Peter Zijlstra
4b96a29ba8
mm: sched: numa: Implement slow start for working set sampling ...
Add a 1 second delay before starting to scan the working set of
a task and starting to balance it amongst nodes.
[ note that before the constant per task WSS sampling rate patch
the initial scan would happen much later still, in effect that
patch caused this regression. ]
The theory is that short-run tasks benefit very little from NUMA
placement: they come and go, and they better stick to the node
they were started on. As tasks mature and rebalance to other CPUs
and nodes, so does their NUMA placement have to change and so
does it start to matter more and more.
In practice this change fixes an observable kbuild regression:
# [ a perf stat --null --repeat 10 test of ten bzImage builds to /dev/shm ]
!NUMA:
45.291088843 seconds time elapsed ( +- 0.40% )
45.154231752 seconds time elapsed ( +- 0.36% )
+NUMA, no slow start:
46.172308123 seconds time elapsed ( +- 0.30% )
46.343168745 seconds time elapsed ( +- 0.25% )
+NUMA, 1 sec slow start:
45.224189155 seconds time elapsed ( +- 0.25% )
45.160866532 seconds time elapsed ( +- 0.17% )
and it also fixes an observable perf bench (hackbench) regression:
# perf stat --null --repeat 10 perf bench sched messaging
-NUMA:
-NUMA: 0.246225691 seconds time elapsed ( +- 1.31% )
+NUMA no slow start: 0.252620063 seconds time elapsed ( +- 1.13% )
+NUMA 1sec delay: 0.248076230 seconds time elapsed ( +- 1.35% )
The implementation is simple and straightforward, most of the patch
deals with adding the /proc/sys/kernel/numa_balancing_scan_delay_ms tunable
knob.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
[ Wrote the changelog, ran measurements, tuned the default. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Rik van Riel <riel@redhat.com>
2012-12-11 14:42:47 +00:00
KGDB/KDB fixes and cleanups
2012-10-13 11:16:58 +09:00
Merge branch 'uprobes/core' of git://git.kernel.org/pub/scm/linux/kernel/git/oleg/misc into perf/urgent
2012-10-21 18:18:17 +02:00
irqdomain: augment add_simple() to allocate descs
2012-10-10 08:57:26 +02:00
Merge branch 'pm-qos'
2012-09-17 20:25:51 +02:00
mm: sched: numa: Implement slow start for working set sampling
2012-12-11 14:42:47 +00:00
Merge branch 'timers-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
2012-10-12 22:17:48 +09:00
Merge branch 'tip/perf/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace into perf/urgent
2012-10-21 19:53:34 +02:00
vfs: make path_openat take a struct filename pointer
2012-10-12 20:15:09 -04:00
[SCSI] async: make async_synchronize_full() flush all work regardless of domain
2012-07-20 09:07:37 +01:00
fs: handle failed audit_log_start properly
2012-10-09 23:33:37 -04:00
audit: optimize audit_compare_dname_path
2012-10-12 00:32:02 -04:00
audit: clean up refcounting in audit-tree
2012-08-15 12:55:22 +02:00
audit: optimize audit_compare_dname_path
2012-10-12 00:32:02 -04:00
audit: optimize audit_compare_dname_path
2012-10-12 00:32:02 -04:00
audit: make audit_inode take struct filename
2012-10-12 20:15:09 -04:00
userns: Teach inode_capable to understand inodes whose uids map to other namespaces.
2012-05-15 14:59:24 -07:00
Revert "cgroup: Remove task_lock() from cgroup_post_fork()"
2012-10-19 14:09:35 -07:00
cgroup: mark subsystems with broken hierarchy support and whine if cgroups are nested for them
2012-09-14 12:01:16 -07:00
new helper: sigsuspend()
2012-05-21 23:52:30 -04:00
CPU hotplug, debug: detect imbalance between get_online_cpus() and put_online_cpus()
2012-10-09 16:22:15 +09:00
kernel/cpu_pm.c: fix various typos
2012-05-31 17:49:27 -07:00
cpusets: Remove/update outdated comments
2012-07-24 13:53:28 +02:00
Merge branch 'modsplit-Oct31_2011' of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/linux
2011-11-06 19:44:47 -08:00
userns: Make credential debugging user namespace safe.
2012-08-23 22:54:18 -07:00
Remove all #inclusions of asm/system.h
2012-03-28 18:30:03 +01:00
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
2012-10-02 20:25:04 -07:00
extable: Skip sorting if sorted at build time.
2012-04-19 15:06:55 -07:00
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/signal
2012-10-10 12:02:25 +09:00
PM / Freezer: Remove references to TIF_FREEZE in comments
2012-03-04 23:08:54 +01:00
futex: Handle futex_pi OWNER_DIED take over correctly
2012-11-01 12:06:54 +01:00
futex: Mark get_robust_list as deprecated
2012-03-29 11:37:17 +02:00
userns: Convert in_group_p and in_egroup_p to use kgid_t
2012-05-03 03:29:33 -07:00
hrtimer: Update hrtimer base offsets each hrtimer_interrupt
2012-07-11 23:34:39 +02:00
hung task debugging: Inject NMI when hung and going to panic
2012-04-25 12:39:25 +02:00
irq_work: fix compile failure on tile from missing include
2012-04-13 13:15:16 -04:00
itimer: Use printk_once instead of WARN_ONCE
2012-04-10 11:00:30 +02:00
jump_label: Export jump_label_rate_limit()
2012-08-06 19:00:35 +03:00
vsprintf: fix %ps on non symbols when using kallsyms
2012-05-29 16:22:32 -07:00
syscalls, x86: add __NR_kcmp syscall
2012-05-31 17:49:32 -07:00
locking: Adjust spin lock inlining Kconfig options
2012-09-13 17:56:13 +02:00
locking/kconfig: Simplify INLINE_SPIN_UNLOCK usage
2012-03-23 13:18:57 +01:00
kdump: remove unneeded include
2012-10-06 03:05:19 +09:00
[media] kernel:kfifo: export __kfifo_max_r symbol
2012-04-11 18:24:37 -03:00
infrastructure for saner ret_from_kernel_thread semantics
2012-10-12 13:35:07 -04:00
kprobes/x86: Fix to support jprobes on ftrace-based kprobe
2012-09-13 22:52:11 -04:00
kernel: ksysfs.c is implicitly using stat.h
2011-10-31 09:20:13 -04:00
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/signal
2012-10-13 10:05:52 +09:00
kernel: Map most files to use export.h instead of module.h
2011-10-31 09:20:12 -04:00
brlocks/lglocks: turn into functions
2012-05-29 23:28:41 -04:00
lockdep: Check if nested lock is actually held
2012-09-13 17:00:44 +02:00
kernel: Map most files to use export.h instead of module.h
2011-10-31 09:20:12 -04:00
Makefile: Documentation for external tool should be correct
2012-10-25 16:00:53 -07:00
MODSIGN: Provide module signing public keys to the kernel
2012-10-10 20:01:22 +10:30
MODSIGN: Move the magic string to the end of a module and eliminate the search
2012-10-19 17:30:40 -07:00
module: fix out-by-one error in kallsyms
2012-10-31 13:56:37 +10:30
module_signing: fix printk format warning
2012-10-22 08:56:34 +03:00
kernel: Map most files to use export.h instead of module.h
2011-10-31 09:20:12 -04:00
sched/rt: Use schedule_preempt_disabled()
2012-03-01 10:28:03 +01:00
kernel: Map most files to use export.h instead of module.h
2011-10-31 09:20:12 -04:00
kernel: Map most files to use export.h instead of module.h
2011-10-31 09:20:12 -04:00
padata: Fix cpu hotplug
2012-03-29 19:52:46 +08:00
panic: fix a possible deadlock in panic()
2012-07-30 17:25:13 -07:00
params: replace printk(KERN_<LVL>...) with pr_<lvl>(...)
2012-05-04 17:28:18 -07:00
net ip6 flowlabel: Make owner a union of struct pid * and kuid_t
2012-08-14 21:49:25 -07:00
pidns: limit the nesting depth of pid namespaces
2012-10-25 14:37:53 -07:00
[S390] cputime: add sparse checking and cleanup
2011-12-15 14:56:19 +01:00
kernel: Map most files to use export.h instead of module.h
2011-10-31 09:20:12 -04:00
printk: Fix scheduling-while-atomic problem in console_cpu_notify()
2012-10-16 18:17:44 -07:00
kernel: Map most files to use export.h instead of module.h
2011-10-31 09:20:12 -04:00
ptrace: mark __ptrace_may_access() static
2012-08-03 14:47:17 +10:00
range: fix bogus misuse of module.h to get printk()
2011-10-31 09:20:11 -04:00
rcu: Allow nesting of rcu_idle_enter() and rcu_idle_exit()
2012-02-21 09:06:12 -08:00
rcu: Add PROVE_RCU_DELAY to provoke difficult races
2012-09-23 07:42:49 -07:00
rcu: Move TINY_RCU quiescent state out of extended quiescent state
2012-09-23 07:42:52 -07:00
rcu: Move TINY_PREEMPT_RCU away from raw_local_irq_save()
2012-09-23 07:42:51 -07:00
rcu: Prevent initialization race in rcutorture kthreads
2012-09-23 07:42:23 -07:00
rcu: Grace-period initialization excludes only RCU notifier
2012-10-08 09:06:38 -07:00
rcu: Grace-period initialization excludes only RCU notifier
2012-10-08 09:06:38 -07:00
rcu: Make RCU_FAST_NO_HZ handle adaptive ticks
2012-09-26 15:44:02 +02:00
Merge remote-tracking branch 'tip/smp/hotplug' into next.2012.09.25b
2012-09-25 10:01:45 -07:00
splice: fix racy pipe->buffers uses
2012-06-13 21:16:42 +02:00
rescounters: add res_counter_uncharge_until()
2012-05-29 16:22:27 -07:00
kernel/resource.c: fix stack overflow in __reserve_region_with_split()
2012-10-06 03:05:31 +09:00
lockdep, rtmutex, bug: Show taint flags on error
2011-12-06 08:16:49 +01:00
rtmutex-tester: convert sysdev_class to a regular subsystem
2011-12-14 14:54:22 -08:00
Revert "rcu: Permit rt_mutex_unlock() with irqs disabled"
2011-12-11 10:33:18 -08:00
Remove all #inclusions of asm/system.h
2012-03-28 18:30:03 +01:00
seccomp: fix build warnings when there is no CONFIG_SECCOMP_FILTER
2012-04-18 12:24:52 +10:00
semaphore: fix improper comment reference to mutex
2012-04-05 17:15:55 -07:00
coredump: pass siginfo_t* to do_coredump() and below, not merely signr
2012-10-06 03:05:16 +09:00
smp: Remove ipi_call_lock[_irq]()/ipi_call_unlock[_irq]()
2012-06-05 17:27:14 +02:00
hotplug: Fix UP bug in smpboot hotplug code
2012-08-13 17:01:07 +02:00
smpboot: Provide infrastructure for percpu hotplug threads
2012-08-13 17:01:07 +02:00
Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
2012-10-01 10:43:39 -07:00
locking/kconfig: Simplify INLINE_SPIN_UNLOCK usage
2012-03-23 13:18:57 +01:00
workqueue: deprecate system_nrt[_freezable]_wq
2012-08-20 14:51:24 -07:00
kernel: Map most files to use export.h instead of module.h
2011-10-31 09:20:12 -04:00
Merge branch 'modsplit-Oct31_2011' of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/linux
2011-11-06 19:44:47 -08:00
use clamp_t in UNAME26 fix
2012-10-19 18:51:17 -07:00
syscalls, x86: add __NR_kcmp syscall
2012-05-31 17:49:32 -07:00
mm: sched: numa: Implement slow start for working set sampling
2012-12-11 14:42:47 +00:00
mm: prepare for removal of obsolete /proc/sys/vm/nr_pdflush_threads
2012-07-31 18:42:40 -07:00
task_work: task_work_add() should not succeed after exit_task_work()
2012-09-13 16:47:34 +02:00
taskstats: cgroupstats_user_cmd() may leak on error
2012-10-06 03:05:31 +09:00
time: Move update_vsyscall definitions to timekeeper_internal.h
2012-09-24 12:38:06 -04:00
timers: Fix endless looping between cascade() and internal_add_timer()
2012-10-09 21:27:14 +02:00
static keys: Introduce 'struct static_key', static_key_true()/false() and static_key_slow_[inc|dec]()
2012-02-24 10:05:59 +01:00
userns: Convert taskstats to handle the user and pid namespaces.
2012-09-18 01:01:32 -07:00
userns: Convert setting and getting uid and gid system calls to use kuid and kgid
2012-05-03 03:28:41 -07:00
kernel: Map most files to use export.h instead of module.h
2011-10-31 09:20:12 -04:00
kernel: Map most files to use export.h instead of module.h
2011-10-31 09:20:12 -04:00
userns: Add kprojid_t and associated infrastructure in projid.h
2012-09-18 01:01:37 -07:00
userns: Add kprojid_t and associated infrastructure in projid.h
2012-09-18 01:01:37 -07:00
userns: Use cred->user_ns instead of cred->user->user_ns
2012-04-07 16:55:51 -07:00
Merge branch 'modsplit-Oct31_2011' of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/linux
2011-11-06 19:44:47 -08:00
lockdep/waitqueues: Add better annotation
2011-12-21 10:07:39 +01:00
watchdog: Use hotplug thread infrastructure
2012-08-13 17:01:07 +02:00
workqueue: cancel_delayed_work() should return %false if work item is idle
2012-10-24 12:38:16 -07:00
4b96a29ba8