vulkan_common: enable OpenGL interop on other Unices

Merge pull request #6102 from ogniK5377/fd-pass
nvdrv: Pass device fd and handle device create methods for device opening and closing
2021-03-30 00:25:25 +00:00 · 2021-03-29 12:53:55 -07:00 · 2021-03-28 13:59:05 -07:00 · 2021-03-28 12:13:39 -07:00 · 2021-03-27 21:34:20 +00:00 · 2021-03-27 10:29:43 -07:00
90 changed files with 1126 additions and 410 deletions
@@ -27,7 +27,7 @@
    url = https://github.com/ReinUsesLisp/sirit
 [submodule "mbedtls"]
    path = externals/mbedtls
-    url = https://github.com/DarkLordZach/mbedtls
+    url = https://github.com/yuzu-emu/mbedtls
 [submodule "libzip"]
    path = externals/libzip/libzip
    url = https://github.com/nih-at/libzip.git
@@ -1,3 +1,8 @@
+# Ensure libusb compiles with UTF-8 encoding on MSVC
+if(MSVC)
+    add_compile_options(/utf-8)
+endif()
+
 add_library(usb STATIC EXCLUDE_FROM_ALL
    libusb/libusb/core.c
    libusb/libusb/core.c
@@ -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
@@ -11,7 +11,7 @@

 namespace Common {

-constexpr std::size_t default_stack_size = 256 * 1024;
+constexpr std::size_t default_stack_size = 512 * 1024;

 struct Fiber::FiberImpl {
    FiberImpl() : stack{default_stack_size}, rewind_stack{default_stack_size} {}
@@ -116,16 +116,19 @@ void Fiber::Rewind() {
    boost::context::detail::jump_fcontext(impl->rewind_context, this);
 }

-void Fiber::YieldTo(std::shared_ptr<Fiber> from, std::shared_ptr<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.get());
-    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();
+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::shared_ptr<Fiber> Fiber::ThreadToFiber() {
@@ -41,7 +41,7 @@ public:

    /// Yields control from Fiber 'from' to Fiber 'to'
    /// Fiber 'from' must be the currently running fiber.
-    static void YieldTo(std::shared_ptr<Fiber> from, std::shared_ptr<Fiber> to);
+    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);
@@ -306,13 +306,18 @@ void ARM_Dynarmic_32::ClearExclusiveState() {

 void ARM_Dynarmic_32::PageTableChanged(Common::PageTable& page_table,
                                       std::size_t new_address_space_size_in_bits) {
+    ThreadContext32 ctx{};
+    SaveContext(ctx);
+
    auto key = std::make_pair(&page_table, new_address_space_size_in_bits);
    auto iter = jit_cache.find(key);
    if (iter != jit_cache.end()) {
        jit = iter->second;
+        LoadContext(ctx);
        return;
    }
    jit = MakeJit(page_table, new_address_space_size_in_bits);
+    LoadContext(ctx);
    jit_cache.emplace(key, jit);
 }

@@ -348,13 +348,18 @@ void ARM_Dynarmic_64::ClearExclusiveState() {

 void ARM_Dynarmic_64::PageTableChanged(Common::PageTable& page_table,
                                       std::size_t new_address_space_size_in_bits) {
+    ThreadContext64 ctx{};
+    SaveContext(ctx);
+
    auto key = std::make_pair(&page_table, new_address_space_size_in_bits);
    auto iter = jit_cache.find(key);
    if (iter != jit_cache.end()) {
        jit = iter->second;
+        LoadContext(ctx);
        return;
    }
    jit = MakeJit(page_table, new_address_space_size_in_bits);
+    LoadContext(ctx);
    jit_cache.emplace(key, jit);
 }

@@ -299,28 +299,17 @@ struct System::Impl {
            gpu_core->WaitIdle();
        }

-        // Shutdown emulation session
        services.reset();
        service_manager.reset();
        cheat_engine.reset();
        telemetry_session.reset();
-
-        // Close all CPU/threading state
        cpu_manager.Shutdown();
-
-        // Release the Time Manager's resources
        time_manager.Shutdown();
-
-        // Shutdown kernel and core timing
        core_timing.Shutdown();
-        kernel.Shutdown();
-
-        // Close app loader
        app_loader.reset();
        gpu_core.reset();
        perf_stats.reset();
-
-        // Clear all applets
+        kernel.Shutdown();
        applet_manager.ClearAll();

        LOG_DEBUG(Core, "Shutdown OK");
@@ -148,7 +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);
+        Common::Fiber::YieldTo(current_thread->GetHostContext(), *core_data[core].host_context);
        ASSERT(scheduler.ContextSwitchPending());
        ASSERT(core == kernel.GetCurrentHostThreadID());
        scheduler.RescheduleCurrentCore();
@@ -245,7 +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);
+        Common::Fiber::YieldTo(current_thread->GetHostContext(), *core_data[0].host_context);
        ASSERT(scheduler.ContextSwitchPending());
        ASSERT(core == kernel.GetCurrentHostThreadID());
        scheduler.RescheduleCurrentCore();
@@ -271,7 +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
@@ -363,7 +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, current_thread->GetHostContext());
+        Common::Fiber::YieldTo(data.host_context, *current_thread->GetHostContext());
        data.is_running = false;
        data.is_paused = true;
        data.exit_barrier->Wait();
@@ -105,8 +105,6 @@ void AESCipher<Key, KeySize>::Transcode(const u8* src, std::size_t size, u8* des
            }
        }
    }
-
-    mbedtls_cipher_finish(context, nullptr, nullptr);
 }

 template <typename Key, std::size_t KeySize>
@@ -14,7 +14,7 @@ namespace NgWord1Data {
 constexpr std::size_t NUMBER_WORD_TXT_FILES = 0x10;

 // Should this archive replacement mysteriously not work on a future game, consider updating.
-constexpr std::array<u8, 4> VERSION_DAT{0x0, 0x0, 0x0, 0x19}; // 5.1.0 System Version
+constexpr std::array<u8, 4> VERSION_DAT{0x0, 0x0, 0x0, 0x20}; // 11.0.1 System Version

 constexpr std::array<u8, 30> WORD_TXT{
    0xFE, 0xFF, 0x00, 0x5E, 0x00, 0x76, 0x00, 0x65, 0x00, 0x72, 0x00, 0x79, 0x00, 0x62, 0x00,
@@ -43,7 +43,7 @@ namespace NgWord2Data {
 constexpr std::size_t NUMBER_AC_NX_FILES = 0x10;

 // Should this archive replacement mysteriously not work on a future game, consider updating.
-constexpr std::array<u8, 4> VERSION_DAT{0x0, 0x0, 0x0, 0x15}; // 5.1.0 System Version
+constexpr std::array<u8, 4> VERSION_DAT{0x0, 0x0, 0x0, 0x1A}; // 11.0.1 System Version

 constexpr std::array<u8, 0x2C> AC_NX_DATA{
    0x1F, 0x8B, 0x08, 0x08, 0xD5, 0x2C, 0x09, 0x5C, 0x04, 0x00, 0x61, 0x63, 0x72, 0x61, 0x77,
@@ -14,15 +14,15 @@ namespace SystemVersionData {

 constexpr u8 VERSION_MAJOR = 11;
 constexpr u8 VERSION_MINOR = 0;
-constexpr u8 VERSION_MICRO = 0;
+constexpr u8 VERSION_MICRO = 1;

-constexpr u8 REVISION_MAJOR = 5;
+constexpr u8 REVISION_MAJOR = 1;
 constexpr u8 REVISION_MINOR = 0;

 constexpr char PLATFORM_STRING[] = "NX";
-constexpr char VERSION_HASH[] = "34197eba8810e2edd5e9dfcfbde7b340882e856d";
-constexpr char DISPLAY_VERSION[] = "11.0.0";
-constexpr char DISPLAY_TITLE[] = "NintendoSDK Firmware for NX 11.0.0-5.0";
+constexpr char VERSION_HASH[] = "69103fcb2004dace877094c2f8c29e6113be5dbf";
+constexpr char DISPLAY_VERSION[] = "11.0.1";
+constexpr char DISPLAY_TITLE[] = "NintendoSDK Firmware for NX 11.0.1-1.0";

 } // namespace SystemVersionData

@@ -734,7 +734,7 @@ void KScheduler::ScheduleImpl() {
    }
    guard.unlock();

-    Common::Fiber::YieldTo(*old_context, switch_fiber);
+    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;
                }
            }
-            std::shared_ptr<Common::Fiber>* next_context;
-            if (next_thread != nullptr) {
-                next_context = &next_thread->GetHostContext();
-            } else {
-                next_context = &idle_thread->GetHostContext();
-            }
-            Common::Fiber::YieldTo(switch_fiber, *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();
 }

@@ -995,22 +995,11 @@ 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*)> 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);
-}
-
-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);
@@ -1027,12 +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_shared<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();
 }
@@ -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;
    }
@@ -68,9 +68,9 @@ struct KernelCore::Impl {
        InitializePhysicalCores();
        InitializeSystemResourceLimit(kernel, system);
        InitializeMemoryLayout();
-        InitializePreemption(kernel);
        InitializeSchedulers();
        InitializeSuspendThreads();
+        InitializePreemption(kernel);
    }

    void InitializeCores() {
@@ -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();
    }
@@ -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();

@@ -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);
@@ -133,7 +133,7 @@ private:
    void GetBlockedUserListIds(Kernel::HLERequestContext& ctx) {
        // This is safe to stub, as there should be no adverse consequences from reporting no
        // blocked users.
-        LOG_WARNING(Service_ACC, "(STUBBED) called");
+        LOG_WARNING(Service_Friend, "(STUBBED) called");
        IPC::ResponseBuilder rb{ctx, 3};
        rb.Push(RESULT_SUCCESS);
        rb.Push<u32>(0); // Indicates there are no blocked users
@@ -141,14 +141,14 @@ private:

    void DeclareCloseOnlinePlaySession(Kernel::HLERequestContext& ctx) {
        // Stub used by Splatoon 2
-        LOG_WARNING(Service_ACC, "(STUBBED) called");
+        LOG_WARNING(Service_Friend, "(STUBBED) called");
        IPC::ResponseBuilder rb{ctx, 2};
        rb.Push(RESULT_SUCCESS);
    }

    void UpdateUserPresence(Kernel::HLERequestContext& ctx) {
        // Stub used by Retro City Rampage
-        LOG_WARNING(Service_ACC, "(STUBBED) called");
+        LOG_WARNING(Service_Friend, "(STUBBED) called");
        IPC::ResponseBuilder rb{ctx, 2};
        rb.Push(RESULT_SUCCESS);
    }
@@ -159,7 +159,7 @@ private:
        const auto uuid = rp.PopRaw<Common::UUID>();
        [[maybe_unused]] const auto filter = rp.PopRaw<SizedFriendFilter>();
        const auto pid = rp.Pop<u64>();
-        LOG_WARNING(Service_ACC, "(STUBBED) called, offset={}, uuid={}, pid={}", friend_offset,
+        LOG_WARNING(Service_Friend, "(STUBBED) called, offset={}, uuid={}, pid={}", friend_offset,
                    uuid.Format(), pid);

        IPC::ResponseBuilder rb{ctx, 3};
@@ -191,7 +191,7 @@ public:

 private:
    void GetEvent(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_ACC, "called");
+        LOG_DEBUG(Service_Friend, "called");

        IPC::ResponseBuilder rb{ctx, 2, 1};
        rb.Push(RESULT_SUCCESS);
@@ -199,7 +199,7 @@ private:
    }

    void Clear(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_ACC, "called");
+        LOG_DEBUG(Service_Friend, "called");
        while (!notifications.empty()) {
            notifications.pop();
        }
@@ -210,10 +210,10 @@ private:
    }

    void Pop(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_ACC, "called");
+        LOG_DEBUG(Service_Friend, "called");

        if (notifications.empty()) {
-            LOG_ERROR(Service_ACC, "No notifications in queue!");
+            LOG_ERROR(Service_Friend, "No notifications in queue!");
            IPC::ResponseBuilder rb{ctx, 2};
            rb.Push(ERR_NO_NOTIFICATIONS);
            return;
@@ -231,7 +231,8 @@ private:
            break;
        default:
            // HOS seems not have an error case for an unknown notification
-            LOG_WARNING(Service_ACC, "Unknown notification {:08X}", notification.notification_type);
+            LOG_WARNING(Service_Friend, "Unknown notification {:08X}",
+                        notification.notification_type);
            break;
        }

@@ -269,14 +270,14 @@ void Module::Interface::CreateFriendService(Kernel::HLERequestContext& ctx) {
    IPC::ResponseBuilder rb{ctx, 2, 0, 1};
    rb.Push(RESULT_SUCCESS);
    rb.PushIpcInterface<IFriendService>(system);
-    LOG_DEBUG(Service_ACC, "called");
+    LOG_DEBUG(Service_Friend, "called");
 }

 void Module::Interface::CreateNotificationService(Kernel::HLERequestContext& ctx) {
    IPC::RequestParser rp{ctx};
    auto uuid = rp.PopRaw<Common::UUID>();

-    LOG_DEBUG(Service_ACC, "called, uuid={}", uuid.Format());
+    LOG_DEBUG(Service_Friend, "called, uuid={}", uuid.Format());

    IPC::ResponseBuilder rb{ctx, 2, 0, 1};
    rb.Push(RESULT_SUCCESS);
@@ -31,7 +31,7 @@ public:
     * @param output A buffer where the output data will be written to.
     * @returns The result code of the ioctl.
     */
-    virtual NvResult Ioctl1(Ioctl command, const std::vector<u8>& input,
+    virtual NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                            std::vector<u8>& output) = 0;

    /**
@@ -42,7 +42,7 @@ public:
     * @param output A buffer where the output data will be written to.
     * @returns The result code of the ioctl.
     */
-    virtual NvResult Ioctl2(Ioctl command, const std::vector<u8>& input,
+    virtual NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                            const std::vector<u8>& inline_input, std::vector<u8>& output) = 0;

    /**
@@ -53,8 +53,20 @@ public:
     * @param inline_output A buffer where the inlined output data will be written to.
     * @returns The result code of the ioctl.
     */
-    virtual NvResult Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                            std::vector<u8>& inline_output) = 0;
+    virtual NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                            std::vector<u8>& output, std::vector<u8>& inline_output) = 0;
+
+    /**
+     * Called once a device is openned
+     * @param fd The device fd
+     */
+    virtual void OnOpen(DeviceFD fd) = 0;
+
+    /**
+     * Called once a device is closed
+     * @param fd The device fd
+     */
+    virtual void OnClose(DeviceFD fd) = 0;

 protected:
    Core::System& system;
@@ -18,24 +18,27 @@ nvdisp_disp0::nvdisp_disp0(Core::System& system, std::shared_ptr<nvmap> nvmap_de
    : nvdevice(system), nvmap_dev(std::move(nvmap_dev)) {}
 nvdisp_disp0 ::~nvdisp_disp0() = default;

-NvResult nvdisp_disp0::Ioctl1(Ioctl command, const std::vector<u8>& input,
+NvResult nvdisp_disp0::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                              std::vector<u8>& output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

-NvResult nvdisp_disp0::Ioctl2(Ioctl command, const std::vector<u8>& input,
+NvResult nvdisp_disp0::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                              const std::vector<u8>& inline_input, std::vector<u8>& output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

-NvResult nvdisp_disp0::Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                              std::vector<u8>& inline_output) {
+NvResult nvdisp_disp0::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                              std::vector<u8>& output, std::vector<u8>& inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

+void nvdisp_disp0::OnOpen(DeviceFD fd) {}
+void nvdisp_disp0::OnClose(DeviceFD fd) {}
+
 void nvdisp_disp0::flip(u32 buffer_handle, u32 offset, u32 format, u32 width, u32 height,
                        u32 stride, NVFlinger::BufferQueue::BufferTransformFlags transform,
                        const Common::Rectangle<int>& crop_rect) {
@@ -20,11 +20,15 @@ public:
    explicit nvdisp_disp0(Core::System& system, std::shared_ptr<nvmap> nvmap_dev);
    ~nvdisp_disp0() override;

-    NvResult Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) override;
-    NvResult Ioctl2(Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                    std::vector<u8>& inline_output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+
+    void OnOpen(DeviceFD fd) override;
+    void OnClose(DeviceFD fd) override;

    /// Performs a screen flip, drawing the buffer pointed to by the handle.
    void flip(u32 buffer_handle, u32 offset, u32 format, u32 width, u32 height, u32 stride,
@@ -21,7 +21,7 @@ nvhost_as_gpu::nvhost_as_gpu(Core::System& system, std::shared_ptr<nvmap> nvmap_
    : nvdevice(system), nvmap_dev(std::move(nvmap_dev)) {}
 nvhost_as_gpu::~nvhost_as_gpu() = default;

-NvResult nvhost_as_gpu::Ioctl1(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_as_gpu::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                               std::vector<u8>& output) {
    switch (command.group) {
    case 'A':
@@ -39,7 +39,7 @@ NvResult nvhost_as_gpu::Ioctl1(Ioctl command, const std::vector<u8>& input,
        case 0x8:
            return GetVARegions(input, output);
        case 0x9:
-            return InitalizeEx(input, output);
+            return AllocAsEx(input, output);
        case 0x14:
            return Remap(input, output);
        default:
@@ -54,14 +54,14 @@ NvResult nvhost_as_gpu::Ioctl1(Ioctl command, const std::vector<u8>& input,
    return NvResult::NotImplemented;
 }

-NvResult nvhost_as_gpu::Ioctl2(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_as_gpu::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                               const std::vector<u8>& inline_input, std::vector<u8>& output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

-NvResult nvhost_as_gpu::Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                               std::vector<u8>& inline_output) {
+NvResult nvhost_as_gpu::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                               std::vector<u8>& output, std::vector<u8>& inline_output) {
    switch (command.group) {
    case 'A':
        switch (command.cmd) {
@@ -78,11 +78,19 @@ NvResult nvhost_as_gpu::Ioctl3(Ioctl command, const std::vector<u8>& input, std:
    return NvResult::NotImplemented;
 }

-NvResult nvhost_as_gpu::InitalizeEx(const std::vector<u8>& input, std::vector<u8>& output) {
-    IoctlInitalizeEx params{};
+void nvhost_as_gpu::OnOpen(DeviceFD fd) {}
+void nvhost_as_gpu::OnClose(DeviceFD fd) {}
+
+NvResult nvhost_as_gpu::AllocAsEx(const std::vector<u8>& input, std::vector<u8>& output) {
+    IoctlAllocAsEx params{};
    std::memcpy(&params, input.data(), input.size());

    LOG_WARNING(Service_NVDRV, "(STUBBED) called, big_page_size=0x{:X}", params.big_page_size);
+    if (params.big_page_size == 0) {
+        params.big_page_size = DEFAULT_BIG_PAGE_SIZE;
+    }
+
+    big_page_size = params.big_page_size;

    return NvResult::Success;
 }
@@ -276,13 +284,18 @@ NvResult nvhost_as_gpu::GetVARegions(const std::vector<u8>& input, std::vector<u
                params.buf_size);

    params.buf_size = 0x30;
-    params.regions[0].offset = 0x04000000;
-    params.regions[0].page_size = 0x1000;
-    params.regions[0].pages = 0x3fbfff;

-    params.regions[1].offset = 0x04000000;
-    params.regions[1].page_size = 0x10000;
-    params.regions[1].pages = 0x1bffff;
+    params.small = IoctlVaRegion{
+        .offset = 0x04000000,
+        .page_size = DEFAULT_SMALL_PAGE_SIZE,
+        .pages = 0x3fbfff,
+    };
+
+    params.big = IoctlVaRegion{
+        .offset = 0x04000000,
+        .page_size = big_page_size,
+        .pages = 0x1bffff,
+    };

    // TODO(ogniK): This probably can stay stubbed but should add support way way later

@@ -299,18 +312,25 @@ NvResult nvhost_as_gpu::GetVARegions(const std::vector<u8>& input, std::vector<u
                params.buf_size);

    params.buf_size = 0x30;
-    params.regions[0].offset = 0x04000000;
-    params.regions[0].page_size = 0x1000;
-    params.regions[0].pages = 0x3fbfff;

-    params.regions[1].offset = 0x04000000;
-    params.regions[1].page_size = 0x10000;
-    params.regions[1].pages = 0x1bffff;
+    params.small = IoctlVaRegion{
+        .offset = 0x04000000,
+        .page_size = 0x1000,
+        .pages = 0x3fbfff,
+    };
+
+    params.big = IoctlVaRegion{
+        .offset = 0x04000000,
+        .page_size = big_page_size,
+        .pages = 0x1bffff,
+    };

    // TODO(ogniK): This probably can stay stubbed but should add support way way later

    std::memcpy(output.data(), &params, output.size());
-    std::memcpy(inline_output.data(), &params.regions, inline_output.size());
+    std::memcpy(inline_output.data(), &params.small, sizeof(IoctlVaRegion));
+    std::memcpy(inline_output.data() + sizeof(IoctlVaRegion), &params.big, sizeof(IoctlVaRegion));
+
    return NvResult::Success;
 }

@@ -16,6 +16,9 @@

 namespace Service::Nvidia::Devices {

+constexpr u32 DEFAULT_BIG_PAGE_SIZE = 1 << 16;
+constexpr u32 DEFAULT_SMALL_PAGE_SIZE = 1 << 12;
+
 class nvmap;

 enum class AddressSpaceFlags : u32 {
@@ -30,11 +33,15 @@ public:
    explicit nvhost_as_gpu(Core::System& system, std::shared_ptr<nvmap> nvmap_dev);
    ~nvhost_as_gpu() override;

-    NvResult Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) override;
-    NvResult Ioctl2(Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                    std::vector<u8>& inline_output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+
+    void OnOpen(DeviceFD fd) override;
+    void OnClose(DeviceFD fd) override;

 private:
    class BufferMap final {
@@ -76,16 +83,16 @@ private:
        bool is_allocated{};
    };

-    struct IoctlInitalizeEx {
-        u32_le big_page_size{}; // depends on GPU's available_big_page_sizes; 0=default
-        s32_le as_fd{};         // ignored; passes 0
-        u32_le flags{};         // passes 0
-        u32_le reserved{};      // ignored; passes 0
-        u64_le unk0{};
-        u64_le unk1{};
-        u64_le unk2{};
+    struct IoctlAllocAsEx {
+        u32_le flags{}; // usually passes 1
+        s32_le as_fd{}; // ignored; passes 0
+        u32_le big_page_size{};
+        u32_le reserved{}; // ignored; passes 0
+        u64_le va_range_start{};
+        u64_le va_range_end{};
+        u64_le va_range_split{};
    };
-    static_assert(sizeof(IoctlInitalizeEx) == 40, "IoctlInitalizeEx is incorrect size");
+    static_assert(sizeof(IoctlAllocAsEx) == 40, "IoctlAllocAsEx is incorrect size");

    struct IoctlAllocSpace {
        u32_le pages{};
@@ -149,14 +156,16 @@ private:
        u64_le buf_addr{}; // (contained output user ptr on linux, ignored)
        u32_le buf_size{}; // forced to 2*sizeof(struct va_region)
        u32_le reserved{};
-        IoctlVaRegion regions[2]{};
+        IoctlVaRegion small{};
+        IoctlVaRegion big{};
    };
    static_assert(sizeof(IoctlGetVaRegions) == 16 + sizeof(IoctlVaRegion) * 2,
                  "IoctlGetVaRegions is incorrect size");

    s32 channel{};
+    u32 big_page_size{DEFAULT_BIG_PAGE_SIZE};

-    NvResult InitalizeEx(const std::vector<u8>& input, std::vector<u8>& output);
+    NvResult AllocAsEx(const std::vector<u8>& input, std::vector<u8>& output);
    NvResult AllocateSpace(const std::vector<u8>& input, std::vector<u8>& output);
    NvResult Remap(const std::vector<u8>& input, std::vector<u8>& output);
    NvResult MapBufferEx(const std::vector<u8>& input, std::vector<u8>& output);
@@ -20,7 +20,8 @@ nvhost_ctrl::nvhost_ctrl(Core::System& system, EventInterface& events_interface,
    : nvdevice(system), events_interface{events_interface}, syncpoint_manager{syncpoint_manager} {}
 nvhost_ctrl::~nvhost_ctrl() = default;

-NvResult nvhost_ctrl::Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                             std::vector<u8>& output) {
    switch (command.group) {
    case 0x0:
        switch (command.cmd) {
@@ -46,18 +47,21 @@ NvResult nvhost_ctrl::Ioctl1(Ioctl command, const std::vector<u8>& input, std::v
    return NvResult::NotImplemented;
 }

-NvResult nvhost_ctrl::Ioctl2(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_ctrl::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                             const std::vector<u8>& inline_input, std::vector<u8>& output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

-NvResult nvhost_ctrl::Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                             std::vector<u8>& inline_outpu) {
+NvResult nvhost_ctrl::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                             std::vector<u8>& output, std::vector<u8>& inline_outpu) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

+void nvhost_ctrl::OnOpen(DeviceFD fd) {}
+void nvhost_ctrl::OnClose(DeviceFD fd) {}
+
 NvResult nvhost_ctrl::NvOsGetConfigU32(const std::vector<u8>& input, std::vector<u8>& output) {
    IocGetConfigParams params{};
    std::memcpy(&params, input.data(), sizeof(params));
@@ -18,11 +18,15 @@ public:
                         SyncpointManager& syncpoint_manager);
    ~nvhost_ctrl() override;

-    NvResult Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) override;
-    NvResult Ioctl2(Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                    std::vector<u8>& inline_output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+
+    void OnOpen(DeviceFD fd) override;
+    void OnClose(DeviceFD fd) override;

 private:
    struct IocSyncptReadParams {
@@ -15,7 +15,7 @@ namespace Service::Nvidia::Devices {
 nvhost_ctrl_gpu::nvhost_ctrl_gpu(Core::System& system) : nvdevice(system) {}
 nvhost_ctrl_gpu::~nvhost_ctrl_gpu() = default;

-NvResult nvhost_ctrl_gpu::Ioctl1(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_ctrl_gpu::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                                 std::vector<u8>& output) {
    switch (command.group) {
    case 'G':
@@ -47,13 +47,13 @@ NvResult nvhost_ctrl_gpu::Ioctl1(Ioctl command, const std::vector<u8>& input,
    return NvResult::NotImplemented;
 }

-NvResult nvhost_ctrl_gpu::Ioctl2(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_ctrl_gpu::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                                 const std::vector<u8>& inline_input, std::vector<u8>& output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

-NvResult nvhost_ctrl_gpu::Ioctl3(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_ctrl_gpu::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                                 std::vector<u8>& output, std::vector<u8>& inline_output) {
    switch (command.group) {
    case 'G':
@@ -73,6 +73,9 @@ NvResult nvhost_ctrl_gpu::Ioctl3(Ioctl command, const std::vector<u8>& input,
    return NvResult::NotImplemented;
 }

+void nvhost_ctrl_gpu::OnOpen(DeviceFD fd) {}
+void nvhost_ctrl_gpu::OnClose(DeviceFD fd) {}
+
 NvResult nvhost_ctrl_gpu::GetCharacteristics(const std::vector<u8>& input,
                                             std::vector<u8>& output) {
    LOG_DEBUG(Service_NVDRV, "called");
@@ -16,11 +16,15 @@ public:
    explicit nvhost_ctrl_gpu(Core::System& system);
    ~nvhost_ctrl_gpu() override;

-    NvResult Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) override;
-    NvResult Ioctl2(Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                    std::vector<u8>& inline_output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+
+    void OnOpen(DeviceFD fd) override;
+    void OnClose(DeviceFD fd) override;

 private:
    struct IoctlGpuCharacteristics {
@@ -23,7 +23,8 @@ nvhost_gpu::nvhost_gpu(Core::System& system, std::shared_ptr<nvmap> nvmap_dev,

 nvhost_gpu::~nvhost_gpu() = default;

-NvResult nvhost_gpu::Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                            std::vector<u8>& output) {
    switch (command.group) {
    case 0x0:
        switch (command.cmd) {
@@ -74,7 +75,7 @@ NvResult nvhost_gpu::Ioctl1(Ioctl command, const std::vector<u8>& input, std::ve
    return NvResult::NotImplemented;
 };

-NvResult nvhost_gpu::Ioctl2(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_gpu::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                            const std::vector<u8>& inline_input, std::vector<u8>& output) {
    switch (command.group) {
    case 'H':
@@ -88,12 +89,15 @@ NvResult nvhost_gpu::Ioctl2(Ioctl command, const std::vector<u8>& input,
    return NvResult::NotImplemented;
 }

-NvResult nvhost_gpu::Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                            std::vector<u8>& inline_output) {
+NvResult nvhost_gpu::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                            std::vector<u8>& output, std::vector<u8>& inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

+void nvhost_gpu::OnOpen(DeviceFD fd) {}
+void nvhost_gpu::OnClose(DeviceFD fd) {}
+
 NvResult nvhost_gpu::SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& output) {
    IoctlSetNvmapFD params{};
    std::memcpy(&params, input.data(), input.size());
@@ -26,11 +26,15 @@ public:
                        SyncpointManager& syncpoint_manager);
    ~nvhost_gpu() override;

-    NvResult Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) override;
-    NvResult Ioctl2(Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                    std::vector<u8>& inline_output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+
+    void OnOpen(DeviceFD fd) override;
+    void OnClose(DeviceFD fd) override;

 private:
    enum class CtxObjects : u32_le {
@@ -16,7 +16,7 @@ nvhost_nvdec::nvhost_nvdec(Core::System& system, std::shared_ptr<nvmap> nvmap_de
    : nvhost_nvdec_common(system, std::move(nvmap_dev), syncpoint_manager) {}
 nvhost_nvdec::~nvhost_nvdec() = default;

-NvResult nvhost_nvdec::Ioctl1(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_nvdec::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                              std::vector<u8>& output) {
    switch (command.group) {
    case 0x0:
@@ -57,16 +57,19 @@ NvResult nvhost_nvdec::Ioctl1(Ioctl command, const std::vector<u8>& input,
    return NvResult::NotImplemented;
 }

-NvResult nvhost_nvdec::Ioctl2(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_nvdec::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                              const std::vector<u8>& inline_input, std::vector<u8>& output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

-NvResult nvhost_nvdec::Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                              std::vector<u8>& inline_output) {
+NvResult nvhost_nvdec::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                              std::vector<u8>& output, std::vector<u8>& inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

+void nvhost_nvdec::OnOpen(DeviceFD fd) {}
+void nvhost_nvdec::OnClose(DeviceFD fd) {}
+
 } // namespace Service::Nvidia::Devices
@@ -15,11 +15,15 @@ public:
                          SyncpointManager& syncpoint_manager);
    ~nvhost_nvdec() override;

-    NvResult Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) override;
-    NvResult Ioctl2(Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                    std::vector<u8>& inline_output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+
+    void OnOpen(DeviceFD fd) override;
+    void OnClose(DeviceFD fd) override;
 };

 } // namespace Service::Nvidia::Devices
@@ -13,7 +13,7 @@ namespace Service::Nvidia::Devices {
 nvhost_nvjpg::nvhost_nvjpg(Core::System& system) : nvdevice(system) {}
 nvhost_nvjpg::~nvhost_nvjpg() = default;

-NvResult nvhost_nvjpg::Ioctl1(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_nvjpg::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                              std::vector<u8>& output) {
    switch (command.group) {
    case 'H':
@@ -32,18 +32,21 @@ NvResult nvhost_nvjpg::Ioctl1(Ioctl command, const std::vector<u8>& input,
    return NvResult::NotImplemented;
 }

-NvResult nvhost_nvjpg::Ioctl2(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_nvjpg::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                              const std::vector<u8>& inline_input, std::vector<u8>& output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

-NvResult nvhost_nvjpg::Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                              std::vector<u8>& inline_output) {
+NvResult nvhost_nvjpg::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                              std::vector<u8>& output, std::vector<u8>& inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

+void nvhost_nvjpg::OnOpen(DeviceFD fd) {}
+void nvhost_nvjpg::OnClose(DeviceFD fd) {}
+
 NvResult nvhost_nvjpg::SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& output) {
    IoctlSetNvmapFD params{};
    std::memcpy(&params, input.data(), input.size());
@@ -16,11 +16,15 @@ public:
    explicit nvhost_nvjpg(Core::System& system);
    ~nvhost_nvjpg() override;

-    NvResult Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) override;
-    NvResult Ioctl2(Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                    std::vector<u8>& inline_output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+
+    void OnOpen(DeviceFD fd) override;
+    void OnClose(DeviceFD fd) override;

 private:
    struct IoctlSetNvmapFD {
@@ -16,7 +16,8 @@ nvhost_vic::nvhost_vic(Core::System& system, std::shared_ptr<nvmap> nvmap_dev,

 nvhost_vic::~nvhost_vic() = default;

-NvResult nvhost_vic::Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_vic::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                            std::vector<u8>& output) {
    switch (command.group) {
    case 0x0:
        switch (command.cmd) {
@@ -55,16 +56,19 @@ NvResult nvhost_vic::Ioctl1(Ioctl command, const std::vector<u8>& input, std::ve
    return NvResult::NotImplemented;
 }

-NvResult nvhost_vic::Ioctl2(Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_vic::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                            const std::vector<u8>& inline_input, std::vector<u8>& output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

-NvResult nvhost_vic::Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                            std::vector<u8>& inline_output) {
+NvResult nvhost_vic::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                            std::vector<u8>& output, std::vector<u8>& inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

+void nvhost_vic::OnOpen(DeviceFD fd) {}
+void nvhost_vic::OnClose(DeviceFD fd) {}
+
 } // namespace Service::Nvidia::Devices
@@ -14,10 +14,14 @@ public:
                        SyncpointManager& syncpoint_manager);
    ~nvhost_vic();

-    NvResult Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) override;
-    NvResult Ioctl2(Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                    std::vector<u8>& inline_output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+
+    void OnOpen(DeviceFD fd) override;
+    void OnClose(DeviceFD fd) override;
 };
 } // namespace Service::Nvidia::Devices
@@ -19,7 +19,8 @@ nvmap::nvmap(Core::System& system) : nvdevice(system) {

 nvmap::~nvmap() = default;

-NvResult nvmap::Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvmap::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                       std::vector<u8>& output) {
    switch (command.group) {
    case 0x1:
        switch (command.cmd) {
@@ -47,18 +48,21 @@ NvResult nvmap::Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<
    return NvResult::NotImplemented;
 }

-NvResult nvmap::Ioctl2(Ioctl command, const std::vector<u8>& input,
+NvResult nvmap::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                       const std::vector<u8>& inline_input, std::vector<u8>& output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

-NvResult nvmap::Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                       std::vector<u8>& inline_output) {
+NvResult nvmap::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                       std::vector<u8>& output, std::vector<u8>& inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }

+void nvmap::OnOpen(DeviceFD fd) {}
+void nvmap::OnClose(DeviceFD fd) {}
+
 VAddr nvmap::GetObjectAddress(u32 handle) const {
    auto object = GetObject(handle);
    ASSERT(object);
@@ -19,11 +19,15 @@ public:
    explicit nvmap(Core::System& system);
    ~nvmap() override;

-    NvResult Ioctl1(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) override;
-    NvResult Ioctl2(Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output,
-                    std::vector<u8>& inline_output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+
+    void OnOpen(DeviceFD fd) override;
+    void OnClose(DeviceFD fd) override;

    /// Returns the allocated address of an nvmap object given its handle.
    VAddr GetObjectAddress(u32 handle) const;
@@ -89,6 +89,8 @@ DeviceFD Module::Open(const std::string& device_name) {
    auto device = devices[device_name];
    const DeviceFD fd = next_fd++;

+    device->OnOpen(fd);
+
    open_files[fd] = std::move(device);

    return fd;
@@ -108,7 +110,7 @@ NvResult Module::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input
        return NvResult::NotImplemented;
    }

-    return itr->second->Ioctl1(command, input, output);
+    return itr->second->Ioctl1(fd, command, input, output);
 }

 NvResult Module::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
@@ -125,7 +127,7 @@ NvResult Module::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input
        return NvResult::NotImplemented;
    }

-    return itr->second->Ioctl2(command, input, inline_input, output);
+    return itr->second->Ioctl2(fd, command, input, inline_input, output);
 }

 NvResult Module::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
@@ -142,7 +144,7 @@ NvResult Module::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input
        return NvResult::NotImplemented;
    }

-    return itr->second->Ioctl3(command, input, output, inline_output);
+    return itr->second->Ioctl3(fd, command, input, output, inline_output);
 }

 NvResult Module::Close(DeviceFD fd) {
@@ -158,6 +160,8 @@ NvResult Module::Close(DeviceFD fd) {
        return NvResult::NotImplemented;
    }

+    itr->second->OnClose(fd);
+
    open_files.erase(itr);

    return NvResult::Success;
@@ -42,7 +42,9 @@ void BSD::PollWork::Execute(BSD* bsd) {
 }

 void BSD::PollWork::Response(Kernel::HLERequestContext& ctx) {
-    ctx.WriteBuffer(write_buffer);
+    if (write_buffer.size() > 0) {
+        ctx.WriteBuffer(write_buffer);
+    }

    IPC::ResponseBuilder rb{ctx, 4};
    rb.Push(RESULT_SUCCESS);
@@ -55,7 +57,9 @@ void BSD::AcceptWork::Execute(BSD* bsd) {
 }

 void BSD::AcceptWork::Response(Kernel::HLERequestContext& ctx) {
-    ctx.WriteBuffer(write_buffer);
+    if (write_buffer.size() > 0) {
+        ctx.WriteBuffer(write_buffer);
+    }

    IPC::ResponseBuilder rb{ctx, 5};
    rb.Push(RESULT_SUCCESS);
@@ -43,6 +43,11 @@ void InstallInterfaces(SM::ServiceManager& service_manager, Core::System& system
    auto module = std::make_shared<Module>();
    std::make_shared<CSRNG>(system, module)->InstallAsService(service_manager);
    std::make_shared<SPL>(system, module)->InstallAsService(service_manager);
+    std::make_shared<SPL_MIG>(system, module)->InstallAsService(service_manager);
+    std::make_shared<SPL_FS>(system, module)->InstallAsService(service_manager);
+    std::make_shared<SPL_SSL>(system, module)->InstallAsService(service_manager);
+    std::make_shared<SPL_ES>(system, module)->InstallAsService(service_manager);
+    std::make_shared<SPL_MANU>(system, module)->InstallAsService(service_manager);
 }

 } // namespace Service::SPL
@@ -8,6 +8,79 @@ namespace Service::SPL {

 SPL::SPL(Core::System& system_, std::shared_ptr<Module> module_)
    : Interface(system_, std::move(module_), "spl:") {
+    // clang-format off
+    static const FunctionInfo functions[] = {
+        {0, nullptr, "GetConfig"},
+        {1, nullptr, "ModularExponentiate"},
+        {5, nullptr, "SetConfig"},
+        {7, &SPL::GetRandomBytes, "GetRandomBytes"},
+        {11, nullptr, "IsDevelopment"},
+        {24, nullptr, "SetBootReason"},
+        {25, nullptr, "GetBootReason"},
+    };
+    // clang-format on
+
+    RegisterHandlers(functions);
+}
+
+SPL_MIG::SPL_MIG(Core::System& system_, std::shared_ptr<Module> module_)
+    : Interface(system_, std::move(module_), "spl:mig") {
+    // clang-format off
+    static const FunctionInfo functions[] = {
+        {0, nullptr, "GetConfig"},
+        {1, nullptr, "ModularExponentiate"},
+        {2, nullptr, "GenerateAesKek"},
+        {3, nullptr, "LoadAesKey"},
+        {4, nullptr, "GenerateAesKey"},
+        {5, nullptr, "SetConfig"},
+        {7, &SPL::GetRandomBytes, "GenerateRandomBytes"},
+        {11, nullptr, "IsDevelopment"},
+        {14, nullptr, "DecryptAesKey"},
+        {15, nullptr, "CryptAesCtr"},
+        {16, nullptr, "ComputeCmac"},
+        {21, nullptr, "AllocateAesKeyslot"},
+        {22, nullptr, "DeallocateAesKeySlot"},
+        {23, nullptr, "GetAesKeyslotAvailableEvent"},
+        {24, nullptr, "SetBootReason"},
+        {25, nullptr, "GetBootReason"},
+    };
+    // clang-format on
+
+    RegisterHandlers(functions);
+}
+
+SPL_FS::SPL_FS(Core::System& system_, std::shared_ptr<Module> module_)
+    : Interface(system_, std::move(module_), "spl:fs") {
+    // clang-format off
+    static const FunctionInfo functions[] = {
+        {0, nullptr, "GetConfig"},
+        {1, nullptr, "ModularExponentiate"},
+        {2, nullptr, "GenerateAesKek"},
+        {3, nullptr, "LoadAesKey"},
+        {4, nullptr, "GenerateAesKey"},
+        {5, nullptr, "SetConfig"},
+        {7, &SPL::GetRandomBytes, "GenerateRandomBytes"},
+        {11, nullptr, "IsDevelopment"},
+        {12, nullptr, "GenerateSpecificAesKey"},
+        {14, nullptr, "DecryptAesKey"},
+        {15, nullptr, "CryptAesCtr"},
+        {16, nullptr, "ComputeCmac"},
+        {19, nullptr, "LoadTitleKey"},
+        {21, nullptr, "AllocateAesKeyslot"},
+        {22, nullptr, "DeallocateAesKeySlot"},
+        {23, nullptr, "GetAesKeyslotAvailableEvent"},
+        {24, nullptr, "SetBootReason"},
+        {25, nullptr, "GetBootReason"},
+        {31, nullptr, "GetPackage2Hash"},
+    };
+    // clang-format on
+
+    RegisterHandlers(functions);
+}
+
+SPL_SSL::SPL_SSL(Core::System& system_, std::shared_ptr<Module> module_)
+    : Interface(system_, std::move(module_), "spl:ssl") {
+    // clang-format off
    static const FunctionInfo functions[] = {
        {0, nullptr, "GetConfig"},
        {1, nullptr, "ModularExponentiate"},
@@ -16,18 +89,11 @@ SPL::SPL(Core::System& system_, std::shared_ptr<Module> module_)
        {4, nullptr, "GenerateAesKey"},
        {5, nullptr, "SetConfig"},
        {7, &SPL::GetRandomBytes, "GetRandomBytes"},
-        {9, nullptr, "ImportLotusKey"},
-        {10, nullptr, "DecryptLotusMessage"},
        {11, nullptr, "IsDevelopment"},
-        {12, nullptr, "GenerateSpecificAesKey"},
        {13, nullptr, "DecryptDeviceUniqueData"},
        {14, nullptr, "DecryptAesKey"},
        {15, nullptr, "CryptAesCtr"},
        {16, nullptr, "ComputeCmac"},
-        {17, nullptr, "ImportEsKey"},
-        {18, nullptr, "UnwrapTitleKey"},
-        {19, nullptr, "LoadTitleKey"},
-        {20, nullptr, "PrepareEsCommonKey"},
        {21, nullptr, "AllocateAesKeyslot"},
        {22, nullptr, "DeallocateAesKeySlot"},
        {23, nullptr, "GetAesKeyslotAvailableEvent"},
@@ -35,15 +101,83 @@ SPL::SPL(Core::System& system_, std::shared_ptr<Module> module_)
        {25, nullptr, "GetBootReason"},
        {26, nullptr, "DecryptAndStoreSslClientCertKey"},
        {27, nullptr, "ModularExponentiateWithSslClientCertKey"},
+    };
+    // clang-format on
+
+    RegisterHandlers(functions);
+}
+
+SPL_ES::SPL_ES(Core::System& system_, std::shared_ptr<Module> module_)
+    : Interface(system_, std::move(module_), "spl:es") {
+    // clang-format off
+    static const FunctionInfo functions[] = {
+        {0, nullptr, "GetConfig"},
+        {1, nullptr, "ModularExponentiate"},
+        {2, nullptr, "GenerateAesKek"},
+        {3, nullptr, "LoadAesKey"},
+        {4, nullptr, "GenerateAesKey"},
+        {5, nullptr, "SetConfig"},
+        {7, &SPL::GetRandomBytes, "GenerateRandomBytes"},
+        {11, nullptr, "IsDevelopment"},
+        {13, nullptr, "DecryptDeviceUniqueData"},
+        {14, nullptr, "DecryptAesKey"},
+        {15, nullptr, "CryptAesCtr"},
+        {16, nullptr, "ComputeCmac"},
+        {18, nullptr, "UnwrapTitleKey"},
+        {20, nullptr, "PrepareEsCommonKey"},
+        {21, nullptr, "AllocateAesKeyslot"},
+        {22, nullptr, "DeallocateAesKeySlot"},
+        {23, nullptr, "GetAesKeyslotAvailableEvent"},
+        {24, nullptr, "SetBootReason"},
+        {25, nullptr, "GetBootReason"},
        {28, nullptr, "DecryptAndStoreDrmDeviceCertKey"},
        {29, nullptr, "ModularExponentiateWithDrmDeviceCertKey"},
-        {30, nullptr, "ReencryptDeviceUniqueData "},
-        {31, nullptr, "PrepareEsArchiveKey"}, // This is also GetPackage2Hash?
+        {31, nullptr, "PrepareEsArchiveKey"},
        {32, nullptr, "LoadPreparedAesKey"},
    };
+    // clang-format on
+
+    RegisterHandlers(functions);
+}
+
+SPL_MANU::SPL_MANU(Core::System& system_, std::shared_ptr<Module> module_)
+    : Interface(system_, std::move(module_), "spl:manu") {
+    // clang-format off
+    static const FunctionInfo functions[] = {
+        {0, nullptr, "GetConfig"},
+        {1, nullptr, "ModularExponentiate"},
+        {2, nullptr, "GenerateAesKek"},
+        {3, nullptr, "LoadAesKey"},
+        {4, nullptr, "GenerateAesKey"},
+        {5, nullptr, "SetConfig"},
+        {7, &SPL::GetRandomBytes, "GetRandomBytes"},
+        {11, nullptr, "IsDevelopment"},
+        {13, nullptr, "DecryptDeviceUniqueData"},
+        {14, nullptr, "DecryptAesKey"},
+        {15, nullptr, "CryptAesCtr"},
+        {16, nullptr, "ComputeCmac"},
+        {21, nullptr, "AllocateAesKeyslot"},
+        {22, nullptr, "DeallocateAesKeySlot"},
+        {23, nullptr, "GetAesKeyslotAvailableEvent"},
+        {24, nullptr, "SetBootReason"},
+        {25, nullptr, "GetBootReason"},
+        {30, nullptr, "ReencryptDeviceUniqueData"},
+    };
+    // clang-format on
+
    RegisterHandlers(functions);
 }

 SPL::~SPL() = default;

+SPL_MIG::~SPL_MIG() = default;
+
+SPL_FS::~SPL_FS() = default;
+
+SPL_SSL::~SPL_SSL() = default;
+
+SPL_ES::~SPL_ES() = default;
+
+SPL_MANU::~SPL_MANU() = default;
+
 } // namespace Service::SPL
@@ -18,4 +18,34 @@ public:
    ~SPL() override;
 };

+class SPL_MIG final : public Module::Interface {
+public:
+    explicit SPL_MIG(Core::System& system_, std::shared_ptr<Module> module_);
+    ~SPL_MIG() override;
+};
+
+class SPL_FS final : public Module::Interface {
+public:
+    explicit SPL_FS(Core::System& system_, std::shared_ptr<Module> module_);
+    ~SPL_FS() override;
+};
+
+class SPL_SSL final : public Module::Interface {
+public:
+    explicit SPL_SSL(Core::System& system_, std::shared_ptr<Module> module_);
+    ~SPL_SSL() override;
+};
+
+class SPL_ES final : public Module::Interface {
+public:
+    explicit SPL_ES(Core::System& system_, std::shared_ptr<Module> module_);
+    ~SPL_ES() override;
+};
+
+class SPL_MANU final : public Module::Interface {
+public:
+    explicit SPL_MANU(Core::System& system_, std::shared_ptr<Module> module_);
+    ~SPL_MANU() override;
+};
+
 } // namespace Service::SPL
@@ -140,6 +140,8 @@ ResultCode Module::Interface::GetClockSnapshotFromSystemClockContextInternal(

    const auto current_time_point{
        time_manager.GetStandardSteadyClockCore().GetCurrentTimePoint(system)};
+    clock_snapshot.steady_clock_time_point = current_time_point;
+
    if (const ResultCode result{Clock::ClockSnapshot::GetCurrentTime(
            clock_snapshot.user_time, current_time_point, clock_snapshot.user_context)};
        result != RESULT_SUCCESS) {
@@ -341,12 +343,18 @@ void Module::Interface::CalculateStandardUserSystemClockDifferenceByUser(
 void Module::Interface::CalculateSpanBetween(Kernel::HLERequestContext& ctx) {
    LOG_DEBUG(Service_Time, "called");

-    IPC::RequestParser rp{ctx};
-    const auto snapshot_a = rp.PopRaw<Clock::ClockSnapshot>();
-    const auto snapshot_b = rp.PopRaw<Clock::ClockSnapshot>();
+    Clock::ClockSnapshot snapshot_a;
+    Clock::ClockSnapshot snapshot_b;
+
+    const auto snapshot_a_data = ctx.ReadBuffer(0);
+    const auto snapshot_b_data = ctx.ReadBuffer(1);
+
+    std::memcpy(&snapshot_a, snapshot_a_data.data(), sizeof(Clock::ClockSnapshot));
+    std::memcpy(&snapshot_b, snapshot_b_data.data(), sizeof(Clock::ClockSnapshot));

    Clock::TimeSpanType time_span_type{};
    s64 span{};
+
    if (const ResultCode result{snapshot_a.steady_clock_time_point.GetSpanBetween(
            snapshot_b.steady_clock_time_point, span)};
        result != RESULT_SUCCESS) {
@@ -1217,6 +1217,32 @@ private:
        }
    }

+    void GetIndirectLayerImageMap(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        const auto width = rp.Pop<s64>();
+        const auto height = rp.Pop<s64>();
+        const auto indirect_layer_consumer_handle = rp.Pop<u64>();
+        const auto applet_resource_user_id = rp.Pop<u64>();
+
+        LOG_WARNING(Service_VI,
+                    "(STUBBED) called, width={}, height={}, indirect_layer_consumer_handle={}, "
+                    "applet_resource_user_id={}",
+                    width, height, indirect_layer_consumer_handle, applet_resource_user_id);
+
+        std::vector<u8> out_buffer(0x46);
+        ctx.WriteBuffer(out_buffer);
+
+        // TODO: Figure out what these are
+
+        constexpr s64 unknown_result_1 = 0;
+        constexpr s64 unknown_result_2 = 0;
+
+        IPC::ResponseBuilder rb{ctx, 6};
+        rb.Push(unknown_result_1);
+        rb.Push(unknown_result_2);
+        rb.Push(RESULT_SUCCESS);
+    }
+
    void GetIndirectLayerImageRequiredMemoryInfo(Kernel::HLERequestContext& ctx) {
        IPC::RequestParser rp{ctx};
        const auto width = rp.Pop<u64>();
@@ -1276,7 +1302,7 @@ IApplicationDisplayService::IApplicationDisplayService(Core::System& system_,
        {2031, &IApplicationDisplayService::DestroyStrayLayer, "DestroyStrayLayer"},
        {2101, &IApplicationDisplayService::SetLayerScalingMode, "SetLayerScalingMode"},
        {2102, &IApplicationDisplayService::ConvertScalingMode, "ConvertScalingMode"},
-        {2450, nullptr, "GetIndirectLayerImageMap"},
+        {2450, &IApplicationDisplayService::GetIndirectLayerImageMap, "GetIndirectLayerImageMap"},
        {2451, nullptr, "GetIndirectLayerImageCropMap"},
        {2460, &IApplicationDisplayService::GetIndirectLayerImageRequiredMemoryInfo,
         "GetIndirectLayerImageRequiredMemoryInfo"},
@@ -12,20 +12,39 @@ namespace InputCommon {

 class KeyButton final : public Input::ButtonDevice {
 public:
-    explicit KeyButton(std::shared_ptr<KeyButtonList> key_button_list_)
-        : key_button_list(std::move(key_button_list_)) {}
+    explicit KeyButton(std::shared_ptr<KeyButtonList> key_button_list_, bool toggle_)
+        : key_button_list(std::move(key_button_list_)), toggle(toggle_) {}

    ~KeyButton() override;

    bool GetStatus() const override {
+        if (toggle) {
+            return toggled_status.load(std::memory_order_relaxed);
+        }
        return status.load();
    }

+    void ToggleButton() {
+        if (lock) {
+            return;
+        }
+        lock = true;
+        const bool old_toggle_status = toggled_status.load();
+        toggled_status.store(!old_toggle_status);
+    }
+
+    void UnlockButton() {
+        lock = false;
+    }
+
    friend class KeyButtonList;

 private:
    std::shared_ptr<KeyButtonList> key_button_list;
    std::atomic<bool> status{false};
+    std::atomic<bool> toggled_status{false};
+    bool lock{false};
+    const bool toggle;
 };

 struct KeyButtonPair {
@@ -51,6 +70,11 @@ public:
        for (const KeyButtonPair& pair : list) {
            if (pair.key_code == key_code) {
                pair.key_button->status.store(pressed);
+                if (pressed) {
+                    pair.key_button->ToggleButton();
+                } else {
+                    pair.key_button->UnlockButton();
+                }
            }
        }
    }
@@ -75,7 +99,8 @@ KeyButton::~KeyButton() {

 std::unique_ptr<Input::ButtonDevice> Keyboard::Create(const Common::ParamPackage& params) {
    const int key_code = params.Get("code", 0);
-    std::unique_ptr<KeyButton> button = std::make_unique<KeyButton>(key_button_list);
+    const bool toggle = params.Get("toggle", false);
+    std::unique_ptr<KeyButton> button = std::make_unique<KeyButton>(key_button_list, toggle);
    key_button_list->AddKeyButton(key_code, button.get());
    return button;
 }
@@ -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;
@@ -152,11 +152,52 @@ 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;
 }

+bool Mouse::ToggleButton(std::size_t button_) {
+    if (button_ >= mouse_info.size()) {
+        return false;
+    }
+    const auto button = 1U << button_;
+    const bool button_state = (toggle_buttons & button) != 0;
+    const bool button_lock = (lock_buttons & button) != 0;
+
+    if (button_lock) {
+        return button_state;
+    }
+
+    lock_buttons |= static_cast<u16>(button);
+
+    if (button_state) {
+        toggle_buttons &= static_cast<u16>(0xFF - button);
+    } else {
+        toggle_buttons |= static_cast<u16>(button);
+    }
+
+    return !button_state;
+}
+
+bool Mouse::UnlockButton(std::size_t button_) {
+    if (button_ >= mouse_info.size()) {
+        return false;
+    }
+
+    const auto button = 1U << button_;
+    const bool button_state = (toggle_buttons & button) != 0;
+
+    lock_buttons &= static_cast<u16>(0xFF - button);
+
+    return button_state;
+}
+
 Common::SPSCQueue<MouseStatus>& Mouse::GetMouseQueue() {
    return mouse_queue;
 }
@@ -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 +67,10 @@ 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_);

    [[nodiscard]] Common::SPSCQueue<MouseStatus>& GetMouseQueue();
    [[nodiscard]] const Common::SPSCQueue<MouseStatus>& GetMouseQueue() const;
@@ -92,9 +97,11 @@ private:
    };

    u16 buttons{};
+    u16 toggle_buttons{};
+    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};
@@ -14,16 +14,25 @@ namespace InputCommon {

 class MouseButton final : public Input::ButtonDevice {
 public:
-    explicit MouseButton(u32 button_, const MouseInput::Mouse* mouse_input_)
-        : button(button_), mouse_input(mouse_input_) {}
+    explicit MouseButton(u32 button_, bool toggle_, MouseInput::Mouse* mouse_input_)
+        : button(button_), toggle(toggle_), mouse_input(mouse_input_) {}

    bool GetStatus() const override {
-        return mouse_input->GetMouseState(button).pressed;
+        const bool button_state = mouse_input->GetMouseState(button).pressed;
+        if (!toggle) {
+            return button_state;
+        }
+
+        if (button_state) {
+            return mouse_input->ToggleButton(button);
+        }
+        return mouse_input->UnlockButton(button);
    }

 private:
    const u32 button;
-    const MouseInput::Mouse* mouse_input;
+    const bool toggle;
+    MouseInput::Mouse* mouse_input;
 };

 MouseButtonFactory::MouseButtonFactory(std::shared_ptr<MouseInput::Mouse> mouse_input_)
@@ -32,8 +41,9 @@ MouseButtonFactory::MouseButtonFactory(std::shared_ptr<MouseInput::Mouse> mouse_
 std::unique_ptr<Input::ButtonDevice> MouseButtonFactory::Create(
    const Common::ParamPackage& params) {
    const auto button_id = params.Get("button", 0);
+    const auto toggle = params.Get("toggle", false);

-    return std::make_unique<MouseButton>(button_id, mouse_input.get());
+    return std::make_unique<MouseButton>(button_id, toggle, mouse_input.get());
 }

 Common::ParamPackage MouseButtonFactory::GetNextInput() const {
@@ -5,6 +5,7 @@
 #include <chrono>
 #include <cstring>
 #include <functional>
+#include <random>
 #include <thread>
 #include <boost/asio.hpp>
 #include "common/logging/log.h"
@@ -26,10 +27,10 @@ class Socket {
 public:
    using clock = std::chrono::system_clock;

-    explicit Socket(const std::string& host, u16 port, std::size_t pad_index_, u32 client_id_,
+    explicit Socket(const std::string& host, u16 port, std::size_t pad_index_,
                    SocketCallback callback_)
        : callback(std::move(callback_)), timer(io_service),
-          socket(io_service, udp::endpoint(udp::v4(), 0)), client_id(client_id_),
+          socket(io_service, udp::endpoint(udp::v4(), 0)), client_id(GenerateRandomClientId()),
          pad_index(pad_index_) {
        boost::system::error_code ec{};
        auto ipv4 = boost::asio::ip::make_address_v4(host, ec);
@@ -63,6 +64,11 @@ public:
    }

 private:
+    u32 GenerateRandomClientId() const {
+        std::random_device device;
+        return device();
+    }
+
    void HandleReceive(const boost::system::error_code&, std::size_t bytes_transferred) {
        if (auto type = Response::Validate(receive_buffer.data(), bytes_transferred)) {
            switch (*type) {
@@ -115,7 +121,7 @@ private:
    boost::asio::basic_waitable_timer<clock> timer;
    udp::socket socket;

-    u32 client_id{};
+    const u32 client_id;
    std::size_t pad_index{};

    static constexpr std::size_t PORT_INFO_SIZE = sizeof(Message<Request::PortInfo>);
@@ -203,7 +209,7 @@ void Client::ReloadSockets() {
                LOG_ERROR(Input, "Duplicated UDP servers found");
                continue;
            }
-            StartCommunication(client++, udp_input_address, udp_input_port, pad, 24872);
+            StartCommunication(client++, udp_input_address, udp_input_port, pad);
        }
    }
 }
@@ -277,7 +283,7 @@ void Client::OnPadData(Response::PadData data, std::size_t client) {
 }

 void Client::StartCommunication(std::size_t client, const std::string& host, u16 port,
-                                std::size_t pad_index, u32 client_id) {
+                                std::size_t pad_index) {
    SocketCallback callback{[this](Response::Version version) { OnVersion(version); },
                            [this](Response::PortInfo info) { OnPortInfo(info); },
                            [this, client](Response::PadData data) { OnPadData(data, client); }};
@@ -287,7 +293,7 @@ void Client::StartCommunication(std::size_t client, const std::string& host, u16
    clients[client].port = port;
    clients[client].pad_index = pad_index;
    clients[client].active = 0;
-    clients[client].socket = std::make_unique<Socket>(host, port, pad_index, client_id, callback);
+    clients[client].socket = std::make_unique<Socket>(host, port, pad_index, callback);
    clients[client].thread = std::thread{SocketLoop, clients[client].socket.get()};
    // Set motion parameters
    // SetGyroThreshold value should be dependent on GyroscopeZeroDriftMode
@@ -416,7 +422,7 @@ const Common::SPSCQueue<UDPPadStatus>& Client::GetPadQueue() const {
    return pad_queue;
 }

-void TestCommunication(const std::string& host, u16 port, std::size_t pad_index, u32 client_id,
+void TestCommunication(const std::string& host, u16 port, std::size_t pad_index,
                       const std::function<void()>& success_callback,
                       const std::function<void()>& failure_callback) {
    std::thread([=] {
@@ -426,7 +432,7 @@ void TestCommunication(const std::string& host, u16 port, std::size_t pad_index,
            .port_info = [](Response::PortInfo) {},
            .pad_data = [&](Response::PadData) { success_event.Set(); },
        };
-        Socket socket{host, port, pad_index, client_id, std::move(callback)};
+        Socket socket{host, port, pad_index, std::move(callback)};
        std::thread worker_thread{SocketLoop, &socket};
        const bool result = success_event.WaitFor(std::chrono::seconds(5));
        socket.Stop();
@@ -440,7 +446,7 @@ void TestCommunication(const std::string& host, u16 port, std::size_t pad_index,
 }

 CalibrationConfigurationJob::CalibrationConfigurationJob(
-    const std::string& host, u16 port, std::size_t pad_index, u32 client_id,
+    const std::string& host, u16 port, std::size_t pad_index,
    std::function<void(Status)> status_callback,
    std::function<void(u16, u16, u16, u16)> data_callback) {

@@ -485,7 +491,7 @@ CalibrationConfigurationJob::CalibrationConfigurationJob(
                                        complete_event.Set();
                                    }
                                }};
-        Socket socket{host, port, pad_index, client_id, std::move(callback)};
+        Socket socket{host, port, pad_index, std::move(callback)};
        std::thread worker_thread{SocketLoop, &socket};
        complete_event.Wait();
        socket.Stop();
@@ -126,7 +126,7 @@ private:
    void OnPortInfo(Response::PortInfo);
    void OnPadData(Response::PadData, std::size_t client);
    void StartCommunication(std::size_t client, const std::string& host, u16 port,
-                            std::size_t pad_index, u32 client_id);
+                            std::size_t pad_index);
    void UpdateYuzuSettings(std::size_t client, const Common::Vec3<float>& acc,
                            const Common::Vec3<float>& gyro);

@@ -165,7 +165,7 @@ public:
     * @param data_callback Called when calibration data is ready
     */
    explicit CalibrationConfigurationJob(const std::string& host, u16 port, std::size_t pad_index,
-                                         u32 client_id, std::function<void(Status)> status_callback,
+                                         std::function<void(Status)> status_callback,
                                         std::function<void(u16, u16, u16, u16)> data_callback);
    ~CalibrationConfigurationJob();
    void Stop();
@@ -174,7 +174,7 @@ private:
    Common::Event complete_event;
 };

-void TestCommunication(const std::string& host, u16 port, std::size_t pad_index, u32 client_id,
+void TestCommunication(const std::string& host, u16 port, std::size_t pad_index,
                       const std::function<void()>& success_callback,
                       const std::function<void()>& failure_callback);

@@ -67,7 +67,7 @@ void TestControl1::DoWork() {
        value++;
    }
    results[id] = value;
-    Fiber::YieldTo(work_fibers[id], thread_fibers[id]);
+    Fiber::YieldTo(work_fibers[id], *thread_fibers[id]);
 }

 void TestControl1::ExecuteThread(u32 id) {
@@ -76,7 +76,7 @@ void TestControl1::ExecuteThread(u32 id) {
    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], work_fibers[id]);
+    Fiber::YieldTo(thread_fibers[id], *work_fibers[id]);
    thread_fibers[id]->Exit();
 }

@@ -117,11 +117,11 @@ public:
        for (u32 i = 0; i < 12000; i++) {
            value1 += i;
        }
-        Fiber::YieldTo(fiber1, fiber3);
+        Fiber::YieldTo(fiber1, *fiber3);
        const u32 id = thread_ids.Get();
        assert1 = id == 1;
        value2 += 5000;
-        Fiber::YieldTo(fiber1, thread_fibers[id]);
+        Fiber::YieldTo(fiber1, *thread_fibers[id]);
    }

    void DoWork2() {
@@ -129,7 +129,7 @@ public:
            ;
        value2 = 2000;
        trap = false;
-        Fiber::YieldTo(fiber2, fiber1);
+        Fiber::YieldTo(fiber2, *fiber1);
        assert3 = false;
    }

@@ -137,19 +137,19 @@ public:
        const u32 id = thread_ids.Get();
        assert2 = id == 0;
        value1 += 1000;
-        Fiber::YieldTo(fiber3, thread_fibers[id]);
+        Fiber::YieldTo(fiber3, *thread_fibers[id]);
    }

    void ExecuteThread(u32 id);

    void CallFiber1() {
        const u32 id = thread_ids.Get();
-        Fiber::YieldTo(thread_fibers[id], fiber1);
+        Fiber::YieldTo(thread_fibers[id], *fiber1);
    }

    void CallFiber2() {
        const u32 id = thread_ids.Get();
-        Fiber::YieldTo(thread_fibers[id], fiber2);
+        Fiber::YieldTo(thread_fibers[id], *fiber2);
    }

    void Exit();
@@ -241,23 +241,23 @@ public:

    void DoWork1() {
        value1 += 1;
-        Fiber::YieldTo(fiber1, fiber2);
+        Fiber::YieldTo(fiber1, *fiber2);
        const u32 id = thread_ids.Get();
        value3 += 1;
-        Fiber::YieldTo(fiber1, thread_fibers[id]);
+        Fiber::YieldTo(fiber1, *thread_fibers[id]);
    }

    void DoWork2() {
        value2 += 1;
        const u32 id = thread_ids.Get();
-        Fiber::YieldTo(fiber2, thread_fibers[id]);
+        Fiber::YieldTo(fiber2, *thread_fibers[id]);
    }

    void ExecuteThread(u32 id);

    void CallFiber1() {
        const u32 id = thread_ids.Get();
-        Fiber::YieldTo(thread_fibers[id], fiber1);
+        Fiber::YieldTo(thread_fibers[id], *fiber1);
    }

    void Exit();
@@ -332,7 +332,7 @@ public:

    void Execute() {
        thread_fiber = Fiber::ThreadToFiber();
-        Fiber::YieldTo(thread_fiber, fiber1);
+        Fiber::YieldTo(thread_fiber, *fiber1);
        thread_fiber->Exit();
    }

@@ -340,7 +340,7 @@ public:
        fiber1->SetRewindPoint(std::function<void(void*)>{WorkControl4}, this);
        if (rewinded) {
            goal_reached = true;
-            Fiber::YieldTo(fiber1, thread_fiber);
+            Fiber::YieldTo(fiber1, *thread_fiber);
        }
        rewinded = true;
        fiber1->Rewind();
@@ -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>
@@ -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
@@ -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
@@ -459,7 +459,7 @@ void GPU::ProcessSemaphoreAcquire() {
 }

 void GPU::Start() {
-    gpu_thread.StartThread(*renderer, renderer->Context(), *dma_pusher, *cdma_pusher);
+    gpu_thread.StartThread(*renderer, renderer->Context(), *dma_pusher);
    cpu_context = renderer->GetRenderWindow().CreateSharedContext();
    cpu_context->MakeCurrent();
 }
@@ -19,7 +19,7 @@ namespace VideoCommon::GPUThread {
 /// Runs the GPU thread
 static void RunThread(Core::System& system, VideoCore::RendererBase& renderer,
                      Core::Frontend::GraphicsContext& context, Tegra::DmaPusher& dma_pusher,
-                      SynchState& state, Tegra::CDmaPusher& cdma_pusher) {
+                      SynchState& state) {
    std::string name = "yuzu:GPU";
    MicroProfileOnThreadCreate(name.c_str());
    SCOPE_EXIT({ MicroProfileOnThreadExit(); });
@@ -46,9 +46,6 @@ static void RunThread(Core::System& system, VideoCore::RendererBase& renderer,
        if (auto* submit_list = std::get_if<SubmitListCommand>(&next.data)) {
            dma_pusher.Push(std::move(submit_list->entries));
            dma_pusher.DispatchCalls();
-        } else if (auto* command_list = std::get_if<SubmitChCommandEntries>(&next.data)) {
-            // NVDEC
-            cdma_pusher.ProcessEntries(std::move(command_list->entries));
        } else if (const auto* data = std::get_if<SwapBuffersCommand>(&next.data)) {
            renderer.SwapBuffers(data->framebuffer ? &*data->framebuffer : nullptr);
        } else if (std::holds_alternative<OnCommandListEndCommand>(next.data)) {
@@ -83,20 +80,16 @@ ThreadManager::~ThreadManager() {

 void ThreadManager::StartThread(VideoCore::RendererBase& renderer,
                                Core::Frontend::GraphicsContext& context,
-                                Tegra::DmaPusher& dma_pusher, Tegra::CDmaPusher& cdma_pusher) {
+                                Tegra::DmaPusher& dma_pusher) {
    rasterizer = renderer.ReadRasterizer();
    thread = std::thread(RunThread, std::ref(system), std::ref(renderer), std::ref(context),
-                         std::ref(dma_pusher), std::ref(state), std::ref(cdma_pusher));
+                         std::ref(dma_pusher), std::ref(state));
 }

 void ThreadManager::SubmitList(Tegra::CommandList&& entries) {
    PushCommand(SubmitListCommand(std::move(entries)));
 }

-void ThreadManager::SubmitCommandBuffer(Tegra::ChCommandHeaderList&& entries) {
-    PushCommand(SubmitChCommandEntries(std::move(entries)));
-}
-
 void ThreadManager::SwapBuffers(const Tegra::FramebufferConfig* framebuffer) {
    PushCommand(SwapBuffersCommand(framebuffer ? std::make_optional(*framebuffer) : std::nullopt));
 }
@@ -43,14 +43,6 @@ struct SubmitListCommand final {
    Tegra::CommandList entries;
 };

-/// Command to signal to the GPU thread that a cdma command list is ready for processing
-struct SubmitChCommandEntries final {
-    explicit SubmitChCommandEntries(Tegra::ChCommandHeaderList&& entries_)
-        : entries{std::move(entries_)} {}
-
-    Tegra::ChCommandHeaderList entries;
-};
-
 /// Command to signal to the GPU thread that a swap buffers is pending
 struct SwapBuffersCommand final {
    explicit SwapBuffersCommand(std::optional<const Tegra::FramebufferConfig> framebuffer_)
@@ -91,9 +83,9 @@ struct OnCommandListEndCommand final {};
 struct GPUTickCommand final {};

 using CommandData =
-    std::variant<EndProcessingCommand, SubmitListCommand, SubmitChCommandEntries,
-                 SwapBuffersCommand, FlushRegionCommand, InvalidateRegionCommand,
-                 FlushAndInvalidateRegionCommand, OnCommandListEndCommand, GPUTickCommand>;
+    std::variant<EndProcessingCommand, SubmitListCommand, SwapBuffersCommand, FlushRegionCommand,
+                 InvalidateRegionCommand, FlushAndInvalidateRegionCommand, OnCommandListEndCommand,
+                 GPUTickCommand>;

 struct CommandDataContainer {
    CommandDataContainer() = default;
@@ -123,14 +115,11 @@ public:

    /// Creates and starts the GPU thread.
    void StartThread(VideoCore::RendererBase& renderer, Core::Frontend::GraphicsContext& context,
-                     Tegra::DmaPusher& dma_pusher, Tegra::CDmaPusher& cdma_pusher);
+                     Tegra::DmaPusher& dma_pusher);

    /// Push GPU command entries to be processed
    void SubmitList(Tegra::CommandList&& entries);

-    /// Push GPU CDMA command buffer entries to be processed
-    void SubmitCommandBuffer(Tegra::ChCommandHeaderList&& entries);
-
    /// Swap buffers (render frame)
    void SwapBuffers(const Tegra::FramebufferConfig* framebuffer);

@@ -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
@@ -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);
+}
@@ -2,63 +2,43 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.

-#include <mutex>
-
-#include <boost/icl/interval_map.hpp>
-#include <boost/range/iterator_range.hpp>
-
 #include "common/assert.h"
 #include "common/common_types.h"
+#include "common/div_ceil.h"
 #include "core/memory.h"
 #include "video_core/rasterizer_accelerated.h"

 namespace VideoCore {

-namespace {
-
-template <typename Map, typename Interval>
-constexpr auto RangeFromInterval(Map& map, const Interval& interval) {
-    return boost::make_iterator_range(map.equal_range(interval));
-}
-
-} // Anonymous namespace
-
 RasterizerAccelerated::RasterizerAccelerated(Core::Memory::Memory& cpu_memory_)
    : cpu_memory{cpu_memory_} {}

 RasterizerAccelerated::~RasterizerAccelerated() = default;

 void RasterizerAccelerated::UpdatePagesCachedCount(VAddr addr, u64 size, int delta) {
-    std::lock_guard lock{pages_mutex};
-    const u64 page_start{addr >> Core::Memory::PAGE_BITS};
-    const u64 page_end{(addr + size + Core::Memory::PAGE_SIZE - 1) >> Core::Memory::PAGE_BITS};
+    const auto page_end = Common::DivCeil(addr + size, Core::Memory::PAGE_SIZE);
+    for (auto page = addr >> Core::Memory::PAGE_BITS; page != page_end; ++page) {
+        auto& count = cached_pages.at(page >> 3).Count(page);

-    // Interval maps will erase segments if count reaches 0, so if delta is negative we have to
-    // subtract after iterating
-    const auto pages_interval = CachedPageMap::interval_type::right_open(page_start, page_end);
-    if (delta > 0) {
-        cached_pages.add({pages_interval, delta});
-    }
-
-    for (const auto& pair : RangeFromInterval(cached_pages, pages_interval)) {
-        const auto interval = pair.first & pages_interval;
-        const int count = pair.second;
-
-        const VAddr interval_start_addr = boost::icl::first(interval) << Core::Memory::PAGE_BITS;
-        const VAddr interval_end_addr = boost::icl::last_next(interval) << Core::Memory::PAGE_BITS;
-        const u64 interval_size = interval_end_addr - interval_start_addr;
-
-        if (delta > 0 && count == delta) {
-            cpu_memory.RasterizerMarkRegionCached(interval_start_addr, interval_size, true);
-        } else if (delta < 0 && count == -delta) {
-            cpu_memory.RasterizerMarkRegionCached(interval_start_addr, interval_size, false);
+        if (delta > 0) {
+            ASSERT_MSG(count < UINT8_MAX, "Count may overflow!");
+        } else if (delta < 0) {
+            ASSERT_MSG(count > 0, "Count may underflow!");
        } else {
-            ASSERT(count >= 0);
+            ASSERT_MSG(true, "Delta must be non-zero!");
        }
-    }

-    if (delta < 0) {
-        cached_pages.add({pages_interval, delta});
+        // Adds or subtracts 1, as count is a unsigned 8-bit value
+        count += static_cast<u8>(delta);
+
+        // Assume delta is either -1 or 1
+        if (count == 0) {
+            cpu_memory.RasterizerMarkRegionCached(page << Core::Memory::PAGE_BITS,
+                                                  Core::Memory::PAGE_SIZE, false);
+        } else if (count == 1 && delta > 0) {
+            cpu_memory.RasterizerMarkRegionCached(page << Core::Memory::PAGE_BITS,
+                                                  Core::Memory::PAGE_SIZE, true);
+        }
    }
 }

@@ -4,9 +4,8 @@

 #pragma once

-#include <mutex>
-
-#include <boost/icl/interval_map.hpp>
+#include <array>
+#include <atomic>

 #include "common/common_types.h"
 #include "video_core/rasterizer_interface.h"
@@ -26,10 +25,24 @@ public:
    void UpdatePagesCachedCount(VAddr addr, u64 size, int delta) override;

 private:
-    using CachedPageMap = boost::icl::interval_map<u64, int>;
-    CachedPageMap cached_pages;
-    std::mutex pages_mutex;
+    class CacheEntry final {
+    public:
+        CacheEntry() = default;

+        std::atomic_uint8_t& Count(std::size_t page) {
+            return values[page & 7];
+        }
+
+        const std::atomic_uint8_t& Count(std::size_t page) const {
+            return values[page & 7];
+        }
+
+    private:
+        std::array<std::atomic_uint8_t, 8> values{};
+    };
+    static_assert(sizeof(CacheEntry) == 8, "CacheEntry should be 8 bytes!");
+
+    std::array<CacheEntry, 0x800000> cached_pages;
    Core::Memory::Memory& cpu_memory;
 };

@@ -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) {
@@ -210,6 +210,12 @@ Device::Device() {
    const bool is_amd = vendor == "ATI Technologies Inc.";
    const bool is_intel = vendor == "Intel";

+#ifdef __unix__
+    const bool is_linux = true;
+#else
+    const bool is_linux = false;
+#endif
+
    bool disable_fast_buffer_sub_data = false;
    if (is_nvidia && version == "4.6.0 NVIDIA 443.24") {
        LOG_WARNING(
@@ -249,7 +255,9 @@ Device::Device() {
                           GLAD_GL_NV_gpu_program5 && GLAD_GL_NV_compute_program5 &&
                           GLAD_GL_NV_transform_feedback && GLAD_GL_NV_transform_feedback2;

-    use_asynchronous_shaders = Settings::values.use_asynchronous_shaders.GetValue();
+    // Blocks AMD and Intel OpenGL drivers on Windows from using asynchronous shader compilation.
+    use_asynchronous_shaders = Settings::values.use_asynchronous_shaders.GetValue() &&
+                               !(is_amd || (is_intel && !is_linux));
    use_driver_cache = is_nvidia;

    LOG_INFO(Render_OpenGL, "Renderer_VariableAOFFI: {}", has_variable_aoffi);
@@ -261,6 +269,10 @@ Device::Device() {
    if (Settings::values.use_assembly_shaders.GetValue() && !use_assembly_shaders) {
        LOG_ERROR(Render_OpenGL, "Assembly shaders enabled but not supported");
    }
+
+    if (Settings::values.use_asynchronous_shaders.GetValue() && !use_asynchronous_shaders) {
+        LOG_WARNING(Render_OpenGL, "Asynchronous shader compilation enabled but not supported");
+    }
 }

 Device::Device(std::nullptr_t) {
@@ -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();
    }
@@ -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;
    }
@@ -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
@@ -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);
@@ -10,7 +10,7 @@

 namespace Vulkan {

-constexpr size_t COMMAND_BUFFER_POOL_SIZE = 0x1000;
+constexpr size_t COMMAND_BUFFER_POOL_SIZE = 4;

 struct CommandPool::Pool {
    vk::CommandPool handle;
@@ -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 {
@@ -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");
@@ -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();
@@ -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);
        }
    });
@@ -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,
@@ -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);
@@ -51,7 +51,7 @@ constexpr std::array REQUIRED_EXTENSIONS{
 #ifdef _WIN32
    VK_KHR_EXTERNAL_MEMORY_WIN32_EXTENSION_NAME,
 #endif
-#ifdef __linux__
+#ifdef __unix__
    VK_KHR_EXTERNAL_MEMORY_FD_EXTENSION_NAME,
 #endif
 };
@@ -62,7 +62,7 @@ public:
        : memory{std::move(memory_)}, allocation_size{allocation_size_}, property_flags{properties},
          shifted_memory_type{1U << type} {}

-#if defined(_WIN32) || defined(__linux__)
+#if defined(_WIN32) || defined(__unix__)
    ~MemoryAllocation() {
        if (owning_opengl_handle != 0) {
            glDeleteMemoryObjectsEXT(1, &owning_opengl_handle);
@@ -114,7 +114,7 @@ public:
        }
        return owning_opengl_handle;
    }
-#elif __linux__
+#elif __unix__
    [[nodiscard]] u32 ExportOpenGLHandle() {
        if (!owning_opengl_handle) {
            glCreateMemoryObjectsEXT(1, &owning_opengl_handle);
@@ -165,7 +165,7 @@ private:
    const u32 shifted_memory_type;              ///< Shifted Vulkan memory type.
    std::vector<Range> commits;                 ///< All commit ranges done from this allocation.
    std::span<u8> memory_mapped_span; ///< Memory mapped span. Empty if not queried before.
-#if defined(_WIN32) || defined(__linux__)
+#if defined(_WIN32) || defined(__unix__)
    u32 owning_opengl_handle{}; ///< Owning OpenGL memory object handle.
 #endif
 };
@@ -249,7 +249,7 @@ void MemoryAllocator::AllocMemory(VkMemoryPropertyFlags flags, u32 type_mask, u6
        .pNext = nullptr,
 #ifdef _WIN32
        .handleTypes = VK_EXTERNAL_MEMORY_HANDLE_TYPE_OPAQUE_WIN32_BIT,
-#elif __linux__
+#elif __unix__
        .handleTypes = VK_EXTERNAL_MEMORY_HANDLE_TYPE_OPAQUE_FD_BIT,
 #else
        .handleTypes = 0,
@@ -376,11 +376,34 @@ void GRenderWindow::closeEvent(QCloseEvent* event) {
 }

 void GRenderWindow::keyPressEvent(QKeyEvent* event) {
-    input_subsystem->GetKeyboard()->PressKey(event->key());
+    if (!event->isAutoRepeat()) {
+        input_subsystem->GetKeyboard()->PressKey(event->key());
+    }
 }

 void GRenderWindow::keyReleaseEvent(QKeyEvent* event) {
-    input_subsystem->GetKeyboard()->ReleaseKey(event->key());
+    if (!event->isAutoRepeat()) {
+        input_subsystem->GetKeyboard()->ReleaseKey(event->key());
+    }
+}
+
+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) {
@@ -391,7 +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);
@@ -425,7 +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);
@@ -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;
@@ -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();

@@ -648,7 +648,7 @@ void Config::ReadDebuggingValues() {
 void Config::ReadServiceValues() {
    qt_config->beginGroup(QStringLiteral("Services"));
    Settings::values.bcat_backend =
-        ReadSetting(QStringLiteral("bcat_backend"), QStringLiteral("null"))
+        ReadSetting(QStringLiteral("bcat_backend"), QStringLiteral("none"))
            .toString()
            .toStdString();
    Settings::values.bcat_boxcat_local =
@@ -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();
@@ -1239,7 +1238,7 @@ void Config::SaveDebuggingValues() {
 void Config::SaveServiceValues() {
    qt_config->beginGroup(QStringLiteral("Services"));
    WriteSetting(QStringLiteral("bcat_backend"),
-                 QString::fromStdString(Settings::values.bcat_backend), QStringLiteral("null"));
+                 QString::fromStdString(Settings::values.bcat_backend), QStringLiteral("none"));
    WriteSetting(QStringLiteral("bcat_boxcat_local"), Settings::values.bcat_boxcat_local, false);
    qt_config->endGroup();
 }
@@ -103,7 +103,10 @@ void ConfigureFilesystem::SetDirectory(DirectoryTarget target, QLineEdit* edit)
        str = QFileDialog::getOpenFileName(this, caption, QFileInfo(edit->text()).dir().path(),
                                           QStringLiteral("NX Gamecard;*.xci"));
    } else {
-        str = QFileDialog::getExistingDirectory(this, caption, edit->text()) + QDir::separator();
+        str = QFileDialog::getExistingDirectory(this, caption, edit->text());
+        if (!str.isNull() && str.back() != QDir::separator()) {
+            str.append(QDir::separator());
+        }
    }

    if (str.isEmpty())
@@ -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"
@@ -104,7 +105,9 @@ QString ButtonToText(const Common::ParamPackage& param) {
    }

    if (param.Get("engine", "") == "keyboard") {
-        return GetKeyName(param.Get("code", 0));
+        const QString button_str = GetKeyName(param.Get("code", 0));
+        const QString toggle = QString::fromStdString(param.Get("toggle", false) ? "~" : "");
+        return QObject::tr("%1%2").arg(toggle, button_str);
    }

    if (param.Get("engine", "") == "gcpad") {
@@ -156,7 +159,8 @@ QString ButtonToText(const Common::ParamPackage& param) {
    if (param.Get("engine", "") == "mouse") {
        if (param.Has("button")) {
            const QString button_str = QString::number(int(param.Get("button", 0)));
-            return QObject::tr("Click %1").arg(button_str);
+            const QString toggle = QString::fromStdString(param.Get("toggle", false) ? "~" : "");
+            return QObject::tr("%1Click %2").arg(toggle, button_str);
        }
        return GetKeyName(param.Get("code", 0));
    }
@@ -300,6 +304,11 @@ ConfigureInputPlayer::ConfigureInputPlayer(QWidget* parent, std::size_t player_i
                        buttons_param[button_id].Clear();
                        button_map[button_id]->setText(tr("[not set]"));
                    });
+                    context_menu.addAction(tr("Toggle button"), [&] {
+                        const bool toggle_value = !buttons_param[button_id].Get("toggle", false);
+                        buttons_param[button_id].Set("toggle", toggle_value);
+                        button_map[button_id]->setText(ButtonToText(buttons_param[button_id]));
+                    });
                    context_menu.exec(button_map[button_id]->mapToGlobal(menu_location));
                    ui->controllerFrame->SetPlayerInput(player_index, buttons_param, analogs_param);
                });
@@ -412,6 +421,15 @@ ConfigureInputPlayer::ConfigureInputPlayer(QWidget* parent, std::size_t player_i
                        analogs_param[analog_id].Set("modifier", "");
                        analog_map_modifier_button[analog_id]->setText(tr("[not set]"));
                    });
+                    context_menu.addAction(tr("Toggle button"), [&] {
+                        Common::ParamPackage modifier_param =
+                            Common::ParamPackage{analogs_param[analog_id].Get("modifier", "")};
+                        const bool toggle_value = !modifier_param.Get("toggle", false);
+                        modifier_param.Set("toggle", toggle_value);
+                        analogs_param[analog_id].Set("modifier", modifier_param.Serialize());
+                        analog_map_modifier_button[analog_id]->setText(
+                            ButtonToText(modifier_param));
+                    });
                    context_menu.exec(
                        analog_map_modifier_button[analog_id]->mapToGlobal(menu_location));
                });
@@ -1345,7 +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) {
@@ -24,7 +24,7 @@

 CalibrationConfigurationDialog::CalibrationConfigurationDialog(QWidget* parent,
                                                               const std::string& host, u16 port,
-                                                               u8 pad_index, u16 client_id)
+                                                               u8 pad_index)
    : QDialog(parent) {
    layout = new QVBoxLayout;
    status_label = new QLabel(tr("Communicating with the server..."));
@@ -41,7 +41,7 @@ CalibrationConfigurationDialog::CalibrationConfigurationDialog(QWidget* parent,

    using namespace InputCommon::CemuhookUDP;
    job = std::make_unique<CalibrationConfigurationJob>(
-        host, port, pad_index, client_id,
+        host, port, pad_index,
        [this](CalibrationConfigurationJob::Status status) {
            QString text;
            switch (status) {
@@ -218,7 +218,6 @@ void ConfigureMotionTouch::OnCemuhookUDPTest() {
    udp_test_in_progress = true;
    InputCommon::CemuhookUDP::TestCommunication(
        ui->udp_server->text().toStdString(), static_cast<u16>(ui->udp_port->text().toInt()), 0,
-        24872,
        [this] {
            LOG_INFO(Frontend, "UDP input test success");
            QMetaObject::invokeMethod(this, "ShowUDPTestResult", Q_ARG(bool, true));
@@ -233,8 +232,7 @@ void ConfigureMotionTouch::OnConfigureTouchCalibration() {
    ui->touch_calibration_config->setEnabled(false);
    ui->touch_calibration_config->setText(tr("Configuring"));
    CalibrationConfigurationDialog dialog(this, ui->udp_server->text().toStdString(),
-                                          static_cast<u16>(ui->udp_port->text().toUInt()), 0,
-                                          24872);
+                                          static_cast<u16>(ui->udp_port->text().toUInt()), 0);
    dialog.exec();
    if (dialog.completed) {
        min_x = dialog.min_x;
@@ -30,7 +30,7 @@ class CalibrationConfigurationDialog : public QDialog {
    Q_OBJECT
 public:
    explicit CalibrationConfigurationDialog(QWidget* parent, const std::string& host, u16 port,
-                                            u8 pad_index, u16 client_id);
+                                            u8 pad_index);
    ~CalibrationConfigurationDialog() override;

 private:
@@ -60,6 +60,7 @@ static FileSys::VirtualFile VfsDirectoryCreateFileWrapper(const FileSys::Virtual
 #include <QPushButton>
 #include <QShortcut>
 #include <QStatusBar>
+#include <QString>
 #include <QSysInfo>
 #include <QUrl>
 #include <QtConcurrent/QtConcurrent>
@@ -854,8 +855,7 @@ 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 +1208,14 @@ 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 +2375,15 @@ 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 +2621,7 @@ 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 +2631,7 @@ 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();
    }
 }
@@ -3058,6 +3062,14 @@ int main(int argc, char* argv[]) {
    chdir(bin_path.c_str());
 #endif

+#ifdef __linux__
+    // Set the DISPLAY variable in order to open web browsers
+    // TODO (lat9nq): Find a better solution for AppImages to start external applications
+    if (QString::fromLocal8Bit(qgetenv("DISPLAY")).isEmpty()) {
+        qputenv("DISPLAY", ":0");
+    }
+#endif
+
    // Enables the core to make the qt created contexts current on std::threads
    QCoreApplication::setAttribute(Qt::AA_DontCheckOpenGLContextThreadAffinity);
    QApplication app(argc, argv);
@@ -457,7 +457,7 @@ void Config::ReadValues() {
    Settings::values.yuzu_token = sdl2_config->Get("WebService", "yuzu_token", "");

    // Services
-    Settings::values.bcat_backend = sdl2_config->Get("Services", "bcat_backend", "null");
+    Settings::values.bcat_backend = sdl2_config->Get("Services", "bcat_backend", "none");
    Settings::values.bcat_boxcat_local =
        sdl2_config->GetBoolean("Services", "bcat_boxcat_local", false);
 }
@@ -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);
        }
    }
 }
@@ -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);
Author	SHA1	Message	Date
Jan Beich	9b50b23a50	vulkan_common: enable OpenGL interop on other Unices	2021-03-30 00:25:25 +00:00
bunnei	fb7dcbf7af	Merge pull request #6102 from ogniK5377/fd-pass nvdrv: Pass device fd and handle device create methods for device opening and closing	2021-03-29 12:53:55 -07:00
bunnei	a63dcb6d56	Merge pull request #6118 from MerryMage/dynarmic externals: Update dynarmic to c28f13af	2021-03-28 13:59:05 -07:00
bunnei	f8561c7a65	Merge pull request #6115 from bunnei/fix-kernel-init hle: kernel: Initialize preemption task after schedulers.	2021-03-28 12:13:39 -07:00
MerryMage	a527e5f0cd	externals: Update dynarmic to c28f13af AVX-512 bugfixes	2021-03-27 21:34:20 +00:00
bunnei	8a5794c4db	Merge pull request #6114 from Morph1984/friend-log service: friend: Change logging class from ACC to Friend	2021-03-27 10:29:43 -07:00
bunnei	cb7f2e5616	hle: kernel: Initialize preemption task after schedulers. - Fixes a startup crash that occurs if CoreTiming tries to preempt before kernel initialization completes.	2021-03-27 10:04:13 -07:00
Morph	3b98fab850	service: friend: Change logging class from ACC to Friend	2021-03-27 11:11:07 -04:00
Chloe Marcec	99fdfa1fcd	nvdrv: Pass device fd and handle device create methods for device opening and closing We pass the fd to the ioctl as well as alert the device when it's opened or closed to allow for fd unique actions to take place	2021-03-25 12:56:42 +11:00
bunnei	65774084fd	Merge pull request #6101 from ogniK5377/alloc-as-ex nvdrv: Change InitializeEx to AllocAsEx	2021-03-24 17:37:01 -07:00
bunnei	b83eb4dd18	Merge pull request #6108 from jbeich/freebsd-async-shaders gl_device: unblock async shaders on other Unix systems	2021-03-24 16:52:27 -07:00
Jan Beich	8c016b02e7	gl_device: unblock async shaders on other Unix systems Mesa is the primary OpenGL provider on all FreeDesktop systems. For example, iris is used on Intel GPU + FreeBSD by default.	2021-03-24 19:59:20 +00:00
bunnei	716285fab8	Merge pull request #6100 from bunnei/arm-fix core: arm_dynarmic: Ensure JIT state is saved/restored on page table changes.	2021-03-23 17:08:37 -07:00
bunnei	dde074eaab	Merge pull request #6092 from ivan-boikov/cancel-dir-select Fix cancelation of getExistingDirectory dialog	2021-03-22 17:55:18 -07:00
Chloe Marcec	875183e7c5	nvdrv: Change InitializeEx to AllocAsEx Wee also report the correct "big page size" now in GetVARegions & fix up the struct for IoctlAllocAsEx	2021-03-22 19:59:30 +11:00
LC	a50133fc5e	Merge pull request #6095 from lat9nq/async-shader-block gl_device: Block async shaders on AMD and Intel	2021-03-21 23:29:49 -04:00
LC	e274e38205	Merge pull request #6078 from Morph1984/libusb-utf8 externals/libusb/CMakeLists: Add /utf-8 compile option for MSVC	2021-03-21 23:29:01 -04:00
bunnei	a35717b245	core: arm_dynarmic: Ensure JIT state is saved/restored on page table changes. - We re-create the JIT here without preserving any state.	2021-03-21 15:25:25 -07:00
lat9nq	538f097f97	gl_device: Block async shaders on AMD and Intel Currently, the Windows versions of the Intel OpenGL driver and the AMD proprietary OpenGL driver do not properly support (or in fact degrade) when asynchronous shader compilation is enabled. This blocks specifically those drivers from using this feature. This affects AMDGPU-PRO on Linux, and AMD's and Intel's OpenGL drivers on Windows.	2021-03-21 01:25:45 -04:00
ivan-boikov	afa4bcbb3b	Fix cancelation of choose directory dialog	2021-03-20 15:52:22 +01:00
bunnei	2e85ee250d	Merge pull request #6052 from Morph1984/vi-getindirectlayerimagemap IApplicationDisplayService: Stub GetIndirectLayerImageMap	2021-03-19 22:59:25 -07:00
bunnei	cb48ed2e1a	Merge pull request #6056 from zkitX/spl-updates service: Refactor spl	2021-03-18 15:49:41 -07:00
bunnei	4aa8189328	Merge pull request #6055 from MerryMage/exceed-the-limit [testing] fiber: Double default stack size	2021-03-17 20:46:45 -07:00
Morph	ec514a4d1b	IApplicationDisplayService: Stub GetIndirectLayerImageMap Used by games invoking the inline software keyboard such as GNOSIA	2021-03-17 03:25:12 -04:00
bunnei	e1f7938a3b	Merge pull request #6070 from Morph1984/sysver-11.0.1 system_version: Update to 11.0.1	2021-03-16 18:13:12 -07:00
bunnei	ab102787fa	Merge pull request #6083 from Morph1984/bsd-avoid-writebuffer bsd: Avoid writing empty buffers	2021-03-16 11:59:05 -07:00
Morph	8441094ba3	bsd: Avoid writing empty buffers Silences log spam on empty buffer writes	2021-03-16 12:50:44 -04:00
bunnei	0687a8370d	Merge pull request #6069 from Morph1984/ngWord system_archive: Update NgWord archive version	2021-03-15 15:44:04 -07:00
Morph	df9899eed6	externals/libusb/CMakeLists: Add /utf-8 compile option for MSVC	2021-03-15 10:51:46 -04:00
Morph	824e53149d	system_version: Update to 11.0.1	2021-03-14 08:47:36 -04:00
Morph	9761618a8d	system_archive: Update NgWord archive version	2021-03-14 08:33:48 -04:00
bunnei	d3a4a192fe	Merge pull request #6054 from Morph1984/time-GetClockSnapshot time: Assign the current time point to the ClockSnapshot	2021-03-13 23:15:54 -08:00
bunnei	3b85ac2ac4	Merge pull request #6053 from Morph1984/time-CalculateSpanBetween time: Fix CalculateSpanBetween implementation	2021-03-12 23:31:09 -08:00
bunnei	4735d18bb9	Merge pull request #6028 from bunnei/raster-cache video_core: rasterizer_accelerated: Use a flat array instead of interval_map for cached pages.	2021-03-12 21:57:27 -08:00
bunnei	a9d24b0df3	video_core: rasterizer_accelerated: Fix un/signed mismatch.	2021-03-12 21:52:49 -08:00
bunnei	5dae45b958	Merge pull request #5327 from AniLeo/master qt: Set DISPLAY env var when not present	2021-03-12 12:10:28 -08:00
bunnei	827dcad26e	Merge pull request #6040 from german77/toggleKeyboard Enable toggle buttons for keyboard and mouse	2021-03-11 11:00:44 -08:00
zkitx	4439801c0f	Fix casing on DeallocateAesKeySlot	2021-03-11 02:46:22 -05:00
zkitx	ad653550eb	Update SPL to fit N's service refactor (4.0.0+) which split into new services.	2021-03-11 02:36:48 -05:00
MerryMage	59173ff7a7	fiber: Double default stack size Stack overflow occurs with some guest applications	2021-03-10 20:37:56 +00:00
Morph	87cfe5b1da	time: Fix CalculateSpanBetween implementation CalculateSpanBetween passes in the ClockSnapshots through 2 input buffers and not as raw arguments. Fix this by reading the 2 input buffers instead of popping raw arguments. Partially fixes Super Smash Bros. Ultimate's Spirit Board	2021-03-10 11:42:59 -05:00
Morph	2490ffbbce	time: Assign the current time point to the ClockSnapshot Fixes the timer in Super Smash Bros Ultimate's Spirit Board.	2021-03-10 11:40:51 -05:00
Rodrigo Locatti	daf5c5060b	Merge pull request #5891 from ameerj/bgra-ogl renderer_opengl: Use compute shaders to swizzle BGR textures on copy	2021-03-09 02:47:51 -03:00
bunnei	d1a7b2eca7	Merge pull request #6021 from ReinUsesLisp/skip-cache-heuristic buffer_cache: Heuristically decide to skip cache on uniform buffers	2021-03-08 17:48:55 -08:00
bunnei	9c4c9f1e7d	Merge pull request #5990 from german77/mousePanningV2 InputCommon: Mouse fixes	2021-03-08 14:50:58 -08:00
bunnei	69ce5e41eb	Merge pull request #6041 from bunnei/fiber-leaks common: fiber: Use weak_ptr when yielding.	2021-03-08 13:51:37 -08:00
Morph	9d77ae39de	Merge pull request #6047 from lioncash/dynarmic external: Update dynarmic	2021-03-08 00:52:19 -05:00
Lioncash	de21c9e330	external: Update dynarmic	2021-03-08 00:44:54 -05:00
bunnei	8f7eb194af	common: Fiber: use a reference for YieldTo. - Fixes another small leak.	2021-03-07 13:46:53 -08:00
Ani	cd7abba1a9	qt: Set DISPLAY env var when not present Fixes web browser opening (Help > Open Mods Page, Help > Open Quickstart Guide)	2021-03-07 15:56:22 +00:00
german77	41e94b7b99	Enable mouse toggle buttons	2021-03-06 13:27:02 -06:00
german	4bcc5bacff	Add toggle button option for normal buttons	2021-03-06 07:36:41 -06:00
bunnei	68ffac250a	common: fiber: Use weak_ptr when yielding. - Avoids a memory leak, as taking a strong reference of the fiber here causes a circular reference. - Supersedes #6006 with a more narrow fix.	2021-03-05 22:10:03 -08:00
german	1f228c51ca	Enable button toggle for keyboard in the modifier button	2021-03-05 19:21:04 -06:00
bunnei	4cf5b860bd	Merge pull request #6036 from bunnei/thread-leak hle: kernel: KThread: Rework dummy threads & fix memory leak.	2021-03-05 17:15:35 -08:00
bunnei	47af34003b	hle: kernel: KThread: Rework dummy threads & fix memory leak. - Dummy threads are created on thread local storage for all host threads. - Fixes a leak by removing creation of fibers, which are not applicable here.	2021-03-05 17:10:57 -08:00
LC	97415ad07a	Merge pull request #6029 from Morph1984/compile-utf8 CMakeLists: Add /utf-8 compile option for MSVC	2021-03-05 20:09:37 -05:00
bunnei	7b29a8ce4e	Merge pull request #6039 from yuzu-emu/revert-6006-fiber-unique-ptr Revert "core: Switch to unique_ptr for usage of Common::Fiber."	2021-03-05 17:08:48 -08:00
bunnei	a5ab85ac37	Revert "core: Switch to unique_ptr for usage of Common::Fiber."	2021-03-05 17:08:17 -08:00
bunnei	9d010be483	Merge pull request #6034 from Morph1984/mbedtls externals: Update mbedtls to 2.16.9	2021-03-05 15:48:28 -08:00
bunnei	34a3ee1631	Merge pull request #6006 from bunnei/fiber-unique-ptr core: Switch to unique_ptr for usage of Common::Fiber.	2021-03-04 23:59:06 -08:00
Morph	96c9e67b1b	aes_util: Remove malformed mbedtls_cipher_finish function call	2021-03-05 02:05:05 -05:00
Morph	6faabd6d69	externals: Update mbedtls to 2.16.9 mbedtls 2.16 is the last version which has licensing for GPL 2.0. This updates mbedtls to our own fork of mbedtls 2.16	2021-03-05 02:05:05 -05:00
Morph	e7038344aa	CMakeLists: Add /utf-8 compile option for MSVC Ensures that the source and execution character sets are in UTF-8	2021-03-05 01:46:56 -05:00
ameerj	5213f70230	texture_cache: Blacklist BGRA8 copies and views on OpenGL In order to force the BGRA8 conversion on Nvidia using OpenGL, we need to forbid texture copies and views with other formats. This commit also adds a boolean relating to this, as this needs to be done only for the OpenGL api, Vulkan must remain unchanged.	2021-03-04 14:14:49 -05:00
ameerj	0639244d85	renderer_opengl: Swizzle BGR textures on copy OpenGL does not natively support BGR internal formats, which causes many BGR textures to render incorrectly, with Red and Blue channels swapped. This commit aims to address this by swizzling the blue and red channels on texture copies when a BGR format is encountered.	2021-03-04 14:14:19 -05:00
bunnei	b8b5891585	Merge pull request #5989 from ReinUsesLisp/cmdpool vk_command_pool: Reduce the command pool size from 4096 to 4	2021-03-04 11:07:31 -08:00
bunnei	394475c4e3	Merge pull request #6004 from german77/udprandom InputCommon: Use an unique client id for each udp socket instance	2021-03-03 15:45:32 -08:00
bunnei	50ee9c46ab	video_core: rasterizer_accelerated: Fix delta check ordering.	2021-03-02 17:48:02 -08:00
bunnei	6ab839462c	video_core: rasterizer_accelerated: Improve error handling & fix implicit conversion.	2021-03-02 17:44:02 -08:00
bunnei	f8bfec3109	Merge pull request #5815 from comex/net-error-reform Network error handling reform	2021-03-02 17:08:47 -08:00
bunnei	94da1e8a7e	video_core: rasterizer_accelerated: Use a flat array instead of interval_map for cached pages. - Uses a fixed 64MB for the cache instead of an ever growing map. - Slightly faster by using atomics instead of a single mutex for access. - Thanks for Rodrigo for the idea.	2021-03-02 16:57:53 -08:00
LC	4a45012f35	Merge pull request #6020 from bunnei/shutdown-crash-2 core: Shutdown: Move kernel cleanup to later in shutdown.	2021-03-02 09:28:28 -05:00
ReinUsesLisp	5ad62e7bfc	buffer_cache: Heuristically decide to skip cache on uniform buffers Some games benefit from skipping caches (Pokémon Sword), and others don't (Animal Crossing: New Horizons). Add an heuristic to decide this at runtime. The cache hit ratio has to be ~98% or better to not skip the cache. There are 16 frames of buffer.	2021-03-02 02:44:19 -03:00
bunnei	925671071c	core: Shutdown: Move kernel cleanup to later in shutdown. - Fixes a shutdown crash due to a race condition with GPU still accessing memory.	2021-03-01 21:42:06 -08:00
bunnei	cd25817938	Merge pull request #6019 from Kelebek1/bcat [Service::nifm] Fix bcat_backend's default initialisation	2021-03-01 19:27:06 -08:00
Kelebek1	c7a7e47615	Fix default bcat_backend init	2021-03-02 03:20:16 +00:00
german	9b3af0027b	inputCommon: Use an unique client id for each socket instance	2021-03-01 09:19:33 -06:00
Morph	ac8b1445ff	Merge pull request #6016 from ameerj/remove-async-nvdec gpu_thread: Remove Async NVDEC placeholders	2021-03-01 04:22:46 -05:00
ameerj	52e9d7fa49	gpu_thread: Remove Async NVDEC placeholders This commit removes early placeholders for an implementation of async nvdec. With recent changes to the source code, the placeholders are no longer accurate, and can cause a nullptr dereference due to the nature of the cdma_pusher lifetime.	2021-02-28 22:03:00 -05:00
german77	4738e14cb0	inputCommon: Mouse fixes	2021-02-27 17:53:10 -06:00
bunnei	51fb0a6f96	core: Switch to unique_ptr for usage of Common::Fiber. - With using unique_ptr instead of shared_ptr, we have more explicit ownership of the context. - Fixes a memory leak due to circular reference of the shared pointer.	2021-02-27 11:56:04 -08:00
ReinUsesLisp	aae399c1a8	vk_command_pool: Reduce the command pool size from 4096 to 4 This allows drivers to reuse memory more easily and preallocate less. The optimal number has been measured booting Pokémon Sword.	2021-02-23 19:08:24 -03:00