Rik van Riel
7e2703e609
sched/numa: Normalize faults_cpu stats and weigh by CPU use ...
Tracing the code that decides the active nodes has made it abundantly clear
that the naive implementation of the faults_from code has issues.
Specifically, the garbage collector in some workloads will access orders
of magnitudes more memory than the threads that do all the active work.
This resulted in the node with the garbage collector being marked the only
active node in the group.
This issue is avoided if we weigh the statistics by CPU use of each task in
the numa group, instead of by how many faults each thread has occurred.
To achieve this, we normalize the number of faults to the fraction of faults
that occurred on each node, and then multiply that fraction by the fraction
of CPU time the task has used since the last time task_numa_placement was
invoked.
This way the nodes in the active node mask will be the ones where the tasks
from the numa group are most actively running, and the influence of eg. the
garbage collector and other do-little threads is properly minimized.
On a 4 node system, using CPU use statistics calculated over a longer interval
results in about 1% fewer page migrations with two 32-warehouse specjbb runs
on a 4 node system, and about 5% fewer page migrations, as well as 1% better
throughput, with two 8-warehouse specjbb runs, as compared with the shorter
term statistics kept by the scheduler.
Signed-off-by: Rik van Riel <riel@redhat.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Chegu Vinod <chegu_vinod@hp.com>
Link: http://lkml.kernel.org/r/1390860228-21539-7-git-send-email-riel@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2014-01-28 15:03:10 +01:00
sched/preempt: Fix up missed PREEMPT_NEED_RESCHED folding
2014-01-13 17:38:55 +01:00
kdb: Add support for external NMI handler to call KGDB/KDB
2013-10-03 18:47:54 +02:00
Merge branch 'perf/urgent' into perf/core
2014-01-16 09:33:30 +01:00
gcov: reuse kbasename helper
2013-11-13 12:09:34 +09:00
Merge branch 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
2013-12-02 10:15:39 -08:00
Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
2014-01-20 10:42:08 -08:00
PM / sleep: Fix memory leak in pm_vt_switch_unregister().
2013-12-22 00:56:35 +01:00
printk.c: comments should refer to /proc/vmcore instead of /proc/vmcoreinfo
2013-11-13 12:09:14 +09:00
Merge branch 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
2014-01-20 10:25:12 -08:00
sched/numa: Normalize faults_cpu stats and weigh by CPU use
2014-01-28 15:03:10 +01:00
Merge branch 'timers-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
2014-01-20 11:34:26 -08:00
sched/clock, x86: Use a static_key for sched_clock_stable
2014-01-13 15:13:13 +01:00
Ignore generated file kernel/x509_certificate_list
2013-12-10 18:21:34 +00:00
Merge git://git.infradead.org/users/eparis/audit
2013-11-21 19:18:14 -08:00
audit: call audit_bprm() only once to add AUDIT_EXECVE information
2013-11-05 11:15:03 -05:00
audit: do not reject all AUDIT_INODE filter types
2013-11-05 11:09:16 -05:00
audit: fix type of sessionid in audit_set_loginuid()
2013-11-06 11:47:24 -05:00
mm: do not allocate page->ptl dynamically, if spinlock_t fits to long
2013-12-20 12:25:45 -08:00
xfs: update for v3.12-rc1
2013-09-09 11:19:09 -07:00
cgroup: don't recycle cgroup id until all csses' have been destroyed
2013-12-17 08:11:52 -05:00
cgroup: make css_for_each_descendant() and friends include the origin css in the iteration
2013-08-08 20:11:27 -04:00
context_tracking: Wrap static key check into more intuitive function name
2013-12-02 20:43:14 +01:00
Merge branch 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
2013-11-14 16:55:11 +09:00
cpuset: Fix memory allocator deadlock
2013-11-27 13:52:47 -05:00
kernel/delayacct.c: remove redundant checking in __delayacct_add_tsk()
2013-11-13 12:09:12 +09:00
switch elf_core_write_extra_phdrs() to dump_emit()
2013-11-09 00:16:23 -05:00
ptrace: revert "Prepare to fix racy accesses on task breakpoints"
2013-07-09 10:33:26 -07:00
kernel/extable: fix address-checks for core_kernel and init areas
2013-11-28 09:49:41 -08:00
Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
2014-01-20 10:42:08 -08:00
libata, freezer: avoid block device removal while system is frozen
2013-12-19 13:50:32 -05:00
Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
2014-01-20 10:42:08 -08:00
userns: Kill nsown_capable it makes the wrong thing easy
2013-08-30 23:44:11 -07:00
sched/deadline: Add SCHED_DEADLINE structures & implementation
2014-01-13 13:41:06 +01:00
Here are the 3.13 KVM changes. There was a lot of work on the PPC
2013-11-15 13:51:36 +09:00
static_key: WARN on usage before jump_label_init was called
2013-10-19 19:45:35 -04:00
kernel: remove CONFIG_USE_GENERIC_SMP_HELPERS
2013-11-15 09:32:22 +09:00
kexec: migrate to reboot cpu
2013-12-18 19:04:50 -08:00
kernel/kmod.c: check for NULL in call_usermodehelper_exec()
2013-09-30 14:31:02 -07:00
kprobes: use KSYM_NAME_LEN to size identifier buffers
2013-11-13 12:09:26 +09:00
kernel: replace strict_strto*() with kstrto*()
2013-09-12 15:38:03 -07:00
kthread: make kthread_create() killable
2013-11-13 12:08:59 +09:00
KEYS: Remove files generated when SYSTEM_TRUSTED_KEYRING=y
2013-12-13 15:59:11 +00:00
KEYS: Separate the kernel signature checking keyring from module signing
2013-09-25 17:17:01 +01:00
Mainly boring here, too. rmmod --wait finally removed, though.
2013-11-15 13:27:50 +09:00
keys: change asymmetric keys to use common hash definitions
2013-10-25 17:15:18 -04:00
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace
2013-09-07 14:35:32 -07:00
padata: make the sequence counter an atomic_t
2013-10-30 12:02:58 +08:00
panic: Make panic_timeout configurable
2013-11-26 12:12:26 +01:00
kernel/params: fix handling of signed integer types
2013-09-28 12:35:52 -07:00
pidns: fix free_pid() to handle the first fork failure
2013-09-30 14:31:03 -07:00
pid_namespace: make freeing struct pid_namespace rcu-delayed
2013-10-24 23:43:29 -04:00
posix-timers: Convert abuses of BUG_ON to WARN_ON
2013-12-09 16:56:29 +01:00
kernel: delete __cpuinit usage from all core kernel files
2013-07-14 19:36:59 -04:00
exec/ptrace: fix get_dumpable() incorrect tests
2013-11-13 12:09:33 +09:00
kexec: migrate to reboot cpu
2013-12-18 19:04:50 -08:00
kernel: delete __cpuinit usage from all core kernel files
2013-07-14 19:36:59 -04:00
memcg: reduce function dereference
2013-09-12 15:38:02 -07:00
constify copy_siginfo_to_user{,32}()
2013-11-09 00:16:29 -05:00
kernel: fix generic_exec_single indentation
2013-11-15 09:32:22 +09:00
kernel: delete __cpuinit usage from all core kernel files
2013-07-14 19:36:59 -04:00
Merge branch 'timers-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
2014-01-20 11:34:26 -08:00
stop_machine: Fix race between stop_two_cpus() and stop_cpus()
2013-11-11 12:43:38 +01:00
kernel/sys.c: remove obsolete #include <linux/kexec.h>
2013-11-13 12:09:13 +09:00
sched/numa, mm: Remove p->numa_migrate_deferred
2014-01-28 13:17:04 +01:00
kernel/sysctl_binary.c: use scnprintf() instead of snprintf()
2013-11-13 12:09:33 +09:00
KEYS: correct alignment of system_certificate_list content in assembly file
2013-12-10 18:25:28 +00:00
KEYS: correct alignment of system_certificate_list content in assembly file
2013-12-10 18:25:28 +00:00
task_work: documentation
2013-09-11 15:58:27 -07:00
genetlink: only pass array to genl_register_family_with_ops()
2013-11-19 16:39:05 -05:00
timer: Convert kmalloc_node(...GFP_ZERO...) to kzalloc_node(...)
2013-11-19 14:59:50 +01:00
userns: Kill nsown_capable it makes the wrong thing easy
2013-08-30 23:44:11 -07:00
kernel: provide a __smp_call_function_single stub for !CONFIG_SMP
2013-11-15 09:32:22 +09:00
KEYS: fix uninitialized persistent_keyring_register_sem
2013-12-13 15:59:11 +00:00
KEYS: Add per-user_namespace registers for persistent per-UID kerberos caches
2013-09-24 10:35:19 +01:00
userns: Kill nsown_capable it makes the wrong thing easy
2013-08-30 23:44:11 -07:00
watchdog: update watchdog_thresh properly
2013-09-24 17:00:25 -07:00
PCI updates for v3.13:
2013-12-15 11:45:27 -08:00
7e2703e609