service/prepo: Add missing header guard

Touchscreen Support

src/common/logging/backend.cpp

@@ -5,6 +5,7 @@
 #include <algorithm>
 #include <array>
 #include <chrono>
+#include <climits>
 #include <condition_variable>
 #include <memory>
 #include <thread>
@@ -83,8 +84,10 @@ private:
                 }
             };
             while (true) {
-                std::unique_lock<std::mutex> lock(message_mutex);
-                message_cv.wait(lock, [&] { return !running || message_queue.Pop(entry); });
+                {
+                    std::unique_lock<std::mutex> lock(message_mutex);
+                    message_cv.wait(lock, [&] { return !running || message_queue.Pop(entry); });
+                }
                 if (!running) {
                     break;
                 }
@@ -92,7 +95,7 @@ private:
             }
             // Drain the logging queue. Only writes out up to MAX_LOGS_TO_WRITE to prevent a case
             // where a system is repeatedly spamming logs even on close.
-            constexpr int MAX_LOGS_TO_WRITE = 100;
+            const int MAX_LOGS_TO_WRITE = filter.IsDebug() ? INT_MAX : 100;
             int logs_written = 0;
             while (logs_written++ < MAX_LOGS_TO_WRITE && message_queue.Pop(entry)) {
                 write_logs(entry);
@@ -282,4 +285,4 @@ void FmtLogMessageImpl(Class log_class, Level log_level, const char* filename,
 
     Impl::Instance().PushEntry(std::move(entry));
 }
-} // namespace Log
+} // namespace Log

src/common/logging/filter.cpp

@@ -94,4 +94,11 @@ bool Filter::ParseFilterRule(const std::string::const_iterator begin,
 bool Filter::CheckMessage(Class log_class, Level level) const {
     return static_cast<u8>(level) >= static_cast<u8>(class_levels[static_cast<size_t>(log_class)]);
 }
+
+bool Filter::IsDebug() const {
+    return std::any_of(class_levels.begin(), class_levels.end(), [](const Level& l) {
+        return static_cast<u8>(l) <= static_cast<u8>(Level::Debug);
+    });
+}
+
 } // namespace Log

src/common/logging/filter.h

@@ -47,6 +47,9 @@ public:
     /// Matches class/level combination against the filter, returning true if it passed.
     bool CheckMessage(Class log_class, Level level) const;
 
+    /// Returns true if any logging classes are set to debug
+    bool IsDebug() const;
+
 private:
     std::array<Level, (size_t)Class::Count> class_levels;
 };

src/common/telemetry.h

@@ -52,27 +52,14 @@ public:
 template <typename T>
 class Field : public FieldInterface {
 public:
-    Field(FieldType type, std::string name, const T& value)
-        : name(std::move(name)), type(type), value(value) {}
-
-    Field(FieldType type, std::string name, T&& value)
+    Field(FieldType type, std::string name, T value)
         : name(std::move(name)), type(type), value(std::move(value)) {}
 
-    Field(const Field& other) : Field(other.type, other.name, other.value) {}
+    Field(const Field&) = default;
+    Field& operator=(const Field&) = default;
 
-    Field& operator=(const Field& other) {
-        type = other.type;
-        name = other.name;
-        value = other.value;
-        return *this;
-    }
-
-    Field& operator=(Field&& other) {
-        type = other.type;
-        name = std::move(other.name);
-        value = std::move(other.value);
-        return *this;
-    }
+    Field(Field&&) = default;
+    Field& operator=(Field&& other) = default;
 
     void Accept(VisitorInterface& visitor) const override;
 
@@ -94,11 +81,11 @@ public:
         return value;
     }
 
-    inline bool operator==(const Field<T>& other) {
+    bool operator==(const Field& other) const {
         return (type == other.type) && (name == other.name) && (value == other.value);
     }
 
-    inline bool operator!=(const Field<T>& other) {
+    bool operator!=(const Field& other) const {
         return !(*this == other);
     }
 

src/core/arm/arm_interface.h

@@ -116,6 +116,8 @@ public:
      */
     virtual void LoadContext(const ThreadContext& ctx) = 0;
 
+    virtual void ClearExclusiveState() = 0;
+
     /// Prepare core for thread reschedule (if needed to correctly handle state)
     virtual void PrepareReschedule() = 0;
 };

src/core/arm/dynarmic/arm_dynarmic.cpp

@@ -226,6 +226,10 @@ void ARM_Dynarmic::ClearInstructionCache() {
     jit->ClearCache();
 }
 
+void ARM_Dynarmic::ClearExclusiveState() {
+    jit->ClearExclusiveState();
+}
+
 void ARM_Dynarmic::PageTableChanged() {
     jit = MakeJit(cb);
     current_page_table = Memory::GetCurrentPageTable();

src/core/arm/dynarmic/arm_dynarmic.h

@@ -39,6 +39,7 @@ public:
     void LoadContext(const ThreadContext& ctx) override;
 
     void PrepareReschedule() override;
+    void ClearExclusiveState() override;
 
     void ClearInstructionCache() override;
     void PageTableChanged() override;

src/core/arm/unicorn/arm_unicorn.cpp

@@ -263,6 +263,8 @@ void ARM_Unicorn::PrepareReschedule() {
     CHECKED(uc_emu_stop(uc));
 }
 
+void ARM_Unicorn::ClearExclusiveState() {}
+
 void ARM_Unicorn::ClearInstructionCache() {}
 
 void ARM_Unicorn::RecordBreak(GDBStub::BreakpointAddress bkpt) {

src/core/arm/unicorn/arm_unicorn.h

@@ -31,6 +31,7 @@ public:
     void SaveContext(ThreadContext& ctx) override;
     void LoadContext(const ThreadContext& ctx) override;
     void PrepareReschedule() override;
+    void ClearExclusiveState() override;
     void ExecuteInstructions(int num_instructions);
     void Run() override;
     void Step() override;

src/core/hle/kernel/address_arbiter.cpp

@@ -115,7 +115,7 @@ ResultCode ModifyByWaitingCountAndSignalToAddressIfEqual(VAddr address, s32 valu
     s32 updated_value;
     if (waiting_threads.size() == 0) {
         updated_value = value - 1;
-    } else if (num_to_wake <= 0 || waiting_threads.size() <= num_to_wake) {
+    } else if (num_to_wake <= 0 || waiting_threads.size() <= static_cast<u32>(num_to_wake)) {
         updated_value = value + 1;
     } else {
         updated_value = value;
@@ -140,7 +140,9 @@ ResultCode WaitForAddressIfLessThan(VAddr address, s32 value, s64 timeout, bool
 
     s32 cur_value = static_cast<s32>(Memory::Read32(address));
     if (cur_value < value) {
-        Memory::Write32(address, static_cast<u32>(cur_value - 1));
+        if (should_decrement) {
+            Memory::Write32(address, static_cast<u32>(cur_value - 1));
+        }
     } else {
         return ERR_INVALID_STATE;
     }

src/core/hle/kernel/hle_ipc.cpp

@@ -29,7 +29,8 @@ void SessionRequestHandler::ClientDisconnected(SharedPtr<ServerSession> server_s
 
 SharedPtr<Event> HLERequestContext::SleepClientThread(SharedPtr<Thread> thread,
                                                       const std::string& reason, u64 timeout,
-                                                      WakeupCallback&& callback) {
+                                                      WakeupCallback&& callback,
+                                                      Kernel::SharedPtr<Kernel::Event> event) {
 
     // Put the client thread to sleep until the wait event is signaled or the timeout expires.
     thread->wakeup_callback =
@@ -41,7 +42,12 @@ SharedPtr<Event> HLERequestContext::SleepClientThread(SharedPtr<Thread> thread,
         return true;
     };
 
-    auto event = Kernel::Event::Create(Kernel::ResetType::OneShot, "HLE Pause Event: " + reason);
+    if (!event) {
+        // Create event if not provided
+        event = Kernel::Event::Create(Kernel::ResetType::OneShot, "HLE Pause Event: " + reason);
+    }
+
+    event->Clear();
     thread->status = THREADSTATUS_WAIT_HLE_EVENT;
     thread->wait_objects = {event};
     event->AddWaitingThread(thread);

src/core/hle/kernel/hle_ipc.h

@@ -118,10 +118,12 @@ public:
      * @param callback Callback to be invoked when the thread is resumed. This callback must write
      * the entire command response once again, regardless of the state of it before this function
      * was called.
+     * @param event Event to use to wake up the thread. If unspecified, an event will be created.
      * @returns Event that when signaled will resume the thread and call the callback function.
      */
     SharedPtr<Event> SleepClientThread(SharedPtr<Thread> thread, const std::string& reason,
-                                       u64 timeout, WakeupCallback&& callback);
+                                       u64 timeout, WakeupCallback&& callback,
+                                       Kernel::SharedPtr<Kernel::Event> event = nullptr);
 
     void ParseCommandBuffer(u32_le* src_cmdbuf, bool incoming);
 

src/core/hle/kernel/scheduler.cpp

@@ -85,6 +85,7 @@ void Scheduler::SwitchContext(Thread* new_thread) {
 
         cpu_core->LoadContext(new_thread->context);
         cpu_core->SetTlsAddress(new_thread->GetTLSAddress());
+        cpu_core->ClearExclusiveState();
     } else {
         current_thread = nullptr;
         // Note: We do not reset the current process and current page table when idling because

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

@@ -5,6 +5,7 @@
 #include <atomic>
 #include "common/logging/log.h"
 #include "core/core_timing.h"
+#include "core/frontend/emu_window.h"
 #include "core/frontend/input.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/kernel/client_port.h"
@@ -18,9 +19,9 @@ namespace Service::HID {
 
 // Updating period for each HID device.
 // TODO(shinyquagsire23): These need better values.
-constexpr u64 pad_update_ticks = CoreTiming::BASE_CLOCK_RATE / 10000;
-constexpr u64 accelerometer_update_ticks = CoreTiming::BASE_CLOCK_RATE / 10000;
-constexpr u64 gyroscope_update_ticks = CoreTiming::BASE_CLOCK_RATE / 10000;
+constexpr u64 pad_update_ticks = CoreTiming::BASE_CLOCK_RATE / 100;
+constexpr u64 accelerometer_update_ticks = CoreTiming::BASE_CLOCK_RATE / 100;
+constexpr u64 gyroscope_update_ticks = CoreTiming::BASE_CLOCK_RATE / 100;
 
 class IAppletResource final : public ServiceFramework<IAppletResource> {
 public:
@@ -63,7 +64,8 @@ private:
         std::transform(Settings::values.analogs.begin() + Settings::NativeAnalog::STICK_HID_BEGIN,
                        Settings::values.analogs.begin() + Settings::NativeAnalog::STICK_HID_END,
                        sticks.begin(), Input::CreateDevice<Input::AnalogDevice>);
-        // TODO(shinyquagsire23): gyro, touch, mouse, keyboard
+        touch_device = Input::CreateDevice<Input::TouchDevice>(Settings::values.touch_device);
+        // TODO(shinyquagsire23): gyro, mouse, keyboard
     }
 
     void UpdatePadCallback(u64 userdata, int cycles_late) {
@@ -151,8 +153,6 @@ private:
             }
         }
 
-        // TODO(bunnei): Properly implement the touch screen, the below will just write empty data
-
         TouchScreen& touchscreen = mem.touchscreen;
         const u64 last_entry = touchscreen.header.latest_entry;
         const u64 curr_entry = (last_entry + 1) % touchscreen.entries.size();
@@ -164,7 +164,26 @@ private:
         touchscreen.header.max_entry_index = touchscreen.entries.size();
         touchscreen.header.timestamp = timestamp;
         touchscreen.entries[curr_entry].header.timestamp = sample_counter;
-        touchscreen.entries[curr_entry].header.num_touches = 0;
+
+        TouchScreenEntryTouch touch_entry{};
+        auto [x, y, pressed] = touch_device->GetStatus();
+        touch_entry.timestamp = timestamp;
+        touch_entry.x = static_cast<u16>(x * Layout::ScreenUndocked::Width);
+        touch_entry.y = static_cast<u16>(y * Layout::ScreenUndocked::Height);
+        touch_entry.touch_index = 0;
+
+        // TODO(DarkLordZach): Maybe try to derive these from EmuWindow?
+        touch_entry.diameter_x = 15;
+        touch_entry.diameter_y = 15;
+        touch_entry.angle = 0;
+
+        // TODO(DarkLordZach): Implement multi-touch support
+        if (pressed) {
+            touchscreen.entries[curr_entry].header.num_touches = 1;
+            touchscreen.entries[curr_entry].touches[0] = touch_entry;
+        } else {
+            touchscreen.entries[curr_entry].header.num_touches = 0;
+        }
 
         // TODO(shinyquagsire23): Properly implement mouse
         Mouse& mouse = mem.mouse;
@@ -250,6 +269,7 @@ private:
     std::array<std::unique_ptr<Input::ButtonDevice>, Settings::NativeButton::NUM_BUTTONS_HID>
         buttons;
     std::array<std::unique_ptr<Input::AnalogDevice>, Settings::NativeAnalog::NUM_STICKS_HID> sticks;
+    std::unique_ptr<Input::TouchDevice> touch_device;
 };
 
 class IActiveVibrationDeviceList final : public ServiceFramework<IActiveVibrationDeviceList> {

src/core/hle/service/nvdrv/devices/nvdisp_disp0.cpp

@@ -18,7 +18,8 @@ u32 nvdisp_disp0::ioctl(Ioctl command, const std::vector<u8>& input, std::vector
 }
 
 void nvdisp_disp0::flip(u32 buffer_handle, u32 offset, u32 format, u32 width, u32 height,
-                        u32 stride, NVFlinger::BufferQueue::BufferTransformFlags transform) {
+                        u32 stride, NVFlinger::BufferQueue::BufferTransformFlags transform,
+                        const MathUtil::Rectangle<int>& crop_rect) {
     VAddr addr = nvmap_dev->GetObjectAddress(buffer_handle);
     LOG_WARNING(Service,
                 "Drawing from address {:X} offset {:08X} Width {} Height {} Stride {} Format {}",
@@ -26,7 +27,8 @@ void nvdisp_disp0::flip(u32 buffer_handle, u32 offset, u32 format, u32 width, u3
 
     using PixelFormat = Tegra::FramebufferConfig::PixelFormat;
     const Tegra::FramebufferConfig framebuffer{
-        addr, offset, width, height, stride, static_cast<PixelFormat>(format), transform};
+        addr,      offset,   width, height, stride, static_cast<PixelFormat>(format),
+        transform, crop_rect};
 
     Core::System::GetInstance().perf_stats.EndGameFrame();
 

src/core/hle/service/nvdrv/devices/nvdisp_disp0.h

@@ -7,6 +7,7 @@
 #include <memory>
 #include <vector>
 #include "common/common_types.h"
+#include "common/math_util.h"
 #include "core/hle/service/nvdrv/devices/nvdevice.h"
 #include "core/hle/service/nvflinger/buffer_queue.h"
 
@@ -23,7 +24,8 @@ public:
 
     /// Performs a screen flip, drawing the buffer pointed to by the handle.
     void flip(u32 buffer_handle, u32 offset, u32 format, u32 width, u32 height, u32 stride,
-              NVFlinger::BufferQueue::BufferTransformFlags transform);
+              NVFlinger::BufferQueue::BufferTransformFlags transform,
+              const MathUtil::Rectangle<int>& crop_rect);
 
 private:
     std::shared_ptr<nvmap> nvmap_dev;

src/core/hle/service/nvflinger/buffer_queue.cpp

@@ -13,8 +13,8 @@ namespace Service {
 namespace NVFlinger {
 
 BufferQueue::BufferQueue(u32 id, u64 layer_id) : id(id), layer_id(layer_id) {
-    native_handle = Kernel::Event::Create(Kernel::ResetType::OneShot, "BufferQueue NativeHandle");
-    native_handle->Signal();
+    buffer_wait_event =
+        Kernel::Event::Create(Kernel::ResetType::Sticky, "BufferQueue NativeHandle");
 }
 
 void BufferQueue::SetPreallocatedBuffer(u32 slot, IGBPBuffer& igbp_buffer) {
@@ -26,10 +26,7 @@ void BufferQueue::SetPreallocatedBuffer(u32 slot, IGBPBuffer& igbp_buffer) {
     LOG_WARNING(Service, "Adding graphics buffer {}", slot);
 
     queue.emplace_back(buffer);
-
-    if (buffer_wait_event) {
-        buffer_wait_event->Signal();
-    }
+    buffer_wait_event->Signal();
 }
 
 boost::optional<u32> BufferQueue::DequeueBuffer(u32 width, u32 height) {
@@ -48,8 +45,6 @@ boost::optional<u32> BufferQueue::DequeueBuffer(u32 width, u32 height) {
         return boost::none;
     }
 
-    buffer_wait_event = nullptr;
-
     itr->status = Buffer::Status::Dequeued;
     return itr->slot;
 }
@@ -62,13 +57,15 @@ const IGBPBuffer& BufferQueue::RequestBuffer(u32 slot) const {
     return itr->igbp_buffer;
 }
 
-void BufferQueue::QueueBuffer(u32 slot, BufferTransformFlags transform) {
+void BufferQueue::QueueBuffer(u32 slot, BufferTransformFlags transform,
+                              const MathUtil::Rectangle<int>& crop_rect) {
     auto itr = std::find_if(queue.begin(), queue.end(),
                             [&](const Buffer& buffer) { return buffer.slot == slot; });
     ASSERT(itr != queue.end());
     ASSERT(itr->status == Buffer::Status::Dequeued);
     itr->status = Buffer::Status::Queued;
     itr->transform = transform;
+    itr->crop_rect = crop_rect;
 }
 
 boost::optional<const BufferQueue::Buffer&> BufferQueue::AcquireBuffer() {
@@ -88,9 +85,7 @@ void BufferQueue::ReleaseBuffer(u32 slot) {
     ASSERT(itr->status == Buffer::Status::Acquired);
     itr->status = Buffer::Status::Free;
 
-    if (buffer_wait_event) {
-        buffer_wait_event->Signal();
-    }
+    buffer_wait_event->Signal();
 }
 
 u32 BufferQueue::Query(QueryType type) {
@@ -106,10 +101,5 @@ u32 BufferQueue::Query(QueryType type) {
     return 0;
 }
 
-void BufferQueue::SetBufferWaitEvent(Kernel::SharedPtr<Kernel::Event>&& wait_event) {
-    ASSERT_MSG(!buffer_wait_event, "buffer_wait_event only supports a single waiting thread!");
-    buffer_wait_event = std::move(wait_event);
-}
-
 } // namespace NVFlinger
 } // namespace Service

src/core/hle/service/nvflinger/buffer_queue.h

@@ -6,6 +6,7 @@
 
 #include <vector>
 #include <boost/optional.hpp>
+#include "common/math_util.h"
 #include "common/swap.h"
 #include "core/hle/kernel/event.h"
 
@@ -68,23 +69,24 @@ public:
         Status status = Status::Free;
         IGBPBuffer igbp_buffer;
         BufferTransformFlags transform;
+        MathUtil::Rectangle<int> crop_rect;
     };
 
     void SetPreallocatedBuffer(u32 slot, IGBPBuffer& buffer);
     boost::optional<u32> DequeueBuffer(u32 width, u32 height);
     const IGBPBuffer& RequestBuffer(u32 slot) const;
-    void QueueBuffer(u32 slot, BufferTransformFlags transform);
+    void QueueBuffer(u32 slot, BufferTransformFlags transform,
+                     const MathUtil::Rectangle<int>& crop_rect);
     boost::optional<const Buffer&> AcquireBuffer();
     void ReleaseBuffer(u32 slot);
     u32 Query(QueryType type);
-    void SetBufferWaitEvent(Kernel::SharedPtr<Kernel::Event>&& wait_event);
 
     u32 GetId() const {
         return id;
     }
 
-    Kernel::SharedPtr<Kernel::Event> GetNativeHandle() const {
-        return native_handle;
+    Kernel::SharedPtr<Kernel::Event> GetBufferWaitEvent() const {
+        return buffer_wait_event;
     }
 
 private:
@@ -92,9 +94,6 @@ private:
     u64 layer_id;
 
     std::vector<Buffer> queue;
-    Kernel::SharedPtr<Kernel::Event> native_handle;
-
-    /// Used to signal waiting thread when no buffers are available
     Kernel::SharedPtr<Kernel::Event> buffer_wait_event;
 };
 

src/core/hle/service/nvflinger/nvflinger.cpp

@@ -149,12 +149,10 @@ void NVFlinger::Compose() {
         ASSERT(nvdisp);
 
         nvdisp->flip(igbp_buffer.gpu_buffer_id, igbp_buffer.offset, igbp_buffer.format,
-                     igbp_buffer.width, igbp_buffer.height, igbp_buffer.stride, buffer->transform);
+                     igbp_buffer.width, igbp_buffer.height, igbp_buffer.stride, buffer->transform,
+                     buffer->crop_rect);
 
         buffer_queue->ReleaseBuffer(buffer->slot);
-
-        // TODO(Subv): Figure out when we should actually signal this event.
-        buffer_queue->GetNativeHandle()->Signal();
     }
 }
 

src/core/hle/service/prepo/prepo.h

@@ -2,6 +2,8 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#pragma once
+
 #include <memory>
 #include <string>
 #include "core/hle/kernel/event.h"

src/core/hle/service/vi/vi.cpp

@@ -7,6 +7,7 @@
 #include <memory>
 #include <boost/optional.hpp>
 #include "common/alignment.h"
+#include "common/math_util.h"
 #include "common/scope_exit.h"
 #include "core/core_timing.h"
 #include "core/hle/ipc_helpers.h"
@@ -27,8 +28,8 @@ struct DisplayInfo {
     char display_name[0x40]{"Default"};
     u64 unknown_1{1};
     u64 unknown_2{1};
-    u64 width{1920};
-    u64 height{1080};
+    u64 width{1280};
+    u64 height{720};
 };
 static_assert(sizeof(DisplayInfo) == 0x60, "DisplayInfo has wrong size");
 
@@ -327,8 +328,8 @@ public:
 
 protected:
     void SerializeData() override {
-        // TODO(Subv): Figure out what this value means, writing non-zero here will make libnx try
-        // to read an IGBPBuffer object from the parcel.
+        // TODO(Subv): Figure out what this value means, writing non-zero here will make libnx
+        // try to read an IGBPBuffer object from the parcel.
         Write<u32_le>(1);
         WriteObject(buffer);
         Write<u32_le>(0);
@@ -360,8 +361,8 @@ public:
         INSERT_PADDING_WORDS(3);
         u32_le timestamp;
         s32_le is_auto_timestamp;
-        s32_le crop_left;
         s32_le crop_top;
+        s32_le crop_left;
         s32_le crop_right;
         s32_le crop_bottom;
         s32_le scaling_mode;
@@ -370,6 +371,10 @@ public:
         INSERT_PADDING_WORDS(2);
         u32_le fence_is_valid;
         std::array<Fence, 2> fences;
+
+        MathUtil::Rectangle<int> GetCropRect() const {
+            return {crop_left, crop_top, crop_right, crop_bottom};
+        }
     };
     static_assert(sizeof(Data) == 80, "ParcelData has wrong size");
 
@@ -495,7 +500,7 @@ private:
                 ctx.WriteBuffer(response.Serialize());
             } else {
                 // Wait the current thread until a buffer becomes available
-                auto wait_event = ctx.SleepClientThread(
+                ctx.SleepClientThread(
                     Kernel::GetCurrentThread(), "IHOSBinderDriver::DequeueBuffer", -1,
                     [=](Kernel::SharedPtr<Kernel::Thread> thread, Kernel::HLERequestContext& ctx,
                         ThreadWakeupReason reason) {
@@ -506,8 +511,8 @@ private:
                         ctx.WriteBuffer(response.Serialize());
                         IPC::ResponseBuilder rb{ctx, 2};
                         rb.Push(RESULT_SUCCESS);
-                    });
-                buffer_queue->SetBufferWaitEvent(std::move(wait_event));
+                    },
+                    buffer_queue->GetBufferWaitEvent());
             }
         } else if (transaction == TransactionId::RequestBuffer) {
             IGBPRequestBufferRequestParcel request{ctx.ReadBuffer()};
@@ -519,7 +524,8 @@ private:
         } else if (transaction == TransactionId::QueueBuffer) {
             IGBPQueueBufferRequestParcel request{ctx.ReadBuffer()};
 
-            buffer_queue->QueueBuffer(request.data.slot, request.data.transform);
+            buffer_queue->QueueBuffer(request.data.slot, request.data.transform,
+                                      request.data.GetCropRect());
 
             IGBPQueueBufferResponseParcel response{1280, 720};
             ctx.WriteBuffer(response.Serialize());
@@ -532,7 +538,7 @@ private:
             IGBPQueryResponseParcel response{value};
             ctx.WriteBuffer(response.Serialize());
         } else if (transaction == TransactionId::CancelBuffer) {
-            LOG_WARNING(Service_VI, "(STUBBED) called, transaction=CancelBuffer");
+            LOG_CRITICAL(Service_VI, "(STUBBED) called, transaction=CancelBuffer");
         } else {
             ASSERT_MSG(false, "Unimplemented");
         }
@@ -565,7 +571,7 @@ private:
         LOG_WARNING(Service_VI, "(STUBBED) called id={}, unknown={:08X}", id, unknown);
         IPC::ResponseBuilder rb{ctx, 2, 1};
         rb.Push(RESULT_SUCCESS);
-        rb.PushCopyObjects(buffer_queue->GetNativeHandle());
+        rb.PushCopyObjects(buffer_queue->GetBufferWaitEvent());
     }
 
     std::shared_ptr<NVFlinger::NVFlinger> nv_flinger;

src/video_core/engines/maxwell_3d.h

@@ -520,7 +520,18 @@ public:
                     u32 enable[NumRenderTargets];
                 } blend;
 
-                INSERT_PADDING_WORDS(0xB);
+                struct {
+                    u32 enable;
+                    u32 front_op_fail;
+                    u32 front_op_zfail;
+                    u32 front_op_zpass;
+                    u32 front_func_func;
+                    u32 front_func_ref;
+                    u32 front_func_mask;
+                    u32 front_mask;
+                } stencil;
+
+                INSERT_PADDING_WORDS(0x3);
 
                 union {
                     BitField<4, 1, u32> triangle_rast_flip;
@@ -556,7 +567,17 @@ public:
                     }
                 } tic;
 
-                INSERT_PADDING_WORDS(0x21);
+                INSERT_PADDING_WORDS(0x5);
+
+                struct {
+                    u32 enable;
+                    u32 back_op_fail;
+                    u32 back_op_zfail;
+                    u32 back_op_zpass;
+                    u32 back_func_func;
+                } stencil_two_side;
+
+                INSERT_PADDING_WORDS(0x17);
 
                 union {
                     BitField<2, 1, u32> coord_origin;
@@ -851,10 +872,12 @@ ASSERT_REG_POSITION(depth_write_enabled, 0x4BA);
 ASSERT_REG_POSITION(d3d_cull_mode, 0x4C2);
 ASSERT_REG_POSITION(depth_test_func, 0x4C3);
 ASSERT_REG_POSITION(blend, 0x4CF);
+ASSERT_REG_POSITION(stencil, 0x4E0);
 ASSERT_REG_POSITION(screen_y_control, 0x4EB);
 ASSERT_REG_POSITION(vb_element_base, 0x50D);
 ASSERT_REG_POSITION(tsc, 0x557);
 ASSERT_REG_POSITION(tic, 0x55D);
+ASSERT_REG_POSITION(stencil_two_side, 0x565);
 ASSERT_REG_POSITION(point_coord_replace, 0x581);
 ASSERT_REG_POSITION(code_address, 0x582);
 ASSERT_REG_POSITION(draw, 0x585);

src/video_core/gpu.h

@@ -67,6 +67,7 @@ struct FramebufferConfig {
 
     using TransformFlags = Service::NVFlinger::BufferQueue::BufferTransformFlags;
     TransformFlags transform_flags;
+    MathUtil::Rectangle<int> crop_rect;
 };
 
 namespace Engines {

src/video_core/renderer_opengl/gl_rasterizer_cache.cpp

@@ -105,6 +105,7 @@ static constexpr std::array<FormatTuple, SurfaceParams::MaxPixelFormat> tex_form
     {GL_COMPRESSED_RGBA_BPTC_UNORM_ARB, GL_RGB, GL_UNSIGNED_INT_8_8_8_8, ComponentType::UNorm,
      true},                                                             // BC7U
     {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_4X4
+    {GL_RG8, GL_RG, GL_UNSIGNED_BYTE, ComponentType::UNorm, false},     // G8R8
 
     // DepthStencil formats
     {GL_DEPTH24_STENCIL8, GL_DEPTH_STENCIL, GL_UNSIGNED_INT_24_8, ComponentType::UNorm,
@@ -196,8 +197,9 @@ static constexpr std::array<void (*)(u32, u32, u32, u8*, Tegra::GPUVAddr),
         MortonCopy<true, PixelFormat::DXT1>,         MortonCopy<true, PixelFormat::DXT23>,
         MortonCopy<true, PixelFormat::DXT45>,        MortonCopy<true, PixelFormat::DXN1>,
         MortonCopy<true, PixelFormat::BC7U>,         MortonCopy<true, PixelFormat::ASTC_2D_4X4>,
-        MortonCopy<true, PixelFormat::Z24S8>,        MortonCopy<true, PixelFormat::S8Z24>,
-        MortonCopy<true, PixelFormat::Z32F>,         MortonCopy<true, PixelFormat::Z16>,
+        MortonCopy<true, PixelFormat::G8R8>,         MortonCopy<true, PixelFormat::Z24S8>,
+        MortonCopy<true, PixelFormat::S8Z24>,        MortonCopy<true, PixelFormat::Z32F>,
+        MortonCopy<true, PixelFormat::Z16>,
 };
 
 static constexpr std::array<void (*)(u32, u32, u32, u8*, Tegra::GPUVAddr),
@@ -217,7 +219,8 @@ static constexpr std::array<void (*)(u32, u32, u32, u8*, Tegra::GPUVAddr),
         nullptr,
         nullptr,
         nullptr,
-        MortonCopy<false, PixelFormat::ABGR8>,
+        nullptr,
+        MortonCopy<false, PixelFormat::G8R8>,
         MortonCopy<false, PixelFormat::Z24S8>,
         MortonCopy<false, PixelFormat::S8Z24>,
         MortonCopy<false, PixelFormat::Z32F>,
@@ -274,10 +277,10 @@ static void ConvertS8Z24ToZ24S8(std::vector<u8>& data, u32 width, u32 height) {
 
     S8Z24 input_pixel{};
     Z24S8 output_pixel{};
-
+    const auto bpp{CachedSurface::GetGLBytesPerPixel(PixelFormat::S8Z24)};
     for (size_t y = 0; y < height; ++y) {
         for (size_t x = 0; x < width; ++x) {
-            const size_t offset{4 * (y * width + x)};
+            const size_t offset{bpp * (y * width + x)};
             std::memcpy(&input_pixel, &data[offset], sizeof(S8Z24));
             output_pixel.s8.Assign(input_pixel.s8);
             output_pixel.z24.Assign(input_pixel.z24);
@@ -285,6 +288,19 @@ static void ConvertS8Z24ToZ24S8(std::vector<u8>& data, u32 width, u32 height) {
         }
     }
 }
+
+static void ConvertG8R8ToR8G8(std::vector<u8>& data, u32 width, u32 height) {
+    const auto bpp{CachedSurface::GetGLBytesPerPixel(PixelFormat::G8R8)};
+    for (size_t y = 0; y < height; ++y) {
+        for (size_t x = 0; x < width; ++x) {
+            const size_t offset{bpp * (y * width + x)};
+            const u8 temp{data[offset]};
+            data[offset] = data[offset + 1];
+            data[offset + 1] = temp;
+        }
+    }
+}
+
 /**
  * Helper function to perform software conversion (as needed) when loading a buffer from Switch
  * memory. This is for Maxwell pixel formats that cannot be represented as-is in OpenGL or with
@@ -305,6 +321,11 @@ static void ConvertFormatAsNeeded_LoadGLBuffer(std::vector<u8>& data, PixelForma
         // Convert the S8Z24 depth format to Z24S8, as OpenGL does not support S8Z24.
         ConvertS8Z24ToZ24S8(data, width, height);
         break;
+
+    case PixelFormat::G8R8:
+        // Convert the G8R8 color format to R8G8, as OpenGL does not support G8R8.
+        ConvertG8R8ToR8G8(data, width, height);
+        break;
     }
 }
 

src/video_core/renderer_opengl/gl_rasterizer_cache.h

@@ -37,14 +37,15 @@ struct SurfaceParams {
         DXN1 = 11, // This is also known as BC4
         BC7U = 12,
         ASTC_2D_4X4 = 13,
+        G8R8 = 14,
 
         MaxColorFormat,
 
         // DepthStencil formats
-        Z24S8 = 14,
-        S8Z24 = 15,
-        Z32F = 16,
-        Z16 = 17,
+        Z24S8 = 15,
+        S8Z24 = 16,
+        Z32F = 17,
+        Z16 = 18,
 
         MaxDepthStencilFormat,
 
@@ -96,6 +97,7 @@ struct SurfaceParams {
             4, // DXN1
             4, // BC7U
             4, // ASTC_2D_4X4
+            1, // G8R8
             1, // Z24S8
             1, // S8Z24
             1, // Z32F
@@ -125,6 +127,7 @@ struct SurfaceParams {
             64,  // DXN1
             128, // BC7U
             32,  // ASTC_2D_4X4
+            16,  // G8R8
             32,  // Z24S8
             32,  // S8Z24
             32,  // Z32F
@@ -186,6 +189,8 @@ struct SurfaceParams {
             return PixelFormat::A1B5G5R5;
         case Tegra::Texture::TextureFormat::R8:
             return PixelFormat::R8;
+        case Tegra::Texture::TextureFormat::G8R8:
+            return PixelFormat::G8R8;
         case Tegra::Texture::TextureFormat::R16_G16_B16_A16:
             return PixelFormat::RGBA16F;
         case Tegra::Texture::TextureFormat::BF10GF11RF11:
@@ -223,6 +228,8 @@ struct SurfaceParams {
             return Tegra::Texture::TextureFormat::A1B5G5R5;
         case PixelFormat::R8:
             return Tegra::Texture::TextureFormat::R8;
+        case PixelFormat::G8R8:
+            return Tegra::Texture::TextureFormat::G8R8;
         case PixelFormat::RGBA16F:
             return Tegra::Texture::TextureFormat::R16_G16_B16_A16;
         case PixelFormat::R11FG11FB10F:

src/video_core/renderer_opengl/renderer_opengl.cpp

@@ -154,6 +154,7 @@ void RendererOpenGL::LoadFBToScreenInfo(const Tegra::FramebufferConfig& framebuf
 
     // Framebuffer orientation handling
     framebuffer_transform_flags = framebuffer.transform_flags;
+    framebuffer_crop_rect = framebuffer.crop_rect;
 
     // Ensure no bad interactions with GL_UNPACK_ALIGNMENT, which by default
     // only allows rows to have a memory alignement of 4.
@@ -320,11 +321,24 @@ void RendererOpenGL::DrawScreenTriangles(const ScreenInfo& screen_info, float x,
         }
     }
 
+    ASSERT_MSG(framebuffer_crop_rect.top == 0, "Unimplemented");
+    ASSERT_MSG(framebuffer_crop_rect.left == 0, "Unimplemented");
+
+    // Scale the output by the crop width/height. This is commonly used with 1280x720 rendering
+    // (e.g. handheld mode) on a 1920x1080 framebuffer.
+    f32 scale_u = 1.f, scale_v = 1.f;
+    if (framebuffer_crop_rect.GetWidth() > 0) {
+        scale_u = static_cast<f32>(framebuffer_crop_rect.GetWidth()) / screen_info.texture.width;
+    }
+    if (framebuffer_crop_rect.GetHeight() > 0) {
+        scale_v = static_cast<f32>(framebuffer_crop_rect.GetHeight()) / screen_info.texture.height;
+    }
+
     std::array<ScreenRectVertex, 4> vertices = {{
-        ScreenRectVertex(x, y, texcoords.top, left),
-        ScreenRectVertex(x + w, y, texcoords.bottom, left),
-        ScreenRectVertex(x, y + h, texcoords.top, right),
-        ScreenRectVertex(x + w, y + h, texcoords.bottom, right),
+        ScreenRectVertex(x, y, texcoords.top * scale_u, left * scale_v),
+        ScreenRectVertex(x + w, y, texcoords.bottom * scale_u, left * scale_v),
+        ScreenRectVertex(x, y + h, texcoords.top * scale_u, right * scale_v),
+        ScreenRectVertex(x + w, y + h, texcoords.bottom * scale_u, right * scale_v),
     }};
 
     state.texture_units[0].texture_2d = screen_info.display_texture;

src/video_core/renderer_opengl/renderer_opengl.h

@@ -97,4 +97,5 @@ private:
 
     /// Used for transforming the framebuffer orientation
     Tegra::FramebufferConfig::TransformFlags framebuffer_transform_flags;
+    MathUtil::Rectangle<int> framebuffer_crop_rect;
 };

src/video_core/textures/astc.cpp

@@ -25,16 +25,15 @@
 
 class BitStream {
 public:
-    BitStream(unsigned char* ptr, int nBits = 0, int start_offset = 0)
-        : m_BitsWritten(0), m_BitsRead(0), m_NumBits(nBits), m_CurByte(ptr),
-          m_NextBit(start_offset % 8), done(false) {}
+    explicit BitStream(unsigned char* ptr, int nBits = 0, int start_offset = 0)
+        : m_NumBits(nBits), m_CurByte(ptr), m_NextBit(start_offset % 8) {}
+
+    ~BitStream() = default;
 
     int GetBitsWritten() const {
         return m_BitsWritten;
     }
 
-    ~BitStream() {}
-
     void WriteBitsR(unsigned int val, unsigned int nBits) {
         for (unsigned int i = 0; i < nBits; i++) {
             WriteBit((val >> (nBits - i - 1)) & 1);
@@ -95,33 +94,28 @@ private:
         done = done || ++m_BitsWritten >= m_NumBits;
     }
 
-    int m_BitsWritten;
+    int m_BitsWritten = 0;
     const int m_NumBits;
     unsigned char* m_CurByte;
-    int m_NextBit;
-    int m_BitsRead;
+    int m_NextBit = 0;
+    int m_BitsRead = 0;
 
-    bool done;
+    bool done = false;
 };
 
 template <typename IntType>
 class Bits {
-private:
-    const IntType& m_Bits;
-
-    // Don't copy
-    Bits() {}
-    Bits(const Bits&) {}
-    Bits& operator=(const Bits&) {}
-
 public:
-    explicit Bits(IntType& v) : m_Bits(v) {}
+    explicit Bits(const IntType& v) : m_Bits(v) {}
 
-    uint8_t operator[](uint32_t bitPos) {
+    Bits(const Bits&) = delete;
+    Bits& operator=(const Bits&) = delete;
+
+    uint8_t operator[](uint32_t bitPos) const {
         return static_cast<uint8_t>((m_Bits >> bitPos) & 1);
     }
 
-    IntType operator()(uint32_t start, uint32_t end) {
+    IntType operator()(uint32_t start, uint32_t end) const {
         if (start == end) {
             return (*this)[start];
         } else if (start > end) {
@@ -133,6 +127,9 @@ public:
         uint64_t mask = (1 << (end - start + 1)) - 1;
         return (m_Bits >> start) & mask;
     }
+
+private:
+    const IntType& m_Bits;
 };
 
 enum EIntegerEncoding { eIntegerEncoding_JustBits, eIntegerEncoding_Quint, eIntegerEncoding_Trit };
@@ -186,12 +183,12 @@ public:
         m_QuintValue = val;
     }
 
-    bool MatchesEncoding(const IntegerEncodedValue& other) {
+    bool MatchesEncoding(const IntegerEncodedValue& other) const {
         return m_Encoding == other.m_Encoding && m_NumBits == other.m_NumBits;
     }
 
     // Returns the number of bits required to encode nVals values.
-    uint32_t GetBitLength(uint32_t nVals) {
+    uint32_t GetBitLength(uint32_t nVals) const {
         uint32_t totalBits = m_NumBits * nVals;
         if (m_Encoding == eIntegerEncoding_Trit) {
             totalBits += (nVals * 8 + 4) / 5;
@@ -382,19 +379,15 @@ private:
 namespace ASTCC {
 
 struct TexelWeightParams {
-    uint32_t m_Width;
-    uint32_t m_Height;
-    bool m_bDualPlane;
-    uint32_t m_MaxWeight;
-    bool m_bError;
-    bool m_bVoidExtentLDR;
-    bool m_bVoidExtentHDR;
+    uint32_t m_Width = 0;
+    uint32_t m_Height = 0;
+    bool m_bDualPlane = false;
+    uint32_t m_MaxWeight = 0;
+    bool m_bError = false;
+    bool m_bVoidExtentLDR = false;
+    bool m_bVoidExtentHDR = false;
 
-    TexelWeightParams() {
-        memset(this, 0, sizeof(*this));
-    }
-
-    uint32_t GetPackedBitSize() {
+    uint32_t GetPackedBitSize() const {
         // How many indices do we have?
         uint32_t nIdxs = m_Height * m_Width;
         if (m_bDualPlane) {
@@ -413,7 +406,7 @@ struct TexelWeightParams {
     }
 };
 
-TexelWeightParams DecodeBlockInfo(BitStream& strm) {
+static TexelWeightParams DecodeBlockInfo(BitStream& strm) {
     TexelWeightParams params;
 
     // Read the entire block mode all at once
@@ -612,8 +605,8 @@ TexelWeightParams DecodeBlockInfo(BitStream& strm) {
     return params;
 }
 
-void FillVoidExtentLDR(BitStream& strm, uint32_t* const outBuf, uint32_t blockWidth,
-                       uint32_t blockHeight) {
+static void FillVoidExtentLDR(BitStream& strm, uint32_t* const outBuf, uint32_t blockWidth,
+                              uint32_t blockHeight) {
     // Don't actually care about the void extent, just read the bits...
     for (int i = 0; i < 4; ++i) {
         strm.ReadBits(13);
@@ -628,23 +621,25 @@ void FillVoidExtentLDR(BitStream& strm, uint32_t* const outBuf, uint32_t blockWi
     uint32_t rgba = (r >> 8) | (g & 0xFF00) | (static_cast<uint32_t>(b) & 0xFF00) << 8 |
                     (static_cast<uint32_t>(a) & 0xFF00) << 16;
 
-    for (uint32_t j = 0; j < blockHeight; j++)
+    for (uint32_t j = 0; j < blockHeight; j++) {
         for (uint32_t i = 0; i < blockWidth; i++) {
             outBuf[j * blockWidth + i] = rgba;
         }
+    }
 }
 
-void FillError(uint32_t* outBuf, uint32_t blockWidth, uint32_t blockHeight) {
-    for (uint32_t j = 0; j < blockHeight; j++)
+static void FillError(uint32_t* outBuf, uint32_t blockWidth, uint32_t blockHeight) {
+    for (uint32_t j = 0; j < blockHeight; j++) {
         for (uint32_t i = 0; i < blockWidth; i++) {
             outBuf[j * blockWidth + i] = 0xFFFF00FF;
         }
+    }
 }
 
 // Replicates low numBits such that [(toBit - 1):(toBit - 1 - fromBit)]
 // is the same as [(numBits - 1):0] and repeats all the way down.
 template <typename IntType>
-IntType Replicate(const IntType& val, uint32_t numBits, uint32_t toBit) {
+static IntType Replicate(const IntType& val, uint32_t numBits, uint32_t toBit) {
     if (numBits == 0)
         return 0;
     if (toBit == 0)
@@ -668,27 +663,15 @@ IntType Replicate(const IntType& val, uint32_t numBits, uint32_t toBit) {
 
 class Pixel {
 protected:
-    typedef int16_t ChannelType;
-    uint8_t m_BitDepth[4];
-    int16_t color[4];
+    using ChannelType = int16_t;
+    uint8_t m_BitDepth[4] = {8, 8, 8, 8};
+    int16_t color[4] = {};
 
 public:
-    Pixel() {
-        for (int i = 0; i < 4; i++) {
-            m_BitDepth[i] = 8;
-            color[i] = 0;
-        }
-    }
-
-    Pixel(ChannelType a, ChannelType r, ChannelType g, ChannelType b, unsigned bitDepth = 8) {
-        for (int i = 0; i < 4; i++)
-            m_BitDepth[i] = bitDepth;
-
-        color[0] = a;
-        color[1] = r;
-        color[2] = g;
-        color[3] = b;
-    }
+    Pixel() = default;
+    Pixel(ChannelType a, ChannelType r, ChannelType g, ChannelType b, unsigned bitDepth = 8)
+        : m_BitDepth{uint8_t(bitDepth), uint8_t(bitDepth), uint8_t(bitDepth), uint8_t(bitDepth)},
+          color{a, r, g, b} {}
 
     // Changes the depth of each pixel. This scales the values to
     // the appropriate bit depth by either truncating the least
@@ -807,8 +790,8 @@ public:
     }
 };
 
-void DecodeColorValues(uint32_t* out, uint8_t* data, uint32_t* modes, const uint32_t nPartitions,
-                       const uint32_t nBitsForColorData) {
+static void DecodeColorValues(uint32_t* out, uint8_t* data, const uint32_t* modes,
+                              const uint32_t nPartitions, const uint32_t nBitsForColorData) {
     // First figure out how many color values we have
     uint32_t nValues = 0;
     for (uint32_t i = 0; i < nPartitions; i++) {
@@ -844,8 +827,7 @@ void DecodeColorValues(uint32_t* out, uint8_t* data, uint32_t* modes, const uint
     // Once we have the decoded values, we need to dequantize them to the 0-255 range
     // This procedure is outlined in ASTC spec C.2.13
     uint32_t outIdx = 0;
-    std::vector<IntegerEncodedValue>::const_iterator itr;
-    for (itr = decodedColorValues.begin(); itr != decodedColorValues.end(); itr++) {
+    for (auto itr = decodedColorValues.begin(); itr != decodedColorValues.end(); ++itr) {
         // Have we already decoded all that we need?
         if (outIdx >= nValues) {
             break;
@@ -978,7 +960,7 @@ void DecodeColorValues(uint32_t* out, uint8_t* data, uint32_t* modes, const uint
     }
 }
 
-uint32_t UnquantizeTexelWeight(const IntegerEncodedValue& val) {
+static uint32_t UnquantizeTexelWeight(const IntegerEncodedValue& val) {
     uint32_t bitval = val.GetBitValue();
     uint32_t bitlen = val.BaseBitLength();
 
@@ -1067,17 +1049,18 @@ uint32_t UnquantizeTexelWeight(const IntegerEncodedValue& val) {
     return result;
 }
 
-void UnquantizeTexelWeights(uint32_t out[2][144], std::vector<IntegerEncodedValue>& weights,
-                            const TexelWeightParams& params, const uint32_t blockWidth,
-                            const uint32_t blockHeight) {
+static void UnquantizeTexelWeights(uint32_t out[2][144],
+                                   const std::vector<IntegerEncodedValue>& weights,
+                                   const TexelWeightParams& params, const uint32_t blockWidth,
+                                   const uint32_t blockHeight) {
     uint32_t weightIdx = 0;
     uint32_t unquantized[2][144];
-    std::vector<IntegerEncodedValue>::const_iterator itr;
-    for (itr = weights.begin(); itr != weights.end(); itr++) {
+
+    for (auto itr = weights.begin(); itr != weights.end(); ++itr) {
         unquantized[0][weightIdx] = UnquantizeTexelWeight(*itr);
 
         if (params.m_bDualPlane) {
-            itr++;
+            ++itr;
             unquantized[1][weightIdx] = UnquantizeTexelWeight(*itr);
             if (itr == weights.end()) {
                 break;
@@ -1261,8 +1244,8 @@ static inline uint32_t Select2DPartition(int32_t seed, int32_t x, int32_t y, int
 }
 
 // Section C.2.14
-void ComputeEndpoints(Pixel& ep1, Pixel& ep2, const uint32_t*& colorValues,
-                      uint32_t colorEndpointMode) {
+static void ComputeEndpoints(Pixel& ep1, Pixel& ep2, const uint32_t*& colorValues,
+                             uint32_t colorEndpointMode) {
 #define READ_UINT_VALUES(N)                                                                        \
     uint32_t v[N];                                                                                 \
     for (uint32_t i = 0; i < N; i++) {                                                             \
@@ -1382,8 +1365,8 @@ void ComputeEndpoints(Pixel& ep1, Pixel& ep2, const uint32_t*& colorValues,
 #undef READ_INT_VALUES
 }
 
-void DecompressBlock(uint8_t inBuf[16], const uint32_t blockWidth, const uint32_t blockHeight,
-                     uint32_t* outBuf) {
+static void DecompressBlock(uint8_t inBuf[16], const uint32_t blockWidth,
+                            const uint32_t blockHeight, uint32_t* outBuf) {
     BitStream strm(inBuf);
     TexelWeightParams weightParams = DecodeBlockInfo(strm);
 
@@ -1617,8 +1600,7 @@ namespace Tegra::Texture::ASTC {
 std::vector<uint8_t> Decompress(std::vector<uint8_t>& data, uint32_t width, uint32_t height,
                                 uint32_t block_width, uint32_t block_height) {
     uint32_t blockIdx = 0;
-    std::vector<uint8_t> outData;
-    outData.resize(height * width * 4);
+    std::vector<uint8_t> outData(height * width * 4);
     for (uint32_t j = 0; j < height; j += block_height) {
         for (uint32_t i = 0; i < width; i += block_width) {
 

src/video_core/textures/decoders.cpp

@@ -62,6 +62,7 @@ u32 BytesPerPixel(TextureFormat format) {
         return 4;
     case TextureFormat::A1B5G5R5:
     case TextureFormat::B5G6R5:
+    case TextureFormat::G8R8:
         return 2;
     case TextureFormat::R8:
         return 1;
@@ -112,6 +113,7 @@ std::vector<u8> UnswizzleTexture(VAddr address, TextureFormat format, u32 width,
     case TextureFormat::A1B5G5R5:
     case TextureFormat::B5G6R5:
     case TextureFormat::R8:
+    case TextureFormat::G8R8:
     case TextureFormat::R16_G16_B16_A16:
     case TextureFormat::R32_G32_B32_A32:
     case TextureFormat::BF10GF11RF11:
@@ -167,6 +169,7 @@ std::vector<u8> DecodeTexture(const std::vector<u8>& texture_data, TextureFormat
     case TextureFormat::A1B5G5R5:
     case TextureFormat::B5G6R5:
     case TextureFormat::R8:
+    case TextureFormat::G8R8:
     case TextureFormat::BF10GF11RF11:
     case TextureFormat::R32_G32_B32_A32:
         // TODO(Subv): For the time being just forward the same data without any decoding.

src/yuzu/configuration/config.cpp

@@ -97,7 +97,7 @@ void Config::ReadValues() {
     qt_config->endGroup();
 
     qt_config->beginGroup("System");
-    Settings::values.use_docked_mode = qt_config->value("use_docked_mode", true).toBool();
+    Settings::values.use_docked_mode = qt_config->value("use_docked_mode", false).toBool();
     qt_config->endGroup();
 
     qt_config->beginGroup("Miscellaneous");

src/yuzu/game_list.cpp

@@ -17,10 +17,7 @@
 #include "game_list_p.h"
 #include "ui_settings.h"
 
-GameList::SearchField::KeyReleaseEater::KeyReleaseEater(GameList* gamelist) {
-    this->gamelist = gamelist;
-    edit_filter_text_old = "";
-}
+GameList::SearchField::KeyReleaseEater::KeyReleaseEater(GameList* gamelist) : gamelist{gamelist} {}
 
 // EventFilter in order to process systemkeys while editing the searchfield
 bool GameList::SearchField::KeyReleaseEater::eventFilter(QObject* obj, QEvent* event) {
@@ -141,10 +138,12 @@ GameList::SearchField::SearchField(GameList* parent) : QWidget{parent} {
  * @param userinput String containing all words getting checked
  * @return true if the haystack contains all words of userinput
  */
-bool GameList::containsAllWords(QString haystack, QString userinput) {
-    QStringList userinput_split = userinput.split(" ", QString::SplitBehavior::SkipEmptyParts);
+bool GameList::ContainsAllWords(const QString& haystack, const QString& userinput) const {
+    const QStringList userinput_split =
+        userinput.split(' ', QString::SplitBehavior::SkipEmptyParts);
+
     return std::all_of(userinput_split.begin(), userinput_split.end(),
-                       [haystack](QString s) { return haystack.contains(s); });
+                       [&haystack](const QString& s) { return haystack.contains(s); });
 }
 
 // Event in order to filter the gamelist after editing the searchfield
@@ -178,7 +177,7 @@ void GameList::onTextChanged(const QString& newText) {
             // The search is case insensitive because of toLower()
             // I decided not to use Qt::CaseInsensitive in containsAllWords to prevent
             // multiple conversions of edit_filter_text for each game in the gamelist
-            if (containsAllWords(file_name.append(" ").append(file_title), edit_filter_text) ||
+            if (ContainsAllWords(file_name.append(' ').append(file_title), edit_filter_text) ||
                 (file_programmid.count() == 16 && edit_filter_text.contains(file_programmid))) {
                 tree_view->setRowHidden(i, root_index, false);
                 ++result_count;

src/yuzu/game_list.h

@@ -89,7 +89,7 @@ private:
 
     void PopupContextMenu(const QPoint& menu_location);
     void RefreshGameDirectory();
-    bool containsAllWords(QString haystack, QString userinput);
+    bool ContainsAllWords(const QString& haystack, const QString& userinput) const;
 
     SearchField* search_field;
     GMainWindow* main_window = nullptr;

src/yuzu_cmd/config.cpp

@@ -110,7 +110,7 @@ void Config::ReadValues() {
         sdl2_config->GetBoolean("Data Storage", "use_virtual_sd", true);
 
     // System
-    Settings::values.use_docked_mode = sdl2_config->GetBoolean("System", "use_docked_mode", true);
+    Settings::values.use_docked_mode = sdl2_config->GetBoolean("System", "use_docked_mode", false);
 
     // Miscellaneous
     Settings::values.log_filter = sdl2_config->Get("Miscellaneous", "log_filter", "*:Trace");

src/yuzu_cmd/default_ini.h

@@ -163,7 +163,7 @@ use_virtual_sd =
 
 [System]
 # Whether the system is docked
-# 1 (default): Yes, 0: No
+# 1: Yes, 0 (default): No
 use_docked_mode =
 
 # The system region that yuzu will use during emulation