Merge branch 'linus' into stackprotector

15 files changed, 475 insertions, 723 deletions

diff --git a/kernel/capability.c b/kernel/capability.c
index 39e8193b41e..cfbe4429948 100644
--- a/kernel/capability.c
+++ b/kernel/capability.c
@@ -53,6 +53,69 @@ static void warn_legacy_capability_use(void)
 }
 
 /*
+ * Version 2 capabilities worked fine, but the linux/capability.h file
+ * that accompanied their introduction encouraged their use without
+ * the necessary user-space source code changes. As such, we have
+ * created a version 3 with equivalent functionality to version 2, but
+ * with a header change to protect legacy source code from using
+ * version 2 when it wanted to use version 1. If your system has code
+ * that trips the following warning, it is using version 2 specific
+ * capabilities and may be doing so insecurely.
+ *
+ * The remedy is to either upgrade your version of libcap (to 2.10+,
+ * if the application is linked against it), or recompile your
+ * application with modern kernel headers and this warning will go
+ * away.
+ */
+
+static void warn_deprecated_v2(void)
+{
+	static int warned;
+
+	if (!warned) {
+		char name[sizeof(current->comm)];
+
+		printk(KERN_INFO "warning: `%s' uses deprecated v2"
+		       " capabilities in a way that may be insecure.\n",
+		       get_task_comm(name, current));
+		warned = 1;
+	}
+}
+
+/*
+ * Version check. Return the number of u32s in each capability flag
+ * array, or a negative value on error.
+ */
+static int cap_validate_magic(cap_user_header_t header, unsigned *tocopy)
+{
+	__u32 version;
+
+	if (get_user(version, &header->version))
+		return -EFAULT;
+
+	switch (version) {
+	case _LINUX_CAPABILITY_VERSION_1:
+		warn_legacy_capability_use();
+		*tocopy = _LINUX_CAPABILITY_U32S_1;
+		break;
+	case _LINUX_CAPABILITY_VERSION_2:
+		warn_deprecated_v2();
+		/*
+		 * fall through - v3 is otherwise equivalent to v2.
+		 */
+	case _LINUX_CAPABILITY_VERSION_3:
+		*tocopy = _LINUX_CAPABILITY_U32S_3;
+		break;
+	default:
+		if (put_user((u32)_KERNEL_CAPABILITY_VERSION, &header->version))
+			return -EFAULT;
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+/*
  * For sys_getproccap() and sys_setproccap(), any of the three
  * capability set pointers may be NULL -- indicating that that set is
  * uninteresting and/or not to be changed.
@@ -71,27 +134,13 @@ asmlinkage long sys_capget(cap_user_header_t header, cap_user_data_t dataptr)
 {
 	int ret = 0;
 	pid_t pid;
-	__u32 version;
 	struct task_struct *target;
 	unsigned tocopy;
 	kernel_cap_t pE, pI, pP;
 
-	if (get_user(version, &header->version))
-		return -EFAULT;
-
-	switch (version) {
-	case _LINUX_CAPABILITY_VERSION_1:
-		warn_legacy_capability_use();
-		tocopy = _LINUX_CAPABILITY_U32S_1;
-		break;
-	case _LINUX_CAPABILITY_VERSION_2:
-		tocopy = _LINUX_CAPABILITY_U32S_2;
-		break;
-	default:
-		if (put_user(_LINUX_CAPABILITY_VERSION, &header->version))
-			return -EFAULT;
-		return -EINVAL;
-	}
+	ret = cap_validate_magic(header, &tocopy);
+	if (ret != 0)
+		return ret;
 
 	if (get_user(pid, &header->pid))
 		return -EFAULT;
@@ -118,7 +167,7 @@ out:
 	spin_unlock(&task_capability_lock);
 
 	if (!ret) {
-		struct __user_cap_data_struct kdata[_LINUX_CAPABILITY_U32S];
+		struct __user_cap_data_struct kdata[_KERNEL_CAPABILITY_U32S];
 		unsigned i;
 
 		for (i = 0; i < tocopy; i++) {
@@ -128,7 +177,7 @@ out:
 		}
 
 		/*
-		 * Note, in the case, tocopy < _LINUX_CAPABILITY_U32S,
+		 * Note, in the case, tocopy < _KERNEL_CAPABILITY_U32S,
 		 * we silently drop the upper capabilities here. This
 		 * has the effect of making older libcap
 		 * implementations implicitly drop upper capability
@@ -240,30 +289,16 @@ static inline int cap_set_all(kernel_cap_t *effective,
  */
 asmlinkage long sys_capset(cap_user_header_t header, const cap_user_data_t data)
 {
-	struct __user_cap_data_struct kdata[_LINUX_CAPABILITY_U32S];
+	struct __user_cap_data_struct kdata[_KERNEL_CAPABILITY_U32S];
 	unsigned i, tocopy;
 	kernel_cap_t inheritable, permitted, effective;
-	__u32 version;
 	struct task_struct *target;
 	int ret;
 	pid_t pid;
 
-	if (get_user(version, &header->version))
-		return -EFAULT;
-
-	switch (version) {
-	case _LINUX_CAPABILITY_VERSION_1:
-		warn_legacy_capability_use();
-		tocopy = _LINUX_CAPABILITY_U32S_1;
-		break;
-	case _LINUX_CAPABILITY_VERSION_2:
-		tocopy = _LINUX_CAPABILITY_U32S_2;
-		break;
-	default:
-		if (put_user(_LINUX_CAPABILITY_VERSION, &header->version))
-			return -EFAULT;
-		return -EINVAL;
-	}
+	ret = cap_validate_magic(header, &tocopy);
+	if (ret != 0)
+		return ret;
 
 	if (get_user(pid, &header->pid))
 		return -EFAULT;
@@ -281,7 +316,7 @@ asmlinkage long sys_capset(cap_user_header_t header, const cap_user_data_t data)
 		permitted.cap[i] = kdata[i].permitted;
 		inheritable.cap[i] = kdata[i].inheritable;
 	}
-	while (i < _LINUX_CAPABILITY_U32S) {
+	while (i < _KERNEL_CAPABILITY_U32S) {
 		effective.cap[i] = 0;
 		permitted.cap[i] = 0;
 		inheritable.cap[i] = 0;
diff --git a/kernel/cpuset.c b/kernel/cpuset.c
index 86ea9e34e32..9fceb97e989 100644
--- a/kernel/cpuset.c
+++ b/kernel/cpuset.c
@@ -797,8 +797,10 @@ static int update_cpumask(struct cpuset *cs, char *buf)
 		retval = cpulist_parse(buf, trialcs.cpus_allowed);
 		if (retval < 0)
 			return retval;
+
+		if (!cpus_subset(trialcs.cpus_allowed, cpu_online_map))
+			return -EINVAL;
 	}
-	cpus_and(trialcs.cpus_allowed, trialcs.cpus_allowed, cpu_online_map);
 	retval = validate_change(cs, &trialcs);
 	if (retval < 0)
 		return retval;
@@ -932,9 +934,11 @@ static int update_nodemask(struct cpuset *cs, char *buf)
 		retval = nodelist_parse(buf, trialcs.mems_allowed);
 		if (retval < 0)
 			goto done;
+
+		if (!nodes_subset(trialcs.mems_allowed,
+				node_states[N_HIGH_MEMORY]))
+			return -EINVAL;
 	}
-	nodes_and(trialcs.mems_allowed, trialcs.mems_allowed,
-						node_states[N_HIGH_MEMORY]);
 	oldmem = cs->mems_allowed;
 	if (nodes_equal(oldmem, trialcs.mems_allowed)) {
 		retval = 0;		/* Too easy - nothing to do */
@@ -1033,8 +1037,8 @@ int current_cpuset_is_being_rebound(void)
 
 static int update_relax_domain_level(struct cpuset *cs, s64 val)
 {
-	if ((int)val < 0)
-		val = -1;
+	if (val < -1 || val >= SD_LV_MAX)
+		return -EINVAL;
 
 	if (val != cs->relax_domain_level) {
 		cs->relax_domain_level = val;
@@ -1886,6 +1890,12 @@ static void common_cpu_mem_hotplug_unplug(void)
 	top_cpuset.mems_allowed = node_states[N_HIGH_MEMORY];
 	scan_for_empty_cpusets(&top_cpuset);
 
+	/*
+	 * Scheduler destroys domains on hotplug events.
+	 * Rebuild them based on the current settings.
+	 */
+	rebuild_sched_domains();
+
 	cgroup_unlock();
 }
 
diff --git a/kernel/futex.c b/kernel/futex.c
index 449def8074f..7d1136e97c1 100644
--- a/kernel/futex.c
+++ b/kernel/futex.c
@@ -1096,21 +1096,64 @@ static void unqueue_me_pi(struct futex_q *q)
  * private futexes.
  */
 static int fixup_pi_state_owner(u32 __user *uaddr, struct futex_q *q,
-				struct task_struct *newowner)
+				struct task_struct *newowner,
+				struct rw_semaphore *fshared)
 {
 	u32 newtid = task_pid_vnr(newowner) | FUTEX_WAITERS;
 	struct futex_pi_state *pi_state = q->pi_state;
+	struct task_struct *oldowner = pi_state->owner;
 	u32 uval, curval, newval;
-	int ret;
+	int ret, attempt = 0;
 
 	/* Owner died? */
+	if (!pi_state->owner)
+		newtid |= FUTEX_OWNER_DIED;
+
+	/*
+	 * We are here either because we stole the rtmutex from the
+	 * pending owner or we are the pending owner which failed to
+	 * get the rtmutex. We have to replace the pending owner TID
+	 * in the user space variable. This must be atomic as we have
+	 * to preserve the owner died bit here.
+	 *
+	 * Note: We write the user space value _before_ changing the
+	 * pi_state because we can fault here. Imagine swapped out
+	 * pages or a fork, which was running right before we acquired
+	 * mmap_sem, that marked all the anonymous memory readonly for
+	 * cow.
+	 *
+	 * Modifying pi_state _before_ the user space value would
+	 * leave the pi_state in an inconsistent state when we fault
+	 * here, because we need to drop the hash bucket lock to
+	 * handle the fault. This might be observed in the PID check
+	 * in lookup_pi_state.
+	 */
+retry:
+	if (get_futex_value_locked(&uval, uaddr))
+		goto handle_fault;
+
+	while (1) {
+		newval = (uval & FUTEX_OWNER_DIED) | newtid;
+
+		curval = cmpxchg_futex_value_locked(uaddr, uval, newval);
+
+		if (curval == -EFAULT)
+			goto handle_fault;
+		if (curval == uval)
+			break;
+		uval = curval;
+	}
+
+	/*
+	 * We fixed up user space. Now we need to fix the pi_state
+	 * itself.
+	 */
 	if (pi_state->owner != NULL) {
 		spin_lock_irq(&pi_state->owner->pi_lock);
 		WARN_ON(list_empty(&pi_state->list));
 		list_del_init(&pi_state->list);
 		spin_unlock_irq(&pi_state->owner->pi_lock);
-	} else
-		newtid |= FUTEX_OWNER_DIED;
+	}
 
 	pi_state->owner = newowner;
 
@@ -1118,26 +1161,35 @@ static int fixup_pi_state_owner(u32 __user *uaddr, struct futex_q *q,
 	WARN_ON(!list_empty(&pi_state->list));
 	list_add(&pi_state->list, &newowner->pi_state_list);
 	spin_unlock_irq(&newowner->pi_lock);
+	return 0;
 
 	/*
-	 * We own it, so we have to replace the pending owner
-	 * TID. This must be atomic as we have preserve the
-	 * owner died bit here.
+	 * To handle the page fault we need to drop the hash bucket
+	 * lock here. That gives the other task (either the pending
+	 * owner itself or the task which stole the rtmutex) the
+	 * chance to try the fixup of the pi_state. So once we are
+	 * back from handling the fault we need to check the pi_state
+	 * after reacquiring the hash bucket lock and before trying to
+	 * do another fixup. When the fixup has been done already we
+	 * simply return.
 	 */
-	ret = get_futex_value_locked(&uval, uaddr);
+handle_fault:
+	spin_unlock(q->lock_ptr);
 
-	while (!ret) {
-		newval = (uval & FUTEX_OWNER_DIED) | newtid;
+	ret = futex_handle_fault((unsigned long)uaddr, fshared, attempt++);
 
-		curval = cmpxchg_futex_value_locked(uaddr, uval, newval);
+	spin_lock(q->lock_ptr);
 
-		if (curval == -EFAULT)
-			ret = -EFAULT;
-		if (curval == uval)
-			break;
-		uval = curval;
-	}
-	return ret;
+	/*
+	 * Check if someone else fixed it for us:
+	 */
+	if (pi_state->owner != oldowner)
+		return 0;
+
+	if (ret)
+		return ret;
+
+	goto retry;
 }
 
 /*
@@ -1507,7 +1559,7 @@ static int futex_lock_pi(u32 __user *uaddr, struct rw_semaphore *fshared,
 		 * that case:
 		 */
 		if (q.pi_state->owner != curr)
-			ret = fixup_pi_state_owner(uaddr, &q, curr);
+			ret = fixup_pi_state_owner(uaddr, &q, curr, fshared);
 	} else {
 		/*
 		 * Catch the rare case, where the lock was released
@@ -1539,7 +1591,8 @@ static int futex_lock_pi(u32 __user *uaddr, struct rw_semaphore *fshared,
 				int res;
 
 				owner = rt_mutex_owner(&q.pi_state->pi_mutex);
-				res = fixup_pi_state_owner(uaddr, &q, owner);
+				res = fixup_pi_state_owner(uaddr, &q, owner,
+							   fshared);
 
 				/* propagate -EFAULT, if the fixup failed */
 				if (res)
diff --git a/kernel/kgdb.c b/kernel/kgdb.c
index 14787de568b..3ec23c3ec97 100644
--- a/kernel/kgdb.c
+++ b/kernel/kgdb.c
@@ -52,6 +52,7 @@
 #include <asm/byteorder.h>
 #include <asm/atomic.h>
 #include <asm/system.h>
+#include <asm/unaligned.h>
 
 static int kgdb_break_asap;
 
@@ -227,8 +228,6 @@ void __weak kgdb_disable_hw_debug(struct pt_regs *regs)
  * GDB remote protocol parser:
  */
 
-static const char	hexchars[] = "0123456789abcdef";
-
 static int hex(char ch)
 {
 	if ((ch >= 'a') && (ch <= 'f'))
@@ -316,8 +315,8 @@ static void put_packet(char *buffer)
 		}
 
 		kgdb_io_ops->write_char('#');
-		kgdb_io_ops->write_char(hexchars[checksum >> 4]);
-		kgdb_io_ops->write_char(hexchars[checksum & 0xf]);
+		kgdb_io_ops->write_char(hex_asc_hi(checksum));
+		kgdb_io_ops->write_char(hex_asc_lo(checksum));
 		if (kgdb_io_ops->flush)
 			kgdb_io_ops->flush();
 
@@ -478,8 +477,8 @@ static void error_packet(char *pkt, int error)
 {
 	error = -error;
 	pkt[0] = 'E';
-	pkt[1] = hexchars[(error / 10)];
-	pkt[2] = hexchars[(error % 10)];
+	pkt[1] = hex_asc[(error / 10)];
+	pkt[2] = hex_asc[(error % 10)];
 	pkt[3] = '\0';
 }
 
@@ -510,10 +509,7 @@ static void int_to_threadref(unsigned char *id, int value)
 	scan = (unsigned char *)id;
 	while (i--)
 		*scan++ = 0;
-	*scan++ = (value >> 24) & 0xff;
-	*scan++ = (value >> 16) & 0xff;
-	*scan++ = (value >> 8) & 0xff;
-	*scan++ = (value & 0xff);
+	put_unaligned_be32(value, scan);
 }
 
 static struct task_struct *getthread(struct pt_regs *regs, int tid)
@@ -1503,7 +1499,8 @@ int kgdb_nmicallback(int cpu, void *regs)
 	return 1;
 }
 
-void kgdb_console_write(struct console *co, const char *s, unsigned count)
+static void kgdb_console_write(struct console *co, const char *s,
+   unsigned count)
 {
 	unsigned long flags;
 
diff --git a/kernel/kprobes.c b/kernel/kprobes.c
index 1e0250cb948..d4998f81e22 100644
--- a/kernel/kprobes.c
+++ b/kernel/kprobes.c
@@ -699,8 +699,9 @@ static int __register_kprobes(struct kprobe **kps, int num,
 		return -EINVAL;
 	for (i = 0; i < num; i++) {
 		ret = __register_kprobe(kps[i], called_from);
-		if (ret < 0 && i > 0) {
-			unregister_kprobes(kps, i);
+		if (ret < 0) {
+			if (i > 0)
+				unregister_kprobes(kps, i);
 			break;
 		}
 	}
@@ -776,8 +777,9 @@ static int __register_jprobes(struct jprobe **jps, int num,
 			jp->kp.break_handler = longjmp_break_handler;
 			ret = __register_kprobe(&jp->kp, called_from);
 		}
-		if (ret < 0 && i > 0) {
-			unregister_jprobes(jps, i);
+		if (ret < 0) {
+			if (i > 0)
+				unregister_jprobes(jps, i);
 			break;
 		}
 	}
@@ -920,8 +922,9 @@ static int __register_kretprobes(struct kretprobe **rps, int num,
 		return -EINVAL;
 	for (i = 0; i < num; i++) {
 		ret = __register_kretprobe(rps[i], called_from);
-		if (ret < 0 && i > 0) {
-			unregister_kretprobes(rps, i);
+		if (ret < 0) {
+			if (i > 0)
+				unregister_kretprobes(rps, i);
 			break;
 		}
 	}
diff --git a/kernel/rcupreempt.c b/kernel/rcupreempt.c
index e1cdf196a51..5e02b774070 100644
--- a/kernel/rcupreempt.c
+++ b/kernel/rcupreempt.c
@@ -217,8 +217,6 @@ long rcu_batches_completed(void)
 }
 EXPORT_SYMBOL_GPL(rcu_batches_completed);
 
-EXPORT_SYMBOL_GPL(rcu_batches_completed_bh);
-
 void __rcu_read_lock(void)
 {
 	int idx;
diff --git a/kernel/relay.c b/kernel/relay.c
index bc24dcdc570..7de644cdec4 100644
--- a/kernel/relay.c
+++ b/kernel/relay.c
@@ -1191,7 +1191,7 @@ static ssize_t relay_file_splice_read(struct file *in,
 	ret = 0;
 	spliced = 0;
 
-	while (len) {
+	while (len && !spliced) {
 		ret = subbuf_splice_actor(in, ppos, pipe, len, flags, &nonpad_ret);
 		if (ret < 0)
 			break;
diff --git a/kernel/sched.c b/kernel/sched.c
index a964ed94509..0cdb50260db 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -136,7 +136,7 @@ static inline void sg_inc_cpu_power(struct sched_group *sg, u32 val)
 
 static inline int rt_policy(int policy)
 {
-	if (unlikely(policy == SCHED_FIFO) || unlikely(policy == SCHED_RR))
+	if (unlikely(policy == SCHED_FIFO || policy == SCHED_RR))
 		return 1;
 	return 0;
 }
@@ -312,12 +312,15 @@ static DEFINE_SPINLOCK(task_group_lock);
 #endif
 
 /*
- * A weight of 0, 1 or ULONG_MAX can cause arithmetics problems.
+ * A weight of 0 or 1 can cause arithmetics problems.
+ * A weight of a cfs_rq is the sum of weights of which entities
+ * are queued on this cfs_rq, so a weight of a entity should not be
+ * too large, so as the shares value of a task group.
  * (The default weight is 1024 - so there's no practical
  *  limitation from this.)
  */
 #define MIN_SHARES	2
-#define MAX_SHARES	(ULONG_MAX - 1)
+#define MAX_SHARES	(1UL << 18)
 
 static int init_task_group_load = INIT_TASK_GROUP_LOAD;
 #endif
@@ -398,43 +401,6 @@ struct cfs_rq {
 	 */
 	struct list_head leaf_cfs_rq_list;
 	struct task_group *tg;	/* group that "owns" this runqueue */
-
-#ifdef CONFIG_SMP
-	unsigned long task_weight;
-	unsigned long shares;
-	/*
-	 * We need space to build a sched_domain wide view of the full task
-	 * group tree, in order to avoid depending on dynamic memory allocation
-	 * during the load balancing we place this in the per cpu task group
-	 * hierarchy. This limits the load balancing to one instance per cpu,
-	 * but more should not be needed anyway.
-	 */
-	struct aggregate_struct {
-		/*
-		 *   load = weight(cpus) * f(tg)
-		 *
-		 * Where f(tg) is the recursive weight fraction assigned to
-		 * this group.
-		 */
-		unsigned long load;
-
-		/*
-		 * part of the group weight distributed to this span.
-		 */
-		unsigned long shares;
-
-		/*
-		 * The sum of all runqueue weights within this span.
-		 */
-		unsigned long rq_weight;
-
-		/*
-		 * Weight contributed by tasks; this is the part we can
-		 * influence by moving tasks around.
-		 */
-		unsigned long task_weight;
-	} aggregate;
-#endif
 #endif
 };
 
@@ -1161,6 +1127,7 @@ static enum hrtimer_restart hrtick(struct hrtimer *timer)
 	return HRTIMER_NORESTART;
 }
 
+#ifdef CONFIG_SMP
 static void hotplug_hrtick_disable(int cpu)
 {
 	struct rq *rq = cpu_rq(cpu);
@@ -1216,6 +1183,7 @@ static void init_hrtick(void)
 {
 	hotcpu_notifier(hotplug_hrtick, 0);
 }
+#endif /* CONFIG_SMP */
 
 static void init_rq_hrtick(struct rq *rq)
 {
@@ -1368,17 +1336,19 @@ static void __resched_task(struct task_struct *p, int tif_bit)
  */
 #define SRR(x, y) (((x) + (1UL << ((y) - 1))) >> (y))
 
-/*
- * delta *= weight / lw
- */
 static unsigned long
 calc_delta_mine(unsigned long delta_exec, unsigned long weight,
 		struct load_weight *lw)
 {
 	u64 tmp;
 
-	if (!lw->inv_weight)
-		lw->inv_weight = 1 + (WMULT_CONST-lw->weight/2)/(lw->weight+1);
+	if (!lw->inv_weight) {
+		if (BITS_PER_LONG > 32 && unlikely(lw->weight >= WMULT_CONST))
+			lw->inv_weight = 1;
+		else
+			lw->inv_weight = 1 + (WMULT_CONST-lw->weight/2)
+				/ (lw->weight+1);
+	}
 
 	tmp = (u64)delta_exec * weight;
 	/*
@@ -1393,6 +1363,12 @@ calc_delta_mine(unsigned long delta_exec, unsigned long weight,
 	return (unsigned long)min(tmp, (u64)(unsigned long)LONG_MAX);
 }
 
+static inline unsigned long
+calc_delta_fair(unsigned long delta_exec, struct load_weight *lw)
+{
+	return calc_delta_mine(delta_exec, NICE_0_LOAD, lw);
+}
+
 static inline void update_load_add(struct load_weight *lw, unsigned long inc)
 {
 	lw->weight += inc;
@@ -1505,326 +1481,6 @@ static unsigned long source_load(int cpu, int type);
 static unsigned long target_load(int cpu, int type);
 static unsigned long cpu_avg_load_per_task(int cpu);
 static int task_hot(struct task_struct *p, u64 now, struct sched_domain *sd);
-
-#ifdef CONFIG_FAIR_GROUP_SCHED
-
-/*
- * Group load balancing.
- *
- * We calculate a few balance domain wide aggregate numbers; load and weight.
- * Given the pictures below, and assuming each item has equal weight:
- *
- *         root          1 - thread
- *         / | \         A - group
- *        A  1  B
- *       /|\   / \
- *      C 2 D 3   4
- *      |   |
- *      5   6
- *
- * load:
- *    A and B get 1/3-rd of the total load. C and D get 1/3-rd of A's 1/3-rd,
- *    which equals 1/9-th of the total load.
- *
- * shares:
- *    The weight of this group on the selected cpus.
- *
- * rq_weight:
- *    Direct sum of all the cpu's their rq weight, e.g. A would get 3 while
- *    B would get 2.
- *
- * task_weight:
- *    Part of the rq_weight contributed by tasks; all groups except B would
- *    get 1, B gets 2.
- */
-
-static inline struct aggregate_struct *
-aggregate(struct task_group *tg, struct sched_domain *sd)
-{
-	return &tg->cfs_rq[sd->first_cpu]->aggregate;
-}
-
-typedef void (*aggregate_func)(struct task_group *, struct sched_domain *);
-
-/*
- * Iterate the full tree, calling @down when first entering a node and @up when
- * leaving it for the final time.
- */
-static
-void aggregate_walk_tree(aggregate_func down, aggregate_func up,
-			 struct sched_domain *sd)
-{
-	struct task_group *parent, *child;
-
-	rcu_read_lock();
-	parent = &root_task_group;
-down:
-	(*down)(parent, sd);
-	list_for_each_entry_rcu(child, &parent->children, siblings) {
-		parent = child;
-		goto down;
-
-up:
-		continue;
-	}
-	(*up)(parent, sd);
-
-	child = parent;
-	parent = parent->parent;
-	if (parent)
-		goto up;
-	rcu_read_unlock();
-}
-
-/*
- * Calculate the aggregate runqueue weight.
- */
-static
-void aggregate_group_weight(struct task_group *tg, struct sched_domain *sd)
-{
-	unsigned long rq_weight = 0;
-	unsigned long task_weight = 0;
-	int i;
-
-	for_each_cpu_mask(i, sd->span) {
-		rq_weight += tg->cfs_rq[i]->load.weight;
-		task_weight += tg->cfs_rq[i]->task_weight;
-	}
-
-	aggregate(tg, sd)->rq_weight = rq_weight;
-	aggregate(tg, sd)->task_weight = task_weight;
-}
-
-/*
- * Compute the weight of this group on the given cpus.
- */
-static
-void aggregate_group_shares(struct task_group *tg, struct sched_domain *sd)
-{
-	unsigned long shares = 0;
-	int i;
-
-	for_each_cpu_mask(i, sd->span)
-		shares += tg->cfs_rq[i]->shares;
-
-	if ((!shares && aggregate(tg, sd)->rq_weight) || shares > tg->shares)
-		shares = tg->shares;
-
-	aggregate(tg, sd)->shares = shares;
-}
-
-/*
- * Compute the load fraction assigned to this group, relies on the aggregate
- * weight and this group's parent's load, i.e. top-down.
- */
-static
-void aggregate_group_load(struct task_group *tg, struct sched_domain *sd)
-{
-	unsigned long load;
-
-	if (!tg->parent) {
-		int i;
-
-		load = 0;
-		for_each_cpu_mask(i, sd->span)
-			load += cpu_rq(i)->load.weight;
-
-	} else {
-		load = aggregate(tg->parent, sd)->load;
-
-		/*
-		 * shares is our weight in the parent's rq so
-		 * shares/parent->rq_weight gives our fraction of the load
-		 */
-		load *= aggregate(tg, sd)->shares;
-		load /= aggregate(tg->parent, sd)->rq_weight + 1;
-	}
-
-	aggregate(tg, sd)->load = load;
-}
-
-static void __set_se_shares(struct sched_entity *se, unsigned long shares);
-
-/*
- * Calculate and set the cpu's group shares.
- */
-static void
-__update_group_shares_cpu(struct task_group *tg, struct sched_domain *sd,
-			  int tcpu)
-{
-	int boost = 0;
-	unsigned long shares;
-	unsigned long rq_weight;
-
-	if (!tg->se[tcpu])
-		return;
-
-	rq_weight = tg->cfs_rq[tcpu]->load.weight;
-
-	/*
-	 * If there are currently no tasks on the cpu pretend there is one of
-	 * average load so that when a new task gets to run here it will not
-	 * get delayed by group starvation.
-	 */
-	if (!rq_weight) {
-		boost = 1;
-		rq_weight = NICE_0_LOAD;
-	}
-
-	/*
-	 *           \Sum shares * rq_weight
-	 * shares =  -----------------------
-	 *               \Sum rq_weight
-	 *
-	 */
-	shares = aggregate(tg, sd)->shares * rq_weight;
-	shares /= aggregate(tg, sd)->rq_weight + 1;
-
-	/*
-	 * record the actual number of shares, not the boosted amount.
-	 */
-	tg->cfs_rq[tcpu]->shares = boost ? 0 : shares;
-
-	if (shares < MIN_SHARES)
-		shares = MIN_SHARES;
-	else if (shares > MAX_SHARES)
-		shares = MAX_SHARES;
-
-	__set_se_shares(tg->se[tcpu], shares);
-}
-
-/*
- * Re-adjust the weights on the cpu the task came from and on the cpu the
- * task went to.
- */
-static void
-__move_group_shares(struct task_group *tg, struct sched_domain *sd,
-		    int scpu, int dcpu)
-{
-	unsigned long shares;
-
-	shares = tg->cfs_rq[scpu]->shares + tg->cfs_rq[dcpu]->shares;
-
-	__update_group_shares_cpu(tg, sd, scpu);
-	__update_group_shares_cpu(tg, sd, dcpu);
-
-	/*
-	 * ensure we never loose shares due to rounding errors in the
-	 * above redistribution.
-	 */
-	shares -= tg->cfs_rq[scpu]->shares + tg->cfs_rq[dcpu]->shares;
-	if (shares)
-		tg->cfs_rq[dcpu]->shares += shares;
-}
-
-/*
- * Because changing a group's shares changes the weight of the super-group
- * we need to walk up the tree and change all shares until we hit the root.
- */
-static void
-move_group_shares(struct task_group *tg, struct sched_domain *sd,
-		  int scpu, int dcpu)
-{
-	while (tg) {
-		__move_group_shares(tg, sd, scpu, dcpu);
-		tg = tg->parent;
-	}
-}
-
-static
-void aggregate_group_set_shares(struct task_group *tg, struct sched_domain *sd)
-{
-	unsigned long shares = aggregate(tg, sd)->shares;
-	int i;
-
-	for_each_cpu_mask(i, sd->span) {
-		struct rq *rq = cpu_rq(i);
-		unsigned long flags;
-
-		spin_lock_irqsave(&rq->lock, flags);
-		__update_group_shares_cpu(tg, sd, i);
-		spin_unlock_irqrestore(&rq->lock, flags);
-	}
-
-	aggregate_group_shares(tg, sd);
-
-	/*
-	 * ensure we never loose shares due to rounding errors in the
-	 * above redistribution.
-	 */
-	shares -= aggregate(tg, sd)->shares;
-	if (shares) {
-		tg->cfs_rq[sd->first_cpu]->shares += shares;
-		aggregate(tg, sd)->shares += shares;
-	}
-}
-
-/*
- * Calculate the accumulative weight and recursive load of each task group
- * while walking down the tree.
- */
-static
-void aggregate_get_down(struct task_group *tg, struct sched_domain *sd)
-{
-	aggregate_group_weight(tg, sd);
-	aggregate_group_shares(tg, sd);
-	aggregate_group_load(tg, sd);
-}
-
-/*
- * Rebalance the cpu shares while walking back up the tree.
- */
-static
-void aggregate_get_up(struct task_group *tg, struct sched_domain *sd)
-{
-	aggregate_group_set_shares(tg, sd);
-}
-
-static DEFINE_PER_CPU(spinlock_t, aggregate_lock);
-
-static void __init init_aggregate(void)
-{
-	int i;
-
-	for_each_possible_cpu(i)
-		spin_lock_init(&per_cpu(aggregate_lock, i));
-}
-
-static int get_aggregate(struct sched_domain *sd)
-{
-	if (!spin_trylock(&per_cpu(aggregate_lock, sd->first_cpu)))
-		return 0;
-
-	aggregate_walk_tree(aggregate_get_down, aggregate_get_up, sd);
-	return 1;
-}
-
-static void put_aggregate(struct sched_domain *sd)
-{
-	spin_unlock(&per_cpu(aggregate_lock, sd->first_cpu));
-}
-
-static void cfs_rq_set_shares(struct cfs_rq *cfs_rq, unsigned long shares)
-{
-	cfs_rq->shares = shares;
-}
-
-#else
-
-static inline void init_aggregate(void)
-{
-}
-
-static inline int get_aggregate(struct sched_domain *sd)
-{
-	return 0;
-}
-
-static inline void put_aggregate(struct sched_domain *sd)
-{
-}
-#endif
-
 #else /* CONFIG_SMP */
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
@@ -1845,14 +1501,26 @@ static void cfs_rq_set_shares(struct cfs_rq *cfs_rq, unsigned long shares)
 
 #define sched_class_highest (&rt_sched_class)
 
-static void inc_nr_running(struct rq *rq)
+static inline void inc_load(struct rq *rq, const struct task_struct *p)
+{
+	update_load_add(&rq->load, p->se.load.weight);
+}
+
+static inline void dec_load(struct rq *rq, const struct task_struct *p)
+{
+	update_load_sub(&rq->load, p->se.load.weight);
+}
+
+static void inc_nr_running(struct task_struct *p, struct rq *rq)
 {
 	rq->nr_running++;
+	inc_load(rq, p);
 }
 
-static void dec_nr_running(struct rq *rq)
+static void dec_nr_running(struct task_struct *p, struct rq *rq)
 {
 	rq->nr_running--;
+	dec_load(rq, p);
 }
 
 static void set_load_weight(struct task_struct *p)
@@ -1944,7 +1612,7 @@ static void activate_task(struct rq *rq, struct task_struct *p, int wakeup)
 		rq->nr_uninterruptible--;
 
 	enqueue_task(rq, p, wakeup);
-	inc_nr_running(rq);
+	inc_nr_running(p, rq);
 }
 
 /*
@@ -1956,7 +1624,7 @@ static void deactivate_task(struct rq *rq, struct task_struct *p, int sleep)
 		rq->nr_uninterruptible++;
 
 	dequeue_task(rq, p, sleep);
-	dec_nr_running(rq);
+	dec_nr_running(p, rq);
 }
 
 /**
@@ -2609,7 +2277,7 @@ void wake_up_new_task(struct task_struct *p, unsigned long clone_flags)
 		 * management (if any):
 		 */
 		p->sched_class->task_new(rq, p);
-		inc_nr_running(rq);
+		inc_nr_running(p, rq);
 	}
 	check_preempt_curr(rq, p);
 #ifdef CONFIG_SMP
@@ -3600,12 +3268,9 @@ static int load_balance(int this_cpu, struct rq *this_rq,
 	unsigned long imbalance;
 	struct rq *busiest;
 	unsigned long flags;
-	int unlock_aggregate;
 
 	cpus_setall(*cpus);
 
-	unlock_aggregate = get_aggregate(sd);
-
 	/*
 	 * When power savings policy is enabled for the parent domain, idle
 	 * sibling can pick up load irrespective of busy siblings. In this case,
@@ -3721,9 +3386,8 @@ redo:
 
 	if (!ld_moved && !sd_idle && sd->flags & SD_SHARE_CPUPOWER &&
 	    !test_sd_parent(sd, SD_POWERSAVINGS_BALANCE))
-		ld_moved = -1;
-
-	goto out;
+		return -1;
+	return ld_moved;
 
 out_balanced:
 	schedstat_inc(sd, lb_balanced[idle]);
@@ -3738,13 +3402,8 @@ out_one_pinned:
 
 	if (!sd_idle && sd->flags & SD_SHARE_CPUPOWER &&
 	    !test_sd_parent(sd, SD_POWERSAVINGS_BALANCE))
-		ld_moved = -1;
-	else
-		ld_moved = 0;
-out:
-	if (unlock_aggregate)
-		put_aggregate(sd);
-	return ld_moved;
+		return -1;
+	return 0;
 }
 
 /*
@@ -4430,7 +4089,7 @@ static inline void schedule_debug(struct task_struct