leandrof/linux
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
d78045306c
linux/kernel
History
Waiman Long d78045306c locking/pvqspinlock, x86: Optimize the PV unlock code path 
The unlock function in queued spinlocks was optimized for better
performance on bare metal systems at the expense of virtualized guests.

For x86-64 systems, the unlock call needs to go through a
PV_CALLEE_SAVE_REGS_THUNK() which saves and restores 8 64-bit
registers before calling the real __pv_queued_spin_unlock()
function. The thunk code may also be in a separate cacheline from
__pv_queued_spin_unlock().

This patch optimizes the PV unlock code path by:

 1) Moving the unlock slowpath code from the fastpath into a separate
    __pv_queued_spin_unlock_slowpath() function to make the fastpath
    as simple as possible..

 2) For x86-64, hand-coded an assembly function to combine the register
    saving thunk code with the fastpath code. Only registers that
    are used in the fastpath will be saved and restored. If the
    fastpath fails, the slowpath function will be called via another
    PV_CALLEE_SAVE_REGS_THUNK(). For 32-bit, it falls back to the C
    __pv_queued_spin_unlock() code as the thunk saves and restores
    only one 32-bit register.

With a microbenchmark of 5M lock-unlock loop, the table below shows
the execution times before and after the patch with different number
of threads in a VM running on a 32-core Westmere-EX box with x86-64
4.2-rc1 based kernels:

  Threads	Before patch	After patch	% Change
  -------	------------	-----------	--------
     1		   134.1 ms	  119.3 ms	  -11%
     2		   1286  ms	   953  ms	  -26%
     3		   3715  ms	  3480  ms	  -6.3%
     4		   4092  ms	  3764  ms	  -8.0%

Signed-off-by: Waiman Long <Waiman.Long@hpe.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Douglas Hatch <doug.hatch@hpe.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Scott J Norton <scott.norton@hpe.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1447114167-47185-5-git-send-email-Waiman.Long@hpe.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-11-23 10:02:02 +01:00
..
bpf bpf, verifier: annotate verbose printer with __printf 2015-11-03 11:29:56 -05:00
configs kconfig: add xenconfig defconfig helper 2015-06-16 11:04:29 +01:00
debug debug: prevent entering debug mode on panic/exception. 2015-02-19 12:39:03 -06:00
events Merge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2015-11-15 09:36:24 -08:00
gcov gcov: add support for GCC 5.1 2015-06-30 19:44:57 -07:00
irq Merge branches 'irq-urgent-for-linus' and 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2015-11-15 09:30:48 -08:00
livepatch livepatch: x86: fix relocation computation with kASLR 2015-11-11 17:36:04 +01:00
locking locking/pvqspinlock, x86: Optimize the PV unlock code path 2015-11-23 10:02:02 +01:00
power mm, page_alloc: rename __GFP_WAIT to __GFP_RECLAIM 2015-11-06 17:50:42 -08:00
printk printk: prevent userland from spoofing kernel messages 2015-11-06 17:50:42 -08:00
rcu Merge branches 'doc.2015.10.06a', 'percpu-rwsem.2015.10.06a' and 'torture.2015.10.06a' into HEAD 2015-10-07 16:06:25 -07:00
sched Merge branch 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2015-11-15 09:35:33 -08:00
time Merge branches 'irq-urgent-for-linus' and 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2015-11-15 09:30:48 -08:00
trace This contains three more clean up patches. 2015-11-12 16:22:54 -08:00
.gitignore certs: add .gitignore to stop git nagging about x509_certificate_list 2015-10-21 15:18:35 +01:00
acct.c acct: check FMODE_CAN_WRITE 2015-04-11 22:27:55 -04:00
async.c
audit_fsnotify.c audit: clean simple fsnotify implementation 2015-08-06 16:14:53 -04:00
audit_tree.c audit: audit_tree_match can be boolean 2015-11-04 08:23:51 -05:00
audit_watch.c Merge branch 'upstream' of git://git.infradead.org/users/pcmoore/audit 2015-09-08 13:34:59 -07:00
audit.c mm, page_alloc: distinguish between being unable to sleep, unwilling to sleep and avoiding waking kswapd 2015-11-06 17:50:42 -08:00
audit.h audit: audit_tree_match can be boolean 2015-11-04 08:23:51 -05:00
auditfilter.c audit: fix comment block whitespace 2015-11-04 08:23:51 -05:00
auditsc.c Merge branch 'upstream' of git://git.infradead.org/users/pcmoore/audit 2015-09-08 13:34:59 -07:00
backtracetest.c
bounds.c
capability.c kernel: conditionally support non-root users, groups and capabilities 2015-04-15 16:35:22 -07:00
cgroup_freezer.c cgroup: allow a cgroup subsystem to reject a fork 2015-07-14 17:29:23 -04:00
cgroup_pids.c cgroup: add cgroup_subsys->free() method and use it to fix pids controller 2015-10-15 16:41:53 -04:00
cgroup.c mm, page_alloc: distinguish between being unable to sleep, unwilling to sleep and avoiding waking kswapd 2015-11-06 17:50:42 -08:00
compat.c compat: cleanup coding in compat_get_bitmap() and compat_put_bitmap() 2015-06-04 23:57:18 +02:00
configs.c
context_tracking.c context_tracking: avoid irq_save/irq_restore on guest entry and exit 2015-11-10 12:06:23 +01:00
cpu_pm.c kernel/cpu_pm: fix cpu_cluster_pm_exit comment 2015-09-03 02:42:20 +02:00
cpu.c Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2015-11-03 18:03:50 -08:00
cpuset.c Merge branch 'akpm' (patches from Andrew) 2015-11-05 23:10:54 -08:00
crash_dump.c
cred.c kernel/cred.c: remove unnecessary kdebug atomic reads 2015-09-10 13:29:01 -07:00
delayacct.c
dma.c
elfcore.c
exec_domain.c Remove rest of exec domains. 2015-04-12 21:03:31 +02:00
exit.c Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2015-11-03 18:03:50 -08:00
extable.c kernel/extable.c: remove duplicated include 2015-09-10 13:29:01 -07:00
fork.c Merge branch 'akpm' (patches from Andrew) 2015-11-05 23:10:54 -08:00
freezer.c
futex_compat.c
futex.c driver core update for 4.4-rc1 2015-11-04 21:50:37 -08:00
groups.c kernel: conditionally support non-root users, groups and capabilities 2015-04-15 16:35:22 -07:00
hung_task.c kernel/hung_task.c: change hung_task.c to use for_each_process_thread() 2015-04-15 16:35:22 -07:00
irq_work.c
jump_label.c locking/static_keys: Add selftest 2015-08-03 11:34:16 +02:00
kallsyms.c
kcmp.c
Kconfig.freezer
Kconfig.hz
Kconfig.locks locking/qrwlock: Rename QUEUE_RWLOCK to QUEUED_RWLOCKS 2015-05-12 09:46:00 +02:00
Kconfig.preempt
kexec_core.c kexec: use file name as the output message prefix 2015-11-06 17:50:42 -08:00
kexec_file.c kexec: use file name as the output message prefix 2015-11-06 17:50:42 -08:00
kexec_internal.h kexec: split kexec_file syscall code to kexec_file.c 2015-09-10 13:29:01 -07:00
kexec.c kexec: use file name as the output message prefix 2015-11-06 17:50:42 -08:00
kmod.c kmod: don't run async usermode helper as a child of kworker thread 2015-10-23 17:55:10 +09:00
kprobes.c perf/x86/hw_breakpoints: Disallow kernel breakpoints unless kprobe-safe 2015-08-04 10:16:54 +02:00
ksysfs.c kexec: split kexec_load syscall from kexec core code 2015-09-10 13:29:01 -07:00
kthread.c kernel/kthread.c:kthread_create_on_node(): clarify documentation 2015-09-04 16:54:41 -07:00
latencytop.c
Makefile sys_membarrier(): system-wide memory barrier (generic, x86) 2015-09-11 15:21:34 -07:00
membarrier.c sys_membarrier(): system-wide memory barrier (generic, x86) 2015-09-11 15:21:34 -07:00
memremap.c libnvdimm for 4.4: 2015-11-10 12:07:22 -08:00
module_signing.c KEYS: Merge the type-specific data with the payload data 2015-10-21 15:18:36 +01:00
module-internal.h
module.c module: Fix locking in symbol_put_addr() 2015-08-24 10:37:01 +09:30
notifier.c Merge branch 'x86-asm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2015-09-01 08:40:25 -07:00
nsproxy.c
padata.c padata: use %*pb[l] to print bitmaps including cpumasks and nodemasks 2015-02-13 21:21:38 -08:00
panic.c kernel/panic.c: turn off locks debug before releasing console lock 2015-11-20 16:17:32 -08:00
params.c Nothing exciting, minor tweaks and cleanups. 2015-11-09 15:53:39 -08:00
pid_namespace.c
pid.c rcu: Rename rcu_lockdep_assert() to RCU_LOCKDEP_WARN() 2015-07-22 15:27:32 -07:00
profile.c mm: rename alloc_pages_exact_node() to __alloc_pages_node() 2015-09-08 15:35:28 -07:00
ptrace.c seccomp, ptrace: add support for dumping seccomp filters 2015-10-27 19:55:13 -07:00
range.c
reboot.c kexec: split kexec_load syscall from kexec core code 2015-09-10 13:29:01 -07:00
relay.c kernel/relay.c: use kvfree() in relay_free_page_array() 2015-06-30 19:44:59 -07:00
resource.c mm: enhance region_is_ram() to region_intersects() 2015-08-10 23:07:05 -04:00
seccomp.c seccomp, ptrace: add support for dumping seccomp filters 2015-10-27 19:55:13 -07:00
signal.c kernel/signal.c: unexport sigsuspend() 2015-11-20 16:17:32 -08:00
smp.c mm, page_alloc: distinguish between being unable to sleep, unwilling to sleep and avoiding waking kswapd 2015-11-06 17:50:42 -08:00
smpboot.c stop_machine: Kill smp_hotplug_thread->pre_unpark, introduce stop_machine_unpark() 2015-10-20 10:23:55 +02:00
smpboot.h
softirq.c Merge branch 'locking-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2015-02-09 15:24:03 -08:00
stacktrace.c
stop_machine.c sched: Move cpu_active() tests from stop_two_cpus() into migrate_swap_stop() 2015-10-20 10:25:56 +02:00
sys_ni.c mm: mlock: add new mlock system call 2015-11-05 19:34:48 -08:00
sys.c pidns: fix set/getpriority and ioprio_set/get in PRIO_USER mode 2015-11-06 17:50:42 -08:00
sysctl_binary.c
sysctl.c kernel/watchdog.c: add sysctl knob hardlockup_panic 2015-11-05 19:34:48 -08:00
task_work.c task_work: remove fifo ordering guarantee 2015-09-05 13:46:58 -07:00
taskstats.c
test_kprobes.c
torture.c torture: Consolidate cond_resched_rcu_qs() into stutter_wait() 2015-10-06 11:25:01 -07:00
tracepoint.c tracepoint: Give priority to probes of tracepoints 2015-10-25 21:33:54 -04:00
tsacct.c
uid16.c
up.c
user_namespace.c capabilities: ambient capabilities 2015-09-04 16:54:41 -07:00
user-return-notifier.c
user.c
utsname_sysctl.c
utsname.c
watchdog.c kernel/watchdog.c: fix race between proc_watchdog_thresh() and watchdog_timer_fn() 2015-11-05 19:34:48 -08:00
workqueue_internal.h
workqueue.c Merge branch 'for-4.4' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq 2015-11-05 14:16:27 -08:00