Pull PSI updates from Ingo Molnar:

 - Various performance optimizations, resulting in a 4%-9% speedup in
   the mmtests/config-scheduler-perfpipe micro-benchmark.

 - New interface to turn PSI on/off on a per cgroup level.

* tag 'sched-psi-2022-10-14' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched/psi: Per-cgroup PSI accounting disable/re-enable interface
  sched/psi: Cache parent psi_group to speed up group iteration
  sched/psi: Consolidate cgroup_psi()
  sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure
  sched/psi: Remove NR_ONCPU task accounting
  sched/psi: Optimize task switch inside shared cgroups again
  sched/psi: Move private helpers to sched/stats.h
  sched/psi: Save percpu memory when !psi_cgroups_enabled
  sched/psi: Don't create cgroup PSI files when psi_disabled
  sched/psi: Fix periodic aggregation shut off

PSI accounts stalls for each cgroup separately and aggregates it
at each level of the hierarchy. This may cause non-negligible overhead
for some workloads when under deep level of the hierarchy.

commit 3958e2d0c34e ("cgroup: make per-cgroup pressure stall tracking configurable")
make PSI to skip per-cgroup stall accounting, only account system-wide
to avoid this each level overhead.

But for our use case, we also want leaf cgroup PSI stats accounted for
userspace adjustment on that cgroup, apart from only system-wide adjustment.

So this patch introduce a per-cgroup PSI accounting disable/re-enable
interface "cgroup.pressure", which is a read-write single value file that
allowed values are "0" and "1", the defaults is "1" so per-cgroup
PSI stats is enabled by default.

Implementation details:

It should be relatively straight-forward to disable and re-enable
state aggregation, time tracking, averaging on a per-cgroup level,
if we can live with losing history from while it was disabled.
I.e. the avgs will restart from 0, total= will have gaps.

But it's hard or complex to stop/restart groupc->tasks[] updates,
which is not implemented in this patch. So we always update
groupc->tasks[] and PSI_ONCPU bit in psi_group_change() even when
the cgroup PSI stats is disabled.

Suggested-by: Johannes Weiner <hannes@cmpxchg.org>
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Link: https://lkml.kernel.org/r/20220907090332.2078-1-zhouchengming@bytedance.com

CFTYPE_PRESSURE is used to flag PSI related files so that they are not
created if PSI is disabled during boot. It's a bit weird to use a generic
flag to mark a specific file type. Let's instead move the PSI files into its
own cftypes array and add/rm them conditionally. This is a bit more code but
cleaner.

No userland visible changes.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>

Let's track whether a cftype is currently added or not using a new flag
__CFTYPE_ADDED so that duplicate operations can be failed safely and
consistently allow using empty cftypes.

Signed-off-by: Tejun Heo <tj@kernel.org>

Every cgroup knows all its ancestors through its ->ancestor_ids[]. There's
no advantage to remembering the IDs instead of the pointers directly and
this makes the array useless for finding an actual ancestor cgroup forcing
cgroup_ancestor() to iteratively walk up the hierarchy instead. Let's
replace cgroup->ancestor_ids[] with ->ancestors[] and remove the walking-up
from cgroup_ancestor().

While at it, improve comments around cgroup_root->cgrp_ancestor_storage.

This patch shouldn't cause user-visible behavior differences.

v2: Update cgroup_ancestor() to use ->ancestors[].

v3: cgroup_root->cgrp_ancestor_storage's type is updated to match
    cgroup->ancestors[]. Better comments.

Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Namhyung Kim <namhyung@kernel.org>

Pull cgroup updates from Tejun Heo:
 "Several core optimizations:

   - threadgroup_rwsem write locking is skipped when configuring
     controllers in empty subtrees.

     Combined with CLONE_INTO_CGROUP, this allows the common static
     usage pattern to not grab threadgroup_rwsem at all (glibc still
     doesn't seem ready for CLONE_INTO_CGROUP unfortunately).

   - threadgroup_rwsem used to be put into non-percpu mode by default
     due to latency concerns in specific use cases. There's no reason
     for everyone else to pay for it. Make the behavior optional.

   - psi no longer allocates memory when disabled.

  ... along with some code cleanups"

* tag 'cgroup-for-5.20' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
  cgroup: Skip subtree root in cgroup_update_dfl_csses()
  cgroup: remove "no" prefixed mount options
  cgroup: Make !percpu threadgroup_rwsem operations optional
  cgroup: Add "no" prefixed mount options
  cgroup: Elide write-locking threadgroup_rwsem when updating csses on an empty subtree
  cgroup.c: remove redundant check for mixable cgroup in cgroup_migrate_vet_dst
  cgroup.c: add helper __cset_cgroup_from_root to cleanup duplicated codes
  psi: dont alloc memory for psi by default

Pull scheduler updates from Ingo Molnar:
"Load-balancing improvements:

   - Improve NUMA balancing on AMD Zen systems for affine workloads.

   - Improve the handling of reduced-capacity CPUs in load-balancing.

   - Energy Model improvements: fix & refine all the energy fairness
     metrics (PELT), and remove the conservative threshold requiring 6%
     energy savings to migrate a task. Doing this improves power
     efficiency for most workloads, and also increases the reliability
     of energy-efficiency scheduling.

   - Optimize/tweak select_idle_cpu() to spend (much) less time
     searching for an idle CPU on overloaded systems. There's reports of
     several milliseconds spent there on large systems with large
     workloads ...

     [ Since the search logic changed, there might be behavioral side
       effects. ]

   - Improve NUMA imbalance behavior. On certain systems with spare
     capacity, initial placement of tasks is non-deterministic, and such
     an artificial placement imbalance can persist for a long time,
     hurting (and sometimes helping) performance.

     The fix is to make fork-time task placement consistent with runtime
     NUMA balancing placement.

     Note that some performance regressions were reported against this,
     caused by workloads that are not memory bandwith limited, which
     benefit from the artificial locality of the placement bug(s). Mel
     Gorman's conclusion, with which we concur, was that consistency is
     better than random workload benefits from non-deterministic bugs:

        "Given there is no crystal ball and it's a tradeoff, I think
         it's better to be consistent and use similar logic at both fork
         time and runtime even if it doesn't have universal benefit."

   - Improve core scheduling by fixing a bug in
     sched_core_update_cookie() that caused unnecessary forced idling.

   - Improve wakeup-balancing by allowing same-LLC wakeup of idle CPUs
     for newly woken tasks.

   - Fix a newidle balancing bug that introduced unnecessary wakeup
     latencies.

  ABI improvements/fixes:

   - Do not check capabilities and do not issue capability check denial
     messages when a scheduler syscall doesn't require privileges. (Such
     as increasing niceness.)

   - Add forced-idle accounting to cgroups too.

   - Fix/improve the RSEQ ABI to not just silently accept unknown flags.
     (No existing tooling is known to have learned to rely on the
     previous behavior.)

   - Depreciate the (unused) RSEQ_CS_FLAG_NO_RESTART_ON_* flags.

  Optimizations:

   - Optimize & simplify leaf_cfs_rq_list()

   - Micro-optimize set_nr_{and_not,if}_polling() via try_cmpxchg().

  Misc fixes & cleanups:

   - Fix the RSEQ self-tests on RISC-V and Glibc 2.35 systems.

   - Fix a full-NOHZ bug that can in some cases result in the tick not
     being re-enabled when the last SCHED_RT task is gone from a
     runqueue but there's still SCHED_OTHER tasks around.

   - Various PREEMPT_RT related fixes.

   - Misc cleanups & smaller fixes"

* tag 'sched-core-2022-08-01' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (32 commits)
  rseq: Kill process when unknown flags are encountered in ABI structures
  rseq: Deprecate RSEQ_CS_FLAG_NO_RESTART_ON_* flags
  sched/core: Fix the bug that task won't enqueue into core tree when update cookie
  nohz/full, sched/rt: Fix missed tick-reenabling bug in dequeue_task_rt()
  sched/core: Always flush pending blk_plug
  sched/fair: fix case with reduced capacity CPU
  sched/core: Use try_cmpxchg in set_nr_{and_not,if}_polling
  sched/core: add forced idle accounting for cgroups
  sched/fair: Remove the energy margin in feec()
  sched/fair: Remove task_util from effective utilization in feec()
  sched/fair: Use the same cpumask per-PD throughout find_energy_efficient_cpu()
  sched/fair: Rename select_idle_mask to select_rq_mask
  sched, drivers: Remove max param from effective_cpu_util()/sched_cpu_util()
  sched/fair: Decay task PELT values during wakeup migration
  sched/fair: Provide u64 read for 32-bits arch helper
  sched/fair: Introduce SIS_UTIL to search idle CPU based on sum of util_avg
  sched: only perform capability check on privileged operation
  sched: Remove unused function group_first_cpu()
  sched/fair: Remove redundant word " *"
  selftests/rseq: check if libc rseq support is registered
  ...

3942a9bd7b58 ("locking, rcu, cgroup: Avoid synchronize_sched() in
__cgroup_procs_write()") disabled percpu operations on threadgroup_rwsem
because the impiled synchronize_rcu() on write locking was pushing up the
latencies too much for android which constantly moves processes between
cgroups.

This makes the hotter paths - fork and exit - slower as they're always
forced into the slow path. There is no reason to force this on everyone
especially given that more common static usage pattern can now completely
avoid write-locking the rwsem. Write-locking is elided when turning on and
off controllers on empty sub-trees and CLONE_INTO_CGROUP enables seeding a
cgroup without grabbing the rwsem.

Restore the default percpu operations and introduce the mount option
"favordynmods" and config option CGROUP_FAVOR_DYNMODS for users who need
lower latencies for the dynamic operations.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Michal Koutn� <mkoutny@suse.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Dmitry Shmidt <dimitrysh@google.com>
Cc: Oleg Nesterov <oleg@redhat.com>

4feee7d1260 previously added per-task forced idle accounting. This patch
extends this to also include cgroups.

rstat is used for cgroup accounting, except for the root, which uses
kcpustat in order to bypass the need for doing an rstat flush when
reading root stats.

Only cgroup v2 is supported. Similar to the task accounting, the cgroup
accounting requires that schedstats is enabled.

Signed-off-by: Josh Don <joshdon@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lkml.kernel.org/r/20220629211426.3329954-1-joshdon@google.com

Each cset (css_set) is pinned by its tasks. When we're moving tasks around
across csets for a migration, we need to hold the source and destination
csets to ensure that they don't go away while we're moving tasks about. This
is done by linking cset->mg_preload_node on either the
mgctx->preloaded_src_csets or mgctx->preloaded_dst_csets list. Using the
same cset->mg_preload_node for both the src and dst lists was deemed okay as
a cset can't be both the source and destination at the same time.

Unfortunately, this overloading becomes problematic when multiple tasks are
involved in a migration and some of them are identity noop migrations while
others are actually moving across cgroups. For example, this can happen with
the following sequence on cgroup1:

 #1> mkdir -p /sys/fs/cgroup/misc/a/b
 #2> echo $$ > /sys/fs/cgroup/misc/a/cgroup.procs
 #3> RUN_A_COMMAND_WHICH_CREATES_MULTIPLE_THREADS &
 #4> PID=$!
 #5> echo $PID > /sys/fs/cgroup/misc/a/b/tasks
 #6> echo $PID > /sys/fs/cgroup/misc/a/cgroup.procs

the process including the group leader back into a. In this final migration,
non-leader threads would be doing identity migration while the group leader
is doing an actual one.

After #3, let's say the whole process was in cset A, and that after #4, the
leader moves to cset B. Then, during #6, the following happens:

 1. cgroup_migrate_add_src() is called on B for the leader.

 2. cgroup_migrate_add_src() is called on A for the other threads.

 3. cgroup_migrate_prepare_dst() is called. It scans the src list.

 4. It notices that B wants to migrate to A, so it tries to A to the dst
    list but realizes that its ->mg_preload_node is already busy.

 5. and then it notices A wants to migrate to A as it's an identity
    migration, it culls it by list_del_init()'ing its ->mg_preload_node and
    putting references accordingly.

 6. The rest of migration takes place with B on the src list but nothing on
    the dst list.

This means that A isn't held while migration is in progress. If all tasks
leave A before the migration finishes and the incoming task pins it, the
cset will be destroyed leading to use-after-free.

This is caused by overloading cset->mg_preload_node for both src and dst
preload lists. We wanted to exclude the cset from the src list but ended up
inadvertently excluding it from the dst list too.

This patch fixes the issue by separating out cset->mg_preload_node into
->mg_src_preload_node and ->mg_dst_preload_node, so that the src and dst
preloadings don't interfere with each other.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Mukesh Ojha <quic_mojha@quicinc.com>
Reported-by: shisiyuan <shisiyuan19870131@gmail.com>
Link: http://lkml.kernel.org/r/1654187688-27411-1-git-send-email-shisiyuan@xiaomi.com
Link: https://www.spinics.net/lists/cgroups/msg33313.html
Fixes: f817de98513d ("cgroup: prepare migration path for unified hierarchy")
Cc: stable@vger.kernel.org # v3.16+