common/xbyak_api: Make BuildRegSet() constexpr

This allows us to eliminate any static constructors that would have been
emitted due to the function not being constexpr.

externals/FidelityFX-FSR/ffx-fsr/ffx_fsr1.h

@@ -747,12 +747,12 @@ AF1 sharpness){
   // Immediate constants for peak range.
   AF2 peakC=AF2(1.0,-1.0*4.0);
   // Limiters, these need to be high precision RCPs.
-  AF1 hitMinR=mn4R*ARcpF1(AF1_(4.0)*mx4R);
-  AF1 hitMinG=mn4G*ARcpF1(AF1_(4.0)*mx4G);
-  AF1 hitMinB=mn4B*ARcpF1(AF1_(4.0)*mx4B);
-  AF1 hitMaxR=(peakC.x-mx4R)*ARcpF1(AF1_(4.0)*mn4R+peakC.y);
-  AF1 hitMaxG=(peakC.x-mx4G)*ARcpF1(AF1_(4.0)*mn4G+peakC.y);
-  AF1 hitMaxB=(peakC.x-mx4B)*ARcpF1(AF1_(4.0)*mn4B+peakC.y);
+  AF1 hitMinR=min(mn4R,eR)*ARcpF1(AF1_(4.0)*mx4R);
+  AF1 hitMinG=min(mn4G,eG)*ARcpF1(AF1_(4.0)*mx4G);
+  AF1 hitMinB=min(mn4B,eB)*ARcpF1(AF1_(4.0)*mx4B);
+  AF1 hitMaxR=(peakC.x-max(mx4R,eR))*ARcpF1(AF1_(4.0)*mn4R+peakC.y);
+  AF1 hitMaxG=(peakC.x-max(mx4G,eG))*ARcpF1(AF1_(4.0)*mn4G+peakC.y);
+  AF1 hitMaxB=(peakC.x-max(mx4B,eB))*ARcpF1(AF1_(4.0)*mn4B+peakC.y);
   AF1 lobeR=max(-hitMinR,hitMaxR);
   AF1 lobeG=max(-hitMinG,hitMaxG);
   AF1 lobeB=max(-hitMinB,hitMaxB);
@@ -845,12 +845,12 @@ AF1 sharpness){
   // Immediate constants for peak range.
   AH2 peakC=AH2(1.0,-1.0*4.0);
   // Limiters, these need to be high precision RCPs.
-  AH1 hitMinR=mn4R*ARcpH1(AH1_(4.0)*mx4R);
-  AH1 hitMinG=mn4G*ARcpH1(AH1_(4.0)*mx4G);
-  AH1 hitMinB=mn4B*ARcpH1(AH1_(4.0)*mx4B);
-  AH1 hitMaxR=(peakC.x-mx4R)*ARcpH1(AH1_(4.0)*mn4R+peakC.y);
-  AH1 hitMaxG=(peakC.x-mx4G)*ARcpH1(AH1_(4.0)*mn4G+peakC.y);
-  AH1 hitMaxB=(peakC.x-mx4B)*ARcpH1(AH1_(4.0)*mn4B+peakC.y);
+  AH1 hitMinR=min(mn4R,eR)*ARcpH1(AH1_(4.0)*mx4R);
+  AH1 hitMinG=min(mn4G,eG)*ARcpH1(AH1_(4.0)*mx4G);
+  AH1 hitMinB=min(mn4B,eB)*ARcpH1(AH1_(4.0)*mx4B);
+  AH1 hitMaxR=(peakC.x-max(mx4R,eR))*ARcpH1(AH1_(4.0)*mn4R+peakC.y);
+  AH1 hitMaxG=(peakC.x-max(mx4G,eG))*ARcpH1(AH1_(4.0)*mn4G+peakC.y);
+  AH1 hitMaxB=(peakC.x-max(mx4B,eB))*ARcpH1(AH1_(4.0)*mn4B+peakC.y);
   AH1 lobeR=max(-hitMinR,hitMaxR);
   AH1 lobeG=max(-hitMinG,hitMaxG);
   AH1 lobeB=max(-hitMinB,hitMaxB);
@@ -963,12 +963,12 @@ AF1 sharpness){
   // Immediate constants for peak range.
   AH2 peakC=AH2(1.0,-1.0*4.0);
   // Limiters, these need to be high precision RCPs.
-  AH2 hitMinR=mn4R*ARcpH2(AH2_(4.0)*mx4R);
-  AH2 hitMinG=mn4G*ARcpH2(AH2_(4.0)*mx4G);
-  AH2 hitMinB=mn4B*ARcpH2(AH2_(4.0)*mx4B);
-  AH2 hitMaxR=(peakC.x-mx4R)*ARcpH2(AH2_(4.0)*mn4R+peakC.y);
-  AH2 hitMaxG=(peakC.x-mx4G)*ARcpH2(AH2_(4.0)*mn4G+peakC.y);
-  AH2 hitMaxB=(peakC.x-mx4B)*ARcpH2(AH2_(4.0)*mn4B+peakC.y);
+  AH2 hitMinR=min(mn4R,eR)*ARcpH2(AH2_(4.0)*mx4R);
+  AH2 hitMinG=min(mn4G,eG)*ARcpH2(AH2_(4.0)*mx4G);
+  AH2 hitMinB=min(mn4B,eB)*ARcpH2(AH2_(4.0)*mx4B);
+  AH2 hitMaxR=(peakC.x-max(mx4R,eR))*ARcpH2(AH2_(4.0)*mn4R+peakC.y);
+  AH2 hitMaxG=(peakC.x-max(mx4G,eG))*ARcpH2(AH2_(4.0)*mn4G+peakC.y);
+  AH2 hitMaxB=(peakC.x-max(mx4B,eB))*ARcpH2(AH2_(4.0)*mn4B+peakC.y);
   AH2 lobeR=max(-hitMinR,hitMaxR);
   AH2 lobeG=max(-hitMinG,hitMaxG);
   AH2 lobeB=max(-hitMinB,hitMaxB);

externals/ffmpeg/CMakeLists.txt

@@ -95,7 +95,8 @@ if (NOT WIN32)
         # ffnvenc could load CUDA libraries at the runtime using dlopen/dlsym or LoadLibrary/GetProcAddress
         # here we handle the hard-linking senario where CUDA is linked during compilation
         if (CUDA_FOUND)
-            list(APPEND FFmpeg_HWACCEL_FLAGS --extra-cflags=-I${CUDA_INCLUDE_DIRS})
+            # This line causes build error if CUDA_INCLUDE_DIRS is anything but a single non-empty value
+            #list(APPEND FFmpeg_HWACCEL_FLAGS --extra-cflags=-I${CUDA_INCLUDE_DIRS})
             list(APPEND FFmpeg_HWACCEL_LIBRARIES ${CUDA_LIBRARIES})
             list(APPEND FFmpeg_HWACCEL_INCLUDE_DIRS ${CUDA_INCLUDE_DIRS})
             list(APPEND FFmpeg_HWACCEL_LDFLAGS ${CUDA_LDFLAGS})
@@ -119,6 +120,8 @@ if (NOT WIN32)
 
     # `configure` parameters builds only exactly what yuzu needs from FFmpeg
     # `--disable-vdpau` is needed to avoid linking issues
+    set(FFmpeg_CC ${CMAKE_C_COMPILER_LAUNCHER}  ${CMAKE_C_COMPILER})
+    set(FFmpeg_CXX ${CMAKE_CXX_COMPILER_LAUNCHER}  ${CMAKE_CXX_COMPILER})
     add_custom_command(
         OUTPUT
             ${FFmpeg_MAKEFILE}
@@ -137,12 +140,14 @@ if (NOT WIN32)
                 --enable-decoder=h264
                 --enable-decoder=vp8
                 --enable-decoder=vp9
-                --cc="${CMAKE_C_COMPILER}"
-                --cxx="${CMAKE_CXX_COMPILER}"
+                --cc="${FFmpeg_CC}"
+                --cxx="${FFmpeg_CXX}"
                 ${FFmpeg_HWACCEL_FLAGS}
         WORKING_DIRECTORY
             ${FFmpeg_BUILD_DIR}
     )
+    unset(FFmpeg_CC)
+    unset(FFmpeg_CXX)
     unset(FFmpeg_HWACCEL_FLAGS)
 
     # Workaround for Ubuntu 18.04's older version of make not being able to call make as a child

src/audio_core/stream.cpp

@@ -79,8 +79,8 @@ static void VolumeAdjustSamples(std::vector<s16>& samples, float game_volume) {
         return;
     }
 
-    // Implementation of a volume slider with a dynamic range of 60 dB
-    const float volume_scale_factor = volume == 0 ? 0 : std::exp(6.90775f * volume) * 0.001f;
+    // Perceived volume is not the same as the volume level
+    const float volume_scale_factor = (0.85f * ((volume * volume) - volume)) + volume;
     for (auto& sample : samples) {
         sample = static_cast<s16>(sample * volume_scale_factor);
     }

src/common/bit_util.h

@@ -45,6 +45,12 @@ template <typename T>
     return static_cast<u32>(log2_f + static_cast<u64>((value ^ (1ULL << log2_f)) != 0ULL));
 }
 
+template <typename T>
+requires std::is_unsigned_v<T>
+[[nodiscard]] constexpr bool IsPow2(T value) {
+    return std::has_single_bit(value);
+}
+
 template <typename T>
 requires std::is_integral_v<T>
 [[nodiscard]] T NextPow2(T value) {

src/common/fiber.cpp

@@ -124,7 +124,10 @@ void Fiber::YieldTo(std::weak_ptr<Fiber> weak_from, Fiber& to) {
 
     // "from" might no longer be valid if the thread was killed
     if (auto from = weak_from.lock()) {
-        ASSERT(from->impl->previous_fiber != nullptr);
+        if (from->impl->previous_fiber == nullptr) {
+            ASSERT_MSG(false, "previous_fiber is nullptr!");
+            return;
+        }
         from->impl->previous_fiber->impl->context = transfer.fctx;
         from->impl->previous_fiber->impl->guard.unlock();
         from->impl->previous_fiber.reset();

src/common/input.h

@@ -209,6 +209,13 @@ enum class ButtonNames {
     Triangle,
     Share,
     Options,
+
+    // Mouse buttons
+    ButtonMouseWheel,
+    ButtonBackward,
+    ButtonForward,
+    ButtonTask,
+    ButtonExtra,
 };
 
 // Callback data consisting of an input type and the equivalent data status

src/common/settings.h

@@ -554,6 +554,7 @@ struct Values {
     Setting<bool> use_docked_mode{true, "use_docked_mode"};
 
     BasicSetting<bool> enable_raw_input{false, "enable_raw_input"};
+    BasicSetting<bool> controller_navigation{true, "controller_navigation"};
 
     Setting<bool> vibration_enabled{true, "vibration_enabled"};
     Setting<bool> enable_accurate_vibrations{false, "enable_accurate_vibrations"};

src/common/x64/xbyak_abi.h

@@ -37,12 +37,12 @@ constexpr Xbyak::Reg IndexToReg(size_t reg_index) {
     }
 }
 
-inline std::bitset<32> BuildRegSet(std::initializer_list<Xbyak::Reg> regs) {
-    std::bitset<32> bits;
+constexpr std::bitset<32> BuildRegSet(std::initializer_list<Xbyak::Reg> regs) {
+    size_t bits = 0;
     for (const Xbyak::Reg& reg : regs) {
-        bits[RegToIndex(reg)] = true;
+        bits |= size_t{1} << RegToIndex(reg);
     }
-    return bits;
+    return {bits};
 }
 
 constexpr inline std::bitset<32> ABI_ALL_GPRS(0x0000FFFF);
@@ -57,7 +57,7 @@ constexpr inline Xbyak::Reg ABI_PARAM2 = Xbyak::util::rdx;
 constexpr inline Xbyak::Reg ABI_PARAM3 = Xbyak::util::r8;
 constexpr inline Xbyak::Reg ABI_PARAM4 = Xbyak::util::r9;
 
-const std::bitset<32> ABI_ALL_CALLER_SAVED = BuildRegSet({
+constexpr inline std::bitset<32> ABI_ALL_CALLER_SAVED = BuildRegSet({
     // GPRs
     Xbyak::util::rcx,
     Xbyak::util::rdx,
@@ -74,7 +74,7 @@ const std::bitset<32> ABI_ALL_CALLER_SAVED = BuildRegSet({
     Xbyak::util::xmm5,
 });
 
-const std::bitset<32> ABI_ALL_CALLEE_SAVED = BuildRegSet({
+constexpr inline std::bitset<32> ABI_ALL_CALLEE_SAVED = BuildRegSet({
     // GPRs
     Xbyak::util::rbx,
     Xbyak::util::rsi,
@@ -108,7 +108,7 @@ constexpr inline Xbyak::Reg ABI_PARAM2 = Xbyak::util::rsi;
 constexpr inline Xbyak::Reg ABI_PARAM3 = Xbyak::util::rdx;
 constexpr inline Xbyak::Reg ABI_PARAM4 = Xbyak::util::rcx;
 
-const std::bitset<32> ABI_ALL_CALLER_SAVED = BuildRegSet({
+constexpr inline std::bitset<32> ABI_ALL_CALLER_SAVED = BuildRegSet({
     // GPRs
     Xbyak::util::rcx,
     Xbyak::util::rdx,
@@ -137,7 +137,7 @@ const std::bitset<32> ABI_ALL_CALLER_SAVED = BuildRegSet({
     Xbyak::util::xmm15,
 });
 
-const std::bitset<32> ABI_ALL_CALLEE_SAVED = BuildRegSet({
+constexpr inline std::bitset<32> ABI_ALL_CALLEE_SAVED = BuildRegSet({
     // GPRs
     Xbyak::util::rbx,
     Xbyak::util::rbp,

src/core/CMakeLists.txt

@@ -247,6 +247,9 @@ add_library(core STATIC
     hle/kernel/k_trace.h
     hle/kernel/k_transfer_memory.cpp
     hle/kernel/k_transfer_memory.h
+    hle/kernel/k_worker_task.h
+    hle/kernel/k_worker_task_manager.cpp
+    hle/kernel/k_worker_task_manager.h
     hle/kernel/k_writable_event.cpp
     hle/kernel/k_writable_event.h
     hle/kernel/kernel.cpp

src/core/hid/emulated_console.cpp

@@ -158,7 +158,7 @@ void EmulatedConsole::SetMotion(const Common::Input::CallbackStatus& callback) {
     auto& motion = console.motion_state;
     motion.accel = emulated.GetAcceleration();
     motion.gyro = emulated.GetGyroscope();
-    motion.rotation = emulated.GetGyroscope();
+    motion.rotation = emulated.GetRotations();
     motion.orientation = emulated.GetOrientation();
     motion.quaternion = emulated.GetQuaternion();
     motion.gyro_bias = emulated.GetGyroBias();

src/core/hid/emulated_controller.cpp

@@ -145,7 +145,7 @@ void EmulatedController::LoadDevices() {
                    motion_devices.begin(), Common::Input::CreateDevice<Common::Input::InputDevice>);
     std::transform(trigger_params.begin(), trigger_params.end(), trigger_devices.begin(),
                    Common::Input::CreateDevice<Common::Input::InputDevice>);
-    std::transform(battery_params.begin(), battery_params.begin(), battery_devices.end(),
+    std::transform(battery_params.begin(), battery_params.end(), battery_devices.begin(),
                    Common::Input::CreateDevice<Common::Input::InputDevice>);
     std::transform(output_params.begin(), output_params.end(), output_devices.begin(),
                    Common::Input::CreateDevice<Common::Input::OutputDevice>);

src/core/hid/input_converter.cpp

@@ -114,7 +114,7 @@ Common::Input::MotionStatus TransformToMotion(const Common::Input::CallbackStatu
         if (TransformToButton(callback).value) {
             std::random_device device;
             std::mt19937 gen(device());
-            std::uniform_int_distribution<s16> distribution(-1000, 1000);
+            std::uniform_int_distribution<s16> distribution(-5000, 5000);
             status.accel.x.raw_value = static_cast<f32>(distribution(gen)) * 0.001f;
             status.accel.y.raw_value = static_cast<f32>(distribution(gen)) * 0.001f;
             status.accel.z.raw_value = static_cast<f32>(distribution(gen)) * 0.001f;

src/core/hle/kernel/hle_ipc.h

@@ -341,10 +341,6 @@ public:
         return *thread;
     }
 
-    bool IsThreadWaiting() const {
-        return is_thread_waiting;
-    }
-
 private:
     friend class IPC::ResponseBuilder;
 
@@ -379,7 +375,6 @@ private:
     u32 domain_offset{};
 
     std::shared_ptr<SessionRequestManager> manager;
-    bool is_thread_waiting{};
 
     KernelCore& kernel;
     Core::Memory::Memory& memory;

src/core/hle/kernel/k_affinity_mask.h

@@ -31,8 +31,6 @@ public:
     }
 
     constexpr void SetAffinity(s32 core, bool set) {
-        ASSERT(0 <= core && core < static_cast<s32>(Core::Hardware::NUM_CPU_CORES));
-
         if (set) {
             this->mask |= GetCoreBit(core);
         } else {

src/core/hle/kernel/k_memory_manager.cpp

@@ -8,12 +8,16 @@
 #include "common/assert.h"
 #include "common/common_types.h"
 #include "common/scope_exit.h"
+#include "core/core.h"
+#include "core/device_memory.h"
 #include "core/hle/kernel/k_memory_manager.h"
 #include "core/hle/kernel/k_page_linked_list.h"
 #include "core/hle/kernel/svc_results.h"
 
 namespace Kernel {
 
+KMemoryManager::KMemoryManager(Core::System& system_) : system{system_} {}
+
 std::size_t KMemoryManager::Impl::Initialize(Pool new_pool, u64 start_address, u64 end_address) {
     const auto size{end_address - start_address};
 
@@ -81,7 +85,7 @@ VAddr KMemoryManager::AllocateAndOpenContinuous(std::size_t num_pages, std::size
 }
 
 ResultCode KMemoryManager::Allocate(KPageLinkedList& page_list, std::size_t num_pages, Pool pool,
-                                    Direction dir) {
+                                    Direction dir, u32 heap_fill_value) {
     ASSERT(page_list.GetNumPages() == 0);
 
     // Early return if we're allocating no pages
@@ -139,6 +143,12 @@ ResultCode KMemoryManager::Allocate(KPageLinkedList& page_list, std::size_t num_
         }
     }
 
+    // Clear allocated memory.
+    for (const auto& it : page_list.Nodes()) {
+        std::memset(system.DeviceMemory().GetPointer(it.GetAddress()), heap_fill_value,
+                    it.GetSize());
+    }
+
     // Only succeed if we allocated as many pages as we wanted
     if (num_pages) {
         return ResultOutOfMemory;
@@ -146,11 +156,12 @@ ResultCode KMemoryManager::Allocate(KPageLinkedList& page_list, std::size_t num_
 
     // We succeeded!
     group_guard.Cancel();
+
     return ResultSuccess;
 }
 
 ResultCode KMemoryManager::Free(KPageLinkedList& page_list, std::size_t num_pages, Pool pool,
-                                Direction dir) {
+                                Direction dir, u32 heap_fill_value) {
     // Early return if we're freeing no pages
     if (!num_pages) {
         return ResultSuccess;

src/core/hle/kernel/k_memory_manager.h

@@ -12,6 +12,10 @@
 #include "core/hle/kernel/k_page_heap.h"
 #include "core/hle/result.h"
 
+namespace Core {
+class System;
+}
+
 namespace Kernel {
 
 class KPageLinkedList;
@@ -42,7 +46,7 @@ public:
         Mask = (0xF << Shift),
     };
 
-    KMemoryManager() = default;
+    explicit KMemoryManager(Core::System& system_);
 
     constexpr std::size_t GetSize(Pool pool) const {
         return managers[static_cast<std::size_t>(pool)].GetSize();
@@ -51,10 +55,10 @@ public:
     void InitializeManager(Pool pool, u64 start_address, u64 end_address);
 
     VAddr AllocateAndOpenContinuous(size_t num_pages, size_t align_pages, u32 option);
-    ResultCode Allocate(KPageLinkedList& page_list, std::size_t num_pages, Pool pool,
-                        Direction dir = Direction::FromFront);
-    ResultCode Free(KPageLinkedList& page_list, std::size_t num_pages, Pool pool,
-                    Direction dir = Direction::FromFront);
+    ResultCode Allocate(KPageLinkedList& page_list, std::size_t num_pages, Pool pool, Direction dir,
+                        u32 heap_fill_value = 0);
+    ResultCode Free(KPageLinkedList& page_list, std::size_t num_pages, Pool pool, Direction dir,
+                    u32 heap_fill_value = 0);
 
     static constexpr std::size_t MaxManagerCount = 10;
 
@@ -129,6 +133,7 @@ private:
     };
 
 private:
+    Core::System& system;
     std::array<std::mutex, static_cast<std::size_t>(Pool::Count)> pool_locks;
     std::array<Impl, MaxManagerCount> managers;
 };

src/core/hle/kernel/k_page_table.cpp

@@ -289,8 +289,8 @@ ResultCode KPageTable::MapProcessCode(VAddr addr, std::size_t num_pages, KMemory
     }
 
     KPageLinkedList page_linked_list;
-    CASCADE_CODE(
-        system.Kernel().MemoryManager().Allocate(page_linked_list, num_pages, memory_pool));
+    CASCADE_CODE(system.Kernel().MemoryManager().Allocate(page_linked_list, num_pages, memory_pool,
+                                                          allocation_option));
     CASCADE_CODE(Operate(addr, num_pages, page_linked_list, OperationType::MapGroup));
 
     block_manager->Update(addr, num_pages, state, perm);
@@ -457,8 +457,8 @@ ResultCode KPageTable::MapPhysicalMemory(VAddr addr, std::size_t size) {
 
     KPageLinkedList page_linked_list;
 
-    CASCADE_CODE(
-        system.Kernel().MemoryManager().Allocate(page_linked_list, remaining_pages, memory_pool));
+    CASCADE_CODE(system.Kernel().MemoryManager().Allocate(page_linked_list, remaining_pages,
+                                                          memory_pool, allocation_option));
 
     // We succeeded, so commit the memory reservation.
     memory_reservation.Commit();
@@ -541,7 +541,8 @@ ResultCode KPageTable::UnmapMemory(VAddr addr, std::size_t size) {
     }
 
     const std::size_t num_pages{size / PageSize};
-    system.Kernel().MemoryManager().Free(page_linked_list, num_pages, memory_pool);
+    system.Kernel().MemoryManager().Free(page_linked_list, num_pages, memory_pool,
+                                         allocation_option);
 
     block_manager->Update(addr, num_pages, KMemoryState::Free);
 
@@ -960,7 +961,7 @@ ResultCode KPageTable::SetHeapSize(VAddr* out, std::size_t size) {
     // Allocate pages for the heap extension.
     KPageLinkedList page_linked_list;
     R_TRY(system.Kernel().MemoryManager().Allocate(page_linked_list, allocation_size / PageSize,
-                                                   memory_pool));
+                                                   memory_pool, allocation_option));
 
     // Map the pages.
     {
@@ -1027,8 +1028,8 @@ ResultVal<VAddr> KPageTable::AllocateAndMapMemory(std::size_t needed_num_pages,
         CASCADE_CODE(Operate(addr, needed_num_pages, perm, OperationType::Map, map_addr));
     } else {
         KPageLinkedList page_group;
-        CASCADE_CODE(
-            system.Kernel().MemoryManager().Allocate(page_group, needed_num_pages, memory_pool));
+        CASCADE_CODE(system.Kernel().MemoryManager().Allocate(page_group, needed_num_pages,
+                                                              memory_pool, allocation_option));
         CASCADE_CODE(Operate(addr, needed_num_pages, page_group, OperationType::MapGroup));
     }
 

src/core/hle/kernel/k_page_table.h

@@ -303,6 +303,7 @@ private:
     bool is_aslr_enabled{};
 
     KMemoryManager::Pool memory_pool{KMemoryManager::Pool::Application};
+    KMemoryManager::Direction allocation_option{KMemoryManager::Direction::FromFront};
 
     Common::PageTable page_table_impl;
 

src/core/hle/kernel/k_priority_queue.h

@@ -45,6 +45,7 @@ concept KPriorityQueueMember = !std::is_reference_v<T> && requires(T & t) {
 
     { t.GetActiveCore() } -> Common::ConvertibleTo<s32>;
     { t.GetPriority() } -> Common::ConvertibleTo<s32>;
+    { t.IsDummyThread() } -> Common::ConvertibleTo<bool>;
 };
 
 template <typename Member, size_t NumCores_, int LowestPriority, int HighestPriority>
@@ -349,24 +350,49 @@ public:
 
     // Mutators.
     constexpr void PushBack(Member* member) {
+        // This is for host (dummy) threads that we do not want to enter the priority queue.
+        if (member->IsDummyThread()) {
+            return;
+        }
+
         this->PushBack(member->GetPriority(), member);
     }
 
     constexpr void Remove(Member* member) {
+        // This is for host (dummy) threads that we do not want to enter the priority queue.
+        if (member->IsDummyThread()) {
+            return;
+        }
+
         this->Remove(member->GetPriority(), member);
     }
 
     constexpr void MoveToScheduledFront(Member* member) {
+        // This is for host (dummy) threads that we do not want to enter the priority queue.
+        if (member->IsDummyThread()) {
+            return;
+        }
+
         this->scheduled_queue.MoveToFront(member->GetPriority(), member->GetActiveCore(), member);
     }
 
     constexpr KThread* MoveToScheduledBack(Member* member) {
+        // This is for host (dummy) threads that we do not want to enter the priority queue.
+        if (member->IsDummyThread()) {
+            return {};
+        }
+
         return this->scheduled_queue.MoveToBack(member->GetPriority(), member->GetActiveCore(),
                                                 member);
     }
 
     // First class fancy operations.
     constexpr void ChangePriority(s32 prev_priority, bool is_running, Member* member) {
+        // This is for host (dummy) threads that we do not want to enter the priority queue.
+        if (member->IsDummyThread()) {
+            return;
+        }
+
         ASSERT(IsValidPriority(prev_priority));
 
         // Remove the member from the queues.
@@ -383,6 +409,11 @@ public:
 
     constexpr void ChangeAffinityMask(s32 prev_core, const AffinityMaskType& prev_affinity,
                                       Member* member) {
+        // This is for host (dummy) threads that we do not want to enter the priority queue.
+        if (member->IsDummyThread()) {
+            return;
+        }
+
         // Get the new information.
         const s32 priority = member->GetPriority();
         const AffinityMaskType& new_affinity = member->GetAffinityMask();
@@ -412,6 +443,11 @@ public:
     }
 
     constexpr void ChangeCore(s32 prev_core, Member* member, bool to_front = false) {
+        // This is for host (dummy) threads that we do not want to enter the priority queue.
+        if (member->IsDummyThread()) {
+            return;
+        }
+
         // Get the new information.
         const s32 new_core = member->GetActiveCore();
         const s32 priority = member->GetPriority();

src/core/hle/kernel/k_process.cpp

@@ -146,27 +146,35 @@ ResultCode KProcess::Initialize(KProcess* process, Core::System& system, std::st
     // Open a reference to the resource limit.
     process->resource_limit->Open();
 
+    // Clear remaining fields.
+    process->num_running_threads = 0;
+    process->is_signaled = false;
+    process->exception_thread = nullptr;
+    process->is_suspended = false;
+    process->schedule_count = 0;
+
     return ResultSuccess;
 }
 
+void KProcess::DoWorkerTaskImpl() {
+    UNIMPLEMENTED();
+}
+
 KResourceLimit* KProcess::GetResourceLimit() const {
     return resource_limit;
 }
 
-void KProcess::IncrementThreadCount() {
-    ASSERT(num_threads >= 0);
-    num_created_threads++;
-
-    if (const auto count = ++num_threads; count > peak_num_threads) {
-        peak_num_threads = count;
-    }
+void KProcess::IncrementRunningThreadCount() {
+    ASSERT(num_running_threads.load() >= 0);
+    ++num_running_threads;
 }
 
-void KProcess::DecrementThreadCount() {
-    ASSERT(num_threads > 0);
+void KProcess::DecrementRunningThreadCount() {
+    ASSERT(num_running_threads.load() > 0);
 
-    if (const auto count = --num_threads; count == 0) {
-        LOG_WARNING(Kernel, "Process termination is not fully implemented.");
+    if (const auto prev = num_running_threads--; prev == 1) {
+        // TODO(bunnei): Process termination to be implemented when multiprocess is supported.
+        UNIMPLEMENTED_MSG("KProcess termination is not implemennted!");
     }
 }
 
@@ -477,7 +485,7 @@ void KProcess::Finalize() {
     }
 
     // Perform inherited finalization.
-    KAutoObjectWithSlabHeapAndContainer<KProcess, KSynchronizationObject>::Finalize();
+    KAutoObjectWithSlabHeapAndContainer<KProcess, KWorkerTask>::Finalize();
 }
 
 /**

src/core/hle/kernel/k_process.h

@@ -15,6 +15,7 @@
 #include "core/hle/kernel/k_condition_variable.h"
 #include "core/hle/kernel/k_handle_table.h"
 #include "core/hle/kernel/k_synchronization_object.h"
+#include "core/hle/kernel/k_worker_task.h"
 #include "core/hle/kernel/process_capability.h"
 #include "core/hle/kernel/slab_helpers.h"
 #include "core/hle/result.h"
@@ -62,8 +63,7 @@ enum class ProcessStatus {
     DebugBreak,
 };
 
-class KProcess final
-    : public KAutoObjectWithSlabHeapAndContainer<KProcess, KSynchronizationObject> {
+class KProcess final : public KAutoObjectWithSlabHeapAndContainer<KProcess, KWorkerTask> {
     KERNEL_AUTOOBJECT_TRAITS(KProcess, KSynchronizationObject);
 
 public:
@@ -235,8 +235,8 @@ public:
         ++schedule_count;
     }
 
-    void IncrementThreadCount();
-    void DecrementThreadCount();
+    void IncrementRunningThreadCount();
+    void DecrementRunningThreadCount();
 
     void SetRunningThread(s32 core, KThread* thread, u64 idle_count) {
         running_threads[core] = thread;
@@ -345,6 +345,8 @@ public:
 
     bool IsSignaled() const override;
 
+    void DoWorkerTaskImpl();
+
     void PinCurrentThread(s32 core_id);
     void UnpinCurrentThread(s32 core_id);
     void UnpinThread(KThread* thread);
@@ -471,9 +473,7 @@ private:
     bool is_suspended{};
     bool is_initialized{};
 
-    std::atomic<s32> num_created_threads{};
-    std::atomic<u16> num_threads{};
-    u16 peak_num_threads{};
+    std::atomic<u16> num_running_threads{};
 
     std::array<KThread*, Core::Hardware::NUM_CPU_CORES> running_threads{};
     std::array<u64, Core::Hardware::NUM_CPU_CORES> running_thread_idle_counts{};

src/core/hle/kernel/k_scheduler.cpp

@@ -49,8 +49,6 @@ void KScheduler::RescheduleCores(KernelCore& kernel, u64 cores_pending_reschedul
         if (!must_context_switch || core != current_core) {
             auto& phys_core = kernel.PhysicalCore(core);
             phys_core.Interrupt();
-        } else {
-            must_context_switch = true;
         }
         cores_pending_reschedule &= ~(1ULL << core);
     }
@@ -408,6 +406,9 @@ void KScheduler::EnableScheduling(KernelCore& kernel, u64 cores_needing_scheduli
     } else {
         RescheduleCores(kernel, cores_needing_scheduling);
     }
+
+    // Special case to ensure dummy threads that are waiting block.
+    current_thread->IfDummyThreadTryWait();
 }
 
 u64 KScheduler::UpdateHighestPriorityThreads(KernelCore& kernel) {
@@ -741,6 +742,12 @@ void KScheduler::ScheduleImpl() {
         next_thread = idle_thread;
     }
 
+    // We never want to schedule a dummy thread, as these are only used by host threads for locking.
+    if (next_thread->GetThreadType() == ThreadType::Dummy) {
+        ASSERT_MSG(false, "Dummy threads should never be scheduled!");
+        next_thread = idle_thread;
+    }
+
     // If we're not actually switching thread, there's nothing to do.
     if (next_thread == current_thread.load()) {
         previous_thread->EnableDispatch();

src/core/hle/kernel/k_server_session.cpp

@@ -8,7 +8,6 @@
 #include "common/assert.h"
 #include "common/common_types.h"
 #include "common/logging/log.h"
-#include "common/scope_exit.h"
 #include "core/core_timing.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/kernel/hle_ipc.h"
@@ -123,20 +122,10 @@ ResultCode KServerSession::QueueSyncRequest(KThread* thread, Core::Memory::Memor
 
     context->PopulateFromIncomingCommandBuffer(kernel.CurrentProcess()->GetHandleTable(), cmd_buf);
 
-    // In the event that something fails here, stub a result to prevent the game from crashing.
-    // This is a work-around in the event that somehow we process a service request after the
-    // session has been closed by the game. This has been observed to happen rarely in Pokemon
-    // Sword/Shield and is likely a result of us using host threads/scheduling for services.
-    // TODO(bunnei): Find a better solution here.
-    auto error_guard = SCOPE_GUARD({ CompleteSyncRequest(*context); });
-
     // Ensure we have a session request handler
     if (manager->HasSessionRequestHandler(*context)) {
         if (auto strong_ptr = manager->GetServiceThread().lock()) {
             strong_ptr->QueueSyncRequest(*parent, std::move(context));
-
-            // We succeeded.
-            error_guard.Cancel();
         } else {
             ASSERT_MSG(false, "strong_ptr is nullptr!");
         }
@@ -171,13 +160,8 @@ ResultCode KServerSession::CompleteSyncRequest(HLERequestContext& context) {
         convert_to_domain = false;
     }
 
-    // Some service requests require the thread to block
-    {
-        KScopedSchedulerLock lock(kernel);
-        if (!context.IsThreadWaiting()) {
-            context.GetThread().EndWait(result);
-        }
-    }
+    // The calling thread is waiting for this request to complete, so wake it up.
+    context.GetThread().EndWait(result);
 
     return result;
 }

src/core/hle/kernel/k_thread.cpp

@@ -30,6 +30,7 @@
 #include "core/hle/kernel/k_system_control.h"
 #include "core/hle/kernel/k_thread.h"
 #include "core/hle/kernel/k_thread_queue.h"
+#include "core/hle/kernel/k_worker_task_manager.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/svc_results.h"
 #include "core/hle/kernel/time_manager.h"
@@ -105,7 +106,7 @@ KThread::~KThread() = default;
 ResultCode KThread::Initialize(KThreadFunction func, uintptr_t arg, VAddr user_stack_top, s32 prio,
                                s32 virt_core, KProcess* owner, ThreadType type) {
     // Assert parameters are valid.
-    ASSERT((type == ThreadType::Main) ||
+    ASSERT((type == ThreadType::Main) || (type == ThreadType::Dummy) ||
            (Svc::HighestThreadPriority <= prio && prio <= Svc::LowestThreadPriority));
     ASSERT((owner != nullptr) || (type != ThreadType::User));
     ASSERT(0 <= virt_core && virt_core < static_cast<s32>(Common::BitSize<u64>()));
@@ -139,7 +140,7 @@ ResultCode KThread::Initialize(KThreadFunction func, uintptr_t arg, VAddr user_s
         UNREACHABLE_MSG("KThread::Initialize: Unknown ThreadType {}", static_cast<u32>(type));
         break;
     }
-    thread_type_for_debugging = type;
+    thread_type = type;
 
     // Set the ideal core ID and affinity mask.
     virtual_ideal_core_id = virt_core;
@@ -214,7 +215,6 @@ ResultCode KThread::Initialize(KThreadFunction func, uintptr_t arg, VAddr user_s
 
         parent = owner;
         parent->Open();
-        parent->IncrementThreadCount();
     }
 
     // Initialize thread context.
@@ -261,7 +261,7 @@ ResultCode KThread::InitializeThread(KThread* thread, KThreadFunction func, uint
 }
 
 ResultCode KThread::InitializeDummyThread(KThread* thread) {
-    return thread->Initialize({}, {}, {}, DefaultThreadPriority, 3, {}, ThreadType::Dummy);
+    return thread->Initialize({}, {}, {}, DummyThreadPriority, 3, {}, ThreadType::Dummy);
 }
 
 ResultCode KThread::InitializeIdleThread(Core::System& system, KThread* thread, s32 virt_core) {
@@ -326,13 +326,8 @@ void KThread::Finalize() {
         }
     }
 
-    // Decrement the parent process's thread count.
-    if (parent != nullptr) {
-        parent->DecrementThreadCount();
-    }
-
     // Perform inherited finalization.
-    KAutoObjectWithSlabHeapAndContainer<KThread, KSynchronizationObject>::Finalize();
+    KSynchronizationObject::Finalize();
 }
 
 bool KThread::IsSignaled() const {
@@ -376,11 +371,28 @@ void KThread::StartTermination() {
 
     // Register terminated dpc flag.
     RegisterDpc(DpcFlag::Terminated);
+}
+
+void KThread::FinishTermination() {
+    // Ensure that the thread is not executing on any core.
+    if (parent != nullptr) {
+        for (std::size_t i = 0; i < static_cast<std::size_t>(Core::Hardware::NUM_CPU_CORES); ++i) {
+            KThread* core_thread{};
+            do {
+                core_thread = kernel.Scheduler(i).GetCurrentThread();
+            } while (core_thread == this);
+        }
+    }
 
     // Close the thread.
     this->Close();
 }
 
+void KThread::DoWorkerTaskImpl() {
+    // Finish the termination that was begun by Exit().
+    this->FinishTermination();
+}
+
 void KThread::Pin(s32 current_core) {
     ASSERT(kernel.GlobalSchedulerContext().IsLocked());
 
@@ -417,12 +429,7 @@ void KThread::Pin(s32 current_core) {
                                          static_cast<u32>(ThreadState::SuspendShift)));
 
         // Update our state.
-        const ThreadState old_state = thread_state;
-        thread_state = static_cast<ThreadState>(GetSuspendFlags() |
-                                                static_cast<u32>(old_state & ThreadState::Mask));
-        if (thread_state != old_state) {
-            KScheduler::OnThreadStateChanged(kernel, this, old_state);
-        }
+        UpdateState();
     }
 
     // TODO(bunnei): Update our SVC access permissions.
@@ -463,20 +470,13 @@ void KThread::Unpin() {
     }
 
     // Allow performing thread suspension (if termination hasn't been requested).
-    {
+    if (!IsTerminationRequested()) {
         // Update our allow flags.
-        if (!IsTerminationRequested()) {
-            suspend_allowed_flags |= (1 << (static_cast<u32>(SuspendType::Thread) +
-                                            static_cast<u32>(ThreadState::SuspendShift)));
-        }
+        suspend_allowed_flags |= (1 << (static_cast<u32>(SuspendType::Thread) +
+                                        static_cast<u32>(ThreadState::SuspendShift)));
 
         // Update our state.
-        const ThreadState old_state = thread_state;
-        thread_state = static_cast<ThreadState>(GetSuspendFlags() |
-                                                static_cast<u32>(old_state & ThreadState::Mask));
-        if (thread_state != old_state) {
-            KScheduler::OnThreadStateChanged(kernel, this, old_state);
-        }
+        UpdateState();
     }
 
     // TODO(bunnei): Update our SVC access permissions.
@@ -689,12 +689,7 @@ void KThread::Resume(SuspendType type) {
         ~(1u << (static_cast<u32>(ThreadState::SuspendShift) + static_cast<u32>(type)));
 
     // Update our state.
-    const ThreadState old_state = thread_state;
-    thread_state = static_cast<ThreadState>(GetSuspendFlags() |
-                                            static_cast<u32>(old_state & ThreadState::Mask));
-    if (thread_state != old_state) {
-        KScheduler::OnThreadStateChanged(kernel, this, old_state);
-    }
+    this->UpdateState();
 }
 
 void KThread::WaitCancel() {
@@ -721,19 +716,22 @@ void KThread::TrySuspend() {
     ASSERT(GetNumKernelWaiters() == 0);
 
     // Perform the suspend.
-    Suspend();
+    this->UpdateState();
 }
 
-void KThread::Suspend() {
+void KThread::UpdateState() {
     ASSERT(kernel.GlobalSchedulerContext().IsLocked());
-    ASSERT(IsSuspendRequested());
 
     // Set our suspend flags in state.
     const auto old_state = thread_state;
-    thread_state = static_cast<ThreadState>(GetSuspendFlags()) | (old_state & ThreadState::Mask);
+    const auto new_state =
+        static_cast<ThreadState>(this->GetSuspendFlags()) | (old_state & ThreadState::Mask);
+    thread_state = new_state;
 
     // Note the state change in scheduler.
-    KScheduler::OnThreadStateChanged(kernel, this, old_state);
+    if (new_state != old_state) {
+        KScheduler::OnThreadStateChanged(kernel, this, old_state);
+    }
 }
 
 void KThread::Continue() {
@@ -998,13 +996,16 @@ ResultCode KThread::Run() {
 
         // If the current thread has been asked to suspend, suspend it and retry.
         if (GetCurrentThread(kernel).IsSuspended()) {
-            GetCurrentThread(kernel).Suspend();
+            GetCurrentThread(kernel).UpdateState();
             continue;
         }
 
         // If we're not a kernel thread and we've been asked to suspend, suspend ourselves.
-        if (IsUserThread() && IsSuspended()) {
-            Suspend();
+        if (KProcess* owner = this->GetOwnerProcess(); owner != nullptr) {
+            if (IsUserThread() && IsSuspended()) {
+                this->UpdateState();
+            }
+            owner->IncrementRunningThreadCount();
         }
 
         // Set our state and finish.
@@ -1019,21 +1020,29 @@ ResultCode KThread::Run() {
 void KThread::Exit() {
     ASSERT(this == GetCurrentThreadPointer(kernel));
 
-    // Release the thread resource hint from parent.
+    // Release the thread resource hint, running thread count from parent.
     if (parent != nullptr) {
         parent->GetResourceLimit()->Release(Kernel::LimitableResource::Threads, 0, 1);
         resource_limit_release_hint = true;
+        parent->DecrementRunningThreadCount();
     }
 
     // Perform termination.
     {
         KScopedSchedulerLock sl{kernel};
 
+        // Disallow all suspension.
+        suspend_allowed_flags = 0;
+        this->UpdateState();
+
         // Disallow all suspension.
         suspend_allowed_flags = 0;
 
         // Start termination.
         StartTermination();
+
+        // Register the thread as a work task.
+        KWorkerTaskManager::AddTask(kernel, KWorkerTaskManager::WorkerType::Exit, this);
     }
 }
 
@@ -1061,12 +1070,46 @@ ResultCode KThread::Sleep(s64 timeout) {
     return ResultSuccess;
 }
 
+void KThread::IfDummyThreadTryWait() {
+    if (!IsDummyThread()) {
+        return;
+    }
+
+    if (GetState() != ThreadState::Waiting) {
+        return;
+    }
+
+    // Block until we can grab the lock.
+    KScopedSpinLock lk{dummy_wait_lock};
+}
+
+void KThread::IfDummyThreadBeginWait() {
+    if (!IsDummyThread()) {
+        return;
+    }
+
+    // Ensure the thread will block when IfDummyThreadTryWait is called.
+    dummy_wait_lock.Lock();
+}
+
+void KThread::IfDummyThreadEndWait() {
+    if (!IsDummyThread()) {
+        return;
+    }
+
+    // Ensure the thread will no longer block.
+    dummy_wait_lock.Unlock();
+}
+
 void KThread::BeginWait(KThreadQueue* queue) {
     // Set our state as waiting.
     SetState(ThreadState::Waiting);
 
     // Set our wait queue.
     wait_queue = queue;
+
+    // Special case for dummy threads to ensure they block.
+    IfDummyThreadBeginWait();
 }
 
 void KThread::NotifyAvailable(KSynchronizationObject* signaled_object, ResultCode wait_result_) {
@@ -1085,7 +1128,16 @@ void KThread::EndWait(ResultCode wait_result_) {
 
     // If we're waiting, notify our queue that we're available.
     if (GetState() == ThreadState::Waiting) {
+        if (wait_queue == nullptr) {
+            // This should never happen, but avoid a hard crash below to get this logged.
+            ASSERT_MSG(false, "wait_queue is nullptr!");
+            return;
+        }
+
         wait_queue->EndWait(this, wait_result_);
+
+        // Special case for dummy threads to wakeup if necessary.
+        IfDummyThreadEndWait();
     }
 }
 

src/core/hle/kernel/k_thread.h

@@ -19,6 +19,7 @@
 #include "core/hle/kernel/k_light_lock.h"
 #include "core/hle/kernel/k_spin_lock.h"
 #include "core/hle/kernel/k_synchronization_object.h"
+#include "core/hle/kernel/k_worker_task.h"
 #include "core/hle/kernel/slab_helpers.h"
 #include "core/hle/kernel/svc_common.h"
 #include "core/hle/kernel/svc_types.h"
@@ -100,7 +101,7 @@ enum class ThreadWaitReasonForDebugging : u32 {
 [[nodiscard]] KThread& GetCurrentThread(KernelCore& kernel);
 [[nodiscard]] s32 GetCurrentCoreId(KernelCore& kernel);
 
-class KThread final : public KAutoObjectWithSlabHeapAndContainer<KThread, KSynchronizationObject>,
+class KThread final : public KAutoObjectWithSlabHeapAndContainer<KThread, KWorkerTask>,
                       public boost::intrusive::list_base_hook<> {
     KERNEL_AUTOOBJECT_TRAITS(KThread, KSynchronizationObject);
 
@@ -111,6 +112,7 @@ private:
 public:
     static constexpr s32 DefaultThreadPriority = 44;
     static constexpr s32 IdleThreadPriority = Svc::LowestThreadPriority + 1;
+    static constexpr s32 DummyThreadPriority = Svc::LowestThreadPriority + 2;
 
     explicit KThread(KernelCore& kernel_);
     ~KThread() override;
@@ -192,9 +194,9 @@ public:
 
     void TrySuspend();
 
-    void Continue();
+    void UpdateState();
 
-    void Suspend();
+    void Continue();
 
     constexpr void SetSyncedIndex(s32 index) {
         synced_index = index;
@@ -385,6 +387,8 @@ public:
 
     void OnTimer();
 
+    void DoWorkerTaskImpl();
+
     static void PostDestroy(uintptr_t arg);
 
     [[nodiscard]] static ResultCode InitializeDummyThread(KThread* thread);
@@ -550,8 +554,12 @@ public:
         return wait_reason_for_debugging;
     }
 
-    [[nodiscard]] ThreadType GetThreadTypeForDebugging() const {
-        return thread_type_for_debugging;
+    [[nodiscard]] ThreadType GetThreadType() const {
+        return thread_type;
+    }
+
+    [[nodiscard]] bool IsDummyThread() const {
+        return GetThreadType() == ThreadType::Dummy;
     }
 
     void SetWaitObjectsForDebugging(const std::span<KSynchronizationObject*>& objects) {
@@ -628,6 +636,14 @@ public:
         return condvar_key;
     }
 
+    // Dummy threads (used for HLE host threads) cannot wait based on the guest scheduler, and
+    // therefore will not block on guest kernel synchronization primitives. These methods handle
+    // blocking as needed.
+
+    void IfDummyThreadTryWait();
+    void IfDummyThreadBeginWait();
+    void IfDummyThreadEndWait();
+
 private:
     static constexpr size_t PriorityInheritanceCountMax = 10;
     union SyncObjectBuffer {
@@ -679,6 +695,8 @@ private:
 
     void StartTermination();
 
+    void FinishTermination();
+
     [[nodiscard]] ResultCode Initialize(KThreadFunction func, uintptr_t arg, VAddr user_stack_top,
                                         s32 prio, s32 virt_core, KProcess* owner, ThreadType type);
 
@@ -744,16 +762,17 @@ private:
     bool resource_limit_release_hint{};
     StackParameters stack_parameters{};
     KSpinLock context_guard{};
+    KSpinLock dummy_wait_lock{};
 
     // For emulation
     std::shared_ptr<Common::Fiber> host_context{};
     bool is_single_core{};
+    ThreadType thread_type{};
 
     // For debugging
     std::vector<KSynchronizationObject*> wait_objects_for_debugging;
     VAddr mutex_wait_address_for_debugging{};
     ThreadWaitReasonForDebugging wait_reason_for_debugging{};
-    ThreadType thread_type_for_debugging{};
 
 public:
     using ConditionVariableThreadTreeType = ConditionVariableThreadTree;

src/core/hle/kernel/k_worker_task.h

@@ -0,0 +1,18 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include "core/hle/kernel/k_synchronization_object.h"
+
+namespace Kernel {
+
+class KWorkerTask : public KSynchronizationObject {
+public:
+    explicit KWorkerTask(KernelCore& kernel_);
+
+    void DoWorkerTask();
+};
+
+} // namespace Kernel

src/core/hle/kernel/k_worker_task_manager.cpp

@@ -0,0 +1,42 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "common/assert.h"
+#include "core/hle/kernel/k_process.h"
+#include "core/hle/kernel/k_thread.h"
+#include "core/hle/kernel/k_worker_task.h"
+#include "core/hle/kernel/k_worker_task_manager.h"
+#include "core/hle/kernel/kernel.h"
+
+namespace Kernel {
+
+KWorkerTask::KWorkerTask(KernelCore& kernel_) : KSynchronizationObject{kernel_} {}
+
+void KWorkerTask::DoWorkerTask() {
+    if (auto* const thread = this->DynamicCast<KThread*>(); thread != nullptr) {
+        return thread->DoWorkerTaskImpl();
+    } else {
+        auto* const process = this->DynamicCast<KProcess*>();
+        ASSERT(process != nullptr);
+
+        return process->DoWorkerTaskImpl();
+    }
+}
+
+KWorkerTaskManager::KWorkerTaskManager() : m_waiting_thread(1, "yuzu:KWorkerTaskManager") {}
+
+void KWorkerTaskManager::AddTask(KernelCore& kernel, WorkerType type, KWorkerTask* task) {
+    ASSERT(type <= WorkerType::Count);
+    kernel.WorkerTaskManager().AddTask(kernel, task);
+}
+
+void KWorkerTaskManager::AddTask(KernelCore& kernel, KWorkerTask* task) {
+    KScopedSchedulerLock sl(kernel);
+    m_waiting_thread.QueueWork([task]() {
+        // Do the task.
+        task->DoWorkerTask();
+    });
+}
+
+} // namespace Kernel

src/core/hle/kernel/k_worker_task_manager.h

@@ -0,0 +1,33 @@
+// Copyright 2022 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include "common/common_types.h"
+#include "common/thread_worker.h"
+
+namespace Kernel {
+
+class KernelCore;
+class KWorkerTask;
+
+class KWorkerTaskManager final {
+public:
+    enum class WorkerType : u32 {
+        Exit,
+        Count,
+    };
+
+    KWorkerTaskManager();
+
+    static void AddTask(KernelCore& kernel_, WorkerType type, KWorkerTask* task);
+
+private:
+    void AddTask(KernelCore& kernel, KWorkerTask* task);
+
+private:
+    Common::ThreadWorker m_waiting_thread;
+};
+
+} // namespace Kernel

src/core/hle/kernel/kernel.cpp

@@ -37,6 +37,7 @@
 #include "core/hle/kernel/k_shared_memory.h"
 #include "core/hle/kernel/k_slab_heap.h"
 #include "core/hle/kernel/k_thread.h"
+#include "core/hle/kernel/k_worker_task_manager.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/physical_core.h"
 #include "core/hle/kernel/service_thread.h"
@@ -300,12 +301,10 @@ struct KernelCore::Impl {
     // Gets the dummy KThread for the caller, allocating a new one if this is the first time
     KThread* GetHostDummyThread() {
         auto make_thread = [this]() {
-            std::lock_guard lk(dummy_thread_lock);
-            auto& thread = dummy_threads.emplace_back(std::make_unique<KThread>(system.Kernel()));
-            KAutoObject::Create(thread.get());
-            ASSERT(KThread::InitializeDummyThread(thread.get()).IsSuccess());
+            KThread* thread = KThread::Create(system.Kernel());
+            ASSERT(KThread::InitializeDummyThread(thread).IsSuccess());
             thread->SetName(fmt::format("DummyThread:{}", GetHostThreadId()));
-            return thread.get();
+            return thread;
         };
 
         thread_local KThread* saved_thread = make_thread();
@@ -630,7 +629,7 @@ struct KernelCore::Impl {
         const auto application_pool = memory_layout.GetKernelApplicationPoolRegionPhysicalExtents();
 
         // Initialize memory managers
-        memory_manager = std::make_unique<KMemoryManager>();
+        memory_manager = std::make_unique<KMemoryManager>(system);
         memory_manager->InitializeManager(KMemoryManager::Pool::Application,
                                           application_pool.GetAddress(),
                                           application_pool.GetEndAddress());
@@ -730,7 +729,6 @@ struct KernelCore::Impl {
     std::mutex server_sessions_lock;
     std::mutex registered_objects_lock;
     std::mutex registered_in_use_objects_lock;
-    std::mutex dummy_thread_lock;
 
     std::atomic<u32> next_object_id{0};
     std::atomic<u64> next_kernel_process_id{KProcess::InitialKIPIDMin};
@@ -787,9 +785,6 @@ struct KernelCore::Impl {
     std::array<Core::CPUInterruptHandler, Core::Hardware::NUM_CPU_CORES> interrupts{};
     std::array<std::unique_ptr<Kernel::KScheduler>, Core::Hardware::NUM_CPU_CORES> schedulers{};
 
-    // Specifically tracked to be automatically destroyed with kernel
-    std::vector<std::unique_ptr<KThread>> dummy_threads;
-
     bool is_multicore{};
     std::atomic_bool is_shutting_down{};
     bool is_phantom_mode_for_singlecore{};
@@ -797,6 +792,8 @@ struct KernelCore::Impl {
 
     std::array<u64, Core::Hardware::NUM_CPU_CORES> svc_ticks{};
 
+    KWorkerTaskManager worker_task_manager;
+
     // System context
     Core::System& system;
 };
@@ -1137,6 +1134,14 @@ const Init::KSlabResourceCounts& KernelCore::SlabResourceCounts() const {
     return impl->slab_resource_counts;
 }
 
+KWorkerTaskManager& KernelCore::WorkerTaskManager() {
+    return impl->worker_task_manager;
+}
+
+const KWorkerTaskManager& KernelCore::WorkerTaskManager() const {
+    return impl->worker_task_manager;
+}
+
 bool KernelCore::IsPhantomModeForSingleCore() const {
     return impl->IsPhantomModeForSingleCore();
 }

src/core/hle/kernel/kernel.h

@@ -52,6 +52,7 @@ class KSharedMemory;
 class KSharedMemoryInfo;
 class KThread;
 class KTransferMemory;
+class KWorkerTaskManager;
 class KWritableEvent;
 class KCodeMemory;
 class PhysicalCore;
@@ -343,6 +344,12 @@ public:
     /// Gets the current slab resource counts.
     const Init::KSlabResourceCounts& SlabResourceCounts() const;
 
+    /// Gets the current worker task manager, used for dispatching KThread/KProcess tasks.
+    KWorkerTaskManager& WorkerTaskManager();
+
+    /// Gets the current worker task manager, used for dispatching KThread/KProcess tasks.
+    const KWorkerTaskManager& WorkerTaskManager() const;
+
 private:
     friend class KProcess;
     friend class KThread;

src/core/hle/kernel/service_thread.cpp

@@ -12,6 +12,7 @@
 #include "common/scope_exit.h"
 #include "common/thread.h"
 #include "core/hle/kernel/k_session.h"
+#include "core/hle/kernel/k_thread.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/service_thread.h"
 
@@ -50,6 +51,10 @@ ServiceThread::Impl::Impl(KernelCore& kernel, std::size_t num_threads, const std
 
             kernel.RegisterHostThread();
 
+            // Ensure the dummy thread allocated for this host thread is closed on exit.
+            auto* dummy_thread = kernel.GetCurrentEmuThread();
+            SCOPE_EXIT({ dummy_thread->Close(); });
+
             while (true) {
                 std::function<void()> task;
 

src/core/hle/service/acc/acc_su.cpp

@@ -37,8 +37,8 @@ ACC_SU::ACC_SU(std::shared_ptr<Module> module_, std::shared_ptr<ProfileManager>
         {130, nullptr, "ActivateOpenContextRetention"},
         {140, &ACC_SU::ListQualifiedUsers, "ListQualifiedUsers"},
         {150, nullptr, "AuthenticateApplicationAsync"},
-        {151, nullptr, "Unknown151"},
-        {152, nullptr, "Unknown152"},
+        {151, nullptr, "EnsureSignedDeviceIdentifierCacheForNintendoAccountAsync"},
+        {152, nullptr, "LoadSignedDeviceIdentifierCacheForNintendoAccount"},
         {190, nullptr, "GetUserLastOpenedApplication"},
         {191, nullptr, "ActivateOpenContextHolder"},
         {200, nullptr, "BeginUserRegistration"},

src/core/hle/service/acc/acc_u1.cpp

@@ -37,8 +37,8 @@ ACC_U1::ACC_U1(std::shared_ptr<Module> module_, std::shared_ptr<ProfileManager>
         {130, nullptr, "ActivateOpenContextRetention"},
         {140, &ACC_U1::ListQualifiedUsers, "ListQualifiedUsers"},
         {150, nullptr, "AuthenticateApplicationAsync"},
-        {151, nullptr, "Unknown151"},
-        {152, nullptr, "Unknown152"},
+        {151, nullptr, "EnsureSignedDeviceIdentifierCacheForNintendoAccountAsync"},
+        {152, nullptr, "LoadSignedDeviceIdentifierCacheForNintendoAccount"},
         {190, nullptr, "GetUserLastOpenedApplication"},
         {191, nullptr, "ActivateOpenContextHolder"},
         {997, nullptr, "DebugInvalidateTokenCacheForUser"},

src/core/hle/service/am/omm.cpp

@@ -37,6 +37,7 @@ OMM::OMM(Core::System& system_) : ServiceFramework{system_, "omm"} {
         {25, nullptr, "SetApplicationCecSettingsAndNotifyChanged"},
         {26, nullptr, "GetOperationModeSystemInfo"},
         {27, nullptr, "GetAppletFullAwakingSystemEvent"},
+        {28, nullptr, "CreateCradleFirmwareUpdater"},
     };
     // clang-format on
 

src/core/hle/service/apm/apm_interface.cpp

@@ -17,7 +17,7 @@ public:
         static const FunctionInfo functions[] = {
             {0, &ISession::SetPerformanceConfiguration, "SetPerformanceConfiguration"},
             {1, &ISession::GetPerformanceConfiguration, "GetPerformanceConfiguration"},
-            {2, nullptr, "SetCpuOverclockEnabled"},
+            {2, &ISession::SetCpuOverclockEnabled, "SetCpuOverclockEnabled"},
         };
         RegisterHandlers(functions);
     }
@@ -47,6 +47,18 @@ private:
         rb.PushEnum(controller.GetCurrentPerformanceConfiguration(mode));
     }
 
+    void SetCpuOverclockEnabled(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+
+        const auto cpu_overclock_enabled = rp.Pop<bool>();
+
+        LOG_WARNING(Service_APM, "(STUBBED) called, cpu_overclock_enabled={}",
+                    cpu_overclock_enabled);
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(ResultSuccess);
+    }
+
     Controller& controller;
 };
 

src/core/hle/service/audio/audctl.cpp

@@ -41,14 +41,14 @@ AudCtl::AudCtl(Core::System& system_) : ServiceFramework{system_, "audctl"} {
         {27, nullptr, "SetVolumeMappingTableForDev"},
         {28, nullptr, "GetAudioOutputChannelCountForPlayReport"},
         {29, nullptr, "BindAudioOutputChannelCountUpdateEventForPlayReport"},
-        {30, nullptr, "Unknown30"},
-        {31, nullptr, "Unknown31"},
-        {32, nullptr, "Unknown32"},
-        {33, nullptr, "Unknown33"},
-        {34, nullptr, "Unknown34"},
-        {10000, nullptr, "Unknown10000"},
-        {10001, nullptr, "Unknown10001"},
-        {10002, nullptr, "Unknown10002"},
+        {30, nullptr, "SetSpeakerAutoMuteEnabled"},
+        {31, nullptr, "IsSpeakerAutoMuteEnabled"},
+        {32, nullptr, "GetActiveOutputTarget"},
+        {33, nullptr, "GetTargetDeviceInfo"},
+        {34, nullptr, "AcquireTargetNotification"},
+        {10000, nullptr, "NotifyAudioOutputTargetForPlayReport"},
+        {10001, nullptr, "NotifyAudioOutputChannelCountForPlayReport"},
+        {10002, nullptr, "NotifyUnsupportedUsbOutputDeviceAttachedForPlayReport"},
     };
     // clang-format on
 

src/core/hle/service/btm/btm.cpp

@@ -201,6 +201,22 @@ public:
             {62, nullptr, "Unknown62"},
             {63, nullptr, "Unknown63"},
             {64, nullptr, "Unknown64"},
+            {65, nullptr, "Unknown65"},
+            {66, nullptr, "Unknown66"},
+            {67, nullptr, "Unknown67"},
+            {68, nullptr, "Unknown68"},
+            {69, nullptr, "Unknown69"},
+            {70, nullptr, "Unknown70"},
+            {71, nullptr, "Unknown71"},
+            {72, nullptr, "Unknown72"},
+            {73, nullptr, "Unknown73"},
+            {74, nullptr, "Unknown74"},
+            {75, nullptr, "Unknown75"},
+            {76, nullptr, "Unknown76"},
+            {100, nullptr, "Unknown100"},
+            {101, nullptr, "Unknown101"},
+            {110, nullptr, "Unknown102"},
+            {111, nullptr, "Unknown103"},
         };
         // clang-format on
 
@@ -249,6 +265,20 @@ public:
             {7, nullptr, "AcquireRadioEvent"},
             {8, nullptr, "AcquireGamepadPairingEvent"},
             {9, nullptr, "IsGamepadPairingStarted"},
+            {10, nullptr, "StartAudioDeviceDiscovery"},
+            {11, nullptr, "StopAudioDeviceDiscovery"},
+            {12, nullptr, "IsDiscoveryingAudioDevice"},
+            {13, nullptr, "GetDiscoveredAudioDevice"},
+            {14, nullptr, "AcquireAudioDeviceConnectionEvent"},
+            {15, nullptr, "ConnectAudioDevice"},
+            {16, nullptr, "IsConnectingAudioDevice"},
+            {17, nullptr, "GetConnectedAudioDevices"},
+            {18, nullptr, "DisconnectAudioDevice"},
+            {19, nullptr, "AcquirePairedAudioDeviceInfoChangedEvent"},
+            {20, nullptr, "GetPairedAudioDevices"},
+            {21, nullptr, "RemoveAudioDevicePairing"},
+            {22, nullptr, "RequestAudioDeviceConnectionRejection"},
+            {23, nullptr, "CancelAudioDeviceConnectionRejection"}
         };
         // clang-format on
 

src/core/hle/service/filesystem/fsp_srv.cpp

@@ -744,6 +744,7 @@ FSP_SRV::FSP_SRV(Core::System& system_)
         {203, &FSP_SRV::OpenPatchDataStorageByCurrentProcess, "OpenPatchDataStorageByCurrentProcess"},
         {204, nullptr, "OpenDataFileSystemByProgramIndex"},
         {205, &FSP_SRV::OpenDataStorageWithProgramIndex, "OpenDataStorageWithProgramIndex"},
+        {206, nullptr, "OpenDataStorageByPath"},
         {400, nullptr, "OpenDeviceOperator"},
         {500, nullptr, "OpenSdCardDetectionEventNotifier"},
         {501, nullptr, "OpenGameCardDetectionEventNotifier"},
@@ -796,6 +797,8 @@ FSP_SRV::FSP_SRV(Core::System& system_)
         {1014, nullptr, "OutputMultiProgramTagAccessLog"},
         {1016, nullptr, "FlushAccessLogOnSdCard"},
         {1017, nullptr, "OutputApplicationInfoAccessLog"},
+        {1018, nullptr, "SetDebugOption"},
+        {1019, nullptr, "UnsetDebugOption"},
         {1100, nullptr, "OverrideSaveDataTransferTokenSignVerificationKey"},
         {1110, nullptr, "CorruptSaveDataFileSystemBySaveDataSpaceId2"},
         {1200, &FSP_SRV::OpenMultiCommitManager, "OpenMultiCommitManager"},

src/core/hle/service/friend/friend.cpp

@@ -27,13 +27,13 @@ public:
             {10101, &IFriendService::GetFriendList, "GetFriendList"},
             {10102, nullptr, "UpdateFriendInfo"},
             {10110, nullptr, "GetFriendProfileImage"},
-            {10120, nullptr, "Unknown10120"},
-            {10121, nullptr, "Unknown10121"},
+            {10120, nullptr, "IsFriendListCacheAvailable"},
+            {10121, nullptr, "EnsureFriendListAvailable"},
             {10200, nullptr, "SendFriendRequestForApplication"},
             {10211, nullptr, "AddFacedFriendRequestForApplication"},
             {10400, &IFriendService::GetBlockedUserListIds, "GetBlockedUserListIds"},
-            {10420, nullptr, "Unknown10420"},
-            {10421, nullptr, "Unknown10421"},
+            {10420, nullptr, "IsBlockedUserListCacheAvailable"},
+            {10421, nullptr, "EnsureBlockedUserListAvailable"},
             {10500, nullptr, "GetProfileList"},
             {10600, nullptr, "DeclareOpenOnlinePlaySession"},
             {10601, &IFriendService::DeclareCloseOnlinePlaySession, "DeclareCloseOnlinePlaySession"},
@@ -103,8 +103,8 @@ public:
             {30900, nullptr, "SendFriendInvitation"},
             {30910, nullptr, "ReadFriendInvitation"},
             {30911, nullptr, "ReadAllFriendInvitations"},
-            {40100, nullptr, "Unknown40100"},
-            {40400, nullptr, "Unknown40400"},
+            {40100, nullptr, "DeleteFriendListCache"},
+            {40400, nullptr, "DeleteBlockedUserListCache"},
             {49900, nullptr, "DeleteNetworkServiceAccountCache"},
         };
         // clang-format on

src/core/hle/service/hid/hid.cpp

@@ -1176,7 +1176,8 @@ void Hid::SetNpadAnalogStickUseCenterClamp(Kernel::HLERequestContext& ctx) {
 
     const auto parameters{rp.PopRaw<Parameters>()};
 
-    applet_resource->GetController<Controller_NPad>(HidController::NPad)
+    GetAppletResource()
+        ->GetController<Controller_NPad>(HidController::NPad)
         .SetAnalogStickUseCenterClamp(parameters.analog_stick_use_center_clamp);
 
     LOG_WARNING(Service_HID,

src/core/hle/service/nim/nim.cpp

@@ -211,6 +211,11 @@ public:
             {127, nullptr, "Unknown127"},
             {128, nullptr, "Unknown128"},
             {129, nullptr, "Unknown129"},
+            {130, nullptr, "Unknown130"},
+            {131, nullptr, "Unknown131"},
+            {132, nullptr, "Unknown132"},
+            {133, nullptr, "Unknown133"},
+            {134, nullptr, "Unknown134"},
         };
         // clang-format on
 
@@ -287,6 +292,7 @@ public:
             {502, nullptr, "RequestDownloadTicketForPrepurchasedContents"},
             {503, nullptr, "RequestSyncTicket"},
             {504, nullptr, "RequestDownloadTicketForPrepurchasedContents2"},
+            {505, nullptr, "RequestDownloadTicketForPrepurchasedContentsForAccount"},
         };
         // clang-format on
 

src/core/hle/service/ns/ns.cpp

@@ -158,6 +158,7 @@ IApplicationManagerInterface::IApplicationManagerInterface(Core::System& system_
         {605, nullptr, "ListApplicationContentMetaStatusWithRightsCheck"},
         {606, nullptr, "GetContentMetaStorage"},
         {607, nullptr, "ListAvailableAddOnContent"},
+        {609, nullptr, "ListAvailabilityAssuredAddOnContent"},
         {700, nullptr, "PushDownloadTaskList"},
         {701, nullptr, "ClearTaskStatusList"},
         {702, nullptr, "RequestDownloadTaskList"},
@@ -289,6 +290,11 @@ IApplicationManagerInterface::IApplicationManagerInterface(Core::System& system_
         {2514, nullptr, "ClearTaskOfAsyncTaskManager"},
         {2515, nullptr, "CleanupAllPlaceHolderAndFragmentsIfNoTask"},
         {2516, nullptr, "EnsureApplicationCertificate"},
+        {2517, nullptr, "CreateApplicationInstance"},
+        {2518, nullptr, "UpdateQualificationForDebug"},
+        {2519, nullptr, "IsQualificationTransitionSupported"},
+        {2520, nullptr, "IsQualificationTransitionSupportedByProcessId"},
+        {2521, nullptr, "GetRightsUserChangedEvent"},
         {2800, nullptr, "GetApplicationIdOfPreomia"},
         {3000, nullptr, "RegisterDeviceLockKey"},
         {3001, nullptr, "UnregisterDeviceLockKey"},

src/core/hle/service/set/set_sys.cpp

@@ -307,6 +307,8 @@ SET_SYS::SET_SYS(Core::System& system_) : ServiceFramework{system_, "set:sys"} {
         {202, nullptr, "SetFieldTestingFlag"},
         {203, nullptr, "GetPanelCrcMode"},
         {204, nullptr, "SetPanelCrcMode"},
+        {205, nullptr, "GetNxControllerSettingsEx"},
+        {206, nullptr, "SetNxControllerSettingsEx"},
     };
     // clang-format on
 

src/core/hle/service/usb/usb.cpp

@@ -17,19 +17,19 @@ public:
     explicit IDsInterface(Core::System& system_) : ServiceFramework{system_, "IDsInterface"} {
         // clang-format off
         static const FunctionInfo functions[] = {
-            {0, nullptr, "GetDsEndpoint"},
-            {1, nullptr, "GetSetupEvent"},
-            {2, nullptr, "Unknown2"},
-            {3, nullptr, "EnableInterface"},
-            {4, nullptr, "DisableInterface"},
-            {5, nullptr, "CtrlInPostBufferAsync"},
-            {6, nullptr, "CtrlOutPostBufferAsync"},
-            {7, nullptr, "GetCtrlInCompletionEvent"},
-            {8, nullptr, "GetCtrlInReportData"},
-            {9, nullptr, "GetCtrlOutCompletionEvent"},
-            {10, nullptr, "GetCtrlOutReportData"},
-            {11, nullptr, "StallCtrl"},
-            {12, nullptr, "AppendConfigurationData"},
+            {0, nullptr, "BindDevice"},
+            {1, nullptr, "BindClientProcess"},
+            {2, nullptr, "AddInterface"},
+            {3, nullptr, "GetStateChangeEvent"},
+            {4, nullptr, "GetState"},
+            {5, nullptr, "ClearDeviceData"},
+            {6, nullptr, "AddUsbStringDescriptor"},
+            {7, nullptr, "DeleteUsbStringDescriptor"},
+            {8, nullptr, "SetUsbDeviceDescriptor"},
+            {9, nullptr, "SetBinaryObjectStore"},
+            {10, nullptr, "Enable"},
+            {11, nullptr, "Disable"},
+            {12, nullptr, "Unknown12"},
         };
         // clang-format on
 
@@ -42,19 +42,7 @@ public:
     explicit USB_DS(Core::System& system_) : ServiceFramework{system_, "usb:ds"} {
         // clang-format off
         static const FunctionInfo functions[] = {
-            {0, nullptr, "BindDevice"},
-            {1, nullptr, "BindClientProcess"},
-            {2, nullptr, "GetDsInterface"},
-            {3, nullptr, "GetStateChangeEvent"},
-            {4, nullptr, "GetState"},
-            {5, nullptr, "ClearDeviceData"},
-            {6, nullptr, "AddUsbStringDescriptor"},
-            {7, nullptr, "DeleteUsbStringDescriptor"},
-            {8, nullptr, "SetUsbDeviceDescriptor"},
-            {9, nullptr, "SetBinaryObjectStore"},
-            {10, nullptr, "Enable"},
-            {11, nullptr, "Disable"},
-            {12, nullptr, "Unknown12"},
+            {0, nullptr, "OpenDsService"},
         };
         // clang-format on
 
@@ -120,7 +108,7 @@ public:
             {5, nullptr, "DestroyInterfaceAvailableEvent"},
             {6, nullptr, "GetInterfaceStateChangeEvent"},
             {7, nullptr, "AcquireUsbIf"},
-            {8, nullptr, "Unknown8"},
+            {8, nullptr, "ResetDevice"},
         };
         // clang-format on
 

src/core/hle/service/wlan/wlan.cpp

@@ -53,6 +53,7 @@ public:
             {35, nullptr, "Unknown35"},
             {36, nullptr, "Unknown36"},
             {37, nullptr, "Unknown37"},
+            {38, nullptr, "Unknown38"},
         };
         // clang-format on
 
@@ -117,7 +118,6 @@ public:
             {49, nullptr, "Unknown49"},
             {50, nullptr, "Unknown50"},
             {51, nullptr, "Unknown51"},
-            {52, nullptr, "Unknown52"},
         };
         // clang-format on
 

src/input_common/drivers/mouse.cpp

@@ -16,6 +16,7 @@ constexpr int mouse_axis_x = 0;
 constexpr int mouse_axis_y = 1;
 constexpr int wheel_axis_x = 2;
 constexpr int wheel_axis_y = 3;
+constexpr int motion_wheel_y = 4;
 constexpr int touch_axis_x = 10;
 constexpr int touch_axis_y = 11;
 constexpr PadIdentifier identifier = {
@@ -30,8 +31,9 @@ Mouse::Mouse(std::string input_engine_) : InputEngine(std::move(input_engine_))
     PreSetAxis(identifier, mouse_axis_y);
     PreSetAxis(identifier, wheel_axis_x);
     PreSetAxis(identifier, wheel_axis_y);
+    PreSetAxis(identifier, motion_wheel_y);
     PreSetAxis(identifier, touch_axis_x);
-    PreSetAxis(identifier, touch_axis_x);
+    PreSetAxis(identifier, touch_axis_y);
     update_thread = std::jthread([this](std::stop_token stop_token) { UpdateThread(stop_token); });
 }
 
@@ -48,6 +50,8 @@ void Mouse::UpdateThread(std::stop_token stop_token) {
             SetAxis(identifier, mouse_axis_y, -last_mouse_change.y * sensitivity);
         }
 
+        SetAxis(identifier, motion_wheel_y, 0.0f);
+
         if (mouse_panning_timout++ > 20) {
             StopPanning();
         }
@@ -136,6 +140,7 @@ void Mouse::MouseWheelChange(int x, int y) {
     wheel_position.y += y;
     SetAxis(identifier, wheel_axis_x, static_cast<f32>(wheel_position.x));
     SetAxis(identifier, wheel_axis_y, static_cast<f32>(wheel_position.y));
+    SetAxis(identifier, motion_wheel_y, static_cast<f32>(y) / 100.0f);
 }
 
 void Mouse::ReleaseAllButtons() {
@@ -171,13 +176,39 @@ AnalogMapping Mouse::GetAnalogMappingForDevice(
     return mapping;
 }
 
+Common::Input::ButtonNames Mouse::GetUIButtonName(const Common::ParamPackage& params) const {
+    const auto button = static_cast<MouseButton>(params.Get("button", 0));
+    switch (button) {
+    case MouseButton::Left:
+        return Common::Input::ButtonNames::ButtonLeft;
+    case MouseButton::Right:
+        return Common::Input::ButtonNames::ButtonRight;
+    case MouseButton::Wheel:
+        return Common::Input::ButtonNames::ButtonMouseWheel;
+    case MouseButton::Backward:
+        return Common::Input::ButtonNames::ButtonBackward;
+    case MouseButton::Forward:
+        return Common::Input::ButtonNames::ButtonForward;
+    case MouseButton::Task:
+        return Common::Input::ButtonNames::ButtonTask;
+    case MouseButton::Extra:
+        return Common::Input::ButtonNames::ButtonExtra;
+    case MouseButton::Undefined:
+    default:
+        return Common::Input::ButtonNames::Undefined;
+    }
+}
+
 Common::Input::ButtonNames Mouse::GetUIName(const Common::ParamPackage& params) const {
     if (params.Has("button")) {
-        return Common::Input::ButtonNames::Value;
+        return GetUIButtonName(params);
     }
     if (params.Has("axis")) {
         return Common::Input::ButtonNames::Value;
     }
+    if (params.Has("axis_x") && params.Has("axis_y") && params.Has("axis_z")) {
+        return Common::Input::ButtonNames::Engine;
+    }
 
     return Common::Input::ButtonNames::Invalid;
 }

src/input_common/drivers/mouse.h

@@ -69,6 +69,8 @@ private:
     void UpdateThread(std::stop_token stop_token);
     void StopPanning();
 
+    Common::Input::ButtonNames GetUIButtonName(const Common::ParamPackage& params) const;
+
     Common::Vec2<int> mouse_origin;
     Common::Vec2<int> last_mouse_position;
     Common::Vec2<float> last_mouse_change;

src/input_common/drivers/sdl_driver.cpp

@@ -198,9 +198,9 @@ public:
         if (sdl_controller) {
             switch (SDL_GameControllerGetType(sdl_controller.get())) {
             case SDL_CONTROLLER_TYPE_XBOX360:
-                return "XBox 360 Controller";
+                return "Xbox 360 Controller";
             case SDL_CONTROLLER_TYPE_XBOXONE:
-                return "XBox One Controller";
+                return "Xbox One Controller";
             case SDL_CONTROLLER_TYPE_PS3:
                 return "DualShock 3 Controller";
             case SDL_CONTROLLER_TYPE_PS4:

src/input_common/drivers/udp_client.cpp

@@ -339,7 +339,7 @@ void UDPClient::StartCommunication(std::size_t client, const std::string& host,
     }
 }
 
-const PadIdentifier UDPClient::GetPadIdentifier(std::size_t pad_index) const {
+PadIdentifier UDPClient::GetPadIdentifier(std::size_t pad_index) const {
     const std::size_t client = pad_index / PADS_PER_CLIENT;
     return {
         .guid = clients[client].uuid,
@@ -348,9 +348,9 @@ const PadIdentifier UDPClient::GetPadIdentifier(std::size_t pad_index) const {
     };
 }
 
-const Common::UUID UDPClient::GetHostUUID(const std::string host) const {
-    const auto ip = boost::asio::ip::address_v4::from_string(host);
-    const auto hex_host = fmt::format("{:06x}", ip.to_ulong());
+Common::UUID UDPClient::GetHostUUID(const std::string& host) const {
+    const auto ip = boost::asio::ip::make_address_v4(host);
+    const auto hex_host = fmt::format("{:06x}", ip.to_uint());
     return Common::UUID{hex_host};
 }
 

src/input_common/drivers/udp_client.h

@@ -145,8 +145,8 @@ 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);
-    const PadIdentifier GetPadIdentifier(std::size_t pad_index) const;
-    const Common::UUID GetHostUUID(const std::string host) const;
+    PadIdentifier GetPadIdentifier(std::size_t pad_index) const;
+    Common::UUID GetHostUUID(const std::string& host) const;
 
     Common::Input::ButtonNames GetUIButtonName(const Common::ParamPackage& params) const;
 

src/input_common/input_engine.cpp

@@ -173,7 +173,7 @@ void InputEngine::ResetButtonState() {
             SetButton(controller.first, button.first, false);
         }
         for (const auto& button : controller.second.hat_buttons) {
-            SetHatButton(controller.first, button.first, false);
+            SetHatButton(controller.first, button.first, 0);
         }
     }
 }

src/input_common/input_engine.h

@@ -16,7 +16,7 @@
 
 // Pad Identifier of data source
 struct PadIdentifier {
-    Common::UUID guid{};
+    Common::UUID guid{Common::INVALID_UUID};
     std::size_t port{};
     std::size_t pad{};
 
@@ -89,7 +89,7 @@ struct UpdateCallback {
 
 // Triggered if data changed on the controller and the engine is on configuring mode
 struct MappingCallback {
-    std::function<void(MappingData)> on_data;
+    std::function<void(const MappingData&)> on_data;
 };
 
 // Input Identifier of data source

src/input_common/input_mapping.cpp

@@ -2,14 +2,13 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included
 
-#include "common/common_types.h"
 #include "common/settings.h"
 #include "input_common/input_engine.h"
 #include "input_common/input_mapping.h"
 
 namespace InputCommon {
 
-MappingFactory::MappingFactory() {}
+MappingFactory::MappingFactory() = default;
 
 void MappingFactory::BeginMapping(Polling::InputType type) {
     is_enabled = true;
@@ -19,7 +18,7 @@ void MappingFactory::BeginMapping(Polling::InputType type) {
     second_axis = -1;
 }
 
-[[nodiscard]] const Common::ParamPackage MappingFactory::GetNextInput() {
+Common::ParamPackage MappingFactory::GetNextInput() {
     Common::ParamPackage input;
     input_queue.Pop(input);
     return input;
@@ -57,7 +56,7 @@ void MappingFactory::StopMapping() {
 void MappingFactory::RegisterButton(const MappingData& data) {
     Common::ParamPackage new_input;
     new_input.Set("engine", data.engine);
-    if (data.pad.guid != Common::UUID{}) {
+    if (data.pad.guid.IsValid()) {
         new_input.Set("guid", data.pad.guid.Format());
     }
     new_input.Set("port", static_cast<int>(data.pad.port));
@@ -93,7 +92,7 @@ void MappingFactory::RegisterButton(const MappingData& data) {
 void MappingFactory::RegisterStick(const MappingData& data) {
     Common::ParamPackage new_input;
     new_input.Set("engine", data.engine);
-    if (data.pad.guid != Common::UUID{}) {
+    if (data.pad.guid.IsValid()) {
         new_input.Set("guid", data.pad.guid.Format());
     }
     new_input.Set("port", static_cast<int>(data.pad.port));
@@ -138,11 +137,24 @@ void MappingFactory::RegisterStick(const MappingData& data) {
 void MappingFactory::RegisterMotion(const MappingData& data) {
     Common::ParamPackage new_input;
     new_input.Set("engine", data.engine);
-    if (data.pad.guid != Common::UUID{}) {
+    if (data.pad.guid.IsValid()) {
         new_input.Set("guid", data.pad.guid.Format());
     }
     new_input.Set("port", static_cast<int>(data.pad.port));
     new_input.Set("pad", static_cast<int>(data.pad.pad));
+
+    // If engine is mouse map the mouse position as 3 axis motion
+    if (data.engine == "mouse") {
+        new_input.Set("axis_x", 1);
+        new_input.Set("invert_x", "-");
+        new_input.Set("axis_y", 0);
+        new_input.Set("axis_z", 4);
+        new_input.Set("range", 1.0f);
+        new_input.Set("deadzone", 0.0f);
+        input_queue.Push(new_input);
+        return;
+    }
+
     switch (data.type) {
     case EngineInputType::Button:
     case EngineInputType::HatButton:

src/input_common/input_mapping.h

@@ -3,8 +3,14 @@
 // Refer to the license.txt file included
 
 #pragma once
+
+#include "common/param_package.h"
 #include "common/threadsafe_queue.h"
 
+namespace InputCommon::Polling {
+enum class InputType;
+}
+
 namespace InputCommon {
 class InputEngine;
 struct MappingData;
@@ -20,7 +26,7 @@ public:
     void BeginMapping(Polling::InputType type);
 
     /// Returns an input event with mapping information from the input_queue
-    [[nodiscard]] const Common::ParamPackage GetNextInput();
+    [[nodiscard]] Common::ParamPackage GetNextInput();
 
     /**
      * Registers mapping input data from the driver

src/input_common/main.cpp

@@ -27,7 +27,7 @@ namespace InputCommon {
 struct InputSubsystem::Impl {
     void Initialize() {
         mapping_factory = std::make_shared<MappingFactory>();
-        MappingCallback mapping_callback{[this](MappingData data) { RegisterInput(data); }};
+        MappingCallback mapping_callback{[this](const MappingData& data) { RegisterInput(data); }};
 
         keyboard = std::make_shared<Keyboard>("keyboard");
         keyboard->SetMappingCallback(mapping_callback);
@@ -284,7 +284,7 @@ struct InputSubsystem::Impl {
 #endif
     }
 
-    void RegisterInput(MappingData data) {
+    void RegisterInput(const MappingData& data) {
         mapping_factory->RegisterInput(data);
     }
 
@@ -394,7 +394,7 @@ void InputSubsystem::BeginMapping(Polling::InputType type) {
     impl->mapping_factory->BeginMapping(type);
 }
 
-const Common::ParamPackage InputSubsystem::GetNextInput() const {
+Common::ParamPackage InputSubsystem::GetNextInput() const {
     return impl->mapping_factory->GetNextInput();
 }
 

src/input_common/main.h

@@ -126,7 +126,7 @@ public:
     void BeginMapping(Polling::InputType type);
 
     /// Returns an input event with mapping information.
-    [[nodiscard]] const Common::ParamPackage GetNextInput() const;
+    [[nodiscard]] Common::ParamPackage GetNextInput() const;
 
     /// Stop polling from all backends.
     void StopMapping() const;

src/shader_recompiler/backend/glsl/glsl_emit_context.cpp

@@ -458,9 +458,10 @@ void EmitContext::DefineGenericOutput(size_t index, u32 invocations) {
         std::string definition{fmt::format("layout(location={}", index)};
         const u32 remainder{4 - element};
         const TransformFeedbackVarying* xfb_varying{};
-        if (!runtime_info.xfb_varyings.empty()) {
-            xfb_varying = &runtime_info.xfb_varyings[base_index + element];
-            xfb_varying = xfb_varying && xfb_varying->components > 0 ? xfb_varying : nullptr;
+        const size_t xfb_varying_index{base_index + element};
+        if (xfb_varying_index < runtime_info.xfb_varyings.size()) {
+            xfb_varying = &runtime_info.xfb_varyings[xfb_varying_index];
+            xfb_varying = xfb_varying->components > 0 ? xfb_varying : nullptr;
         }
         const u32 num_components{xfb_varying ? xfb_varying->components : remainder};
         if (element > 0) {

src/shader_recompiler/backend/spirv/spirv_emit_context.cpp

@@ -164,9 +164,10 @@ void DefineGenericOutput(EmitContext& ctx, size_t index, std::optional<u32> invo
     while (element < 4) {
         const u32 remainder{4 - element};
         const TransformFeedbackVarying* xfb_varying{};
-        if (!ctx.runtime_info.xfb_varyings.empty()) {
-            xfb_varying = &ctx.runtime_info.xfb_varyings[base_attr_index + element];
-            xfb_varying = xfb_varying && xfb_varying->components > 0 ? xfb_varying : nullptr;
+        const size_t xfb_varying_index{base_attr_index + element};
+        if (xfb_varying_index < ctx.runtime_info.xfb_varyings.size()) {
+            xfb_varying = &ctx.runtime_info.xfb_varyings[xfb_varying_index];
+            xfb_varying = xfb_varying->components > 0 ? xfb_varying : nullptr;
         }
         const u32 num_components{xfb_varying ? xfb_varying->components : remainder};
 

src/shader_recompiler/frontend/maxwell/translate_program.h

@@ -21,7 +21,6 @@ namespace Shader::Maxwell {
 [[nodiscard]] IR::Program MergeDualVertexPrograms(IR::Program& vertex_a, IR::Program& vertex_b,
                                                   Environment& env_vertex_b);
 
-[[nodiscard]] void ConvertLegacyToGeneric(IR::Program& program,
-                                          const Shader::RuntimeInfo& runtime_info);
+void ConvertLegacyToGeneric(IR::Program& program, const RuntimeInfo& runtime_info);
 
 } // namespace Shader::Maxwell

src/video_core/command_classes/codecs/codec.h

@@ -66,7 +66,7 @@ private:
     bool initialized{};
     NvdecCommon::VideoCodec current_codec{NvdecCommon::VideoCodec::None};
 
-    AVCodec* av_codec{nullptr};
+    const AVCodec* av_codec{nullptr};
     AVCodecContext* av_codec_ctx{nullptr};
     AVBufferRef* av_gpu_decoder{nullptr};
 

src/video_core/engines/maxwell_3d.cpp

@@ -240,7 +240,7 @@ void Maxwell3D::CallMacroMethod(u32 method, const std::vector<u32>& parameters)
         ((method - MacroRegistersStart) >> 1) % static_cast<u32>(macro_positions.size());
 
     // Execute the current macro.
-    macro_engine->Execute(*this, macro_positions[entry], parameters);
+    macro_engine->Execute(macro_positions[entry], parameters);
     if (mme_draw.current_mode != MMEDrawMode::Undefined) {
         FlushMMEInlineDraw();
     }

src/video_core/gpu.h

@@ -12,9 +12,6 @@
 #include "video_core/framebuffer_config.h"
 
 namespace Core {
-namespace Frontend {
-class EmuWindow;
-}
 class System;
 } // namespace Core
 
@@ -25,7 +22,6 @@ class ShaderNotify;
 
 namespace Tegra {
 class DmaPusher;
-class CDmaPusher;
 struct CommandList;
 
 enum class RenderTargetFormat : u32 {
@@ -88,15 +84,9 @@ enum class DepthFormat : u32 {
     D32_FLOAT_S8X24_UINT = 0x19,
 };
 
-struct CommandListHeader;
-class DebugContext;
-
 namespace Engines {
-class Fermi2D;
 class Maxwell3D;
-class MaxwellDMA;
 class KeplerCompute;
-class KeplerMemory;
 } // namespace Engines
 
 enum class EngineID {
@@ -190,12 +180,6 @@ public:
     /// Returns a const reference to the GPU DMA pusher.
     [[nodiscard]] const Tegra::DmaPusher& DmaPusher() const;
 
-    /// Returns a reference to the GPU CDMA pusher.
-    [[nodiscard]] Tegra::CDmaPusher& CDmaPusher();
-
-    /// Returns a const reference to the GPU CDMA pusher.
-    [[nodiscard]] const Tegra::CDmaPusher& CDmaPusher() const;
-
     /// Returns a reference to the underlying renderer.
     [[nodiscard]] VideoCore::RendererBase& Renderer();
 

src/video_core/host_shaders/astc_decoder.comp

@@ -155,9 +155,6 @@ uint SwizzleOffset(uvec2 pos) {
 // Replicates low num_bits such that [(to_bit - 1):(to_bit - 1 - from_bit)]
 // is the same as [(num_bits - 1):0] and repeats all the way down.
 uint Replicate(uint val, uint num_bits, uint to_bit) {
-    if (num_bits == 0 || to_bit == 0) {
-        return 0;
-    }
     const uint v = val & uint((1 << num_bits) - 1);
     uint res = v;
     uint reslen = num_bits;
@@ -187,42 +184,57 @@ uint ReplicateBitTo9(uint value) {
     return REPLICATE_1_BIT_TO_9_TABLE[value];
 }
 
-uint FastReplicateTo8(uint value, uint num_bits) {
-    switch (num_bits) {
-    case 1:
-        return REPLICATE_1_BIT_TO_8_TABLE[value];
-    case 2:
-        return REPLICATE_2_BIT_TO_8_TABLE[value];
-    case 3:
-        return REPLICATE_3_BIT_TO_8_TABLE[value];
-    case 4:
-        return REPLICATE_4_BIT_TO_8_TABLE[value];
-    case 5:
-        return REPLICATE_5_BIT_TO_8_TABLE[value];
-    case 6:
-        return REPLICATE_6_BIT_TO_8_TABLE[value];
-    case 7:
-        return REPLICATE_7_BIT_TO_8_TABLE[value];
-    case 8:
+uint FastReplicate(uint value, uint num_bits, uint to_bit) {
+    if (num_bits == 0) {
+        return 0;
+    }
+    if (num_bits == to_bit) {
         return value;
     }
-    return Replicate(value, num_bits, 8);
+    if (to_bit == 6) {
+        switch (num_bits) {
+        case 1:
+            return REPLICATE_1_BIT_TO_6_TABLE[value];
+        case 2:
+            return REPLICATE_2_BIT_TO_6_TABLE[value];
+        case 3:
+            return REPLICATE_3_BIT_TO_6_TABLE[value];
+        case 4:
+            return REPLICATE_4_BIT_TO_6_TABLE[value];
+        case 5:
+            return REPLICATE_5_BIT_TO_6_TABLE[value];
+        default:
+            break;
+        }
+    } else { /* if (to_bit == 8) */
+        switch (num_bits) {
+        case 1:
+            return REPLICATE_1_BIT_TO_8_TABLE[value];
+        case 2:
+            return REPLICATE_2_BIT_TO_8_TABLE[value];
+        case 3:
+            return REPLICATE_3_BIT_TO_8_TABLE[value];
+        case 4:
+            return REPLICATE_4_BIT_TO_8_TABLE[value];
+        case 5:
+            return REPLICATE_5_BIT_TO_8_TABLE[value];
+        case 6:
+            return REPLICATE_6_BIT_TO_8_TABLE[value];
+        case 7:
+            return REPLICATE_7_BIT_TO_8_TABLE[value];
+        default:
+            break;
+        }
+    }
+    return Replicate(value, num_bits, to_bit);
+}
+
+uint FastReplicateTo8(uint value, uint num_bits) {
+    return FastReplicate(value, num_bits, 8);
 }
 
 uint FastReplicateTo6(uint value, uint num_bits) {
-    switch (num_bits) {
-    case 1:
-        return REPLICATE_1_BIT_TO_6_TABLE[value];
-    case 2:
-        return REPLICATE_2_BIT_TO_6_TABLE[value];
-    case 3:
-        return REPLICATE_3_BIT_TO_6_TABLE[value];
-    case 4:
-        return REPLICATE_4_BIT_TO_6_TABLE[value];
-    case 5:
-        return REPLICATE_5_BIT_TO_6_TABLE[value];
-    }
-    return Replicate(value, num_bits, 6);
+    return FastReplicate(value, num_bits, 6);
 }
 
 uint Div3Floor(uint v) {

src/video_core/macro/macro.cpp

@@ -2,12 +2,13 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <cstring>
 #include <optional>
+
 #include <boost/container_hash/hash.hpp>
+
 #include "common/assert.h"
-#include "common/logging/log.h"
 #include "common/settings.h"
-#include "video_core/engines/maxwell_3d.h"
 #include "video_core/macro/macro.h"
 #include "video_core/macro/macro_hle.h"
 #include "video_core/macro/macro_interpreter.h"
@@ -24,8 +25,7 @@ void MacroEngine::AddCode(u32 method, u32 data) {
     uploaded_macro_code[method].push_back(data);
 }
 
-void MacroEngine::Execute(Engines::Maxwell3D& maxwell3d, u32 method,
-                          const std::vector<u32>& parameters) {
+void MacroEngine::Execute(u32 method, const std::vector<u32>& parameters) {
     auto compiled_macro = macro_cache.find(method);
     if (compiled_macro != macro_cache.end()) {
         const auto& cache_info = compiled_macro->second;
@@ -66,10 +66,9 @@ void MacroEngine::Execute(Engines::Maxwell3D& maxwell3d, u32 method,
             cache_info.lle_program = Compile(code);
         }
 
-        auto hle_program = hle_macros->GetHLEProgram(cache_info.hash);
-        if (hle_program.has_value()) {
+        if (auto hle_program = hle_macros->GetHLEProgram(cache_info.hash)) {
             cache_info.has_hle_program = true;
-            cache_info.hle_program = std::move(hle_program.value());
+            cache_info.hle_program = std::move(hle_program);
             cache_info.hle_program->Execute(parameters, method);
         } else {
             cache_info.lle_program->Execute(parameters, method);

src/video_core/macro/macro.h

@@ -119,7 +119,7 @@ public:
     void AddCode(u32 method, u32 data);
 
     // Compiles the macro if its not in the cache, and executes the compiled macro
-    void Execute(Engines::Maxwell3D& maxwell3d, u32 method, const std::vector<u32>& parameters);
+    void Execute(u32 method, const std::vector<u32>& parameters);
 
 protected:
     virtual std::unique_ptr<CachedMacro> Compile(const std::vector<u32>& code) = 0;

src/video_core/macro/macro_hle.cpp

@@ -5,12 +5,15 @@
 #include <array>
 #include <vector>
 #include "video_core/engines/maxwell_3d.h"
+#include "video_core/macro/macro.h"
 #include "video_core/macro/macro_hle.h"
 #include "video_core/rasterizer_interface.h"
 
 namespace Tegra {
-
 namespace {
+
+using HLEFunction = void (*)(Engines::Maxwell3D& maxwell3d, const std::vector<u32>& parameters);
+
 // HLE'd functions
 void HLE_771BB18C62444DA0(Engines::Maxwell3D& maxwell3d, const std::vector<u32>& parameters) {
     const u32 instance_count = parameters[2] & maxwell3d.GetRegisterValue(0xD1B);
@@ -77,7 +80,6 @@ void HLE_0217920100488FF7(Engines::Maxwell3D& maxwell3d, const std::vector<u32>&
     maxwell3d.CallMethodFromMME(0x8e5, 0x0);
     maxwell3d.mme_draw.current_mode = Engines::Maxwell3D::MMEDrawMode::Undefined;
 }
-} // Anonymous namespace
 
 constexpr std::array<std::pair<u64, HLEFunction>, 3> hle_funcs{{
     {0x771BB18C62444DA0, &HLE_771BB18C62444DA0},
@@ -85,25 +87,31 @@ constexpr std::array<std::pair<u64, HLEFunction>, 3> hle_funcs{{
     {0x0217920100488FF7, &HLE_0217920100488FF7},
 }};
 
+class HLEMacroImpl final : public CachedMacro {
+public:
+    explicit HLEMacroImpl(Engines::Maxwell3D& maxwell3d_, HLEFunction func_)
+        : maxwell3d{maxwell3d_}, func{func_} {}
+
+    void Execute(const std::vector<u32>& parameters, u32 method) override {
+        func(maxwell3d, parameters);
+    }
+
+private:
+    Engines::Maxwell3D& maxwell3d;
+    HLEFunction func;
+};
+} // Anonymous namespace
+
 HLEMacro::HLEMacro(Engines::Maxwell3D& maxwell3d_) : maxwell3d{maxwell3d_} {}
 HLEMacro::~HLEMacro() = default;
 
-std::optional<std::unique_ptr<CachedMacro>> HLEMacro::GetHLEProgram(u64 hash) const {
+std::unique_ptr<CachedMacro> HLEMacro::GetHLEProgram(u64 hash) const {
     const auto it = std::find_if(hle_funcs.cbegin(), hle_funcs.cend(),
                                  [hash](const auto& pair) { return pair.first == hash; });
     if (it == hle_funcs.end()) {
-        return std::nullopt;
+        return nullptr;
     }
     return std::make_unique<HLEMacroImpl>(maxwell3d, it->second);
 }
 
-HLEMacroImpl::~HLEMacroImpl() = default;
-
-HLEMacroImpl::HLEMacroImpl(Engines::Maxwell3D& maxwell3d_, HLEFunction func_)
-    : maxwell3d{maxwell3d_}, func{func_} {}
-
-void HLEMacroImpl::Execute(const std::vector<u32>& parameters, u32 method) {
-    func(maxwell3d, parameters);
-}
-
 } // namespace Tegra

src/video_core/macro/macro_hle.h

@@ -5,10 +5,7 @@
 #pragma once
 
 #include <memory>
-#include <optional>
-#include <vector>
 #include "common/common_types.h"
-#include "video_core/macro/macro.h"
 
 namespace Tegra {
 
@@ -16,29 +13,17 @@ namespace Engines {
 class Maxwell3D;
 }
 
-using HLEFunction = void (*)(Engines::Maxwell3D& maxwell3d, const std::vector<u32>& parameters);
-
 class HLEMacro {
 public:
     explicit HLEMacro(Engines::Maxwell3D& maxwell3d_);
     ~HLEMacro();
 
-    std::optional<std::unique_ptr<CachedMacro>> GetHLEProgram(u64 hash) const;
+    // Allocates and returns a cached macro if the hash matches a known function.
+    // Returns nullptr otherwise.
+    [[nodiscard]] std::unique_ptr<CachedMacro> GetHLEProgram(u64 hash) const;
 
 private:
     Engines::Maxwell3D& maxwell3d;
 };
 
-class HLEMacroImpl : public CachedMacro {
-public:
-    explicit HLEMacroImpl(Engines::Maxwell3D& maxwell3d, HLEFunction func);
-    ~HLEMacroImpl();
-
-    void Execute(const std::vector<u32>& parameters, u32 method) override;
-
-private:
-    Engines::Maxwell3D& maxwell3d;
-    HLEFunction func;
-};
-
 } // namespace Tegra

src/video_core/macro/macro_interpreter.cpp

@@ -2,6 +2,9 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <array>
+#include <optional>
+
 #include "common/assert.h"
 #include "common/logging/log.h"
 #include "common/microprofile.h"
@@ -11,16 +14,81 @@
 MICROPROFILE_DEFINE(MacroInterp, "GPU", "Execute macro interpreter", MP_RGB(128, 128, 192));
 
 namespace Tegra {
-MacroInterpreter::MacroInterpreter(Engines::Maxwell3D& maxwell3d_)
-    : MacroEngine{maxwell3d_}, maxwell3d{maxwell3d_} {}
+namespace {
+class MacroInterpreterImpl final : public CachedMacro {
+public:
+    explicit MacroInterpreterImpl(Engines::Maxwell3D& maxwell3d_, const std::vector<u32>& code_)
+        : maxwell3d{maxwell3d_}, code{code_} {}
 
-std::unique_ptr<CachedMacro> MacroInterpreter::Compile(const std::vector<u32>& code) {
-    return std::make_unique<MacroInterpreterImpl>(maxwell3d, code);
-}
+    void Execute(const std::vector<u32>& params, u32 method) override;
 
-MacroInterpreterImpl::MacroInterpreterImpl(Engines::Maxwell3D& maxwell3d_,
-                                           const std::vector<u32>& code_)
-    : maxwell3d{maxwell3d_}, code{code_} {}
+private:
+    /// Resets the execution engine state, zeroing registers, etc.
+    void Reset();
+
+    /**
+     * Executes a single macro instruction located at the current program counter. Returns whether
+     * the interpreter should keep running.
+     *
+     * @param is_delay_slot Whether the current step is being executed due to a delay slot in a
+     *                      previous instruction.
+     */
+    bool Step(bool is_delay_slot);
+
+    /// Calculates the result of an ALU operation. src_a OP src_b;
+    u32 GetALUResult(Macro::ALUOperation operation, u32 src_a, u32 src_b);
+
+    /// Performs the result operation on the input result and stores it in the specified register
+    /// (if necessary).
+    void ProcessResult(Macro::ResultOperation operation, u32 reg, u32 result);
+
+    /// Evaluates the branch condition and returns whether the branch should be taken or not.
+    bool EvaluateBranchCondition(Macro::BranchCondition cond, u32 value) const;
+
+    /// Reads an opcode at the current program counter location.
+    Macro::Opcode GetOpcode() const;
+
+    /// Returns the specified register's value. Register 0 is hardcoded to always return 0.
+    u32 GetRegister(u32 register_id) const;
+
+    /// Sets the register to the input value.
+    void SetRegister(u32 register_id, u32 value);
+
+    /// Sets the method address to use for the next Send instruction.
+    void SetMethodAddress(u32 address);
+
+    /// Calls a GPU Engine method with the input parameter.
+    void Send(u32 value);
+
+    /// Reads a GPU register located at the method address.
+    u32 Read(u32 method) const;
+
+    /// Returns the next parameter in the parameter queue.
+    u32 FetchParameter();
+
+    Engines::Maxwell3D& maxwell3d;
+
+    /// Current program counter
+    u32 pc{};
+    /// Program counter to execute at after the delay slot is executed.
+    std::optional<u32> delayed_pc;
+
+    /// General purpose macro registers.
+    std::array<u32, Macro::NUM_MACRO_REGISTERS> registers = {};
+
+    /// Method address to use for the next Send instruction.
+    Macro::MethodAddress method_address = {};
+
+    /// Input parameters of the current macro.
+    std::unique_ptr<u32[]> parameters;
+    std::size_t num_parameters = 0;
+    std::size_t parameters_capacity = 0;
+    /// Index of the next parameter that will be fetched by the 'parm' instruction.
+    u32 next_parameter_index = 0;
+
+    bool carry_flag = false;
+    const std::vector<u32>& code;
+};
 
 void MacroInterpreterImpl::Execute(const std::vector<u32>& params, u32 method) {
     MICROPROFILE_SCOPE(MacroInterp);
@@ -283,5 +351,13 @@ u32 MacroInterpreterImpl::FetchParameter() {
     ASSERT(next_parameter_index < num_parameters);
     return parameters[next_parameter_index++];
 }
+} // Anonymous namespace
+
+MacroInterpreter::MacroInterpreter(Engines::Maxwell3D& maxwell3d_)
+    : MacroEngine{maxwell3d_}, maxwell3d{maxwell3d_} {}
+
+std::unique_ptr<CachedMacro> MacroInterpreter::Compile(const std::vector<u32>& code) {
+    return std::make_unique<MacroInterpreterImpl>(maxwell3d, code);
+}
 
 } // namespace Tegra

src/video_core/macro/macro_interpreter.h

@@ -3,10 +3,9 @@
 // Refer to the license.txt file included.
 
 #pragma once
-#include <array>
-#include <optional>
+
 #include <vector>
-#include "common/bit_field.h"
+
 #include "common/common_types.h"
 #include "video_core/macro/macro.h"
 
@@ -26,77 +25,4 @@ private:
     Engines::Maxwell3D& maxwell3d;
 };
 
-class MacroInterpreterImpl : public CachedMacro {
-public:
-    explicit MacroInterpreterImpl(Engines::Maxwell3D& maxwell3d_, const std::vector<u32>& code_);
-    void Execute(const std::vector<u32>& params, u32 method) override;
-
-private:
-    /// Resets the execution engine state, zeroing registers, etc.
-    void Reset();
-
-    /**
-     * Executes a single macro instruction located at the current program counter. Returns whether
-     * the interpreter should keep running.
-     *
-     * @param is_delay_slot Whether the current step is being executed due to a delay slot in a
-     *                      previous instruction.
-     */
-    bool Step(bool is_delay_slot);
-
-    /// Calculates the result of an ALU operation. src_a OP src_b;
-    u32 GetALUResult(Macro::ALUOperation operation, u32 src_a, u32 src_b);
-
-    /// Performs the result operation on the input result and stores it in the specified register
-    /// (if necessary).
-    void ProcessResult(Macro::ResultOperation operation, u32 reg, u32 result);
-
-    /// Evaluates the branch condition and returns whether the branch should be taken or not.
-    bool EvaluateBranchCondition(Macro::BranchCondition cond, u32 value) const;
-
-    /// Reads an opcode at the current program counter location.
-    Macro::Opcode GetOpcode() const;
-
-    /// Returns the specified register's value. Register 0 is hardcoded to always return 0.
-    u32 GetRegister(u32 register_id) const;
-
-    /// Sets the register to the input value.
-    void SetRegister(u32 register_id, u32 value);
-
-    /// Sets the method address to use for the next Send instruction.
-    void SetMethodAddress(u32 address);
-
-    /// Calls a GPU Engine method with the input parameter.
-    void Send(u32 value);
-
-    /// Reads a GPU register located at the method address.
-    u32 Read(u32 method) const;
-
-    /// Returns the next parameter in the parameter queue.
-    u32 FetchParameter();
-
-    Engines::Maxwell3D& maxwell3d;
-
-    /// Current program counter
-    u32 pc;
-    /// Program counter to execute at after the delay slot is executed.
-    std::optional<u32> delayed_pc;
-
-    /// General purpose macro registers.
-    std::array<u32, Macro::NUM_MACRO_REGISTERS> registers = {};
-
-    /// Method address to use for the next Send instruction.
-    Macro::MethodAddress method_address = {};
-
-    /// Input parameters of the current macro.
-    std::unique_ptr<u32[]> parameters;
-    std::size_t num_parameters = 0;
-    std::size_t parameters_capacity = 0;
-    /// Index of the next parameter that will be fetched by the 'parm' instruction.
-    u32 next_parameter_index = 0;
-
-    bool carry_flag = false;
-    const std::vector<u32>& code;
-};
-
 } // namespace Tegra

src/video_core/macro/macro_jit_x64.cpp

@@ -2,9 +2,17 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <array>
+#include <bitset>
+#include <optional>
+
+#include <xbyak/xbyak.h>
+
 #include "common/assert.h"
+#include "common/bit_field.h"
 #include "common/logging/log.h"
 #include "common/microprofile.h"
+#include "common/x64/xbyak_abi.h"
 #include "common/x64/xbyak_util.h"
 #include "video_core/engines/maxwell_3d.h"
 #include "video_core/macro/macro_interpreter.h"
@@ -14,13 +22,14 @@ MICROPROFILE_DEFINE(MacroJitCompile, "GPU", "Compile macro JIT", MP_RGB(173, 255
 MICROPROFILE_DEFINE(MacroJitExecute, "GPU", "Execute macro JIT", MP_RGB(255, 255, 0));
 
 namespace Tegra {
+namespace {
 constexpr Xbyak::Reg64 STATE = Xbyak::util::rbx;
 constexpr Xbyak::Reg32 RESULT = Xbyak::util::ebp;
 constexpr Xbyak::Reg64 PARAMETERS = Xbyak::util::r12;
 constexpr Xbyak::Reg32 METHOD_ADDRESS = Xbyak::util::r14d;
 constexpr Xbyak::Reg64 BRANCH_HOLDER = Xbyak::util::r15;
 
-static const std::bitset<32> PERSISTENT_REGISTERS = Common::X64::BuildRegSet({
+constexpr std::bitset<32> PERSISTENT_REGISTERS = Common::X64::BuildRegSet({
     STATE,
     RESULT,
     PARAMETERS,
@@ -28,19 +37,75 @@ static const std::bitset<32> PERSISTENT_REGISTERS = Common::X64::BuildRegSet({
     BRANCH_HOLDER,
 });
 
-MacroJITx64::MacroJITx64(Engines::Maxwell3D& maxwell3d_)
-    : MacroEngine{maxwell3d_}, maxwell3d{maxwell3d_} {}
+// Arbitrarily chosen based on current booting games.
+constexpr size_t MAX_CODE_SIZE = 0x10000;
 
-std::unique_ptr<CachedMacro> MacroJITx64::Compile(const std::vector<u32>& code) {
-    return std::make_unique<MacroJITx64Impl>(maxwell3d, code);
+std::bitset<32> PersistentCallerSavedRegs() {
+    return PERSISTENT_REGISTERS & Common::X64::ABI_ALL_CALLER_SAVED;
 }
 
-MacroJITx64Impl::MacroJITx64Impl(Engines::Maxwell3D& maxwell3d_, const std::vector<u32>& code_)
-    : CodeGenerator{MAX_CODE_SIZE}, code{code_}, maxwell3d{maxwell3d_} {
-    Compile();
-}
+class MacroJITx64Impl final : public Xbyak::CodeGenerator, public CachedMacro {
+public:
+    explicit MacroJITx64Impl(Engines::Maxwell3D& maxwell3d_, const std::vector<u32>& code_)
+        : CodeGenerator{MAX_CODE_SIZE}, code{code_}, maxwell3d{maxwell3d_} {
+        Compile();
+    }
 
-MacroJITx64Impl::~MacroJITx64Impl() = default;
+    void Execute(const std::vector<u32>& parameters, u32 method) override;
+
+    void Compile_ALU(Macro::Opcode opcode);
+    void Compile_AddImmediate(Macro::Opcode opcode);
+    void Compile_ExtractInsert(Macro::Opcode opcode);
+    void Compile_ExtractShiftLeftImmediate(Macro::Opcode opcode);
+    void Compile_ExtractShiftLeftRegister(Macro::Opcode opcode);
+    void Compile_Read(Macro::Opcode opcode);
+    void Compile_Branch(Macro::Opcode opcode);
+
+private:
+    void Optimizer_ScanFlags();
+
+    void Compile();
+    bool Compile_NextInstruction();
+
+    Xbyak::Reg32 Compile_FetchParameter();
+    Xbyak::Reg32 Compile_GetRegister(u32 index, Xbyak::Reg32 dst);
+
+    void Compile_ProcessResult(Macro::ResultOperation operation, u32 reg);
+    void Compile_Send(Xbyak::Reg32 value);
+
+    Macro::Opcode GetOpCode() const;
+
+    struct JITState {
+        Engines::Maxwell3D* maxwell3d{};
+        std::array<u32, Macro::NUM_MACRO_REGISTERS> registers{};
+        u32 carry_flag{};
+    };
+    static_assert(offsetof(JITState, maxwell3d) == 0, "Maxwell3D is not at 0x0");
+    using ProgramType = void (*)(JITState*, const u32*);
+
+    struct OptimizerState {
+        bool can_skip_carry{};
+        bool has_delayed_pc{};
+        bool zero_reg_skip{};
+        bool skip_dummy_addimmediate{};
+        bool optimize_for_method_move{};
+        bool enable_asserts{};
+    };
+    OptimizerState optimizer{};
+
+    std::optional<Macro::Opcode> next_opcode{};
+    ProgramType program{nullptr};
+
+    std::array<Xbyak::Label, MAX_CODE_SIZE> labels;
+    std::array<Xbyak::Label, MAX_CODE_SIZE> delay_skip;
+    Xbyak::Label end_of_code{};
+
+    bool is_delay_slot{};
+    u32 pc{};
+
+    const std::vector<u32>& code;
+    Engines::Maxwell3D& maxwell3d;
+};
 
 void MacroJITx64Impl::Execute(const std::vector<u32>& parameters, u32 method) {
     MICROPROFILE_SCOPE(MacroJitExecute);
@@ -307,11 +372,11 @@ void MacroJITx64Impl::Compile_Read(Macro::Opcode opcode) {
     Compile_ProcessResult(opcode.result_operation, opcode.dst);
 }
 
-static void Send(Engines::Maxwell3D* maxwell3d, Macro::MethodAddress method_address, u32 value) {
+void Send(Engines::Maxwell3D* maxwell3d, Macro::MethodAddress method_address, u32 value) {
     maxwell3d->CallMethodFromMME(method_address.address, value);
 }
 
-void Tegra::MacroJITx64Impl::Compile_Send(Xbyak::Reg32 value) {
+void MacroJITx64Impl::Compile_Send(Xbyak::Reg32 value) {
     Common::X64::ABI_PushRegistersAndAdjustStack(*this, PersistentCallerSavedRegs(), 0);
     mov(Common::X64::ABI_PARAM1, qword[STATE]);
     mov(Common::X64::ABI_PARAM2, METHOD_ADDRESS);
@@ -338,7 +403,7 @@ void Tegra::MacroJITx64Impl::Compile_Send(Xbyak::Reg32 value) {
     L(dont_process);
 }
 
-void Tegra::MacroJITx64Impl::Compile_Branch(Macro::Opcode opcode) {
+void MacroJITx64Impl::Compile_Branch(Macro::Opcode opcode) {
     ASSERT_MSG(!is_delay_slot, "Executing a branch in a delay slot is not valid");
     const s32 jump_address =
         static_cast<s32>(pc) + static_cast<s32>(opcode.GetBranchTarget() / sizeof(s32));
@@ -392,7 +457,7 @@ void Tegra::MacroJITx64Impl::Compile_Branch(Macro::Opcode opcode) {
     L(end);
 }
 
-void Tegra::MacroJITx64Impl::Optimizer_ScanFlags() {
+void MacroJITx64Impl::Optimizer_ScanFlags() {
     optimizer.can_skip_carry = true;
     optimizer.has_delayed_pc = false;
     for (auto raw_op : code) {
@@ -534,7 +599,7 @@ bool MacroJITx64Impl::Compile_NextInstruction() {
     return true;
 }
 
-Xbyak::Reg32 Tegra::MacroJITx64Impl::Compile_FetchParameter() {
+Xbyak::Reg32 MacroJITx64Impl::Compile_FetchParameter() {
     mov(eax, dword[PARAMETERS]);
     add(PARAMETERS, sizeof(u32));
     return eax;
@@ -611,9 +676,12 @@ Macro::Opcode MacroJITx64Impl::GetOpCode() const {
     ASSERT(pc < code.size());
     return {code[pc]};
 }
+} // Anonymous namespace
 
-std::bitset<32> MacroJITx64Impl::PersistentCallerSavedRegs() const {
-    return PERSISTENT_REGISTERS & Common::X64::ABI_ALL_CALLER_SAVED;
+MacroJITx64::MacroJITx64(Engines::Maxwell3D& maxwell3d_)
+    : MacroEngine{maxwell3d_}, maxwell3d{maxwell3d_} {}
+
+std::unique_ptr<CachedMacro> MacroJITx64::Compile(const std::vector<u32>& code) {
+    return std::make_unique<MacroJITx64Impl>(maxwell3d, code);
 }
-
 } // namespace Tegra

src/video_core/macro/macro_jit_x64.h

@@ -4,12 +4,7 @@
 
 #pragma once
 
-#include <array>
-#include <bitset>
-#include <xbyak/xbyak.h>
-#include "common/bit_field.h"
 #include "common/common_types.h"
-#include "common/x64/xbyak_abi.h"
 #include "video_core/macro/macro.h"
 
 namespace Tegra {
@@ -18,9 +13,6 @@ namespace Engines {
 class Maxwell3D;
 }
 
-/// MAX_CODE_SIZE is arbitrarily chosen based on current booting games
-constexpr size_t MAX_CODE_SIZE = 0x10000;
-
 class MacroJITx64 final : public MacroEngine {
 public:
     explicit MacroJITx64(Engines::Maxwell3D& maxwell3d_);
@@ -32,67 +24,4 @@ private:
     Engines::Maxwell3D& maxwell3d;
 };
 
-class MacroJITx64Impl : public Xbyak::CodeGenerator, public CachedMacro {
-public:
-    explicit MacroJITx64Impl(Engines::Maxwell3D& maxwell3d_, const std::vector<u32>& code_);
-    ~MacroJITx64Impl();
-
-    void Execute(const std::vector<u32>& parameters, u32 method) override;
-
-    void Compile_ALU(Macro::Opcode opcode);
-    void Compile_AddImmediate(Macro::Opcode opcode);
-    void Compile_ExtractInsert(Macro::Opcode opcode);
-    void Compile_ExtractShiftLeftImmediate(Macro::Opcode opcode);
-    void Compile_ExtractShiftLeftRegister(Macro::Opcode opcode);
-    void Compile_Read(Macro::Opcode opcode);
-    void Compile_Branch(Macro::Opcode opcode);
-
-private:
-    void Optimizer_ScanFlags();
-
-    void Compile();
-    bool Compile_NextInstruction();
-
-    Xbyak::Reg32 Compile_FetchParameter();
-    Xbyak::Reg32 Compile_GetRegister(u32 index, Xbyak::Reg32 dst);
-
-    void Compile_ProcessResult(Macro::ResultOperation operation, u32 reg);
-    void Compile_Send(Xbyak::Reg32 value);
-
-    Macro::Opcode GetOpCode() const;
-    std::bitset<32> PersistentCallerSavedRegs() const;
-
-    struct JITState {
-        Engines::Maxwell3D* maxwell3d{};
-        std::array<u32, Macro::NUM_MACRO_REGISTERS> registers{};
-        u32 carry_flag{};
-    };
-    static_assert(offsetof(JITState, maxwell3d) == 0, "Maxwell3D is not at 0x0");
-    using ProgramType = void (*)(JITState*, const u32*);
-
-    struct OptimizerState {
-        bool can_skip_carry{};
-        bool has_delayed_pc{};
-        bool zero_reg_skip{};
-        bool skip_dummy_addimmediate{};
-        bool optimize_for_method_move{};
-        bool enable_asserts{};
-    };
-    OptimizerState optimizer{};
-
-    std::optional<Macro::Opcode> next_opcode{};
-    ProgramType program{nullptr};
-
-    std::array<Xbyak::Label, MAX_CODE_SIZE> labels;
-    std::array<Xbyak::Label, MAX_CODE_SIZE> delay_skip;
-    Xbyak::Label end_of_code{};
-
-    bool is_delay_slot{};
-    u32 pc{};
-    std::optional<u32> delayed_pc;
-
-    const std::vector<u32>& code;
-    Engines::Maxwell3D& maxwell3d;
-};
-
 } // namespace Tegra

src/video_core/renderer_vulkan/vk_fsr.cpp

@@ -214,7 +214,7 @@ VkImageView FSR::Draw(VKScheduler& scheduler, size_t image_index, VkImageView im
 
         {
             VkImageMemoryBarrier fsr_write_barrier = base_barrier;
-            fsr_write_barrier.image = *images[image_index],
+            fsr_write_barrier.image = *images[image_index];
             fsr_write_barrier.oldLayout = VK_IMAGE_LAYOUT_UNDEFINED;
 
             cmdbuf.PipelineBarrier(VK_PIPELINE_STAGE_ALL_COMMANDS_BIT,

src/video_core/vulkan_common/vulkan_device.cpp

@@ -1038,7 +1038,7 @@ std::vector<const char*> Device::LoadExtensions(bool requires_surface) {
     }
     if (has_ext_shader_atomic_int64) {
         VkPhysicalDeviceShaderAtomicInt64Features atomic_int64;
-        atomic_int64.sType = VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_EXTENDED_DYNAMIC_STATE_FEATURES_EXT;
+        atomic_int64.sType = VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SHADER_ATOMIC_INT64_FEATURES;
         atomic_int64.pNext = nullptr;
         features.pNext = &atomic_int64;
         physical.GetFeatures2KHR(features);

src/yuzu/configuration/config.cpp

@@ -65,18 +65,18 @@ const std::array<int, 2> Config::default_stick_mod = {
 // This must be in alphabetical order according to action name as it must have the same order as
 // UISetting::values.shortcuts, which is alphabetically ordered.
 // clang-format off
-const std::array<UISettings::Shortcut, 21> Config::default_hotkeys{{
+const std::array<UISettings::Shortcut, 20> Config::default_hotkeys{{
+    {QStringLiteral("Audio Mute/Unmute"),        QStringLiteral("Main Window"), {QStringLiteral("Ctrl+M"),  QStringLiteral("Home+Dpad_Right"), Qt::WindowShortcut}},
+    {QStringLiteral("Audio Volume Down"),        QStringLiteral("Main Window"), {QStringLiteral("-"),       QStringLiteral("Home+Dpad_Down"), Qt::ApplicationShortcut}},
+    {QStringLiteral("Audio Volume Up"),          QStringLiteral("Main Window"), {QStringLiteral("+"),       QStringLiteral("Home+Dpad_Up"), Qt::ApplicationShortcut}},
     {QStringLiteral("Capture Screenshot"),       QStringLiteral("Main Window"), {QStringLiteral("Ctrl+P"),  QStringLiteral("Screenshot"), Qt::WidgetWithChildrenShortcut}},
     {QStringLiteral("Change Docked Mode"),       QStringLiteral("Main Window"), {QStringLiteral("F10"),     QStringLiteral("Home+X"), Qt::ApplicationShortcut}},
     {QStringLiteral("Continue/Pause Emulation"), QStringLiteral("Main Window"), {QStringLiteral("F4"),      QStringLiteral("Home+Plus"), Qt::WindowShortcut}},
-    {QStringLiteral("Decrease Speed Limit"),     QStringLiteral("Main Window"), {QStringLiteral("-"),       QStringLiteral(""), Qt::ApplicationShortcut}},
     {QStringLiteral("Exit Fullscreen"),          QStringLiteral("Main Window"), {QStringLiteral("Esc"),     QStringLiteral(""), Qt::WindowShortcut}},
     {QStringLiteral("Exit yuzu"),                QStringLiteral("Main Window"), {QStringLiteral("Ctrl+Q"),  QStringLiteral("Home+Minus"), Qt::WindowShortcut}},
     {QStringLiteral("Fullscreen"),               QStringLiteral("Main Window"), {QStringLiteral("F11"),     QStringLiteral("Home+B"), Qt::WindowShortcut}},
-    {QStringLiteral("Increase Speed Limit"),     QStringLiteral("Main Window"), {QStringLiteral("+"),       QStringLiteral(""), Qt::ApplicationShortcut}},
     {QStringLiteral("Load Amiibo"),              QStringLiteral("Main Window"), {QStringLiteral("F2"),      QStringLiteral("Home+A"), Qt::WidgetWithChildrenShortcut}},
     {QStringLiteral("Load File"),                QStringLiteral("Main Window"), {QStringLiteral("Ctrl+O"),  QStringLiteral(""), Qt::WidgetWithChildrenShortcut}},
-    {QStringLiteral("Mute Audio"),               QStringLiteral("Main Window"), {QStringLiteral("Ctrl+M"),  QStringLiteral(""), Qt::WindowShortcut}},
     {QStringLiteral("Restart Emulation"),        QStringLiteral("Main Window"), {QStringLiteral("F6"),      QStringLiteral(""), Qt::WindowShortcut}},
     {QStringLiteral("Stop Emulation"),           QStringLiteral("Main Window"), {QStringLiteral("F5"),      QStringLiteral(""), Qt::WindowShortcut}},
     {QStringLiteral("TAS Start/Stop"),           QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F5"), QStringLiteral(""), Qt::ApplicationShortcut}},
@@ -85,7 +85,6 @@ const std::array<UISettings::Shortcut, 21> Config::default_hotkeys{{
     {QStringLiteral("Toggle Filter Bar"),        QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F"),  QStringLiteral(""), Qt::WindowShortcut}},
     {QStringLiteral("Toggle Framerate Limit"),   QStringLiteral("Main Window"), {QStringLiteral("Ctrl+U"),  QStringLiteral("Home+Y"), Qt::ApplicationShortcut}},
     {QStringLiteral("Toggle Mouse Panning"),     QStringLiteral("Main Window"), {QStringLiteral("Ctrl+F9"), QStringLiteral(""), Qt::ApplicationShortcut}},
-    {QStringLiteral("Toggle Speed Limit"),       QStringLiteral("Main Window"), {QStringLiteral("Ctrl+Z"),  QStringLiteral(""), Qt::ApplicationShortcut}},
     {QStringLiteral("Toggle Status Bar"),        QStringLiteral("Main Window"), {QStringLiteral("Ctrl+S"),  QStringLiteral(""), Qt::WindowShortcut}},
 }};
 // clang-format on
@@ -394,6 +393,8 @@ void Config::ReadControlValues() {
     ReadGlobalSetting(Settings::values.enable_accurate_vibrations);
     ReadGlobalSetting(Settings::values.motion_enabled);
 
+    ReadBasicSetting(Settings::values.controller_navigation);
+
     qt_config->endGroup();
 }
 
@@ -743,7 +744,9 @@ void Config::ReadUIValues() {
     qt_config->beginGroup(QStringLiteral("UI"));
 
     UISettings::values.theme =
-        ReadSetting(QStringLiteral("theme"), QString::fromUtf8(UISettings::themes[0].second))
+        ReadSetting(
+            QStringLiteral("theme"),
+            QString::fromUtf8(UISettings::themes[static_cast<size_t>(default_theme)].second))
             .toString();
     ReadBasicSetting(UISettings::values.enable_discord_presence);
     ReadBasicSetting(UISettings::values.select_user_on_boot);
@@ -1000,6 +1003,7 @@ void Config::SaveControlValues() {
     WriteBasicSetting(Settings::values.keyboard_enabled);
     WriteBasicSetting(Settings::values.emulate_analog_keyboard);
     WriteBasicSetting(Settings::values.mouse_panning_sensitivity);
+    WriteBasicSetting(Settings::values.controller_navigation);
 
     WriteBasicSetting(Settings::values.tas_enable);
     WriteBasicSetting(Settings::values.tas_loop);
@@ -1271,7 +1275,7 @@ void Config::SaveUIValues() {
     qt_config->beginGroup(QStringLiteral("UI"));
 
     WriteSetting(QStringLiteral("theme"), UISettings::values.theme,
-                 QString::fromUtf8(UISettings::themes[0].second));
+                 QString::fromUtf8(UISettings::themes[static_cast<size_t>(default_theme)].second));
     WriteBasicSetting(UISettings::values.enable_discord_presence);
     WriteBasicSetting(UISettings::values.select_user_on_boot);
 

src/yuzu/configuration/config.h

@@ -46,7 +46,15 @@ public:
         default_mouse_buttons;
     static const std::array<int, Settings::NativeKeyboard::NumKeyboardKeys> default_keyboard_keys;
     static const std::array<int, Settings::NativeKeyboard::NumKeyboardMods> default_keyboard_mods;
-    static const std::array<UISettings::Shortcut, 21> default_hotkeys;
+    static const std::array<UISettings::Shortcut, 20> default_hotkeys;
+
+    static constexpr UISettings::Theme default_theme{
+#ifdef _WIN32
+        UISettings::Theme::DarkColorful
+#else
+        UISettings::Theme::DefaultColorful
+#endif
+    };
 
 private:
     void Initialize(const std::string& config_name);

src/yuzu/configuration/configure_graphics.ui

@@ -429,7 +429,7 @@
              </item>
              <item>
               <property name="text">
-               <string>AMD FidelityFX™️ Super Resolution [Vulkan Only]</string>
+               <string>AMD FidelityFX™️ Super Resolution (Vulkan Only)</string>
               </property>
              </item>
             </widget>

src/yuzu/configuration/configure_input_advanced.cpp

@@ -131,6 +131,7 @@ void ConfigureInputAdvanced::ApplyConfiguration() {
     Settings::values.touchscreen.enabled = ui->touchscreen_enabled->isChecked();
     Settings::values.enable_raw_input = ui->enable_raw_input->isChecked();
     Settings::values.enable_udp_controller = ui->enable_udp_controller->isChecked();
+    Settings::values.controller_navigation = ui->controller_navigation->isChecked();
 }
 
 void ConfigureInputAdvanced::LoadConfiguration() {
@@ -162,6 +163,7 @@ void ConfigureInputAdvanced::LoadConfiguration() {
     ui->touchscreen_enabled->setChecked(Settings::values.touchscreen.enabled);
     ui->enable_raw_input->setChecked(Settings::values.enable_raw_input.GetValue());
     ui->enable_udp_controller->setChecked(Settings::values.enable_udp_controller.GetValue());
+    ui->controller_navigation->setChecked(Settings::values.controller_navigation.GetValue());
 
     UpdateUIEnabled();
 }

src/yuzu/configuration/configure_input_advanced.ui

@@ -2655,6 +2655,19 @@
                      </widget>
                    </item>
                    <item row="4" column="0">
+                     <widget class="QCheckBox" name="controller_navigation">
+                       <property name="minimumSize">
+                         <size>
+                           <width>0</width>
+                           <height>23</height>
+                         </size>
+                       </property>
+                       <property name="text">
+                         <string>Controller navigation</string>
+                       </property>
+                     </widget>
+                   </item>
+                   <item row="5" column="0">
                      <widget class="QCheckBox" name="mouse_panning">
                        <property name="minimumSize">
                          <size>
@@ -2667,7 +2680,7 @@
                        </property>
                      </widget>
                    </item>
-                   <item row="4" column="2">
+                   <item row="5" column="2">
                      <widget class="QSpinBox" name="mouse_panning_sensitivity">
                        <property name="toolTip">
                          <string>Mouse sensitivity</string>
@@ -2689,14 +2702,14 @@
                        </property>
                      </widget>
                    </item>
-                   <item row="5" column="0">
+                   <item row="6" column="0">
                      <widget class="QLabel" name="motion_touch">
                        <property name="text">
                          <string>Motion / Touch</string>
                        </property>
                      </widget>
                    </item>
-                   <item row="5" column="2">
+                   <item row="6" column="2">
                      <widget class="QPushButton" name="buttonMotionTouch">
                        <property name="text">
                          <string>Configure</string>

src/yuzu/configuration/configure_input_player.cpp

@@ -102,6 +102,16 @@ QString GetButtonName(Common::Input::ButtonNames button_name) {
         return QObject::tr("Share");
     case Common::Input::ButtonNames::Options:
         return QObject::tr("Options");
+    case Common::Input::ButtonNames::ButtonMouseWheel:
+        return QObject::tr("Wheel", "Indicates the mouse wheel");
+    case Common::Input::ButtonNames::ButtonBackward:
+        return QObject::tr("Backward");
+    case Common::Input::ButtonNames::ButtonForward:
+        return QObject::tr("Forward");
+    case Common::Input::ButtonNames::ButtonTask:
+        return QObject::tr("Task");
+    case Common::Input::ButtonNames::ButtonExtra:
+        return QObject::tr("Extra");
     default:
         return QObject::tr("[undefined]");
     }
@@ -137,7 +147,7 @@ QString ConfigureInputPlayer::ButtonToText(const Common::ParamPackage& param) {
     // Retrieve the names from Qt
     if (param.Get("engine", "") == "keyboard") {
         const QString button_str = GetKeyName(param.Get("code", 0));
-        return QObject::tr("%1%2").arg(toggle, button_str);
+        return QObject::tr("%1%2%3").arg(toggle, inverted, button_str);
     }
 
     if (common_button_name == Common::Input::ButtonNames::Invalid) {
@@ -331,7 +341,7 @@ ConfigureInputPlayer::ConfigureInputPlayer(QWidget* parent, std::size_t player_i
                         emulated_controller->SetButtonParam(button_id, {});
                         button_map[button_id]->setText(tr("[not set]"));
                     });
-                    if (param.Has("button") || param.Has("hat")) {
+                    if (param.Has("code") || param.Has("button") || param.Has("hat")) {
                         context_menu.addAction(tr("Toggle button"), [&] {
                             const bool toggle_value = !param.Get("toggle", false);
                             param.Set("toggle", toggle_value);
@@ -339,8 +349,8 @@ ConfigureInputPlayer::ConfigureInputPlayer(QWidget* parent, std::size_t player_i
                             emulated_controller->SetButtonParam(button_id, param);
                         });
                         context_menu.addAction(tr("Invert button"), [&] {
-                            const bool toggle_value = !param.Get("inverted", false);
-                            param.Set("inverted", toggle_value);
+                            const bool invert_value = !param.Get("inverted", false);
+                            param.Set("inverted", invert_value);
                             button_map[button_id]->setText(ButtonToText(param));
                             emulated_controller->SetButtonParam(button_id, param);
                         });
@@ -500,28 +510,37 @@ ConfigureInputPlayer::ConfigureInputPlayer(QWidget* parent, std::size_t player_i
 
         analog_map_modifier_button[analog_id]->setContextMenuPolicy(Qt::CustomContextMenu);
 
-        connect(analog_map_modifier_button[analog_id], &QPushButton::customContextMenuRequested,
-                [=, this](const QPoint& menu_location) {
-                    QMenu context_menu;
-                    Common::ParamPackage param = emulated_controller->GetStickParam(analog_id);
-                    context_menu.addAction(tr("Clear"), [&] {
-                        param.Set("modifier", "");
-                        analog_map_modifier_button[analog_id]->setText(tr("[not set]"));
-                        emulated_controller->SetStickParam(analog_id, param);
-                    });
-                    context_menu.addAction(tr("Toggle button"), [&] {
-                        Common::ParamPackage modifier_param =
-                            Common::ParamPackage{param.Get("modifier", "")};
-                        const bool toggle_value = !modifier_param.Get("toggle", false);
-                        modifier_param.Set("toggle", toggle_value);
-                        param.Set("modifier", modifier_param.Serialize());
-                        analog_map_modifier_button[analog_id]->setText(
-                            ButtonToText(modifier_param));
-                        emulated_controller->SetStickParam(analog_id, param);
-                    });
-                    context_menu.exec(
-                        analog_map_modifier_button[analog_id]->mapToGlobal(menu_location));
+        connect(
+            analog_map_modifier_button[analog_id], &QPushButton::customContextMenuRequested,
+            [=, this](const QPoint& menu_location) {
+                QMenu context_menu;
+                Common::ParamPackage param = emulated_controller->GetStickParam(analog_id);
+                context_menu.addAction(tr("Clear"), [&] {
+                    param.Set("modifier", "");
+                    analog_map_modifier_button[analog_id]->setText(tr("[not set]"));
+                    emulated_controller->SetStickParam(analog_id, param);
                 });
+                context_menu.addAction(tr("Toggle button"), [&] {
+                    Common::ParamPackage modifier_param =
+                        Common::ParamPackage{param.Get("modifier", "")};
+                    const bool toggle_value = !modifier_param.Get("toggle", false);
+                    modifier_param.Set("toggle", toggle_value);
+                    param.Set("modifier", modifier_param.Serialize());
+                    analog_map_modifier_button[analog_id]->setText(ButtonToText(modifier_param));
+                    emulated_controller->SetStickParam(analog_id, param);
+                });
+                context_menu.addAction(tr("Invert button"), [&] {
+                    Common::ParamPackage modifier_param =
+                        Common::ParamPackage{param.Get("modifier", "")};
+                    const bool invert_value = !modifier_param.Get("inverted", false);
+                    modifier_param.Set("inverted", invert_value);
+                    param.Set("modifier", modifier_param.Serialize());
+                    analog_map_modifier_button[analog_id]->setText(ButtonToText(modifier_param));
+                    emulated_controller->SetStickParam(analog_id, param);
+                });
+                context_menu.exec(
+                    analog_map_modifier_button[analog_id]->mapToGlobal(menu_location));
+            });
 
         connect(analog_map_range_spinbox[analog_id], qOverload<int>(&QSpinBox::valueChanged),
                 [=, this] {

src/yuzu/debugger/wait_tree.cpp

@@ -95,7 +95,7 @@ std::vector<std::unique_ptr<WaitTreeThread>> WaitTreeItem::MakeThreadItemList(
     std::size_t row = 0;
     auto add_threads = [&](const std::vector<Kernel::KThread*>& threads) {
         for (std::size_t i = 0; i < threads.size(); ++i) {
-            if (threads[i]->GetThreadTypeForDebugging() == Kernel::ThreadType::User) {
+            if (threads[i]->GetThreadType() == Kernel::ThreadType::User) {
                 item_list.push_back(std::make_unique<WaitTreeThread>(*threads[i], system));
                 item_list.back()->row = row;
             }
@@ -153,7 +153,7 @@ QString WaitTreeCallstack::GetText() const {
 std::vector<std::unique_ptr<WaitTreeItem>> WaitTreeCallstack::GetChildren() const {
     std::vector<std::unique_ptr<WaitTreeItem>> list;
 
-    if (thread.GetThreadTypeForDebugging() != Kernel::ThreadType::User) {
+    if (thread.GetThreadType() != Kernel::ThreadType::User) {
         return list;
     }
 

src/yuzu/hotkeys.cpp

@@ -190,6 +190,9 @@ void ControllerShortcut::ControllerUpdateEvent(Core::HID::ControllerTriggerType
     if (type != Core::HID::ControllerTriggerType::Button) {
         return;
     }
+    if (!Settings::values.controller_navigation) {
+        return;
+    }
     if (button_sequence.npad.raw == Core::HID::NpadButton::None &&
         button_sequence.capture.raw == 0 && button_sequence.home.raw == 0) {
         return;

src/yuzu/main.cpp

@@ -1008,33 +1008,24 @@ void GMainWindow::InitializeHotkeys() {
             ToggleFullscreen();
         }
     });
-    connect_shortcut(QStringLiteral("Toggle Speed Limit"), [&] {
-        Settings::values.use_speed_limit.SetValue(!Settings::values.use_speed_limit.GetValue());
-        UpdateStatusBar();
-    });
-    constexpr u16 SPEED_LIMIT_STEP = 5;
-    connect_shortcut(QStringLiteral("Increase Speed Limit"), [&] {
-        if (Settings::values.speed_limit.GetValue() < 9999 - SPEED_LIMIT_STEP) {
-            Settings::values.speed_limit.SetValue(SPEED_LIMIT_STEP +
-                                                  Settings::values.speed_limit.GetValue());
-            UpdateStatusBar();
-        }
-    });
-    connect_shortcut(QStringLiteral("Decrease Speed Limit"), [&] {
-        if (Settings::values.speed_limit.GetValue() > SPEED_LIMIT_STEP) {
-            Settings::values.speed_limit.SetValue(Settings::values.speed_limit.GetValue() -
-                                                  SPEED_LIMIT_STEP);
-            UpdateStatusBar();
-        }
-    });
     connect_shortcut(QStringLiteral("Change Docked Mode"), [&] {
         Settings::values.use_docked_mode.SetValue(!Settings::values.use_docked_mode.GetValue());
         OnDockedModeChanged(!Settings::values.use_docked_mode.GetValue(),
                             Settings::values.use_docked_mode.GetValue(), *system);
         dock_status_button->setChecked(Settings::values.use_docked_mode.GetValue());
     });
-    connect_shortcut(QStringLiteral("Mute Audio"),
+    connect_shortcut(QStringLiteral("Audio Mute/Unmute"),
                      [] { Settings::values.audio_muted = !Settings::values.audio_muted; });
+    connect_shortcut(QStringLiteral("Audio Volume Down"), [] {
+        const auto current_volume = static_cast<int>(Settings::values.volume.GetValue());
+        const auto new_volume = std::max(current_volume - 5, 0);
+        Settings::values.volume.SetValue(static_cast<u8>(new_volume));
+    });
+    connect_shortcut(QStringLiteral("Audio Volume Up"), [] {
+        const auto current_volume = static_cast<int>(Settings::values.volume.GetValue());
+        const auto new_volume = std::min(current_volume + 5, 100);
+        Settings::values.volume.SetValue(static_cast<u8>(new_volume));
+    });
     connect_shortcut(QStringLiteral("Toggle Framerate Limit"), [] {
         Settings::values.disable_fps_limit.SetValue(!Settings::values.disable_fps_limit.GetValue());
     });

src/yuzu/uisettings.h

@@ -29,6 +29,15 @@ struct Shortcut {
     ContextualShortcut shortcut;
 };
 
+enum class Theme {
+    Default,
+    DefaultColorful,
+    Dark,
+    DarkColorful,
+    MidnightBlue,
+    MidnightBlueColorful,
+};
+
 using Themes = std::array<std::pair<const char*, const char*>, 6>;
 extern const Themes themes;
 

src/yuzu/util/controller_navigation.cpp

@@ -40,6 +40,9 @@ void ControllerNavigation::TriggerButton(Settings::NativeButton::Values native_b
 
 void ControllerNavigation::ControllerUpdateEvent(Core::HID::ControllerTriggerType type) {
     std::lock_guard lock{mutex};
+    if (!Settings::values.controller_navigation) {
+        return;
+    }
     if (type == Core::HID::ControllerTriggerType::Button) {
         ControllerUpdateButton();
         return;