19 files changed, 136 insertions, 158 deletions

diff --git a/kernel/cpuset.c b/kernel/cpuset.c
index 647c77a88fcb..a85678865c5e 100644
--- a/kernel/cpuset.c
+++ b/kernel/cpuset.c
@@ -568,7 +568,7 @@ update_domain_attr_tree(struct sched_domain_attr *dattr, struct cpuset *c)
  * load balancing domains (sched domains) as specified by that partial
  * partition.
  *
- * See "What is sched_load_balance" in Documentation/cpusets.txt
+ * See "What is sched_load_balance" in Documentation/cgroups/cpusets.txt
  * for a background explanation of this.
  *
  * Does not return errors, on the theory that the callers of this
diff --git a/kernel/dma-coherent.c b/kernel/dma-coherent.c
index 038707404b76..962a3b574f21 100644
--- a/kernel/dma-coherent.c
+++ b/kernel/dma-coherent.c
@@ -98,7 +98,7 @@ EXPORT_SYMBOL(dma_mark_declared_memory_occupied);
  * @size:       size of requested memory area
  * @dma_handle: This will be filled with the correct dma handle
  * @ret:        This pointer will be filled with the virtual address
- *              to allocated area.
+ *              to allocated area.
  *
  * This function should be only called from per-arch dma_alloc_coherent()
  * to support allocation from per-device coherent memory pools.
@@ -118,31 +118,32 @@ int dma_alloc_from_coherent(struct device *dev, ssize_t size,
         mem = dev->dma_mem;
         if (!mem)
                 return 0;
-        if (unlikely(size > mem->size))
-                return 0;
+
+        *ret = NULL;
+
+        if (unlikely(size > (mem->size << PAGE_SHIFT)))
+                goto err;
 
         pageno = bitmap_find_free_region(mem->bitmap, mem->size, order);
-        if (pageno >= 0) {
-                /*
-                 * Memory was found in the per-device arena.
-                 */
-                *dma_handle = mem->device_base + (pageno << PAGE_SHIFT);
-                *ret = mem->virt_base + (pageno << PAGE_SHIFT);
-                memset(*ret, 0, size);
-        } else if (mem->flags & DMA_MEMORY_EXCLUSIVE) {
-                /*
-                 * The per-device arena is exhausted and we are not
-                 * permitted to fall back to generic memory.
-                 */
-                *ret = NULL;
-        } else {
-                /*
-                 * The per-device arena is exhausted and we are
-                 * permitted to fall back to generic memory.
-                 */
-                 return 0;
-        }
+        if (unlikely(pageno < 0))
+                goto err;
+
+        /*
+         * Memory was found in the per-device area.
+         */
+        *dma_handle = mem->device_base + (pageno << PAGE_SHIFT);
+        *ret = mem->virt_base + (pageno << PAGE_SHIFT);
+        memset(*ret, 0, size);
+
         return 1;
+
+err:
+        /*
+         * In the case where the allocation can not be satisfied from the
+         * per-device area, try to fall back to generic memory if the
+         * constraints allow it.
+         */
+        return mem->flags & DMA_MEMORY_EXCLUSIVE;
 }
 EXPORT_SYMBOL(dma_alloc_from_coherent);
 
diff --git a/kernel/fork.c b/kernel/fork.c
index bf0cef8bbdf2..242a706e7721 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -817,17 +817,17 @@ static void posix_cpu_timers_init_group(struct signal_struct *sig)
 static int copy_signal(unsigned long clone_flags, struct task_struct *tsk)
 {
         struct signal_struct *sig;
-        int ret;
 
         if (clone_flags & CLONE_THREAD) {
-                ret = thread_group_cputime_clone_thread(current);
-                if (likely(!ret)) {
-                        atomic_inc(&current->signal->count);
-                        atomic_inc(&current->signal->live);
-                }
-                return ret;
+                atomic_inc(&current->signal->count);
+                atomic_inc(&current->signal->live);
+                return 0;
         }
         sig = kmem_cache_alloc(signal_cachep, GFP_KERNEL);
+
+        if (sig)
+                posix_cpu_timers_init_group(sig);
+
         tsk->signal = sig;
         if (!sig)
                 return -ENOMEM;
@@ -864,8 +864,6 @@ static int copy_signal(unsigned long clone_flags, struct task_struct *tsk)
         memcpy(sig->rlim, current->signal->rlim, sizeof sig->rlim);
         task_unlock(current->group_leader);
 
-        posix_cpu_timers_init_group(sig);
-
         acct_init_pacct(&sig->pacct);
 
         tty_audit_fork(sig);
diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c
index 2dc30c59c5fd..f33afb0407bc 100644
--- a/kernel/hrtimer.c
+++ b/kernel/hrtimer.c
@@ -614,7 +614,9 @@ void clock_was_set(void)
  */
 void hres_timers_resume(void)
 {
-        /* Retrigger the CPU local events: */
+        WARN_ONCE(!irqs_disabled(),
+                  KERN_INFO "hres_timers_resume() called with IRQs enabled!");
+
         retrigger_next_event(NULL);
 }
 
diff --git a/kernel/kallsyms.c b/kernel/kallsyms.c
index e694afa0eb8c..7b8b0f21a5b1 100644
--- a/kernel/kallsyms.c
+++ b/kernel/kallsyms.c
@@ -30,19 +30,20 @@
 #define all_var 0
 #endif
 
-extern const unsigned long kallsyms_addresses[];
-extern const u8 kallsyms_names[];
+/* These will be re-linked against their real values during the second link stage */
+extern const unsigned long kallsyms_addresses[] __attribute__((weak));
+extern const u8 kallsyms_names[] __attribute__((weak));
 
 /* tell the compiler that the count isn't in the small data section if the arch
  * has one (eg: FRV)
  */
 extern const unsigned long kallsyms_num_syms
-        __attribute__((__section__(".rodata")));
+__attribute__((weak, section(".rodata")));
 
-extern const u8 kallsyms_token_table[];
-extern const u16 kallsyms_token_index[];
+extern const u8 kallsyms_token_table[] __attribute__((weak));
+extern const u16 kallsyms_token_index[] __attribute__((weak));
 
-extern const unsigned long kallsyms_markers[];
+extern const unsigned long kallsyms_markers[] __attribute__((weak));
 
 static inline int is_kernel_inittext(unsigned long addr)
 {
@@ -167,6 +168,9 @@ static unsigned long get_symbol_pos(unsigned long addr,
         unsigned long symbol_start = 0, symbol_end = 0;
         unsigned long i, low, high, mid;
 
+        /* This kernel should never had been booted. */
+        BUG_ON(!kallsyms_addresses);
+
         /* do a binary search on the sorted kallsyms_addresses array */
         low = 0;
         high = kallsyms_num_syms;
diff --git a/kernel/kprobes.c b/kernel/kprobes.c
index 1b9cbdc0127a..7ba8cd9845cb 100644
--- a/kernel/kprobes.c
+++ b/kernel/kprobes.c
@@ -123,7 +123,7 @@ static int collect_garbage_slots(void);
 static int __kprobes check_safety(void)
 {
         int ret = 0;
-#if defined(CONFIG_PREEMPT) && defined(CONFIG_PM)
+#if defined(CONFIG_PREEMPT) && defined(CONFIG_FREEZER)
         ret = freeze_processes();
         if (ret == 0) {
                 struct task_struct *p, *q;
diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c
index 157de3a47832..fa07da94d7be 100644
--- a/kernel/posix-cpu-timers.c
+++ b/kernel/posix-cpu-timers.c
@@ -10,76 +10,6 @@
 #include <linux/kernel_stat.h>
 
 /*
- * Allocate the thread_group_cputime structure appropriately and fill in the
- * current values of the fields.  Called from copy_signal() via
- * thread_group_cputime_clone_thread() when adding a second or subsequent
- * thread to a thread group.  Assumes interrupts are enabled when called.
- */
-int thread_group_cputime_alloc(struct task_struct *tsk)
-{
-        struct signal_struct *sig = tsk->signal;
-        struct task_cputime *cputime;
-
-        /*
-         * If we have multiple threads and we don't already have a
-         * per-CPU task_cputime struct (checked in the caller), allocate
-         * one and fill it in with the times accumulated so far.  We may
-         * race with another thread so recheck after we pick up the sighand
-         * lock.
-         */
-        cputime = alloc_percpu(struct task_cputime);
-        if (cputime == NULL)
-                return -ENOMEM;
-        spin_lock_irq(&tsk->sighand->siglock);
-        if (sig->cputime.totals) {
-                spin_unlock_irq(&tsk->sighand->siglock);
-                free_percpu(cputime);
-                return 0;
-        }
-        sig->cputime.totals = cputime;
-        cputime = per_cpu_ptr(sig->cputime.totals, smp_processor_id());
-        cputime->utime = tsk->utime;
-        cputime->stime = tsk->stime;
-        cputime->sum_exec_runtime = tsk->se.sum_exec_runtime;
-        spin_unlock_irq(&tsk->sighand->siglock);
-        return 0;
-}
-
-/**
- * thread_group_cputime - Sum the thread group time fields across all CPUs.
- *
- * @tsk:        The task we use to identify the thread group.
- * @times:      task_cputime structure in which we return the summed fields.
- *
- * Walk the list of CPUs to sum the per-CPU time fields in the thread group
- * time structure.
- */
-void thread_group_cputime(
-        struct task_struct *tsk,
-        struct task_cputime *times)
-{
-        struct task_cputime *totals, *tot;
-        int i;
-
-        totals = tsk->signal->cputime.totals;
-        if (!totals) {
-                times->utime = tsk->utime;
-                times->stime = tsk->stime;
-                times->sum_exec_runtime = tsk->se.sum_exec_runtime;
-                return;
-        }
-
-        times->stime = times->utime = cputime_zero;
-        times->sum_exec_runtime = 0;
-        for_each_possible_cpu(i) {
-                tot = per_cpu_ptr(totals, i);
-                times->utime = cputime_add(times->utime, tot->utime);
-                times->stime = cputime_add(times->stime, tot->stime);
-                times->sum_exec_runtime += tot->sum_exec_runtime;
-        }
-}
-
-/*
  * Called after updating RLIMIT_CPU to set timer expiration if necessary.
  */
 void update_rlimit_cpu(unsigned long rlim_new)
diff --git a/kernel/power/Makefile b/kernel/power/Makefile
index 597823b5b700..d7a10167a25b 100644
--- a/kernel/power/Makefile
+++ b/kernel/power/Makefile
@@ -4,7 +4,8 @@ EXTRA_CFLAGS	+=	-DDEBUG
 endif
 
 obj-y                           := main.o
-obj-$(CONFIG_PM_SLEEP)          += process.o console.o
+obj-$(CONFIG_PM_SLEEP)          += console.o
+obj-$(CONFIG_FREEZER)           += process.o
 obj-$(CONFIG_HIBERNATION)       += swsusp.o disk.o snapshot.o swap.o user.o
 
 obj-$(CONFIG_MAGIC_SYSRQ)       += poweroff.o
diff --git a/kernel/rcuclassic.c b/kernel/rcuclassic.c
index 490934fc7ac3..bd5a9003497c 100644
--- a/kernel/rcuclassic.c
+++ b/kernel/rcuclassic.c
@@ -716,7 +716,7 @@ void rcu_check_callbacks(int cpu, int user)
         raise_rcu_softirq();
 }
 
-static void rcu_init_percpu_data(int cpu, struct rcu_ctrlblk *rcp,
+static void __cpuinit rcu_init_percpu_data(int cpu, struct rcu_ctrlblk *rcp,
                                                 struct rcu_data *rdp)
 {
         unsigned long flags;
diff --git a/kernel/rcutree.c b/kernel/rcutree.c
index f2d8638e6c60..b2fd602a6f6f 100644
--- a/kernel/rcutree.c
+++ b/kernel/rcutree.c
@@ -1314,7 +1314,7 @@ int rcu_needs_cpu(int cpu)
  * access due to the fact that this CPU cannot possibly have any RCU
  * callbacks in flight yet.
  */
-static void
+static void __cpuinit
 rcu_init_percpu_data(int cpu, struct rcu_state *rsp)
 {
         unsigned long flags;
diff --git a/kernel/relay.c b/kernel/relay.c
index 09ac2008f77b..9d79b7854fa6 100644
--- a/kernel/relay.c
+++ b/kernel/relay.c
@@ -663,8 +663,10 @@ int relay_late_setup_files(struct rchan *chan,
 
         mutex_lock(&relay_channels_mutex);
         /* Is chan already set up? */
-        if (unlikely(chan->has_base_filename))
+        if (unlikely(chan->has_base_filename)) {
+                mutex_unlock(&relay_channels_mutex);
                 return -EEXIST;
+        }
         chan->has_base_filename = 1;
         chan->parent = parent;
         curr_cpu = get_cpu();
diff --git a/kernel/resource.c b/kernel/resource.c
index ca6a1536b205..fd5d7d574bb9 100644
--- a/kernel/resource.c
+++ b/kernel/resource.c
@@ -620,6 +620,7 @@ resource_size_t resource_alignment(struct resource *res)
  * @start: resource start address
  * @n: resource region size
  * @name: reserving caller's ID string
+ * @flags: IO resource flags
  */
 struct resource * __request_region(struct resource *parent,
                                    resource_size_t start, resource_size_t n,
diff --git a/kernel/sched.c b/kernel/sched.c
index eb1931eef587..52bbf1c842a8 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -1323,8 +1323,8 @@ static inline void update_load_sub(struct load_weight *lw, unsigned long dec)
  * slice expiry etc.
  */
 
-#define WEIGHT_IDLEPRIO         2
-#define WMULT_IDLEPRIO          (1 << 31)
+#define WEIGHT_IDLEPRIO                3
+#define WMULT_IDLEPRIO         1431655765
 
 /*
  * Nice levels are multiplicative, with a gentle 10% change for every
@@ -4440,7 +4440,7 @@ void __kprobes sub_preempt_count(int val)
         /*
          * Underflow?
          */
-       if (DEBUG_LOCKS_WARN_ON(val > preempt_count() - (!!kernel_locked())))
+        if (DEBUG_LOCKS_WARN_ON(val > preempt_count()))
                 return;
         /*
          * Is the spinlock portion underflowing?
@@ -9050,6 +9050,13 @@ static int tg_schedulable(struct task_group *tg, void *data)
                 runtime = d->rt_runtime;
         }
 
+#ifdef CONFIG_USER_SCHED
+        if (tg == &root_task_group) {
+                period = global_rt_period();
+                runtime = global_rt_runtime();
+        }
+#endif
+
         /*
          * Cannot have more runtime than the period.
          */
diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index 8e1352c75557..5cc1c162044f 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -283,7 +283,7 @@ static void update_min_vruntime(struct cfs_rq *cfs_rq)
                                                    struct sched_entity,
                                                    run_node);
 
-                if (vruntime == cfs_rq->min_vruntime)
+                if (!cfs_rq->curr)
                         vruntime = se->vruntime;
                 else
                         vruntime = min_vruntime(vruntime, se->vruntime);
@@ -429,7 +429,10 @@ static u64 sched_slice(struct cfs_rq *cfs_rq, struct sched_entity *se)
         u64 slice = __sched_period(cfs_rq->nr_running + !se->on_rq);
 
         for_each_sched_entity(se) {
-                struct load_weight *load = &cfs_rq->load;
+                struct load_weight *load;
+
+                cfs_rq = cfs_rq_of(se);
+                load = &cfs_rq->load;
 
                 if (unlikely(!se->on_rq)) {
                         struct load_weight lw = cfs_rq->load;
@@ -677,9 +680,13 @@ place_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int initial)
                         unsigned long thresh = sysctl_sched_latency;
 
                         /*
-                         * convert the sleeper threshold into virtual time
+                         * Convert the sleeper threshold into virtual time.
+                         * SCHED_IDLE is a special sub-class.  We care about
+                         * fairness only relative to other SCHED_IDLE tasks,
+                         * all of which have the same weight.
                          */
-                        if (sched_feat(NORMALIZED_SLEEPER))
+                        if (sched_feat(NORMALIZED_SLEEPER) &&
+                                        task_of(se)->policy != SCHED_IDLE)
                                 thresh = calc_delta_fair(thresh, se);
 
                         vruntime -= thresh;
@@ -1340,14 +1347,18 @@ wakeup_preempt_entity(struct sched_entity *curr, struct sched_entity *se)
 
 static void set_last_buddy(struct sched_entity *se)
 {
-        for_each_sched_entity(se)
-                cfs_rq_of(se)->last = se;
+        if (likely(task_of(se)->policy != SCHED_IDLE)) {
+                for_each_sched_entity(se)
+                        cfs_rq_of(se)->last = se;
+        }
 }
 
 static void set_next_buddy(struct sched_entity *se)
 {
-        for_each_sched_entity(se)
-                cfs_rq_of(se)->next = se;
+        if (likely(task_of(se)->policy != SCHED_IDLE)) {
+                for_each_sched_entity(se)
+                        cfs_rq_of(se)->next = se;
+        }
 }
 
 /*
@@ -1393,12 +1404,18 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int sync)
                 return;
 
         /*
-         * Batch tasks do not preempt (their preemption is driven by
+         * Batch and idle tasks do not preempt (their preemption is driven by
          * the tick):
          */
-        if (unlikely(p->policy == SCHED_BATCH))
+        if (unlikely(p->policy != SCHED_NORMAL))
                 return;
 
+        /* Idle tasks are by definition preempted by everybody. */
+        if (unlikely(curr->policy == SCHED_IDLE)) {
+                resched_task(curr);
+                return;
+        }
+
         if (!sched_feat(WAKEUP_PREEMPT))
                 return;
 
diff --git a/kernel/sched_stats.h b/kernel/sched_stats.h
index f2773b5d1226..8ab0cef8ecab 100644
--- a/kernel/sched_stats.h
+++ b/kernel/sched_stats.h
@@ -296,6 +296,7 @@ sched_info_switch(struct task_struct *prev, struct task_struct *next)
 static inline void account_group_user_time(struct task_struct *tsk,
                                            cputime_t cputime)
 {
+        struct task_cputime *times;
         struct signal_struct *sig;
 
         /* tsk == current, ensure it is safe to use ->signal */
@@ -303,13 +304,11 @@ static inline void account_group_user_time(struct task_struct *tsk,
                 return;
 
         sig = tsk->signal;
-        if (sig->cputime.totals) {
-                struct task_cputime *times;
+        times = &sig->cputime.totals;
 
-                times = per_cpu_ptr(sig->cputime.totals, get_cpu());
-                times->utime = cputime_add(times->utime, cputime);
-                put_cpu_no_resched();
-        }
+        spin_lock(&times->lock);
+        times->utime = cputime_add(times->utime, cputime);
+        spin_unlock(&times->lock);
 }
 
 /**
@@ -325,6 +324,7 @@ static inline void account_group_user_time(struct task_struct *tsk,
 static inline void account_group_system_time(struct task_struct *tsk,
                                              cputime_t cputime)
 {
+        struct task_cputime *times;
         struct signal_struct *sig;
 
         /* tsk == current, ensure it is safe to use ->signal */
@@ -332,13 +332,11 @@ static inline void account_group_system_time(struct task_struct *tsk,
                 return;
 
         sig = tsk->signal;
-        if (sig->cputime.totals) {
-                struct task_cputime *times;
+        times = &sig->cputime.totals;
 
-                times = per_cpu_ptr(sig->cputime.totals, get_cpu());
-                times->stime = cputime_add(times->stime, cputime);
-                put_cpu_no_resched();
-        }
+        spin_lock(&times->lock);
+        times->stime = cputime_add(times->stime, cputime);
+        spin_unlock(&times->lock);
 }
 
 /**
@@ -354,6 +352,7 @@ static inline void account_group_system_time(struct task_struct *tsk,
 static inline void account_group_exec_runtime(struct task_struct *tsk,
                                               unsigned long long ns)
 {
+        struct task_cputime *times;
         struct signal_struct *sig;
 
         sig = tsk->signal;
@@ -362,11 +361,9 @@ static inline void account_group_exec_runtime(struct task_struct *tsk,
         if (unlikely(!sig))
                 return;
 
-        if (sig->cputime.totals) {
-                struct task_cputime *times;
+        times = &sig->cputime.totals;
 
-                times = per_cpu_ptr(sig->cputime.totals, get_cpu());
-                times->sum_exec_runtime += ns;
-                put_cpu_no_resched();
-        }
+        spin_lock(&times->lock);
+        times->sum_exec_runtime += ns;
+        spin_unlock(&times->lock);
 }
diff --git a/kernel/softlockup.c b/kernel/softlockup.c
index d9188c66278a..85d5a2455103 100644
--- a/kernel/softlockup.c
+++ b/kernel/softlockup.c
@@ -16,6 +16,7 @@
 #include <linux/lockdep.h>
 #include <linux/notifier.h>
 #include <linux/module.h>
+#include <linux/sysctl.h>
 
 #include <asm/irq_regs.h>
 
@@ -88,6 +89,14 @@ void touch_all_softlockup_watchdogs(void)
 }
 EXPORT_SYMBOL(touch_all_softlockup_watchdogs);
 
+int proc_dosoftlockup_thresh(struct ctl_table *table, int write,
+                             struct file *filp, void __user *buffer,
+                             size_t *lenp, loff_t *ppos)
+{
+        touch_all_softlockup_watchdogs();
+        return proc_dointvec_minmax(table, write, filp, buffer, lenp, ppos);
+}
+
 /*
  * This callback runs from the timer interrupt, and checks
  * whether the watchdog thread has hung or not:
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index 3e38b74b6124..790f9d785663 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -144,6 +144,7 @@ extern int acct_parm[];
 
 #ifdef CONFIG_IA64
 extern int no_unaligned_warning;
+extern int unaligned_dump_stack;
 #endif
 
 #ifdef CONFIG_RT_MUTEXES
@@ -781,6 +782,14 @@ static struct ctl_table kern_table[] = {
                 .mode           = 0644,
                 .proc_handler   = &proc_dointvec,
         },
+        {
+                .ctl_name       = CTL_UNNUMBERED,
+                .procname       = "unaligned-dump-stack",
+                .data           = &unaligned_dump_stack,
+                .maxlen         = sizeof (int),
+                .mode           = 0644,
+                .proc_handler   = &proc_dointvec,
+        },
 #endif
 #ifdef CONFIG_DETECT_SOFTLOCKUP
         {
@@ -800,7 +809,7 @@ static struct ctl_table kern_table[] = {
                 .data           = &softlockup_thresh,
                 .maxlen         = sizeof(int),
                 .mode           = 0644,
-                .proc_handler   = &proc_dointvec_minmax,
+                .proc_handler   = &proc_dosoftlockup_thresh,
                 .strategy       = &sysctl_intvec,
                 .extra1         = &neg_one,
                 .extra2         = &sixty,
diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
index 1b6c05bd0d0a..d3f1ef4d5cbe 100644
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@ -134,7 +134,7 @@ __setup("nohz=", setup_tick_nohz);
  * value. We do this unconditionally on any cpu, as we don't know whether the
  * cpu, which has the update task assigned is in a long sleep.
  */
-void tick_nohz_update_jiffies(void)
+static void tick_nohz_update_jiffies(void)
 {
         int cpu = smp_processor_id();
         struct tick_sched *ts = &per_cpu(tick_cpu_sched, cpu);
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 2f445833ae37..1f0c509b40d3 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -971,6 +971,8 @@ undo:
 }
 
 #ifdef CONFIG_SMP
+static struct workqueue_struct *work_on_cpu_wq __read_mostly;
+
 struct work_for_cpu {
         struct work_struct work;
         long (*fn)(void *);
@@ -991,8 +993,8 @@ static void do_work_for_cpu(struct work_struct *w)
  * @fn: the function to run
  * @arg: the function arg
  *
- * This will return -EINVAL in the cpu is not online, or the return value
- * of @fn otherwise.
+ * This will return the value @fn returns.
+ * It is up to the caller to ensure that the cpu doesn't go offline.
  */
 long work_on_cpu(unsigned int cpu, long (*fn)(void *), void *arg)
 {
@@ -1001,14 +1003,8 @@ long work_on_cpu(unsigned int cpu, long (*fn)(void *), void *arg)
         INIT_WORK(&wfc.work, do_work_for_cpu);
         wfc.fn = fn;
         wfc.arg = arg;
-        get_online_cpus();
-        if (unlikely(!cpu_online(cpu)))
-                wfc.ret = -EINVAL;
-        else {
-                schedule_work_on(cpu, &wfc.work);
-                flush_work(&wfc.work);
-        }
-        put_online_cpus();
+        queue_work_on(cpu, work_on_cpu_wq, &wfc.work);
+        flush_work(&wfc.work);
 
         return wfc.ret;
 }
@@ -1025,4 +1021,8 @@ void __init init_workqueues(void)
         hotcpu_notifier(workqueue_cpu_callback, 0);
         keventd_wq = create_workqueue("events");
         BUG_ON(!keventd_wq);
+#ifdef CONFIG_SMP
+        work_on_cpu_wq = create_workqueue("work_on_cpu");
+        BUG_ON(!work_on_cpu_wq);
+#endif
 }