The estimated utilization for a task:
util_est = max(util_avg, est.enqueue, est.ewma)
is defined based on:
- util_avg: the PELT defined utilization
- est.enqueued: the util_avg at the end of the last activation
- est.ewma: a exponential moving average on the est.enqueued samples
According to this definition, when a task suddenly changes its bandwidth
requirements from small to big, the EWMA will need to collect multiple
samples before converging up to track the new big utilization.
This slow convergence towards bigger utilization values is not
aligned to the default scheduler behavior, which is to optimize for
performance. Moreover, the est.ewma component fails to compensate for
temporarely utilization drops which spans just few est.enqueued samples.
To let util_est do a better job in the scenario depicted above, change
its definition by making util_est directly follow upward motion and
only decay the est.ewma on downward.
Signed-off-by: Patrick Bellasi <patrick.bellasi@matbug.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Douglas Raillard <douglas.raillard@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Quentin Perret <qperret@google.com>
Cc: Rafael J . Wysocki <rafael.j.wysocki@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20191023205630.14469-1-patrick.bellasi@matbug.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
While the static key is correctly initialized as being disabled, it will
remain forever enabled once turned on. This means that if we start with an
asymmetric system and hotplug out enough CPUs to end up with an SMP system,
the static key will remain set - which is obviously wrong. We should detect
this and turn off things like misfit migration and capacity aware wakeups.
As Quentin pointed out, having separate root domains makes this slightly
trickier. We could have exclusive cpusets that create an SMP island - IOW,
the domains within this root domain will not see any asymmetry. This means
we can't just disable the key on domain destruction, we need to count how
many asymmetric root domains we have.
Consider the following example using Juno r0 which is 2+4 big.LITTLE, where
two identical cpusets are created: they both span both big and LITTLE CPUs:
asym0 asym1
[ ][ ]
L L B L L B
$ cgcreate -g cpuset:asym0
$ cgset -r cpuset.cpus=0,1,3 asym0
$ cgset -r cpuset.mems=0 asym0
$ cgset -r cpuset.cpu_exclusive=1 asym0
$ cgcreate -g cpuset:asym1
$ cgset -r cpuset.cpus=2,4,5 asym1
$ cgset -r cpuset.mems=0 asym1
$ cgset -r cpuset.cpu_exclusive=1 asym1
$ cgset -r cpuset.sched_load_balance=0 .
(the CPU numbering may look odd because on the Juno LITTLEs are CPUs 0,3-5
and bigs are CPUs 1-2)
If we make one of those SMP (IOW remove asymmetry) by e.g. hotplugging its
big core, we would end up with an SMP cpuset and an asymmetric cpuset - the
static key must remain set, because we still have one asymmetric root domain.
With the above example, this could be done with:
$ echo 0 > /sys/devices/system/cpu/cpu2/online
Which would result in:
asym0 asym1
[ ][ ]
L L B L L
When both SMP and asymmetric cpusets are present, all CPUs will observe
sched_asym_cpucapacity being set (it is system-wide), but not all CPUs
observe asymmetry in their sched domain hierarchy:
per_cpu(sd_asym_cpucapacity, <any CPU in asym0>) == <some SD at DIE level>
per_cpu(sd_asym_cpucapacity, <any CPU in asym1>) == NULL
Change the simple key enablement to an increment, and decrement the key
counter when destroying domains that cover asymmetric CPUs.
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>
Cc: Dietmar.Eggemann@arm.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hannes@cmpxchg.org
Cc: lizefan@huawei.com
Cc: morten.rasmussen@arm.com
Cc: qperret@google.com
Cc: tj@kernel.org
Cc: vincent.guittot@linaro.org
Fixes: df054e8445 ("sched/topology: Add static_key for asymmetric CPU capacity optimizations")
Link: https://lkml.kernel.org/r/20191023153745.19515-3-valentin.schneider@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Remove the nr_exclusive argument from __wake_up_sync_key() and derived
functions as everything seems to set it to 1. Note also that if it wasn't
set to 1, it would clear WF_SYNC anyway.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
The load_balance() algorithm contains some heuristics which have become
meaningless since the rework of the scheduler's metrics like the
introduction of PELT.
Furthermore, load is an ill-suited metric for solving certain task
placement imbalance scenarios.
For instance, in the presence of idle CPUs, we should simply try to get at
least one task per CPU, whereas the current load-based algorithm can actually
leave idle CPUs alone simply because the load is somewhat balanced.
The current algorithm ends up creating virtual and meaningless values like
the avg_load_per_task or tweaks the state of a group to make it overloaded
whereas it's not, in order to try to migrate tasks.
load_balance() should better qualify the imbalance of the group and clearly
define what has to be moved to fix this imbalance.
The type of sched_group has been extended to better reflect the type of
imbalance. We now have:
group_has_spare
group_fully_busy
group_misfit_task
group_asym_packing
group_imbalanced
group_overloaded
Based on the type of sched_group, load_balance now sets what it wants to
move in order to fix the imbalance. It can be some load as before but also
some utilization, a number of task or a type of task:
migrate_task
migrate_util
migrate_load
migrate_misfit
This new load_balance() algorithm fixes several pending wrong tasks
placement:
- the 1 task per CPU case with asymmetric system
- the case of cfs task preempted by other class
- the case of tasks not evenly spread on groups with spare capacity
Also the load balance decisions have been consolidated in the 3 functions
below after removing the few bypasses and hacks of the current code:
- update_sd_pick_busiest() select the busiest sched_group.
- find_busiest_group() checks if there is an imbalance between local and
busiest group.
- calculate_imbalance() decides what have to be moved.
Finally, the now unused field total_running of struct sd_lb_stats has been
removed.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Ben Segall <bsegall@google.com>
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: Mike Galbraith <efault@gmx.de>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hdanton@sina.com
Cc: parth@linux.ibm.com
Cc: pauld@redhat.com
Cc: quentin.perret@arm.com
Cc: riel@surriel.com
Cc: srikar@linux.vnet.ibm.com
Cc: valentin.schneider@arm.com
Link: https://lkml.kernel.org/r/1571405198-27570-5-git-send-email-vincent.guittot@linaro.org
[ Small readability and spelling updates. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull scheduler fixes from Ingo Molnar:
"Two fixes: a guest-cputime accounting fix, and a cgroup bandwidth
quota precision fix"
* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/vtime: Fix guest/system mis-accounting on task switch
sched/fair: Scale bandwidth quota and period without losing quota/period ratio precision
On context switch we are locking the vtime seqcount of the scheduling-out
task twice:
* On vtime_task_switch_common(), when we flush the pending vtime through
vtime_account_system()
* On arch_vtime_task_switch() to reset the vtime state.
This is pointless as these actions can be performed without the need
to unlock/lock in the middle. The reason these steps are separated is to
consolidate a very small amount of common code between
CONFIG_VIRT_CPU_ACCOUNTING_GEN and CONFIG_VIRT_CPU_ACCOUNTING_NATIVE.
Performance in this fast path is definitely a priority over artificial
code factorization so split the task switch code between GEN and
NATIVE and mutualize the parts than can run under a single seqcount
locked block.
As a side effect, vtime_account_idle() becomes included in the seqcount
protection. This happens to be a welcome preparation in order to
properly support kcpustat under vtime in the future and fetch
CPUTIME_IDLE without race.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Wanpeng Li <wanpengli@tencent.com>
Cc: Yauheni Kaliuta <yauheni.kaliuta@redhat.com>
Link: https://lkml.kernel.org/r/20191003161745.28464-3-frederic@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
vtime_account_system() assumes that the target task to account cputime
to is always the current task. This is most often true indeed except on
task switch where we call:
vtime_common_task_switch(prev)
vtime_account_system(prev)
Here prev is the scheduling-out task where we account the cputime to. It
doesn't match current that is already the scheduling-in task at this
stage of the context switch.
So we end up checking the wrong task flags to determine if we are
accounting guest or system time to the previous task.
As a result the wrong task is used to check if the target is running in
guest mode. We may then spuriously account or leak either system or
guest time on task switch.
Fix this assumption and also turn vtime_guest_enter/exit() to use the
task passed in parameter as well to avoid future similar issues.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Wanpeng Li <wanpengli@tencent.com>
Fixes: 2a42eb9594 ("sched/cputime: Accumulate vtime on top of nsec clocksource")
Link: https://lkml.kernel.org/r/20190925214242.21873-1-frederic@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The quota/period ratio is used to ensure a child task group won't get
more bandwidth than the parent task group, and is calculated as:
normalized_cfs_quota() = [(quota_us << 20) / period_us]
If the quota/period ratio was changed during this scaling due to
precision loss, it will cause inconsistency between parent and child
task groups.
See below example:
A userspace container manager (kubelet) does three operations:
1) Create a parent cgroup, set quota to 1,000us and period to 10,000us.
2) Create a few children cgroups.
3) Set quota to 1,000us and period to 10,000us on a child cgroup.
These operations are expected to succeed. However, if the scaling of
147/128 happens before step 3, quota and period of the parent cgroup
will be changed:
new_quota: 1148437ns, 1148us
new_period: 11484375ns, 11484us
And when step 3 comes in, the ratio of the child cgroup will be
104857, which will be larger than the parent cgroup ratio (104821),
and will fail.
Scaling them by a factor of 2 will fix the problem.
Tested-by: Phil Auld <pauld@redhat.com>
Signed-off-by: Xuewei Zhang <xueweiz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Phil Auld <pauld@redhat.com>
Cc: Anton Blanchard <anton@ozlabs.org>
Cc: Ben Segall <bsegall@google.com>
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: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Fixes: 2e8e192263 ("sched/fair: Limit sched_cfs_period_timer() loop to avoid hard lockup")
Link: https://lkml.kernel.org/r/20191004001243.140897-1-xueweiz@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull copy_struct_from_user() helper from Christian Brauner:
"This contains the copy_struct_from_user() helper which got split out
from the openat2() patchset. It is a generic interface designed to
copy a struct from userspace.
The helper will be especially useful for structs versioned by size of
which we have quite a few. This allows for backwards compatibility,
i.e. an extended struct can be passed to an older kernel, or a legacy
struct can be passed to a newer kernel. For the first case (extended
struct, older kernel) the new fields in an extended struct can be set
to zero and the struct safely passed to an older kernel.
The most obvious benefit is that this helper lets us get rid of
duplicate code present in at least sched_setattr(), perf_event_open(),
and clone3(). More importantly it will also help to ensure that users
implementing versioning-by-size end up with the same core semantics.
This point is especially crucial since we have at least one case where
versioning-by-size is used but with slighly different semantics:
sched_setattr(), perf_event_open(), and clone3() all do do similar
checks to copy_struct_from_user() while rt_sigprocmask(2) always
rejects differently-sized struct arguments.
With this pull request we also switch over sched_setattr(),
perf_event_open(), and clone3() to use the new helper"
* tag 'copy-struct-from-user-v5.4-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux:
usercopy: Add parentheses around assignment in test_copy_struct_from_user
perf_event_open: switch to copy_struct_from_user()
sched_setattr: switch to copy_struct_from_user()
clone3: switch to copy_struct_from_user()
lib: introduce copy_struct_from_user() helper
Switch sched_setattr() syscall from it's own copying struct sched_attr
from userspace to the new dedicated copy_struct_from_user() helper.
The change is very straightforward, and helps unify the syscall
interface for struct-from-userspace syscalls. Ideally we could also
unify sched_getattr(2)-style syscalls as well, but unfortunately the
correct semantics for such syscalls are much less clear (see [1] for
more detail). In future we could come up with a more sane idea for how
the syscall interface should look.
[1]: commit 1251201c0d ("sched/core: Fix uclamp ABI bug, clean up and
robustify sched_read_attr() ABI logic and code")
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Christian Brauner <christian.brauner@ubuntu.com>
[christian.brauner@ubuntu.com: improve commit message]
Link: https://lore.kernel.org/r/20191001011055.19283-4-cyphar@cyphar.com
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Pull scheduler fixes from Ingo Molnar:
- Apply a number of membarrier related fixes and cleanups, which fixes
a use-after-free race in the membarrier code
- Introduce proper RCU protection for tasks on the runqueue - to get
rid of the subtle task_rcu_dereference() interface that was easy to
get wrong
- Misc fixes, but also an EAS speedup
* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/fair: Avoid redundant EAS calculation
sched/core: Remove double update_max_interval() call on CPU startup
sched/core: Fix preempt_schedule() interrupt return comment
sched/fair: Fix -Wunused-but-set-variable warnings
sched/core: Fix migration to invalid CPU in __set_cpus_allowed_ptr()
sched/membarrier: Return -ENOMEM to userspace on memory allocation failure
sched/membarrier: Skip IPIs when mm->mm_users == 1
selftests, sched/membarrier: Add multi-threaded test
sched/membarrier: Fix p->mm->membarrier_state racy load
sched/membarrier: Call sync_core only before usermode for same mm
sched/membarrier: Remove redundant check
sched/membarrier: Fix private expedited registration check
tasks, sched/core: RCUify the assignment of rq->curr
tasks, sched/core: With a grace period after finish_task_switch(), remove unnecessary code
tasks, sched/core: Ensure tasks are available for a grace period after leaving the runqueue
tasks: Add a count of task RCU users
sched/core: Convert vcpu_is_preempted() from macro to an inline function
sched/fair: Remove unused cfs_rq_clock_task() function
Commit:
de53fd7aed ("sched/fair: Fix low cpu usage with high throttling by removing expiration of cpu-local slices")
introduced a few compilation warnings:
kernel/sched/fair.c: In function '__refill_cfs_bandwidth_runtime':
kernel/sched/fair.c:4365:6: warning: variable 'now' set but not used [-Wunused-but-set-variable]
kernel/sched/fair.c: In function 'start_cfs_bandwidth':
kernel/sched/fair.c:4992:6: warning: variable 'overrun' set but not used [-Wunused-but-set-variable]
Also, __refill_cfs_bandwidth_runtime() does no longer update the
expiration time, so fix the comments accordingly.
Signed-off-by: Qian Cai <cai@lca.pw>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Ben Segall <bsegall@google.com>
Reviewed-by: Dave Chiluk <chiluk+linux@indeed.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: pauld@redhat.com
Fixes: de53fd7aed ("sched/fair: Fix low cpu usage with high throttling by removing expiration of cpu-local slices")
Link: https://lkml.kernel.org/r/1566326455-8038-1-git-send-email-cai@lca.pw
Signed-off-by: Ingo Molnar <mingo@kernel.org>
An oops can be triggered in the scheduler when running qemu on arm64:
Unable to handle kernel paging request at virtual address ffff000008effe40
Internal error: Oops: 96000007 [#1] SMP
Process migration/0 (pid: 12, stack limit = 0x00000000084e3736)
pstate: 20000085 (nzCv daIf -PAN -UAO)
pc : __ll_sc___cmpxchg_case_acq_4+0x4/0x20
lr : move_queued_task.isra.21+0x124/0x298
...
Call trace:
__ll_sc___cmpxchg_case_acq_4+0x4/0x20
__migrate_task+0xc8/0xe0
migration_cpu_stop+0x170/0x180
cpu_stopper_thread+0xec/0x178
smpboot_thread_fn+0x1ac/0x1e8
kthread+0x134/0x138
ret_from_fork+0x10/0x18
__set_cpus_allowed_ptr() will choose an active dest_cpu in affinity mask to
migrage the process if process is not currently running on any one of the
CPUs specified in affinity mask. __set_cpus_allowed_ptr() will choose an
invalid dest_cpu (dest_cpu >= nr_cpu_ids, 1024 in my virtual machine) if
CPUS in an affinity mask are deactived by cpu_down after cpumask_intersects
check. cpumask_test_cpu() of dest_cpu afterwards is overflown and may pass if
corresponding bit is coincidentally set. As a consequence, kernel will
access an invalid rq address associate with the invalid CPU in
migration_cpu_stop->__migrate_task->move_queued_task and the Oops occurs.
The reproduce the crash:
1) A process repeatedly binds itself to cpu0 and cpu1 in turn by calling
sched_setaffinity.
2) A shell script repeatedly does "echo 0 > /sys/devices/system/cpu/cpu1/online"
and "echo 1 > /sys/devices/system/cpu/cpu1/online" in turn.
3) Oops appears if the invalid CPU is set in memory after tested cpumask.
Signed-off-by: KeMeng Shi <shikemeng@huawei.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/1568616808-16808-1-git-send-email-shikemeng@huawei.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The membarrier_state field is located within the mm_struct, which
is not guaranteed to exist when used from runqueue-lock-free iteration
on runqueues by the membarrier system call.
Copy the membarrier_state from the mm_struct into the scheduler runqueue
when the scheduler switches between mm.
When registering membarrier for mm, after setting the registration bit
in the mm membarrier state, issue a synchronize_rcu() to ensure the
scheduler observes the change. In order to take care of the case
where a runqueue keeps executing the target mm without swapping to
other mm, iterate over each runqueue and issue an IPI to copy the
membarrier_state from the mm_struct into each runqueue which have the
same mm which state has just been modified.
Move the mm membarrier_state field closer to pgd in mm_struct to use
a cache line already touched by the scheduler switch_mm.
The membarrier_execve() (now membarrier_exec_mmap) hook now needs to
clear the runqueue's membarrier state in addition to clear the mm
membarrier state, so move its implementation into the scheduler
membarrier code so it can access the runqueue structure.
Add memory barrier in membarrier_exec_mmap() prior to clearing
the membarrier state, ensuring memory accesses executed prior to exec
are not reordered with the stores clearing the membarrier state.
As suggested by Linus, move all membarrier.c RCU read-side locks outside
of the for each cpu loops.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Chris Metcalf <cmetcalf@ezchip.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Kirill Tkhai <tkhai@yandex.ru>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Paul E. McKenney <paulmck@linux.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King - ARM Linux admin <linux@armlinux.org.uk>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20190919173705.2181-5-mathieu.desnoyers@efficios.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The current task on the runqueue is currently read with rcu_dereference().
To obtain ordinary RCU semantics for an rcu_dereference() of rq->curr it needs
to be paired with rcu_assign_pointer() of rq->curr. Which provides the
memory barrier necessary to order assignments to the task_struct
and the assignment to rq->curr.
Unfortunately the assignment of rq->curr in __schedule is a hot path,
and it has already been show that additional barriers in that code
will reduce the performance of the scheduler. So I will attempt to
describe below why you can effectively have ordinary RCU semantics
without any additional barriers.
The assignment of rq->curr in init_idle is a slow path called once
per cpu and that can use rcu_assign_pointer() without any concerns.
As I write this there are effectively two users of rcu_dereference() on
rq->curr. There is the membarrier code in kernel/sched/membarrier.c
that only looks at "->mm" after the rcu_dereference(). Then there is
task_numa_compare() in kernel/sched/fair.c. My best reading of the
code shows that task_numa_compare only access: "->flags",
"->cpus_ptr", "->numa_group", "->numa_faults[]",
"->total_numa_faults", and "->se.cfs_rq".
The code in __schedule() essentially does:
rq_lock(...);
smp_mb__after_spinlock();
next = pick_next_task(...);
rq->curr = next;
context_switch(prev, next);
At the start of the function the rq_lock/smp_mb__after_spinlock
pair provides a full memory barrier. Further there is a full memory barrier
in context_switch().
This means that any task that has already run and modified itself (the
common case) has already seen two memory barriers before __schedule()
runs and begins executing. A task that modifies itself then sees a
third full memory barrier pair with the rq_lock();
For a brand new task that is enqueued with wake_up_new_task() there
are the memory barriers present from the taking and release the
pi_lock and the rq_lock as the processes is enqueued as well as the
full memory barrier at the start of __schedule() assuming __schedule()
happens on the same cpu.
This means that by the time we reach the assignment of rq->curr
except for values on the task struct modified in pick_next_task
the code has the same guarantees as if it used rcu_assign_pointer().
Reading through all of the implementations of pick_next_task it
appears pick_next_task is limited to modifying the task_struct fields
"->se", "->rt", "->dl". These fields are the sched_entity structures
of the varies schedulers.
Further "->se.cfs_rq" is only changed in cgroup attach/move operations
initialized by userspace.
Unless I have missed something this means that in practice that the
users of "rcu_dereference(rq->curr)" get normal RCU semantics of
rcu_dereference() for the fields the care about, despite the
assignment of rq->curr in __schedule() ot using rcu_assign_pointer.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Chris Metcalf <cmetcalf@ezchip.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Kirill Tkhai <tkhai@yandex.ru>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Paul E. McKenney <paulmck@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King - ARM Linux admin <linux@armlinux.org.uk>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20190903200603.GW2349@hirez.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Remove work arounds that were written before there was a grace period
after tasks left the runqueue in finish_task_switch().
In particular now that there tasks exiting the runqueue exprience
a RCU grace period none of the work performed by task_rcu_dereference()
excpet the rcu_dereference() is necessary so replace task_rcu_dereference()
with rcu_dereference().
Remove the code in rcuwait_wait_event() that checks to ensure the current
task has not exited. It is no longer necessary as it is guaranteed
that any running task will experience a RCU grace period after it
leaves the run queueue.
Remove the comment in rcuwait_wake_up() as it is no longer relevant.
Ref: 8f95c90ceb ("sched/wait, RCU: Introduce rcuwait machinery")
Ref: 150593bf86 ("sched/api: Introduce task_rcu_dereference() and try_get_task_struct()")
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Chris Metcalf <cmetcalf@ezchip.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Kirill Tkhai <tkhai@yandex.ru>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Paul E. McKenney <paulmck@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King - ARM Linux admin <linux@armlinux.org.uk>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/87lfurdpk9.fsf_-_@x220.int.ebiederm.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
In the ordinary case today the RCU grace period for a task_struct is
triggered when another process wait's for it's zombine and causes the
kernel to call release_task(). As the waiting task has to receive a
signal and then act upon it before this happens, typically this will
occur after the original task as been removed from the runqueue.
Unfortunaty in some cases such as self reaping tasks it can be shown
that release_task() will be called starting the grace period for
task_struct long before the task leaves the runqueue.
Therefore use put_task_struct_rcu_user() in finish_task_switch() to
guarantee that the there is a RCU lifetime after the task
leaves the runqueue.
Besides the change in the start of the RCU grace period for the
task_struct this change may cause perf_event_delayed_put and
trace_sched_process_free. The function perf_event_delayed_put boils
down to just a WARN_ON for cases that I assume never show happen. So
I don't see any problem with delaying it.
The function trace_sched_process_free is a trace point and thus
visible to user space. Occassionally userspace has the strangest
dependencies so this has a miniscule chance of causing a regression.
This change only changes the timing of when the tracepoint is called.
The change in timing arguably gives userspace a more accurate picture
of what is going on. So I don't expect there to be a regression.
In the case where a task self reaps we are pretty much guaranteed that
the RCU grace period is delayed. So we should get quite a bit of
coverage in of this worst case for the change in a normal threaded
workload. So I expect any issues to turn up quickly or not at all.
I have lightly tested this change and everything appears to work
fine.
Inspired-by: Linus Torvalds <torvalds@linux-foundation.org>
Inspired-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Chris Metcalf <cmetcalf@ezchip.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Kirill Tkhai <tkhai@yandex.ru>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul E. McKenney <paulmck@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King - ARM Linux admin <linux@armlinux.org.uk>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/87r24jdpl5.fsf_-_@x220.int.ebiederm.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Patch series "mm: remove quicklist page table caches".
A while ago Nicholas proposed to remove quicklist page table caches [1].
I've rebased his patch on the curren upstream and switched ia64 and sh to
use generic versions of PTE allocation.
[1] https://lore.kernel.org/linux-mm/20190711030339.20892-1-npiggin@gmail.com
This patch (of 3):
Remove page table allocator "quicklists". These have been around for a
long time, but have not got much traction in the last decade and are only
used on ia64 and sh architectures.
The numbers in the initial commit look interesting but probably don't
apply anymore. If anybody wants to resurrect this it's in the git
history, but it's unhelpful to have this code and divergent allocator
behaviour for minor archs.
Also it might be better to instead make more general improvements to page
allocator if this is still so slow.
Link: http://lkml.kernel.org/r/1565250728-21721-2-git-send-email-rppt@linux.ibm.com
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull hmm updates from Jason Gunthorpe:
"This is more cleanup and consolidation of the hmm APIs and the very
strongly related mmu_notifier interfaces. Many places across the tree
using these interfaces are touched in the process. Beyond that a
cleanup to the page walker API and a few memremap related changes
round out the series:
- General improvement of hmm_range_fault() and related APIs, more
documentation, bug fixes from testing, API simplification &
consolidation, and unused API removal
- Simplify the hmm related kconfigs to HMM_MIRROR and DEVICE_PRIVATE,
and make them internal kconfig selects
- Hoist a lot of code related to mmu notifier attachment out of
drivers by using a refcount get/put attachment idiom and remove the
convoluted mmu_notifier_unregister_no_release() and related APIs.
- General API improvement for the migrate_vma API and revision of its
only user in nouveau
- Annotate mmu_notifiers with lockdep and sleeping region debugging
Two series unrelated to HMM or mmu_notifiers came along due to
dependencies:
- Allow pagemap's memremap_pages family of APIs to work without
providing a struct device
- Make walk_page_range() and related use a constant structure for
function pointers"
* tag 'for-linus-hmm' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (75 commits)
libnvdimm: Enable unit test infrastructure compile checks
mm, notifier: Catch sleeping/blocking for !blockable
kernel.h: Add non_block_start/end()
drm/radeon: guard against calling an unpaired radeon_mn_unregister()
csky: add missing brackets in a macro for tlb.h
pagewalk: use lockdep_assert_held for locking validation
pagewalk: separate function pointers from iterator data
mm: split out a new pagewalk.h header from mm.h
mm/mmu_notifiers: annotate with might_sleep()
mm/mmu_notifiers: prime lockdep
mm/mmu_notifiers: add a lockdep map for invalidate_range_start/end
mm/mmu_notifiers: remove the __mmu_notifier_invalidate_range_start/end exports
mm/hmm: hmm_range_fault() infinite loop
mm/hmm: hmm_range_fault() NULL pointer bug
mm/hmm: fix hmm_range_fault()'s handling of swapped out pages
mm/mmu_notifiers: remove unregister_no_release
RDMA/odp: remove ib_ucontext from ib_umem
RDMA/odp: use mmu_notifier_get/put for 'struct ib_ucontext_per_mm'
RDMA/mlx5: Use odp instead of mr->umem in pagefault_mr
RDMA/mlx5: Use ib_umem_start instead of umem.address
...
Pull power management updates from Rafael Wysocki:
"These include a rework of the main suspend-to-idle code flow (related
to the handling of spurious wakeups), a switch over of several users
of cpufreq notifiers to QoS-based limits, a new devfreq driver for
Tegra20, a new cpuidle driver and governor for virtualized guests, an
extension of the wakeup sources framework to expose wakeup sources as
device objects in sysfs, and more.
Specifics:
- Rework the main suspend-to-idle control flow to avoid repeating
"noirq" device resume and suspend operations in case of spurious
wakeups from the ACPI EC and decouple the ACPI EC wakeups support
from the LPS0 _DSM support (Rafael Wysocki).
- Extend the wakeup sources framework to expose wakeup sources as
device objects in sysfs (Tri Vo, Stephen Boyd).
- Expose system suspend statistics in sysfs (Kalesh Singh).
- Introduce a new haltpoll cpuidle driver and a new matching governor
for virtualized guests wanting to do guest-side polling in the idle
loop (Marcelo Tosatti, Joao Martins, Wanpeng Li, Stephen Rothwell).
- Fix the menu and teo cpuidle governors to allow the scheduler tick
to be stopped if PM QoS is used to limit the CPU idle state exit
latency in some cases (Rafael Wysocki).
- Increase the resolution of the play_idle() argument to microseconds
for more fine-grained injection of CPU idle cycles (Daniel
Lezcano).
- Switch over some users of cpuidle notifiers to the new QoS-based
frequency limits and drop the CPUFREQ_ADJUST and CPUFREQ_NOTIFY
policy notifier events (Viresh Kumar).
- Add new cpufreq driver based on nvmem for sun50i (Yangtao Li).
- Add support for MT8183 and MT8516 to the mediatek cpufreq driver
(Andrew-sh.Cheng, Fabien Parent).
- Add i.MX8MN support to the imx-cpufreq-dt cpufreq driver (Anson
Huang).
- Add qcs404 to cpufreq-dt-platdev blacklist (Jorge Ramirez-Ortiz).
- Update the qcom cpufreq driver (among other things, to make it
easier to extend and to use kryo cpufreq for other nvmem-based
SoCs) and add qcs404 support to it (Niklas Cassel, Douglas
RAILLARD, Sibi Sankar, Sricharan R).
- Fix assorted issues and make assorted minor improvements in the
cpufreq code (Colin Ian King, Douglas RAILLARD, Florian Fainelli,
Gustavo Silva, Hariprasad Kelam).
- Add new devfreq driver for NVidia Tegra20 (Dmitry Osipenko, Arnd
Bergmann).
- Add new Exynos PPMU events to devfreq events and extend that
mechanism (Lukasz Luba).
- Fix and clean up the exynos-bus devfreq driver (Kamil Konieczny).
- Improve devfreq documentation and governor code, fix spelling typos
in devfreq (Ezequiel Garcia, Krzysztof Kozlowski, Leonard Crestez,
MyungJoo Ham, Gaël PORTAY).
- Add regulators enable and disable to the OPP (operating performance
points) framework (Kamil Konieczny).
- Update the OPP framework to support multiple opp-suspend properties
(Anson Huang).
- Fix assorted issues and make assorted minor improvements in the OPP
code (Niklas Cassel, Viresh Kumar, Yue Hu).
- Clean up the generic power domains (genpd) framework (Ulf Hansson).
- Clean up assorted pieces of power management code and documentation
(Akinobu Mita, Amit Kucheria, Chuhong Yuan).
- Update the pm-graph tool to version 5.5 including multiple fixes
and improvements (Todd Brandt).
- Update the cpupower utility (Benjamin Weis, Geert Uytterhoeven,
Sébastien Szymanski)"
* tag 'pm-5.4-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (126 commits)
cpuidle-haltpoll: Enable kvm guest polling when dedicated physical CPUs are available
cpuidle-haltpoll: do not set an owner to allow modunload
cpuidle-haltpoll: return -ENODEV on modinit failure
cpuidle-haltpoll: set haltpoll as preferred governor
cpuidle: allow governor switch on cpuidle_register_driver()
PM: runtime: Documentation: add runtime_status ABI document
pm-graph: make setVal unbuffered again for python2 and python3
powercap: idle_inject: Use higher resolution for idle injection
cpuidle: play_idle: Increase the resolution to usec
cpuidle-haltpoll: vcpu hotplug support
cpufreq: Add qcs404 to cpufreq-dt-platdev blacklist
cpufreq: qcom: Add support for qcs404 on nvmem driver
cpufreq: qcom: Refactor the driver to make it easier to extend
cpufreq: qcom: Re-organise kryo cpufreq to use it for other nvmem based qcom socs
dt-bindings: opp: Add qcom-opp bindings with properties needed for CPR
dt-bindings: opp: qcom-nvmem: Support pstates provided by a power domain
Documentation: cpufreq: Update policy notifier documentation
cpufreq: Remove CPUFREQ_ADJUST and CPUFREQ_NOTIFY policy notifier events
PM / Domains: Verify PM domain type in dev_pm_genpd_set_performance_state()
PM / Domains: Simplify genpd_lookup_dev()
...
Pull core timer updates from Thomas Gleixner:
"Timers and timekeeping updates:
- A large overhaul of the posix CPU timer code which is a preparation
for moving the CPU timer expiry out into task work so it can be
properly accounted on the task/process.
An update to the bogus permission checks will come later during the
merge window as feedback was not complete before heading of for
travel.
- Switch the timerqueue code to use cached rbtrees and get rid of the
homebrewn caching of the leftmost node.
- Consolidate hrtimer_init() + hrtimer_init_sleeper() calls into a
single function
- Implement the separation of hrtimers to be forced to expire in hard
interrupt context even when PREEMPT_RT is enabled and mark the
affected timers accordingly.
- Implement a mechanism for hrtimers and the timer wheel to protect
RT against priority inversion and live lock issues when a (hr)timer
which should be canceled is currently executing the callback.
Instead of infinitely spinning, the task which tries to cancel the
timer blocks on a per cpu base expiry lock which is held and
released by the (hr)timer expiry code.
- Enable the Hyper-V TSC page based sched_clock for Hyper-V guests
resulting in faster access to timekeeping functions.
- Updates to various clocksource/clockevent drivers and their device
tree bindings.
- The usual small improvements all over the place"
* 'timers-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (101 commits)
posix-cpu-timers: Fix permission check regression
posix-cpu-timers: Always clear head pointer on dequeue
hrtimer: Add a missing bracket and hide `migration_base' on !SMP
posix-cpu-timers: Make expiry_active check actually work correctly
posix-timers: Unbreak CONFIG_POSIX_TIMERS=n build
tick: Mark sched_timer to expire in hard interrupt context
hrtimer: Add kernel doc annotation for HRTIMER_MODE_HARD
x86/hyperv: Hide pv_ops access for CONFIG_PARAVIRT=n
posix-cpu-timers: Utilize timerqueue for storage
posix-cpu-timers: Move state tracking to struct posix_cputimers
posix-cpu-timers: Deduplicate rlimit handling
posix-cpu-timers: Remove pointless comparisons
posix-cpu-timers: Get rid of 64bit divisions
posix-cpu-timers: Consolidate timer expiry further
posix-cpu-timers: Get rid of zero checks
rlimit: Rewrite non-sensical RLIMIT_CPU comment
posix-cpu-timers: Respect INFINITY for hard RTTIME limit
posix-cpu-timers: Switch thread group sampling to array
posix-cpu-timers: Restructure expiry array
posix-cpu-timers: Remove cputime_expires
...
* pm-cpufreq: (36 commits)
cpufreq: Add qcs404 to cpufreq-dt-platdev blacklist
cpufreq: qcom: Add support for qcs404 on nvmem driver
cpufreq: qcom: Refactor the driver to make it easier to extend
cpufreq: qcom: Re-organise kryo cpufreq to use it for other nvmem based qcom socs
dt-bindings: opp: Add qcom-opp bindings with properties needed for CPR
dt-bindings: opp: qcom-nvmem: Support pstates provided by a power domain
Documentation: cpufreq: Update policy notifier documentation
cpufreq: Remove CPUFREQ_ADJUST and CPUFREQ_NOTIFY policy notifier events
sched/cpufreq: Align trace event behavior of fast switching
ACPI: cpufreq: Switch to QoS requests instead of cpufreq notifier
video: pxafb: Remove cpufreq policy notifier
video: sa1100fb: Remove cpufreq policy notifier
arch_topology: Use CPUFREQ_CREATE_POLICY instead of CPUFREQ_NOTIFY
cpufreq: powerpc_cbe: Switch to QoS requests for freq limits
cpufreq: powerpc: macintosh: Switch to QoS requests for freq limits
cpufreq: Print driver name if cpufreq_suspend() fails
cpufreq: mediatek: Add support for mt8183
cpufreq: mediatek: change to regulator_get_optional
cpufreq: imx-cpufreq-dt: Add i.MX8MN support
cpufreq: Use imx-cpufreq-dt for i.MX8MN's speed grading
...
* pm-cpuidle:
cpuidle-haltpoll: Enable kvm guest polling when dedicated physical CPUs are available
cpuidle-haltpoll: do not set an owner to allow modunload
cpuidle-haltpoll: return -ENODEV on modinit failure
cpuidle-haltpoll: set haltpoll as preferred governor
cpuidle: allow governor switch on cpuidle_register_driver()
powercap: idle_inject: Use higher resolution for idle injection
cpuidle: play_idle: Increase the resolution to usec
cpuidle-haltpoll: vcpu hotplug support
cpuidle: teo: Get rid of redundant check in teo_update()
cpuidle: teo: Allow tick to be stopped if PM QoS is used
cpuidle: menu: Allow tick to be stopped if PM QoS is used
cpuidle: header file stubs must be "static inline"
cpuidle-haltpoll: disable host side polling when kvm virtualized
cpuidle: add haltpoll governor
governors: unify last_state_idx
cpuidle: add poll_limit_ns to cpuidle_device structure
add cpuidle-haltpoll driver
Pull scheduler updates from Ingo Molnar:
- MAINTAINERS: Add Mark Rutland as perf submaintainer, Juri Lelli and
Vincent Guittot as scheduler submaintainers. Add Dietmar Eggemann,
Steven Rostedt, Ben Segall and Mel Gorman as scheduler reviewers.
As perf and the scheduler is getting bigger and more complex,
document the status quo of current responsibilities and interests,
and spread the review pain^H^H^H^H fun via an increase in the Cc:
linecount generated by scripts/get_maintainer.pl. :-)
- Add another series of patches that brings the -rt (PREEMPT_RT) tree
closer to mainline: split the monolithic CONFIG_PREEMPT dependencies
into a new CONFIG_PREEMPTION category that will allow the eventual
introduction of CONFIG_PREEMPT_RT. Still a few more hundred patches
to go though.
- Extend the CPU cgroup controller with uclamp.min and uclamp.max to
allow the finer shaping of CPU bandwidth usage.
- Micro-optimize energy-aware wake-ups from O(CPUS^2) to O(CPUS).
- Improve the behavior of high CPU count, high thread count
applications running under cpu.cfs_quota_us constraints.
- Improve balancing with SCHED_IDLE (SCHED_BATCH) tasks present.
- Improve CPU isolation housekeeping CPU allocation NUMA locality.
- Fix deadline scheduler bandwidth calculations and logic when cpusets
rebuilds the topology, or when it gets deadline-throttled while it's
being offlined.
- Convert the cpuset_mutex to percpu_rwsem, to allow it to be used from
setscheduler() system calls without creating global serialization.
Add new synchronization between cpuset topology-changing events and
the deadline acceptance tests in setscheduler(), which were broken
before.
- Rework the active_mm state machine to be less confusing and more
optimal.
- Rework (simplify) the pick_next_task() slowpath.
- Improve load-balancing on AMD EPYC systems.
- ... and misc cleanups, smaller fixes and improvements - please see
the Git log for more details.
* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (53 commits)
sched/psi: Correct overly pessimistic size calculation
sched/fair: Speed-up energy-aware wake-ups
sched/uclamp: Always use 'enum uclamp_id' for clamp_id values
sched/uclamp: Update CPU's refcount on TG's clamp changes
sched/uclamp: Use TG's clamps to restrict TASK's clamps
sched/uclamp: Propagate system defaults to the root group
sched/uclamp: Propagate parent clamps
sched/uclamp: Extend CPU's cgroup controller
sched/topology: Improve load balancing on AMD EPYC systems
arch, ia64: Make NUMA select SMP
sched, perf: MAINTAINERS update, add submaintainers and reviewers
sched/fair: Use rq_lock/unlock in online_fair_sched_group
cpufreq: schedutil: fix equation in comment
sched: Rework pick_next_task() slow-path
sched: Allow put_prev_task() to drop rq->lock
sched/fair: Expose newidle_balance()
sched: Add task_struct pointer to sched_class::set_curr_task
sched: Rework CPU hotplug task selection
sched/{rt,deadline}: Fix set_next_task vs pick_next_task
sched: Fix kerneldoc comment for ia64_set_curr_task
...
