1
linux/kernel
Paul E. McKenney 1423cc033d rcu: Delay rcu_barrier() wait until beginning of next CPU-hotunplug operation.
Ingo Molnar reported this lockup:

 [  200.380003] Hangcheck: hangcheck value past margin!
 [  248.192003] INFO: task S99local:2974 blocked for more than 120 seconds.
 [  248.194532] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
 [  248.202330] S99local      D 0000000c  6256  2974   2687 0x00000000
 [  248.208929]  9c7ebe90 00000086 6b67ef8b 0000000c 9f25a610 81a69869 00000001 820b6990
 [  248.216123]  820b6990 820b6990 9c6e4c20 9c6e4eb4 82c78990 00000000 6b993559 0000000c
 [  248.220616]  9c7ebe90 8105f22a 9c6e4eb4 9c6e4c20 00000001 9c7ebe98 9c7ebeb4 81a65cb3
 [  248.229990] Call Trace:
 [  248.234049]  [<81a69869>] ? _spin_unlock_irqrestore+0x22/0x37
 [  248.239769]  [<8105f22a>] ? prepare_to_wait+0x48/0x4e
 [  248.244796]  [<81a65cb3>] rcu_barrier_cpu_hotplug+0xaa/0xc9
 [  248.250343]  [<8105f029>] ? autoremove_wake_function+0x0/0x38
 [  248.256063]  [<81062cf2>] notifier_call_chain+0x49/0x71
 [  248.261263]  [<81062da0>] raw_notifier_call_chain+0x11/0x13
 [  248.266809]  [<81a0b475>] _cpu_down+0x272/0x288
 [  248.271316]  [<81a0b4d5>] cpu_down+0x4a/0xa2
 [  248.275563]  [<81a0c48a>] store_online+0x2a/0x5e
 [  248.280156]  [<81a0c460>] ? store_online+0x0/0x5e
 [  248.284836]  [<814ddc35>] sysdev_store+0x20/0x28
 [  248.289429]  [<8112e403>] sysfs_write_file+0xb8/0xe3
 [  248.294369]  [<8112e34b>] ? sysfs_write_file+0x0/0xe3
 [  248.299396]  [<810e4c8f>] vfs_write+0x91/0x120
 [  248.303817]  [<810e4dc1>] sys_write+0x40/0x65
 [  248.308150]  [<81002d73>] sysenter_do_call+0x12/0x28

This change moves an RCU grace period delay off of the
critical path for CPU-hotunplug operations.

Since RCU callback migration is only performed on
CPU-hotunplug operations, and since the rcu_barrier() race is
provoked only by consecutive CPU-hotunplug operations, it is
not necessary to delay the end of a given CPU-hotunplug
operation.

We can instead choose to delay the beginning of the next
CPU-hotunplug operation.

Reported-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Josh Triplett <josht@linux.vnet.ibm.com>
Cc: laijs@cn.fujitsu.com
Cc: dipankar@in.ibm.com
Cc: akpm@linux-foundation.org
Cc: mathieu.desnoyers@polymtl.ca
Cc: dvhltc@us.ibm.com
Cc: niv@us.ibm.com
Cc: peterz@infradead.org
Cc: rostedt@goodmis.org
Cc: hugh.dickins@tiscali.co.uk
Cc: benh@kernel.crashing.org
LKML-Reference: <20090819060614.GA14383@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-19 13:58:54 +02:00
..
gcov gcov: enable GCOV_PROFILE_ALL for x86_64 2009-06-18 13:03:58 -07:00
irq genirq: prevent wakeup of freed irq thread 2009-08-13 23:09:27 +02:00
power headers: smp_lock.h redux 2009-07-12 12:22:34 -07:00
time clocksource: Prevent NULL pointer dereference 2009-07-19 17:15:54 +02:00
trace Remove double removal of blktrace directory 2009-08-12 18:50:08 +02:00
.gitignore
acct.c bsdacct: fix access to invalid filp in acct_on() 2009-06-30 18:56:00 -07:00
async.c async: Fix lack of boot-time console due to insufficient synchronization 2009-06-08 12:31:53 -07:00
audit_tree.c Fix rule eviction order for AUDIT_DIR 2009-06-24 00:02:38 -04:00
audit_watch.c Audit: clean up all op= output to include string quoting 2009-06-24 00:00:52 -04:00
audit.c Fix rule eviction order for AUDIT_DIR 2009-06-24 00:02:38 -04:00
audit.h Fix rule eviction order for AUDIT_DIR 2009-06-24 00:02:38 -04:00
auditfilter.c Audit: clean up all op= output to include string quoting 2009-06-24 00:00:52 -04:00
auditsc.c Fix rule eviction order for AUDIT_DIR 2009-06-24 00:02:38 -04:00
backtracetest.c
bounds.c
capability.c
cgroup_debug.c
cgroup_freezer.c
cgroup.c cgroup avoid permanent sleep at rmdir 2009-07-29 19:10:35 -07:00
compat.c signals: implement sys_rt_tgsigqueueinfo 2009-04-30 19:24:24 +02:00
configs.c
cpu.c mm/init: cpu_hotplug_init() must be initialized before SLAB 2009-06-22 21:18:12 -07:00
cpuset.c cpuset,mm: update tasks' mems_allowed in time 2009-06-16 19:47:31 -07:00
cred-internals.h
cred.c CRED: Rename cred_exec_mutex to reflect that it's a guard against ptrace 2009-05-11 08:15:36 +10:00
delayacct.c
dma-coherent.c
dma.c
exec_domain.c
exit.c headers: mnt_namespace.h redux 2009-07-08 09:31:56 -07:00
extable.c
fork.c execve: must clear current->clear_child_tid 2009-08-07 10:39:56 -07:00
freezer.c sched: fix nr_uninterruptible accounting of frozen tasks really 2009-07-18 14:19:53 +02:00
futex_compat.c futex: Fix compat_futex to be same as futex for REQUEUE_PI 2009-08-10 15:41:12 +02:00
futex.c futex: Fix handling of bad requeue syscall pairing 2009-08-10 20:38:11 +02:00
groups.c groups: move code to kernel/groups.c 2009-06-16 19:47:48 -07:00
hrtimer.c hrtimer: Fix migration expiry check 2009-07-10 17:32:55 +02:00
hung_task.c
itimer.c
kallsyms.c kernel/kallsyms.c: replace deprecated __initcall with device_initcall and fix whitespace 2009-06-09 22:37:52 +02:00
Kconfig.freezer
Kconfig.hz
Kconfig.preempt
kexec.c kexec: fix omitting offset in extended crashkernel syntax 2009-07-29 19:10:34 -07:00
kfifo.c kernel/kfifo.c: replace conditional test with is_power_of_2() 2009-06-16 19:47:47 -07:00
kgdb.c sysrq, intel_fb: fix sysrq g collision 2009-05-15 07:56:24 -05:00
kmod.c headers: mnt_namespace.h redux 2009-07-08 09:31:56 -07:00
kprobes.c kprobes: Use kernel_text_address() for checking probe address 2009-07-30 16:44:06 -07:00
ksysfs.c
kthread.c update the comment in kthread_stop() 2009-07-27 12:15:46 -07:00
latencytop.c
lockdep_internals.h lockdep: increase MAX_LOCKDEP_ENTRIES and MAX_LOCKDEP_CHAINS 2009-05-12 19:59:52 +02:00
lockdep_proc.c lockdep: Fix file mode of lock_stat 2009-08-07 11:58:38 +02:00
lockdep_states.h
lockdep.c Merge branch 'linus' into tracing/core 2009-05-07 11:17:34 +02:00
Makefile Merge commit 'v2.6.31-rc6' into core/rcu 2009-08-15 18:56:13 +02:00
marker.c
module.c module: use MODULE_SYMBOL_PREFIX with module_layout 2009-07-27 12:15:45 -07:00
mutex-debug.c
mutex-debug.h
mutex.c Merge branch 'linus' into perfcounters/core 2009-06-11 17:55:42 +02:00
mutex.h
notifier.c
ns_cgroup.c
nsproxy.c nsproxy: extract create_nsproxy() 2009-06-18 13:03:56 -07:00
panic.c trace: stop tracer in oops_enter() 2009-07-24 15:30:45 -04:00
params.c module_param: allow 'bool' module_params to be bool, not just int. 2009-06-12 21:46:58 +09:30
perf_counter.c perf_counter: Report the cloning task as parent on perf_counter_fork() 2009-08-13 16:17:15 +02:00
pid_namespace.c pidns: rewrite copy_pid_ns() 2009-06-18 13:03:55 -07:00
pid.c kmemleak: Remove alloc_bootmem annotations introduced in the past 2009-07-09 17:07:02 +01:00
pm_qos_params.c
posix-cpu-timers.c posix_cpu_timers_exit_group(): Do not use thread_group_cputimer() 2009-08-08 18:30:25 +02:00
posix-timers.c posix-timers: Fix oops in clock_nanosleep() with CLOCK_MONOTONIC_RAW 2009-08-04 10:16:41 +02:00
printk.c printk: Add KERN_DEFAULT printk log-level 2009-06-16 11:02:28 -07:00
profile.c profile: suppress warning about large allocations when profile=1 is specified 2009-07-29 19:10:36 -07:00
ptrace.c cred_guard_mutex: do not return -EINTR to user-space 2009-07-06 13:57:04 -07:00
rcupdate.c rcu: Delay rcu_barrier() wait until beginning of next CPU-hotunplug operation. 2009-08-19 13:58:54 +02:00
rcupreempt_trace.c rcu: Make rcupreempt_trace.c look at offline CPUs 2009-08-15 19:02:09 +02:00
rcupreempt.c rcu: Fix typo in rcu_irq_exit() comment header 2009-08-15 19:02:09 +02:00
rcutorture.c rcu: Add synchronize_sched_expedited() torture tests 2009-07-03 10:02:29 +02:00
rcutree_trace.c
rcutree.c rcu: Simplify RCU CPU-hotplug notification 2009-08-15 19:02:08 +02:00
rcutree.h
relay.c
res_counter.c memcg: add interface to reset limits 2009-06-18 13:03:48 -07:00
resource.c kernel/resource.c: fix sign extension in reserve_setup() 2009-06-30 18:56:00 -07:00
rtmutex_common.h
rtmutex-debug.c
rtmutex-debug.h
rtmutex-tester.c
rtmutex.c rtmutex: Avoid deadlock in rt_mutex_start_proxy_lock() 2009-08-06 05:50:21 +02:00
rtmutex.h
rwsem.c
sched_clock.c sched: Fix fallback sched_clock()'s offset when using jiffies 2009-05-09 10:08:19 +02:00
sched_cpupri.c sched: Fix race in cpupri introduced by cpumask_var changes 2009-08-02 14:23:29 +02:00
sched_cpupri.h
sched_debug.c sched: Hide runqueues from direct refer at source code level 2009-06-17 18:29:42 +02:00
sched_fair.c sched: Fix latencytop and sleep profiling vs group scheduling 2009-08-02 14:10:12 +02:00
sched_features.h
sched_idletask.c sched, timers: move calc_load() to scheduler 2009-05-15 15:32:45 +02:00
sched_rt.c sched_rt: Fix overload bug on rt group scheduling 2009-07-10 10:43:29 +02:00
sched_stats.h
sched.c Merge commit 'v2.6.31-rc6' into core/rcu 2009-08-15 18:56:13 +02:00
seccomp.c
semaphore.c
signal.c do_sigaltstack: small cleanups 2009-08-01 11:18:56 -07:00
slow-work.c slow-work: use round_jiffies() for thread pool's cull and OOM timers 2009-06-16 19:47:49 -07:00
smp.c generic-ipi: fix hotplug_cfd() 2009-08-07 10:39:55 -07:00
softirq.c softirq: introduce tasklet_hrtimer infrastructure 2009-07-22 17:01:17 +02:00
softlockup.c
spinlock.c
srcu.c
stacktrace.c
stop_machine.c
sys_ni.c
sys.c groups: move code to kernel/groups.c 2009-06-16 19:47:48 -07:00
sysctl_check.c
sysctl.c Merge branch 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip 2009-06-28 11:05:28 -07:00
taskstats.c
test_kprobes.c
time.c
timeconst.pl
timer.c timer: Avoid reading uninitialized data 2009-07-18 23:11:43 +02:00
tracepoint.c
tsacct.c
uid16.c
up.c
user_namespace.c
user.c sched: delayed cleanup of user_struct 2009-06-15 21:30:23 -07:00
utsname_sysctl.c
utsname.c utsns: extract creeate_uts_ns() 2009-06-18 13:03:55 -07:00
wait.c locking, sched: Give waitqueue spinlocks their own lockdep classes 2009-08-10 14:43:09 +02:00
workqueue.c ftrace, workqueuetrace: make workqueue tracepoints use TRACE_EVENT macro 2009-06-02 01:10:40 +02:00