Displaying "PREEMPT" on kernel headers when CONFIG_PREEMPT_DYNAMIC=y
can be misleading for anybody involved in remote debugging because it
is then not guaranteed that there is an actual preemption behaviour. It
depends on default Kconfig or boot defined choices.

Therefore, tell about PREEMPT_DYNAMIC on static kernel headers and leave
the search for the actual preemption behaviour to browsing dmesg.

Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20220217111240.GA742892@lothringen

While investigating the sparse warning reported by the LKP bot [1],
observed that we have a redundant variable "top" in the function
build_sched_domains that was introduced in the recent commit
e496132e ("sched/fair: Adjust the allowed NUMA imbalance when
SD_NUMA spans multiple LLCs")

The existing variable "sd" suffices which allows us to remove the
redundant variable "top" while annotating the other variable "top_p"
with the "__rcu" annotation to silence the sparse warning.

[1] https://lore.kernel.org/lkml/202202170853.9vofgC3O-lkp@intel.com/

Fixes: e496132e

 ("sched/fair: Adjust the allowed NUMA imbalance when SD_NUMA spans multiple LLCs")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: K Prateek Nayak <kprateek.nayak@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lore.kernel.org/r/20220218162743.1134-1-kprateek.nayak@amd.com

The `struct rq *rq` parameter isn't used. Remove it.

Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lore.kernel.org/r/20220302183433.333029-7-dietmar.eggemann@arm.com

The need_pull_[rt|dl]_task() and pull_[rt|dl]_task() functions are not
used on a !CONFIG_SMP system. Remove them.

Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lore.kernel.org/r/20220302183433.333029-6-dietmar.eggemann@arm.com

Deploy __node_2_pdl(node), __node_2_dle(node) and rb_first_cached()
consistently throughout the sched class source file which makes the
code at least easier to read.

Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lore.kernel.org/r/20220302183433.333029-5-dietmar.eggemann@arm.com

Both functions are doing almost the same, that is checking if admission
control is still respected.

With exclusive cpusets, dl_task_can_attach() checks if the destination
cpuset (i.e. its root domain) has enough CPU capacity to accommodate the
task.
dl_cpu_busy() checks if there is enough CPU capacity in the cpuset in
case the CPU is hot-plugged out.

dl_task_can_attach() is used to check if a task can be admitted while
dl_cpu_busy() is used to check if a CPU can be hotplugged out.

Make dl_cpu_busy() able to deal with a task and use it instead of
dl_task_can_attach() in task_can_attach().

Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lore.kernel.org/r/20220302183433.333029-4-dietmar.eggemann@arm.com

Move the deadline bandwidth management (admission control) functions
__dl_add(), __dl_sub() and __dl_overflow() as well as the bandwidth
reclaim function __dl_update() from private task scheduler header file
to the deadline sched class source file.
The functions are only used internally so they don't have to be
exported.

Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lore.kernel.org/r/20220302183433.333029-3-dietmar.eggemann@arm.com

Since commit 1724813d

 ("sched/deadline: Remove the sysctl_sched_dl
knobs") the default deadline bandwidth control structure has no purpose.
Remove it.

Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lore.kernel.org/r/20220302183433.333029-2-dietmar.eggemann@arm.com

TASK_RTLOCK_WAIT currently isn't part of TASK_REPORT, thus a task blocking
on an rtlock will appear as having a task state == 0, IOW TASK_RUNNING.

The actual state is saved in p->saved_state, but reading it after reading
p->__state has a few issues:
o that could still be TASK_RUNNING in the case of e.g. rt_spin_lock
o ttwu_state_match() might have changed that to TASK_RUNNING

As pointed out by Eric, adding TASK_RTLOCK_WAIT to TASK_REPORT implies
exposing a new state to userspace tools which way not know what to do with
them. The only information that needs to be conveyed here is that a task is
waiting on an rt_mutex, which matches TASK_UNINTERRUPTIBLE - there's no
need for a new state.

Reported-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Link: https://lore.kernel.org/r/20220120162520.570782-3-valentin.schneider@arm.com

As of commit

  c6e7bd7a

 ("sched/core: Optimize ttwu() spinning on p->on_cpu")

the following sequence becomes possible:

		      p->__state = TASK_INTERRUPTIBLE;
		      __schedule()
			deactivate_task(p);
  ttwu()
    READ !p->on_rq
    p->__state=TASK_WAKING
			trace_sched_switch()
			  __trace_sched_switch_state()
			    task_state_index()
			      return 0;

TASK_WAKING isn't in TASK_REPORT, so the task appears as TASK_RUNNING in
the trace event.

Prevent this by pushing the value read from __schedule() down the trace
event.

Reported-by: Abhijeet Dharmapurikar <adharmap@quicinc.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Link: https://lore.kernel.org/r/20220120162520.570782-2-valentin.schneider@arm.com

John reported that push_rt_task() can end up invoking
find_lowest_rq(rq->curr) when curr is not an RT task (in this case a CFS
one), which causes mayhem down convert_prio().

This can happen when current gets demoted to e.g. CFS when releasing an
rt_mutex, and the local CPU gets hit with an rto_push_work irqwork before
getting the chance to reschedule. Exactly who triggers this work isn't
entirely clear to me - switched_from_rt() only invokes rt_queue_pull_task()
if there are no RT tasks on the local RQ, which means the local CPU can't
be in the rto_mask.

My current suspected sequence is something along the lines of the below,
with the demoted task being current.

  mark_wakeup_next_waiter()
    rt_mutex_adjust_prio()
      rt_mutex_setprio() // deboost originally-CFS task
	check_class_changed()
	  switched_from_rt() // Only rt_queue_pull_task() if !rq->rt.rt_nr_running
	  switched_to_fair() // Sets need_resched
      __balance_callbacks() // if pull_rt_task(), tell_cpu_to_push() can't select local CPU per the above
      raw_spin_rq_unlock(rq)

       // need_resched is set, so task_woken_rt() can't
       // invoke push_rt_tasks(). Best I can come up with is
       // local CPU has rt_nr_migratory >= 2 after the demotion, so stays
       // in the rto_mask, and then:

       <some other CPU running rto_push_irq_work_func() queues rto_push_work on this CPU>
	 push_rt_task()
	   // breakage follows here as rq->curr is CFS

Move an existing check to check rq->curr vs the next pushable task's
priority before getting anywhere near find_lowest_rq(). While at it, add an
explicit sched_class of rq->curr check prior to invoking
find_lowest_rq(rq->curr). Align the DL logic to also reschedule regardless
of next_task's migratability.

Fixes: a7c81556

 ("sched: Fix migrate_disable() vs rt/dl balancing")
Reported-by: John Keeping <john@metanate.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Tested-by: John Keeping <john@metanate.com>
Link: https://lore.kernel.org/r/20220127154059.974729-1-valentin.schneider@arm.com

The cpuacct_account_field() and it's cgroup v2 wrapper
cgroup_account_cputime_field() is only called from cputime
in task_group_account_field(), which is already in RCU read-side
critical section. So remove these redundant RCU read lock.

Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20220220051426.5274-3-zhouchengming@bytedance.com

Since cpuacct_charge() is called from the scheduler update_curr(),
we must already have rq lock held, then the RCU read lock can
be optimized away.

And do the same thing in it's wrapper cgroup_account_cputime(),
but we can't use lockdep_assert_rq_held() there, which defined
in kernel/sched/sched.h.

Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20220220051426.5274-2-zhouchengming@bytedance.com

The cpuacct_account_field() is always called by the current task
itself, so it's ok to use __this_cpu_add() to charge the tick time.

But cpuacct_charge() maybe called by update_curr() in load_balance()
on a random CPU, different from the CPU on which the task is running.
So __this_cpu_add() will charge that cputime to a random incorrect CPU.

Fixes: 73e6aafd

 ("sched/cpuacct: Simplify the cpuacct code")
Reported-by: Minye Zhu <zhuminye@bytedance.com>
Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20220220051426.5274-1-zhouchengming@bytedance.com

New conflicts in sched/core due to the following upstream fixes:

  44585f7b ("psi: fix "defined but not used" warnings when CONFIG_PROC_FS=n")
  a06247c6

 ("psi: Fix uaf issue when psi trigger is destroyed while being polled")

Conflicts:
	include/linux/psi_types.h
	kernel/sched/psi.c

Signed-off-by: Ingo Molnar <mingo@kernel.org>

Pull locking fix from Borislav Petkov:
 "Fix a NULL ptr dereference when dumping lockdep chains through
  /proc/lockdep_chains"

* tag 'locking_urgent_for_v5.17_rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  lockdep: Correct lock_classes index mapping

Pull x86 fixes from Borislav Petkov:

 - Fix the ptrace regset xfpregs_set() callback to behave according to
   the ABI

 - Handle poisoned pages properly in the SGX reclaimer code

* tag 'x86_urgent_for_v5.17_rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/ptrace: Fix xfpregs_set()'s incorrect xmm clearing
  x86/sgx: Fix missing poison handling in reclaimer

Pull scheduler fix from Borislav Petkov:
 "Fix task exposure order when forking tasks"

* tag 'sched_urgent_for_v5.17_rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched: Fix yet more sched_fork() races

Pull EDAC fix from Borislav Petkov:
 "Fix a long-standing struct alignment bug in the EDAC struct allocation
  code"

* tag 'edac_urgent_for_v5.17_rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras:
  EDAC: Fix calculation of returned address and next offset in edac_align_ptr()

Pull SCSI fixes from James Bottomley:
 "Three fixes, all in drivers.

  The ufs and qedi fixes are minor; the lpfc one is a bit bigger because
  it involves adding a heuristic to detect and deal with common but not
  standards compliant behaviour"

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
  scsi: ufs: core: Fix divide by zero in ufshcd_map_queues()
  scsi: lpfc: Fix pt2pt NVMe PRLI reject LOGO loop
  scsi: qedi: Fix ABBA deadlock in qedi_process_tmf_resp() and qedi_process_cmd_cleanup_resp()

Pull dmaengine fixes from Vinod Koul:
 "A bunch of driver fixes for:

   - ptdma error handling in init

   - lock fix in at_hdmac

   - error path and error num fix for sh dma

   - pm balance fix for stm32"

* tag 'dmaengine-fix-5.17' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine:
  dmaengine: shdma: Fix runtime PM imbalance on error
  dmaengine: sh: rcar-dmac: Check for error num after dma_set_max_seg_size
  dmaengine: stm32-dmamux: Fix PM disable depth imbalance in stm32_dmamux_probe
  dmaengine: sh: rcar-dmac: Check for error num after setting mask
  dmaengine: at_xdmac: Fix missing unlock in at_xdmac_tasklet()
  dmaengine: ptdma: Fix the error handling path in pt_core_init()

Pull i2c fixes from Wolfram Sang:
 "Some driver updates, a MAINTAINERS fix, and additions to COMPILE_TEST
  (so we won't miss build problems again)"

* 'i2c/for-current' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
  MAINTAINERS: remove duplicate entry for i2c-qcom-geni
  i2c: brcmstb: fix support for DSL and CM variants
  i2c: qup: allow COMPILE_TEST
  i2c: imx: allow COMPILE_TEST
  i2c: cadence: allow COMPILE_TEST
  i2c: qcom-cci: don't put a device tree node before i2c_add_adapter()
  i2c: qcom-cci: don't delete an unregistered adapter
  i2c: bcm2835: Avoid clock stretching timeouts

Pull input fixes from Dmitry Torokhov:

 - a fix for Synaptics touchpads in RMI4 mode failing to suspend/resume
   properly because I2C client devices are now being suspended and
   resumed asynchronously which changed the ordering

 - a change to make sure we do not set right and middle buttons
   capabilities on touchpads that are "buttonpads" (i.e. do not have
   separate physical buttons)

 - a change to zinitix touchscreen driver adding more compatible
   strings/IDs

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: psmouse - set up dependency between PS/2 and SMBus companions
  Input: zinitix - add new compatible strings
  Input: clear BTN_RIGHT/MIDDLE on buttonpads

Pull power supply fixes from Sebastian Reichel:
 "Three regression fixes for the 5.17 cycle:

   - build warning fix for power-supply documentation

   - pointer size fix in cw2015 battery driver

   - OOM handling in bq256xx charger driver"

* tag 'for-v5.17-rc' of git://git.kernel.org/pub/scm/linux/kernel/git/sre/linux-power-supply:
  power: supply: bq256xx: Handle OOM correctly
  power: supply: core: fix application of sizeof to pointer
  power: supply: fix table problem in sysfs-class-power

Merge tag 'fs.mount_setattr.v5.17-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux

Pull mount_setattr test/doc fixes from Christian Brauner:
 "This contains a fix for one of the selftests for the mount_setattr
  syscall to create idmapped mounts, an entry for idmapped mounts for
  maintainers, and missing kernel documentation for the helper we split
  out some time ago to get and yield write access to a mount when
  changing mount properties"

* tag 'fs.mount_setattr.v5.17-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux:
  fs: add kernel doc for mnt_{hold,unhold}_writers()
  MAINTAINERS: add entry for idmapped mounts
  tests: fix idmapped mount_setattr test

Pull pidfd fix from Christian Brauner:
 "This fixes a problem reported by lockdep when installing a pidfd via
  fd_install() with siglock and the tasklisk write lock held in
  copy_process() when calling clone()/clone3() with CLONE_PIDFD.

  Originally a pidfd was created prior to holding any of these locks but
  this required a call to ksys_close(). So quite some time ago in
  6fd2fe49 ("copy_process(): don't use ksys_close() on cleanups") we
  switched to a get_unused_fd_flags() + fd_install() model.

  As part of that we moved fd_install() as late as possible. This was
  done for two main reasons. First, because we needed to ensure that we
  call fd_install() past the point of no return as once that's called
  the fd is live in the task's file table. Second, because we tried to
  ensure that the fd is visible in /proc/<pid>/fd/<pidfd> right when the
  task is visible.

  This fix moves the fd_install() to an even later point which means
  that a task will be visible in proc while the pidfd isn't yet under
  /proc/<pid>/fd/<pidfd>.

  While this is a user visible change it's very unlikely that this will
  have any impact. Nobody should be relying on that and if they do we
  need to come up with something better but again, it's doubtful this is
  relevant"

* tag 'pidfd.v5.17-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux:
  copy_process(): Move fd_install() out of sighand->siglock critical section

Merge branch 'ucount-rlimit-fixes-for-v5.17' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace

Pull ucounts fixes from Eric Biederman:
 "Michal Koutný recently found some bugs in the enforcement of
  RLIMIT_NPROC in the recent ucount rlimit implementation.

  In this set of patches I have developed a very conservative approach
  changing only what is necessary to fix the bugs that I can see
  clearly. Cleanups and anything that is making the code more consistent
  can follow after we have the code working as it has historically.

  The problem is not so much inconsistencies (although those exist) but
  that it is very difficult to figure out what the code should be doing
  in the case of RLIMIT_NPROC.

  All other rlimits are only enforced where the resource is acquired
  (allocated). RLIMIT_NPROC by necessity needs to be enforced in an
  additional location, and our current implementation stumbled it's way
  into that implementation"

* 'ucount-rlimit-fixes-for-v5.17' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace:
  ucounts: Handle wrapping in is_ucounts_overlimit
  ucounts: Move RLIMIT_NPROC handling after set_user
  ucounts: Base set_cred_ucounts changes on the real user
  ucounts: Enforce RLIMIT_NPROC not RLIMIT_NPROC+1
  rlimit: Fix RLIMIT_NPROC enforcement failure caused by capability calls in set_user

The driver is already covered in the ARM/QUALCOMM section. Also, Akash
Asthana's email bounces meanwhile and Mukesh Savaliya has never
responded to mails regarding this driver.

Signed-off-by: Wolfram Sang <wsa@kernel.org>
Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Wolfram Sang <wsa@kernel.org>

This patch enables support for PREEMPT_DYNAMIC on arm64, allowing the
preemption model to be chosen at boot time.

Specifically, this patch selects HAVE_PREEMPT_DYNAMIC_KEY, so that each
preemption function is an out-of-line call with an early return
depending upon a static key. This leaves almost all the codegen up to
the compiler, and side-steps a number of pain points with static calls
(e.g. interaction with CFI schemes). This should have no worse overhead
than using non-inline static calls, as those use out-of-line trampolines
with early returns.

For example, the dynamic_cond_resched() wrapper looks as follows when
enabled. When disabled, the first `B` is replaced with a `NOP`,
resulting in an early return.

| <dynamic_cond_resched>:
|        bti     c
|        b       <dynamic_cond_resched+0x10>     // or `nop`
|        mov     w0, #0x0
|        ret
|        mrs     x0, sp_el0
|        ldr     x0, [x0, #8]
|        cbnz    x0, <dynamic_cond_resched+0x8>
|        paciasp
|        stp     x29, x30, [sp, #-16]!
|        mov     x29, sp
|        bl      <preempt_schedule_common>
|        mov     w0, #0x1
|        ldp     x29, x30, [sp], #16
|        autiasp
|        ret

... compared to the regular form of the function:

| <__cond_resched>:
|        bti     c
|        mrs     x0, sp_el0
|        ldr     x1, [x0, #8]
|        cbz     x1, <__cond_resched+0x18>
|        mov     w0, #0x0
|        ret
|        paciasp
|        stp     x29, x30, [sp, #-16]!
|        mov     x29, sp
|        bl      <preempt_schedule_common>
|        mov     w0, #0x1
|        ldp     x29, x30, [sp], #16
|        autiasp
|        ret

Since arm64 does not yet use the generic entry code, we must define our
own `sk_dynamic_irqentry_exit_cond_resched`, which will be
enabled/disabled by the common code in kernel/sched/core.c. All other
preemption functions and associated static keys are defined there.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20220214165216.2231574-8-mark.rutland@arm.com

For historical reasons, the decision of whether or not to preempt is
spread across arm64_preempt_schedule_irq() and __el1_irq(), and it would
be clearer if this were all in one place.

Also, arm64_preempt_schedule_irq() calls lockdep_assert_irqs_disabled(),
but this is redundant, as we have a subsequent identical assertion in
__exit_to_kernel_mode(), and preempt_schedule_irq() will
BUG_ON(!irqs_disabled()) anyway.

This patch removes the redundant assertion and centralizes the
preemption decision making within arm64_preempt_schedule_irq().

Other than the slight change to assertion behaviour, there should be no
functional change as a result of this patch.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20220214165216.2231574-7-mark.rutland@arm.com

Where an architecture selects HAVE_STATIC_CALL but not
HAVE_STATIC_CALL_INLINE, each static call has an out-of-line trampoline
which will either branch to a callee or return to the caller.

On such architectures, a number of constraints can conspire to make
those trampolines more complicated and potentially less useful than we'd
like. For example:

* Hardware and software control flow integrity schemes can require the
  addition of "landing pad" instructions (e.g. `BTI` for arm64), which
  will also be present at the "real" callee.

* Limited branch ranges can require that trampolines generate or load an
  address into a register and perform an indirect branch (or at least
  have a slow path that does so). This loses some of the benefits of
  having a direct branch.

* Interaction with SW CFI schemes can be complicated and fragile, e.g.
  requiring that we can recognise idiomatic codegen and remove
  indirections understand, at least until clang proves more helpful
  mechanisms for dealing with this.

For PREEMPT_DYNAMIC, we don't need the full power of static calls, as we
really only need to enable/disable specific preemption functions. We can
achieve the same effect without a number of the pain points above by
using static keys to fold early returns into the preemption functions
themselves rather than in an out-of-line trampoline, effectively
inlining the trampoline into the start of the function.

For arm64, this results in good code generation. For example, the
dynamic_cond_resched() wrapper looks as follows when enabled. When
disabled, the first `B` is replaced with a `NOP`, resulting in an early
return.

| <dynamic_cond_resched>:
|        bti     c
|        b       <dynamic_cond_resched+0x10>     // or `nop`
|        mov     w0, #0x0
|        ret
|        mrs     x0, sp_el0
|        ldr     x0, [x0, #8]
|        cbnz    x0, <dynamic_cond_resched+0x8>
|        paciasp
|        stp     x29, x30, [sp, #-16]!
|        mov     x29, sp
|        bl      <preempt_schedule_common>
|        mov     w0, #0x1
|        ldp     x29, x30, [sp], #16
|        autiasp
|        ret

... compared to the regular form of the function:

| <__cond_resched>:
|        bti     c
|        mrs     x0, sp_el0
|        ldr     x1, [x0, #8]
|        cbz     x1, <__cond_resched+0x18>
|        mov     w0, #0x0
|        ret
|        paciasp
|        stp     x29, x30, [sp, #-16]!
|        mov     x29, sp
|        bl      <preempt_schedule_common>
|        mov     w0, #0x1
|        ldp     x29, x30, [sp], #16
|        autiasp
|        ret

Any architecture which implements static keys should be able to use this
to implement PREEMPT_DYNAMIC with similar cost to non-inlined static
calls. Since this is likely to have greater overhead than (inlined)
static calls, PREEMPT_DYNAMIC is only defaulted to enabled when
HAVE_PREEMPT_DYNAMIC_CALL is selected.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20220214165216.2231574-6-mark.rutland@arm.com

Now that the enabled/disabled states for the preemption functions are
declared alongside their definitions, the core PREEMPT_DYNAMIC logic is
no longer tied to GENERIC_ENTRY, and can safely be selected so long as
an architecture provides enabled/disabled states for
irqentry_exit_cond_resched().

Make it possible to select HAVE_PREEMPT_DYNAMIC without GENERIC_ENTRY.

For existing users of HAVE_PREEMPT_DYNAMIC there should be no functional
change as a result of this patch.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20220214165216.2231574-5-mark.rutland@arm.com

Currently callers of irqentry_exit_cond_resched() need to be aware of
whether the function should be indirected via a static call, leading to
ugly ifdeffery in callers.

Save them the hassle with a static inline wrapper that does the right
thing. The raw_irqentry_exit_cond_resched() will also be useful in
subsequent patches which will add conditional wrappers for preemption
functions.

Note: in arch/x86/entry/common.c, xen_pv_evtchn_do_upcall() always calls
irqentry_exit_cond_resched() directly, even when PREEMPT_DYNAMIC is in
use. I believe this is a latent bug (which this patch corrects), but I'm
not entirely certain this wasn't deliberate.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20220214165216.2231574-4-mark.rutland@arm.com

Currently sched_dynamic_update needs to open-code the enabled/disabled
function names for each preemption model it supports, when in practice
this is a boolean enabled/disabled state for each function.

Make this clearer and avoid repetition by defining the enabled/disabled
states at the function definition, and using helper macros to perform the
static_call_update(). Where x86 currently overrides the enabled
function, it is made to provide both the enabled and disabled states for
consistency, with defaults provided by the core code otherwise.

In subsequent patches this will allow us to support PREEMPT_DYNAMIC
without static calls.

There should be no functional change as a result of this patch.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20220214165216.2231574-3-mark.rutland@arm.com

The PREEMPT_DYNAMIC logic in kernel/sched/core.c patches static calls
for a bunch of preemption functions. While most are defined prior to
this, the definition of cond_resched() is later in the file, and so we
only have its declarations from include/linux/sched.h.

In subsequent patches we'd like to define some macros alongside the
definition of each of the preemption functions, which we can use within
sched_dynamic_update(). For this to be possible, the PREEMPT_DYNAMIC
logic needs to be placed after the various preemption functions.

As a preparatory step, this patch moves the PREEMPT_DYNAMIC logic after
the various preemption functions, with no other changes -- this is
purely a move.

There should be no functional change as a result of this patch.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Acked-by: Frederic Weisbecker <frederic@kernel.org>
Link: https://lore.kernel.org/r/20220214165216.2231574-2-mark.rutland@arm.com

Where commit 4ef0c5c6 ("kernel/sched: Fix sched_fork() access an
invalid sched_task_group") fixed a fork race vs cgroup, it opened up a
race vs syscalls by not placing the task on the runqueue before it
gets exposed through the pidhash.

Commit 13765de8 ("sched/fair: Fix fault in reweight_entity") is
trying to fix a single instance of this, instead fix the whole class
of issues, effectively reverting this commit.

Fixes: 4ef0c5c6

 ("kernel/sched: Fix sched_fork() access an invalid sched_task_group")
Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Tadeusz Struk <tadeusz.struk@linaro.org>
Tested-by: Zhang Qiao <zhangqiao22@huawei.com>
Tested-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Link: https://lkml.kernel.org/r/YgoeCbwj5mbCR0qA@hirez.programming.kicks-ass.net

Pull NFS client bugfixes from Anna Schumaker:

 - Fix unnecessary changeattr revalidations

 - Fix resolving symlinks during directory lookups

 - Don't report writeback errors in nfs_getattr()

* tag 'nfs-for-5.17-3' of git://git.linux-nfs.org/projects/anna/linux-nfs:
  NFS: Do not report writeback errors in nfs_getattr()
  NFS: LOOKUP_DIRECTORY is also ok with symlinks
  NFS: Remove an incorrect revalidation in nfs4_update_changeattr_locked()

Pull ACPI fixes from Rafael Wysocki:
 "These make an excess warning message go away and fix a recently
  introduced boot failure on a vintage machine.

  Specifics:

   - Change the log level of the "table not found" message in
     acpi_table_parse_entries_array() to debug to prevent it from
     showing up in the logs unnecessarily (Dan Williams)

   - Add a C-state limit quirk for 32-bit ThinkPad T40 to prevent it
     from crashing on boot after recent changes in the ACPI processor
     driver (Woody Suwalski)"

* tag 'acpi-5.17-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  ACPI: processor: idle: fix lockup regression on 32-bit ThinkPad T40
  ACPI: tables: Quiet ACPI table not found warning

Pull RISC-V fixes from Palmer Dabbelt:
 "A set of three fixes, all aimed at fixing some fallout from the recent
  sparse hart ID support"

* tag 'riscv-for-linus-5.17-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:
  RISC-V: Fix IPI/RFENCE hmask on non-monotonic hartid ordering
  RISC-V: Fix handling of empty cpu masks
  RISC-V: Fix hartid mask handling for hartid 31 and up