linux/Documentation
Shaohua Li 09223371de rcu: Use softirq to address performance regression
Commit a26ac2455ffcf3(rcu: move TREE_RCU from softirq to kthread)
introduced performance regression. In an AIM7 test, this commit degraded
performance by about 40%.

The commit runs rcu callbacks in a kthread instead of softirq. We observed
high rate of context switch which is caused by this. Out test system has
64 CPUs and HZ is 1000, so we saw more than 64k context switch per second
which is caused by RCU's per-CPU kthread.  A trace showed that most of
the time the RCU per-CPU kthread doesn't actually handle any callbacks,
but instead just does a very small amount of work handling grace periods.
This means that RCU's per-CPU kthreads are making the scheduler do quite
a bit of work in order to allow a very small amount of RCU-related
processing to be done.

Alex Shi's analysis determined that this slowdown is due to lock
contention within the scheduler.  Unfortunately, as Peter Zijlstra points
out, the scheduler's real-time semantics require global action, which
means that this contention is inherent in real-time scheduling.  (Yes,
perhaps someone will come up with a workaround -- otherwise, -rt is not
going to do well on large SMP systems -- but this patch will work around
this issue in the meantime.  And "the meantime" might well be forever.)

This patch therefore re-introduces softirq processing to RCU, but only
for core RCU work.  RCU callbacks are still executed in kthread context,
so that only a small amount of RCU work runs in softirq context in the
common case.  This should minimize ksoftirqd execution, allowing us to
skip boosting of ksoftirqd for CONFIG_RCU_BOOST=y kernels.

Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Tested-by: "Alex,Shi" <alex.shi@intel.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-06-14 15:25:39 -07:00
..
ABI Merge branch 'linux-next' of git://git.kernel.org/pub/scm/linux/kernel/git/jlbec/ocfs2 2011-05-26 10:55:15 -07:00
accounting getdelays: show average CPU/IO/SWAP/RECLAIM delays 2011-05-26 17:12:34 -07:00
acpi ACPI: Split out custom_method functionality into an own driver 2011-05-29 01:50:40 -04:00
aoe Documentation: update broken web addresses. 2010-08-04 15:21:40 +02:00
arm Merge branch 'for-rmk' of git://git.kernel.org/pub/scm/linux/kernel/git/kgene/linux-samsung into devel-stable 2011-05-25 21:47:48 +01:00
auxdisplay includecheck fix: Documentation, cfag12864b-example.c 2009-09-24 07:20:57 -07:00
blackfin Blackfin: document SPI CS limitations with CPHA=0 2010-08-06 12:55:52 -04:00
block Fix common misspellings 2011-03-31 11:26:23 -03:00
blockdev cciss: add cciss_tape_cmds module paramter 2011-05-06 08:23:59 -06:00
cdrom Documentation: update broken web addresses. 2010-08-04 15:21:40 +02:00
cgroups cgroup: remove the ns_cgroup 2011-05-26 17:12:34 -07:00
connector Documentation/: it's -> its where appropriate 2010-04-23 02:09:52 +02:00
console doc: fix console doc typo 2010-02-24 13:51:32 +01:00
cpu-freq [CPUFREQ] Add documentation for sampling_down_factor 2011-03-16 17:54:31 -04:00
cpuidle
cris
crypto
development-process docs: update the development process document 2011-03-25 14:30:31 -06:00
device-mapper Fix common misspellings 2011-03-31 11:26:23 -03:00
devicetree Merge branches 'devel', 'devel-stable' and 'fixes' into for-linus 2011-05-27 22:59:57 +01:00
DocBook Merge git://git.infradead.org/mtd-2.6 2011-05-27 20:06:53 -07:00
driver-model driver core: remove the driver-model structures from the documentation 2011-05-06 18:01:05 -07:00
dvb Merge branch 'for-linus2' of git://git.profusion.mobi/users/lucas/linux-2.6 2011-04-07 11:14:49 -07:00
early-userspace
fault-injection lkdtm: add debugfs access and loosen KPROBE ties 2010-03-06 11:26:32 -08:00
fb Fix common misspellings 2011-03-31 11:26:23 -03:00
filesystems rcu: Use softirq to address performance regression 2011-06-14 15:25:39 -07:00
firmware_class firmware: Update hotplug script 2010-08-05 13:53:34 -07:00
frv
hid HID: Move hiddev.txt to the new Documentation/hid directory 2011-03-22 11:43:51 +01:00
hwmon hwmon: New driver for the SMSC EMC6W201 2011-05-25 20:43:33 +02:00
i2c i2c-i801: SMBus patch for Intel Panther Point DeviceIDs 2011-05-24 20:58:49 +02:00
i2o Fix common misspellings 2011-03-31 11:26:23 -03:00
ia64 Fix common misspellings 2011-03-31 11:26:23 -03:00
ide
infiniband Documentation: update broken web addresses. 2010-08-04 15:21:40 +02:00
input Input: elantech - describe further the protocol 2011-05-16 22:48:24 -07:00
ioctl Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/cjb/mmc 2011-05-25 16:55:55 -07:00
isdn Fix common misspellings 2011-03-31 11:26:23 -03:00
ja_JP HOWTO: sync up Documentaion/ja_JP/HOWTO 2011-04-19 16:53:45 -07:00
kbuild Merge branch 'kconfig-for-40' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild-2.6 2011-05-25 16:54:01 -07:00
kdump kdump: update kexec-tools URL and Vivek's email 2010-11-25 14:36:38 +01:00
ko_KR Docs/Kconfig: Update: osdl.org -> linuxfoundation.org 2010-11-15 23:50:13 +01:00
laptops acer-wmi: Delete out-of-date documentation 2011-05-27 12:36:46 -04:00
leds Documentation: consolidate leds files to leds/ subdir 2011-04-04 17:51:47 -07:00
m68k
make kbuild: introduce HDR_ARCH_LIST for headers_install_all 2010-12-14 22:16:19 +01:00
mips Fix common misspellings 2011-03-31 11:26:23 -03:00
misc-devices Fix common misspellings 2011-03-31 11:26:23 -03:00
mmc mmc: MMC boot partitions support. 2011-05-24 21:01:21 -04:00
mn10300
mtd Documentation: update broken web addresses. 2010-08-04 15:21:40 +02:00
namespaces
netlabel Documentation/: it's -> its where appropriate 2010-04-23 02:09:52 +02:00
networking Merge branch 'docs-move' of git://git.kernel.org/pub/scm/linux/kernel/git/rdunlap/linux-docs 2011-05-27 10:25:02 -07:00
nfc NFC: Driver for NXP Semiconductors PN544 NFC chip. 2011-01-13 08:03:19 -08:00
parisc
PCI Fix common misspellings 2011-03-31 11:26:23 -03:00
pcmcia pcmcia: use autoconfiguration feature for ioports and iomem 2010-09-29 17:20:24 +02:00
power regulator: Remove supply_regulator_dev from machine configuration 2011-05-27 10:34:37 +01:00
powerpc Fix common misspellings 2011-03-31 11:26:23 -03:00
pps pps: add parallel port PPS signal generator 2011-01-13 08:03:21 -08:00
prctl
pti Kernel documentation for the PTI feature. 2011-05-13 16:31:00 -07:00
ptp ptp: Added a brand new class driver for ptp clocks. 2011-05-23 13:01:00 -07:00
rapidio rapidio: add RapidIO documentation 2011-03-23 19:46:41 -07:00
RCU rcu: Decrease memory-barrier usage based on semi-formal proof 2011-05-26 09:42:23 -07:00
s390 Documentation: fix minor typos/spelling 2011-04-04 17:51:47 -07:00
scheduler sched, doc: Beef up load balancing description 2011-03-31 13:00:35 +02:00
scsi [SCSI] megaraid_sas: Version and Changelog update 2011-05-24 12:36:06 -04:00
security Create Documentation/security/, 2011-05-19 15:59:38 -07:00
serial Fix common misspellings 2011-03-31 11:26:23 -03:00
sh sh: clkfwk: Kill off unused clk_set_rate_ex(). 2010-11-15 18:25:12 +09:00
sound Merge branch 'topic/hda' into for-linus 2011-05-22 10:01:35 +02:00
sparc
spi Fix common misspellings 2011-03-31 11:26:23 -03:00
sysctl coredump: add support for exe_file in core name 2011-05-26 17:12:36 -07:00
target [SCSI] tcm_mod_builder.py: Fix generated *_drop_nodeacl() handler 2011-03-23 11:36:45 -05:00
telephony Fix common misspellings 2011-03-31 11:26:23 -03:00
thermal thermal: Add event notification to thermal framework 2011-01-12 00:08:35 -05:00
timers doc: timers-howto: fix a typo ("unsgined") 2011-05-17 10:52:02 +02:00
trace sched: Get rid of lock_depth 2011-04-24 13:18:38 +02:00
usb Merge branch 'usb-next' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb-2.6 2011-05-23 12:33:02 -07:00
video4linux [media] uvcvideo: Add driver documentation 2011-05-20 09:30:36 -03:00
virtual lguest: remove support for VIRTIO_F_NOTIFY_ON_EMPTY. 2011-05-30 11:14:12 +09:30
vm Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/djm/tmem 2011-05-26 10:50:56 -07:00
w1 Fix common misspellings 2011-03-31 11:26:23 -03:00
watchdog Fix common misspellings 2011-03-31 11:26:23 -03:00
wimax
x86 x86, gart: Rename pci-gart_64.c to amd_gart_64.c 2011-05-10 17:22:06 +02:00
zh_CN Translated Documentation/email-clients.txt 2011-05-06 17:59:20 -07:00
.gitignore add random binaries to .gitignore 2010-04-08 11:34:34 +02:00
00-INDEX Merge branch 'docs-security' into docs-move 2011-05-20 09:10:04 -07:00
applying-patches.txt
atomic_ops.txt Documentation/atomic_ops.txt: avoid volatile in sample code 2011-05-26 17:12:34 -07:00
bad_memory.txt
basic_profiling.txt
binfmt_misc.txt Documentation: update broken web addresses. 2010-08-04 15:21:40 +02:00
braille-console.txt
bt8xxgpio.txt
btmrvl.txt
BUG-HUNTING
bus-virt-phys-mapping.txt documentation: fix almost duplicate filenames (IO/io-mapping.txt) 2010-07-20 17:49:30 +00:00
cachetlb.txt mm: convert mm->cpu_vm_cpumask into cpumask_var_t 2011-05-25 08:39:21 -07:00
Changes Documentation/Changes: minor corrections 2011-03-22 17:44:17 -07:00
circular-buffers.txt Document Linux's circular buffering capabilities 2010-03-24 16:31:22 -07:00
coccinelle.txt scripts/coccinelle: update for compatability with Coccinelle 0.2.4 2010-12-03 12:27:01 +01:00
CodingStyle Documentation/CodingStyle: flesh out if-else examples 2011-03-22 17:44:16 -07:00
cpu-hotplug.txt Fix common misspellings 2011-03-31 11:26:23 -03:00
cpu-load.txt
cputopology.txt topology/sysfs: Provide book id and siblings attributes 2010-09-09 20:41:25 +02:00
dcdbas.txt
debugging-modules.txt
debugging-via-ohci1394.txt ieee1394: update URLs in debugging-via-ohci1394.txt 2009-10-03 09:28:11 +02:00
dell_rbu.txt Fix common misspellings 2011-03-31 11:26:23 -03:00
devices.txt Merge branch 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-2.6 2010-10-28 09:35:11 -07:00
DMA-API-HOWTO.txt Documentation: DMA-API-HOWTO.txt: rename ARCH_KMALLOC_MINALIGN to ARCH_DMA_MINALIGN 2010-08-14 11:56:46 -07:00
DMA-API.txt dma-mapping: remove dma_is_consistent API 2010-08-11 08:59:21 -07:00
DMA-attributes.txt
DMA-ISA-LPC.txt
dmaengine.txt dmaengine: Add API documentation for slave dma usage 2011-05-25 14:49:20 -07:00
dontdiff Documentation: update dontdiff file 2011-05-23 15:14:11 -07:00
dynamic-debug-howto.txt Merge branch 'docs-next' of git://git.lwn.net/linux-2.6 2011-03-27 19:46:59 -07:00
edac.txt Merge branch 'for-linus2' of git://git.profusion.mobi/users/lucas/linux-2.6 2011-04-07 11:14:49 -07:00
eisa.txt Fix common misspellings 2011-03-31 11:26:23 -03:00
email-clients.txt Documentation/email-clients.txt: update Thunderbird docs with wordwrap plugin 2011-01-13 08:03:15 -08:00
feature-removal-schedule.txt Merge branch 'idle-release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-idle-2.6 2011-05-29 11:18:09 -07:00
flexible-arrays.txt flex_array: flex_array_prealloc takes a number of elements, not an end 2011-04-28 16:12:47 -04:00
futex-requeue-pi.txt
gcov.txt trivial: fix typo in CONFIG_DEBUG_FS in gcov doc 2009-09-21 15:14:56 +02:00
gpio.txt Revert "gpiolib: annotate gpio-intialization with __must_check" 2011-01-13 17:26:46 -08:00
highuid.txt
HOWTO Documentation: update LXR web link 2011-05-23 15:14:11 -07:00
hw_random.txt
hwspinlock.txt drivers: hwspinlock: add framework 2011-02-17 09:52:03 -08:00
init.txt init/main.c: improve usability in case of init binary failure 2010-03-06 11:26:29 -08:00
initrd.txt
intel_txt.txt Documentation: update broken web addresses. 2010-08-04 15:21:40 +02:00
Intel-IOMMU.txt intel-iommu: Kill DMAR_BROKEN_GFX_WA option. 2009-09-19 09:37:23 -07:00
io_ordering.txt
io-mapping.txt
iostats.txt Documentation/iostats.txt: bit-size reference etc. 2011-03-23 20:44:18 +01:00
IPMI.txt IPMI: Add the document description of ipmi_get_smi_info 2010-12-14 00:22:00 -05:00
IRQ-affinity.txt bitmap, irq: add smp_affinity_list interface to /proc/irq 2011-05-25 08:39:45 -07:00
IRQ.txt
irqflags-tracing.txt Fix common misspellings 2011-03-31 11:26:23 -03:00
isapnp.txt
java.txt
kernel-doc-nano-HOWTO.txt docbook: warn on unused doc entries 2010-09-11 16:49:21 -07:00
kernel-docs.txt Documentation: update kernel-docs.txt 2011-01-06 09:59:38 -08:00
kernel-parameters.txt Merge git://git.infradead.org/iommu-2.6 2011-06-02 05:48:50 +09:00
kmemcheck.txt
kmemleak.txt Documentation: update kmemleak arch. info 2011-04-04 17:51:46 -07:00
kobject.txt kobject: documentation: Update to refer to kset-example.c. 2010-03-19 07:12:20 -07:00
kprobes.txt tree-wide: fix comment/printk typos 2010-11-01 15:38:34 -04:00
kref.txt kref: Fix typo in kref documentation 2011-03-07 13:20:05 -08:00
ldm.txt Documentation: update broken web addresses. 2010-08-04 15:21:40 +02:00
local_ops.txt
lockdep-design.txt
lockstat.txt Documentation: Add statistics about nested locks 2011-05-28 17:03:29 +02:00
logo.gif
logo.txt
magic-number.txt take coda-private headers out of include/linux 2011-01-12 20:02:48 -05:00
Makefile [media] Remove the old V4L1 v4lgrab.c file 2010-12-29 08:17:12 -02:00
ManagementStyle
mca.txt
md.txt md: Update documentation for sync_min and sync_max entries 2011-04-20 15:40:01 +10:00
media-framework.txt Fix common misspellings 2011-03-31 11:26:23 -03:00
memory-barriers.txt smp: Document transitivity for memory barriers. 2011-03-04 08:05:49 -08:00
memory-hotplug.txt memory hotplug: Allow memory blocks to span multiple memory sections 2011-02-03 16:08:57 -08:00
memory.txt Documentation/memory.txt: remove some very outdated recommendations 2009-09-22 07:17:26 -07:00
mono.txt
mutex-design.txt mutex: Fix annotations to include it in kernel-locking docbook 2010-09-03 08:19:51 +02:00
nmi_watchdog.txt
nommu-mmap.txt nommu: fix malloc performance by adding uninitialized flag 2009-12-15 08:53:24 -08:00
numastat.txt mm: fix NUMA accounting in numastat.txt 2009-09-22 07:17:39 -07:00
oops-tracing.txt panic: Add taint flag TAINT_FIRMWARE_WORKAROUND ('I') 2010-05-19 08:37:43 +01:00
padata.txt Documentation/padata.txt: fix typos etc. 2010-08-11 08:59:18 -07:00
parport-lowlevel.txt
parport.txt
pi-futex.txt
pnp.txt doc: capitalization and other minor fixes in pnp doc 2010-02-05 12:22:44 +01:00
preempt-locking.txt
printk-formats.txt
prio_tree.txt
rbtree.txt Documentation: remove anticipatory scheduler info 2010-11-11 12:09:59 +01:00
rfkill.txt Document the rfkill sysfs ABI 2010-03-10 17:09:33 -05:00
robust-futex-ABI.txt
robust-futexes.txt
rt-mutex-design.txt variable name fix to Documentation/rt-mutex-design.txt 2010-06-05 17:39:09 +02:00
rt-mutex.txt
rtc.txt RTC: Fix up rtc.txt documentation to reflect changes to generic rtc layer 2011-03-09 11:25:10 -08:00
SAK.txt
SecurityBugs Fix common misspellings 2011-03-31 11:26:23 -03:00
serial-console.txt
sgi-ioc4.txt
sgi-visws.txt
SM501.txt
sparse.txt update email address 2010-07-19 10:56:54 +02:00
spinlocks.txt locking: Remove deprecated lock initializers 2011-01-27 12:30:38 +01:00
stable_api_nonsense.txt doc: stable_api_nonsense.txt: fix paragraph to make more sense. 2011-03-30 12:02:05 +02:00
stable_kernel_rules.txt Documentation: -stable rules: upstream commit ID requirement reworded 2010-04-22 15:24:56 -07:00
SubmitChecklist Documentation: update SubmitChecklist for O=objdir and kconfig testing 2010-05-24 07:31:20 -07:00
SubmittingDrivers Fix common misspellings 2011-03-31 11:26:23 -03:00
SubmittingPatches Documentation: update GregKH links 2011-05-23 15:14:11 -07:00
svga.txt
sysfs-rules.txt Fix typos in comments 2010-03-16 11:47:56 +01:00
sysrq.txt documentation: update sysrq.txt magic sysrq keys 2010-10-26 17:32:41 -07:00
unaligned-memory-access.txt
unicode.txt
unshare.txt
VGA-softcursor.txt
vgaarbiter.txt Documentation: fix vgaarbiter.txt typos etc. 2011-05-23 15:14:11 -07:00
video-output.txt
volatile-considered-harmful.txt Documentation/volatile-considered-harmful.txt: correct cpu_relax() documentation 2010-03-24 16:31:20 -07:00
workqueue.txt workqueue: Document debugging tricks 2011-03-31 13:40:42 +02:00
xz.txt decompressors: add XZ decompressor module 2011-01-13 08:03:24 -08:00
zorro.txt