Kernel: Optimize futexes

Add support for processor local futexes. These work the exact same way
as global ones, but only lock a process specific lock and use a process
specific hash map.

Also reduce the time futex lock is held. There was no need to hold the
global lock while validating addresses in the process' address space.

userspace/libraries/LibC/pthread.cpp

@@ -884,9 +884,11 @@ int pthread_mutex_unlock(pthread_mutex_t* mutex)
 	mutex->lock_depth--;
 	if (mutex->lock_depth == 0)
 	{
+		const int op = FUTEX_WAKE | (mutex->attr.shared ? 0 : FUTEX_PRIVATE);
+
 		BAN::atomic_store(mutex->futex, 0, BAN::memory_order_release);
 		if (BAN::atomic_load(mutex->waiters))
-			futex(FUTEX_WAKE, &mutex->futex, 1, nullptr);
+			futex(op, &mutex->futex, 1, nullptr);
 	}
 
 	return 0;

userspace/libraries/LibC/semaphore.cpp

@@ -28,8 +28,9 @@ int sem_getvalue(sem_t* __restrict sem, int* __restrict sval)
 int sem_post(sem_t* sem)
 {
 	const auto old = BAN::atomic_fetch_add(sem->value, 1);
+	const int op = FUTEX_WAKE | (sem->shared ? 0 : FUTEX_PRIVATE);
 	if (old == 0)
-		futex(FUTEX_WAKE, &sem->value, 1, nullptr);
+		futex(op, &sem->value, 1, nullptr);
 	return 0;
 }