38  *  - Bit 0: RWSEM_READER_OWNED - rwsem may be owned by readers (just a hint)
55  * is involved. Ideally we would like to track all the readers that own
109  * 1) rwsem_mark_wake() for readers		-- set, clear
294  * The lock is owned by readers when
299  * Having some reader bits set is not enough to guarantee a readers owned
300  * lock as the readers may be in the process of backing out from the count
348 	RWSEM_WAKE_READERS,	/* Wake readers only */
360  * Magic number to batch-wakeup waiting readers, even when writers are
407  * Implies rwsem_del_waiter() for all woken readers.
431 			 * Readers, on the other hand, will block as they  in rwsem_mark_wake()
449 	 * We prefer to do the first reader grant before counting readers  in rwsem_mark_wake()
476 		 * indication that readers now have the lock.  in rwsem_mark_wake()
485 	 * Grant up to MAX_READERS_WAKEUP read locks to all the readers in the  in rwsem_mark_wake()
488 	 * number of readers before waking any processes up.  in rwsem_mark_wake()
491 	 * reader phase (first waiter is a reader), all readers are eligible  in rwsem_mark_wake()
516 		 * Limit # of readers that can be woken up per wakeup call.  in rwsem_mark_wake()
795  * The more readers own the rwsem, the longer it will take for them to
801  * The limit is capped to a maximum of 25us (30 readers). This is just
807 	int readers = count >> RWSEM_READER_SHIFT;  in rwsem_rspin_threshold()  local
810 	if (readers > 30)  in rwsem_rspin_threshold()
811 		readers = 30;  in rwsem_rspin_threshold()
812 	delta = (20 + readers) * NSEC_PER_USEC / 2;  in rwsem_rspin_threshold()
832 	 *  2) readers own the lock and spinning time has exceeded limit.  in rwsem_optimistic_spin()
1002 	 * To prevent a constant stream of readers from starving a sleeping  in rwsem_down_read_slowpath()
1004 	 * very likely owned by readers.  in rwsem_down_read_slowpath()
1018 		 * Wake up other readers in the wait queue if it is  in rwsem_down_read_slowpath()
1135 			 * when a large number of readers are to be woken up.  in rwsem_down_write_slowpath()
1220  * - just wake up any readers at the front of the queue
1511 	int count = atomic_read(&sem->rwbase.readers);  in is_rwsem_reader_owned()