1================================== 2RT-mutex subsystem with PI support 3================================== 4 5RT-mutexes with priority inheritance are used to support PI-futexes, 6which enable pthread_mutex_t priority inheritance attributes 7(PTHREAD_PRIO_INHERIT). [See Documentation/pi-futex.txt for more details 8about PI-futexes.] 9 10This technology was developed in the -rt tree and streamlined for 11pthread_mutex support. 12 13Basic principles: 14----------------- 15 16RT-mutexes extend the semantics of simple mutexes by the priority 17inheritance protocol. 18 19A low priority owner of a rt-mutex inherits the priority of a higher 20priority waiter until the rt-mutex is released. If the temporarily 21boosted owner blocks on a rt-mutex itself it propagates the priority 22boosting to the owner of the other rt_mutex it gets blocked on. The 23priority boosting is immediately removed once the rt_mutex has been 24unlocked. 25 26This approach allows us to shorten the block of high-prio tasks on 27mutexes which protect shared resources. Priority inheritance is not a 28magic bullet for poorly designed applications, but it allows 29well-designed applications to use userspace locks in critical parts of 30an high priority thread, without losing determinism. 31 32The enqueueing of the waiters into the rtmutex waiter tree is done in 33priority order. For same priorities FIFO order is chosen. For each 34rtmutex, only the top priority waiter is enqueued into the owner's 35priority waiters tree. This tree too queues in priority order. Whenever 36the top priority waiter of a task changes (for example it timed out or 37got a signal), the priority of the owner task is readjusted. The 38priority enqueueing is handled by "pi_waiters". 39 40RT-mutexes are optimized for fastpath operations and have no internal 41locking overhead when locking an uncontended mutex or unlocking a mutex 42without waiters. The optimized fastpath operations require cmpxchg 43support. [If that is not available then the rt-mutex internal spinlock 44is used] 45 46The state of the rt-mutex is tracked via the owner field of the rt-mutex 47structure: 48 49lock->owner holds the task_struct pointer of the owner. Bit 0 is used to 50keep track of the "lock has waiters" state: 51 52 ============ ======= ================================================ 53 owner bit0 Notes 54 ============ ======= ================================================ 55 NULL 0 lock is free (fast acquire possible) 56 NULL 1 lock is free and has waiters and the top waiter 57 is going to take the lock [1]_ 58 taskpointer 0 lock is held (fast release possible) 59 taskpointer 1 lock is held and has waiters [2]_ 60 ============ ======= ================================================ 61 62The fast atomic compare exchange based acquire and release is only 63possible when bit 0 of lock->owner is 0. 64 65.. [1] It also can be a transitional state when grabbing the lock 66 with ->wait_lock is held. To prevent any fast path cmpxchg to the lock, 67 we need to set the bit0 before looking at the lock, and the owner may 68 be NULL in this small time, hence this can be a transitional state. 69 70.. [2] There is a small time when bit 0 is set but there are no 71 waiters. This can happen when grabbing the lock in the slow path. 72 To prevent a cmpxchg of the owner releasing the lock, we need to 73 set this bit before looking at the lock. 74 75BTW, there is still technically a "Pending Owner", it's just not called 76that anymore. The pending owner happens to be the top_waiter of a lock 77that has no owner and has been woken up to grab the lock. 78