From 64ac24e738823161693bf791f87adc802cf529ff Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 7 Mar 2008 21:55:58 -0500 Subject: Generic semaphore implementation Semaphores are no longer performance-critical, so a generic C implementation is better for maintainability, debuggability and extensibility. Thanks to Peter Zijlstra for fixing the lockdep warning. Thanks to Harvey Harrison for pointing out that the unlikely() was unnecessary. Signed-off-by: Matthew Wilcox Acked-by: Ingo Molnar --- kernel/semaphore.c | 187 +++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 187 insertions(+) create mode 100644 kernel/semaphore.c (limited to 'kernel/semaphore.c') diff --git a/kernel/semaphore.c b/kernel/semaphore.c new file mode 100644 index 00000000000..d5a72702f26 --- /dev/null +++ b/kernel/semaphore.c @@ -0,0 +1,187 @@ +/* + * Copyright (c) 2008 Intel Corporation + * Author: Matthew Wilcox + * + * Distributed under the terms of the GNU GPL, version 2 + */ + +#include +#include +#include +#include +#include +#include + +/* + * Some notes on the implementation: + * + * down_trylock() and up() can be called from interrupt context. + * So we have to disable interrupts when taking the lock. + * + * The ->count variable, if positive, defines how many more tasks can + * acquire the semaphore. If negative, it represents how many tasks are + * waiting on the semaphore (*). If zero, no tasks are waiting, and no more + * tasks can acquire the semaphore. + * + * (*) Except for the window between one task calling up() and the task + * sleeping in a __down_common() waking up. In order to avoid a third task + * coming in and stealing the second task's wakeup, we leave the ->count + * negative. If we have a more complex situation, the ->count may become + * zero or negative (eg a semaphore with count = 2, three tasks attempt to + * acquire it, one sleeps, two finish and call up(), the second task to call + * up() notices that the list is empty and just increments count). + */ + +static noinline void __down(struct semaphore *sem); +static noinline int __down_interruptible(struct semaphore *sem); +static noinline void __up(struct semaphore *sem); + +void down(struct semaphore *sem) +{ + unsigned long flags; + + spin_lock_irqsave(&sem->lock, flags); + if (unlikely(sem->count-- <= 0)) + __down(sem); + spin_unlock_irqrestore(&sem->lock, flags); +} +EXPORT_SYMBOL(down); + +int down_interruptible(struct semaphore *sem) +{ + unsigned long flags; + int result = 0; + + spin_lock_irqsave(&sem->lock, flags); + if (unlikely(sem->count-- <= 0)) + result = __down_interruptible(sem); + spin_unlock_irqrestore(&sem->lock, flags); + + return result; +} +EXPORT_SYMBOL(down_interruptible); + +/** + * down_trylock - try to acquire the semaphore, without waiting + * @sem: the semaphore to be acquired + * + * Try to acquire the semaphore atomically. Returns 0 if the mutex has + * been acquired successfully and 1 if it is contended. + * + * NOTE: This return value is inverted from both spin_trylock and + * mutex_trylock! Be careful about this when converting code. + * + * Unlike mutex_trylock, this function can be used from interrupt context, + * and the semaphore can be released by any task or interrupt. + */ +int down_trylock(struct semaphore *sem) +{ + unsigned long flags; + int count; + + spin_lock_irqsave(&sem->lock, flags); + count = sem->count - 1; + if (likely(count >= 0)) + sem->count = count; + spin_unlock_irqrestore(&sem->lock, flags); + + return (count < 0); +} +EXPORT_SYMBOL(down_trylock); + +void up(struct semaphore *sem) +{ + unsigned long flags; + + spin_lock_irqsave(&sem->lock, flags); + if (likely(sem->count >= 0)) + sem->count++; + else + __up(sem); + spin_unlock_irqrestore(&sem->lock, flags); +} +EXPORT_SYMBOL(up); + +/* Functions for the contended case */ + +struct semaphore_waiter { + struct list_head list; + struct task_struct *task; + int up; +}; + +/* + * Wake up a process waiting on a semaphore. We need to call this from both + * __up and __down_common as it's possible to race a task into the semaphore + * if it comes in at just the right time between two tasks calling up() and + * a third task waking up. This function assumes the wait_list is already + * checked for being non-empty. + */ +static noinline void __sched __up_down_common(struct semaphore *sem) +{ + struct semaphore_waiter *waiter = list_first_entry(&sem->wait_list, + struct semaphore_waiter, list); + list_del(&waiter->list); + waiter->up = 1; + wake_up_process(waiter->task); +} + +/* + * Because this function is inlined, the 'state' parameter will be constant, + * and thus optimised away by the compiler. + */ +static inline int __sched __down_common(struct semaphore *sem, long state) +{ + int result = 0; + struct task_struct *task = current; + struct semaphore_waiter waiter; + + list_add_tail(&waiter.list, &sem->wait_list); + waiter.task = task; + waiter.up = 0; + + for (;;) { + if (state == TASK_INTERRUPTIBLE && signal_pending(task)) + goto interrupted; + __set_task_state(task, state); + spin_unlock_irq(&sem->lock); + schedule(); + spin_lock_irq(&sem->lock); + if (waiter.up) + goto woken; + } + + interrupted: + list_del(&waiter.list); + result = -EINTR; + woken: + /* + * Account for the process which woke us up. For the case where + * we're interrupted, we need to increment the count on our own + * behalf. I don't believe we can hit the case where the + * sem->count hits zero, *and* there's a second task sleeping, + * but it doesn't hurt, that's not a commonly exercised path and + * it's not a performance path either. + */ + if (unlikely((++sem->count >= 0) && !list_empty(&sem->wait_list))) + __up_down_common(sem); + return result; +} + +static noinline void __sched __down(struct semaphore *sem) +{ + __down_common(sem, TASK_UNINTERRUPTIBLE); +} + +static noinline int __sched __down_interruptible(struct semaphore *sem) +{ + return __down_common(sem, TASK_INTERRUPTIBLE); +} + +static noinline void __sched __up(struct semaphore *sem) +{ + if (unlikely(list_empty(&sem->wait_list))) + sem->count++; + else + __up_down_common(sem); +} -- cgit v1.2.3-18-g5258 From f06d96865861c3dd01520f47e2e61c899db1631f Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 14 Mar 2008 13:19:33 -0400 Subject: Introduce down_killable() down_killable() is the functional counterpart of mutex_lock_killable. Signed-off-by: Matthew Wilcox --- kernel/semaphore.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) (limited to 'kernel/semaphore.c') diff --git a/kernel/semaphore.c b/kernel/semaphore.c index d5a72702f26..2da2aed950f 100644 --- a/kernel/semaphore.c +++ b/kernel/semaphore.c @@ -34,6 +34,7 @@ static noinline void __down(struct semaphore *sem); static noinline int __down_interruptible(struct semaphore *sem); +static noinline int __down_killable(struct semaphore *sem); static noinline void __up(struct semaphore *sem); void down(struct semaphore *sem) @@ -61,6 +62,20 @@ int down_interruptible(struct semaphore *sem) } EXPORT_SYMBOL(down_interruptible); +int down_killable(struct semaphore *sem) +{ + unsigned long flags; + int result = 0; + + spin_lock_irqsave(&sem->lock, flags); + if (unlikely(sem->count-- <= 0)) + result = __down_killable(sem); + spin_unlock_irqrestore(&sem->lock, flags); + + return result; +} +EXPORT_SYMBOL(down_killable); + /** * down_trylock - try to acquire the semaphore, without waiting * @sem: the semaphore to be acquired @@ -143,6 +158,8 @@ static inline int __sched __down_common(struct semaphore *sem, long state) for (;;) { if (state == TASK_INTERRUPTIBLE && signal_pending(task)) goto interrupted; + if (state == TASK_KILLABLE && fatal_signal_pending(task)) + goto interrupted; __set_task_state(task, state); spin_unlock_irq(&sem->lock); schedule(); @@ -178,6 +195,11 @@ static noinline int __sched __down_interruptible(struct semaphore *sem) return __down_common(sem, TASK_INTERRUPTIBLE); } +static noinline int __sched __down_killable(struct semaphore *sem) +{ + return __down_common(sem, TASK_KILLABLE); +} + static noinline void __sched __up(struct semaphore *sem) { if (unlikely(list_empty(&sem->wait_list))) -- cgit v1.2.3-18-g5258 From f1241c87a16c4fe9f4f51d6ed3589f031c505e8d Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 14 Mar 2008 13:43:13 -0400 Subject: Add down_timeout and change ACPI to use it ACPI currently emulates a timeout for semaphores with calls to down_trylock and sleep. This produces horrible behaviour in terms of fairness and excessive wakeups. Now that we have a unified semaphore implementation, adding a real down_trylock is almost trivial. Signed-off-by: Matthew Wilcox --- kernel/semaphore.c | 42 +++++++++++++++++++++++++++++++++++------- 1 file changed, 35 insertions(+), 7 deletions(-) (limited to 'kernel/semaphore.c') diff --git a/kernel/semaphore.c b/kernel/semaphore.c index 2da2aed950f..5a12a855898 100644 --- a/kernel/semaphore.c +++ b/kernel/semaphore.c @@ -35,6 +35,7 @@ static noinline void __down(struct semaphore *sem); static noinline int __down_interruptible(struct semaphore *sem); static noinline int __down_killable(struct semaphore *sem); +static noinline int __down_timeout(struct semaphore *sem, long jiffies); static noinline void __up(struct semaphore *sem); void down(struct semaphore *sem) @@ -104,6 +105,20 @@ int down_trylock(struct semaphore *sem) } EXPORT_SYMBOL(down_trylock); +int down_timeout(struct semaphore *sem, long jiffies) +{ + unsigned long flags; + int result = 0; + + spin_lock_irqsave(&sem->lock, flags); + if (unlikely(sem->count-- <= 0)) + result = __down_timeout(sem, jiffies); + spin_unlock_irqrestore(&sem->lock, flags); + + return result; +} +EXPORT_SYMBOL(down_timeout); + void up(struct semaphore *sem) { unsigned long flags; @@ -142,10 +157,12 @@ static noinline void __sched __up_down_common(struct semaphore *sem) } /* - * Because this function is inlined, the 'state' parameter will be constant, - * and thus optimised away by the compiler. + * Because this function is inlined, the 'state' parameter will be + * constant, and thus optimised away by the compiler. Likewise the + * 'timeout' parameter for the cases without timeouts. */ -static inline int __sched __down_common(struct semaphore *sem, long state) +static inline int __sched __down_common(struct semaphore *sem, long state, + long timeout) { int result = 0; struct task_struct *task = current; @@ -160,14 +177,20 @@ static inline int __sched __down_common(struct semaphore *sem, long state) goto interrupted; if (state == TASK_KILLABLE && fatal_signal_pending(task)) goto interrupted; + if (timeout <= 0) + goto timed_out; __set_task_state(task, state); spin_unlock_irq(&sem->lock); - schedule(); + timeout = schedule_timeout(timeout); spin_lock_irq(&sem->lock); if (waiter.up) goto woken; } + timed_out: + list_del(&waiter.list); + result = -ETIME; + goto woken; interrupted: list_del(&waiter.list); result = -EINTR; @@ -187,17 +210,22 @@ static inline int __sched __down_common(struct semaphore *sem, long state) static noinline void __sched __down(struct semaphore *sem) { - __down_common(sem, TASK_UNINTERRUPTIBLE); + __down_common(sem, TASK_UNINTERRUPTIBLE, MAX_SCHEDULE_TIMEOUT); } static noinline int __sched __down_interruptible(struct semaphore *sem) { - return __down_common(sem, TASK_INTERRUPTIBLE); + return __down_common(sem, TASK_INTERRUPTIBLE, MAX_SCHEDULE_TIMEOUT); } static noinline int __sched __down_killable(struct semaphore *sem) { - return __down_common(sem, TASK_KILLABLE); + return __down_common(sem, TASK_KILLABLE, MAX_SCHEDULE_TIMEOUT); +} + +static noinline int __sched __down_timeout(struct semaphore *sem, long jiffies) +{ + return __down_common(sem, TASK_UNINTERRUPTIBLE, jiffies); } static noinline void __sched __up(struct semaphore *sem) -- cgit v1.2.3-18-g5258 From b17170b2fac96705db3188f093f89e8e838418e4 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 14 Mar 2008 14:35:22 -0400 Subject: Simplify semaphore implementation By removing the negative values of 'count' and relying on the wait_list to indicate whether we have any waiters, we can simplify the implementation by removing the protection against an unlikely race condition. Thanks to David Howells for his suggestions. Signed-off-by: Matthew Wilcox --- kernel/semaphore.c | 78 +++++++++++++++++------------------------------------- 1 file changed, 24 insertions(+), 54 deletions(-) (limited to 'kernel/semaphore.c') diff --git a/kernel/semaphore.c b/kernel/semaphore.c index 5a12a855898..bef977b1696 100644 --- a/kernel/semaphore.c +++ b/kernel/semaphore.c @@ -18,18 +18,8 @@ * down_trylock() and up() can be called from interrupt context. * So we have to disable interrupts when taking the lock. * - * The ->count variable, if positive, defines how many more tasks can - * acquire the semaphore. If negative, it represents how many tasks are - * waiting on the semaphore (*). If zero, no tasks are waiting, and no more - * tasks can acquire the semaphore. - * - * (*) Except for the window between one task calling up() and the task - * sleeping in a __down_common() waking up. In order to avoid a third task - * coming in and stealing the second task's wakeup, we leave the ->count - * negative. If we have a more complex situation, the ->count may become - * zero or negative (eg a semaphore with count = 2, three tasks attempt to - * acquire it, one sleeps, two finish and call up(), the second task to call - * up() notices that the list is empty and just increments count). + * The ->count variable defines how many more tasks can acquire the + * semaphore. If it's zero, there may be tasks waiting on the list. */ static noinline void __down(struct semaphore *sem); @@ -43,7 +33,9 @@ void down(struct semaphore *sem) unsigned long flags; spin_lock_irqsave(&sem->lock, flags); - if (unlikely(sem->count-- <= 0)) + if (likely(sem->count > 0)) + sem->count--; + else __down(sem); spin_unlock_irqrestore(&sem->lock, flags); } @@ -55,7 +47,9 @@ int down_interruptible(struct semaphore *sem) int result = 0; spin_lock_irqsave(&sem->lock, flags); - if (unlikely(sem->count-- <= 0)) + if (likely(sem->count > 0)) + sem->count--; + else result = __down_interruptible(sem); spin_unlock_irqrestore(&sem->lock, flags); @@ -69,7 +63,9 @@ int down_killable(struct semaphore *sem) int result = 0; spin_lock_irqsave(&sem->lock, flags); - if (unlikely(sem->count-- <= 0)) + if (likely(sem->count > 0)) + sem->count--; + else result = __down_killable(sem); spin_unlock_irqrestore(&sem->lock, flags); @@ -111,7 +107,9 @@ int down_timeout(struct semaphore *sem, long jiffies) int result = 0; spin_lock_irqsave(&sem->lock, flags); - if (unlikely(sem->count-- <= 0)) + if (likely(sem->count > 0)) + sem->count--; + else result = __down_timeout(sem, jiffies); spin_unlock_irqrestore(&sem->lock, flags); @@ -124,7 +122,7 @@ void up(struct semaphore *sem) unsigned long flags; spin_lock_irqsave(&sem->lock, flags); - if (likely(sem->count >= 0)) + if (likely(list_empty(&sem->wait_list))) sem->count++; else __up(sem); @@ -140,22 +138,6 @@ struct semaphore_waiter { int up; }; -/* - * Wake up a process waiting on a semaphore. We need to call this from both - * __up and __down_common as it's possible to race a task into the semaphore - * if it comes in at just the right time between two tasks calling up() and - * a third task waking up. This function assumes the wait_list is already - * checked for being non-empty. - */ -static noinline void __sched __up_down_common(struct semaphore *sem) -{ - struct semaphore_waiter *waiter = list_first_entry(&sem->wait_list, - struct semaphore_waiter, list); - list_del(&waiter->list); - waiter->up = 1; - wake_up_process(waiter->task); -} - /* * Because this function is inlined, the 'state' parameter will be * constant, and thus optimised away by the compiler. Likewise the @@ -164,7 +146,6 @@ static noinline void __sched __up_down_common(struct semaphore *sem) static inline int __sched __down_common(struct semaphore *sem, long state, long timeout) { - int result = 0; struct task_struct *task = current; struct semaphore_waiter waiter; @@ -184,28 +165,16 @@ static inline int __sched __down_common(struct semaphore *sem, long state, timeout = schedule_timeout(timeout); spin_lock_irq(&sem->lock); if (waiter.up) - goto woken; + return 0; } timed_out: list_del(&waiter.list); - result = -ETIME; - goto woken; + return -ETIME; + interrupted: list_del(&waiter.list); - result = -EINTR; - woken: - /* - * Account for the process which woke us up. For the case where - * we're interrupted, we need to increment the count on our own - * behalf. I don't believe we can hit the case where the - * sem->count hits zero, *and* there's a second task sleeping, - * but it doesn't hurt, that's not a commonly exercised path and - * it's not a performance path either. - */ - if (unlikely((++sem->count >= 0) && !list_empty(&sem->wait_list))) - __up_down_common(sem); - return result; + return -EINTR; } static noinline void __sched __down(struct semaphore *sem) @@ -230,8 +199,9 @@ static noinline int __sched __down_timeout(struct semaphore *sem, long jiffies) static noinline void __sched __up(struct semaphore *sem) { - if (unlikely(list_empty(&sem->wait_list))) - sem->count++; - else - __up_down_common(sem); + struct semaphore_waiter *waiter = list_first_entry(&sem->wait_list, + struct semaphore_waiter, list); + list_del(&waiter->list); + waiter->up = 1; + wake_up_process(waiter->task); } -- cgit v1.2.3-18-g5258 From 714493cd5468f42ca3c4f730a9c17c203abd5059 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 11 Apr 2008 15:23:52 -0400 Subject: Improve semaphore documentation Move documentation from semaphore.h to semaphore.c as requested by Andrew Morton. Also reformat to kernel-doc style and add some more notes about the implementation. Signed-off-by: Matthew Wilcox --- kernel/semaphore.c | 79 ++++++++++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 68 insertions(+), 11 deletions(-) (limited to 'kernel/semaphore.c') diff --git a/kernel/semaphore.c b/kernel/semaphore.c index bef977b1696..5c2942e768c 100644 --- a/kernel/semaphore.c +++ b/kernel/semaphore.c @@ -3,6 +3,26 @@ * Author: Matthew Wilcox * * Distributed under the terms of the GNU GPL, version 2 + * + * This file implements counting semaphores. + * A counting semaphore may be acquired 'n' times before sleeping. + * See mutex.c for single-acquisition sleeping locks which enforce + * rules which allow code to be debugged more easily. + */ + +/* + * Some notes on the implementation: + * + * The spinlock controls access to the other members of the semaphore. + * down_trylock() and up() can be called from interrupt context, so we + * have to disable interrupts when taking the lock. It turns out various + * parts of the kernel expect to be able to use down() on a semaphore in + * interrupt context when they know it will succeed, so we have to use + * irqsave variants for down(), down_interruptible() and down_killable() + * too. + * + * The ->count variable represents how many more tasks can acquire this + * semaphore. If it's zero, there may be tasks waiting on the wait_list. */ #include @@ -12,22 +32,23 @@ #include #include -/* - * Some notes on the implementation: - * - * down_trylock() and up() can be called from interrupt context. - * So we have to disable interrupts when taking the lock. - * - * The ->count variable defines how many more tasks can acquire the - * semaphore. If it's zero, there may be tasks waiting on the list. - */ - static noinline void __down(struct semaphore *sem); static noinline int __down_interruptible(struct semaphore *sem); static noinline int __down_killable(struct semaphore *sem); static noinline int __down_timeout(struct semaphore *sem, long jiffies); static noinline void __up(struct semaphore *sem); +/** + * down - acquire the semaphore + * @sem: the semaphore to be acquired + * + * Acquires the semaphore. If no more tasks are allowed to acquire the + * semaphore, calling this function will put the task to sleep until the + * semaphore is released. + * + * Use of this function is deprecated, please use down_interruptible() or + * down_killable() instead. + */ void down(struct semaphore *sem) { unsigned long flags; @@ -41,6 +62,15 @@ void down(struct semaphore *sem) } EXPORT_SYMBOL(down); +/** + * down_interruptible - acquire the semaphore unless interrupted + * @sem: the semaphore to be acquired + * + * Attempts to acquire the semaphore. If no more tasks are allowed to + * acquire the semaphore, calling this function will put the task to sleep. + * If the sleep is interrupted by a signal, this function will return -EINTR. + * If the semaphore is successfully acquired, this function returns 0. + */ int down_interruptible(struct semaphore *sem) { unsigned long flags; @@ -57,6 +87,16 @@ int down_interruptible(struct semaphore *sem) } EXPORT_SYMBOL(down_interruptible); +/** + * down_killable - acquire the semaphore unless killed + * @sem: the semaphore to be acquired + * + * Attempts to acquire the semaphore. If no more tasks are allowed to + * acquire the semaphore, calling this function will put the task to sleep. + * If the sleep is interrupted by a fatal signal, this function will return + * -EINTR. If the semaphore is successfully acquired, this function returns + * 0. + */ int down_killable(struct semaphore *sem) { unsigned long flags; @@ -78,7 +118,7 @@ EXPORT_SYMBOL(down_killable); * @sem: the semaphore to be acquired * * Try to acquire the semaphore atomically. Returns 0 if the mutex has - * been acquired successfully and 1 if it is contended. + * been acquired successfully or 1 if it it cannot be acquired. * * NOTE: This return value is inverted from both spin_trylock and * mutex_trylock! Be careful about this when converting code. @@ -101,6 +141,16 @@ int down_trylock(struct semaphore *sem) } EXPORT_SYMBOL(down_trylock); +/** + * down_timeout - acquire the semaphore within a specified time + * @sem: the semaphore to be acquired + * @jiffies: how long to wait before failing + * + * Attempts to acquire the semaphore. If no more tasks are allowed to + * acquire the semaphore, calling this function will put the task to sleep. + * If the semaphore is not released within the specified number of jiffies, + * this function returns -ETIME. It returns 0 if the semaphore was acquired. + */ int down_timeout(struct semaphore *sem, long jiffies) { unsigned long flags; @@ -117,6 +167,13 @@ int down_timeout(struct semaphore *sem, long jiffies) } EXPORT_SYMBOL(down_timeout); +/** + * up - release the semaphore + * @sem: the semaphore to release + * + * Release the semaphore. Unlike mutexes, up() may be called from any + * context and even by tasks which have never called down(). + */ void up(struct semaphore *sem) { unsigned long flags; -- cgit v1.2.3-18-g5258