Kernel: Implemented mutex priority inheritance.

Verified with a hwtest and implemented based on reverse engineering.

Thread A's priority will get bumped to the highest priority among all the threads that are waiting for a mutex that A holds.
Once A releases the mutex and ownership is transferred to B, A's priority will return to normal and B's priority will be bumped.
This commit is contained in:
Subv 2018-04-20 20:15:16 -05:00
parent a70ed9c8ae
commit 46572d027d
4 changed files with 94 additions and 10 deletions

View file

@ -109,6 +109,15 @@ public:
*/
void BoostPriority(u32 priority);
/// Adds a thread to the list of threads that are waiting for a lock held by this thread.
void AddMutexWaiter(SharedPtr<Thread> thread);
/// Removes a thread from the list of threads that are waiting for a lock held by this thread.
void RemoveMutexWaiter(SharedPtr<Thread> thread);
/// Recalculates the current priority taking into account priority inheritance.
void UpdatePriority();
/**
* Gets the thread's thread ID
* @return The thread's ID
@ -205,6 +214,12 @@ public:
// passed to WaitSynchronization1/N.
std::vector<SharedPtr<WaitObject>> wait_objects;
/// List of threads that are waiting for a mutex that is held by this thread.
std::vector<SharedPtr<Thread>> wait_mutex_threads;
/// Thread that owns the lock that this thread is waiting for.
SharedPtr<Thread> lock_owner;
// If waiting on a ConditionVariable, this is the ConditionVariable address
VAddr condvar_wait_address;
VAddr mutex_wait_address; ///< If waiting on a Mutex, this is the mutex address