kernel_optimize_test/kernel
Eric Dumazet 5517d86bea Speed up divides by cpu_power in scheduler
I noticed expensive divides done in try_to_wakeup() and
find_busiest_group() on a bi dual core Opteron machine (total of 4 cores),
moderatly loaded (15.000 context switch per second)

oprofile numbers :

CPU: AMD64 processors, speed 2600.05 MHz (estimated)
Counted CPU_CLK_UNHALTED events (Cycles outside of halt state) with a unit
mask of 0x00 (No unit mask) count 50000
samples  %        symbol name
...
613914    1.0498  try_to_wake_up
    834  0.0013 :ffffffff80227ae1:   div    %rcx
77513  0.1191 :ffffffff80227ae4:   mov    %rax,%r11

608893    1.0413  find_busiest_group
   1841  0.0031 :ffffffff802260bf:       div    %rdi
140109  0.2394 :ffffffff802260c2:       test   %sil,%sil

Some of these divides can use the reciprocal divides we introduced some
time ago (currently used in slab AFAIK)

We can assume a load will fit in a 32bits number, because with a
SCHED_LOAD_SCALE=128 value, its still a theorical limit of 33554432

When/if we reach this limit one day, probably cpus will have a fast
hardware divide and we can zap the reciprocal divide trick.

Ingo suggested to rename cpu_power to __cpu_power to make clear it should
not be modified without changing its reciprocal value too.

I did not convert the divide in cpu_avg_load_per_task(), because tracking
nr_running changes may be not worth it ?  We could use a static table of 32
reciprocal values but it would add a conditional branch and table lookup.

[akpm@linux-foundation.org: !SMP build fix]
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-05-08 11:15:17 -07:00
..
irq Pad irq_desc to internode cacheline size 2007-05-08 11:15:09 -07:00
power header cleaning: don't include smp_lock.h when not used 2007-05-08 11:15:07 -07:00
time sched: dynticks idle load balancing 2007-05-08 11:15:17 -07:00
.gitignore
acct.c [PATCH] kernel: change uses of f_{dentry, vfsmnt} to use f_path 2006-12-08 08:28:42 -08:00
audit.c audit: add spaces on either side of case "..." operator. 2007-05-08 11:15:09 -07:00
audit.h [PATCH] audit: AUDIT_PERM support 2006-09-11 13:32:30 -04:00
auditfilter.c [PATCH] minor update to rule add/delete messages (ver 2) 2007-02-17 21:30:09 -05:00
auditsc.c [PATCH] fix deadlock in audit_log_task_context() 2007-03-14 15:27:48 -07:00
capability.c [PATCH] pid: replace do/while_each_task_pid with do/while_each_pid_task 2007-02-12 09:48:32 -08:00
compat.c [PATCH] Common compat_sys_sysinfo 2007-02-11 10:51:32 -08:00
configs.c [PATCH] struct seq_operations and struct file_operations constification 2006-12-07 08:39:46 -08:00
cpu.c [PATCH] Fix microcode-related suspend problem 2007-04-02 10:06:09 -07:00
cpuset.c cpusets: allow empty {cpus,mems}_allowed to be set for unpopulated cpuset 2007-05-08 11:15:14 -07:00
delayacct.c KMEM_CACHE(): simplify slab cache creation 2007-05-07 12:12:55 -07:00
die_notifier.c move die notifier handling to common code 2007-05-08 11:15:04 -07:00
dma.c [PATCH] struct seq_operations and struct file_operations constification 2006-12-07 08:39:46 -08:00
exec_domain.c
exit.c header cleaning: don't include smp_lock.h when not used 2007-05-08 11:15:07 -07:00
extable.c
fork.c header cleaning: don't include smp_lock.h when not used 2007-05-08 11:15:07 -07:00
futex_compat.c [PATCH] __user annotations: futex 2006-10-10 15:37:22 -07:00
futex.c futex: restartable futex_wait 2007-05-08 11:15:03 -07:00
hrtimer.c export hrtimer_forward 2007-05-08 11:15:15 -07:00
itimer.c The scheduled -EINVAL for invalid timevals in setitimer 2007-05-08 11:15:13 -07:00
kallsyms.c kallsyms: cleanup: use seq_release_private() where appropriate 2007-05-08 11:15:09 -07:00
Kconfig.hz [PATCH] HZ: 300Hz support 2006-12-07 08:39:36 -08:00
Kconfig.preempt
kexec.c kdump/kexec: calculate note size at compile time 2007-05-08 11:15:07 -07:00
kfifo.c [PATCH] Numerous fixes to kernel-doc info in source files. 2007-02-11 10:51:32 -08:00
kmod.c Fix kevent's childs priority greediness 2007-05-08 11:15:07 -07:00
kprobes.c Simplify kallsyms_lookup() 2007-05-08 11:15:08 -07:00
ksysfs.c remove "struct subsystem" as it is no longer needed 2007-05-02 18:57:59 -07:00
kthread.c [PATCH] Numerous fixes to kernel-doc info in source files. 2007-02-11 10:51:32 -08:00
latency.c [PATCH] severing module.h->sched.h 2006-12-04 02:00:22 -05:00
lockdep_internals.h [PATCH] lockdep: more chains 2006-12-07 08:39:43 -08:00
lockdep_proc.c [PATCH] remove many unneeded #includes of sched.h 2007-02-14 08:09:54 -08:00
lockdep.c lockdep: removed unused ip argument in mark_lock & mark_held_locks 2007-05-08 11:15:13 -07:00
Makefile move die notifier handling to common code 2007-05-08 11:15:04 -07:00
module.c Fix race between cat /proc/slab_allocators and rmmod 2007-05-08 11:15:08 -07:00
mutex-debug.c [PATCH] remove many unneeded #includes of sched.h 2007-02-14 08:09:54 -08:00
mutex-debug.h
mutex.c [PATCH] lockdep: avoid lockdep warning in md 2006-12-08 08:28:39 -08:00
mutex.h [PATCH] lockdep: prove mutex locking correctness 2006-07-03 15:27:04 -07:00
nsproxy.c Merge sys_clone()/sys_unshare() nsproxy and namespace handling 2007-05-08 11:15:00 -07:00
panic.c [PATCH] Add TAINT_USER and ability to set taint flags from userspace 2007-02-11 10:51:29 -08:00
params.c kernel/params.c: fix lying comment for param_array() 2007-05-08 11:15:08 -07:00
pid.c Merge sys_clone()/sys_unshare() nsproxy and namespace handling 2007-05-08 11:15:00 -07:00
posix-cpu-timers.c Introduce a handy list_first_entry macro 2007-05-08 11:15:11 -07:00
posix-timers.c header cleaning: don't include smp_lock.h when not used 2007-05-08 11:15:07 -07:00
printk.c header cleaning: don't include smp_lock.h when not used 2007-05-08 11:15:07 -07:00
profile.c [PATCH] proc: remove useless (and buggy) ->nlink settings 2007-02-11 10:51:32 -08:00
ptrace.c [PATCH] pidspace: is_init() 2006-09-29 09:18:12 -07:00
rcupdate.c [PATCH] rcu: add a prefetch() in rcu_do_batch() 2006-12-07 08:39:40 -08:00
rcutorture.c rcutorture: Mark rcu_torture_init as __init 2007-05-08 11:15:00 -07:00
relay.c [PATCH] kernel-doc fixes for 2.6.20-git15 (non-drivers) 2007-03-01 14:53:37 -08:00
resource.c libata/IDE: remove combined mode quirk 2007-04-28 14:15:59 -04:00
rtmutex_common.h
rtmutex-debug.c Remove all inclusions of <linux/config.h> 2006-10-04 03:38:54 -04:00
rtmutex-debug.h
rtmutex-tester.c [PATCH] Add include/linux/freezer.h and move definitions from sched.h 2006-12-07 08:39:27 -08:00
rtmutex.c [PATCH] hrtimers: namespace and enum cleanup 2007-02-16 08:13:58 -08:00
rtmutex.h
rwsem.c Lockdep treats down_write_trylock like regular down_write 2007-05-08 11:15:09 -07:00
sched.c Speed up divides by cpu_power in scheduler 2007-05-08 11:15:17 -07:00
seccomp.c
signal.c header cleaning: don't include smp_lock.h when not used 2007-05-08 11:15:07 -07:00
softirq.c [PATCH] tick-management: dyntick / highres functionality 2007-02-16 08:13:59 -08:00
softlockup.c add touch_all_softlockup_watchdogs() 2007-05-08 11:15:06 -07:00
spinlock.c [PATCH] lockdep: spin_lock_irqsave_nested() 2006-11-25 13:28:34 -08:00
srcu.c [PATCH] SRCU: report out-of-memory errors 2006-10-04 07:55:30 -07:00
stacktrace.c [PATCH] lockdep: stacktrace subsystem, core 2006-07-03 15:27:02 -07:00
stop_machine.c Use stop_machine_run in the Intel RNG driver 2007-05-08 11:15:00 -07:00
sys_ni.c [PATCH] Create compat_sys_migrate_pages 2006-11-03 12:27:59 -08:00
sys.c CPU time limit patch / setrlimit(RLIMIT_CPU, 0) cheat fix 2007-05-08 11:15:12 -07:00
sysctl.c proc: maps protection 2007-05-08 11:15:02 -07:00
taskstats.c KMEM_CACHE(): simplify slab cache creation 2007-05-07 12:12:55 -07:00
time.c header cleaning: don't include smp_lock.h when not used 2007-05-08 11:15:07 -07:00
timer.c Introduce a handy list_first_entry macro 2007-05-08 11:15:11 -07:00
tsacct.c [PATCH] time: x86_64: split x86_64/kernel/time.c up 2007-02-16 08:14:00 -08:00
uid16.c header cleaning: don't include smp_lock.h when not used 2007-05-08 11:15:07 -07:00
user.c [PATCH] slab: remove kmem_cache_t 2006-12-07 08:39:25 -08:00
utsname_sysctl.c [PATCH] sysctl: remove insert_at_head from register_sysctl 2007-02-14 08:09:59 -08:00
utsname.c Merge sys_clone()/sys_unshare() nsproxy and namespace handling 2007-05-08 11:15:00 -07:00
wait.c [PATCH] uninline init_waitqueue_head() 2006-07-10 13:24:25 -07:00
workqueue.c [PATCH] Add debugging feature /proc/timer_stat 2007-02-16 08:13:59 -08:00