linux-stable/kernel
Suresh Siddha aae6d3ddd8 sched: Use group weight, idle cpu metrics to fix imbalances during idle
Currently we consider a sched domain to be well balanced when the imbalance
is less than the domain's imablance_pct. As the number of cores and threads
are increasing, current values of imbalance_pct (for example 25% for a
NUMA domain) are not enough to detect imbalances like:

a) On a WSM-EP system (two sockets, each having 6 cores and 12 logical threads),
24 cpu-hogging tasks get scheduled as 13 on one socket and 11 on another
socket. Leading to an idle HT cpu.

b) On a hypothetial 2 socket NHM-EX system (each socket having 8 cores and
16 logical threads), 16 cpu-hogging tasks can get scheduled as 9 on one
socket and 7 on another socket. Leaving one core in a socket idle
whereas in another socket we have a core having both its HT siblings busy.

While this issue can be fixed by decreasing the domain's imbalance_pct
(by making it a function of number of logical cpus in the domain), it
can potentially cause more task migrations across sched groups in an
overloaded case.

Fix this by using imbalance_pct only during newly_idle and busy
load balancing. And during idle load balancing, check if there
is an imbalance in number of idle cpu's across the busiest and this
sched_group or if the busiest group has more tasks than its weight that
the idle cpu in this_group can pull.

Reported-by: Nikhil Rao <ncrao@google.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1284760952.2676.11.camel@sbsiddha-MOBL3.sc.intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-11-10 23:13:56 +01:00
..
debug
gcov
irq Merge branches 'irq-core-for-linus' and 'core-locking-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip 2010-10-31 20:40:24 -04:00
power
time
trace
.gitignore
acct.c
async.c
audit.c audit: Use rcu for task lookup protection 2010-10-30 08:45:42 -04:00
audit.h audit: make functions static 2010-10-30 01:42:19 -04:00
audit_tree.c in untag_chunk() we need to do alloc_chunk() a bit earlier 2010-10-30 02:18:32 -04:00
audit_watch.c audit: make functions static 2010-10-30 01:42:19 -04:00
auditfilter.c Audit: add support to match lsm labels on user audit messages 2010-10-30 01:41:57 -04:00
auditsc.c audit mmap 2010-10-30 08:45:43 -04:00
backtracetest.c
bounds.c
capability.c
cgroup.c
cgroup_freezer.c
compat.c
configs.c
cpu.c
cpuset.c
cred.c
delayacct.c
dma.c
elfcore.c
exec_domain.c
exit.c posix-cpu-timers: workaround to suppress the problems with mt exec 2010-11-05 14:16:03 -07:00
extable.c
fork.c
freezer.c
futex.c
futex_compat.c
groups.c
hrtimer.c
hung_task.c
hw_breakpoint.c
irq_work.c
itimer.c
jump_label.c
kallsyms.c
Kconfig.freezer
Kconfig.hz
Kconfig.locks
Kconfig.preempt
kexec.c
kfifo.c
kmod.c
kprobes.c
ksysfs.c
kthread.c
latencytop.c
lockdep.c
lockdep_internals.h
lockdep_proc.c
lockdep_states.h
Makefile
module.c
mutex-debug.c
mutex-debug.h
mutex.c
mutex.h
notifier.c
ns_cgroup.c
nsproxy.c
padata.c
panic.c
params.c
perf_event.c
pid.c
pid_namespace.c
pm_qos_params.c
posix-cpu-timers.c
posix-timers.c
printk.c
profile.c
ptrace.c
range.c
rcupdate.c
rcutiny.c
rcutiny_plugin.h
rcutorture.c
rcutree.c
rcutree.h
rcutree_plugin.h
rcutree_trace.c
relay.c Clean up relay_alloc_page_array() slightly by using vzalloc rather than vmalloc and memset 2010-11-05 08:21:34 -07:00
res_counter.c
resource.c
rtmutex-debug.c
rtmutex-debug.h
rtmutex-tester.c
rtmutex.c
rtmutex.h
rtmutex_common.h
rwsem.c
sched.c sched: Use group weight, idle cpu metrics to fix imbalances during idle 2010-11-10 23:13:56 +01:00
sched_clock.c
sched_cpupri.c
sched_cpupri.h
sched_debug.c
sched_fair.c sched: Use group weight, idle cpu metrics to fix imbalances during idle 2010-11-10 23:13:56 +01:00
sched_features.h
sched_idletask.c
sched_rt.c
sched_stats.h
sched_stoptask.c
seccomp.c
semaphore.c
signal.c
smp.c
softirq.c
spinlock.c
srcu.c
stacktrace.c
stop_machine.c
sys.c
sys_ni.c
sysctl.c
sysctl_binary.c
sysctl_check.c
taskstats.c
test_kprobes.c
time.c
timeconst.pl
timer.c
tracepoint.c
tsacct.c
uid16.c
up.c
user-return-notifier.c
user.c
user_namespace.c
utsname.c
utsname_sysctl.c
wait.c
watchdog.c watchdog: Fix section mismatch and potential undefined behavior. 2010-11-05 17:45:35 -07:00
workqueue.c
workqueue_sched.h