From: Thomas Gleixner <tglx@linutronix.de>
Date: Thu, 12 Oct 2017 17:11:19 +0200
Subject: rtmutex: add sleeping lock implementation
Git-repo: git://git.kernel.org/pub/scm/linux/kernel/git/rt/linux-rt-devel.git
Git-commit: da77ceac3d20f27310a07a7c346a4ee6b40d6c28
Patch-mainline: Queued in subsystem maintainer repository
References: SLE Realtime Extension
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Daniel Wagner <dwagner@suse.de>
---
include/linux/kernel.h | 5
include/linux/rtmutex.h | 21 +
include/linux/sched.h | 8
include/linux/sched/wake_q.h | 13 +
include/linux/spinlock_rt.h | 156 +++++++++++++
include/linux/spinlock_types_rt.h | 48 ++++
kernel/fork.c | 1
kernel/futex.c | 11
kernel/locking/rtmutex.c | 437 ++++++++++++++++++++++++++++++++++----
kernel/locking/rtmutex_common.h | 14 -
kernel/sched/core.c | 39 ++-
11 files changed, 694 insertions(+), 59 deletions(-)
create mode 100644 include/linux/spinlock_rt.h
create mode 100644 include/linux/spinlock_types_rt.h
--- a/include/linux/kernel.h
+++ b/include/linux/kernel.h
@@ -234,6 +234,10 @@ extern void __cant_sleep(const char *fil
*/
# define might_sleep() \
do { __might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0)
+
+# define might_sleep_no_state_check() \
+ do { ___might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0)
+
/**
* cant_sleep - annotation for functions that cannot sleep
*
@@ -265,6 +269,7 @@ extern void __cant_sleep(const char *fil
static inline void __might_sleep(const char *file, int line,
int preempt_offset) { }
# define might_sleep() do { might_resched(); } while (0)
+# define might_sleep_no_state_check() do { might_resched(); } while (0)
# define cant_sleep() do { } while (0)
# define sched_annotate_sleep() do { } while (0)
# define non_block_start() do { } while (0)
--- a/include/linux/rtmutex.h
+++ b/include/linux/rtmutex.h
@@ -14,11 +14,15 @@
#define __LINUX_RT_MUTEX_H
#include <linux/linkage.h>
-#include <linux/rbtree.h>
#include <linux/spinlock_types_raw.h>
+#include <linux/rbtree.h>
extern int max_lock_depth; /* for sysctl */
+#ifdef CONFIG_DEBUG_MUTEXES
+#include <linux/debug_locks.h>
+#endif
+
/**
* The rt_mutex structure
*
@@ -31,8 +35,8 @@ struct rt_mutex {
raw_spinlock_t wait_lock;
struct rb_root_cached waiters;
struct task_struct *owner;
-#ifdef CONFIG_DEBUG_RT_MUTEXES
int save_state;
+#ifdef CONFIG_DEBUG_RT_MUTEXES
const char *name, *file;
int line;
void *magic;
@@ -82,16 +86,23 @@ do { \
#define __DEP_MAP_RT_MUTEX_INITIALIZER(mutexname)
#endif
-#define __RT_MUTEX_INITIALIZER(mutexname) \
- { .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \
+#define __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \
+ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \
, .waiters = RB_ROOT_CACHED \
, .owner = NULL \
__DEBUG_RT_MUTEX_INITIALIZER(mutexname) \
- __DEP_MAP_RT_MUTEX_INITIALIZER(mutexname)}
+ __DEP_MAP_RT_MUTEX_INITIALIZER(mutexname)
+
+#define __RT_MUTEX_INITIALIZER(mutexname) \
+ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) }
#define DEFINE_RT_MUTEX(mutexname) \
struct rt_mutex mutexname = __RT_MUTEX_INITIALIZER(mutexname)
+#define __RT_MUTEX_INITIALIZER_SAVE_STATE(mutexname) \
+ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \
+ , .save_state = 1 }
+
/**
* rt_mutex_is_locked - is the mutex locked
* @lock: the mutex to be queried
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -137,6 +137,9 @@ struct task_group;
smp_store_mb(current->state, (state_value)); \
} while (0)
+#define __set_current_state_no_track(state_value) \
+ current->state = (state_value);
+
#define set_special_state(state_value) \
do { \
unsigned long flags; /* may shadow */ \
@@ -146,6 +149,7 @@ struct task_group;
current->state = (state_value); \
raw_spin_unlock_irqrestore(¤t->pi_lock, flags); \
} while (0)
+
#else
/*
* set_current_state() includes a barrier so that the write of current->state
@@ -190,6 +194,9 @@ struct task_group;
#define set_current_state(state_value) \
smp_store_mb(current->state, (state_value))
+#define __set_current_state_no_track(state_value) \
+ __set_current_state(state_value)
+
/*
* set_special_state() should be used for those states when the blocking task
* can not use the regular condition based wait-loop. In that case we must
@@ -965,6 +972,7 @@ struct task_struct {
raw_spinlock_t pi_lock;
struct wake_q_node wake_q;
+ struct wake_q_node wake_q_sleeper;
#ifdef CONFIG_RT_MUTEXES
/* PI waiters blocked on a rt_mutex held by this task: */
--- a/include/linux/sched/wake_q.h
+++ b/include/linux/sched/wake_q.h
@@ -58,6 +58,17 @@ static inline bool wake_q_empty(struct w
extern void wake_q_add(struct wake_q_head *head, struct task_struct *task);
extern void wake_q_add_safe(struct wake_q_head *head, struct task_struct *task);
-extern void wake_up_q(struct wake_q_head *head);
+extern void wake_q_add_sleeper(struct wake_q_head *head, struct task_struct *task);
+extern void __wake_up_q(struct wake_q_head *head, bool sleeper);
+
+static inline void wake_up_q(struct wake_q_head *head)
+{
+ __wake_up_q(head, false);
+}
+
+static inline void wake_up_q_sleeper(struct wake_q_head *head)
+{
+ __wake_up_q(head, true);
+}
#endif /* _LINUX_SCHED_WAKE_Q_H */
--- /dev/null
+++ b/include/linux/spinlock_rt.h
@@ -0,0 +1,156 @@
+#ifndef __LINUX_SPINLOCK_RT_H
+#define __LINUX_SPINLOCK_RT_H
+
+#ifndef __LINUX_SPINLOCK_H
+#error Do not include directly. Use spinlock.h
+#endif
+
+#include <linux/bug.h>
+
+extern void
+__rt_spin_lock_init(spinlock_t *lock, const char *name, struct lock_class_key *key);
+
+#define spin_lock_init(slock) \
+do { \
+ static struct lock_class_key __key; \
+ \
+ rt_mutex_init(&(slock)->lock); \
+ __rt_spin_lock_init(slock, #slock, &__key); \
+} while (0)
+
+extern void __lockfunc rt_spin_lock(spinlock_t *lock);
+extern unsigned long __lockfunc rt_spin_lock_trace_flags(spinlock_t *lock);
+extern void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass);
+extern void __lockfunc rt_spin_unlock(spinlock_t *lock);
+extern void __lockfunc rt_spin_unlock_wait(spinlock_t *lock);
+extern int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags);
+extern int __lockfunc rt_spin_trylock_bh(spinlock_t *lock);
+extern int __lockfunc rt_spin_trylock(spinlock_t *lock);
+extern int atomic_dec_and_spin_lock(atomic_t *atomic, spinlock_t *lock);
+
+/*
+ * lockdep-less calls, for derived types like rwlock:
+ * (for trylock they can use rt_mutex_trylock() directly.
+ * Migrate disable handling must be done at the call site.
+ */
+extern void __lockfunc __rt_spin_lock(struct rt_mutex *lock);
+extern void __lockfunc __rt_spin_trylock(struct rt_mutex *lock);
+extern void __lockfunc __rt_spin_unlock(struct rt_mutex *lock);
+
+#define spin_lock(lock) rt_spin_lock(lock)
+
+#define spin_lock_bh(lock) \
+ do { \
+ local_bh_disable(); \
+ rt_spin_lock(lock); \
+ } while (0)
+
+#define spin_lock_irq(lock) spin_lock(lock)
+
+#define spin_do_trylock(lock) __cond_lock(lock, rt_spin_trylock(lock))
+
+#define spin_trylock(lock) \
+({ \
+ int __locked; \
+ __locked = spin_do_trylock(lock); \
+ __locked; \
+})
+
+#ifdef CONFIG_LOCKDEP
+# define spin_lock_nested(lock, subclass) \
+ do { \
+ rt_spin_lock_nested(lock, subclass); \
+ } while (0)
+
+#define spin_lock_bh_nested(lock, subclass) \
+ do { \
+ local_bh_disable(); \
+ rt_spin_lock_nested(lock, subclass); \
+ } while (0)
+
+# define spin_lock_irqsave_nested(lock, flags, subclass) \
+ do { \
+ typecheck(unsigned long, flags); \
+ flags = 0; \
+ rt_spin_lock_nested(lock, subclass); \
+ } while (0)
+#else
+# define spin_lock_nested(lock, subclass) spin_lock(lock)
+# define spin_lock_bh_nested(lock, subclass) spin_lock_bh(lock)
+
+# define spin_lock_irqsave_nested(lock, flags, subclass) \
+ do { \
+ typecheck(unsigned long, flags); \
+ flags = 0; \
+ spin_lock(lock); \
+ } while (0)
+#endif
+
+#define spin_lock_irqsave(lock, flags) \
+ do { \
+ typecheck(unsigned long, flags); \
+ flags = 0; \
+ spin_lock(lock); \
+ } while (0)
+
+static inline unsigned long spin_lock_trace_flags(spinlock_t *lock)
+{
+ unsigned long flags = 0;
+#ifdef CONFIG_TRACE_IRQFLAGS
+ flags = rt_spin_lock_trace_flags(lock);
+#else
+ spin_lock(lock); /* lock_local */
+#endif
+ return flags;
+}
+
+/* FIXME: we need rt_spin_lock_nest_lock */
+#define spin_lock_nest_lock(lock, nest_lock) spin_lock_nested(lock, 0)
+
+#define spin_unlock(lock) rt_spin_unlock(lock)
+
+#define spin_unlock_bh(lock) \
+ do { \
+ rt_spin_unlock(lock); \
+ local_bh_enable(); \
+ } while (0)
+
+#define spin_unlock_irq(lock) spin_unlock(lock)
+
+#define spin_unlock_irqrestore(lock, flags) \
+ do { \
+ typecheck(unsigned long, flags); \
+ (void) flags; \
+ spin_unlock(lock); \
+ } while (0)
+
+#define spin_trylock_bh(lock) __cond_lock(lock, rt_spin_trylock_bh(lock))
+#define spin_trylock_irq(lock) spin_trylock(lock)
+
+#define spin_trylock_irqsave(lock, flags) \
+ rt_spin_trylock_irqsave(lock, &(flags))
+
+#define spin_unlock_wait(lock) rt_spin_unlock_wait(lock)
+
+#ifdef CONFIG_GENERIC_LOCKBREAK
+# define spin_is_contended(lock) ((lock)->break_lock)
+#else
+# define spin_is_contended(lock) (((void)(lock), 0))
+#endif
+
+static inline int spin_can_lock(spinlock_t *lock)
+{
+ return !rt_mutex_is_locked(&lock->lock);
+}
+
+static inline int spin_is_locked(spinlock_t *lock)
+{
+ return rt_mutex_is_locked(&lock->lock);
+}
+
+static inline void assert_spin_locked(spinlock_t *lock)
+{
+ BUG_ON(!spin_is_locked(lock));
+}
+
+#endif
--- /dev/null
+++ b/include/linux/spinlock_types_rt.h
@@ -0,0 +1,48 @@
+#ifndef __LINUX_SPINLOCK_TYPES_RT_H
+#define __LINUX_SPINLOCK_TYPES_RT_H
+
+#ifndef __LINUX_SPINLOCK_TYPES_H
+#error "Do not include directly. Include spinlock_types.h instead"
+#endif
+
+#include <linux/cache.h>
+
+/*
+ * PREEMPT_RT: spinlocks - an RT mutex plus lock-break field:
+ */
+typedef struct spinlock {
+ struct rt_mutex lock;
+ unsigned int break_lock;
+#ifdef CONFIG_DEBUG_LOCK_ALLOC
+ struct lockdep_map dep_map;
+#endif
+} spinlock_t;
+
+#ifdef CONFIG_DEBUG_RT_MUTEXES
+# define __RT_SPIN_INITIALIZER(name) \
+ { \
+ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \
+ .save_state = 1, \
+ .file = __FILE__, \
+ .line = __LINE__ , \
+ }
+#else
+# define __RT_SPIN_INITIALIZER(name) \
+ { \
+ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \
+ .save_state = 1, \
+ }
+#endif
+
+/*
+.wait_list = PLIST_HEAD_INIT_RAW((name).lock.wait_list, (name).lock.wait_lock)
+*/
+
+#define __SPIN_LOCK_UNLOCKED(name) \
+ { .lock = __RT_SPIN_INITIALIZER(name.lock), \
+ SPIN_DEP_MAP_INIT(name) }
+
+#define DEFINE_SPINLOCK(name) \
+ spinlock_t name = __SPIN_LOCK_UNLOCKED(name)
+
+#endif
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -937,6 +937,7 @@ static struct task_struct *dup_task_stru
tsk->splice_pipe = NULL;
tsk->task_frag.page = NULL;
tsk->wake_q.next = NULL;
+ tsk->wake_q_sleeper.next = NULL;
account_kernel_stack(tsk, 1);
--- a/kernel/futex.c
+++ b/kernel/futex.c
@@ -1496,6 +1496,7 @@ static int wake_futex_pi(u32 __user *uad
struct task_struct *new_owner;
bool postunlock = false;
DEFINE_WAKE_Q(wake_q);
+ DEFINE_WAKE_Q(wake_sleeper_q);
int ret = 0;
new_owner = rt_mutex_next_owner(&pi_state->pi_mutex);
@@ -1545,14 +1546,15 @@ static int wake_futex_pi(u32 __user *uad
* not fail.
*/
pi_state_update_owner(pi_state, new_owner);
- postunlock = __rt_mutex_futex_unlock(&pi_state->pi_mutex, &wake_q);
+ postunlock = __rt_mutex_futex_unlock(&pi_state->pi_mutex, &wake_q,
+ &wake_sleeper_q);
}
out_unlock:
raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock);
if (postunlock)
- rt_mutex_postunlock(&wake_q);
+ rt_mutex_postunlock(&wake_q, &wake_sleeper_q);
return ret;
}
@@ -2857,7 +2859,7 @@ retry_private:
goto no_block;
}
- rt_mutex_init_waiter(&rt_waiter);
+ rt_mutex_init_waiter(&rt_waiter, false);
/*
* On PREEMPT_RT_FULL, when hb->lock becomes an rt_mutex, we must not
@@ -3203,7 +3205,7 @@ static int futex_wait_requeue_pi(u32 __u
* The waiter is allocated on our stack, manipulated by the requeue
* code while we sleep on uaddr.
*/
- rt_mutex_init_waiter(&rt_waiter);
+ rt_mutex_init_waiter(&rt_waiter, false);
ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2, FUTEX_WRITE);
if (unlikely(ret != 0))
--- a/kernel/locking/rtmutex.c
+++ b/kernel/locking/rtmutex.c
@@ -8,6 +8,11 @@
* Copyright (C) 2005-2006 Timesys Corp., Thomas Gleixner <tglx@timesys.com>
* Copyright (C) 2005 Kihon Technologies Inc., Steven Rostedt
* Copyright (C) 2006 Esben Nielsen
+ * Adaptive Spinlocks:
+ * Copyright (C) 2008 Novell, Inc., Gregory Haskins, Sven Dietrich,
+ * and Peter Morreale,
+ * Adaptive Spinlocks simplification:
+ * Copyright (C) 2008 Red Hat, Inc., Steven Rostedt <srostedt@redhat.com>
*
* See Documentation/locking/rt-mutex-design.rst for details.
*/
@@ -235,7 +240,7 @@ static inline bool unlock_rt_mutex_safe(
* Only use with rt_mutex_waiter_{less,equal}()
*/
#define task_to_waiter(p) \
- &(struct rt_mutex_waiter){ .prio = (p)->prio, .deadline = (p)->dl.deadline }
+ &(struct rt_mutex_waiter){ .prio = (p)->prio, .deadline = (p)->dl.deadline, .task = (p) }
static inline int
rt_mutex_waiter_less(struct rt_mutex_waiter *left,
@@ -275,6 +280,27 @@ rt_mutex_waiter_equal(struct rt_mutex_wa
return 1;
}
+#define STEAL_NORMAL 0
+#define STEAL_LATERAL 1
+
+static inline int
+rt_mutex_steal(struct rt_mutex *lock, struct rt_mutex_waiter *waiter, int mode)
+{
+ struct rt_mutex_waiter *top_waiter = rt_mutex_top_waiter(lock);
+
+ if (waiter == top_waiter || rt_mutex_waiter_less(waiter, top_waiter))
+ return 1;
+
+ /*
+ * Note that RT tasks are excluded from lateral-steals
+ * to prevent the introduction of an unbounded latency.
+ */
+ if (mode == STEAL_NORMAL || rt_task(waiter->task))
+ return 0;
+
+ return rt_mutex_waiter_equal(waiter, top_waiter);
+}
+
static void
rt_mutex_enqueue(struct rt_mutex *lock, struct rt_mutex_waiter *waiter)
{
@@ -379,6 +405,14 @@ static bool rt_mutex_cond_detect_deadloc
return debug_rt_mutex_detect_deadlock(waiter, chwalk);
}
+static void rt_mutex_wake_waiter(struct rt_mutex_waiter *waiter)
+{
+ if (waiter->savestate)
+ wake_up_lock_sleeper(waiter->task);
+ else
+ wake_up_process(waiter->task);
+}
+
/*
* Max number of times we'll walk the boosting chain:
*/
@@ -703,13 +737,16 @@ static int rt_mutex_adjust_prio_chain(st
* follow here. This is the end of the chain we are walking.
*/
if (!rt_mutex_owner(lock)) {
+ struct rt_mutex_waiter *lock_top_waiter;
+
/*
* If the requeue [7] above changed the top waiter,
* then we need to wake the new top waiter up to try
* to get the lock.
*/
- if (prerequeue_top_waiter != rt_mutex_top_waiter(lock))
- wake_up_process(rt_mutex_top_waiter(lock)->task);
+ lock_top_waiter = rt_mutex_top_waiter(lock);
+ if (prerequeue_top_waiter != lock_top_waiter)
+ rt_mutex_wake_waiter(lock_top_waiter);
raw_spin_unlock_irq(&lock->wait_lock);
return 0;
}
@@ -810,9 +847,11 @@ static int rt_mutex_adjust_prio_chain(st
* @task: The task which wants to acquire the lock
* @waiter: The waiter that is queued to the lock's wait tree if the
* callsite called task_blocked_on_lock(), otherwise NULL
+ * @mode: Lock steal mode (STEAL_NORMAL, STEAL_LATERAL)
*/
-static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
- struct rt_mutex_waiter *waiter)
+static int __try_to_take_rt_mutex(struct rt_mutex *lock,
+ struct task_struct *task,
+ struct rt_mutex_waiter *waiter, int mode)
{
lockdep_assert_held(&lock->wait_lock);
@@ -848,12 +887,11 @@ static int try_to_take_rt_mutex(struct r
*/
if (waiter) {
/*
- * If waiter is not the highest priority waiter of
- * @lock, give up.
+ * If waiter is not the highest priority waiter of @lock,
+ * or its peer when lateral steal is allowed, give up.
*/
- if (waiter != rt_mutex_top_waiter(lock))
+ if (!rt_mutex_steal(lock, waiter, mode))
return 0;
-
/*
* We can acquire the lock. Remove the waiter from the
* lock waiters tree.
@@ -871,14 +909,12 @@ static int try_to_take_rt_mutex(struct r
*/
if (rt_mutex_has_waiters(lock)) {
/*
- * If @task->prio is greater than or equal to
- * the top waiter priority (kernel view),
- * @task lost.
+ * If @task->prio is greater than the top waiter
+ * priority (kernel view), or equal to it when a
+ * lateral steal is forbidden, @task lost.
*/
- if (!rt_mutex_waiter_less(task_to_waiter(task),
- rt_mutex_top_waiter(lock)))
+ if (!rt_mutex_steal(lock, task_to_waiter(task), mode))
return 0;
-
/*
* The current top waiter stays enqueued. We
* don't have to change anything in the lock
@@ -925,6 +961,295 @@ takeit:
return 1;
}
+#ifdef CONFIG_PREEMPT_RT
+/*
+ * preemptible spin_lock functions:
+ */
+static inline void rt_spin_lock_fastlock(struct rt_mutex *lock,
+ void (*slowfn)(struct rt_mutex *lock))
+{
+ might_sleep_no_state_check();
+
+ if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current)))
+ return;
+ else
+ slowfn(lock);
+}
+
+static inline void rt_spin_lock_fastunlock(struct rt_mutex *lock,
+ void (*slowfn)(struct rt_mutex *lock))
+{
+ if (likely(rt_mutex_cmpxchg_release(lock, current, NULL)))
+ return;
+ else
+ slowfn(lock);
+}
+#ifdef CONFIG_SMP
+/*
+ * Note that owner is a speculative pointer and dereferencing relies
+ * on rcu_read_lock() and the check against the lock owner.
+ */
+static int adaptive_wait(struct rt_mutex *lock,
+ struct task_struct *owner)
+{
+ int res = 0;
+
+ rcu_read_lock();
+ for (;;) {
+ if (owner != rt_mutex_owner(lock))
+ break;
+ /*
+ * Ensure that owner->on_cpu is dereferenced _after_
+ * checking the above to be valid.
+ */
+ barrier();
+ if (!owner->on_cpu) {
+ res = 1;
+ break;
+ }
+ cpu_relax();
+ }
+ rcu_read_unlock();
+ return res;
+}
+#else
+static int adaptive_wait(struct rt_mutex *lock,
+ struct task_struct *orig_owner)
+{
+ return 1;
+}
+#endif
+
+static int task_blocks_on_rt_mutex(struct rt_mutex *lock,
+ struct rt_mutex_waiter *waiter,
+ struct task_struct *task,
+ enum rtmutex_chainwalk chwalk);
+/*
+ * Slow path lock function spin_lock style: this variant is very
+ * careful not to miss any non-lock wakeups.
+ *
+ * We store the current state under p->pi_lock in p->saved_state and
+ * the try_to_wake_up() code handles this accordingly.
+ */
+void __sched rt_spin_lock_slowlock_locked(struct rt_mutex *lock,
+ struct rt_mutex_waiter *waiter,
+ unsigned long flags)
+{
+ struct task_struct *lock_owner, *self = current;
+ struct rt_mutex_waiter *top_waiter;
+ int ret;
+
+ if (__try_to_take_rt_mutex(lock, self, NULL, STEAL_LATERAL))
+ return;
+
+ BUG_ON(rt_mutex_owner(lock) == self);
+
+ /*
+ * We save whatever state the task is in and we'll restore it
+ * after acquiring the lock taking real wakeups into account
+ * as well. We are serialized via pi_lock against wakeups. See
+ * try_to_wake_up().
+ */
+ raw_spin_lock(&self->pi_lock);
+ self->saved_state = self->state;
+ __set_current_state_no_track(TASK_UNINTERRUPTIBLE);
+ raw_spin_unlock(&self->pi_lock);
+
+ ret = task_blocks_on_rt_mutex(lock, waiter, self, RT_MUTEX_MIN_CHAINWALK);
+ BUG_ON(ret);
+
+ for (;;) {
+ /* Try to acquire the lock again. */
+ if (__try_to_take_rt_mutex(lock, self, waiter, STEAL_LATERAL))
+ break;
+
+ top_waiter = rt_mutex_top_waiter(lock);
+ lock_owner = rt_mutex_owner(lock);
+
+ raw_spin_unlock_irqrestore(&lock->wait_lock, flags);
+
+ debug_rt_mutex_print_deadlock(waiter);
+
+ if (top_waiter != waiter || adaptive_wait(lock, lock_owner))
+ schedule();
+
+ raw_spin_lock_irqsave(&lock->wait_lock, flags);
+
+ raw_spin_lock(&self->pi_lock);
+ __set_current_state_no_track(TASK_UNINTERRUPTIBLE);
+ raw_spin_unlock(&self->pi_lock);
+ }
+
+ /*
+ * Restore the task state to current->saved_state. We set it
+ * to the original state above and the try_to_wake_up() code
+ * has possibly updated it when a real (non-rtmutex) wakeup
+ * happened while we were blocked. Clear saved_state so
+ * try_to_wakeup() does not get confused.
+ */
+ raw_spin_lock(&self->pi_lock);
+ __set_current_state_no_track(self->saved_state);
+ self->saved_state = TASK_RUNNING;
+ raw_spin_unlock(&self->pi_lock);
+
+ /*
+ * try_to_take_rt_mutex() sets the waiter bit
+ * unconditionally. We might have to fix that up:
+ */
+ fixup_rt_mutex_waiters(lock);
+
+ BUG_ON(rt_mutex_has_waiters(lock) && waiter == rt_mutex_top_waiter(lock));
+ BUG_ON(!RB_EMPTY_NODE(&waiter->tree_entry));
+}
+
+static void noinline __sched rt_spin_lock_slowlock(struct rt_mutex *lock)
+{
+ struct rt_mutex_waiter waiter;
+ unsigned long flags;
+
+ rt_mutex_init_waiter(&waiter, true);
+
+ raw_spin_lock_irqsave(&lock->wait_lock, flags);
+ rt_spin_lock_slowlock_locked(lock, &waiter, flags);
+ raw_spin_unlock_irqrestore(&lock->wait_lock, flags);
+ debug_rt_mutex_free_waiter(&waiter);
+}
+
+static bool __sched __rt_mutex_unlock_common(struct rt_mutex *lock,
+ struct wake_q_head *wake_q,
+ struct wake_q_head *wq_sleeper);
+/*
+ * Slow path to release a rt_mutex spin_lock style
+ */
+void __sched rt_spin_lock_slowunlock(struct rt_mutex *lock)
+{
+ unsigned long flags;
+ DEFINE_WAKE_Q(wake_q);
+ DEFINE_WAKE_Q(wake_sleeper_q);
+ bool postunlock;
+
+ raw_spin_lock_irqsave(&lock->wait_lock, flags);
+ postunlock = __rt_mutex_unlock_common(lock, &wake_q, &wake_sleeper_q);
+ raw_spin_unlock_irqrestore(&lock->wait_lock, flags);
+
+ if (postunlock)
+ rt_mutex_postunlock(&wake_q, &wake_sleeper_q);
+}
+
+void __lockfunc rt_spin_lock(spinlock_t *lock)
+{
+ migrate_disable();
+ spin_acquire(&lock->dep_map, 0, 0, _RET_IP_);
+ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock);
+}
+EXPORT_SYMBOL(rt_spin_lock);
+
+void __lockfunc __rt_spin_lock(struct rt_mutex *lock)
+{
+ rt_spin_lock_fastlock(lock, rt_spin_lock_slowlock);
+}
+
+#ifdef CONFIG_DEBUG_LOCK_ALLOC
+void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass)
+{
+ migrate_disable();
+ spin_acquire(&lock->dep_map, subclass, 0, _RET_IP_);
+ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock);
+}
+EXPORT_SYMBOL(rt_spin_lock_nested);
+#endif
+
+void __lockfunc rt_spin_unlock(spinlock_t *lock)
+{
+ spin_release(&lock->dep_map, _RET_IP_);
+ rt_spin_lock_fastunlock(&lock->lock, rt_spin_lock_slowunlock);
+ migrate_enable();
+}
+EXPORT_SYMBOL(rt_spin_unlock);
+
+void __lockfunc __rt_spin_unlock(struct rt_mutex *lock)
+{
+ rt_spin_lock_fastunlock(lock, rt_spin_lock_slowunlock);
+}
+EXPORT_SYMBOL(__rt_spin_unlock);
+
+/*
+ * Wait for the lock to get unlocked: instead of polling for an unlock
+ * (like raw spinlocks do), we lock and unlock, to force the kernel to
+ * schedule if there's contention:
+ */
+void __lockfunc rt_spin_unlock_wait(spinlock_t *lock)
+{
+ spin_lock(lock);
+ spin_unlock(lock);
+}
+EXPORT_SYMBOL(rt_spin_unlock_wait);
+
+int __lockfunc rt_spin_trylock(spinlock_t *lock)
+{
+ int ret;
+
+ migrate_disable();
+ ret = __rt_mutex_trylock(&lock->lock);
+ if (ret)
+ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_);
+ else
+ migrate_enable();
+ return ret;
+}
+EXPORT_SYMBOL(rt_spin_trylock);
+
+int __lockfunc rt_spin_trylock_bh(spinlock_t *lock)
+{
+ int ret;
+
+ local_bh_disable();
+ ret = __rt_mutex_trylock(&lock->lock);
+ if (ret) {
+ migrate_disable();
+ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_);
+ } else
+ local_bh_enable();
+ return ret;
+}
+EXPORT_SYMBOL(rt_spin_trylock_bh);
+
+int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags)
+{
+ int ret;
+
+ *flags = 0;
+ ret = __rt_mutex_trylock(&lock->lock);
+ if (ret) {
+ migrate_disable();
+ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_);
+ }
+ return ret;
+}
+EXPORT_SYMBOL(rt_spin_trylock_irqsave);
+
+void
+__rt_spin_lock_init(spinlock_t *lock, const char *name, struct lock_class_key *key)
+{
+#ifdef CONFIG_DEBUG_LOCK_ALLOC
+ /*
+ * Make sure we are not reinitializing a held lock:
+ */
+ debug_check_no_locks_freed((void *)lock, sizeof(*lock));
+ lockdep_init_map(&lock->dep_map, name, key, 0);
+#endif
+}
+EXPORT_SYMBOL(__rt_spin_lock_init);
+
+#endif /* PREEMPT_RT */
+
+static inline int
+try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
+ struct rt_mutex_waiter *waiter)
+{
+ return __try_to_take_rt_mutex(lock, task, waiter, STEAL_NORMAL);
+}
+
/*
* Task blocks on lock.
*
@@ -1038,6 +1363,7 @@ static int task_blocks_on_rt_mutex(struc
* Called with lock->wait_lock held and interrupts disabled.
*/
static void mark_wakeup_next_waiter(struct wake_q_head *wake_q,
+ struct wake_q_head *wake_sleeper_q,
struct rt_mutex *lock)
{
struct rt_mutex_waiter *waiter;
@@ -1077,7 +1403,10 @@ static void mark_wakeup_next_waiter(stru
* Pairs with preempt_enable() in rt_mutex_postunlock();
*/
preempt_disable();
- wake_q_add(wake_q, waiter->task);
+ if (waiter->savestate)
+ wake_q_add_sleeper(wake_sleeper_q, waiter->task);
+ else
+ wake_q_add(wake_q, waiter->task);
raw_spin_unlock(¤t->pi_lock);
}
@@ -1161,21 +1490,22 @@ void rt_mutex_adjust_pi(struct task_stru
return;
}
next_lock = waiter->lock;
- raw_spin_unlock_irqrestore(&task->pi_lock, flags);
/* gets dropped in rt_mutex_adjust_prio_chain()! */
get_task_struct(task);
+ raw_spin_unlock_irqrestore(&task->pi_lock, flags);
rt_mutex_adjust_prio_chain(task, RT_MUTEX_MIN_CHAINWALK, NULL,
next_lock, NULL, task);
}
-void rt_mutex_init_waiter(struct rt_mutex_waiter *waiter)
+void rt_mutex_init_waiter(struct rt_mutex_waiter *waiter, bool savestate)
{
debug_rt_mutex_init_waiter(waiter);
RB_CLEAR_NODE(&waiter->pi_tree_entry);
RB_CLEAR_NODE(&waiter->tree_entry);
waiter->task = NULL;
+ waiter->savestate = savestate;
}
/**
@@ -1292,7 +1622,7 @@ rt_mutex_slowlock(struct rt_mutex *lock,
unsigned long flags;
int ret = 0;
- rt_mutex_init_waiter(&waiter);
+ rt_mutex_init_waiter(&waiter, false);
/*
* Technically we could use raw_spin_[un]lock_irq() here, but this can
@@ -1365,7 +1695,8 @@ static inline int rt_mutex_slowtrylock(s
* Return whether the current task needs to call rt_mutex_postunlock().
*/
static bool __sched rt_mutex_slowunlock(struct rt_mutex *lock,
- struct wake_q_head *wake_q)
+ struct wake_q_head *wake_q,
+ struct wake_q_head *wake_sleeper_q)
{
unsigned long flags;
@@ -1419,7 +1750,7 @@ static bool __sched rt_mutex_slowunlock(
*
* Queue the next waiter for wakeup once we release the wait_lock.
*/
- mark_wakeup_next_waiter(wake_q, lock);
+ mark_wakeup_next_waiter(wake_q, wake_sleeper_q, lock);
raw_spin_unlock_irqrestore(&lock->wait_lock, flags);
return true; /* call rt_mutex_postunlock() */
@@ -1471,9 +1802,11 @@ rt_mutex_fasttrylock(struct rt_mutex *lo
/*
* Performs the wakeup of the the top-waiter and re-enables preemption.
*/
-void rt_mutex_postunlock(struct wake_q_head *wake_q)
+void rt_mutex_postunlock(struct wake_q_head *wake_q,
+ struct wake_q_head *wake_sleeper_q)
{
wake_up_q(wake_q);
+ wake_up_q_sleeper(wake_sleeper_q);
/* Pairs with preempt_disable() in rt_mutex_slowunlock() */
preempt_enable();
@@ -1482,15 +1815,17 @@ void rt_mutex_postunlock(struct wake_q_h
static inline void
rt_mutex_fastunlock(struct rt_mutex *lock,
bool (*slowfn)(struct rt_mutex *lock,
- struct wake_q_head *wqh))
+ struct wake_q_head *wqh,
+ struct wake_q_head *wq_sleeper))
{
DEFINE_WAKE_Q(wake_q);
+ DEFINE_WAKE_Q(wake_sleeper_q);
if (likely(rt_mutex_cmpxchg_release(lock, current, NULL)))
return;
- if (slowfn(lock, &wake_q))
- rt_mutex_postunlock(&wake_q);
+ if (slowfn(lock, &wake_q, &wake_sleeper_q))
+ rt_mutex_postunlock(&wake_q, &wake_sleeper_q);
}
int __sched __rt_mutex_lock_state(struct rt_mutex *lock, int state)
@@ -1513,7 +1848,7 @@ static inline int __sched rt_mutex_lock_
mutex_acquire(&lock->dep_map, subclass, 0, _RET_IP_);
ret = __rt_mutex_lock_state(lock, state);
if (ret)
- mutex_release(&lock->dep_map, 1, _RET_IP_);
+ mutex_release(&lock->dep_map, _RET_IP_);
return ret;
}
@@ -1668,16 +2003,13 @@ void __sched __rt_mutex_unlock(struct rt
void __sched rt_mutex_unlock(struct rt_mutex *lock)
{
mutex_release(&lock->dep_map, _RET_IP_);
- rt_mutex_fastunlock(lock, rt_mutex_slowunlock);
+ __rt_mutex_unlock(lock);
}
EXPORT_SYMBOL_GPL(rt_mutex_unlock);
-/**
- * Futex variant, that since futex variants do not use the fast-path, can be
- * simple and will not need to retry.
- */
-bool __sched __rt_mutex_futex_unlock(struct rt_mutex *lock,
- struct wake_q_head *wake_q)
+static bool __sched __rt_mutex_unlock_common(struct rt_mutex *lock,
+ struct wake_q_head *wake_q,
+ struct wake_q_head *wq_sleeper)
{
lockdep_assert_held(&lock->wait_lock);
@@ -1694,23 +2026,35 @@ bool __sched __rt_mutex_futex_unlock(str
* avoid inversion prior to the wakeup. preempt_disable()
* therein pairs with rt_mutex_postunlock().
*/
- mark_wakeup_next_waiter(wake_q, lock);
+ mark_wakeup_next_waiter(wake_q, wq_sleeper, lock);
return true; /* call postunlock() */
}
+/**
+ * Futex variant, that since futex variants do not use the fast-path, can be
+ * simple and will not need to retry.
+ */
+bool __sched __rt_mutex_futex_unlock(struct rt_mutex *lock,
+ struct wake_q_head *wake_q,
+ struct wake_q_head *wq_sleeper)
+{
+ return __rt_mutex_unlock_common(lock, wake_q, wq_sleeper);
+}
+
void __sched rt_mutex_futex_unlock(struct rt_mutex *lock)
{
DEFINE_WAKE_Q(wake_q);
+ DEFINE_WAKE_Q(wake_sleeper_q);
unsigned long flags;
bool postunlock;
raw_spin_lock_irqsave(&lock->wait_lock, flags);
- postunlock = __rt_mutex_futex_unlock(lock, &wake_q);
+ postunlock = __rt_mutex_futex_unlock(lock, &wake_q, &wake_sleeper_q);
raw_spin_unlock_irqrestore(&lock->wait_lock, flags);
if (postunlock)
- rt_mutex_postunlock(&wake_q);
+ rt_mutex_postunlock(&wake_q, &wake_sleeper_q);
}
/**
@@ -1749,7 +2093,7 @@ void __rt_mutex_init(struct rt_mutex *lo
if (name && key)
debug_rt_mutex_init(lock, name, key);
}
-EXPORT_SYMBOL_GPL(__rt_mutex_init);
+EXPORT_SYMBOL(__rt_mutex_init);
/**
* rt_mutex_init_proxy_locked - initialize and lock a rt_mutex on behalf of a
@@ -1943,6 +2287,7 @@ int rt_mutex_wait_proxy_lock(struct rt_m
struct hrtimer_sleeper *to,
struct rt_mutex_waiter *waiter)
{
+ struct task_struct *tsk = current;
int ret;
raw_spin_lock_irq(&lock->wait_lock);
@@ -1954,6 +2299,24 @@ int rt_mutex_wait_proxy_lock(struct rt_m
* have to fix that up.
*/
fixup_rt_mutex_waiters(lock);
+ /*
+ * RT has a problem here when the wait got interrupted by a timeout
+ * or a signal. task->pi_blocked_on is still set. The task must
+ * acquire the hash bucket lock when returning from this function.
+ *
+ * If the hash bucket lock is contended then the
+ * BUG_ON(rt_mutex_real_waiter(task->pi_blocked_on)) in
+ * task_blocks_on_rt_mutex() will trigger. This can be avoided by
+ * clearing task->pi_blocked_on which removes the task from the
+ * boosting chain of the rtmutex. That's correct because the task
+ * is not longer blocked on it.
+ */
+ if (ret) {
+ raw_spin_lock(&tsk->pi_lock);
+ tsk->pi_blocked_on = NULL;
+ raw_spin_unlock(&tsk->pi_lock);
+ }
+
raw_spin_unlock_irq(&lock->wait_lock);
return ret;
--- a/kernel/locking/rtmutex_common.h
+++ b/kernel/locking/rtmutex_common.h
@@ -30,6 +30,7 @@ struct rt_mutex_waiter {
struct rb_node pi_tree_entry;
struct task_struct *task;
struct rt_mutex *lock;
+ bool savestate;
#ifdef CONFIG_DEBUG_RT_MUTEXES
unsigned long ip;
struct pid *deadlock_task_pid;
@@ -138,7 +139,7 @@ extern struct task_struct *rt_mutex_next
extern void rt_mutex_init_proxy_locked(struct rt_mutex *lock,
struct task_struct *proxy_owner);
extern void rt_mutex_proxy_unlock(struct rt_mutex *lock);
-extern void rt_mutex_init_waiter(struct rt_mutex_waiter *waiter);
+extern void rt_mutex_init_waiter(struct rt_mutex_waiter *waiter, bool savetate);
extern int __rt_mutex_start_proxy_lock(struct rt_mutex *lock,
struct rt_mutex_waiter *waiter,
struct task_struct *task);
@@ -156,9 +157,12 @@ extern int __rt_mutex_futex_trylock(stru
extern void rt_mutex_futex_unlock(struct rt_mutex *lock);
extern bool __rt_mutex_futex_unlock(struct rt_mutex *lock,
- struct wake_q_head *wqh);
+ struct wake_q_head *wqh,
+ struct wake_q_head *wq_sleeper);
+
+extern void rt_mutex_postunlock(struct wake_q_head *wake_q,
+ struct wake_q_head *wake_sleeper_q);
-extern void rt_mutex_postunlock(struct wake_q_head *wake_q);
/* RW semaphore special interface */
extern int __rt_mutex_lock_state(struct rt_mutex *lock, int state);
@@ -168,6 +172,10 @@ int __sched rt_mutex_slowlock_locked(str
struct hrtimer_sleeper *timeout,
enum rtmutex_chainwalk chwalk,
struct rt_mutex_waiter *waiter);
+void __sched rt_spin_lock_slowlock_locked(struct rt_mutex *lock,
+ struct rt_mutex_waiter *waiter,
+ unsigned long flags);
+void __sched rt_spin_lock_slowunlock(struct rt_mutex *lock);
#ifdef CONFIG_DEBUG_RT_MUTEXES
# include "rtmutex-debug.h"
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -520,9 +520,15 @@ static bool set_nr_if_polling(struct tas
#endif
#endif
-static bool __wake_q_add(struct wake_q_head *head, struct task_struct *task)
+static bool __wake_q_add(struct wake_q_head *head, struct task_struct *task,
+ bool sleeper)
{
- struct wake_q_node *node = &task->wake_q;
+ struct wake_q_node *node;
+
+ if (sleeper)
+ node = &task->wake_q_sleeper;
+ else
+ node = &task->wake_q;
/*
* Atomically grab the task, if ->wake_q is !nil already it means
@@ -558,7 +564,13 @@ static bool __wake_q_add(struct wake_q_h
*/
void wake_q_add(struct wake_q_head *head, struct task_struct *task)
{
- if (__wake_q_add(head, task))
+ if (__wake_q_add(head, task, false))
+ get_task_struct(task);
+}
+
+void wake_q_add_sleeper(struct wake_q_head *head, struct task_struct *task)
+{
+ if (__wake_q_add(head, task, true))
get_task_struct(task);
}
@@ -581,28 +593,39 @@ void wake_q_add(struct wake_q_head *head
*/
void wake_q_add_safe(struct wake_q_head *head, struct task_struct *task)
{
- if (!__wake_q_add(head, task))
+ if (!__wake_q_add(head, task, false))
put_task_struct(task);
}
-void wake_up_q(struct wake_q_head *head)
+void __wake_up_q(struct wake_q_head *head, bool sleeper)
{
struct wake_q_node *node = head->first;
while (node != WAKE_Q_TAIL) {
struct task_struct *task;
- task = container_of(node, struct task_struct, wake_q);
+ if (sleeper)
+ task = container_of(node, struct task_struct, wake_q_sleeper);
+ else
+ task = container_of(node, struct task_struct, wake_q);
+
BUG_ON(!task);
/* Task can safely be re-inserted now: */
node = node->next;
- task->wake_q.next = NULL;
+ if (sleeper)
+ task->wake_q_sleeper.next = NULL;
+ else
+ task->wake_q.next = NULL;
/*
* wake_up_process() executes a full barrier, which pairs with
* the queueing in wake_q_add() so as not to miss wakeups.
*/
- wake_up_process(task);
+ if (sleeper)
+ wake_up_lock_sleeper(task);
+ else
+ wake_up_process(task);
+
put_task_struct(task);
}
}