android_kernel_xiaomi_sm8350/include
Ingo Molnar 9fcd18c9e6 sched: re-tune balancing
Impact: improve wakeup affinity on NUMA systems, tweak SMP systems

Given the fixes+tweaks to the wakeup-buddy code, re-tweak the domain
balancing defaults on NUMA and SMP systems.

Turn on SD_WAKE_AFFINE which was off on x86 NUMA - there's no reason
why we would not want to have wakeup affinity across nodes as well.
(we already do this in the standard NUMA template.)

lat_ctx on a NUMA box is particularly happy about this change:

before:

 |   phoenix:~/l> ./lat_ctx -s 0 2
 |   "size=0k ovr=2.60
 |   2 5.70

after:

 |   phoenix:~/l> ./lat_ctx -s 0 2
 |   "size=0k ovr=2.65
 |   2 2.07

a 2.75x speedup.

pipe-test is similarly happy about it too:

 |  phoenix:~/sched-tests> ./pipe-test
 |   18.26 usecs/loop.
 |   14.70 usecs/loop.
 |   14.38 usecs/loop.
 |   10.55 usecs/loop.              # +WAKE_AFFINE on domain0+domain1
 |   8.63 usecs/loop.
 |   8.59 usecs/loop.
 |   9.03 usecs/loop.
 |   8.94 usecs/loop.
 |   8.96 usecs/loop.
 |   8.63 usecs/loop.

Also:

 - disable SD_BALANCE_NEWIDLE on NUMA and SMP domains (keep it for siblings)
 - enable SD_WAKE_BALANCE on SMP domains

Sysbench+postgresql improves all around the board, quite significantly:

           .28-rc3-11474e2c  .28-rc3-11474e2c-tune
-------------------------------------------------
    1:             571              688    +17.08%
    2:            1236             1206    -2.55%
    4:            2381             2642    +9.89%
    8:            4958             5164    +3.99%
   16:            9580             9574    -0.07%
   32:            7128             8118    +12.20%
   64:            7342             8266    +11.18%
  128:            7342             8064    +8.95%
  256:            7519             7884    +4.62%
  512:            7350             7731    +4.93%
-------------------------------------------------
  SUM:           55412            59341    +6.62%

So it's a win both for the runup portion, the peak area and the tail.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-11-05 18:04:38 +01:00
..
acpi Merge branch 'linus' into test 2008-10-23 00:11:07 -04:00
asm-arm
asm-frv ide: fix support for IDE PCI controllers using MMIO on frv 2008-10-17 18:09:14 +02:00
asm-generic asm-generic: define DIE_OOPS in asm-generic 2008-10-27 11:39:03 +01:00
asm-h8300 h8300: update timer handler - new files 2008-10-16 11:21:29 -07:00
asm-m32r [PATCH] remove unused ibcs2/PER_SVR4 in SET_PERSONALITY 2008-10-16 15:40:05 +02:00
asm-m68k proc: move /proc/hardware to m68k-specific code 2008-10-23 14:24:03 +04:00
asm-mn10300 [PATCH] remove unused ibcs2/PER_SVR4 in SET_PERSONALITY 2008-10-16 15:40:05 +02:00
asm-x86 x86: add iomap_atomic*()/iounmap_atomic() on 32-bit using fixmaps 2008-10-31 10:12:38 +01:00
asm-xtensa Merge git://git.kernel.org/pub/scm/linux/kernel/git/czankel/xtensa-2.6 2008-10-23 09:16:56 -07:00
crypto
drm i915: Add GEM ioctl to get available aperture size. 2008-11-03 10:56:49 +10:00
keys
linux sched: re-tune balancing 2008-11-05 18:04:38 +01:00
math-emu math-emu: Fix thinko in _FP_DIV 2008-10-22 22:09:59 -07:00
media V4L/DVB (9335): videobuf: split unregister bus creating self-contained frontend de-allocator 2008-10-21 14:32:08 -02:00
mtd
net netns: add register_pernet_gen_subsys/unregister_pernet_gen_subsys 2008-10-30 23:55:16 -07:00
pcmcia
rdma
rxrpc
scsi scsi: make sure that scsi_init_shared_tag_map() doesn't overwrite existing map 2008-10-27 19:25:30 +01:00
sound Merge branches 'topic/fix/misc' and 'topic/fix/asoc' into for-linus 2008-10-30 14:26:24 +01:00
trace
video video: add sh_mobile_lcdc platform flags 2008-10-20 11:38:50 +09:00
xen
Kbuild