reported-by: Marcel Ziswiler <marcel.ziswiler@codethink.co.uk>
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Marcel Ziswiler <marcel.ziswiler@codethink.co.uk> # nuc & rock5b
Link: https://lore.kernel.org/r/20250627115118.438797-3-juri.lelli@redhat.com

sched/deadline: Initialize dl_servers after SMP

dl-servers are currently initialized too early at boot when CPUs are not
fully up (only boot CPU is). This results in miscalculation of per
runqueue DEADLINE variables like extra_bw (which needs a stable CPU
count).

Move initialization of dl-servers later on after SMP has been
initialized and CPUs are all online, so that CPU count is stable and
DEADLINE variables can be computed correctly.

Fixes: d741f297bceaf ("sched/fair: Fair server interface")
Reported-by: Marcel Ziswiler <marcel.ziswiler@codethink.co.uk>
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Waiman Long <longman@redhat.com>
Tested-by: Marcel Ziswiler <marcel.ziswiler@codethink.co.uk> # nuc & rock5b
Link: https://lore.kernel.org/r/20250627115118.438797-2-juri.lelli@redhat.com

lib/smp_processor_id: Make migration check unconditional of SMP

Commit cac5cefbade90 ("sched/smp: Make SMP unconditional")
migrate_disable() even on UP builds.
Commit 06ddd17521bf1 ("sched/smp: Always define is_percpu_thread() and
scheduler_ipi()") made is_percpu_thread() check the affinity mask
instead replying always true for UP mask.

As a consequence smp_processor_id() now complains if invoked within a
migrate_disable() section because is_percpu_thread() checks its mask and
the migration check is left out.

Make migration check unconditional of SMP.

Fixes: cac5cefbade90 ("sched/smp: Make SMP unconditional")
Closes: https://lore.kernel.org/oe-lkp/202507100448.6b88d6f1-lkp@intel.com
Reported-by: kernel test robot <oliver.sang@intel.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Chen Yu <yu.c.chen@intel.com>
Link: https://lore.kernel.org/r/20250710082748.-DPO1rjO@linutronix.de

sched/fair: Always trigger resched at the end of a protected period

Always trigger a resched after a protected period even if the entity is
still eligible. It can happen that an entity remains eligible at the end
of the protected period but must let an entity with a shorter slice to run
in order to keep its lag shorter than slice. This is particulalry true
with run to parity which tries to maximize the lag.

Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250708165630.1948751-7-vincent.guittot@linaro.org

sched/fair: Fix entity's lag with run to parity

When an entity is enqueued without preempting current, we must ensure
that the slice protection is updated to take into account the slice
duration of the newly enqueued task so that its lag will not exceed
its slice (+ tick).

Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250708165630.1948751-6-vincent.guittot@linaro.org

sched/fair: Limit run to parity to the min slice of enqueued entities

Run to parity ensures that current will get a chance to run its full
slice in one go but this can create large latency and/or lag for
entities with shorter slice that have exhausted their previous slice
and wait to run their next slice.

Clamp the run to parity to the shortest slice of all enqueued entities.

Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250708165630.1948751-5-vincent.guittot@linaro.org

sched/fair: Remove spurious shorter slice preemption

Even if the waking task can preempt current, it might not be the one
selected by pick_task_fair. Check that the waking task will be selected
if we cancel the slice protection before doing so.

Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250708165630.1948751-4-vincent.guittot@linaro.org

sched/fair: Fix NO_RUN_TO_PARITY case

EEVDF expects the scheduler to allocate a time quantum to the selected
entity and then pick a new entity for next quantum.
Although this notion of time quantum is not strictly doable in our case,
we can ensure a minimum runtime for each task most of the time and pick a
new entity after a minimum time has elapsed.
Reuse the slice protection of run to parity to ensure such runtime
quantum.

Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250708165630.1948751-3-vincent.guittot@linaro.org

sched/fair: Use protect_slice() instead of direct comparison

Replace the test by the relevant protect_slice() function.

Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dhaval Giani (AMD) <dhaval@gianis.ca>
Link: https://lkml.kernel.org/r/20250708165630.1948751-2-vincent.guittot@linaro.org

sched/deadline: Less agressive dl_server handling

Chris reported that commit 5f6bd380c7bd ("sched/rt: Remove default
bandwidth control") caused a significant dip in his favourite
benchmark of the day. Simply disabling dl_server cured things.

His workload hammers the 0->1, 1->0 transitions, and the
dl_server_{start,stop}() overhead kills it -- fairly obviously a bad
idea in hind sight and all that.

Change things around to only disable the dl_server when there has not
been a fair task around for a whole period. Since the default period
is 1 second, this ensures the benchmark never trips this, overhead
gone.

Fixes: 557a6bfc662c ("sched/fair: Add trivial fair server")
Reported-by: Chris Mason <clm@meta.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Juri Lelli <juri.lelli@redhat.com>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lkml.kernel.org/r/20250702121158.465086194@infradead.org

sched/psi: Optimize psi_group_change() cpu_clock() usage

Dietmar reported that commit 3840cbe24cf0 ("sched: psi: fix bogus
pressure spikes from aggregation race") caused a regression for him on
a high context switch rate benchmark (schbench) due to the now
repeating cpu_clock() calls.

In particular the problem is that get_recent_times() will extrapolate
the current state to 'now'. But if an update uses a timestamp from
before the start of the update, it is possible to get two reads
with inconsistent results. It is effectively back-dating an update.

(note that this all hard-relies on the clock being synchronized across
CPUs -- if this is not the case, all bets are off).

Combine this problem with the fact that there are per-group-per-cpu
seqcounts, the commit in question pushed the clock read into the group
iteration, causing tree-depth cpu_clock() calls. On architectures
where cpu_clock() has appreciable overhead, this hurts.

Instead move to a per-cpu seqcount, which allows us to have a single
clock read for all group updates, increasing internal consistency and
lowering update overhead. This comes at the cost of a longer update
side (proportional to the tree depth) which can cause the read side to
retry more often.

Fixes: 3840cbe24cf0 ("sched: psi: fix bogus pressure spikes from aggregation race")
Reported-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Tested-by: Dietmar Eggemann <dietmar.eggemann@arm.com>,
Link: https://lkml.kernel.org/20250522084844.GC31726@noisy.programming.kicks-ass.net

sched/fair: Bump sd->max_newidle_lb_cost when newidle balance fails

schbench (https://github.com/masoncl/schbench.git) is showing a
regression from previous production kernels that bisected down to:

sched/fair: Remove sysctl_sched_migration_cost condition (c5b0a7eefc)

The schbench command line was:

schbench -L -m 4 -M auto -t 256 -n 0 -r 0 -s 0

This creates 4 message threads pinned to CPUs 0-3, and 256x4 worker
threads spread across the rest of the CPUs.  Neither the worker threads
or the message threads do any work, they just wake each other up and go
back to sleep as soon as possible.

The end result is the first 4 CPUs are pegged waking up those 1024
workers, and the rest of the CPUs are constantly banging in and out of
idle.  If I take a v6.9 Linus kernel and revert that one commit,
performance goes from 3.4M RPS to 5.4M RPS.

schedstat shows there are ~100x  more new idle balance operations, and
profiling shows the worker threads are spending ~20% of their CPU time
on new idle balance.  schedstats also shows that almost all of these new
idle balance attemps are failing to find busy groups.

The fix used here is to crank up the cost of the newidle balance whenever it
fails.  Since we don't want sd->max_newidle_lb_cost to grow out of
control, this also changes update_newidle_cost() to use
sysctl_sched_migration_cost as the upper limit on max_newidle_lb_cost.

Signed-off-by: Chris Mason <clm@fb.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20250626144017.1510594-2-clm@fb.com

Merge tag 'rust-sched.2025.06.24' of git://git./linux/kernel/git/boqun/linux into sched/core

Rust task & schedule changes for v6.17:

- Make Task, CondVar and PollCondVar methods inline to avoid unnecessary
  function calls

- Add might_sleep() support for Rust code: Rust's "#[track_caller]"
  mechanism is used so that Rust's might_sleep() doesn't need to be
  defined as a macro

Signed-off-by: Peter Zijlstra <peterz@infradead.org>
# -----BEGIN PGP SIGNATURE-----
#
# iQEzBAABCAAdFiEEj5IosQTPz8XU1wRHSXnow7UH+rgFAmhbLOgACgkQSXnow7UH
# +riscwf/e/+KmJTox5/JOqs6yxxQdCHMaGnMK62E5AII7NsiUI8+XB9z6efzCMmy
# kS2W7aCmBZX67Y1B/xRL/ArHMBAJBi/CrCedZJcmzfB9aMa4Lj4mgiPkbUXkxE6Q
# F5CDQK21ftu+0Q7Hhlq92ec17ZWodOvNxCFBBmjtQqUvBzj0dY45jcG7brs+N+1Z
# t9UO3YokzukNqpIXTpG0HFP+XNafWWCgn9iIQ44lRxIaAoPI44uJjh1OXLTrZ1M9
# EMWYIrsY3b71Im78l6pzr+UOzJdJLI+QCBiz7ySLYz3kZ5dEfFdJOsumbc0G8A69
# VSGDFPEbJxZYuMxrH0E44XmxH4rJdA==
# =gD/Y
# -----END PGP SIGNATURE-----
# gpg: Signature made Wed 25 Jun 2025 12:55:36 AM CEST
# gpg:                using RSA key 8F9228B104CFCFC5D4D704474979E8C3B507FAB8
# gpg: Can't check signature: No public key

rust: task: Add Rust version of might_sleep()

Add a helper function equivalent to the C's might_sleep(), which
serves as a debugging aid and a potential scheduling point.

Note that this function can only be used in a nonatomic context.

This will be used by Rust version of read_poll_timeout().

[boqun: Use file_from_location() to get a C string instead of changing
__might_sleep()]

Signed-off-by: FUJITA Tomonori <fujita.tomonori@gmail.com>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Link: https://lore.kernel.org/r/20250619151007.61767-3-boqun.feng@gmail.com

rust: Introduce file_from_location()

Most of kernel debugging facilities take a nul-terminated string for
file names for a callsite (generated from __FILE__), however the Rust
courterpart, Location, would return a Rust string (not nul-terminated)
from method .file(). And such a string cannot be passed to C debugging
function directly.

There is ongoing work to support a Location::file_with_nul() [1], which
returns a nul-terminated string from a Location. Since it's still
working in progress, and it will take some time before the feature
finally gets stabilized and the kernel's minimal rustc version might
also take a while to bump to a version that at least has that feature,
introduce a file_from_location() function, which returns a warning
string if Location::file_with_nul() is not available.

This should work in most cases because as for now the known usage of
Location::file_with_nul() is only in debugging code (e.g. might_sleep())
and there might be other information reported by the debugging code that
could help locate the problematic function, so missing the file name is
fine at the moment.

Link: https://github.com/rust-lang/rust/issues/141727
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Link: https://lore.kernel.org/r/20250619151007.61767-2-boqun.feng@gmail.com

rust: task: Mark Task methods inline

When building the kernel using the llvm-18.1.3-rust-1.85.0-x86_64
toolchain provided by kernel.org, the following symbols are generated:

$ nm vmlinux | grep ' _R'.*Task | rustfilt
... T <kernel::task::Task>::get_pid_ns
... T <kernel::task::Task>::tgid_nr_ns
... T <kernel::task::Task>::current_pid_ns
... T <kernel::task::Task>::signal_pending
... T <kernel::task::Task>::uid
... T <kernel::task::Task>::euid
... T <kernel::task::Task>::current
... T <kernel::task::Task>::wake_up
... T <kernel::task::Task as kernel::types::AlwaysRefCounted>::dec_ref
... T <kernel::task::Task as kernel::types::AlwaysRefCounted>::inc_ref

These Rust symbols are trivial wrappers around the C functions. It
doesn't make sense to go through a trivial wrapper for these functions,
so mark them inline.

[boqun: Capitalize the title, reword a bit to avoid listing all the C
functions as the code already shows them and remove the addresses of the
symbols in the commit log as they are different from build to build.]

Link: https://github.com/Rust-for-Linux/linux/issues/1145
Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Reviewed-by: Christian Schrefl <chrisi.schrefl@gmail.com>
Reviewed-by: Charalampos Mitrodimas <charmitro@posteo.net>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Signed-off-by: Panagiotis Foliadis <pfoliadis@posteo.net>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Link: https://lore.kernel.org/r/20250315-inline-c-wrappers-v3-1-048e43fcef7d@posteo.net

rust: sync: Mark PollCondVar::drop() inline

When building the kernel using the llvm-18.1.3-rust-1.85.0-x86_64
with ARCH=arm64, the following symbols are generated:

$nm vmlinux | grep ' _R'.*PollCondVar  | rustfilt
... T <kernel::sync::poll::PollCondVar as kernel::init::PinnedDrop>::drop
...

This Rust symbol is trivial wrappers around the C functions
__wake_up_pollfree() and synchronize_rcu(). It doesn't make sense to go
through a trivial wrapper for its functions, so mark it inline.

[boqun: Reword the commit title and re-format the commit log per tip
tree's requirement, remove unnecessary information from "nm vmlinux"
result.]

Link: https://github.com/Rust-for-Linux/linux/issues/1145
Suggested-by: Alice Ryhl <aliceryhl@google.com>
Co-developed-by: Grace Deng <Grace.Deng006@Gmail.com>
Signed-off-by: Grace Deng <Grace.Deng006@Gmail.com>
Signed-off-by: Kunwu Chan <kunwu.chan@hotmail.com>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Link: https://lore.kernel.org/r/20250317025205.2366518-1-kunwu.chan@linux.dev

rust: sync: Mark CondVar::notify_*() inline

When build the kernel using the llvm-18.1.3-rust-1.85.0-x86_64
with ARCH=arm64, the following symbols are generated:

$nm vmlinux | grep ' _R'.*CondVar | rustfilt
... T <kernel::sync::condvar::CondVar>::notify_all
... T <kernel::sync::condvar::CondVar>::notify_one
... T <kernel::sync::condvar::CondVar>::notify_sync
...

These notify_*() symbols are trivial wrappers around the C functions
__wake_up() and __wake_up_sync(). It doesn't make sense to go through
a trivial wrapper for these functions, so mark them inline.

[boqun: Reword the commit title for consistency and reformat the commit
log.]

Suggested-by: Alice Ryhl <aliceryhl@google.com>
Link: https://github.com/Rust-for-Linux/linux/issues/1145
Co-developed-by: Grace Deng <Grace.Deng006@Gmail.com>
Signed-off-by: Grace Deng <Grace.Deng006@Gmail.com>
Signed-off-by: Kunwu Chan <kunwu.chan@hotmail.com>
Reviewed-by: Benno Lossin <benno.lossin@proton.me>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Link: https://lore.kernel.org/r/20250324061835.1693125-1-kunwu.chan@linux.dev

sched/core: Reorganize cgroup bandwidth control interface file writes

- Move input parameter validation from tg_set_cfs_bandwidth() to the new
  outer function tg_set_bandwidth(). The outer function handles parameters
  in usecs, validates them and calls tg_set_cfs_bandwidth() which converts
  them into nsecs. This matches tg_bandwidth() on the read side.

- max/min_cfs_* consts are now used by tg_set_bandwidth(). Relocate, convert
  into usecs and drop "cfs" from the names.

- Reimplement cpu_cfs_{period|quote|burst}_write_*() using tg_bandwidth()
  and tg_set_bandwidth() and replace "cfs" in the names with "bw".

- Update cpu_max_write() to use tg_set_bandiwdth(). cpu_period_quota_parse()
  is updated to drop nsec conversion accordingly. This aligns the behavior
  with cfs_period_quota_print().

- Drop now unused tg_set_cfs_{period|quota|burst}().

- While at it, for consistency, rename default_cfs_period() to
  default_bw_period_us() and make it return usecs.

This is to prepare for adding bandwidth control support to sched_ext.
tg_set_bandwidth() will be used as the muxing point. No functional changes
intended.

Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250614012346.2358261-5-tj@kernel.org

sched/core: Reorganize cgroup bandwidth control interface file reads

- Update tg_get_cfs_*() to return u64 values. These are now used as the low
  level accessors to the fair's bandwidth configuration parameters.
  Translation to usecs takes place in these functions.

- Add tg_bandwidth() which reads all three bandwidth parameters using
  tg_get_cfs_*().

- Reimplement cgroup interface read functions using tg_bandwidth(). Drop cfs
  from the function names.

This is to prepare for adding bandwidth control support to sched_ext.
tg_bandwidth() will be used as the muxing point similar to tg_weight(). No
functional changes.

Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250614012346.2358261-4-tj@kernel.org

sched/core: Relocate tg_get_cfs_*() and cpu_cfs_*_read_*()

Collect the getters, relocate the trivial interface file wrappers, and put
all of them in period, quota, burst order to prepare for future changes.
Pure reordering. No functional changes.

Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250614012346.2358261-3-tj@kernel.org

sched/fair: Move max_cfs_quota_period decl and default_cfs_period() def from fair.c to sched.h

max_cfs_quota_period is defined in core.c but has a declaration in fair.c.
Move the declaration to kernel/sched/sched.h. Also, move
default_cfs_period() from fair.c to sched.h. No functional changes.

Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250614012346.2358261-2-tj@kernel.org

sched/smp: Use the SMP version of double_rq_clock_clear_update()

Simplify the scheduler by making CONFIG_SMP=y code in
double_rq_clock_clear_update() unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-44-mingo@kernel.org

sched/smp: Use the SMP version of add_nr_running()

Simplify the scheduler by making CONFIG_SMP=y code in
add_nr_running() unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-43-mingo@kernel.org

sched/smp: Use the SMP version of ENQUEUE_MIGRATED

Simplify the scheduler by making the CONFIG_SMP-only ENQUEUE_MIGRATED
flag unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-42-mingo@kernel.org

sched/smp: Use the SMP version of WF_ and SD_ flag sanity checks

Simplify the scheduler by making CONFIG_SMP=y asserts related
to WF_ and SD_ flags unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-41-mingo@kernel.org

sched/smp: Use the SMP version of task_on_cpu()

Simplify the scheduler by making CONFIG_SMP=y code in task_on_cpu()
unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-40-mingo@kernel.org

sched/smp: Use the SMP version of rq_pin_lock()

Simplify the scheduler by making a CONFIG_SMP-only warning
in rq_pin_lock() unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-39-mingo@kernel.org

sched/smp: Use the SMP version of is_migration_disabled()

Simplify the scheduler by making the CONFIG_SMP-only code in
is_migration_disabled() unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-38-mingo@kernel.org

sched/smp: Use the SMP version of cpu_of()

Simplify the scheduler by making CONFIG_SMP=y code in cpu_of()
unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-37-mingo@kernel.org

sched/smp: Use the SMP version of the stop-CPU scheduling class

Simplify the scheduler by making CONFIG_SMP=y code in the stop-CPU
scheduling class unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-36-mingo@kernel.org

sched/smp: Use the SMP version of the idle scheduling class

Simplify the scheduler by making CONFIG_SMP=y code in the
idle scheduling classunconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-35-mingo@kernel.org

sched/smp: Use the SMP version of sched_update_asym_prefer_cpu()

Simplify the scheduler by making CONFIG_SMP=y code in
sched_update_asym_prefer_cpu() unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-34-mingo@kernel.org

sched/smp: Use the SMP version of the scheduler syscalls

Simplify the scheduler by making CONFIG_SMP=y code in
idle_cpu(), __sched_setscheduler() and sched_setaffinity()
unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-33-mingo@kernel.org

sched/smp: Use the SMP version of schedstats

Simplify the scheduler by making CONFIG_SMP=y schedstats
debugging output unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-32-mingo@kernel.org

sched/smp: Use the SMP version of scheduler debugging data

Simplify the scheduler by making CONFIG_SMP=y debug output
unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-31-mingo@kernel.org

sched/smp: Use the SMP version of the deadline scheduling class

Simplify the scheduler by making CONFIG_SMP=y code
in prio_changed_dl() unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-30-mingo@kernel.org

sched/smp: Use the SMP version of the RT scheduling class

Simplify the scheduler by making CONFIG_SMP=y primitives and data
structures unconditional in the RT policies scheduler.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-29-mingo@kernel.org

sched/smp: Use the SMP version of idle_thread_set_boot_cpu()

Simplify the scheduler by making the CONFIG_SMP=y version of
idle_thread_set_boot_cpu() unconditional.

Note that idle_thread_set_boot_cpu() is already conditional
on CONFIG_GENERIC_SMP_IDLE_THREAD, which most architectures
select unconditionally on both UP and SMP kernels.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-28-mingo@kernel.org

sched/smp: Use the SMP version of sched_exec()

Simplify the scheduler making CONFIG_SMP=y sched_exec()
code unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-27-mingo@kernel.org

sched/smp: Use the SMP version of wake_up_new_task()

Simplify the scheduler by making CONFIG_SMP=y code in wake_up_new_task()
unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-26-mingo@kernel.org

sched/smp: Use the SMP version of __task_needs_rq_lock()

Simplify the scheduler by making CONFIG_SMP=y code in
__task_needs_rq_lock() unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-25-mingo@kernel.org

sched/smp: Use the SMP version of try_to_wake_up()

Simplify the scheduler by making CONFIG_SMP=y logic within
try_to_wake_up() unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-24-mingo@kernel.org

sched/smp: Always define rq->hrtick_csd

Simplify the scheduler by making CONFIG_SMP=y data structure
of rq->hrtick_csd unconditional.

Adjust hrtick_start() accordingly, which was split due to the
::hrtick_csd asymmetry and use the SMP version there too.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-23-mingo@kernel.org

sched/smp: Always define is_percpu_thread() and scheduler_ipi()

Simplify the scheduler by making the CONFIG_SMP=y primitives
of is_percpu_thread() and scheduler_ipi() unconditional.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-22-mingo@kernel.org

sched/smp: Make SMP unconditional

Simplify the scheduler by making CONFIG_SMP=y primitives and data
structures unconditional.

Introduce transitory wrappers for functionality not yet converted to SMP.

Note that this patch is pretty large, because there's no clear separation
between various aspects of the SMP scheduler, it's basically a huge block
of #ifdef CONFIG_SMP. A fair amount of it has to be switched on for it to
boot and work on UP systems.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-21-mingo@kernel.org

sched/smp: Always define sched_domains_mutex_lock()/unlock(), def_root_domain and sched_domains_mutex

Simplify the scheduler by making CONFIG_SMP=y primitives and data
structures unconditional.

Unconditionally build kernel/sched/topology.c and the main sched-domains
locking primitives.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-20-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/topology.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

 - Fix whitespace noise and other inconsistencies.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-19-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/syscalls.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-18-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/stats.[ch]

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-17-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/sched.h

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

 - Fix whitespace noise and other inconsistencies.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-16-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/rt.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

 - Fix whitespace noise and other inconsistencies.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-15-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/psi.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-14-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/pelt.[ch]

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

 - Fix whitespace noise and other inconsistencies.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-13-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/loadavg.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

 - Fix whitespace noise and other inconsistencies.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-12-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/idle.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

 - Fix whitespace noise and other inconsistencies.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-11-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/fair.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

 - Fix whitespace noise and other inconsistencies.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-10-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/debug.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

 - Fix whitespace noise and other inconsistencies.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-9-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/deadline.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

 - Fix whitespace noise and other inconsistencies.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-8-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/cputime.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-7-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/cpupri.h

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-6-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/cpufreq_schedutil.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-5-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/core.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

#if CONFIG_FOO
...
#else /* !CONFIG_FOO: */
...
#endif /* !CONFIG_FOO */

 - Apply this simplification:

-#if defined(CONFIG_FOO)
+#ifdef CONFIG_FOO

 - Fix whitespace noise.

 - Use vertical alignment to better visualize nested #ifdef blocks,
   where appropriate:

#ifdef CONFIG_FOO
# ifdef CONFIG_BAR
...
# endif
#endif

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-4-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/clock.c

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-3-mingo@kernel.org

sched: Clean up and standardize #if/#else/#endif markers in sched/autogroup.[ch]

 - Use the standard #ifdef marker format for larger blocks,
   where appropriate:

        #if CONFIG_FOO
        ...
        #else /* !CONFIG_FOO: */
        ...
        #endif /* !CONFIG_FOO */

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <vschneid@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250528080924.2273858-2-mingo@kernel.org

sched/eevdf: Correct the comment in place_entity

Correct "l" to "vl_i".

Signed-off-by: wang wei <a929244872@163.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20250605152931.22804-1-a929244872@163.com

sched: Make clangd usable

Due to the weird Makefile setup of sched the various files do not
compile as stand alone units. The new generation of editors are trying
to do just this -- mostly to offer fancy things like completions but
also better syntax highlighting and code navigation.

Specifically, I've been playing around with neovim and clangd.

Setting up clangd on the kernel source is a giant pain in the arse
(this really should be improved), but once you do manage, you run into
dumb stuff like the above.

Fix up the scheduler files to at least pretend to work.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lkml.kernel.org/r/20250523164348.GN39944@noisy.programming.kicks-ass.net

Linux 6.16-rc1

Merge tag 'turbostat-2025.06.08' of git://git./linux/kernel/git/lenb/linux

Pull turbostat updates from Len Brown:

 - Add initial DMR support, which required smarter RAPL probe

 - Fix AMD MSR RAPL energy reporting

 - Add RAPL power limit configuration output

 - Minor fixes

* tag 'turbostat-2025.06.08' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux:
  tools/power turbostat: version 2025.06.08
  tools/power turbostat: Add initial support for BartlettLake
  tools/power turbostat: Add initial support for DMR
  tools/power turbostat: Dump RAPL sysfs info
  tools/power turbostat: Avoid probing the same perf counters
  tools/power turbostat: Allow probing RAPL with platform_features->rapl_msrs cleared
  tools/power turbostat: Clean up add perf/msr counter logic
  tools/power turbostat: Introduce add_msr_counter()
  tools/power turbostat: Remove add_msr_perf_counter_()
  tools/power turbostat: Remove add_cstate_perf_counter_()
  tools/power turbostat: Remove add_rapl_perf_counter_()
  tools/power turbostat: Quit early for unsupported RAPL counters
  tools/power turbostat: Always check rapl_joules flag
  tools/power turbostat: Fix AMD package-energy reporting
  tools/power turbostat: Fix RAPL_GFX_ALL typo
  tools/power turbostat: Add Android support for MSR device handling
  tools/power turbostat.8: pm_domain wording fix
  tools/power turbostat.8: fix typo: idle_pct should be pct_idle

Merge tag 'timers-cleanups-2025-06-08' of git://git./linux/kernel/git/tip/tip

Pull timer cleanup from Thomas Gleixner:
 "The delayed from_timer() API cleanup:

  The renaming to the timer_*() namespace was delayed due massive
  conflicts against Linux-next. Now that everything is upstream finish
  the conversion"

* tag 'timers-cleanups-2025-06-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  treewide, timers: Rename from_timer() to timer_container_of()

Merge tag 'x86-urgent-2025-06-08' of git://git./linux/kernel/git/tip/tip

Pull x86 fixes from Thomas Gleixner:
 "A small set of x86 fixes:

   - Cure IO bitmap inconsistencies

     A failed fork cleans up all resources of the newly created thread
     via exit_thread(). exit_thread() invokes io_bitmap_exit() which
     does the IO bitmap cleanups, which unfortunately assume that the
     cleanup is related to the current task, which is obviously bogus.

     Make it work correctly

   - A lockdep fix in the resctrl code removed the clearing of the
     command buffer in two places, which keeps stale error messages
     around. Bring them back.

   - Remove unused trace events"

* tag 'x86-urgent-2025-06-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  fs/resctrl: Restore the rdt_last_cmd_clear() calls after acquiring rdtgroup_mutex
  x86/iopl: Cure TIF_IO_BITMAP inconsistencies
  x86/fpu: Remove unused trace events

Merge tag 'timers-urgent-2025-06-08' of git://git./linux/kernel/git/tip/tip

Pull timer fix from Thomas Gleixner:
 "Add the missing seq_file forward declaration in the timer namespace
  header"

* tag 'timers-urgent-2025-06-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  timens: Add struct seq_file forward declaration

tools/power turbostat: version 2025.06.08

Add initial DMR support, which required smarter RAPL probe
Fix AMD MSR RAPL energy reporting
Add RAPL power limit configuration output
Minor fixes

Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Add initial support for BartlettLake

Add initial support for BartlettLake.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Add initial support for DMR

Add initial support for DMR.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Dump RAPL sysfs info

for example:

intel-rapl:1: psys 28.0s:100W 976.0us:100W
intel-rapl:0: package-0 28.0s:57W,max:15W 2.4ms:57W
intel-rapl:0/intel-rapl:0:0: core disabled
intel-rapl:0/intel-rapl:0:1: uncore disabled
intel-rapl-mmio:0: package-0 28.0s:28W,max:15W 2.4ms:57W

[lenb: simplified format]

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
squish me

Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Avoid probing the same perf counters

For the RAPL package energy status counter, Intel and AMD share the same
perf_subsys and perf_name, but with different MSR addresses.

Both rapl_counter_arch_infos[0] and rapl_counter_arch_infos[1] are
introduced to describe this counter for different Vendors.

As a result, the perf counter is probed twice, and causes a failure in
in get_rapl_counters() because expected_read_size and actual_read_size
don't match.

Fix the problem by skipping the already probed counter.

Note, this is not a perfect fix. For example, if different
vendors/platforms use the same MSR value for different purpose, the code
can be fooled when it probes a rapl_counter_arch_infos[] entry that does
not belong to the running Vendor/Platform.

In a long run, better to put rapl_counter_arch_infos[] into the
platform_features so that this becomes Vendor/Platform specific.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Allow probing RAPL with platform_features->rapl_msrs cleared

platform_features->rapl_msrs describes the RAPL MSRs supported. While
RAPL Perf counters can be exposed from different kernel backend drivers,
e.g. RAPL MSR I/F driver, or RAPL TPMI I/F driver.

Thus, turbostat should first blindly probe all the available RAPL Perf
counters, and falls back to the RAPL MSR counters if they are listed in
platform_features->rapl_msrs.

With this, platforms that don't have RAPL MSRs can clear the
platform_features->rapl_msrs bits and use RAPL Perf counters only.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Clean up add perf/msr counter logic

Increase the code readability by moving the no_perf/no_msr flag and the
cai->perf_name/cai->msr sanity checks into the counter probe functions.

No functional change.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Introduce add_msr_counter()

probe_rapl_msr() is reused for probing RAPL MSR counters, cstate MSR
counters and MPERF/APERF/SMI MSR counters, thus its name is misleading.

Similar to add_perf_counter(), introduce add_msr_counter() to probe a
counter via MSR. Introduce wrapper function add_rapl_msr_counter() at
the same time to add extra check for Zero return value for specified
RAPL counters.

No functional change intended.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Remove add_msr_perf_counter_()

As the only caller of add_msr_perf_counter_(), add_msr_perf_counter()
just gives extra debug output on top. There is no need to keep both
functions.

Remove add_msr_perf_counter_() and move all the logic to
add_msr_perf_counter().

No functional change.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Remove add_cstate_perf_counter_()

As the only caller of add_cstate_perf_counter_(),
add_cstate_perf_counter() just gives extra debug output on top. There is
no need to keep both functions.

Remove add_cstate_perf_counter_() and move all the logic to
add_cstate_perf_counter().

No functional change.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Remove add_rapl_perf_counter_()

As the only caller of add_rapl_perf_counter_(), add_rapl_perf_counter()
just gives extra debug output on top. There is no need to keep both
functions.

Remove add_rapl_perf_counter_() and move all the logic to
add_rapl_perf_counter().

No functional change.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Quit early for unsupported RAPL counters

Quit early for unsupported RAPL counters.

No functional change.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Always check rapl_joules flag

rapl_joules bit should always be checked even if
platform_features->rapl_msrs is not set or no_msr flag is used.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Fix AMD package-energy reporting

commit 05a2f07db888 ("tools/power turbostat: read RAPL counters via
perf") that adds support to read RAPL counters via perf defines the
notion of a RAPL domain_id which is set to physical_core_id on
platforms which support per_core_rapl counters (Eg: AMD processors
Family 17h onwards) and is set to the physical_package_id on all the
other platforms.

However, the physical_core_id is only unique within a package and on
platforms with multiple packages more than one core can have the same
physical_core_id and thus the same domain_id. (For eg, the first cores
of each package have the physical_core_id = 0). This results in all
these cores with the same physical_core_id using the same entry in the
rapl_counter_info_perdomain[]. Since rapl_perf_init() skips the
perf-initialization for cores whose domain_ids have already been
visited, cores that have the same physical_core_id always read the
perf file corresponding to the physical_core_id of the first package
and thus the package-energy is incorrectly reported to be the same
value for different packages.

Note: This issue only arises when RAPL counters are read via perf and
not when they are read via MSRs since in the latter case the MSRs are
read separately on each core.

Fix this issue by associating each CPU with rapl_core_id which is
unique across all the packages in the system.

Fixes: 05a2f07db888 ("tools/power turbostat: read RAPL counters via perf")
Signed-off-by: Gautham R. Shenoy <gautham.shenoy@amd.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Fix RAPL_GFX_ALL typo

Fix typo in the currently unused RAPL_GFX_ALL macro definition.

Signed-off-by: Kaushlendra Kumar <kaushlendra.kumar@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat: Add Android support for MSR device handling

It uses /dev/msrN device paths on Android instead of /dev/cpu/N/msr,
updates error messages and permission checks to reflect the Android
device path, and wraps platform-specific code with #if defined(ANDROID)
to ensure correct behavior on both Android and non-Android systems.
These changes improve compatibility and usability of turbostat on
Android devices.

Signed-off-by: Kaushlendra Kumar <kaushlendra.kumar@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat.8: pm_domain wording fix

turbostat.8: clarify that uncore "domains" are Power Management domains,
aka pm_domains.

Signed-off-by: Len Brown <len.brown@intel.com>

tools/power turbostat.8: fix typo: idle_pct should be pct_idle

idle_pct should be pct_idle

Signed-off-by: Len Brown <len.brown@intel.com>

Merge tag 'perf-urgent-2025-06-08' of git://git./linux/kernel/git/tip/tip

Pull x86 perf fix from Thomas Gleixner:
 "A single fix for the x86 performance counters on Intel CPUs:

  The MSR offset calculations for fixed performance counters are stored
  at the wrong index in the configuration array causing the general
  purpose counter MSR offset to be overwritten, so both the general
  purpose and the fixed counters offsets are incorrect.

  Correct the array index calculation to fix that"

* tag 'perf-urgent-2025-06-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/x86/intel: Fix incorrect MSR index calculations in intel_pmu_config_acr()

Merge tag 'irq-urgent-2025-06-08' of git://git./linux/kernel/git/tip/tip

Pull irq fix from Thomas Gleixner:
 "A single fix for the PCI/MSI code:

  The conversion to per device MSI domains created a MSI domain with
  size 1 instead of sizing it to the maximum possible number of MSI
  interrupts for the device. This "worked" as the subsequent allocations
  resized the domain, but the recent change to move the prepare() call
  into the domain creation path broke this works by chance mechanism.

  Size the domain properly at creation time"

* tag 'irq-urgent-2025-06-08' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  PCI/MSI: Size device MSI domain with the maximum number of vectors