38 files changed, 470 insertions, 168 deletions

diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
index 1cfd3bbc9..8bd7e5f72 100644
--- a/src/CMakeLists.txt
+++ b/src/CMakeLists.txt
@@ -27,6 +27,7 @@ if (MSVC)
     # /Zo                 - Enhanced debug info for optimized builds
     # /permissive-        - Enables stricter C++ standards conformance checks
     # /EHsc               - C++-only exception handling semantics
+    # /utf-8              - Set source and execution character sets to UTF-8
     # /volatile:iso       - Use strict standards-compliant volatile semantics.
     # /Zc:externConstexpr - Allow extern constexpr variables to have external linkage, like the standard mandates
     # /Zc:inline          - Let codegen omit inline functions in object files
@@ -38,6 +39,7 @@ if (MSVC)
         /permissive-
         /EHsc
         /std:c++latest
+        /utf-8
         /volatile:iso
         /Zc:externConstexpr
         /Zc:inline
diff --git a/src/common/fiber.cpp b/src/common/fiber.cpp
index b8e98b12a..39532ff58 100644
--- a/src/common/fiber.cpp
+++ b/src/common/fiber.cpp
@@ -24,7 +24,7 @@ struct Fiber::FiberImpl {
     std::function<void(void*)> rewind_point;
     void* rewind_parameter{};
     void* start_parameter{};
-    Fiber* previous_fiber;
+    std::shared_ptr<Fiber> previous_fiber;
     bool is_thread_fiber{};
     bool released{};
 
@@ -47,7 +47,7 @@ void Fiber::Start(boost::context::detail::transfer_t& transfer) {
     ASSERT(impl->previous_fiber != nullptr);
     impl->previous_fiber->impl->context = transfer.fctx;
     impl->previous_fiber->impl->guard.unlock();
-    impl->previous_fiber = nullptr;
+    impl->previous_fiber.reset();
     impl->entry_point(impl->start_parameter);
     UNREACHABLE();
 }
@@ -116,20 +116,23 @@ void Fiber::Rewind() {
     boost::context::detail::jump_fcontext(impl->rewind_context, this);
 }
 
-void Fiber::YieldTo(Fiber* from, Fiber* to) {
-    ASSERT_MSG(from != nullptr, "Yielding fiber is null!");
-    ASSERT_MSG(to != nullptr, "Next fiber is null!");
-    to->impl->guard.lock();
-    to->impl->previous_fiber = from;
-    auto transfer = boost::context::detail::jump_fcontext(to->impl->context, to);
-    ASSERT(from->impl->previous_fiber != nullptr);
-    from->impl->previous_fiber->impl->context = transfer.fctx;
-    from->impl->previous_fiber->impl->guard.unlock();
-    from->impl->previous_fiber = nullptr;
+void Fiber::YieldTo(std::weak_ptr<Fiber> weak_from, Fiber& to) {
+    to.impl->guard.lock();
+    to.impl->previous_fiber = weak_from.lock();
+
+    auto transfer = boost::context::detail::jump_fcontext(to.impl->context, &to);
+
+    // "from" might no longer be valid if the thread was killed
+    if (auto from = weak_from.lock()) {
+        ASSERT(from->impl->previous_fiber != nullptr);
+        from->impl->previous_fiber->impl->context = transfer.fctx;
+        from->impl->previous_fiber->impl->guard.unlock();
+        from->impl->previous_fiber.reset();
+    }
 }
 
-std::unique_ptr<Fiber> Fiber::ThreadToFiber() {
-    std::unique_ptr<Fiber> fiber = std::unique_ptr<Fiber>{new Fiber()};
+std::shared_ptr<Fiber> Fiber::ThreadToFiber() {
+    std::shared_ptr<Fiber> fiber = std::shared_ptr<Fiber>{new Fiber()};
     fiber->impl->guard.lock();
     fiber->impl->is_thread_fiber = true;
     return fiber;
diff --git a/src/common/fiber.h b/src/common/fiber.h
index 6924f7996..f2a8ff29a 100644
--- a/src/common/fiber.h
+++ b/src/common/fiber.h
@@ -41,8 +41,8 @@ public:
 
     /// Yields control from Fiber 'from' to Fiber 'to'
     /// Fiber 'from' must be the currently running fiber.
-    static void YieldTo(Fiber* from, Fiber* to);
-    [[nodiscard]] static std::unique_ptr<Fiber> ThreadToFiber();
+    static void YieldTo(std::weak_ptr<Fiber> weak_from, Fiber& to);
+    [[nodiscard]] static std::shared_ptr<Fiber> ThreadToFiber();
 
     void SetRewindPoint(std::function<void(void*)>&& rewind_func, void* rewind_param);
 
diff --git a/src/core/cpu_manager.cpp b/src/core/cpu_manager.cpp
index c35438c6f..bdb374792 100644
--- a/src/core/cpu_manager.cpp
+++ b/src/core/cpu_manager.cpp
@@ -111,7 +111,7 @@ void CpuManager::MultiCoreRunGuestThread() {
     auto& kernel = system.Kernel();
     kernel.CurrentScheduler()->OnThreadStart();
     auto* thread = kernel.CurrentScheduler()->GetCurrentThread();
-    auto host_context = thread->GetHostContext();
+    auto& host_context = thread->GetHostContext();
     host_context->SetRewindPoint(GuestRewindFunction, this);
     MultiCoreRunGuestLoop();
 }
@@ -148,8 +148,7 @@ void CpuManager::MultiCoreRunSuspendThread() {
         auto core = kernel.GetCurrentHostThreadID();
         auto& scheduler = *kernel.CurrentScheduler();
         Kernel::KThread* current_thread = scheduler.GetCurrentThread();
-        Common::Fiber::YieldTo(current_thread->GetHostContext(),
-                               core_data[core].host_context.get());
+        Common::Fiber::YieldTo(current_thread->GetHostContext(), *core_data[core].host_context);
         ASSERT(scheduler.ContextSwitchPending());
         ASSERT(core == kernel.GetCurrentHostThreadID());
         scheduler.RescheduleCurrentCore();
@@ -202,7 +201,7 @@ void CpuManager::SingleCoreRunGuestThread() {
     auto& kernel = system.Kernel();
     kernel.CurrentScheduler()->OnThreadStart();
     auto* thread = kernel.CurrentScheduler()->GetCurrentThread();
-    auto host_context = thread->GetHostContext();
+    auto& host_context = thread->GetHostContext();
     host_context->SetRewindPoint(GuestRewindFunction, this);
     SingleCoreRunGuestLoop();
 }
@@ -246,7 +245,7 @@ void CpuManager::SingleCoreRunSuspendThread() {
         auto core = kernel.GetCurrentHostThreadID();
         auto& scheduler = *kernel.CurrentScheduler();
         Kernel::KThread* current_thread = scheduler.GetCurrentThread();
-        Common::Fiber::YieldTo(current_thread->GetHostContext(), core_data[0].host_context.get());
+        Common::Fiber::YieldTo(current_thread->GetHostContext(), *core_data[0].host_context);
         ASSERT(scheduler.ContextSwitchPending());
         ASSERT(core == kernel.GetCurrentHostThreadID());
         scheduler.RescheduleCurrentCore();
@@ -272,7 +271,7 @@ void CpuManager::PreemptSingleCore(bool from_running_enviroment) {
         scheduler.Unload(scheduler.GetCurrentThread());
 
         auto& next_scheduler = kernel.Scheduler(current_core);
-        Common::Fiber::YieldTo(current_thread->GetHostContext(), next_scheduler.ControlContext());
+        Common::Fiber::YieldTo(current_thread->GetHostContext(), *next_scheduler.ControlContext());
     }
 
     // May have changed scheduler
@@ -364,7 +363,7 @@ void CpuManager::RunThread(std::size_t core) {
 
         auto current_thread = system.Kernel().CurrentScheduler()->GetCurrentThread();
         data.is_running = true;
-        Common::Fiber::YieldTo(data.host_context.get(), current_thread->GetHostContext());
+        Common::Fiber::YieldTo(data.host_context, *current_thread->GetHostContext());
         data.is_running = false;
         data.is_paused = true;
         data.exit_barrier->Wait();
diff --git a/src/core/cpu_manager.h b/src/core/cpu_manager.h
index 5ea149f1f..17420c941 100644
--- a/src/core/cpu_manager.h
+++ b/src/core/cpu_manager.h
@@ -83,7 +83,7 @@ private:
     void RunThread(std::size_t core);
 
     struct CoreData {
-        std::unique_ptr<Common::Fiber> host_context;
+        std::shared_ptr<Common::Fiber> host_context;
         std::unique_ptr<Common::Event> enter_barrier;
         std::unique_ptr<Common::Event> exit_barrier;
         std::atomic<bool> is_running;
diff --git a/src/core/hle/kernel/k_scheduler.cpp b/src/core/hle/kernel/k_scheduler.cpp
index 465036f3d..e7de48476 100644
--- a/src/core/hle/kernel/k_scheduler.cpp
+++ b/src/core/hle/kernel/k_scheduler.cpp
@@ -608,7 +608,7 @@ void KScheduler::YieldToAnyThread(KernelCore& kernel) {
 }
 
 KScheduler::KScheduler(Core::System& system, s32 core_id) : system(system), core_id(core_id) {
-    switch_fiber = std::make_unique<Common::Fiber>(OnSwitch, this);
+    switch_fiber = std::make_shared<Common::Fiber>(OnSwitch, this);
     state.needs_scheduling.store(true);
     state.interrupt_task_thread_runnable = false;
     state.should_count_idle = false;
@@ -726,15 +726,15 @@ void KScheduler::ScheduleImpl() {
     // Save context for previous thread
     Unload(previous_thread);
 
-    Common::Fiber* old_context;
+    std::shared_ptr<Common::Fiber>* old_context;
     if (previous_thread != nullptr) {
-        old_context = previous_thread->GetHostContext();
+        old_context = &previous_thread->GetHostContext();
     } else {
-        old_context = idle_thread->GetHostContext();
+        old_context = &idle_thread->GetHostContext();
     }
     guard.unlock();
 
-    Common::Fiber::YieldTo(old_context, switch_fiber.get());
+    Common::Fiber::YieldTo(*old_context, *switch_fiber);
     /// When a thread wakes up, the scheduler may have changed to other in another core.
     auto& next_scheduler = *system.Kernel().CurrentScheduler();
     next_scheduler.SwitchContextStep2();
@@ -769,13 +769,8 @@ void KScheduler::SwitchToCurrent() {
                     break;
                 }
             }
-            Common::Fiber* next_context;
-            if (next_thread != nullptr) {
-                next_context = next_thread->GetHostContext();
-            } else {
-                next_context = idle_thread->GetHostContext();
-            }
-            Common::Fiber::YieldTo(switch_fiber.get(), next_context);
+            auto thread = next_thread ? next_thread : idle_thread;
+            Common::Fiber::YieldTo(switch_fiber, *thread->GetHostContext());
         } while (!is_switch_pending());
     }
 }
@@ -800,9 +795,9 @@ void KScheduler::Initialize() {
     std::string name = "Idle Thread Id:" + std::to_string(core_id);
     std::function<void(void*)> init_func = Core::CpuManager::GetIdleThreadStartFunc();
     void* init_func_parameter = system.GetCpuManager().GetStartFuncParamater();
-    auto thread_res = KThread::Create(system, ThreadType::Main, name, 0,
-                                      KThread::IdleThreadPriority, 0, static_cast<u32>(core_id), 0,
-                                      nullptr, std::move(init_func), init_func_parameter);
+    auto thread_res = KThread::CreateThread(
+        system, ThreadType::Main, name, 0, KThread::IdleThreadPriority, 0,
+        static_cast<u32>(core_id), 0, nullptr, std::move(init_func), init_func_parameter);
     idle_thread = thread_res.Unwrap().get();
 }
 
diff --git a/src/core/hle/kernel/k_scheduler.h b/src/core/hle/kernel/k_scheduler.h
index a4285c595..f595b9a5c 100644
--- a/src/core/hle/kernel/k_scheduler.h
+++ b/src/core/hle/kernel/k_scheduler.h
@@ -68,12 +68,12 @@ public:
 
     void OnThreadStart();
 
-    [[nodiscard]] Common::Fiber* ControlContext() {
-        return switch_fiber.get();
+    [[nodiscard]] std::shared_ptr<Common::Fiber>& ControlContext() {
+        return switch_fiber;
     }
 
-    [[nodiscard]] const Common::Fiber* ControlContext() const {
-        return switch_fiber.get();
+    [[nodiscard]] const std::shared_ptr<Common::Fiber>& ControlContext() const {
+        return switch_fiber;
     }
 
     [[nodiscard]] u64 UpdateHighestPriorityThread(KThread* highest_thread);
@@ -178,7 +178,7 @@ private:
 
     KThread* idle_thread;
 
-    std::unique_ptr<Common::Fiber> switch_fiber{};
+    std::shared_ptr<Common::Fiber> switch_fiber{};
 
     struct SchedulingState {
         std::atomic<bool> needs_scheduling;
diff --git a/src/core/hle/kernel/k_thread.cpp b/src/core/hle/kernel/k_thread.cpp
index f49e31b72..e0f53287c 100644
--- a/src/core/hle/kernel/k_thread.cpp
+++ b/src/core/hle/kernel/k_thread.cpp
@@ -991,22 +991,15 @@ void KThread::SetState(ThreadState state) {
     }
 }
 
-ResultVal<std::shared_ptr<KThread>> KThread::Create(Core::System& system, ThreadType type_flags,
-                                                    std::string name, VAddr entry_point,
-                                                    u32 priority, u64 arg, s32 processor_id,
-                                                    VAddr stack_top, Process* owner_process) {
-    std::function<void(void*)> init_func = Core::CpuManager::GetGuestThreadStartFunc();
-    void* init_func_parameter = system.GetCpuManager().GetStartFuncParamater();
-    return Create(system, type_flags, name, entry_point, priority, arg, processor_id, stack_top,
-                  owner_process, std::move(init_func), init_func_parameter);
+std::shared_ptr<Common::Fiber>& KThread::GetHostContext() {
+    return host_context;
 }
 
-ResultVal<std::shared_ptr<KThread>> KThread::Create(Core::System& system, ThreadType type_flags,
-                                                    std::string name, VAddr entry_point,
-                                                    u32 priority, u64 arg, s32 processor_id,
-                                                    VAddr stack_top, Process* owner_process,
-                                                    std::function<void(void*)>&& thread_start_func,
-                                                    void* thread_start_parameter) {
+ResultVal<std::shared_ptr<KThread>> KThread::CreateThread(Core::System& system,
+                                                          ThreadType type_flags, std::string name,
+                                                          VAddr entry_point, u32 priority, u64 arg,
+                                                          s32 processor_id, VAddr stack_top,
+                                                          Process* owner_process) {
     auto& kernel = system.Kernel();
 
     std::shared_ptr<KThread> thread = std::make_shared<KThread>(kernel);
@@ -1023,12 +1016,35 @@ ResultVal<std::shared_ptr<KThread>> KThread::Create(Core::System& system, Thread
     auto& scheduler = kernel.GlobalSchedulerContext();
     scheduler.AddThread(thread);
 
-    thread->host_context =
-        std::make_unique<Common::Fiber>(std::move(thread_start_func), thread_start_parameter);
-
     return MakeResult<std::shared_ptr<KThread>>(std::move(thread));
 }
 
+ResultVal<std::shared_ptr<KThread>> KThread::CreateThread(
+    Core::System& system, ThreadType type_flags, std::string name, VAddr entry_point, u32 priority,
+    u64 arg, s32 processor_id, VAddr stack_top, Process* owner_process,
+    std::function<void(void*)>&& thread_start_func, void* thread_start_parameter) {
+    auto thread_result = CreateThread(system, type_flags, name, entry_point, priority, arg,
+                                      processor_id, stack_top, owner_process);
+
+    if (thread_result.Succeeded()) {
+        (*thread_result)->host_context =
+            std::make_shared<Common::Fiber>(std::move(thread_start_func), thread_start_parameter);
+    }
+
+    return thread_result;
+}
+
+ResultVal<std::shared_ptr<KThread>> KThread::CreateUserThread(
+    Core::System& system, ThreadType type_flags, std::string name, VAddr entry_point, u32 priority,
+    u64 arg, s32 processor_id, VAddr stack_top, Process* owner_process) {
+    std::function<void(void*)> init_func = Core::CpuManager::GetGuestThreadStartFunc();
+
+    void* init_func_parameter = system.GetCpuManager().GetStartFuncParamater();
+
+    return CreateThread(system, type_flags, name, entry_point, priority, arg, processor_id,
+                        stack_top, owner_process, std::move(init_func), init_func_parameter);
+}
+
 KThread* GetCurrentThreadPointer(KernelCore& kernel) {
     return kernel.GetCurrentEmuThread();
 }
diff --git a/src/core/hle/kernel/k_thread.h b/src/core/hle/kernel/k_thread.h
index a2893d939..1c19b23dc 100644
--- a/src/core/hle/kernel/k_thread.h
+++ b/src/core/hle/kernel/k_thread.h
@@ -116,7 +116,7 @@ public:
     using WaiterList = boost::intrusive::list<KThread>;
 
     /**
-     * Creates and returns a new thread. The new thread is immediately scheduled
+     * Creates and returns a new thread.
      * @param system The instance of the whole system
      * @param name The friendly name desired for the thread
      * @param entry_point The address at which the thread should start execution
@@ -127,12 +127,12 @@ public:
      * @param owner_process The parent process for the thread, if null, it's a kernel thread
      * @return A shared pointer to the newly created thread
      */
-    [[nodiscard]] static ResultVal<std::shared_ptr<KThread>> Create(
+    [[nodiscard]] static ResultVal<std::shared_ptr<KThread>> CreateThread(
         Core::System& system, ThreadType type_flags, std::string name, VAddr entry_point,
         u32 priority, u64 arg, s32 processor_id, VAddr stack_top, Process* owner_process);
 
     /**
-     * Creates and returns a new thread. The new thread is immediately scheduled
+     * Creates and returns a new thread, with a specified entry point.
      * @param system The instance of the whole system
      * @param name The friendly name desired for the thread
      * @param entry_point The address at which the thread should start execution
@@ -145,11 +145,27 @@ public:
      * @param thread_start_parameter The parameter which will passed to host context on init
      * @return A shared pointer to the newly created thread
      */
-    [[nodiscard]] static ResultVal<std::shared_ptr<KThread>> Create(
+    [[nodiscard]] static ResultVal<std::shared_ptr<KThread>> CreateThread(
         Core::System& system, ThreadType type_flags, std::string name, VAddr entry_point,
         u32 priority, u64 arg, s32 processor_id, VAddr stack_top, Process* owner_process,
         std::function<void(void*)>&& thread_start_func, void* thread_start_parameter);
 
+    /**
+     * Creates and returns a new thread for the emulated "user" process.
+     * @param system The instance of the whole system
+     * @param name The friendly name desired for the thread
+     * @param entry_point The address at which the thread should start execution
+     * @param priority The thread's priority
+     * @param arg User data to pass to the thread
+     * @param processor_id The ID(s) of the processors on which the thread is desired to be run
+     * @param stack_top The address of the thread's stack top
+     * @param owner_process The parent process for the thread, if null, it's a kernel thread
+     * @return A shared pointer to the newly created thread
+     */
+    [[nodiscard]] static ResultVal<std::shared_ptr<KThread>> CreateUserThread(
+        Core::System& system, ThreadType type_flags, std::string name, VAddr entry_point,
+        u32 priority, u64 arg, s32 processor_id, VAddr stack_top, Process* owner_process);
+
     [[nodiscard]] std::string GetName() const override {
         return name;
     }
@@ -293,13 +309,7 @@ public:
         return thread_context_64;
     }
 
-    [[nodiscard]] Common::Fiber* GetHostContext() {
-        return host_context.get();
-    }
-
-    [[nodiscard]] const Common::Fiber* GetHostContext() const {
-        return host_context.get();
-    }
+    [[nodiscard]] std::shared_ptr<Common::Fiber>& GetHostContext();
 
     [[nodiscard]] ThreadState GetState() const {
         return thread_state & ThreadState::Mask;
@@ -725,7 +735,7 @@ private:
     Common::SpinLock context_guard{};
 
     // For emulation
-    std::unique_ptr<Common::Fiber> host_context{};
+    std::shared_ptr<Common::Fiber> host_context{};
 
     // For debugging
     std::vector<KSynchronizationObject*> wait_objects_for_debugging;
diff --git a/src/core/hle/kernel/kernel.cpp b/src/core/hle/kernel/kernel.cpp
index 331cf3a60..780008b08 100644
--- a/src/core/hle/kernel/kernel.cpp
+++ b/src/core/hle/kernel/kernel.cpp
@@ -181,9 +181,9 @@ struct KernelCore::Impl {
             std::string name = "Suspend Thread Id:" + std::to_string(i);
             std::function<void(void*)> init_func = Core::CpuManager::GetSuspendThreadStartFunc();
             void* init_func_parameter = system.GetCpuManager().GetStartFuncParamater();
-            auto thread_res = KThread::Create(system, ThreadType::HighPriority, std::move(name), 0,
-                                              0, 0, static_cast<u32>(i), 0, nullptr,
-                                              std::move(init_func), init_func_parameter);
+            auto thread_res = KThread::CreateThread(
+                system, ThreadType::HighPriority, std::move(name), 0, 0, 0, static_cast<u32>(i), 0,
+                nullptr, std::move(init_func), init_func_parameter);
 
             suspend_threads[i] = std::move(thread_res).Unwrap();
         }
@@ -221,10 +221,9 @@ struct KernelCore::Impl {
     // Gets the dummy KThread for the caller, allocating a new one if this is the first time
     KThread* GetHostDummyThread() {
         const thread_local auto thread =
-            KThread::Create(
+            KThread::CreateThread(
                 system, ThreadType::Main, fmt::format("DummyThread:{}", GetHostThreadId()), 0,
-                KThread::DefaultThreadPriority, 0, static_cast<u32>(3), 0, nullptr,
-                []([[maybe_unused]] void* arg) { UNREACHABLE(); }, nullptr)
+                KThread::DefaultThreadPriority, 0, static_cast<u32>(3), 0, nullptr)
                 .Unwrap();
         return thread.get();
     }
diff --git a/src/core/hle/kernel/process.cpp b/src/core/hle/kernel/process.cpp
index 73b85d6f9..9d5956ead 100644
--- a/src/core/hle/kernel/process.cpp
+++ b/src/core/hle/kernel/process.cpp
@@ -40,8 +40,9 @@ namespace {
 void SetupMainThread(Core::System& system, Process& owner_process, u32 priority, VAddr stack_top) {
     const VAddr entry_point = owner_process.PageTable().GetCodeRegionStart();
     ASSERT(owner_process.GetResourceLimit()->Reserve(LimitableResource::Threads, 1));
-    auto thread_res = KThread::Create(system, ThreadType::User, "main", entry_point, priority, 0,
-                                      owner_process.GetIdealCoreId(), stack_top, &owner_process);
+    auto thread_res =
+        KThread::CreateUserThread(system, ThreadType::User, "main", entry_point, priority, 0,
+                                  owner_process.GetIdealCoreId(), stack_top, &owner_process);
 
     std::shared_ptr<KThread> thread = std::move(thread_res).Unwrap();
 
diff --git a/src/core/hle/kernel/svc.cpp b/src/core/hle/kernel/svc.cpp
index d04116115..326d3b9ec 100644
--- a/src/core/hle/kernel/svc.cpp
+++ b/src/core/hle/kernel/svc.cpp
@@ -1532,8 +1532,9 @@ static ResultCode CreateThread(Core::System& system, Handle* out_handle, VAddr e
     std::shared_ptr<KThread> thread;
     {
         KScopedLightLock lk{process.GetStateLock()};
-        CASCADE_RESULT(thread, KThread::Create(system, ThreadType::User, "", entry_point, priority,
-                                               arg, core_id, stack_bottom, &process));
+        CASCADE_RESULT(thread,
+                       KThread::CreateUserThread(system, ThreadType::User, "", entry_point,
+                                                 priority, arg, core_id, stack_bottom, &process));
     }
 
     const auto new_thread_handle = process.GetHandleTable().Create(thread);
@@ -2626,7 +2627,8 @@ void Call(Core::System& system, u32 immediate) {
     kernel.ExitSVCProfile();
 
     if (!thread->IsCallingSvc()) {
-        thread->GetHostContext()->Rewind();
+        auto* host_context = thread->GetHostContext().get();
+        host_context->Rewind();
     }
 
     system.EnterDynarmicProfile();
diff --git a/src/input_common/mouse/mouse_input.cpp b/src/input_common/mouse/mouse_input.cpp
index 6818d5eee..329e416c7 100644
--- a/src/input_common/mouse/mouse_input.cpp
+++ b/src/input_common/mouse/mouse_input.cpp
@@ -59,7 +59,7 @@ void Mouse::UpdateYuzuSettings() {
     });
 }
 
-void Mouse::PressButton(int x, int y, int button_) {
+void Mouse::PressButton(int x, int y, MouseButton button_) {
     const auto button_index = static_cast<std::size_t>(button_);
     if (button_index >= mouse_info.size()) {
         return;
@@ -67,7 +67,7 @@ void Mouse::PressButton(int x, int y, int button_) {
 
     const auto button = 1U << button_index;
     buttons |= static_cast<u16>(button);
-    last_button = static_cast<MouseButton>(button_index);
+    last_button = button_;
 
     mouse_info[button_index].mouse_origin = Common::MakeVec(x, y);
     mouse_info[button_index].last_mouse_position = Common::MakeVec(x, y);
@@ -129,7 +129,7 @@ void Mouse::MouseMove(int x, int y, int center_x, int center_y) {
     }
 }
 
-void Mouse::ReleaseButton(int button_) {
+void Mouse::ReleaseButton(MouseButton button_) {
     const auto button_index = static_cast<std::size_t>(button_);
     if (button_index >= mouse_info.size()) {
         return;
@@ -152,6 +152,11 @@ void Mouse::BeginConfiguration() {
 
 void Mouse::EndConfiguration() {
     buttons = 0;
+    for (MouseInfo& info : mouse_info) {
+        info.tilt_speed = 0;
+        info.data.pressed = false;
+        info.data.axis = {0, 0};
+    }
     last_button = MouseButton::Undefined;
     mouse_queue.Clear();
     configuring = false;
diff --git a/src/input_common/mouse/mouse_input.h b/src/input_common/mouse/mouse_input.h
index a75f0ecb5..750d9b011 100644
--- a/src/input_common/mouse/mouse_input.h
+++ b/src/input_common/mouse/mouse_input.h
@@ -18,10 +18,12 @@ namespace MouseInput {
 
 enum class MouseButton {
     Left,
-    Wheel,
     Right,
-    Forward,
+    Wheel,
     Backward,
+    Forward,
+    Task,
+    Extra,
     Undefined,
 };
 
@@ -51,7 +53,7 @@ public:
      * @param y the y-coordinate of the cursor
      * @param button_ the button pressed
      */
-    void PressButton(int x, int y, int button_);
+    void PressButton(int x, int y, MouseButton button_);
 
     /**
      * Signals that mouse has moved.
@@ -65,7 +67,7 @@ public:
     /**
      * Signals that a motion sensor tilt has ended.
      */
-    void ReleaseButton(int button_);
+    void ReleaseButton(MouseButton button_);
 
     [[nodiscard]] bool ToggleButton(std::size_t button_);
     [[nodiscard]] bool UnlockButton(std::size_t button_);
@@ -99,7 +101,7 @@ private:
     u16 lock_buttons{};
     std::thread update_thread;
     MouseButton last_button{MouseButton::Undefined};
-    std::array<MouseInfo, 5> mouse_info;
+    std::array<MouseInfo, 7> mouse_info;
     Common::SPSCQueue<MouseStatus> mouse_queue;
     bool configuring{false};
     bool update_thread_running{true};
diff --git a/src/tests/common/fibers.cpp b/src/tests/common/fibers.cpp
index b2ded2065..751cbe196 100644
--- a/src/tests/common/fibers.cpp
+++ b/src/tests/common/fibers.cpp
@@ -67,15 +67,16 @@ void TestControl1::DoWork() {
         value++;
     }
     results[id] = value;
-    Fiber::YieldTo(work_fibers[id].get(), thread_fibers[id].get());
+    Fiber::YieldTo(work_fibers[id], *thread_fibers[id]);
 }
 
 void TestControl1::ExecuteThread(u32 id) {
     thread_ids.Register(id);
-    thread_fibers[id] = Fiber::ThreadToFiber();
+    auto thread_fiber = Fiber::ThreadToFiber();
+    thread_fibers[id] = thread_fiber;
     work_fibers[id] = std::make_shared<Fiber>(std::function<void(void*)>{WorkControl1}, this);
     items[id] = rand() % 256;
-    Fiber::YieldTo(thread_fibers[id].get(), work_fibers[id].get());
+    Fiber::YieldTo(thread_fibers[id], *work_fibers[id]);
     thread_fibers[id]->Exit();
 }
 
@@ -116,11 +117,11 @@ public:
         for (u32 i = 0; i < 12000; i++) {
             value1 += i;
         }
-        Fiber::YieldTo(fiber1.get(), fiber3.get());
+        Fiber::YieldTo(fiber1, *fiber3);
         const u32 id = thread_ids.Get();
         assert1 = id == 1;
         value2 += 5000;
-        Fiber::YieldTo(fiber1.get(), thread_fibers[id].get());
+        Fiber::YieldTo(fiber1, *thread_fibers[id]);
     }
 
     void DoWork2() {
@@ -128,7 +129,7 @@ public:
             ;
         value2 = 2000;
         trap = false;
-        Fiber::YieldTo(fiber2.get(), fiber1.get());
+        Fiber::YieldTo(fiber2, *fiber1);
         assert3 = false;
     }
 
@@ -136,19 +137,19 @@ public:
         const u32 id = thread_ids.Get();
         assert2 = id == 0;
         value1 += 1000;
-        Fiber::YieldTo(fiber3.get(), thread_fibers[id].get());
+        Fiber::YieldTo(fiber3, *thread_fibers[id]);
     }
 
     void ExecuteThread(u32 id);
 
     void CallFiber1() {
         const u32 id = thread_ids.Get();
-        Fiber::YieldTo(thread_fibers[id].get(), fiber1.get());
+        Fiber::YieldTo(thread_fibers[id], *fiber1);
     }
 
     void CallFiber2() {
         const u32 id = thread_ids.Get();
-        Fiber::YieldTo(thread_fibers[id].get(), fiber2.get());
+        Fiber::YieldTo(thread_fibers[id], *fiber2);
     }
 
     void Exit();
@@ -184,7 +185,8 @@ static void WorkControl2_3(void* control) {
 
 void TestControl2::ExecuteThread(u32 id) {
     thread_ids.Register(id);
-    thread_fibers[id] = Fiber::ThreadToFiber();
+    auto thread_fiber = Fiber::ThreadToFiber();
+    thread_fibers[id] = thread_fiber;
 }
 
 void TestControl2::Exit() {
@@ -239,23 +241,23 @@ public:
 
     void DoWork1() {
         value1 += 1;
-        Fiber::YieldTo(fiber1.get(), fiber2.get());
+        Fiber::YieldTo(fiber1, *fiber2);
         const u32 id = thread_ids.Get();
         value3 += 1;
-        Fiber::YieldTo(fiber1.get(), thread_fibers[id].get());
+        Fiber::YieldTo(fiber1, *thread_fibers[id]);
     }
 
     void DoWork2() {
         value2 += 1;
         const u32 id = thread_ids.Get();
-        Fiber::YieldTo(fiber2.get(), thread_fibers[id].get());
+        Fiber::YieldTo(fiber2, *thread_fibers[id]);
     }
 
     void ExecuteThread(u32 id);
 
     void CallFiber1() {
         const u32 id = thread_ids.Get();
-        Fiber::YieldTo(thread_fibers[id].get(), fiber1.get());
+        Fiber::YieldTo(thread_fibers[id], *fiber1);
     }
 
     void Exit();
@@ -264,7 +266,7 @@ public:
     u32 value2{};
     u32 value3{};
     ThreadIds thread_ids;
-    std::vector<std::unique_ptr<Common::Fiber>> thread_fibers;
+    std::vector<std::shared_ptr<Common::Fiber>> thread_fibers;
     std::shared_ptr<Common::Fiber> fiber1;
     std::shared_ptr<Common::Fiber> fiber2;
 };
@@ -281,7 +283,8 @@ static void WorkControl3_2(void* control) {
 
 void TestControl3::ExecuteThread(u32 id) {
     thread_ids.Register(id);
-    thread_fibers[id] = Fiber::ThreadToFiber();
+    auto thread_fiber = Fiber::ThreadToFiber();
+    thread_fibers[id] = thread_fiber;
 }
 
 void TestControl3::Exit() {
@@ -329,7 +332,7 @@ public:
 
     void Execute() {
         thread_fiber = Fiber::ThreadToFiber();
-        Fiber::YieldTo(thread_fiber.get(), fiber1.get());
+        Fiber::YieldTo(thread_fiber, *fiber1);
         thread_fiber->Exit();
     }
 
@@ -337,7 +340,7 @@ public:
         fiber1->SetRewindPoint(std::function<void(void*)>{WorkControl4}, this);
         if (rewinded) {
             goal_reached = true;
-            Fiber::YieldTo(fiber1.get(), thread_fiber.get());
+            Fiber::YieldTo(fiber1, *thread_fiber);
         }
         rewinded = true;
         fiber1->Rewind();
diff --git a/src/video_core/buffer_cache/buffer_cache.h b/src/video_core/buffer_cache/buffer_cache.h
index 2a6844ab1..4de1e37e5 100644
--- a/src/video_core/buffer_cache/buffer_cache.h
+++ b/src/video_core/buffer_cache/buffer_cache.h
@@ -9,6 +9,7 @@
 #include <deque>
 #include <memory>
 #include <mutex>
+#include <numeric>
 #include <span>
 #include <unordered_map>
 #include <vector>
@@ -91,7 +92,7 @@ class BufferCache {
     };
 
 public:
-    static constexpr u32 SKIP_CACHE_SIZE = 4096;
+    static constexpr u32 DEFAULT_SKIP_CACHE_SIZE = 4096;
 
     explicit BufferCache(VideoCore::RasterizerInterface& rasterizer_,
                          Tegra::Engines::Maxwell3D& maxwell3d_,
@@ -240,9 +241,9 @@ private:
     template <bool insert>
     void ChangeRegister(BufferId buffer_id);
 
-    void SynchronizeBuffer(Buffer& buffer, VAddr cpu_addr, u32 size);
+    bool SynchronizeBuffer(Buffer& buffer, VAddr cpu_addr, u32 size);
 
-    void SynchronizeBufferImpl(Buffer& buffer, VAddr cpu_addr, u32 size);
+    bool SynchronizeBufferImpl(Buffer& buffer, VAddr cpu_addr, u32 size);
 
     void UploadMemory(Buffer& buffer, u64 total_size_bytes, u64 largest_copy,
                       std::span<BufferCopy> copies);
@@ -297,6 +298,11 @@ private:
 
     std::array<u32, NUM_STAGES> fast_bound_uniform_buffers{};
 
+    std::array<u32, 16> uniform_cache_hits{};
+    std::array<u32, 16> uniform_cache_shots{};
+
+    u32 uniform_buffer_skip_cache_size = DEFAULT_SKIP_CACHE_SIZE;
+
     bool has_deleted_buffers = false;
 
     std::conditional_t<HAS_PERSISTENT_UNIFORM_BUFFER_BINDINGS, std::array<u32, NUM_STAGES>, Empty>
@@ -328,6 +334,19 @@ BufferCache<P>::BufferCache(VideoCore::RasterizerInterface& rasterizer_,
 
 template <class P>
 void BufferCache<P>::TickFrame() {
+    // Calculate hits and shots and move hit bits to the right
+    const u32 hits = std::reduce(uniform_cache_hits.begin(), uniform_cache_hits.end());
+    const u32 shots = std::reduce(uniform_cache_shots.begin(), uniform_cache_shots.end());
+    std::copy_n(uniform_cache_hits.begin(), uniform_cache_hits.size() - 1,
+                uniform_cache_hits.begin() + 1);
+    std::copy_n(uniform_cache_shots.begin(), uniform_cache_shots.size() - 1,
+                uniform_cache_shots.begin() + 1);
+    uniform_cache_hits[0] = 0;
+    uniform_cache_shots[0] = 0;
+
+    const bool skip_preferred = hits * 256 < shots * 251;
+    uniform_buffer_skip_cache_size = skip_preferred ? DEFAULT_SKIP_CACHE_SIZE : 0;
+
     delayed_destruction_ring.Tick();
 }
 
@@ -671,7 +690,7 @@ void BufferCache<P>::BindHostGraphicsUniformBuffer(size_t stage, u32 index, u32
     const VAddr cpu_addr = binding.cpu_addr;
     const u32 size = binding.size;
     Buffer& buffer = slot_buffers[binding.buffer_id];
-    if (size <= SKIP_CACHE_SIZE && !buffer.IsRegionGpuModified(cpu_addr, size)) {
+    if (size <= uniform_buffer_skip_cache_size && !buffer.IsRegionGpuModified(cpu_addr, size)) {
         if constexpr (IS_OPENGL) {
             if (runtime.HasFastBufferSubData()) {
                 // Fast path for Nvidia
@@ -692,7 +711,12 @@ void BufferCache<P>::BindHostGraphicsUniformBuffer(size_t stage, u32 index, u32
         return;
     }
     // Classic cached path
-    SynchronizeBuffer(buffer, cpu_addr, size);
+    const bool sync_cached = SynchronizeBuffer(buffer, cpu_addr, size);
+    if (sync_cached) {
+        ++uniform_cache_hits[0];
+    }
+    ++uniform_cache_shots[0];
+
     if (!needs_bind && !HasFastUniformBufferBound(stage, binding_index)) {
         // Skip binding if it's not needed and if the bound buffer is not the fast version
         // This exists to avoid instances where the fast buffer is bound and a GPU write happens
@@ -1106,15 +1130,15 @@ void BufferCache<P>::ChangeRegister(BufferId buffer_id) {
 }
 
 template <class P>
-void BufferCache<P>::SynchronizeBuffer(Buffer& buffer, VAddr cpu_addr, u32 size) {
+bool BufferCache<P>::SynchronizeBuffer(Buffer& buffer, VAddr cpu_addr, u32 size) {
     if (buffer.CpuAddr() == 0) {
-        return;
+        return true;
     }
-    SynchronizeBufferImpl(buffer, cpu_addr, size);
+    return SynchronizeBufferImpl(buffer, cpu_addr, size);
 }
 
 template <class P>
-void BufferCache<P>::SynchronizeBufferImpl(Buffer& buffer, VAddr cpu_addr, u32 size) {
+bool BufferCache<P>::SynchronizeBufferImpl(Buffer& buffer, VAddr cpu_addr, u32 size) {
     boost::container::small_vector<BufferCopy, 4> copies;
     u64 total_size_bytes = 0;
     u64 largest_copy = 0;
@@ -1128,10 +1152,11 @@ void BufferCache<P>::SynchronizeBufferImpl(Buffer& buffer, VAddr cpu_addr, u32 s
         largest_copy = std::max(largest_copy, range_size);
     });
     if (total_size_bytes == 0) {
-        return;
+        return true;
     }
     const std::span<BufferCopy> copies_span(copies.data(), copies.size());
     UploadMemory(buffer, total_size_bytes, largest_copy, copies_span);
+    return false;
 }
 
 template <class P>
diff --git a/src/video_core/compatible_formats.cpp b/src/video_core/compatible_formats.cpp
index acf2668dc..8317d0636 100644
--- a/src/video_core/compatible_formats.cpp
+++ b/src/video_core/compatible_formats.cpp
@@ -48,6 +48,15 @@ constexpr std::array VIEW_CLASS_32_BITS{
     PixelFormat::A2B10G10R10_UINT,
 };
 
+constexpr std::array VIEW_CLASS_32_BITS_NO_BGR{
+    PixelFormat::R16G16_FLOAT,      PixelFormat::B10G11R11_FLOAT,  PixelFormat::R32_FLOAT,
+    PixelFormat::A2B10G10R10_UNORM, PixelFormat::R16G16_UINT,      PixelFormat::R32_UINT,
+    PixelFormat::R16G16_SINT,       PixelFormat::R32_SINT,         PixelFormat::A8B8G8R8_UNORM,
+    PixelFormat::R16G16_UNORM,      PixelFormat::A8B8G8R8_SNORM,   PixelFormat::R16G16_SNORM,
+    PixelFormat::A8B8G8R8_SRGB,     PixelFormat::E5B9G9R9_FLOAT,   PixelFormat::A8B8G8R8_UINT,
+    PixelFormat::A8B8G8R8_SINT,     PixelFormat::A2B10G10R10_UINT,
+};
+
 // TODO: How should we handle 24 bits?
 
 constexpr std::array VIEW_CLASS_16_BITS{
@@ -205,7 +214,6 @@ constexpr Table MakeViewTable() {
     EnableRange(view, VIEW_CLASS_128_BITS);
     EnableRange(view, VIEW_CLASS_96_BITS);
     EnableRange(view, VIEW_CLASS_64_BITS);
-    EnableRange(view, VIEW_CLASS_32_BITS);
     EnableRange(view, VIEW_CLASS_16_BITS);
     EnableRange(view, VIEW_CLASS_8_BITS);
     EnableRange(view, VIEW_CLASS_RGTC1_RED);
@@ -231,20 +239,47 @@ constexpr Table MakeCopyTable() {
     EnableRange(copy, COPY_CLASS_64_BITS);
     return copy;
 }
+
+constexpr Table MakeNativeBgrViewTable() {
+    Table copy = MakeViewTable();
+    EnableRange(copy, VIEW_CLASS_32_BITS);
+    return copy;
+}
+
+constexpr Table MakeNonNativeBgrViewTable() {
+    Table copy = MakeViewTable();
+    EnableRange(copy, VIEW_CLASS_32_BITS_NO_BGR);
+    return copy;
+}
+
+constexpr Table MakeNativeBgrCopyTable() {
+    Table copy = MakeCopyTable();
+    EnableRange(copy, VIEW_CLASS_32_BITS);
+    return copy;
+}
+
+constexpr Table MakeNonNativeBgrCopyTable() {
+    Table copy = MakeCopyTable();
+    EnableRange(copy, VIEW_CLASS_32_BITS);
+    return copy;
+}
 } // Anonymous namespace
 
-bool IsViewCompatible(PixelFormat format_a, PixelFormat format_b, bool broken_views) {
+bool IsViewCompatible(PixelFormat format_a, PixelFormat format_b, bool broken_views,
+                      bool native_bgr) {
     if (broken_views) {
         // If format views are broken, only accept formats that are identical.
         return format_a == format_b;
     }
-    static constexpr Table TABLE = MakeViewTable();
-    return IsSupported(TABLE, format_a, format_b);
+    static constexpr Table BGR_TABLE = MakeNativeBgrViewTable();
+    static constexpr Table NO_BGR_TABLE = MakeNonNativeBgrViewTable();
+    return IsSupported(native_bgr ? BGR_TABLE : NO_BGR_TABLE, format_a, format_b);
 }
 
-bool IsCopyCompatible(PixelFormat format_a, PixelFormat format_b) {
-    static constexpr Table TABLE = MakeCopyTable();
-    return IsSupported(TABLE, format_a, format_b);
+bool IsCopyCompatible(PixelFormat format_a, PixelFormat format_b, bool native_bgr) {
+    static constexpr Table BGR_TABLE = MakeNativeBgrCopyTable();
+    static constexpr Table NO_BGR_TABLE = MakeNonNativeBgrCopyTable();
+    return IsSupported(native_bgr ? BGR_TABLE : NO_BGR_TABLE, format_a, format_b);
 }
 
 } // namespace VideoCore::Surface
diff --git a/src/video_core/compatible_formats.h b/src/video_core/compatible_formats.h
index 9a0522988..55745e042 100644
--- a/src/video_core/compatible_formats.h
+++ b/src/video_core/compatible_formats.h
@@ -8,8 +8,9 @@
 
 namespace VideoCore::Surface {
 
-bool IsViewCompatible(PixelFormat format_a, PixelFormat format_b, bool broken_views);
+bool IsViewCompatible(PixelFormat format_a, PixelFormat format_b, bool broken_views,
+                      bool native_bgr);
 
-bool IsCopyCompatible(PixelFormat format_a, PixelFormat format_b);
+bool IsCopyCompatible(PixelFormat format_a, PixelFormat format_b, bool native_bgr);
 
 } // namespace VideoCore::Surface
diff --git a/src/video_core/host_shaders/CMakeLists.txt b/src/video_core/host_shaders/CMakeLists.txt
index 970120acc..3494318ca 100644
--- a/src/video_core/host_shaders/CMakeLists.txt
+++ b/src/video_core/host_shaders/CMakeLists.txt
@@ -5,6 +5,7 @@ set(SHADER_FILES
     convert_float_to_depth.frag
     full_screen_triangle.vert
     opengl_copy_bc4.comp
+    opengl_copy_bgra.comp
     opengl_present.frag
     opengl_present.vert
     pitch_unswizzle.comp
diff --git a/src/video_core/host_shaders/opengl_copy_bgra.comp b/src/video_core/host_shaders/opengl_copy_bgra.comp
new file mode 100644
index 000000000..2571a4abf
--- /dev/null
+++ b/src/video_core/host_shaders/opengl_copy_bgra.comp
@@ -0,0 +1,15 @@
+// Copyright 2021 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#version 430 core
+
+layout (local_size_x = 4, local_size_y = 4) in;
+
+layout(binding = 0, rgba8) readonly uniform image2DArray bgr_input;
+layout(binding = 1, rgba8) writeonly uniform image2DArray bgr_output;
+
+void main() {
+    vec4 color = imageLoad(bgr_input, ivec3(gl_GlobalInvocationID));
+    imageStore(bgr_output, ivec3(gl_GlobalInvocationID), color.bgra);
+}
diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.cpp b/src/video_core/renderer_opengl/gl_buffer_cache.cpp
index 6da3906a4..c225d1fc9 100644
--- a/src/video_core/renderer_opengl/gl_buffer_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_buffer_cache.cpp
@@ -73,7 +73,8 @@ BufferCacheRuntime::BufferCacheRuntime(const Device& device_)
     for (auto& stage_uniforms : fast_uniforms) {
         for (OGLBuffer& buffer : stage_uniforms) {
             buffer.Create();
-            glNamedBufferData(buffer.handle, BufferCache::SKIP_CACHE_SIZE, nullptr, GL_STREAM_DRAW);
+            glNamedBufferData(buffer.handle, BufferCache::DEFAULT_SKIP_CACHE_SIZE, nullptr,
+                              GL_STREAM_DRAW);
         }
     }
     for (auto& stage_uniforms : copy_uniforms) {
diff --git a/src/video_core/renderer_opengl/gl_texture_cache.cpp b/src/video_core/renderer_opengl/gl_texture_cache.cpp
index 12434db67..e028677e9 100644
--- a/src/video_core/renderer_opengl/gl_texture_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_texture_cache.cpp
@@ -96,7 +96,7 @@ constexpr std::array<FormatTuple, MaxPixelFormat> FORMAT_TABLE = {{
     {GL_COMPRESSED_RGB_BPTC_UNSIGNED_FLOAT},                          // BC6H_UFLOAT
     {GL_COMPRESSED_RGB_BPTC_SIGNED_FLOAT},                            // BC6H_SFLOAT
     {GL_COMPRESSED_RGBA_ASTC_4x4_KHR},                                // ASTC_2D_4X4_UNORM
-    {GL_RGBA8, GL_BGRA, GL_UNSIGNED_BYTE},                            // B8G8R8A8_UNORM
+    {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE},                            // B8G8R8A8_UNORM
     {GL_RGBA32F, GL_RGBA, GL_FLOAT},                                  // R32G32B32A32_FLOAT
     {GL_RGBA32I, GL_RGBA_INTEGER, GL_INT},                            // R32G32B32A32_SINT
     {GL_RG32F, GL_RG, GL_FLOAT},                                      // R32G32_FLOAT
@@ -125,7 +125,7 @@ constexpr std::array<FormatTuple, MaxPixelFormat> FORMAT_TABLE = {{
     {GL_COMPRESSED_RGBA_ASTC_8x8_KHR},                                // ASTC_2D_8X8_UNORM
     {GL_COMPRESSED_RGBA_ASTC_8x5_KHR},                                // ASTC_2D_8X5_UNORM
     {GL_COMPRESSED_RGBA_ASTC_5x4_KHR},                                // ASTC_2D_5X4_UNORM
-    {GL_SRGB8_ALPHA8, GL_BGRA, GL_UNSIGNED_BYTE},                     // B8G8R8A8_UNORM
+    {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_BYTE},                     // B8G8R8A8_SRGB
     {GL_COMPRESSED_SRGB_ALPHA_S3TC_DXT1_EXT},                         // BC1_RGBA_SRGB
     {GL_COMPRESSED_SRGB_ALPHA_S3TC_DXT3_EXT},                         // BC2_SRGB
     {GL_COMPRESSED_SRGB_ALPHA_S3TC_DXT5_EXT},                         // BC3_SRGB
@@ -396,6 +396,17 @@ void AttachTexture(GLuint fbo, GLenum attachment, const ImageView* image_view) {
     }
 }
 
+[[nodiscard]] bool IsPixelFormatBGR(PixelFormat format) {
+    switch (format) {
+    case PixelFormat::B5G6R5_UNORM:
+    case PixelFormat::B8G8R8A8_UNORM:
+    case PixelFormat::B8G8R8A8_SRGB:
+        return true;
+    default:
+        return false;
+    }
+}
+
 } // Anonymous namespace
 
 ImageBufferMap::~ImageBufferMap() {
@@ -512,6 +523,9 @@ bool TextureCacheRuntime::CanImageBeCopied(const Image& dst, const Image& src) {
     if (dst.info.type == ImageType::e3D && dst.info.format == PixelFormat::BC4_UNORM) {
         return false;
     }
+    if (IsPixelFormatBGR(dst.info.format) || IsPixelFormatBGR(src.info.format)) {
+        return false;
+    }
     return true;
 }
 
@@ -520,6 +534,8 @@ void TextureCacheRuntime::EmulateCopyImage(Image& dst, Image& src,
     if (dst.info.type == ImageType::e3D && dst.info.format == PixelFormat::BC4_UNORM) {
         ASSERT(src.info.type == ImageType::e3D);
         util_shaders.CopyBC4(dst, src, copies);
+    } else if (IsPixelFormatBGR(dst.info.format) || IsPixelFormatBGR(src.info.format)) {
+        util_shaders.CopyBGR(dst, src, copies);
     } else {
         UNREACHABLE();
     }
diff --git a/src/video_core/renderer_opengl/gl_texture_cache.h b/src/video_core/renderer_opengl/gl_texture_cache.h
index a6172f009..3fbaa102f 100644
--- a/src/video_core/renderer_opengl/gl_texture_cache.h
+++ b/src/video_core/renderer_opengl/gl_texture_cache.h
@@ -86,6 +86,11 @@ public:
 
     FormatProperties FormatInfo(VideoCommon::ImageType type, GLenum internal_format) const;
 
+    bool HasNativeBgr() const noexcept {
+        // OpenGL does not have native support for the BGR internal format
+        return false;
+    }
+
     bool HasBrokenTextureViewFormats() const noexcept {
         return has_broken_texture_view_formats;
     }
diff --git a/src/video_core/renderer_opengl/util_shaders.cpp b/src/video_core/renderer_opengl/util_shaders.cpp
index 31ec68505..2fe4799bc 100644
--- a/src/video_core/renderer_opengl/util_shaders.cpp
+++ b/src/video_core/renderer_opengl/util_shaders.cpp
@@ -14,6 +14,7 @@
 #include "video_core/host_shaders/block_linear_unswizzle_2d_comp.h"
 #include "video_core/host_shaders/block_linear_unswizzle_3d_comp.h"
 #include "video_core/host_shaders/opengl_copy_bc4_comp.h"
+#include "video_core/host_shaders/opengl_copy_bgra_comp.h"
 #include "video_core/host_shaders/pitch_unswizzle_comp.h"
 #include "video_core/renderer_opengl/gl_resource_manager.h"
 #include "video_core/renderer_opengl/gl_shader_manager.h"
@@ -48,6 +49,11 @@ OGLProgram MakeProgram(std::string_view source) {
     return program;
 }
 
+size_t NumPixelsInCopy(const VideoCommon::ImageCopy& copy) {
+    return static_cast<size_t>(copy.extent.width * copy.extent.height *
+                               copy.src_subresource.num_layers);
+}
+
 } // Anonymous namespace
 
 UtilShaders::UtilShaders(ProgramManager& program_manager_)
@@ -55,6 +61,7 @@ UtilShaders::UtilShaders(ProgramManager& program_manager_)
       block_linear_unswizzle_2d_program(MakeProgram(BLOCK_LINEAR_UNSWIZZLE_2D_COMP)),
       block_linear_unswizzle_3d_program(MakeProgram(BLOCK_LINEAR_UNSWIZZLE_3D_COMP)),
       pitch_unswizzle_program(MakeProgram(PITCH_UNSWIZZLE_COMP)),
+      copy_bgra_program(MakeProgram(OPENGL_COPY_BGRA_COMP)),
       copy_bc4_program(MakeProgram(OPENGL_COPY_BC4_COMP)) {
     const auto swizzle_table = Tegra::Texture::MakeSwizzleTable();
     swizzle_table_buffer.Create();
@@ -205,6 +212,43 @@ void UtilShaders::CopyBC4(Image& dst_image, Image& src_image, std::span<const Im
     program_manager.RestoreGuestCompute();
 }
 
+void UtilShaders::CopyBGR(Image& dst_image, Image& src_image,
+                          std::span<const VideoCommon::ImageCopy> copies) {
+    static constexpr GLuint BINDING_INPUT_IMAGE = 0;
+    static constexpr GLuint BINDING_OUTPUT_IMAGE = 1;
+    static constexpr VideoCommon::Offset3D zero_offset{0, 0, 0};
+    const u32 bytes_per_block = BytesPerBlock(dst_image.info.format);
+    switch (bytes_per_block) {
+    case 2:
+        // BGR565 copy
+        for (const ImageCopy& copy : copies) {
+            ASSERT(copy.src_offset == zero_offset);
+            ASSERT(copy.dst_offset == zero_offset);
+            bgr_copy_pass.Execute(dst_image, src_image, copy);
+        }
+        break;
+    case 4: {
+        // BGRA8 copy
+        program_manager.BindHostCompute(copy_bgra_program.handle);
+        constexpr GLenum FORMAT = GL_RGBA8;
+        for (const ImageCopy& copy : copies) {
+            ASSERT(copy.src_offset == zero_offset);
+            ASSERT(copy.dst_offset == zero_offset);
+            glBindImageTexture(BINDING_INPUT_IMAGE, src_image.StorageHandle(),
+                               copy.src_subresource.base_level, GL_FALSE, 0, GL_READ_ONLY, FORMAT);
+            glBindImageTexture(BINDING_OUTPUT_IMAGE, dst_image.StorageHandle(),
+                               copy.dst_subresource.base_level, GL_FALSE, 0, GL_WRITE_ONLY, FORMAT);
+            glDispatchCompute(copy.extent.width, copy.extent.height, copy.extent.depth);
+        }
+        program_manager.RestoreGuestCompute();
+        break;
+    }
+    default:
+        UNREACHABLE();
+        break;
+    }
+}
+
 GLenum StoreFormat(u32 bytes_per_block) {
     switch (bytes_per_block) {
     case 1:
@@ -222,4 +266,36 @@ GLenum StoreFormat(u32 bytes_per_block) {
     return GL_R8UI;
 }
 
+void Bgr565CopyPass::Execute(const Image& dst_image, const Image& src_image,
+                             const ImageCopy& copy) {
+    if (CopyBufferCreationNeeded(copy)) {
+        CreateNewCopyBuffer(copy, GL_TEXTURE_2D_ARRAY, GL_RGB565);
+    }
+    // Copy from source to PBO
+    glPixelStorei(GL_PACK_ALIGNMENT, 1);
+    glPixelStorei(GL_PACK_ROW_LENGTH, copy.extent.width);
+    glBindBuffer(GL_PIXEL_PACK_BUFFER, bgr16_pbo.handle);
+    glGetTextureSubImage(src_image.Handle(), 0, 0, 0, 0, copy.extent.width, copy.extent.height,
+                         copy.src_subresource.num_layers, GL_RGB, GL_UNSIGNED_SHORT_5_6_5,
+                         static_cast<GLsizei>(bgr16_pbo_size), nullptr);
+
+    // Copy from PBO to destination in reverse order
+    glPixelStorei(GL_UNPACK_ALIGNMENT, 1);
+    glPixelStorei(GL_UNPACK_ROW_LENGTH, copy.extent.width);
+    glBindBuffer(GL_PIXEL_UNPACK_BUFFER, bgr16_pbo.handle);
+    glTextureSubImage3D(dst_image.Handle(), 0, 0, 0, 0, copy.extent.width, copy.extent.height,
+                        copy.dst_subresource.num_layers, GL_RGB, GL_UNSIGNED_SHORT_5_6_5_REV,
+                        nullptr);
+}
+
+bool Bgr565CopyPass::CopyBufferCreationNeeded(const ImageCopy& copy) {
+    return bgr16_pbo_size < NumPixelsInCopy(copy) * sizeof(u16);
+}
+
+void Bgr565CopyPass::CreateNewCopyBuffer(const ImageCopy& copy, GLenum target, GLuint format) {
+    bgr16_pbo.Create();
+    bgr16_pbo_size = NumPixelsInCopy(copy) * sizeof(u16);
+    glNamedBufferData(bgr16_pbo.handle, bgr16_pbo_size, nullptr, GL_STREAM_COPY);
+}
+
 } // namespace OpenGL
diff --git a/src/video_core/renderer_opengl/util_shaders.h b/src/video_core/renderer_opengl/util_shaders.h
index 7b1d16b09..93b009743 100644
--- a/src/video_core/renderer_opengl/util_shaders.h
+++ b/src/video_core/renderer_opengl/util_shaders.h
@@ -19,6 +19,22 @@ class ProgramManager;
 
 struct ImageBufferMap;
 
+class Bgr565CopyPass {
+public:
+    Bgr565CopyPass() = default;
+    ~Bgr565CopyPass() = default;
+
+    void Execute(const Image& dst_image, const Image& src_image,
+                 const VideoCommon::ImageCopy& copy);
+
+private:
+    [[nodiscard]] bool CopyBufferCreationNeeded(const VideoCommon::ImageCopy& copy);
+    void CreateNewCopyBuffer(const VideoCommon::ImageCopy& copy, GLenum target, GLuint format);
+
+    OGLBuffer bgr16_pbo;
+    size_t bgr16_pbo_size{};
+};
+
 class UtilShaders {
 public:
     explicit UtilShaders(ProgramManager& program_manager);
@@ -36,6 +52,9 @@ public:
     void CopyBC4(Image& dst_image, Image& src_image,
                  std::span<const VideoCommon::ImageCopy> copies);
 
+    void CopyBGR(Image& dst_image, Image& src_image,
+                 std::span<const VideoCommon::ImageCopy> copies);
+
 private:
     ProgramManager& program_manager;
 
@@ -44,7 +63,10 @@ private:
     OGLProgram block_linear_unswizzle_2d_program;
     OGLProgram block_linear_unswizzle_3d_program;
     OGLProgram pitch_unswizzle_program;
+    OGLProgram copy_bgra_program;
     OGLProgram copy_bc4_program;
+
+    Bgr565CopyPass bgr_copy_pass;
 };
 
 GLenum StoreFormat(u32 bytes_per_block);
diff --git a/src/video_core/renderer_vulkan/vk_texture_cache.h b/src/video_core/renderer_vulkan/vk_texture_cache.h
index b08c23459..3aee27ce0 100644
--- a/src/video_core/renderer_vulkan/vk_texture_cache.h
+++ b/src/video_core/renderer_vulkan/vk_texture_cache.h
@@ -93,6 +93,11 @@ struct TextureCacheRuntime {
         // No known Vulkan driver has broken image views
         return false;
     }
+
+    bool HasNativeBgr() const noexcept {
+        // All known Vulkan drivers can natively handle BGR textures
+        return true;
+    }
 };
 
 class Image : public VideoCommon::ImageBase {
diff --git a/src/video_core/texture_cache/image_base.cpp b/src/video_core/texture_cache/image_base.cpp
index 959b3f115..9914926b3 100644
--- a/src/video_core/texture_cache/image_base.cpp
+++ b/src/video_core/texture_cache/image_base.cpp
@@ -120,9 +120,10 @@ void AddImageAlias(ImageBase& lhs, ImageBase& rhs, ImageId lhs_id, ImageId rhs_i
     if (lhs.info.type == ImageType::Linear) {
         base = SubresourceBase{.level = 0, .layer = 0};
     } else {
-        // We are passing relaxed formats as an option, having broken views or not won't matter
+        // We are passing relaxed formats as an option, having broken views/bgr or not won't matter
         static constexpr bool broken_views = false;
-        base = FindSubresource(rhs.info, lhs, rhs.gpu_addr, OPTIONS, broken_views);
+        static constexpr bool native_bgr = true;
+        base = FindSubresource(rhs.info, lhs, rhs.gpu_addr, OPTIONS, broken_views, native_bgr);
     }
     if (!base) {
         LOG_ERROR(HW_GPU, "Image alias should have been flipped");
diff --git a/src/video_core/texture_cache/image_view_base.cpp b/src/video_core/texture_cache/image_view_base.cpp
index 18f72e508..f89a40b4c 100644
--- a/src/video_core/texture_cache/image_view_base.cpp
+++ b/src/video_core/texture_cache/image_view_base.cpp
@@ -24,7 +24,7 @@ ImageViewBase::ImageViewBase(const ImageViewInfo& info, const ImageInfo& image_i
           .height = std::max(image_info.size.height >> range.base.level, 1u),
           .depth = std::max(image_info.size.depth >> range.base.level, 1u),
       } {
-    ASSERT_MSG(VideoCore::Surface::IsViewCompatible(image_info.format, info.format, false),
+    ASSERT_MSG(VideoCore::Surface::IsViewCompatible(image_info.format, info.format, false, true),
                "Image view format {} is incompatible with image format {}", info.format,
                image_info.format);
     const bool is_async = Settings::values.use_asynchronous_gpu_emulation.GetValue();
diff --git a/src/video_core/texture_cache/texture_cache.h b/src/video_core/texture_cache/texture_cache.h
index b1da69971..98e33c3a0 100644
--- a/src/video_core/texture_cache/texture_cache.h
+++ b/src/video_core/texture_cache/texture_cache.h
@@ -876,6 +876,7 @@ ImageId TextureCache<P>::FindImage(const ImageInfo& info, GPUVAddr gpu_addr,
         return ImageId{};
     }
     const bool broken_views = runtime.HasBrokenTextureViewFormats();
+    const bool native_bgr = runtime.HasNativeBgr();
     ImageId image_id;
     const auto lambda = [&](ImageId existing_image_id, ImageBase& existing_image) {
         if (info.type == ImageType::Linear || existing_image.info.type == ImageType::Linear) {
@@ -885,11 +886,12 @@ ImageId TextureCache<P>::FindImage(const ImageInfo& info, GPUVAddr gpu_addr,
             if (existing_image.gpu_addr == gpu_addr && existing.type == info.type &&
                 existing.pitch == info.pitch &&
                 IsPitchLinearSameSize(existing, info, strict_size) &&
-                IsViewCompatible(existing.format, info.format, broken_views)) {
+                IsViewCompatible(existing.format, info.format, broken_views, native_bgr)) {
                 image_id = existing_image_id;
                 return true;
             }
-        } else if (IsSubresource(info, existing_image, gpu_addr, options, broken_views)) {
+        } else if (IsSubresource(info, existing_image, gpu_addr, options, broken_views,
+                                 native_bgr)) {
             image_id = existing_image_id;
             return true;
         }
@@ -920,6 +922,7 @@ ImageId TextureCache<P>::JoinImages(const ImageInfo& info, GPUVAddr gpu_addr, VA
     ImageInfo new_info = info;
     const size_t size_bytes = CalculateGuestSizeInBytes(new_info);
     const bool broken_views = runtime.HasBrokenTextureViewFormats();
+    const bool native_bgr = runtime.HasNativeBgr();
     std::vector<ImageId> overlap_ids;
     std::vector<ImageId> left_aliased_ids;
     std::vector<ImageId> right_aliased_ids;
@@ -935,8 +938,8 @@ ImageId TextureCache<P>::JoinImages(const ImageInfo& info, GPUVAddr gpu_addr, VA
             return;
         }
         static constexpr bool strict_size = true;
-        const std::optional<OverlapResult> solution =
-            ResolveOverlap(new_info, gpu_addr, cpu_addr, overlap, strict_size, broken_views);
+        const std::optional<OverlapResult> solution = ResolveOverlap(
+            new_info, gpu_addr, cpu_addr, overlap, strict_size, broken_views, native_bgr);
         if (solution) {
             gpu_addr = solution->gpu_addr;
             cpu_addr = solution->cpu_addr;
@@ -946,10 +949,10 @@ ImageId TextureCache<P>::JoinImages(const ImageInfo& info, GPUVAddr gpu_addr, VA
         }
         static constexpr auto options = RelaxedOptions::Size | RelaxedOptions::Format;
         const ImageBase new_image_base(new_info, gpu_addr, cpu_addr);
-        if (IsSubresource(new_info, overlap, gpu_addr, options, broken_views)) {
+        if (IsSubresource(new_info, overlap, gpu_addr, options, broken_views, native_bgr)) {
             left_aliased_ids.push_back(overlap_id);
         } else if (IsSubresource(overlap.info, new_image_base, overlap.gpu_addr, options,
-                                 broken_views)) {
+                                 broken_views, native_bgr)) {
             right_aliased_ids.push_back(overlap_id);
         }
     });
diff --git a/src/video_core/texture_cache/util.cpp b/src/video_core/texture_cache/util.cpp
index a0bc1f7b6..2c42d1449 100644
--- a/src/video_core/texture_cache/util.cpp
+++ b/src/video_core/texture_cache/util.cpp
@@ -1035,13 +1035,13 @@ bool IsPitchLinearSameSize(const ImageInfo& lhs, const ImageInfo& rhs, bool stri
 
 std::optional<OverlapResult> ResolveOverlap(const ImageInfo& new_info, GPUVAddr gpu_addr,
                                             VAddr cpu_addr, const ImageBase& overlap,
-                                            bool strict_size, bool broken_views) {
+                                            bool strict_size, bool broken_views, bool native_bgr) {
     ASSERT(new_info.type != ImageType::Linear);
     ASSERT(overlap.info.type != ImageType::Linear);
     if (!IsLayerStrideCompatible(new_info, overlap.info)) {
         return std::nullopt;
     }
-    if (!IsViewCompatible(overlap.info.format, new_info.format, broken_views)) {
+    if (!IsViewCompatible(overlap.info.format, new_info.format, broken_views, native_bgr)) {
         return std::nullopt;
     }
     if (gpu_addr == overlap.gpu_addr) {
@@ -1085,14 +1085,14 @@ bool IsLayerStrideCompatible(const ImageInfo& lhs, const ImageInfo& rhs) {
 
 std::optional<SubresourceBase> FindSubresource(const ImageInfo& candidate, const ImageBase& image,
                                                GPUVAddr candidate_addr, RelaxedOptions options,
-                                               bool broken_views) {
+                                               bool broken_views, bool native_bgr) {
     const std::optional<SubresourceBase> base = image.TryFindBase(candidate_addr);
     if (!base) {
         return std::nullopt;
     }
     const ImageInfo& existing = image.info;
     if (False(options & RelaxedOptions::Format)) {
-        if (!IsViewCompatible(existing.format, candidate.format, broken_views)) {
+        if (!IsViewCompatible(existing.format, candidate.format, broken_views, native_bgr)) {
             return std::nullopt;
         }
     }
@@ -1129,8 +1129,9 @@ std::optional<SubresourceBase> FindSubresource(const ImageInfo& candidate, const
 }
 
 bool IsSubresource(const ImageInfo& candidate, const ImageBase& image, GPUVAddr candidate_addr,
-                   RelaxedOptions options, bool broken_views) {
-    return FindSubresource(candidate, image, candidate_addr, options, broken_views).has_value();
+                   RelaxedOptions options, bool broken_views, bool native_bgr) {
+    return FindSubresource(candidate, image, candidate_addr, options, broken_views, native_bgr)
+        .has_value();
 }
 
 void DeduceBlitImages(ImageInfo& dst_info, ImageInfo& src_info, const ImageBase* dst,
diff --git a/src/video_core/texture_cache/util.h b/src/video_core/texture_cache/util.h
index 52a9207d6..4d0072867 100644
--- a/src/video_core/texture_cache/util.h
+++ b/src/video_core/texture_cache/util.h
@@ -87,7 +87,8 @@ void SwizzleImage(Tegra::MemoryManager& gpu_memory, GPUVAddr gpu_addr, const Ima
 [[nodiscard]] std::optional<OverlapResult> ResolveOverlap(const ImageInfo& new_info,
                                                           GPUVAddr gpu_addr, VAddr cpu_addr,
                                                           const ImageBase& overlap,
-                                                          bool strict_size, bool broken_views);
+                                                          bool strict_size, bool broken_views,
+                                                          bool native_bgr);
 
 [[nodiscard]] bool IsLayerStrideCompatible(const ImageInfo& lhs, const ImageInfo& rhs);
 
@@ -95,11 +96,11 @@ void SwizzleImage(Tegra::MemoryManager& gpu_memory, GPUVAddr gpu_addr, const Ima
                                                              const ImageBase& image,
                                                              GPUVAddr candidate_addr,
                                                              RelaxedOptions options,
-                                                             bool broken_views);
+                                                             bool broken_views, bool native_bgr);
 
 [[nodiscard]] bool IsSubresource(const ImageInfo& candidate, const ImageBase& image,
-                                 GPUVAddr candidate_addr, RelaxedOptions options,
-                                 bool broken_views);
+                                 GPUVAddr candidate_addr, RelaxedOptions options, bool broken_views,
+                                 bool native_bgr);
 
 void DeduceBlitImages(ImageInfo& dst_info, ImageInfo& src_info, const ImageBase* dst,
                       const ImageBase* src);
diff --git a/src/yuzu/bootmanager.cpp b/src/yuzu/bootmanager.cpp
index 50938fb04..15c09e0ad 100644
--- a/src/yuzu/bootmanager.cpp
+++ b/src/yuzu/bootmanager.cpp
@@ -387,6 +387,25 @@ void GRenderWindow::keyReleaseEvent(QKeyEvent* event) {
     }
 }
 
+MouseInput::MouseButton GRenderWindow::QtButtonToMouseButton(Qt::MouseButton button) {
+    switch (button) {
+    case Qt::LeftButton:
+        return MouseInput::MouseButton::Left;
+    case Qt::RightButton:
+        return MouseInput::MouseButton::Right;
+    case Qt::MiddleButton:
+        return MouseInput::MouseButton::Wheel;
+    case Qt::BackButton:
+        return MouseInput::MouseButton::Backward;
+    case Qt::ForwardButton:
+        return MouseInput::MouseButton::Forward;
+    case Qt::TaskButton:
+        return MouseInput::MouseButton::Task;
+    default:
+        return MouseInput::MouseButton::Extra;
+    }
+}
+
 void GRenderWindow::mousePressEvent(QMouseEvent* event) {
     // Touch input is handled in TouchBeginEvent
     if (event->source() == Qt::MouseEventSynthesizedBySystem) {
@@ -395,7 +414,8 @@ void GRenderWindow::mousePressEvent(QMouseEvent* event) {
 
     auto pos = event->pos();
     const auto [x, y] = ScaleTouch(pos);
-    input_subsystem->GetMouse()->PressButton(x, y, event->button());
+    const auto button = QtButtonToMouseButton(event->button());
+    input_subsystem->GetMouse()->PressButton(x, y, button);
 
     if (event->button() == Qt::LeftButton) {
         this->TouchPressed(x, y, 0);
@@ -429,7 +449,8 @@ void GRenderWindow::mouseReleaseEvent(QMouseEvent* event) {
         return;
     }
 
-    input_subsystem->GetMouse()->ReleaseButton(event->button());
+    const auto button = QtButtonToMouseButton(event->button());
+    input_subsystem->GetMouse()->ReleaseButton(button);
 
     if (event->button() == Qt::LeftButton) {
         this->TouchReleased(0);
diff --git a/src/yuzu/bootmanager.h b/src/yuzu/bootmanager.h
index b5ec7de07..acfe2bc8c 100644
--- a/src/yuzu/bootmanager.h
+++ b/src/yuzu/bootmanager.h
@@ -28,6 +28,10 @@ namespace InputCommon {
 class InputSubsystem;
 }
 
+namespace MouseInput {
+enum class MouseButton;
+}
+
 namespace VideoCore {
 enum class LoadCallbackStage;
 }
@@ -149,6 +153,9 @@ public:
     void keyPressEvent(QKeyEvent* event) override;
     void keyReleaseEvent(QKeyEvent* event) override;
 
+    /// Converts a Qt mouse button into MouseInput mouse button
+    static MouseInput::MouseButton QtButtonToMouseButton(Qt::MouseButton button);
+
     void mousePressEvent(QMouseEvent* event) override;
     void mouseMoveEvent(QMouseEvent* event) override;
     void mouseReleaseEvent(QMouseEvent* event) override;
diff --git a/src/yuzu/configuration/config.cpp b/src/yuzu/configuration/config.cpp
index b319d69fc..1bac57bb2 100644
--- a/src/yuzu/configuration/config.cpp
+++ b/src/yuzu/configuration/config.cpp
@@ -235,7 +235,7 @@ const std::array<UISettings::Shortcut, 17> Config::default_hotkeys{{
     {QStringLiteral("Restart Emulation"),        QStringLiteral("Main Window"), {QStringLiteral("F6"), Qt::WindowShortcut}},
     {QStringLiteral("Stop Emulation"),           QStringLiteral("Main Window"), {QStringLiteral("F5"), Qt::WindowShortcut}},
     {QStringLiteral("Toggle Filter Bar"),        QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F"), Qt::WindowShortcut}},
-    {QStringLiteral("Toggle Mouse Panning"),     QStringLiteral("Main Window"), {QStringLiteral("F9"), Qt::ApplicationShortcut}},
+    {QStringLiteral("Toggle Mouse Panning"),     QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F9"), Qt::ApplicationShortcut}},
     {QStringLiteral("Toggle Speed Limit"),       QStringLiteral("Main Window"), {QStringLiteral("Ctrl+Z"), Qt::ApplicationShortcut}},
     {QStringLiteral("Toggle Status Bar"),        QStringLiteral("Main Window"), {QStringLiteral("Ctrl+S"), Qt::WindowShortcut}},
 }};
@@ -508,7 +508,7 @@ void Config::ReadControlValues() {
 
     Settings::values.emulate_analog_keyboard =
         ReadSetting(QStringLiteral("emulate_analog_keyboard"), false).toBool();
-    Settings::values.mouse_panning = ReadSetting(QStringLiteral("mouse_panning"), false).toBool();
+    Settings::values.mouse_panning = false;
     Settings::values.mouse_panning_sensitivity =
         ReadSetting(QStringLiteral("mouse_panning_sensitivity"), 1).toFloat();
 
@@ -1182,7 +1182,6 @@ void Config::SaveControlValues() {
     WriteSetting(QStringLiteral("keyboard_enabled"), Settings::values.keyboard_enabled, false);
     WriteSetting(QStringLiteral("emulate_analog_keyboard"),
                  Settings::values.emulate_analog_keyboard, false);
-    WriteSetting(QStringLiteral("mouse_panning"), Settings::values.mouse_panning, false);
     WriteSetting(QStringLiteral("mouse_panning_sensitivity"),
                  Settings::values.mouse_panning_sensitivity, 1.0f);
     qt_config->endGroup();
diff --git a/src/yuzu/configuration/configure_input_player.cpp b/src/yuzu/configuration/configure_input_player.cpp
index 55adbd53d..c9318c562 100644
--- a/src/yuzu/configuration/configure_input_player.cpp
+++ b/src/yuzu/configuration/configure_input_player.cpp
@@ -21,6 +21,7 @@
 #include "input_common/mouse/mouse_poller.h"
 #include "input_common/udp/udp.h"
 #include "ui_configure_input_player.h"
+#include "yuzu/bootmanager.h"
 #include "yuzu/configuration/config.h"
 #include "yuzu/configuration/configure_input_player.h"
 #include "yuzu/configuration/configure_input_player_widget.h"
@@ -1362,7 +1363,8 @@ void ConfigureInputPlayer::mousePressEvent(QMouseEvent* event) {
         return;
     }
 
-    input_subsystem->GetMouse()->PressButton(0, 0, event->button());
+    const auto button = GRenderWindow::QtButtonToMouseButton(event->button());
+    input_subsystem->GetMouse()->PressButton(0, 0, button);
 }
 
 void ConfigureInputPlayer::keyPressEvent(QKeyEvent* event) {
diff --git a/src/yuzu/main.cpp b/src/yuzu/main.cpp
index 0ba7c07cc..56d892a31 100644
--- a/src/yuzu/main.cpp
+++ b/src/yuzu/main.cpp
@@ -854,8 +854,7 @@ void GMainWindow::InitializeHotkeys() {
     connect(hotkey_registry.GetHotkey(main_window, QStringLiteral("Toggle Mouse Panning"), this),
             &QShortcut::activated, this, [&] {
                 Settings::values.mouse_panning = !Settings::values.mouse_panning;
-                if (UISettings::values.hide_mouse || Settings::values.mouse_panning) {
-                    mouse_hide_timer.start();
+                if (Settings::values.mouse_panning) {
                     render_window->installEventFilter(render_window);
                     render_window->setAttribute(Qt::WA_Hover, true);
                 }
@@ -1208,11 +1207,14 @@ void GMainWindow::BootGame(const QString& filename, std::size_t program_index) {
     renderer_status_button->setDisabled(true);
 
     if (UISettings::values.hide_mouse || Settings::values.mouse_panning) {
-        mouse_hide_timer.start();
         render_window->installEventFilter(render_window);
         render_window->setAttribute(Qt::WA_Hover, true);
     }
 
+    if (UISettings::values.hide_mouse) {
+        mouse_hide_timer.start();
+    }
+
     std::string title_name;
     std::string title_version;
     const auto res = system.GetGameName(title_name);
@@ -2372,12 +2374,15 @@ void GMainWindow::OnConfigure() {
     if ((UISettings::values.hide_mouse || Settings::values.mouse_panning) && emulation_running) {
         render_window->installEventFilter(render_window);
         render_window->setAttribute(Qt::WA_Hover, true);
-        mouse_hide_timer.start();
     } else {
         render_window->removeEventFilter(render_window);
         render_window->setAttribute(Qt::WA_Hover, false);
     }
 
+    if (UISettings::values.hide_mouse) {
+        mouse_hide_timer.start();
+    }
+
     UpdateStatusButtons();
 }
 
@@ -2615,8 +2620,7 @@ void GMainWindow::UpdateUISettings() {
 }
 
 void GMainWindow::HideMouseCursor() {
-    if (emu_thread == nullptr ||
-        (!UISettings::values.hide_mouse && !Settings::values.mouse_panning)) {
+    if (emu_thread == nullptr && UISettings::values.hide_mouse) {
         mouse_hide_timer.stop();
         ShowMouseCursor();
         return;
@@ -2626,8 +2630,7 @@ void GMainWindow::HideMouseCursor() {
 
 void GMainWindow::ShowMouseCursor() {
     render_window->unsetCursor();
-    if (emu_thread != nullptr &&
-        (UISettings::values.hide_mouse || Settings::values.mouse_panning)) {
+    if (emu_thread != nullptr && UISettings::values.hide_mouse) {
         mouse_hide_timer.start();
     }
 }
diff --git a/src/yuzu_cmd/emu_window/emu_window_sdl2.cpp b/src/yuzu_cmd/emu_window/emu_window_sdl2.cpp
index 7e391ab89..ce8b7c218 100644
--- a/src/yuzu_cmd/emu_window/emu_window_sdl2.cpp
+++ b/src/yuzu_cmd/emu_window/emu_window_sdl2.cpp
@@ -35,18 +35,36 @@ void EmuWindow_SDL2::OnMouseMotion(s32 x, s32 y) {
     input_subsystem->GetMouse()->MouseMove(x, y, 0, 0);
 }
 
+MouseInput::MouseButton EmuWindow_SDL2::SDLButtonToMouseButton(u32 button) const {
+    switch (button) {
+    case SDL_BUTTON_LEFT:
+        return MouseInput::MouseButton::Left;
+    case SDL_BUTTON_RIGHT:
+        return MouseInput::MouseButton::Right;
+    case SDL_BUTTON_MIDDLE:
+        return MouseInput::MouseButton::Wheel;
+    case SDL_BUTTON_X1:
+        return MouseInput::MouseButton::Backward;
+    case SDL_BUTTON_X2:
+        return MouseInput::MouseButton::Forward;
+    default:
+        return MouseInput::MouseButton::Undefined;
+    }
+}
+
 void EmuWindow_SDL2::OnMouseButton(u32 button, u8 state, s32 x, s32 y) {
+    const auto mouse_button = SDLButtonToMouseButton(button);
     if (button == SDL_BUTTON_LEFT) {
         if (state == SDL_PRESSED) {
             TouchPressed((unsigned)std::max(x, 0), (unsigned)std::max(y, 0), 0);
         } else {
             TouchReleased(0);
         }
-    } else if (button == SDL_BUTTON_RIGHT) {
+    } else {
         if (state == SDL_PRESSED) {
-            input_subsystem->GetMouse()->PressButton(x, y, button);
+            input_subsystem->GetMouse()->PressButton(x, y, mouse_button);
         } else {
-            input_subsystem->GetMouse()->ReleaseButton(button);
+            input_subsystem->GetMouse()->ReleaseButton(mouse_button);
         }
     }
 }
diff --git a/src/yuzu_cmd/emu_window/emu_window_sdl2.h b/src/yuzu_cmd/emu_window/emu_window_sdl2.h
index 51a12a6a9..0e17bbca7 100644
--- a/src/yuzu_cmd/emu_window/emu_window_sdl2.h
+++ b/src/yuzu_cmd/emu_window/emu_window_sdl2.h
@@ -18,6 +18,10 @@ namespace InputCommon {
 class InputSubsystem;
 }
 
+namespace MouseInput {
+enum class MouseButton;
+}
+
 class EmuWindow_SDL2 : public Core::Frontend::EmuWindow {
 public:
     explicit EmuWindow_SDL2(InputCommon::InputSubsystem* input_subsystem);
@@ -42,6 +46,9 @@ protected:
     /// Called by WaitEvent when the mouse moves.
     void OnMouseMotion(s32 x, s32 y);
 
+    /// Converts a SDL mouse button into MouseInput mouse button
+    MouseInput::MouseButton SDLButtonToMouseButton(u32 button) const;
+
     /// Called by WaitEvent when a mouse button is pressed or released
     void OnMouseButton(u32 button, u8 state, s32 x, s32 y);