15 files changed, 508 insertions, 1182 deletions

diff --git a/src/core/CMakeLists.txt b/src/core/CMakeLists.txt
index d29d4573e..1b8ad476e 100644
--- a/src/core/CMakeLists.txt
+++ b/src/core/CMakeLists.txt
@@ -142,8 +142,6 @@ add_library(core STATIC
     hardware_interrupt_manager.h
     hle/ipc.h
     hle/ipc_helpers.h
-    hle/kernel/address_arbiter.cpp
-    hle/kernel/address_arbiter.h
     hle/kernel/client_port.cpp
     hle/kernel/client_port.h
     hle/kernel/client_session.cpp
@@ -189,8 +187,6 @@ add_library(core STATIC
     hle/kernel/memory/slab_heap.h
     hle/kernel/memory/system_control.cpp
     hle/kernel/memory/system_control.h
-    hle/kernel/mutex.cpp
-    hle/kernel/mutex.h
     hle/kernel/object.cpp
     hle/kernel/object.h
     hle/kernel/physical_core.cpp
diff --git a/src/core/hle/kernel/address_arbiter.cpp b/src/core/hle/kernel/address_arbiter.cpp
deleted file mode 100644
index fe8675186..000000000
--- a/src/core/hle/kernel/address_arbiter.cpp
+++ /dev/null
@@ -1,317 +0,0 @@
-// Copyright 2018 yuzu emulator team
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#include <algorithm>
-#include <vector>
-
-#include "common/assert.h"
-#include "common/common_types.h"
-#include "core/arm/exclusive_monitor.h"
-#include "core/core.h"
-#include "core/hle/kernel/address_arbiter.h"
-#include "core/hle/kernel/errors.h"
-#include "core/hle/kernel/handle_table.h"
-#include "core/hle/kernel/k_scheduler.h"
-#include "core/hle/kernel/k_scoped_scheduler_lock_and_sleep.h"
-#include "core/hle/kernel/kernel.h"
-#include "core/hle/kernel/thread.h"
-#include "core/hle/kernel/time_manager.h"
-#include "core/hle/result.h"
-#include "core/memory.h"
-
-namespace Kernel {
-
-// Wake up num_to_wake (or all) threads in a vector.
-void AddressArbiter::WakeThreads(const std::vector<std::shared_ptr<Thread>>& waiting_threads,
-                                 s32 num_to_wake) {
-    // Only process up to 'target' threads, unless 'target' is <= 0, in which case process
-    // them all.
-    std::size_t last = waiting_threads.size();
-    if (num_to_wake > 0) {
-        last = std::min(last, static_cast<std::size_t>(num_to_wake));
-    }
-
-    // Signal the waiting threads.
-    for (std::size_t i = 0; i < last; i++) {
-        waiting_threads[i]->SetSynchronizationResults(nullptr, RESULT_SUCCESS);
-        RemoveThread(waiting_threads[i]);
-        waiting_threads[i]->WaitForArbitration(false);
-        waiting_threads[i]->Wakeup();
-    }
-}
-
-AddressArbiter::AddressArbiter(Core::System& system) : system{system} {}
-AddressArbiter::~AddressArbiter() = default;
-
-ResultCode AddressArbiter::SignalToAddress(VAddr address, SignalType type, s32 value,
-                                           s32 num_to_wake) {
-    switch (type) {
-    case SignalType::Signal:
-        return SignalToAddressOnly(address, num_to_wake);
-    case SignalType::IncrementAndSignalIfEqual:
-        return IncrementAndSignalToAddressIfEqual(address, value, num_to_wake);
-    case SignalType::ModifyByWaitingCountAndSignalIfEqual:
-        return ModifyByWaitingCountAndSignalToAddressIfEqual(address, value, num_to_wake);
-    default:
-        return ERR_INVALID_ENUM_VALUE;
-    }
-}
-
-ResultCode AddressArbiter::SignalToAddressOnly(VAddr address, s32 num_to_wake) {
-    KScopedSchedulerLock lock(system.Kernel());
-    const std::vector<std::shared_ptr<Thread>> waiting_threads =
-        GetThreadsWaitingOnAddress(address);
-    WakeThreads(waiting_threads, num_to_wake);
-    return RESULT_SUCCESS;
-}
-
-ResultCode AddressArbiter::IncrementAndSignalToAddressIfEqual(VAddr address, s32 value,
-                                                              s32 num_to_wake) {
-    KScopedSchedulerLock lock(system.Kernel());
-    auto& memory = system.Memory();
-
-    // Ensure that we can write to the address.
-    if (!memory.IsValidVirtualAddress(address)) {
-        return ERR_INVALID_ADDRESS_STATE;
-    }
-
-    const std::size_t current_core = system.CurrentCoreIndex();
-    auto& monitor = system.Monitor();
-    u32 current_value;
-    do {
-        current_value = monitor.ExclusiveRead32(current_core, address);
-
-        if (current_value != static_cast<u32>(value)) {
-            return ERR_INVALID_STATE;
-        }
-        current_value++;
-    } while (!monitor.ExclusiveWrite32(current_core, address, current_value));
-
-    return SignalToAddressOnly(address, num_to_wake);
-}
-
-ResultCode AddressArbiter::ModifyByWaitingCountAndSignalToAddressIfEqual(VAddr address, s32 value,
-                                                                         s32 num_to_wake) {
-    KScopedSchedulerLock lock(system.Kernel());
-    auto& memory = system.Memory();
-
-    // Ensure that we can write to the address.
-    if (!memory.IsValidVirtualAddress(address)) {
-        return ERR_INVALID_ADDRESS_STATE;
-    }
-
-    // Get threads waiting on the address.
-    const std::vector<std::shared_ptr<Thread>> waiting_threads =
-        GetThreadsWaitingOnAddress(address);
-
-    const std::size_t current_core = system.CurrentCoreIndex();
-    auto& monitor = system.Monitor();
-    s32 updated_value;
-    do {
-        updated_value = monitor.ExclusiveRead32(current_core, address);
-
-        if (updated_value != value) {
-            return ERR_INVALID_STATE;
-        }
-        // Determine the modified value depending on the waiting count.
-        if (num_to_wake <= 0) {
-            if (waiting_threads.empty()) {
-                updated_value = value + 1;
-            } else {
-                updated_value = value - 1;
-            }
-        } else {
-            if (waiting_threads.empty()) {
-                updated_value = value + 1;
-            } else if (waiting_threads.size() <= static_cast<u32>(num_to_wake)) {
-                updated_value = value - 1;
-            } else {
-                updated_value = value;
-            }
-        }
-    } while (!monitor.ExclusiveWrite32(current_core, address, updated_value));
-
-    WakeThreads(waiting_threads, num_to_wake);
-    return RESULT_SUCCESS;
-}
-
-ResultCode AddressArbiter::WaitForAddress(VAddr address, ArbitrationType type, s32 value,
-                                          s64 timeout_ns) {
-    switch (type) {
-    case ArbitrationType::WaitIfLessThan:
-        return WaitForAddressIfLessThan(address, value, timeout_ns, false);
-    case ArbitrationType::DecrementAndWaitIfLessThan:
-        return WaitForAddressIfLessThan(address, value, timeout_ns, true);
-    case ArbitrationType::WaitIfEqual:
-        return WaitForAddressIfEqual(address, value, timeout_ns);
-    default:
-        return ERR_INVALID_ENUM_VALUE;
-    }
-}
-
-ResultCode AddressArbiter::WaitForAddressIfLessThan(VAddr address, s32 value, s64 timeout,
-                                                    bool should_decrement) {
-    auto& memory = system.Memory();
-    auto& kernel = system.Kernel();
-    Thread* current_thread = kernel.CurrentScheduler()->GetCurrentThread();
-
-    Handle event_handle = InvalidHandle;
-    {
-        KScopedSchedulerLockAndSleep lock(kernel, event_handle, current_thread, timeout);
-
-        if (current_thread->IsTerminationRequested()) {
-            lock.CancelSleep();
-            return ERR_THREAD_TERMINATING;
-        }
-
-        // Ensure that we can read the address.
-        if (!memory.IsValidVirtualAddress(address)) {
-            lock.CancelSleep();
-            return ERR_INVALID_ADDRESS_STATE;
-        }
-
-        s32 current_value = static_cast<s32>(memory.Read32(address));
-        if (current_value >= value) {
-            lock.CancelSleep();
-            return ERR_INVALID_STATE;
-        }
-
-        current_thread->SetSynchronizationResults(nullptr, RESULT_TIMEOUT);
-
-        s32 decrement_value;
-
-        const std::size_t current_core = system.CurrentCoreIndex();
-        auto& monitor = system.Monitor();
-        do {
-            current_value = static_cast<s32>(monitor.ExclusiveRead32(current_core, address));
-            if (should_decrement) {
-                decrement_value = current_value - 1;
-            } else {
-                decrement_value = current_value;
-            }
-        } while (
-            !monitor.ExclusiveWrite32(current_core, address, static_cast<u32>(decrement_value)));
-
-        // Short-circuit without rescheduling, if timeout is zero.
-        if (timeout == 0) {
-            lock.CancelSleep();
-            return RESULT_TIMEOUT;
-        }
-
-        current_thread->SetArbiterWaitAddress(address);
-        InsertThread(SharedFrom(current_thread));
-        current_thread->SetState(ThreadState::Waiting);
-        current_thread->WaitForArbitration(true);
-    }
-
-    if (event_handle != InvalidHandle) {
-        auto& time_manager = kernel.TimeManager();
-        time_manager.UnscheduleTimeEvent(event_handle);
-    }
-
-    {
-        KScopedSchedulerLock lock(kernel);
-        if (current_thread->IsWaitingForArbitration()) {
-            RemoveThread(SharedFrom(current_thread));
-            current_thread->WaitForArbitration(false);
-        }
-    }
-
-    return current_thread->GetSignalingResult();
-}
-
-ResultCode AddressArbiter::WaitForAddressIfEqual(VAddr address, s32 value, s64 timeout) {
-    auto& memory = system.Memory();
-    auto& kernel = system.Kernel();
-    Thread* current_thread = kernel.CurrentScheduler()->GetCurrentThread();
-
-    Handle event_handle = InvalidHandle;
-    {
-        KScopedSchedulerLockAndSleep lock(kernel, event_handle, current_thread, timeout);
-
-        if (current_thread->IsTerminationRequested()) {
-            lock.CancelSleep();
-            return ERR_THREAD_TERMINATING;
-        }
-
-        // Ensure that we can read the address.
-        if (!memory.IsValidVirtualAddress(address)) {
-            lock.CancelSleep();
-            return ERR_INVALID_ADDRESS_STATE;
-        }
-
-        s32 current_value = static_cast<s32>(memory.Read32(address));
-        if (current_value != value) {
-            lock.CancelSleep();
-            return ERR_INVALID_STATE;
-        }
-
-        // Short-circuit without rescheduling, if timeout is zero.
-        if (timeout == 0) {
-            lock.CancelSleep();
-            return RESULT_TIMEOUT;
-        }
-
-        current_thread->SetSynchronizationResults(nullptr, RESULT_TIMEOUT);
-        current_thread->SetArbiterWaitAddress(address);
-        InsertThread(SharedFrom(current_thread));
-        current_thread->SetState(ThreadState::Waiting);
-        current_thread->WaitForArbitration(true);
-    }
-
-    if (event_handle != InvalidHandle) {
-        auto& time_manager = kernel.TimeManager();
-        time_manager.UnscheduleTimeEvent(event_handle);
-    }
-
-    {
-        KScopedSchedulerLock lock(kernel);
-        if (current_thread->IsWaitingForArbitration()) {
-            RemoveThread(SharedFrom(current_thread));
-            current_thread->WaitForArbitration(false);
-        }
-    }
-
-    return current_thread->GetSignalingResult();
-}
-
-void AddressArbiter::InsertThread(std::shared_ptr<Thread> thread) {
-    const VAddr arb_addr = thread->GetArbiterWaitAddress();
-    std::list<std::shared_ptr<Thread>>& thread_list = arb_threads[arb_addr];
-
-    const auto iter =
-        std::find_if(thread_list.cbegin(), thread_list.cend(), [&thread](const auto& entry) {
-            return entry->GetPriority() >= thread->GetPriority();
-        });
-
-    if (iter == thread_list.cend()) {
-        thread_list.push_back(std::move(thread));
-    } else {
-        thread_list.insert(iter, std::move(thread));
-    }
-}
-
-void AddressArbiter::RemoveThread(std::shared_ptr<Thread> thread) {
-    const VAddr arb_addr = thread->GetArbiterWaitAddress();
-    std::list<std::shared_ptr<Thread>>& thread_list = arb_threads[arb_addr];
-
-    const auto iter = std::find_if(thread_list.cbegin(), thread_list.cend(),
-                                   [&thread](const auto& entry) { return thread == entry; });
-
-    if (iter != thread_list.cend()) {
-        thread_list.erase(iter);
-    }
-}
-
-std::vector<std::shared_ptr<Thread>> AddressArbiter::GetThreadsWaitingOnAddress(
-    VAddr address) const {
-    const auto iter = arb_threads.find(address);
-    if (iter == arb_threads.cend()) {
-        return {};
-    }
-
-    const std::list<std::shared_ptr<Thread>>& thread_list = iter->second;
-    return {thread_list.cbegin(), thread_list.cend()};
-}
-} // namespace Kernel
diff --git a/src/core/hle/kernel/address_arbiter.h b/src/core/hle/kernel/address_arbiter.h
deleted file mode 100644
index b91edc67d..000000000
--- a/src/core/hle/kernel/address_arbiter.h
+++ /dev/null
@@ -1,91 +0,0 @@
-// Copyright 2018 yuzu emulator team
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#pragma once
-
-#include <list>
-#include <memory>
-#include <unordered_map>
-#include <vector>
-
-#include "common/common_types.h"
-
-union ResultCode;
-
-namespace Core {
-class System;
-}
-
-namespace Kernel {
-
-class Thread;
-
-class AddressArbiter {
-public:
-    enum class ArbitrationType {
-        WaitIfLessThan = 0,
-        DecrementAndWaitIfLessThan = 1,
-        WaitIfEqual = 2,
-    };
-
-    enum class SignalType {
-        Signal = 0,
-        IncrementAndSignalIfEqual = 1,
-        ModifyByWaitingCountAndSignalIfEqual = 2,
-    };
-
-    explicit AddressArbiter(Core::System& system);
-    ~AddressArbiter();
-
-    AddressArbiter(const AddressArbiter&) = delete;
-    AddressArbiter& operator=(const AddressArbiter&) = delete;
-
-    AddressArbiter(AddressArbiter&&) = default;
-    AddressArbiter& operator=(AddressArbiter&&) = delete;
-
-    /// Signals an address being waited on with a particular signaling type.
-    ResultCode SignalToAddress(VAddr address, SignalType type, s32 value, s32 num_to_wake);
-
-    /// Waits on an address with a particular arbitration type.
-    ResultCode WaitForAddress(VAddr address, ArbitrationType type, s32 value, s64 timeout_ns);
-
-private:
-    /// Signals an address being waited on.
-    ResultCode SignalToAddressOnly(VAddr address, s32 num_to_wake);
-
-    /// Signals an address being waited on and increments its value if equal to the value argument.
-    ResultCode IncrementAndSignalToAddressIfEqual(VAddr address, s32 value, s32 num_to_wake);
-
-    /// Signals an address being waited on and modifies its value based on waiting thread count if
-    /// equal to the value argument.
-    ResultCode ModifyByWaitingCountAndSignalToAddressIfEqual(VAddr address, s32 value,
-                                                             s32 num_to_wake);
-
-    /// Waits on an address if the value passed is less than the argument value,
-    /// optionally decrementing.
-    ResultCode WaitForAddressIfLessThan(VAddr address, s32 value, s64 timeout,
-                                        bool should_decrement);
-
-    /// Waits on an address if the value passed is equal to the argument value.
-    ResultCode WaitForAddressIfEqual(VAddr address, s32 value, s64 timeout);
-
-    /// Wake up num_to_wake (or all) threads in a vector.
-    void WakeThreads(const std::vector<std::shared_ptr<Thread>>& waiting_threads, s32 num_to_wake);
-
-    /// Insert a thread into the address arbiter container
-    void InsertThread(std::shared_ptr<Thread> thread);
-
-    /// Removes a thread from the address arbiter container
-    void RemoveThread(std::shared_ptr<Thread> thread);
-
-    // Gets the threads waiting on an address.
-    std::vector<std::shared_ptr<Thread>> GetThreadsWaitingOnAddress(VAddr address) const;
-
-    /// List of threads waiting for a address arbiter
-    std::unordered_map<VAddr, std::list<std::shared_ptr<Thread>>> arb_threads;
-
-    Core::System& system;
-};
-
-} // namespace Kernel
diff --git a/src/core/hle/kernel/k_synchronization_object.cpp b/src/core/hle/kernel/k_synchronization_object.cpp
index 64c566caa..11b989ecd 100644
--- a/src/core/hle/kernel/k_synchronization_object.cpp
+++ b/src/core/hle/kernel/k_synchronization_object.cpp
@@ -72,7 +72,7 @@ ResultCode KSynchronizationObject::Wait(KernelCore& kernel, s32* out_index,
         }
 
         // For debugging only
-        thread->SetWaitObjectsForDebugging(objects, num_objects);
+        thread->SetWaitObjectsForDebugging({objects, static_cast<std::size_t>(num_objects)});
 
         // Mark the thread as waiting.
         thread->SetCancellable();
@@ -86,7 +86,7 @@ ResultCode KSynchronizationObject::Wait(KernelCore& kernel, s32* out_index,
     thread->ClearCancellable();
 
     // For debugging only
-    thread->SetWaitObjectsForDebugging(nullptr, 0);
+    thread->SetWaitObjectsForDebugging({});
 
     // Cancel the timer as needed.
     if (timer != InvalidHandle) {
diff --git a/src/core/hle/kernel/mutex.cpp b/src/core/hle/kernel/mutex.cpp
deleted file mode 100644
index 8a0faacf8..000000000
--- a/src/core/hle/kernel/mutex.cpp
+++ /dev/null
@@ -1,170 +0,0 @@
-// Copyright 2014 Citra Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#include <memory>
-#include <utility>
-#include <vector>
-
-#include "common/assert.h"
-#include "common/logging/log.h"
-#include "core/core.h"
-#include "core/hle/kernel/errors.h"
-#include "core/hle/kernel/handle_table.h"
-#include "core/hle/kernel/k_scheduler.h"
-#include "core/hle/kernel/kernel.h"
-#include "core/hle/kernel/mutex.h"
-#include "core/hle/kernel/object.h"
-#include "core/hle/kernel/process.h"
-#include "core/hle/kernel/thread.h"
-#include "core/hle/result.h"
-#include "core/memory.h"
-
-namespace Kernel {
-
-/// Returns the number of threads that are waiting for a mutex, and the highest priority one among
-/// those.
-static std::pair<std::shared_ptr<Thread>, u32> GetHighestPriorityMutexWaitingThread(
-    const std::shared_ptr<Thread>& current_thread, VAddr mutex_addr) {
-
-    std::shared_ptr<Thread> highest_priority_thread;
-    u32 num_waiters = 0;
-
-    for (const auto& thread : current_thread->GetMutexWaitingThreads()) {
-        if (thread->GetMutexWaitAddress() != mutex_addr)
-            continue;
-
-        ++num_waiters;
-        if (highest_priority_thread == nullptr ||
-            thread->GetPriority() < highest_priority_thread->GetPriority()) {
-            highest_priority_thread = thread;
-        }
-    }
-
-    return {highest_priority_thread, num_waiters};
-}
-
-/// Update the mutex owner field of all threads waiting on the mutex to point to the new owner.
-static void TransferMutexOwnership(VAddr mutex_addr, std::shared_ptr<Thread> current_thread,
-                                   std::shared_ptr<Thread> new_owner) {
-    current_thread->RemoveMutexWaiter(new_owner);
-    const auto threads = current_thread->GetMutexWaitingThreads();
-    for (const auto& thread : threads) {
-        if (thread->GetMutexWaitAddress() != mutex_addr)
-            continue;
-
-        ASSERT(thread->GetLockOwner() == current_thread.get());
-        current_thread->RemoveMutexWaiter(thread);
-        if (new_owner != thread)
-            new_owner->AddMutexWaiter(thread);
-    }
-}
-
-Mutex::Mutex(Core::System& system) : system{system} {}
-Mutex::~Mutex() = default;
-
-ResultCode Mutex::TryAcquire(VAddr address, Handle holding_thread_handle,
-                             Handle requesting_thread_handle) {
-    // The mutex address must be 4-byte aligned
-    if ((address % sizeof(u32)) != 0) {
-        LOG_ERROR(Kernel, "Address is not 4-byte aligned! address={:016X}", address);
-        return ERR_INVALID_ADDRESS;
-    }
-
-    auto& kernel = system.Kernel();
-    std::shared_ptr<Thread> current_thread =
-        SharedFrom(kernel.CurrentScheduler()->GetCurrentThread());
-    {
-        KScopedSchedulerLock lock(kernel);
-        // The mutex address must be 4-byte aligned
-        if ((address % sizeof(u32)) != 0) {
-            return ERR_INVALID_ADDRESS;
-        }
-
-        const auto& handle_table = kernel.CurrentProcess()->GetHandleTable();
-        std::shared_ptr<Thread> holding_thread = handle_table.Get<Thread>(holding_thread_handle);
-        std::shared_ptr<Thread> requesting_thread =
-            handle_table.Get<Thread>(requesting_thread_handle);
-
-        // TODO(Subv): It is currently unknown if it is possible to lock a mutex in behalf of
-        // another thread.
-        ASSERT(requesting_thread == current_thread);
-
-        current_thread->SetSynchronizationResults(nullptr, RESULT_SUCCESS);
-
-        const u32 addr_value = system.Memory().Read32(address);
-
-        // If the mutex isn't being held, just return success.
-        if (addr_value != (holding_thread_handle | Mutex::MutexHasWaitersFlag)) {
-            return RESULT_SUCCESS;
-        }
-
-        if (holding_thread == nullptr) {
-            return ERR_INVALID_HANDLE;
-        }
-
-        // Wait until the mutex is released
-        current_thread->SetMutexWaitAddress(address);
-        current_thread->SetWaitHandle(requesting_thread_handle);
-
-        current_thread->SetState(ThreadState::Waiting);
-
-        // Update the lock holder thread's priority to prevent priority inversion.
-        holding_thread->AddMutexWaiter(current_thread);
-    }
-
-    {
-        KScopedSchedulerLock lock(kernel);
-        auto* owner = current_thread->GetLockOwner();
-        if (owner != nullptr) {
-            owner->RemoveMutexWaiter(current_thread);
-        }
-    }
-    return current_thread->GetSignalingResult();
-}
-
-std::pair<ResultCode, std::shared_ptr<Thread>> Mutex::Unlock(std::shared_ptr<Thread> owner,
-                                                             VAddr address) {
-    // The mutex address must be 4-byte aligned
-    if ((address % sizeof(u32)) != 0) {
-        LOG_ERROR(Kernel, "Address is not 4-byte aligned! address={:016X}", address);
-        return {ERR_INVALID_ADDRESS, nullptr};
-    }
-
-    auto [new_owner, num_waiters] = GetHighestPriorityMutexWaitingThread(owner, address);
-    if (new_owner == nullptr) {
-        system.Memory().Write32(address, 0);
-        return {RESULT_SUCCESS, nullptr};
-    }
-    // Transfer the ownership of the mutex from the previous owner to the new one.
-    TransferMutexOwnership(address, owner, new_owner);
-    u32 mutex_value = new_owner->GetWaitHandle();
-    if (num_waiters >= 2) {
-        // Notify the guest that there are still some threads waiting for the mutex
-        mutex_value |= Mutex::MutexHasWaitersFlag;
-    }
-    new_owner->SetSynchronizationResults(nullptr, RESULT_SUCCESS);
-    new_owner->SetLockOwner(nullptr);
-    new_owner->Wakeup();
-
-    system.Memory().Write32(address, mutex_value);
-    return {RESULT_SUCCESS, new_owner};
-}
-
-ResultCode Mutex::Release(VAddr address) {
-    auto& kernel = system.Kernel();
-    KScopedSchedulerLock lock(kernel);
-
-    std::shared_ptr<Thread> current_thread =
-        SharedFrom(kernel.CurrentScheduler()->GetCurrentThread());
-
-    auto [result, new_owner] = Unlock(current_thread, address);
-
-    if (result != RESULT_SUCCESS && new_owner != nullptr) {
-        new_owner->SetSynchronizationResults(nullptr, result);
-    }
-
-    return result;
-}
-
-} // namespace Kernel
diff --git a/src/core/hle/kernel/mutex.h b/src/core/hle/kernel/mutex.h
deleted file mode 100644
index 3b81dc3df..000000000
--- a/src/core/hle/kernel/mutex.h
+++ /dev/null
@@ -1,42 +0,0 @@
-// Copyright 2014 Citra Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#pragma once
-
-#include "common/common_types.h"
-
-union ResultCode;
-
-namespace Core {
-class System;
-}
-
-namespace Kernel {
-
-class Mutex final {
-public:
-    explicit Mutex(Core::System& system);
-    ~Mutex();
-
-    /// Flag that indicates that a mutex still has threads waiting for it.
-    static constexpr u32 MutexHasWaitersFlag = 0x40000000;
-    /// Mask of the bits in a mutex address value that contain the mutex owner.
-    static constexpr u32 MutexOwnerMask = 0xBFFFFFFF;
-
-    /// Attempts to acquire a mutex at the specified address.
-    ResultCode TryAcquire(VAddr address, Handle holding_thread_handle,
-                          Handle requesting_thread_handle);
-
-    /// Unlocks a mutex for owner at address
-    std::pair<ResultCode, std::shared_ptr<Thread>> Unlock(std::shared_ptr<Thread> owner,
-                                                          VAddr address);
-
-    /// Releases the mutex at the specified address.
-    ResultCode Release(VAddr address);
-
-private:
-    Core::System& system;
-};
-
-} // namespace Kernel
diff --git a/src/core/hle/kernel/process.cpp b/src/core/hle/kernel/process.cpp
index a306c7c73..37b77fa6e 100644
--- a/src/core/hle/kernel/process.cpp
+++ b/src/core/hle/kernel/process.cpp
@@ -162,48 +162,6 @@ u64 Process::GetTotalPhysicalMemoryUsedWithoutSystemResource() const {
     return GetTotalPhysicalMemoryUsed() - GetSystemResourceUsage();
 }
 
-void Process::InsertConditionVariableThread(std::shared_ptr<Thread> thread) {
-    VAddr cond_var_addr = thread->GetCondVarWaitAddress();
-    std::list<std::shared_ptr<Thread>>& thread_list = cond_var_threads[cond_var_addr];
-    auto it = thread_list.begin();
-    while (it != thread_list.end()) {
-        const std::shared_ptr<Thread> current_thread = *it;
-        if (current_thread->GetPriority() > thread->GetPriority()) {
-            thread_list.insert(it, thread);
-            return;
-        }
-        ++it;
-    }
-    thread_list.push_back(thread);
-}
-
-void Process::RemoveConditionVariableThread(std::shared_ptr<Thread> thread) {
-    VAddr cond_var_addr = thread->GetCondVarWaitAddress();
-    std::list<std::shared_ptr<Thread>>& thread_list = cond_var_threads[cond_var_addr];
-    auto it = thread_list.begin();
-    while (it != thread_list.end()) {
-        const std::shared_ptr<Thread> current_thread = *it;
-        if (current_thread.get() == thread.get()) {
-            thread_list.erase(it);
-            return;
-        }
-        ++it;
-    }
-}
-
-std::vector<std::shared_ptr<Thread>> Process::GetConditionVariableThreads(
-    const VAddr cond_var_addr) {
-    std::vector<std::shared_ptr<Thread>> result{};
-    std::list<std::shared_ptr<Thread>>& thread_list = cond_var_threads[cond_var_addr];
-    auto it = thread_list.begin();
-    while (it != thread_list.end()) {
-        std::shared_ptr<Thread> current_thread = *it;
-        result.push_back(current_thread);
-        ++it;
-    }
-    return result;
-}
-
 void Process::RegisterThread(const Thread* thread) {
     thread_list.push_back(thread);
 }
@@ -412,9 +370,9 @@ bool Process::IsSignaled() const {
 }
 
 Process::Process(Core::System& system)
-    : KSynchronizationObject{system.Kernel()}, page_table{std::make_unique<Memory::PageTable>(
-                                                   system)},
-      handle_table{system.Kernel()}, address_arbiter{system}, mutex{system}, system{system} {}
+    : KSynchronizationObject{system.Kernel()},
+      page_table{std::make_unique<Memory::PageTable>(system)}, handle_table{system.Kernel()},
+      address_arbiter{system}, condition_var{system}, system{system} {}
 
 Process::~Process() = default;
 
diff --git a/src/core/hle/kernel/process.h b/src/core/hle/kernel/process.h
index 901f1ff27..564e1f27d 100644
--- a/src/core/hle/kernel/process.h
+++ b/src/core/hle/kernel/process.h
@@ -11,10 +11,10 @@
 #include <unordered_map>
 #include <vector>
 #include "common/common_types.h"
-#include "core/hle/kernel/address_arbiter.h"
 #include "core/hle/kernel/handle_table.h"
+#include "core/hle/kernel/k_address_arbiter.h"
+#include "core/hle/kernel/k_condition_variable.h"
 #include "core/hle/kernel/k_synchronization_object.h"
-#include "core/hle/kernel/mutex.h"
 #include "core/hle/kernel/process_capability.h"
 #include "core/hle/result.h"
 
@@ -123,24 +123,30 @@ public:
         return handle_table;
     }
 
-    /// Gets a reference to the process' address arbiter.
-    AddressArbiter& GetAddressArbiter() {
-        return address_arbiter;
+    ResultCode SignalToAddress(VAddr address) {
+        return condition_var.SignalToAddress(address);
     }
 
-    /// Gets a const reference to the process' address arbiter.
-    const AddressArbiter& GetAddressArbiter() const {
-        return address_arbiter;
+    ResultCode WaitForAddress(Handle handle, VAddr address, u32 tag) {
+        return condition_var.WaitForAddress(handle, address, tag);
     }
 
-    /// Gets a reference to the process' mutex lock.
-    Mutex& GetMutex() {
-        return mutex;
+    void SignalConditionVariable(u64 cv_key, int32_t count) {
+        return condition_var.Signal(cv_key, count);
     }
 
-    /// Gets a const reference to the process' mutex lock
-    const Mutex& GetMutex() const {
-        return mutex;
+    ResultCode WaitConditionVariable(VAddr address, u64 cv_key, u32 tag, s64 ns) {
+        return condition_var.Wait(address, cv_key, tag, ns);
+    }
+
+    ResultCode SignalAddressArbiter(VAddr address, Svc::SignalType signal_type, s32 value,
+                                    s32 count) {
+        return address_arbiter.SignalToAddress(address, signal_type, value, count);
+    }
+
+    ResultCode WaitAddressArbiter(VAddr address, Svc::ArbitrationType arb_type, s32 value,
+                                  s64 timeout) {
+        return address_arbiter.WaitForAddress(address, arb_type, value, timeout);
     }
 
     /// Gets the address to the process' dedicated TLS region.
@@ -250,15 +256,6 @@ public:
         return thread_list;
     }
 
-    /// Insert a thread into the condition variable wait container
-    void InsertConditionVariableThread(std::shared_ptr<Thread> thread);
-
-    /// Remove a thread from the condition variable wait container
-    void RemoveConditionVariableThread(std::shared_ptr<Thread> thread);
-
-    /// Obtain all condition variable threads waiting for some address
-    std::vector<std::shared_ptr<Thread>> GetConditionVariableThreads(VAddr cond_var_addr);
-
     /// Registers a thread as being created under this process,
     /// adding it to this process' thread list.
     void RegisterThread(const Thread* thread);
@@ -369,12 +366,12 @@ private:
     HandleTable handle_table;
 
     /// Per-process address arbiter.
-    AddressArbiter address_arbiter;
+    KAddressArbiter address_arbiter;
 
     /// The per-process mutex lock instance used for handling various
     /// forms of services, such as lock arbitration, and condition
     /// variable related facilities.
-    Mutex mutex;
+    KConditionVariable condition_var;
 
     /// Address indicating the location of the process' dedicated TLS region.
     VAddr tls_region_address = 0;
@@ -385,9 +382,6 @@ private:
     /// List of threads that are running with this process as their owner.
     std::list<const Thread*> thread_list;
 
-    /// List of threads waiting for a condition variable
-    std::unordered_map<VAddr, std::list<std::shared_ptr<Thread>>> cond_var_threads;
-
     /// Address of the top of the main thread's stack
     VAddr main_thread_stack_top{};
 
diff --git a/src/core/hle/kernel/svc.cpp b/src/core/hle/kernel/svc.cpp
index 304b8727d..99bb4ea20 100644
--- a/src/core/hle/kernel/svc.cpp
+++ b/src/core/hle/kernel/svc.cpp
@@ -10,6 +10,7 @@
 
 #include "common/alignment.h"
 #include "common/assert.h"
+#include "common/common_funcs.h"
 #include "common/fiber.h"
 #include "common/logging/log.h"
 #include "common/microprofile.h"
@@ -19,24 +20,26 @@
 #include "core/core_timing.h"
 #include "core/core_timing_util.h"
 #include "core/cpu_manager.h"
-#include "core/hle/kernel/address_arbiter.h"
 #include "core/hle/kernel/client_port.h"
 #include "core/hle/kernel/client_session.h"
 #include "core/hle/kernel/errors.h"
 #include "core/hle/kernel/handle_table.h"
+#include "core/hle/kernel/k_address_arbiter.h"
+#include "core/hle/kernel/k_condition_variable.h"
 #include "core/hle/kernel/k_scheduler.h"
 #include "core/hle/kernel/k_scoped_scheduler_lock_and_sleep.h"
 #include "core/hle/kernel/k_synchronization_object.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/memory/memory_block.h"
+#include "core/hle/kernel/memory/memory_layout.h"
 #include "core/hle/kernel/memory/page_table.h"
-#include "core/hle/kernel/mutex.h"
 #include "core/hle/kernel/physical_core.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/kernel/readable_event.h"
 #include "core/hle/kernel/resource_limit.h"
 #include "core/hle/kernel/shared_memory.h"
 #include "core/hle/kernel/svc.h"
+#include "core/hle/kernel/svc_results.h"
 #include "core/hle/kernel/svc_types.h"
 #include "core/hle/kernel/svc_wrap.h"
 #include "core/hle/kernel/thread.h"
@@ -347,12 +350,6 @@ static ResultCode SendSyncRequest(Core::System& system, Handle handle) {
         session->SendSyncRequest(SharedFrom(thread), system.Memory(), system.CoreTiming());
     }
 
-    Handle event_handle = thread->GetHLETimeEvent();
-    if (event_handle != InvalidHandle) {
-        auto& time_manager = kernel.TimeManager();
-        time_manager.UnscheduleTimeEvent(event_handle);
-    }
-
     return thread->GetSignalingResult();
 }
 
@@ -491,56 +488,37 @@ static ResultCode CancelSynchronization32(Core::System& system, Handle thread_ha
     return CancelSynchronization(system, thread_handle);
 }
 
-/// Attempts to locks a mutex, creating it if it does not already exist
-static ResultCode ArbitrateLock(Core::System& system, Handle holding_thread_handle,
-                                VAddr mutex_addr, Handle requesting_thread_handle) {
-    LOG_TRACE(Kernel_SVC,
-              "called holding_thread_handle=0x{:08X}, mutex_addr=0x{:X}, "
-              "requesting_current_thread_handle=0x{:08X}",
-              holding_thread_handle, mutex_addr, requesting_thread_handle);
-
-    if (Core::Memory::IsKernelVirtualAddress(mutex_addr)) {
-        LOG_ERROR(Kernel_SVC, "Mutex Address is a kernel virtual address, mutex_addr={:016X}",
-                  mutex_addr);
-        return ERR_INVALID_ADDRESS_STATE;
-    }
+/// Attempts to locks a mutex
+static ResultCode ArbitrateLock(Core::System& system, Handle thread_handle, VAddr address,
+                                u32 tag) {
+    LOG_TRACE(Kernel_SVC, "called thread_handle=0x{:08X}, address=0x{:X}, tag=0x{:08X}",
+              thread_handle, address, tag);
 
-    if (!Common::IsWordAligned(mutex_addr)) {
-        LOG_ERROR(Kernel_SVC, "Mutex Address is not word aligned, mutex_addr={:016X}", mutex_addr);
-        return ERR_INVALID_ADDRESS;
-    }
+    // Validate the input address.
+    R_UNLESS(!Memory::IsKernelAddress(address), Svc::ResultInvalidCurrentMemory);
+    R_UNLESS(Common::IsAligned(address, sizeof(u32)), Svc::ResultInvalidAddress);
 
-    auto* const current_process = system.Kernel().CurrentProcess();
-    return current_process->GetMutex().TryAcquire(mutex_addr, holding_thread_handle,
-                                                  requesting_thread_handle);
+    return system.Kernel().CurrentProcess()->WaitForAddress(thread_handle, address, tag);
 }
 
-static ResultCode ArbitrateLock32(Core::System& system, Handle holding_thread_handle,
-                                  u32 mutex_addr, Handle requesting_thread_handle) {
-    return ArbitrateLock(system, holding_thread_handle, mutex_addr, requesting_thread_handle);
+static ResultCode ArbitrateLock32(Core::System& system, Handle thread_handle, u32 address,
+                                  u32 tag) {
+    return ArbitrateLock(system, thread_handle, address, tag);
 }
 
 /// Unlock a mutex
-static ResultCode ArbitrateUnlock(Core::System& system, VAddr mutex_addr) {
-    LOG_TRACE(Kernel_SVC, "called mutex_addr=0x{:X}", mutex_addr);
+static ResultCode ArbitrateUnlock(Core::System& system, VAddr address) {
+    LOG_TRACE(Kernel_SVC, "called address=0x{:X}", address);
 
-    if (Core::Memory::IsKernelVirtualAddress(mutex_addr)) {
-        LOG_ERROR(Kernel_SVC, "Mutex Address is a kernel virtual address, mutex_addr={:016X}",
-                  mutex_addr);
-        return ERR_INVALID_ADDRESS_STATE;
-    }
+    // Validate the input address.
+    R_UNLESS(!Memory::IsKernelAddress(address), Svc::ResultInvalidCurrentMemory);
+    R_UNLESS(Common::IsAligned(address, sizeof(u32)), Svc::ResultInvalidAddress);
 
-    if (!Common::IsWordAligned(mutex_addr)) {
-        LOG_ERROR(Kernel_SVC, "Mutex Address is not word aligned, mutex_addr={:016X}", mutex_addr);
-        return ERR_INVALID_ADDRESS;
-    }
-
-    auto* const current_process = system.Kernel().CurrentProcess();
-    return current_process->GetMutex().Release(mutex_addr);
+    return system.Kernel().CurrentProcess()->SignalToAddress(address);
 }
 
-static ResultCode ArbitrateUnlock32(Core::System& system, u32 mutex_addr) {
-    return ArbitrateUnlock(system, mutex_addr);
+static ResultCode ArbitrateUnlock32(Core::System& system, u32 address) {
+    return ArbitrateUnlock(system, address);
 }
 
 enum class BreakType : u32 {
@@ -1167,7 +1145,7 @@ static ResultCode SetThreadPriority(Core::System& system, Handle handle, u32 pri
         return ERR_INVALID_HANDLE;
     }
 
-    thread->SetPriority(priority);
+    thread->SetBasePriority(priority);
 
     return RESULT_SUCCESS;
 }
@@ -1607,223 +1585,135 @@ static void SleepThread32(Core::System& system, u32 nanoseconds_low, u32 nanosec
 }
 
 /// Wait process wide key atomic
-static ResultCode WaitProcessWideKeyAtomic(Core::System& system, VAddr mutex_addr,
-                                           VAddr condition_variable_addr, Handle thread_handle,
-                                           s64 nano_seconds) {
-    LOG_TRACE(
-        Kernel_SVC,
-        "called mutex_addr={:X}, condition_variable_addr={:X}, thread_handle=0x{:08X}, timeout={}",
-        mutex_addr, condition_variable_addr, thread_handle, nano_seconds);
-
-    if (Core::Memory::IsKernelVirtualAddress(mutex_addr)) {
-        LOG_ERROR(
-            Kernel_SVC,
-            "Given mutex address must not be within the kernel address space. address=0x{:016X}",
-            mutex_addr);
-        return ERR_INVALID_ADDRESS_STATE;
-    }
-
-    if (!Common::IsWordAligned(mutex_addr)) {
-        LOG_ERROR(Kernel_SVC, "Given mutex address must be word-aligned. address=0x{:016X}",
-                  mutex_addr);
-        return ERR_INVALID_ADDRESS;
-    }
-
-    ASSERT(condition_variable_addr == Common::AlignDown(condition_variable_addr, 4));
-    auto& kernel = system.Kernel();
-    Handle event_handle;
-    Thread* current_thread = kernel.CurrentScheduler()->GetCurrentThread();
-    auto* const current_process = kernel.CurrentProcess();
-    {
-        KScopedSchedulerLockAndSleep lock(kernel, event_handle, current_thread, nano_seconds);
-        const auto& handle_table = current_process->GetHandleTable();
-        std::shared_ptr<Thread> thread = handle_table.Get<Thread>(thread_handle);
-        ASSERT(thread);
-
-        current_thread->SetSynchronizationResults(nullptr, RESULT_TIMEOUT);
-
-        if (thread->IsTerminationRequested()) {
-            lock.CancelSleep();
-            return ERR_THREAD_TERMINATING;
-        }
-
-        const auto release_result = current_process->GetMutex().Release(mutex_addr);
-        if (release_result.IsError()) {
-            lock.CancelSleep();
-            return release_result;
-        }
-
-        if (nano_seconds == 0) {
-            lock.CancelSleep();
-            return RESULT_TIMEOUT;
-        }
-
-        current_thread->SetCondVarWaitAddress(condition_variable_addr);
-        current_thread->SetMutexWaitAddress(mutex_addr);
-        current_thread->SetWaitHandle(thread_handle);
-        current_thread->SetState(ThreadState::Waiting);
-        current_thread->SetWaitingCondVar(true);
-        current_process->InsertConditionVariableThread(SharedFrom(current_thread));
-    }
-
-    if (event_handle != InvalidHandle) {
-        auto& time_manager = kernel.TimeManager();
-        time_manager.UnscheduleTimeEvent(event_handle);
-    }
-
-    {
-        KScopedSchedulerLock lock(kernel);
-
-        auto* owner = current_thread->GetLockOwner();
-        if (owner != nullptr) {
-            owner->RemoveMutexWaiter(SharedFrom(current_thread));
+static ResultCode WaitProcessWideKeyAtomic(Core::System& system, VAddr address, VAddr cv_key,
+                                           u32 tag, s64 timeout_ns) {
+    LOG_TRACE(Kernel_SVC, "called address={:X}, cv_key={:X}, tag=0x{:08X}, timeout_ns={}", address,
+              cv_key, tag, timeout_ns);
+
+    // Validate input.
+    R_UNLESS(!Memory::IsKernelAddress(address), Svc::ResultInvalidCurrentMemory);
+    R_UNLESS(Common::IsAligned(address, sizeof(int32_t)), Svc::ResultInvalidAddress);
+
+    // Convert timeout from nanoseconds to ticks.
+    s64 timeout{};
+    if (timeout_ns > 0) {
+        const s64 offset_tick(timeout_ns);
+        if (offset_tick > 0) {
+            timeout = offset_tick + 2;
+            if (timeout <= 0) {
+                timeout = std::numeric_limits<s64>::max();
+            }
+        } else {
+            timeout = std::numeric_limits<s64>::max();
         }
-
-        current_process->RemoveConditionVariableThread(SharedFrom(current_thread));
+    } else {
+        timeout = timeout_ns;
     }
-    // Note: Deliberately don't attempt to inherit the lock owner's priority.
 
-    return current_thread->GetSignalingResult();
+    // Wait on the condition variable.
+    return system.Kernel().CurrentProcess()->WaitConditionVariable(
+        address, Common::AlignDown(cv_key, sizeof(u32)), tag, timeout);
 }
 
-static ResultCode WaitProcessWideKeyAtomic32(Core::System& system, u32 mutex_addr,
-                                             u32 condition_variable_addr, Handle thread_handle,
-                                             u32 nanoseconds_low, u32 nanoseconds_high) {
-    const auto nanoseconds = static_cast<s64>(nanoseconds_low | (u64{nanoseconds_high} << 32));
-    return WaitProcessWideKeyAtomic(system, mutex_addr, condition_variable_addr, thread_handle,
-                                    nanoseconds);
+static ResultCode WaitProcessWideKeyAtomic32(Core::System& system, u32 address, u32 cv_key, u32 tag,
+                                             u32 timeout_ns_low, u32 timeout_ns_high) {
+    const auto timeout_ns = static_cast<s64>(timeout_ns_low | (u64{timeout_ns_high} << 32));
+    return WaitProcessWideKeyAtomic(system, address, cv_key, tag, timeout_ns);
 }
 
 /// Signal process wide key
-static void SignalProcessWideKey(Core::System& system, VAddr condition_variable_addr, s32 target) {
-    LOG_TRACE(Kernel_SVC, "called, condition_variable_addr=0x{:X}, target=0x{:08X}",
-              condition_variable_addr, target);
+static void SignalProcessWideKey(Core::System& system, VAddr cv_key, s32 count) {
+    LOG_TRACE(Kernel_SVC, "called, cv_key=0x{:X}, count=0x{:08X}", cv_key, count);
+
+    // Signal the condition variable.
+    return system.Kernel().CurrentProcess()->SignalConditionVariable(
+        Common::AlignDown(cv_key, sizeof(u32)), count);
+}
 
-    ASSERT(condition_variable_addr == Common::AlignDown(condition_variable_addr, 4));
+static void SignalProcessWideKey32(Core::System& system, u32 cv_key, s32 count) {
+    SignalProcessWideKey(system, cv_key, count);
+}
 
-    // Retrieve a list of all threads that are waiting for this condition variable.
-    auto& kernel = system.Kernel();
-    KScopedSchedulerLock lock(kernel);
-    auto* const current_process = kernel.CurrentProcess();
-    std::vector<std::shared_ptr<Thread>> waiting_threads =
-        current_process->GetConditionVariableThreads(condition_variable_addr);
-
-    // Only process up to 'target' threads, unless 'target' is less equal 0, in which case process
-    // them all.
-    std::size_t last = waiting_threads.size();
-    if (target > 0) {
-        last = std::min(waiting_threads.size(), static_cast<std::size_t>(target));
-    }
-    for (std::size_t index = 0; index < last; ++index) {
-        auto& thread = waiting_threads[index];
-
-        ASSERT(thread->GetCondVarWaitAddress() == condition_variable_addr);
-
-        // liberate Cond Var Thread.
-        current_process->RemoveConditionVariableThread(thread);
-
-        const std::size_t current_core = system.CurrentCoreIndex();
-        auto& monitor = system.Monitor();
-
-        // Atomically read the value of the mutex.
-        u32 mutex_val = 0;
-        u32 update_val = 0;
-        const VAddr mutex_address = thread->GetMutexWaitAddress();
-        do {
-            // If the mutex is not yet acquired, acquire it.
-            mutex_val = monitor.ExclusiveRead32(current_core, mutex_address);
-
-            if (mutex_val != 0) {
-                update_val = mutex_val | Mutex::MutexHasWaitersFlag;
-            } else {
-                update_val = thread->GetWaitHandle();
-            }
-        } while (!monitor.ExclusiveWrite32(current_core, mutex_address, update_val));
-        monitor.ClearExclusive();
-        if (mutex_val == 0) {
-            // We were able to acquire the mutex, resume this thread.
-            auto* const lock_owner = thread->GetLockOwner();
-            if (lock_owner != nullptr) {
-                lock_owner->RemoveMutexWaiter(thread);
-            }
+namespace {
 
-            thread->SetLockOwner(nullptr);
-            thread->SetSynchronizationResults(nullptr, RESULT_SUCCESS);
-            thread->Wakeup();
-        } else {
-            // The mutex is already owned by some other thread, make this thread wait on it.
-            const Handle owner_handle = static_cast<Handle>(mutex_val & Mutex::MutexOwnerMask);
-            const auto& handle_table = system.Kernel().CurrentProcess()->GetHandleTable();
-            auto owner = handle_table.Get<Thread>(owner_handle);
-            ASSERT(owner);
-            thread->SetWaitingCondVar(false);
-
-            owner->AddMutexWaiter(thread);
-        }
+constexpr bool IsValidSignalType(Svc::SignalType type) {
+    switch (type) {
+    case Svc::SignalType::Signal:
+    case Svc::SignalType::SignalAndIncrementIfEqual:
+    case Svc::SignalType::SignalAndModifyByWaitingCountIfEqual:
+        return true;
+    default:
+        return false;
     }
 }
 
-static void SignalProcessWideKey32(Core::System& system, u32 condition_variable_addr, s32 target) {
-    SignalProcessWideKey(system, condition_variable_addr, target);
+constexpr bool IsValidArbitrationType(Svc::ArbitrationType type) {
+    switch (type) {
+    case Svc::ArbitrationType::WaitIfLessThan:
+    case Svc::ArbitrationType::DecrementAndWaitIfLessThan:
+    case Svc::ArbitrationType::WaitIfEqual:
+        return true;
+    default:
+        return false;
+    }
 }
 
-// Wait for an address (via Address Arbiter)
-static ResultCode WaitForAddress(Core::System& system, VAddr address, u32 type, s32 value,
-                                 s64 timeout) {
-    LOG_TRACE(Kernel_SVC, "called, address=0x{:X}, type=0x{:X}, value=0x{:X}, timeout={}", address,
-              type, value, timeout);
-
-    // If the passed address is a kernel virtual address, return invalid memory state.
-    if (Core::Memory::IsKernelVirtualAddress(address)) {
-        LOG_ERROR(Kernel_SVC, "Address is a kernel virtual address, address={:016X}", address);
-        return ERR_INVALID_ADDRESS_STATE;
-    }
+} // namespace
 
-    // If the address is not properly aligned to 4 bytes, return invalid address.
-    if (!Common::IsWordAligned(address)) {
-        LOG_ERROR(Kernel_SVC, "Address is not word aligned, address={:016X}", address);
-        return ERR_INVALID_ADDRESS;
+// Wait for an address (via Address Arbiter)
+static ResultCode WaitForAddress(Core::System& system, VAddr address, Svc::ArbitrationType arb_type,
+                                 s32 value, s64 timeout_ns) {
+    LOG_TRACE(Kernel_SVC, "called, address=0x{:X}, arb_type=0x{:X}, value=0x{:X}, timeout_ns={}",
+              address, arb_type, value, timeout_ns);
+
+    // Validate input.
+    R_UNLESS(!Memory::IsKernelAddress(address), Svc::ResultInvalidCurrentMemory);
+    R_UNLESS(Common::IsAligned(address, sizeof(int32_t)), Svc::ResultInvalidAddress);
+    R_UNLESS(IsValidArbitrationType(arb_type), Svc::ResultInvalidEnumValue);
+
+    // Convert timeout from nanoseconds to ticks.
+    s64 timeout{};
+    if (timeout_ns > 0) {
+        const s64 offset_tick(timeout_ns);
+        if (offset_tick > 0) {
+            timeout = offset_tick + 2;
+            if (timeout <= 0) {
+                timeout = std::numeric_limits<s64>::max();
+            }
+        } else {
+            timeout = std::numeric_limits<s64>::max();
+        }
+    } else {
+        timeout = timeout_ns;
     }
 
-    const auto arbitration_type = static_cast<AddressArbiter::ArbitrationType>(type);
-    auto& address_arbiter = system.Kernel().CurrentProcess()->GetAddressArbiter();
-    const ResultCode result =
-        address_arbiter.WaitForAddress(address, arbitration_type, value, timeout);
-    return result;
+    return system.Kernel().CurrentProcess()->WaitAddressArbiter(address, arb_type, value, timeout);
 }
 
-static ResultCode WaitForAddress32(Core::System& system, u32 address, u32 type, s32 value,
-                                   u32 timeout_low, u32 timeout_high) {
-    const auto timeout = static_cast<s64>(timeout_low | (u64{timeout_high} << 32));
-    return WaitForAddress(system, address, type, value, timeout);
+static ResultCode WaitForAddress32(Core::System& system, u32 address, Svc::ArbitrationType arb_type,
+                                   s32 value, u32 timeout_ns_low, u32 timeout_ns_high) {
+    const auto timeout = static_cast<s64>(timeout_ns_low | (u64{timeout_ns_high} << 32));
+    return WaitForAddress(system, address, arb_type, value, timeout);
 }
 
 // Signals to an address (via Address Arbiter)
-static ResultCode SignalToAddress(Core::System& system, VAddr address, u32 type, s32 value,
-                                  s32 num_to_wake) {
-    LOG_TRACE(Kernel_SVC, "called, address=0x{:X}, type=0x{:X}, value=0x{:X}, num_to_wake=0x{:X}",
-              address, type, value, num_to_wake);
-
-    // If the passed address is a kernel virtual address, return invalid memory state.
-    if (Core::Memory::IsKernelVirtualAddress(address)) {
-        LOG_ERROR(Kernel_SVC, "Address is a kernel virtual address, address={:016X}", address);
-        return ERR_INVALID_ADDRESS_STATE;
-    }
+static ResultCode SignalToAddress(Core::System& system, VAddr address, Svc::SignalType signal_type,
+                                  s32 value, s32 count) {
+    LOG_TRACE(Kernel_SVC, "called, address=0x{:X}, signal_type=0x{:X}, value=0x{:X}, count=0x{:X}",
+              address, signal_type, value, count);
 
-    // If the address is not properly aligned to 4 bytes, return invalid address.
-    if (!Common::IsWordAligned(address)) {
-        LOG_ERROR(Kernel_SVC, "Address is not word aligned, address={:016X}", address);
-        return ERR_INVALID_ADDRESS;
-    }
+    // Validate input.
+    R_UNLESS(!Memory::IsKernelAddress(address), Svc::ResultInvalidCurrentMemory);
+    R_UNLESS(Common::IsAligned(address, sizeof(s32)), Svc::ResultInvalidAddress);
+    R_UNLESS(IsValidSignalType(signal_type), Svc::ResultInvalidEnumValue);
 
-    const auto signal_type = static_cast<AddressArbiter::SignalType>(type);
-    auto& address_arbiter = system.Kernel().CurrentProcess()->GetAddressArbiter();
-    return address_arbiter.SignalToAddress(address, signal_type, value, num_to_wake);
+    return system.Kernel().CurrentProcess()->SignalAddressArbiter(address, signal_type, value,
+                                                                  count);
 }
 
-static ResultCode SignalToAddress32(Core::System& system, u32 address, u32 type, s32 value,
-                                    s32 num_to_wake) {
-    return SignalToAddress(system, address, type, value, num_to_wake);
+static ResultCode SignalToAddress32(Core::System& system, u32 address, Svc::SignalType signal_type,
+                                    s32 value, s32 count) {
+    return SignalToAddress(system, address, signal_type, value, count);
 }
 
 static void KernelDebug([[maybe_unused]] Core::System& system,
diff --git a/src/core/hle/kernel/svc_common.h b/src/core/hle/kernel/svc_common.h
index 7734bb236..4af049551 100644
--- a/src/core/hle/kernel/svc_common.h
+++ b/src/core/hle/kernel/svc_common.h
@@ -8,6 +8,7 @@
 
 namespace Kernel::Svc {
 
+constexpr s32 ArgumentHandleCountMax = 0x40;
 constexpr u32 HandleWaitMask{1u << 30};
 
 } // namespace Kernel::Svc
diff --git a/src/core/hle/kernel/svc_wrap.h b/src/core/hle/kernel/svc_wrap.h
index f94c487ba..a32750ed7 100644
--- a/src/core/hle/kernel/svc_wrap.h
+++ b/src/core/hle/kernel/svc_wrap.h
@@ -7,6 +7,7 @@
 #include "common/common_types.h"
 #include "core/arm/arm_interface.h"
 #include "core/core.h"
+#include "core/hle/kernel/svc_types.h"
 #include "core/hle/result.h"
 
 namespace Kernel {
@@ -277,18 +278,22 @@ void SvcWrap64(Core::System& system) {
     FuncReturn(system, retval);
 }
 
-template <ResultCode func(Core::System&, u64, u32, s32, s64)>
+// Used by WaitForAddress
+template <ResultCode func(Core::System&, u64, Svc::ArbitrationType, s32, s64)>
 void SvcWrap64(Core::System& system) {
-    FuncReturn(system, func(system, Param(system, 0), static_cast<u32>(Param(system, 1)),
-                            static_cast<s32>(Param(system, 2)), static_cast<s64>(Param(system, 3)))
-                           .raw);
+    FuncReturn(system,
+               func(system, Param(system, 0), static_cast<Svc::ArbitrationType>(Param(system, 1)),
+                    static_cast<s32>(Param(system, 2)), static_cast<s64>(Param(system, 3)))
+                   .raw);
 }
 
-template <ResultCode func(Core::System&, u64, u32, s32, s32)>
+// Used by SignalToAddress
+template <ResultCode func(Core::System&, u64, Svc::SignalType, s32, s32)>
 void SvcWrap64(Core::System& system) {
-    FuncReturn(system, func(system, Param(system, 0), static_cast<u32>(Param(system, 1)),
-                            static_cast<s32>(Param(system, 2)), static_cast<s32>(Param(system, 3)))
-                           .raw);
+    FuncReturn(system,
+               func(system, Param(system, 0), static_cast<Svc::SignalType>(Param(system, 1)),
+                    static_cast<s32>(Param(system, 2)), static_cast<s32>(Param(system, 3)))
+                   .raw);
 }
 
 ////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -504,22 +509,23 @@ void SvcWrap32(Core::System& system) {
 }
 
 // Used by WaitForAddress32
-template <ResultCode func(Core::System&, u32, u32, s32, u32, u32)>
+template <ResultCode func(Core::System&, u32, Svc::ArbitrationType, s32, u32, u32)>
 void SvcWrap32(Core::System& system) {
     const u32 retval = func(system, static_cast<u32>(Param(system, 0)),
-                            static_cast<u32>(Param(system, 1)), static_cast<s32>(Param(system, 2)),
-                            static_cast<u32>(Param(system, 3)), static_cast<u32>(Param(system, 4)))
+                            static_cast<Svc::ArbitrationType>(Param(system, 1)),
+                            static_cast<s32>(Param(system, 2)), static_cast<u32>(Param(system, 3)),
+                            static_cast<u32>(Param(system, 4)))
                            .raw;
     FuncReturn(system, retval);
 }
 
 // Used by SignalToAddress32
-template <ResultCode func(Core::System&, u32, u32, s32, s32)>
+template <ResultCode func(Core::System&, u32, Svc::SignalType, s32, s32)>
 void SvcWrap32(Core::System& system) {
-    const u32 retval =
-        func(system, static_cast<u32>(Param(system, 0)), static_cast<u32>(Param(system, 1)),
-             static_cast<s32>(Param(system, 2)), static_cast<s32>(Param(system, 3)))
-            .raw;
+    const u32 retval = func(system, static_cast<u32>(Param(system, 0)),
+                            static_cast<Svc::SignalType>(Param(system, 1)),
+                            static_cast<s32>(Param(system, 2)), static_cast<s32>(Param(system, 3)))
+                           .raw;
     FuncReturn(system, retval);
 }
 
diff --git a/src/core/hle/kernel/thread.cpp b/src/core/hle/kernel/thread.cpp
index 33a4e1fa3..eda56c31c 100644
--- a/src/core/hle/kernel/thread.cpp
+++ b/src/core/hle/kernel/thread.cpp
@@ -17,9 +17,11 @@
 #include "core/hardware_properties.h"
 #include "core/hle/kernel/errors.h"
 #include "core/hle/kernel/handle_table.h"
+#include "core/hle/kernel/k_condition_variable.h"
 #include "core/hle/kernel/k_scheduler.h"
 #include "core/hle/kernel/k_scoped_scheduler_lock_and_sleep.h"
 #include "core/hle/kernel/kernel.h"
+#include "core/hle/kernel/memory/memory_layout.h"
 #include "core/hle/kernel/object.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/kernel/thread.h"
@@ -62,24 +64,6 @@ void Thread::Stop() {
 
 void Thread::Wakeup() {
     KScopedSchedulerLock lock(kernel);
-    switch (thread_state) {
-    case ThreadState::Runnable:
-        // If the thread is waiting on multiple wait objects, it might be awoken more than once
-        // before actually resuming. We can ignore subsequent wakeups if the thread status has
-        // already been set to ThreadStatus::Ready.
-        return;
-    case ThreadState::Terminated:
-        // This should never happen, as threads must complete before being stopped.
-        DEBUG_ASSERT_MSG(false, "Thread with object id {} cannot be resumed because it's DEAD.",
-                         GetObjectId());
-        return;
-    }
-
-    SetState(ThreadState::Runnable);
-}
-
-void Thread::OnWakeUp() {
-    KScopedSchedulerLock lock(kernel);
     SetState(ThreadState::Runnable);
 }
 
@@ -167,15 +151,14 @@ ResultVal<std::shared_ptr<Thread>> Thread::Create(Core::System& system, ThreadTy
     thread->stack_top = stack_top;
     thread->disable_count = 1;
     thread->tpidr_el0 = 0;
-    thread->nominal_priority = thread->current_priority = priority;
+    thread->current_priority = priority;
+    thread->base_priority = priority;
+    thread->lock_owner = nullptr;
     thread->schedule_count = -1;
     thread->last_scheduled_tick = 0;
     thread->processor_id = processor_id;
     thread->ideal_core = processor_id;
     thread->affinity_mask.SetAffinity(processor_id, true);
-    thread->mutex_wait_address = 0;
-    thread->condvar_wait_address = 0;
-    thread->wait_handle = 0;
     thread->name = std::move(name);
     thread->global_handle = kernel.GlobalHandleTable().Create(thread).Unwrap();
     thread->owner_process = owner_process;
@@ -205,12 +188,17 @@ ResultVal<std::shared_ptr<Thread>> Thread::Create(Core::System& system, ThreadTy
     return MakeResult<std::shared_ptr<Thread>>(std::move(thread));
 }
 
-void Thread::SetPriority(u32 priority) {
-    KScopedSchedulerLock lock(kernel);
+void Thread::SetBasePriority(u32 priority) {
     ASSERT_MSG(priority <= THREADPRIO_LOWEST && priority >= THREADPRIO_HIGHEST,
                "Invalid priority value.");
-    nominal_priority = priority;
-    UpdatePriority();
+
+    KScopedSchedulerLock lock(kernel);
+
+    // Change our base priority.
+    base_priority = priority;
+
+    // Perform a priority restoration.
+    RestorePriority(kernel, this);
 }
 
 void Thread::SetSynchronizationResults(KSynchronizationObject* object, ResultCode result) {
@@ -224,95 +212,146 @@ VAddr Thread::GetCommandBufferAddress() const {
     return GetTLSAddress() + command_header_offset;
 }
 
-void Thread::SetState(ThreadState new_status) {
-    if (new_status == thread_state) {
-        return;
+void Thread::SetState(ThreadState state) {
+    KScopedSchedulerLock sl(kernel);
+
+    SetMutexWaitAddressForDebugging(0);
+    const ThreadState old_state = thread_state;
+    thread_state =
+        static_cast<ThreadState>((old_state & ~ThreadState::Mask) | (state & ThreadState::Mask));
+    if (thread_state != old_state) {
+        KScheduler::OnThreadStateChanged(kernel, this, old_state);
     }
+}
+
+void Thread::AddWaiterImpl(Thread* thread) {
+    ASSERT(kernel.GlobalSchedulerContext().IsLocked());
 
-    if (new_status != ThreadState::Waiting) {
-        SetWaitingCondVar(false);
+    // Find the right spot to insert the waiter.
+    auto it = waiter_list.begin();
+    while (it != waiter_list.end()) {
+        if (it->GetPriority() > thread->GetPriority()) {
+            break;
+        }
+        it++;
     }
 
-    SetSchedulingStatus(new_status);
+    // Keep track of how many kernel waiters we have.
+    if (Memory::IsKernelAddressKey(thread->GetAddressKey())) {
+        ASSERT((num_kernel_waiters++) >= 0);
+    }
 
-    thread_state = new_status;
+    // Insert the waiter.
+    waiter_list.insert(it, *thread);
+    thread->SetLockOwner(this);
 }
 
-void Thread::AddMutexWaiter(std::shared_ptr<Thread> thread) {
-    if (thread->lock_owner.get() == this) {
-        // If the thread is already waiting for this thread to release the mutex, ensure that the
-        // waiters list is consistent and return without doing anything.
-        const auto iter = std::find(wait_mutex_threads.begin(), wait_mutex_threads.end(), thread);
-        ASSERT(iter != wait_mutex_threads.end());
-        return;
+void Thread::RemoveWaiterImpl(Thread* thread) {
+    ASSERT(kernel.GlobalSchedulerContext().IsLocked());
+
+    // Keep track of how many kernel waiters we have.
+    if (Memory::IsKernelAddressKey(thread->GetAddressKey())) {
+        ASSERT((num_kernel_waiters--) > 0);
     }
 
-    // A thread can't wait on two different mutexes at the same time.
-    ASSERT(thread->lock_owner == nullptr);
+    // Remove the waiter.
+    waiter_list.erase(waiter_list.iterator_to(*thread));
+    thread->SetLockOwner(nullptr);
+}
 
-    // Ensure that the thread is not already in the list of mutex waiters
-    const auto iter = std::find(wait_mutex_threads.begin(), wait_mutex_threads.end(), thread);
-    ASSERT(iter == wait_mutex_threads.end());
+void Thread::RestorePriority(KernelCore& kernel, Thread* thread) {
+    ASSERT(kernel.GlobalSchedulerContext().IsLocked());
 
-    // Keep the list in an ordered fashion
-    const auto insertion_point = std::find_if(
-        wait_mutex_threads.begin(), wait_mutex_threads.end(),
-        [&thread](const auto& entry) { return entry->GetPriority() > thread->GetPriority(); });
-    wait_mutex_threads.insert(insertion_point, thread);
-    thread->lock_owner = SharedFrom(this);
+    while (true) {
+        // We want to inherit priority where possible.
+        s32 new_priority = thread->GetBasePriority();
+        if (thread->HasWaiters()) {
+            new_priority = std::min(new_priority, thread->waiter_list.front().GetPriority());
+        }
 
-    UpdatePriority();
-}
+        // If the priority we would inherit is not different from ours, don't do anything.
+        if (new_priority == thread->GetPriority()) {
+            return;
+        }
 
-void Thread::RemoveMutexWaiter(std::shared_ptr<Thread> thread) {
-    ASSERT(thread->lock_owner.get() == this);
+        // Ensure we don't violate condition variable red black tree invariants.
+        if (auto* cv_tree = thread->GetConditionVariableTree(); cv_tree != nullptr) {
+            BeforeUpdatePriority(kernel, cv_tree, thread);
+        }
 
-    // Ensure that the thread is in the list of mutex waiters
-    const auto iter = std::find(wait_mutex_threads.begin(), wait_mutex_threads.end(), thread);
-    ASSERT(iter != wait_mutex_threads.end());
+        // Change the priority.
+        const s32 old_priority = thread->GetPriority();
+        thread->SetPriority(new_priority);
 
-    wait_mutex_threads.erase(iter);
+        // Restore the condition variable, if relevant.
+        if (auto* cv_tree = thread->GetConditionVariableTree(); cv_tree != nullptr) {
+            AfterUpdatePriority(kernel, cv_tree, thread);
+        }
 
-    thread->lock_owner = nullptr;
-    UpdatePriority();
-}
+        // Update the scheduler.
+        KScheduler::OnThreadPriorityChanged(kernel, thread, old_priority);
 
-void Thread::UpdatePriority() {
-    // If any of the threads waiting on the mutex have a higher priority
-    // (taking into account priority inheritance), then this thread inherits
-    // that thread's priority.
-    u32 new_priority = nominal_priority;
-    if (!wait_mutex_threads.empty()) {
-        if (wait_mutex_threads.front()->current_priority < new_priority) {
-            new_priority = wait_mutex_threads.front()->current_priority;
+        // Keep the lock owner up to date.
+        Thread* lock_owner = thread->GetLockOwner();
+        if (lock_owner == nullptr) {
+            return;
         }
-    }
 
-    if (new_priority == current_priority) {
-        return;
+        // Update the thread in the lock owner's sorted list, and continue inheriting.
+        lock_owner->RemoveWaiterImpl(thread);
+        lock_owner->AddWaiterImpl(thread);
+        thread = lock_owner;
     }
+}
 
-    if (GetState() == ThreadState::Waiting && is_waiting_on_condvar) {
-        owner_process->RemoveConditionVariableThread(SharedFrom(this));
-    }
+void Thread::AddWaiter(Thread* thread) {
+    AddWaiterImpl(thread);
+    RestorePriority(kernel, this);
+}
 
-    SetCurrentPriority(new_priority);
+void Thread::RemoveWaiter(Thread* thread) {
+    RemoveWaiterImpl(thread);
+    RestorePriority(kernel, this);
+}
 
-    if (GetState() == ThreadState::Waiting && is_waiting_on_condvar) {
-        owner_process->InsertConditionVariableThread(SharedFrom(this));
-    }
+Thread* Thread::RemoveWaiterByKey(s32* out_num_waiters, VAddr key) {
+    ASSERT(kernel.GlobalSchedulerContext().IsLocked());
 
-    if (!lock_owner) {
-        return;
+    s32 num_waiters{};
+    Thread* next_lock_owner{};
+    auto it = waiter_list.begin();
+    while (it != waiter_list.end()) {
+        if (it->GetAddressKey() == key) {
+            Thread* thread = std::addressof(*it);
+
+            // Keep track of how many kernel waiters we have.
+            if (Memory::IsKernelAddressKey(thread->GetAddressKey())) {
+                ASSERT((num_kernel_waiters--) > 0);
+            }
+            it = waiter_list.erase(it);
+
+            // Update the next lock owner.
+            if (next_lock_owner == nullptr) {
+                next_lock_owner = thread;
+                next_lock_owner->SetLockOwner(nullptr);
+            } else {
+                next_lock_owner->AddWaiterImpl(thread);
+            }
+            num_waiters++;
+        } else {
+            it++;
+        }
     }
 
-    // Ensure that the thread is within the correct location in the waiting list.
-    auto old_owner = lock_owner;
-    lock_owner->RemoveMutexWaiter(SharedFrom(this));
-    old_owner->AddMutexWaiter(SharedFrom(this));
+    // Do priority updates, if we have a next owner.
+    if (next_lock_owner) {
+        RestorePriority(kernel, this);
+        RestorePriority(kernel, next_lock_owner);
+    }
 
-    // Recursively update the priority of the thread that depends on the priority of this one.
-    lock_owner->UpdatePriority();
+    // Return output.
+    *out_num_waiters = num_waiters;
+    return next_lock_owner;
 }
 
 ResultCode Thread::SetActivity(ThreadActivity value) {
@@ -372,18 +411,6 @@ void Thread::RemoveSchedulingFlag(ThreadSchedFlags flag) {
     KScheduler::OnThreadStateChanged(kernel, this, old_state);
 }
 
-void Thread::SetSchedulingStatus(ThreadState new_status) {
-    const auto old_state = GetRawState();
-    thread_state = (thread_state & ThreadState::HighMask) | new_status;
-    KScheduler::OnThreadStateChanged(kernel, this, old_state);
-}
-
-void Thread::SetCurrentPriority(u32 new_priority) {
-    const u32 old_priority = std::exchange(current_priority, new_priority);
-    KScheduler::OnThreadPriorityChanged(kernel, this, kernel.CurrentScheduler()->GetCurrentThread(),
-                                        old_priority);
-}
-
 ResultCode Thread::SetCoreAndAffinityMask(s32 new_core, u64 new_affinity_mask) {
     KScopedSchedulerLock lock(kernel);
     const auto HighestSetCore = [](u64 mask, u32 max_cores) {
diff --git a/src/core/hle/kernel/thread.h b/src/core/hle/kernel/thread.h
index 06dd2ef2d..820ea524f 100644
--- a/src/core/hle/kernel/thread.h
+++ b/src/core/hle/kernel/thread.h
@@ -6,16 +6,21 @@
 
 #include <array>
 #include <functional>
+#include <span>
 #include <string>
 #include <utility>
 #include <vector>
 
+#include <boost/intrusive/list.hpp>
+
 #include "common/common_types.h"
+#include "common/intrusive_red_black_tree.h"
 #include "common/spin_lock.h"
 #include "core/arm/arm_interface.h"
 #include "core/hle/kernel/k_affinity_mask.h"
 #include "core/hle/kernel/k_synchronization_object.h"
 #include "core/hle/kernel/object.h"
+#include "core/hle/kernel/svc_common.h"
 #include "core/hle/result.h"
 
 namespace Common {
@@ -89,8 +94,6 @@ enum class ThreadState : u16 {
     InitSuspended = (1 << (4 + SuspendShift)),
 
     SuspendFlagMask = ((1 << 5) - 1) << SuspendShift,
-
-    HighMask = 0xfff0,
 };
 DECLARE_ENUM_FLAG_OPERATORS(ThreadState);
 
@@ -111,7 +114,10 @@ enum class ThreadSchedFlags : u32 {
     KernelInitPauseFlag = 1 << 8,
 };
 
-class Thread final : public KSynchronizationObject {
+class Thread final : public KSynchronizationObject, public boost::intrusive::list_base_hook<> {
+    friend class KScheduler;
+    friend class Process;
+
 public:
     explicit Thread(KernelCore& kernel);
     ~Thread() override;
@@ -180,49 +186,46 @@ public:
      * Gets the thread's current priority
      * @return The current thread's priority
      */
-    u32 GetPriority() const {
+    [[nodiscard]] s32 GetPriority() const {
         return current_priority;
     }
 
     /**
+     * Sets the thread's current priority.
+     * @param priority The new priority.
+     */
+    void SetPriority(s32 priority) {
+        current_priority = priority;
+    }
+
+    /**
      * Gets the thread's nominal priority.
      * @return The current thread's nominal priority.
      */
-    u32 GetNominalPriority() const {
-        return nominal_priority;
+    [[nodiscard]] s32 GetBasePriority() const {
+        return base_priority;
     }
 
     /**
-     * Sets the thread's current priority
-     * @param priority The new priority
+     * Sets the thread's nominal priority.
+     * @param priority The new priority.
      */
-    void SetPriority(u32 priority);
-
-    /// Adds a thread to the list of threads that are waiting for a lock held by this thread.
-    void AddMutexWaiter(std::shared_ptr<Thread> thread);
-
-    /// Removes a thread from the list of threads that are waiting for a lock held by this thread.
-    void RemoveMutexWaiter(std::shared_ptr<Thread> thread);
-
-    /// Recalculates the current priority taking into account priority inheritance.
-    void UpdatePriority();
+    void SetBasePriority(u32 priority);
 
     /// Changes the core that the thread is running or scheduled to run on.
-    ResultCode SetCoreAndAffinityMask(s32 new_core, u64 new_affinity_mask);
+    [[nodiscard]] ResultCode SetCoreAndAffinityMask(s32 new_core, u64 new_affinity_mask);
 
     /**
      * Gets the thread's thread ID
      * @return The thread's ID
      */
-    u64 GetThreadID() const {
+    [[nodiscard]] u64 GetThreadID() const {
         return thread_id;
     }
 
     /// Resumes a thread from waiting
     void Wakeup();
 
-    void OnWakeUp();
-
     ResultCode Start();
 
     virtual bool IsSignaled() const override;
@@ -242,7 +245,7 @@ public:
     }
 
     ResultCode GetWaitResult(KSynchronizationObject** out) const {
-        *out = this->signaling_object;
+        *out = signaling_object;
         return signaling_result;
     }
 
@@ -328,18 +331,14 @@ public:
         return thread_state;
     }
 
-    void SetState(ThreadState new_state);
-
-    void SetWaitingCondVar(bool value) {
-        is_waiting_on_condvar = value;
-    }
+    void SetState(ThreadState state);
 
     s64 GetLastScheduledTick() const {
-        return this->last_scheduled_tick;
+        return last_scheduled_tick;
     }
 
     void SetLastScheduledTick(s64 tick) {
-        this->last_scheduled_tick = tick;
+        last_scheduled_tick = tick;
     }
 
     u64 GetTotalCPUTimeTicks() const {
@@ -379,55 +378,13 @@ public:
     }
 
     Thread* GetLockOwner() const {
-        return lock_owner.get();
-    }
-
-    void SetLockOwner(std::shared_ptr<Thread> owner) {
-        lock_owner = std::move(owner);
-    }
-
-    VAddr GetCondVarWaitAddress() const {
-        return condvar_wait_address;
-    }
-
-    void SetCondVarWaitAddress(VAddr address) {
-        condvar_wait_address = address;
-    }
-
-    VAddr GetMutexWaitAddress() const {
-        return mutex_wait_address;
+        return lock_owner;
     }
 
-    void SetMutexWaitAddress(VAddr address) {
-        mutex_wait_address = address;
+    void SetLockOwner(Thread* owner) {
+        lock_owner = owner;
     }
 
-    Handle GetWaitHandle() const {
-        return wait_handle;
-    }
-
-    void SetWaitHandle(Handle handle) {
-        wait_handle = handle;
-    }
-
-    VAddr GetArbiterWaitAddress() const {
-        return arb_wait_address;
-    }
-
-    void SetArbiterWaitAddress(VAddr address) {
-        arb_wait_address = address;
-    }
-
-    void SetHLETimeEvent(Handle time_event) {
-        hle_time_event = time_event;
-    }
-
-    Handle GetHLETimeEvent() const {
-        return hle_time_event;
-    }
-
-    bool InvokeHLECallback(std::shared_ptr<Thread> thread);
-
     u32 GetIdealCore() const {
         return ideal_core;
     }
@@ -442,11 +399,11 @@ public:
     ResultCode Sleep(s64 nanoseconds);
 
     s64 GetYieldScheduleCount() const {
-        return this->schedule_count;
+        return schedule_count;
     }
 
     void SetYieldScheduleCount(s64 count) {
-        this->schedule_count = count;
+        schedule_count = count;
     }
 
     bool IsRunning() const {
@@ -469,14 +426,6 @@ public:
         return global_handle;
     }
 
-    bool IsWaitingForArbitration() const {
-        return waiting_for_arbitration;
-    }
-
-    void WaitForArbitration(bool set) {
-        waiting_for_arbitration = set;
-    }
-
     bool IsCancellable() const {
         return is_cancellable;
     }
@@ -490,7 +439,7 @@ public:
     }
 
     bool IsTerminationRequested() const {
-        return will_be_terminated || GetState() == ThreadState::Terminated;
+        return will_be_terminated || GetRawState() == ThreadState::Terminated;
     }
 
     bool IsPaused() const {
@@ -522,21 +471,21 @@ public:
         constexpr QueueEntry() = default;
 
         constexpr void Initialize() {
-            this->prev = nullptr;
-            this->next = nullptr;
+            prev = nullptr;
+            next = nullptr;
         }
 
         constexpr Thread* GetPrev() const {
-            return this->prev;
+            return prev;
         }
         constexpr Thread* GetNext() const {
-            return this->next;
+            return next;
         }
         constexpr void SetPrev(Thread* thread) {
-            this->prev = thread;
+            prev = thread;
         }
         constexpr void SetNext(Thread* thread) {
-            this->next = thread;
+            next = thread;
         }
 
     private:
@@ -545,11 +494,11 @@ public:
     };
 
     QueueEntry& GetPriorityQueueEntry(s32 core) {
-        return this->per_core_priority_queue_entry[core];
+        return per_core_priority_queue_entry[core];
     }
 
     const QueueEntry& GetPriorityQueueEntry(s32 core) const {
-        return this->per_core_priority_queue_entry[core];
+        return per_core_priority_queue_entry[core];
     }
 
     s32 GetDisableDispatchCount() const {
@@ -566,27 +515,155 @@ public:
         disable_count--;
     }
 
-    void SetWaitObjectsForDebugging(KSynchronizationObject** objects, s32 num_objects) {
+    void SetWaitObjectsForDebugging(const std::span<KSynchronizationObject*>& objects) {
         wait_objects_for_debugging.clear();
-        wait_objects_for_debugging.reserve(num_objects);
-        for (auto i = 0; i < num_objects; ++i) {
-            wait_objects_for_debugging.emplace_back(objects[i]);
+        wait_objects_for_debugging.reserve(objects.size());
+        for (const auto& object : objects) {
+            wait_objects_for_debugging.emplace_back(object);
         }
     }
 
-    const std::vector<KSynchronizationObject*>& GetWaitObjectsForDebugging() const {
+    [[nodiscard]] const std::vector<KSynchronizationObject*>& GetWaitObjectsForDebugging() const {
         return wait_objects_for_debugging;
     }
 
+    void SetMutexWaitAddressForDebugging(VAddr address) {
+        mutex_wait_address_for_debugging = address;
+    }
+
+    [[nodiscard]] VAddr GetMutexWaitAddressForDebugging() const {
+        return mutex_wait_address_for_debugging;
+    }
+
+    void AddWaiter(Thread* thread);
+
+    void RemoveWaiter(Thread* thread);
+
+    [[nodiscard]] Thread* RemoveWaiterByKey(s32* out_num_waiters, VAddr key);
+
+    [[nodiscard]] VAddr GetAddressKey() const {
+        return address_key;
+    }
+
+    [[nodiscard]] u32 GetAddressKeyValue() const {
+        return address_key_value;
+    }
+
+    void SetAddressKey(VAddr key) {
+        address_key = key;
+    }
+
+    void SetAddressKey(VAddr key, u32 val) {
+        address_key = key;
+        address_key_value = val;
+    }
+
 private:
-    friend class GlobalSchedulerContext;
-    friend class KScheduler;
-    friend class Process;
+    static constexpr size_t PriorityInheritanceCountMax = 10;
+    union SyncObjectBuffer {
+        std::array<KSynchronizationObject*, Svc::ArgumentHandleCountMax> sync_objects{};
+        std::array<Handle,
+                   Svc::ArgumentHandleCountMax*(sizeof(KSynchronizationObject*) / sizeof(Handle))>
+            handles;
+        constexpr SyncObjectBuffer() {}
+    };
+    static_assert(sizeof(SyncObjectBuffer::sync_objects) == sizeof(SyncObjectBuffer::handles));
+
+    struct ConditionVariableComparator {
+        struct LightCompareType {
+            u64 cv_key{};
+            s32 priority{};
+
+            [[nodiscard]] constexpr u64 GetConditionVariableKey() const {
+                return cv_key;
+            }
+
+            [[nodiscard]] constexpr s32 GetPriority() const {
+                return priority;
+            }
+        };
+
+        template <typename T>
+        requires(
+            std::same_as<T, Thread> ||
+            std::same_as<T, LightCompareType>) static constexpr int Compare(const T& lhs,
+                                                                            const Thread& rhs) {
+            const uintptr_t l_key = lhs.GetConditionVariableKey();
+            const uintptr_t r_key = rhs.GetConditionVariableKey();
+
+            if (l_key < r_key) {
+                // Sort first by key
+                return -1;
+            } else if (l_key == r_key && lhs.GetPriority() < rhs.GetPriority()) {
+                // And then by priority.
+                return -1;
+            } else {
+                return 1;
+            }
+        }
+    };
+
+    Common::IntrusiveRedBlackTreeNode condvar_arbiter_tree_node{};
+
+    using ConditionVariableThreadTreeTraits =
+        Common::IntrusiveRedBlackTreeMemberTraitsDeferredAssert<&Thread::condvar_arbiter_tree_node>;
+    using ConditionVariableThreadTree =
+        ConditionVariableThreadTreeTraits::TreeType<ConditionVariableComparator>;
+
+public:
+    using ConditionVariableThreadTreeType = ConditionVariableThreadTree;
 
-    void SetSchedulingStatus(ThreadState new_status);
+    [[nodiscard]] uintptr_t GetConditionVariableKey() const {
+        return condvar_key;
+    }
+
+    [[nodiscard]] uintptr_t GetAddressArbiterKey() const {
+        return condvar_key;
+    }
+
+    void SetConditionVariable(ConditionVariableThreadTree* tree, VAddr address, uintptr_t cv_key,
+                              u32 value) {
+        condvar_tree = tree;
+        condvar_key = cv_key;
+        address_key = address;
+        address_key_value = value;
+    }
+
+    void ClearConditionVariable() {
+        condvar_tree = nullptr;
+    }
+
+    [[nodiscard]] bool IsWaitingForConditionVariable() const {
+        return condvar_tree != nullptr;
+    }
+
+    void SetAddressArbiter(ConditionVariableThreadTree* tree, uintptr_t address) {
+        condvar_tree = tree;
+        condvar_key = address;
+    }
+
+    void ClearAddressArbiter() {
+        condvar_tree = nullptr;
+    }
+
+    [[nodiscard]] bool IsWaitingForAddressArbiter() const {
+        return condvar_tree != nullptr;
+    }
+
+    [[nodiscard]] ConditionVariableThreadTree* GetConditionVariableTree() const {
+        return condvar_tree;
+    }
+
+    [[nodiscard]] bool HasWaiters() const {
+        return !waiter_list.empty();
+    }
+
+private:
     void AddSchedulingFlag(ThreadSchedFlags flag);
     void RemoveSchedulingFlag(ThreadSchedFlags flag);
-    void SetCurrentPriority(u32 new_priority);
+    void AddWaiterImpl(Thread* thread);
+    void RemoveWaiterImpl(Thread* thread);
+    static void RestorePriority(KernelCore& kernel, Thread* thread);
 
     Common::SpinLock context_guard{};
     ThreadContext32 context_32{};
@@ -606,11 +683,11 @@ private:
     /// Nominal thread priority, as set by the emulated application.
     /// The nominal priority is the thread priority without priority
     /// inheritance taken into account.
-    u32 nominal_priority = 0;
+    s32 base_priority{};
 
     /// Current thread priority. This may change over the course of the
     /// thread's lifetime in order to facilitate priority inheritance.
-    u32 current_priority = 0;
+    s32 current_priority{};
 
     u64 total_cpu_time_ticks = 0; ///< Total CPU running ticks.
     s64 schedule_count{};
@@ -628,6 +705,9 @@ private:
     /// passed to WaitSynchronization. This is used for debugging only.
     std::vector<KSynchronizationObject*> wait_objects_for_debugging;
 
+    /// The current mutex wait address. This is used for debugging only.
+    VAddr mutex_wait_address_for_debugging{};
+
     KSynchronizationObject* signaling_object;
     ResultCode signaling_result{RESULT_SUCCESS};
 
@@ -635,25 +715,11 @@ private:
     MutexWaitingThreads wait_mutex_threads;
 
     /// Thread that owns the lock that this thread is waiting for.
-    std::shared_ptr<Thread> lock_owner;
-
-    /// If waiting on a ConditionVariable, this is the ConditionVariable address
-    VAddr condvar_wait_address = 0;
-    bool is_waiting_on_condvar{};
-    /// If waiting on a Mutex, this is the mutex address
-    VAddr mutex_wait_address = 0;
-    /// The handle used to wait for the mutex.
-    Handle wait_handle = 0;
-
-    /// If waiting for an AddressArbiter, this is the address being waited on.
-    VAddr arb_wait_address{0};
-    bool waiting_for_arbitration{};
+    Thread* lock_owner{};
 
     /// Handle used as userdata to reference this object when inserting into the CoreTiming queue.
     Handle global_handle = 0;
 
-    Handle hle_time_event;
-
     KScheduler* scheduler = nullptr;
 
     std::array<QueueEntry, Core::Hardware::NUM_CPU_CORES> per_core_priority_queue_entry{};
@@ -679,6 +745,16 @@ private:
 
     bool signaled{};
 
+    ConditionVariableThreadTree* condvar_tree{};
+    uintptr_t condvar_key{};
+    VAddr address_key{};
+    u32 address_key_value{};
+    s32 num_kernel_waiters{};
+
+    using WaiterList = boost::intrusive::list<Thread>;
+    WaiterList waiter_list{};
+    WaiterList pinned_waiter_list{};
+
     std::string name;
 };
 
diff --git a/src/core/hle/kernel/time_manager.cpp b/src/core/hle/kernel/time_manager.cpp
index b58a76dba..832edd629 100644
--- a/src/core/hle/kernel/time_manager.cpp
+++ b/src/core/hle/kernel/time_manager.cpp
@@ -18,12 +18,10 @@ TimeManager::TimeManager(Core::System& system_) : system{system_} {
     time_manager_event_type = Core::Timing::CreateEvent(
         "Kernel::TimeManagerCallback",
         [this](std::uintptr_t thread_handle, std::chrono::nanoseconds) {
-            const KScopedSchedulerLock lock(system.Kernel());
-            const auto proper_handle = static_cast<Handle>(thread_handle);
-
             std::shared_ptr<Thread> thread;
             {
                 std::lock_guard lock{mutex};
+                const auto proper_handle = static_cast<Handle>(thread_handle);
                 if (cancelled_events[proper_handle]) {
                     return;
                 }
@@ -32,7 +30,7 @@ TimeManager::TimeManager(Core::System& system_) : system{system_} {
 
             if (thread) {
                 // Thread can be null if process has exited
-                thread->OnWakeUp();
+                thread->Wakeup();
             }
         });
 }
diff --git a/src/yuzu/debugger/wait_tree.cpp b/src/yuzu/debugger/wait_tree.cpp
index acf6b7ab5..3ded85720 100644
--- a/src/yuzu/debugger/wait_tree.cpp
+++ b/src/yuzu/debugger/wait_tree.cpp
@@ -15,9 +15,9 @@
 #include "core/hle/kernel/handle_table.h"
 #include "core/hle/kernel/k_scheduler.h"
 #include "core/hle/kernel/k_synchronization_object.h"
-#include "core/hle/kernel/mutex.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/kernel/readable_event.h"
+#include "core/hle/kernel/svc_common.h"
 #include "core/hle/kernel/thread.h"
 #include "core/memory.h"
 
@@ -116,7 +116,7 @@ QString WaitTreeText::GetText() const {
 WaitTreeMutexInfo::WaitTreeMutexInfo(VAddr mutex_address, const Kernel::HandleTable& handle_table)
     : mutex_address(mutex_address) {
     mutex_value = Core::System::GetInstance().Memory().Read32(mutex_address);
-    owner_handle = static_cast<Kernel::Handle>(mutex_value & Kernel::Mutex::MutexOwnerMask);
+    owner_handle = static_cast<Kernel::Handle>(mutex_value & Kernel::Svc::HandleWaitMask);
     owner = handle_table.Get<Kernel::Thread>(owner_handle);
 }
 
@@ -127,7 +127,7 @@ QString WaitTreeMutexInfo::GetText() const {
 }
 
 std::vector<std::unique_ptr<WaitTreeItem>> WaitTreeMutexInfo::GetChildren() const {
-    const bool has_waiters = (mutex_value & Kernel::Mutex::MutexHasWaitersFlag) != 0;
+    const bool has_waiters = (mutex_value & Kernel::Svc::HandleWaitMask) != 0;
 
     std::vector<std::unique_ptr<WaitTreeItem>> list;
     list.push_back(std::make_unique<WaitTreeText>(tr("has waiters: %1").arg(has_waiters)));
@@ -324,11 +324,11 @@ std::vector<std::unique_ptr<WaitTreeItem>> WaitTreeThread::GetChildren() const {
     list.push_back(std::make_unique<WaitTreeText>(tr("thread id = %1").arg(thread.GetThreadID())));
     list.push_back(std::make_unique<WaitTreeText>(tr("priority = %1(current) / %2(normal)")
                                                       .arg(thread.GetPriority())
-                                                      .arg(thread.GetNominalPriority())));
+                                                      .arg(thread.GetBasePriority())));
     list.push_back(std::make_unique<WaitTreeText>(
         tr("last running ticks = %1").arg(thread.GetLastScheduledTick())));
 
-    const VAddr mutex_wait_address = thread.GetMutexWaitAddress();
+    const VAddr mutex_wait_address = thread.GetMutexWaitAddressForDebugging();
     if (mutex_wait_address != 0) {
         const auto& handle_table = thread.GetOwnerProcess()->GetHandleTable();
         list.push_back(std::make_unique<WaitTreeMutexInfo>(mutex_wait_address, handle_table));