summaryrefslogtreecommitdiff
path: root/kernel
diff options
context:
space:
mode:
Diffstat (limited to 'kernel')
-rw-r--r--kernel/cpuset.c82
-rw-r--r--kernel/exit.c8
-rw-r--r--kernel/fork.c38
-rw-r--r--kernel/lockdep.c203
-rw-r--r--kernel/nsproxy.c4
-rw-r--r--kernel/power/process.c21
-rw-r--r--kernel/relay.c4
-rw-r--r--kernel/sched.c515
-rw-r--r--kernel/signal.c4
-rw-r--r--kernel/sysctl.c71
-rw-r--r--kernel/time/clocksource.c8
-rw-r--r--kernel/timer.c162
-rw-r--r--kernel/tsacct.c9
13 files changed, 725 insertions, 404 deletions
diff --git a/kernel/cpuset.c b/kernel/cpuset.c
index 2c3b4431472b..232aed2b10f9 100644
--- a/kernel/cpuset.c
+++ b/kernel/cpuset.c
@@ -2342,32 +2342,48 @@ static const struct cpuset *nearest_exclusive_ancestor(const struct cpuset *cs)
}
/**
- * cpuset_zone_allowed - Can we allocate memory on zone z's memory node?
+ * cpuset_zone_allowed_softwall - Can we allocate on zone z's memory node?
* @z: is this zone on an allowed node?
- * @gfp_mask: memory allocation flags (we use __GFP_HARDWALL)
+ * @gfp_mask: memory allocation flags
*
- * If we're in interrupt, yes, we can always allocate. If zone
+ * If we're in interrupt, yes, we can always allocate. If
+ * __GFP_THISNODE is set, yes, we can always allocate. If zone
* z's node is in our tasks mems_allowed, yes. If it's not a
* __GFP_HARDWALL request and this zone's nodes is in the nearest
* mem_exclusive cpuset ancestor to this tasks cpuset, yes.
* Otherwise, no.
*
+ * If __GFP_HARDWALL is set, cpuset_zone_allowed_softwall()
+ * reduces to cpuset_zone_allowed_hardwall(). Otherwise,
+ * cpuset_zone_allowed_softwall() might sleep, and might allow a zone
+ * from an enclosing cpuset.
+ *
+ * cpuset_zone_allowed_hardwall() only handles the simpler case of
+ * hardwall cpusets, and never sleeps.
+ *
+ * The __GFP_THISNODE placement logic is really handled elsewhere,
+ * by forcibly using a zonelist starting at a specified node, and by
+ * (in get_page_from_freelist()) refusing to consider the zones for
+ * any node on the zonelist except the first. By the time any such
+ * calls get to this routine, we should just shut up and say 'yes'.
+ *
* GFP_USER allocations are marked with the __GFP_HARDWALL bit,
* and do not allow allocations outside the current tasks cpuset.
* GFP_KERNEL allocations are not so marked, so can escape to the
- * nearest mem_exclusive ancestor cpuset.
+ * nearest enclosing mem_exclusive ancestor cpuset.
*
- * Scanning up parent cpusets requires callback_mutex. The __alloc_pages()
- * routine only calls here with __GFP_HARDWALL bit _not_ set if
- * it's a GFP_KERNEL allocation, and all nodes in the current tasks
- * mems_allowed came up empty on the first pass over the zonelist.
- * So only GFP_KERNEL allocations, if all nodes in the cpuset are
- * short of memory, might require taking the callback_mutex mutex.
+ * Scanning up parent cpusets requires callback_mutex. The
+ * __alloc_pages() routine only calls here with __GFP_HARDWALL bit
+ * _not_ set if it's a GFP_KERNEL allocation, and all nodes in the
+ * current tasks mems_allowed came up empty on the first pass over
+ * the zonelist. So only GFP_KERNEL allocations, if all nodes in the
+ * cpuset are short of memory, might require taking the callback_mutex
+ * mutex.
*
* The first call here from mm/page_alloc:get_page_from_freelist()
- * has __GFP_HARDWALL set in gfp_mask, enforcing hardwall cpusets, so
- * no allocation on a node outside the cpuset is allowed (unless in
- * interrupt, of course).
+ * has __GFP_HARDWALL set in gfp_mask, enforcing hardwall cpusets,
+ * so no allocation on a node outside the cpuset is allowed (unless
+ * in interrupt, of course).
*
* The second pass through get_page_from_freelist() doesn't even call
* here for GFP_ATOMIC calls. For those calls, the __alloc_pages()
@@ -2380,12 +2396,12 @@ static const struct cpuset *nearest_exclusive_ancestor(const struct cpuset *cs)
* GFP_USER - only nodes in current tasks mems allowed ok.
*
* Rule:
- * Don't call cpuset_zone_allowed() if you can't sleep, unless you
+ * Don't call cpuset_zone_allowed_softwall if you can't sleep, unless you
* pass in the __GFP_HARDWALL flag set in gfp_flag, which disables
* the code that might scan up ancestor cpusets and sleep.
- **/
+ */
-int __cpuset_zone_allowed(struct zone *z, gfp_t gfp_mask)
+int __cpuset_zone_allowed_softwall(struct zone *z, gfp_t gfp_mask)
{
int node; /* node that zone z is on */
const struct cpuset *cs; /* current cpuset ancestors */
@@ -2415,6 +2431,40 @@ int __cpuset_zone_allowed(struct zone *z, gfp_t gfp_mask)
return allowed;
}
+/*
+ * cpuset_zone_allowed_hardwall - Can we allocate on zone z's memory node?
+ * @z: is this zone on an allowed node?
+ * @gfp_mask: memory allocation flags
+ *
+ * If we're in interrupt, yes, we can always allocate.
+ * If __GFP_THISNODE is set, yes, we can always allocate. If zone
+ * z's node is in our tasks mems_allowed, yes. Otherwise, no.
+ *
+ * The __GFP_THISNODE placement logic is really handled elsewhere,
+ * by forcibly using a zonelist starting at a specified node, and by
+ * (in get_page_from_freelist()) refusing to consider the zones for
+ * any node on the zonelist except the first. By the time any such
+ * calls get to this routine, we should just shut up and say 'yes'.
+ *
+ * Unlike the cpuset_zone_allowed_softwall() variant, above,
+ * this variant requires that the zone be in the current tasks
+ * mems_allowed or that we're in interrupt. It does not scan up the
+ * cpuset hierarchy for the nearest enclosing mem_exclusive cpuset.
+ * It never sleeps.
+ */
+
+int __cpuset_zone_allowed_hardwall(struct zone *z, gfp_t gfp_mask)
+{
+ int node; /* node that zone z is on */
+
+ if (in_interrupt() || (gfp_mask & __GFP_THISNODE))
+ return 1;
+ node = zone_to_nid(z);
+ if (node_isset(node, current->mems_allowed))
+ return 1;
+ return 0;
+}
+
/**
* cpuset_lock - lock out any changes to cpuset structures
*
diff --git a/kernel/exit.c b/kernel/exit.c
index 03e64fe4a14a..122fadb972fc 100644
--- a/kernel/exit.c
+++ b/kernel/exit.c
@@ -425,7 +425,7 @@ static void close_files(struct files_struct * files)
for (;;) {
unsigned long set;
i = j * __NFDBITS;
- if (i >= fdt->max_fdset || i >= fdt->max_fds)
+ if (i >= fdt->max_fds)
break;
set = fdt->open_fds->fds_bits[j++];
while (set) {
@@ -466,11 +466,9 @@ void fastcall put_files_struct(struct files_struct *files)
* you can free files immediately.
*/
fdt = files_fdtable(files);
- if (fdt == &files->fdtab)
- fdt->free_files = files;
- else
+ if (fdt != &files->fdtab)
kmem_cache_free(files_cachep, files);
- free_fdtable(fdt);
+ call_rcu(&fdt->rcu, free_fdtable_rcu);
}
}
diff --git a/kernel/fork.c b/kernel/fork.c
index 8c859eef8e6a..d16c566eb645 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -36,6 +36,7 @@
#include <linux/syscalls.h>
#include <linux/jiffies.h>
#include <linux/futex.h>
+#include <linux/task_io_accounting_ops.h>
#include <linux/rcupdate.h>
#include <linux/ptrace.h>
#include <linux/mount.h>
@@ -613,7 +614,7 @@ static inline int copy_fs(unsigned long clone_flags, struct task_struct * tsk)
static int count_open_files(struct fdtable *fdt)
{
- int size = fdt->max_fdset;
+ int size = fdt->max_fds;
int i;
/* Find the last open fd */
@@ -640,12 +641,10 @@ static struct files_struct *alloc_files(void)
newf->next_fd = 0;
fdt = &newf->fdtab;
fdt->max_fds = NR_OPEN_DEFAULT;
- fdt->max_fdset = EMBEDDED_FD_SET_SIZE;
fdt->close_on_exec = (fd_set *)&newf->close_on_exec_init;
fdt->open_fds = (fd_set *)&newf->open_fds_init;
fdt->fd = &newf->fd_array[0];
INIT_RCU_HEAD(&fdt->rcu);
- fdt->free_files = NULL;
fdt->next = NULL;
rcu_assign_pointer(newf->fdt, fdt);
out:
@@ -661,7 +660,7 @@ static struct files_struct *dup_fd(struct files_struct *oldf, int *errorp)
{
struct files_struct *newf;
struct file **old_fds, **new_fds;
- int open_files, size, i, expand;
+ int open_files, size, i;
struct fdtable *old_fdt, *new_fdt;
*errorp = -ENOMEM;
@@ -672,25 +671,14 @@ static struct files_struct *dup_fd(struct files_struct *oldf, int *errorp)
spin_lock(&oldf->file_lock);
old_fdt = files_fdtable(oldf);
new_fdt = files_fdtable(newf);
- size = old_fdt->max_fdset;
open_files = count_open_files(old_fdt);
- expand = 0;
/*
- * Check whether we need to allocate a larger fd array or fd set.
- * Note: we're not a clone task, so the open count won't change.
+ * Check whether we need to allocate a larger fd array and fd set.
+ * Note: we're not a clone task, so the open count won't change.
*/
- if (open_files > new_fdt->max_fdset) {
- new_fdt->max_fdset = 0;
- expand = 1;
- }
if (open_files > new_fdt->max_fds) {
new_fdt->max_fds = 0;
- expand = 1;
- }
-
- /* if the old fdset gets grown now, we'll only copy up to "size" fds */
- if (expand) {
spin_unlock(&oldf->file_lock);
spin_lock(&newf->file_lock);
*errorp = expand_files(newf, open_files-1);
@@ -710,8 +698,10 @@ static struct files_struct *dup_fd(struct files_struct *oldf, int *errorp)
old_fds = old_fdt->fd;
new_fds = new_fdt->fd;
- memcpy(new_fdt->open_fds->fds_bits, old_fdt->open_fds->fds_bits, open_files/8);
- memcpy(new_fdt->close_on_exec->fds_bits, old_fdt->close_on_exec->fds_bits, open_files/8);
+ memcpy(new_fdt->open_fds->fds_bits,
+ old_fdt->open_fds->fds_bits, open_files/8);
+ memcpy(new_fdt->close_on_exec->fds_bits,
+ old_fdt->close_on_exec->fds_bits, open_files/8);
for (i = open_files; i != 0; i--) {
struct file *f = *old_fds++;
@@ -736,22 +726,19 @@ static struct files_struct *dup_fd(struct files_struct *oldf, int *errorp)
/* This is long word aligned thus could use a optimized version */
memset(new_fds, 0, size);
- if (new_fdt->max_fdset > open_files) {
- int left = (new_fdt->max_fdset-open_files)/8;
+ if (new_fdt->max_fds > open_files) {
+ int left = (new_fdt->max_fds-open_files)/8;
int start = open_files / (8 * sizeof(unsigned long));
memset(&new_fdt->open_fds->fds_bits[start], 0, left);
memset(&new_fdt->close_on_exec->fds_bits[start], 0, left);
}
-out:
return newf;
out_release:
- free_fdset (new_fdt->close_on_exec, new_fdt->max_fdset);
- free_fdset (new_fdt->open_fds, new_fdt->max_fdset);
- free_fd_array(new_fdt->fd, new_fdt->max_fds);
kmem_cache_free(files_cachep, newf);
+out:
return NULL;
}
@@ -1055,6 +1042,7 @@ static struct task_struct *copy_process(unsigned long clone_flags,
p->wchar = 0; /* I/O counter: bytes written */
p->syscr = 0; /* I/O counter: read syscalls */
p->syscw = 0; /* I/O counter: write syscalls */
+ task_io_accounting_init(p);
acct_clear_integrals(p);
p->it_virt_expires = cputime_zero;
diff --git a/kernel/lockdep.c b/kernel/lockdep.c
index b02032476dc2..01e750559034 100644
--- a/kernel/lockdep.c
+++ b/kernel/lockdep.c
@@ -43,13 +43,49 @@
#include "lockdep_internals.h"
/*
- * hash_lock: protects the lockdep hashes and class/list/hash allocators.
+ * lockdep_lock: protects the lockdep graph, the hashes and the
+ * class/list/hash allocators.
*
* This is one of the rare exceptions where it's justified
* to use a raw spinlock - we really dont want the spinlock
- * code to recurse back into the lockdep code.
+ * code to recurse back into the lockdep code...
*/
-static raw_spinlock_t hash_lock = (raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED;
+static raw_spinlock_t lockdep_lock = (raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED;
+
+static int graph_lock(void)
+{
+ __raw_spin_lock(&lockdep_lock);
+ /*
+ * Make sure that if another CPU detected a bug while
+ * walking the graph we dont change it (while the other
+ * CPU is busy printing out stuff with the graph lock
+ * dropped already)
+ */
+ if (!debug_locks) {
+ __raw_spin_unlock(&lockdep_lock);
+ return 0;
+ }
+ return 1;
+}
+
+static inline int graph_unlock(void)
+{
+ __raw_spin_unlock(&lockdep_lock);
+ return 0;
+}
+
+/*
+ * Turn lock debugging off and return with 0 if it was off already,
+ * and also release the graph lock:
+ */
+static inline int debug_locks_off_graph_unlock(void)
+{
+ int ret = debug_locks_off();
+
+ __raw_spin_unlock(&lockdep_lock);
+
+ return ret;
+}
static int lockdep_initialized;
@@ -57,14 +93,15 @@ unsigned long nr_list_entries;
static struct lock_list list_entries[MAX_LOCKDEP_ENTRIES];
/*
- * Allocate a lockdep entry. (assumes hash_lock held, returns
+ * Allocate a lockdep entry. (assumes the graph_lock held, returns
* with NULL on failure)
*/
static struct lock_list *alloc_list_entry(void)
{
if (nr_list_entries >= MAX_LOCKDEP_ENTRIES) {
- __raw_spin_unlock(&hash_lock);
- debug_locks_off();
+ if (!debug_locks_off_graph_unlock())
+ return NULL;
+
printk("BUG: MAX_LOCKDEP_ENTRIES too low!\n");
printk("turning off the locking correctness validator.\n");
return NULL;
@@ -145,9 +182,7 @@ EXPORT_SYMBOL(lockdep_on);
*/
#define VERBOSE 0
-#ifdef VERBOSE
-# define VERY_VERBOSE 0
-#endif
+#define VERY_VERBOSE 0
#if VERBOSE
# define HARDIRQ_VERBOSE 1
@@ -172,8 +207,8 @@ static int class_filter(struct lock_class *class)
!strcmp(class->name, "&struct->lockfield"))
return 1;
#endif
- /* Allow everything else. 0 would be filter everything else */
- return 1;
+ /* Filter everything else. 1 would be to allow everything else */
+ return 0;
}
#endif
@@ -207,7 +242,7 @@ static int softirq_verbose(struct lock_class *class)
/*
* Stack-trace: tightly packed array of stack backtrace
- * addresses. Protected by the hash_lock.
+ * addresses. Protected by the graph_lock.
*/
unsigned long nr_stack_trace_entries;
static unsigned long stack_trace[MAX_STACK_TRACE_ENTRIES];
@@ -226,18 +261,15 @@ static int save_trace(struct stack_trace *trace)
trace->max_entries = trace->nr_entries;
nr_stack_trace_entries += trace->nr_entries;
- if (DEBUG_LOCKS_WARN_ON(nr_stack_trace_entries > MAX_STACK_TRACE_ENTRIES)) {
- __raw_spin_unlock(&hash_lock);
- return 0;
- }
if (nr_stack_trace_entries == MAX_STACK_TRACE_ENTRIES) {
- __raw_spin_unlock(&hash_lock);
- if (debug_locks_off()) {
- printk("BUG: MAX_STACK_TRACE_ENTRIES too low!\n");
- printk("turning off the locking correctness validator.\n");
- dump_stack();
- }
+ if (!debug_locks_off_graph_unlock())
+ return 0;
+
+ printk("BUG: MAX_STACK_TRACE_ENTRIES too low!\n");
+ printk("turning off the locking correctness validator.\n");
+ dump_stack();
+
return 0;
}
@@ -526,9 +558,7 @@ print_circular_bug_header(struct lock_list *entry, unsigned int depth)
{
struct task_struct *curr = current;
- __raw_spin_unlock(&hash_lock);
- debug_locks_off();
- if (debug_locks_silent)
+ if (!debug_locks_off_graph_unlock() || debug_locks_silent)
return 0;
printk("\n=======================================================\n");
@@ -556,12 +586,10 @@ static noinline int print_circular_bug_tail(void)
if (debug_locks_silent)
return 0;
- /* hash_lock unlocked by the header */
- __raw_spin_lock(&hash_lock);
this.class = check_source->class;
if (!save_trace(&this.trace))
return 0;
- __raw_spin_unlock(&hash_lock);
+
print_circular_bug_entry(&this, 0);
printk("\nother info that might help us debug this:\n\n");
@@ -577,8 +605,10 @@ static noinline int print_circular_bug_tail(void)
static int noinline print_infinite_recursion_bug(void)
{
- __raw_spin_unlock(&hash_lock);
- DEBUG_LOCKS_WARN_ON(1);
+ if (!debug_locks_off_graph_unlock())
+ return 0;
+
+ WARN_ON(1);
return 0;
}
@@ -713,9 +743,7 @@ print_bad_irq_dependency(struct task_struct *curr,
enum lock_usage_bit bit2,
const char *irqclass)
{
- __raw_spin_unlock(&hash_lock);
- debug_locks_off();
- if (debug_locks_silent)
+ if (!debug_locks_off_graph_unlock() || debug_locks_silent)
return 0;
printk("\n======================================================\n");
@@ -796,9 +824,7 @@ static int
print_deadlock_bug(struct task_struct *curr, struct held_lock *prev,
struct held_lock *next)
{
- debug_locks_off();
- __raw_spin_unlock(&hash_lock);
- if (debug_locks_silent)
+ if (!debug_locks_off_graph_unlock() || debug_locks_silent)
return 0;
printk("\n=============================================\n");
@@ -974,14 +1000,14 @@ check_prev_add(struct task_struct *curr, struct held_lock *prev,
* Debugging printouts:
*/
if (verbose(prev->class) || verbose(next->class)) {
- __raw_spin_unlock(&hash_lock);
+ graph_unlock();
printk("\n new dependency: ");
print_lock_name(prev->class);
printk(" => ");
print_lock_name(next->class);
printk("\n");
dump_stack();
- __raw_spin_lock(&hash_lock);
+ return graph_lock();
}
return 1;
}
@@ -1046,8 +1072,10 @@ check_prevs_add(struct task_struct *curr, struct held_lock *next)
}
return 1;
out_bug:
- __raw_spin_unlock(&hash_lock);
- DEBUG_LOCKS_WARN_ON(1);
+ if (!debug_locks_off_graph_unlock())
+ return 0;
+
+ WARN_ON(1);
return 0;
}
@@ -1201,7 +1229,10 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force)
hash_head = classhashentry(key);
raw_local_irq_save(flags);
- __raw_spin_lock(&hash_lock);
+ if (!graph_lock()) {
+ raw_local_irq_restore(flags);
+ return NULL;
+ }
/*
* We have to do the hash-walk again, to avoid races
* with another CPU:
@@ -1214,9 +1245,12 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force)
* the hash:
*/
if (nr_lock_classes >= MAX_LOCKDEP_KEYS) {
- __raw_spin_unlock(&hash_lock);
+ if (!debug_locks_off_graph_unlock()) {
+ raw_local_irq_restore(flags);
+ return NULL;
+ }
raw_local_irq_restore(flags);
- debug_locks_off();
+
printk("BUG: MAX_LOCKDEP_KEYS too low!\n");
printk("turning off the locking correctness validator.\n");
return NULL;
@@ -1237,18 +1271,23 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force)
list_add_tail_rcu(&class->hash_entry, hash_head);
if (verbose(class)) {
- __raw_spin_unlock(&hash_lock);
+ graph_unlock();
raw_local_irq_restore(flags);
+
printk("\nnew class %p: %s", class->key, class->name);
if (class->name_version > 1)
printk("#%d", class->name_version);
printk("\n");
dump_stack();
+
raw_local_irq_save(flags);
- __raw_spin_lock(&hash_lock);
+ if (!graph_lock()) {
+ raw_local_irq_restore(flags);
+ return NULL;
+ }
}
out_unlock_set:
- __raw_spin_unlock(&hash_lock);
+ graph_unlock();
raw_local_irq_restore(flags);
if (!subclass || force)
@@ -1264,7 +1303,7 @@ out_unlock_set:
* add it and return 0 - in this case the new dependency chain is
* validated. If the key is already hashed, return 1.
*/
-static inline int lookup_chain_cache(u64 chain_key)
+static inline int lookup_chain_cache(u64 chain_key, struct lock_class *class)
{
struct list_head *hash_head = chainhashentry(chain_key);
struct lock_chain *chain;
@@ -1278,34 +1317,32 @@ static inline int lookup_chain_cache(u64 chain_key)
if (chain->chain_key == chain_key) {
cache_hit:
debug_atomic_inc(&chain_lookup_hits);
- /*
- * In the debugging case, force redundant checking
- * by returning 1:
- */
-#ifdef CONFIG_DEBUG_LOCKDEP
- __raw_spin_lock(&hash_lock);
- return 1;
-#endif
+ if (very_verbose(class))
+ printk("\nhash chain already cached, key: %016Lx tail class: [%p] %s\n", chain_key, class->key, class->name);
return 0;
}
}
+ if (very_verbose(class))
+ printk("\nnew hash chain, key: %016Lx tail class: [%p] %s\n", chain_key, class->key, class->name);
/*
* Allocate a new chain entry from the static array, and add
* it to the hash:
*/
- __raw_spin_lock(&hash_lock);
+ if (!graph_lock())
+ return 0;
/*
* We have to walk the chain again locked - to avoid duplicates:
*/
list_for_each_entry(chain, hash_head, entry) {
if (chain->chain_key == chain_key) {
- __raw_spin_unlock(&hash_lock);
+ graph_unlock();
goto cache_hit;
}
}
if (unlikely(nr_lock_chains >= MAX_LOCKDEP_CHAINS)) {
- __raw_spin_unlock(&hash_lock);
- debug_locks_off();
+ if (!debug_locks_off_graph_unlock())
+ return 0;
+
printk("BUG: MAX_LOCKDEP_CHAINS too low!\n");
printk("turning off the locking correctness validator.\n");
return 0;
@@ -1381,9 +1418,7 @@ print_irq_inversion_bug(struct task_struct *curr, struct lock_class *other,
struct held_lock *this, int forwards,
const char *irqclass)
{
- __raw_spin_unlock(&hash_lock);
- debug_locks_off();
- if (debug_locks_silent)
+ if (!debug_locks_off_graph_unlock() || debug_locks_silent)
return 0;
printk("\n=========================================================\n");
@@ -1453,7 +1488,7 @@ check_usage_backwards(struct task_struct *curr, struct held_lock *this,
return print_irq_inversion_bug(curr, backwards_match, this, 0, irqclass);
}
-static inline void print_irqtrace_events(struct task_struct *curr)
+void print_irqtrace_events(struct task_struct *curr)
{
printk("irq event stamp: %u\n", curr->irq_events);
printk("hardirqs last enabled at (%u): ", curr->hardirq_enable_event);
@@ -1466,19 +1501,13 @@ static inline void print_irqtrace_events(struct task_struct *curr)
print_ip_sym(curr->softirq_disable_ip);
}
-#else
-static inline void print_irqtrace_events(struct task_struct *curr)
-{
-}
#endif
static int
print_usage_bug(struct task_struct *curr, struct held_lock *this,
enum lock_usage_bit prev_bit, enum lock_usage_bit new_bit)
{
- __raw_spin_unlock(&hash_lock);
- debug_locks_off();
- if (debug_locks_silent)
+ if (!debug_locks_off_graph_unlock() || debug_locks_silent)
return 0;
printk("\n=================================\n");
@@ -1539,12 +1568,13 @@ static int mark_lock(struct task_struct *curr, struct held_lock *this,
if (likely(this->class->usage_mask & new_mask))
return 1;
- __raw_spin_lock(&hash_lock);
+ if (!graph_lock())
+ return 0;
/*
* Make sure we didnt race:
*/
if (unlikely(this->class->usage_mask & new_mask)) {
- __raw_spin_unlock(&hash_lock);
+ graph_unlock();
return 1;
}
@@ -1730,16 +1760,16 @@ static int mark_lock(struct task_struct *curr, struct held_lock *this,
debug_atomic_dec(&nr_unused_locks);
break;
default:
- __raw_spin_unlock(&hash_lock);
- debug_locks_off();
+ if (!debug_locks_off_graph_unlock())
+ return 0;
WARN_ON(1);
return 0;
}
- __raw_spin_unlock(&hash_lock);
+ graph_unlock();
/*
- * We must printk outside of the hash_lock:
+ * We must printk outside of the graph_lock:
*/
if (ret == 2) {
printk("\nmarked lock as {%s}:\n", usage_str[new_bit]);
@@ -2137,9 +2167,9 @@ out_calc_hash:
* We look up the chain_key and do the O(N^2) check and update of
* the dependencies only if this is a new dependency chain.
* (If lookup_chain_cache() returns with 1 it acquires
- * hash_lock for us)
+ * graph_lock for us)
*/
- if (!trylock && (check == 2) && lookup_chain_cache(chain_key)) {
+ if (!trylock && (check == 2) && lookup_chain_cache(chain_key, class)) {
/*
* Check whether last held lock:
*
@@ -2170,7 +2200,7 @@ out_calc_hash:
if (!chain_head && ret != 2)
if (!check_prevs_add(curr, hlock))
return 0;
- __raw_spin_unlock(&hash_lock);
+ graph_unlock();
}
curr->lockdep_depth++;
check_chain_key(curr);
@@ -2433,6 +2463,7 @@ EXPORT_SYMBOL_GPL(lock_release);
void lockdep_reset(void)
{
unsigned long flags;
+ int i;
raw_local_irq_save(flags);
current->curr_chain_key = 0;
@@ -2443,6 +2474,8 @@ void lockdep_reset(void)
nr_softirq_chains = 0;
nr_process_chains = 0;
debug_locks = 1;
+ for (i = 0; i < CHAINHASH_SIZE; i++)
+ INIT_LIST_HEAD(chainhash_table + i);
raw_local_irq_restore(flags);
}
@@ -2479,7 +2512,7 @@ void lockdep_free_key_range(void *start, unsigned long size)
int i;
raw_local_irq_save(flags);
- __raw_spin_lock(&hash_lock);
+ graph_lock();
/*
* Unhash all classes that were created by this module:
@@ -2493,7 +2526,7 @@ void lockdep_free_key_range(void *start, unsigned long size)
zap_class(class);
}
- __raw_spin_unlock(&hash_lock);
+ graph_unlock();
raw_local_irq_restore(flags);
}
@@ -2521,20 +2554,20 @@ void lockdep_reset_lock(struct lockdep_map *lock)
* Debug check: in the end all mapped classes should
* be gone.
*/
- __raw_spin_lock(&hash_lock);
+ graph_lock();
for (i = 0; i < CLASSHASH_SIZE; i++) {
head = classhash_table + i;
if (list_empty(head))
continue;
list_for_each_entry_safe(class, next, head, hash_entry) {
if (unlikely(class == lock->class_cache)) {
- __raw_spin_unlock(&hash_lock);
- DEBUG_LOCKS_WARN_ON(1);
+ if (debug_locks_off_graph_unlock())
+ WARN_ON(1);
goto out_restore;
}
}
}
- __raw_spin_unlock(&hash_lock);
+ graph_unlock();
out_restore:
raw_local_irq_restore(flags);
diff --git a/kernel/nsproxy.c b/kernel/nsproxy.c
index e2ce748e96af..f5b9ee6f6bbb 100644
--- a/kernel/nsproxy.c
+++ b/kernel/nsproxy.c
@@ -46,10 +46,8 @@ static inline struct nsproxy *clone_namespaces(struct nsproxy *orig)
struct nsproxy *ns;
ns = kmemdup(orig, sizeof(struct nsproxy), GFP_KERNEL);
- if (ns) {
+ if (ns)
atomic_set(&ns->count, 1);
- ns->id = -1;
- }
return ns;
}
diff --git a/kernel/power/process.c b/kernel/power/process.c
index 99eeb119b06d..6d566bf7085c 100644
--- a/kernel/power/process.c
+++ b/kernel/power/process.c
@@ -28,8 +28,7 @@ static inline int freezeable(struct task_struct * p)
if ((p == current) ||
(p->flags & PF_NOFREEZE) ||
(p->exit_state == EXIT_ZOMBIE) ||
- (p->exit_state == EXIT_DEAD) ||
- (p->state == TASK_STOPPED))
+ (p->exit_state == EXIT_DEAD))
return 0;
return 1;
}
@@ -61,10 +60,16 @@ static inline void freeze_process(struct task_struct *p)
unsigned long flags;
if (!freezing(p)) {
- freeze(p);
- spin_lock_irqsave(&p->sighand->siglock, flags);
- signal_wake_up(p, 0);
- spin_unlock_irqrestore(&p->sighand->siglock, flags);
+ rmb();
+ if (!frozen(p)) {
+ if (p->state == TASK_STOPPED)
+ force_sig_specific(SIGSTOP, p);
+
+ freeze(p);
+ spin_lock_irqsave(&p->sighand->siglock, flags);
+ signal_wake_up(p, p->state == TASK_STOPPED);
+ spin_unlock_irqrestore(&p->sighand->siglock, flags);
+ }
}
}
@@ -103,9 +108,7 @@ static unsigned int try_to_freeze_tasks(int freeze_user_space)
if (frozen(p))
continue;
- if (p->state == TASK_TRACED &&
- (frozen(p->parent) ||
- p->parent->state == TASK_STOPPED)) {
+ if (p->state == TASK_TRACED && frozen(p->parent)) {
cancel_freezing(p);
continue;
}
diff --git a/kernel/relay.c b/kernel/relay.c
index 818e514729cf..a4701e7ba7d0 100644
--- a/kernel/relay.c
+++ b/kernel/relay.c
@@ -138,7 +138,7 @@ depopulate:
*/
struct rchan_buf *relay_create_buf(struct rchan *chan)
{
- struct rchan_buf *buf = kcalloc(1, sizeof(struct rchan_buf), GFP_KERNEL);
+ struct rchan_buf *buf = kzalloc(sizeof(struct rchan_buf), GFP_KERNEL);
if (!buf)
return NULL;
@@ -479,7 +479,7 @@ struct rchan *relay_open(const char *base_filename,
if (!(subbuf_size && n_subbufs))
return NULL;
- chan = kcalloc(1, sizeof(struct rchan), GFP_KERNEL);
+ chan = kzalloc(sizeof(struct rchan), GFP_KERNEL);
if (!chan)
return NULL;
diff --git a/kernel/sched.c b/kernel/sched.c
index f385eff4682d..5cd833bc2173 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -225,8 +225,10 @@ struct rq {
unsigned long nr_uninterruptible;
unsigned long expired_timestamp;
- unsigned long long timestamp_last_tick;
+ /* Cached timestamp set by update_cpu_clock() */
+ unsigned long long most_recent_timestamp;
struct task_struct *curr, *idle;
+ unsigned long next_balance;
struct mm_struct *prev_mm;
struct prio_array *active, *expired, arrays[2];
int best_expired_prio;
@@ -426,7 +428,7 @@ static inline void task_rq_unlock(struct rq *rq, unsigned long *flags)
* bump this up when changing the output format or the meaning of an existing
* format, so that tools can adapt (or abort)
*/
-#define SCHEDSTAT_VERSION 12
+#define SCHEDSTAT_VERSION 14
static int show_schedstat(struct seq_file *seq, void *v)
{
@@ -464,7 +466,8 @@ static int show_schedstat(struct seq_file *seq, void *v)
seq_printf(seq, "domain%d %s", dcnt++, mask_str);
for (itype = SCHED_IDLE; itype < MAX_IDLE_TYPES;
itype++) {
- seq_printf(seq, " %lu %lu %lu %lu %lu %lu %lu %lu",
+ seq_printf(seq, " %lu %lu %lu %lu %lu %lu %lu "
+ "%lu",
sd->lb_cnt[itype],
sd->lb_balanced[itype],
sd->lb_failed[itype],
@@ -474,11 +477,13 @@ static int show_schedstat(struct seq_file *seq, void *v)
sd->lb_nobusyq[itype],
sd->lb_nobusyg[itype]);
}
- seq_printf(seq, " %lu %lu %lu %lu %lu %lu %lu %lu %lu %lu %lu %lu\n",
+ seq_printf(seq, " %lu %lu %lu %lu %lu %lu %lu %lu %lu"
+ " %lu %lu %lu\n",
sd->alb_cnt, sd->alb_failed, sd->alb_pushed,
sd->sbe_cnt, sd->sbe_balanced, sd->sbe_pushed,
sd->sbf_cnt, sd->sbf_balanced, sd->sbf_pushed,
- sd->ttwu_wake_remote, sd->ttwu_move_affine, sd->ttwu_move_balance);
+ sd->ttwu_wake_remote, sd->ttwu_move_affine,
+ sd->ttwu_move_balance);
}
preempt_enable();
#endif
@@ -547,7 +552,7 @@ rq_sched_info_depart(struct rq *rq, unsigned long delta_jiffies)
#endif
/*
- * rq_lock - lock a given runqueue and disable interrupts.
+ * this_rq_lock - lock this runqueue and disable interrupts.
*/
static inline struct rq *this_rq_lock(void)
__acquires(rq->lock)
@@ -938,13 +943,16 @@ static void activate_task(struct task_struct *p, struct rq *rq, int local)
{
unsigned long long now;
+ if (rt_task(p))
+ goto out;
+
now = sched_clock();
#ifdef CONFIG_SMP
if (!local) {
/* Compensate for drifting sched_clock */
struct rq *this_rq = this_rq();
- now = (now - this_rq->timestamp_last_tick)
- + rq->timestamp_last_tick;
+ now = (now - this_rq->most_recent_timestamp)
+ + rq->most_recent_timestamp;
}
#endif
@@ -959,8 +967,7 @@ static void activate_task(struct task_struct *p, struct rq *rq, int local)
(now - p->timestamp) >> 20);
}
- if (!rt_task(p))
- p->prio = recalc_task_prio(p, now);
+ p->prio = recalc_task_prio(p, now);
/*
* This checks to make sure it's not an uninterruptible task
@@ -985,7 +992,7 @@ static void activate_task(struct task_struct *p, struct rq *rq, int local)
}
}
p->timestamp = now;
-
+out:
__activate_task(p, rq);
}
@@ -1450,7 +1457,9 @@ static int try_to_wake_up(struct task_struct *p, unsigned int state, int sync)
if (this_sd->flags & SD_WAKE_AFFINE) {
unsigned long tl = this_load;
- unsigned long tl_per_task = cpu_avg_load_per_task(this_cpu);
+ unsigned long tl_per_task;
+
+ tl_per_task = cpu_avg_load_per_task(this_cpu);
/*
* If sync wakeup then subtract the (maximum possible)
@@ -1688,8 +1697,8 @@ void fastcall wake_up_new_task(struct task_struct *p, unsigned long clone_flags)
* Not the local CPU - must adjust timestamp. This should
* get optimised away in the !CONFIG_SMP case.
*/
- p->timestamp = (p->timestamp - this_rq->timestamp_last_tick)
- + rq->timestamp_last_tick;
+ p->timestamp = (p->timestamp - this_rq->most_recent_timestamp)
+ + rq->most_recent_timestamp;
__activate_task(p, rq);
if (TASK_PREEMPTS_CURR(p, rq))
resched_task(rq->curr);
@@ -1952,6 +1961,7 @@ static void double_rq_lock(struct rq *rq1, struct rq *rq2)
__acquires(rq1->lock)
__acquires(rq2->lock)
{
+ BUG_ON(!irqs_disabled());
if (rq1 == rq2) {
spin_lock(&rq1->lock);
__acquire(rq2->lock); /* Fake it out ;) */
@@ -1991,6 +2001,11 @@ static void double_lock_balance(struct rq *this_rq, struct rq *busiest)
__acquires(busiest->lock)
__acquires(this_rq->lock)
{
+ if (unlikely(!irqs_disabled())) {
+ /* printk() doesn't work good under rq->lock */
+ spin_unlock(&this_rq->lock);
+ BUG_ON(1);
+ }
if (unlikely(!spin_trylock(&busiest->lock))) {
if (busiest < this_rq) {
spin_unlock(&this_rq->lock);
@@ -2061,8 +2076,8 @@ static void pull_task(struct rq *src_rq, struct prio_array *src_array,
set_task_cpu(p, this_cpu);
inc_nr_running(p, this_rq);
enqueue_task(p, this_array);
- p->timestamp = (p->timestamp - src_rq->timestamp_last_tick)
- + this_rq->timestamp_last_tick;
+ p->timestamp = (p->timestamp - src_rq->most_recent_timestamp)
+ + this_rq->most_recent_timestamp;
/*
* Note that idle threads have a prio of MAX_PRIO, for this test
* to be always true for them.
@@ -2098,10 +2113,15 @@ int can_migrate_task(struct task_struct *p, struct rq *rq, int this_cpu,
* 2) too many balance attempts have failed.
*/
- if (sd->nr_balance_failed > sd->cache_nice_tries)
+ if (sd->nr_balance_failed > sd->cache_nice_tries) {
+#ifdef CONFIG_SCHEDSTATS
+ if (task_hot(p, rq->most_recent_timestamp, sd))
+ schedstat_inc(sd, lb_hot_gained[idle]);
+#endif
return 1;
+ }
- if (task_hot(p, rq->timestamp_last_tick, sd))
+ if (task_hot(p, rq->most_recent_timestamp, sd))
return 0;
return 1;
}
@@ -2199,11 +2219,6 @@ skip_queue:
goto skip_bitmap;
}
-#ifdef CONFIG_SCHEDSTATS
- if (task_hot(tmp, busiest->timestamp_last_tick, sd))
- schedstat_inc(sd, lb_hot_gained[idle]);
-#endif
-
pull_task(busiest, array, tmp, this_rq, dst_array, this_cpu);
pulled++;
rem_load_move -= tmp->load_weight;
@@ -2241,7 +2256,7 @@ out:
static struct sched_group *
find_busiest_group(struct sched_domain *sd, int this_cpu,
unsigned long *imbalance, enum idle_type idle, int *sd_idle,
- cpumask_t *cpus)
+ cpumask_t *cpus, int *balance)
{
struct sched_group *busiest = NULL, *this = NULL, *group = sd->groups;
unsigned long max_load, avg_load, total_load, this_load, total_pwr;
@@ -2270,10 +2285,14 @@ find_busiest_group(struct sched_domain *sd, int this_cpu,
unsigned long load, group_capacity;
int local_group;
int i;
+ unsigned int balance_cpu = -1, first_idle_cpu = 0;
unsigned long sum_nr_running, sum_weighted_load;
local_group = cpu_isset(this_cpu, group->cpumask);
+ if (local_group)
+ balance_cpu = first_cpu(group->cpumask);
+
/* Tally up the load of all CPUs in the group */
sum_weighted_load = sum_nr_running = avg_load = 0;
@@ -2289,9 +2308,14 @@ find_busiest_group(struct sched_domain *sd, int this_cpu,
*sd_idle = 0;
/* Bias balancing toward cpus of our domain */
- if (local_group)
+ if (local_group) {
+ if (idle_cpu(i) && !first_idle_cpu) {
+ first_idle_cpu = 1;
+ balance_cpu = i;
+ }
+
load = target_load(i, load_idx);
- else
+ } else
load = source_load(i, load_idx);
avg_load += load;
@@ -2299,6 +2323,16 @@ find_busiest_group(struct sched_domain *sd, int this_cpu,
sum_weighted_load += rq->raw_weighted_load;
}
+ /*
+ * First idle cpu or the first cpu(busiest) in this sched group
+ * is eligible for doing load balancing at this and above
+ * domains.
+ */
+ if (local_group && balance_cpu != this_cpu && balance) {
+ *balance = 0;
+ goto ret;
+ }
+
total_load += avg_load;
total_pwr += group->cpu_power;
@@ -2458,18 +2492,21 @@ small_imbalance:
pwr_now /= SCHED_LOAD_SCALE;
/* Amount of load we'd subtract */
- tmp = busiest_load_per_task*SCHED_LOAD_SCALE/busiest->cpu_power;
+ tmp = busiest_load_per_task * SCHED_LOAD_SCALE /
+ busiest->cpu_power;
if (max_load > tmp)
pwr_move += busiest->cpu_power *
min(busiest_load_per_task, max_load - tmp);
/* Amount of load we'd add */
- if (max_load*busiest->cpu_power <
- busiest_load_per_task*SCHED_LOAD_SCALE)
- tmp = max_load*busiest->cpu_power/this->cpu_power;
+ if (max_load * busiest->cpu_power <
+ busiest_load_per_task * SCHED_LOAD_SCALE)
+ tmp = max_load * busiest->cpu_power / this->cpu_power;
else
- tmp = busiest_load_per_task*SCHED_LOAD_SCALE/this->cpu_power;
- pwr_move += this->cpu_power*min(this_load_per_task, this_load + tmp);
+ tmp = busiest_load_per_task * SCHED_LOAD_SCALE /
+ this->cpu_power;
+ pwr_move += this->cpu_power *
+ min(this_load_per_task, this_load + tmp);
pwr_move /= SCHED_LOAD_SCALE;
/* Move if we gain throughput */
@@ -2490,8 +2527,8 @@ out_balanced:
*imbalance = min_load_per_task;
return group_min;
}
-ret:
#endif
+ret:
*imbalance = 0;
return NULL;
}
@@ -2540,17 +2577,17 @@ static inline unsigned long minus_1_or_zero(unsigned long n)
/*
* Check this_cpu to ensure it is balanced within domain. Attempt to move
* tasks if there is an imbalance.
- *
- * Called with this_rq unlocked.
*/
static int load_balance(int this_cpu, struct rq *this_rq,
- struct sched_domain *sd, enum idle_type idle)
+ struct sched_domain *sd, enum idle_type idle,
+ int *balance)
{
int nr_moved, all_pinned = 0, active_balance = 0, sd_idle = 0;
struct sched_group *group;
unsigned long imbalance;
struct rq *busiest;
cpumask_t cpus = CPU_MASK_ALL;
+ unsigned long flags;
/*
* When power savings policy is enabled for the parent domain, idle
@@ -2566,7 +2603,11 @@ static int load_balance(int this_cpu, struct rq *this_rq,
redo:
group = find_busiest_group(sd, this_cpu, &imbalance, idle, &sd_idle,
- &cpus);
+ &cpus, balance);
+
+ if (*balance == 0)
+ goto out_balanced;
+
if (!group) {
schedstat_inc(sd, lb_nobusyg[idle]);
goto out_balanced;
@@ -2590,11 +2631,13 @@ redo:
* still unbalanced. nr_moved simply stays zero, so it is
* correctly treated as an imbalance.
*/
+ local_irq_save(flags);
double_rq_lock(this_rq, busiest);
nr_moved = move_tasks(this_rq, this_cpu, busiest,
minus_1_or_zero(busiest->nr_running),
imbalance, sd, idle, &all_pinned);
double_rq_unlock(this_rq, busiest);
+ local_irq_restore(flags);
/* All tasks on this runqueue were pinned by CPU affinity */
if (unlikely(all_pinned)) {
@@ -2611,13 +2654,13 @@ redo:
if (unlikely(sd->nr_balance_failed > sd->cache_nice_tries+2)) {
- spin_lock(&busiest->lock);
+ spin_lock_irqsave(&busiest->lock, flags);
/* don't kick the migration_thread, if the curr
* task on busiest cpu can't be moved to this_cpu
*/
if (!cpu_isset(this_cpu, busiest->curr->cpus_allowed)) {
- spin_unlock(&busiest->lock);
+ spin_unlock_irqrestore(&busiest->lock, flags);
all_pinned = 1;
goto out_one_pinned;
}
@@ -2627,7 +2670,7 @@ redo:
busiest->push_cpu = this_cpu;
active_balance = 1;
}
- spin_unlock(&busiest->lock);
+ spin_unlock_irqrestore(&busiest->lock, flags);
if (active_balance)
wake_up_process(busiest->migration_thread);
@@ -2706,7 +2749,7 @@ load_balance_newidle(int this_cpu, struct rq *this_rq, struct sched_domain *sd)
schedstat_inc(sd, lb_cnt[NEWLY_IDLE]);
redo:
group = find_busiest_group(sd, this_cpu, &imbalance, NEWLY_IDLE,
- &sd_idle, &cpus);
+ &sd_idle, &cpus, NULL);
if (!group) {
schedstat_inc(sd, lb_nobusyg[NEWLY_IDLE]);
goto out_balanced;
@@ -2766,14 +2809,28 @@ out_balanced:
static void idle_balance(int this_cpu, struct rq *this_rq)
{
struct sched_domain *sd;
+ int pulled_task = 0;
+ unsigned long next_balance = jiffies + 60 * HZ;
for_each_domain(this_cpu, sd) {
if (sd->flags & SD_BALANCE_NEWIDLE) {
/* If we've pulled tasks over stop searching: */
- if (load_balance_newidle(this_cpu, this_rq, sd))
+ pulled_task = load_balance_newidle(this_cpu,
+ this_rq, sd);
+ if (time_after(next_balance,
+ sd->last_balance + sd->balance_interval))
+ next_balance = sd->last_balance
+ + sd->balance_interval;
+ if (pulled_task)
break;
}
}
+ if (!pulled_task)
+ /*
+ * We are going idle. next_balance may be set based on
+ * a busy processor. So reset next_balance.
+ */
+ this_rq->next_balance = next_balance;
}
/*
@@ -2826,26 +2883,9 @@ static void active_load_balance(struct rq *busiest_rq, int busiest_cpu)
spin_unlock(&target_rq->lock);
}
-/*
- * rebalance_tick will get called every timer tick, on every CPU.
- *
- * It checks each scheduling domain to see if it is due to be balanced,
- * and initiates a balancing operation if so.
- *
- * Balancing parameters are set up in arch_init_sched_domains.
- */
-
-/* Don't have all balancing operations going off at once: */
-static inline unsigned long cpu_offset(int cpu)
+static void update_load(struct rq *this_rq)
{
- return jiffies + cpu * HZ / NR_CPUS;
-}
-
-static void
-rebalance_tick(int this_cpu, struct rq *this_rq, enum idle_type idle)
-{
- unsigned long this_load, interval, j = cpu_offset(this_cpu);
- struct sched_domain *sd;
+ unsigned long this_load;
int i, scale;
this_load = this_rq->raw_weighted_load;
@@ -2865,6 +2905,32 @@ rebalance_tick(int this_cpu, struct rq *this_rq, enum idle_type idle)
new_load += scale-1;
this_rq->cpu_load[i] = (old_load*(scale-1) + new_load) / scale;
}
+}
+
+/*
+ * run_rebalance_domains is triggered when needed from the scheduler tick.
+ *
+ * It checks each scheduling domain to see if it is due to be balanced,
+ * and initiates a balancing operation if so.
+ *
+ * Balancing parameters are set up in arch_init_sched_domains.
+ */
+static DEFINE_SPINLOCK(balancing);
+
+static void run_rebalance_domains(struct softirq_action *h)
+{
+ int this_cpu = smp_processor_id(), balance = 1;
+ struct rq *this_rq = cpu_rq(this_cpu);
+ unsigned long interval;
+ struct sched_domain *sd;
+ /*
+ * We are idle if there are no processes running. This
+ * is valid even if we are the idle process (SMT).
+ */
+ enum idle_type idle = !this_rq->nr_running ?
+ SCHED_IDLE : NOT_IDLE;
+ /* Earliest time when we have to call run_rebalance_domains again */
+ unsigned long next_balance = jiffies + 60*HZ;
for_each_domain(this_cpu, sd) {
if (!(sd->flags & SD_LOAD_BALANCE))
@@ -2879,8 +2945,13 @@ rebalance_tick(int this_cpu, struct rq *this_rq, enum idle_type idle)
if (unlikely(!interval))
interval = 1;
- if (j - sd->last_balance >= interval) {
- if (load_balance(this_cpu, this_rq, sd, idle)) {
+ if (sd->flags & SD_SERIALIZE) {
+ if (!spin_trylock(&balancing))
+ goto out;
+ }
+
+ if (time_after_eq(jiffies, sd->last_balance + interval)) {
+ if (load_balance(this_cpu, this_rq, sd, idle, &balance)) {
/*
* We've pulled tasks over so either we're no
* longer idle, or one of our SMT siblings is
@@ -2888,39 +2959,48 @@ rebalance_tick(int this_cpu, struct rq *this_rq, enum idle_type idle)
*/
idle = NOT_IDLE;
}
- sd->last_balance += interval;
+ sd->last_balance = jiffies;
}
+ if (sd->flags & SD_SERIALIZE)
+ spin_unlock(&balancing);
+out:
+ if (time_after(next_balance, sd->last_balance + interval))
+ next_balance = sd->last_balance + interval;
+
+ /*
+ * Stop the load balance at this level. There is another
+ * CPU in our sched group which is doing load balancing more
+ * actively.
+ */
+ if (!balance)
+ break;
}
+ this_rq->next_balance = next_balance;
}
#else
/*
* on UP we do not need to balance between CPUs:
*/
-static inline void rebalance_tick(int cpu, struct rq *rq, enum idle_type idle)
-{
-}
static inline void idle_balance(int cpu, struct rq *rq)
{
}
#endif
-static inline int wake_priority_sleeper(struct rq *rq)
+static inline void wake_priority_sleeper(struct rq *rq)
{
- int ret = 0;
-
#ifdef CONFIG_SCHED_SMT
+ if (!rq->nr_running)
+ return;
+
spin_lock(&rq->lock);
/*
* If an SMT sibling task has been put to sleep for priority
* reasons reschedule the idle task to see if it can now run.
*/
- if (rq->nr_running) {
+ if (rq->nr_running)
resched_task(rq->idle);
- ret = 1;
- }
spin_unlock(&rq->lock);
#endif
- return ret;
}
DEFINE_PER_CPU(struct kernel_stat, kstat);
@@ -2934,7 +3014,8 @@ EXPORT_PER_CPU_SYMBOL(kstat);
static inline void
update_cpu_clock(struct task_struct *p, struct rq *rq, unsigned long long now)
{
- p->sched_time += now - max(p->timestamp, rq->timestamp_last_tick);
+ p->sched_time += now - p->last_ran;
+ p->last_ran = rq->most_recent_timestamp = now;
}
/*
@@ -2947,8 +3028,7 @@ unsigned long long current_sched_time(const struct task_struct *p)
unsigned long flags;
local_irq_save(flags);
- ns = max(p->timestamp, task_rq(p)->timestamp_last_tick);
- ns = p->sched_time + sched_clock() - ns;
+ ns = p->sched_time + sched_clock() - p->last_ran;
local_irq_restore(flags);
return ns;
@@ -3048,35 +3128,12 @@ void account_steal_time(struct task_struct *p, cputime_t steal)
cpustat->steal = cputime64_add(cpustat->steal, tmp);
}
-/*
- * This function gets called by the timer code, with HZ frequency.
- * We call it with interrupts disabled.
- *
- * It also gets called by the fork code, when changing the parent's
- * timeslices.
- */
-void scheduler_tick(void)
+static void task_running_tick(struct rq *rq, struct task_struct *p)
{
- unsigned long long now = sched_clock();
- struct task_struct *p = current;
- int cpu = smp_processor_id();
- struct rq *rq = cpu_rq(cpu);
-
- update_cpu_clock(p, rq, now);
-
- rq->timestamp_last_tick = now;
-
- if (p == rq->idle) {
- if (wake_priority_sleeper(rq))
- goto out;
- rebalance_tick(cpu, rq, SCHED_IDLE);
- return;
- }
-
- /* Task might have expired already, but not scheduled off yet */
if (p->array != rq->active) {
+ /* Task has expired but was not scheduled yet */
set_tsk_need_resched(p);
- goto out;
+ return;
}
spin_lock(&rq->lock);
/*
@@ -3144,8 +3201,34 @@ void scheduler_tick(void)
}
out_unlock:
spin_unlock(&rq->lock);
-out:
- rebalance_tick(cpu, rq, NOT_IDLE);
+}
+
+/*
+ * This function gets called by the timer code, with HZ frequency.
+ * We call it with interrupts disabled.
+ *
+ * It also gets called by the fork code, when changing the parent's
+ * timeslices.
+ */
+void scheduler_tick(void)
+{
+ unsigned long long now = sched_clock();
+ struct task_struct *p = current;
+ int cpu = smp_processor_id();
+ struct rq *rq = cpu_rq(cpu);
+
+ update_cpu_clock(p, rq, now);
+
+ if (p == rq->idle)
+ /* Task on the idle queue */
+ wake_priority_sleeper(rq);
+ else
+ task_running_tick(rq, p);
+#ifdef CONFIG_SMP
+ update_load(rq);
+ if (time_after_eq(jiffies, rq->next_balance))
+ raise_softirq(SCHED_SOFTIRQ);
+#endif
}
#ifdef CONFIG_SCHED_SMT
@@ -3291,7 +3374,8 @@ void fastcall add_preempt_count(int val)
/*
* Spinlock count overflowing soon?
*/
- DEBUG_LOCKS_WARN_ON((preempt_count() & PREEMPT_MASK) >= PREEMPT_MASK-10);
+ DEBUG_LOCKS_WARN_ON((preempt_count() & PREEMPT_MASK) >=
+ PREEMPT_MASK - 10);
}
EXPORT_SYMBOL(add_preempt_count);
@@ -3345,6 +3429,8 @@ asmlinkage void __sched schedule(void)
"%s/0x%08x/%d\n",
current->comm, preempt_count(), current->pid);
debug_show_held_locks(current);
+ if (irqs_disabled())
+ print_irqtrace_events(current);
dump_stack();
}
profile_hit(SCHED_PROFILING, __builtin_return_address(0));
@@ -4990,8 +5076,8 @@ static int __migrate_task(struct task_struct *p, int src_cpu, int dest_cpu)
* afterwards, and pretending it was a local activate.
* This way is cleaner and logically correct.
*/
- p->timestamp = p->timestamp - rq_src->timestamp_last_tick
- + rq_dest->timestamp_last_tick;
+ p->timestamp = p->timestamp - rq_src->most_recent_timestamp
+ + rq_dest->most_recent_timestamp;
deactivate_task(p, rq_src);
__activate_task(p, rq_dest);
if (TASK_PREEMPTS_CURR(p, rq_dest))
@@ -5067,7 +5153,10 @@ wait_to_die:
}
#ifdef CONFIG_HOTPLUG_CPU
-/* Figure out where task on dead CPU should go, use force if neccessary. */
+/*
+ * Figure out where task on dead CPU should go, use force if neccessary.
+ * NOTE: interrupts should be disabled by the caller
+ */
static void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p)
{
unsigned long flags;
@@ -5187,6 +5276,7 @@ void idle_task_exit(void)
mmdrop(mm);
}
+/* called under rq->lock with disabled interrupts */
static void migrate_dead(unsigned int dead_cpu, struct task_struct *p)
{
struct rq *rq = cpu_rq(dead_cpu);
@@ -5203,10 +5293,11 @@ static void migrate_dead(unsigned int dead_cpu, struct task_struct *p)
* Drop lock around migration; if someone else moves it,
* that's OK. No task can be added to this CPU, so iteration is
* fine.
+ * NOTE: interrupts should be left disabled --dev@
*/
- spin_unlock_irq(&rq->lock);
+ spin_unlock(&rq->lock);
move_task_off_dead_cpu(dead_cpu, p);
- spin_lock_irq(&rq->lock);
+ spin_lock(&rq->lock);
put_task_struct(p);
}
@@ -5359,16 +5450,19 @@ static void sched_domain_debug(struct sched_domain *sd, int cpu)
if (!(sd->flags & SD_LOAD_BALANCE)) {
printk("does not load-balance\n");
if (sd->parent)
- printk(KERN_ERR "ERROR: !SD_LOAD_BALANCE domain has parent");
+ printk(KERN_ERR "ERROR: !SD_LOAD_BALANCE domain"
+ " has parent");
break;
}
printk("span %s\n", str);
if (!cpu_isset(cpu, sd->span))
- printk(KERN_ERR "ERROR: domain->span does not contain CPU%d\n", cpu);
+ printk(KERN_ERR "ERROR: domain->span does not contain "
+ "CPU%d\n", cpu);
if (!cpu_isset(cpu, group->cpumask))
- printk(KERN_ERR "ERROR: domain->groups does not contain CPU%d\n", cpu);
+ printk(KERN_ERR "ERROR: domain->groups does not contain"
+ " CPU%d\n", cpu);
printk(KERN_DEBUG);
for (i = 0; i < level + 2; i++)
@@ -5383,7 +5477,8 @@ static void sched_domain_debug(struct sched_domain *sd, int cpu)
if (!group->cpu_power) {
printk("\n");
- printk(KERN_ERR "ERROR: domain->cpu_power not set\n");
+ printk(KERN_ERR "ERROR: domain->cpu_power not "
+ "set\n");
}
if (!cpus_weight(group->cpumask)) {
@@ -5406,15 +5501,17 @@ static void sched_domain_debug(struct sched_domain *sd, int cpu)
printk("\n");
if (!cpus_equal(sd->span, groupmask))
- printk(KERN_ERR "ERROR: groups don't span domain->span\n");
+ printk(KERN_ERR "ERROR: groups don't span "
+ "domain->span\n");
level++;
sd = sd->parent;
+ if (!sd)
+ continue;
- if (sd) {
- if (!cpus_subset(groupmask, sd->span))
- printk(KERN_ERR "ERROR: parent span is not a superset of domain->span\n");
- }
+ if (!cpus_subset(groupmask, sd->span))
+ printk(KERN_ERR "ERROR: parent span is not a superset "
+ "of domain->span\n");
} while (sd);
}
@@ -5528,28 +5625,27 @@ static int __init isolated_cpu_setup(char *str)
__setup ("isolcpus=", isolated_cpu_setup);
/*
- * init_sched_build_groups takes an array of groups, the cpumask we wish
- * to span, and a pointer to a function which identifies what group a CPU
- * belongs to. The return value of group_fn must be a valid index into the
- * groups[] array, and must be >= 0 and < NR_CPUS (due to the fact that we
- * keep track of groups covered with a cpumask_t).
+ * init_sched_build_groups takes the cpumask we wish to span, and a pointer
+ * to a function which identifies what group(along with sched group) a CPU
+ * belongs to. The return value of group_fn must be a >= 0 and < NR_CPUS
+ * (due to the fact that we keep track of groups covered with a cpumask_t).
*
* init_sched_build_groups will build a circular linked list of the groups
* covered by the given span, and will set each group's ->cpumask correctly,
* and ->cpu_power to 0.
*/
static void
-init_sched_build_groups(struct sched_group groups[], cpumask_t span,
- const cpumask_t *cpu_map,
- int (*group_fn)(int cpu, const cpumask_t *cpu_map))
+init_sched_build_groups(cpumask_t span, const cpumask_t *cpu_map,
+ int (*group_fn)(int cpu, const cpumask_t *cpu_map,
+ struct sched_group **sg))
{
struct sched_group *first = NULL, *last = NULL;
cpumask_t covered = CPU_MASK_NONE;
int i;
for_each_cpu_mask(i, span) {
- int group = group_fn(i, cpu_map);
- struct sched_group *sg = &groups[group];
+ struct sched_group *sg;
+ int group = group_fn(i, cpu_map, &sg);
int j;
if (cpu_isset(i, covered))
@@ -5559,7 +5655,7 @@ init_sched_build_groups(struct sched_group groups[], cpumask_t span,
sg->cpu_power = 0;
for_each_cpu_mask(j, span) {
- if (group_fn(j, cpu_map) != group)
+ if (group_fn(j, cpu_map, NULL) != group)
continue;
cpu_set(j, covered);
@@ -5733,8 +5829,9 @@ __setup("max_cache_size=", setup_max_cache_size);
*/
static void touch_cache(void *__cache, unsigned long __size)
{
- unsigned long size = __size/sizeof(long), chunk1 = size/3,
- chunk2 = 2*size/3;
+ unsigned long size = __size / sizeof(long);
+ unsigned long chunk1 = size / 3;
+ unsigned long chunk2 = 2 * size / 3;
unsigned long *cache = __cache;
int i;
@@ -5843,11 +5940,11 @@ measure_cost(int cpu1, int cpu2, void *cache, unsigned int size)
*/
measure_one(cache, size, cpu1, cpu2);
for (i = 0; i < ITERATIONS; i++)
- cost1 += measure_one(cache, size - i*1024, cpu1, cpu2);
+ cost1 += measure_one(cache, size - i * 1024, cpu1, cpu2);
measure_one(cache, size, cpu2, cpu1);
for (i = 0; i < ITERATIONS; i++)
- cost1 += measure_one(cache, size - i*1024, cpu2, cpu1);
+ cost1 += measure_one(cache, size - i * 1024, cpu2, cpu1);
/*
* (We measure the non-migrating [cached] cost on both
@@ -5857,17 +5954,17 @@ measure_cost(int cpu1, int cpu2, void *cache, unsigned int size)
measure_one(cache, size, cpu1, cpu1);
for (i = 0; i < ITERATIONS; i++)
- cost2 += measure_one(cache, size - i*1024, cpu1, cpu1);
+ cost2 += measure_one(cache, size - i * 1024, cpu1, cpu1);
measure_one(cache, size, cpu2, cpu2);
for (i = 0; i < ITERATIONS; i++)
- cost2 += measure_one(cache, size - i*1024, cpu2, cpu2);
+ cost2 += measure_one(cache, size - i * 1024, cpu2, cpu2);
/*
* Get the per-iteration migration cost:
*/
- do_div(cost1, 2*ITERATIONS);
- do_div(cost2, 2*ITERATIONS);
+ do_div(cost1, 2 * ITERATIONS);
+ do_div(cost2, 2 * ITERATIONS);
return cost1 - cost2;
}
@@ -5905,7 +6002,7 @@ static unsigned long long measure_migration_cost(int cpu1, int cpu2)
*/
cache = vmalloc(max_size);
if (!cache) {
- printk("could not vmalloc %d bytes for cache!\n", 2*max_size);
+ printk("could not vmalloc %d bytes for cache!\n", 2 * max_size);
return 1000000; /* return 1 msec on very small boxen */
}
@@ -5930,7 +6027,8 @@ static unsigned long long measure_migration_cost(int cpu1, int cpu2)
avg_fluct = (avg_fluct + fluct)/2;
if (migration_debug)
- printk("-> [%d][%d][%7d] %3ld.%ld [%3ld.%ld] (%ld): (%8Ld %8Ld)\n",
+ printk("-> [%d][%d][%7d] %3ld.%ld [%3ld.%ld] (%ld): "
+ "(%8Ld %8Ld)\n",
cpu1, cpu2, size,
(long)cost / 1000000,
((long)cost / 100000) % 10,
@@ -6025,20 +6123,18 @@ static void calibrate_migration_costs(const cpumask_t *cpu_map)
-1
#endif
);
- if (system_state == SYSTEM_BOOTING) {
- if (num_online_cpus() > 1) {
- printk("migration_cost=");
- for (distance = 0; distance <= max_distance; distance++) {
- if (distance)
- printk(",");
- printk("%ld", (long)migration_cost[distance] / 1000);
- }
- printk("\n");
+ if (system_state == SYSTEM_BOOTING && num_online_cpus() > 1) {
+ printk("migration_cost=");
+ for (distance = 0; distance <= max_distance; distance++) {
+ if (distance)
+ printk(",");
+ printk("%ld", (long)migration_cost[distance] / 1000);
}
+ printk("\n");
}
j1 = jiffies;
if (migration_debug)
- printk("migration: %ld seconds\n", (j1-j0)/HZ);
+ printk("migration: %ld seconds\n", (j1-j0) / HZ);
/*
* Move back to the original CPU. NUMA-Q gets confused
@@ -6135,10 +6231,13 @@ int sched_smt_power_savings = 0, sched_mc_power_savings = 0;
*/
#ifdef CONFIG_SCHED_SMT
static DEFINE_PER_CPU(struct sched_domain, cpu_domains);
-static struct sched_group sched_group_cpus[NR_CPUS];
+static DEFINE_PER_CPU(struct sched_group, sched_group_cpus);
-static int cpu_to_cpu_group(int cpu, const cpumask_t *cpu_map)
+static int cpu_to_cpu_group(int cpu, const cpumask_t *cpu_map,
+ struct sched_group **sg)
{
+ if (sg)
+ *sg = &per_cpu(sched_group_cpus, cpu);
return cpu;
}
#endif
@@ -6148,39 +6247,52 @@ static int cpu_to_cpu_group(int cpu, const cpumask_t *cpu_map)
*/
#ifdef CONFIG_SCHED_MC
static DEFINE_PER_CPU(struct sched_domain, core_domains);
-static struct sched_group sched_group_core[NR_CPUS];
+static DEFINE_PER_CPU(struct sched_group, sched_group_core);
#endif
#if defined(CONFIG_SCHED_MC) && defined(CONFIG_SCHED_SMT)
-static int cpu_to_core_group(int cpu, const cpumask_t *cpu_map)
+static int cpu_to_core_group(int cpu, const cpumask_t *cpu_map,
+ struct sched_group **sg)
{
+ int group;
cpumask_t mask = cpu_sibling_map[cpu];
cpus_and(mask, mask, *cpu_map);
- return first_cpu(mask);
+ group = first_cpu(mask);
+ if (sg)
+ *sg = &per_cpu(sched_group_core, group);
+ return group;
}
#elif defined(CONFIG_SCHED_MC)
-static int cpu_to_core_group(int cpu, const cpumask_t *cpu_map)
+static int cpu_to_core_group(int cpu, const cpumask_t *cpu_map,
+ struct sched_group **sg)
{
+ if (sg)
+ *sg = &per_cpu(sched_group_core, cpu);
return cpu;
}
#endif
static DEFINE_PER_CPU(struct sched_domain, phys_domains);
-static struct sched_group sched_group_phys[NR_CPUS];
+static DEFINE_PER_CPU(struct sched_group, sched_group_phys);
-static int cpu_to_phys_group(int cpu, const cpumask_t *cpu_map)
+static int cpu_to_phys_group(int cpu, const cpumask_t *cpu_map,
+ struct sched_group **sg)
{
+ int group;
#ifdef CONFIG_SCHED_MC
cpumask_t mask = cpu_coregroup_map(cpu);
cpus_and(mask, mask, *cpu_map);
- return first_cpu(mask);
+ group = first_cpu(mask);
#elif defined(CONFIG_SCHED_SMT)
cpumask_t mask = cpu_sibling_map[cpu];
cpus_and(mask, mask, *cpu_map);
- return first_cpu(mask);
+ group = first_cpu(mask);
#else
- return cpu;
+ group = cpu;
#endif
+ if (sg)
+ *sg = &per_cpu(sched_group_phys, group);
+ return group;
}
#ifdef CONFIG_NUMA
@@ -6193,12 +6305,22 @@ static DEFINE_PER_CPU(struct sched_domain, node_domains);
static struct sched_group **sched_group_nodes_bycpu[NR_CPUS];
static DEFINE_PER_CPU(struct sched_domain, allnodes_domains);
-static struct sched_group *sched_group_allnodes_bycpu[NR_CPUS];
+static DEFINE_PER_CPU(struct sched_group, sched_group_allnodes);
-static int cpu_to_allnodes_group(int cpu, const cpumask_t *cpu_map)
+static int cpu_to_allnodes_group(int cpu, const cpumask_t *cpu_map,
+ struct sched_group **sg)
{
- return cpu_to_node(cpu);
+ cpumask_t nodemask = node_to_cpumask(cpu_to_node(cpu));
+ int group;
+
+ cpus_and(nodemask, nodemask, *cpu_map);
+ group = first_cpu(nodemask);
+
+ if (sg)
+ *sg = &per_cpu(sched_group_allnodes, group);
+ return group;
}
+
static void init_numa_sched_groups_power(struct sched_group *group_head)
{
struct sched_group *sg = group_head;
@@ -6234,16 +6356,9 @@ static void free_sched_groups(const cpumask_t *cpu_map)
int cpu, i;
for_each_cpu_mask(cpu, *cpu_map) {
- struct sched_group *sched_group_allnodes
- = sched_group_allnodes_bycpu[cpu];
struct sched_group **sched_group_nodes
= sched_group_nodes_bycpu[cpu];
- if (sched_group_allnodes) {
- kfree(sched_group_allnodes);
- sched_group_allnodes_bycpu[cpu] = NULL;
- }
-
if (!sched_group_nodes)
continue;
@@ -6337,7 +6452,7 @@ static int build_sched_domains(const cpumask_t *cpu_map)
struct sched_domain *sd;
#ifdef CONFIG_NUMA
struct sched_group **sched_group_nodes = NULL;
- struct sched_group *sched_group_allnodes = NULL;
+ int sd_allnodes = 0;
/*
* Allocate the per-node list of sched groups
@@ -6355,7 +6470,6 @@ static int build_sched_domains(const cpumask_t *cpu_map)
* Set up domains for cpus specified by the cpu_map.
*/
for_each_cpu_mask(i, *cpu_map) {
- int group;
struct sched_domain *sd = NULL, *p;
cpumask_t nodemask = node_to_cpumask(cpu_to_node(i));
@@ -6364,26 +6478,12 @@ static int build_sched_domains(const cpumask_t *cpu_map)
#ifdef CONFIG_NUMA
if (cpus_weight(*cpu_map)
> SD_NODES_PER_DOMAIN*cpus_weight(nodemask)) {
- if (!sched_group_allnodes) {
- sched_group_allnodes
- = kmalloc_node(sizeof(struct sched_group)
- * MAX_NUMNODES,
- GFP_KERNEL,
- cpu_to_node(i));
- if (!sched_group_allnodes) {
- printk(KERN_WARNING
- "Can not alloc allnodes sched group\n");
- goto error;
- }
- sched_group_allnodes_bycpu[i]
- = sched_group_allnodes;
- }
sd = &per_cpu(allnodes_domains, i);
*sd = SD_ALLNODES_INIT;
sd->span = *cpu_map;
- group = cpu_to_allnodes_group(i, cpu_map);
- sd->groups = &sched_group_allnodes[group];
+ cpu_to_allnodes_group(i, cpu_map, &sd->groups);
p = sd;
+ sd_allnodes = 1;
} else
p = NULL;
@@ -6398,36 +6498,33 @@ static int build_sched_domains(const cpumask_t *cpu_map)
p = sd;
sd = &per_cpu(phys_domains, i);
- group = cpu_to_phys_group(i, cpu_map);
*sd = SD_CPU_INIT;
sd->span = nodemask;
sd->parent = p;
if (p)
p->child = sd;
- sd->groups = &sched_group_phys[group];
+ cpu_to_phys_group(i, cpu_map, &sd->groups);
#ifdef CONFIG_SCHED_MC
p = sd;
sd = &per_cpu(core_domains, i);
- group = cpu_to_core_group(i, cpu_map);
*sd = SD_MC_INIT;
sd->span = cpu_coregroup_map(i);
cpus_and(sd->span, sd->span, *cpu_map);
sd->parent = p;
p->child = sd;
- sd->groups = &sched_group_core[group];
+ cpu_to_core_group(i, cpu_map, &sd->groups);
#endif
#ifdef CONFIG_SCHED_SMT
p = sd;
sd = &per_cpu(cpu_domains, i);
- group = cpu_to_cpu_group(i, cpu_map);
*sd = SD_SIBLING_INIT;
sd->span = cpu_sibling_map[i];
cpus_and(sd->span, sd->span, *cpu_map);
sd->parent = p;
p->child = sd;
- sd->groups = &sched_group_cpus[group];
+ cpu_to_cpu_group(i, cpu_map, &sd->groups);
#endif
}
@@ -6439,8 +6536,7 @@ static int build_sched_domains(const cpumask_t *cpu_map)
if (i != first_cpu(this_sibling_map))
continue;
- init_sched_build_groups(sched_group_cpus, this_sibling_map,
- cpu_map, &cpu_to_cpu_group);
+ init_sched_build_groups(this_sibling_map, cpu_map, &cpu_to_cpu_group);
}
#endif
@@ -6451,8 +6547,7 @@ static int build_sched_domains(const cpumask_t *cpu_map)
cpus_and(this_core_map, this_core_map, *cpu_map);
if (i != first_cpu(this_core_map))
continue;
- init_sched_build_groups(sched_group_core, this_core_map,
- cpu_map, &cpu_to_core_group);
+ init_sched_build_groups(this_core_map, cpu_map, &cpu_to_core_group);
}
#endif
@@ -6465,15 +6560,13 @@ static int build_sched_domains(const cpumask_t *cpu_map)
if (cpus_empty(nodemask))
continue;
- init_sched_build_groups(sched_group_phys, nodemask,
- cpu_map, &cpu_to_phys_group);
+ init_sched_build_groups(nodemask, cpu_map, &cpu_to_phys_group);
}
#ifdef CONFIG_NUMA
/* Set up node groups */
- if (sched_group_allnodes)
- init_sched_build_groups(sched_group_allnodes, *cpu_map,
- cpu_map, &cpu_to_allnodes_group);
+ if (sd_allnodes)
+ init_sched_build_groups(*cpu_map, cpu_map, &cpu_to_allnodes_group);
for (i = 0; i < MAX_NUMNODES; i++) {
/* Set up node groups */
@@ -6565,10 +6658,10 @@ static int build_sched_domains(const cpumask_t *cpu_map)
for (i = 0; i < MAX_NUMNODES; i++)
init_numa_sched_groups_power(sched_group_nodes[i]);
- if (sched_group_allnodes) {
- int group = cpu_to_allnodes_group(first_cpu(*cpu_map), cpu_map);
- struct sched_group *sg = &sched_group_allnodes[group];
+ if (sd_allnodes) {
+ struct sched_group *sg;
+ cpu_to_allnodes_group(first_cpu(*cpu_map), cpu_map, &sg);
init_numa_sched_groups_power(sg);
}
#endif
@@ -6847,6 +6940,10 @@ void __init sched_init(void)
set_load_weight(&init_task);
+#ifdef CONFIG_SMP
+ open_softirq(SCHED_SOFTIRQ, run_rebalance_domains, NULL);
+#endif
+
#ifdef CONFIG_RT_MUTEXES
plist_head_init(&init_task.pi_waiters, &init_task.pi_lock);
#endif
@@ -6882,6 +6979,8 @@ void __might_sleep(char *file, int line)
printk("in_atomic():%d, irqs_disabled():%d\n",
in_atomic(), irqs_disabled());
debug_show_held_locks(current);
+ if (irqs_disabled())
+ print_irqtrace_events(current);
dump_stack();
}
#endif
diff --git a/kernel/signal.c b/kernel/signal.c
index 1921ffdc5e77..5630255d2e2a 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -1705,7 +1705,9 @@ finish_stop(int stop_count)
read_unlock(&tasklist_lock);
}
- schedule();
+ do {
+ schedule();
+ } while (try_to_freeze());
/*
* Now we don't run again until continued.
*/
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index 025fcb3c66f8..600b33358ded 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -65,7 +65,6 @@ extern int sysctl_overcommit_memory;
extern int sysctl_overcommit_ratio;
extern int sysctl_panic_on_oom;
extern int max_threads;
-extern int sysrq_enabled;
extern int core_uses_pid;
extern int suid_dumpable;
extern char core_pattern[];
@@ -133,7 +132,7 @@ extern int max_lock_depth;
#ifdef CONFIG_SYSCTL_SYSCALL
static int parse_table(int __user *, int, void __user *, size_t __user *,
- void __user *, size_t, ctl_table *, void **);
+ void __user *, size_t, ctl_table *);
#endif
static int proc_do_uts_string(ctl_table *table, int write, struct file *filp,
@@ -141,12 +140,12 @@ static int proc_do_uts_string(ctl_table *table, int write, struct file *filp,
static int sysctl_uts_string(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context);
+ void __user *newval, size_t newlen);
#ifdef CONFIG_SYSVIPC
static int sysctl_ipc_data(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context);
+ void __user *newval, size_t newlen);
#endif
#ifdef CONFIG_PROC_SYSCTL
@@ -543,7 +542,7 @@ static ctl_table kern_table[] = {
{
.ctl_name = KERN_SYSRQ,
.procname = "sysrq",
- .data = &sysrq_enabled,
+ .data = &__sysrq_enabled,
.maxlen = sizeof (int),
.mode = 0644,
.proc_handler = &proc_dointvec,
@@ -1243,7 +1242,6 @@ int do_sysctl(int __user *name, int nlen, void __user *oldval, size_t __user *ol
do {
struct ctl_table_header *head =
list_entry(tmp, struct ctl_table_header, ctl_entry);
- void *context = NULL;
if (!use_table(head))
continue;
@@ -1251,9 +1249,7 @@ int do_sysctl(int __user *name, int nlen, void __user *oldval, size_t __user *ol
spin_unlock(&sysctl_lock);
error = parse_table(name, nlen, oldval, oldlenp,
- newval, newlen, head->ctl_table,
- &context);
- kfree(context);
+ newval, newlen, head->ctl_table);
spin_lock(&sysctl_lock);
unuse_table(head);
@@ -1309,7 +1305,7 @@ static inline int ctl_perm(ctl_table *table, int op)
static int parse_table(int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
void __user *newval, size_t newlen,
- ctl_table *table, void **context)
+ ctl_table *table)
{
int n;
repeat:
@@ -1329,7 +1325,7 @@ repeat:
error = table->strategy(
table, name, nlen,
oldval, oldlenp,
- newval, newlen, context);
+ newval, newlen);
if (error)
return error;
}
@@ -1340,7 +1336,7 @@ repeat:
}
error = do_sysctl_strategy(table, name, nlen,
oldval, oldlenp,
- newval, newlen, context);
+ newval, newlen);
return error;
}
}
@@ -1351,7 +1347,7 @@ repeat:
int do_sysctl_strategy (ctl_table *table,
int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
int op = 0, rc;
size_t len;
@@ -1365,7 +1361,7 @@ int do_sysctl_strategy (ctl_table *table,
if (table->strategy) {
rc = table->strategy(table, name, nlen, oldval, oldlenp,
- newval, newlen, context);
+ newval, newlen);
if (rc < 0)
return rc;
if (rc > 0)
@@ -1931,9 +1927,6 @@ int proc_dointvec(ctl_table *table, int write, struct file *filp,
#define OP_SET 0
#define OP_AND 1
-#define OP_OR 2
-#define OP_MAX 3
-#define OP_MIN 4
static int do_proc_dointvec_bset_conv(int *negp, unsigned long *lvalp,
int *valp,
@@ -1945,13 +1938,6 @@ static int do_proc_dointvec_bset_conv(int *negp, unsigned long *lvalp,
switch(op) {
case OP_SET: *valp = val; break;
case OP_AND: *valp &= val; break;
- case OP_OR: *valp |= val; break;
- case OP_MAX: if(*valp < val)
- *valp = val;
- break;
- case OP_MIN: if(*valp > val)
- *valp = val;
- break;
}
} else {
int val = *valp;
@@ -2408,6 +2394,17 @@ static int proc_do_ipc_string(ctl_table *table, int write, struct file *filp,
{
return -ENOSYS;
}
+static int proc_ipc_dointvec(ctl_table *table, int write, struct file *filp,
+ void __user *buffer, size_t *lenp, loff_t *ppos)
+{
+ return -ENOSYS;
+}
+static int proc_ipc_doulongvec_minmax(ctl_table *table, int write,
+ struct file *filp, void __user *buffer,
+ size_t *lenp, loff_t *ppos)
+{
+ return -ENOSYS;
+}
#endif
int proc_dointvec(ctl_table *table, int write, struct file *filp,
@@ -2472,7 +2469,7 @@ int proc_doulongvec_ms_jiffies_minmax(ctl_table *table, int write,
/* The generic string strategy routine: */
int sysctl_string(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
if (!table->data || !table->maxlen)
return -ENOTDIR;
@@ -2518,7 +2515,7 @@ int sysctl_string(ctl_table *table, int __user *name, int nlen,
*/
int sysctl_intvec(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
if (newval && newlen) {
@@ -2554,7 +2551,7 @@ int sysctl_intvec(ctl_table *table, int __user *name, int nlen,
/* Strategy function to convert jiffies to seconds */
int sysctl_jiffies(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
if (oldval) {
size_t olen;
@@ -2582,7 +2579,7 @@ int sysctl_jiffies(ctl_table *table, int __user *name, int nlen,
/* Strategy function to convert jiffies to seconds */
int sysctl_ms_jiffies(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
if (oldval) {
size_t olen;
@@ -2611,7 +2608,7 @@ int sysctl_ms_jiffies(ctl_table *table, int __user *name, int nlen,
/* The generic string strategy routine: */
static int sysctl_uts_string(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
struct ctl_table uts_table;
int r, write;
@@ -2619,7 +2616,7 @@ static int sysctl_uts_string(ctl_table *table, int __user *name, int nlen,
memcpy(&uts_table, table, sizeof(uts_table));
uts_table.data = get_uts(table, write);
r = sysctl_string(&uts_table, name, nlen,
- oldval, oldlenp, newval, newlen, context);
+ oldval, oldlenp, newval, newlen);
put_uts(table, write, uts_table.data);
return r;
}
@@ -2628,7 +2625,7 @@ static int sysctl_uts_string(ctl_table *table, int __user *name, int nlen,
/* The generic sysctl ipc data routine. */
static int sysctl_ipc_data(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
size_t len;
void *data;
@@ -2703,41 +2700,41 @@ out:
int sysctl_string(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
return -ENOSYS;
}
int sysctl_intvec(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
return -ENOSYS;
}
int sysctl_jiffies(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
return -ENOSYS;
}
int sysctl_ms_jiffies(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
return -ENOSYS;
}
static int sysctl_uts_string(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
return -ENOSYS;
}
static int sysctl_ipc_data(ctl_table *table, int __user *name, int nlen,
void __user *oldval, size_t __user *oldlenp,
- void __user *newval, size_t newlen, void **context)
+ void __user *newval, size_t newlen)
{
return -ENOSYS;
}
diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c
index 74eca5939bd9..22504afc0d34 100644
--- a/kernel/time/clocksource.c
+++ b/kernel/time/clocksource.c
@@ -156,7 +156,7 @@ int clocksource_register(struct clocksource *c)
/* check if clocksource is already registered */
if (is_registered_source(c)) {
printk("register_clocksource: Cannot register %s. "
- "Already registered!", c->name);
+ "Already registered!", c->name);
ret = -EBUSY;
} else {
/* register it */
@@ -186,6 +186,7 @@ void clocksource_reselect(void)
}
EXPORT_SYMBOL(clocksource_reselect);
+#ifdef CONFIG_SYSFS
/**
* sysfs_show_current_clocksources - sysfs interface for current clocksource
* @dev: unused
@@ -275,10 +276,10 @@ sysfs_show_available_clocksources(struct sys_device *dev, char *buf)
* Sysfs setup bits:
*/
static SYSDEV_ATTR(current_clocksource, 0600, sysfs_show_current_clocksources,
- sysfs_override_clocksource);
+ sysfs_override_clocksource);
static SYSDEV_ATTR(available_clocksource, 0600,
- sysfs_show_available_clocksources, NULL);
+ sysfs_show_available_clocksources, NULL);
static struct sysdev_class clocksource_sysclass = {
set_kset_name("clocksource"),
@@ -307,6 +308,7 @@ static int __init init_clocksource_sysfs(void)
}
device_initcall(init_clocksource_sysfs);
+#endif /* CONFIG_SYSFS */
/**
* boot_override_clocksource - boot clock override
diff --git a/kernel/timer.c b/kernel/timer.c
index c1c7fbcffec1..feddf817baa5 100644
--- a/kernel/timer.c
+++ b/kernel/timer.c
@@ -80,6 +80,138 @@ tvec_base_t boot_tvec_bases;
EXPORT_SYMBOL(boot_tvec_bases);
static DEFINE_PER_CPU(tvec_base_t *, tvec_bases) = &boot_tvec_bases;
+/**
+ * __round_jiffies - function to round jiffies to a full second
+ * @j: the time in (absolute) jiffies that should be rounded
+ * @cpu: the processor number on which the timeout will happen
+ *
+ * __round_jiffies rounds an absolute time in the future (in jiffies)
+ * up or down to (approximately) full seconds. This is useful for timers
+ * for which the exact time they fire does not matter too much, as long as
+ * they fire approximately every X seconds.
+ *
+ * By rounding these timers to whole seconds, all such timers will fire
+ * at the same time, rather than at various times spread out. The goal
+ * of this is to have the CPU wake up less, which saves power.
+ *
+ * The exact rounding is skewed for each processor to avoid all
+ * processors firing at the exact same time, which could lead
+ * to lock contention or spurious cache line bouncing.
+ *
+ * The return value is the rounded version of the "j" parameter.
+ */
+unsigned long __round_jiffies(unsigned long j, int cpu)
+{
+ int rem;
+ unsigned long original = j;
+
+ /*
+ * We don't want all cpus firing their timers at once hitting the
+ * same lock or cachelines, so we skew each extra cpu with an extra
+ * 3 jiffies. This 3 jiffies came originally from the mm/ code which
+ * already did this.
+ * The skew is done by adding 3*cpunr, then round, then subtract this
+ * extra offset again.
+ */
+ j += cpu * 3;
+
+ rem = j % HZ;
+
+ /*
+ * If the target jiffie is just after a whole second (which can happen
+ * due to delays of the timer irq, long irq off times etc etc) then
+ * we should round down to the whole second, not up. Use 1/4th second
+ * as cutoff for this rounding as an extreme upper bound for this.
+ */
+ if (rem < HZ/4) /* round down */
+ j = j - rem;
+ else /* round up */
+ j = j - rem + HZ;
+
+ /* now that we have rounded, subtract the extra skew again */
+ j -= cpu * 3;
+
+ if (j <= jiffies) /* rounding ate our timeout entirely; */
+ return original;
+ return j;
+}
+EXPORT_SYMBOL_GPL(__round_jiffies);
+
+/**
+ * __round_jiffies_relative - function to round jiffies to a full second
+ * @j: the time in (relative) jiffies that should be rounded
+ * @cpu: the processor number on which the timeout will happen
+ *
+ * __round_jiffies_relative rounds a time delta in the future (in jiffies)
+ * up or down to (approximately) full seconds. This is useful for timers
+ * for which the exact time they fire does not matter too much, as long as
+ * they fire approximately every X seconds.
+ *
+ * By rounding these timers to whole seconds, all such timers will fire
+ * at the same time, rather than at various times spread out. The goal
+ * of this is to have the CPU wake up less, which saves power.
+ *
+ * The exact rounding is skewed for each processor to avoid all
+ * processors firing at the exact same time, which could lead
+ * to lock contention or spurious cache line bouncing.
+ *
+ * The return value is the rounded version of the "j" parameter.
+ */
+unsigned long __round_jiffies_relative(unsigned long j, int cpu)
+{
+ /*
+ * In theory the following code can skip a jiffy in case jiffies
+ * increments right between the addition and the later subtraction.
+ * However since the entire point of this function is to use approximate
+ * timeouts, it's entirely ok to not handle that.
+ */
+ return __round_jiffies(j + jiffies, cpu) - jiffies;
+}
+EXPORT_SYMBOL_GPL(__round_jiffies_relative);
+
+/**
+ * round_jiffies - function to round jiffies to a full second
+ * @j: the time in (absolute) jiffies that should be rounded
+ *
+ * round_jiffies rounds an absolute time in the future (in jiffies)
+ * up or down to (approximately) full seconds. This is useful for timers
+ * for which the exact time they fire does not matter too much, as long as
+ * they fire approximately every X seconds.
+ *
+ * By rounding these timers to whole seconds, all such timers will fire
+ * at the same time, rather than at various times spread out. The goal
+ * of this is to have the CPU wake up less, which saves power.
+ *
+ * The return value is the rounded version of the "j" parameter.
+ */
+unsigned long round_jiffies(unsigned long j)
+{
+ return __round_jiffies(j, raw_smp_processor_id());
+}
+EXPORT_SYMBOL_GPL(round_jiffies);
+
+/**
+ * round_jiffies_relative - function to round jiffies to a full second
+ * @j: the time in (relative) jiffies that should be rounded
+ *
+ * round_jiffies_relative rounds a time delta in the future (in jiffies)
+ * up or down to (approximately) full seconds. This is useful for timers
+ * for which the exact time they fire does not matter too much, as long as
+ * they fire approximately every X seconds.
+ *
+ * By rounding these timers to whole seconds, all such timers will fire
+ * at the same time, rather than at various times spread out. The goal
+ * of this is to have the CPU wake up less, which saves power.
+ *
+ * The return value is the rounded version of the "j" parameter.
+ */
+unsigned long round_jiffies_relative(unsigned long j)
+{
+ return __round_jiffies_relative(j, raw_smp_processor_id());
+}
+EXPORT_SYMBOL_GPL(round_jiffies_relative);
+
+
static inline void set_running_timer(tvec_base_t *base,
struct timer_list *timer)
{
@@ -714,7 +846,7 @@ static int change_clocksource(void)
clock = new;
clock->cycle_last = now;
printk(KERN_INFO "Time: %s clocksource has been installed.\n",
- clock->name);
+ clock->name);
return 1;
} else if (clock->update_callback) {
return clock->update_callback();
@@ -722,7 +854,10 @@ static int change_clocksource(void)
return 0;
}
#else
-#define change_clocksource() (0)
+static inline int change_clocksource(void)
+{
+ return 0;
+}
#endif
/**
@@ -820,7 +955,8 @@ device_initcall(timekeeping_init_device);
* If the error is already larger, we look ahead even further
* to compensate for late or lost adjustments.
*/
-static __always_inline int clocksource_bigadjust(s64 error, s64 *interval, s64 *offset)
+static __always_inline int clocksource_bigadjust(s64 error, s64 *interval,
+ s64 *offset)
{
s64 tick_error, i;
u32 look_ahead, adj;
@@ -844,7 +980,8 @@ static __always_inline int clocksource_bigadjust(s64 error, s64 *interval, s64 *
* Now calculate the error in (1 << look_ahead) ticks, but first
* remove the single look ahead already included in the error.
*/
- tick_error = current_tick_length() >> (TICK_LENGTH_SHIFT - clock->shift + 1);
+ tick_error = current_tick_length() >>
+ (TICK_LENGTH_SHIFT - clock->shift + 1);
tick_error -= clock->xtime_interval >> 1;
error = ((error - tick_error) >> look_ahead) + tick_error;
@@ -896,7 +1033,8 @@ static void clocksource_adjust(struct clocksource *clock, s64 offset)
clock->mult += adj;
clock->xtime_interval += interval;
clock->xtime_nsec -= offset;
- clock->error -= (interval - offset) << (TICK_LENGTH_SHIFT - clock->shift);
+ clock->error -= (interval - offset) <<
+ (TICK_LENGTH_SHIFT - clock->shift);
}
/**
@@ -1008,11 +1146,15 @@ static inline void calc_load(unsigned long ticks)
unsigned long active_tasks; /* fixed-point */
static int count = LOAD_FREQ;
- active_tasks = count_active_tasks();
- for (count -= ticks; count < 0; count += LOAD_FREQ) {
- CALC_LOAD(avenrun[0], EXP_1, active_tasks);
- CALC_LOAD(avenrun[1], EXP_5, active_tasks);
- CALC_LOAD(avenrun[2], EXP_15, active_tasks);
+ count -= ticks;
+ if (unlikely(count < 0)) {
+ active_tasks = count_active_tasks();
+ do {
+ CALC_LOAD(avenrun[0], EXP_1, active_tasks);
+ CALC_LOAD(avenrun[1], EXP_5, active_tasks);
+ CALC_LOAD(avenrun[2], EXP_15, active_tasks);
+ count += LOAD_FREQ;
+ } while (count < 0);
}
}
diff --git a/kernel/tsacct.c b/kernel/tsacct.c
index 96f77013d3f0..baacc3691415 100644
--- a/kernel/tsacct.c
+++ b/kernel/tsacct.c
@@ -96,6 +96,15 @@ void xacct_add_tsk(struct taskstats *stats, struct task_struct *p)
stats->write_char = p->wchar;
stats->read_syscalls = p->syscr;
stats->write_syscalls = p->syscw;
+#ifdef CONFIG_TASK_IO_ACCOUNTING
+ stats->read_bytes = p->ioac.read_bytes;
+ stats->write_bytes = p->ioac.write_bytes;
+ stats->cancelled_write_bytes = p->ioac.cancelled_write_bytes;
+#else
+ stats->read_bytes = 0;
+ stats->write_bytes = 0;
+ stats->cancelled_write_bytes = 0;
+#endif
}
#undef KB
#undef MB
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-17 23:43:31 +0000