205 files changed, 7375 insertions, 3047 deletions

diff --git a/src/audio_core/audio_renderer.h b/src/audio_core/audio_renderer.h
index 046417da3..71ba4be40 100644
--- a/src/audio_core/audio_renderer.h
+++ b/src/audio_core/audio_renderer.h
@@ -143,7 +143,7 @@ struct AuxInfo {
     std::array<u8, 24> output_mix_buffers;
     u32_le mix_buffer_count;
     u32_le sample_rate; // Stored in the aux buffer currently
-    u32_le sampe_count;
+    u32_le sample_count;
     u64_le send_buffer_info;
     u64_le send_buffer_base;
 
diff --git a/src/audio_core/cubeb_sink.cpp b/src/audio_core/cubeb_sink.cpp
index 392039688..d31a1c844 100644
--- a/src/audio_core/cubeb_sink.cpp
+++ b/src/audio_core/cubeb_sink.cpp
@@ -121,7 +121,8 @@ CubebSink::CubebSink(std::string target_device_name) {
             const auto collection_end{collection.device + collection.count};
             const auto device{
                 std::find_if(collection.device, collection_end, [&](const cubeb_device_info& info) {
-                    return target_device_name == info.friendly_name;
+                    return info.friendly_name != nullptr &&
+                           target_device_name == info.friendly_name;
                 })};
             if (device != collection_end) {
                 output_device = device->devid;
diff --git a/src/audio_core/stream.cpp b/src/audio_core/stream.cpp
index 742a5e0a0..f35628e45 100644
--- a/src/audio_core/stream.cpp
+++ b/src/audio_core/stream.cpp
@@ -11,7 +11,6 @@
 #include "audio_core/stream.h"
 #include "common/assert.h"
 #include "common/logging/log.h"
-#include "common/microprofile.h"
 #include "core/core_timing.h"
 #include "core/core_timing_util.h"
 #include "core/settings.h"
@@ -104,10 +103,7 @@ void Stream::PlayNextBuffer() {
     CoreTiming::ScheduleEventThreadsafe(GetBufferReleaseCycles(*active_buffer), release_event, {});
 }
 
-MICROPROFILE_DEFINE(AudioOutput, "Audio", "ReleaseActiveBuffer", MP_RGB(100, 100, 255));
-
 void Stream::ReleaseActiveBuffer() {
-    MICROPROFILE_SCOPE(AudioOutput);
     ASSERT(active_buffer);
     released_buffers.push(std::move(active_buffer));
     release_callback();
diff --git a/src/audio_core/time_stretch.cpp b/src/audio_core/time_stretch.cpp
index d72d67994..2fe0b3aef 100644
--- a/src/audio_core/time_stretch.cpp
+++ b/src/audio_core/time_stretch.cpp
@@ -10,8 +10,7 @@
 
 namespace AudioCore {
 
-TimeStretcher::TimeStretcher(u32 sample_rate, u32 channel_count)
-    : m_sample_rate(sample_rate), m_channel_count(channel_count) {
+TimeStretcher::TimeStretcher(u32 sample_rate, u32 channel_count) : m_sample_rate{sample_rate} {
     m_sound_touch.setChannels(channel_count);
     m_sound_touch.setSampleRate(sample_rate);
     m_sound_touch.setPitch(1.0);
@@ -33,10 +32,10 @@ std::size_t TimeStretcher::Process(const s16* in, std::size_t num_in, s16* out,
     // We were given actual_samples number of samples, and num_samples were requested from us.
     double current_ratio = static_cast<double>(num_in) / static_cast<double>(num_out);
 
-    const double max_latency = 1.0; // seconds
+    const double max_latency = 0.25; // seconds
     const double max_backlog = m_sample_rate * max_latency;
     const double backlog_fullness = m_sound_touch.numSamples() / max_backlog;
-    if (backlog_fullness > 5.0) {
+    if (backlog_fullness > 4.0) {
         // Too many samples in backlog: Don't push anymore on
         num_in = 0;
     }
@@ -50,7 +49,7 @@ std::size_t TimeStretcher::Process(const s16* in, std::size_t num_in, s16* out,
 
     // This low-pass filter smoothes out variance in the calculated stretch ratio.
     // The time-scale determines how responsive this filter is.
-    constexpr double lpf_time_scale = 2.0; // seconds
+    constexpr double lpf_time_scale = 0.712; // seconds
     const double lpf_gain = 1.0 - std::exp(-time_delta / lpf_time_scale);
     m_stretch_ratio += lpf_gain * (current_ratio - m_stretch_ratio);
 
diff --git a/src/audio_core/time_stretch.h b/src/audio_core/time_stretch.h
index decd760f1..bb2270b96 100644
--- a/src/audio_core/time_stretch.h
+++ b/src/audio_core/time_stretch.h
@@ -27,7 +27,6 @@ public:
 
 private:
     u32 m_sample_rate;
-    u32 m_channel_count;
     soundtouch::SoundTouch m_sound_touch;
     double m_stretch_ratio = 1.0;
 };
diff --git a/src/common/bit_field.h b/src/common/bit_field.h
index bf803da8d..21e07925d 100644
--- a/src/common/bit_field.h
+++ b/src/common/bit_field.h
@@ -117,21 +117,21 @@ private:
     // We don't delete it because we want BitField to be trivially copyable.
     constexpr BitField& operator=(const BitField&) = default;
 
-    // StorageType is T for non-enum types and the underlying type of T if
+    // UnderlyingType is T for non-enum types and the underlying type of T if
     // T is an enumeration. Note that T is wrapped within an enable_if in the
     // former case to workaround compile errors which arise when using
     // std::underlying_type<T>::type directly.
-    using StorageType = typename std::conditional_t<std::is_enum<T>::value, std::underlying_type<T>,
-                                                    std::enable_if<true, T>>::type;
+    using UnderlyingType = typename std::conditional_t<std::is_enum_v<T>, std::underlying_type<T>,
+                                                       std::enable_if<true, T>>::type;
 
-    // Unsigned version of StorageType
-    using StorageTypeU = std::make_unsigned_t<StorageType>;
+    // We store the value as the unsigned type to avoid undefined behaviour on value shifting
+    using StorageType = std::make_unsigned_t<UnderlyingType>;
 
 public:
     /// Constants to allow limited introspection of fields if needed
     static constexpr std::size_t position = Position;
     static constexpr std::size_t bits = Bits;
-    static constexpr StorageType mask = (((StorageTypeU)~0) >> (8 * sizeof(T) - bits)) << position;
+    static constexpr StorageType mask = (((StorageType)~0) >> (8 * sizeof(T) - bits)) << position;
 
     /**
      * Formats a value by masking and shifting it according to the field parameters. A value
@@ -148,11 +148,12 @@ public:
      * union in a constexpr context.
      */
     static constexpr FORCE_INLINE T ExtractValue(const StorageType& storage) {
-        if (std::numeric_limits<T>::is_signed) {
+        if constexpr (std::numeric_limits<UnderlyingType>::is_signed) {
             std::size_t shift = 8 * sizeof(T) - bits;
-            return (T)((storage << (shift - position)) >> shift);
+            return static_cast<T>(static_cast<UnderlyingType>(storage << (shift - position)) >>
+                                  shift);
         } else {
-            return (T)((storage & mask) >> position);
+            return static_cast<T>((storage & mask) >> position);
         }
     }
 
diff --git a/src/common/logging/backend.cpp b/src/common/logging/backend.cpp
index 9f5918851..5753b871a 100644
--- a/src/common/logging/backend.cpp
+++ b/src/common/logging/backend.cpp
@@ -12,7 +12,8 @@
 #include <thread>
 #include <vector>
 #ifdef _WIN32
-#include <share.h> // For _SH_DENYWR
+#include <share.h>   // For _SH_DENYWR
+#include <windows.h> // For OutputDebugStringA
 #else
 #define _SH_DENYWR 0
 #endif
@@ -139,12 +140,18 @@ void FileBackend::Write(const Entry& entry) {
     if (!file.IsOpen() || bytes_written > MAX_BYTES_WRITTEN) {
         return;
     }
-    bytes_written += file.WriteString(FormatLogMessage(entry) + '\n');
+    bytes_written += file.WriteString(FormatLogMessage(entry).append(1, '\n'));
     if (entry.log_level >= Level::Error) {
         file.Flush();
     }
 }
 
+void DebuggerBackend::Write(const Entry& entry) {
+#ifdef _WIN32
+    ::OutputDebugStringA(FormatLogMessage(entry).append(1, '\n').c_str());
+#endif
+}
+
 /// Macro listing all log classes. Code should define CLS and SUB as desired before invoking this.
 #define ALL_LOG_CLASSES()                                                                          \
     CLS(Log)                                                                                       \
@@ -196,6 +203,7 @@ void FileBackend::Write(const Entry& entry) {
     SUB(Service, NFP)                                                                              \
     SUB(Service, NIFM)                                                                             \
     SUB(Service, NIM)                                                                              \
+    SUB(Service, NPNS)                                                                             \
     SUB(Service, NS)                                                                               \
     SUB(Service, NVDRV)                                                                            \
     SUB(Service, PCIE)                                                                             \
@@ -204,10 +212,12 @@ void FileBackend::Write(const Entry& entry) {
     SUB(Service, PM)                                                                               \
     SUB(Service, PREPO)                                                                            \
     SUB(Service, PSC)                                                                              \
+    SUB(Service, PSM)                                                                              \
     SUB(Service, SET)                                                                              \
     SUB(Service, SM)                                                                               \
     SUB(Service, SPL)                                                                              \
     SUB(Service, SSL)                                                                              \
+    SUB(Service, TCAP)                                                                             \
     SUB(Service, Time)                                                                             \
     SUB(Service, USB)                                                                              \
     SUB(Service, VI)                                                                               \
diff --git a/src/common/logging/backend.h b/src/common/logging/backend.h
index 11edbf1b6..91bb0c309 100644
--- a/src/common/logging/backend.h
+++ b/src/common/logging/backend.h
@@ -103,6 +103,20 @@ private:
     std::size_t bytes_written;
 };
 
+/**
+ * Backend that writes to Visual Studio's output window
+ */
+class DebuggerBackend : public Backend {
+public:
+    static const char* Name() {
+        return "debugger";
+    }
+    const char* GetName() const override {
+        return Name();
+    }
+    void Write(const Entry& entry) override;
+};
+
 void AddBackend(std::unique_ptr<Backend> backend);
 
 void RemoveBackend(std::string_view backend_name);
diff --git a/src/common/logging/log.h b/src/common/logging/log.h
index c9161155a..d4ec31ec3 100644
--- a/src/common/logging/log.h
+++ b/src/common/logging/log.h
@@ -83,6 +83,7 @@ enum class Class : ClassType {
     Service_NFP,       ///< The NFP service
     Service_NIFM,      ///< The NIFM (Network interface) service
     Service_NIM,       ///< The NIM service
+    Service_NPNS,      ///< The NPNS service
     Service_NS,        ///< The NS services
     Service_NVDRV,     ///< The NVDRV (Nvidia driver) service
     Service_PCIE,      ///< The PCIe service
@@ -96,6 +97,7 @@ enum class Class : ClassType {
     Service_SM,        ///< The SM (Service manager) service
     Service_SPL,       ///< The SPL service
     Service_SSL,       ///< The SSL service
+    Service_TCAP,      ///< The TCAP service.
     Service_Time,      ///< The time service
     Service_USB,       ///< The USB (Universal Serial Bus) service
     Service_VI,        ///< The VI (Video interface) service
diff --git a/src/common/string_util.cpp b/src/common/string_util.cpp
index 731d1db34..14f7037d8 100644
--- a/src/common/string_util.cpp
+++ b/src/common/string_util.cpp
@@ -4,11 +4,10 @@
 
 #include <algorithm>
 #include <cctype>
-#include <cerrno>
 #include <codecvt>
-#include <cstdio>
 #include <cstdlib>
-#include <cstring>
+#include <locale>
+#include <sstream>
 #include "common/common_paths.h"
 #include "common/logging/log.h"
 #include "common/string_util.h"
@@ -33,24 +32,6 @@ std::string ToUpper(std::string str) {
     return str;
 }
 
-// For Debugging. Read out an u8 array.
-std::string ArrayToString(const u8* data, std::size_t size, int line_len, bool spaces) {
-    std::ostringstream oss;
-    oss << std::setfill('0') << std::hex;
-
-    for (int line = 0; size; ++data, --size) {
-        oss << std::setw(2) << (int)*data;
-
-        if (line_len == ++line) {
-            oss << '\n';
-            line = 0;
-        } else if (spaces)
-            oss << ' ';
-    }
-
-    return oss.str();
-}
-
 std::string StringFromBuffer(const std::vector<u8>& data) {
     return std::string(data.begin(), std::find(data.begin(), data.end(), '\0'));
 }
@@ -75,40 +56,6 @@ std::string StripQuotes(const std::string& s) {
         return s;
 }
 
-bool TryParse(const std::string& str, u32* const output) {
-    char* endptr = nullptr;
-
-    // Reset errno to a value other than ERANGE
-    errno = 0;
-
-    unsigned long value = strtoul(str.c_str(), &endptr, 0);
-
-    if (!endptr || *endptr)
-        return false;
-
-    if (errno == ERANGE)
-        return false;
-
-#if ULONG_MAX > UINT_MAX
-    if (value >= 0x100000000ull && value <= 0xFFFFFFFF00000000ull)
-        return false;
-#endif
-
-    *output = static_cast<u32>(value);
-    return true;
-}
-
-bool TryParse(const std::string& str, bool* const output) {
-    if ("1" == str || "true" == ToLower(str))
-        *output = true;
-    else if ("0" == str || "false" == ToLower(str))
-        *output = false;
-    else
-        return false;
-
-    return true;
-}
-
 std::string StringFromBool(bool value) {
     return value ? "True" : "False";
 }
diff --git a/src/common/string_util.h b/src/common/string_util.h
index 32bf6a19c..08f96533b 100644
--- a/src/common/string_util.h
+++ b/src/common/string_util.h
@@ -5,8 +5,6 @@
 #pragma once
 
 #include <cstddef>
-#include <iomanip>
-#include <sstream>
 #include <string>
 #include <vector>
 #include "common/common_types.h"
@@ -19,44 +17,13 @@ std::string ToLower(std::string str);
 /// Make a string uppercase
 std::string ToUpper(std::string str);
 
-std::string ArrayToString(const u8* data, std::size_t size, int line_len = 20, bool spaces = true);
-
 std::string StringFromBuffer(const std::vector<u8>& data);
 
 std::string StripSpaces(const std::string& s);
 std::string StripQuotes(const std::string& s);
 
-// Thousand separator. Turns 12345678 into 12,345,678
-template <typename I>
-std::string ThousandSeparate(I value, int spaces = 0) {
-    std::ostringstream oss;
-
-// std::locale("") seems to be broken on many platforms
-#if defined _WIN32 || (defined __linux__ && !defined __clang__)
-    oss.imbue(std::locale(""));
-#endif
-    oss << std::setw(spaces) << value;
-
-    return oss.str();
-}
-
 std::string StringFromBool(bool value);
 
-bool TryParse(const std::string& str, bool* output);
-bool TryParse(const std::string& str, u32* output);
-
-template <typename N>
-static bool TryParse(const std::string& str, N* const output) {
-    std::istringstream iss(str);
-
-    N tmp = 0;
-    if (iss >> tmp) {
-        *output = tmp;
-        return true;
-    } else
-        return false;
-}
-
 std::string TabsToSpaces(int tab_size, std::string in);
 
 void SplitString(const std::string& str, char delim, std::vector<std::string>& output);
diff --git a/src/common/telemetry.h b/src/common/telemetry.h
index 8d6ab986b..854a73fae 100644
--- a/src/common/telemetry.h
+++ b/src/common/telemetry.h
@@ -153,6 +153,7 @@ struct VisitorInterface : NonCopyable {
 
     /// Completion method, called once all fields have been visited
     virtual void Complete() = 0;
+    virtual bool SubmitTestcase() = 0;
 };
 
 /**
@@ -178,6 +179,9 @@ struct NullVisitor : public VisitorInterface {
     void Visit(const Field<std::chrono::microseconds>& /*field*/) override {}
 
     void Complete() override {}
+    bool SubmitTestcase() override {
+        return false;
+    }
 };
 
 /// Appends build-specific information to the given FieldCollection,
diff --git a/src/core/core.cpp b/src/core/core.cpp
index 7cb86ed92..6d5b5a2d0 100644
--- a/src/core/core.cpp
+++ b/src/core/core.cpp
@@ -185,7 +185,7 @@ struct System::Impl {
             LOG_CRITICAL(Core, "Failed to obtain loader for {}!", filepath);
             return ResultStatus::ErrorGetLoader;
         }
-        std::pair<boost::optional<u32>, Loader::ResultStatus> system_mode =
+        std::pair<std::optional<u32>, Loader::ResultStatus> system_mode =
             app_loader->LoadKernelSystemMode();
 
         if (system_mode.second != Loader::ResultStatus::Success) {
@@ -312,6 +312,10 @@ Cpu& System::CurrentCpuCore() {
     return impl->CurrentCpuCore();
 }
 
+const Cpu& System::CurrentCpuCore() const {
+    return impl->CurrentCpuCore();
+}
+
 System::ResultStatus System::RunLoop(bool tight_loop) {
     return impl->RunLoop(tight_loop);
 }
@@ -342,7 +346,11 @@ PerfStatsResults System::GetAndResetPerfStats() {
     return impl->GetAndResetPerfStats();
 }
 
-Core::TelemetrySession& System::TelemetrySession() const {
+TelemetrySession& System::TelemetrySession() {
+    return *impl->telemetry_session;
+}
+
+const TelemetrySession& System::TelemetrySession() const {
     return *impl->telemetry_session;
 }
 
@@ -350,7 +358,11 @@ ARM_Interface& System::CurrentArmInterface() {
     return CurrentCpuCore().ArmInterface();
 }
 
-std::size_t System::CurrentCoreIndex() {
+const ARM_Interface& System::CurrentArmInterface() const {
+    return CurrentCpuCore().ArmInterface();
+}
+
+std::size_t System::CurrentCoreIndex() const {
     return CurrentCpuCore().CoreIndex();
 }
 
@@ -358,6 +370,10 @@ Kernel::Scheduler& System::CurrentScheduler() {
     return CurrentCpuCore().Scheduler();
 }
 
+const Kernel::Scheduler& System::CurrentScheduler() const {
+    return CurrentCpuCore().Scheduler();
+}
+
 Kernel::Scheduler& System::Scheduler(std::size_t core_index) {
     return CpuCore(core_index).Scheduler();
 }
@@ -378,6 +394,10 @@ ARM_Interface& System::ArmInterface(std::size_t core_index) {
     return CpuCore(core_index).ArmInterface();
 }
 
+const ARM_Interface& System::ArmInterface(std::size_t core_index) const {
+    return CpuCore(core_index).ArmInterface();
+}
+
 Cpu& System::CpuCore(std::size_t core_index) {
     ASSERT(core_index < NUM_CPU_CORES);
     return *impl->cpu_cores[core_index];
@@ -392,6 +412,10 @@ ExclusiveMonitor& System::Monitor() {
     return *impl->cpu_exclusive_monitor;
 }
 
+const ExclusiveMonitor& System::Monitor() const {
+    return *impl->cpu_exclusive_monitor;
+}
+
 Tegra::GPU& System::GPU() {
     return *impl->gpu_core;
 }
diff --git a/src/core/core.h b/src/core/core.h
index 173be45f8..cfacceb81 100644
--- a/src/core/core.h
+++ b/src/core/core.h
@@ -129,11 +129,11 @@ public:
      */
     bool IsPoweredOn() const;
 
-    /**
-     * Returns a reference to the telemetry session for this emulation session.
-     * @returns Reference to the telemetry session.
-     */
-    Core::TelemetrySession& TelemetrySession() const;
+    /// Gets a reference to the telemetry session for this emulation session.
+    Core::TelemetrySession& TelemetrySession();
+
+    /// Gets a reference to the telemetry session for this emulation session.
+    const Core::TelemetrySession& TelemetrySession() const;
 
     /// Prepare the core emulation for a reschedule
     void PrepareReschedule();
@@ -144,24 +144,36 @@ public:
     /// Gets an ARM interface to the CPU core that is currently running
     ARM_Interface& CurrentArmInterface();
 
+    /// Gets an ARM interface to the CPU core that is currently running
+    const ARM_Interface& CurrentArmInterface() const;
+
     /// Gets the index of the currently running CPU core
-    std::size_t CurrentCoreIndex();
+    std::size_t CurrentCoreIndex() const;
 
     /// Gets the scheduler for the CPU core that is currently running
     Kernel::Scheduler& CurrentScheduler();
 
-    /// Gets an ARM interface to the CPU core with the specified index
+    /// Gets the scheduler for the CPU core that is currently running
+    const Kernel::Scheduler& CurrentScheduler() const;
+
+    /// Gets a reference to an ARM interface for the CPU core with the specified index
     ARM_Interface& ArmInterface(std::size_t core_index);
 
+    /// Gets a const reference to an ARM interface from the CPU core with the specified index
+    const ARM_Interface& ArmInterface(std::size_t core_index) const;
+
     /// Gets a CPU interface to the CPU core with the specified index
     Cpu& CpuCore(std::size_t core_index);
 
     /// Gets a CPU interface to the CPU core with the specified index
     const Cpu& CpuCore(std::size_t core_index) const;
 
-    /// Gets the exclusive monitor
+    /// Gets a reference to the exclusive monitor
     ExclusiveMonitor& Monitor();
 
+    /// Gets a constant reference to the exclusive monitor
+    const ExclusiveMonitor& Monitor() const;
+
     /// Gets a mutable reference to the GPU interface
     Tegra::GPU& GPU();
 
@@ -230,6 +242,9 @@ private:
     /// Returns the currently running CPU core
     Cpu& CurrentCpuCore();
 
+    /// Returns the currently running CPU core
+    const Cpu& CurrentCpuCore() const;
+
     /**
      * Initialize the emulated system.
      * @param emu_window Reference to the host-system window used for video output and keyboard
diff --git a/src/core/crypto/key_manager.cpp b/src/core/crypto/key_manager.cpp
index fd0786068..904afa039 100644
--- a/src/core/crypto/key_manager.cpp
+++ b/src/core/crypto/key_manager.cpp
@@ -141,28 +141,28 @@ Key128 DeriveKeyblobMACKey(const Key128& keyblob_key, const Key128& mac_source)
     return mac_key;
 }
 
-boost::optional<Key128> DeriveSDSeed() {
+std::optional<Key128> DeriveSDSeed() {
     const FileUtil::IOFile save_43(FileUtil::GetUserPath(FileUtil::UserPath::NANDDir) +
                                        "/system/save/8000000000000043",
                                    "rb+");
     if (!save_43.IsOpen())
-        return boost::none;
+        return {};
 
     const FileUtil::IOFile sd_private(
         FileUtil::GetUserPath(FileUtil::UserPath::SDMCDir) + "/Nintendo/Contents/private", "rb+");
     if (!sd_private.IsOpen())
-        return boost::none;
+        return {};
 
     std::array<u8, 0x10> private_seed{};
     if (sd_private.ReadBytes(private_seed.data(), private_seed.size()) != private_seed.size()) {
-        return boost::none;
+        return {};
     }
 
     std::array<u8, 0x10> buffer{};
     std::size_t offset = 0;
     for (; offset + 0x10 < save_43.GetSize(); ++offset) {
         if (!save_43.Seek(offset, SEEK_SET)) {
-            return boost::none;
+            return {};
         }
 
         save_43.ReadBytes(buffer.data(), buffer.size());
@@ -172,12 +172,12 @@ boost::optional<Key128> DeriveSDSeed() {
     }
 
     if (!save_43.Seek(offset + 0x10, SEEK_SET)) {
-        return boost::none;
+        return {};
     }
 
     Key128 seed{};
     if (save_43.ReadBytes(seed.data(), seed.size()) != seed.size()) {
-        return boost::none;
+        return {};
     }
     return seed;
 }
@@ -291,26 +291,26 @@ static std::array<u8, target_size> MGF1(const std::array<u8, in_size>& seed) {
 }
 
 template <size_t size>
-static boost::optional<u64> FindTicketOffset(const std::array<u8, size>& data) {
+static std::optional<u64> FindTicketOffset(const std::array<u8, size>& data) {
     u64 offset = 0;
     for (size_t i = 0x20; i < data.size() - 0x10; ++i) {
         if (data[i] == 0x1) {
             offset = i + 1;
             break;
         } else if (data[i] != 0x0) {
-            return boost::none;
+            return {};
         }
     }
 
     return offset;
 }
 
-boost::optional<std::pair<Key128, Key128>> ParseTicket(const TicketRaw& ticket,
-                                                       const RSAKeyPair<2048>& key) {
+std::optional<std::pair<Key128, Key128>> ParseTicket(const TicketRaw& ticket,
+                                                     const RSAKeyPair<2048>& key) {
     u32 cert_authority;
     std::memcpy(&cert_authority, ticket.data() + 0x140, sizeof(cert_authority));
     if (cert_authority == 0)
-        return boost::none;
+        return {};
     if (cert_authority != Common::MakeMagic('R', 'o', 'o', 't')) {
         LOG_INFO(Crypto,
                  "Attempting to parse ticket with non-standard certificate authority {:08X}.",
@@ -321,7 +321,7 @@ boost::optional<std::pair<Key128, Key128>> ParseTicket(const TicketRaw& ticket,
     std::memcpy(rights_id.data(), ticket.data() + 0x2A0, sizeof(Key128));
 
     if (rights_id == Key128{})
-        return boost::none;
+        return {};
 
     Key128 key_temp{};
 
@@ -356,17 +356,17 @@ boost::optional<std::pair<Key128, Key128>> ParseTicket(const TicketRaw& ticket,
     std::memcpy(m_2.data(), rsa_step.data() + 0x21, m_2.size());
 
     if (m_0 != 0)
-        return boost::none;
+        return {};
 
     m_1 = m_1 ^ MGF1<0x20>(m_2);
     m_2 = m_2 ^ MGF1<0xDF>(m_1);
 
     const auto offset = FindTicketOffset(m_2);
-    if (offset == boost::none)
-        return boost::none;
-    ASSERT(offset.get() > 0);
+    if (!offset)
+        return {};
+    ASSERT(*offset > 0);
 
-    std::memcpy(key_temp.data(), m_2.data() + offset.get(), key_temp.size());
+    std::memcpy(key_temp.data(), m_2.data() + *offset, key_temp.size());
 
     return std::make_pair(rights_id, key_temp);
 }
@@ -395,7 +395,7 @@ static bool ValidCryptoRevisionString(std::string_view base, size_t begin, size_
     if (base.size() < begin + length)
         return false;
     return std::all_of(base.begin() + begin, base.begin() + begin + length,
-                       [](u8 c) { return std::isdigit(c); });
+                       [](u8 c) { return std::isxdigit(c); });
 }
 
 void KeyManager::LoadFromFile(const std::string& filename, bool is_title_keys) {
@@ -661,8 +661,8 @@ void KeyManager::DeriveSDSeedLazy() {
         return;
 
     const auto res = DeriveSDSeed();
-    if (res != boost::none)
-        SetKey(S128KeyType::SDSeed, res.get());
+    if (res)
+        SetKey(S128KeyType::SDSeed, *res);
 }
 
 static Key128 CalculateCMAC(const u8* source, size_t size, const Key128& key) {
@@ -713,7 +713,6 @@ void KeyManager::DeriveBase() {
 
     const auto sbk = GetKey(S128KeyType::SecureBoot);
     const auto tsec = GetKey(S128KeyType::TSEC);
-    const auto master_source = GetKey(S128KeyType::Source, static_cast<u64>(SourceKeyType::Master));
 
     for (size_t i = 0; i < revisions.size(); ++i) {
         if (!revisions[i])
@@ -890,9 +889,9 @@ void KeyManager::DeriveETicket(PartitionDataManager& data) {
 
     for (const auto& raw : res) {
         const auto pair = ParseTicket(raw, rsa_key);
-        if (pair == boost::none)
+        if (!pair)
             continue;
-        const auto& [rid, key] = pair.value();
+        const auto& [rid, key] = *pair;
         u128 rights_id;
         std::memcpy(rights_id.data(), rid.data(), rid.size());
         SetKey(S128KeyType::Titlekey, key, rights_id[1], rights_id[0]);
diff --git a/src/core/crypto/key_manager.h b/src/core/crypto/key_manager.h
index cccb3c0ae..22f268c65 100644
--- a/src/core/crypto/key_manager.h
+++ b/src/core/crypto/key_manager.h
@@ -6,9 +6,10 @@
 
 #include <array>
 #include <map>
+#include <optional>
 #include <string>
+
 #include <boost/container/flat_map.hpp>
-#include <boost/optional.hpp>
 #include <fmt/format.h>
 #include "common/common_types.h"
 #include "core/crypto/partition_data_manager.h"
@@ -191,14 +192,14 @@ Key128 DeriveMasterKey(const std::array<u8, 0x90>& keyblob, const Key128& master
 std::array<u8, 0x90> DecryptKeyblob(const std::array<u8, 0xB0>& encrypted_keyblob,
                                     const Key128& key);
 
-boost::optional<Key128> DeriveSDSeed();
+std::optional<Key128> DeriveSDSeed();
 Loader::ResultStatus DeriveSDKeys(std::array<Key256, 2>& sd_keys, KeyManager& keys);
 
 std::vector<TicketRaw> GetTicketblob(const FileUtil::IOFile& ticket_save);
 
 // Returns a pair of {rights_id, titlekey}. Fails if the ticket has no certificate authority (offset
 // 0x140-0x144 is zero)
-boost::optional<std::pair<Key128, Key128>> ParseTicket(
-    const TicketRaw& ticket, const RSAKeyPair<2048>& eticket_extended_key);
+std::optional<std::pair<Key128, Key128>> ParseTicket(const TicketRaw& ticket,
+                                                     const RSAKeyPair<2048>& eticket_extended_key);
 
 } // namespace Core::Crypto
diff --git a/src/core/file_sys/bis_factory.cpp b/src/core/file_sys/bis_factory.cpp
index 76a2b7e86..e29f70b3a 100644
--- a/src/core/file_sys/bis_factory.cpp
+++ b/src/core/file_sys/bis_factory.cpp
@@ -8,8 +8,9 @@
 
 namespace FileSys {
 
-BISFactory::BISFactory(VirtualDir nand_root_, VirtualDir load_root_)
+BISFactory::BISFactory(VirtualDir nand_root_, VirtualDir load_root_, VirtualDir dump_root_)
     : nand_root(std::move(nand_root_)), load_root(std::move(load_root_)),
+      dump_root(std::move(dump_root_)),
       sysnand_cache(std::make_unique<RegisteredCache>(
           GetOrCreateDirectoryRelative(nand_root, "/system/Contents/registered"))),
       usrnand_cache(std::make_unique<RegisteredCache>(
@@ -32,4 +33,10 @@ VirtualDir BISFactory::GetModificationLoadRoot(u64 title_id) const {
     return GetOrCreateDirectoryRelative(load_root, fmt::format("/{:016X}", title_id));
 }
 
+VirtualDir BISFactory::GetModificationDumpRoot(u64 title_id) const {
+    if (title_id == 0)
+        return nullptr;
+    return GetOrCreateDirectoryRelative(dump_root, fmt::format("/{:016X}", title_id));
+}
+
 } // namespace FileSys
diff --git a/src/core/file_sys/bis_factory.h b/src/core/file_sys/bis_factory.h
index 364d309bd..453c11ad2 100644
--- a/src/core/file_sys/bis_factory.h
+++ b/src/core/file_sys/bis_factory.h
@@ -17,17 +17,19 @@ class RegisteredCache;
 /// registered caches.
 class BISFactory {
 public:
-    explicit BISFactory(VirtualDir nand_root, VirtualDir load_root);
+    explicit BISFactory(VirtualDir nand_root, VirtualDir load_root, VirtualDir dump_root);
     ~BISFactory();
 
     RegisteredCache* GetSystemNANDContents() const;
     RegisteredCache* GetUserNANDContents() const;
 
     VirtualDir GetModificationLoadRoot(u64 title_id) const;
+    VirtualDir GetModificationDumpRoot(u64 title_id) const;
 
 private:
     VirtualDir nand_root;
     VirtualDir load_root;
+    VirtualDir dump_root;
 
     std::unique_ptr<RegisteredCache> sysnand_cache;
     std::unique_ptr<RegisteredCache> usrnand_cache;
diff --git a/src/core/file_sys/card_image.cpp b/src/core/file_sys/card_image.cpp
index ecdd7505b..2c145bd09 100644
--- a/src/core/file_sys/card_image.cpp
+++ b/src/core/file_sys/card_image.cpp
@@ -168,10 +168,6 @@ VirtualDir XCI::GetParentDirectory() const {
     return file->GetContainingDirectory();
 }
 
-bool XCI::ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) {
-    return false;
-}
-
 Loader::ResultStatus XCI::AddNCAFromPartition(XCIPartition part) {
     if (partitions[static_cast<std::size_t>(part)] == nullptr) {
         return Loader::ResultStatus::ErrorXCIMissingPartition;
@@ -180,7 +176,7 @@ Loader::ResultStatus XCI::AddNCAFromPartition(XCIPartition part) {
     for (const VirtualFile& file : partitions[static_cast<std::size_t>(part)]->GetFiles()) {
         if (file->GetExtension() != "nca")
             continue;
-        auto nca = std::make_shared<NCA>(file);
+        auto nca = std::make_shared<NCA>(file, nullptr, 0, keys);
         // TODO(DarkLordZach): Add proper Rev1+ Support
         if (nca->IsUpdate())
             continue;
diff --git a/src/core/file_sys/card_image.h b/src/core/file_sys/card_image.h
index 48cbef666..25f5914b6 100644
--- a/src/core/file_sys/card_image.h
+++ b/src/core/file_sys/card_image.h
@@ -9,6 +9,7 @@
 #include <vector>
 #include "common/common_types.h"
 #include "common/swap.h"
+#include "core/crypto/key_manager.h"
 #include "core/file_sys/vfs.h"
 
 namespace Loader {
@@ -94,9 +95,6 @@ public:
 
     VirtualDir GetParentDirectory() const override;
 
-protected:
-    bool ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) override;
-
 private:
     Loader::ResultStatus AddNCAFromPartition(XCIPartition part);
 
@@ -110,5 +108,7 @@ private:
     std::shared_ptr<NSP> secure_partition;
     std::shared_ptr<NCA> program;
     std::vector<std::shared_ptr<NCA>> ncas;
+
+    Core::Crypto::KeyManager keys;
 };
 } // namespace FileSys
diff --git a/src/core/file_sys/content_archive.cpp b/src/core/file_sys/content_archive.cpp
index 6c356d85d..19b6f8600 100644
--- a/src/core/file_sys/content_archive.cpp
+++ b/src/core/file_sys/content_archive.cpp
@@ -4,10 +4,9 @@
 
 #include <algorithm>
 #include <cstring>
+#include <optional>
 #include <utility>
 
-#include <boost/optional.hpp>
-
 #include "common/logging/log.h"
 #include "core/crypto/aes_util.h"
 #include "core/crypto/ctr_encryption_layer.h"
@@ -102,8 +101,9 @@ static bool IsValidNCA(const NCAHeader& header) {
     return header.magic == Common::MakeMagic('N', 'C', 'A', '3');
 }
 
-NCA::NCA(VirtualFile file_, VirtualFile bktr_base_romfs_, u64 bktr_base_ivfc_offset)
-    : file(std::move(file_)), bktr_base_romfs(std::move(bktr_base_romfs_)) {
+NCA::NCA(VirtualFile file_, VirtualFile bktr_base_romfs_, u64 bktr_base_ivfc_offset,
+         Core::Crypto::KeyManager keys_)
+    : file(std::move(file_)), bktr_base_romfs(std::move(bktr_base_romfs_)), keys(std::move(keys_)) {
     if (file == nullptr) {
         status = Loader::ResultStatus::ErrorNullFile;
         return;
@@ -306,18 +306,18 @@ bool NCA::ReadRomFSSection(const NCASectionHeader& section, const NCASectionTabl
         subsection_buckets.back().entries.push_back({section.bktr.relocation.offset, {0}, ctr_low});
         subsection_buckets.back().entries.push_back({size, {0}, 0});
 
-        boost::optional<Core::Crypto::Key128> key = boost::none;
+        std::optional<Core::Crypto::Key128> key = {};
         if (encrypted) {
             if (has_rights_id) {
                 status = Loader::ResultStatus::Success;
                 key = GetTitlekey();
-                if (key == boost::none) {
+                if (!key) {
                     status = Loader::ResultStatus::ErrorMissingTitlekey;
                     return false;
                 }
             } else {
                 key = GetKeyAreaKey(NCASectionCryptoType::BKTR);
-                if (key == boost::none) {
+                if (!key) {
                     status = Loader::ResultStatus::ErrorMissingKeyAreaKey;
                     return false;
                 }
@@ -332,7 +332,7 @@ bool NCA::ReadRomFSSection(const NCASectionHeader& section, const NCASectionTabl
         auto bktr = std::make_shared<BKTR>(
             bktr_base_romfs, std::make_shared<OffsetVfsFile>(file, romfs_size, base_offset),
             relocation_block, relocation_buckets, subsection_block, subsection_buckets, encrypted,
-            encrypted ? key.get() : Core::Crypto::Key128{}, base_offset, bktr_base_ivfc_offset,
+            encrypted ? *key : Core::Crypto::Key128{}, base_offset, bktr_base_ivfc_offset,
             section.raw.section_ctr);
 
         // BKTR applies to entire IVFC, so make an offset version to level 6
@@ -388,11 +388,11 @@ u8 NCA::GetCryptoRevision() const {
     return master_key_id;
 }
 
-boost::optional<Core::Crypto::Key128> NCA::GetKeyAreaKey(NCASectionCryptoType type) const {
+std::optional<Core::Crypto::Key128> NCA::GetKeyAreaKey(NCASectionCryptoType type) const {
     const auto master_key_id = GetCryptoRevision();
 
     if (!keys.HasKey(Core::Crypto::S128KeyType::KeyArea, master_key_id, header.key_index))
-        return boost::none;
+        return {};
 
     std::vector<u8> key_area(header.key_area.begin(), header.key_area.end());
     Core::Crypto::AESCipher<Core::Crypto::Key128> cipher(
@@ -416,25 +416,25 @@ boost::optional<Core::Crypto::Key128> NCA::GetKeyAreaKey(NCASectionCryptoType ty
     return out;
 }
 
-boost::optional<Core::Crypto::Key128> NCA::GetTitlekey() {
+std::optional<Core::Crypto::Key128> NCA::GetTitlekey() {
     const auto master_key_id = GetCryptoRevision();
 
     u128 rights_id{};
     memcpy(rights_id.data(), header.rights_id.data(), 16);
     if (rights_id == u128{}) {
         status = Loader::ResultStatus::ErrorInvalidRightsID;
-        return boost::none;
+        return {};
     }
 
     auto titlekey = keys.GetKey(Core::Crypto::S128KeyType::Titlekey, rights_id[1], rights_id[0]);
     if (titlekey == Core::Crypto::Key128{}) {
         status = Loader::ResultStatus::ErrorMissingTitlekey;
-        return boost::none;
+        return {};
     }
 
     if (!keys.HasKey(Core::Crypto::S128KeyType::Titlekek, master_key_id)) {
         status = Loader::ResultStatus::ErrorMissingTitlekek;
-        return boost::none;
+        return {};
     }
 
     Core::Crypto::AESCipher<Core::Crypto::Key128> cipher(
@@ -458,25 +458,25 @@ VirtualFile NCA::Decrypt(const NCASectionHeader& s_header, VirtualFile in, u64 s
     case NCASectionCryptoType::BKTR:
         LOG_DEBUG(Crypto, "called with mode=CTR, starting_offset={:016X}", starting_offset);
         {
-            boost::optional<Core::Crypto::Key128> key = boost::none;
+            std::optional<Core::Crypto::Key128> key = {};
             if (has_rights_id) {
                 status = Loader::ResultStatus::Success;
                 key = GetTitlekey();
-                if (key == boost::none) {
+                if (!key) {
                     if (status == Loader::ResultStatus::Success)
                         status = Loader::ResultStatus::ErrorMissingTitlekey;
                     return nullptr;
                 }
             } else {
                 key = GetKeyAreaKey(NCASectionCryptoType::CTR);
-                if (key == boost::none) {
+                if (!key) {
                     status = Loader::ResultStatus::ErrorMissingKeyAreaKey;
                     return nullptr;
                 }
             }
 
-            auto out = std::make_shared<Core::Crypto::CTREncryptionLayer>(
-                std::move(in), key.value(), starting_offset);
+            auto out = std::make_shared<Core::Crypto::CTREncryptionLayer>(std::move(in), *key,
+                                                                          starting_offset);
             std::vector<u8> iv(16);
             for (u8 i = 0; i < 8; ++i)
                 iv[i] = s_header.raw.section_ctr[0x8 - i - 1];
@@ -546,7 +546,4 @@ u64 NCA::GetBaseIVFCOffset() const {
     return ivfc_offset;
 }
 
-bool NCA::ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) {
-    return false;
-}
 } // namespace FileSys
diff --git a/src/core/file_sys/content_archive.h b/src/core/file_sys/content_archive.h
index 1c903cd3f..99294cbb4 100644
--- a/src/core/file_sys/content_archive.h
+++ b/src/core/file_sys/content_archive.h
@@ -6,9 +6,10 @@
 
 #include <array>
 #include <memory>
+#include <optional>
 #include <string>
 #include <vector>
-#include <boost/optional.hpp>
+
 #include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/swap.h"
@@ -78,7 +79,8 @@ inline bool IsDirectoryExeFS(const std::shared_ptr<VfsDirectory>& pfs) {
 class NCA : public ReadOnlyVfsDirectory {
 public:
     explicit NCA(VirtualFile file, VirtualFile bktr_base_romfs = nullptr,
-                 u64 bktr_base_ivfc_offset = 0);
+                 u64 bktr_base_ivfc_offset = 0,
+                 Core::Crypto::KeyManager keys = Core::Crypto::KeyManager());
     ~NCA() override;
 
     Loader::ResultStatus GetStatus() const;
@@ -100,9 +102,6 @@ public:
     // Returns the base ivfc offset used in BKTR patching.
     u64 GetBaseIVFCOffset() const;
 
-protected:
-    bool ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) override;
-
 private:
     bool CheckSupportedNCA(const NCAHeader& header);
     bool HandlePotentialHeaderDecryption();
@@ -114,8 +113,8 @@ private:
     bool ReadPFS0Section(const NCASectionHeader& section, const NCASectionTableEntry& entry);
 
     u8 GetCryptoRevision() const;
-    boost::optional<Core::Crypto::Key128> GetKeyAreaKey(NCASectionCryptoType type) const;
-    boost::optional<Core::Crypto::Key128> GetTitlekey();
+    std::optional<Core::Crypto::Key128> GetKeyAreaKey(NCASectionCryptoType type) const;
+    std::optional<Core::Crypto::Key128> GetTitlekey();
     VirtualFile Decrypt(const NCASectionHeader& header, VirtualFile in, u64 starting_offset);
 
     std::vector<VirtualDir> dirs;
diff --git a/src/core/file_sys/control_metadata.cpp b/src/core/file_sys/control_metadata.cpp
index a012c2be9..c8fa912bf 100644
--- a/src/core/file_sys/control_metadata.cpp
+++ b/src/core/file_sys/control_metadata.cpp
@@ -66,4 +66,10 @@ std::string NACP::GetVersionString() const {
     return Common::StringFromFixedZeroTerminatedBuffer(raw->version_string.data(),
                                                        raw->version_string.size());
 }
+
+std::vector<u8> NACP::GetRawBytes() const {
+    std::vector<u8> out(sizeof(RawNACP));
+    std::memcpy(out.data(), raw.get(), sizeof(RawNACP));
+    return out;
+}
 } // namespace FileSys
diff --git a/src/core/file_sys/control_metadata.h b/src/core/file_sys/control_metadata.h
index 141f7e056..bfaad46b4 100644
--- a/src/core/file_sys/control_metadata.h
+++ b/src/core/file_sys/control_metadata.h
@@ -81,6 +81,7 @@ public:
     u64 GetTitleId() const;
     u64 GetDLCBaseTitleId() const;
     std::string GetVersionString() const;
+    std::vector<u8> GetRawBytes() const;
 
 private:
     std::unique_ptr<RawNACP> raw;
diff --git a/src/core/file_sys/errors.h b/src/core/file_sys/errors.h
index fea0593c7..e4a4ee4ab 100644
--- a/src/core/file_sys/errors.h
+++ b/src/core/file_sys/errors.h
@@ -8,25 +8,10 @@
 
 namespace FileSys {
 
-namespace ErrCodes {
-enum {
-    NotFound = 1,
-    TitleNotFound = 1002,
-    SdCardNotFound = 2001,
-    RomFSNotFound = 2520,
-};
-}
-
-constexpr ResultCode ERROR_PATH_NOT_FOUND(ErrorModule::FS, ErrCodes::NotFound);
-
-// TODO(bunnei): Replace these with correct errors for Switch OS
-constexpr ResultCode ERROR_INVALID_PATH(-1);
-constexpr ResultCode ERROR_UNSUPPORTED_OPEN_FLAGS(-1);
-constexpr ResultCode ERROR_INVALID_OPEN_FLAGS(-1);
-constexpr ResultCode ERROR_FILE_NOT_FOUND(-1);
-constexpr ResultCode ERROR_UNEXPECTED_FILE_OR_DIRECTORY(-1);
-constexpr ResultCode ERROR_DIRECTORY_ALREADY_EXISTS(-1);
-constexpr ResultCode ERROR_FILE_ALREADY_EXISTS(-1);
-constexpr ResultCode ERROR_DIRECTORY_NOT_EMPTY(-1);
+constexpr ResultCode ERROR_PATH_NOT_FOUND{ErrorModule::FS, 1};
+constexpr ResultCode ERROR_ENTITY_NOT_FOUND{ErrorModule::FS, 1002};
+constexpr ResultCode ERROR_SD_CARD_NOT_FOUND{ErrorModule::FS, 2001};
+constexpr ResultCode ERROR_INVALID_OFFSET{ErrorModule::FS, 6061};
+constexpr ResultCode ERROR_INVALID_SIZE{ErrorModule::FS, 6062};
 
 } // namespace FileSys
diff --git a/src/core/file_sys/fsmitm_romfsbuild.h b/src/core/file_sys/fsmitm_romfsbuild.h
index 3d377b0af..a62502193 100644
--- a/src/core/file_sys/fsmitm_romfsbuild.h
+++ b/src/core/file_sys/fsmitm_romfsbuild.h
@@ -27,7 +27,6 @@
 #include <map>
 #include <memory>
 #include <string>
-#include <boost/detail/container_fwd.hpp>
 #include "common/common_types.h"
 #include "core/file_sys/vfs.h"
 
diff --git a/src/core/file_sys/ips_layer.cpp b/src/core/file_sys/ips_layer.cpp
index 554eae9bc..485c4913a 100644
--- a/src/core/file_sys/ips_layer.cpp
+++ b/src/core/file_sys/ips_layer.cpp
@@ -99,16 +99,16 @@ VirtualFile PatchIPS(const VirtualFile& in, const VirtualFile& ips) {
             u16 rle_size{};
             if (ips->ReadObject(&rle_size, offset) != sizeof(u16))
                 return nullptr;
-            rle_size = Common::swap16(data_size);
+            rle_size = Common::swap16(rle_size);
             offset += sizeof(u16);
 
             const auto data = ips->ReadByte(offset++);
-            if (data == boost::none)
+            if (!data)
                 return nullptr;
 
             if (real_offset + rle_size > in_data.size())
                 rle_size = static_cast<u16>(in_data.size() - real_offset);
-            std::memset(in_data.data() + real_offset, data.get(), rle_size);
+            std::memset(in_data.data() + real_offset, *data, rle_size);
         } else { // Standard Patch
             auto read = data_size;
             if (real_offset + read > in_data.size())
diff --git a/src/core/file_sys/partition_filesystem.cpp b/src/core/file_sys/partition_filesystem.cpp
index 5791c76ff..a5259a593 100644
--- a/src/core/file_sys/partition_filesystem.cpp
+++ b/src/core/file_sys/partition_filesystem.cpp
@@ -83,7 +83,7 @@ std::vector<std::shared_ptr<VfsFile>> PartitionFilesystem::GetFiles() const {
 }
 
 std::vector<std::shared_ptr<VfsDirectory>> PartitionFilesystem::GetSubdirectories() const {
-    return pfs_dirs;
+    return {};
 }
 
 std::string PartitionFilesystem::GetName() const {
@@ -103,18 +103,4 @@ void PartitionFilesystem::PrintDebugInfo() const {
                   pfs_files[i]->GetName(), pfs_files[i]->GetSize());
     }
 }
-
-bool PartitionFilesystem::ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) {
-    const auto iter = std::find(pfs_files.begin(), pfs_files.end(), file);
-    if (iter == pfs_files.end())
-        return false;
-
-    const std::ptrdiff_t offset = std::distance(pfs_files.begin(), iter);
-    pfs_files[offset] = std::move(pfs_files.back());
-    pfs_files.pop_back();
-
-    pfs_dirs.emplace_back(std::move(dir));
-
-    return true;
-}
 } // namespace FileSys
diff --git a/src/core/file_sys/partition_filesystem.h b/src/core/file_sys/partition_filesystem.h
index 739c63a7f..248fdfdeb 100644
--- a/src/core/file_sys/partition_filesystem.h
+++ b/src/core/file_sys/partition_filesystem.h
@@ -35,9 +35,6 @@ public:
     std::shared_ptr<VfsDirectory> GetParentDirectory() const override;
     void PrintDebugInfo() const;
 
-protected:
-    bool ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) override;
-
 private:
     struct Header {
         u32_le magic;
@@ -84,7 +81,6 @@ private:
     std::size_t content_offset = 0;
 
     std::vector<VirtualFile> pfs_files;
-    std::vector<VirtualDir> pfs_dirs;
 };
 
 } // namespace FileSys
diff --git a/src/core/file_sys/patch_manager.cpp b/src/core/file_sys/patch_manager.cpp
index 0117cb0bf..8d062eb3e 100644
--- a/src/core/file_sys/patch_manager.cpp
+++ b/src/core/file_sys/patch_manager.cpp
@@ -19,6 +19,7 @@
 #include "core/file_sys/vfs_vector.h"
 #include "core/hle/service/filesystem/filesystem.h"
 #include "core/loader/loader.h"
+#include "core/settings.h"
 
 namespace FileSys {
 
@@ -61,13 +62,12 @@ VirtualDir PatchManager::PatchExeFS(VirtualDir exefs) const {
     // Game Updates
     const auto update_tid = GetUpdateTitleID(title_id);
     const auto update = installed->GetEntry(update_tid, ContentRecordType::Program);
-    if (update != nullptr) {
-        if (update->GetStatus() == Loader::ResultStatus::ErrorMissingBKTRBaseRomFS &&
-            update->GetExeFS() != nullptr) {
-            LOG_INFO(Loader, "    ExeFS: Update ({}) applied successfully",
-                     FormatTitleVersion(installed->GetEntryVersion(update_tid).get_value_or(0)));
-            exefs = update->GetExeFS();
-        }
+
+    if (update != nullptr && update->GetExeFS() != nullptr &&
+        update->GetStatus() == Loader::ResultStatus::ErrorMissingBKTRBaseRomFS) {
+        LOG_INFO(Loader, "    ExeFS: Update ({}) applied successfully",
+                 FormatTitleVersion(installed->GetEntryVersion(update_tid).value_or(0)));
+        exefs = update->GetExeFS();
     }
 
     return exefs;
@@ -120,6 +120,18 @@ std::vector<u8> PatchManager::PatchNSO(const std::vector<u8>& nso) const {
     const auto build_id_raw = Common::HexArrayToString(header.build_id);
     const auto build_id = build_id_raw.substr(0, build_id_raw.find_last_not_of('0') + 1);
 
+    if (Settings::values.dump_nso) {
+        LOG_INFO(Loader, "Dumping NSO for build_id={}, title_id={:016X}", build_id, title_id);
+        const auto dump_dir = Service::FileSystem::GetModificationDumpRoot(title_id);
+        if (dump_dir != nullptr) {
+            const auto nso_dir = GetOrCreateDirectoryRelative(dump_dir, "/nso");
+            const auto file = nso_dir->CreateFile(fmt::format("{}.nso", build_id));
+
+            file->Resize(nso.size());
+            file->WriteBytes(nso);
+        }
+    }
+
     LOG_INFO(Loader, "Patching NSO for build_id={}", build_id);
 
     const auto load_dir = Service::FileSystem::GetModificationLoadRoot(title_id);
@@ -168,7 +180,8 @@ bool PatchManager::HasNSOPatch(const std::array<u8, 32>& build_id_) const {
 
 static void ApplyLayeredFS(VirtualFile& romfs, u64 title_id, ContentRecordType type) {
     const auto load_dir = Service::FileSystem::GetModificationLoadRoot(title_id);
-    if (type != ContentRecordType::Program || load_dir == nullptr || load_dir->GetSize() <= 0) {
+    if ((type != ContentRecordType::Program && type != ContentRecordType::Data) ||
+        load_dir == nullptr || load_dir->GetSize() <= 0) {
         return;
     }
 
@@ -218,7 +231,7 @@ VirtualFile PatchManager::PatchRomFS(VirtualFile romfs, u64 ivfc_offset, Content
                                         title_id, static_cast<u8>(type))
                                 .c_str();
 
-    if (type == ContentRecordType::Program)
+    if (type == ContentRecordType::Program || type == ContentRecordType::Data)
         LOG_INFO(Loader, log_string);
     else
         LOG_DEBUG(Loader, log_string);
@@ -236,7 +249,7 @@ VirtualFile PatchManager::PatchRomFS(VirtualFile romfs, u64 ivfc_offset, Content
         if (new_nca->GetStatus() == Loader::ResultStatus::Success &&
             new_nca->GetRomFS() != nullptr) {
             LOG_INFO(Loader, "    RomFS: Update ({}) applied successfully",
-                     FormatTitleVersion(installed->GetEntryVersion(update_tid).get_value_or(0)));
+                     FormatTitleVersion(installed->GetEntryVersion(update_tid).value_or(0)));
             romfs = new_nca->GetRomFS();
         }
     } else if (update_raw != nullptr) {
@@ -280,12 +293,11 @@ std::map<std::string, std::string, std::less<>> PatchManager::GetPatchVersionNam
     } else {
         if (installed->HasEntry(update_tid, ContentRecordType::Program)) {
             const auto meta_ver = installed->GetEntryVersion(update_tid);
-            if (meta_ver == boost::none || meta_ver.get() == 0) {
+            if (meta_ver.value_or(0) == 0) {
                 out.insert_or_assign("Update", "");
             } else {
                 out.insert_or_assign(
-                    "Update",
-                    FormatTitleVersion(meta_ver.get(), TitleVersionFormat::ThreeElements));
+                    "Update", FormatTitleVersion(*meta_ver, TitleVersionFormat::ThreeElements));
             }
         } else if (update_raw != nullptr) {
             out.insert_or_assign("Update", "PACKED");
diff --git a/src/core/file_sys/registered_cache.cpp b/src/core/file_sys/registered_cache.cpp
index 1febb398e..a3f8f2f73 100644
--- a/src/core/file_sys/registered_cache.cpp
+++ b/src/core/file_sys/registered_cache.cpp
@@ -2,6 +2,7 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <algorithm>
 #include <regex>
 #include <mbedtls/sha256.h>
 #include "common/assert.h"
@@ -30,6 +31,14 @@ bool operator<(const RegisteredCacheEntry& lhs, const RegisteredCacheEntry& rhs)
     return (lhs.title_id < rhs.title_id) || (lhs.title_id == rhs.title_id && lhs.type < rhs.type);
 }
 
+bool operator==(const RegisteredCacheEntry& lhs, const RegisteredCacheEntry& rhs) {
+    return std::tie(lhs.title_id, lhs.type) == std::tie(rhs.title_id, rhs.type);
+}
+
+bool operator!=(const RegisteredCacheEntry& lhs, const RegisteredCacheEntry& rhs) {
+    return !operator==(lhs, rhs);
+}
+
 static bool FollowsTwoDigitDirFormat(std::string_view name) {
     static const std::regex two_digit_regex("000000[0-9A-F]{2}", std::regex_constants::ECMAScript |
                                                                      std::regex_constants::icase);
@@ -97,9 +106,12 @@ static ContentRecordType GetCRTypeFromNCAType(NCAContentType type) {
 
 VirtualFile RegisteredCache::OpenFileOrDirectoryConcat(const VirtualDir& dir,
                                                        std::string_view path) const {
-    if (dir->GetFileRelative(path) != nullptr)
-        return dir->GetFileRelative(path);
-    if (dir->GetDirectoryRelative(path) != nullptr) {
+    const auto file = dir->GetFileRelative(path);
+    if (file != nullptr)
+        return file;
+
+    const auto nca_dir = dir->GetDirectoryRelative(path);
+    if (nca_dir != nullptr) {
         const auto nca_dir = dir->GetDirectoryRelative(path);
         VirtualFile file = nullptr;
 
@@ -150,28 +162,28 @@ VirtualFile RegisteredCache::GetFileAtID(NcaID id) const {
     return file;
 }
 
-static boost::optional<NcaID> CheckMapForContentRecord(
+static std::optional<NcaID> CheckMapForContentRecord(
     const boost::container::flat_map<u64, CNMT>& map, u64 title_id, ContentRecordType type) {
     if (map.find(title_id) == map.end())
-        return boost::none;
+        return {};
 
     const auto& cnmt = map.at(title_id);
 
     const auto iter = std::find_if(cnmt.GetContentRecords().begin(), cnmt.GetContentRecords().end(),
                                    [type](const ContentRecord& rec) { return rec.type == type; });
     if (iter == cnmt.GetContentRecords().end())
-        return boost::none;
+        return {};
 
-    return boost::make_optional(iter->nca_id);
+    return std::make_optional(iter->nca_id);
 }
 
-boost::optional<NcaID> RegisteredCache::GetNcaIDFromMetadata(u64 title_id,
-                                                             ContentRecordType type) const {
+std::optional<NcaID> RegisteredCache::GetNcaIDFromMetadata(u64 title_id,
+                                                           ContentRecordType type) const {
     if (type == ContentRecordType::Meta && meta_id.find(title_id) != meta_id.end())
         return meta_id.at(title_id);
 
     const auto res1 = CheckMapForContentRecord(yuzu_meta, title_id, type);
-    if (res1 != boost::none)
+    if (res1)
         return res1;
     return CheckMapForContentRecord(meta, title_id, type);
 }
@@ -216,7 +228,7 @@ void RegisteredCache::ProcessFiles(const std::vector<NcaID>& ids) {
 
         if (file == nullptr)
             continue;
-        const auto nca = std::make_shared<NCA>(parser(file, id));
+        const auto nca = std::make_shared<NCA>(parser(file, id), nullptr, 0, keys);
         if (nca->GetStatus() != Loader::ResultStatus::Success ||
             nca->GetType() != NCAContentType::Meta) {
             continue;
@@ -274,17 +286,14 @@ bool RegisteredCache::HasEntry(RegisteredCacheEntry entry) const {
 
 VirtualFile RegisteredCache::GetEntryUnparsed(u64 title_id, ContentRecordType type) const {
     const auto id = GetNcaIDFromMetadata(title_id, type);
-    if (id == boost::none)
-        return nullptr;
-
-    return GetFileAtID(id.get());
+    return id ? GetFileAtID(*id) : nullptr;
 }
 
 VirtualFile RegisteredCache::GetEntryUnparsed(RegisteredCacheEntry entry) const {
     return GetEntryUnparsed(entry.title_id, entry.type);
 }
 
-boost::optional<u32> RegisteredCache::GetEntryVersion(u64 title_id) const {
+std::optional<u32> RegisteredCache::GetEntryVersion(u64 title_id) const {
     const auto meta_iter = meta.find(title_id);
     if (meta_iter != meta.end())
         return meta_iter->second.GetTitleVersion();
@@ -293,15 +302,12 @@ boost::optional<u32> RegisteredCache::GetEntryVersion(u64 title_id) const {
     if (yuzu_meta_iter != yuzu_meta.end())
         return yuzu_meta_iter->second.GetTitleVersion();
 
-    return boost::none;
+    return {};
 }
 
 VirtualFile RegisteredCache::GetEntryRaw(u64 title_id, ContentRecordType type) const {
     const auto id = GetNcaIDFromMetadata(title_id, type);
-    if (id == boost::none)
-        return nullptr;
-
-    return parser(GetFileAtID(id.get()), id.get());
+    return id ? parser(GetFileAtID(*id), *id) : nullptr;
 }
 
 VirtualFile RegisteredCache::GetEntryRaw(RegisteredCacheEntry entry) const {
@@ -312,7 +318,7 @@ std::unique_ptr<NCA> RegisteredCache::GetEntry(u64 title_id, ContentRecordType t
     const auto raw = GetEntryRaw(title_id, type);
     if (raw == nullptr)
         return nullptr;
-    return std::make_unique<NCA>(raw);
+    return std::make_unique<NCA>(raw, nullptr, 0, keys);
 }
 
 std::unique_ptr<NCA> RegisteredCache::GetEntry(RegisteredCacheEntry entry) const {
@@ -355,8 +361,8 @@ std::vector<RegisteredCacheEntry> RegisteredCache::ListEntries() const {
 }
 
 std::vector<RegisteredCacheEntry> RegisteredCache::ListEntriesFilter(
-    boost::optional<TitleType> title_type, boost::optional<ContentRecordType> record_type,
-    boost::optional<u64> title_id) const {
+    std::optional<TitleType> title_type, std::optional<ContentRecordType> record_type,
+    std::optional<u64> title_id) const {
     std::vector<RegisteredCacheEntry> out;
     IterateAllMetadata<RegisteredCacheEntry>(
         out,
@@ -364,11 +370,11 @@ std::vector<RegisteredCacheEntry> RegisteredCache::ListEntriesFilter(
             return RegisteredCacheEntry{c.GetTitleID(), r.type};
         },
         [&title_type, &record_type, &title_id](const CNMT& c, const ContentRecord& r) {
-            if (title_type != boost::none && title_type.get() != c.GetType())
+            if (title_type && *title_type != c.GetType())
                 return false;
-            if (record_type != boost::none && record_type.get() != r.type)
+            if (record_type && *record_type != r.type)
                 return false;
-            if (title_id != boost::none && title_id.get() != c.GetTitleID())
+            if (title_id && *title_id != c.GetTitleID())
                 return false;
             return true;
         });
@@ -450,7 +456,7 @@ InstallResult RegisteredCache::InstallEntry(std::shared_ptr<NCA> nca, TitleType
 
 InstallResult RegisteredCache::RawInstallNCA(std::shared_ptr<NCA> nca, const VfsCopyFunction& copy,
                                              bool overwrite_if_exists,
-                                             boost::optional<NcaID> override_id) {
+                                             std::optional<NcaID> override_id) {
     const auto in = nca->GetBaseFile();
     Core::Crypto::SHA256Hash hash{};
 
@@ -459,12 +465,12 @@ InstallResult RegisteredCache::RawInstallNCA(std::shared_ptr<NCA> nca, const Vfs
     // game is massive), we're going to cheat and only hash the first MB of the NCA.
     // Also, for XCIs the NcaID matters, so if the override id isn't none, use that.
     NcaID id{};
-    if (override_id == boost::none) {
+    if (override_id) {
+        id = *override_id;
+    } else {
         const auto& data = in->ReadBytes(0x100000);
         mbedtls_sha256(data.data(), data.size(), hash.data(), 0);
         memcpy(id.data(), hash.data(), 16);
-    } else {
-        id = override_id.get();
     }
 
     std::string path = GetRelativePathFromNcaID(id, false, true);
@@ -534,14 +540,14 @@ bool RegisteredCacheUnion::HasEntry(RegisteredCacheEntry entry) const {
     return HasEntry(entry.title_id, entry.type);
 }
 
-boost::optional<u32> RegisteredCacheUnion::GetEntryVersion(u64 title_id) const {
+std::optional<u32> RegisteredCacheUnion::GetEntryVersion(u64 title_id) const {
     for (const auto& c : caches) {
         const auto res = c->GetEntryVersion(title_id);
-        if (res != boost::none)
+        if (res)
             return res;
     }
 
-    return boost::none;
+    return {};
 }
 
 VirtualFile RegisteredCacheUnion::GetEntryUnparsed(u64 title_id, ContentRecordType type) const {
@@ -593,12 +599,15 @@ std::vector<RegisteredCacheEntry> RegisteredCacheUnion::ListEntries() const {
             },
             [](const CNMT& c, const ContentRecord& r) { return true; });
     }
+
+    std::sort(out.begin(), out.end());
+    out.erase(std::unique(out.begin(), out.end()), out.end());
     return out;
 }
 
 std::vector<RegisteredCacheEntry> RegisteredCacheUnion::ListEntriesFilter(
-    boost::optional<TitleType> title_type, boost::optional<ContentRecordType> record_type,
-    boost::optional<u64> title_id) const {
+    std::optional<TitleType> title_type, std::optional<ContentRecordType> record_type,
+    std::optional<u64> title_id) const {
     std::vector<RegisteredCacheEntry> out;
     for (const auto& c : caches) {
         c->IterateAllMetadata<RegisteredCacheEntry>(
@@ -607,15 +616,18 @@ std::vector<RegisteredCacheEntry> RegisteredCacheUnion::ListEntriesFilter(
                 return RegisteredCacheEntry{c.GetTitleID(), r.type};
             },
             [&title_type, &record_type, &title_id](const CNMT& c, const ContentRecord& r) {
-                if (title_type != boost::none && title_type.get() != c.GetType())
+                if (title_type && *title_type != c.GetType())
                     return false;
-                if (record_type != boost::none && record_type.get() != r.type)
+                if (record_type && *record_type != r.type)
                     return false;
-                if (title_id != boost::none && title_id.get() != c.GetTitleID())
+                if (title_id && *title_id != c.GetTitleID())
                     return false;
                 return true;
             });
     }
+
+    std::sort(out.begin(), out.end());
+    out.erase(std::unique(out.begin(), out.end()), out.end());
     return out;
 }
 } // namespace FileSys
diff --git a/src/core/file_sys/registered_cache.h b/src/core/file_sys/registered_cache.h
index 5ddacba47..6b89db8de 100644
--- a/src/core/file_sys/registered_cache.h
+++ b/src/core/file_sys/registered_cache.h
@@ -12,6 +12,7 @@
 #include <vector>
 #include <boost/container/flat_map.hpp>
 #include "common/common_types.h"
+#include "core/crypto/key_manager.h"
 #include "core/file_sys/vfs.h"
 
 namespace FileSys {
@@ -50,6 +51,10 @@ constexpr u64 GetUpdateTitleID(u64 base_title_id) {
 // boost flat_map requires operator< for O(log(n)) lookups.
 bool operator<(const RegisteredCacheEntry& lhs, const RegisteredCacheEntry& rhs);
 
+// std unique requires operator== to identify duplicates.
+bool operator==(const RegisteredCacheEntry& lhs, const RegisteredCacheEntry& rhs);
+bool operator!=(const RegisteredCacheEntry& lhs, const RegisteredCacheEntry& rhs);
+
 /*
  * A class that catalogues NCAs in the registered directory structure.
  * Nintendo's registered format follows this structure:
@@ -60,8 +65,8 @@ bool operator<(const RegisteredCacheEntry& lhs, const RegisteredCacheEntry& rhs)
  *         | 00
  *         | 01 <- Actual content split along 4GB boundaries. (optional)
  *
- * (This impl also supports substituting the nca dir for an nca file, as that's more convenient when
- * 4GB splitting can be ignored.)
+ * (This impl also supports substituting the nca dir for an nca file, as that's more convenient
+ * when 4GB splitting can be ignored.)
  */
 class RegisteredCache {
     friend class RegisteredCacheUnion;
@@ -80,7 +85,7 @@ public:
     bool HasEntry(u64 title_id, ContentRecordType type) const;
     bool HasEntry(RegisteredCacheEntry entry) const;
 
-    boost::optional<u32> GetEntryVersion(u64 title_id) const;
+    std::optional<u32> GetEntryVersion(u64 title_id) const;
 
     VirtualFile GetEntryUnparsed(u64 title_id, ContentRecordType type) const;
     VirtualFile GetEntryUnparsed(RegisteredCacheEntry entry) const;
@@ -92,11 +97,10 @@ public:
     std::unique_ptr<NCA> GetEntry(RegisteredCacheEntry entry) const;
 
     std::vector<RegisteredCacheEntry> ListEntries() const;
-    // If a parameter is not boost::none, it will be filtered for from all entries.
+    // If a parameter is not std::nullopt, it will be filtered for from all entries.
     std::vector<RegisteredCacheEntry> ListEntriesFilter(
-        boost::optional<TitleType> title_type = boost::none,
-        boost::optional<ContentRecordType> record_type = boost::none,
-        boost::optional<u64> title_id = boost::none) const;
+        std::optional<TitleType> title_type = {}, std::optional<ContentRecordType> record_type = {},
+        std::optional<u64> title_id = {}) const;
 
     // Raw copies all the ncas from the xci/nsp to the csache. Does some quick checks to make sure
     // there is a meta NCA and all of them are accessible.
@@ -121,16 +125,17 @@ private:
     std::vector<NcaID> AccumulateFiles() const;
     void ProcessFiles(const std::vector<NcaID>& ids);
     void AccumulateYuzuMeta();
-    boost::optional<NcaID> GetNcaIDFromMetadata(u64 title_id, ContentRecordType type) const;
+    std::optional<NcaID> GetNcaIDFromMetadata(u64 title_id, ContentRecordType type) const;
     VirtualFile GetFileAtID(NcaID id) const;
     VirtualFile OpenFileOrDirectoryConcat(const VirtualDir& dir, std::string_view path) const;
     InstallResult RawInstallNCA(std::shared_ptr<NCA> nca, const VfsCopyFunction& copy,
-                                bool overwrite_if_exists,
-                                boost::optional<NcaID> override_id = boost::none);
+                                bool overwrite_if_exists, std::optional<NcaID> override_id = {});
     bool RawInstallYuzuMeta(const CNMT& cnmt);
 
     VirtualDir dir;
     RegisteredCacheParsingFunction parser;
+    Core::Crypto::KeyManager keys;
+
     // maps tid -> NcaID of meta
     boost::container::flat_map<u64, NcaID> meta_id;
     // maps tid -> meta
@@ -149,7 +154,7 @@ public:
     bool HasEntry(u64 title_id, ContentRecordType type) const;
     bool HasEntry(RegisteredCacheEntry entry) const;
 
-    boost::optional<u32> GetEntryVersion(u64 title_id) const;
+    std::optional<u32> GetEntryVersion(u64 title_id) const;
 
     VirtualFile GetEntryUnparsed(u64 title_id, ContentRecordType type) const;
     VirtualFile GetEntryUnparsed(RegisteredCacheEntry entry) const;
@@ -161,11 +166,10 @@ public:
     std::unique_ptr<NCA> GetEntry(RegisteredCacheEntry entry) const;
 
     std::vector<RegisteredCacheEntry> ListEntries() const;
-    // If a parameter is not boost::none, it will be filtered for from all entries.
+    // If a parameter is not std::nullopt, it will be filtered for from all entries.
     std::vector<RegisteredCacheEntry> ListEntriesFilter(
-        boost::optional<TitleType> title_type = boost::none,
-        boost::optional<ContentRecordType> record_type = boost::none,
-        boost::optional<u64> title_id = boost::none) const;
+        std::optional<TitleType> title_type = {}, std::optional<ContentRecordType> record_type = {},
+        std::optional<u64> title_id = {}) const;
 
 private:
     std::vector<RegisteredCache*> caches;
diff --git a/src/core/file_sys/savedata_factory.cpp b/src/core/file_sys/savedata_factory.cpp
index ef1aaebbb..5434f2149 100644
--- a/src/core/file_sys/savedata_factory.cpp
+++ b/src/core/file_sys/savedata_factory.cpp
@@ -83,28 +83,32 @@ ResultVal<VirtualDir> SaveDataFactory::Open(SaveDataSpaceId space, SaveDataDescr
     return MakeResult<VirtualDir>(std::move(out));
 }
 
-std::string SaveDataFactory::GetFullPath(SaveDataSpaceId space, SaveDataType type, u64 title_id,
-                                         u128 user_id, u64 save_id) {
-    // According to switchbrew, if a save is of type SaveData and the title id field is 0, it should
-    // be interpreted as the title id of the current process.
-    if (type == SaveDataType::SaveData && title_id == 0)
-        title_id = Core::CurrentProcess()->GetTitleID();
-
-    std::string out;
+VirtualDir SaveDataFactory::GetSaveDataSpaceDirectory(SaveDataSpaceId space) const {
+    return dir->GetDirectoryRelative(GetSaveDataSpaceIdPath(space));
+}
 
+std::string SaveDataFactory::GetSaveDataSpaceIdPath(SaveDataSpaceId space) {
     switch (space) {
     case SaveDataSpaceId::NandSystem:
-        out = "/system/";
-        break;
+        return "/system/";
     case SaveDataSpaceId::NandUser:
-        out = "/user/";
-        break;
+        return "/user/";
     case SaveDataSpaceId::TemporaryStorage:
-        out = "/temp/";
-        break;
+        return "/temp/";
     default:
         ASSERT_MSG(false, "Unrecognized SaveDataSpaceId: {:02X}", static_cast<u8>(space));
+        return "/unrecognized/"; ///< To prevent corruption when ignoring asserts.
     }
+}
+
+std::string SaveDataFactory::GetFullPath(SaveDataSpaceId space, SaveDataType type, u64 title_id,
+                                         u128 user_id, u64 save_id) {
+    // According to switchbrew, if a save is of type SaveData and the title id field is 0, it should
+    // be interpreted as the title id of the current process.
+    if (type == SaveDataType::SaveData && title_id == 0)
+        title_id = Core::CurrentProcess()->GetTitleID();
+
+    std::string out = GetSaveDataSpaceIdPath(space);
 
     switch (type) {
     case SaveDataType::SystemSaveData:
diff --git a/src/core/file_sys/savedata_factory.h b/src/core/file_sys/savedata_factory.h
index d69ef6741..2a0088040 100644
--- a/src/core/file_sys/savedata_factory.h
+++ b/src/core/file_sys/savedata_factory.h
@@ -52,6 +52,9 @@ public:
 
     ResultVal<VirtualDir> Open(SaveDataSpaceId space, SaveDataDescriptor meta);
 
+    VirtualDir GetSaveDataSpaceDirectory(SaveDataSpaceId space) const;
+
+    static std::string GetSaveDataSpaceIdPath(SaveDataSpaceId space);
     static std::string GetFullPath(SaveDataSpaceId space, SaveDataType type, u64 title_id,
                                    u128 user_id, u64 save_id);
 
diff --git a/src/core/file_sys/submission_package.cpp b/src/core/file_sys/submission_package.cpp
index ab5dc900c..e1a4210db 100644
--- a/src/core/file_sys/submission_package.cpp
+++ b/src/core/file_sys/submission_package.cpp
@@ -205,10 +205,6 @@ VirtualDir NSP::GetParentDirectory() const {
     return file->GetContainingDirectory();
 }
 
-bool NSP::ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) {
-    return false;
-}
-
 void NSP::InitializeExeFSAndRomFS(const std::vector<VirtualFile>& files) {
     exefs = pfs;
 
@@ -256,7 +252,7 @@ void NSP::ReadNCAs(const std::vector<VirtualFile>& files) {
                     continue;
                 }
 
-                auto next_nca = std::make_shared<NCA>(next_file);
+                auto next_nca = std::make_shared<NCA>(next_file, nullptr, 0, keys);
                 if (next_nca->GetType() == NCAContentType::Program)
                     program_status[cnmt.GetTitleID()] = next_nca->GetStatus();
                 if (next_nca->GetStatus() == Loader::ResultStatus::Success ||
diff --git a/src/core/file_sys/submission_package.h b/src/core/file_sys/submission_package.h
index da3dc5e9f..9a28ed5bb 100644
--- a/src/core/file_sys/submission_package.h
+++ b/src/core/file_sys/submission_package.h
@@ -55,9 +55,6 @@ public:
 
     VirtualDir GetParentDirectory() const override;
 
-protected:
-    bool ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) override;
-
 private:
     void InitializeExeFSAndRomFS(const std::vector<VirtualFile>& files);
     void ReadNCAs(const std::vector<VirtualFile>& files);
@@ -73,6 +70,8 @@ private:
     std::map<u64, std::map<ContentRecordType, std::shared_ptr<NCA>>> ncas;
     std::vector<VirtualFile> ticket_files;
 
+    Core::Crypto::KeyManager keys;
+
     VirtualFile romfs;
     VirtualDir exefs;
 };
diff --git a/src/core/file_sys/vfs.cpp b/src/core/file_sys/vfs.cpp
index bfe50da73..7b584de7f 100644
--- a/src/core/file_sys/vfs.cpp
+++ b/src/core/file_sys/vfs.cpp
@@ -167,13 +167,13 @@ std::string VfsFile::GetExtension() const {
 
 VfsDirectory::~VfsDirectory() = default;
 
-boost::optional<u8> VfsFile::ReadByte(std::size_t offset) const {
+std::optional<u8> VfsFile::ReadByte(std::size_t offset) const {
     u8 out{};
     std::size_t size = Read(&out, 1, offset);
     if (size == 1)
         return out;
 
-    return boost::none;
+    return {};
 }
 
 std::vector<u8> VfsFile::ReadBytes(std::size_t size, std::size_t offset) const {
@@ -472,10 +472,14 @@ bool VfsRawCopy(const VirtualFile& src, const VirtualFile& dest, std::size_t blo
     std::vector<u8> temp(std::min(block_size, src->GetSize()));
     for (std::size_t i = 0; i < src->GetSize(); i += block_size) {
         const auto read = std::min(block_size, src->GetSize() - i);
-        const auto block = src->Read(temp.data(), read, i);
 
-        if (dest->Write(temp.data(), read, i) != read)
+        if (src->Read(temp.data(), read, i) != read) {
             return false;
+        }
+
+        if (dest->Write(temp.data(), read, i) != read) {
+            return false;
+        }
     }
 
     return true;
diff --git a/src/core/file_sys/vfs.h b/src/core/file_sys/vfs.h
index 7f0d520ca..002f99d4e 100644
--- a/src/core/file_sys/vfs.h
+++ b/src/core/file_sys/vfs.h
@@ -4,13 +4,15 @@
 
 #pragma once
 
+#include <functional>
 #include <map>
 #include <memory>
+#include <optional>
 #include <string>
 #include <string_view>
 #include <type_traits>
 #include <vector>
-#include <boost/optional.hpp>
+
 #include "common/common_types.h"
 #include "core/file_sys/vfs_types.h"
 
@@ -103,8 +105,8 @@ public:
     // into file. Returns number of bytes successfully written.
     virtual std::size_t Write(const u8* data, std::size_t length, std::size_t offset = 0) = 0;
 
-    // Reads exactly one byte at the offset provided, returning boost::none on error.
-    virtual boost::optional<u8> ReadByte(std::size_t offset = 0) const;
+    // Reads exactly one byte at the offset provided, returning std::nullopt on error.
+    virtual std::optional<u8> ReadByte(std::size_t offset = 0) const;
     // Reads size bytes starting at offset in file into a vector.
     virtual std::vector<u8> ReadBytes(std::size_t size, std::size_t offset = 0) const;
     // Reads all the bytes from the file into a vector. Equivalent to 'file->Read(file->GetSize(),
@@ -262,36 +264,8 @@ public:
     // item name -> type.
     virtual std::map<std::string, VfsEntryType, std::less<>> GetEntries() const;
 
-    // Interprets the file with name file instead as a directory of type directory.
-    // The directory must have a constructor that takes a single argument of type
-    // std::shared_ptr<VfsFile>. Allows to reinterpret container files (i.e NCA, zip, XCI, etc) as a
-    // subdirectory in one call.
-    template <typename Directory>
-    bool InterpretAsDirectory(std::string_view file) {
-        auto file_p = GetFile(file);
-
-        if (file_p == nullptr) {
-            return false;
-        }
-
-        return ReplaceFileWithSubdirectory(file_p, std::make_shared<Directory>(file_p));
-    }
-
-    bool InterpretAsDirectory(const std::function<VirtualDir(VirtualFile)>& function,
-                              const std::string& file) {
-        auto file_p = GetFile(file);
-        if (file_p == nullptr)
-            return false;
-        return ReplaceFileWithSubdirectory(file_p, function(file_p));
-    }
-
     // Returns the full path of this directory as a string, recursively
     virtual std::string GetFullPath() const;
-
-protected:
-    // Backend for InterpretAsDirectory.
-    // Removes all references to file and adds a reference to dir in the directory's implementation.
-    virtual bool ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) = 0;
 };
 
 // A convenience partial-implementation of VfsDirectory that stubs out methods that should only work
diff --git a/src/core/file_sys/vfs_layered.cpp b/src/core/file_sys/vfs_layered.cpp
index bfee01725..338e398da 100644
--- a/src/core/file_sys/vfs_layered.cpp
+++ b/src/core/file_sys/vfs_layered.cpp
@@ -126,7 +126,4 @@ bool LayeredVfsDirectory::Rename(std::string_view name_) {
     return true;
 }
 
-bool LayeredVfsDirectory::ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) {
-    return false;
-}
 } // namespace FileSys
diff --git a/src/core/file_sys/vfs_layered.h b/src/core/file_sys/vfs_layered.h
index d85310f57..8a25c3428 100644
--- a/src/core/file_sys/vfs_layered.h
+++ b/src/core/file_sys/vfs_layered.h
@@ -39,9 +39,6 @@ public:
     bool DeleteFile(std::string_view name) override;
     bool Rename(std::string_view name) override;
 
-protected:
-    bool ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) override;
-
 private:
     std::vector<VirtualDir> dirs;
     std::string name;
diff --git a/src/core/file_sys/vfs_offset.cpp b/src/core/file_sys/vfs_offset.cpp
index a4c6719a0..c96f88488 100644
--- a/src/core/file_sys/vfs_offset.cpp
+++ b/src/core/file_sys/vfs_offset.cpp
@@ -57,11 +57,11 @@ std::size_t OffsetVfsFile::Write(const u8* data, std::size_t length, std::size_t
     return file->Write(data, TrimToFit(length, r_offset), offset + r_offset);
 }
 
-boost::optional<u8> OffsetVfsFile::ReadByte(std::size_t r_offset) const {
+std::optional<u8> OffsetVfsFile::ReadByte(std::size_t r_offset) const {
     if (r_offset < size)
         return file->ReadByte(offset + r_offset);
 
-    return boost::none;
+    return {};
 }
 
 std::vector<u8> OffsetVfsFile::ReadBytes(std::size_t r_size, std::size_t r_offset) const {
diff --git a/src/core/file_sys/vfs_offset.h b/src/core/file_sys/vfs_offset.h
index 8062702a7..f7b7a3256 100644
--- a/src/core/file_sys/vfs_offset.h
+++ b/src/core/file_sys/vfs_offset.h
@@ -29,7 +29,7 @@ public:
     bool IsReadable() const override;
     std::size_t Read(u8* data, std::size_t length, std::size_t offset) const override;
     std::size_t Write(const u8* data, std::size_t length, std::size_t offset) override;
-    boost::optional<u8> ReadByte(std::size_t offset) const override;
+    std::optional<u8> ReadByte(std::size_t offset) const override;
     std::vector<u8> ReadBytes(std::size_t size, std::size_t offset) const override;
     std::vector<u8> ReadAllBytes() const override;
     bool WriteByte(u8 data, std::size_t offset) override;
diff --git a/src/core/file_sys/vfs_real.cpp b/src/core/file_sys/vfs_real.cpp
index 9defad04c..e21300a7c 100644
--- a/src/core/file_sys/vfs_real.cpp
+++ b/src/core/file_sys/vfs_real.cpp
@@ -430,7 +430,4 @@ std::map<std::string, VfsEntryType, std::less<>> RealVfsDirectory::GetEntries()
     return out;
 }
 
-bool RealVfsDirectory::ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) {
-    return false;
-}
 } // namespace FileSys
diff --git a/src/core/file_sys/vfs_real.h b/src/core/file_sys/vfs_real.h
index 5b61db90d..a0a857a31 100644
--- a/src/core/file_sys/vfs_real.h
+++ b/src/core/file_sys/vfs_real.h
@@ -100,9 +100,6 @@ public:
     std::string GetFullPath() const override;
     std::map<std::string, VfsEntryType, std::less<>> GetEntries() const override;
 
-protected:
-    bool ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) override;
-
 private:
     RealVfsDirectory(RealVfsFilesystem& base, const std::string& path, Mode perms = Mode::Read);
 
diff --git a/src/core/file_sys/vfs_static.h b/src/core/file_sys/vfs_static.h
index 44fab51d1..9f5a90b1b 100644
--- a/src/core/file_sys/vfs_static.h
+++ b/src/core/file_sys/vfs_static.h
@@ -53,10 +53,10 @@ public:
         return 0;
     }
 
-    boost::optional<u8> ReadByte(std::size_t offset) const override {
+    std::optional<u8> ReadByte(std::size_t offset) const override {
         if (offset < size)
             return value;
-        return boost::none;
+        return {};
     }
 
     std::vector<u8> ReadBytes(std::size_t length, std::size_t offset) const override {
diff --git a/src/core/file_sys/vfs_vector.cpp b/src/core/file_sys/vfs_vector.cpp
index 389c7e003..808f31e81 100644
--- a/src/core/file_sys/vfs_vector.cpp
+++ b/src/core/file_sys/vfs_vector.cpp
@@ -132,11 +132,4 @@ void VectorVfsDirectory::AddFile(VirtualFile file) {
 void VectorVfsDirectory::AddDirectory(VirtualDir dir) {
     dirs.push_back(std::move(dir));
 }
-
-bool VectorVfsDirectory::ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) {
-    if (!DeleteFile(file->GetName()))
-        return false;
-    dirs.emplace_back(std::move(dir));
-    return true;
-}
 } // namespace FileSys
diff --git a/src/core/file_sys/vfs_vector.h b/src/core/file_sys/vfs_vector.h
index 48a414c98..3e3f790c3 100644
--- a/src/core/file_sys/vfs_vector.h
+++ b/src/core/file_sys/vfs_vector.h
@@ -57,9 +57,6 @@ public:
     virtual void AddFile(VirtualFile file);
     virtual void AddDirectory(VirtualDir dir);
 
-protected:
-    bool ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) override;
-
 private:
     std::vector<VirtualFile> files;
     std::vector<VirtualDir> dirs;
diff --git a/src/core/file_sys/xts_archive.cpp b/src/core/file_sys/xts_archive.cpp
index b2b164368..eec51c64e 100644
--- a/src/core/file_sys/xts_archive.cpp
+++ b/src/core/file_sys/xts_archive.cpp
@@ -163,7 +163,4 @@ std::shared_ptr<VfsDirectory> NAX::GetParentDirectory() const {
     return file->GetContainingDirectory();
 }
 
-bool NAX::ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) {
-    return false;
-}
 } // namespace FileSys
diff --git a/src/core/file_sys/xts_archive.h b/src/core/file_sys/xts_archive.h
index 8fedd8585..7704dee90 100644
--- a/src/core/file_sys/xts_archive.h
+++ b/src/core/file_sys/xts_archive.h
@@ -51,9 +51,6 @@ public:
 
     std::shared_ptr<VfsDirectory> GetParentDirectory() const override;
 
-protected:
-    bool ReplaceFileWithSubdirectory(VirtualFile file, VirtualDir dir) override;
-
 private:
     Loader::ResultStatus Parse(std::string_view path);
 
diff --git a/src/core/hle/ipc.h b/src/core/hle/ipc.h
index 419f45896..ed84197b3 100644
--- a/src/core/hle/ipc.h
+++ b/src/core/hle/ipc.h
@@ -14,11 +14,6 @@ namespace IPC {
 /// Size of the command buffer area, in 32-bit words.
 constexpr std::size_t COMMAND_BUFFER_LENGTH = 0x100 / sizeof(u32);
 
-// These errors are commonly returned by invalid IPC translations, so alias them here for
-// convenience.
-// TODO(yuriks): These will probably go away once translation is implemented inside the kernel.
-constexpr auto ERR_INVALID_HANDLE = Kernel::ERR_INVALID_HANDLE_OS;
-
 enum class ControlCommand : u32 {
     ConvertSessionToDomain = 0,
     ConvertDomainToSession = 1,
diff --git a/src/core/hle/ipc_helpers.h b/src/core/hle/ipc_helpers.h
index a4bfe2eb0..0a7142ada 100644
--- a/src/core/hle/ipc_helpers.h
+++ b/src/core/hle/ipc_helpers.h
@@ -117,8 +117,7 @@ public:
 
         AlignWithPadding();
 
-        const bool request_has_domain_header{context.GetDomainMessageHeader() != nullptr};
-        if (context.Session()->IsDomain() && request_has_domain_header) {
+        if (context.Session()->IsDomain() && context.HasDomainMessageHeader()) {
             IPC::DomainMessageHeader domain_header{};
             domain_header.num_objects = num_domain_objects;
             PushRaw(domain_header);
diff --git a/src/core/hle/kernel/errors.h b/src/core/hle/kernel/errors.h
index 885259618..8b58d701d 100644
--- a/src/core/hle/kernel/errors.h
+++ b/src/core/hle/kernel/errors.h
@@ -8,71 +8,28 @@
 
 namespace Kernel {
 
-namespace ErrCodes {
-enum {
-    // TODO(Subv): Remove these 3DS OS error codes.
-    SessionClosedByRemote = 26,
-    NoPendingSessions = 35,
-    InvalidBufferDescriptor = 48,
-
-    // Confirmed Switch OS error codes
-    MaxConnectionsReached = 7,
-    InvalidSize = 101,
-    InvalidAddress = 102,
-    HandleTableFull = 105,
-    InvalidMemoryState = 106,
-    InvalidMemoryPermissions = 108,
-    InvalidMemoryRange = 110,
-    InvalidThreadPriority = 112,
-    InvalidProcessorId = 113,
-    InvalidHandle = 114,
-    InvalidCombination = 116,
-    Timeout = 117,
-    SynchronizationCanceled = 118,
-    TooLarge = 119,
-    InvalidEnumValue = 120,
-    NoSuchEntry = 121,
-    AlreadyRegistered = 122,
-    InvalidState = 125,
-    ResourceLimitExceeded = 132,
-};
-}
-
-// WARNING: The kernel is quite inconsistent in it's usage of errors code. Make sure to always
-// double check that the code matches before re-using the constant.
-
-// TODO(bunnei): Replace -1 with correct errors for Switch OS
-constexpr ResultCode ERR_HANDLE_TABLE_FULL(ErrorModule::Kernel, ErrCodes::HandleTableFull);
-constexpr ResultCode ERR_SESSION_CLOSED_BY_REMOTE(-1);
-constexpr ResultCode ERR_PORT_NAME_TOO_LONG(ErrorModule::Kernel, ErrCodes::TooLarge);
-constexpr ResultCode ERR_MAX_CONNECTIONS_REACHED(ErrorModule::Kernel,
-                                                 ErrCodes::MaxConnectionsReached);
-constexpr ResultCode ERR_INVALID_ENUM_VALUE(ErrorModule::Kernel, ErrCodes::InvalidEnumValue);
-constexpr ResultCode ERR_INVALID_ENUM_VALUE_FND(-1);
-constexpr ResultCode ERR_INVALID_COMBINATION(-1);
-constexpr ResultCode ERR_INVALID_COMBINATION_KERNEL(ErrorModule::Kernel,
-                                                    ErrCodes::InvalidCombination);
-constexpr ResultCode ERR_OUT_OF_MEMORY(-1);
-constexpr ResultCode ERR_INVALID_ADDRESS(ErrorModule::Kernel, ErrCodes::InvalidAddress);
-constexpr ResultCode ERR_INVALID_ADDRESS_STATE(ErrorModule::Kernel, ErrCodes::InvalidMemoryState);
-constexpr ResultCode ERR_INVALID_MEMORY_PERMISSIONS(ErrorModule::Kernel,
-                                                    ErrCodes::InvalidMemoryPermissions);
-constexpr ResultCode ERR_INVALID_MEMORY_RANGE(ErrorModule::Kernel, ErrCodes::InvalidMemoryRange);
-constexpr ResultCode ERR_INVALID_HANDLE(ErrorModule::Kernel, ErrCodes::InvalidHandle);
-constexpr ResultCode ERR_INVALID_PROCESSOR_ID(ErrorModule::Kernel, ErrCodes::InvalidProcessorId);
-constexpr ResultCode ERR_INVALID_SIZE(ErrorModule::Kernel, ErrCodes::InvalidSize);
-constexpr ResultCode ERR_ALREADY_REGISTERED(ErrorModule::Kernel, ErrCodes::AlreadyRegistered);
-constexpr ResultCode ERR_INVALID_STATE(ErrorModule::Kernel, ErrCodes::InvalidState);
-constexpr ResultCode ERR_INVALID_THREAD_PRIORITY(ErrorModule::Kernel,
-                                                 ErrCodes::InvalidThreadPriority);
-constexpr ResultCode ERR_INVALID_POINTER(-1);
-constexpr ResultCode ERR_INVALID_OBJECT_ADDR(-1);
-constexpr ResultCode ERR_NOT_AUTHORIZED(-1);
-/// Alternate code returned instead of ERR_INVALID_HANDLE in some code paths.
-constexpr ResultCode ERR_INVALID_HANDLE_OS(-1);
-constexpr ResultCode ERR_NOT_FOUND(ErrorModule::Kernel, ErrCodes::NoSuchEntry);
-constexpr ResultCode RESULT_TIMEOUT(ErrorModule::Kernel, ErrCodes::Timeout);
-/// Returned when Accept() is called on a port with no sessions to be accepted.
-constexpr ResultCode ERR_NO_PENDING_SESSIONS(-1);
+// Confirmed Switch kernel error codes
+
+constexpr ResultCode ERR_MAX_CONNECTIONS_REACHED{ErrorModule::Kernel, 7};
+constexpr ResultCode ERR_INVALID_SIZE{ErrorModule::Kernel, 101};
+constexpr ResultCode ERR_INVALID_ADDRESS{ErrorModule::Kernel, 102};
+constexpr ResultCode ERR_HANDLE_TABLE_FULL{ErrorModule::Kernel, 105};
+constexpr ResultCode ERR_INVALID_ADDRESS_STATE{ErrorModule::Kernel, 106};
+constexpr ResultCode ERR_INVALID_MEMORY_PERMISSIONS{ErrorModule::Kernel, 108};
+constexpr ResultCode ERR_INVALID_MEMORY_RANGE{ErrorModule::Kernel, 110};
+constexpr ResultCode ERR_INVALID_PROCESSOR_ID{ErrorModule::Kernel, 113};
+constexpr ResultCode ERR_INVALID_THREAD_PRIORITY{ErrorModule::Kernel, 112};
+constexpr ResultCode ERR_INVALID_HANDLE{ErrorModule::Kernel, 114};
+constexpr ResultCode ERR_INVALID_POINTER{ErrorModule::Kernel, 115};
+constexpr ResultCode ERR_INVALID_COMBINATION{ErrorModule::Kernel, 116};
+constexpr ResultCode RESULT_TIMEOUT{ErrorModule::Kernel, 117};
+constexpr ResultCode ERR_SYNCHRONIZATION_CANCELED{ErrorModule::Kernel, 118};
+constexpr ResultCode ERR_OUT_OF_RANGE{ErrorModule::Kernel, 119};
+constexpr ResultCode ERR_INVALID_ENUM_VALUE{ErrorModule::Kernel, 120};
+constexpr ResultCode ERR_NOT_FOUND{ErrorModule::Kernel, 121};
+constexpr ResultCode ERR_ALREADY_REGISTERED{ErrorModule::Kernel, 122};
+constexpr ResultCode ERR_SESSION_CLOSED_BY_REMOTE{ErrorModule::Kernel, 123};
+constexpr ResultCode ERR_INVALID_STATE{ErrorModule::Kernel, 125};
+constexpr ResultCode ERR_RESOURCE_LIMIT_EXCEEDED{ErrorModule::Kernel, 132};
 
 } // namespace Kernel
diff --git a/src/core/hle/kernel/hle_ipc.h b/src/core/hle/kernel/hle_ipc.h
index f01491daa..a38e34b74 100644
--- a/src/core/hle/kernel/hle_ipc.h
+++ b/src/core/hle/kernel/hle_ipc.h
@@ -161,8 +161,12 @@ public:
         return buffer_c_desciptors;
     }
 
-    const std::shared_ptr<IPC::DomainMessageHeader>& GetDomainMessageHeader() const {
-        return domain_message_header;
+    const IPC::DomainMessageHeader* GetDomainMessageHeader() const {
+        return domain_message_header.get();
+    }
+
+    bool HasDomainMessageHeader() const {
+        return domain_message_header != nullptr;
     }
 
     /// Helper function to read a buffer using the appropriate buffer descriptor
diff --git a/src/core/hle/kernel/kernel.cpp b/src/core/hle/kernel/kernel.cpp
index 4b6b32dd5..1fd4ba5d2 100644
--- a/src/core/hle/kernel/kernel.cpp
+++ b/src/core/hle/kernel/kernel.cpp
@@ -32,7 +32,7 @@ namespace Kernel {
  */
 static void ThreadWakeupCallback(u64 thread_handle, [[maybe_unused]] int cycles_late) {
     const auto proper_handle = static_cast<Handle>(thread_handle);
-    auto& system = Core::System::GetInstance();
+    const auto& system = Core::System::GetInstance();
 
     // Lock the global kernel mutex when we enter the kernel HLE.
     std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
@@ -90,7 +90,7 @@ static void ThreadWakeupCallback(u64 thread_handle, [[maybe_unused]] int cycles_
 /// The timer callback event, called when a timer is fired
 static void TimerCallback(u64 timer_handle, int cycles_late) {
     const auto proper_handle = static_cast<Handle>(timer_handle);
-    auto& system = Core::System::GetInstance();
+    const auto& system = Core::System::GetInstance();
     SharedPtr<Timer> timer = system.Kernel().RetrieveTimerFromCallbackHandleTable(proper_handle);
 
     if (timer == nullptr) {
diff --git a/src/core/hle/kernel/mutex.cpp b/src/core/hle/kernel/mutex.cpp
index dd541ffcc..0743670ad 100644
--- a/src/core/hle/kernel/mutex.cpp
+++ b/src/core/hle/kernel/mutex.cpp
@@ -6,8 +6,6 @@
 #include <utility>
 #include <vector>
 
-#include <boost/range/algorithm_ext/erase.hpp>
-
 #include "common/assert.h"
 #include "core/core.h"
 #include "core/hle/kernel/errors.h"
diff --git a/src/core/hle/kernel/process.cpp b/src/core/hle/kernel/process.cpp
index 073dd5a7d..a257c3726 100644
--- a/src/core/hle/kernel/process.cpp
+++ b/src/core/hle/kernel/process.cpp
@@ -5,11 +5,9 @@
 #include <algorithm>
 #include <memory>
 #include "common/assert.h"
-#include "common/common_funcs.h"
 #include "common/logging/log.h"
 #include "core/core.h"
 #include "core/file_sys/program_metadata.h"
-#include "core/hle/kernel/errors.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/kernel/resource_limit.h"
@@ -17,6 +15,7 @@
 #include "core/hle/kernel/thread.h"
 #include "core/hle/kernel/vm_manager.h"
 #include "core/memory.h"
+#include "core/settings.h"
 
 namespace Kernel {
 
@@ -35,6 +34,11 @@ SharedPtr<Process> Process::Create(KernelCore& kernel, std::string&& name) {
     process->process_id = kernel.CreateNewProcessID();
     process->svc_access_mask.set();
 
+    std::mt19937 rng(Settings::values.rng_seed.value_or(0));
+    std::uniform_int_distribution<u64> distribution;
+    std::generate(process->random_entropy.begin(), process->random_entropy.end(),
+                  [&] { return distribution(rng); });
+
     kernel.AppendNewProcess(process);
     return process;
 }
@@ -232,86 +236,24 @@ void Process::LoadModule(CodeSet module_, VAddr base_addr) {
     MapSegment(module_.CodeSegment(), VMAPermission::ReadExecute, MemoryState::CodeStatic);
     MapSegment(module_.RODataSegment(), VMAPermission::Read, MemoryState::CodeMutable);
     MapSegment(module_.DataSegment(), VMAPermission::ReadWrite, MemoryState::CodeMutable);
+
+    // Clear instruction cache in CPU JIT
+    Core::System::GetInstance().ArmInterface(0).ClearInstructionCache();
+    Core::System::GetInstance().ArmInterface(1).ClearInstructionCache();
+    Core::System::GetInstance().ArmInterface(2).ClearInstructionCache();
+    Core::System::GetInstance().ArmInterface(3).ClearInstructionCache();
 }
 
 ResultVal<VAddr> Process::HeapAllocate(VAddr target, u64 size, VMAPermission perms) {
-    if (target < vm_manager.GetHeapRegionBaseAddress() ||
-        target + size > vm_manager.GetHeapRegionEndAddress() || target + size < target) {
-        return ERR_INVALID_ADDRESS;
-    }
-
-    if (heap_memory == nullptr) {
-        // Initialize heap
-        heap_memory = std::make_shared<std::vector<u8>>();
-        heap_start = heap_end = target;
-    } else {
-        vm_manager.UnmapRange(heap_start, heap_end - heap_start);
-    }
-
-    // If necessary, expand backing vector to cover new heap extents.
-    if (target < heap_start) {
-        heap_memory->insert(begin(*heap_memory), heap_start - target, 0);
-        heap_start = target;
-        vm_manager.RefreshMemoryBlockMappings(heap_memory.get());
-    }
-    if (target + size > heap_end) {
-        heap_memory->insert(end(*heap_memory), (target + size) - heap_end, 0);
-        heap_end = target + size;
-        vm_manager.RefreshMemoryBlockMappings(heap_memory.get());
-    }
-    ASSERT(heap_end - heap_start == heap_memory->size());
-
-    CASCADE_RESULT(auto vma, vm_manager.MapMemoryBlock(target, heap_memory, target - heap_start,
-                                                       size, MemoryState::Heap));
-    vm_manager.Reprotect(vma, perms);
-
-    heap_used = size;
-
-    return MakeResult<VAddr>(heap_end - size);
+    return vm_manager.HeapAllocate(target, size, perms);
 }
 
 ResultCode Process::HeapFree(VAddr target, u32 size) {
-    if (target < vm_manager.GetHeapRegionBaseAddress() ||
-        target + size > vm_manager.GetHeapRegionEndAddress() || target + size < target) {
-        return ERR_INVALID_ADDRESS;
-    }
-
-    if (size == 0) {
-        return RESULT_SUCCESS;
-    }
-
-    ResultCode result = vm_manager.UnmapRange(target, size);
-    if (result.IsError())
-        return result;
-
-    heap_used -= size;
-
-    return RESULT_SUCCESS;
+    return vm_manager.HeapFree(target, size);
 }
 
-ResultCode Process::MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size) {
-    auto vma = vm_manager.FindVMA(src_addr);
-
-    ASSERT_MSG(vma != vm_manager.vma_map.end(), "Invalid memory address");
-    ASSERT_MSG(vma->second.backing_block, "Backing block doesn't exist for address");
-
-    // The returned VMA might be a bigger one encompassing the desired address.
-    auto vma_offset = src_addr - vma->first;
-    ASSERT_MSG(vma_offset + size <= vma->second.size,
-               "Shared memory exceeds bounds of mapped block");
-
-    const std::shared_ptr<std::vector<u8>>& backing_block = vma->second.backing_block;
-    std::size_t backing_block_offset = vma->second.offset + vma_offset;
-
-    CASCADE_RESULT(auto new_vma,
-                   vm_manager.MapMemoryBlock(dst_addr, backing_block, backing_block_offset, size,
-                                             MemoryState::Mapped));
-    // Protect mirror with permissions from old region
-    vm_manager.Reprotect(new_vma, vma->second.permissions);
-    // Remove permissions from old region
-    vm_manager.Reprotect(vma, VMAPermission::None);
-
-    return RESULT_SUCCESS;
+ResultCode Process::MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size, MemoryState state) {
+    return vm_manager.MirrorMemory(dst_addr, src_addr, size, state);
 }
 
 ResultCode Process::UnmapMemory(VAddr dst_addr, VAddr /*src_addr*/, u64 size) {
diff --git a/src/core/hle/kernel/process.h b/src/core/hle/kernel/process.h
index 148478488..230e395ff 100644
--- a/src/core/hle/kernel/process.h
+++ b/src/core/hle/kernel/process.h
@@ -8,6 +8,7 @@
 #include <bitset>
 #include <cstddef>
 #include <memory>
+#include <random>
 #include <string>
 #include <vector>
 #include <boost/container/static_vector.hpp>
@@ -119,6 +120,8 @@ struct CodeSet final {
 
 class Process final : public Object {
 public:
+    static constexpr std::size_t RANDOM_ENTROPY_SIZE = 4;
+
     static SharedPtr<Process> Create(KernelCore& kernel, std::string&& name);
 
     std::string GetTypeName() const override {
@@ -202,6 +205,21 @@ public:
         return is_64bit_process;
     }
 
+    /// Gets the total running time of the process instance in ticks.
+    u64 GetCPUTimeTicks() const {
+        return total_process_running_time_ticks;
+    }
+
+    /// Updates the total running time, adding the given ticks to it.
+    void UpdateCPUTimeTicks(u64 ticks) {
+        total_process_running_time_ticks += ticks;
+    }
+
+    /// Gets 8 bytes of random data for svcGetInfo RandomEntropy
+    u64 GetRandomEntropy(std::size_t index) const {
+        return random_entropy.at(index);
+    }
+
     /**
      * Loads process-specifics configuration info with metadata provided
      * by an executable.
@@ -241,7 +259,8 @@ public:
     ResultVal<VAddr> HeapAllocate(VAddr target, u64 size, VMAPermission perms);
     ResultCode HeapFree(VAddr target, u32 size);
 
-    ResultCode MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size);
+    ResultCode MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size,
+                            MemoryState state = MemoryState::Mapped);
 
     ResultCode UnmapMemory(VAddr dst_addr, VAddr src_addr, u64 size);
 
@@ -282,17 +301,6 @@ private:
     u32 allowed_thread_priority_mask = 0xFFFFFFFF;
     u32 is_virtual_address_memory_enabled = 0;
 
-    // Memory used to back the allocations in the regular heap. A single vector is used to cover
-    // the entire virtual address space extents that bound the allocations, including any holes.
-    // This makes deallocation and reallocation of holes fast and keeps process memory contiguous
-    // in the emulator address space, allowing Memory::GetPointer to be reasonably safe.
-    std::shared_ptr<std::vector<u8>> heap_memory;
-
-    // The left/right bounds of the address space covered by heap_memory.
-    VAddr heap_start = 0;
-    VAddr heap_end = 0;
-    u64 heap_used = 0;
-
     /// The Thread Local Storage area is allocated as processes create threads,
     /// each TLS area is 0x200 bytes, so one page (0x1000) is split up in 8 parts, and each part
     /// holds the TLS for a specific thread. This vector contains which parts are in use for each
@@ -305,9 +313,15 @@ private:
     /// specified by metadata provided to the process during loading.
     bool is_64bit_process = true;
 
+    /// Total running time for the process in ticks.
+    u64 total_process_running_time_ticks = 0;
+
     /// Per-process handle table for storing created object handles in.
     HandleTable handle_table;
 
+    /// Random values for svcGetInfo RandomEntropy
+    std::array<u64, RANDOM_ENTROPY_SIZE> random_entropy;
+
     std::string name;
 };
 
diff --git a/src/core/hle/kernel/scheduler.cpp b/src/core/hle/kernel/scheduler.cpp
index 1342c597e..5a5f4cef1 100644
--- a/src/core/hle/kernel/scheduler.cpp
+++ b/src/core/hle/kernel/scheduler.cpp
@@ -9,6 +9,7 @@
 #include "common/logging/log.h"
 #include "core/arm/arm_interface.h"
 #include "core/core.h"
+#include "core/core_timing.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/kernel/scheduler.h"
@@ -34,6 +35,10 @@ Thread* Scheduler::GetCurrentThread() const {
     return current_thread.get();
 }
 
+u64 Scheduler::GetLastContextSwitchTicks() const {
+    return last_context_switch_time;
+}
+
 Thread* Scheduler::PopNextReadyThread() {
     Thread* next = nullptr;
     Thread* thread = GetCurrentThread();
@@ -54,7 +59,10 @@ Thread* Scheduler::PopNextReadyThread() {
 }
 
 void Scheduler::SwitchContext(Thread* new_thread) {
-    Thread* previous_thread = GetCurrentThread();
+    Thread* const previous_thread = GetCurrentThread();
+    Process* const previous_process = Core::CurrentProcess();
+
+    UpdateLastContextSwitchTime(previous_thread, previous_process);
 
     // Save context for previous thread
     if (previous_thread) {
@@ -78,8 +86,6 @@ void Scheduler::SwitchContext(Thread* new_thread) {
         // Cancel any outstanding wakeup events for this thread
         new_thread->CancelWakeupTimer();
 
-        auto* const previous_process = Core::CurrentProcess();
-
         current_thread = new_thread;
 
         ready_queue.remove(new_thread->GetPriority(), new_thread);
@@ -102,6 +108,22 @@ void Scheduler::SwitchContext(Thread* new_thread) {
     }
 }
 
+void Scheduler::UpdateLastContextSwitchTime(Thread* thread, Process* process) {
+    const u64 prev_switch_ticks = last_context_switch_time;
+    const u64 most_recent_switch_ticks = CoreTiming::GetTicks();
+    const u64 update_ticks = most_recent_switch_ticks - prev_switch_ticks;
+
+    if (thread != nullptr) {
+        thread->UpdateCPUTimeTicks(update_ticks);
+    }
+
+    if (process != nullptr) {
+        process->UpdateCPUTimeTicks(update_ticks);
+    }
+
+    last_context_switch_time = most_recent_switch_ticks;
+}
+
 void Scheduler::Reschedule() {
     std::lock_guard<std::mutex> lock(scheduler_mutex);
 
diff --git a/src/core/hle/kernel/scheduler.h b/src/core/hle/kernel/scheduler.h
index 2c94641ec..c63032b7d 100644
--- a/src/core/hle/kernel/scheduler.h
+++ b/src/core/hle/kernel/scheduler.h
@@ -17,6 +17,8 @@ class ARM_Interface;
 
 namespace Kernel {
 
+class Process;
+
 class Scheduler final {
 public:
     explicit Scheduler(Core::ARM_Interface& cpu_core);
@@ -31,6 +33,9 @@ public:
     /// Gets the current running thread
     Thread* GetCurrentThread() const;
 
+    /// Gets the timestamp for the last context switch in ticks.
+    u64 GetLastContextSwitchTicks() const;
+
     /// Adds a new thread to the scheduler
     void AddThread(SharedPtr<Thread> thread, u32 priority);
 
@@ -64,6 +69,19 @@ private:
      */
     void SwitchContext(Thread* new_thread);
 
+    /**
+     * Called on every context switch to update the internal timestamp
+     * This also updates the running time ticks for the given thread and
+     * process using the following difference:
+     *
+     * ticks += most_recent_ticks - last_context_switch_ticks
+     *
+     * The internal tick timestamp for the scheduler is simply the
+     * most recent tick count retrieved. No special arithmetic is
+     * applied to it.
+     */
+    void UpdateLastContextSwitchTime(Thread* thread, Process* process);
+
     /// Lists all thread ids that aren't deleted/etc.
     std::vector<SharedPtr<Thread>> thread_list;
 
@@ -73,6 +91,7 @@ private:
     SharedPtr<Thread> current_thread = nullptr;
 
     Core::ARM_Interface& cpu_core;
+    u64 last_context_switch_time = 0;
 
     static std::mutex scheduler_mutex;
 };
diff --git a/src/core/hle/kernel/server_port.cpp b/src/core/hle/kernel/server_port.cpp
index 3792e3e18..d6ceeb2da 100644
--- a/src/core/hle/kernel/server_port.cpp
+++ b/src/core/hle/kernel/server_port.cpp
@@ -18,7 +18,7 @@ ServerPort::~ServerPort() = default;
 
 ResultVal<SharedPtr<ServerSession>> ServerPort::Accept() {
     if (pending_sessions.empty()) {
-        return ERR_NO_PENDING_SESSIONS;
+        return ERR_NOT_FOUND;
     }
 
     auto session = std::move(pending_sessions.back());
@@ -28,7 +28,7 @@ ResultVal<SharedPtr<ServerSession>> ServerPort::Accept() {
 
 bool ServerPort::ShouldWait(Thread* thread) const {
     // If there are no pending sessions, we wait until a new one is added.
-    return pending_sessions.size() == 0;
+    return pending_sessions.empty();
 }
 
 void ServerPort::Acquire(Thread* thread) {
diff --git a/src/core/hle/kernel/server_session.cpp b/src/core/hle/kernel/server_session.cpp
index 5fc320403..80897f3a4 100644
--- a/src/core/hle/kernel/server_session.cpp
+++ b/src/core/hle/kernel/server_session.cpp
@@ -63,7 +63,7 @@ void ServerSession::Acquire(Thread* thread) {
 }
 
 ResultCode ServerSession::HandleDomainSyncRequest(Kernel::HLERequestContext& context) {
-    auto& domain_message_header = context.GetDomainMessageHeader();
+    auto* const domain_message_header = context.GetDomainMessageHeader();
     if (domain_message_header) {
         // Set domain handlers in HLE context, used for domain objects (IPC interfaces) as inputs
         context.SetDomainRequestHandlers(domain_request_handlers);
@@ -111,7 +111,7 @@ ResultCode ServerSession::HandleSyncRequest(SharedPtr<Thread> thread) {
 
     ResultCode result = RESULT_SUCCESS;
     // If the session has been converted to a domain, handle the domain request
-    if (IsDomain() && context.GetDomainMessageHeader()) {
+    if (IsDomain() && context.HasDomainMessageHeader()) {
         result = HandleDomainSyncRequest(context);
         // If there is no domain header, the regular session handler is used
     } else if (hle_handler != nullptr) {
diff --git a/src/core/hle/kernel/shared_memory.cpp b/src/core/hle/kernel/shared_memory.cpp
index d061e6155..a016a86b6 100644
--- a/src/core/hle/kernel/shared_memory.cpp
+++ b/src/core/hle/kernel/shared_memory.cpp
@@ -80,20 +80,19 @@ SharedPtr<SharedMemory> SharedMemory::CreateForApplet(
 
 ResultCode SharedMemory::Map(Process* target_process, VAddr address, MemoryPermission permissions,
                              MemoryPermission other_permissions) {
-
-    MemoryPermission own_other_permissions =
+    const MemoryPermission own_other_permissions =
         target_process == owner_process ? this->permissions : this->other_permissions;
 
     // Automatically allocated memory blocks can only be mapped with other_permissions = DontCare
     if (base_address == 0 && other_permissions != MemoryPermission::DontCare) {
-        return ERR_INVALID_COMBINATION;
+        return ERR_INVALID_MEMORY_PERMISSIONS;
     }
 
     // Error out if the requested permissions don't match what the creator process allows.
     if (static_cast<u32>(permissions) & ~static_cast<u32>(own_other_permissions)) {
         LOG_ERROR(Kernel, "cannot map id={}, address=0x{:X} name={}, permissions don't match",
                   GetObjectId(), address, name);
-        return ERR_INVALID_COMBINATION;
+        return ERR_INVALID_MEMORY_PERMISSIONS;
     }
 
     // Error out if the provided permissions are not compatible with what the creator process needs.
diff --git a/src/core/hle/kernel/svc.cpp b/src/core/hle/kernel/svc.cpp
index 67ea67666..75dbfc31d 100644
--- a/src/core/hle/kernel/svc.cpp
+++ b/src/core/hle/kernel/svc.cpp
@@ -34,6 +34,7 @@
 #include "core/hle/lock.h"
 #include "core/hle/result.h"
 #include "core/hle/service/service.h"
+#include "core/settings.h"
 
 namespace Kernel {
 namespace {
@@ -122,6 +123,48 @@ static ResultCode SetHeapSize(VAddr* heap_addr, u64 heap_size) {
     return RESULT_SUCCESS;
 }
 
+static ResultCode SetMemoryPermission(VAddr addr, u64 size, u32 prot) {
+    LOG_TRACE(Kernel_SVC, "called, addr=0x{:X}, size=0x{:X}, prot=0x{:X}", addr, size, prot);
+
+    if (!Common::Is4KBAligned(addr)) {
+        return ERR_INVALID_ADDRESS;
+    }
+
+    if (size == 0 || !Common::Is4KBAligned(size)) {
+        return ERR_INVALID_SIZE;
+    }
+
+    if (!IsValidAddressRange(addr, size)) {
+        return ERR_INVALID_ADDRESS_STATE;
+    }
+
+    const auto permission = static_cast<MemoryPermission>(prot);
+    if (permission != MemoryPermission::None && permission != MemoryPermission::Read &&
+        permission != MemoryPermission::ReadWrite) {
+        return ERR_INVALID_MEMORY_PERMISSIONS;
+    }
+
+    auto* const current_process = Core::CurrentProcess();
+    auto& vm_manager = current_process->VMManager();
+
+    if (!IsInsideAddressSpace(vm_manager, addr, size)) {
+        return ERR_INVALID_ADDRESS_STATE;
+    }
+
+    const VMManager::VMAHandle iter = vm_manager.FindVMA(addr);
+    if (iter == vm_manager.vma_map.end()) {
+        return ERR_INVALID_ADDRESS_STATE;
+    }
+
+    LOG_WARNING(Kernel_SVC, "Uniformity check on protected memory is not implemented.");
+    // TODO: Performs a uniformity check to make sure only protected memory is changed (it doesn't
+    // make sense to allow changing permissions on kernel memory itself, etc).
+
+    const auto converted_permissions = SharedMemory::ConvertPermissions(permission);
+
+    return vm_manager.ReprotectRange(addr, size, converted_permissions);
+}
+
 static ResultCode SetMemoryAttribute(VAddr addr, u64 size, u32 state0, u32 state1) {
     LOG_WARNING(Kernel_SVC,
                 "(STUBBED) called, addr=0x{:X}, size=0x{:X}, state0=0x{:X}, state1=0x{:X}", addr,
@@ -171,7 +214,7 @@ static ResultCode ConnectToNamedPort(Handle* out_handle, VAddr port_name_address
     // Read 1 char beyond the max allowed port name to detect names that are too long.
     std::string port_name = Memory::ReadCString(port_name_address, PortNameMaxLength + 1);
     if (port_name.size() > PortNameMaxLength) {
-        return ERR_PORT_NAME_TOO_LONG;
+        return ERR_OUT_OF_RANGE;
     }
 
     LOG_TRACE(Kernel_SVC, "called port_name={}", port_name);
@@ -267,8 +310,9 @@ static ResultCode WaitSynchronization(Handle* index, VAddr handles_address, u64
 
     static constexpr u64 MaxHandles = 0x40;
 
-    if (handle_count > MaxHandles)
-        return ResultCode(ErrorModule::Kernel, ErrCodes::TooLarge);
+    if (handle_count > MaxHandles) {
+        return ERR_OUT_OF_RANGE;
+    }
 
     auto* const thread = GetCurrentThread();
 
@@ -333,8 +377,7 @@ static ResultCode CancelSynchronization(Handle thread_handle) {
     }
 
     ASSERT(thread->GetStatus() == ThreadStatus::WaitSynchAny);
-    thread->SetWaitSynchronizationResult(
-        ResultCode(ErrorModule::Kernel, ErrCodes::SynchronizationCanceled));
+    thread->SetWaitSynchronizationResult(ERR_SYNCHRONIZATION_CANCELED);
     thread->ResumeFromWait();
     return RESULT_SUCCESS;
 }
@@ -375,9 +418,19 @@ static ResultCode ArbitrateUnlock(VAddr mutex_addr) {
     return Mutex::Release(mutex_addr);
 }
 
+enum class BreakType : u32 {
+    Panic = 0,
+    AssertionFailed = 1,
+    PreNROLoad = 3,
+    PostNROLoad = 4,
+    PreNROUnload = 5,
+    PostNROUnload = 6,
+};
+
 struct BreakReason {
     union {
         u32 raw;
+        BitField<0, 30, BreakType> break_type;
         BitField<31, 1, u32> signal_debugger;
     };
 };
@@ -385,16 +438,80 @@ struct BreakReason {
 /// Break program execution
 static void Break(u32 reason, u64 info1, u64 info2) {
     BreakReason break_reason{reason};
-    if (break_reason.signal_debugger) {
-        LOG_ERROR(
+    bool has_dumped_buffer{};
+
+    const auto handle_debug_buffer = [&](VAddr addr, u64 sz) {
+        if (sz == 0 || addr == 0 || has_dumped_buffer) {
+            return;
+        }
+
+        // This typically is an error code so we're going to assume this is the case
+        if (sz == sizeof(u32)) {
+            LOG_CRITICAL(Debug_Emulated, "debug_buffer_err_code={:X}", Memory::Read32(addr));
+        } else {
+            // We don't know what's in here so we'll hexdump it
+            std::vector<u8> debug_buffer(sz);
+            Memory::ReadBlock(addr, debug_buffer.data(), sz);
+            std::string hexdump;
+            for (std::size_t i = 0; i < debug_buffer.size(); i++) {
+                hexdump += fmt::format("{:02X} ", debug_buffer[i]);
+                if (i != 0 && i % 16 == 0) {
+                    hexdump += '\n';
+                }
+            }
+            LOG_CRITICAL(Debug_Emulated, "debug_buffer=\n{}", hexdump);
+        }
+        has_dumped_buffer = true;
+    };
+    switch (break_reason.break_type) {
+    case BreakType::Panic:
+        LOG_CRITICAL(Debug_Emulated, "Signalling debugger, PANIC! info1=0x{:016X}, info2=0x{:016X}",
+                     info1, info2);
+        handle_debug_buffer(info1, info2);
+        break;
+    case BreakType::AssertionFailed:
+        LOG_CRITICAL(Debug_Emulated,
+                     "Signalling debugger, Assertion failed! info1=0x{:016X}, info2=0x{:016X}",
+                     info1, info2);
+        handle_debug_buffer(info1, info2);
+        break;
+    case BreakType::PreNROLoad:
+        LOG_WARNING(
             Debug_Emulated,
-            "Emulated program broke execution! reason=0x{:016X}, info1=0x{:016X}, info2=0x{:016X}",
-            reason, info1, info2);
-    } else {
+            "Signalling debugger, Attempting to load an NRO at 0x{:016X} with size 0x{:016X}",
+            info1, info2);
+        break;
+    case BreakType::PostNROLoad:
+        LOG_WARNING(Debug_Emulated,
+                    "Signalling debugger, Loaded an NRO at 0x{:016X} with size 0x{:016X}", info1,
+                    info2);
+        break;
+    case BreakType::PreNROUnload:
+        LOG_WARNING(
+            Debug_Emulated,
+            "Signalling debugger, Attempting to unload an NRO at 0x{:016X} with size 0x{:016X}",
+            info1, info2);
+        break;
+    case BreakType::PostNROUnload:
+        LOG_WARNING(Debug_Emulated,
+                    "Signalling debugger, Unloaded an NRO at 0x{:016X} with size 0x{:016X}", info1,
+                    info2);
+        break;
+    default:
+        LOG_WARNING(
+            Debug_Emulated,
+            "Signalling debugger, Unknown break reason {}, info1=0x{:016X}, info2=0x{:016X}",
+            static_cast<u32>(break_reason.break_type.Value()), info1, info2);
+        handle_debug_buffer(info1, info2);
+        break;
+    }
+
+    if (!break_reason.signal_debugger) {
         LOG_CRITICAL(
             Debug_Emulated,
             "Emulated program broke execution! reason=0x{:016X}, info1=0x{:016X}, info2=0x{:016X}",
             reason, info1, info2);
+        handle_debug_buffer(info1, info2);
         ASSERT(false);
 
         Core::CurrentProcess()->PrepareForTermination();
@@ -421,6 +538,37 @@ static ResultCode GetInfo(u64* result, u64 info_id, u64 handle, u64 info_sub_id)
     LOG_TRACE(Kernel_SVC, "called info_id=0x{:X}, info_sub_id=0x{:X}, handle=0x{:08X}", info_id,
               info_sub_id, handle);
 
+    enum class GetInfoType : u64 {
+        // 1.0.0+
+        AllowedCpuIdBitmask = 0,
+        AllowedThreadPrioBitmask = 1,
+        MapRegionBaseAddr = 2,
+        MapRegionSize = 3,
+        HeapRegionBaseAddr = 4,
+        HeapRegionSize = 5,
+        TotalMemoryUsage = 6,
+        TotalHeapUsage = 7,
+        IsCurrentProcessBeingDebugged = 8,
+        ResourceHandleLimit = 9,
+        IdleTickCount = 10,
+        RandomEntropy = 11,
+        PerformanceCounter = 0xF0000002,
+        // 2.0.0+
+        ASLRRegionBaseAddr = 12,
+        ASLRRegionSize = 13,
+        NewMapRegionBaseAddr = 14,
+        NewMapRegionSize = 15,
+        // 3.0.0+
+        IsVirtualAddressMemoryEnabled = 16,
+        PersonalMmHeapUsage = 17,
+        TitleId = 18,
+        // 4.0.0+
+        PrivilegedProcessId = 19,
+        // 5.0.0+
+        UserExceptionContextAddr = 20,
+        ThreadTickCount = 0xF0000002,
+    };
+
     const auto* current_process = Core::CurrentProcess();
     const auto& vm_manager = current_process->VMManager();
 
@@ -453,7 +601,16 @@ static ResultCode GetInfo(u64* result, u64 info_id, u64 handle, u64 info_sub_id)
         *result = 0;
         break;
     case GetInfoType::RandomEntropy:
-        *result = 0;
+        if (handle != 0) {
+            return ERR_INVALID_HANDLE;
+        }
+
+        if (info_sub_id >= Process::RANDOM_ENTROPY_SIZE) {
+            return ERR_INVALID_COMBINATION;
+        }
+
+        *result = current_process->GetRandomEntropy(info_sub_id);
+        return RESULT_SUCCESS;
         break;
     case GetInfoType::ASLRRegionBaseAddr:
         *result = vm_manager.GetASLRRegionBaseAddress();
@@ -483,6 +640,36 @@ static ResultCode GetInfo(u64* result, u64 info_id, u64 handle, u64 info_sub_id)
                     "(STUBBED) Attempted to query user exception context address, returned 0");
         *result = 0;
         break;
+    case GetInfoType::ThreadTickCount: {
+        constexpr u64 num_cpus = 4;
+        if (info_sub_id != 0xFFFFFFFFFFFFFFFF && info_sub_id >= num_cpus) {
+            return ERR_INVALID_COMBINATION;
+        }
+
+        const auto thread =
+            current_process->GetHandleTable().Get<Thread>(static_cast<Handle>(handle));
+        if (!thread) {
+            return ERR_INVALID_HANDLE;
+        }
+
+        const auto& system = Core::System::GetInstance();
+        const auto& scheduler = system.CurrentScheduler();
+        const auto* const current_thread = scheduler.GetCurrentThread();
+        const bool same_thread = current_thread == thread;
+
+        const u64 prev_ctx_ticks = scheduler.GetLastContextSwitchTicks();
+        u64 out_ticks = 0;
+        if (same_thread && info_sub_id == 0xFFFFFFFFFFFFFFFF) {
+            const u64 thread_ticks = current_thread->GetTotalCPUTimeTicks();
+
+            out_ticks = thread_ticks + (CoreTiming::GetTicks() - prev_ctx_ticks);
+        } else if (same_thread && info_sub_id == system.CurrentCoreIndex()) {
+            out_ticks = CoreTiming::GetTicks() - prev_ctx_ticks;
+        }
+
+        *result = out_ticks;
+        break;
+    }
     default:
         UNIMPLEMENTED();
     }
@@ -548,16 +735,17 @@ static ResultCode SetThreadPriority(Handle handle, u32 priority) {
     }
 
     const auto* const current_process = Core::CurrentProcess();
-    SharedPtr<Thread> thread = current_process->GetHandleTable().Get<Thread>(handle);
-    if (!thread) {
-        return ERR_INVALID_HANDLE;
-    }
 
     // Note: The kernel uses the current process's resource limit instead of
     // the one from the thread owner's resource limit.
     const ResourceLimit& resource_limit = current_process->GetResourceLimit();
     if (resource_limit.GetMaxResourceValue(ResourceType::Priority) > priority) {
-        return ERR_NOT_AUTHORIZED;
+        return ERR_INVALID_THREAD_PRIORITY;
+    }
+
+    SharedPtr<Thread> thread = current_process->GetHandleTable().Get<Thread>(handle);
+    if (!thread) {
+        return ERR_INVALID_HANDLE;
     }
 
     thread->SetPriority(priority);
@@ -699,7 +887,7 @@ static ResultCode CreateThread(Handle* out_handle, VAddr entry_point, u64 arg, V
     auto* const current_process = Core::CurrentProcess();
     const ResourceLimit& resource_limit = current_process->GetResourceLimit();
     if (resource_limit.GetMaxResourceValue(ResourceType::Priority) > priority) {
-        return ERR_NOT_AUTHORIZED;
+        return ERR_INVALID_THREAD_PRIORITY;
     }
 
     if (processor_id == THREADPROCESSORID_DEFAULT) {
@@ -993,7 +1181,7 @@ static ResultCode CloseHandle(Handle handle) {
 
 /// Reset an event
 static ResultCode ResetSignal(Handle handle) {
-    LOG_WARNING(Kernel_SVC, "(STUBBED) called handle 0x{:08X}", handle);
+    LOG_DEBUG(Kernel_SVC, "called handle 0x{:08X}", handle);
 
     const auto& handle_table = Core::CurrentProcess()->GetHandleTable();
     auto event = handle_table.Get<Event>(handle);
@@ -1048,7 +1236,7 @@ static ResultCode SetThreadCoreMask(Handle thread_handle, u32 core, u64 mask) {
     }
 
     if (mask == 0) {
-        return ResultCode(ErrorModule::Kernel, ErrCodes::InvalidCombination);
+        return ERR_INVALID_COMBINATION;
     }
 
     /// This value is used to only change the affinity mask without changing the current ideal core.
@@ -1057,12 +1245,12 @@ static ResultCode SetThreadCoreMask(Handle thread_handle, u32 core, u64 mask) {
     if (core == OnlyChangeMask) {
         core = thread->GetIdealCore();
     } else if (core >= Core::NUM_CPU_CORES && core != static_cast<u32>(-1)) {
-        return ResultCode(ErrorModule::Kernel, ErrCodes::InvalidProcessorId);
+        return ERR_INVALID_PROCESSOR_ID;
     }
 
     // Error out if the input core isn't enabled in the input mask.
     if (core < Core::NUM_CPU_CORES && (mask & (1ull << core)) == 0) {
-        return ResultCode(ErrorModule::Kernel, ErrCodes::InvalidCombination);
+        return ERR_INVALID_COMBINATION;
     }
 
     thread->ChangeCore(core, mask);
@@ -1151,7 +1339,7 @@ struct FunctionDef {
 static const FunctionDef SVC_Table[] = {
     {0x00, nullptr, "Unknown"},
     {0x01, SvcWrap<SetHeapSize>, "SetHeapSize"},
-    {0x02, nullptr, "SetMemoryPermission"},
+    {0x02, SvcWrap<SetMemoryPermission>, "SetMemoryPermission"},
     {0x03, SvcWrap<SetMemoryAttribute>, "SetMemoryAttribute"},
     {0x04, SvcWrap<MapMemory>, "MapMemory"},
     {0x05, SvcWrap<UnmapMemory>, "UnmapMemory"},
diff --git a/src/core/hle/kernel/svc.h b/src/core/hle/kernel/svc.h
index 554a5e328..b06aac4ec 100644
--- a/src/core/hle/kernel/svc.h
+++ b/src/core/hle/kernel/svc.h
@@ -24,37 +24,6 @@ struct PageInfo {
     u64 flags;
 };
 
-/// Values accepted by svcGetInfo
-enum class GetInfoType : u64 {
-    // 1.0.0+
-    AllowedCpuIdBitmask = 0,
-    AllowedThreadPrioBitmask = 1,
-    MapRegionBaseAddr = 2,
-    MapRegionSize = 3,
-    HeapRegionBaseAddr = 4,
-    HeapRegionSize = 5,
-    TotalMemoryUsage = 6,
-    TotalHeapUsage = 7,
-    IsCurrentProcessBeingDebugged = 8,
-    ResourceHandleLimit = 9,
-    IdleTickCount = 10,
-    RandomEntropy = 11,
-    PerformanceCounter = 0xF0000002,
-    // 2.0.0+
-    ASLRRegionBaseAddr = 12,
-    ASLRRegionSize = 13,
-    NewMapRegionBaseAddr = 14,
-    NewMapRegionSize = 15,
-    // 3.0.0+
-    IsVirtualAddressMemoryEnabled = 16,
-    PersonalMmHeapUsage = 17,
-    TitleId = 18,
-    // 4.0.0+
-    PrivilegedProcessId = 19,
-    // 5.0.0+
-    UserExceptionContextAddr = 20,
-};
-
 void CallSVC(u32 immediate);
 
 } // namespace Kernel
diff --git a/src/core/hle/kernel/svc_wrap.h b/src/core/hle/kernel/svc_wrap.h
index b09753c80..233a99fb0 100644
--- a/src/core/hle/kernel/svc_wrap.h
+++ b/src/core/hle/kernel/svc_wrap.h
@@ -121,6 +121,11 @@ void SvcWrap() {
     FuncReturn(func(Param(0), Param(1), Param(2)).raw);
 }
 
+template <ResultCode func(u64, u64, u32)>
+void SvcWrap() {
+    FuncReturn(func(Param(0), Param(1), static_cast<u32>(Param(2))).raw);
+}
+
 template <ResultCode func(u32, u64, u64, u32)>
 void SvcWrap() {
     FuncReturn(
diff --git a/src/core/hle/kernel/thread.cpp b/src/core/hle/kernel/thread.cpp
index 59bc9e0af..4ffb76818 100644
--- a/src/core/hle/kernel/thread.cpp
+++ b/src/core/hle/kernel/thread.cpp
@@ -4,9 +4,9 @@
 
 #include <algorithm>
 #include <cinttypes>
+#include <optional>
 #include <vector>
 
-#include <boost/optional.hpp>
 #include <boost/range/algorithm_ext/erase.hpp>
 
 #include "common/assert.h"
@@ -94,7 +94,7 @@ void Thread::CancelWakeupTimer() {
     CoreTiming::UnscheduleEventThreadsafe(kernel.ThreadWakeupCallbackEventType(), callback_handle);
 }
 
-static boost::optional<s32> GetNextProcessorId(u64 mask) {
+static std::optional<s32> GetNextProcessorId(u64 mask) {
     for (s32 index = 0; index < Core::NUM_CPU_CORES; ++index) {
         if (mask & (1ULL << index)) {
             if (!Core::System::GetInstance().Scheduler(index).GetCurrentThread()) {
@@ -142,36 +142,7 @@ void Thread::ResumeFromWait() {
 
     status = ThreadStatus::Ready;
 
-    boost::optional<s32> new_processor_id = GetNextProcessorId(affinity_mask);
-    if (!new_processor_id) {
-        new_processor_id = processor_id;
-    }
-    if (ideal_core != -1 &&
-        Core::System::GetInstance().Scheduler(ideal_core).GetCurrentThread() == nullptr) {
-        new_processor_id = ideal_core;
-    }
-
-    ASSERT(*new_processor_id < 4);
-
-    // Add thread to new core's scheduler
-    auto* next_scheduler = &Core::System::GetInstance().Scheduler(*new_processor_id);
-
-    if (*new_processor_id != processor_id) {
-        // Remove thread from previous core's scheduler
-        scheduler->RemoveThread(this);
-        next_scheduler->AddThread(this, current_priority);
-    }
-
-    processor_id = *new_processor_id;
-
-    // If the thread was ready, unschedule from the previous core and schedule on the new core
-    scheduler->UnscheduleThread(this, current_priority);
-    next_scheduler->ScheduleThread(this, current_priority);
-
-    // Change thread's scheduler
-    scheduler = next_scheduler;
-
-    Core::System::GetInstance().CpuCore(processor_id).PrepareReschedule();
+    ChangeScheduler();
 }
 
 /**
@@ -364,42 +335,45 @@ void Thread::UpdatePriority() {
 void Thread::ChangeCore(u32 core, u64 mask) {
     ideal_core = core;
     affinity_mask = mask;
+    ChangeScheduler();
+}
 
+void Thread::ChangeScheduler() {
     if (status != ThreadStatus::Ready) {
         return;
     }
 
-    boost::optional<s32> new_processor_id{GetNextProcessorId(affinity_mask)};
+    auto& system = Core::System::GetInstance();
+    std::optional<s32> new_processor_id{GetNextProcessorId(affinity_mask)};
 
     if (!new_processor_id) {
         new_processor_id = processor_id;
     }
-    if (ideal_core != -1 &&
-        Core::System::GetInstance().Scheduler(ideal_core).GetCurrentThread() == nullptr) {
+    if (ideal_core != -1 && system.Scheduler(ideal_core).GetCurrentThread() == nullptr) {
         new_processor_id = ideal_core;
     }
 
     ASSERT(*new_processor_id < 4);
 
     // Add thread to new core's scheduler
-    auto* next_scheduler = &Core::System::GetInstance().Scheduler(*new_processor_id);
+    auto& next_scheduler = system.Scheduler(*new_processor_id);
 
     if (*new_processor_id != processor_id) {
         // Remove thread from previous core's scheduler
         scheduler->RemoveThread(this);
-        next_scheduler->AddThread(this, current_priority);
+        next_scheduler.AddThread(this, current_priority);
     }
 
     processor_id = *new_processor_id;
 
     // If the thread was ready, unschedule from the previous core and schedule on the new core
     scheduler->UnscheduleThread(this, current_priority);
-    next_scheduler->ScheduleThread(this, current_priority);
+    next_scheduler.ScheduleThread(this, current_priority);
 
     // Change thread's scheduler
-    scheduler = next_scheduler;
+    scheduler = &next_scheduler;
 
-    Core::System::GetInstance().CpuCore(processor_id).PrepareReschedule();
+    system.CpuCore(processor_id).PrepareReschedule();
 }
 
 bool Thread::AllWaitObjectsReady() {
diff --git a/src/core/hle/kernel/thread.h b/src/core/hle/kernel/thread.h
index f4d7bd235..d384d50db 100644
--- a/src/core/hle/kernel/thread.h
+++ b/src/core/hle/kernel/thread.h
@@ -258,6 +258,14 @@ public:
         return last_running_ticks;
     }
 
+    u64 GetTotalCPUTimeTicks() const {
+        return total_cpu_time_ticks;
+    }
+
+    void UpdateCPUTimeTicks(u64 ticks) {
+        total_cpu_time_ticks += ticks;
+    }
+
     s32 GetProcessorID() const {
         return processor_id;
     }
@@ -366,6 +374,8 @@ private:
     explicit Thread(KernelCore& kernel);
     ~Thread() override;
 
+    void ChangeScheduler();
+
     Core::ARM_Interface::ThreadContext context{};
 
     u32 thread_id = 0;
@@ -378,7 +388,8 @@ private:
     u32 nominal_priority = 0; ///< Nominal thread priority, as set by the emulated application
     u32 current_priority = 0; ///< Current thread priority, can be temporarily changed
 
-    u64 last_running_ticks = 0; ///< CPU tick when thread was last running
+    u64 total_cpu_time_ticks = 0; ///< Total CPU running ticks.
+    u64 last_running_ticks = 0;   ///< CPU tick when thread was last running
 
     s32 processor_id = 0;
 
diff --git a/src/core/hle/kernel/vm_manager.cpp b/src/core/hle/kernel/vm_manager.cpp
index e1a34eef1..100f8f6bf 100644
--- a/src/core/hle/kernel/vm_manager.cpp
+++ b/src/core/hle/kernel/vm_manager.cpp
@@ -143,6 +143,26 @@ ResultVal<VMManager::VMAHandle> VMManager::MapBackingMemory(VAddr target, u8* me
     return MakeResult<VMAHandle>(MergeAdjacent(vma_handle));
 }
 
+ResultVal<VAddr> VMManager::FindFreeRegion(u64 size) const {
+    // Find the first Free VMA.
+    const VAddr base = GetASLRRegionBaseAddress();
+    const VMAHandle vma_handle = std::find_if(vma_map.begin(), vma_map.end(), [&](const auto& vma) {
+        if (vma.second.type != VMAType::Free)
+            return false;
+
+        const VAddr vma_end = vma.second.base + vma.second.size;
+        return vma_end > base && vma_end >= base + size;
+    });
+
+    if (vma_handle == vma_map.end()) {
+        // TODO(Subv): Find the correct error code here.
+        return ResultCode(-1);
+    }
+
+    const VAddr target = std::max(base, vma_handle->second.base);
+    return MakeResult<VAddr>(target);
+}
+
 ResultVal<VMManager::VMAHandle> VMManager::MapMMIO(VAddr target, PAddr paddr, u64 size,
                                                    MemoryState state,
                                                    Memory::MemoryHookPointer mmio_handler) {
@@ -223,6 +243,85 @@ ResultCode VMManager::ReprotectRange(VAddr target, u64 size, VMAPermission new_p
     return RESULT_SUCCESS;
 }
 
+ResultVal<VAddr> VMManager::HeapAllocate(VAddr target, u64 size, VMAPermission perms) {
+    if (target < GetHeapRegionBaseAddress() || target + size > GetHeapRegionEndAddress() ||
+        target + size < target) {
+        return ERR_INVALID_ADDRESS;
+    }
+
+    if (heap_memory == nullptr) {
+        // Initialize heap
+        heap_memory = std::make_shared<std::vector<u8>>();
+        heap_start = heap_end = target;
+    } else {
+        UnmapRange(heap_start, heap_end - heap_start);
+    }
+
+    // If necessary, expand backing vector to cover new heap extents.
+    if (target < heap_start) {
+        heap_memory->insert(begin(*heap_memory), heap_start - target, 0);
+        heap_start = target;
+        RefreshMemoryBlockMappings(heap_memory.get());
+    }
+    if (target + size > heap_end) {
+        heap_memory->insert(end(*heap_memory), (target + size) - heap_end, 0);
+        heap_end = target + size;
+        RefreshMemoryBlockMappings(heap_memory.get());
+    }
+    ASSERT(heap_end - heap_start == heap_memory->size());
+
+    CASCADE_RESULT(auto vma, MapMemoryBlock(target, heap_memory, target - heap_start, size,
+                                            MemoryState::Heap));
+    Reprotect(vma, perms);
+
+    heap_used = size;
+
+    return MakeResult<VAddr>(heap_end - size);
+}
+
+ResultCode VMManager::HeapFree(VAddr target, u64 size) {
+    if (target < GetHeapRegionBaseAddress() || target + size > GetHeapRegionEndAddress() ||
+        target + size < target) {
+        return ERR_INVALID_ADDRESS;
+    }
+
+    if (size == 0) {
+        return RESULT_SUCCESS;
+    }
+
+    const ResultCode result = UnmapRange(target, size);
+    if (result.IsError()) {
+        return result;
+    }
+
+    heap_used -= size;
+    return RESULT_SUCCESS;
+}
+
+ResultCode VMManager::MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size, MemoryState state) {
+    const auto vma = FindVMA(src_addr);
+
+    ASSERT_MSG(vma != vma_map.end(), "Invalid memory address");
+    ASSERT_MSG(vma->second.backing_block, "Backing block doesn't exist for address");
+
+    // The returned VMA might be a bigger one encompassing the desired address.
+    const auto vma_offset = src_addr - vma->first;
+    ASSERT_MSG(vma_offset + size <= vma->second.size,
+               "Shared memory exceeds bounds of mapped block");
+
+    const std::shared_ptr<std::vector<u8>>& backing_block = vma->second.backing_block;
+    const std::size_t backing_block_offset = vma->second.offset + vma_offset;
+
+    CASCADE_RESULT(auto new_vma,
+                   MapMemoryBlock(dst_addr, backing_block, backing_block_offset, size, state));
+    // Protect mirror with permissions from old region
+    Reprotect(new_vma, vma->second.permissions);
+    // Remove permissions from old region
+    Reprotect(vma, VMAPermission::None);
+
+    return RESULT_SUCCESS;
+}
+
 void VMManager::RefreshMemoryBlockMappings(const std::vector<u8>* block) {
     // If this ever proves to have a noticeable performance impact, allow users of the function to
     // specify a specific range of addresses to limit the scan to.
@@ -475,8 +574,7 @@ u64 VMManager::GetTotalMemoryUsage() const {
 }
 
 u64 VMManager::GetTotalHeapUsage() const {
-    LOG_WARNING(Kernel, "(STUBBED) called");
-    return 0x0;
+    return heap_used;
 }
 
 VAddr VMManager::GetAddressSpaceBaseAddress() const {
diff --git a/src/core/hle/kernel/vm_manager.h b/src/core/hle/kernel/vm_manager.h
index 84c890224..d522404fe 100644
--- a/src/core/hle/kernel/vm_manager.h
+++ b/src/core/hle/kernel/vm_manager.h
@@ -158,6 +158,14 @@ public:
     ResultVal<VMAHandle> MapBackingMemory(VAddr target, u8* memory, u64 size, MemoryState state);
 
     /**
+     * Finds the first free address that can hold a region of the desired size.
+     *
+     * @param size Size of the desired region.
+     * @return The found free address.
+     */
+    ResultVal<VAddr> FindFreeRegion(u64 size) const;
+
+    /**
      * Maps a memory-mapped IO region at a given address.
      *
      * @param target The guest address to start the mapping at.
@@ -178,6 +186,12 @@ public:
     /// Changes the permissions of a range of addresses, splitting VMAs as necessary.
     ResultCode ReprotectRange(VAddr target, u64 size, VMAPermission new_perms);
 
+    ResultVal<VAddr> HeapAllocate(VAddr target, u64 size, VMAPermission perms);
+    ResultCode HeapFree(VAddr target, u64 size);
+
+    ResultCode MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size,
+                            MemoryState state = MemoryState::Mapped);
+
     /**
      * Scans all VMAs and updates the page table range of any that use the given vector as backing
      * memory. This should be called after any operation that causes reallocation of the vector.
@@ -335,5 +349,15 @@ private:
 
     VAddr tls_io_region_base = 0;
     VAddr tls_io_region_end = 0;
+
+    // Memory used to back the allocations in the regular heap. A single vector is used to cover
+    // the entire virtual address space extents that bound the allocations, including any holes.
+    // This makes deallocation and reallocation of holes fast and keeps process memory contiguous
+    // in the emulator address space, allowing Memory::GetPointer to be reasonably safe.
+    std::shared_ptr<std::vector<u8>> heap_memory;
+    // The left/right bounds of the address space covered by heap_memory.
+    VAddr heap_start = 0;
+    VAddr heap_end = 0;
+    u64 heap_used = 0;
 };
 } // namespace Kernel
diff --git a/src/core/hle/result.h b/src/core/hle/result.h
index c6b18cfba..bfb77cc31 100644
--- a/src/core/hle/result.h
+++ b/src/core/hle/result.h
@@ -19,8 +19,6 @@
 enum class ErrorDescription : u32 {
     Success = 0,
     RemoteProcessDead = 301,
-    InvalidOffset = 6061,
-    InvalidLength = 6062,
 };
 
 /**
diff --git a/src/core/hle/service/acc/acc.cpp b/src/core/hle/service/acc/acc.cpp
index e61748ca3..c629f9357 100644
--- a/src/core/hle/service/acc/acc.cpp
+++ b/src/core/hle/service/acc/acc.cpp
@@ -2,9 +2,13 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <algorithm>
 #include <array>
+#include "common/common_paths.h"
 #include "common/common_types.h"
+#include "common/file_util.h"
 #include "common/logging/log.h"
+#include "common/string_util.h"
 #include "common/swap.h"
 #include "core/core_timing.h"
 #include "core/hle/ipc_helpers.h"
@@ -16,6 +20,7 @@
 #include "core/hle/service/acc/profile_manager.h"
 
 namespace Service::Account {
+
 // TODO: RE this structure
 struct UserData {
     INSERT_PADDING_WORDS(1);
@@ -27,6 +32,29 @@ struct UserData {
 };
 static_assert(sizeof(UserData) == 0x80, "UserData structure has incorrect size");
 
+// Smallest JPEG https://github.com/mathiasbynens/small/blob/master/jpeg.jpg
+// used as a backup should the one on disk not exist
+constexpr u32 backup_jpeg_size = 107;
+constexpr std::array<u8, backup_jpeg_size> backup_jpeg{{
+    0xff, 0xd8, 0xff, 0xdb, 0x00, 0x43, 0x00, 0x03, 0x02, 0x02, 0x02, 0x02, 0x02, 0x03, 0x02, 0x02,
+    0x02, 0x03, 0x03, 0x03, 0x03, 0x04, 0x06, 0x04, 0x04, 0x04, 0x04, 0x04, 0x08, 0x06, 0x06, 0x05,
+    0x06, 0x09, 0x08, 0x0a, 0x0a, 0x09, 0x08, 0x09, 0x09, 0x0a, 0x0c, 0x0f, 0x0c, 0x0a, 0x0b, 0x0e,
+    0x0b, 0x09, 0x09, 0x0d, 0x11, 0x0d, 0x0e, 0x0f, 0x10, 0x10, 0x11, 0x10, 0x0a, 0x0c, 0x12, 0x13,
+    0x12, 0x10, 0x13, 0x0f, 0x10, 0x10, 0x10, 0xff, 0xc9, 0x00, 0x0b, 0x08, 0x00, 0x01, 0x00, 0x01,
+    0x01, 0x01, 0x11, 0x00, 0xff, 0xcc, 0x00, 0x06, 0x00, 0x10, 0x10, 0x05, 0xff, 0xda, 0x00, 0x08,
+    0x01, 0x01, 0x00, 0x00, 0x3f, 0x00, 0xd2, 0xcf, 0x20, 0xff, 0xd9,
+}};
+
+static std::string GetImagePath(UUID uuid) {
+    return FileUtil::GetUserPath(FileUtil::UserPath::NANDDir) +
+           "/system/save/8000000000000010/su/avators/" + uuid.FormatSwitch() + ".jpg";
+}
+
+static constexpr u32 SanitizeJPEGSize(std::size_t size) {
+    constexpr std::size_t max_jpeg_image_size = 0x20000;
+    return static_cast<u32>(std::min(size, max_jpeg_image_size));
+}
+
 class IProfile final : public ServiceFramework<IProfile> {
 public:
     explicit IProfile(UUID user_id, ProfileManager& profile_manager)
@@ -73,32 +101,42 @@ private:
     }
 
     void LoadImage(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_ACC, "(STUBBED) called");
-        // smallest jpeg https://github.com/mathiasbynens/small/blob/master/jpeg.jpg
-        // TODO(mailwl): load actual profile image from disk, width 256px, max size 0x20000
-        constexpr u32 jpeg_size = 107;
-        static constexpr std::array<u8, jpeg_size> jpeg{
-            0xff, 0xd8, 0xff, 0xdb, 0x00, 0x43, 0x00, 0x03, 0x02, 0x02, 0x02, 0x02, 0x02, 0x03,
-            0x02, 0x02, 0x02, 0x03, 0x03, 0x03, 0x03, 0x04, 0x06, 0x04, 0x04, 0x04, 0x04, 0x04,
-            0x08, 0x06, 0x06, 0x05, 0x06, 0x09, 0x08, 0x0a, 0x0a, 0x09, 0x08, 0x09, 0x09, 0x0a,
-            0x0c, 0x0f, 0x0c, 0x0a, 0x0b, 0x0e, 0x0b, 0x09, 0x09, 0x0d, 0x11, 0x0d, 0x0e, 0x0f,
-            0x10, 0x10, 0x11, 0x10, 0x0a, 0x0c, 0x12, 0x13, 0x12, 0x10, 0x13, 0x0f, 0x10, 0x10,
-            0x10, 0xff, 0xc9, 0x00, 0x0b, 0x08, 0x00, 0x01, 0x00, 0x01, 0x01, 0x01, 0x11, 0x00,
-            0xff, 0xcc, 0x00, 0x06, 0x00, 0x10, 0x10, 0x05, 0xff, 0xda, 0x00, 0x08, 0x01, 0x01,
-            0x00, 0x00, 0x3f, 0x00, 0xd2, 0xcf, 0x20, 0xff, 0xd9,
-        };
-        ctx.WriteBuffer(jpeg);
+        LOG_DEBUG(Service_ACC, "called");
+
         IPC::ResponseBuilder rb{ctx, 3};
         rb.Push(RESULT_SUCCESS);
-        rb.Push<u32>(jpeg_size);
+
+        const FileUtil::IOFile image(GetImagePath(user_id), "rb");
+        if (!image.IsOpen()) {
+            LOG_WARNING(Service_ACC,
+                        "Failed to load user provided image! Falling back to built-in backup...");
+            ctx.WriteBuffer(backup_jpeg);
+            rb.Push<u32>(backup_jpeg_size);
+            return;
+        }
+
+        const u32 size = SanitizeJPEGSize(image.GetSize());
+        std::vector<u8> buffer(size);
+        image.ReadBytes(buffer.data(), buffer.size());
+
+        ctx.WriteBuffer(buffer.data(), buffer.size());
+        rb.Push<u32>(size);
     }
 
     void GetImageSize(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_ACC, "(STUBBED) called");
-        constexpr u32 jpeg_size = 107;
+        LOG_DEBUG(Service_ACC, "called");
         IPC::ResponseBuilder rb{ctx, 3};
         rb.Push(RESULT_SUCCESS);
-        rb.Push<u32>(jpeg_size);
+
+        const FileUtil::IOFile image(GetImagePath(user_id), "rb");
+
+        if (!image.IsOpen()) {
+            LOG_WARNING(Service_ACC,
+                        "Failed to load user provided image! Falling back to built-in backup...");
+            rb.Push<u32>(backup_jpeg_size);
+        } else {
+            rb.Push<u32>(SanitizeJPEGSize(image.GetSize()));
+        }
     }
 
     const ProfileManager& profile_manager;
@@ -204,6 +242,30 @@ void Module::Interface::GetBaasAccountManagerForApplication(Kernel::HLERequestCo
     LOG_DEBUG(Service_ACC, "called");
 }
 
+void Module::Interface::TrySelectUserWithoutInteraction(Kernel::HLERequestContext& ctx) {
+    LOG_DEBUG(Service_ACC, "called");
+    // A u8 is passed into this function which we can safely ignore. It's to determine if we have
+    // access to use the network or not by the looks of it
+    IPC::ResponseBuilder rb{ctx, 6};
+    if (profile_manager->GetUserCount() != 1) {
+        rb.Push(RESULT_SUCCESS);
+        rb.PushRaw<u128>(INVALID_UUID);
+        return;
+    }
+
+    const auto user_list = profile_manager->GetAllUsers();
+    if (std::all_of(user_list.begin(), user_list.end(),
+                    [](const auto& user) { return user.uuid == INVALID_UUID; })) {
+        rb.Push(ResultCode(-1)); // TODO(ogniK): Find the correct error code
+        rb.PushRaw<u128>(INVALID_UUID);
+        return;
+    }
+
+    // Select the first user we have
+    rb.Push(RESULT_SUCCESS);
+    rb.PushRaw<u128>(profile_manager->GetUser(0)->uuid);
+}
+
 Module::Interface::Interface(std::shared_ptr<Module> module,
                              std::shared_ptr<ProfileManager> profile_manager, const char* name)
     : ServiceFramework(name), module(std::move(module)),
diff --git a/src/core/hle/service/acc/acc.h b/src/core/hle/service/acc/acc.h
index c7ed74351..89b2104fa 100644
--- a/src/core/hle/service/acc/acc.h
+++ b/src/core/hle/service/acc/acc.h
@@ -27,6 +27,7 @@ public:
         void InitializeApplicationInfo(Kernel::HLERequestContext& ctx);
         void GetBaasAccountManagerForApplication(Kernel::HLERequestContext& ctx);
         void IsUserRegistrationRequestPermitted(Kernel::HLERequestContext& ctx);
+        void TrySelectUserWithoutInteraction(Kernel::HLERequestContext& ctx);
 
     protected:
         std::shared_ptr<Module> module;
diff --git a/src/core/hle/service/acc/acc_su.cpp b/src/core/hle/service/acc/acc_su.cpp
index ad455c3a7..5e2030355 100644
--- a/src/core/hle/service/acc/acc_su.cpp
+++ b/src/core/hle/service/acc/acc_su.cpp
@@ -17,7 +17,7 @@ ACC_SU::ACC_SU(std::shared_ptr<Module> module, std::shared_ptr<ProfileManager> p
         {5, &ACC_SU::GetProfile, "GetProfile"},
         {6, nullptr, "GetProfileDigest"},
         {50, &ACC_SU::IsUserRegistrationRequestPermitted, "IsUserRegistrationRequestPermitted"},
-        {51, nullptr, "TrySelectUserWithoutInteraction"},
+        {51, &ACC_SU::TrySelectUserWithoutInteraction, "TrySelectUserWithoutInteraction"},
         {60, nullptr, "ListOpenContextStoredUsers"},
         {100, nullptr, "GetUserRegistrationNotifier"},
         {101, nullptr, "GetUserStateChangeNotifier"},
diff --git a/src/core/hle/service/acc/acc_u0.cpp b/src/core/hle/service/acc/acc_u0.cpp
index 72d4adf35..a4d705b45 100644
--- a/src/core/hle/service/acc/acc_u0.cpp
+++ b/src/core/hle/service/acc/acc_u0.cpp
@@ -17,7 +17,7 @@ ACC_U0::ACC_U0(std::shared_ptr<Module> module, std::shared_ptr<ProfileManager> p
         {5, &ACC_U0::GetProfile, "GetProfile"},
         {6, nullptr, "GetProfileDigest"},
         {50, &ACC_U0::IsUserRegistrationRequestPermitted, "IsUserRegistrationRequestPermitted"},
-        {51, nullptr, "TrySelectUserWithoutInteraction"},
+        {51, &ACC_U0::TrySelectUserWithoutInteraction, "TrySelectUserWithoutInteraction"},
         {60, nullptr, "ListOpenContextStoredUsers"},
         {100, &ACC_U0::InitializeApplicationInfo, "InitializeApplicationInfo"},
         {101, &ACC_U0::GetBaasAccountManagerForApplication, "GetBaasAccountManagerForApplication"},
diff --git a/src/core/hle/service/acc/acc_u1.cpp b/src/core/hle/service/acc/acc_u1.cpp
index d480f08e5..8fffc93b5 100644
--- a/src/core/hle/service/acc/acc_u1.cpp
+++ b/src/core/hle/service/acc/acc_u1.cpp
@@ -17,7 +17,7 @@ ACC_U1::ACC_U1(std::shared_ptr<Module> module, std::shared_ptr<ProfileManager> p
         {5, &ACC_U1::GetProfile, "GetProfile"},
         {6, nullptr, "GetProfileDigest"},
         {50, &ACC_U1::IsUserRegistrationRequestPermitted, "IsUserRegistrationRequestPermitted"},
-        {51, nullptr, "TrySelectUserWithoutInteraction"},
+        {51, &ACC_U1::TrySelectUserWithoutInteraction, "TrySelectUserWithoutInteraction"},
         {60, nullptr, "ListOpenContextStoredUsers"},
         {100, nullptr, "GetUserRegistrationNotifier"},
         {101, nullptr, "GetUserStateChangeNotifier"},
diff --git a/src/core/hle/service/acc/profile_manager.cpp b/src/core/hle/service/acc/profile_manager.cpp
index bcb3475db..968263846 100644
--- a/src/core/hle/service/acc/profile_manager.cpp
+++ b/src/core/hle/service/acc/profile_manager.cpp
@@ -2,42 +2,83 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <cstring>
 #include <random>
-#include <boost/optional.hpp>
+
+#include <fmt/format.h>
+
+#include "common/file_util.h"
 #include "core/hle/service/acc/profile_manager.h"
 #include "core/settings.h"
 
 namespace Service::Account {
+
+struct UserRaw {
+    UUID uuid;
+    UUID uuid2;
+    u64 timestamp;
+    ProfileUsername username;
+    INSERT_PADDING_BYTES(0x80);
+};
+static_assert(sizeof(UserRaw) == 0xC8, "UserRaw has incorrect size.");
+
+struct ProfileDataRaw {
+    INSERT_PADDING_BYTES(0x10);
+    std::array<UserRaw, MAX_USERS> users;
+};
+static_assert(sizeof(ProfileDataRaw) == 0x650, "ProfileDataRaw has incorrect size.");
+
 // TODO(ogniK): Get actual error codes
 constexpr ResultCode ERROR_TOO_MANY_USERS(ErrorModule::Account, -1);
 constexpr ResultCode ERROR_USER_ALREADY_EXISTS(ErrorModule::Account, -2);
 constexpr ResultCode ERROR_ARGUMENT_IS_NULL(ErrorModule::Account, 20);
 
-const UUID& UUID::Generate() {
+constexpr char ACC_SAVE_AVATORS_BASE_PATH[] = "/system/save/8000000000000010/su/avators/";
+
+UUID UUID::Generate() {
     std::random_device device;
     std::mt19937 gen(device());
     std::uniform_int_distribution<u64> distribution(1, std::numeric_limits<u64>::max());
-    uuid[0] = distribution(gen);
-    uuid[1] = distribution(gen);
-    return *this;
+    return UUID{distribution(gen), distribution(gen)};
+}
+
+std::string UUID::Format() const {
+    return fmt::format("0x{:016X}{:016X}", uuid[1], uuid[0]);
+}
+
+std::string UUID::FormatSwitch() const {
+    std::array<u8, 16> s{};
+    std::memcpy(s.data(), uuid.data(), sizeof(u128));
+    return fmt::format("{:02x}{:02x}{:02x}{:02x}-{:02x}{:02x}-{:02x}{:02x}-{:02x}{:02x}-{:02x}{"
+                       ":02x}{:02x}{:02x}{:02x}{:02x}",
+                       s[0], s[1], s[2], s[3], s[4], s[5], s[6], s[7], s[8], s[9], s[10], s[11],
+                       s[12], s[13], s[14], s[15]);
 }
 
 ProfileManager::ProfileManager() {
-    // TODO(ogniK): Create the default user we have for now until loading/saving users is added
-    auto user_uuid = UUID{1, 0};
-    ASSERT(CreateNewUser(user_uuid, Settings::values.username).IsSuccess());
-    OpenUser(user_uuid);
+    ParseUserSaveFile();
+
+    if (user_count == 0)
+        CreateNewUser(UUID::Generate(), "yuzu");
+
+    auto current = std::clamp<int>(Settings::values.current_user, 0, MAX_USERS - 1);
+    if (UserExistsIndex(current))
+        current = 0;
+
+    OpenUser(*GetUser(current));
 }
 
-ProfileManager::~ProfileManager() = default;
+ProfileManager::~ProfileManager() {
+    WriteUserSaveFile();
+}
 
 /// After a users creation it needs to be "registered" to the system. AddToProfiles handles the
 /// internal management of the users profiles
-boost::optional<std::size_t> ProfileManager::AddToProfiles(const ProfileInfo& user) {
+std::optional<std::size_t> ProfileManager::AddToProfiles(const ProfileInfo& profile) {
     if (user_count >= MAX_USERS) {
-        return boost::none;
+        return {};
     }
-    profiles[user_count] = user;
+    profiles[user_count] = profile;
     return user_count++;
 }
 
@@ -56,7 +97,7 @@ bool ProfileManager::RemoveProfileAtIndex(std::size_t index) {
 
 /// Helper function to register a user to the system
 ResultCode ProfileManager::AddUser(const ProfileInfo& user) {
-    if (AddToProfiles(user) == boost::none) {
+    if (!AddToProfiles(user)) {
         return ERROR_TOO_MANY_USERS;
     }
     return RESULT_SUCCESS;
@@ -101,31 +142,40 @@ ResultCode ProfileManager::CreateNewUser(UUID uuid, const std::string& username)
     return CreateNewUser(uuid, username_output);
 }
 
+std::optional<UUID> ProfileManager::GetUser(std::size_t index) const {
+    if (index >= MAX_USERS) {
+        return {};
+    }
+
+    return profiles[index].user_uuid;
+}
+
 /// Returns a users profile index based on their user id.
-boost::optional<std::size_t> ProfileManager::GetUserIndex(const UUID& uuid) const {
+std::optional<std::size_t> ProfileManager::GetUserIndex(const UUID& uuid) const {
     if (!uuid) {
-        return boost::none;
+        return {};
     }
-    auto iter = std::find_if(profiles.begin(), profiles.end(),
-                             [&uuid](const ProfileInfo& p) { return p.user_uuid == uuid; });
+
+    const auto iter = std::find_if(profiles.begin(), profiles.end(),
+                                   [&uuid](const ProfileInfo& p) { return p.user_uuid == uuid; });
     if (iter == profiles.end()) {
-        return boost::none;
+        return {};
     }
+
     return static_cast<std::size_t>(std::distance(profiles.begin(), iter));
 }
 
 /// Returns a users profile index based on their profile
-boost::optional<std::size_t> ProfileManager::GetUserIndex(const ProfileInfo& user) const {
+std::optional<std::size_t> ProfileManager::GetUserIndex(const ProfileInfo& user) const {
     return GetUserIndex(user.user_uuid);
 }
 
 /// Returns the data structure used by the switch when GetProfileBase is called on acc:*
-bool ProfileManager::GetProfileBase(boost::optional<std::size_t> index,
-                                    ProfileBase& profile) const {
-    if (index == boost::none || index >= MAX_USERS) {
+bool ProfileManager::GetProfileBase(std::optional<std::size_t> index, ProfileBase& profile) const {
+    if (!index || index >= MAX_USERS) {
         return false;
     }
-    const auto& prof_info = profiles[index.get()];
+    const auto& prof_info = profiles[*index];
     profile.user_uuid = prof_info.user_uuid;
     profile.username = prof_info.username;
     profile.timestamp = prof_info.creation_time;
@@ -134,7 +184,7 @@ bool ProfileManager::GetProfileBase(boost::optional<std::size_t> index,
 
 /// Returns the data structure used by the switch when GetProfileBase is called on acc:*
 bool ProfileManager::GetProfileBase(UUID uuid, ProfileBase& profile) const {
-    auto idx = GetUserIndex(uuid);
+    const auto idx = GetUserIndex(uuid);
     return GetProfileBase(idx, profile);
 }
 
@@ -161,26 +211,34 @@ std::size_t ProfileManager::GetOpenUserCount() const {
 
 /// Checks if a user id exists in our profile manager
 bool ProfileManager::UserExists(UUID uuid) const {
-    return (GetUserIndex(uuid) != boost::none);
+    return GetUserIndex(uuid).has_value();
+}
+
+bool ProfileManager::UserExistsIndex(std::size_t index) const {
+    if (index >= MAX_USERS)
+        return false;
+    return profiles[index].user_uuid.uuid != INVALID_UUID;
 }
 
 /// Opens a specific user
 void ProfileManager::OpenUser(UUID uuid) {
-    auto idx = GetUserIndex(uuid);
-    if (idx == boost::none) {
+    const auto idx = GetUserIndex(uuid);
+    if (!idx) {
         return;
     }
-    profiles[idx.get()].is_open = true;
+
+    profiles[*idx].is_open = true;
     last_opened_user = uuid;
 }
 
 /// Closes a specific user
 void ProfileManager::CloseUser(UUID uuid) {
-    auto idx = GetUserIndex(uuid);
-    if (idx == boost::none) {
+    const auto idx = GetUserIndex(uuid);
+    if (!idx) {
         return;
     }
-    profiles[idx.get()].is_open = false;
+
+    profiles[*idx].is_open = false;
 }
 
 /// Gets all valid user ids on the system
@@ -210,10 +268,10 @@ UUID ProfileManager::GetLastOpenedUser() const {
 }
 
 /// Return the users profile base and the unknown arbitary data.
-bool ProfileManager::GetProfileBaseAndData(boost::optional<std::size_t> index, ProfileBase& profile,
+bool ProfileManager::GetProfileBaseAndData(std::optional<std::size_t> index, ProfileBase& profile,
                                            ProfileData& data) const {
     if (GetProfileBase(index, profile)) {
-        data = profiles[index.get()].data;
+        data = profiles[*index].data;
         return true;
     }
     return false;
@@ -222,7 +280,7 @@ bool ProfileManager::GetProfileBaseAndData(boost::optional<std::size_t> index, P
 /// Return the users profile base and the unknown arbitary data.
 bool ProfileManager::GetProfileBaseAndData(UUID uuid, ProfileBase& profile,
                                            ProfileData& data) const {
-    auto idx = GetUserIndex(uuid);
+    const auto idx = GetUserIndex(uuid);
     return GetProfileBaseAndData(idx, profile, data);
 }
 
@@ -239,4 +297,97 @@ bool ProfileManager::CanSystemRegisterUser() const {
     // emulate qlaunch. Update this to dynamically change.
 }
 
+bool ProfileManager::RemoveUser(UUID uuid) {
+    const auto index = GetUserIndex(uuid);
+    if (!index) {
+        return false;
+    }
+
+    profiles[*index] = ProfileInfo{};
+    std::stable_partition(profiles.begin(), profiles.end(),
+                          [](const ProfileInfo& profile) { return profile.user_uuid; });
+    return true;
+}
+
+bool ProfileManager::SetProfileBase(UUID uuid, const ProfileBase& profile_new) {
+    const auto index = GetUserIndex(uuid);
+    if (!index || profile_new.user_uuid == UUID(INVALID_UUID)) {
+        return false;
+    }
+
+    auto& profile = profiles[*index];
+    profile.user_uuid = profile_new.user_uuid;
+    profile.username = profile_new.username;
+    profile.creation_time = profile_new.timestamp;
+
+    return true;
+}
+
+void ProfileManager::ParseUserSaveFile() {
+    FileUtil::IOFile save(FileUtil::GetUserPath(FileUtil::UserPath::NANDDir) +
+                              ACC_SAVE_AVATORS_BASE_PATH + "profiles.dat",
+                          "rb");
+
+    if (!save.IsOpen()) {
+        LOG_WARNING(Service_ACC, "Failed to load profile data from save data... Generating new "
+                                 "user 'yuzu' with random UUID.");
+        return;
+    }
+
+    ProfileDataRaw data;
+    if (save.ReadBytes(&data, sizeof(ProfileDataRaw)) != sizeof(ProfileDataRaw)) {
+        LOG_WARNING(Service_ACC, "profiles.dat is smaller than expected... Generating new user "
+                                 "'yuzu' with random UUID.");
+        return;
+    }
+
+    for (const auto& user : data.users) {
+        if (user.uuid == UUID(INVALID_UUID)) {
+            continue;
+        }
+
+        AddUser({user.uuid, user.username, user.timestamp, {}, false});
+    }
+
+    std::stable_partition(profiles.begin(), profiles.end(),
+                          [](const ProfileInfo& profile) { return profile.user_uuid; });
+}
+
+void ProfileManager::WriteUserSaveFile() {
+    ProfileDataRaw raw{};
+
+    for (std::size_t i = 0; i < MAX_USERS; ++i) {
+        raw.users[i].username = profiles[i].username;
+        raw.users[i].uuid2 = profiles[i].user_uuid;
+        raw.users[i].uuid = profiles[i].user_uuid;
+        raw.users[i].timestamp = profiles[i].creation_time;
+    }
+
+    const auto raw_path =
+        FileUtil::GetUserPath(FileUtil::UserPath::NANDDir) + "/system/save/8000000000000010";
+    if (FileUtil::Exists(raw_path) && !FileUtil::IsDirectory(raw_path))
+        FileUtil::Delete(raw_path);
+
+    const auto path = FileUtil::GetUserPath(FileUtil::UserPath::NANDDir) +
+                      ACC_SAVE_AVATORS_BASE_PATH + "profiles.dat";
+
+    if (!FileUtil::CreateFullPath(path)) {
+        LOG_WARNING(Service_ACC, "Failed to create full path of profiles.dat. Create the directory "
+                                 "nand/system/save/8000000000000010/su/avators to mitigate this "
+                                 "issue.");
+        return;
+    }
+
+    FileUtil::IOFile save(path, "wb");
+
+    if (!save.IsOpen()) {
+        LOG_WARNING(Service_ACC, "Failed to write save data to file... No changes to user data "
+                                 "made in current session will be saved.");
+        return;
+    }
+
+    save.Resize(sizeof(ProfileDataRaw));
+    save.WriteBytes(&raw, sizeof(ProfileDataRaw));
+}
+
 }; // namespace Service::Account
diff --git a/src/core/hle/service/acc/profile_manager.h b/src/core/hle/service/acc/profile_manager.h
index bffd4cf4d..d2d8e6c6b 100644
--- a/src/core/hle/service/acc/profile_manager.h
+++ b/src/core/hle/service/acc/profile_manager.h
@@ -5,8 +5,8 @@
 #pragma once
 
 #include <array>
+#include <optional>
 
-#include "boost/optional.hpp"
 #include "common/common_types.h"
 #include "common/swap.h"
 #include "core/hle/result.h"
@@ -36,19 +36,20 @@ struct UUID {
     }
 
     // TODO(ogniK): Properly generate uuids based on RFC-4122
-    const UUID& Generate();
+    static UUID Generate();
 
     // Set the UUID to {0,0} to be considered an invalid user
     void Invalidate() {
         uuid = INVALID_UUID;
     }
-    std::string Format() const {
-        return fmt::format("0x{:016X}{:016X}", uuid[1], uuid[0]);
-    }
+
+    std::string Format() const;
+    std::string FormatSwitch() const;
 };
 static_assert(sizeof(UUID) == 16, "UUID is an invalid size!");
 
-using ProfileUsername = std::array<u8, 0x20>;
+constexpr std::size_t profile_username_size = 32;
+using ProfileUsername = std::array<u8, profile_username_size>;
 using ProfileData = std::array<u8, MAX_DATA>;
 using UserIDArray = std::array<UUID, MAX_USERS>;
 
@@ -81,18 +82,19 @@ static_assert(sizeof(ProfileBase) == 0x38, "ProfileBase is an invalid size");
 /// objects
 class ProfileManager {
 public:
-    ProfileManager(); // TODO(ogniK): Load from system save
+    ProfileManager();
     ~ProfileManager();
 
     ResultCode AddUser(const ProfileInfo& user);
     ResultCode CreateNewUser(UUID uuid, const ProfileUsername& username);
     ResultCode CreateNewUser(UUID uuid, const std::string& username);
-    boost::optional<std::size_t> GetUserIndex(const UUID& uuid) const;
-    boost::optional<std::size_t> GetUserIndex(const ProfileInfo& user) const;
-    bool GetProfileBase(boost::optional<std::size_t> index, ProfileBase& profile) const;
+    std::optional<UUID> GetUser(std::size_t index) const;
+    std::optional<std::size_t> GetUserIndex(const UUID& uuid) const;
+    std::optional<std::size_t> GetUserIndex(const ProfileInfo& user) const;
+    bool GetProfileBase(std::optional<std::size_t> index, ProfileBase& profile) const;
     bool GetProfileBase(UUID uuid, ProfileBase& profile) const;
     bool GetProfileBase(const ProfileInfo& user, ProfileBase& profile) const;
-    bool GetProfileBaseAndData(boost::optional<std::size_t> index, ProfileBase& profile,
+    bool GetProfileBaseAndData(std::optional<std::size_t> index, ProfileBase& profile,
                                ProfileData& data) const;
     bool GetProfileBaseAndData(UUID uuid, ProfileBase& profile, ProfileData& data) const;
     bool GetProfileBaseAndData(const ProfileInfo& user, ProfileBase& profile,
@@ -100,6 +102,7 @@ public:
     std::size_t GetUserCount() const;
     std::size_t GetOpenUserCount() const;
     bool UserExists(UUID uuid) const;
+    bool UserExistsIndex(std::size_t index) const;
     void OpenUser(UUID uuid);
     void CloseUser(UUID uuid);
     UserIDArray GetOpenUsers() const;
@@ -108,11 +111,17 @@ public:
 
     bool CanSystemRegisterUser() const;
 
+    bool RemoveUser(UUID uuid);
+    bool SetProfileBase(UUID uuid, const ProfileBase& profile_new);
+
 private:
+    void ParseUserSaveFile();
+    void WriteUserSaveFile();
+    std::optional<std::size_t> AddToProfiles(const ProfileInfo& profile);
+    bool RemoveProfileAtIndex(std::size_t index);
+
     std::array<ProfileInfo, MAX_USERS> profiles{};
     std::size_t user_count = 0;
-    boost::optional<std::size_t> AddToProfiles(const ProfileInfo& profile);
-    bool RemoveProfileAtIndex(std::size_t index);
     UUID last_opened_user{INVALID_UUID};
 };
 
diff --git a/src/core/hle/service/am/am.cpp b/src/core/hle/service/am/am.cpp
index ecf72ae24..3758ecae1 100644
--- a/src/core/hle/service/am/am.cpp
+++ b/src/core/hle/service/am/am.cpp
@@ -4,11 +4,13 @@
 
 #include <array>
 #include <cinttypes>
+#include <cstring>
 #include <stack>
 #include "core/core.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/kernel/event.h"
 #include "core/hle/kernel/process.h"
+#include "core/hle/service/acc/profile_manager.h"
 #include "core/hle/service/am/am.h"
 #include "core/hle/service/am/applet_ae.h"
 #include "core/hle/service/am/applet_oe.h"
@@ -26,6 +28,16 @@
 
 namespace Service::AM {
 
+constexpr u32 POP_LAUNCH_PARAMETER_MAGIC = 0xC79497CA;
+
+struct LaunchParameters {
+    u32_le magic;
+    u32_le is_account_selected;
+    u128 current_user;
+    INSERT_PADDING_BYTES(0x70);
+};
+static_assert(sizeof(LaunchParameters) == 0x88);
+
 IWindowController::IWindowController() : ServiceFramework("IWindowController") {
     // clang-format off
     static const FunctionInfo functions[] = {
@@ -191,8 +203,8 @@ ISelfController::ISelfController(std::shared_ptr<NVFlinger::NVFlinger> nvflinger
 ISelfController::~ISelfController() = default;
 
 void ISelfController::SetFocusHandlingMode(Kernel::HLERequestContext& ctx) {
-    // Takes 3 input u8s with each field located immediately after the previous u8, these are
-    // bool flags. No output.
+    // Takes 3 input u8s with each field located immediately after the previous
+    // u8, these are bool flags. No output.
 
     IPC::RequestParser rp{ctx};
 
@@ -246,8 +258,8 @@ void ISelfController::SetOperationModeChangedNotification(Kernel::HLERequestCont
 }
 
 void ISelfController::SetOutOfFocusSuspendingEnabled(Kernel::HLERequestContext& ctx) {
-    // Takes 3 input u8s with each field located immediately after the previous u8, these are
-    // bool flags. No output.
+    // Takes 3 input u8s with each field located immediately after the previous
+    // u8, these are bool flags. No output.
     IPC::RequestParser rp{ctx};
 
     bool enabled = rp.Pop<bool>();
@@ -290,8 +302,8 @@ void ISelfController::SetScreenShotImageOrientation(Kernel::HLERequestContext& c
 }
 
 void ISelfController::CreateManagedDisplayLayer(Kernel::HLERequestContext& ctx) {
-    // TODO(Subv): Find out how AM determines the display to use, for now just create the layer
-    // in the Default display.
+    // TODO(Subv): Find out how AM determines the display to use, for now just
+    // create the layer in the Default display.
     u64 display_id = nvflinger->OpenDisplay("Default");
     u64 layer_id = nvflinger->CreateLayer(display_id);
 
@@ -326,7 +338,54 @@ void ISelfController::GetIdleTimeDetectionExtension(Kernel::HLERequestContext& c
     LOG_WARNING(Service_AM, "(STUBBED) called");
 }
 
-ICommonStateGetter::ICommonStateGetter() : ServiceFramework("ICommonStateGetter") {
+AppletMessageQueue::AppletMessageQueue() {
+    auto& kernel = Core::System::GetInstance().Kernel();
+    on_new_message = Kernel::Event::Create(kernel, Kernel::ResetType::Sticky,
+                                           "AMMessageQueue:OnMessageRecieved");
+    on_operation_mode_changed = Kernel::Event::Create(kernel, Kernel::ResetType::OneShot,
+                                                      "AMMessageQueue:OperationModeChanged");
+}
+
+AppletMessageQueue::~AppletMessageQueue() = default;
+
+const Kernel::SharedPtr<Kernel::Event>& AppletMessageQueue::GetMesssageRecieveEvent() const {
+    return on_new_message;
+}
+
+const Kernel::SharedPtr<Kernel::Event>& AppletMessageQueue::GetOperationModeChangedEvent() const {
+    return on_operation_mode_changed;
+}
+
+void AppletMessageQueue::PushMessage(AppletMessage msg) {
+    messages.push(msg);
+    on_new_message->Signal();
+}
+
+AppletMessageQueue::AppletMessage AppletMessageQueue::PopMessage() {
+    if (messages.empty()) {
+        on_new_message->Clear();
+        return AppletMessage::NoMessage;
+    }
+    auto msg = messages.front();
+    messages.pop();
+    if (messages.empty()) {
+        on_new_message->Clear();
+    }
+    return msg;
+}
+
+std::size_t AppletMessageQueue::GetMessageCount() const {
+    return messages.size();
+}
+
+void AppletMessageQueue::OperationModeChanged() {
+    PushMessage(AppletMessage::OperationModeChanged);
+    PushMessage(AppletMessage::PerformanceModeChanged);
+    on_operation_mode_changed->Signal();
+}
+
+ICommonStateGetter::ICommonStateGetter(std::shared_ptr<AppletMessageQueue> msg_queue)
+    : ServiceFramework("ICommonStateGetter"), msg_queue(std::move(msg_queue)) {
     // clang-format off
     static const FunctionInfo functions[] = {
         {0, &ICommonStateGetter::GetEventHandle, "GetEventHandle"},
@@ -376,21 +435,19 @@ void ICommonStateGetter::GetBootMode(Kernel::HLERequestContext& ctx) {
 }
 
 void ICommonStateGetter::GetEventHandle(Kernel::HLERequestContext& ctx) {
-    event->Signal();
-
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushCopyObjects(event);
+    rb.PushCopyObjects(msg_queue->GetMesssageRecieveEvent());
 
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+    LOG_DEBUG(Service_AM, "called");
 }
 
 void ICommonStateGetter::ReceiveMessage(Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 3};
     rb.Push(RESULT_SUCCESS);
-    rb.Push<u32>(15);
+    rb.PushEnum<AppletMessageQueue::AppletMessage>(msg_queue->PopMessage());
 
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+    LOG_DEBUG(Service_AM, "called");
 }
 
 void ICommonStateGetter::GetCurrentFocusState(Kernel::HLERequestContext& ctx) {
@@ -402,13 +459,11 @@ void ICommonStateGetter::GetCurrentFocusState(Kernel::HLERequestContext& ctx) {
 }
 
 void ICommonStateGetter::GetDefaultDisplayResolutionChangeEvent(Kernel::HLERequestContext& ctx) {
-    event->Signal();
-
     IPC::ResponseBuilder rb{ctx, 2, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushCopyObjects(event);
+    rb.PushCopyObjects(msg_queue->GetOperationModeChangedEvent());
 
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+    LOG_DEBUG(Service_AM, "called");
 }
 
 void ICommonStateGetter::GetDefaultDisplayResolution(Kernel::HLERequestContext& ctx) {
@@ -432,7 +487,7 @@ void ICommonStateGetter::GetOperationMode(Kernel::HLERequestContext& ctx) {
     rb.Push(RESULT_SUCCESS);
     rb.Push(static_cast<u8>(use_docked_mode ? OperationMode::Docked : OperationMode::Handheld));
 
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+    LOG_DEBUG(Service_AM, "called");
 }
 
 void ICommonStateGetter::GetPerformanceMode(Kernel::HLERequestContext& ctx) {
@@ -442,7 +497,7 @@ void ICommonStateGetter::GetPerformanceMode(Kernel::HLERequestContext& ctx) {
     rb.Push(static_cast<u32>(use_docked_mode ? APM::PerformanceMode::Docked
                                              : APM::PerformanceMode::Handheld));
 
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+    LOG_DEBUG(Service_AM, "called");
 }
 
 class IStorageAccessor final : public ServiceFramework<IStorageAccessor> {
@@ -678,7 +733,7 @@ IApplicationFunctions::IApplicationFunctions() : ServiceFramework("IApplicationF
         {70, nullptr, "RequestToShutdown"},
         {71, nullptr, "RequestToReboot"},
         {80, nullptr, "ExitAndRequestToShowThanksMessage"},
-        {90, nullptr, "EnableApplicationCrashReport"},
+        {90, &IApplicationFunctions::EnableApplicationCrashReport, "EnableApplicationCrashReport"},
         {100, nullptr, "InitializeApplicationCopyrightFrameBuffer"},
         {101, nullptr, "SetApplicationCopyrightImage"},
         {102, nullptr, "SetApplicationCopyrightVisibility"},
@@ -697,6 +752,12 @@ IApplicationFunctions::IApplicationFunctions() : ServiceFramework("IApplicationF
 
 IApplicationFunctions::~IApplicationFunctions() = default;
 
+void IApplicationFunctions::EnableApplicationCrashReport(Kernel::HLERequestContext& ctx) {
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(RESULT_SUCCESS);
+    LOG_WARNING(Service_AM, "(STUBBED) called");
+}
+
 void IApplicationFunctions::BeginBlockingHomeButtonShortAndLongPressed(
     Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2};
@@ -724,20 +785,23 @@ void IApplicationFunctions::EndBlockingHomeButton(Kernel::HLERequestContext& ctx
 }
 
 void IApplicationFunctions::PopLaunchParameter(Kernel::HLERequestContext& ctx) {
-    constexpr std::array<u8, 0x88> data{{
-        0xca, 0x97, 0x94, 0xc7, // Magic
-        1,    0,    0,    0,    // IsAccountSelected (bool)
-        1,    0,    0,    0,    // User Id (word 0)
-        0,    0,    0,    0,    // User Id (word 1)
-        0,    0,    0,    0,    // User Id (word 2)
-        0,    0,    0,    0     // User Id (word 3)
-    }};
+    LaunchParameters params{};
+
+    params.magic = POP_LAUNCH_PARAMETER_MAGIC;
+    params.is_account_selected = 1;
 
-    std::vector<u8> buffer(data.begin(), data.end());
+    Account::ProfileManager profile_manager{};
+    const auto uuid = profile_manager.GetUser(Settings::values.current_user);
+    ASSERT(uuid);
+    params.current_user = uuid->uuid;
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
 
     rb.Push(RESULT_SUCCESS);
+
+    std::vector<u8> buffer(sizeof(LaunchParameters));
+    std::memcpy(buffer.data(), &params, buffer.size());
+
     rb.PushIpcInterface<AM::IStorage>(buffer);
 
     LOG_DEBUG(Service_AM, "called");
@@ -763,7 +827,8 @@ void IApplicationFunctions::EnsureSaveData(Kernel::HLERequestContext& ctx) {
 
 void IApplicationFunctions::SetTerminateResult(Kernel::HLERequestContext& ctx) {
     // Takes an input u32 Result, no output.
-    // For example, in some cases official apps use this with error 0x2A2 then uses svcBreak.
+    // For example, in some cases official apps use this with error 0x2A2 then
+    // uses svcBreak.
 
     IPC::RequestParser rp{ctx};
     u32 result = rp.Pop<u32>();
@@ -825,8 +890,12 @@ void IApplicationFunctions::GetPseudoDeviceId(Kernel::HLERequestContext& ctx) {
 
 void InstallInterfaces(SM::ServiceManager& service_manager,
                        std::shared_ptr<NVFlinger::NVFlinger> nvflinger) {
-    std::make_shared<AppletAE>(nvflinger)->InstallAsService(service_manager);
-    std::make_shared<AppletOE>(nvflinger)->InstallAsService(service_manager);
+    auto message_queue = std::make_shared<AppletMessageQueue>();
+    message_queue->PushMessage(AppletMessageQueue::AppletMessage::FocusStateChanged); // Needed on
+                                                                                      // game boot
+
+    std::make_shared<AppletAE>(nvflinger, message_queue)->InstallAsService(service_manager);
+    std::make_shared<AppletOE>(nvflinger, message_queue)->InstallAsService(service_manager);
     std::make_shared<IdleSys>()->InstallAsService(service_manager);
     std::make_shared<OMM>()->InstallAsService(service_manager);
     std::make_shared<SPSM>()->InstallAsService(service_manager);
diff --git a/src/core/hle/service/am/am.h b/src/core/hle/service/am/am.h
index 095f94851..5a3fcba8f 100644
--- a/src/core/hle/service/am/am.h
+++ b/src/core/hle/service/am/am.h
@@ -5,6 +5,7 @@
 #pragma once
 
 #include <memory>
+#include <queue>
 #include "core/hle/service/service.h"
 
 namespace Kernel {
@@ -39,6 +40,31 @@ enum SystemLanguage {
     TraditionalChinese = 16,
 };
 
+class AppletMessageQueue {
+public:
+    enum class AppletMessage : u32 {
+        NoMessage = 0,
+        FocusStateChanged = 15,
+        OperationModeChanged = 30,
+        PerformanceModeChanged = 31,
+    };
+
+    AppletMessageQueue();
+    ~AppletMessageQueue();
+
+    const Kernel::SharedPtr<Kernel::Event>& GetMesssageRecieveEvent() const;
+    const Kernel::SharedPtr<Kernel::Event>& GetOperationModeChangedEvent() const;
+    void PushMessage(AppletMessage msg);
+    AppletMessage PopMessage();
+    std::size_t GetMessageCount() const;
+    void OperationModeChanged();
+
+private:
+    std::queue<AppletMessage> messages;
+    Kernel::SharedPtr<Kernel::Event> on_new_message;
+    Kernel::SharedPtr<Kernel::Event> on_operation_mode_changed;
+};
+
 class IWindowController final : public ServiceFramework<IWindowController> {
 public:
     IWindowController();
@@ -102,7 +128,7 @@ private:
 
 class ICommonStateGetter final : public ServiceFramework<ICommonStateGetter> {
 public:
-    ICommonStateGetter();
+    explicit ICommonStateGetter(std::shared_ptr<AppletMessageQueue> msg_queue);
     ~ICommonStateGetter() override;
 
 private:
@@ -126,6 +152,7 @@ private:
     void GetDefaultDisplayResolution(Kernel::HLERequestContext& ctx);
 
     Kernel::SharedPtr<Kernel::Event> event;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 class ILibraryAppletCreator final : public ServiceFramework<ILibraryAppletCreator> {
@@ -158,6 +185,7 @@ private:
     void EndBlockingHomeButtonShortAndLongPressed(Kernel::HLERequestContext& ctx);
     void BeginBlockingHomeButton(Kernel::HLERequestContext& ctx);
     void EndBlockingHomeButton(Kernel::HLERequestContext& ctx);
+    void EnableApplicationCrashReport(Kernel::HLERequestContext& ctx);
 };
 
 class IHomeMenuFunctions final : public ServiceFramework<IHomeMenuFunctions> {
diff --git a/src/core/hle/service/am/applet_ae.cpp b/src/core/hle/service/am/applet_ae.cpp
index 68ea778e8..ec93e3529 100644
--- a/src/core/hle/service/am/applet_ae.cpp
+++ b/src/core/hle/service/am/applet_ae.cpp
@@ -12,8 +12,10 @@ namespace Service::AM {
 
 class ILibraryAppletProxy final : public ServiceFramework<ILibraryAppletProxy> {
 public:
-    explicit ILibraryAppletProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
-        : ServiceFramework("ILibraryAppletProxy"), nvflinger(std::move(nvflinger)) {
+    explicit ILibraryAppletProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                                 std::shared_ptr<AppletMessageQueue> msg_queue)
+        : ServiceFramework("ILibraryAppletProxy"), nvflinger(std::move(nvflinger)),
+          msg_queue(std::move(msg_queue)) {
         static const FunctionInfo functions[] = {
             {0, &ILibraryAppletProxy::GetCommonStateGetter, "GetCommonStateGetter"},
             {1, &ILibraryAppletProxy::GetSelfController, "GetSelfController"},
@@ -32,7 +34,7 @@ private:
     void GetCommonStateGetter(Kernel::HLERequestContext& ctx) {
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(RESULT_SUCCESS);
-        rb.PushIpcInterface<ICommonStateGetter>();
+        rb.PushIpcInterface<ICommonStateGetter>(msg_queue);
         LOG_DEBUG(Service_AM, "called");
     }
 
@@ -93,12 +95,15 @@ private:
     }
 
     std::shared_ptr<NVFlinger::NVFlinger> nvflinger;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 class ISystemAppletProxy final : public ServiceFramework<ISystemAppletProxy> {
 public:
-    explicit ISystemAppletProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
-        : ServiceFramework("ISystemAppletProxy"), nvflinger(std::move(nvflinger)) {
+    explicit ISystemAppletProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                                std::shared_ptr<AppletMessageQueue> msg_queue)
+        : ServiceFramework("ISystemAppletProxy"), nvflinger(std::move(nvflinger)),
+          msg_queue(std::move(msg_queue)) {
         static const FunctionInfo functions[] = {
             {0, &ISystemAppletProxy::GetCommonStateGetter, "GetCommonStateGetter"},
             {1, &ISystemAppletProxy::GetSelfController, "GetSelfController"},
@@ -119,7 +124,7 @@ private:
     void GetCommonStateGetter(Kernel::HLERequestContext& ctx) {
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(RESULT_SUCCESS);
-        rb.PushIpcInterface<ICommonStateGetter>();
+        rb.PushIpcInterface<ICommonStateGetter>(msg_queue);
         LOG_DEBUG(Service_AM, "called");
     }
 
@@ -186,31 +191,34 @@ private:
         LOG_DEBUG(Service_AM, "called");
     }
     std::shared_ptr<NVFlinger::NVFlinger> nvflinger;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 void AppletAE::OpenSystemAppletProxy(Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushIpcInterface<ISystemAppletProxy>(nvflinger);
+    rb.PushIpcInterface<ISystemAppletProxy>(nvflinger, msg_queue);
     LOG_DEBUG(Service_AM, "called");
 }
 
 void AppletAE::OpenLibraryAppletProxy(Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushIpcInterface<ILibraryAppletProxy>(nvflinger);
+    rb.PushIpcInterface<ILibraryAppletProxy>(nvflinger, msg_queue);
     LOG_DEBUG(Service_AM, "called");
 }
 
 void AppletAE::OpenLibraryAppletProxyOld(Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushIpcInterface<ILibraryAppletProxy>(nvflinger);
+    rb.PushIpcInterface<ILibraryAppletProxy>(nvflinger, msg_queue);
     LOG_DEBUG(Service_AM, "called");
 }
 
-AppletAE::AppletAE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
-    : ServiceFramework("appletAE"), nvflinger(std::move(nvflinger)) {
+AppletAE::AppletAE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                   std::shared_ptr<AppletMessageQueue> msg_queue)
+    : ServiceFramework("appletAE"), nvflinger(std::move(nvflinger)),
+      msg_queue(std::move(msg_queue)) {
     // clang-format off
     static const FunctionInfo functions[] = {
         {100, &AppletAE::OpenSystemAppletProxy, "OpenSystemAppletProxy"},
@@ -228,4 +236,8 @@ AppletAE::AppletAE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
 
 AppletAE::~AppletAE() = default;
 
+const std::shared_ptr<AppletMessageQueue>& AppletAE::GetMessageQueue() const {
+    return msg_queue;
+}
+
 } // namespace Service::AM
diff --git a/src/core/hle/service/am/applet_ae.h b/src/core/hle/service/am/applet_ae.h
index 1ed77baa4..902db2665 100644
--- a/src/core/hle/service/am/applet_ae.h
+++ b/src/core/hle/service/am/applet_ae.h
@@ -17,15 +17,19 @@ namespace AM {
 
 class AppletAE final : public ServiceFramework<AppletAE> {
 public:
-    explicit AppletAE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger);
+    explicit AppletAE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                      std::shared_ptr<AppletMessageQueue> msg_queue);
     ~AppletAE() override;
 
+    const std::shared_ptr<AppletMessageQueue>& GetMessageQueue() const;
+
 private:
     void OpenSystemAppletProxy(Kernel::HLERequestContext& ctx);
     void OpenLibraryAppletProxy(Kernel::HLERequestContext& ctx);
     void OpenLibraryAppletProxyOld(Kernel::HLERequestContext& ctx);
 
     std::shared_ptr<NVFlinger::NVFlinger> nvflinger;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 } // namespace AM
diff --git a/src/core/hle/service/am/applet_oe.cpp b/src/core/hle/service/am/applet_oe.cpp
index 60717afd9..20c8d5fff 100644
--- a/src/core/hle/service/am/applet_oe.cpp
+++ b/src/core/hle/service/am/applet_oe.cpp
@@ -12,8 +12,10 @@ namespace Service::AM {
 
 class IApplicationProxy final : public ServiceFramework<IApplicationProxy> {
 public:
-    explicit IApplicationProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
-        : ServiceFramework("IApplicationProxy"), nvflinger(std::move(nvflinger)) {
+    explicit IApplicationProxy(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                               std::shared_ptr<AppletMessageQueue> msg_queue)
+        : ServiceFramework("IApplicationProxy"), nvflinger(std::move(nvflinger)),
+          msg_queue(std::move(msg_queue)) {
         // clang-format off
         static const FunctionInfo functions[] = {
             {0, &IApplicationProxy::GetCommonStateGetter, "GetCommonStateGetter"},
@@ -70,7 +72,7 @@ private:
     void GetCommonStateGetter(Kernel::HLERequestContext& ctx) {
         IPC::ResponseBuilder rb{ctx, 2, 0, 1};
         rb.Push(RESULT_SUCCESS);
-        rb.PushIpcInterface<ICommonStateGetter>();
+        rb.PushIpcInterface<ICommonStateGetter>(msg_queue);
         LOG_DEBUG(Service_AM, "called");
     }
 
@@ -89,17 +91,20 @@ private:
     }
 
     std::shared_ptr<NVFlinger::NVFlinger> nvflinger;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 void AppletOE::OpenApplicationProxy(Kernel::HLERequestContext& ctx) {
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushIpcInterface<IApplicationProxy>(nvflinger);
+    rb.PushIpcInterface<IApplicationProxy>(nvflinger, msg_queue);
     LOG_DEBUG(Service_AM, "called");
 }
 
-AppletOE::AppletOE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
-    : ServiceFramework("appletOE"), nvflinger(std::move(nvflinger)) {
+AppletOE::AppletOE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                   std::shared_ptr<AppletMessageQueue> msg_queue)
+    : ServiceFramework("appletOE"), nvflinger(std::move(nvflinger)),
+      msg_queue(std::move(msg_queue)) {
     static const FunctionInfo functions[] = {
         {0, &AppletOE::OpenApplicationProxy, "OpenApplicationProxy"},
     };
@@ -108,4 +113,8 @@ AppletOE::AppletOE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger)
 
 AppletOE::~AppletOE() = default;
 
+const std::shared_ptr<AppletMessageQueue>& AppletOE::GetMessageQueue() const {
+    return msg_queue;
+}
+
 } // namespace Service::AM
diff --git a/src/core/hle/service/am/applet_oe.h b/src/core/hle/service/am/applet_oe.h
index 60cfdfd9d..bbd0108ef 100644
--- a/src/core/hle/service/am/applet_oe.h
+++ b/src/core/hle/service/am/applet_oe.h
@@ -17,13 +17,17 @@ namespace AM {
 
 class AppletOE final : public ServiceFramework<AppletOE> {
 public:
-    explicit AppletOE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger);
+    explicit AppletOE(std::shared_ptr<NVFlinger::NVFlinger> nvflinger,
+                      std::shared_ptr<AppletMessageQueue> msg_queue);
     ~AppletOE() override;
 
+    const std::shared_ptr<AppletMessageQueue>& GetMessageQueue() const;
+
 private:
     void OpenApplicationProxy(Kernel::HLERequestContext& ctx);
 
     std::shared_ptr<NVFlinger::NVFlinger> nvflinger;
+    std::shared_ptr<AppletMessageQueue> msg_queue;
 };
 
 } // namespace AM
diff --git a/src/core/hle/service/aoc/aoc_u.cpp b/src/core/hle/service/aoc/aoc_u.cpp
index 428069df2..54305cf05 100644
--- a/src/core/hle/service/aoc/aoc_u.cpp
+++ b/src/core/hle/service/aoc/aoc_u.cpp
@@ -24,8 +24,8 @@ namespace Service::AOC {
 constexpr u64 DLC_BASE_TITLE_ID_MASK = 0xFFFFFFFFFFFFE000;
 constexpr u64 DLC_BASE_TO_AOC_ID = 0x1000;
 
-static bool CheckAOCTitleIDMatchesBase(u64 base, u64 aoc) {
-    return (aoc & DLC_BASE_TITLE_ID_MASK) == base;
+static bool CheckAOCTitleIDMatchesBase(u64 title_id, u64 base) {
+    return (title_id & DLC_BASE_TITLE_ID_MASK) == base;
 }
 
 static std::vector<u64> AccumulateAOCTitleIDs() {
@@ -74,7 +74,7 @@ void AOC_U::CountAddOnContent(Kernel::HLERequestContext& ctx) {
     const auto current = Core::System::GetInstance().CurrentProcess()->GetTitleID();
     rb.Push<u32>(static_cast<u32>(
         std::count_if(add_on_content.begin(), add_on_content.end(),
-                      [&current](u64 tid) { return (tid & DLC_BASE_TITLE_ID_MASK) == current; })));
+                      [current](u64 tid) { return CheckAOCTitleIDMatchesBase(tid, current); })));
 }
 
 void AOC_U::ListAddOnContent(Kernel::HLERequestContext& ctx) {
diff --git a/src/core/hle/service/audio/audren_u.cpp b/src/core/hle/service/audio/audren_u.cpp
index fac6785a5..d3ea57ea7 100644
--- a/src/core/hle/service/audio/audren_u.cpp
+++ b/src/core/hle/service/audio/audren_u.cpp
@@ -28,13 +28,13 @@ public:
             {1, &IAudioRenderer::GetSampleCount, "GetSampleCount"},
             {2, &IAudioRenderer::GetMixBufferCount, "GetMixBufferCount"},
             {3, &IAudioRenderer::GetState, "GetState"},
-            {4, &IAudioRenderer::RequestUpdate, "RequestUpdate"},
+            {4, &IAudioRenderer::RequestUpdateImpl, "RequestUpdate"},
             {5, &IAudioRenderer::Start, "Start"},
             {6, &IAudioRenderer::Stop, "Stop"},
             {7, &IAudioRenderer::QuerySystemEvent, "QuerySystemEvent"},
-            {8, nullptr, "SetRenderingTimeLimit"},
-            {9, nullptr, "GetRenderingTimeLimit"},
-            {10, nullptr, "RequestUpdateAuto"},
+            {8, &IAudioRenderer::SetRenderingTimeLimit, "SetRenderingTimeLimit"},
+            {9, &IAudioRenderer::GetRenderingTimeLimit, "GetRenderingTimeLimit"},
+            {10, &IAudioRenderer::RequestUpdateImpl, "RequestUpdateAuto"},
             {11, nullptr, "ExecuteAudioRendererRendering"},
         };
         // clang-format on
@@ -79,7 +79,7 @@ private:
         LOG_DEBUG(Service_Audio, "called");
     }
 
-    void RequestUpdate(Kernel::HLERequestContext& ctx) {
+    void RequestUpdateImpl(Kernel::HLERequestContext& ctx) {
         ctx.WriteBuffer(renderer->UpdateAudioRenderer(ctx.ReadBuffer()));
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(RESULT_SUCCESS);
@@ -110,8 +110,29 @@ private:
         LOG_WARNING(Service_Audio, "(STUBBED) called");
     }
 
+    void SetRenderingTimeLimit(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        rendering_time_limit_percent = rp.Pop<u32>();
+        ASSERT(rendering_time_limit_percent >= 0 && rendering_time_limit_percent <= 100);
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+
+        LOG_DEBUG(Service_Audio, "called. rendering_time_limit_percent={}",
+                  rendering_time_limit_percent);
+    }
+
+    void GetRenderingTimeLimit(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_Audio, "called");
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push(rendering_time_limit_percent);
+    }
+
     Kernel::SharedPtr<Kernel::Event> system_event;
     std::unique_ptr<AudioCore::AudioRenderer> renderer;
+    u32 rendering_time_limit_percent = 100;
 };
 
 class IAudioDevice final : public ServiceFramework<IAudioDevice> {
diff --git a/src/core/hle/service/audio/hwopus.cpp b/src/core/hle/service/audio/hwopus.cpp
index 7168c6a10..763e619a4 100644
--- a/src/core/hle/service/audio/hwopus.cpp
+++ b/src/core/hle/service/audio/hwopus.cpp
@@ -77,8 +77,8 @@ private:
         IPC::ResponseBuilder rb{ctx, 6};
         rb.Push(RESULT_SUCCESS);
         rb.Push<u32>(consumed);
-        rb.Push<u64>(performance);
         rb.Push<u32>(sample_count);
+        rb.Push<u64>(performance);
         ctx.WriteBuffer(samples.data(), samples.size() * sizeof(s16));
     }
 
@@ -161,7 +161,7 @@ void HwOpus::OpenOpusDecoder(Kernel::HLERequestContext& ctx) {
     ASSERT_MSG(channel_count == 1 || channel_count == 2, "Invalid channel count");
 
     std::size_t worker_sz = WorkerBufferSize(channel_count);
-    ASSERT_MSG(buffer_sz < worker_sz, "Worker buffer too large");
+    ASSERT_MSG(buffer_sz >= worker_sz, "Worker buffer too large");
     std::unique_ptr<OpusDecoder, OpusDeleter> decoder{
         static_cast<OpusDecoder*>(operator new(worker_sz))};
     if (opus_decoder_init(decoder.get(), sample_rate, channel_count)) {
diff --git a/src/core/hle/service/btdrv/btdrv.cpp b/src/core/hle/service/btdrv/btdrv.cpp
index d0a15cc4c..f3bde6d0d 100644
--- a/src/core/hle/service/btdrv/btdrv.cpp
+++ b/src/core/hle/service/btdrv/btdrv.cpp
@@ -2,12 +2,49 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include "common/logging/log.h"
+#include "core/hle/ipc_helpers.h"
+#include "core/hle/kernel/event.h"
+#include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/btdrv/btdrv.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sm/sm.h"
 
 namespace Service::BtDrv {
 
+class Bt final : public ServiceFramework<Bt> {
+public:
+    explicit Bt() : ServiceFramework{"bt"} {
+        // clang-format off
+        static const FunctionInfo functions[] = {
+            {0, nullptr, "Unknown0"},
+            {1, nullptr, "Unknown1"},
+            {2, nullptr, "Unknown2"},
+            {3, nullptr, "Unknown3"},
+            {4, nullptr, "Unknown4"},
+            {5, nullptr, "Unknown5"},
+            {6, nullptr, "Unknown6"},
+            {7, nullptr, "Unknown7"},
+            {8, nullptr, "Unknown8"},
+            {9, &Bt::RegisterEvent, "RegisterEvent"},
+        };
+        // clang-format on
+        RegisterHandlers(functions);
+    }
+
+private:
+    void RegisterEvent(Kernel::HLERequestContext& ctx) {
+        auto& kernel = Core::System::GetInstance().Kernel();
+        register_event =
+            Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "BT:RegisterEvent");
+        IPC::ResponseBuilder rb{ctx, 2, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushCopyObjects(register_event);
+        LOG_WARNING(Service_BTM, "(STUBBED) called");
+    }
+    Kernel::SharedPtr<Kernel::Event> register_event;
+};
+
 class BtDrv final : public ServiceFramework<BtDrv> {
 public:
     explicit BtDrv() : ServiceFramework{"btdrv"} {
@@ -67,6 +104,7 @@ public:
 
 void InstallInterfaces(SM::ServiceManager& sm) {
     std::make_shared<BtDrv>()->InstallAsService(sm);
+    std::make_shared<Bt>()->InstallAsService(sm);
 }
 
 } // namespace Service::BtDrv
diff --git a/src/core/hle/service/btm/btm.cpp b/src/core/hle/service/btm/btm.cpp
index b949bfabd..a02f6b53a 100644
--- a/src/core/hle/service/btm/btm.cpp
+++ b/src/core/hle/service/btm/btm.cpp
@@ -6,13 +6,118 @@
 
 #include "common/logging/log.h"
 #include "core/hle/ipc_helpers.h"
+#include "core/hle/kernel/event.h"
 #include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/btm/btm.h"
 #include "core/hle/service/service.h"
-#include "core/hle/service/sm/sm.h"
 
 namespace Service::BTM {
 
+class IBtmUserCore final : public ServiceFramework<IBtmUserCore> {
+public:
+    explicit IBtmUserCore() : ServiceFramework{"IBtmUserCore"} {
+        // clang-format off
+        static const FunctionInfo functions[] = {
+            {0, &IBtmUserCore::GetScanEvent, "GetScanEvent"},
+            {1, nullptr, "Unknown1"},
+            {2, nullptr, "Unknown2"},
+            {3, nullptr, "Unknown3"},
+            {4, nullptr, "Unknown4"},
+            {5, nullptr, "Unknown5"},
+            {6, nullptr, "Unknown6"},
+            {7, nullptr, "Unknown7"},
+            {8, nullptr, "Unknown8"},
+            {9, nullptr, "Unknown9"},
+            {10, nullptr, "Unknown10"},
+            {17, &IBtmUserCore::GetConnectionEvent, "GetConnectionEvent"},
+            {18, nullptr, "Unknown18"},
+            {19, nullptr, "Unknown19"},
+            {20, nullptr, "Unknown20"},
+            {21, nullptr, "Unknown21"},
+            {22, nullptr, "Unknown22"},
+            {23, nullptr, "Unknown23"},
+            {24, nullptr, "Unknown24"},
+            {25, nullptr, "Unknown25"},
+            {26, &IBtmUserCore::GetDiscoveryEvent, "AcquireBleServiceDiscoveryEventImpl"},
+            {27, nullptr, "Unknown27"},
+            {28, nullptr, "Unknown28"},
+            {29, nullptr, "Unknown29"},
+            {30, nullptr, "Unknown30"},
+            {31, nullptr, "Unknown31"},
+            {32, nullptr, "Unknown32"},
+            {33, &IBtmUserCore::GetConfigEvent, "GetConfigEvent"},
+            {34, nullptr, "Unknown34"},
+            {35, nullptr, "Unknown35"},
+            {36, nullptr, "Unknown36"},
+            {37, nullptr, "Unknown37"},
+        };
+        // clang-format on
+        RegisterHandlers(functions);
+    }
+
+private:
+    void GetScanEvent(Kernel::HLERequestContext& ctx) {
+        auto& kernel = Core::System::GetInstance().Kernel();
+        scan_event =
+            Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "IBtmUserCore:ScanEvent");
+        IPC::ResponseBuilder rb{ctx, 2, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushCopyObjects(scan_event);
+        LOG_WARNING(Service_BTM, "(STUBBED) called");
+    }
+    void GetConnectionEvent(Kernel::HLERequestContext& ctx) {
+        auto& kernel = Core::System::GetInstance().Kernel();
+        connection_event = Kernel::Event::Create(kernel, Kernel::ResetType::OneShot,
+                                                 "IBtmUserCore:ConnectionEvent");
+        IPC::ResponseBuilder rb{ctx, 2, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushCopyObjects(connection_event);
+        LOG_WARNING(Service_BTM, "(STUBBED) called");
+    }
+    void GetDiscoveryEvent(Kernel::HLERequestContext& ctx) {
+        auto& kernel = Core::System::GetInstance().Kernel();
+        service_discovery =
+            Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "IBtmUserCore:Discovery");
+        IPC::ResponseBuilder rb{ctx, 2, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushCopyObjects(service_discovery);
+        LOG_WARNING(Service_BTM, "(STUBBED) called");
+    }
+    void GetConfigEvent(Kernel::HLERequestContext& ctx) {
+        auto& kernel = Core::System::GetInstance().Kernel();
+        config_event =
+            Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "IBtmUserCore:ConfigEvent");
+        IPC::ResponseBuilder rb{ctx, 2, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushCopyObjects(config_event);
+        LOG_WARNING(Service_BTM, "(STUBBED) called");
+    }
+    Kernel::SharedPtr<Kernel::Event> scan_event;
+    Kernel::SharedPtr<Kernel::Event> connection_event;
+    Kernel::SharedPtr<Kernel::Event> service_discovery;
+    Kernel::SharedPtr<Kernel::Event> config_event;
+};
+
+class BTM_USR final : public ServiceFramework<BTM_USR> {
+public:
+    explicit BTM_USR() : ServiceFramework{"btm:u"} {
+        // clang-format off
+        static const FunctionInfo functions[] = {
+            {0, &BTM_USR::GetCoreImpl, "GetCoreImpl"},
+        };
+        // clang-format on
+        RegisterHandlers(functions);
+    }
+
+private:
+    void GetCoreImpl(Kernel::HLERequestContext& ctx) {
+        IPC::ResponseBuilder rb{ctx, 2, 0, 1};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushIpcInterface<IBtmUserCore>();
+        LOG_DEBUG(Service_BTM, "called");
+    }
+};
+
 class BTM final : public ServiceFramework<BTM> {
 public:
     explicit BTM() : ServiceFramework{"btm"} {
@@ -116,6 +221,7 @@ void InstallInterfaces(SM::ServiceManager& sm) {
     std::make_shared<BTM>()->InstallAsService(sm);
     std::make_shared<BTM_DBG>()->InstallAsService(sm);
     std::make_shared<BTM_SYS>()->InstallAsService(sm);
+    std::make_shared<BTM_USR>()->InstallAsService(sm);
 }
 
 } // namespace Service::BTM
diff --git a/src/core/hle/service/filesystem/filesystem.cpp b/src/core/hle/service/filesystem/filesystem.cpp
index e32a7c48e..5d6294016 100644
--- a/src/core/hle/service/filesystem/filesystem.cpp
+++ b/src/core/hle/service/filesystem/filesystem.cpp
@@ -303,25 +303,42 @@ ResultVal<FileSys::VirtualDir> OpenSaveData(FileSys::SaveDataSpaceId space,
               static_cast<u8>(space), save_struct.DebugInfo());
 
     if (save_data_factory == nullptr) {
-        return ResultCode(ErrorModule::FS, FileSys::ErrCodes::TitleNotFound);
+        return FileSys::ERROR_ENTITY_NOT_FOUND;
     }
 
     return save_data_factory->Open(space, save_struct);
 }
 
+ResultVal<FileSys::VirtualDir> OpenSaveDataSpace(FileSys::SaveDataSpaceId space) {
+    LOG_TRACE(Service_FS, "Opening Save Data Space for space_id={:01X}", static_cast<u8>(space));
+
+    if (save_data_factory == nullptr) {
+        return FileSys::ERROR_ENTITY_NOT_FOUND;
+    }
+
+    return MakeResult(save_data_factory->GetSaveDataSpaceDirectory(space));
+}
+
 ResultVal<FileSys::VirtualDir> OpenSDMC() {
     LOG_TRACE(Service_FS, "Opening SDMC");
 
     if (sdmc_factory == nullptr) {
-        return ResultCode(ErrorModule::FS, FileSys::ErrCodes::SdCardNotFound);
+        return FileSys::ERROR_SD_CARD_NOT_FOUND;
     }
 
     return sdmc_factory->Open();
 }
 
-std::unique_ptr<FileSys::RegisteredCacheUnion> GetUnionContents() {
-    return std::make_unique<FileSys::RegisteredCacheUnion>(std::vector<FileSys::RegisteredCache*>{
-        GetSystemNANDContents(), GetUserNANDContents(), GetSDMCContents()});
+std::shared_ptr<FileSys::RegisteredCacheUnion> registered_cache_union;
+
+std::shared_ptr<FileSys::RegisteredCacheUnion> GetUnionContents() {
+    if (registered_cache_union == nullptr) {
+        registered_cache_union =
+            std::make_shared<FileSys::RegisteredCacheUnion>(std::vector<FileSys::RegisteredCache*>{
+                GetSystemNANDContents(), GetUserNANDContents(), GetSDMCContents()});
+    }
+
+    return registered_cache_union;
 }
 
 FileSys::RegisteredCache* GetSystemNANDContents() {
@@ -360,6 +377,15 @@ FileSys::VirtualDir GetModificationLoadRoot(u64 title_id) {
     return bis_factory->GetModificationLoadRoot(title_id);
 }
 
+FileSys::VirtualDir GetModificationDumpRoot(u64 title_id) {
+    LOG_TRACE(Service_FS, "Opening mod dump root for tid={:016X}", title_id);
+
+    if (bis_factory == nullptr)
+        return nullptr;
+
+    return bis_factory->GetModificationDumpRoot(title_id);
+}
+
 void CreateFactories(FileSys::VfsFilesystem& vfs, bool overwrite) {
     if (overwrite) {
         bis_factory = nullptr;
@@ -373,13 +399,21 @@ void CreateFactories(FileSys::VfsFilesystem& vfs, bool overwrite) {
                                           FileSys::Mode::ReadWrite);
     auto load_directory = vfs.OpenDirectory(FileUtil::GetUserPath(FileUtil::UserPath::LoadDir),
                                             FileSys::Mode::ReadWrite);
+    auto dump_directory = vfs.OpenDirectory(FileUtil::GetUserPath(FileUtil::UserPath::DumpDir),
+                                            FileSys::Mode::ReadWrite);
 
-    if (bis_factory == nullptr)
-        bis_factory = std::make_unique<FileSys::BISFactory>(nand_directory, load_directory);
-    if (save_data_factory == nullptr)
+    if (bis_factory == nullptr) {
+        bis_factory =
+            std::make_unique<FileSys::BISFactory>(nand_directory, load_directory, dump_directory);
+    }
+
+    if (save_data_factory == nullptr) {
         save_data_factory = std::make_unique<FileSys::SaveDataFactory>(std::move(nand_directory));
-    if (sdmc_factory == nullptr)
+    }
+
+    if (sdmc_factory == nullptr) {
         sdmc_factory = std::make_unique<FileSys::SDMCFactory>(std::move(sd_directory));
+    }
 }
 
 void InstallInterfaces(SM::ServiceManager& service_manager, FileSys::VfsFilesystem& vfs) {
diff --git a/src/core/hle/service/filesystem/filesystem.h b/src/core/hle/service/filesystem/filesystem.h
index 6ca5c5636..ff9182e84 100644
--- a/src/core/hle/service/filesystem/filesystem.h
+++ b/src/core/hle/service/filesystem/filesystem.h
@@ -45,15 +45,17 @@ ResultVal<FileSys::VirtualFile> OpenRomFS(u64 title_id, FileSys::StorageId stora
                                           FileSys::ContentRecordType type);
 ResultVal<FileSys::VirtualDir> OpenSaveData(FileSys::SaveDataSpaceId space,
                                             FileSys::SaveDataDescriptor save_struct);
+ResultVal<FileSys::VirtualDir> OpenSaveDataSpace(FileSys::SaveDataSpaceId space);
 ResultVal<FileSys::VirtualDir> OpenSDMC();
 
-std::unique_ptr<FileSys::RegisteredCacheUnion> GetUnionContents();
+std::shared_ptr<FileSys::RegisteredCacheUnion> GetUnionContents();
 
 FileSys::RegisteredCache* GetSystemNANDContents();
 FileSys::RegisteredCache* GetUserNANDContents();
 FileSys::RegisteredCache* GetSDMCContents();
 
 FileSys::VirtualDir GetModificationLoadRoot(u64 title_id);
+FileSys::VirtualDir GetModificationDumpRoot(u64 title_id);
 
 // Creates the SaveData, SDMC, and BIS Factories. Should be called once and before any function
 // above is called.
diff --git a/src/core/hle/service/filesystem/fsp_srv.cpp b/src/core/hle/service/filesystem/fsp_srv.cpp
index d5dced429..038dc80b1 100644
--- a/src/core/hle/service/filesystem/fsp_srv.cpp
+++ b/src/core/hle/service/filesystem/fsp_srv.cpp
@@ -11,12 +11,14 @@
 
 #include "common/assert.h"
 #include "common/common_types.h"
+#include "common/hex_util.h"
 #include "common/logging/log.h"
 #include "common/string_util.h"
 #include "core/file_sys/directory.h"
 #include "core/file_sys/errors.h"
 #include "core/file_sys/mode.h"
 #include "core/file_sys/nca_metadata.h"
+#include "core/file_sys/patch_manager.h"
 #include "core/file_sys/savedata_factory.h"
 #include "core/file_sys/vfs.h"
 #include "core/hle/ipc_helpers.h"
@@ -61,12 +63,12 @@ private:
         // Error checking
         if (length < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidLength));
+            rb.Push(FileSys::ERROR_INVALID_SIZE);
             return;
         }
         if (offset < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidOffset));
+            rb.Push(FileSys::ERROR_INVALID_OFFSET);
             return;
         }
 
@@ -106,12 +108,12 @@ private:
         // Error checking
         if (length < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidLength));
+            rb.Push(FileSys::ERROR_INVALID_SIZE);
             return;
         }
         if (offset < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidOffset));
+            rb.Push(FileSys::ERROR_INVALID_OFFSET);
             return;
         }
 
@@ -137,12 +139,12 @@ private:
         // Error checking
         if (length < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidLength));
+            rb.Push(FileSys::ERROR_INVALID_SIZE);
             return;
         }
         if (offset < 0) {
             IPC::ResponseBuilder rb{ctx, 2};
-            rb.Push(ResultCode(ErrorModule::FS, ErrorDescription::InvalidOffset));
+            rb.Push(FileSys::ERROR_INVALID_OFFSET);
             return;
         }
 
@@ -272,8 +274,8 @@ public:
             {0, &IFileSystem::CreateFile, "CreateFile"},
             {1, &IFileSystem::DeleteFile, "DeleteFile"},
             {2, &IFileSystem::CreateDirectory, "CreateDirectory"},
-            {3, nullptr, "DeleteDirectory"},
-            {4, nullptr, "DeleteDirectoryRecursively"},
+            {3, &IFileSystem::DeleteDirectory, "DeleteDirectory"},
+            {4, &IFileSystem::DeleteDirectoryRecursively, "DeleteDirectoryRecursively"},
             {5, &IFileSystem::RenameFile, "RenameFile"},
             {6, nullptr, "RenameDirectory"},
             {7, &IFileSystem::GetEntryType, "GetEntryType"},
@@ -328,6 +330,30 @@ public:
         rb.Push(backend.CreateDirectory(name));
     }
 
+    void DeleteDirectory(Kernel::HLERequestContext& ctx) {
+        const IPC::RequestParser rp{ctx};
+
+        const auto file_buffer = ctx.ReadBuffer();
+        std::string name = Common::StringFromBuffer(file_buffer);
+
+        LOG_DEBUG(Service_FS, "called directory {}", name);
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(backend.DeleteDirectory(name));
+    }
+
+    void DeleteDirectoryRecursively(Kernel::HLERequestContext& ctx) {
+        const IPC::RequestParser rp{ctx};
+
+        const auto file_buffer = ctx.ReadBuffer();
+        std::string name = Common::StringFromBuffer(file_buffer);
+
+        LOG_DEBUG(Service_FS, "called directory {}", name);
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(backend.DeleteDirectoryRecursively(name));
+    }
+
     void RenameFile(Kernel::HLERequestContext& ctx) {
         IPC::RequestParser rp{ctx};
 
@@ -426,7 +452,147 @@ private:
     VfsDirectoryServiceWrapper backend;
 };
 
+class ISaveDataInfoReader final : public ServiceFramework<ISaveDataInfoReader> {
+public:
+    explicit ISaveDataInfoReader(FileSys::SaveDataSpaceId space)
+        : ServiceFramework("ISaveDataInfoReader") {
+        static const FunctionInfo functions[] = {
+            {0, &ISaveDataInfoReader::ReadSaveDataInfo, "ReadSaveDataInfo"},
+        };
+        RegisterHandlers(functions);
+
+        FindAllSaves(space);
+    }
+
+    void ReadSaveDataInfo(Kernel::HLERequestContext& ctx) {
+        // Calculate how many entries we can fit in the output buffer
+        const u64 count_entries = ctx.GetWriteBufferSize() / sizeof(SaveDataInfo);
+
+        // Cap at total number of entries.
+        const u64 actual_entries = std::min(count_entries, info.size() - next_entry_index);
+
+        // Determine data start and end
+        const auto* begin = reinterpret_cast<u8*>(info.data() + next_entry_index);
+        const auto* end = reinterpret_cast<u8*>(info.data() + next_entry_index + actual_entries);
+        const auto range_size = static_cast<std::size_t>(std::distance(begin, end));
+
+        next_entry_index += actual_entries;
+
+        // Write the data to memory
+        ctx.WriteBuffer(begin, range_size);
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u32>(static_cast<u32>(actual_entries));
+    }
+
+private:
+    static u64 stoull_be(std::string_view str) {
+        if (str.size() != 16)
+            return 0;
+
+        const auto bytes = Common::HexStringToArray<0x8>(str);
+        u64 out{};
+        std::memcpy(&out, bytes.data(), sizeof(u64));
+
+        return Common::swap64(out);
+    }
+
+    void FindAllSaves(FileSys::SaveDataSpaceId space) {
+        const auto save_root = OpenSaveDataSpace(space);
+        ASSERT(save_root.Succeeded());
+
+        for (const auto& type : (*save_root)->GetSubdirectories()) {
+            if (type->GetName() == "save") {
+                for (const auto& save_id : type->GetSubdirectories()) {
+                    for (const auto& user_id : save_id->GetSubdirectories()) {
+                        const auto save_id_numeric = stoull_be(save_id->GetName());
+                        auto user_id_numeric = Common::HexStringToArray<0x10>(user_id->GetName());
+                        std::reverse(user_id_numeric.begin(), user_id_numeric.end());
+
+                        if (save_id_numeric != 0) {
+                            // System Save Data
+                            info.emplace_back(SaveDataInfo{
+                                0,
+                                space,
+                                FileSys::SaveDataType::SystemSaveData,
+                                {},
+                                user_id_numeric,
+                                save_id_numeric,
+                                0,
+                                user_id->GetSize(),
+                                {},
+                            });
+
+                            continue;
+                        }
+
+                        for (const auto& title_id : user_id->GetSubdirectories()) {
+                            const auto device =
+                                std::all_of(user_id_numeric.begin(), user_id_numeric.end(),
+                                            [](u8 val) { return val == 0; });
+                            info.emplace_back(SaveDataInfo{
+                                0,
+                                space,
+                                device ? FileSys::SaveDataType::DeviceSaveData
+                                       : FileSys::SaveDataType::SaveData,
+                                {},
+                                user_id_numeric,
+                                save_id_numeric,
+                                stoull_be(title_id->GetName()),
+                                title_id->GetSize(),
+                                {},
+                            });
+                        }
+                    }
+                }
+            } else if (space == FileSys::SaveDataSpaceId::TemporaryStorage) {
+                // Temporary Storage
+                for (const auto& user_id : type->GetSubdirectories()) {
+                    for (const auto& title_id : user_id->GetSubdirectories()) {
+                        if (!title_id->GetFiles().empty() ||
+                            !title_id->GetSubdirectories().empty()) {
+                            auto user_id_numeric =
+                                Common::HexStringToArray<0x10>(user_id->GetName());
+                            std::reverse(user_id_numeric.begin(), user_id_numeric.end());
+
+                            info.emplace_back(SaveDataInfo{
+                                0,
+                                space,
+                                FileSys::SaveDataType::TemporaryStorage,
+                                {},
+                                user_id_numeric,
+                                stoull_be(type->GetName()),
+                                stoull_be(title_id->GetName()),
+                                title_id->GetSize(),
+                                {},
+                            });
+                        }
+                    }
+                }
+            }
+        }
+    }
+
+    struct SaveDataInfo {
+        u64_le save_id_unknown;
+        FileSys::SaveDataSpaceId space;
+        FileSys::SaveDataType type;
+        INSERT_PADDING_BYTES(0x6);
+        std::array<u8, 0x10> user_id;
+        u64_le save_id;
+        u64_le title_id;
+        u64_le save_image_size;
+        INSERT_PADDING_BYTES(0x28);
+    };
+    static_assert(sizeof(SaveDataInfo) == 0x60, "SaveDataInfo has incorrect size.");
+
+    std::vector<SaveDataInfo> info;
+    u64 next_entry_index = 0;
+};
+
 FSP_SRV::FSP_SRV() : ServiceFramework("fsp-srv") {
+    // clang-format off
     static const FunctionInfo functions[] = {
         {0, nullptr, "MountContent"},
         {1, &FSP_SRV::Initialize, "Initialize"},
@@ -460,7 +626,7 @@ FSP_SRV::FSP_SRV() : ServiceFramework("fsp-srv") {
         {58, nullptr, "ReadSaveDataFileSystemExtraData"},
         {59, nullptr, "WriteSaveDataFileSystemExtraData"},
         {60, nullptr, "OpenSaveDataInfoReader"},
-        {61, nullptr, "OpenSaveDataInfoReaderBySaveDataSpaceId"},
+        {61, &FSP_SRV::OpenSaveDataInfoReaderBySaveDataSpaceId, "OpenSaveDataInfoReaderBySaveDataSpaceId"},
         {62, nullptr, "OpenCacheStorageList"},
         {64, nullptr, "OpenSaveDataInternalStorageFileSystem"},
         {65, nullptr, "UpdateSaveDataMacForDebug"},
@@ -519,6 +685,7 @@ FSP_SRV::FSP_SRV() : ServiceFramework("fsp-srv") {
         {1009, nullptr, "GetAndClearMemoryReportInfo"},
         {1100, nullptr, "OverrideSaveDataTransferTokenSignVerificationKey"},
     };
+    // clang-format on
     RegisterHandlers(functions);
 }
 
@@ -577,7 +744,7 @@ void FSP_SRV::MountSaveData(Kernel::HLERequestContext& ctx) {
 
     if (dir.Failed()) {
         IPC::ResponseBuilder rb{ctx, 2, 0, 0};
-        rb.Push(ResultCode(ErrorModule::FS, FileSys::ErrCodes::TitleNotFound));
+        rb.Push(FileSys::ERROR_ENTITY_NOT_FOUND);
         return;
     }
 
@@ -593,6 +760,15 @@ void FSP_SRV::OpenReadOnlySaveDataFileSystem(Kernel::HLERequestContext& ctx) {
     MountSaveData(ctx);
 }
 
+void FSP_SRV::OpenSaveDataInfoReaderBySaveDataSpaceId(Kernel::HLERequestContext& ctx) {
+    IPC::RequestParser rp{ctx};
+    const auto space = rp.PopRaw<FileSys::SaveDataSpaceId>();
+
+    IPC::ResponseBuilder rb{ctx, 2, 0, 1};
+    rb.Push(RESULT_SUCCESS);
+    rb.PushIpcInterface<ISaveDataInfoReader>(std::make_shared<ISaveDataInfoReader>(space));
+}
+
 void FSP_SRV::GetGlobalAccessLogMode(Kernel::HLERequestContext& ctx) {
     LOG_WARNING(Service_FS, "(STUBBED) called");
 
@@ -630,6 +806,7 @@ void FSP_SRV::OpenDataStorageByDataId(Kernel::HLERequestContext& ctx) {
               static_cast<u8>(storage_id), unknown, title_id);
 
     auto data = OpenRomFS(title_id, storage_id, FileSys::ContentRecordType::Data);
+
     if (data.Failed()) {
         // TODO(DarkLordZach): Find the right error code to use here
         LOG_ERROR(Service_FS,
@@ -640,7 +817,9 @@ void FSP_SRV::OpenDataStorageByDataId(Kernel::HLERequestContext& ctx) {
         return;
     }
 
-    IStorage storage(std::move(data.Unwrap()));
+    FileSys::PatchManager pm{title_id};
+
+    IStorage storage(pm.PatchRomFS(std::move(data.Unwrap()), 0, FileSys::ContentRecordType::Data));
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(RESULT_SUCCESS);
@@ -657,7 +836,7 @@ void FSP_SRV::OpenRomStorage(Kernel::HLERequestContext& ctx) {
               static_cast<u8>(storage_id), title_id);
 
     IPC::ResponseBuilder rb{ctx, 2};
-    rb.Push(ResultCode(ErrorModule::FS, FileSys::ErrCodes::TitleNotFound));
+    rb.Push(FileSys::ERROR_ENTITY_NOT_FOUND);
 }
 
 } // namespace Service::FileSystem
diff --git a/src/core/hle/service/filesystem/fsp_srv.h b/src/core/hle/service/filesystem/fsp_srv.h
index 4aa0358cb..e7abec0a3 100644
--- a/src/core/hle/service/filesystem/fsp_srv.h
+++ b/src/core/hle/service/filesystem/fsp_srv.h
@@ -25,6 +25,7 @@ private:
     void CreateSaveData(Kernel::HLERequestContext& ctx);
     void MountSaveData(Kernel::HLERequestContext& ctx);
     void OpenReadOnlySaveDataFileSystem(Kernel::HLERequestContext& ctx);
+    void OpenSaveDataInfoReaderBySaveDataSpaceId(Kernel::HLERequestContext& ctx);
     void GetGlobalAccessLogMode(Kernel::HLERequestContext& ctx);
     void OpenDataStorageByCurrentProcess(Kernel::HLERequestContext& ctx);
     void OpenDataStorageByDataId(Kernel::HLERequestContext& ctx);
diff --git a/src/core/hle/service/hid/controllers/npad.cpp b/src/core/hle/service/hid/controllers/npad.cpp
index b06e65a77..205e4fd14 100644
--- a/src/core/hle/service/hid/controllers/npad.cpp
+++ b/src/core/hle/service/hid/controllers/npad.cpp
@@ -40,6 +40,29 @@ enum class JoystickId : std::size_t {
     Joystick_Right,
 };
 
+static std::size_t NPadIdToIndex(u32 npad_id) {
+    switch (npad_id) {
+    case 0:
+    case 1:
+    case 2:
+    case 3:
+    case 4:
+    case 5:
+    case 6:
+    case 7:
+        return npad_id;
+    case 8:
+    case NPAD_HANDHELD:
+        return 8;
+    case 9:
+    case NPAD_UNKNOWN:
+        return 9;
+    default:
+        UNIMPLEMENTED_MSG("Unknown npad id {}", npad_id);
+        return 0;
+    }
+}
+
 Controller_NPad::Controller_NPad() = default;
 Controller_NPad::~Controller_NPad() = default;
 
@@ -108,9 +131,10 @@ void Controller_NPad::OnInit() {
     styleset_changed_event =
         Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "npad:NpadStyleSetChanged");
 
-    if (!IsControllerActivated())
+    if (!IsControllerActivated()) {
         return;
-    std::size_t controller{};
+    }
+
     if (style.raw == 0) {
         // We want to support all controllers
         style.handheld.Assign(1);
@@ -287,10 +311,11 @@ void Controller_NPad::OnUpdate(u8* data, std::size_t data_len) {
         switch (controller_type) {
         case NPadControllerType::Handheld:
             handheld_entry.connection_status.raw = 0;
-            handheld_entry.connection_status.IsConnected.Assign(1);
-            if (!Settings::values.use_docked_mode) {
-                handheld_entry.connection_status.IsWired.Assign(1);
-            }
+            handheld_entry.connection_status.IsWired.Assign(1);
+            handheld_entry.connection_status.IsLeftJoyConnected.Assign(1);
+            handheld_entry.connection_status.IsRightJoyConnected.Assign(1);
+            handheld_entry.connection_status.IsLeftJoyWired.Assign(1);
+            handheld_entry.connection_status.IsRightJoyWired.Assign(1);
             handheld_entry.pad_states.raw = pad_state.raw;
             handheld_entry.l_stick = lstick_entry;
             handheld_entry.r_stick = rstick_entry;
@@ -309,6 +334,7 @@ void Controller_NPad::OnUpdate(u8* data, std::size_t data_len) {
             dual_entry.pad_states.raw = pad_state.raw;
             dual_entry.l_stick = lstick_entry;
             dual_entry.r_stick = rstick_entry;
+            break;
         case NPadControllerType::JoyLeft:
             left_entry.connection_status.raw = 0;
 
@@ -369,16 +395,30 @@ void Controller_NPad::SetSupportedNPadIdTypes(u8* data, std::size_t length) {
     supported_npad_id_types.clear();
     supported_npad_id_types.resize(length / sizeof(u32));
     std::memcpy(supported_npad_id_types.data(), data, length);
+    bool had_controller_update = false;
     for (std::size_t i = 0; i < connected_controllers.size(); i++) {
         auto& controller = connected_controllers[i];
         if (!controller.is_connected) {
             continue;
         }
         if (!IsControllerSupported(PREFERRED_CONTROLLER)) {
-            controller.type = DecideBestController(PREFERRED_CONTROLLER);
-            InitNewlyAddedControler(i);
+            const auto best_type = DecideBestController(PREFERRED_CONTROLLER);
+            const bool is_handheld = (best_type == NPadControllerType::Handheld ||
+                                      PREFERRED_CONTROLLER == NPadControllerType::Handheld);
+            if (is_handheld) {
+                controller.type = NPadControllerType::None;
+                controller.is_connected = false;
+                AddNewController(best_type);
+            } else {
+                controller.type = best_type;
+                InitNewlyAddedControler(i);
+            }
+            had_controller_update = true;
         }
     }
+    if (had_controller_update) {
+        styleset_changed_event->Signal();
+    }
 }
 
 void Controller_NPad::GetSupportedNpadIdTypes(u32* data, std::size_t max_length) {
@@ -391,8 +431,10 @@ std::size_t Controller_NPad::GetSupportedNPadIdTypesSize() const {
 }
 
 void Controller_NPad::SetHoldType(NpadHoldType joy_hold_type) {
+    styleset_changed_event->Signal();
     hold_type = joy_hold_type;
 }
+
 Controller_NPad::NpadHoldType Controller_NPad::GetHoldType() const {
     return hold_type;
 }
@@ -426,6 +468,9 @@ void Controller_NPad::VibrateController(const std::vector<u32>& controller_ids,
 }
 
 Kernel::SharedPtr<Kernel::Event> Controller_NPad::GetStyleSetChangedEvent() const {
+    // TODO(ogniK): Figure out the best time to signal this event. This event seems that it should
+    // be signalled at least once, and signaled after a new controller is connected?
+    styleset_changed_event->Signal();
     return styleset_changed_event;
 }
 
@@ -451,15 +496,11 @@ void Controller_NPad::AddNewController(NPadControllerType controller) {
 }
 
 void Controller_NPad::ConnectNPad(u32 npad_id) {
-    if (npad_id >= connected_controllers.size())
-        return;
-    connected_controllers[npad_id].is_connected = true;
+    connected_controllers[NPadIdToIndex(npad_id)].is_connected = true;
 }
 
 void Controller_NPad::DisconnectNPad(u32 npad_id) {
-    if (npad_id >= connected_controllers.size())
-        return;
-    connected_controllers[npad_id].is_connected = false;
+    connected_controllers[NPadIdToIndex(npad_id)].is_connected = false;
 }
 
 Controller_NPad::LedPattern Controller_NPad::GetLedPattern(u32 npad_id) {
diff --git a/src/core/hle/service/hid/hid.cpp b/src/core/hle/service/hid/hid.cpp
index a9aa9ec78..39631b14f 100644
--- a/src/core/hle/service/hid/hid.cpp
+++ b/src/core/hle/service/hid/hid.cpp
@@ -96,6 +96,8 @@ public:
         // TODO(shinyquagsire23): Other update callbacks? (accel, gyro?)
 
         CoreTiming::ScheduleEvent(pad_update_ticks, pad_update_event);
+
+        ReloadInputDevices();
     }
 
     void ActivateController(HidController controller) {
@@ -284,10 +286,10 @@ public:
             {519, nullptr, "GetPalmaOperationResult"},
             {520, nullptr, "ReadPalmaPlayLog"},
             {521, nullptr, "ResetPalmaPlayLog"},
-            {522, nullptr, "SetIsPalmaAllConnectable"},
+            {522, &Hid::SetIsPalmaAllConnectable, "SetIsPalmaAllConnectable"},
             {523, nullptr, "SetIsPalmaPairedConnectable"},
             {524, nullptr, "PairPalma"},
-            {525, nullptr, "SetPalmaBoostMode"},
+            {525, &Hid::SetPalmaBoostMode, "SetPalmaBoostMode"},
             {1000, nullptr, "SetNpadCommunicationMode"},
             {1001, nullptr, "GetNpadCommunicationMode"},
         };
@@ -594,6 +596,18 @@ private:
         rb.Push(RESULT_SUCCESS);
         LOG_WARNING(Service_HID, "(STUBBED) called");
     }
+
+    void SetIsPalmaAllConnectable(Kernel::HLERequestContext& ctx) {
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+        LOG_WARNING(Service_HID, "(STUBBED) called");
+    }
+
+    void SetPalmaBoostMode(Kernel::HLERequestContext& ctx) {
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+        LOG_WARNING(Service_HID, "(STUBBED) called");
+    }
 };
 
 class HidDbg final : public ServiceFramework<HidDbg> {
diff --git a/src/core/hle/service/ldr/ldr.cpp b/src/core/hle/service/ldr/ldr.cpp
index ec32faf15..b43f1f054 100644
--- a/src/core/hle/service/ldr/ldr.cpp
+++ b/src/core/hle/service/ldr/ldr.cpp
@@ -3,12 +3,51 @@
 // Refer to the license.txt file included.
 
 #include <memory>
+#include <fmt/format.h>
+#include <mbedtls/sha256.h>
 
+#include "common/alignment.h"
+#include "common/hex_util.h"
+#include "core/hle/ipc_helpers.h"
+#include "core/hle/kernel/process.h"
 #include "core/hle/service/ldr/ldr.h"
 #include "core/hle/service/service.h"
+#include "core/loader/nro.h"
 
 namespace Service::LDR {
 
+namespace ErrCodes {
+enum {
+    InvalidMemoryState = 51,
+    InvalidNRO = 52,
+    InvalidNRR = 53,
+    MissingNRRHash = 54,
+    MaximumNRO = 55,
+    MaximumNRR = 56,
+    AlreadyLoaded = 57,
+    InvalidAlignment = 81,
+    InvalidSize = 82,
+    InvalidNROAddress = 84,
+    InvalidNRRAddress = 85,
+    NotInitialized = 87,
+};
+}
+
+constexpr ResultCode ERROR_INVALID_MEMORY_STATE(ErrorModule::Loader, ErrCodes::InvalidMemoryState);
+constexpr ResultCode ERROR_INVALID_NRO(ErrorModule::Loader, ErrCodes::InvalidNRO);
+constexpr ResultCode ERROR_INVALID_NRR(ErrorModule::Loader, ErrCodes::InvalidNRR);
+constexpr ResultCode ERROR_MISSING_NRR_HASH(ErrorModule::Loader, ErrCodes::MissingNRRHash);
+constexpr ResultCode ERROR_MAXIMUM_NRO(ErrorModule::Loader, ErrCodes::MaximumNRO);
+constexpr ResultCode ERROR_MAXIMUM_NRR(ErrorModule::Loader, ErrCodes::MaximumNRR);
+constexpr ResultCode ERROR_ALREADY_LOADED(ErrorModule::Loader, ErrCodes::AlreadyLoaded);
+constexpr ResultCode ERROR_INVALID_ALIGNMENT(ErrorModule::Loader, ErrCodes::InvalidAlignment);
+constexpr ResultCode ERROR_INVALID_SIZE(ErrorModule::Loader, ErrCodes::InvalidSize);
+constexpr ResultCode ERROR_INVALID_NRO_ADDRESS(ErrorModule::Loader, ErrCodes::InvalidNROAddress);
+constexpr ResultCode ERROR_INVALID_NRR_ADDRESS(ErrorModule::Loader, ErrCodes::InvalidNRRAddress);
+constexpr ResultCode ERROR_NOT_INITIALIZED(ErrorModule::Loader, ErrCodes::NotInitialized);
+
+constexpr u64 MAXIMUM_LOADED_RO = 0x40;
+
 class DebugMonitor final : public ServiceFramework<DebugMonitor> {
 public:
     explicit DebugMonitor() : ServiceFramework{"ldr:dmnt"} {
@@ -59,16 +98,392 @@ public:
     explicit RelocatableObject() : ServiceFramework{"ldr:ro"} {
         // clang-format off
         static const FunctionInfo functions[] = {
-            {0, nullptr, "LoadNro"},
-            {1, nullptr, "UnloadNro"},
-            {2, nullptr, "LoadNrr"},
-            {3, nullptr, "UnloadNrr"},
-            {4, nullptr, "Initialize"},
+            {0, &RelocatableObject::LoadNro, "LoadNro"},
+            {1, &RelocatableObject::UnloadNro, "UnloadNro"},
+            {2, &RelocatableObject::LoadNrr, "LoadNrr"},
+            {3, &RelocatableObject::UnloadNrr, "UnloadNrr"},
+            {4, &RelocatableObject::Initialize, "Initialize"},
         };
         // clang-format on
 
         RegisterHandlers(functions);
     }
+
+    void LoadNrr(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        rp.Skip(2, false);
+        const VAddr nrr_addr{rp.Pop<VAddr>()};
+        const u64 nrr_size{rp.Pop<u64>()};
+
+        if (!initialized) {
+            LOG_ERROR(Service_LDR, "LDR:RO not initialized before use!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_NOT_INITIALIZED);
+            return;
+        }
+
+        if (nrr.size() >= MAXIMUM_LOADED_RO) {
+            LOG_ERROR(Service_LDR, "Loading new NRR would exceed the maximum number of loaded NRRs "
+                                   "(0x40)! Failing...");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_MAXIMUM_NRR);
+            return;
+        }
+
+        // NRR Address does not fall on 0x1000 byte boundary
+        if (!Common::Is4KBAligned(nrr_addr)) {
+            LOG_ERROR(Service_LDR, "NRR Address has invalid alignment (actual {:016X})!", nrr_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_ALIGNMENT);
+            return;
+        }
+
+        // NRR Size is zero or causes overflow
+        if (nrr_addr + nrr_size <= nrr_addr || nrr_size == 0 || !Common::Is4KBAligned(nrr_size)) {
+            LOG_ERROR(Service_LDR, "NRR Size is invalid! (nrr_address={:016X}, nrr_size={:016X})",
+                      nrr_addr, nrr_size);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_SIZE);
+            return;
+        }
+        // Read NRR data from memory
+        std::vector<u8> nrr_data(nrr_size);
+        Memory::ReadBlock(nrr_addr, nrr_data.data(), nrr_size);
+        NRRHeader header;
+        std::memcpy(&header, nrr_data.data(), sizeof(NRRHeader));
+
+        if (header.magic != Common::MakeMagic('N', 'R', 'R', '0')) {
+            LOG_ERROR(Service_LDR, "NRR did not have magic 'NRR0' (actual {:08X})!", header.magic);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_NRR);
+            return;
+        }
+
+        if (header.size != nrr_size) {
+            LOG_ERROR(Service_LDR,
+                      "NRR header reported size did not match LoadNrr parameter size! "
+                      "(header_size={:016X}, loadnrr_size={:016X})",
+                      header.size, nrr_size);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_SIZE);
+            return;
+        }
+
+        if (Core::CurrentProcess()->GetTitleID() != header.title_id) {
+            LOG_ERROR(Service_LDR,
+                      "Attempting to load NRR with title ID other than current process. (actual "
+                      "{:016X})!",
+                      header.title_id);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_NRR);
+            return;
+        }
+
+        std::vector<SHA256Hash> hashes;
+
+        // Copy all hashes in the NRR (specified by hash count/hash offset) into vector.
+        for (std::size_t i = header.hash_offset;
+             i < (header.hash_offset + (header.hash_count * sizeof(SHA256Hash))); i += 8) {
+            SHA256Hash hash;
+            std::memcpy(hash.data(), nrr_data.data() + i, sizeof(SHA256Hash));
+            hashes.emplace_back(hash);
+        }
+
+        nrr.insert_or_assign(nrr_addr, std::move(hashes));
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void UnloadNrr(Kernel::HLERequestContext& ctx) {
+        if (!initialized) {
+            LOG_ERROR(Service_LDR, "LDR:RO not initialized before use!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_NOT_INITIALIZED);
+            return;
+        }
+
+        IPC::RequestParser rp{ctx};
+        rp.Skip(2, false);
+        const auto nrr_addr{rp.Pop<VAddr>()};
+
+        if (!Common::Is4KBAligned(nrr_addr)) {
+            LOG_ERROR(Service_LDR, "NRR Address has invalid alignment (actual {:016X})!", nrr_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_ALIGNMENT);
+            return;
+        }
+
+        const auto iter = nrr.find(nrr_addr);
+        if (iter == nrr.end()) {
+            LOG_ERROR(Service_LDR,
+                      "Attempting to unload NRR which has not been loaded! (addr={:016X})",
+                      nrr_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_NRR_ADDRESS);
+            return;
+        }
+
+        nrr.erase(iter);
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void LoadNro(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        rp.Skip(2, false);
+        const VAddr nro_addr{rp.Pop<VAddr>()};
+        const u64 nro_size{rp.Pop<u64>()};
+        const VAddr bss_addr{rp.Pop<VAddr>()};
+        const u64 bss_size{rp.Pop<u64>()};
+
+        if (!initialized) {
+            LOG_ERROR(Service_LDR, "LDR:RO not initialized before use!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_NOT_INITIALIZED);
+            return;
+        }
+
+        if (nro.size() >= MAXIMUM_LOADED_RO) {
+            LOG_ERROR(Service_LDR, "Loading new NRO would exceed the maximum number of loaded NROs "
+                                   "(0x40)! Failing...");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_MAXIMUM_NRO);
+            return;
+        }
+
+        // NRO Address does not fall on 0x1000 byte boundary
+        if (!Common::Is4KBAligned(nro_addr)) {
+            LOG_ERROR(Service_LDR, "NRO Address has invalid alignment (actual {:016X})!", nro_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_ALIGNMENT);
+            return;
+        }
+
+        // NRO Size or BSS Size is zero or causes overflow
+        const auto nro_size_valid =
+            nro_size != 0 && nro_addr + nro_size > nro_addr && Common::Is4KBAligned(nro_size);
+        const auto bss_size_valid =
+            nro_size + bss_size >= nro_size && (bss_size == 0 || bss_addr + bss_size > bss_addr);
+
+        if (!nro_size_valid || !bss_size_valid) {
+            LOG_ERROR(Service_LDR,
+                      "NRO Size or BSS Size is invalid! (nro_address={:016X}, nro_size={:016X}, "
+                      "bss_address={:016X}, bss_size={:016X})",
+                      nro_addr, nro_size, bss_addr, bss_size);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_SIZE);
+            return;
+        }
+
+        // Read NRO data from memory
+        std::vector<u8> nro_data(nro_size);
+        Memory::ReadBlock(nro_addr, nro_data.data(), nro_size);
+
+        SHA256Hash hash{};
+        mbedtls_sha256(nro_data.data(), nro_data.size(), hash.data(), 0);
+
+        // NRO Hash is already loaded
+        if (std::any_of(nro.begin(), nro.end(), [&hash](const std::pair<VAddr, NROInfo>& info) {
+                return info.second.hash == hash;
+            })) {
+            LOG_ERROR(Service_LDR, "NRO is already loaded!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_ALREADY_LOADED);
+            return;
+        }
+
+        // NRO Hash is not in any loaded NRR
+        if (!IsValidNROHash(hash)) {
+            LOG_ERROR(Service_LDR,
+                      "NRO hash is not present in any currently loaded NRRs (hash={})!",
+                      Common::HexArrayToString(hash));
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_MISSING_NRR_HASH);
+            return;
+        }
+
+        NROHeader header;
+        std::memcpy(&header, nro_data.data(), sizeof(NROHeader));
+
+        if (!IsValidNRO(header, nro_size, bss_size)) {
+            LOG_ERROR(Service_LDR, "NRO was invalid!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_NRO);
+            return;
+        }
+
+        // Load NRO as new executable module
+        auto* process = Core::CurrentProcess();
+        auto& vm_manager = process->VMManager();
+        auto map_address = vm_manager.FindFreeRegion(nro_size + bss_size);
+
+        if (!map_address.Succeeded() ||
+            *map_address + nro_size + bss_size > vm_manager.GetAddressSpaceEndAddress()) {
+
+            LOG_ERROR(Service_LDR,
+                      "General error while allocation memory or no available memory to allocate!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_MEMORY_STATE);
+            return;
+        }
+
+        ASSERT(process->MirrorMemory(*map_address, nro_addr, nro_size,
+                                     Kernel::MemoryState::ModuleCodeStatic) == RESULT_SUCCESS);
+        ASSERT(process->UnmapMemory(nro_addr, 0, nro_size) == RESULT_SUCCESS);
+
+        if (bss_size > 0) {
+            ASSERT(process->MirrorMemory(*map_address + nro_size, bss_addr, bss_size,
+                                         Kernel::MemoryState::ModuleCodeStatic) == RESULT_SUCCESS);
+            ASSERT(process->UnmapMemory(bss_addr, 0, bss_size) == RESULT_SUCCESS);
+        }
+
+        vm_manager.ReprotectRange(*map_address, header.text_size,
+                                  Kernel::VMAPermission::ReadExecute);
+        vm_manager.ReprotectRange(*map_address + header.ro_offset, header.ro_size,
+                                  Kernel::VMAPermission::Read);
+        vm_manager.ReprotectRange(*map_address + header.rw_offset, header.rw_size,
+                                  Kernel::VMAPermission::ReadWrite);
+
+        Core::System::GetInstance().ArmInterface(0).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(1).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(2).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(3).ClearInstructionCache();
+
+        nro.insert_or_assign(*map_address, NROInfo{hash, nro_size + bss_size});
+
+        IPC::ResponseBuilder rb{ctx, 4};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push(*map_address);
+    }
+
+    void UnloadNro(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        rp.Skip(2, false);
+        const VAddr mapped_addr{rp.PopRaw<VAddr>()};
+        const VAddr heap_addr{rp.PopRaw<VAddr>()};
+
+        if (!initialized) {
+            LOG_ERROR(Service_LDR, "LDR:RO not initialized before use!");
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_NOT_INITIALIZED);
+            return;
+        }
+
+        if (!Common::Is4KBAligned(mapped_addr) || !Common::Is4KBAligned(heap_addr)) {
+            LOG_ERROR(Service_LDR,
+                      "NRO/BSS Address has invalid alignment (actual nro_addr={:016X}, "
+                      "bss_addr={:016X})!",
+                      mapped_addr, heap_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_ALIGNMENT);
+            return;
+        }
+
+        const auto iter = nro.find(mapped_addr);
+        if (iter == nro.end()) {
+            LOG_ERROR(Service_LDR,
+                      "The NRO attempting to unmap was not mapped or has an invalid address "
+                      "(actual {:016X})!",
+                      mapped_addr);
+            IPC::ResponseBuilder rb{ctx, 2};
+            rb.Push(ERROR_INVALID_NRO_ADDRESS);
+            return;
+        }
+
+        auto* process = Core::CurrentProcess();
+        auto& vm_manager = process->VMManager();
+        const auto& nro_size = iter->second.size;
+
+        ASSERT(process->MirrorMemory(heap_addr, mapped_addr, nro_size,
+                                     Kernel::MemoryState::ModuleCodeStatic) == RESULT_SUCCESS);
+        ASSERT(process->UnmapMemory(mapped_addr, 0, nro_size) == RESULT_SUCCESS);
+
+        Core::System::GetInstance().ArmInterface(0).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(1).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(2).ClearInstructionCache();
+        Core::System::GetInstance().ArmInterface(3).ClearInstructionCache();
+
+        nro.erase(iter);
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void Initialize(Kernel::HLERequestContext& ctx) {
+        initialized = true;
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+        LOG_WARNING(Service_LDR, "(STUBBED) called");
+    }
+
+private:
+    using SHA256Hash = std::array<u8, 0x20>;
+
+    struct NROHeader {
+        u32_le entrypoint_insn;
+        u32_le mod_offset;
+        INSERT_PADDING_WORDS(2);
+        u32_le magic;
+        INSERT_PADDING_WORDS(1);
+        u32_le nro_size;
+        INSERT_PADDING_WORDS(1);
+        u32_le text_offset;
+        u32_le text_size;
+        u32_le ro_offset;
+        u32_le ro_size;
+        u32_le rw_offset;
+        u32_le rw_size;
+        u32_le bss_size;
+        INSERT_PADDING_WORDS(1);
+        std::array<u8, 0x20> build_id;
+        INSERT_PADDING_BYTES(0x20);
+    };
+    static_assert(sizeof(NROHeader) == 0x80, "NROHeader has invalid size.");
+
+    struct NRRHeader {
+        u32_le magic;
+        INSERT_PADDING_BYTES(0x1C);
+        u64_le title_id_mask;
+        u64_le title_id_pattern;
+        std::array<u8, 0x100> modulus;
+        std::array<u8, 0x100> signature_1;
+        std::array<u8, 0x100> signature_2;
+        u64_le title_id;
+        u32_le size;
+        INSERT_PADDING_BYTES(4);
+        u32_le hash_offset;
+        u32_le hash_count;
+        INSERT_PADDING_BYTES(8);
+    };
+    static_assert(sizeof(NRRHeader) == 0x350, "NRRHeader has incorrect size.");
+
+    struct NROInfo {
+        SHA256Hash hash;
+        u64 size;
+    };
+
+    bool initialized = false;
+
+    std::map<VAddr, NROInfo> nro;
+    std::map<VAddr, std::vector<SHA256Hash>> nrr;
+
+    bool IsValidNROHash(const SHA256Hash& hash) {
+        return std::any_of(
+            nrr.begin(), nrr.end(), [&hash](const std::pair<VAddr, std::vector<SHA256Hash>>& p) {
+                return std::find(p.second.begin(), p.second.end(), hash) != p.second.end();
+            });
+    }
+
+    static bool IsValidNRO(const NROHeader& header, u64 nro_size, u64 bss_size) {
+        return header.magic == Common::MakeMagic('N', 'R', 'O', '0') &&
+               header.nro_size == nro_size && header.bss_size == bss_size &&
+               header.ro_offset == header.text_offset + header.text_size &&
+               header.rw_offset == header.ro_offset + header.ro_size &&
+               nro_size == header.rw_offset + header.rw_size &&
+               Common::Is4KBAligned(header.text_size) && Common::Is4KBAligned(header.ro_size) &&
+               Common::Is4KBAligned(header.rw_size);
+    }
 };
 
 void InstallInterfaces(SM::ServiceManager& sm) {
diff --git a/src/core/hle/service/nfc/nfc.cpp b/src/core/hle/service/nfc/nfc.cpp
index 8fec97db8..30e542542 100644
--- a/src/core/hle/service/nfc/nfc.cpp
+++ b/src/core/hle/service/nfc/nfc.cpp
@@ -10,12 +10,13 @@
 #include "core/hle/service/nfc/nfc.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sm/sm.h"
+#include "core/settings.h"
 
 namespace Service::NFC {
 
 class IAm final : public ServiceFramework<IAm> {
 public:
-    explicit IAm() : ServiceFramework{"IAm"} {
+    explicit IAm() : ServiceFramework{"NFC::IAm"} {
         // clang-format off
         static const FunctionInfo functions[] = {
             {0, nullptr, "Initialize"},
@@ -52,7 +53,7 @@ private:
 
 class MFIUser final : public ServiceFramework<MFIUser> {
 public:
-    explicit MFIUser() : ServiceFramework{"IUser"} {
+    explicit MFIUser() : ServiceFramework{"NFC::MFIUser"} {
         // clang-format off
         static const FunctionInfo functions[] = {
             {0, nullptr, "Initialize"},
@@ -100,13 +101,13 @@ private:
 
 class IUser final : public ServiceFramework<IUser> {
 public:
-    explicit IUser() : ServiceFramework{"IUser"} {
+    explicit IUser() : ServiceFramework{"NFC::IUser"} {
         // clang-format off
         static const FunctionInfo functions[] = {
-            {0, nullptr, "Initialize"},
-            {1, nullptr, "Finalize"},
-            {2, nullptr, "GetState"},
-            {3, nullptr, "IsNfcEnabled"},
+            {0, &IUser::InitializeOld, "InitializeOld"},
+            {1, &IUser::FinalizeOld, "FinalizeOld"},
+            {2, &IUser::GetStateOld, "GetStateOld"},
+            {3, &IUser::IsNfcEnabledOld, "IsNfcEnabledOld"},
             {400, nullptr, "Initialize"},
             {401, nullptr, "Finalize"},
             {402, nullptr, "GetState"},
@@ -130,11 +131,47 @@ public:
 
         RegisterHandlers(functions);
     }
+
+private:
+    enum class NfcStates : u32 {
+        Finalized = 6,
+    };
+
+    void InitializeOld(Kernel::HLERequestContext& ctx) {
+        IPC::ResponseBuilder rb{ctx, 2, 0};
+        rb.Push(RESULT_SUCCESS);
+
+        // We don't deal with hardware initialization so we can just stub this.
+        LOG_DEBUG(Service_NFC, "called");
+    }
+
+    void IsNfcEnabledOld(Kernel::HLERequestContext& ctx) {
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushRaw<u8>(Settings::values.enable_nfc);
+
+        LOG_DEBUG(Service_NFC, "IsNfcEnabledOld");
+    }
+
+    void GetStateOld(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_NFC, "(STUBBED) called");
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushEnum(NfcStates::Finalized); // TODO(ogniK): Figure out if this matches nfp
+    }
+
+    void FinalizeOld(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_NFC, "(STUBBED) called");
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
 };
 
 class NFC_U final : public ServiceFramework<NFC_U> {
 public:
-    explicit NFC_U() : ServiceFramework{"nfc:u"} {
+    explicit NFC_U() : ServiceFramework{"nfc:user"} {
         // clang-format off
         static const FunctionInfo functions[] = {
             {0, &NFC_U::CreateUserInterface, "CreateUserInterface"},
diff --git a/src/core/hle/service/nfp/nfp.cpp b/src/core/hle/service/nfp/nfp.cpp
index 39c0c1e63..1d6e7756f 100644
--- a/src/core/hle/service/nfp/nfp.cpp
+++ b/src/core/hle/service/nfp/nfp.cpp
@@ -2,56 +2,67 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <atomic>
+
 #include "common/logging/log.h"
 #include "core/core.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/kernel/event.h"
+#include "core/hle/lock.h"
 #include "core/hle/service/hid/hid.h"
 #include "core/hle/service/nfp/nfp.h"
 #include "core/hle/service/nfp/nfp_user.h"
 
 namespace Service::NFP {
 
+namespace ErrCodes {
+constexpr ResultCode ERR_TAG_FAILED(ErrorModule::NFP,
+                                    -1); // TODO(ogniK): Find the actual error code
+}
+
 Module::Interface::Interface(std::shared_ptr<Module> module, const char* name)
-    : ServiceFramework(name), module(std::move(module)) {}
+    : ServiceFramework(name), module(std::move(module)) {
+    auto& kernel = Core::System::GetInstance().Kernel();
+    nfc_tag_load =
+        Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "IUser:NFCTagDetected");
+}
 
 Module::Interface::~Interface() = default;
 
 class IUser final : public ServiceFramework<IUser> {
 public:
-    IUser() : ServiceFramework("IUser") {
+    IUser(Module::Interface& nfp_interface)
+        : ServiceFramework("NFP::IUser"), nfp_interface(nfp_interface) {
         static const FunctionInfo functions[] = {
             {0, &IUser::Initialize, "Initialize"},
-            {1, nullptr, "Finalize"},
+            {1, &IUser::Finalize, "Finalize"},
             {2, &IUser::ListDevices, "ListDevices"},
-            {3, nullptr, "StartDetection"},
-            {4, nullptr, "StopDetection"},
-            {5, nullptr, "Mount"},
-            {6, nullptr, "Unmount"},
-            {7, nullptr, "OpenApplicationArea"},
-            {8, nullptr, "GetApplicationArea"},
+            {3, &IUser::StartDetection, "StartDetection"},
+            {4, &IUser::StopDetection, "StopDetection"},
+            {5, &IUser::Mount, "Mount"},
+            {6, &IUser::Unmount, "Unmount"},
+            {7, &IUser::OpenApplicationArea, "OpenApplicationArea"},
+            {8, &IUser::GetApplicationArea, "GetApplicationArea"},
             {9, nullptr, "SetApplicationArea"},
             {10, nullptr, "Flush"},
             {11, nullptr, "Restore"},
             {12, nullptr, "CreateApplicationArea"},
-            {13, nullptr, "GetTagInfo"},
-            {14, nullptr, "GetRegisterInfo"},
-            {15, nullptr, "GetCommonInfo"},
-            {16, nullptr, "GetModelInfo"},
+            {13, &IUser::GetTagInfo, "GetTagInfo"},
+            {14, &IUser::GetRegisterInfo, "GetRegisterInfo"},
+            {15, &IUser::GetCommonInfo, "GetCommonInfo"},
+            {16, &IUser::GetModelInfo, "GetModelInfo"},
             {17, &IUser::AttachActivateEvent, "AttachActivateEvent"},
             {18, &IUser::AttachDeactivateEvent, "AttachDeactivateEvent"},
             {19, &IUser::GetState, "GetState"},
             {20, &IUser::GetDeviceState, "GetDeviceState"},
             {21, &IUser::GetNpadId, "GetNpadId"},
-            {22, nullptr, "GetApplicationArea2"},
+            {22, &IUser::GetApplicationAreaSize, "GetApplicationAreaSize"},
             {23, &IUser::AttachAvailabilityChangeEvent, "AttachAvailabilityChangeEvent"},
             {24, nullptr, "RecreateApplicationArea"},
         };
         RegisterHandlers(functions);
 
         auto& kernel = Core::System::GetInstance().Kernel();
-        activate_event =
-            Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "IUser:ActivateEvent");
         deactivate_event =
             Kernel::Event::Create(kernel, Kernel::ResetType::OneShot, "IUser:DeactivateEvent");
         availability_change_event = Kernel::Event::Create(kernel, Kernel::ResetType::OneShot,
@@ -59,6 +70,17 @@ public:
     }
 
 private:
+    struct TagInfo {
+        std::array<u8, 10> uuid;
+        u8 uuid_length; // TODO(ogniK): Figure out if this is actual the uuid length or does it
+                        // mean something else
+        INSERT_PADDING_BYTES(0x15);
+        u32_le protocol;
+        u32_le tag_type;
+        INSERT_PADDING_BYTES(0x2c);
+    };
+    static_assert(sizeof(TagInfo) == 0x54, "TagInfo is an invalid size");
+
     enum class State : u32 {
         NonInitialized = 0,
         Initialized = 1,
@@ -66,15 +88,40 @@ private:
 
     enum class DeviceState : u32 {
         Initialized = 0,
+        SearchingForTag = 1,
+        TagFound = 2,
+        TagRemoved = 3,
+        TagNearby = 4,
+        Unknown5 = 5,
+        Finalized = 6
+    };
+
+    struct CommonInfo {
+        u16_be last_write_year;
+        u8 last_write_month;
+        u8 last_write_day;
+        u16_be write_counter;
+        u16_be version;
+        u32_be application_area_size;
+        INSERT_PADDING_BYTES(0x34);
     };
+    static_assert(sizeof(CommonInfo) == 0x40, "CommonInfo is an invalid size");
 
     void Initialize(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_NFP, "(STUBBED) called");
+        IPC::ResponseBuilder rb{ctx, 2, 0};
+        rb.Push(RESULT_SUCCESS);
 
         state = State::Initialized;
 
-        IPC::ResponseBuilder rb{ctx, 2};
+        LOG_DEBUG(Service_NFC, "called");
+    }
+
+    void GetState(Kernel::HLERequestContext& ctx) {
+        IPC::ResponseBuilder rb{ctx, 3, 0};
         rb.Push(RESULT_SUCCESS);
+        rb.PushRaw<u32>(static_cast<u32>(state));
+
+        LOG_DEBUG(Service_NFC, "called");
     }
 
     void ListDevices(Kernel::HLERequestContext& ctx) {
@@ -83,80 +130,219 @@ private:
 
         ctx.WriteBuffer(&device_handle, sizeof(device_handle));
 
-        LOG_WARNING(Service_NFP, "(STUBBED) called, array_size={}", array_size);
+        LOG_DEBUG(Service_NFP, "called, array_size={}", array_size);
 
         IPC::ResponseBuilder rb{ctx, 3};
         rb.Push(RESULT_SUCCESS);
-        rb.Push<u32>(0);
+        rb.Push<u32>(1);
     }
 
-    void AttachActivateEvent(Kernel::HLERequestContext& ctx) {
+    void GetNpadId(Kernel::HLERequestContext& ctx) {
         IPC::RequestParser rp{ctx};
         const u64 dev_handle = rp.Pop<u64>();
-        LOG_WARNING(Service_NFP, "(STUBBED) called, dev_handle=0x{:X}", dev_handle);
+        LOG_DEBUG(Service_NFP, "called, dev_handle=0x{:X}", dev_handle);
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u32>(npad_id);
+    }
 
+    void AttachActivateEvent(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        const u64 dev_handle = rp.Pop<u64>();
+        LOG_DEBUG(Service_NFP, "called, dev_handle=0x{:X}", dev_handle);
         IPC::ResponseBuilder rb{ctx, 2, 1};
         rb.Push(RESULT_SUCCESS);
-        rb.PushCopyObjects(activate_event);
+        rb.PushCopyObjects(nfp_interface.GetNFCEvent());
+        has_attached_handle = true;
     }
 
     void AttachDeactivateEvent(Kernel::HLERequestContext& ctx) {
         IPC::RequestParser rp{ctx};
         const u64 dev_handle = rp.Pop<u64>();
-        LOG_WARNING(Service_NFP, "(STUBBED) called, dev_handle=0x{:X}", dev_handle);
+        LOG_DEBUG(Service_NFP, "called, dev_handle=0x{:X}", dev_handle);
 
         IPC::ResponseBuilder rb{ctx, 2, 1};
         rb.Push(RESULT_SUCCESS);
         rb.PushCopyObjects(deactivate_event);
     }
 
-    void GetState(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_NFP, "(STUBBED) called");
-        IPC::ResponseBuilder rb{ctx, 3};
+    void StopDetection(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_NFP, "called");
+        switch (device_state) {
+        case DeviceState::TagFound:
+        case DeviceState::TagNearby:
+            deactivate_event->Signal();
+            device_state = DeviceState::Initialized;
+            break;
+        case DeviceState::SearchingForTag:
+        case DeviceState::TagRemoved:
+            device_state = DeviceState::Initialized;
+            break;
+        }
+        IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(RESULT_SUCCESS);
-        rb.Push<u32>(static_cast<u32>(state));
     }
 
     void GetDeviceState(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_NFP, "(STUBBED) called");
+        LOG_DEBUG(Service_NFP, "called");
+        auto nfc_event = nfp_interface.GetNFCEvent();
+        if (!nfc_event->ShouldWait(Kernel::GetCurrentThread()) && !has_attached_handle) {
+            device_state = DeviceState::TagFound;
+            nfc_event->Clear();
+        }
+
         IPC::ResponseBuilder rb{ctx, 3};
         rb.Push(RESULT_SUCCESS);
         rb.Push<u32>(static_cast<u32>(device_state));
     }
 
-    void GetNpadId(Kernel::HLERequestContext& ctx) {
-        IPC::RequestParser rp{ctx};
-        const u64 dev_handle = rp.Pop<u64>();
-        LOG_WARNING(Service_NFP, "(STUBBED) called, dev_handle=0x{:X}", dev_handle);
-        IPC::ResponseBuilder rb{ctx, 3};
+    void StartDetection(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_NFP, "called");
+
+        if (device_state == DeviceState::Initialized || device_state == DeviceState::TagRemoved) {
+            device_state = DeviceState::SearchingForTag;
+        }
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void GetTagInfo(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_NFP, "called");
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        auto amiibo = nfp_interface.GetAmiiboBuffer();
+        TagInfo tag_info{};
+        tag_info.uuid = amiibo.uuid;
+        tag_info.uuid_length = static_cast<u8>(tag_info.uuid.size());
+
+        tag_info.protocol = 1; // TODO(ogniK): Figure out actual values
+        tag_info.tag_type = 2;
+        ctx.WriteBuffer(&tag_info, sizeof(TagInfo));
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void Mount(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_NFP, "called");
+
+        device_state = DeviceState::TagNearby;
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void GetModelInfo(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_NFP, "called");
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        auto amiibo = nfp_interface.GetAmiiboBuffer();
+        ctx.WriteBuffer(&amiibo.model_info, sizeof(amiibo.model_info));
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void Unmount(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_NFP, "called");
+
+        device_state = DeviceState::TagFound;
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void Finalize(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_NFP, "called");
+
+        device_state = DeviceState::Finalized;
+
+        IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(RESULT_SUCCESS);
-        rb.Push<u32>(npad_id);
     }
 
     void AttachAvailabilityChangeEvent(Kernel::HLERequestContext& ctx) {
-        IPC::RequestParser rp{ctx};
-        const u64 dev_handle = rp.Pop<u64>();
-        LOG_WARNING(Service_NFP, "(STUBBED) called, dev_handle=0x{:X}", dev_handle);
+        LOG_WARNING(Service_NFP, "(STUBBED) called");
 
         IPC::ResponseBuilder rb{ctx, 2, 1};
         rb.Push(RESULT_SUCCESS);
         rb.PushCopyObjects(availability_change_event);
     }
 
-    const u64 device_handle{0xDEAD};
-    const u32 npad_id{0}; // This is the first player controller id
+    void GetRegisterInfo(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_NFP, "(STUBBED) called");
+
+        // TODO(ogniK): Pull Mii and owner data from amiibo
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void GetCommonInfo(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_NFP, "(STUBBED) called");
+
+        // TODO(ogniK): Pull common information from amiibo
+
+        CommonInfo common_info{};
+        common_info.application_area_size = 0;
+        ctx.WriteBuffer(&common_info, sizeof(CommonInfo));
+
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void OpenApplicationArea(Kernel::HLERequestContext& ctx) {
+        LOG_DEBUG(Service_NFP, "called");
+        // We don't need to worry about this since we can just open the file
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_SUCCESS);
+    }
+
+    void GetApplicationAreaSize(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_NFP, "(STUBBED) called");
+        // We don't need to worry about this since we can just open the file
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushRaw<u32>(0); // This is from the GetCommonInfo stub
+    }
+
+    void GetApplicationArea(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_NFP, "(STUBBED) called");
+
+        // TODO(ogniK): Pull application area from amiibo
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushRaw<u32>(0); // This is from the GetCommonInfo stub
+    }
+
+    bool has_attached_handle{};
+    const u64 device_handle{Common::MakeMagic('Y', 'U', 'Z', 'U')};
+    const u32 npad_id{0}; // Player 1 controller
     State state{State::NonInitialized};
     DeviceState device_state{DeviceState::Initialized};
-    Kernel::SharedPtr<Kernel::Event> activate_event;
     Kernel::SharedPtr<Kernel::Event> deactivate_event;
     Kernel::SharedPtr<Kernel::Event> availability_change_event;
+    const Module::Interface& nfp_interface;
 };
 
 void Module::Interface::CreateUserInterface(Kernel::HLERequestContext& ctx) {
     LOG_DEBUG(Service_NFP, "called");
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
     rb.Push(RESULT_SUCCESS);
-    rb.PushIpcInterface<IUser>();
+    rb.PushIpcInterface<IUser>(*this);
+}
+
+bool Module::Interface::LoadAmiibo(const std::vector<u8>& buffer) {
+    std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
+    if (buffer.size() < sizeof(AmiiboFile)) {
+        return false;
+    }
+
+    std::memcpy(&amiibo, buffer.data(), sizeof(amiibo));
+    nfc_tag_load->Signal();
+    return true;
+}
+const Kernel::SharedPtr<Kernel::Event>& Module::Interface::GetNFCEvent() const {
+    return nfc_tag_load;
+}
+const Module::Interface::AmiiboFile& Module::Interface::GetAmiiboBuffer() const {
+    return amiibo;
 }
 
 void InstallInterfaces(SM::ServiceManager& service_manager) {
diff --git a/src/core/hle/service/nfp/nfp.h b/src/core/hle/service/nfp/nfp.h
index 77df343c4..5c0ae8a54 100644
--- a/src/core/hle/service/nfp/nfp.h
+++ b/src/core/hle/service/nfp/nfp.h
@@ -4,6 +4,9 @@
 
 #pragma once
 
+#include <array>
+#include <vector>
+#include "core/hle/kernel/event.h"
 #include "core/hle/service/service.h"
 
 namespace Service::NFP {
@@ -15,7 +18,27 @@ public:
         explicit Interface(std::shared_ptr<Module> module, const char* name);
         ~Interface() override;
 
+        struct ModelInfo {
+            std::array<u8, 0x8> amiibo_identification_block;
+            INSERT_PADDING_BYTES(0x38);
+        };
+        static_assert(sizeof(ModelInfo) == 0x40, "ModelInfo is an invalid size");
+
+        struct AmiiboFile {
+            std::array<u8, 10> uuid;
+            INSERT_PADDING_BYTES(0x4a);
+            ModelInfo model_info;
+        };
+        static_assert(sizeof(AmiiboFile) == 0x94, "AmiiboFile is an invalid size");
+
         void CreateUserInterface(Kernel::HLERequestContext& ctx);
+        bool LoadAmiibo(const std::vector<u8>& buffer);
+        const Kernel::SharedPtr<Kernel::Event>& GetNFCEvent() const;
+        const AmiiboFile& GetAmiiboBuffer() const;
+
+    private:
+        Kernel::SharedPtr<Kernel::Event> nfc_tag_load{};
+        AmiiboFile amiibo{};
 
     protected:
         std::shared_ptr<Module> module;
diff --git a/src/core/hle/service/ns/ns.cpp b/src/core/hle/service/ns/ns.cpp
index 07c1381fe..1d2978f24 100644
--- a/src/core/hle/service/ns/ns.cpp
+++ b/src/core/hle/service/ns/ns.cpp
@@ -2,6 +2,9 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include "common/logging/log.h"
+#include "core/file_sys/control_metadata.h"
+#include "core/file_sys/patch_manager.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/service/ns/ns.h"
@@ -118,7 +121,7 @@ public:
             {305, nullptr, "TerminateSystemApplet"},
             {306, nullptr, "LaunchOverlayApplet"},
             {307, nullptr, "TerminateOverlayApplet"},
-            {400, nullptr, "GetApplicationControlData"},
+            {400, &IApplicationManagerInterface::GetApplicationControlData, "GetApplicationControlData"},
             {401, nullptr, "InvalidateAllApplicationControlCache"},
             {402, nullptr, "RequestDownloadApplicationControlData"},
             {403, nullptr, "GetMaxApplicationControlCacheCount"},
@@ -243,6 +246,65 @@ public:
 
         RegisterHandlers(functions);
     }
+
+    void GetApplicationControlData(Kernel::HLERequestContext& ctx) {
+        IPC::RequestParser rp{ctx};
+        const auto flag = rp.PopRaw<u64>();
+        LOG_DEBUG(Service_NS, "called with flag={:016X}", flag);
+
+        const auto title_id = rp.PopRaw<u64>();
+
+        const auto size = ctx.GetWriteBufferSize();
+
+        const FileSys::PatchManager pm{title_id};
+        const auto control = pm.GetControlMetadata();
+
+        std::vector<u8> out;
+
+        if (control.first != nullptr) {
+            if (size < 0x4000) {
+                LOG_ERROR(Service_NS,
+                          "output buffer is too small! (actual={:016X}, expected_min=0x4000)",
+                          size);
+                IPC::ResponseBuilder rb{ctx, 2};
+                // TODO(DarkLordZach): Find a better error code for this.
+                rb.Push(ResultCode(-1));
+                return;
+            }
+
+            out.resize(0x4000);
+            const auto bytes = control.first->GetRawBytes();
+            std::memcpy(out.data(), bytes.data(), bytes.size());
+        } else {
+            LOG_WARNING(Service_NS, "missing NACP data for title_id={:016X}, defaulting to zeros.",
+                        title_id);
+            out.resize(std::min<u64>(0x4000, size));
+        }
+
+        if (control.second != nullptr) {
+            if (size < 0x4000 + control.second->GetSize()) {
+                LOG_ERROR(Service_NS,
+                          "output buffer is too small! (actual={:016X}, expected_min={:016X})",
+                          size, 0x4000 + control.second->GetSize());
+                IPC::ResponseBuilder rb{ctx, 2};
+                // TODO(DarkLordZach): Find a better error code for this.
+                rb.Push(ResultCode(-1));
+                return;
+            }
+
+            out.resize(0x4000 + control.second->GetSize());
+            control.second->Read(out.data() + 0x4000, control.second->GetSize());
+        } else {
+            LOG_WARNING(Service_NS, "missing icon data for title_id={:016X}, defaulting to zeros.",
+                        title_id);
+        }
+
+        ctx.WriteBuffer(out);
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.Push<u32>(static_cast<u32>(out.size()));
+    }
 };
 
 class IApplicationVersionInterface final : public ServiceFramework<IApplicationVersionInterface> {
diff --git a/src/core/hle/service/ns/pl_u.cpp b/src/core/hle/service/ns/pl_u.cpp
index 44accecb7..1066bf505 100644
--- a/src/core/hle/service/ns/pl_u.cpp
+++ b/src/core/hle/service/ns/pl_u.cpp
@@ -351,6 +351,14 @@ void PL_U::GetSharedFontInOrderOfPriority(Kernel::HLERequestContext& ctx) {
         font_sizes.push_back(region.size);
     }
 
+    // Resize buffers if game requests smaller size output.
+    font_codes.resize(
+        std::min<std::size_t>(font_codes.size(), ctx.GetWriteBufferSize(0) / sizeof(u32)));
+    font_offsets.resize(
+        std::min<std::size_t>(font_offsets.size(), ctx.GetWriteBufferSize(1) / sizeof(u32)));
+    font_sizes.resize(
+        std::min<std::size_t>(font_sizes.size(), ctx.GetWriteBufferSize(2) / sizeof(u32)));
+
     ctx.WriteBuffer(font_codes, 0);
     ctx.WriteBuffer(font_offsets, 1);
     ctx.WriteBuffer(font_sizes, 2);
diff --git a/src/core/hle/service/nvflinger/buffer_queue.cpp b/src/core/hle/service/nvflinger/buffer_queue.cpp
index fd98d541d..630ebbfc7 100644
--- a/src/core/hle/service/nvflinger/buffer_queue.cpp
+++ b/src/core/hle/service/nvflinger/buffer_queue.cpp
@@ -31,7 +31,7 @@ void BufferQueue::SetPreallocatedBuffer(u32 slot, const IGBPBuffer& igbp_buffer)
     buffer_wait_event->Signal();
 }
 
-boost::optional<u32> BufferQueue::DequeueBuffer(u32 width, u32 height) {
+std::optional<u32> BufferQueue::DequeueBuffer(u32 width, u32 height) {
     auto itr = std::find_if(queue.begin(), queue.end(), [&](const Buffer& buffer) {
         // Only consider free buffers. Buffers become free once again after they've been Acquired
         // and Released by the compositor, see the NVFlinger::Compose method.
@@ -44,7 +44,7 @@ boost::optional<u32> BufferQueue::DequeueBuffer(u32 width, u32 height) {
     });
 
     if (itr == queue.end()) {
-        return boost::none;
+        return {};
     }
 
     itr->status = Buffer::Status::Dequeued;
@@ -70,12 +70,12 @@ void BufferQueue::QueueBuffer(u32 slot, BufferTransformFlags transform,
     itr->crop_rect = crop_rect;
 }
 
-boost::optional<const BufferQueue::Buffer&> BufferQueue::AcquireBuffer() {
+std::optional<std::reference_wrapper<const BufferQueue::Buffer>> BufferQueue::AcquireBuffer() {
     auto itr = std::find_if(queue.begin(), queue.end(), [](const Buffer& buffer) {
         return buffer.status == Buffer::Status::Queued;
     });
     if (itr == queue.end())
-        return boost::none;
+        return {};
     itr->status = Buffer::Status::Acquired;
     return *itr;
 }
diff --git a/src/core/hle/service/nvflinger/buffer_queue.h b/src/core/hle/service/nvflinger/buffer_queue.h
index 50b767732..8cff5eb71 100644
--- a/src/core/hle/service/nvflinger/buffer_queue.h
+++ b/src/core/hle/service/nvflinger/buffer_queue.h
@@ -4,8 +4,9 @@
 
 #pragma once
 
+#include <optional>
 #include <vector>
-#include <boost/optional.hpp>
+
 #include "common/common_funcs.h"
 #include "common/math_util.h"
 #include "common/swap.h"
@@ -57,9 +58,9 @@ public:
         /// Rotate source image 90 degrees clockwise
         Rotate90 = 0x04,
         /// Rotate source image 180 degrees
-        Roate180 = 0x03,
+        Rotate180 = 0x03,
         /// Rotate source image 270 degrees clockwise
-        Roate270 = 0x07,
+        Rotate270 = 0x07,
     };
 
     struct Buffer {
@@ -73,11 +74,11 @@ public:
     };
 
     void SetPreallocatedBuffer(u32 slot, const IGBPBuffer& igbp_buffer);
-    boost::optional<u32> DequeueBuffer(u32 width, u32 height);
+    std::optional<u32> DequeueBuffer(u32 width, u32 height);
     const IGBPBuffer& RequestBuffer(u32 slot) const;
     void QueueBuffer(u32 slot, BufferTransformFlags transform,
                      const MathUtil::Rectangle<int>& crop_rect);
-    boost::optional<const Buffer&> AcquireBuffer();
+    std::optional<std::reference_wrapper<const Buffer>> AcquireBuffer();
     void ReleaseBuffer(u32 slot);
     u32 Query(QueryType type);
 
diff --git a/src/core/hle/service/nvflinger/nvflinger.cpp b/src/core/hle/service/nvflinger/nvflinger.cpp
index d47b6f659..214e6d1b3 100644
--- a/src/core/hle/service/nvflinger/nvflinger.cpp
+++ b/src/core/hle/service/nvflinger/nvflinger.cpp
@@ -3,7 +3,7 @@
 // Refer to the license.txt file included.
 
 #include <algorithm>
-#include <boost/optional.hpp>
+#include <optional>
 
 #include "common/alignment.h"
 #include "common/assert.h"
@@ -134,7 +134,7 @@ void NVFlinger::Compose() {
 
         MicroProfileFlip();
 
-        if (buffer == boost::none) {
+        if (!buffer) {
             auto& system_instance = Core::System::GetInstance();
 
             // There was no queued buffer to draw, render previous frame
@@ -143,7 +143,7 @@ void NVFlinger::Compose() {
             continue;
         }
 
-        auto& igbp_buffer = buffer->igbp_buffer;
+        auto& igbp_buffer = buffer->get().igbp_buffer;
 
         // Now send the buffer to the GPU for drawing.
         // TODO(Subv): Support more than just disp0. The display device selection is probably based
@@ -152,10 +152,10 @@ 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,
-                     buffer->crop_rect);
+                     igbp_buffer.width, igbp_buffer.height, igbp_buffer.stride,
+                     buffer->get().transform, buffer->get().crop_rect);
 
-        buffer_queue->ReleaseBuffer(buffer->slot);
+        buffer_queue->ReleaseBuffer(buffer->get().slot);
     }
 }
 
diff --git a/src/core/hle/service/service.cpp b/src/core/hle/service/service.cpp
index a4cf45267..1ec340466 100644
--- a/src/core/hle/service/service.cpp
+++ b/src/core/hle/service/service.cpp
@@ -80,8 +80,8 @@ namespace Service {
  * Creates a function string for logging, complete with the name (or header code, depending
  * on what's passed in) the port name, and all the cmd_buff arguments.
  */
-static std::string MakeFunctionString(const char* name, const char* port_name,
-                                      const u32* cmd_buff) {
+[[maybe_unused]] static std::string MakeFunctionString(const char* name, const char* port_name,
+                                                       const u32* cmd_buff) {
     // Number of params == bits 0-5 + bits 6-11
     int num_params = (cmd_buff[0] & 0x3F) + ((cmd_buff[0] >> 6) & 0x3F);
 
diff --git a/src/core/hle/service/spl/module.cpp b/src/core/hle/service/spl/module.cpp
index 44a6717d0..b2de2a818 100644
--- a/src/core/hle/service/spl/module.cpp
+++ b/src/core/hle/service/spl/module.cpp
@@ -3,18 +3,23 @@
 // Refer to the license.txt file included.
 
 #include <algorithm>
+#include <chrono>
 #include <cstdlib>
+#include <ctime>
+#include <functional>
 #include <vector>
 #include "common/logging/log.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/service/spl/csrng.h"
 #include "core/hle/service/spl/module.h"
 #include "core/hle/service/spl/spl.h"
+#include "core/settings.h"
 
 namespace Service::SPL {
 
 Module::Interface::Interface(std::shared_ptr<Module> module, const char* name)
-    : ServiceFramework(name), module(std::move(module)) {}
+    : ServiceFramework(name), module(std::move(module)),
+      rng(Settings::values.rng_seed.value_or(std::time(nullptr))) {}
 
 Module::Interface::~Interface() = default;
 
@@ -23,8 +28,9 @@ void Module::Interface::GetRandomBytes(Kernel::HLERequestContext& ctx) {
 
     std::size_t size = ctx.GetWriteBufferSize();
 
+    std::uniform_int_distribution<u16> distribution(0, std::numeric_limits<u8>::max());
     std::vector<u8> data(size);
-    std::generate(data.begin(), data.end(), std::rand);
+    std::generate(data.begin(), data.end(), [&] { return static_cast<u8>(distribution(rng)); });
 
     ctx.WriteBuffer(data);
 
diff --git a/src/core/hle/service/spl/module.h b/src/core/hle/service/spl/module.h
index 48fda6099..afa1f0295 100644
--- a/src/core/hle/service/spl/module.h
+++ b/src/core/hle/service/spl/module.h
@@ -4,6 +4,7 @@
 
 #pragma once
 
+#include <random>
 #include "core/hle/service/service.h"
 
 namespace Service::SPL {
@@ -19,6 +20,9 @@ public:
 
     protected:
         std::shared_ptr<Module> module;
+
+    private:
+        std::mt19937 rng;
     };
 };
 
diff --git a/src/core/hle/service/time/interface.cpp b/src/core/hle/service/time/interface.cpp
index 18a5d71d5..b3a196f65 100644
--- a/src/core/hle/service/time/interface.cpp
+++ b/src/core/hle/service/time/interface.cpp
@@ -21,9 +21,10 @@ Time::Time(std::shared_ptr<Module> time, const char* name)
         {102, nullptr, "GetStandardUserSystemClockInitialYear"},
         {200, nullptr, "IsStandardNetworkSystemClockAccuracySufficient"},
         {300, nullptr, "CalculateMonotonicSystemClockBaseTimePoint"},
-        {400, nullptr, "GetClockSnapshot"},
+        {400, &Time::GetClockSnapshot, "GetClockSnapshot"},
         {401, nullptr, "GetClockSnapshotFromSystemClockContext"},
-        {500, nullptr, "CalculateStandardUserSystemClockDifferenceByUser"},
+        {500, &Time::CalculateStandardUserSystemClockDifferenceByUser,
+         "CalculateStandardUserSystemClockDifferenceByUser"},
         {501, nullptr, "CalculateSpanBetween"},
     };
     RegisterHandlers(functions);
diff --git a/src/core/hle/service/time/time.cpp b/src/core/hle/service/time/time.cpp
index 28fd8debc..e561a0c52 100644
--- a/src/core/hle/service/time/time.cpp
+++ b/src/core/hle/service/time/time.cpp
@@ -15,6 +15,44 @@
 
 namespace Service::Time {
 
+static void PosixToCalendar(u64 posix_time, CalendarTime& calendar_time,
+                            CalendarAdditionalInfo& additional_info,
+                            [[maybe_unused]] const TimeZoneRule& /*rule*/) {
+    const std::time_t time(posix_time);
+    const std::tm* tm = std::localtime(&time);
+    if (tm == nullptr) {
+        calendar_time = {};
+        additional_info = {};
+        return;
+    }
+    calendar_time.year = tm->tm_year + 1900;
+    calendar_time.month = tm->tm_mon + 1;
+    calendar_time.day = tm->tm_mday;
+    calendar_time.hour = tm->tm_hour;
+    calendar_time.minute = tm->tm_min;
+    calendar_time.second = tm->tm_sec;
+
+    additional_info.day_of_week = tm->tm_wday;
+    additional_info.day_of_year = tm->tm_yday;
+    std::memcpy(additional_info.name.data(), "UTC", sizeof("UTC"));
+    additional_info.utc_offset = 0;
+}
+
+static u64 CalendarToPosix(const CalendarTime& calendar_time,
+                           [[maybe_unused]] const TimeZoneRule& /*rule*/) {
+    std::tm time{};
+    time.tm_year = calendar_time.year - 1900;
+    time.tm_mon = calendar_time.month - 1;
+    time.tm_mday = calendar_time.day;
+
+    time.tm_hour = calendar_time.hour;
+    time.tm_min = calendar_time.minute;
+    time.tm_sec = calendar_time.second;
+
+    std::time_t epoch_time = std::mktime(&time);
+    return static_cast<u64>(epoch_time);
+}
+
 class ISystemClock final : public ServiceFramework<ISystemClock> {
 public:
     ISystemClock() : ServiceFramework("ISystemClock") {
@@ -80,8 +118,8 @@ public:
             {5, nullptr, "GetTimeZoneRuleVersion"},
             {100, &ITimeZoneService::ToCalendarTime, "ToCalendarTime"},
             {101, &ITimeZoneService::ToCalendarTimeWithMyRule, "ToCalendarTimeWithMyRule"},
-            {201, nullptr, "ToPosixTime"},
-            {202, nullptr, "ToPosixTimeWithMyRule"},
+            {201, &ITimeZoneService::ToPosixTime, "ToPosixTime"},
+            {202, &ITimeZoneService::ToPosixTimeWithMyRule, "ToPosixTimeWithMyRule"},
         };
         RegisterHandlers(functions);
     }
@@ -151,24 +189,29 @@ private:
         rb.PushRaw(additional_info);
     }
 
-    void PosixToCalendar(u64 posix_time, CalendarTime& calendar_time,
-                         CalendarAdditionalInfo& additional_info, const TimeZoneRule& /*rule*/) {
-        std::time_t t(posix_time);
-        std::tm* tm = std::localtime(&t);
-        if (!tm) {
-            return;
-        }
-        calendar_time.year = tm->tm_year + 1900;
-        calendar_time.month = tm->tm_mon + 1;
-        calendar_time.day = tm->tm_mday;
-        calendar_time.hour = tm->tm_hour;
-        calendar_time.minute = tm->tm_min;
-        calendar_time.second = tm->tm_sec;
-
-        additional_info.day_of_week = tm->tm_wday;
-        additional_info.day_of_year = tm->tm_yday;
-        std::memcpy(additional_info.name.data(), "UTC", sizeof("UTC"));
-        additional_info.utc_offset = 0;
+    void ToPosixTime(Kernel::HLERequestContext& ctx) {
+        // TODO(ogniK): Figure out how to handle multiple times
+        LOG_WARNING(Service_Time, "(STUBBED) called");
+        IPC::RequestParser rp{ctx};
+        auto calendar_time = rp.PopRaw<CalendarTime>();
+        auto posix_time = CalendarToPosix(calendar_time, {});
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushRaw<u32>(1); // Amount of times we're returning
+        ctx.WriteBuffer(&posix_time, sizeof(u64));
+    }
+
+    void ToPosixTimeWithMyRule(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_Time, "(STUBBED) called");
+        IPC::RequestParser rp{ctx};
+        auto calendar_time = rp.PopRaw<CalendarTime>();
+        auto posix_time = CalendarToPosix(calendar_time, {});
+
+        IPC::ResponseBuilder rb{ctx, 3};
+        rb.Push(RESULT_SUCCESS);
+        rb.PushRaw<u32>(1); // Amount of times we're returning
+        ctx.WriteBuffer(&posix_time, sizeof(u64));
     }
 };
 
@@ -207,6 +250,70 @@ void Module::Interface::GetStandardLocalSystemClock(Kernel::HLERequestContext& c
     LOG_DEBUG(Service_Time, "called");
 }
 
+void Module::Interface::GetClockSnapshot(Kernel::HLERequestContext& ctx) {
+    LOG_DEBUG(Service_Time, "called");
+
+    IPC::RequestParser rp{ctx};
+    auto unknown_u8 = rp.PopRaw<u8>();
+
+    ClockSnapshot clock_snapshot{};
+
+    const s64 time_since_epoch{std::chrono::duration_cast<std::chrono::seconds>(
+                                   std::chrono::system_clock::now().time_since_epoch())
+                                   .count()};
+    CalendarTime calendar_time{};
+    const std::time_t time(time_since_epoch);
+    const std::tm* tm = std::localtime(&time);
+    if (tm == nullptr) {
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(ResultCode(-1)); // TODO(ogniK): Find appropriate error code
+        return;
+    }
+    SteadyClockTimePoint steady_clock_time_point{CoreTiming::cyclesToMs(CoreTiming::GetTicks()) /
+                                                 1000};
+
+    LocationName location_name{"UTC"};
+    calendar_time.year = tm->tm_year + 1900;
+    calendar_time.month = tm->tm_mon + 1;
+    calendar_time.day = tm->tm_mday;
+    calendar_time.hour = tm->tm_hour;
+    calendar_time.minute = tm->tm_min;
+    calendar_time.second = tm->tm_sec;
+    clock_snapshot.system_posix_time = time_since_epoch;
+    clock_snapshot.network_posix_time = time_since_epoch;
+    clock_snapshot.system_calendar_time = calendar_time;
+    clock_snapshot.network_calendar_time = calendar_time;
+
+    CalendarAdditionalInfo additional_info{};
+    PosixToCalendar(time_since_epoch, calendar_time, additional_info, {});
+
+    clock_snapshot.system_calendar_info = additional_info;
+    clock_snapshot.network_calendar_info = additional_info;
+
+    clock_snapshot.steady_clock_timepoint = steady_clock_time_point;
+    clock_snapshot.location_name = location_name;
+    clock_snapshot.clock_auto_adjustment_enabled = 1;
+    clock_snapshot.ipc_u8 = unknown_u8;
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(RESULT_SUCCESS);
+    ctx.WriteBuffer(&clock_snapshot, sizeof(ClockSnapshot));
+}
+
+void Module::Interface::CalculateStandardUserSystemClockDifferenceByUser(
+    Kernel::HLERequestContext& ctx) {
+    LOG_DEBUG(Service_Time, "called");
+
+    IPC::RequestParser rp{ctx};
+    const auto snapshot_a = rp.PopRaw<ClockSnapshot>();
+    const auto snapshot_b = rp.PopRaw<ClockSnapshot>();
+    const u64 difference =
+        snapshot_b.user_clock_context.offset - snapshot_a.user_clock_context.offset;
+
+    IPC::ResponseBuilder rb{ctx, 4};
+    rb.Push(RESULT_SUCCESS);
+    rb.PushRaw<u64>(difference);
+}
+
 Module::Interface::Interface(std::shared_ptr<Module> time, const char* name)
     : ServiceFramework(name), time(std::move(time)) {}
 
diff --git a/src/core/hle/service/time/time.h b/src/core/hle/service/time/time.h
index 5659ecad3..ea43fbea7 100644
--- a/src/core/hle/service/time/time.h
+++ b/src/core/hle/service/time/time.h
@@ -5,6 +5,7 @@
 #pragma once
 
 #include <array>
+#include "common/common_funcs.h"
 #include "core/hle/service/service.h"
 
 namespace Service::Time {
@@ -53,6 +54,23 @@ struct SystemClockContext {
 static_assert(sizeof(SystemClockContext) == 0x20,
               "SystemClockContext structure has incorrect size");
 
+struct ClockSnapshot {
+    SystemClockContext user_clock_context;
+    SystemClockContext network_clock_context;
+    s64_le system_posix_time;
+    s64_le network_posix_time;
+    CalendarTime system_calendar_time;
+    CalendarTime network_calendar_time;
+    CalendarAdditionalInfo system_calendar_info;
+    CalendarAdditionalInfo network_calendar_info;
+    SteadyClockTimePoint steady_clock_timepoint;
+    LocationName location_name;
+    u8 clock_auto_adjustment_enabled;
+    u8 ipc_u8;
+    INSERT_PADDING_BYTES(2);
+};
+static_assert(sizeof(ClockSnapshot) == 0xd0, "ClockSnapshot is an invalid size");
+
 class Module final {
 public:
     class Interface : public ServiceFramework<Interface> {
@@ -65,6 +83,8 @@ public:
         void GetStandardSteadyClock(Kernel::HLERequestContext& ctx);
         void GetTimeZoneService(Kernel::HLERequestContext& ctx);
         void GetStandardLocalSystemClock(Kernel::HLERequestContext& ctx);
+        void GetClockSnapshot(Kernel::HLERequestContext& ctx);
+        void CalculateStandardUserSystemClockDifferenceByUser(Kernel::HLERequestContext& ctx);
 
     protected:
         std::shared_ptr<Module> time;
diff --git a/src/core/hle/service/usb/usb.cpp b/src/core/hle/service/usb/usb.cpp
index e7fb5a419..f0a831d45 100644
--- a/src/core/hle/service/usb/usb.cpp
+++ b/src/core/hle/service/usb/usb.cpp
@@ -67,15 +67,15 @@ public:
     explicit IClientEpSession() : ServiceFramework{"IClientEpSession"} {
         // clang-format off
         static const FunctionInfo functions[] = {
-            {0, nullptr, "Unknown1"},
-            {1, nullptr, "Unknown2"},
-            {2, nullptr, "Unknown3"},
-            {3, nullptr, "Unknown4"},
+            {0, nullptr, "Open"},
+            {1, nullptr, "Close"},
+            {2, nullptr, "Unknown1"},
+            {3, nullptr, "Populate"},
             {4, nullptr, "PostBufferAsync"},
-            {5, nullptr, "Unknown5"},
-            {6, nullptr, "Unknown6"},
-            {7, nullptr, "Unknown7"},
-            {8, nullptr, "Unknown8"},
+            {5, nullptr, "GetXferReport"},
+            {6, nullptr, "Unknown2"},
+            {7, nullptr, "Unknown3"},
+            {8, nullptr, "Unknown4"},
         };
         // clang-format on
 
@@ -89,15 +89,15 @@ public:
         // clang-format off
         static const FunctionInfo functions[] = {
             {0, nullptr, "Unknown1"},
-            {1, nullptr, "Unknown2"},
-            {2, nullptr, "Unknown3"},
-            {3, nullptr, "Unknown4"},
-            {4, nullptr, "Unknown5"},
+            {1, nullptr, "SetInterface"},
+            {2, nullptr, "GetInterface"},
+            {3, nullptr, "GetAlternateInterface"},
+            {4, nullptr, "GetCurrentFrame"},
             {5, nullptr, "CtrlXferAsync"},
-            {6, nullptr, "Unknown6"},
+            {6, nullptr, "Unknown2"},
             {7, nullptr, "GetCtrlXferReport"},
-            {8, nullptr, "Unknown7"},
-            {9, nullptr, "GetClientEpSession"},
+            {8, nullptr, "ResetDevice"},
+            {9, nullptr, "OpenUsbEp"},
         };
         // clang-format on
 
@@ -111,13 +111,14 @@ public:
         // clang-format off
         static const FunctionInfo functions[] = {
             {0, nullptr, "BindClientProcess"},
-            {1, nullptr, "Unknown1"},
-            {2, nullptr, "Unknown2"},
-            {3, nullptr, "Unknown3"},
-            {4, nullptr, "Unknown4"},
-            {5, nullptr, "Unknown5"},
+            {1, nullptr, "QueryAllInterfaces"},
+            {2, nullptr, "QueryAvailableInterfaces"},
+            {3, nullptr, "QueryAcquiredInterfaces"},
+            {4, nullptr, "CreateInterfaceAvailableEvent"},
+            {5, nullptr, "DestroyInterfaceAvailableEvent"},
             {6, nullptr, "GetInterfaceStateChangeEvent"},
-            {7, nullptr, "GetClientIfSession"},
+            {7, nullptr, "AcquireUsbIf"},
+            {8, nullptr, "Unknown1"},
         };
         // clang-format on
 
@@ -131,11 +132,11 @@ public:
         // clang-format off
         static const FunctionInfo functions[] = {
             {0, nullptr, "BindNoticeEvent"},
-            {1, nullptr, "Unknown1"},
+            {1, nullptr, "UnbindNoticeEvent"},
             {2, nullptr, "GetStatus"},
             {3, nullptr, "GetNotice"},
-            {4, nullptr, "Unknown2"},
-            {5, nullptr, "Unknown3"},
+            {4, nullptr, "EnablePowerRequestNotice"},
+            {5, nullptr, "DisablePowerRequestNotice"},
             {6, nullptr, "ReplyPowerRequest"},
         };
         // clang-format on
diff --git a/src/core/hle/service/vi/vi.cpp b/src/core/hle/service/vi/vi.cpp
index 184537daa..d25fdb1fe 100644
--- a/src/core/hle/service/vi/vi.cpp
+++ b/src/core/hle/service/vi/vi.cpp
@@ -6,9 +6,10 @@
 #include <array>
 #include <cstring>
 #include <memory>
+#include <optional>
 #include <type_traits>
 #include <utility>
-#include <boost/optional.hpp>
+
 #include "common/alignment.h"
 #include "common/assert.h"
 #include "common/common_funcs.h"
@@ -236,6 +237,22 @@ private:
     Data data{};
 };
 
+/// Represents a parcel containing one int '0' as its data
+/// Used by DetachBuffer and Disconnect
+class IGBPEmptyResponseParcel : public Parcel {
+protected:
+    void SerializeData() override {
+        Write(data);
+    }
+
+private:
+    struct Data {
+        u32_le unk_0;
+    };
+
+    Data data{};
+};
+
 class IGBPSetPreallocatedBufferRequestParcel : public Parcel {
 public:
     explicit IGBPSetPreallocatedBufferRequestParcel(std::vector<u8> buffer)
@@ -506,9 +523,9 @@ private:
             IGBPDequeueBufferRequestParcel request{ctx.ReadBuffer()};
             const u32 width{request.data.width};
             const u32 height{request.data.height};
-            boost::optional<u32> slot = buffer_queue->DequeueBuffer(width, height);
+            std::optional<u32> slot = buffer_queue->DequeueBuffer(width, height);
 
-            if (slot != boost::none) {
+            if (slot) {
                 // Buffer is available
                 IGBPDequeueBufferResponseParcel response{*slot};
                 ctx.WriteBuffer(response.Serialize());
@@ -520,7 +537,7 @@ private:
                         Kernel::ThreadWakeupReason reason) {
                         // Repeat TransactParcel DequeueBuffer when a buffer is available
                         auto buffer_queue = nv_flinger->GetBufferQueue(id);
-                        boost::optional<u32> slot = buffer_queue->DequeueBuffer(width, height);
+                        std::optional<u32> slot = buffer_queue->DequeueBuffer(width, height);
                         IGBPDequeueBufferResponseParcel response{*slot};
                         ctx.WriteBuffer(response.Serialize());
                         IPC::ResponseBuilder rb{ctx, 2};
@@ -553,6 +570,12 @@ private:
             ctx.WriteBuffer(response.Serialize());
         } else if (transaction == TransactionId::CancelBuffer) {
             LOG_CRITICAL(Service_VI, "(STUBBED) called, transaction=CancelBuffer");
+        } else if (transaction == TransactionId::Disconnect ||
+                   transaction == TransactionId::DetachBuffer) {
+            const auto buffer = ctx.ReadBuffer();
+
+            IGBPEmptyResponseParcel response{};
+            ctx.WriteBuffer(response.Serialize());
         } else {
             ASSERT_MSG(false, "Unimplemented");
         }
diff --git a/src/core/loader/loader.h b/src/core/loader/loader.h
index e562b3a04..7686634bf 100644
--- a/src/core/loader/loader.h
+++ b/src/core/loader/loader.h
@@ -6,10 +6,11 @@
 
 #include <iosfwd>
 #include <memory>
+#include <optional>
 #include <string>
 #include <utility>
 #include <vector>
-#include <boost/optional.hpp>
+
 #include "common/common_types.h"
 #include "core/file_sys/vfs.h"
 
@@ -145,7 +146,7 @@ public:
      * information.
      * @returns A pair with the optional system mode, and and the status.
      */
-    virtual std::pair<boost::optional<u32>, ResultStatus> LoadKernelSystemMode() {
+    virtual std::pair<std::optional<u32>, ResultStatus> LoadKernelSystemMode() {
         // 96MB allocated to the application.
         return std::make_pair(2, ResultStatus::Success);
     }
diff --git a/src/core/loader/nro.cpp b/src/core/loader/nro.cpp
index 243b499f2..fbbd6b0de 100644
--- a/src/core/loader/nro.cpp
+++ b/src/core/loader/nro.cpp
@@ -12,10 +12,12 @@
 #include "common/swap.h"
 #include "core/core.h"
 #include "core/file_sys/control_metadata.h"
+#include "core/file_sys/romfs_factory.h"
 #include "core/file_sys/vfs_offset.h"
 #include "core/gdbstub/gdbstub.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/kernel/vm_manager.h"
+#include "core/hle/service/filesystem/filesystem.h"
 #include "core/loader/nro.h"
 #include "core/loader/nso.h"
 #include "core/memory.h"
@@ -127,18 +129,23 @@ static constexpr u32 PageAlignSize(u32 size) {
     return (size + Memory::PAGE_MASK) & ~Memory::PAGE_MASK;
 }
 
-bool AppLoader_NRO::LoadNro(const FileSys::VfsFile& file, VAddr load_base) {
-    // Read NSO header
-    NroHeader nro_header{};
-    if (sizeof(NroHeader) != file.ReadObject(&nro_header)) {
+/*static*/ bool AppLoader_NRO::LoadNro(const std::vector<u8>& data, const std::string& name,
+                                       VAddr load_base) {
+
+    if (data.size() < sizeof(NroHeader)) {
         return {};
     }
+
+    // Read NSO header
+    NroHeader nro_header{};
+    std::memcpy(&nro_header, data.data(), sizeof(NroHeader));
     if (nro_header.magic != Common::MakeMagic('N', 'R', 'O', '0')) {
         return {};
     }
 
     // Build program image
-    std::vector<u8> program_image = file.ReadBytes(PageAlignSize(nro_header.file_size));
+    std::vector<u8> program_image(PageAlignSize(nro_header.file_size));
+    std::memcpy(program_image.data(), data.data(), program_image.size());
     if (program_image.size() != PageAlignSize(nro_header.file_size)) {
         return {};
     }
@@ -163,17 +170,20 @@ bool AppLoader_NRO::LoadNro(const FileSys::VfsFile& file, VAddr load_base) {
                     arg_data.size());
     }
 
-    // Read MOD header
-    ModHeader mod_header{};
     // Default .bss to NRO header bss size if MOD0 section doesn't exist
     u32 bss_size{PageAlignSize(nro_header.bss_size)};
+
+    // Read MOD header
+    ModHeader mod_header{};
     std::memcpy(&mod_header, program_image.data() + nro_header.module_header_offset,
                 sizeof(ModHeader));
+
     const bool has_mod_header{mod_header.magic == Common::MakeMagic('M', 'O', 'D', '0')};
     if (has_mod_header) {
         // Resize program image to include .bss section and page align each section
         bss_size = PageAlignSize(mod_header.bss_end_offset - mod_header.bss_start_offset);
     }
+
     codeset.DataSegment().size += bss_size;
     program_image.resize(static_cast<u32>(program_image.size()) + bss_size);
 
@@ -182,11 +192,15 @@ bool AppLoader_NRO::LoadNro(const FileSys::VfsFile& file, VAddr load_base) {
     Core::CurrentProcess()->LoadModule(std::move(codeset), load_base);
 
     // Register module with GDBStub
-    GDBStub::RegisterModule(file.GetName(), load_base, load_base);
+    GDBStub::RegisterModule(name, load_base, load_base);
 
     return true;
 }
 
+bool AppLoader_NRO::LoadNro(const FileSys::VfsFile& file, VAddr load_base) {
+    return AppLoader_NRO::LoadNro(file.ReadAllBytes(), file.GetName(), load_base);
+}
+
 ResultStatus AppLoader_NRO::Load(Kernel::Process& process) {
     if (is_loaded) {
         return ResultStatus::ErrorAlreadyLoaded;
@@ -199,6 +213,9 @@ ResultStatus AppLoader_NRO::Load(Kernel::Process& process) {
         return ResultStatus::ErrorLoadingNRO;
     }
 
+    if (romfs != nullptr)
+        Service::FileSystem::RegisterRomFS(std::make_unique<FileSys::RomFSFactory>(*this));
+
     process.Run(base_address, Kernel::THREADPRIO_DEFAULT, Memory::DEFAULT_STACK_SIZE);
 
     is_loaded = true;
diff --git a/src/core/loader/nro.h b/src/core/loader/nro.h
index 50ee5a78a..3e6959302 100644
--- a/src/core/loader/nro.h
+++ b/src/core/loader/nro.h
@@ -5,6 +5,7 @@
 #pragma once
 
 #include <string>
+#include <vector>
 #include "common/common_types.h"
 #include "core/loader/linker.h"
 #include "core/loader/loader.h"
@@ -40,6 +41,8 @@ public:
     ResultStatus ReadTitle(std::string& title) override;
     bool IsRomFSUpdatable() const override;
 
+    static bool LoadNro(const std::vector<u8>& data, const std::string& name, VAddr load_base);
+
 private:
     bool LoadNro(const FileSys::VfsFile& file, VAddr load_base);
 
diff --git a/src/core/loader/nso.cpp b/src/core/loader/nso.cpp
index 68efca5c0..aaf006309 100644
--- a/src/core/loader/nso.cpp
+++ b/src/core/loader/nso.cpp
@@ -154,7 +154,7 @@ std::optional<VAddr> AppLoader_NSO::LoadModule(const FileSys::VfsFile& file, VAd
     program_image.resize(image_size);
 
     // Apply patches if necessary
-    if (pm && pm->HasNSOPatch(nso_header.build_id)) {
+    if (pm && (pm->HasNSOPatch(nso_header.build_id) || Settings::values.dump_nso)) {
         std::vector<u8> pi_header(program_image.size() + 0x100);
         std::memcpy(pi_header.data(), &nso_header, sizeof(NsoHeader));
         std::memcpy(pi_header.data() + 0x100, program_image.data(), program_image.size());
diff --git a/src/core/loader/nsp.cpp b/src/core/loader/nsp.cpp
index 13e57848d..080d89904 100644
--- a/src/core/loader/nsp.cpp
+++ b/src/core/loader/nsp.cpp
@@ -36,6 +36,16 @@ AppLoader_NSP::AppLoader_NSP(FileSys::VirtualFile file)
 
     std::tie(nacp_file, icon_file) =
         FileSys::PatchManager(nsp->GetProgramTitleID()).ParseControlNCA(*control_nca);
+
+    if (nsp->IsExtractedType()) {
+        secondary_loader = std::make_unique<AppLoader_DeconstructedRomDirectory>(nsp->GetExeFS());
+    } else {
+        if (title_id == 0)
+            return;
+
+        secondary_loader = std::make_unique<AppLoader_NCA>(
+            nsp->GetNCAFile(title_id, FileSys::ContentRecordType::Program));
+    }
 }
 
 AppLoader_NSP::~AppLoader_NSP() = default;
@@ -67,26 +77,19 @@ ResultStatus AppLoader_NSP::Load(Kernel::Process& process) {
         return ResultStatus::ErrorAlreadyLoaded;
     }
 
-    if (nsp->IsExtractedType()) {
-        secondary_loader = std::make_unique<AppLoader_DeconstructedRomDirectory>(nsp->GetExeFS());
-    } else {
-        if (title_id == 0)
-            return ResultStatus::ErrorNSPMissingProgramNCA;
-
-        secondary_loader = std::make_unique<AppLoader_NCA>(
-            nsp->GetNCAFile(title_id, FileSys::ContentRecordType::Program));
+    if (title_id == 0)
+        return ResultStatus::ErrorNSPMissingProgramNCA;
 
-        if (nsp->GetStatus() != ResultStatus::Success)
-            return nsp->GetStatus();
+    if (nsp->GetStatus() != ResultStatus::Success)
+        return nsp->GetStatus();
 
-        if (nsp->GetProgramStatus(title_id) != ResultStatus::Success)
-            return nsp->GetProgramStatus(title_id);
+    if (nsp->GetProgramStatus(title_id) != ResultStatus::Success)
+        return nsp->GetProgramStatus(title_id);
 
-        if (nsp->GetNCA(title_id, FileSys::ContentRecordType::Program) == nullptr) {
-            if (!Core::Crypto::KeyManager::KeyFileExists(false))
-                return ResultStatus::ErrorMissingProductionKeyFile;
-            return ResultStatus::ErrorNSPMissingProgramNCA;
-        }
+    if (nsp->GetNCA(title_id, FileSys::ContentRecordType::Program) == nullptr) {
+        if (!Core::Crypto::KeyManager::KeyFileExists(false))
+            return ResultStatus::ErrorMissingProductionKeyFile;
+        return ResultStatus::ErrorNSPMissingProgramNCA;
     }
 
     const auto result = secondary_loader->Load(process);
diff --git a/src/core/memory.cpp b/src/core/memory.cpp
index 014298ed6..70abd856a 100644
--- a/src/core/memory.cpp
+++ b/src/core/memory.cpp
@@ -4,9 +4,9 @@
 
 #include <algorithm>
 #include <cstring>
+#include <optional>
 #include <utility>
 
-#include <boost/optional.hpp>
 #include "common/assert.h"
 #include "common/common_types.h"
 #include "common/logging/log.h"
diff --git a/src/core/memory_hook.h b/src/core/memory_hook.h
index 0269c7ff1..940777107 100644
--- a/src/core/memory_hook.h
+++ b/src/core/memory_hook.h
@@ -5,7 +5,8 @@
 #pragma once
 
 #include <memory>
-#include <boost/optional.hpp>
+#include <optional>
+
 #include "common/common_types.h"
 
 namespace Memory {
@@ -18,19 +19,19 @@ namespace Memory {
  *
  * A hook may be mapped to multiple regions of memory.
  *
- * If a boost::none or false is returned from a function, the read/write request is passed through
+ * If a std::nullopt or false is returned from a function, the read/write request is passed through
  * to the underlying memory region.
  */
 class MemoryHook {
 public:
     virtual ~MemoryHook();
 
-    virtual boost::optional<bool> IsValidAddress(VAddr addr) = 0;
+    virtual std::optional<bool> IsValidAddress(VAddr addr) = 0;
 
-    virtual boost::optional<u8> Read8(VAddr addr) = 0;
-    virtual boost::optional<u16> Read16(VAddr addr) = 0;
-    virtual boost::optional<u32> Read32(VAddr addr) = 0;
-    virtual boost::optional<u64> Read64(VAddr addr) = 0;
+    virtual std::optional<u8> Read8(VAddr addr) = 0;
+    virtual std::optional<u16> Read16(VAddr addr) = 0;
+    virtual std::optional<u32> Read32(VAddr addr) = 0;
+    virtual std::optional<u64> Read64(VAddr addr) = 0;
 
     virtual bool ReadBlock(VAddr src_addr, void* dest_buffer, std::size_t size) = 0;
 
diff --git a/src/core/perf_stats.cpp b/src/core/perf_stats.cpp
index 7d95816fe..c716a462b 100644
--- a/src/core/perf_stats.cpp
+++ b/src/core/perf_stats.cpp
@@ -74,10 +74,6 @@ double PerfStats::GetLastFrameTimeScale() {
 }
 
 void FrameLimiter::DoFrameLimiting(microseconds current_system_time_us) {
-    // Max lag caused by slow frames. Can be adjusted to compensate for too many slow frames. Higher
-    // values increase the time needed to recover and limit framerate again after spikes.
-    constexpr microseconds MAX_LAG_TIME_US = 25000us;
-
     if (!Settings::values.use_frame_limit) {
         return;
     }
diff --git a/src/core/settings.h b/src/core/settings.h
index 8f2da01c8..e424479f2 100644
--- a/src/core/settings.h
+++ b/src/core/settings.h
@@ -6,6 +6,7 @@
 
 #include <array>
 #include <atomic>
+#include <optional>
 #include <string>
 #include "common/common_types.h"
 
@@ -113,8 +114,10 @@ static const std::array<const char*, NumAnalogs> mapping = {{
 struct Values {
     // System
     bool use_docked_mode;
-    std::string username;
-    int language_index;
+    bool enable_nfc;
+    std::optional<u32> rng_seed;
+    s32 current_user;
+    s32 language_index;
 
     // Controls
     std::array<std::string, NativeButton::NumButtons> buttons;
@@ -156,6 +159,7 @@ struct Values {
     bool use_gdbstub;
     u16 gdbstub_port;
     std::string program_args;
+    bool dump_nso;
 
     // WebService
     bool enable_telemetry;
diff --git a/src/core/telemetry_session.cpp b/src/core/telemetry_session.cpp
index 0de13edd3..a3b08c740 100644
--- a/src/core/telemetry_session.cpp
+++ b/src/core/telemetry_session.cpp
@@ -184,4 +184,13 @@ TelemetrySession::~TelemetrySession() {
     backend = nullptr;
 }
 
+bool TelemetrySession::SubmitTestcase() {
+#ifdef ENABLE_WEB_SERVICE
+    field_collection.Accept(*backend);
+    return backend->SubmitTestcase();
+#else
+    return false;
+#endif
+}
+
 } // namespace Core
diff --git a/src/core/telemetry_session.h b/src/core/telemetry_session.h
index 2a4845797..023612b79 100644
--- a/src/core/telemetry_session.h
+++ b/src/core/telemetry_session.h
@@ -31,6 +31,12 @@ public:
         field_collection.AddField(type, name, std::move(value));
     }
 
+    /**
+     * Submits a Testcase.
+     * @returns A bool indicating whether the submission succeeded
+     */
+    bool SubmitTestcase();
+
 private:
     Telemetry::FieldCollection field_collection; ///< Tracks all added fields for the session
     std::unique_ptr<Telemetry::VisitorInterface> backend; ///< Backend interface that logs fields
diff --git a/src/tests/core/arm/arm_test_common.cpp b/src/tests/core/arm/arm_test_common.cpp
index 37e15bad0..9b8a44fa1 100644
--- a/src/tests/core/arm/arm_test_common.cpp
+++ b/src/tests/core/arm/arm_test_common.cpp
@@ -64,11 +64,11 @@ void TestEnvironment::ClearWriteRecords() {
 
 TestEnvironment::TestMemory::~TestMemory() {}
 
-boost::optional<bool> TestEnvironment::TestMemory::IsValidAddress(VAddr addr) {
+std::optional<bool> TestEnvironment::TestMemory::IsValidAddress(VAddr addr) {
     return true;
 }
 
-boost::optional<u8> TestEnvironment::TestMemory::Read8(VAddr addr) {
+std::optional<u8> TestEnvironment::TestMemory::Read8(VAddr addr) {
     const auto iter = data.find(addr);
 
     if (iter == data.end()) {
@@ -79,15 +79,15 @@ boost::optional<u8> TestEnvironment::TestMemory::Read8(VAddr addr) {
     return iter->second;
 }
 
-boost::optional<u16> TestEnvironment::TestMemory::Read16(VAddr addr) {
+std::optional<u16> TestEnvironment::TestMemory::Read16(VAddr addr) {
     return *Read8(addr) | static_cast<u16>(*Read8(addr + 1)) << 8;
 }
 
-boost::optional<u32> TestEnvironment::TestMemory::Read32(VAddr addr) {
+std::optional<u32> TestEnvironment::TestMemory::Read32(VAddr addr) {
     return *Read16(addr) | static_cast<u32>(*Read16(addr + 2)) << 16;
 }
 
-boost::optional<u64> TestEnvironment::TestMemory::Read64(VAddr addr) {
+std::optional<u64> TestEnvironment::TestMemory::Read64(VAddr addr) {
     return *Read32(addr) | static_cast<u64>(*Read32(addr + 4)) << 32;
 }
 
diff --git a/src/tests/core/arm/arm_test_common.h b/src/tests/core/arm/arm_test_common.h
index 5de8dab4e..0b7539601 100644
--- a/src/tests/core/arm/arm_test_common.h
+++ b/src/tests/core/arm/arm_test_common.h
@@ -64,12 +64,12 @@ private:
 
         ~TestMemory() override;
 
-        boost::optional<bool> IsValidAddress(VAddr addr) override;
+        std::optional<bool> IsValidAddress(VAddr addr) override;
 
-        boost::optional<u8> Read8(VAddr addr) override;
-        boost::optional<u16> Read16(VAddr addr) override;
-        boost::optional<u32> Read32(VAddr addr) override;
-        boost::optional<u64> Read64(VAddr addr) override;
+        std::optional<u8> Read8(VAddr addr) override;
+        std::optional<u16> Read16(VAddr addr) override;
+        std::optional<u32> Read32(VAddr addr) override;
+        std::optional<u64> Read64(VAddr addr) override;
 
         bool ReadBlock(VAddr src_addr, void* dest_buffer, std::size_t size) override;
 
diff --git a/src/video_core/CMakeLists.txt b/src/video_core/CMakeLists.txt
index 09ecc5bad..a780215c1 100644
--- a/src/video_core/CMakeLists.txt
+++ b/src/video_core/CMakeLists.txt
@@ -21,6 +21,7 @@ add_library(video_core STATIC
     macro_interpreter.h
     memory_manager.cpp
     memory_manager.h
+    rasterizer_cache.cpp
     rasterizer_cache.h
     rasterizer_interface.h
     renderer_base.cpp
@@ -33,6 +34,7 @@ add_library(video_core STATIC
     renderer_opengl/gl_rasterizer.h
     renderer_opengl/gl_rasterizer_cache.cpp
     renderer_opengl/gl_rasterizer_cache.h
+    renderer_opengl/gl_resource_manager.cpp
     renderer_opengl/gl_resource_manager.h
     renderer_opengl/gl_shader_cache.cpp
     renderer_opengl/gl_shader_cache.h
@@ -51,6 +53,10 @@ add_library(video_core STATIC
     renderer_opengl/maxwell_to_gl.h
     renderer_opengl/renderer_opengl.cpp
     renderer_opengl/renderer_opengl.h
+    renderer_opengl/utils.cpp
+    renderer_opengl/utils.h
+    surface.cpp
+    surface.h
     textures/astc.cpp
     textures/astc.h
     textures/decoders.cpp
diff --git a/src/video_core/command_processor.cpp b/src/video_core/command_processor.cpp
index f1aa6091b..28e8c13aa 100644
--- a/src/video_core/command_processor.cpp
+++ b/src/video_core/command_processor.cpp
@@ -81,7 +81,7 @@ void GPU::ProcessCommandLists(const std::vector<CommandListHeader>& commands) {
     for (auto entry : commands) {
         Tegra::GPUVAddr address = entry.Address();
         u32 size = entry.sz;
-        const boost::optional<VAddr> head_address = memory_manager->GpuToCpuAddress(address);
+        const std::optional<VAddr> head_address = memory_manager->GpuToCpuAddress(address);
         VAddr current_addr = *head_address;
         while (current_addr < *head_address + size * sizeof(CommandHeader)) {
             const CommandHeader header = {Memory::Read32(current_addr)};
diff --git a/src/video_core/engines/maxwell_3d.cpp b/src/video_core/engines/maxwell_3d.cpp
index bca014a4a..6de07ea56 100644
--- a/src/video_core/engines/maxwell_3d.cpp
+++ b/src/video_core/engines/maxwell_3d.cpp
@@ -3,6 +3,7 @@
 // Refer to the license.txt file included.
 
 #include <cinttypes>
+#include <cstring>
 #include "common/assert.h"
 #include "core/core.h"
 #include "core/core_timing.h"
@@ -19,21 +20,69 @@ namespace Tegra::Engines {
 constexpr u32 MacroRegistersStart = 0xE00;
 
 Maxwell3D::Maxwell3D(VideoCore::RasterizerInterface& rasterizer, MemoryManager& memory_manager)
-    : memory_manager(memory_manager), rasterizer{rasterizer}, macro_interpreter(*this) {}
+    : memory_manager(memory_manager), rasterizer{rasterizer}, macro_interpreter(*this) {
+    InitializeRegisterDefaults();
+}
+
+void Maxwell3D::InitializeRegisterDefaults() {
+    // Initializes registers to their default values - what games expect them to be at boot. This is
+    // for certain registers that may not be explicitly set by games.
+
+    // Reset all registers to zero
+    std::memset(&regs, 0, sizeof(regs));
+
+    // Depth range near/far is not always set, but is expected to be the default 0.0f, 1.0f. This is
+    // needed for ARMS.
+    for (std::size_t viewport{}; viewport < Regs::NumViewports; ++viewport) {
+        regs.viewport[viewport].depth_range_near = 0.0f;
+        regs.viewport[viewport].depth_range_far = 1.0f;
+    }
+    // Doom and Bomberman seems to use the uninitialized registers and just enable blend
+    // so initialize blend registers with sane values
+    regs.blend.equation_rgb = Regs::Blend::Equation::Add;
+    regs.blend.factor_source_rgb = Regs::Blend::Factor::One;
+    regs.blend.factor_dest_rgb = Regs::Blend::Factor::Zero;
+    regs.blend.equation_a = Regs::Blend::Equation::Add;
+    regs.blend.factor_source_a = Regs::Blend::Factor::One;
+    regs.blend.factor_dest_a = Regs::Blend::Factor::Zero;
+    for (std::size_t blend_index = 0; blend_index < Regs::NumRenderTargets; blend_index++) {
+        regs.independent_blend[blend_index].equation_rgb = Regs::Blend::Equation::Add;
+        regs.independent_blend[blend_index].factor_source_rgb = Regs::Blend::Factor::One;
+        regs.independent_blend[blend_index].factor_dest_rgb = Regs::Blend::Factor::Zero;
+        regs.independent_blend[blend_index].equation_a = Regs::Blend::Equation::Add;
+        regs.independent_blend[blend_index].factor_source_a = Regs::Blend::Factor::One;
+        regs.independent_blend[blend_index].factor_dest_a = Regs::Blend::Factor::Zero;
+    }
+    regs.stencil_front_op_fail = Regs::StencilOp::Keep;
+    regs.stencil_front_op_zfail = Regs::StencilOp::Keep;
+    regs.stencil_front_op_zpass = Regs::StencilOp::Keep;
+    regs.stencil_front_func_func = Regs::ComparisonOp::Always;
+    regs.stencil_front_func_mask = 0xFFFFFFFF;
+    regs.stencil_front_mask = 0xFFFFFFFF;
+    regs.stencil_two_side_enable = 1;
+    regs.stencil_back_op_fail = Regs::StencilOp::Keep;
+    regs.stencil_back_op_zfail = Regs::StencilOp::Keep;
+    regs.stencil_back_op_zpass = Regs::StencilOp::Keep;
+    regs.stencil_back_func_func = Regs::ComparisonOp::Always;
+    regs.stencil_back_func_mask = 0xFFFFFFFF;
+    regs.stencil_back_mask = 0xFFFFFFFF;
+}
 
 void Maxwell3D::CallMacroMethod(u32 method, std::vector<u32> parameters) {
     // Reset the current macro.
     executing_macro = 0;
 
-    // The requested macro must have been uploaded already.
-    auto macro_code = uploaded_macros.find(method);
-    if (macro_code == uploaded_macros.end()) {
-        LOG_ERROR(HW_GPU, "Macro {:04X} was not uploaded", method);
+    // Lookup the macro offset
+    const u32 entry{(method - MacroRegistersStart) >> 1};
+    const auto& search{macro_offsets.find(entry)};
+    if (search == macro_offsets.end()) {
+        LOG_CRITICAL(HW_GPU, "macro not found for method 0x{:X}!", method);
+        UNREACHABLE();
         return;
     }
 
     // Execute the current macro.
-    macro_interpreter.Execute(macro_code->second, std::move(parameters));
+    macro_interpreter.Execute(search->second, std::move(parameters));
 }
 
 void Maxwell3D::WriteReg(u32 method, u32 value, u32 remaining_params) {
@@ -72,13 +121,23 @@ void Maxwell3D::WriteReg(u32 method, u32 value, u32 remaining_params) {
         debug_context->OnEvent(Tegra::DebugContext::Event::MaxwellCommandLoaded, nullptr);
     }
 
-    regs.reg_array[method] = value;
+    if (regs.reg_array[method] != value) {
+        regs.reg_array[method] = value;
+        if (method >= MAXWELL3D_REG_INDEX(vertex_attrib_format) &&
+            method < MAXWELL3D_REG_INDEX(vertex_attrib_format) + regs.vertex_attrib_format.size()) {
+            dirty_flags.vertex_attrib_format = true;
+        }
+    }
 
     switch (method) {
     case MAXWELL3D_REG_INDEX(macros.data): {
         ProcessMacroUpload(value);
         break;
     }
+    case MAXWELL3D_REG_INDEX(macros.bind): {
+        ProcessMacroBind(value);
+        break;
+    }
     case MAXWELL3D_REG_INDEX(const_buffer.cb_data[0]):
     case MAXWELL3D_REG_INDEX(const_buffer.cb_data[1]):
     case MAXWELL3D_REG_INDEX(const_buffer.cb_data[2]):
@@ -140,22 +199,25 @@ void Maxwell3D::WriteReg(u32 method, u32 value, u32 remaining_params) {
 }
 
 void Maxwell3D::ProcessMacroUpload(u32 data) {
-    // Store the uploaded macro code to interpret them when they're called.
-    auto& macro = uploaded_macros[regs.macros.entry * 2 + MacroRegistersStart];
-    macro.push_back(data);
+    ASSERT_MSG(regs.macros.upload_address < macro_memory.size(),
+               "upload_address exceeded macro_memory size!");
+    macro_memory[regs.macros.upload_address++] = data;
+}
+
+void Maxwell3D::ProcessMacroBind(u32 data) {
+    macro_offsets[regs.macros.entry] = data;
 }
 
 void Maxwell3D::ProcessQueryGet() {
     GPUVAddr sequence_address = regs.query.QueryAddress();
     // Since the sequence address is given as a GPU VAddr, we have to convert it to an application
     // VAddr before writing.
-    boost::optional<VAddr> address = memory_manager.GpuToCpuAddress(sequence_address);
+    std::optional<VAddr> address = memory_manager.GpuToCpuAddress(sequence_address);
 
     // TODO(Subv): Support the other query units.
     ASSERT_MSG(regs.query.query_get.unit == Regs::QueryUnit::Crop,
                "Units other than CROP are unimplemented");
 
-    u32 value = Memory::Read32(*address);
     u64 result = 0;
 
     // TODO(Subv): Support the other query variables
@@ -268,7 +330,7 @@ void Maxwell3D::ProcessCBData(u32 value) {
     // Don't allow writing past the end of the buffer.
     ASSERT(regs.const_buffer.cb_pos + sizeof(u32) <= regs.const_buffer.cb_size);
 
-    boost::optional<VAddr> address =
+    std::optional<VAddr> address =
         memory_manager.GpuToCpuAddress(buffer_address + regs.const_buffer.cb_pos);
 
     Memory::Write32(*address, value);
@@ -281,7 +343,7 @@ Texture::TICEntry Maxwell3D::GetTICEntry(u32 tic_index) const {
     GPUVAddr tic_base_address = regs.tic.TICAddress();
 
     GPUVAddr tic_address_gpu = tic_base_address + tic_index * sizeof(Texture::TICEntry);
-    boost::optional<VAddr> tic_address_cpu = memory_manager.GpuToCpuAddress(tic_address_gpu);
+    std::optional<VAddr> tic_address_cpu = memory_manager.GpuToCpuAddress(tic_address_gpu);
 
     Texture::TICEntry tic_entry;
     Memory::ReadBlock(*tic_address_cpu, &tic_entry, sizeof(Texture::TICEntry));
@@ -305,7 +367,7 @@ Texture::TSCEntry Maxwell3D::GetTSCEntry(u32 tsc_index) const {
     GPUVAddr tsc_base_address = regs.tsc.TSCAddress();
 
     GPUVAddr tsc_address_gpu = tsc_base_address + tsc_index * sizeof(Texture::TSCEntry);
-    boost::optional<VAddr> tsc_address_cpu = memory_manager.GpuToCpuAddress(tsc_address_gpu);
+    std::optional<VAddr> tsc_address_cpu = memory_manager.GpuToCpuAddress(tsc_address_gpu);
 
     Texture::TSCEntry tsc_entry;
     Memory::ReadBlock(*tsc_address_cpu, &tsc_entry, sizeof(Texture::TSCEntry));
@@ -369,7 +431,7 @@ Texture::FullTextureInfo Maxwell3D::GetStageTexture(Regs::ShaderStage stage,
 
     ASSERT(tex_info_address < tex_info_buffer.address + tex_info_buffer.size);
 
-    boost::optional<VAddr> tex_address_cpu = memory_manager.GpuToCpuAddress(tex_info_address);
+    std::optional<VAddr> tex_address_cpu = memory_manager.GpuToCpuAddress(tex_info_address);
     Texture::TextureHandle tex_handle{Memory::Read32(*tex_address_cpu)};
 
     Texture::FullTextureInfo tex_info{};
diff --git a/src/video_core/engines/maxwell_3d.h b/src/video_core/engines/maxwell_3d.h
index 0e09a7ee5..91ca57883 100644
--- a/src/video_core/engines/maxwell_3d.h
+++ b/src/video_core/engines/maxwell_3d.h
@@ -345,6 +345,14 @@ public:
             Invert = 6,
             IncrWrap = 7,
             DecrWrap = 8,
+            KeepOGL = 0x1E00,
+            ZeroOGL = 0,
+            ReplaceOGL = 0x1E01,
+            IncrOGL = 0x1E02,
+            DecrOGL = 0x1E03,
+            InvertOGL = 0x150A,
+            IncrWrapOGL = 0x8507,
+            DecrWrapOGL = 0x8508,
         };
 
         enum class MemoryLayout : u32 {
@@ -462,6 +470,16 @@ public:
             }
         };
 
+        struct ColorMask {
+            union {
+                u32 raw;
+                BitField<0, 4, u32> R;
+                BitField<4, 4, u32> G;
+                BitField<8, 4, u32> B;
+                BitField<12, 4, u32> A;
+            };
+        };
+
         bool IsShaderConfigEnabled(std::size_t index) const {
             // The VertexB is always enabled.
             if (index == static_cast<std::size_t>(Regs::ShaderProgram::VertexB)) {
@@ -475,12 +493,13 @@ public:
                 INSERT_PADDING_WORDS(0x45);
 
                 struct {
-                    INSERT_PADDING_WORDS(1);
+                    u32 upload_address;
                     u32 data;
                     u32 entry;
+                    u32 bind;
                 } macros;
 
-                INSERT_PADDING_WORDS(0x189);
+                INSERT_PADDING_WORDS(0x188);
 
                 u32 tfb_enabled;
 
@@ -570,7 +589,11 @@ public:
                 u32 stencil_back_mask;
                 u32 stencil_back_func_mask;
 
-                INSERT_PADDING_WORDS(0x13);
+                INSERT_PADDING_WORDS(0xC);
+
+                u32 color_mask_common;
+
+                INSERT_PADDING_WORDS(0x6);
 
                 u32 rt_separate_frag_data;
 
@@ -645,8 +668,14 @@ public:
                 ComparisonOp depth_test_func;
                 float alpha_test_ref;
                 ComparisonOp alpha_test_func;
-
-                INSERT_PADDING_WORDS(0x9);
+                u32 draw_tfb_stride;
+                struct {
+                    float r;
+                    float g;
+                    float b;
+                    float a;
+                } blend_color;
+                INSERT_PADDING_WORDS(0x4);
 
                 struct {
                     u32 separate_alpha;
@@ -723,7 +752,11 @@ public:
                 StencilOp stencil_back_op_zpass;
                 ComparisonOp stencil_back_func_func;
 
-                INSERT_PADDING_WORDS(0x17);
+                INSERT_PADDING_WORDS(0x4);
+
+                u32 framebuffer_srgb;
+
+                INSERT_PADDING_WORDS(0x12);
 
                 union {
                     BitField<2, 1, u32> coord_origin;
@@ -751,7 +784,14 @@ public:
                     };
                 } draw;
 
-                INSERT_PADDING_WORDS(0x6B);
+                INSERT_PADDING_WORDS(0xA);
+
+                struct {
+                    u32 enabled;
+                    u32 index;
+                } primitive_restart;
+
+                INSERT_PADDING_WORDS(0x5F);
 
                 struct {
                     u32 start_addr_high;
@@ -829,8 +869,9 @@ public:
                     BitField<6, 4, u32> RT;
                     BitField<10, 11, u32> layer;
                 } clear_buffers;
-
-                INSERT_PADDING_WORDS(0x4B);
+                INSERT_PADDING_WORDS(0xB);
+                std::array<ColorMask, NumRenderTargets> color_mask;
+                INSERT_PADDING_WORDS(0x38);
 
                 struct {
                     u32 query_address_high;
@@ -971,6 +1012,12 @@ public:
     State state{};
     MemoryManager& memory_manager;
 
+    struct DirtyFlags {
+        bool vertex_attrib_format = true;
+    };
+
+    DirtyFlags dirty_flags;
+
     /// Reads a register value located at the input method address
     u32 GetRegisterValue(u32 method) const;
 
@@ -983,10 +1030,25 @@ public:
     /// Returns the texture information for a specific texture in a specific shader stage.
     Texture::FullTextureInfo GetStageTexture(Regs::ShaderStage stage, std::size_t offset) const;
 
+    /// Memory for macro code - it's undetermined how big this is, however 1MB is much larger than
+    /// we've seen used.
+    using MacroMemory = std::array<u32, 0x40000>;
+
+    /// Gets a reference to macro memory.
+    const MacroMemory& GetMacroMemory() const {
+        return macro_memory;
+    }
+
 private:
+    void InitializeRegisterDefaults();
+
     VideoCore::RasterizerInterface& rasterizer;
 
-    std::unordered_map<u32, std::vector<u32>> uploaded_macros;
+    /// Start offsets of each macro in macro_memory
+    std::unordered_map<u32, u32> macro_offsets;
+
+    /// Memory for macro code
+    MacroMemory macro_memory;
 
     /// Macro method that is currently being executed / being fed parameters.
     u32 executing_macro = 0;
@@ -1009,9 +1071,12 @@ private:
      */
     void CallMacroMethod(u32 method, std::vector<u32> parameters);
 
-    /// Handles writes to the macro uploading registers.
+    /// Handles writes to the macro uploading register.
     void ProcessMacroUpload(u32 data);
 
+    /// Handles writes to the macro bind register.
+    void ProcessMacroBind(u32 data);
+
     /// Handles a write to the CLEAR_BUFFERS register.
     void ProcessClearBuffers();
 
@@ -1045,6 +1110,7 @@ ASSERT_REG_POSITION(scissor_test, 0x380);
 ASSERT_REG_POSITION(stencil_back_func_ref, 0x3D5);
 ASSERT_REG_POSITION(stencil_back_mask, 0x3D6);
 ASSERT_REG_POSITION(stencil_back_func_mask, 0x3D7);
+ASSERT_REG_POSITION(color_mask_common, 0x3E4);
 ASSERT_REG_POSITION(rt_separate_frag_data, 0x3EB);
 ASSERT_REG_POSITION(zeta, 0x3F8);
 ASSERT_REG_POSITION(vertex_attrib_format, 0x458);
@@ -1057,6 +1123,10 @@ ASSERT_REG_POSITION(depth_write_enabled, 0x4BA);
 ASSERT_REG_POSITION(alpha_test_enabled, 0x4BB);
 ASSERT_REG_POSITION(d3d_cull_mode, 0x4C2);
 ASSERT_REG_POSITION(depth_test_func, 0x4C3);
+ASSERT_REG_POSITION(alpha_test_ref, 0x4C4);
+ASSERT_REG_POSITION(alpha_test_func, 0x4C5);
+ASSERT_REG_POSITION(draw_tfb_stride, 0x4C6);
+ASSERT_REG_POSITION(blend_color, 0x4C7);
 ASSERT_REG_POSITION(blend, 0x4CF);
 ASSERT_REG_POSITION(stencil_enable, 0x4E0);
 ASSERT_REG_POSITION(stencil_front_op_fail, 0x4E1);
@@ -1077,14 +1147,17 @@ ASSERT_REG_POSITION(stencil_back_op_fail, 0x566);
 ASSERT_REG_POSITION(stencil_back_op_zfail, 0x567);
 ASSERT_REG_POSITION(stencil_back_op_zpass, 0x568);
 ASSERT_REG_POSITION(stencil_back_func_func, 0x569);
+ASSERT_REG_POSITION(framebuffer_srgb, 0x56E);
 ASSERT_REG_POSITION(point_coord_replace, 0x581);
 ASSERT_REG_POSITION(code_address, 0x582);
 ASSERT_REG_POSITION(draw, 0x585);
+ASSERT_REG_POSITION(primitive_restart, 0x591);
 ASSERT_REG_POSITION(index_array, 0x5F2);
 ASSERT_REG_POSITION(instanced_arrays, 0x620);
 ASSERT_REG_POSITION(cull, 0x646);
 ASSERT_REG_POSITION(logic_op, 0x671);
 ASSERT_REG_POSITION(clear_buffers, 0x674);
+ASSERT_REG_POSITION(color_mask, 0x680);
 ASSERT_REG_POSITION(query, 0x6C0);
 ASSERT_REG_POSITION(vertex_array[0], 0x700);
 ASSERT_REG_POSITION(independent_blend, 0x780);
diff --git a/src/video_core/engines/shader_bytecode.h b/src/video_core/engines/shader_bytecode.h
index 6cd08d28b..83a6fd875 100644
--- a/src/video_core/engines/shader_bytecode.h
+++ b/src/video_core/engines/shader_bytecode.h
@@ -5,12 +5,11 @@
 #pragma once
 
 #include <bitset>
+#include <optional>
 #include <string>
 #include <tuple>
 #include <vector>
 
-#include <boost/optional.hpp>
-
 #include "common/assert.h"
 #include "common/bit_field.h"
 #include "common/common_types.h"
@@ -79,6 +78,7 @@ union Attribute {
     constexpr explicit Attribute(u64 value) : value(value) {}
 
     enum class Index : u64 {
+        PointSize = 6,
         Position = 7,
         Attribute_0 = 8,
         Attribute_31 = 39,
@@ -207,6 +207,16 @@ enum class UniformType : u64 {
     Double = 5,
 };
 
+enum class StoreType : u64 {
+    Unsigned8 = 0,
+    Signed8 = 1,
+    Unsigned16 = 2,
+    Signed16 = 3,
+    Bytes32 = 4,
+    Bytes64 = 5,
+    Bytes128 = 6,
+};
+
 enum class IMinMaxExchange : u64 {
     None = 0,
     XLo = 1,
@@ -568,6 +578,10 @@ union Instruction {
     } fmul32;
 
     union {
+        BitField<52, 1, u64> generates_cc;
+    } op_32;
+
+    union {
         BitField<48, 1, u64> is_signed;
     } shift;
 
@@ -747,6 +761,18 @@ union Instruction {
     } ld_c;
 
     union {
+        BitField<48, 3, StoreType> type;
+    } ldst_sl;
+
+    union {
+        BitField<44, 2, u64> unknown;
+    } ld_l;
+
+    union {
+        BitField<44, 2, u64> unknown;
+    } st_l;
+
+    union {
         BitField<0, 3, u64> pred0;
         BitField<3, 3, u64> pred3;
         BitField<7, 1, u64> abs_a;
@@ -1208,6 +1234,8 @@ union Instruction {
     BitField<61, 1, u64> is_b_imm;
     BitField<60, 1, u64> is_b_gpr;
     BitField<59, 1, u64> is_c_gpr;
+    BitField<20, 24, s64> smem_imm;
+    BitField<0, 5, ControlCode> flow_control_code;
 
     Attribute attribute;
     Sampler sampler;
@@ -1231,8 +1259,12 @@ public:
         BRA,
         PBK,
         LD_A,
+        LD_L,
+        LD_S,
         LD_C,
         ST_A,
+        ST_L,
+        ST_S,
         LDG, // Load from global memory
         STG, // Store in global memory
         TEX,
@@ -1428,7 +1460,7 @@ public:
         Type type;
     };
 
-    static boost::optional<const Matcher&> Decode(Instruction instr) {
+    static std::optional<std::reference_wrapper<const Matcher>> Decode(Instruction instr) {
         static const auto table{GetDecodeTable()};
 
         const auto matches_instruction = [instr](const auto& matcher) {
@@ -1436,7 +1468,8 @@ public:
         };
 
         auto iter = std::find_if(table.begin(), table.end(), matches_instruction);
-        return iter != table.end() ? boost::optional<const Matcher&>(*iter) : boost::none;
+        return iter != table.end() ? std::optional<std::reference_wrapper<const Matcher>>(*iter)
+                                   : std::nullopt;
     }
 
 private:
@@ -1489,8 +1522,12 @@ private:
             INST("111000110100---", Id::BRK, Type::Flow, "BRK"),
             INST("1111000011110---", Id::DEPBAR, Type::Synch, "DEPBAR"),
             INST("1110111111011---", Id::LD_A, Type::Memory, "LD_A"),
+            INST("1110111101001---", Id::LD_S, Type::Memory, "LD_S"),
+            INST("1110111101000---", Id::LD_L, Type::Memory, "LD_L"),
             INST("1110111110010---", Id::LD_C, Type::Memory, "LD_C"),
             INST("1110111111110---", Id::ST_A, Type::Memory, "ST_A"),
+            INST("1110111101011---", Id::ST_S, Type::Memory, "ST_S"),
+            INST("1110111101010---", Id::ST_L, Type::Memory, "ST_L"),
             INST("1110111011010---", Id::LDG, Type::Memory, "LDG"),
             INST("1110111011011---", Id::STG, Type::Memory, "STG"),
             INST("110000----111---", Id::TEX, Type::Memory, "TEX"),
@@ -1626,4 +1663,4 @@ private:
     }
 };
 
-} // namespace Tegra::Shader
\ No newline at end of file
+} // namespace Tegra::Shader
diff --git a/src/video_core/engines/shader_header.h b/src/video_core/engines/shader_header.h
index a885ee3cf..a0e015c4b 100644
--- a/src/video_core/engines/shader_header.h
+++ b/src/video_core/engines/shader_header.h
@@ -96,6 +96,11 @@ struct Header {
             }
         } ps;
     };
+
+    u64 GetLocalMemorySize() {
+        return (common1.shader_local_memory_low_size |
+                (common2.shader_local_memory_high_size << 24));
+    }
 };
 
 static_assert(sizeof(Header) == 0x50, "Incorrect structure size");
diff --git a/src/video_core/macro_interpreter.cpp b/src/video_core/macro_interpreter.cpp
index 377bd66ab..335a8d407 100644
--- a/src/video_core/macro_interpreter.cpp
+++ b/src/video_core/macro_interpreter.cpp
@@ -11,7 +11,7 @@ namespace Tegra {
 
 MacroInterpreter::MacroInterpreter(Engines::Maxwell3D& maxwell3d) : maxwell3d(maxwell3d) {}
 
-void MacroInterpreter::Execute(const std::vector<u32>& code, std::vector<u32> parameters) {
+void MacroInterpreter::Execute(u32 offset, std::vector<u32> parameters) {
     Reset();
     registers[1] = parameters[0];
     this->parameters = std::move(parameters);
@@ -19,7 +19,7 @@ void MacroInterpreter::Execute(const std::vector<u32>& code, std::vector<u32> pa
     // Execute the code until we hit an exit condition.
     bool keep_executing = true;
     while (keep_executing) {
-        keep_executing = Step(code, false);
+        keep_executing = Step(offset, false);
     }
 
     // Assert the the macro used all the input parameters
@@ -29,7 +29,7 @@ void MacroInterpreter::Execute(const std::vector<u32>& code, std::vector<u32> pa
 void MacroInterpreter::Reset() {
     registers = {};
     pc = 0;
-    delayed_pc = boost::none;
+    delayed_pc = {};
     method_address.raw = 0;
     parameters.clear();
     // The next parameter index starts at 1, because $r1 already has the value of the first
@@ -37,17 +37,17 @@ void MacroInterpreter::Reset() {
     next_parameter_index = 1;
 }
 
-bool MacroInterpreter::Step(const std::vector<u32>& code, bool is_delay_slot) {
+bool MacroInterpreter::Step(u32 offset, bool is_delay_slot) {
     u32 base_address = pc;
 
-    Opcode opcode = GetOpcode(code);
+    Opcode opcode = GetOpcode(offset);
     pc += 4;
 
     // Update the program counter if we were delayed
-    if (delayed_pc != boost::none) {
+    if (delayed_pc) {
         ASSERT(is_delay_slot);
         pc = *delayed_pc;
-        delayed_pc = boost::none;
+        delayed_pc = {};
     }
 
     switch (opcode.operation) {
@@ -108,7 +108,7 @@ bool MacroInterpreter::Step(const std::vector<u32>& code, bool is_delay_slot) {
 
             delayed_pc = base_address + opcode.GetBranchTarget();
             // Execute one more instruction due to the delay slot.
-            return Step(code, true);
+            return Step(offset, true);
         }
         break;
     }
@@ -121,17 +121,18 @@ bool MacroInterpreter::Step(const std::vector<u32>& code, bool is_delay_slot) {
         // Exit has a delay slot, execute the next instruction
         // Note: Executing an exit during a branch delay slot will cause the instruction at the
         // branch target to be executed before exiting.
-        Step(code, true);
+        Step(offset, true);
         return false;
     }
 
     return true;
 }
 
-MacroInterpreter::Opcode MacroInterpreter::GetOpcode(const std::vector<u32>& code) const {
+MacroInterpreter::Opcode MacroInterpreter::GetOpcode(u32 offset) const {
+    const auto& macro_memory{maxwell3d.GetMacroMemory()};
     ASSERT((pc % sizeof(u32)) == 0);
-    ASSERT(pc < code.size() * sizeof(u32));
-    return {code[pc / sizeof(u32)]};
+    ASSERT((pc + offset) < macro_memory.size() * sizeof(u32));
+    return {macro_memory[offset + pc / sizeof(u32)]};
 }
 
 u32 MacroInterpreter::GetALUResult(ALUOperation operation, u32 src_a, u32 src_b) const {
diff --git a/src/video_core/macro_interpreter.h b/src/video_core/macro_interpreter.h
index cee0baaf3..62d1ce289 100644
--- a/src/video_core/macro_interpreter.h
+++ b/src/video_core/macro_interpreter.h
@@ -5,8 +5,9 @@
 #pragma once
 
 #include <array>
+#include <optional>
 #include <vector>
-#include <boost/optional.hpp>
+
 #include "common/bit_field.h"
 #include "common/common_types.h"
 
@@ -21,10 +22,10 @@ public:
 
     /**
      * Executes the macro code with the specified input parameters.
-     * @param code The macro byte code to execute
-     * @param parameters The parameters of the macro
+     * @param offset Offset to start execution at.
+     * @param parameters The parameters of the macro.
      */
-    void Execute(const std::vector<u32>& code, std::vector<u32> parameters);
+    void Execute(u32 offset, std::vector<u32> parameters);
 
 private:
     enum class Operation : u32 {
@@ -109,11 +110,11 @@ private:
     /**
      * Executes a single macro instruction located at the current program counter. Returns whether
      * the interpreter should keep running.
-     * @param code The macro code to execute.
+     * @param offset Offset to start execution at.
      * @param is_delay_slot Whether the current step is being executed due to a delay slot in a
      * previous instruction.
      */
-    bool Step(const std::vector<u32>& code, bool is_delay_slot);
+    bool Step(u32 offset, bool is_delay_slot);
 
     /// Calculates the result of an ALU operation. src_a OP src_b;
     u32 GetALUResult(ALUOperation operation, u32 src_a, u32 src_b) const;
@@ -126,7 +127,7 @@ private:
     bool EvaluateBranchCondition(BranchCondition cond, u32 value) const;
 
     /// Reads an opcode at the current program counter location.
-    Opcode GetOpcode(const std::vector<u32>& code) const;
+    Opcode GetOpcode(u32 offset) const;
 
     /// Returns the specified register's value. Register 0 is hardcoded to always return 0.
     u32 GetRegister(u32 register_id) const;
@@ -149,7 +150,7 @@ private:
     Engines::Maxwell3D& maxwell3d;
 
     u32 pc; ///< Current program counter
-    boost::optional<u32>
+    std::optional<u32>
         delayed_pc; ///< Program counter to execute at after the delay slot is executed.
 
     static constexpr std::size_t NumMacroRegisters = 8;
diff --git a/src/video_core/memory_manager.cpp b/src/video_core/memory_manager.cpp
index 022d4ab74..77a20bb84 100644
--- a/src/video_core/memory_manager.cpp
+++ b/src/video_core/memory_manager.cpp
@@ -4,18 +4,21 @@
 
 #include "common/alignment.h"
 #include "common/assert.h"
+#include "common/logging/log.h"
 #include "video_core/memory_manager.h"
 
 namespace Tegra {
 
 GPUVAddr MemoryManager::AllocateSpace(u64 size, u64 align) {
-    boost::optional<GPUVAddr> gpu_addr = FindFreeBlock(size, align);
-    ASSERT(gpu_addr);
+    const std::optional<GPUVAddr> gpu_addr{FindFreeBlock(0, size, align, PageStatus::Unmapped)};
 
-    for (u64 offset = 0; offset < size; offset += PAGE_SIZE) {
-        VAddr& slot = PageSlot(*gpu_addr + offset);
+    ASSERT_MSG(gpu_addr, "unable to find available GPU memory");
+
+    for (u64 offset{}; offset < size; offset += PAGE_SIZE) {
+        VAddr& slot{PageSlot(*gpu_addr + offset)};
 
         ASSERT(slot == static_cast<u64>(PageStatus::Unmapped));
+
         slot = static_cast<u64>(PageStatus::Allocated);
     }
 
@@ -23,10 +26,11 @@ GPUVAddr MemoryManager::AllocateSpace(u64 size, u64 align) {
 }
 
 GPUVAddr MemoryManager::AllocateSpace(GPUVAddr gpu_addr, u64 size, u64 align) {
-    for (u64 offset = 0; offset < size; offset += PAGE_SIZE) {
-        VAddr& slot = PageSlot(gpu_addr + offset);
+    for (u64 offset{}; offset < size; offset += PAGE_SIZE) {
+        VAddr& slot{PageSlot(gpu_addr + offset)};
 
         ASSERT(slot == static_cast<u64>(PageStatus::Unmapped));
+
         slot = static_cast<u64>(PageStatus::Allocated);
     }
 
@@ -34,17 +38,19 @@ GPUVAddr MemoryManager::AllocateSpace(GPUVAddr gpu_addr, u64 size, u64 align) {
 }
 
 GPUVAddr MemoryManager::MapBufferEx(VAddr cpu_addr, u64 size) {
-    boost::optional<GPUVAddr> gpu_addr = FindFreeBlock(size, PAGE_SIZE);
-    ASSERT(gpu_addr);
+    const std::optional<GPUVAddr> gpu_addr{FindFreeBlock(0, size, PAGE_SIZE, PageStatus::Unmapped)};
+
+    ASSERT_MSG(gpu_addr, "unable to find available GPU memory");
 
-    for (u64 offset = 0; offset < size; offset += PAGE_SIZE) {
-        VAddr& slot = PageSlot(*gpu_addr + offset);
+    for (u64 offset{}; offset < size; offset += PAGE_SIZE) {
+        VAddr& slot{PageSlot(*gpu_addr + offset)};
 
         ASSERT(slot == static_cast<u64>(PageStatus::Unmapped));
+
         slot = cpu_addr + offset;
     }
 
-    MappedRegion region{cpu_addr, *gpu_addr, size};
+    const MappedRegion region{cpu_addr, *gpu_addr, size};
     mapped_regions.push_back(region);
 
     return *gpu_addr;
@@ -53,14 +59,31 @@ GPUVAddr MemoryManager::MapBufferEx(VAddr cpu_addr, u64 size) {
 GPUVAddr MemoryManager::MapBufferEx(VAddr cpu_addr, GPUVAddr gpu_addr, u64 size) {
     ASSERT((gpu_addr & PAGE_MASK) == 0);
 
-    for (u64 offset = 0; offset < size; offset += PAGE_SIZE) {
-        VAddr& slot = PageSlot(gpu_addr + offset);
+    if (PageSlot(gpu_addr) != static_cast<u64>(PageStatus::Allocated)) {
+        // Page has been already mapped. In this case, we must find a new area of memory to use that
+        // is different than the specified one. Super Mario Odyssey hits this scenario when changing
+        // areas, but we do not want to overwrite the old pages.
+        // TODO(bunnei): We need to write a hardware test to confirm this behavior.
+
+        LOG_ERROR(HW_GPU, "attempting to map addr 0x{:016X}, which is not available!", gpu_addr);
+
+        const std::optional<GPUVAddr> new_gpu_addr{
+            FindFreeBlock(gpu_addr, size, PAGE_SIZE, PageStatus::Allocated)};
+
+        ASSERT_MSG(new_gpu_addr, "unable to find available GPU memory");
+
+        gpu_addr = *new_gpu_addr;
+    }
+
+    for (u64 offset{}; offset < size; offset += PAGE_SIZE) {
+        VAddr& slot{PageSlot(gpu_addr + offset)};
 
         ASSERT(slot == static_cast<u64>(PageStatus::Allocated));
+
         slot = cpu_addr + offset;
     }
 
-    MappedRegion region{cpu_addr, gpu_addr, size};
+    const MappedRegion region{cpu_addr, gpu_addr, size};
     mapped_regions.push_back(region);
 
     return gpu_addr;
@@ -69,11 +92,12 @@ GPUVAddr MemoryManager::MapBufferEx(VAddr cpu_addr, GPUVAddr gpu_addr, u64 size)
 GPUVAddr MemoryManager::UnmapBuffer(GPUVAddr gpu_addr, u64 size) {
     ASSERT((gpu_addr & PAGE_MASK) == 0);
 
-    for (u64 offset = 0; offset < size; offset += PAGE_SIZE) {
-        VAddr& slot = PageSlot(gpu_addr + offset);
+    for (u64 offset{}; offset < size; offset += PAGE_SIZE) {
+        VAddr& slot{PageSlot(gpu_addr + offset)};
 
         ASSERT(slot != static_cast<u64>(PageStatus::Allocated) &&
                slot != static_cast<u64>(PageStatus::Unmapped));
+
         slot = static_cast<u64>(PageStatus::Unmapped);
     }
 
@@ -97,13 +121,14 @@ GPUVAddr MemoryManager::GetRegionEnd(GPUVAddr region_start) const {
     return {};
 }
 
-boost::optional<GPUVAddr> MemoryManager::FindFreeBlock(u64 size, u64 align) {
-    GPUVAddr gpu_addr = 0;
-    u64 free_space = 0;
+std::optional<GPUVAddr> MemoryManager::FindFreeBlock(GPUVAddr region_start, u64 size, u64 align,
+                                                     PageStatus status) {
+    GPUVAddr gpu_addr{region_start};
+    u64 free_space{};
     align = (align + PAGE_MASK) & ~PAGE_MASK;
 
     while (gpu_addr + free_space < MAX_ADDRESS) {
-        if (!IsPageMapped(gpu_addr + free_space)) {
+        if (PageSlot(gpu_addr + free_space) == static_cast<u64>(status)) {
             free_space += PAGE_SIZE;
             if (free_space >= size) {
                 return gpu_addr;
@@ -118,8 +143,8 @@ boost::optional<GPUVAddr> MemoryManager::FindFreeBlock(u64 size, u64 align) {
     return {};
 }
 
-boost::optional<VAddr> MemoryManager::GpuToCpuAddress(GPUVAddr gpu_addr) {
-    VAddr base_addr = PageSlot(gpu_addr);
+std::optional<VAddr> MemoryManager::GpuToCpuAddress(GPUVAddr gpu_addr) {
+    const VAddr base_addr{PageSlot(gpu_addr)};
 
     if (base_addr == static_cast<u64>(PageStatus::Allocated) ||
         base_addr == static_cast<u64>(PageStatus::Unmapped)) {
@@ -133,19 +158,15 @@ std::vector<GPUVAddr> MemoryManager::CpuToGpuAddress(VAddr cpu_addr) const {
     std::vector<GPUVAddr> results;
     for (const auto& region : mapped_regions) {
         if (cpu_addr >= region.cpu_addr && cpu_addr < (region.cpu_addr + region.size)) {
-            u64 offset = cpu_addr - region.cpu_addr;
+            const u64 offset{cpu_addr - region.cpu_addr};
             results.push_back(region.gpu_addr + offset);
         }
     }
     return results;
 }
 
-bool MemoryManager::IsPageMapped(GPUVAddr gpu_addr) {
-    return PageSlot(gpu_addr) != static_cast<u64>(PageStatus::Unmapped);
-}
-
 VAddr& MemoryManager::PageSlot(GPUVAddr gpu_addr) {
-    auto& block = page_table[(gpu_addr >> (PAGE_BITS + PAGE_TABLE_BITS)) & PAGE_TABLE_MASK];
+    auto& block{page_table[(gpu_addr >> (PAGE_BITS + PAGE_TABLE_BITS)) & PAGE_TABLE_MASK]};
     if (!block) {
         block = std::make_unique<PageBlock>();
         block->fill(static_cast<VAddr>(PageStatus::Unmapped));
diff --git a/src/video_core/memory_manager.h b/src/video_core/memory_manager.h
index caf80093f..4eb338aa2 100644
--- a/src/video_core/memory_manager.h
+++ b/src/video_core/memory_manager.h
@@ -6,10 +6,9 @@
 
 #include <array>
 #include <memory>
+#include <optional>
 #include <vector>
 
-#include <boost/optional.hpp>
-
 #include "common/common_types.h"
 
 namespace Tegra {
@@ -27,7 +26,7 @@ public:
     GPUVAddr MapBufferEx(VAddr cpu_addr, GPUVAddr gpu_addr, u64 size);
     GPUVAddr UnmapBuffer(GPUVAddr gpu_addr, u64 size);
     GPUVAddr GetRegionEnd(GPUVAddr region_start) const;
-    boost::optional<VAddr> GpuToCpuAddress(GPUVAddr gpu_addr);
+    std::optional<VAddr> GpuToCpuAddress(GPUVAddr gpu_addr);
     std::vector<GPUVAddr> CpuToGpuAddress(VAddr cpu_addr) const;
 
     static constexpr u64 PAGE_BITS = 16;
@@ -35,15 +34,15 @@ public:
     static constexpr u64 PAGE_MASK = PAGE_SIZE - 1;
 
 private:
-    boost::optional<GPUVAddr> FindFreeBlock(u64 size, u64 align = 1);
-    bool IsPageMapped(GPUVAddr gpu_addr);
-    VAddr& PageSlot(GPUVAddr gpu_addr);
-
     enum class PageStatus : u64 {
         Unmapped = 0xFFFFFFFFFFFFFFFFULL,
         Allocated = 0xFFFFFFFFFFFFFFFEULL,
     };
 
+    std::optional<GPUVAddr> FindFreeBlock(GPUVAddr region_start, u64 size, u64 align,
+                                          PageStatus status);
+    VAddr& PageSlot(GPUVAddr gpu_addr);
+
     static constexpr u64 MAX_ADDRESS{0x10000000000ULL};
     static constexpr u64 PAGE_TABLE_BITS{10};
     static constexpr u64 PAGE_TABLE_SIZE{1 << PAGE_TABLE_BITS};
diff --git a/src/video_core/rasterizer_cache.cpp b/src/video_core/rasterizer_cache.cpp
new file mode 100644
index 000000000..093b2cdf4
--- /dev/null
+++ b/src/video_core/rasterizer_cache.cpp
@@ -0,0 +1,7 @@
+// Copyright 2018 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "video_core/rasterizer_cache.h"
+
+RasterizerCacheObject::~RasterizerCacheObject() = default;
diff --git a/src/video_core/rasterizer_cache.h b/src/video_core/rasterizer_cache.h
index 0a3b3951e..bcf0c15a4 100644
--- a/src/video_core/rasterizer_cache.h
+++ b/src/video_core/rasterizer_cache.h
@@ -5,18 +5,19 @@
 #pragma once
 
 #include <set>
+#include <unordered_map>
 
 #include <boost/icl/interval_map.hpp>
 #include <boost/range/iterator_range_core.hpp>
 
 #include "common/common_types.h"
-#include "core/core.h"
 #include "core/settings.h"
 #include "video_core/rasterizer_interface.h"
-#include "video_core/renderer_base.h"
 
 class RasterizerCacheObject {
 public:
+    virtual ~RasterizerCacheObject();
+
     /// Gets the address of the shader in guest memory, required for cache management
     virtual VAddr GetAddr() const = 0;
 
@@ -64,6 +65,8 @@ class RasterizerCache : NonCopyable {
     friend class RasterizerCacheObject;
 
 public:
+    explicit RasterizerCache(VideoCore::RasterizerInterface& rasterizer) : rasterizer{rasterizer} {}
+
     /// Write any cached resources overlapping the specified region back to memory
     void FlushRegion(Tegra::GPUVAddr addr, size_t size) {
         const auto& objects{GetSortedObjectsFromRegion(addr, size)};
@@ -86,45 +89,39 @@ public:
 
     /// Invalidates everything in the cache
     void InvalidateAll() {
-        while (object_cache.begin() != object_cache.end()) {
-            Unregister(*object_cache.begin()->second.begin());
+        while (interval_cache.begin() != interval_cache.end()) {
+            Unregister(*interval_cache.begin()->second.begin());
         }
     }
 
 protected:
     /// Tries to get an object from the cache with the specified address
     T TryGet(VAddr addr) const {
-        const ObjectInterval interval{addr};
-        for (auto& pair : boost::make_iterator_range(object_cache.equal_range(interval))) {
-            for (auto& cached_object : pair.second) {
-                if (cached_object->GetAddr() == addr) {
-                    return cached_object;
-                }
-            }
-        }
+        const auto iter = map_cache.find(addr);
+        if (iter != map_cache.end())
+            return iter->second;
         return nullptr;
     }
 
     /// Register an object into the cache
     void Register(const T& object) {
         object->SetIsRegistered(true);
-        object_cache.add({GetInterval(object), ObjectSet{object}});
-        auto& rasterizer = Core::System::GetInstance().Renderer().Rasterizer();
+        interval_cache.add({GetInterval(object), ObjectSet{object}});
+        map_cache.insert({object->GetAddr(), object});
         rasterizer.UpdatePagesCachedCount(object->GetAddr(), object->GetSizeInBytes(), 1);
     }
 
     /// Unregisters an object from the cache
     void Unregister(const T& object) {
         object->SetIsRegistered(false);
-        auto& rasterizer = Core::System::GetInstance().Renderer().Rasterizer();
         rasterizer.UpdatePagesCachedCount(object->GetAddr(), object->GetSizeInBytes(), -1);
-
         // Only flush if use_accurate_gpu_emulation is enabled, as it incurs a performance hit
         if (Settings::values.use_accurate_gpu_emulation) {
             FlushObject(object);
         }
 
-        object_cache.subtract({GetInterval(object), ObjectSet{object}});
+        interval_cache.subtract({GetInterval(object), ObjectSet{object}});
+        map_cache.erase(object->GetAddr());
     }
 
     /// Returns a ticks counter used for tracking when cached objects were last modified
@@ -141,7 +138,7 @@ private:
 
         std::vector<T> objects;
         const ObjectInterval interval{addr, addr + size};
-        for (auto& pair : boost::make_iterator_range(object_cache.equal_range(interval))) {
+        for (auto& pair : boost::make_iterator_range(interval_cache.equal_range(interval))) {
             for (auto& cached_object : pair.second) {
                 if (!cached_object) {
                     continue;
@@ -167,14 +164,17 @@ private:
     }
 
     using ObjectSet = std::set<T>;
-    using ObjectCache = boost::icl::interval_map<VAddr, ObjectSet>;
-    using ObjectInterval = typename ObjectCache::interval_type;
+    using ObjectCache = std::unordered_map<VAddr, T>;
+    using IntervalCache = boost::icl::interval_map<VAddr, ObjectSet>;
+    using ObjectInterval = typename IntervalCache::interval_type;
 
     static auto GetInterval(const T& object) {
         return ObjectInterval::right_open(object->GetAddr(),
                                           object->GetAddr() + object->GetSizeInBytes());
     }
 
-    ObjectCache object_cache; ///< Cache of objects
-    u64 modified_ticks{};     ///< Counter of cache state ticks, used for in-order flushing
+    ObjectCache map_cache;
+    IntervalCache interval_cache; ///< Cache of objects
+    u64 modified_ticks{};         ///< Counter of cache state ticks, used for in-order flushing
+    VideoCore::RasterizerInterface& rasterizer;
 };
diff --git a/src/video_core/renderer_base.cpp b/src/video_core/renderer_base.cpp
index 0df3725c2..1482cdb40 100644
--- a/src/video_core/renderer_base.cpp
+++ b/src/video_core/renderer_base.cpp
@@ -5,7 +5,6 @@
 #include "core/frontend/emu_window.h"
 #include "core/settings.h"
 #include "video_core/renderer_base.h"
-#include "video_core/renderer_opengl/gl_rasterizer.h"
 
 namespace VideoCore {
 
diff --git a/src/video_core/renderer_base.h b/src/video_core/renderer_base.h
index 2cd0738ff..669e26e15 100644
--- a/src/video_core/renderer_base.h
+++ b/src/video_core/renderer_base.h
@@ -6,7 +6,8 @@
 
 #include <atomic>
 #include <memory>
-#include <boost/optional.hpp>
+#include <optional>
+
 #include "common/common_types.h"
 #include "video_core/gpu.h"
 #include "video_core/rasterizer_interface.h"
@@ -28,7 +29,8 @@ public:
     virtual ~RendererBase();
 
     /// Swap buffers (render frame)
-    virtual void SwapBuffers(boost::optional<const Tegra::FramebufferConfig&> framebuffer) = 0;
+    virtual void SwapBuffers(
+        std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) = 0;
 
     /// Initialize the renderer
     virtual bool Init() = 0;
diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.cpp b/src/video_core/renderer_opengl/gl_buffer_cache.cpp
index c142095c5..075192c3f 100644
--- a/src/video_core/renderer_opengl/gl_buffer_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_buffer_cache.cpp
@@ -9,15 +9,17 @@
 #include "core/core.h"
 #include "core/memory.h"
 #include "video_core/renderer_opengl/gl_buffer_cache.h"
+#include "video_core/renderer_opengl/gl_rasterizer.h"
 
 namespace OpenGL {
 
-OGLBufferCache::OGLBufferCache(std::size_t size) : stream_buffer(GL_ARRAY_BUFFER, size) {}
+OGLBufferCache::OGLBufferCache(RasterizerOpenGL& rasterizer, std::size_t size)
+    : RasterizerCache{rasterizer}, stream_buffer(GL_ARRAY_BUFFER, size) {}
 
 GLintptr OGLBufferCache::UploadMemory(Tegra::GPUVAddr gpu_addr, std::size_t size,
                                       std::size_t alignment, bool cache) {
     auto& memory_manager = Core::System::GetInstance().GPU().MemoryManager();
-    const boost::optional<VAddr> cpu_addr{memory_manager.GpuToCpuAddress(gpu_addr)};
+    const std::optional<VAddr> cpu_addr{memory_manager.GpuToCpuAddress(gpu_addr)};
 
     // Cache management is a big overhead, so only cache entries with a given size.
     // TODO: Figure out which size is the best for given games.
diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.h b/src/video_core/renderer_opengl/gl_buffer_cache.h
index be29dc8be..91fca3f6c 100644
--- a/src/video_core/renderer_opengl/gl_buffer_cache.h
+++ b/src/video_core/renderer_opengl/gl_buffer_cache.h
@@ -15,6 +15,8 @@
 
 namespace OpenGL {
 
+class RasterizerOpenGL;
+
 struct CachedBufferEntry final : public RasterizerCacheObject {
     VAddr GetAddr() const override {
         return addr;
@@ -35,7 +37,7 @@ struct CachedBufferEntry final : public RasterizerCacheObject {
 
 class OGLBufferCache final : public RasterizerCache<std::shared_ptr<CachedBufferEntry>> {
 public:
-    explicit OGLBufferCache(std::size_t size);
+    explicit OGLBufferCache(RasterizerOpenGL& rasterizer, std::size_t size);
 
     /// Uploads data from a guest GPU address. Returns host's buffer offset where it's been
     /// allocated.
diff --git a/src/video_core/renderer_opengl/gl_primitive_assembler.cpp b/src/video_core/renderer_opengl/gl_primitive_assembler.cpp
index ee1d9601b..d9ed08437 100644
--- a/src/video_core/renderer_opengl/gl_primitive_assembler.cpp
+++ b/src/video_core/renderer_opengl/gl_primitive_assembler.cpp
@@ -6,6 +6,7 @@
 #include <array>
 #include "common/assert.h"
 #include "common/common_types.h"
+#include "core/core.h"
 #include "core/memory.h"
 #include "video_core/renderer_opengl/gl_buffer_cache.h"
 #include "video_core/renderer_opengl/gl_primitive_assembler.h"
@@ -45,7 +46,7 @@ GLintptr PrimitiveAssembler::MakeQuadIndexed(Tegra::GPUVAddr gpu_addr, std::size
     auto [dst_pointer, index_offset] = buffer_cache.ReserveMemory(map_size);
 
     auto& memory_manager = Core::System::GetInstance().GPU().MemoryManager();
-    const boost::optional<VAddr> cpu_addr{memory_manager.GpuToCpuAddress(gpu_addr)};
+    const std::optional<VAddr> cpu_addr{memory_manager.GpuToCpuAddress(gpu_addr)};
     const u8* source{Memory::GetPointer(*cpu_addr)};
 
     for (u32 primitive = 0; primitive < count / 4; ++primitive) {
diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp
index be51c5215..84bd91eed 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer.cpp
+++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp
@@ -30,10 +30,11 @@
 namespace OpenGL {
 
 using Maxwell = Tegra::Engines::Maxwell3D::Regs;
-using PixelFormat = SurfaceParams::PixelFormat;
-using SurfaceType = SurfaceParams::SurfaceType;
+using PixelFormat = VideoCore::Surface::PixelFormat;
+using SurfaceType = VideoCore::Surface::SurfaceType;
 
-MICROPROFILE_DEFINE(OpenGL_VAO, "OpenGL", "Vertex Array Setup", MP_RGB(128, 128, 192));
+MICROPROFILE_DEFINE(OpenGL_VAO, "OpenGL", "Vertex Format Setup", MP_RGB(128, 128, 192));
+MICROPROFILE_DEFINE(OpenGL_VB, "OpenGL", "Vertex Buffer Setup", MP_RGB(128, 128, 192));
 MICROPROFILE_DEFINE(OpenGL_Shader, "OpenGL", "Shader Setup", MP_RGB(128, 128, 192));
 MICROPROFILE_DEFINE(OpenGL_UBO, "OpenGL", "Const Buffer Setup", MP_RGB(128, 128, 192));
 MICROPROFILE_DEFINE(OpenGL_Index, "OpenGL", "Index Buffer Setup", MP_RGB(128, 128, 192));
@@ -79,7 +80,8 @@ struct DrawParameters {
 };
 
 RasterizerOpenGL::RasterizerOpenGL(Core::Frontend::EmuWindow& window, ScreenInfo& info)
-    : emu_window{window}, screen_info{info}, buffer_cache(STREAM_BUFFER_SIZE) {
+    : res_cache{*this}, shader_cache{*this}, emu_window{window}, screen_info{info},
+      buffer_cache(*this, STREAM_BUFFER_SIZE) {
     // Create sampler objects
     for (std::size_t i = 0; i < texture_samplers.size(); ++i) {
         texture_samplers[i].Create();
@@ -104,9 +106,7 @@ RasterizerOpenGL::RasterizerOpenGL(Core::Frontend::EmuWindow& window, ScreenInfo
     }
 
     ASSERT_MSG(has_ARB_separate_shader_objects, "has_ARB_separate_shader_objects is unsupported");
-
-    // Clipping plane 0 is always enabled for PICA fixed clip plane z <= 0
-    state.clip_distance[0] = true;
+    OpenGLState::ApplyDefaultState();
 
     // Create render framebuffer
     framebuffer.Create();
@@ -115,8 +115,6 @@ RasterizerOpenGL::RasterizerOpenGL(Core::Frontend::EmuWindow& window, ScreenInfo
     state.draw.shader_program = 0;
     state.Apply();
 
-    glEnable(GL_BLEND);
-
     glGetIntegerv(GL_UNIFORM_BUFFER_OFFSET_ALIGNMENT, &uniform_buffer_alignment);
 
     LOG_CRITICAL(Render_OpenGL, "Sync fixed function OpenGL state here!");
@@ -124,18 +122,23 @@ RasterizerOpenGL::RasterizerOpenGL(Core::Frontend::EmuWindow& window, ScreenInfo
 
 RasterizerOpenGL::~RasterizerOpenGL() {}
 
-void RasterizerOpenGL::SetupVertexArrays() {
-    MICROPROFILE_SCOPE(OpenGL_VAO);
-    const auto& gpu = Core::System::GetInstance().GPU().Maxwell3D();
+void RasterizerOpenGL::SetupVertexFormat() {
+    auto& gpu = Core::System::GetInstance().GPU().Maxwell3D();
     const auto& regs = gpu.regs;
 
+    if (!gpu.dirty_flags.vertex_attrib_format)
+        return;
+    gpu.dirty_flags.vertex_attrib_format = false;
+
+    MICROPROFILE_SCOPE(OpenGL_VAO);
+
     auto [iter, is_cache_miss] = vertex_array_cache.try_emplace(regs.vertex_attrib_format);
     auto& VAO = iter->second;
 
     if (is_cache_miss) {
         VAO.Create();
         state.draw.vertex_array = VAO.handle;
-        state.Apply();
+        state.ApplyVertexBufferState();
 
         // The index buffer binding is stored within the VAO. Stupid OpenGL, but easy to work
         // around.
@@ -177,8 +180,13 @@ void RasterizerOpenGL::SetupVertexArrays() {
         }
     }
     state.draw.vertex_array = VAO.handle;
-    state.draw.vertex_buffer = buffer_cache.GetHandle();
-    state.Apply();
+    state.ApplyVertexBufferState();
+}
+
+void RasterizerOpenGL::SetupVertexBuffer() {
+    MICROPROFILE_SCOPE(OpenGL_VB);
+    const auto& gpu = Core::System::GetInstance().GPU().Maxwell3D();
+    const auto& regs = gpu.regs;
 
     // Upload all guest vertex arrays sequentially to our buffer
     for (u32 index = 0; index < Maxwell::NumVertexArrays; ++index) {
@@ -205,6 +213,9 @@ void RasterizerOpenGL::SetupVertexArrays() {
             glVertexBindingDivisor(index, 0);
         }
     }
+
+    // Implicit set by glBindVertexBuffer. Stupid glstate handling...
+    state.draw.vertex_buffer = buffer_cache.GetHandle();
 }
 
 DrawParameters RasterizerOpenGL::SetupDraw() {
@@ -329,8 +340,6 @@ void RasterizerOpenGL::SetupShaders(GLenum primitive_mode) {
             index++;
         }
     }
-
-    state.Apply();
 }
 
 std::size_t RasterizerOpenGL::CalculateVertexArraysSize() const {
@@ -399,9 +408,9 @@ void RasterizerOpenGL::UpdatePagesCachedCount(VAddr addr, u64 size, int delta) {
         cached_pages.add({pages_interval, delta});
 }
 
-void RasterizerOpenGL::ConfigureFramebuffers(bool using_color_fb, bool using_depth_fb,
-                                             bool preserve_contents,
-                                             boost::optional<std::size_t> single_color_target) {
+void RasterizerOpenGL::ConfigureFramebuffers(OpenGLState& current_state, bool using_color_fb,
+                                             bool using_depth_fb, bool preserve_contents,
+                                             std::optional<std::size_t> single_color_target) {
     MICROPROFILE_SCOPE(OpenGL_Framebuffer);
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
 
@@ -416,8 +425,9 @@ void RasterizerOpenGL::ConfigureFramebuffers(bool using_color_fb, bool using_dep
     ASSERT_MSG(regs.rt_separate_frag_data == 0, "Unimplemented");
 
     // Bind the framebuffer surfaces
-    state.draw.draw_framebuffer = framebuffer.handle;
-    state.Apply();
+    current_state.draw.draw_framebuffer = framebuffer.handle;
+    current_state.ApplyFramebufferState();
+    current_state.framebuffer_srgb.enabled = regs.framebuffer_srgb != 0;
 
     if (using_color_fb) {
         if (single_color_target) {
@@ -429,6 +439,9 @@ void RasterizerOpenGL::ConfigureFramebuffers(bool using_color_fb, bool using_dep
                 // Assume that a surface will be written to if it is used as a framebuffer, even if
                 // the shader doesn't actually write to it.
                 color_surface->MarkAsModified(true, res_cache);
+                // Workaround for and issue in nvidia drivers
+                // https://devtalk.nvidia.com/default/topic/776591/opengl/gl_framebuffer_srgb-functions-incorrectly/
+                state.framebuffer_srgb.enabled |= color_surface->GetSurfaceParams().srgb_conversion;
             }
 
             glFramebufferTexture2D(
@@ -446,6 +459,11 @@ void RasterizerOpenGL::ConfigureFramebuffers(bool using_color_fb, bool using_dep
                     // Assume that a surface will be written to if it is used as a framebuffer, even
                     // if the shader doesn't actually write to it.
                     color_surface->MarkAsModified(true, res_cache);
+                    // Enable sRGB only for supported formats
+                    // Workaround for and issue in nvidia drivers
+                    // https://devtalk.nvidia.com/default/topic/776591/opengl/gl_framebuffer_srgb-functions-incorrectly/
+                    state.framebuffer_srgb.enabled |=
+                        color_surface->GetSurfaceParams().srgb_conversion;
                 }
 
                 buffers[index] = GL_COLOR_ATTACHMENT0 + regs.rt_control.GetMap(index);
@@ -487,10 +505,7 @@ void RasterizerOpenGL::ConfigureFramebuffers(bool using_color_fb, bool using_dep
         glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER, GL_DEPTH_STENCIL_ATTACHMENT, GL_TEXTURE_2D, 0,
                                0);
     }
-
-    SyncViewport();
-
-    state.Apply();
+    SyncViewport(current_state);
 }
 
 void RasterizerOpenGL::Clear() {
@@ -503,22 +518,23 @@ void RasterizerOpenGL::Clear() {
     bool use_stencil{};
 
     OpenGLState clear_state;
-    clear_state.draw.draw_framebuffer = framebuffer.handle;
-    clear_state.color_mask.red_enabled = regs.clear_buffers.R ? GL_TRUE : GL_FALSE;
-    clear_state.color_mask.green_enabled = regs.clear_buffers.G ? GL_TRUE : GL_FALSE;
-    clear_state.color_mask.blue_enabled = regs.clear_buffers.B ? GL_TRUE : GL_FALSE;
-    clear_state.color_mask.alpha_enabled = regs.clear_buffers.A ? GL_TRUE : GL_FALSE;
-
     if (regs.clear_buffers.R || regs.clear_buffers.G || regs.clear_buffers.B ||
         regs.clear_buffers.A) {
         use_color = true;
     }
+    if (use_color) {
+        clear_state.color_mask[0].red_enabled = regs.clear_buffers.R ? GL_TRUE : GL_FALSE;
+        clear_state.color_mask[0].green_enabled = regs.clear_buffers.G ? GL_TRUE : GL_FALSE;
+        clear_state.color_mask[0].blue_enabled = regs.clear_buffers.B ? GL_TRUE : GL_FALSE;
+        clear_state.color_mask[0].alpha_enabled = regs.clear_buffers.A ? GL_TRUE : GL_FALSE;
+    }
     if (regs.clear_buffers.Z) {
         ASSERT_MSG(regs.zeta_enable != 0, "Tried to clear Z but buffer is not enabled!");
         use_depth = true;
 
         // Always enable the depth write when clearing the depth buffer. The depth write mask is
-        // ignored when clearing the buffer in the Switch, but OpenGL obeys it so we set it to true.
+        // ignored when clearing the buffer in the Switch, but OpenGL obeys it so we set it to
+        // true.
         clear_state.depth.test_enabled = true;
         clear_state.depth.test_func = GL_ALWAYS;
     }
@@ -535,9 +551,8 @@ void RasterizerOpenGL::Clear() {
 
     ScopeAcquireGLContext acquire_context{emu_window};
 
-    ConfigureFramebuffers(use_color, use_depth || use_stencil, false,
+    ConfigureFramebuffers(clear_state, use_color, use_depth || use_stencil, false,
                           regs.clear_buffers.RT.Value());
-
     clear_state.Apply();
 
     if (use_color) {
@@ -563,13 +578,14 @@ void RasterizerOpenGL::DrawArrays() {
 
     ScopeAcquireGLContext acquire_context{emu_window};
 
-    ConfigureFramebuffers();
-
+    ConfigureFramebuffers(state);
+    SyncColorMask();
     SyncDepthTestState();
     SyncStencilTestState();
     SyncBlendState();
     SyncLogicOpState();
     SyncCullMode();
+    SyncPrimitiveRestart();
     SyncScissorTest();
     // Alpha Testing is synced on shaders.
     SyncTransformFeedback();
@@ -583,7 +599,7 @@ void RasterizerOpenGL::DrawArrays() {
     const bool is_indexed = accelerate_draw == AccelDraw::Indexed;
 
     state.draw.vertex_buffer = buffer_cache.GetHandle();
-    state.Apply();
+    state.ApplyVertexBufferState();
 
     std::size_t buffer_size = CalculateVertexArraysSize();
 
@@ -610,7 +626,8 @@ void RasterizerOpenGL::DrawArrays() {
 
     buffer_cache.Map(buffer_size);
 
-    SetupVertexArrays();
+    SetupVertexFormat();
+    SetupVertexBuffer();
     DrawParameters params = SetupDraw();
     SetupShaders(params.primitive_mode);
 
@@ -690,7 +707,8 @@ bool RasterizerOpenGL::AccelerateDisplay(const Tegra::FramebufferConfig& config,
 
     // Verify that the cached surface is the same size and format as the requested framebuffer
     const auto& params{surface->GetSurfaceParams()};
-    const auto& pixel_format{SurfaceParams::PixelFormatFromGPUPixelFormat(config.pixel_format)};
+    const auto& pixel_format{
+        VideoCore::Surface::PixelFormatFromGPUPixelFormat(config.pixel_format)};
     ASSERT_MSG(params.width == config.width, "Framebuffer width is different");
     ASSERT_MSG(params.height == config.height, "Framebuffer height is different");
     ASSERT_MSG(params.pixel_format == pixel_format, "Framebuffer pixel_format is different");
@@ -713,16 +731,20 @@ void RasterizerOpenGL::SamplerInfo::Create() {
     glSamplerParameteri(sampler.handle, GL_TEXTURE_COMPARE_FUNC, GL_NEVER);
 }
 
-void RasterizerOpenGL::SamplerInfo::SyncWithConfig(const Tegra::Texture::TSCEntry& config) {
+void RasterizerOpenGL::SamplerInfo::SyncWithConfig(const Tegra::Texture::FullTextureInfo& info) {
     const GLuint s = sampler.handle;
-
+    const Tegra::Texture::TSCEntry& config = info.tsc;
     if (mag_filter != config.mag_filter) {
         mag_filter = config.mag_filter;
-        glSamplerParameteri(s, GL_TEXTURE_MAG_FILTER, MaxwellToGL::TextureFilterMode(mag_filter));
+        glSamplerParameteri(
+            s, GL_TEXTURE_MAG_FILTER,
+            MaxwellToGL::TextureFilterMode(mag_filter, Tegra::Texture::TextureMipmapFilter::None));
     }
-    if (min_filter != config.min_filter) {
+    if (min_filter != config.min_filter || mip_filter != config.mip_filter) {
         min_filter = config.min_filter;
-        glSamplerParameteri(s, GL_TEXTURE_MIN_FILTER, MaxwellToGL::TextureFilterMode(min_filter));
+        mip_filter = config.mip_filter;
+        glSamplerParameteri(s, GL_TEXTURE_MIN_FILTER,
+                            MaxwellToGL::TextureFilterMode(min_filter, mip_filter));
     }
 
     if (wrap_u != config.wrap_u) {
@@ -762,6 +784,22 @@ void RasterizerOpenGL::SamplerInfo::SyncWithConfig(const Tegra::Texture::TSCEntr
             glSamplerParameterfv(s, GL_TEXTURE_BORDER_COLOR, border_color.data());
         }
     }
+    if (info.tic.use_header_opt_control == 0) {
+        if (GLAD_GL_ARB_texture_filter_anisotropic) {
+            glSamplerParameterf(s, GL_TEXTURE_MAX_ANISOTROPY,
+                                static_cast<float>(1 << info.tic.max_anisotropy.Value()));
+        } else if (GLAD_GL_EXT_texture_filter_anisotropic) {
+            glSamplerParameterf(s, GL_TEXTURE_MAX_ANISOTROPY_EXT,
+                                static_cast<float>(1 << info.tic.max_anisotropy.Value()));
+        }
+        glSamplerParameterf(s, GL_TEXTURE_MIN_LOD,
+                            static_cast<float>(info.tic.res_min_mip_level.Value()));
+        glSamplerParameterf(s, GL_TEXTURE_MAX_LOD,
+                            static_cast<float>(info.tic.res_max_mip_level.Value() == 0
+                                                   ? 16
+                                                   : info.tic.res_max_mip_level.Value()));
+        glSamplerParameterf(s, GL_TEXTURE_LOD_BIAS, info.tic.mip_lod_bias.Value() / 256.f);
+    }
 }
 
 u32 RasterizerOpenGL::SetupConstBuffers(Maxwell::ShaderStage stage, Shader& shader,
@@ -859,7 +897,7 @@ u32 RasterizerOpenGL::SetupTextures(Maxwell::ShaderStage stage, Shader& shader,
             continue;
         }
 
-        texture_samplers[current_bindpoint].SyncWithConfig(texture.tsc);
+        texture_samplers[current_bindpoint].SyncWithConfig(texture);
         Surface surface = res_cache.GetTextureSurface(texture, entry);
         if (surface != nullptr) {
             state.texture_units[current_bindpoint].texture = surface->Texture().handle;
@@ -881,14 +919,18 @@ u32 RasterizerOpenGL::SetupTextures(Maxwell::ShaderStage stage, Shader& shader,
     return current_unit + static_cast<u32>(entries.size());
 }
 
-void RasterizerOpenGL::SyncViewport() {
+void RasterizerOpenGL::SyncViewport(OpenGLState& current_state) {
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
-    const MathUtil::Rectangle<s32> viewport_rect{regs.viewport_transform[0].GetRect()};
-
-    state.viewport.x = viewport_rect.left;
-    state.viewport.y = viewport_rect.bottom;
-    state.viewport.width = static_cast<GLsizei>(viewport_rect.GetWidth());
-    state.viewport.height = static_cast<GLsizei>(viewport_rect.GetHeight());
+    for (size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
+        const MathUtil::Rectangle<s32> viewport_rect{regs.viewport_transform[i].GetRect()};
+        auto& viewport = current_state.viewports[i];
+        viewport.x = viewport_rect.left;
+        viewport.y = viewport_rect.bottom;
+        viewport.width = static_cast<GLfloat>(viewport_rect.GetWidth());
+        viewport.height = static_cast<GLfloat>(viewport_rect.GetHeight());
+        viewport.depth_range_far = regs.viewport[i].depth_range_far;
+        viewport.depth_range_near = regs.viewport[i].depth_range_near;
+    }
 }
 
 void RasterizerOpenGL::SyncClipEnabled() {
@@ -923,12 +965,11 @@ void RasterizerOpenGL::SyncCullMode() {
     }
 }
 
-void RasterizerOpenGL::SyncDepthScale() {
-    UNREACHABLE();
-}
+void RasterizerOpenGL::SyncPrimitiveRestart() {
+    const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
 
-void RasterizerOpenGL::SyncDepthOffset() {
-    UNREACHABLE();
+    state.primitive_restart.enabled = regs.primitive_restart.enabled;
+    state.primitive_restart.index = regs.primitive_restart.index;
 }
 
 void RasterizerOpenGL::SyncDepthTestState() {
@@ -951,9 +992,6 @@ void RasterizerOpenGL::SyncStencilTestState() {
         return;
     }
 
-    // TODO(bunnei): Verify behavior when this is not set
-    ASSERT(regs.stencil_two_side_enable);
-
     state.stencil.front.test_func = MaxwellToGL::ComparisonOp(regs.stencil_front_func_func);
     state.stencil.front.test_ref = regs.stencil_front_func_ref;
     state.stencil.front.test_mask = regs.stencil_front_func_mask;
@@ -961,36 +999,79 @@ void RasterizerOpenGL::SyncStencilTestState() {
     state.stencil.front.action_depth_fail = MaxwellToGL::StencilOp(regs.stencil_front_op_zfail);
     state.stencil.front.action_depth_pass = MaxwellToGL::StencilOp(regs.stencil_front_op_zpass);
     state.stencil.front.write_mask = regs.stencil_front_mask;
+    if (regs.stencil_two_side_enable) {
+        state.stencil.back.test_func = MaxwellToGL::ComparisonOp(regs.stencil_back_func_func);
+        state.stencil.back.test_ref = regs.stencil_back_func_ref;
+        state.stencil.back.test_mask = regs.stencil_back_func_mask;
+        state.stencil.back.action_stencil_fail = MaxwellToGL::StencilOp(regs.stencil_back_op_fail);
+        state.stencil.back.action_depth_fail = MaxwellToGL::StencilOp(regs.stencil_back_op_zfail);
+        state.stencil.back.action_depth_pass = MaxwellToGL::StencilOp(regs.stencil_back_op_zpass);
+        state.stencil.back.write_mask = regs.stencil_back_mask;
+    } else {
+        state.stencil.back.test_func = GL_ALWAYS;
+        state.stencil.back.test_ref = 0;
+        state.stencil.back.test_mask = 0xFFFFFFFF;
+        state.stencil.back.write_mask = 0xFFFFFFFF;
+        state.stencil.back.action_stencil_fail = GL_KEEP;
+        state.stencil.back.action_depth_fail = GL_KEEP;
+        state.stencil.back.action_depth_pass = GL_KEEP;
+    }
+}
 
-    state.stencil.back.test_func = MaxwellToGL::ComparisonOp(regs.stencil_back_func_func);
-    state.stencil.back.test_ref = regs.stencil_back_func_ref;
-    state.stencil.back.test_mask = regs.stencil_back_func_mask;
-    state.stencil.back.action_stencil_fail = MaxwellToGL::StencilOp(regs.stencil_back_op_fail);
-    state.stencil.back.action_depth_fail = MaxwellToGL::StencilOp(regs.stencil_back_op_zfail);
-    state.stencil.back.action_depth_pass = MaxwellToGL::StencilOp(regs.stencil_back_op_zpass);
-    state.stencil.back.write_mask = regs.stencil_back_mask;
+void RasterizerOpenGL::SyncColorMask() {
+    const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
+    for (size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
+        const auto& source = regs.color_mask[regs.color_mask_common ? 0 : i];
+        auto& dest = state.color_mask[i];
+        dest.red_enabled = (source.R == 0) ? GL_FALSE : GL_TRUE;
+        dest.green_enabled = (source.G == 0) ? GL_FALSE : GL_TRUE;
+        dest.blue_enabled = (source.B == 0) ? GL_FALSE : GL_TRUE;
+        dest.alpha_enabled = (source.A == 0) ? GL_FALSE : GL_TRUE;
+    }
 }
 
 void RasterizerOpenGL::SyncBlendState() {
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
 
-    // TODO(Subv): Support more than just render target 0.
-    state.blend.enabled = regs.blend.enable[0] != 0;
-
-    if (!state.blend.enabled)
+    state.blend_color.red = regs.blend_color.r;
+    state.blend_color.green = regs.blend_color.g;
+    state.blend_color.blue = regs.blend_color.b;
+    state.blend_color.alpha = regs.blend_color.a;
+
+    state.independant_blend.enabled = regs.independent_blend_enable;
+    if (!state.independant_blend.enabled) {
+        auto& blend = state.blend[0];
+        blend.enabled = regs.blend.enable[0] != 0;
+        blend.separate_alpha = regs.blend.separate_alpha;
+        blend.rgb_equation = MaxwellToGL::BlendEquation(regs.blend.equation_rgb);
+        blend.src_rgb_func = MaxwellToGL::BlendFunc(regs.blend.factor_source_rgb);
+        blend.dst_rgb_func = MaxwellToGL::BlendFunc(regs.blend.factor_dest_rgb);
+        if (blend.separate_alpha) {
+            blend.a_equation = MaxwellToGL::BlendEquation(regs.blend.equation_a);
+            blend.src_a_func = MaxwellToGL::BlendFunc(regs.blend.factor_source_a);
+            blend.dst_a_func = MaxwellToGL::BlendFunc(regs.blend.factor_dest_a);
+        }
+        for (size_t i = 1; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
+            state.blend[i].enabled = false;
+        }
         return;
+    }
 
-    ASSERT_MSG(regs.logic_op.enable == 0,
-               "Blending and logic op can't be enabled at the same time.");
-
-    ASSERT_MSG(regs.independent_blend_enable == 1, "Only independent blending is implemented");
-    ASSERT_MSG(!regs.independent_blend[0].separate_alpha, "Unimplemented");
-    state.blend.rgb_equation = MaxwellToGL::BlendEquation(regs.independent_blend[0].equation_rgb);
-    state.blend.src_rgb_func = MaxwellToGL::BlendFunc(regs.independent_blend[0].factor_source_rgb);
-    state.blend.dst_rgb_func = MaxwellToGL::BlendFunc(regs.independent_blend[0].factor_dest_rgb);
-    state.blend.a_equation = MaxwellToGL::BlendEquation(regs.independent_blend[0].equation_a);
-    state.blend.src_a_func = MaxwellToGL::BlendFunc(regs.independent_blend[0].factor_source_a);
-    state.blend.dst_a_func = MaxwellToGL::BlendFunc(regs.independent_blend[0].factor_dest_a);
+    for (size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
+        auto& blend = state.blend[i];
+        blend.enabled = regs.blend.enable[i] != 0;
+        if (!blend.enabled)
+            continue;
+        blend.separate_alpha = regs.independent_blend[i].separate_alpha;
+        blend.rgb_equation = MaxwellToGL::BlendEquation(regs.independent_blend[i].equation_rgb);
+        blend.src_rgb_func = MaxwellToGL::BlendFunc(regs.independent_blend[i].factor_source_rgb);
+        blend.dst_rgb_func = MaxwellToGL::BlendFunc(regs.independent_blend[i].factor_dest_rgb);
+        if (blend.separate_alpha) {
+            blend.a_equation = MaxwellToGL::BlendEquation(regs.independent_blend[i].equation_a);
+            blend.src_a_func = MaxwellToGL::BlendFunc(regs.independent_blend[i].factor_source_a);
+            blend.dst_a_func = MaxwellToGL::BlendFunc(regs.independent_blend[i].factor_dest_a);
+        }
+    }
 }
 
 void RasterizerOpenGL::SyncLogicOpState() {
@@ -1009,19 +1090,19 @@ void RasterizerOpenGL::SyncLogicOpState() {
 }
 
 void RasterizerOpenGL::SyncScissorTest() {
+    // TODO: what is the correct behavior here, a single scissor for all targets
+    // or scissor disabled for the rest of the targets?
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
-
     state.scissor.enabled = (regs.scissor_test.enable != 0);
-    // TODO(Blinkhawk): Figure if the hardware supports scissor testing per viewport and how it's
-    // implemented.
-    if (regs.scissor_test.enable != 0) {
-        const u32 width = regs.scissor_test.max_x - regs.scissor_test.min_x;
-        const u32 height = regs.scissor_test.max_y - regs.scissor_test.min_y;
-        state.scissor.x = regs.scissor_test.min_x;
-        state.scissor.y = regs.scissor_test.min_y;
-        state.scissor.width = width;
-        state.scissor.height = height;
+    if (regs.scissor_test.enable == 0) {
+        return;
     }
+    const u32 width = regs.scissor_test.max_x - regs.scissor_test.min_x;
+    const u32 height = regs.scissor_test.max_y - regs.scissor_test.min_y;
+    state.scissor.x = regs.scissor_test.min_x;
+    state.scissor.y = regs.scissor_test.min_y;
+    state.scissor.width = width;
+    state.scissor.height = height;
 }
 
 void RasterizerOpenGL::SyncTransformFeedback() {
@@ -1046,9 +1127,8 @@ void RasterizerOpenGL::CheckAlphaTests() {
     const auto& regs = Core::System::GetInstance().GPU().Maxwell3D().regs;
 
     if (regs.alpha_test_enabled != 0 && regs.rt_control.count > 1) {
-        LOG_CRITICAL(
-            Render_OpenGL,
-            "Alpha Testing is enabled with Multiple Render Targets, this behavior is undefined.");
+        LOG_CRITICAL(Render_OpenGL, "Alpha Testing is enabled with Multiple Render Targets, "
+                                    "this behavior is undefined.");
         UNREACHABLE();
     }
 }
diff --git a/src/video_core/renderer_opengl/gl_rasterizer.h b/src/video_core/renderer_opengl/gl_rasterizer.h
index 0e90a31f5..8ef0f6c12 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer.h
+++ b/src/video_core/renderer_opengl/gl_rasterizer.h
@@ -8,12 +8,12 @@
 #include <cstddef>
 #include <map>
 #include <memory>
+#include <optional>
 #include <tuple>
 #include <utility>
 #include <vector>
 
 #include <boost/icl/interval_map.hpp>
-#include <boost/optional.hpp>
 #include <boost/range/iterator_range.hpp>
 #include <glad/glad.h>
 
@@ -88,11 +88,12 @@ private:
         /// SamplerInfo struct.
         void Create();
         /// Syncs the sampler object with the config, updating any necessary state.
-        void SyncWithConfig(const Tegra::Texture::TSCEntry& config);
+        void SyncWithConfig(const Tegra::Texture::FullTextureInfo& info);
 
     private:
         Tegra::Texture::TextureFilter mag_filter;
         Tegra::Texture::TextureFilter min_filter;
+        Tegra::Texture::TextureMipmapFilter mip_filter;
         Tegra::Texture::WrapMode wrap_u;
         Tegra::Texture::WrapMode wrap_v;
         Tegra::Texture::WrapMode wrap_p;
@@ -108,9 +109,9 @@ private:
      * @param preserve_contents If true, tries to preserve data from a previously used framebuffer.
      * @param single_color_target Specifies if a single color buffer target should be used.
      */
-    void ConfigureFramebuffers(bool use_color_fb = true, bool using_depth_fb = true,
-                               bool preserve_contents = true,
-                               boost::optional<std::size_t> single_color_target = {});
+    void ConfigureFramebuffers(OpenGLState& current_state, bool use_color_fb = true,
+                               bool using_depth_fb = true, bool preserve_contents = true,
+                               std::optional<std::size_t> single_color_target = {});
 
     /*
      * Configures the current constbuffers to use for the draw command.
@@ -132,8 +133,8 @@ private:
     u32 SetupTextures(Tegra::Engines::Maxwell3D::Regs::ShaderStage stage, Shader& shader,
                       GLenum primitive_mode, u32 current_unit);
 
-    /// Syncs the viewport to match the guest state
-    void SyncViewport();
+    /// Syncs the viewport and depth range to match the guest state
+    void SyncViewport(OpenGLState& current_state);
 
     /// Syncs the clip enabled status to match the guest state
     void SyncClipEnabled();
@@ -144,11 +145,8 @@ private:
     /// Syncs the cull mode to match the guest state
     void SyncCullMode();
 
-    /// Syncs the depth scale to match the guest state
-    void SyncDepthScale();
-
-    /// Syncs the depth offset to match the guest state
-    void SyncDepthOffset();
+    /// Syncs the primitve restart to match the guest state
+    void SyncPrimitiveRestart();
 
     /// Syncs the depth test state to match the guest state
     void SyncDepthTestState();
@@ -171,6 +169,9 @@ private:
     /// Syncs the point state to match the guest state
     void SyncPointState();
 
+    /// Syncs Color Mask
+    void SyncColorMask();
+
     /// Check asserts for alpha testing.
     void CheckAlphaTests();
 
@@ -206,7 +207,8 @@ private:
 
     std::size_t CalculateIndexBufferSize() const;
 
-    void SetupVertexArrays();
+    void SetupVertexFormat();
+    void SetupVertexBuffer();
 
     DrawParameters SetupDraw();
 
diff --git a/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp b/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp
index 9c8925383..9ca82c06c 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp
@@ -15,16 +15,24 @@
 #include "core/memory.h"
 #include "core/settings.h"
 #include "video_core/engines/maxwell_3d.h"
+#include "video_core/renderer_opengl/gl_rasterizer.h"
 #include "video_core/renderer_opengl/gl_rasterizer_cache.h"
+#include "video_core/renderer_opengl/gl_state.h"
+#include "video_core/renderer_opengl/utils.h"
+#include "video_core/surface.h"
 #include "video_core/textures/astc.h"
 #include "video_core/textures/decoders.h"
 #include "video_core/utils.h"
 
 namespace OpenGL {
 
-using SurfaceType = SurfaceParams::SurfaceType;
-using PixelFormat = SurfaceParams::PixelFormat;
-using ComponentType = SurfaceParams::ComponentType;
+using VideoCore::Surface::ComponentTypeFromDepthFormat;
+using VideoCore::Surface::ComponentTypeFromRenderTarget;
+using VideoCore::Surface::ComponentTypeFromTexture;
+using VideoCore::Surface::PixelFormatFromDepthFormat;
+using VideoCore::Surface::PixelFormatFromRenderTargetFormat;
+using VideoCore::Surface::PixelFormatFromTextureFormat;
+using VideoCore::Surface::SurfaceTargetFromTextureType;
 
 struct FormatTuple {
     GLint internal_format;
@@ -34,34 +42,6 @@ struct FormatTuple {
     bool compressed;
 };
 
-static bool IsPixelFormatASTC(PixelFormat format) {
-    switch (format) {
-    case PixelFormat::ASTC_2D_4X4:
-    case PixelFormat::ASTC_2D_5X4:
-    case PixelFormat::ASTC_2D_8X8:
-    case PixelFormat::ASTC_2D_8X5:
-        return true;
-    default:
-        return false;
-    }
-}
-
-static std::pair<u32, u32> GetASTCBlockSize(PixelFormat format) {
-    switch (format) {
-    case PixelFormat::ASTC_2D_4X4:
-        return {4, 4};
-    case PixelFormat::ASTC_2D_5X4:
-        return {5, 4};
-    case PixelFormat::ASTC_2D_8X8:
-        return {8, 8};
-    case PixelFormat::ASTC_2D_8X5:
-        return {8, 5};
-    default:
-        LOG_CRITICAL(HW_GPU, "Unhandled format: {}", static_cast<u32>(format));
-        UNREACHABLE();
-    }
-}
-
 void SurfaceParams::InitCacheParameters(Tegra::GPUVAddr gpu_addr_) {
     auto& memory_manager{Core::System::GetInstance().GPU().MemoryManager()};
     const auto cpu_addr{memory_manager.GpuToCpuAddress(gpu_addr_)};
@@ -78,6 +58,36 @@ void SurfaceParams::InitCacheParameters(Tegra::GPUVAddr gpu_addr_) {
     }
 }
 
+std::size_t SurfaceParams::InnerMipmapMemorySize(u32 mip_level, bool force_gl, bool layer_only,
+                                                 bool uncompressed) const {
+    const u32 tile_x{GetDefaultBlockWidth(pixel_format)};
+    const u32 tile_y{GetDefaultBlockHeight(pixel_format)};
+    const u32 bytes_per_pixel{GetBytesPerPixel(pixel_format)};
+    u32 m_depth = (layer_only ? 1U : depth);
+    u32 m_width = MipWidth(mip_level);
+    u32 m_height = MipHeight(mip_level);
+    m_width = uncompressed ? m_width : std::max(1U, (m_width + tile_x - 1) / tile_x);
+    m_height = uncompressed ? m_height : std::max(1U, (m_height + tile_y - 1) / tile_y);
+    m_depth = std::max(1U, m_depth >> mip_level);
+    u32 m_block_height = MipBlockHeight(mip_level);
+    u32 m_block_depth = MipBlockDepth(mip_level);
+    return Tegra::Texture::CalculateSize(force_gl ? false : is_tiled, bytes_per_pixel, m_width,
+                                         m_height, m_depth, m_block_height, m_block_depth);
+}
+
+std::size_t SurfaceParams::InnerMemorySize(bool force_gl, bool layer_only,
+                                           bool uncompressed) const {
+    std::size_t block_size_bytes = Tegra::Texture::GetGOBSize() * block_height * block_depth;
+    std::size_t size = 0;
+    for (u32 i = 0; i < max_mip_level; i++) {
+        size += InnerMipmapMemorySize(i, force_gl, layer_only, uncompressed);
+    }
+    if (!force_gl && is_tiled) {
+        size = Common::AlignUp(size, block_size_bytes);
+    }
+    return size;
+}
+
 /*static*/ SurfaceParams SurfaceParams::CreateForTexture(
     const Tegra::Texture::FullTextureInfo& config, const GLShader::SamplerEntry& entry) {
     SurfaceParams params{};
@@ -85,8 +95,9 @@ void SurfaceParams::InitCacheParameters(Tegra::GPUVAddr gpu_addr_) {
     params.block_width = params.is_tiled ? config.tic.BlockWidth() : 0,
     params.block_height = params.is_tiled ? config.tic.BlockHeight() : 0,
     params.block_depth = params.is_tiled ? config.tic.BlockDepth() : 0,
-    params.pixel_format =
-        PixelFormatFromTextureFormat(config.tic.format, config.tic.r_type.Value());
+    params.srgb_conversion = config.tic.IsSrgbConversionEnabled();
+    params.pixel_format = PixelFormatFromTextureFormat(config.tic.format, config.tic.r_type.Value(),
+                                                       params.srgb_conversion);
     params.component_type = ComponentTypeFromTexture(config.tic.r_type.Value());
     params.type = GetFormatType(params.pixel_format);
     params.width = Common::AlignUp(config.tic.Width(), GetCompressionFactor(params.pixel_format));
@@ -117,6 +128,13 @@ void SurfaceParams::InitCacheParameters(Tegra::GPUVAddr gpu_addr_) {
             params.target = SurfaceTarget::Texture2D;
         }
         break;
+    case SurfaceTarget::TextureCubeArray:
+        params.depth = config.tic.Depth() * 6;
+        if (!entry.IsArray()) {
+            ASSERT(params.depth == 6);
+            params.target = SurfaceTarget::TextureCubemap;
+        }
+        break;
     default:
         LOG_CRITICAL(HW_GPU, "Unknown depth for target={}", static_cast<u32>(params.target));
         UNREACHABLE();
@@ -124,6 +142,7 @@ void SurfaceParams::InitCacheParameters(Tegra::GPUVAddr gpu_addr_) {
         break;
     }
 
+    params.is_layered = SurfaceTargetIsLayered(params.target);
     params.max_mip_level = config.tic.max_mip_level + 1;
     params.rt = {};
 
@@ -142,6 +161,8 @@ void SurfaceParams::InitCacheParameters(Tegra::GPUVAddr gpu_addr_) {
     params.block_height = 1 << config.memory_layout.block_height;
     params.block_depth = 1 << config.memory_layout.block_depth;
     params.pixel_format = PixelFormatFromRenderTargetFormat(config.format);
+    params.srgb_conversion = config.format == Tegra::RenderTargetFormat::BGRA8_SRGB ||
+                             config.format == Tegra::RenderTargetFormat::RGBA8_SRGB;
     params.component_type = ComponentTypeFromRenderTarget(config.format);
     params.type = GetFormatType(params.pixel_format);
     params.width = config.width;
@@ -149,7 +170,8 @@ void SurfaceParams::InitCacheParameters(Tegra::GPUVAddr gpu_addr_) {
     params.unaligned_height = config.height;
     params.target = SurfaceTarget::Texture2D;
     params.depth = 1;
-    params.max_mip_level = 0;
+    params.max_mip_level = 1;
+    params.is_layered = false;
 
     // Render target specific parameters, not used for caching
     params.rt.index = static_cast<u32>(index);
@@ -176,12 +198,14 @@ void SurfaceParams::InitCacheParameters(Tegra::GPUVAddr gpu_addr_) {
     params.pixel_format = PixelFormatFromDepthFormat(format);
     params.component_type = ComponentTypeFromDepthFormat(format);
     params.type = GetFormatType(params.pixel_format);
+    params.srgb_conversion = false;
     params.width = zeta_width;
     params.height = zeta_height;
     params.unaligned_height = zeta_height;
     params.target = SurfaceTarget::Texture2D;
     params.depth = 1;
-    params.max_mip_level = 0;
+    params.max_mip_level = 1;
+    params.is_layered = false;
     params.rt = {};
 
     params.InitCacheParameters(zeta_address);
@@ -198,6 +222,8 @@ void SurfaceParams::InitCacheParameters(Tegra::GPUVAddr gpu_addr_) {
     params.block_height = params.is_tiled ? std::min(config.BlockHeight(), 32U) : 0,
     params.block_depth = params.is_tiled ? std::min(config.BlockDepth(), 32U) : 0,
     params.pixel_format = PixelFormatFromRenderTargetFormat(config.format);
+    params.srgb_conversion = config.format == Tegra::RenderTargetFormat::BGRA8_SRGB ||
+                             config.format == Tegra::RenderTargetFormat::RGBA8_SRGB;
     params.component_type = ComponentTypeFromRenderTarget(config.format);
     params.type = GetFormatType(params.pixel_format);
     params.width = config.width;
@@ -205,7 +231,7 @@ void SurfaceParams::InitCacheParameters(Tegra::GPUVAddr gpu_addr_) {
     params.unaligned_height = config.height;
     params.target = SurfaceTarget::Texture2D;
     params.depth = 1;
-    params.max_mip_level = 0;
+    params.max_mip_level = 1;
     params.rt = {};
 
     params.InitCacheParameters(config.Address());
@@ -213,7 +239,7 @@ void SurfaceParams::InitCacheParameters(Tegra::GPUVAddr gpu_addr_) {
     return params;
 }
 
-static constexpr std::array<FormatTuple, SurfaceParams::MaxPixelFormat> tex_format_tuples = {{
+static constexpr std::array<FormatTuple, VideoCore::Surface::MaxPixelFormat> tex_format_tuples = {{
     {GL_RGBA8, GL_RGBA, GL_UNSIGNED_INT_8_8_8_8_REV, ComponentType::UNorm, false}, // ABGR8U
     {GL_RGBA8, GL_RGBA, GL_BYTE, ComponentType::SNorm, false},                     // ABGR8S
     {GL_RGBA8UI, GL_RGBA_INTEGER, GL_UNSIGNED_BYTE, ComponentType::UInt, false},   // ABGR8UI
@@ -229,7 +255,7 @@ static constexpr std::array<FormatTuple, SurfaceParams::MaxPixelFormat> tex_form
     {GL_R11F_G11F_B10F, GL_RGB, GL_UNSIGNED_INT_10F_11F_11F_REV, ComponentType::Float,
      false},                                                                     // R11FG11FB10F
     {GL_RGBA32UI, GL_RGBA_INTEGER, GL_UNSIGNED_INT, ComponentType::UInt, false}, // RGBA32UI
-    {GL_COMPRESSED_RGB_S3TC_DXT1_EXT, GL_RGB, GL_UNSIGNED_INT_8_8_8_8, ComponentType::UNorm,
+    {GL_COMPRESSED_RGBA_S3TC_DXT1_EXT, GL_RGBA, GL_UNSIGNED_INT_8_8_8_8, ComponentType::UNorm,
      true}, // DXT1
     {GL_COMPRESSED_RGBA_S3TC_DXT3_EXT, GL_RGBA, GL_UNSIGNED_INT_8_8_8_8, ComponentType::UNorm,
      true}, // DXT23
@@ -263,14 +289,33 @@ static constexpr std::array<FormatTuple, SurfaceParams::MaxPixelFormat> tex_form
     {GL_RG16I, GL_RG_INTEGER, GL_SHORT, ComponentType::SInt, false},           // RG16I
     {GL_RG16_SNORM, GL_RG, GL_SHORT, ComponentType::SNorm, false},             // RG16S
     {GL_RGB32F, GL_RGB, GL_FLOAT, ComponentType::Float, false},                // RGB32F
-    {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_INT_8_8_8_8_REV, ComponentType::UNorm, false}, // SRGBA8
-    {GL_RG8, GL_RG, GL_UNSIGNED_BYTE, ComponentType::UNorm, false},                       // RG8U
-    {GL_RG8, GL_RG, GL_BYTE, ComponentType::SNorm, false},                                // RG8S
-    {GL_RG32UI, GL_RG_INTEGER, GL_UNSIGNED_INT, ComponentType::UInt, false},              // RG32UI
-    {GL_R32UI, GL_RED_INTEGER, GL_UNSIGNED_INT, ComponentType::UInt, false},              // R32UI
-    {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_8X8
-    {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_8X5
-    {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_5X4
+    {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_INT_8_8_8_8_REV, ComponentType::UNorm,
+     false},                                                                   // RGBA8_SRGB
+    {GL_RG8, GL_RG, GL_UNSIGNED_BYTE, ComponentType::UNorm, false},            // RG8U
+    {GL_RG8, GL_RG, GL_BYTE, ComponentType::SNorm, false},                     // RG8S
+    {GL_RG32UI, GL_RG_INTEGER, GL_UNSIGNED_INT, ComponentType::UInt, false},   // RG32UI
+    {GL_R32UI, GL_RED_INTEGER, GL_UNSIGNED_INT, ComponentType::UInt, false},   // R32UI
+    {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false},        // ASTC_2D_8X8
+    {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false},        // ASTC_2D_8X5
+    {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false},        // ASTC_2D_5X4
+    {GL_SRGB8_ALPHA8, GL_BGRA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // BGRA8
+    // Compressed sRGB formats
+    {GL_COMPRESSED_SRGB_ALPHA_S3TC_DXT1_EXT, GL_RGBA, GL_UNSIGNED_INT_8_8_8_8, ComponentType::UNorm,
+     true}, // DXT1_SRGB
+    {GL_COMPRESSED_SRGB_ALPHA_S3TC_DXT3_EXT, GL_RGBA, GL_UNSIGNED_INT_8_8_8_8, ComponentType::UNorm,
+     true}, // DXT23_SRGB
+    {GL_COMPRESSED_SRGB_ALPHA_S3TC_DXT5_EXT, GL_RGBA, GL_UNSIGNED_INT_8_8_8_8, ComponentType::UNorm,
+     true}, // DXT45_SRGB
+    {GL_COMPRESSED_SRGB_ALPHA_BPTC_UNORM_ARB, GL_RGBA, GL_UNSIGNED_INT_8_8_8_8,
+     ComponentType::UNorm, true},                                              // BC7U_SRGB
+    {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_4X4_SRGB
+    {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_8X8_SRGB
+    {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_8X5_SRGB
+    {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_5X4_SRGB
+    {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false},        // ASTC_2D_5X5
+    {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_5X5_SRGB
+    {GL_RGBA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false},        // ASTC_2D_10X8
+    {GL_SRGB8_ALPHA8, GL_RGBA, GL_UNSIGNED_BYTE, ComponentType::UNorm, false}, // ASTC_2D_10X8_SRGB
 
     // Depth formats
     {GL_DEPTH_COMPONENT32F, GL_DEPTH_COMPONENT, GL_FLOAT, ComponentType::Float, false}, // Z32F
@@ -286,20 +331,22 @@ static constexpr std::array<FormatTuple, SurfaceParams::MaxPixelFormat> tex_form
      ComponentType::Float, false}, // Z32FS8
 }};
 
-static GLenum SurfaceTargetToGL(SurfaceParams::SurfaceTarget target) {
+static GLenum SurfaceTargetToGL(SurfaceTarget target) {
     switch (target) {
-    case SurfaceParams::SurfaceTarget::Texture1D:
+    case SurfaceTarget::Texture1D:
         return GL_TEXTURE_1D;
-    case SurfaceParams::SurfaceTarget::Texture2D:
+    case SurfaceTarget::Texture2D:
         return GL_TEXTURE_2D;
-    case SurfaceParams::SurfaceTarget::Texture3D:
+    case SurfaceTarget::Texture3D:
         return GL_TEXTURE_3D;
-    case SurfaceParams::SurfaceTarget::Texture1DArray:
+    case SurfaceTarget::Texture1DArray:
         return GL_TEXTURE_1D_ARRAY;
-    case SurfaceParams::SurfaceTarget::Texture2DArray:
+    case SurfaceTarget::Texture2DArray:
         return GL_TEXTURE_2D_ARRAY;
-    case SurfaceParams::SurfaceTarget::TextureCubemap:
+    case SurfaceTarget::TextureCubemap:
         return GL_TEXTURE_CUBE_MAP;
+    case SurfaceTarget::TextureCubeArray:
+        return GL_TEXTURE_CUBE_MAP_ARRAY_ARB;
     }
     LOG_CRITICAL(Render_OpenGL, "Unimplemented texture target={}", static_cast<u32>(target));
     UNREACHABLE();
@@ -314,57 +361,41 @@ static const FormatTuple& GetFormatTuple(PixelFormat pixel_format, ComponentType
     return format;
 }
 
-MathUtil::Rectangle<u32> SurfaceParams::GetRect() const {
-    u32 actual_height{unaligned_height};
+MathUtil::Rectangle<u32> SurfaceParams::GetRect(u32 mip_level) const {
+    u32 actual_height{std::max(1U, unaligned_height >> mip_level)};
     if (IsPixelFormatASTC(pixel_format)) {
         // ASTC formats must stop at the ATSC block size boundary
         actual_height = Common::AlignDown(actual_height, GetASTCBlockSize(pixel_format).second);
     }
-    return {0, actual_height, width, 0};
-}
-
-/// Returns true if the specified PixelFormat is a BCn format, e.g. DXT or DXN
-static bool IsFormatBCn(PixelFormat format) {
-    switch (format) {
-    case PixelFormat::DXT1:
-    case PixelFormat::DXT23:
-    case PixelFormat::DXT45:
-    case PixelFormat::DXN1:
-    case PixelFormat::DXN2SNORM:
-    case PixelFormat::DXN2UNORM:
-    case PixelFormat::BC7U:
-    case PixelFormat::BC6H_UF16:
-    case PixelFormat::BC6H_SF16:
-        return true;
-    }
-    return false;
+    return {0, actual_height, MipWidth(mip_level), 0};
 }
 
 template <bool morton_to_gl, PixelFormat format>
 void MortonCopy(u32 stride, u32 block_height, u32 height, u32 block_depth, u32 depth, u8* gl_buffer,
                 std::size_t gl_buffer_size, VAddr addr) {
-    constexpr u32 bytes_per_pixel = SurfaceParams::GetBytesPerPixel(format);
+    constexpr u32 bytes_per_pixel = GetBytesPerPixel(format);
 
     // With the BCn formats (DXT and DXN), each 4x4 tile is swizzled instead of just individual
     // pixel values.
-    const u32 tile_size{IsFormatBCn(format) ? 4U : 1U};
+    const u32 tile_size_x{GetDefaultBlockWidth(format)};
+    const u32 tile_size_y{GetDefaultBlockHeight(format)};
 
     if (morton_to_gl) {
-        const std::vector<u8> data = Tegra::Texture::UnswizzleTexture(
-            addr, tile_size, bytes_per_pixel, stride, height, depth, block_height, block_depth);
-        const std::size_t size_to_copy{std::min(gl_buffer_size, data.size())};
-        memcpy(gl_buffer, data.data(), size_to_copy);
+        Tegra::Texture::UnswizzleTexture(gl_buffer, addr, tile_size_x, tile_size_y, bytes_per_pixel,
+                                         stride, height, depth, block_height, block_depth);
     } else {
-        Tegra::Texture::CopySwizzledData(stride / tile_size, height / tile_size, depth,
+        Tegra::Texture::CopySwizzledData((stride + tile_size_x - 1) / tile_size_x,
+                                         (height + tile_size_y - 1) / tile_size_y, depth,
                                          bytes_per_pixel, bytes_per_pixel, Memory::GetPointer(addr),
                                          gl_buffer, false, block_height, block_depth);
     }
 }
 
-static constexpr std::array<void (*)(u32, u32, u32, u32, u32, u8*, std::size_t, VAddr),
-                            SurfaceParams::MaxPixelFormat>
-    morton_to_gl_fns = {
-        // clang-format off
+using GLConversionArray = std::array<void (*)(u32, u32, u32, u32, u32, u8*, std::size_t, VAddr),
+                                     VideoCore::Surface::MaxPixelFormat>;
+
+static constexpr GLConversionArray morton_to_gl_fns = {
+    // clang-format off
         MortonCopy<true, PixelFormat::ABGR8U>,
         MortonCopy<true, PixelFormat::ABGR8S>,
         MortonCopy<true, PixelFormat::ABGR8UI>,
@@ -405,7 +436,7 @@ static constexpr std::array<void (*)(u32, u32, u32, u32, u32, u8*, std::size_t,
         MortonCopy<true, PixelFormat::RG16I>,
         MortonCopy<true, PixelFormat::RG16S>,
         MortonCopy<true, PixelFormat::RGB32F>,
-        MortonCopy<true, PixelFormat::SRGBA8>,
+        MortonCopy<true, PixelFormat::RGBA8_SRGB>,
         MortonCopy<true, PixelFormat::RG8U>,
         MortonCopy<true, PixelFormat::RG8S>,
         MortonCopy<true, PixelFormat::RG32UI>,
@@ -413,18 +444,29 @@ static constexpr std::array<void (*)(u32, u32, u32, u32, u32, u8*, std::size_t,
         MortonCopy<true, PixelFormat::ASTC_2D_8X8>,
         MortonCopy<true, PixelFormat::ASTC_2D_8X5>,
         MortonCopy<true, PixelFormat::ASTC_2D_5X4>,
+        MortonCopy<true, PixelFormat::BGRA8_SRGB>,
+        MortonCopy<true, PixelFormat::DXT1_SRGB>,
+        MortonCopy<true, PixelFormat::DXT23_SRGB>,
+        MortonCopy<true, PixelFormat::DXT45_SRGB>,
+        MortonCopy<true, PixelFormat::BC7U_SRGB>,
+        MortonCopy<true, PixelFormat::ASTC_2D_4X4_SRGB>,
+        MortonCopy<true, PixelFormat::ASTC_2D_8X8_SRGB>,
+        MortonCopy<true, PixelFormat::ASTC_2D_8X5_SRGB>,
+        MortonCopy<true, PixelFormat::ASTC_2D_5X4_SRGB>,
+        MortonCopy<true, PixelFormat::ASTC_2D_5X5>,
+        MortonCopy<true, PixelFormat::ASTC_2D_5X5_SRGB>,
+        MortonCopy<true, PixelFormat::ASTC_2D_10X8>,
+        MortonCopy<true, PixelFormat::ASTC_2D_10X8_SRGB>,
         MortonCopy<true, PixelFormat::Z32F>,
         MortonCopy<true, PixelFormat::Z16>,
         MortonCopy<true, PixelFormat::Z24S8>,
         MortonCopy<true, PixelFormat::S8Z24>,
         MortonCopy<true, PixelFormat::Z32FS8>,
-        // clang-format on
+    // clang-format on
 };
 
-static constexpr std::array<void (*)(u32, u32, u32, u32, u32, u8*, std::size_t, VAddr),
-                            SurfaceParams::MaxPixelFormat>
-    gl_to_morton_fns = {
-        // clang-format off
+static constexpr GLConversionArray gl_to_morton_fns = {
+    // clang-format off
         MortonCopy<false, PixelFormat::ABGR8U>,
         MortonCopy<false, PixelFormat::ABGR8S>,
         MortonCopy<false, PixelFormat::ABGR8UI>,
@@ -466,7 +508,7 @@ static constexpr std::array<void (*)(u32, u32, u32, u32, u32, u8*, std::size_t,
         MortonCopy<false, PixelFormat::RG16I>,
         MortonCopy<false, PixelFormat::RG16S>,
         MortonCopy<false, PixelFormat::RGB32F>,
-        MortonCopy<false, PixelFormat::SRGBA8>,
+        MortonCopy<false, PixelFormat::RGBA8_SRGB>,
         MortonCopy<false, PixelFormat::RG8U>,
         MortonCopy<false, PixelFormat::RG8S>,
         MortonCopy<false, PixelFormat::RG32UI>,
@@ -474,17 +516,61 @@ static constexpr std::array<void (*)(u32, u32, u32, u32, u32, u8*, std::size_t,
         nullptr,
         nullptr,
         nullptr,
+        MortonCopy<false, PixelFormat::BGRA8_SRGB>,
+        MortonCopy<false, PixelFormat::DXT1_SRGB>,
+        MortonCopy<false, PixelFormat::DXT23_SRGB>,
+        MortonCopy<false, PixelFormat::DXT45_SRGB>,
+        MortonCopy<false, PixelFormat::BC7U_SRGB>,
+        nullptr,
+        nullptr,
+        nullptr,
+        nullptr,
+        nullptr,
+        nullptr,
+        nullptr,
+        nullptr,
         MortonCopy<false, PixelFormat::Z32F>,
         MortonCopy<false, PixelFormat::Z16>,
         MortonCopy<false, PixelFormat::Z24S8>,
         MortonCopy<false, PixelFormat::S8Z24>,
         MortonCopy<false, PixelFormat::Z32FS8>,
-        // clang-format on
+    // clang-format on
 };
 
+void SwizzleFunc(const GLConversionArray& functions, const SurfaceParams& params,
+                 std::vector<u8>& gl_buffer, u32 mip_level) {
+    u32 depth = params.MipDepth(mip_level);
+    if (params.target == SurfaceTarget::Texture2D) {
+        // TODO(Blinkhawk): Eliminate this condition once all texture types are implemented.
+        depth = 1U;
+    }
+    if (params.is_layered) {
+        u64 offset = params.GetMipmapLevelOffset(mip_level);
+        u64 offset_gl = 0;
+        const u64 layer_size = params.LayerMemorySize();
+        const u64 gl_size = params.LayerSizeGL(mip_level);
+        for (u32 i = 0; i < params.depth; i++) {
+            functions[static_cast<std::size_t>(params.pixel_format)](
+                params.MipWidth(mip_level), params.MipBlockHeight(mip_level),
+                params.MipHeight(mip_level), params.MipBlockDepth(mip_level), 1,
+                gl_buffer.data() + offset_gl, gl_size, params.addr + offset);
+            offset += layer_size;
+            offset_gl += gl_size;
+        }
+    } else {
+        const u64 offset = params.GetMipmapLevelOffset(mip_level);
+        functions[static_cast<std::size_t>(params.pixel_format)](
+            params.MipWidth(mip_level), params.MipBlockHeight(mip_level),
+            params.MipHeight(mip_level), params.MipBlockDepth(mip_level), depth, gl_buffer.data(),
+            gl_buffer.size(), params.addr + offset);
+    }
+}
+
+MICROPROFILE_DEFINE(OpenGL_BlitSurface, "OpenGL", "BlitSurface", MP_RGB(128, 192, 64));
 static bool BlitSurface(const Surface& src_surface, const Surface& dst_surface,
                         GLuint read_fb_handle, GLuint draw_fb_handle, GLenum src_attachment = 0,
                         GLenum dst_attachment = 0, std::size_t cubemap_face = 0) {
+    MICROPROFILE_SCOPE(OpenGL_BlitSurface);
 
     const auto& src_params{src_surface->GetSurfaceParams()};
     const auto& dst_params{dst_surface->GetSurfaceParams()};
@@ -495,19 +581,21 @@ static bool BlitSurface(const Surface& src_surface, const Surface& dst_surface,
     OpenGLState state;
     state.draw.read_framebuffer = read_fb_handle;
     state.draw.draw_framebuffer = draw_fb_handle;
-    state.Apply();
+    // Set sRGB enabled if the destination surfaces need it
+    state.framebuffer_srgb.enabled = dst_params.srgb_conversion;
+    state.ApplyFramebufferState();
 
     u32 buffers{};
 
     if (src_params.type == SurfaceType::ColorTexture) {
         switch (src_params.target) {
-        case SurfaceParams::SurfaceTarget::Texture2D:
+        case SurfaceTarget::Texture2D:
             glFramebufferTexture2D(GL_READ_FRAMEBUFFER, GL_COLOR_ATTACHMENT0 + src_attachment,
                                    GL_TEXTURE_2D, src_surface->Texture().handle, 0);
             glFramebufferTexture2D(GL_READ_FRAMEBUFFER, GL_DEPTH_STENCIL_ATTACHMENT, GL_TEXTURE_2D,
                                    0, 0);
             break;
-        case SurfaceParams::SurfaceTarget::TextureCubemap:
+        case SurfaceTarget::TextureCubemap:
             glFramebufferTexture2D(
                 GL_READ_FRAMEBUFFER, GL_COLOR_ATTACHMENT0 + src_attachment,
                 static_cast<GLenum>(GL_TEXTURE_CUBE_MAP_POSITIVE_X + cubemap_face),
@@ -516,12 +604,12 @@ static bool BlitSurface(const Surface& src_surface, const Surface& dst_surface,
                 GL_READ_FRAMEBUFFER, GL_DEPTH_STENCIL_ATTACHMENT,
                 static_cast<GLenum>(GL_TEXTURE_CUBE_MAP_POSITIVE_X + cubemap_face), 0, 0);
             break;
-        case SurfaceParams::SurfaceTarget::Texture2DArray:
+        case SurfaceTarget::Texture2DArray:
             glFramebufferTextureLayer(GL_READ_FRAMEBUFFER, GL_COLOR_ATTACHMENT0 + src_attachment,
                                       src_surface->Texture().handle, 0, 0);
             glFramebufferTextureLayer(GL_READ_FRAMEBUFFER, GL_DEPTH_STENCIL_ATTACHMENT, 0, 0, 0);
             break;
-        case SurfaceParams::SurfaceTarget::Texture3D:
+        case SurfaceTarget::Texture3D:
             glFramebufferTexture3D(GL_READ_FRAMEBUFFER, GL_COLOR_ATTACHMENT0 + src_attachment,
                                    SurfaceTargetToGL(src_params.target),
                                    src_surface->Texture().handle, 0, 0);
@@ -537,13 +625,13 @@ static bool BlitSurface(const Surface& src_surface, const Surface& dst_surface,
         }
 
         switch (dst_params.target) {
-        case SurfaceParams::SurfaceTarget::Texture2D:
+        case SurfaceTarget::Texture2D:
             glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER, GL_COLOR_ATTACHMENT0 + dst_attachment,
                                    GL_TEXTURE_2D, dst_surface->Texture().handle, 0);
             glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER, GL_DEPTH_STENCIL_ATTACHMENT, GL_TEXTURE_2D,
                                    0, 0);
             break;
-        case SurfaceParams::SurfaceTarget::TextureCubemap:
+        case SurfaceTarget::TextureCubemap:
             glFramebufferTexture2D(
                 GL_DRAW_FRAMEBUFFER, GL_COLOR_ATTACHMENT0 + dst_attachment,
                 static_cast<GLenum>(GL_TEXTURE_CUBE_MAP_POSITIVE_X + cubemap_face),
@@ -552,13 +640,13 @@ static bool BlitSurface(const Surface& src_surface, const Surface& dst_surface,
                 GL_DRAW_FRAMEBUFFER, GL_DEPTH_STENCIL_ATTACHMENT,
                 static_cast<GLenum>(GL_TEXTURE_CUBE_MAP_POSITIVE_X + cubemap_face), 0, 0);
             break;
-        case SurfaceParams::SurfaceTarget::Texture2DArray:
+        case SurfaceTarget::Texture2DArray:
             glFramebufferTextureLayer(GL_DRAW_FRAMEBUFFER, GL_COLOR_ATTACHMENT0 + dst_attachment,
                                       dst_surface->Texture().handle, 0, 0);
             glFramebufferTextureLayer(GL_DRAW_FRAMEBUFFER, GL_DEPTH_STENCIL_ATTACHMENT, 0, 0, 0);
             break;
 
-        case SurfaceParams::SurfaceTarget::Texture3D:
+        case SurfaceTarget::Texture3D:
             glFramebufferTexture3D(GL_DRAW_FRAMEBUFFER, GL_COLOR_ATTACHMENT0 + dst_attachment,
                                    SurfaceTargetToGL(dst_params.target),
                                    dst_surface->Texture().handle, 0, 0);
@@ -622,18 +710,20 @@ static void FastCopySurface(const Surface& src_surface, const Surface& dst_surfa
                        0, 0, width, height, 1);
 }
 
+MICROPROFILE_DEFINE(OpenGL_CopySurface, "OpenGL", "CopySurface", MP_RGB(128, 192, 64));
 static void CopySurface(const Surface& src_surface, const Surface& dst_surface,
-                        GLuint copy_pbo_handle, GLenum src_attachment = 0,
-                        GLenum dst_attachment = 0, std::size_t cubemap_face = 0) {
+                        const GLuint copy_pbo_handle, const GLenum src_attachment = 0,
+                        const GLenum dst_attachment = 0, const std::size_t cubemap_face = 0) {
+    MICROPROFILE_SCOPE(OpenGL_CopySurface);
     ASSERT_MSG(dst_attachment == 0, "Unimplemented");
 
     const auto& src_params{src_surface->GetSurfaceParams()};
     const auto& dst_params{dst_surface->GetSurfaceParams()};
 
-    auto source_format = GetFormatTuple(src_params.pixel_format, src_params.component_type);
-    auto dest_format = GetFormatTuple(dst_params.pixel_format, dst_params.component_type);
+    const auto source_format = GetFormatTuple(src_params.pixel_format, src_params.component_type);
+    const auto dest_format = GetFormatTuple(dst_params.pixel_format, dst_params.component_type);
 
-    std::size_t buffer_size = std::max(src_params.size_in_bytes, dst_params.size_in_bytes);
+    const std::size_t buffer_size = std::max(src_params.size_in_bytes, dst_params.size_in_bytes);
 
     glBindBuffer(GL_PIXEL_PACK_BUFFER, copy_pbo_handle);
     glBufferData(GL_PIXEL_PACK_BUFFER, buffer_size, nullptr, GL_STREAM_DRAW_ARB);
@@ -657,13 +747,10 @@ static void CopySurface(const Surface& src_surface, const Surface& dst_surface,
             LOG_DEBUG(HW_GPU, "Trying to upload extra texture data from the CPU during "
                               "reinterpretation but the texture is tiled.");
         }
-        std::size_t remaining_size = dst_params.size_in_bytes - src_params.size_in_bytes;
-        std::vector<u8> data(remaining_size);
-        std::memcpy(data.data(), Memory::GetPointer(dst_params.addr + src_params.size_in_bytes),
-                    data.size());
+        const std::size_t remaining_size = dst_params.size_in_bytes - src_params.size_in_bytes;
 
         glBufferSubData(GL_PIXEL_PACK_BUFFER, src_params.size_in_bytes, remaining_size,
-                        data.data());
+                        Memory::GetPointer(dst_params.addr + src_params.size_in_bytes));
     }
 
     glBindBuffer(GL_PIXEL_PACK_BUFFER, 0);
@@ -679,21 +766,22 @@ static void CopySurface(const Surface& src_surface, const Surface& dst_surface,
         UNREACHABLE();
     } else {
         switch (dst_params.target) {
-        case SurfaceParams::SurfaceTarget::Texture1D:
+        case SurfaceTarget::Texture1D:
             glTextureSubImage1D(dst_surface->Texture().handle, 0, 0, width, dest_format.format,
                                 dest_format.type, nullptr);
             break;
-        case SurfaceParams::SurfaceTarget::Texture2D:
+        case SurfaceTarget::Texture2D:
             glTextureSubImage2D(dst_surface->Texture().handle, 0, 0, 0, width, height,
                                 dest_format.format, dest_format.type, nullptr);
             break;
-        case SurfaceParams::SurfaceTarget::Texture3D:
-        case SurfaceParams::SurfaceTarget::Texture2DArray:
+        case SurfaceTarget::Texture3D:
+        case SurfaceTarget::Texture2DArray:
+        case SurfaceTarget::TextureCubeArray:
             glTextureSubImage3D(dst_surface->Texture().handle, 0, 0, 0, 0, width, height,
                                 static_cast<GLsizei>(dst_params.depth), dest_format.format,
                                 dest_format.type, nullptr);
             break;
-        case SurfaceParams::SurfaceTarget::TextureCubemap:
+        case SurfaceTarget::TextureCubemap:
             glTextureSubImage3D(dst_surface->Texture().handle, 0, 0, 0,
                                 static_cast<GLint>(cubemap_face), width, height, 1,
                                 dest_format.format, dest_format.type, nullptr);
@@ -730,35 +818,43 @@ CachedSurface::CachedSurface(const SurfaceParams& params)
     if (!format_tuple.compressed) {
         // Only pre-create the texture for non-compressed textures.
         switch (params.target) {
-        case SurfaceParams::SurfaceTarget::Texture1D:
-            glTexStorage1D(SurfaceTargetToGL(params.target), 1, format_tuple.internal_format,
-                           rect.GetWidth());
+        case SurfaceTarget::Texture1D:
+            glTexStorage1D(SurfaceTargetToGL(params.target), params.max_mip_level,
+                           format_tuple.internal_format, rect.GetWidth());
             break;
-        case SurfaceParams::SurfaceTarget::Texture2D:
-        case SurfaceParams::SurfaceTarget::TextureCubemap:
-            glTexStorage2D(SurfaceTargetToGL(params.target), 1, format_tuple.internal_format,
-                           rect.GetWidth(), rect.GetHeight());
+        case SurfaceTarget::Texture2D:
+        case SurfaceTarget::TextureCubemap:
+            glTexStorage2D(SurfaceTargetToGL(params.target), params.max_mip_level,
+                           format_tuple.internal_format, rect.GetWidth(), rect.GetHeight());
             break;
-        case SurfaceParams::SurfaceTarget::Texture3D:
-        case SurfaceParams::SurfaceTarget::Texture2DArray:
-            glTexStorage3D(SurfaceTargetToGL(params.target), 1, format_tuple.internal_format,
-                           rect.GetWidth(), rect.GetHeight(), params.depth);
+        case SurfaceTarget::Texture3D:
+        case SurfaceTarget::Texture2DArray:
+        case SurfaceTarget::TextureCubeArray:
+            glTexStorage3D(SurfaceTargetToGL(params.target), params.max_mip_level,
+                           format_tuple.internal_format, rect.GetWidth(), rect.GetHeight(),
+                           params.depth);
             break;
         default:
             LOG_CRITICAL(Render_OpenGL, "Unimplemented surface target={}",
                          static_cast<u32>(params.target));
             UNREACHABLE();
-            glTexStorage2D(GL_TEXTURE_2D, 1, format_tuple.internal_format, rect.GetWidth(),
-                           rect.GetHeight());
+            glTexStorage2D(GL_TEXTURE_2D, params.max_mip_level, format_tuple.internal_format,
+                           rect.GetWidth(), rect.GetHeight());
         }
     }
 
     glTexParameteri(SurfaceTargetToGL(params.target), GL_TEXTURE_MIN_FILTER, GL_LINEAR);
+    glTexParameteri(SurfaceTargetToGL(params.target), GL_TEXTURE_MAG_FILTER, GL_LINEAR);
     glTexParameteri(SurfaceTargetToGL(params.target), GL_TEXTURE_WRAP_S, GL_CLAMP_TO_EDGE);
     glTexParameteri(SurfaceTargetToGL(params.target), GL_TEXTURE_WRAP_T, GL_CLAMP_TO_EDGE);
+    glTexParameteri(SurfaceTargetToGL(params.target), GL_TEXTURE_MAX_LEVEL,
+                    params.max_mip_level - 1);
+    if (params.max_mip_level == 1) {
+        glTexParameterf(SurfaceTargetToGL(params.target), GL_TEXTURE_LOD_BIAS, 1000.0);
+    }
 
-    VideoCore::LabelGLObject(GL_TEXTURE, texture.handle, params.addr,
-                             SurfaceParams::SurfaceTargetName(params.target));
+    LabelGLObject(GL_TEXTURE, texture.handle, params.addr,
+                  SurfaceParams::SurfaceTargetName(params.target));
 
     // Clamp size to mapped GPU memory region
     // TODO(bunnei): Super Mario Odyssey maps a 0x40000 byte region and then uses it for a 0x80000
@@ -788,7 +884,7 @@ static void ConvertS8Z24ToZ24S8(std::vector<u8>& data, u32 width, u32 height, bo
 
     S8Z24 s8z24_pixel{};
     Z24S8 z24s8_pixel{};
-    constexpr auto bpp{SurfaceParams::GetBytesPerPixel(PixelFormat::S8Z24)};
+    constexpr auto bpp{GetBytesPerPixel(PixelFormat::S8Z24)};
     for (std::size_t y = 0; y < height; ++y) {
         for (std::size_t x = 0; x < width; ++x) {
             const std::size_t offset{bpp * (y * width + x)};
@@ -808,7 +904,7 @@ static void ConvertS8Z24ToZ24S8(std::vector<u8>& data, u32 width, u32 height, bo
 }
 
 static void ConvertG8R8ToR8G8(std::vector<u8>& data, u32 width, u32 height) {
-    constexpr auto bpp{SurfaceParams::GetBytesPerPixel(PixelFormat::G8R8U)};
+    constexpr auto bpp{GetBytesPerPixel(PixelFormat::G8R8U)};
     for (std::size_t y = 0; y < height; ++y) {
         for (std::size_t x = 0; x < width; ++x) {
             const std::size_t offset{bpp * (y * width + x)};
@@ -825,17 +921,26 @@ static void ConvertG8R8ToR8G8(std::vector<u8>& data, u32 width, u32 height) {
  * typical desktop GPUs.
  */
 static void ConvertFormatAsNeeded_LoadGLBuffer(std::vector<u8>& data, PixelFormat pixel_format,
-                                               u32 width, u32 height) {
+                                               u32 width, u32 height, u32 depth) {
     switch (pixel_format) {
     case PixelFormat::ASTC_2D_4X4:
     case PixelFormat::ASTC_2D_8X8:
     case PixelFormat::ASTC_2D_8X5:
-    case PixelFormat::ASTC_2D_5X4: {
+    case PixelFormat::ASTC_2D_5X4:
+    case PixelFormat::ASTC_2D_5X5:
+    case PixelFormat::ASTC_2D_4X4_SRGB:
+    case PixelFormat::ASTC_2D_8X8_SRGB:
+    case PixelFormat::ASTC_2D_8X5_SRGB:
+    case PixelFormat::ASTC_2D_5X4_SRGB:
+    case PixelFormat::ASTC_2D_5X5_SRGB:
+    case PixelFormat::ASTC_2D_10X8:
+    case PixelFormat::ASTC_2D_10X8_SRGB: {
         // Convert ASTC pixel formats to RGBA8, as most desktop GPUs do not support ASTC.
         u32 block_width{};
         u32 block_height{};
         std::tie(block_width, block_height) = GetASTCBlockSize(pixel_format);
-        data = Tegra::Texture::ASTC::Decompress(data, width, height, block_width, block_height);
+        data =
+            Tegra::Texture::ASTC::Decompress(data, width, height, depth, block_width, block_height);
         break;
     }
     case PixelFormat::S8Z24:
@@ -862,7 +967,13 @@ static void ConvertFormatAsNeeded_FlushGLBuffer(std::vector<u8>& data, PixelForm
     case PixelFormat::G8R8U:
     case PixelFormat::G8R8S:
     case PixelFormat::ASTC_2D_4X4:
-    case PixelFormat::ASTC_2D_8X8: {
+    case PixelFormat::ASTC_2D_8X8:
+    case PixelFormat::ASTC_2D_4X4_SRGB:
+    case PixelFormat::ASTC_2D_8X8_SRGB:
+    case PixelFormat::ASTC_2D_5X5:
+    case PixelFormat::ASTC_2D_5X5_SRGB:
+    case PixelFormat::ASTC_2D_10X8:
+    case PixelFormat::ASTC_2D_10X8_SRGB: {
         LOG_CRITICAL(HW_GPU, "Conversion of format {} after texture flushing is not implemented",
                      static_cast<u32>(pixel_format));
         UNREACHABLE();
@@ -875,34 +986,25 @@ static void ConvertFormatAsNeeded_FlushGLBuffer(std::vector<u8>& data, PixelForm
     }
 }
 
-MICROPROFILE_DEFINE(OpenGL_SurfaceLoad, "OpenGL", "Surface Load", MP_RGB(128, 64, 192));
+MICROPROFILE_DEFINE(OpenGL_SurfaceLoad, "OpenGL", "Surface Load", MP_RGB(128, 192, 64));
 void CachedSurface::LoadGLBuffer() {
     MICROPROFILE_SCOPE(OpenGL_SurfaceLoad);
-
-    gl_buffer.resize(params.size_in_bytes_gl);
+    gl_buffer.resize(params.max_mip_level);
+    for (u32 i = 0; i < params.max_mip_level; i++)
+        gl_buffer[i].resize(params.GetMipmapSizeGL(i));
     if (params.is_tiled) {
-        u32 depth = params.depth;
-        u32 block_depth = params.block_depth;
-
         ASSERT_MSG(params.block_width == 1, "Block width is defined as {} on texture type {}",
                    params.block_width, static_cast<u32>(params.target));
-
-        if (params.target == SurfaceParams::SurfaceTarget::Texture2D) {
-            // TODO(Blinkhawk): Eliminate this condition once all texture types are implemented.
-            depth = 1U;
-            block_depth = 1U;
-        }
-
-        morton_to_gl_fns[static_cast<std::size_t>(params.pixel_format)](
-            params.width, params.block_height, params.height, block_depth, depth, gl_buffer.data(),
-            gl_buffer.size(), params.addr);
+        for (u32 i = 0; i < params.max_mip_level; i++)
+            SwizzleFunc(morton_to_gl_fns, params, gl_buffer[i], i);
     } else {
         const auto texture_src_data{Memory::GetPointer(params.addr)};
         const auto texture_src_data_end{texture_src_data + params.size_in_bytes_gl};
-        gl_buffer.assign(texture_src_data, texture_src_data_end);
+        gl_buffer[0].assign(texture_src_data, texture_src_data_end);
     }
-
-    ConvertFormatAsNeeded_LoadGLBuffer(gl_buffer, params.pixel_format, params.width, params.height);
+    for (u32 i = 0; i < params.max_mip_level; i++)
+        ConvertFormatAsNeeded_LoadGLBuffer(gl_buffer[i], params.pixel_format, params.MipWidth(i),
+                                           params.MipHeight(i), params.MipDepth(i));
 }
 
 MICROPROFILE_DEFINE(OpenGL_SurfaceFlush, "OpenGL", "Surface Flush", MP_RGB(128, 192, 64));
@@ -912,57 +1014,44 @@ void CachedSurface::FlushGLBuffer() {
     ASSERT_MSG(!IsPixelFormatASTC(params.pixel_format), "Unimplemented");
 
     // OpenGL temporary buffer needs to be big enough to store raw texture size
-    gl_buffer.resize(GetSizeInBytes());
+    gl_buffer.resize(1);
+    gl_buffer[0].resize(GetSizeInBytes());
 
     const FormatTuple& tuple = GetFormatTuple(params.pixel_format, params.component_type);
     // Ensure no bad interactions with GL_UNPACK_ALIGNMENT
-    ASSERT(params.width * SurfaceParams::GetBytesPerPixel(params.pixel_format) % 4 == 0);
+    ASSERT(params.width * GetBytesPerPixel(params.pixel_format) % 4 == 0);
     glPixelStorei(GL_PACK_ROW_LENGTH, static_cast<GLint>(params.width));
     ASSERT(!tuple.compressed);
     glBindBuffer(GL_PIXEL_PACK_BUFFER, 0);
-    glGetTextureImage(texture.handle, 0, tuple.format, tuple.type, gl_buffer.size(),
-                      gl_buffer.data());
+    glGetTextureImage(texture.handle, 0, tuple.format, tuple.type,
+                      static_cast<GLsizei>(gl_buffer[0].size()), gl_buffer[0].data());
     glPixelStorei(GL_PACK_ROW_LENGTH, 0);
-    ConvertFormatAsNeeded_FlushGLBuffer(gl_buffer, params.pixel_format, params.width,
+    ConvertFormatAsNeeded_FlushGLBuffer(gl_buffer[0], params.pixel_format, params.width,
                                         params.height);
     ASSERT(params.type != SurfaceType::Fill);
     const u8* const texture_src_data = Memory::GetPointer(params.addr);
     ASSERT(texture_src_data);
     if (params.is_tiled) {
-        u32 depth = params.depth;
-        u32 block_depth = params.block_depth;
-
         ASSERT_MSG(params.block_width == 1, "Block width is defined as {} on texture type {}",
                    params.block_width, static_cast<u32>(params.target));
 
-        if (params.target == SurfaceParams::SurfaceTarget::Texture2D) {
-            // TODO(Blinkhawk): Eliminate this condition once all texture types are implemented.
-            depth = 1U;
-        }
-        gl_to_morton_fns[static_cast<size_t>(params.pixel_format)](
-            params.width, params.block_height, params.height, block_depth, depth, gl_buffer.data(),
-            gl_buffer.size(), GetAddr());
+        SwizzleFunc(gl_to_morton_fns, params, gl_buffer[0], 0);
     } else {
-        std::memcpy(Memory::GetPointer(GetAddr()), gl_buffer.data(), GetSizeInBytes());
+        std::memcpy(Memory::GetPointer(GetAddr()), gl_buffer[0].data(), GetSizeInBytes());
     }
 }
 
-MICROPROFILE_DEFINE(OpenGL_TextureUL, "OpenGL", "Texture Upload", MP_RGB(128, 64, 192));
-void CachedSurface::UploadGLTexture(GLuint read_fb_handle, GLuint draw_fb_handle) {
-    if (params.type == SurfaceType::Fill)
-        return;
-
-    MICROPROFILE_SCOPE(OpenGL_TextureUL);
-
-    const auto& rect{params.GetRect()};
+void CachedSurface::UploadGLMipmapTexture(u32 mip_map, GLuint read_fb_handle,
+                                          GLuint draw_fb_handle) {
+    const auto& rect{params.GetRect(mip_map)};
 
     // Load data from memory to the surface
     const GLint x0 = static_cast<GLint>(rect.left);
     const GLint y0 = static_cast<GLint>(rect.bottom);
     std::size_t buffer_offset =
-        static_cast<std::size_t>(static_cast<std::size_t>(y0) * params.width +
+        static_cast<std::size_t>(static_cast<std::size_t>(y0) * params.MipWidth(mip_map) +
                                  static_cast<std::size_t>(x0)) *
-        SurfaceParams::GetBytesPerPixel(params.pixel_format);
+        GetBytesPerPixel(params.pixel_format);
 
     const FormatTuple& tuple = GetFormatTuple(params.pixel_format, params.component_type);
     const GLuint target_tex = texture.handle;
@@ -978,89 +1067,120 @@ void CachedSurface::UploadGLTexture(GLuint read_fb_handle, GLuint draw_fb_handle
     cur_state.Apply();
 
     // Ensure no bad interactions with GL_UNPACK_ALIGNMENT
-    ASSERT(params.width * SurfaceParams::GetBytesPerPixel(params.pixel_format) % 4 == 0);
-    glPixelStorei(GL_UNPACK_ROW_LENGTH, static_cast<GLint>(params.width));
+    ASSERT(params.MipWidth(mip_map) * GetBytesPerPixel(params.pixel_format) % 4 == 0);
+    glPixelStorei(GL_UNPACK_ROW_LENGTH, static_cast<GLint>(params.MipWidth(mip_map)));
 
+    GLsizei image_size = static_cast<GLsizei>(params.GetMipmapSizeGL(mip_map, false));
     glActiveTexture(GL_TEXTURE0);
     if (tuple.compressed) {
         switch (params.target) {
-        case SurfaceParams::SurfaceTarget::Texture2D:
-            glCompressedTexImage2D(
-                SurfaceTargetToGL(params.target), 0, tuple.internal_format,
-                static_cast<GLsizei>(params.width), static_cast<GLsizei>(params.height), 0,
-                static_cast<GLsizei>(params.size_in_bytes_gl), &gl_buffer[buffer_offset]);
+        case SurfaceTarget::Texture2D:
+            glCompressedTexImage2D(SurfaceTargetToGL(params.target), mip_map, tuple.internal_format,
+                                   static_cast<GLsizei>(params.MipWidth(mip_map)),
+                                   static_cast<GLsizei>(params.MipHeight(mip_map)), 0, image_size,
+                                   &gl_buffer[mip_map][buffer_offset]);
+            break;
+        case SurfaceTarget::Texture3D:
+            glCompressedTexImage3D(SurfaceTargetToGL(params.target), mip_map, tuple.internal_format,
+                                   static_cast<GLsizei>(params.MipWidth(mip_map)),
+                                   static_cast<GLsizei>(params.MipHeight(mip_map)),
+                                   static_cast<GLsizei>(params.MipDepth(mip_map)), 0, image_size,
+                                   &gl_buffer[mip_map][buffer_offset]);
             break;
-        case SurfaceParams::SurfaceTarget::Texture3D:
-        case SurfaceParams::SurfaceTarget::Texture2DArray:
-            glCompressedTexImage3D(
-                SurfaceTargetToGL(params.target), 0, tuple.internal_format,
-                static_cast<GLsizei>(params.width), static_cast<GLsizei>(params.height),
-                static_cast<GLsizei>(params.depth), 0,
-                static_cast<GLsizei>(params.size_in_bytes_gl), &gl_buffer[buffer_offset]);
+        case SurfaceTarget::Texture2DArray:
+        case SurfaceTarget::TextureCubeArray:
+            glCompressedTexImage3D(SurfaceTargetToGL(params.target), mip_map, tuple.internal_format,
+                                   static_cast<GLsizei>(params.MipWidth(mip_map)),
+                                   static_cast<GLsizei>(params.MipHeight(mip_map)),
+                                   static_cast<GLsizei>(params.depth), 0, image_size,
+                                   &gl_buffer[mip_map][buffer_offset]);
             break;
-        case SurfaceParams::SurfaceTarget::TextureCubemap:
+        case SurfaceTarget::TextureCubemap: {
+            GLsizei layer_size = static_cast<GLsizei>(params.LayerSizeGL(mip_map));
             for (std::size_t face = 0; face < params.depth; ++face) {
                 glCompressedTexImage2D(static_cast<GLenum>(GL_TEXTURE_CUBE_MAP_POSITIVE_X + face),
-                                       0, tuple.internal_format, static_cast<GLsizei>(params.width),
-                                       static_cast<GLsizei>(params.height), 0,
-                                       static_cast<GLsizei>(params.SizeInBytesCubeFaceGL()),
-                                       &gl_buffer[buffer_offset]);
-                buffer_offset += params.SizeInBytesCubeFace();
+                                       mip_map, tuple.internal_format,
+                                       static_cast<GLsizei>(params.MipWidth(mip_map)),
+                                       static_cast<GLsizei>(params.MipHeight(mip_map)), 0,
+                                       layer_size, &gl_buffer[mip_map][buffer_offset]);
+                buffer_offset += layer_size;
             }
             break;
+        }
         default:
             LOG_CRITICAL(Render_OpenGL, "Unimplemented surface target={}",
                          static_cast<u32>(params.target));
             UNREACHABLE();
-            glCompressedTexImage2D(
-                GL_TEXTURE_2D, 0, tuple.internal_format, static_cast<GLsizei>(params.width),
-                static_cast<GLsizei>(params.height), 0,
-                static_cast<GLsizei>(params.size_in_bytes_gl), &gl_buffer[buffer_offset]);
+            glCompressedTexImage2D(GL_TEXTURE_2D, mip_map, tuple.internal_format,
+                                   static_cast<GLsizei>(params.MipWidth(mip_map)),
+                                   static_cast<GLsizei>(params.MipHeight(mip_map)), 0,
+                                   static_cast<GLsizei>(params.size_in_bytes_gl),
+                                   &gl_buffer[mip_map][buffer_offset]);
         }
     } else {
 
         switch (params.target) {
-        case SurfaceParams::SurfaceTarget::Texture1D:
-            glTexSubImage1D(SurfaceTargetToGL(params.target), 0, x0,
+        case SurfaceTarget::Texture1D:
+            glTexSubImage1D(SurfaceTargetToGL(params.target), mip_map, x0,
                             static_cast<GLsizei>(rect.GetWidth()), tuple.format, tuple.type,
-                            &gl_buffer[buffer_offset]);
+                            &gl_buffer[mip_map][buffer_offset]);
             break;
-        case SurfaceParams::SurfaceTarget::Texture2D:
-            glTexSubImage2D(SurfaceTargetToGL(params.target), 0, x0, y0,
+        case SurfaceTarget::Texture2D:
+            glTexSubImage2D(SurfaceTargetToGL(params.target), mip_map, x0, y0,
                             static_cast<GLsizei>(rect.GetWidth()),
                             static_cast<GLsizei>(rect.GetHeight()), tuple.format, tuple.type,
-                            &gl_buffer[buffer_offset]);
+                            &gl_buffer[mip_map][buffer_offset]);
             break;
-        case SurfaceParams::SurfaceTarget::Texture3D:
-        case SurfaceParams::SurfaceTarget::Texture2DArray:
-            glTexSubImage3D(SurfaceTargetToGL(params.target), 0, x0, y0, 0,
+        case SurfaceTarget::Texture3D:
+            glTexSubImage3D(SurfaceTargetToGL(params.target), mip_map, x0, y0, 0,
+                            static_cast<GLsizei>(rect.GetWidth()),
+                            static_cast<GLsizei>(rect.GetHeight()), params.MipDepth(mip_map),
+                            tuple.format, tuple.type, &gl_buffer[mip_map][buffer_offset]);
+            break;
+        case SurfaceTarget::Texture2DArray:
+        case SurfaceTarget::TextureCubeArray:
+            glTexSubImage3D(SurfaceTargetToGL(params.target), mip_map, x0, y0, 0,
                             static_cast<GLsizei>(rect.GetWidth()),
                             static_cast<GLsizei>(rect.GetHeight()), params.depth, tuple.format,
-                            tuple.type, &gl_buffer[buffer_offset]);
+                            tuple.type, &gl_buffer[mip_map][buffer_offset]);
             break;
-        case SurfaceParams::SurfaceTarget::TextureCubemap:
+        case SurfaceTarget::TextureCubemap: {
+            std::size_t start = buffer_offset;
             for (std::size_t face = 0; face < params.depth; ++face) {
-                glTexSubImage2D(static_cast<GLenum>(GL_TEXTURE_CUBE_MAP_POSITIVE_X + face), 0, x0,
-                                y0, static_cast<GLsizei>(rect.GetWidth()),
+                glTexSubImage2D(static_cast<GLenum>(GL_TEXTURE_CUBE_MAP_POSITIVE_X + face), mip_map,
+                                x0, y0, static_cast<GLsizei>(rect.GetWidth()),
                                 static_cast<GLsizei>(rect.GetHeight()), tuple.format, tuple.type,
-                                &gl_buffer[buffer_offset]);
-                buffer_offset += params.SizeInBytesCubeFace();
+                                &gl_buffer[mip_map][buffer_offset]);
+                buffer_offset += params.LayerSizeGL(mip_map);
             }
             break;
+        }
         default:
             LOG_CRITICAL(Render_OpenGL, "Unimplemented surface target={}",
                          static_cast<u32>(params.target));
             UNREACHABLE();
-            glTexSubImage2D(GL_TEXTURE_2D, 0, x0, y0, static_cast<GLsizei>(rect.GetWidth()),
+            glTexSubImage2D(GL_TEXTURE_2D, mip_map, x0, y0, static_cast<GLsizei>(rect.GetWidth()),
                             static_cast<GLsizei>(rect.GetHeight()), tuple.format, tuple.type,
-                            &gl_buffer[buffer_offset]);
+                            &gl_buffer[mip_map][buffer_offset]);
         }
     }
 
     glPixelStorei(GL_UNPACK_ROW_LENGTH, 0);
 }
 
-RasterizerCacheOpenGL::RasterizerCacheOpenGL() {
+MICROPROFILE_DEFINE(OpenGL_TextureUL, "OpenGL", "Texture Upload", MP_RGB(128, 192, 64));
+void CachedSurface::UploadGLTexture(GLuint read_fb_handle, GLuint draw_fb_handle) {
+    if (params.type == SurfaceType::Fill)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_TextureUL);
+
+    for (u32 i = 0; i < params.max_mip_level; i++)
+        UploadGLMipmapTexture(i, read_fb_handle, draw_fb_handle);
+}
+
+RasterizerCacheOpenGL::RasterizerCacheOpenGL(RasterizerOpenGL& rasterizer)
+    : RasterizerCache{rasterizer} {
     read_framebuffer.Create();
     draw_framebuffer.Create();
     copy_pbo.Create();
@@ -1179,7 +1299,7 @@ void RasterizerCacheOpenGL::AccurateCopySurface(const Surface& src_surface,
                                                 const Surface& dst_surface) {
     const auto& src_params{src_surface->GetSurfaceParams()};
     const auto& dst_params{dst_surface->GetSurfaceParams()};
-    FlushRegion(src_params.addr, dst_params.size_in_bytes);
+    FlushRegion(src_params.addr, dst_params.MemorySize());
     LoadSurface(dst_surface);
 }
 
@@ -1200,8 +1320,7 @@ Surface RasterizerCacheOpenGL::RecreateSurface(const Surface& old_surface,
     // For compatible surfaces, we can just do fast glCopyImageSubData based copy
     if (old_params.target == new_params.target && old_params.type == new_params.type &&
         old_params.depth == new_params.depth && old_params.depth == 1 &&
-        SurfaceParams::GetFormatBpp(old_params.pixel_format) ==
-            SurfaceParams::GetFormatBpp(new_params.pixel_format)) {
+        GetFormatBpp(old_params.pixel_format) == GetFormatBpp(new_params.pixel_format)) {
         FastCopySurface(old_surface, new_surface);
         return new_surface;
     }
@@ -1214,51 +1333,19 @@ Surface RasterizerCacheOpenGL::RecreateSurface(const Surface& old_surface,
     const bool is_blit{old_params.pixel_format == new_params.pixel_format};
 
     switch (new_params.target) {
-    case SurfaceParams::SurfaceTarget::Texture2D:
+    case SurfaceTarget::Texture2D:
         if (is_blit) {
             BlitSurface(old_surface, new_surface, read_framebuffer.handle, draw_framebuffer.handle);
         } else {
             CopySurface(old_surface, new_surface, copy_pbo.handle);
         }
         break;
-    case SurfaceParams::SurfaceTarget::Texture3D:
+    case SurfaceTarget::TextureCubemap:
+    case SurfaceTarget::Texture3D:
+    case SurfaceTarget::Texture2DArray:
+    case SurfaceTarget::TextureCubeArray:
         AccurateCopySurface(old_surface, new_surface);
         break;
-    case SurfaceParams::SurfaceTarget::TextureCubemap: {
-        if (old_params.rt.array_mode != 1) {
-            // TODO(bunnei): This is used by Breath of the Wild, I'm not sure how to implement this
-            // yet (array rendering used as a cubemap texture).
-            LOG_CRITICAL(HW_GPU, "Unhandled rendertarget array_mode {}", old_params.rt.array_mode);
-            UNREACHABLE();
-            return new_surface;
-        }
-
-        // This seems to be used for render-to-cubemap texture
-        ASSERT_MSG(old_params.target == SurfaceParams::SurfaceTarget::Texture2D, "Unexpected");
-        ASSERT_MSG(old_params.pixel_format == new_params.pixel_format, "Unexpected");
-        ASSERT_MSG(old_params.rt.base_layer == 0, "Unimplemented");
-
-        // TODO(bunnei): Verify the below - this stride seems to be in 32-bit words, not pixels.
-        // Tested with Splatoon 2, Super Mario Odyssey, and Breath of the Wild.
-        const std::size_t byte_stride{old_params.rt.layer_stride * sizeof(u32)};
-
-        for (std::size_t index = 0; index < new_params.depth; ++index) {
-            Surface face_surface{TryGetReservedSurface(old_params)};
-            ASSERT_MSG(face_surface, "Unexpected");
-
-            if (is_blit) {
-                BlitSurface(face_surface, new_surface, read_framebuffer.handle,
-                            draw_framebuffer.handle, face_surface->GetSurfaceParams().rt.index,
-                            new_params.rt.index, index);
-            } else {
-                CopySurface(face_surface, new_surface, copy_pbo.handle,
-                            face_surface->GetSurfaceParams().rt.index, new_params.rt.index, index);
-            }
-
-            old_params.addr += byte_stride;
-        }
-        break;
-    }
     default:
         LOG_CRITICAL(Render_OpenGL, "Unimplemented surface target={}",
                      static_cast<u32>(new_params.target));
diff --git a/src/video_core/renderer_opengl/gl_rasterizer_cache.h b/src/video_core/renderer_opengl/gl_rasterizer_cache.h
index 0dd0d90a3..494f6b903 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer_cache.h
+++ b/src/video_core/renderer_opengl/gl_rasterizer_cache.h
@@ -7,6 +7,7 @@
 #include <array>
 #include <map>
 #include <memory>
+#include <string>
 #include <vector>
 
 #include "common/alignment.h"
@@ -18,6 +19,7 @@
 #include "video_core/rasterizer_cache.h"
 #include "video_core/renderer_opengl/gl_resource_manager.h"
 #include "video_core/renderer_opengl/gl_shader_gen.h"
+#include "video_core/surface.h"
 #include "video_core/textures/decoders.h"
 #include "video_core/textures/texture.h"
 
@@ -27,126 +29,12 @@ class CachedSurface;
 using Surface = std::shared_ptr<CachedSurface>;
 using SurfaceSurfaceRect_Tuple = std::tuple<Surface, Surface, MathUtil::Rectangle<u32>>;
 
-struct SurfaceParams {
-    enum class PixelFormat {
-        ABGR8U = 0,
-        ABGR8S = 1,
-        ABGR8UI = 2,
-        B5G6R5U = 3,
-        A2B10G10R10U = 4,
-        A1B5G5R5U = 5,
-        R8U = 6,
-        R8UI = 7,
-        RGBA16F = 8,
-        RGBA16U = 9,
-        RGBA16UI = 10,
-        R11FG11FB10F = 11,
-        RGBA32UI = 12,
-        DXT1 = 13,
-        DXT23 = 14,
-        DXT45 = 15,
-        DXN1 = 16, // This is also known as BC4
-        DXN2UNORM = 17,
-        DXN2SNORM = 18,
-        BC7U = 19,
-        BC6H_UF16 = 20,
-        BC6H_SF16 = 21,
-        ASTC_2D_4X4 = 22,
-        G8R8U = 23,
-        G8R8S = 24,
-        BGRA8 = 25,
-        RGBA32F = 26,
-        RG32F = 27,
-        R32F = 28,
-        R16F = 29,
-        R16U = 30,
-        R16S = 31,
-        R16UI = 32,
-        R16I = 33,
-        RG16 = 34,
-        RG16F = 35,
-        RG16UI = 36,
-        RG16I = 37,
-        RG16S = 38,
-        RGB32F = 39,
-        SRGBA8 = 40,
-        RG8U = 41,
-        RG8S = 42,
-        RG32UI = 43,
-        R32UI = 44,
-        ASTC_2D_8X8 = 45,
-        ASTC_2D_8X5 = 46,
-        ASTC_2D_5X4 = 47,
-
-        MaxColorFormat,
-
-        // Depth formats
-        Z32F = 48,
-        Z16 = 49,
-
-        MaxDepthFormat,
-
-        // DepthStencil formats
-        Z24S8 = 50,
-        S8Z24 = 51,
-        Z32FS8 = 52,
-
-        MaxDepthStencilFormat,
-
-        Max = MaxDepthStencilFormat,
-        Invalid = 255,
-    };
-
-    static constexpr std::size_t MaxPixelFormat = static_cast<std::size_t>(PixelFormat::Max);
-
-    enum class ComponentType {
-        Invalid = 0,
-        SNorm = 1,
-        UNorm = 2,
-        SInt = 3,
-        UInt = 4,
-        Float = 5,
-    };
-
-    enum class SurfaceType {
-        ColorTexture = 0,
-        Depth = 1,
-        DepthStencil = 2,
-        Fill = 3,
-        Invalid = 4,
-    };
-
-    enum class SurfaceTarget {
-        Texture1D,
-        Texture2D,
-        Texture3D,
-        Texture1DArray,
-        Texture2DArray,
-        TextureCubemap,
-    };
-
-    static SurfaceTarget SurfaceTargetFromTextureType(Tegra::Texture::TextureType texture_type) {
-        switch (texture_type) {
-        case Tegra::Texture::TextureType::Texture1D:
-            return SurfaceTarget::Texture1D;
-        case Tegra::Texture::TextureType::Texture2D:
-        case Tegra::Texture::TextureType::Texture2DNoMipmap:
-            return SurfaceTarget::Texture2D;
-        case Tegra::Texture::TextureType::Texture3D:
-            return SurfaceTarget::Texture3D;
-        case Tegra::Texture::TextureType::TextureCubemap:
-            return SurfaceTarget::TextureCubemap;
-        case Tegra::Texture::TextureType::Texture1DArray:
-            return SurfaceTarget::Texture1DArray;
-        case Tegra::Texture::TextureType::Texture2DArray:
-            return SurfaceTarget::Texture2DArray;
-        default:
-            LOG_CRITICAL(HW_GPU, "Unimplemented texture_type={}", static_cast<u32>(texture_type));
-            UNREACHABLE();
-            return SurfaceTarget::Texture2D;
-        }
-    }
+using SurfaceTarget = VideoCore::Surface::SurfaceTarget;
+using SurfaceType = VideoCore::Surface::SurfaceType;
+using PixelFormat = VideoCore::Surface::PixelFormat;
+using ComponentType = VideoCore::Surface::ComponentType;
 
+struct SurfaceParams {
     static std::string SurfaceTargetName(SurfaceTarget target) {
         switch (target) {
         case SurfaceTarget::Texture1D:
@@ -161,6 +49,8 @@ struct SurfaceParams {
             return "Texture2DArray";
         case SurfaceTarget::TextureCubemap:
             return "TextureCubemap";
+        case SurfaceTarget::TextureCubeArray:
+            return "TextureCubeArray";
         default:
             LOG_CRITICAL(HW_GPU, "Unimplemented surface_target={}", static_cast<u32>(target));
             UNREACHABLE();
@@ -168,552 +58,12 @@ struct SurfaceParams {
         }
     }
 
-    /**
-     * Gets the compression factor for the specified PixelFormat. This applies to just the
-     * "compressed width" and "compressed height", not the overall compression factor of a
-     * compressed image. This is used for maintaining proper surface sizes for compressed
-     * texture formats.
-     */
-    static constexpr u32 GetCompressionFactor(PixelFormat format) {
-        if (format == PixelFormat::Invalid)
-            return 0;
-
-        constexpr std::array<u32, MaxPixelFormat> compression_factor_table = {{
-            1, // ABGR8U
-            1, // ABGR8S
-            1, // ABGR8UI
-            1, // B5G6R5U
-            1, // A2B10G10R10U
-            1, // A1B5G5R5U
-            1, // R8U
-            1, // R8UI
-            1, // RGBA16F
-            1, // RGBA16U
-            1, // RGBA16UI
-            1, // R11FG11FB10F
-            1, // RGBA32UI
-            4, // DXT1
-            4, // DXT23
-            4, // DXT45
-            4, // DXN1
-            4, // DXN2UNORM
-            4, // DXN2SNORM
-            4, // BC7U
-            4, // BC6H_UF16
-            4, // BC6H_SF16
-            4, // ASTC_2D_4X4
-            1, // G8R8U
-            1, // G8R8S
-            1, // BGRA8
-            1, // RGBA32F
-            1, // RG32F
-            1, // R32F
-            1, // R16F
-            1, // R16U
-            1, // R16S
-            1, // R16UI
-            1, // R16I
-            1, // RG16
-            1, // RG16F
-            1, // RG16UI
-            1, // RG16I
-            1, // RG16S
-            1, // RGB32F
-            1, // SRGBA8
-            1, // RG8U
-            1, // RG8S
-            1, // RG32UI
-            1, // R32UI
-            4, // ASTC_2D_8X8
-            4, // ASTC_2D_8X5
-            4, // ASTC_2D_5X4
-            1, // Z32F
-            1, // Z16
-            1, // Z24S8
-            1, // S8Z24
-            1, // Z32FS8
-        }};
-
-        ASSERT(static_cast<std::size_t>(format) < compression_factor_table.size());
-        return compression_factor_table[static_cast<std::size_t>(format)];
-    }
-
-    static constexpr u32 GetFormatBpp(PixelFormat format) {
-        if (format == PixelFormat::Invalid)
-            return 0;
-
-        constexpr std::array<u32, MaxPixelFormat> bpp_table = {{
-            32,  // ABGR8U
-            32,  // ABGR8S
-            32,  // ABGR8UI
-            16,  // B5G6R5U
-            32,  // A2B10G10R10U
-            16,  // A1B5G5R5U
-            8,   // R8U
-            8,   // R8UI
-            64,  // RGBA16F
-            64,  // RGBA16U
-            64,  // RGBA16UI
-            32,  // R11FG11FB10F
-            128, // RGBA32UI
-            64,  // DXT1
-            128, // DXT23
-            128, // DXT45
-            64,  // DXN1
-            128, // DXN2UNORM
-            128, // DXN2SNORM
-            128, // BC7U
-            128, // BC6H_UF16
-            128, // BC6H_SF16
-            32,  // ASTC_2D_4X4
-            16,  // G8R8U
-            16,  // G8R8S
-            32,  // BGRA8
-            128, // RGBA32F
-            64,  // RG32F
-            32,  // R32F
-            16,  // R16F
-            16,  // R16U
-            16,  // R16S
-            16,  // R16UI
-            16,  // R16I
-            32,  // RG16
-            32,  // RG16F
-            32,  // RG16UI
-            32,  // RG16I
-            32,  // RG16S
-            96,  // RGB32F
-            32,  // SRGBA8
-            16,  // RG8U
-            16,  // RG8S
-            64,  // RG32UI
-            32,  // R32UI
-            16,  // ASTC_2D_8X8
-            32,  // ASTC_2D_8X5
-            32,  // ASTC_2D_5X4
-            32,  // Z32F
-            16,  // Z16
-            32,  // Z24S8
-            32,  // S8Z24
-            64,  // Z32FS8
-        }};
-
-        ASSERT(static_cast<std::size_t>(format) < bpp_table.size());
-        return bpp_table[static_cast<std::size_t>(format)];
-    }
-
     u32 GetFormatBpp() const {
-        return GetFormatBpp(pixel_format);
-    }
-
-    static PixelFormat PixelFormatFromDepthFormat(Tegra::DepthFormat format) {
-        switch (format) {
-        case Tegra::DepthFormat::S8_Z24_UNORM:
-            return PixelFormat::S8Z24;
-        case Tegra::DepthFormat::Z24_S8_UNORM:
-            return PixelFormat::Z24S8;
-        case Tegra::DepthFormat::Z32_FLOAT:
-            return PixelFormat::Z32F;
-        case Tegra::DepthFormat::Z16_UNORM:
-            return PixelFormat::Z16;
-        case Tegra::DepthFormat::Z32_S8_X24_FLOAT:
-            return PixelFormat::Z32FS8;
-        default:
-            LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));
-            UNREACHABLE();
-        }
-    }
-
-    static PixelFormat PixelFormatFromRenderTargetFormat(Tegra::RenderTargetFormat format) {
-        switch (format) {
-        // TODO (Hexagon12): Converting SRGBA to RGBA is a hack and doesn't completely correct the
-        // gamma.
-        case Tegra::RenderTargetFormat::RGBA8_SRGB:
-        case Tegra::RenderTargetFormat::RGBA8_UNORM:
-            return PixelFormat::ABGR8U;
-        case Tegra::RenderTargetFormat::RGBA8_SNORM:
-            return PixelFormat::ABGR8S;
-        case Tegra::RenderTargetFormat::RGBA8_UINT:
-            return PixelFormat::ABGR8UI;
-        case Tegra::RenderTargetFormat::BGRA8_SRGB:
-        case Tegra::RenderTargetFormat::BGRA8_UNORM:
-            return PixelFormat::BGRA8;
-        case Tegra::RenderTargetFormat::RGB10_A2_UNORM:
-            return PixelFormat::A2B10G10R10U;
-        case Tegra::RenderTargetFormat::RGBA16_FLOAT:
-            return PixelFormat::RGBA16F;
-        case Tegra::RenderTargetFormat::RGBA16_UNORM:
-            return PixelFormat::RGBA16U;
-        case Tegra::RenderTargetFormat::RGBA16_UINT:
-            return PixelFormat::RGBA16UI;
-        case Tegra::RenderTargetFormat::RGBA32_FLOAT:
-            return PixelFormat::RGBA32F;
-        case Tegra::RenderTargetFormat::RG32_FLOAT:
-            return PixelFormat::RG32F;
-        case Tegra::RenderTargetFormat::R11G11B10_FLOAT:
-            return PixelFormat::R11FG11FB10F;
-        case Tegra::RenderTargetFormat::B5G6R5_UNORM:
-            return PixelFormat::B5G6R5U;
-        case Tegra::RenderTargetFormat::BGR5A1_UNORM:
-            return PixelFormat::A1B5G5R5U;
-        case Tegra::RenderTargetFormat::RGBA32_UINT:
-            return PixelFormat::RGBA32UI;
-        case Tegra::RenderTargetFormat::R8_UNORM:
-            return PixelFormat::R8U;
-        case Tegra::RenderTargetFormat::R8_UINT:
-            return PixelFormat::R8UI;
-        case Tegra::RenderTargetFormat::RG16_FLOAT:
-            return PixelFormat::RG16F;
-        case Tegra::RenderTargetFormat::RG16_UINT:
-            return PixelFormat::RG16UI;
-        case Tegra::RenderTargetFormat::RG16_SINT:
-            return PixelFormat::RG16I;
-        case Tegra::RenderTargetFormat::RG16_UNORM:
-            return PixelFormat::RG16;
-        case Tegra::RenderTargetFormat::RG16_SNORM:
-            return PixelFormat::RG16S;
-        case Tegra::RenderTargetFormat::RG8_UNORM:
-            return PixelFormat::RG8U;
-        case Tegra::RenderTargetFormat::RG8_SNORM:
-            return PixelFormat::RG8S;
-        case Tegra::RenderTargetFormat::R16_FLOAT:
-            return PixelFormat::R16F;
-        case Tegra::RenderTargetFormat::R16_UNORM:
-            return PixelFormat::R16U;
-        case Tegra::RenderTargetFormat::R16_SNORM:
-            return PixelFormat::R16S;
-        case Tegra::RenderTargetFormat::R16_UINT:
-            return PixelFormat::R16UI;
-        case Tegra::RenderTargetFormat::R16_SINT:
-            return PixelFormat::R16I;
-        case Tegra::RenderTargetFormat::R32_FLOAT:
-            return PixelFormat::R32F;
-        case Tegra::RenderTargetFormat::R32_UINT:
-            return PixelFormat::R32UI;
-        case Tegra::RenderTargetFormat::RG32_UINT:
-            return PixelFormat::RG32UI;
-        default:
-            LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));
-            UNREACHABLE();
-        }
-    }
-
-    static PixelFormat PixelFormatFromTextureFormat(Tegra::Texture::TextureFormat format,
-                                                    Tegra::Texture::ComponentType component_type) {
-        // TODO(Subv): Properly implement this
-        switch (format) {
-        case Tegra::Texture::TextureFormat::A8R8G8B8:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::UNORM:
-                return PixelFormat::ABGR8U;
-            case Tegra::Texture::ComponentType::SNORM:
-                return PixelFormat::ABGR8S;
-            case Tegra::Texture::ComponentType::UINT:
-                return PixelFormat::ABGR8UI;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::B5G6R5:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::UNORM:
-                return PixelFormat::B5G6R5U;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::A2B10G10R10:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::UNORM:
-                return PixelFormat::A2B10G10R10U;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::A1B5G5R5:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::UNORM:
-                return PixelFormat::A1B5G5R5U;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::R8:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::UNORM:
-                return PixelFormat::R8U;
-            case Tegra::Texture::ComponentType::UINT:
-                return PixelFormat::R8UI;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::G8R8:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::UNORM:
-                return PixelFormat::G8R8U;
-            case Tegra::Texture::ComponentType::SNORM:
-                return PixelFormat::G8R8S;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::R16_G16_B16_A16:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::UNORM:
-                return PixelFormat::RGBA16U;
-            case Tegra::Texture::ComponentType::FLOAT:
-                return PixelFormat::RGBA16F;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::BF10GF11RF11:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::FLOAT:
-                return PixelFormat::R11FG11FB10F;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::R32_G32_B32_A32:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::FLOAT:
-                return PixelFormat::RGBA32F;
-            case Tegra::Texture::ComponentType::UINT:
-                return PixelFormat::RGBA32UI;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::R32_G32:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::FLOAT:
-                return PixelFormat::RG32F;
-            case Tegra::Texture::ComponentType::UINT:
-                return PixelFormat::RG32UI;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::R32_G32_B32:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::FLOAT:
-                return PixelFormat::RGB32F;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::R16:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::FLOAT:
-                return PixelFormat::R16F;
-            case Tegra::Texture::ComponentType::UNORM:
-                return PixelFormat::R16U;
-            case Tegra::Texture::ComponentType::SNORM:
-                return PixelFormat::R16S;
-            case Tegra::Texture::ComponentType::UINT:
-                return PixelFormat::R16UI;
-            case Tegra::Texture::ComponentType::SINT:
-                return PixelFormat::R16I;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::R32:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::FLOAT:
-                return PixelFormat::R32F;
-            case Tegra::Texture::ComponentType::UINT:
-                return PixelFormat::R32UI;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::ZF32:
-            return PixelFormat::Z32F;
-        case Tegra::Texture::TextureFormat::Z16:
-            return PixelFormat::Z16;
-        case Tegra::Texture::TextureFormat::Z24S8:
-            return PixelFormat::Z24S8;
-        case Tegra::Texture::TextureFormat::DXT1:
-            return PixelFormat::DXT1;
-        case Tegra::Texture::TextureFormat::DXT23:
-            return PixelFormat::DXT23;
-        case Tegra::Texture::TextureFormat::DXT45:
-            return PixelFormat::DXT45;
-        case Tegra::Texture::TextureFormat::DXN1:
-            return PixelFormat::DXN1;
-        case Tegra::Texture::TextureFormat::DXN2:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::UNORM:
-                return PixelFormat::DXN2UNORM;
-            case Tegra::Texture::ComponentType::SNORM:
-                return PixelFormat::DXN2SNORM;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        case Tegra::Texture::TextureFormat::BC7U:
-            return PixelFormat::BC7U;
-        case Tegra::Texture::TextureFormat::BC6H_UF16:
-            return PixelFormat::BC6H_UF16;
-        case Tegra::Texture::TextureFormat::BC6H_SF16:
-            return PixelFormat::BC6H_SF16;
-        case Tegra::Texture::TextureFormat::ASTC_2D_4X4:
-            return PixelFormat::ASTC_2D_4X4;
-        case Tegra::Texture::TextureFormat::ASTC_2D_5X4:
-            return PixelFormat::ASTC_2D_5X4;
-        case Tegra::Texture::TextureFormat::ASTC_2D_8X8:
-            return PixelFormat::ASTC_2D_8X8;
-        case Tegra::Texture::TextureFormat::ASTC_2D_8X5:
-            return PixelFormat::ASTC_2D_8X5;
-        case Tegra::Texture::TextureFormat::R16_G16:
-            switch (component_type) {
-            case Tegra::Texture::ComponentType::FLOAT:
-                return PixelFormat::RG16F;
-            case Tegra::Texture::ComponentType::UNORM:
-                return PixelFormat::RG16;
-            case Tegra::Texture::ComponentType::SNORM:
-                return PixelFormat::RG16S;
-            case Tegra::Texture::ComponentType::UINT:
-                return PixelFormat::RG16UI;
-            case Tegra::Texture::ComponentType::SINT:
-                return PixelFormat::RG16I;
-            }
-            LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}",
-                         static_cast<u32>(component_type));
-            UNREACHABLE();
-        default:
-            LOG_CRITICAL(HW_GPU, "Unimplemented format={}, component_type={}",
-                         static_cast<u32>(format), static_cast<u32>(component_type));
-            UNREACHABLE();
-        }
-    }
-
-    static ComponentType ComponentTypeFromTexture(Tegra::Texture::ComponentType type) {
-        // TODO(Subv): Implement more component types
-        switch (type) {
-        case Tegra::Texture::ComponentType::UNORM:
-            return ComponentType::UNorm;
-        case Tegra::Texture::ComponentType::FLOAT:
-            return ComponentType::Float;
-        case Tegra::Texture::ComponentType::SNORM:
-            return ComponentType::SNorm;
-        case Tegra::Texture::ComponentType::UINT:
-            return ComponentType::UInt;
-        case Tegra::Texture::ComponentType::SINT:
-            return ComponentType::SInt;
-        default:
-            LOG_CRITICAL(HW_GPU, "Unimplemented component type={}", static_cast<u32>(type));
-            UNREACHABLE();
-        }
-    }
-
-    static ComponentType ComponentTypeFromRenderTarget(Tegra::RenderTargetFormat format) {
-        // TODO(Subv): Implement more render targets
-        switch (format) {
-        case Tegra::RenderTargetFormat::RGBA8_UNORM:
-        case Tegra::RenderTargetFormat::RGBA8_SRGB:
-        case Tegra::RenderTargetFormat::BGRA8_UNORM:
-        case Tegra::RenderTargetFormat::BGRA8_SRGB:
-        case Tegra::RenderTargetFormat::RGB10_A2_UNORM:
-        case Tegra::RenderTargetFormat::R8_UNORM:
-        case Tegra::RenderTargetFormat::RG16_UNORM:
-        case Tegra::RenderTargetFormat::R16_UNORM:
-        case Tegra::RenderTargetFormat::B5G6R5_UNORM:
-        case Tegra::RenderTargetFormat::BGR5A1_UNORM:
-        case Tegra::RenderTargetFormat::RG8_UNORM:
-        case Tegra::RenderTargetFormat::RGBA16_UNORM:
-            return ComponentType::UNorm;
-        case Tegra::RenderTargetFormat::RGBA8_SNORM:
-        case Tegra::RenderTargetFormat::RG16_SNORM:
-        case Tegra::RenderTargetFormat::R16_SNORM:
-        case Tegra::RenderTargetFormat::RG8_SNORM:
-            return ComponentType::SNorm;
-        case Tegra::RenderTargetFormat::RGBA16_FLOAT:
-        case Tegra::RenderTargetFormat::R11G11B10_FLOAT:
-        case Tegra::RenderTargetFormat::RGBA32_FLOAT:
-        case Tegra::RenderTargetFormat::RG32_FLOAT:
-        case Tegra::RenderTargetFormat::RG16_FLOAT:
-        case Tegra::RenderTargetFormat::R16_FLOAT:
-        case Tegra::RenderTargetFormat::R32_FLOAT:
-            return ComponentType::Float;
-        case Tegra::RenderTargetFormat::RGBA32_UINT:
-        case Tegra::RenderTargetFormat::RGBA16_UINT:
-        case Tegra::RenderTargetFormat::RG16_UINT:
-        case Tegra::RenderTargetFormat::R8_UINT:
-        case Tegra::RenderTargetFormat::R16_UINT:
-        case Tegra::RenderTargetFormat::RG32_UINT:
-        case Tegra::RenderTargetFormat::R32_UINT:
-        case Tegra::RenderTargetFormat::RGBA8_UINT:
-            return ComponentType::UInt;
-        case Tegra::RenderTargetFormat::RG16_SINT:
-        case Tegra::RenderTargetFormat::R16_SINT:
-            return ComponentType::SInt;
-        default:
-            LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));
-            UNREACHABLE();
-        }
-    }
-
-    static PixelFormat PixelFormatFromGPUPixelFormat(Tegra::FramebufferConfig::PixelFormat format) {
-        switch (format) {
-        case Tegra::FramebufferConfig::PixelFormat::ABGR8:
-            return PixelFormat::ABGR8U;
-        default:
-            LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));
-            UNREACHABLE();
-        }
-    }
-
-    static ComponentType ComponentTypeFromDepthFormat(Tegra::DepthFormat format) {
-        switch (format) {
-        case Tegra::DepthFormat::Z16_UNORM:
-        case Tegra::DepthFormat::S8_Z24_UNORM:
-        case Tegra::DepthFormat::Z24_S8_UNORM:
-            return ComponentType::UNorm;
-        case Tegra::DepthFormat::Z32_FLOAT:
-        case Tegra::DepthFormat::Z32_S8_X24_FLOAT:
-            return ComponentType::Float;
-        default:
-            LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));
-            UNREACHABLE();
-        }
-    }
-
-    static SurfaceType GetFormatType(PixelFormat pixel_format) {
-        if (static_cast<std::size_t>(pixel_format) <
-            static_cast<std::size_t>(PixelFormat::MaxColorFormat)) {
-            return SurfaceType::ColorTexture;
-        }
-
-        if (static_cast<std::size_t>(pixel_format) <
-            static_cast<std::size_t>(PixelFormat::MaxDepthFormat)) {
-            return SurfaceType::Depth;
-        }
-
-        if (static_cast<std::size_t>(pixel_format) <
-            static_cast<std::size_t>(PixelFormat::MaxDepthStencilFormat)) {
-            return SurfaceType::DepthStencil;
-        }
-
-        // TODO(Subv): Implement the other formats
-        ASSERT(false);
-
-        return SurfaceType::Invalid;
-    }
-
-    /// Returns the sizer in bytes of the specified pixel format
-    static constexpr u32 GetBytesPerPixel(PixelFormat pixel_format) {
-        if (pixel_format == SurfaceParams::PixelFormat::Invalid) {
-            return 0;
-        }
-        return GetFormatBpp(pixel_format) / CHAR_BIT;
+        return VideoCore::Surface::GetFormatBpp(pixel_format);
     }
 
     /// Returns the rectangle corresponding to this surface
-    MathUtil::Rectangle<u32> GetRect() const;
+    MathUtil::Rectangle<u32> GetRect(u32 mip_level = 0) const;
 
     /// Returns the total size of this surface in bytes, adjusted for compression
     std::size_t SizeInBytesRaw(bool ignore_tiled = false) const {
@@ -742,6 +92,91 @@ struct SurfaceParams {
         return size_in_bytes_gl / 6;
     }
 
+    /// Returns the exact size of memory occupied by the texture in VRAM, including mipmaps.
+    std::size_t MemorySize() const {
+        std::size_t size = InnerMemorySize(false, is_layered);
+        if (is_layered)
+            return size * depth;
+        return size;
+    }
+
+    /// Returns the exact size of the memory occupied by a layer in a texture in VRAM, including
+    /// mipmaps.
+    std::size_t LayerMemorySize() const {
+        return InnerMemorySize(false, true);
+    }
+
+    /// Returns the size of a layer of this surface in OpenGL.
+    std::size_t LayerSizeGL(u32 mip_level) const {
+        return InnerMipmapMemorySize(mip_level, true, is_layered, false);
+    }
+
+    std::size_t GetMipmapSizeGL(u32 mip_level, bool ignore_compressed = true) const {
+        std::size_t size = InnerMipmapMemorySize(mip_level, true, is_layered, ignore_compressed);
+        if (is_layered)
+            return size * depth;
+        return size;
+    }
+
+    std::size_t GetMipmapLevelOffset(u32 mip_level) const {
+        std::size_t offset = 0;
+        for (u32 i = 0; i < mip_level; i++)
+            offset += InnerMipmapMemorySize(i, false, is_layered);
+        return offset;
+    }
+
+    std::size_t GetMipmapLevelOffsetGL(u32 mip_level) const {
+        std::size_t offset = 0;
+        for (u32 i = 0; i < mip_level; i++)
+            offset += InnerMipmapMemorySize(i, true, is_layered);
+        return offset;
+    }
+
+    u32 MipWidth(u32 mip_level) const {
+        return std::max(1U, width >> mip_level);
+    }
+
+    u32 MipHeight(u32 mip_level) const {
+        return std::max(1U, height >> mip_level);
+    }
+
+    u32 MipDepth(u32 mip_level) const {
+        return is_layered ? depth : std::max(1U, depth >> mip_level);
+    }
+
+    // Auto block resizing algorithm from:
+    // https://cgit.freedesktop.org/mesa/mesa/tree/src/gallium/drivers/nouveau/nv50/nv50_miptree.c
+    u32 MipBlockHeight(u32 mip_level) const {
+        if (mip_level == 0)
+            return block_height;
+        u32 alt_height = MipHeight(mip_level);
+        u32 h = GetDefaultBlockHeight(pixel_format);
+        u32 blocks_in_y = (alt_height + h - 1) / h;
+        u32 bh = 16;
+        while (bh > 1 && blocks_in_y <= bh * 4) {
+            bh >>= 1;
+        }
+        return bh;
+    }
+
+    u32 MipBlockDepth(u32 mip_level) const {
+        if (mip_level == 0)
+            return block_depth;
+        if (is_layered)
+            return 1;
+        u32 depth = MipDepth(mip_level);
+        u32 bd = 32;
+        while (bd > 1 && depth * 2 <= bd) {
+            bd >>= 1;
+        }
+        if (bd == 32) {
+            u32 bh = MipBlockHeight(mip_level);
+            if (bh >= 4)
+                return 16;
+        }
+        return bd;
+    }
+
     /// Creates SurfaceParams from a texture configuration
     static SurfaceParams CreateForTexture(const Tegra::Texture::FullTextureInfo& config,
                                           const GLShader::SamplerEntry& entry);
@@ -782,7 +217,8 @@ struct SurfaceParams {
     u32 unaligned_height;
     SurfaceTarget target;
     u32 max_mip_level;
-
+    bool is_layered;
+    bool srgb_conversion;
     // Parameters used for caching
     VAddr addr;
     Tegra::GPUVAddr gpu_addr;
@@ -797,6 +233,12 @@ struct SurfaceParams {
         u32 layer_stride;
         u32 base_layer;
     } rt;
+
+private:
+    std::size_t InnerMipmapMemorySize(u32 mip_level, bool force_gl = false, bool layer_only = false,
+                                      bool uncompressed = false) const;
+    std::size_t InnerMemorySize(bool force_gl = false, bool layer_only = false,
+                                bool uncompressed = false) const;
 };
 
 }; // namespace OpenGL
@@ -822,6 +264,8 @@ struct hash<SurfaceReserveKey> {
 
 namespace OpenGL {
 
+class RasterizerOpenGL;
+
 class CachedSurface final : public RasterizerCacheObject {
 public:
     CachedSurface(const SurfaceParams& params);
@@ -858,8 +302,10 @@ public:
     void UploadGLTexture(GLuint read_fb_handle, GLuint draw_fb_handle);
 
 private:
+    void UploadGLMipmapTexture(u32 mip_map, GLuint read_fb_handle, GLuint draw_fb_handle);
+
     OGLTexture texture;
-    std::vector<u8> gl_buffer;
+    std::vector<std::vector<u8>> gl_buffer;
     SurfaceParams params;
     GLenum gl_target;
     std::size_t cached_size_in_bytes;
@@ -867,7 +313,7 @@ private:
 
 class RasterizerCacheOpenGL final : public RasterizerCache<Surface> {
 public:
-    RasterizerCacheOpenGL();
+    explicit RasterizerCacheOpenGL(RasterizerOpenGL& rasterizer);
 
     /// Get a surface based on the texture configuration
     Surface GetTextureSurface(const Tegra::Texture::FullTextureInfo& config,
diff --git a/src/video_core/renderer_opengl/gl_resource_manager.cpp b/src/video_core/renderer_opengl/gl_resource_manager.cpp
new file mode 100644
index 000000000..c17d5ac00
--- /dev/null
+++ b/src/video_core/renderer_opengl/gl_resource_manager.cpp
@@ -0,0 +1,186 @@
+// Copyright 2015 Citra Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <utility>
+#include <glad/glad.h>
+#include "common/common_types.h"
+#include "common/microprofile.h"
+#include "video_core/renderer_opengl/gl_resource_manager.h"
+#include "video_core/renderer_opengl/gl_shader_util.h"
+#include "video_core/renderer_opengl/gl_state.h"
+
+MICROPROFILE_DEFINE(OpenGL_ResourceCreation, "OpenGL", "Resource Creation", MP_RGB(128, 128, 192));
+MICROPROFILE_DEFINE(OpenGL_ResourceDeletion, "OpenGL", "Resource Deletion", MP_RGB(128, 128, 192));
+
+namespace OpenGL {
+
+void OGLTexture::Create() {
+    if (handle != 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceCreation);
+    glGenTextures(1, &handle);
+}
+
+void OGLTexture::Release() {
+    if (handle == 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceDeletion);
+    glDeleteTextures(1, &handle);
+    OpenGLState::GetCurState().UnbindTexture(handle).Apply();
+    handle = 0;
+}
+
+void OGLSampler::Create() {
+    if (handle != 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceCreation);
+    glGenSamplers(1, &handle);
+}
+
+void OGLSampler::Release() {
+    if (handle == 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceDeletion);
+    glDeleteSamplers(1, &handle);
+    OpenGLState::GetCurState().ResetSampler(handle).Apply();
+    handle = 0;
+}
+
+void OGLShader::Create(const char* source, GLenum type) {
+    if (handle != 0)
+        return;
+    if (source == nullptr)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceCreation);
+    handle = GLShader::LoadShader(source, type);
+}
+
+void OGLShader::Release() {
+    if (handle == 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceDeletion);
+    glDeleteShader(handle);
+    handle = 0;
+}
+
+void OGLProgram::CreateFromSource(const char* vert_shader, const char* geo_shader,
+                                  const char* frag_shader, bool separable_program) {
+    OGLShader vert, geo, frag;
+    if (vert_shader)
+        vert.Create(vert_shader, GL_VERTEX_SHADER);
+    if (geo_shader)
+        geo.Create(geo_shader, GL_GEOMETRY_SHADER);
+    if (frag_shader)
+        frag.Create(frag_shader, GL_FRAGMENT_SHADER);
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceCreation);
+    Create(separable_program, vert.handle, geo.handle, frag.handle);
+}
+
+void OGLProgram::Release() {
+    if (handle == 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceDeletion);
+    glDeleteProgram(handle);
+    OpenGLState::GetCurState().ResetProgram(handle).Apply();
+    handle = 0;
+}
+
+void OGLPipeline::Create() {
+    if (handle != 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceCreation);
+    glGenProgramPipelines(1, &handle);
+}
+
+void OGLPipeline::Release() {
+    if (handle == 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceDeletion);
+    glDeleteProgramPipelines(1, &handle);
+    OpenGLState::GetCurState().ResetPipeline(handle).Apply();
+    handle = 0;
+}
+
+void OGLBuffer::Create() {
+    if (handle != 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceCreation);
+    glGenBuffers(1, &handle);
+}
+
+void OGLBuffer::Release() {
+    if (handle == 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceDeletion);
+    glDeleteBuffers(1, &handle);
+    OpenGLState::GetCurState().ResetBuffer(handle).Apply();
+    handle = 0;
+}
+
+void OGLSync::Create() {
+    if (handle != 0)
+        return;
+
+    // Don't profile here, this one is expected to happen ingame.
+    handle = glFenceSync(GL_SYNC_GPU_COMMANDS_COMPLETE, 0);
+}
+
+void OGLSync::Release() {
+    if (handle == 0)
+        return;
+
+    // Don't profile here, this one is expected to happen ingame.
+    glDeleteSync(handle);
+    handle = 0;
+}
+
+void OGLVertexArray::Create() {
+    if (handle != 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceCreation);
+    glGenVertexArrays(1, &handle);
+}
+
+void OGLVertexArray::Release() {
+    if (handle == 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceDeletion);
+    glDeleteVertexArrays(1, &handle);
+    OpenGLState::GetCurState().ResetVertexArray(handle).Apply();
+    handle = 0;
+}
+
+void OGLFramebuffer::Create() {
+    if (handle != 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceCreation);
+    glGenFramebuffers(1, &handle);
+}
+
+void OGLFramebuffer::Release() {
+    if (handle == 0)
+        return;
+
+    MICROPROFILE_SCOPE(OpenGL_ResourceDeletion);
+    glDeleteFramebuffers(1, &handle);
+    OpenGLState::GetCurState().ResetFramebuffer(handle).Apply();
+    handle = 0;
+}
+
+} // namespace OpenGL
diff --git a/src/video_core/renderer_opengl/gl_resource_manager.h b/src/video_core/renderer_opengl/gl_resource_manager.h
index 3bc1b83b5..e33f1e973 100644
--- a/src/video_core/renderer_opengl/gl_resource_manager.h
+++ b/src/video_core/renderer_opengl/gl_resource_manager.h
@@ -8,7 +8,6 @@
 #include <glad/glad.h>
 #include "common/common_types.h"
 #include "video_core/renderer_opengl/gl_shader_util.h"
-#include "video_core/renderer_opengl/gl_state.h"
 
 namespace OpenGL {
 
@@ -29,20 +28,10 @@ public:
     }
 
     /// Creates a new internal OpenGL resource and stores the handle
-    void Create() {
-        if (handle != 0)
-            return;
-        glGenTextures(1, &handle);
-    }
+    void Create();
 
     /// Deletes the internal OpenGL resource
-    void Release() {
-        if (handle == 0)
-            return;
-        glDeleteTextures(1, &handle);
-        OpenGLState::GetCurState().UnbindTexture(handle).Apply();
-        handle = 0;
-    }
+    void Release();
 
     GLuint handle = 0;
 };
@@ -64,20 +53,10 @@ public:
     }
 
     /// Creates a new internal OpenGL resource and stores the handle
-    void Create() {
-        if (handle != 0)
-            return;
-        glGenSamplers(1, &handle);
-    }
+    void Create();
 
     /// Deletes the internal OpenGL resource
-    void Release() {
-        if (handle == 0)
-            return;
-        glDeleteSamplers(1, &handle);
-        OpenGLState::GetCurState().ResetSampler(handle).Apply();
-        handle = 0;
-    }
+    void Release();
 
     GLuint handle = 0;
 };
@@ -98,20 +77,9 @@ public:
         return *this;
     }
 
-    void Create(const char* source, GLenum type) {
-        if (handle != 0)
-            return;
-        if (source == nullptr)
-            return;
-        handle = GLShader::LoadShader(source, type);
-    }
+    void Create(const char* source, GLenum type);
 
-    void Release() {
-        if (handle == 0)
-            return;
-        glDeleteShader(handle);
-        handle = 0;
-    }
+    void Release();
 
     GLuint handle = 0;
 };
@@ -141,25 +109,10 @@ public:
 
     /// Creates a new internal OpenGL resource and stores the handle
     void CreateFromSource(const char* vert_shader, const char* geo_shader, const char* frag_shader,
-                          bool separable_program = false) {
-        OGLShader vert, geo, frag;
-        if (vert_shader)
-            vert.Create(vert_shader, GL_VERTEX_SHADER);
-        if (geo_shader)
-            geo.Create(geo_shader, GL_GEOMETRY_SHADER);
-        if (frag_shader)
-            frag.Create(frag_shader, GL_FRAGMENT_SHADER);
-        Create(separable_program, vert.handle, geo.handle, frag.handle);
-    }
+                          bool separable_program = false);
 
     /// Deletes the internal OpenGL resource
-    void Release() {
-        if (handle == 0)
-            return;
-        glDeleteProgram(handle);
-        OpenGLState::GetCurState().ResetProgram(handle).Apply();
-        handle = 0;
-    }
+    void Release();
 
     GLuint handle = 0;
 };
@@ -178,20 +131,10 @@ public:
     }
 
     /// Creates a new internal OpenGL resource and stores the handle
-    void Create() {
-        if (handle != 0)
-            return;
-        glGenProgramPipelines(1, &handle);
-    }
+    void Create();
 
     /// Deletes the internal OpenGL resource
-    void Release() {
-        if (handle == 0)
-            return;
-        glDeleteProgramPipelines(1, &handle);
-        OpenGLState::GetCurState().ResetPipeline(handle).Apply();
-        handle = 0;
-    }
+    void Release();
 
     GLuint handle = 0;
 };
@@ -213,20 +156,10 @@ public:
     }
 
     /// Creates a new internal OpenGL resource and stores the handle
-    void Create() {
-        if (handle != 0)
-            return;
-        glGenBuffers(1, &handle);
-    }
+    void Create();
 
     /// Deletes the internal OpenGL resource
-    void Release() {
-        if (handle == 0)
-            return;
-        glDeleteBuffers(1, &handle);
-        OpenGLState::GetCurState().ResetBuffer(handle).Apply();
-        handle = 0;
-    }
+    void Release();
 
     GLuint handle = 0;
 };
@@ -247,19 +180,10 @@ public:
     }
 
     /// Creates a new internal OpenGL resource and stores the handle
-    void Create() {
-        if (handle != 0)
-            return;
-        handle = glFenceSync(GL_SYNC_GPU_COMMANDS_COMPLETE, 0);
-    }
+    void Create();
 
     /// Deletes the internal OpenGL resource
-    void Release() {
-        if (handle == 0)
-            return;
-        glDeleteSync(handle);
-        handle = 0;
-    }
+    void Release();
 
     GLsync handle = 0;
 };
@@ -281,20 +205,10 @@ public:
     }
 
     /// Creates a new internal OpenGL resource and stores the handle
-    void Create() {
-        if (handle != 0)
-            return;
-        glGenVertexArrays(1, &handle);
-    }
+    void Create();
 
     /// Deletes the internal OpenGL resource
-    void Release() {
-        if (handle == 0)
-            return;
-        glDeleteVertexArrays(1, &handle);
-        OpenGLState::GetCurState().ResetVertexArray(handle).Apply();
-        handle = 0;
-    }
+    void Release();
 
     GLuint handle = 0;
 };
@@ -316,20 +230,10 @@ public:
     }
 
     /// Creates a new internal OpenGL resource and stores the handle
-    void Create() {
-        if (handle != 0)
-            return;
-        glGenFramebuffers(1, &handle);
-    }
+    void Create();
 
     /// Deletes the internal OpenGL resource
-    void Release() {
-        if (handle == 0)
-            return;
-        glDeleteFramebuffers(1, &handle);
-        OpenGLState::GetCurState().ResetFramebuffer(handle).Apply();
-        handle = 0;
-    }
+    void Release();
 
     GLuint handle = 0;
 };
diff --git a/src/video_core/renderer_opengl/gl_shader_cache.cpp b/src/video_core/renderer_opengl/gl_shader_cache.cpp
index 1a03a677f..a85a7c0c5 100644
--- a/src/video_core/renderer_opengl/gl_shader_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_cache.cpp
@@ -6,9 +6,10 @@
 #include "core/core.h"
 #include "core/memory.h"
 #include "video_core/engines/maxwell_3d.h"
+#include "video_core/renderer_opengl/gl_rasterizer.h"
 #include "video_core/renderer_opengl/gl_shader_cache.h"
 #include "video_core/renderer_opengl/gl_shader_manager.h"
-#include "video_core/utils.h"
+#include "video_core/renderer_opengl/utils.h"
 
 namespace OpenGL {
 
@@ -89,7 +90,7 @@ CachedShader::CachedShader(VAddr addr, Maxwell::ShaderProgram program_type)
         shader.Create(program_result.first.c_str(), gl_type);
         program.Create(true, shader.handle);
         SetShaderUniformBlockBindings(program.handle);
-        VideoCore::LabelGLObject(GL_PROGRAM, program.handle, addr);
+        LabelGLObject(GL_PROGRAM, program.handle, addr);
     } else {
         // Store shader's code to lazily build it on draw
         geometry_programs.code = program_result.first;
@@ -120,20 +121,26 @@ GLint CachedShader::GetUniformLocation(const GLShader::SamplerEntry& sampler) {
 }
 
 GLuint CachedShader::LazyGeometryProgram(OGLProgram& target_program,
-                                         const std::string& glsl_topology,
+                                         const std::string& glsl_topology, u32 max_vertices,
                                          const std::string& debug_name) {
     if (target_program.handle != 0) {
         return target_program.handle;
     }
-    const std::string source{geometry_programs.code + "layout (" + glsl_topology + ") in;\n"};
+    std::string source = "#version 430 core\n";
+    source += "layout (" + glsl_topology + ") in;\n";
+    source += "#define MAX_VERTEX_INPUT " + std::to_string(max_vertices) + '\n';
+    source += geometry_programs.code;
+
     OGLShader shader;
     shader.Create(source.c_str(), GL_GEOMETRY_SHADER);
     target_program.Create(true, shader.handle);
     SetShaderUniformBlockBindings(target_program.handle);
-    VideoCore::LabelGLObject(GL_PROGRAM, target_program.handle, addr, debug_name);
+    LabelGLObject(GL_PROGRAM, target_program.handle, addr, debug_name);
     return target_program.handle;
 };
 
+ShaderCacheOpenGL::ShaderCacheOpenGL(RasterizerOpenGL& rasterizer) : RasterizerCache{rasterizer} {}
+
 Shader ShaderCacheOpenGL::GetStageProgram(Maxwell::ShaderProgram program) {
     const VAddr program_addr{GetShaderAddress(program)};
 
diff --git a/src/video_core/renderer_opengl/gl_shader_cache.h b/src/video_core/renderer_opengl/gl_shader_cache.h
index a210f1731..ffbf21831 100644
--- a/src/video_core/renderer_opengl/gl_shader_cache.h
+++ b/src/video_core/renderer_opengl/gl_shader_cache.h
@@ -16,6 +16,8 @@
 namespace OpenGL {
 
 class CachedShader;
+class RasterizerOpenGL;
+
 using Shader = std::shared_ptr<CachedShader>;
 using Maxwell = Tegra::Engines::Maxwell3D::Regs;
 
@@ -46,22 +48,23 @@ public:
         }
         switch (primitive_mode) {
         case GL_POINTS:
-            return LazyGeometryProgram(geometry_programs.points, "points", "ShaderPoints");
+            return LazyGeometryProgram(geometry_programs.points, "points", 1, "ShaderPoints");
         case GL_LINES:
         case GL_LINE_STRIP:
-            return LazyGeometryProgram(geometry_programs.lines, "lines", "ShaderLines");
+            return LazyGeometryProgram(geometry_programs.lines, "lines", 2, "ShaderLines");
         case GL_LINES_ADJACENCY:
         case GL_LINE_STRIP_ADJACENCY:
-            return LazyGeometryProgram(geometry_programs.lines_adjacency, "lines_adjacency",
+            return LazyGeometryProgram(geometry_programs.lines_adjacency, "lines_adjacency", 4,
                                        "ShaderLinesAdjacency");
         case GL_TRIANGLES:
         case GL_TRIANGLE_STRIP:
         case GL_TRIANGLE_FAN:
-            return LazyGeometryProgram(geometry_programs.triangles, "triangles", "ShaderTriangles");
+            return LazyGeometryProgram(geometry_programs.triangles, "triangles", 3,
+                                       "ShaderTriangles");
         case GL_TRIANGLES_ADJACENCY:
         case GL_TRIANGLE_STRIP_ADJACENCY:
             return LazyGeometryProgram(geometry_programs.triangles_adjacency, "triangles_adjacency",
-                                       "ShaderLines");
+                                       6, "ShaderTrianglesAdjacency");
         default:
             UNREACHABLE_MSG("Unknown primitive mode.");
         }
@@ -76,7 +79,7 @@ public:
 private:
     /// Generates a geometry shader or returns one that already exists.
     GLuint LazyGeometryProgram(OGLProgram& target_program, const std::string& glsl_topology,
-                               const std::string& debug_name);
+                               u32 max_vertices, const std::string& debug_name);
 
     VAddr addr;
     Maxwell::ShaderProgram program_type;
@@ -104,6 +107,8 @@ private:
 
 class ShaderCacheOpenGL final : public RasterizerCache<Shader> {
 public:
+    explicit ShaderCacheOpenGL(RasterizerOpenGL& rasterizer);
+
     /// Gets the current specified shader stage program
     Shader GetStageProgram(Maxwell::ShaderProgram program);
 };
diff --git a/src/video_core/renderer_opengl/gl_shader_decompiler.cpp b/src/video_core/renderer_opengl/gl_shader_decompiler.cpp
index fe4d1bd83..5fde22ad4 100644
--- a/src/video_core/renderer_opengl/gl_shader_decompiler.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_decompiler.cpp
@@ -3,11 +3,12 @@
 // Refer to the license.txt file included.
 
 #include <map>
+#include <optional>
 #include <set>
 #include <string>
 #include <string_view>
+#include <unordered_set>
 
-#include <boost/optional.hpp>
 #include <fmt/format.h>
 
 #include "common/assert.h"
@@ -143,7 +144,7 @@ private:
         for (u32 offset = begin; offset != end && offset != PROGRAM_END; ++offset) {
             const Instruction instr = {program_code[offset]};
             if (const auto opcode = OpCode::Decode(instr)) {
-                switch (opcode->GetId()) {
+                switch (opcode->get().GetId()) {
                 case OpCode::Id::EXIT: {
                     // The EXIT instruction can be predicated, which means that the shader can
                     // conditionally end on this instruction. We have to consider the case where the
@@ -276,7 +277,8 @@ public:
     GLSLRegisterManager(ShaderWriter& shader, ShaderWriter& declarations,
                         const Maxwell3D::Regs::ShaderStage& stage, const std::string& suffix,
                         const Tegra::Shader::Header& header)
-        : shader{shader}, declarations{declarations}, stage{stage}, suffix{suffix}, header{header} {
+        : shader{shader}, declarations{declarations}, stage{stage}, suffix{suffix}, header{header},
+          fixed_pipeline_output_attributes_used{}, local_memory_size{0} {
         BuildRegisterList();
         BuildInputList();
     }
@@ -339,10 +341,10 @@ public:
      */
     void SetRegisterToFloat(const Register& reg, u64 elem, const std::string& value,
                             u64 dest_num_components, u64 value_num_components,
-                            bool is_saturated = false, u64 dest_elem = 0) {
+                            bool is_saturated = false, u64 dest_elem = 0, bool precise = false) {
 
         SetRegister(reg, elem, is_saturated ? "clamp(" + value + ", 0.0, 1.0)" : value,
-                    dest_num_components, value_num_components, dest_elem);
+                    dest_num_components, value_num_components, dest_elem, precise);
     }
 
     /**
@@ -366,11 +368,12 @@ public:
         const std::string func{is_signed ? "intBitsToFloat" : "uintBitsToFloat"};
 
         SetRegister(reg, elem, func + '(' + ConvertIntegerSize(value, size) + ')',
-                    dest_num_components, value_num_components, dest_elem);
+                    dest_num_components, value_num_components, dest_elem, false);
 
         if (sets_cc) {
             const std::string zero_condition = "( " + ConvertIntegerSize(value, size) + " == 0 )";
             SetInternalFlag(InternalFlag::ZeroFlag, zero_condition);
+            LOG_WARNING(HW_GPU, "Control Codes Imcomplete.");
         }
     }
 
@@ -414,7 +417,7 @@ public:
             }
         }();
 
-        SetRegister(reg, elem, result, dest_num_components, value_num_components, dest_elem);
+        SetRegister(reg, elem, result, dest_num_components, value_num_components, dest_elem, false);
     }
 
     /**
@@ -428,12 +431,31 @@ public:
      */
     void SetRegisterToInputAttibute(const Register& reg, u64 elem, Attribute::Index attribute,
                                     const Tegra::Shader::IpaMode& input_mode,
-                                    boost::optional<Register> vertex = {}) {
+                                    std::optional<Register> vertex = {}) {
         const std::string dest = GetRegisterAsFloat(reg);
         const std::string src = GetInputAttribute(attribute, input_mode, vertex) + GetSwizzle(elem);
         shader.AddLine(dest + " = " + src + ';');
     }
 
+    std::string GetLocalMemoryAsFloat(const std::string& index) {
+        return "lmem[" + index + ']';
+    }
+
+    std::string GetLocalMemoryAsInteger(const std::string& index, bool is_signed = false) {
+        const std::string func{is_signed ? "floatToIntBits" : "floatBitsToUint"};
+        return func + "(lmem[" + index + "])";
+    }
+
+    void SetLocalMemoryAsFloat(const std::string& index, const std::string& value) {
+        shader.AddLine("lmem[" + index + "] = " + value + ';');
+    }
+
+    void SetLocalMemoryAsInteger(const std::string& index, const std::string& value,
+                                 bool is_signed = false) {
+        const std::string func{is_signed ? "intBitsToFloat" : "uintBitsToFloat"};
+        shader.AddLine("lmem[" + index + "] = " + func + '(' + value + ");");
+    }
+
     std::string GetControlCode(const Tegra::Shader::ControlCode cc) const {
         switch (cc) {
         case Tegra::Shader::ControlCode::NEU:
@@ -472,15 +494,20 @@ public:
             // instruction for now.
             if (stage == Maxwell3D::Regs::ShaderStage::Geometry) {
                 // TODO(Rodrigo): nouveau sets some attributes after setting emitting a geometry
-                // shader. These instructions use a dirty register as buffer index. To avoid some
-                // drivers from complaining for the out of boundary writes, guard them.
-                const std::string buf_index{"min(" + GetRegisterAsInteger(buf_reg) + ", " +
-                                            std::to_string(MAX_GEOMETRY_BUFFERS - 1) + ')'};
+                // shader. These instructions use a dirty register as buffer index, to avoid some
+                // drivers from complaining about out of boundary writes, guard them.
+                const std::string buf_index{"((" + GetRegisterAsInteger(buf_reg) + ") % " +
+                                            std::to_string(MAX_GEOMETRY_BUFFERS) + ')'};
                 shader.AddLine("amem[" + buf_index + "][" +
                                std::to_string(static_cast<u32>(attribute)) + ']' +
                                GetSwizzle(elem) + " = " + src + ';');
             } else {
-                shader.AddLine(dest + GetSwizzle(elem) + " = " + src + ';');
+                if (attribute == Attribute::Index::PointSize) {
+                    fixed_pipeline_output_attributes_used.insert(attribute);
+                    shader.AddLine(dest + " = " + src + ';');
+                } else {
+                    shader.AddLine(dest + GetSwizzle(elem) + " = " + src + ';');
+                }
             }
         }
     }
@@ -524,7 +551,9 @@ public:
 
     /// Add declarations.
     void GenerateDeclarations(const std::string& suffix) {
+        GenerateVertex();
         GenerateRegisters(suffix);
+        GenerateLocalMemory();
         GenerateInternalFlags();
         GenerateInputAttrs();
         GenerateOutputAttrs();
@@ -570,6 +599,10 @@ public:
         return entry.GetName();
     }
 
+    void SetLocalMemory(u64 lmem) {
+        local_memory_size = lmem;
+    }
+
 private:
     /// Generates declarations for registers.
     void GenerateRegisters(const std::string& suffix) {
@@ -580,6 +613,15 @@ private:
         declarations.AddNewLine();
     }
 
+    /// Generates declarations for local memory.
+    void GenerateLocalMemory() {
+        if (local_memory_size > 0) {
+            declarations.AddLine("float lmem[" + std::to_string((local_memory_size - 1 + 4) / 4) +
+                                 "];");
+            declarations.AddNewLine();
+        }
+    }
+
     /// Generates declarations for internal flags.
     void GenerateInternalFlags() {
         for (u32 ii = 0; ii < static_cast<u64>(InternalFlag::Amount); ii++) {
@@ -683,6 +725,20 @@ private:
         declarations.AddNewLine();
     }
 
+    void GenerateVertex() {
+        if (stage != Maxwell3D::Regs::ShaderStage::Vertex)
+            return;
+        declarations.AddLine("out gl_PerVertex {");
+        ++declarations.scope;
+        declarations.AddLine("vec4 gl_Position;");
+        for (auto& o : fixed_pipeline_output_attributes_used) {
+            if (o == Attribute::Index::PointSize)
+                declarations.AddLine("float gl_PointSize;");
+        }
+        --declarations.scope;
+        declarations.AddLine("};");
+    }
+
     /// Generates code representing a temporary (GPR) register.
     std::string GetRegister(const Register& reg, unsigned elem) {
         if (reg == Register::ZeroIndex) {
@@ -702,7 +758,8 @@ private:
      * @param dest_elem Optional, the destination element to use for the operation.
      */
     void SetRegister(const Register& reg, u64 elem, const std::string& value,
-                     u64 dest_num_components, u64 value_num_components, u64 dest_elem) {
+                     u64 dest_num_components, u64 value_num_components, u64 dest_elem,
+                     bool precise) {
         if (reg == Register::ZeroIndex) {
             LOG_CRITICAL(HW_GPU, "Cannot set Register::ZeroIndex");
             UNREACHABLE();
@@ -719,7 +776,18 @@ private:
             src += GetSwizzle(elem);
         }
 
-        shader.AddLine(dest + " = " + src + ';');
+        if (precise && stage != Maxwell3D::Regs::ShaderStage::Fragment) {
+            shader.AddLine('{');
+            ++shader.scope;
+            // This avoids optimizations of constant propagation and keeps the code as the original
+            // Sadly using the precise keyword causes "linking" errors on fragment shaders.
+            shader.AddLine("precise float tmp = " + src + ';');
+            shader.AddLine(dest + " = tmp;");
+            --shader.scope;
+            shader.AddLine('}');
+        } else {
+            shader.AddLine(dest + " = " + src + ';');
+        }
     }
 
     /// Build the GLSL register list.
@@ -740,10 +808,14 @@ private:
     /// Generates code representing an input attribute register.
     std::string GetInputAttribute(Attribute::Index attribute,
                                   const Tegra::Shader::IpaMode& input_mode,
-                                  boost::optional<Register> vertex = {}) {
+                                  std::optional<Register> vertex = {}) {
         auto GeometryPass = [&](const std::string& name) {
             if (stage == Maxwell3D::Regs::ShaderStage::Geometry && vertex) {
-                return "gs_" + name + '[' + GetRegisterAsInteger(vertex.value(), 0, false) + ']';
+                // TODO(Rodrigo): Guard geometry inputs against out of bound reads. Some games set
+                // an 0x80000000 index for those and the shader fails to build. Find out why this
+                // happens and what's its intent.
+                return "gs_" + name + '[' + GetRegisterAsInteger(*vertex, 0, false) +
+                       " % MAX_VERTEX_INPUT]";
             }
             return name;
         };
@@ -836,6 +908,8 @@ private:
     /// Generates code representing the declaration name of an output attribute register.
     std::string GetOutputAttribute(Attribute::Index attribute) {
         switch (attribute) {
+        case Attribute::Index::PointSize:
+            return "gl_PointSize";
         case Attribute::Index::Position:
             return "position";
         default:
@@ -870,6 +944,8 @@ private:
     const Maxwell3D::Regs::ShaderStage& stage;
     const std::string& suffix;
     const Tegra::Shader::Header& header;
+    std::unordered_set<Attribute::Index> fixed_pipeline_output_attributes_used;
+    u64 local_memory_size;
 };
 
 class GLSLGenerator {
@@ -879,6 +955,8 @@ public:
         : subroutines(subroutines), program_code(program_code), main_offset(main_offset),
           stage(stage), suffix(suffix) {
         std::memcpy(&header, program_code.data(), sizeof(Tegra::Shader::Header));
+        local_memory_size = header.GetLocalMemorySize();
+        regs.SetLocalMemory(local_memory_size);
         Generate(suffix);
     }
 
@@ -1392,7 +1470,7 @@ private:
         }
 
         shader.AddLine(
-            fmt::format("// {}: {} (0x{:016x})", offset, opcode->GetName(), instr.value));
+            fmt::format("// {}: {} (0x{:016x})", offset, opcode->get().GetName(), instr.value));
 
         using Tegra::Shader::Pred;
         ASSERT_MSG(instr.pred.full_pred != Pred::NeverExecute,
@@ -1400,7 +1478,7 @@ private:
 
         // Some instructions (like SSY) don't have a predicate field, they are always
         // unconditionally executed.
-        bool can_be_predicated = OpCode::IsPredicatedInstruction(opcode->GetId());
+        bool can_be_predicated = OpCode::IsPredicatedInstruction(opcode->get().GetId());
 
         if (can_be_predicated && instr.pred.pred_index != static_cast<u64>(Pred::UnusedIndex)) {
             shader.AddLine("if (" +
@@ -1410,7 +1488,7 @@ private:
             ++shader.scope;
         }
 
-        switch (opcode->GetType()) {
+        switch (opcode->get().GetType()) {
         case OpCode::Type::Arithmetic: {
             std::string op_a = regs.GetRegisterAsFloat(instr.gpr8);
 
@@ -1427,7 +1505,7 @@ private:
                 }
             }
 
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::MOV_C:
             case OpCode::Id::MOV_R: {
                 // MOV does not have neither 'abs' nor 'neg' bits.
@@ -1449,8 +1527,13 @@ private:
                 ASSERT_MSG(instr.fmul.cc == 0, "FMUL cc is not implemented");
 
                 op_b = GetOperandAbsNeg(op_b, false, instr.fmul.negate_b);
+
                 regs.SetRegisterToFloat(instr.gpr0, 0, op_a + " * " + op_b, 1, 1,
-                                        instr.alu.saturate_d);
+                                        instr.alu.saturate_d, 0, true);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "FMUL Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::FADD_C:
@@ -1458,8 +1541,13 @@ private:
             case OpCode::Id::FADD_IMM: {
                 op_a = GetOperandAbsNeg(op_a, instr.alu.abs_a, instr.alu.negate_a);
                 op_b = GetOperandAbsNeg(op_b, instr.alu.abs_b, instr.alu.negate_b);
+
                 regs.SetRegisterToFloat(instr.gpr0, 0, op_a + " + " + op_b, 1, 1,
-                                        instr.alu.saturate_d);
+                                        instr.alu.saturate_d, 0, true);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "FADD Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::MUFU: {
@@ -1467,31 +1555,31 @@ private:
                 switch (instr.sub_op) {
                 case SubOp::Cos:
                     regs.SetRegisterToFloat(instr.gpr0, 0, "cos(" + op_a + ')', 1, 1,
-                                            instr.alu.saturate_d);
+                                            instr.alu.saturate_d, 0, true);
                     break;
                 case SubOp::Sin:
                     regs.SetRegisterToFloat(instr.gpr0, 0, "sin(" + op_a + ')', 1, 1,
-                                            instr.alu.saturate_d);
+                                            instr.alu.saturate_d, 0, true);
                     break;
                 case SubOp::Ex2:
                     regs.SetRegisterToFloat(instr.gpr0, 0, "exp2(" + op_a + ')', 1, 1,
-                                            instr.alu.saturate_d);
+                                            instr.alu.saturate_d, 0, true);
                     break;
                 case SubOp::Lg2:
                     regs.SetRegisterToFloat(instr.gpr0, 0, "log2(" + op_a + ')', 1, 1,
-                                            instr.alu.saturate_d);
+                                            instr.alu.saturate_d, 0, true);
                     break;
                 case SubOp::Rcp:
                     regs.SetRegisterToFloat(instr.gpr0, 0, "1.0 / " + op_a, 1, 1,
-                                            instr.alu.saturate_d);
+                                            instr.alu.saturate_d, 0, true);
                     break;
                 case SubOp::Rsq:
                     regs.SetRegisterToFloat(instr.gpr0, 0, "inversesqrt(" + op_a + ')', 1, 1,
-                                            instr.alu.saturate_d);
+                                            instr.alu.saturate_d, 0, true);
                     break;
                 case SubOp::Sqrt:
                     regs.SetRegisterToFloat(instr.gpr0, 0, "sqrt(" + op_a + ')', 1, 1,
-                                            instr.alu.saturate_d);
+                                            instr.alu.saturate_d, 0, true);
                     break;
                 default:
                     LOG_CRITICAL(HW_GPU, "Unhandled MUFU sub op: {0:x}",
@@ -1512,7 +1600,11 @@ private:
                 regs.SetRegisterToFloat(instr.gpr0, 0,
                                         '(' + condition + ") ? min(" + parameters + ") : max(" +
                                             parameters + ')',
-                                        1, 1);
+                                        1, 1, false, 0, true);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "FMNMX Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::RRO_C:
@@ -1525,14 +1617,15 @@ private:
                 break;
             }
             default: {
-                LOG_CRITICAL(HW_GPU, "Unhandled arithmetic instruction: {}", opcode->GetName());
+                LOG_CRITICAL(HW_GPU, "Unhandled arithmetic instruction: {}",
+                             opcode->get().GetName());
                 UNREACHABLE();
             }
             }
             break;
         }
         case OpCode::Type::ArithmeticImmediate: {
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::MOV32_IMM: {
                 regs.SetRegisterToFloat(instr.gpr0, 0, GetImmediate32(instr), 1, 1);
                 break;
@@ -1541,7 +1634,11 @@ private:
                 regs.SetRegisterToFloat(instr.gpr0, 0,
                                         regs.GetRegisterAsFloat(instr.gpr8) + " * " +
                                             GetImmediate32(instr),
-                                        1, 1, instr.fmul32.saturate);
+                                        1, 1, instr.fmul32.saturate, 0, true);
+                if (instr.op_32.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "FMUL32 Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::FADD32I: {
@@ -1564,7 +1661,11 @@ private:
                     op_b = "-(" + op_b + ')';
                 }
 
-                regs.SetRegisterToFloat(instr.gpr0, 0, op_a + " + " + op_b, 1, 1);
+                regs.SetRegisterToFloat(instr.gpr0, 0, op_a + " + " + op_b, 1, 1, false, 0, true);
+                if (instr.op_32.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "FADD32 Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             }
@@ -1576,7 +1677,7 @@ private:
             std::string op_a = instr.bfe.negate_a ? "-" : "";
             op_a += regs.GetRegisterAsInteger(instr.gpr8);
 
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::BFE_IMM: {
                 std::string inner_shift =
                     '(' + op_a + " << " + std::to_string(instr.bfe.GetLeftShiftValue()) + ')';
@@ -1585,10 +1686,14 @@ private:
                     std::to_string(instr.bfe.GetLeftShiftValue() + instr.bfe.shift_position) + ')';
 
                 regs.SetRegisterToInteger(instr.gpr0, true, 0, outer_shift, 1, 1);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "BFE Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             default: {
-                LOG_CRITICAL(HW_GPU, "Unhandled BFE instruction: {}", opcode->GetName());
+                LOG_CRITICAL(HW_GPU, "Unhandled BFE instruction: {}", opcode->get().GetName());
                 UNREACHABLE();
             }
             }
@@ -1610,7 +1715,7 @@ private:
                 }
             }
 
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::SHR_C:
             case OpCode::Id::SHR_R:
             case OpCode::Id::SHR_IMM: {
@@ -1622,15 +1727,23 @@ private:
                 // Cast to int is superfluous for arithmetic shift, it's only for a logical shift
                 regs.SetRegisterToInteger(instr.gpr0, true, 0, "int(" + op_a + " >> " + op_b + ')',
                                           1, 1);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "SHR Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::SHL_C:
             case OpCode::Id::SHL_R:
             case OpCode::Id::SHL_IMM:
                 regs.SetRegisterToInteger(instr.gpr0, true, 0, op_a + " << " + op_b, 1, 1);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "SHL Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             default: {
-                LOG_CRITICAL(HW_GPU, "Unhandled shift instruction: {}", opcode->GetName());
+                LOG_CRITICAL(HW_GPU, "Unhandled shift instruction: {}", opcode->get().GetName());
                 UNREACHABLE();
             }
             }
@@ -1640,13 +1753,17 @@ private:
             std::string op_a = regs.GetRegisterAsInteger(instr.gpr8);
             std::string op_b = std::to_string(instr.alu.imm20_32.Value());
 
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::IADD32I:
                 if (instr.iadd32i.negate_a)
                     op_a = "-(" + op_a + ')';
 
                 regs.SetRegisterToInteger(instr.gpr0, true, 0, op_a + " + " + op_b, 1, 1,
                                           instr.iadd32i.saturate != 0);
+                if (instr.op_32.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "IADD32 Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             case OpCode::Id::LOP32I: {
                 if (instr.alu.lop32i.invert_a)
@@ -1658,11 +1775,15 @@ private:
                 WriteLogicOperation(instr.gpr0, instr.alu.lop32i.operation, op_a, op_b,
                                     Tegra::Shader::PredicateResultMode::None,
                                     Tegra::Shader::Pred::UnusedIndex);
+                if (instr.op_32.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "LOP32I Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             default: {
                 LOG_CRITICAL(HW_GPU, "Unhandled ArithmeticIntegerImmediate instruction: {}",
-                             opcode->GetName());
+                             opcode->get().GetName());
                 UNREACHABLE();
             }
             }
@@ -1682,7 +1803,7 @@ private:
                 }
             }
 
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::IADD_C:
             case OpCode::Id::IADD_R:
             case OpCode::Id::IADD_IMM: {
@@ -1694,6 +1815,10 @@ private:
 
                 regs.SetRegisterToInteger(instr.gpr0, true, 0, op_a + " + " + op_b, 1, 1,
                                           instr.alu.saturate_d);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "IADD Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::IADD3_C:
@@ -1718,7 +1843,7 @@ private:
                     }
                 };
 
-                if (opcode->GetId() == OpCode::Id::IADD3_R) {
+                if (opcode->get().GetId() == OpCode::Id::IADD3_R) {
                     apply_height(instr.iadd3.height_a, op_a);
                     apply_height(instr.iadd3.height_b, op_b);
                     apply_height(instr.iadd3.height_c, op_c);
@@ -1734,7 +1859,7 @@ private:
                     op_c = "-(" + op_c + ')';
 
                 std::string result;
-                if (opcode->GetId() == OpCode::Id::IADD3_R) {
+                if (opcode->get().GetId() == OpCode::Id::IADD3_R) {
                     switch (instr.iadd3.mode) {
                     case Tegra::Shader::IAdd3Mode::RightShift:
                         // TODO(tech4me): According to
@@ -1755,6 +1880,11 @@ private:
                 }
 
                 regs.SetRegisterToInteger(instr.gpr0, true, 0, result, 1, 1);
+
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "IADD3 Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::ISCADD_C:
@@ -1770,6 +1900,10 @@ private:
 
                 regs.SetRegisterToInteger(instr.gpr0, true, 0,
                                           "((" + op_a + " << " + shift + ") + " + op_b + ')', 1, 1);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "ISCADD Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::POPC_C:
@@ -1801,6 +1935,10 @@ private:
 
                 WriteLogicOperation(instr.gpr0, instr.alu.lop.operation, op_a, op_b,
                                     instr.alu.lop.pred_result_mode, instr.alu.lop.pred48);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "LOP Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::LOP3_C:
@@ -1809,13 +1947,17 @@ private:
                 const std::string op_c = regs.GetRegisterAsInteger(instr.gpr39);
                 std::string lut;
 
-                if (opcode->GetId() == OpCode::Id::LOP3_R) {
+                if (opcode->get().GetId() == OpCode::Id::LOP3_R) {
                     lut = '(' + std::to_string(instr.alu.lop3.GetImmLut28()) + ')';
                 } else {
                     lut = '(' + std::to_string(instr.alu.lop3.GetImmLut48()) + ')';
                 }
 
                 WriteLop3Instruction(instr.gpr0, op_a, op_b, op_c, lut);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "LOP3 Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::IMNMX_C:
@@ -1830,6 +1972,10 @@ private:
                                           '(' + condition + ") ? min(" + parameters + ") : max(" +
                                               parameters + ')',
                                           1, 1);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "IMNMX Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::LEA_R2:
@@ -1839,7 +1985,7 @@ private:
             case OpCode::Id::LEA_HI: {
                 std::string op_c;
 
-                switch (opcode->GetId()) {
+                switch (opcode->get().GetId()) {
                 case OpCode::Id::LEA_R2: {
                     op_a = regs.GetRegisterAsInteger(instr.gpr20);
                     op_b = regs.GetRegisterAsInteger(instr.gpr39);
@@ -1884,7 +2030,8 @@ private:
                     op_b = regs.GetRegisterAsInteger(instr.gpr8);
                     op_a = std::to_string(instr.lea.imm.entry_a);
                     op_c = std::to_string(instr.lea.imm.entry_b);
-                    LOG_CRITICAL(HW_GPU, "Unhandled LEA subinstruction: {}", opcode->GetName());
+                    LOG_CRITICAL(HW_GPU, "Unhandled LEA subinstruction: {}",
+                                 opcode->get().GetName());
                     UNREACHABLE();
                 }
                 }
@@ -1899,7 +2046,7 @@ private:
             }
             default: {
                 LOG_CRITICAL(HW_GPU, "Unhandled ArithmeticInteger instruction: {}",
-                             opcode->GetName());
+                             opcode->get().GetName());
                 UNREACHABLE();
             }
             }
@@ -1907,20 +2054,21 @@ private:
             break;
         }
         case OpCode::Type::ArithmeticHalf: {
-            if (opcode->GetId() == OpCode::Id::HADD2_C || opcode->GetId() == OpCode::Id::HADD2_R) {
+            if (opcode->get().GetId() == OpCode::Id::HADD2_C ||
+                opcode->get().GetId() == OpCode::Id::HADD2_R) {
                 ASSERT_MSG(instr.alu_half.ftz == 0, "Unimplemented");
             }
             const bool negate_a =
-                opcode->GetId() != OpCode::Id::HMUL2_R && instr.alu_half.negate_a != 0;
+                opcode->get().GetId() != OpCode::Id::HMUL2_R && instr.alu_half.negate_a != 0;
             const bool negate_b =
-                opcode->GetId() != OpCode::Id::HMUL2_C && instr.alu_half.negate_b != 0;
+                opcode->get().GetId() != OpCode::Id::HMUL2_C && instr.alu_half.negate_b != 0;
 
             const std::string op_a =
                 GetHalfFloat(regs.GetRegisterAsInteger(instr.gpr8, 0, false), instr.alu_half.type_a,
                              instr.alu_half.abs_a != 0, negate_a);
 
             std::string op_b;
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::HADD2_C:
             case OpCode::Id::HMUL2_C:
                 op_b = regs.GetUniform(instr.cbuf34.index, instr.cbuf34.offset,
@@ -1938,7 +2086,7 @@ private:
             op_b = GetHalfFloat(op_b, instr.alu_half.type_b, instr.alu_half.abs_b != 0, negate_b);
 
             const std::string result = [&]() {
-                switch (opcode->GetId()) {
+                switch (opcode->get().GetId()) {
                 case OpCode::Id::HADD2_C:
                 case OpCode::Id::HADD2_R:
                     return '(' + op_a + " + " + op_b + ')';
@@ -1946,7 +2094,8 @@ private:
                 case OpCode::Id::HMUL2_R:
                     return '(' + op_a + " * " + op_b + ')';
                 default:
-                    LOG_CRITICAL(HW_GPU, "Unhandled half float instruction: {}", opcode->GetName());
+                    LOG_CRITICAL(HW_GPU, "Unhandled half float instruction: {}",
+                                 opcode->get().GetName());
                     UNREACHABLE();
                     return std::string("0");
                 }
@@ -1957,7 +2106,7 @@ private:
             break;
         }
         case OpCode::Type::ArithmeticHalfImmediate: {
-            if (opcode->GetId() == OpCode::Id::HADD2_IMM) {
+            if (opcode->get().GetId() == OpCode::Id::HADD2_IMM) {
                 ASSERT_MSG(instr.alu_half_imm.ftz == 0, "Unimplemented");
             } else {
                 ASSERT_MSG(instr.alu_half_imm.precision == Tegra::Shader::HalfPrecision::None,
@@ -1971,7 +2120,7 @@ private:
             const std::string op_b = UnpackHalfImmediate(instr, true);
 
             const std::string result = [&]() {
-                switch (opcode->GetId()) {
+                switch (opcode->get().GetId()) {
                 case OpCode::Id::HADD2_IMM:
                     return op_a + " + " + op_b;
                 case OpCode::Id::HMUL2_IMM:
@@ -1997,7 +2146,7 @@ private:
             ASSERT_MSG(instr.ffma.tab5980_1 == 0, "FFMA tab5980_1({}) not implemented",
                        instr.ffma.tab5980_1.Value());
 
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::FFMA_CR: {
                 op_b += regs.GetUniform(instr.cbuf34.index, instr.cbuf34.offset,
                                         GLSLRegister::Type::Float);
@@ -2021,24 +2170,29 @@ private:
                 break;
             }
             default: {
-                LOG_CRITICAL(HW_GPU, "Unhandled FFMA instruction: {}", opcode->GetName());
+                LOG_CRITICAL(HW_GPU, "Unhandled FFMA instruction: {}", opcode->get().GetName());
                 UNREACHABLE();
             }
             }
 
-            regs.SetRegisterToFloat(instr.gpr0, 0, op_a + " * " + op_b + " + " + op_c, 1, 1,
-                                    instr.alu.saturate_d);
+            regs.SetRegisterToFloat(instr.gpr0, 0, "fma(" + op_a + ", " + op_b + ", " + op_c + ')',
+                                    1, 1, instr.alu.saturate_d, 0, true);
+            if (instr.generates_cc) {
+                LOG_CRITICAL(HW_GPU, "FFMA Generates an unhandled Control Code");
+                UNREACHABLE();
+            }
+
             break;
         }
         case OpCode::Type::Hfma2: {
-            if (opcode->GetId() == OpCode::Id::HFMA2_RR) {
+            if (opcode->get().GetId() == OpCode::Id::HFMA2_RR) {
                 ASSERT_MSG(instr.hfma2.rr.precision == Tegra::Shader::HalfPrecision::None,
                            "Unimplemented");
             } else {
                 ASSERT_MSG(instr.hfma2.precision == Tegra::Shader::HalfPrecision::None,
                            "Unimplemented");
             }
-            const bool saturate = opcode->GetId() == OpCode::Id::HFMA2_RR
+            const bool saturate = opcode->get().GetId() == OpCode::Id::HFMA2_RR
                                       ? instr.hfma2.rr.saturate != 0
                                       : instr.hfma2.saturate != 0;
 
@@ -2046,7 +2200,7 @@ private:
                 GetHalfFloat(regs.GetRegisterAsInteger(instr.gpr8, 0, false), instr.hfma2.type_a);
             std::string op_b, op_c;
 
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::HFMA2_CR:
                 op_b = GetHalfFloat(regs.GetUniform(instr.cbuf34.index, instr.cbuf34.offset,
                                                     GLSLRegister::Type::UnsignedInteger),
@@ -2084,7 +2238,7 @@ private:
             break;
         }
         case OpCode::Type::Conversion: {
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::I2I_R: {
                 ASSERT_MSG(!instr.conversion.selector, "Unimplemented");
 
@@ -2132,6 +2286,11 @@ private:
                 }
 
                 regs.SetRegisterToFloat(instr.gpr0, 0, op_a, 1, 1);
+
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "I2F Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::F2F_R: {
@@ -2170,6 +2329,11 @@ private:
                 }
 
                 regs.SetRegisterToFloat(instr.gpr0, 0, op_a, 1, 1, instr.alu.saturate_d);
+
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "F2F Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             case OpCode::Id::F2I_R:
@@ -2219,17 +2383,22 @@ private:
 
                 regs.SetRegisterToInteger(instr.gpr0, instr.conversion.is_output_signed, 0, op_a, 1,
                                           1, false, 0, instr.conversion.dest_size);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "F2I Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
                 break;
             }
             default: {
-                LOG_CRITICAL(HW_GPU, "Unhandled conversion instruction: {}", opcode->GetName());
+                LOG_CRITICAL(HW_GPU, "Unhandled conversion instruction: {}",
+                             opcode->get().GetName());
                 UNREACHABLE();
             }
             }
             break;
         }
         case OpCode::Type::Memory: {
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::LD_A: {
                 // Note: Shouldn't this be interp mode flat? As in no interpolation made.
                 ASSERT_MSG(instr.gpr8.Value() == Register::ZeroIndex,
@@ -2299,6 +2468,39 @@ private:
                 shader.AddLine("}");
                 break;
             }
+            case OpCode::Id::LD_L: {
+                // Add an extra scope and declare the index register inside to prevent
+                // overwriting it in case it is used as an output of the LD instruction.
+                shader.AddLine('{');
+                ++shader.scope;
+
+                std::string op = '(' + regs.GetRegisterAsInteger(instr.gpr8, 0, false) + " + " +
+                                 std::to_string(instr.smem_imm.Value()) + ')';
+
+                shader.AddLine("uint index = (" + op + " / 4);");
+
+                const std::string op_a = regs.GetLocalMemoryAsFloat("index");
+
+                if (instr.ld_l.unknown != 1) {
+                    LOG_CRITICAL(HW_GPU, "LD_L Unhandled mode: {}",
+                                 static_cast<unsigned>(instr.ld_l.unknown.Value()));
+                    UNREACHABLE();
+                }
+
+                switch (instr.ldst_sl.type.Value()) {
+                case Tegra::Shader::StoreType::Bytes32:
+                    regs.SetRegisterToFloat(instr.gpr0, 0, op_a, 1, 1);
+                    break;
+                default:
+                    LOG_CRITICAL(HW_GPU, "LD_L Unhandled type: {}",
+                                 static_cast<unsigned>(instr.ldst_sl.type.Value()));
+                    UNREACHABLE();
+                }
+
+                --shader.scope;
+                shader.AddLine('}');
+                break;
+            }
             case OpCode::Id::ST_A: {
                 ASSERT_MSG(instr.gpr8.Value() == Register::ZeroIndex,
                            "Indirect attribute loads are not supported");
@@ -2327,6 +2529,37 @@ private:
 
                 break;
             }
+            case OpCode::Id::ST_L: {
+                // Add an extra scope and declare the index register inside to prevent
+                // overwriting it in case it is used as an output of the LD instruction.
+                shader.AddLine('{');
+                ++shader.scope;
+
+                std::string op = '(' + regs.GetRegisterAsInteger(instr.gpr8, 0, false) + " + " +
+                                 std::to_string(instr.smem_imm.Value()) + ')';
+
+                shader.AddLine("uint index = (" + op + " / 4);");
+
+                if (instr.st_l.unknown != 0) {
+                    LOG_CRITICAL(HW_GPU, "ST_L Unhandled mode: {}",
+                                 static_cast<unsigned>(instr.st_l.unknown.Value()));
+                    UNREACHABLE();
+                }
+
+                switch (instr.ldst_sl.type.Value()) {
+                case Tegra::Shader::StoreType::Bytes32:
+                    regs.SetLocalMemoryAsFloat("index", regs.GetRegisterAsFloat(instr.gpr0));
+                    break;
+                default:
+                    LOG_CRITICAL(HW_GPU, "ST_L Unhandled type: {}",
+                                 static_cast<unsigned>(instr.ldst_sl.type.Value()));
+                    UNREACHABLE();
+                }
+
+                --shader.scope;
+                shader.AddLine('}');
+                break;
+            }
             case OpCode::Id::TEX: {
                 Tegra::Shader::TextureType texture_type{instr.tex.texture_type};
                 std::string coord;
@@ -2513,12 +2746,12 @@ private:
                 }
                 case 3: {
                     if (is_array) {
-                        UNIMPLEMENTED_MSG("3-coordinate arrays not fully implemented");
-                        const std::string x = regs.GetRegisterAsFloat(instr.gpr8);
-                        const std::string y = regs.GetRegisterAsFloat(instr.gpr20);
-                        coord = "vec2 coords = vec2(" + x + ", " + y + ");";
-                        texture_type = Tegra::Shader::TextureType::Texture2D;
-                        is_array = false;
+                        const std::string index = regs.GetRegisterAsInteger(instr.gpr8);
+                        const std::string x = regs.GetRegisterAsFloat(instr.gpr8.Value() + 1);
+                        const std::string y = regs.GetRegisterAsFloat(instr.gpr8.Value() + 2);
+                        const std::string z = regs.GetRegisterAsFloat(instr.gpr20);
+                        coord =
+                            "vec4 coords = vec4(" + x + ", " + y + ", " + z + ", " + index + ");";
                     } else {
                         const std::string x = regs.GetRegisterAsFloat(instr.gpr8);
                         const std::string y = regs.GetRegisterAsFloat(instr.gpr8.Value() + 1);
@@ -2548,7 +2781,11 @@ private:
                     break;
                 }
                 case Tegra::Shader::TextureProcessMode::LZ: {
-                    texture = "textureLod(" + sampler + ", coords, 0.0)";
+                    if (depth_compare && is_array) {
+                        texture = "texture(" + sampler + ", coords)";
+                    } else {
+                        texture = "textureLod(" + sampler + ", coords, 0.0)";
+                    }
                     break;
                 }
                 case Tegra::Shader::TextureProcessMode::LL: {
@@ -2809,7 +3046,7 @@ private:
                 break;
             }
             default: {
-                LOG_CRITICAL(HW_GPU, "Unhandled memory instruction: {}", opcode->GetName());
+                LOG_CRITICAL(HW_GPU, "Unhandled memory instruction: {}", opcode->get().GetName());
                 UNREACHABLE();
             }
             }
@@ -2903,7 +3140,7 @@ private:
                              instr.hsetp2.abs_a, instr.hsetp2.negate_a);
 
             const std::string op_b = [&]() {
-                switch (opcode->GetId()) {
+                switch (opcode->get().GetId()) {
                 case OpCode::Id::HSETP2_R:
                     return GetHalfFloat(regs.GetRegisterAsInteger(instr.gpr20, 0, false),
                                         instr.hsetp2.type_b, instr.hsetp2.abs_a,
@@ -2962,10 +3199,15 @@ private:
                 regs.SetRegisterToFloat(instr.gpr0, 0, value, 1, 1);
             }
 
+            if (instr.generates_cc) {
+                LOG_CRITICAL(HW_GPU, "PSET Generates an unhandled Control Code");
+                UNREACHABLE();
+            }
+
             break;
         }
         case OpCode::Type::PredicateSetPredicate: {
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::PSETP: {
                 const std::string op_a =
                     GetPredicateCondition(instr.psetp.pred12, instr.psetp.neg_pred12 != 0);
@@ -3011,7 +3253,8 @@ private:
                 break;
             }
             default: {
-                LOG_CRITICAL(HW_GPU, "Unhandled predicate instruction: {}", opcode->GetName());
+                LOG_CRITICAL(HW_GPU, "Unhandled predicate instruction: {}",
+                             opcode->get().GetName());
                 UNREACHABLE();
             }
             }
@@ -3099,7 +3342,7 @@ private:
                              instr.hset2.abs_a != 0, instr.hset2.negate_a != 0);
 
             const std::string op_b = [&]() {
-                switch (opcode->GetId()) {
+                switch (opcode->get().GetId()) {
                 case OpCode::Id::HSET2_R:
                     return GetHalfFloat(regs.GetRegisterAsInteger(instr.gpr20, 0, false),
                                         instr.hset2.type_b, instr.hset2.abs_b != 0,
@@ -3148,7 +3391,7 @@ private:
             const bool is_signed{instr.xmad.sign_a == 1};
 
             bool is_merge{};
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::XMAD_CR: {
                 is_merge = instr.xmad.merge_56;
                 op_b += regs.GetUniform(instr.cbuf34.index, instr.cbuf34.offset,
@@ -3177,7 +3420,7 @@ private:
                 break;
             }
             default: {
-                LOG_CRITICAL(HW_GPU, "Unhandled XMAD instruction: {}", opcode->GetName());
+                LOG_CRITICAL(HW_GPU, "Unhandled XMAD instruction: {}", opcode->get().GetName());
                 UNREACHABLE();
             }
             }
@@ -3226,15 +3469,25 @@ private:
             }
 
             regs.SetRegisterToInteger(instr.gpr0, is_signed, 0, sum, 1, 1);
+            if (instr.generates_cc) {
+                LOG_CRITICAL(HW_GPU, "XMAD Generates an unhandled Control Code");
+                UNREACHABLE();
+            }
             break;
         }
         default: {
-            switch (opcode->GetId()) {
+            switch (opcode->get().GetId()) {
             case OpCode::Id::EXIT: {
                 if (stage == Maxwell3D::Regs::ShaderStage::Fragment) {
                     EmitFragmentOutputsWrite();
                 }
 
+                const Tegra::Shader::ControlCode cc = instr.flow_control_code;
+                if (cc != Tegra::Shader::ControlCode::T) {
+                    LOG_CRITICAL(HW_GPU, "EXIT Control Code used: {}", static_cast<u32>(cc));
+                    UNREACHABLE();
+                }
+
                 switch (instr.flow.cond) {
                 case Tegra::Shader::FlowCondition::Always:
                     shader.AddLine("return true;");
@@ -3264,6 +3517,11 @@ private:
 
                 // Enclose "discard" in a conditional, so that GLSL compilation does not complain
                 // about unexecuted instructions that may follow this.
+                const Tegra::Shader::ControlCode cc = instr.flow_control_code;
+                if (cc != Tegra::Shader::ControlCode::T) {
+                    LOG_CRITICAL(HW_GPU, "KIL Control Code used: {}", static_cast<u32>(cc));
+                    UNREACHABLE();
+                }
                 shader.AddLine("if (true) {");
                 ++shader.scope;
                 shader.AddLine("discard;");
@@ -3321,6 +3579,11 @@ private:
             case OpCode::Id::BRA: {
                 ASSERT_MSG(instr.bra.constant_buffer == 0,
                            "BRA with constant buffers are not implemented");
+                const Tegra::Shader::ControlCode cc = instr.flow_control_code;
+                if (cc != Tegra::Shader::ControlCode::T) {
+                    LOG_CRITICAL(HW_GPU, "BRA Control Code used: {}", static_cast<u32>(cc));
+                    UNREACHABLE();
+                }
                 const u32 target = offset + instr.bra.GetBranchTarget();
                 shader.AddLine("{ jmp_to = " + std::to_string(target) + "u; break; }");
                 break;
@@ -3361,13 +3624,21 @@ private:
             }
             case OpCode::Id::SYNC: {
                 // The SYNC opcode jumps to the address previously set by the SSY opcode
-                ASSERT(instr.flow.cond == Tegra::Shader::FlowCondition::Always);
+                const Tegra::Shader::ControlCode cc = instr.flow_control_code;
+                if (cc != Tegra::Shader::ControlCode::T) {
+                    LOG_CRITICAL(HW_GPU, "SYNC Control Code used: {}", static_cast<u32>(cc));
+                    UNREACHABLE();
+                }
                 EmitPopFromFlowStack();
                 break;
             }
             case OpCode::Id::BRK: {
                 // The BRK opcode jumps to the address previously set by the PBK opcode
-                ASSERT(instr.flow.cond == Tegra::Shader::FlowCondition::Always);
+                const Tegra::Shader::ControlCode cc = instr.flow_control_code;
+                if (cc != Tegra::Shader::ControlCode::T) {
+                    LOG_CRITICAL(HW_GPU, "BRK Control Code used: {}", static_cast<u32>(cc));
+                    UNREACHABLE();
+                }
                 EmitPopFromFlowStack();
                 break;
             }
@@ -3397,6 +3668,11 @@ private:
                 regs.SetRegisterToInteger(instr.gpr0, result_signed, 1, result, 1, 1,
                                           instr.vmad.saturate == 1, 0, Register::Size::Word,
                                           instr.vmad.cc);
+                if (instr.generates_cc) {
+                    LOG_CRITICAL(HW_GPU, "VMAD Generates an unhandled Control Code");
+                    UNREACHABLE();
+                }
+
                 break;
             }
             case OpCode::Id::VSETP: {
@@ -3424,7 +3700,7 @@ private:
                 break;
             }
             default: {
-                LOG_CRITICAL(HW_GPU, "Unhandled instruction: {}", opcode->GetName());
+                LOG_CRITICAL(HW_GPU, "Unhandled instruction: {}", opcode->get().GetName());
                 UNREACHABLE();
             }
             }
@@ -3550,6 +3826,7 @@ private:
     const u32 main_offset;
     Maxwell3D::Regs::ShaderStage stage;
     const std::string& suffix;
+    u64 local_memory_size;
 
     ShaderWriter shader;
     ShaderWriter declarations;
@@ -3564,9 +3841,9 @@ std::string GetCommonDeclarations() {
                        RasterizerOpenGL::MaxConstbufferSize / sizeof(GLvec4));
 }
 
-boost::optional<ProgramResult> DecompileProgram(const ProgramCode& program_code, u32 main_offset,
-                                                Maxwell3D::Regs::ShaderStage stage,
-                                                const std::string& suffix) {
+std::optional<ProgramResult> DecompileProgram(const ProgramCode& program_code, u32 main_offset,
+                                              Maxwell3D::Regs::ShaderStage stage,
+                                              const std::string& suffix) {
     try {
         const auto subroutines =
             ControlFlowAnalyzer(program_code, main_offset, suffix).GetSubroutines();
@@ -3575,7 +3852,7 @@ boost::optional<ProgramResult> DecompileProgram(const ProgramCode& program_code,
     } catch (const DecompileFail& exception) {
         LOG_ERROR(HW_GPU, "Shader decompilation failed: {}", exception.what());
     }
-    return boost::none;
+    return {};
 }
 
 } // namespace OpenGL::GLShader::Decompiler
diff --git a/src/video_core/renderer_opengl/gl_shader_decompiler.h b/src/video_core/renderer_opengl/gl_shader_decompiler.h
index b20cc4bfa..d01a4a7ee 100644
--- a/src/video_core/renderer_opengl/gl_shader_decompiler.h
+++ b/src/video_core/renderer_opengl/gl_shader_decompiler.h
@@ -6,8 +6,8 @@
 
 #include <array>
 #include <functional>
+#include <optional>
 #include <string>
-#include <boost/optional.hpp>
 #include "common/common_types.h"
 #include "video_core/engines/maxwell_3d.h"
 #include "video_core/renderer_opengl/gl_shader_gen.h"
@@ -18,8 +18,8 @@ using Tegra::Engines::Maxwell3D;
 
 std::string GetCommonDeclarations();
 
-boost::optional<ProgramResult> DecompileProgram(const ProgramCode& program_code, u32 main_offset,
-                                                Maxwell3D::Regs::ShaderStage stage,
-                                                const std::string& suffix);
+std::optional<ProgramResult> DecompileProgram(const ProgramCode& program_code, u32 main_offset,
+                                              Maxwell3D::Regs::ShaderStage stage,
+                                              const std::string& suffix);
 
 } // namespace OpenGL::GLShader::Decompiler
diff --git a/src/video_core/renderer_opengl/gl_shader_gen.cpp b/src/video_core/renderer_opengl/gl_shader_gen.cpp
index e883ffb1d..eea090e52 100644
--- a/src/video_core/renderer_opengl/gl_shader_gen.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_gen.cpp
@@ -19,9 +19,6 @@ ProgramResult GenerateVertexShader(const ShaderSetup& setup) {
     out += Decompiler::GetCommonDeclarations();
 
     out += R"(
-out gl_PerVertex {
-    vec4 gl_Position;
-};
 
 layout (location = 0) out vec4 position;
 
@@ -40,7 +37,7 @@ layout(std140) uniform vs_config {
     ProgramResult program =
         Decompiler::DecompileProgram(setup.program.code, PROGRAM_OFFSET,
                                      Maxwell3D::Regs::ShaderStage::Vertex, "vertex")
-            .get_value_or({});
+            .value_or(ProgramResult());
 
     out += program.first;
 
@@ -48,7 +45,7 @@ layout(std140) uniform vs_config {
         ProgramResult program_b =
             Decompiler::DecompileProgram(setup.program.code_b, PROGRAM_OFFSET,
                                          Maxwell3D::Regs::ShaderStage::Vertex, "vertex_b")
-                .get_value_or({});
+                .value_or(ProgramResult());
         out += program_b.first;
     }
 
@@ -85,15 +82,15 @@ void main() {
 }
 
 ProgramResult GenerateGeometryShader(const ShaderSetup& setup) {
-    std::string out = "#version 430 core\n";
-    out += "#extension GL_ARB_separate_shader_objects : enable\n\n";
+    // Version is intentionally skipped in shader generation, it's added by the lazy compilation.
+    std::string out = "#extension GL_ARB_separate_shader_objects : enable\n\n";
     out += Decompiler::GetCommonDeclarations();
     out += "bool exec_geometry();\n";
 
     ProgramResult program =
         Decompiler::DecompileProgram(setup.program.code, PROGRAM_OFFSET,
                                      Maxwell3D::Regs::ShaderStage::Geometry, "geometry")
-            .get_value_or({});
+            .value_or(ProgramResult());
     out += R"(
 out gl_PerVertex {
     vec4 gl_Position;
@@ -127,7 +124,7 @@ ProgramResult GenerateFragmentShader(const ShaderSetup& setup) {
     ProgramResult program =
         Decompiler::DecompileProgram(setup.program.code, PROGRAM_OFFSET,
                                      Maxwell3D::Regs::ShaderStage::Fragment, "fragment")
-            .get_value_or({});
+            .value_or(ProgramResult());
     out += R"(
 layout(location = 0) out vec4 FragColor0;
 layout(location = 1) out vec4 FragColor1;
diff --git a/src/video_core/renderer_opengl/gl_shader_manager.h b/src/video_core/renderer_opengl/gl_shader_manager.h
index 36fe1f04c..2a069cdd8 100644
--- a/src/video_core/renderer_opengl/gl_shader_manager.h
+++ b/src/video_core/renderer_opengl/gl_shader_manager.h
@@ -7,6 +7,7 @@
 #include <glad/glad.h>
 
 #include "video_core/renderer_opengl/gl_resource_manager.h"
+#include "video_core/renderer_opengl/gl_state.h"
 #include "video_core/renderer_opengl/maxwell_to_gl.h"
 
 namespace OpenGL::GLShader {
diff --git a/src/video_core/renderer_opengl/gl_state.cpp b/src/video_core/renderer_opengl/gl_state.cpp
index 1fe26a2a9..98622a058 100644
--- a/src/video_core/renderer_opengl/gl_state.cpp
+++ b/src/video_core/renderer_opengl/gl_state.cpp
@@ -11,9 +11,10 @@
 namespace OpenGL {
 
 OpenGLState OpenGLState::cur_state;
-
+bool OpenGLState::s_rgb_used;
 OpenGLState::OpenGLState() {
     // These all match default OpenGL values
+    framebuffer_srgb.enabled = false;
     cull.enabled = false;
     cull.mode = GL_BACK;
     cull.front_face = GL_CCW;
@@ -22,11 +23,14 @@ OpenGLState::OpenGLState() {
     depth.test_func = GL_LESS;
     depth.write_mask = GL_TRUE;
 
-    color_mask.red_enabled = GL_TRUE;
-    color_mask.green_enabled = GL_TRUE;
-    color_mask.blue_enabled = GL_TRUE;
-    color_mask.alpha_enabled = GL_TRUE;
-
+    primitive_restart.enabled = false;
+    primitive_restart.index = 0;
+    for (auto& item : color_mask) {
+        item.red_enabled = GL_TRUE;
+        item.green_enabled = GL_TRUE;
+        item.blue_enabled = GL_TRUE;
+        item.alpha_enabled = GL_TRUE;
+    }
     stencil.test_enabled = false;
     auto reset_stencil = [](auto& config) {
         config.test_func = GL_ALWAYS;
@@ -39,19 +43,33 @@ OpenGLState::OpenGLState() {
     };
     reset_stencil(stencil.front);
     reset_stencil(stencil.back);
-
-    blend.enabled = true;
-    blend.rgb_equation = GL_FUNC_ADD;
-    blend.a_equation = GL_FUNC_ADD;
-    blend.src_rgb_func = GL_ONE;
-    blend.dst_rgb_func = GL_ZERO;
-    blend.src_a_func = GL_ONE;
-    blend.dst_a_func = GL_ZERO;
-    blend.color.red = 0.0f;
-    blend.color.green = 0.0f;
-    blend.color.blue = 0.0f;
-    blend.color.alpha = 0.0f;
-
+    for (auto& item : viewports) {
+        item.x = 0;
+        item.y = 0;
+        item.width = 0;
+        item.height = 0;
+        item.depth_range_near = 0.0f;
+        item.depth_range_far = 1.0f;
+    }
+    scissor.enabled = false;
+    scissor.x = 0;
+    scissor.y = 0;
+    scissor.width = 0;
+    scissor.height = 0;
+    for (auto& item : blend) {
+        item.enabled = true;
+        item.rgb_equation = GL_FUNC_ADD;
+        item.a_equation = GL_FUNC_ADD;
+        item.src_rgb_func = GL_ONE;
+        item.dst_rgb_func = GL_ZERO;
+        item.src_a_func = GL_ONE;
+        item.dst_a_func = GL_ZERO;
+    }
+    independant_blend.enabled = false;
+    blend_color.red = 0.0f;
+    blend_color.green = 0.0f;
+    blend_color.blue = 0.0f;
+    blend_color.alpha = 0.0f;
     logic_op.enabled = false;
     logic_op.operation = GL_COPY;
 
@@ -67,138 +85,309 @@ OpenGLState::OpenGLState() {
     draw.shader_program = 0;
     draw.program_pipeline = 0;
 
-    scissor.enabled = false;
-    scissor.x = 0;
-    scissor.y = 0;
-    scissor.width = 0;
-    scissor.height = 0;
-
-    viewport.x = 0;
-    viewport.y = 0;
-    viewport.width = 0;
-    viewport.height = 0;
-
     clip_distance = {};
 
     point.size = 1;
 }
 
-void OpenGLState::Apply() const {
+void OpenGLState::ApplyDefaultState() {
+    glDisable(GL_FRAMEBUFFER_SRGB);
+    glDisable(GL_CULL_FACE);
+    glDisable(GL_DEPTH_TEST);
+    glDisable(GL_PRIMITIVE_RESTART);
+    glDisable(GL_STENCIL_TEST);
+    glEnable(GL_BLEND);
+    glDisable(GL_COLOR_LOGIC_OP);
+    glDisable(GL_SCISSOR_TEST);
+}
+
+void OpenGLState::ApplySRgb() const {
+    // sRGB
+    if (framebuffer_srgb.enabled != cur_state.framebuffer_srgb.enabled) {
+        if (framebuffer_srgb.enabled) {
+            // Track if sRGB is used
+            s_rgb_used = true;
+            glEnable(GL_FRAMEBUFFER_SRGB);
+        } else {
+            glDisable(GL_FRAMEBUFFER_SRGB);
+        }
+    }
+}
+
+void OpenGLState::ApplyCulling() const {
     // Culling
-    if (cull.enabled != cur_state.cull.enabled) {
+    const bool cull_changed = cull.enabled != cur_state.cull.enabled;
+    if (cull_changed) {
         if (cull.enabled) {
             glEnable(GL_CULL_FACE);
         } else {
             glDisable(GL_CULL_FACE);
         }
     }
+    if (cull.enabled) {
+        if (cull_changed || cull.mode != cur_state.cull.mode) {
+            glCullFace(cull.mode);
+        }
 
-    if (cull.mode != cur_state.cull.mode) {
-        glCullFace(cull.mode);
+        if (cull_changed || cull.front_face != cur_state.cull.front_face) {
+            glFrontFace(cull.front_face);
+        }
     }
+}
 
-    if (cull.front_face != cur_state.cull.front_face) {
-        glFrontFace(cull.front_face);
+void OpenGLState::ApplyColorMask() const {
+    if (GLAD_GL_ARB_viewport_array) {
+        for (size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
+            const auto& updated = color_mask[i];
+            const auto& current = cur_state.color_mask[i];
+            if (updated.red_enabled != current.red_enabled ||
+                updated.green_enabled != current.green_enabled ||
+                updated.blue_enabled != current.blue_enabled ||
+                updated.alpha_enabled != current.alpha_enabled) {
+                glColorMaski(static_cast<GLuint>(i), updated.red_enabled, updated.green_enabled,
+                             updated.blue_enabled, updated.alpha_enabled);
+            }
+        }
+    } else {
+        const auto& updated = color_mask[0];
+        const auto& current = cur_state.color_mask[0];
+        if (updated.red_enabled != current.red_enabled ||
+            updated.green_enabled != current.green_enabled ||
+            updated.blue_enabled != current.blue_enabled ||
+            updated.alpha_enabled != current.alpha_enabled) {
+            glColorMask(updated.red_enabled, updated.green_enabled, updated.blue_enabled,
+                        updated.alpha_enabled);
+        }
     }
+}
 
+void OpenGLState::ApplyDepth() const {
     // Depth test
-    if (depth.test_enabled != cur_state.depth.test_enabled) {
+    const bool depth_test_changed = depth.test_enabled != cur_state.depth.test_enabled;
+    if (depth_test_changed) {
         if (depth.test_enabled) {
             glEnable(GL_DEPTH_TEST);
         } else {
             glDisable(GL_DEPTH_TEST);
         }
     }
-
-    if (depth.test_func != cur_state.depth.test_func) {
+    if (depth.test_enabled &&
+        (depth_test_changed || depth.test_func != cur_state.depth.test_func)) {
         glDepthFunc(depth.test_func);
     }
-
     // Depth mask
     if (depth.write_mask != cur_state.depth.write_mask) {
         glDepthMask(depth.write_mask);
     }
+}
 
-    // Color mask
-    if (color_mask.red_enabled != cur_state.color_mask.red_enabled ||
-        color_mask.green_enabled != cur_state.color_mask.green_enabled ||
-        color_mask.blue_enabled != cur_state.color_mask.blue_enabled ||
-        color_mask.alpha_enabled != cur_state.color_mask.alpha_enabled) {
-        glColorMask(color_mask.red_enabled, color_mask.green_enabled, color_mask.blue_enabled,
-                    color_mask.alpha_enabled);
+void OpenGLState::ApplyPrimitiveRestart() const {
+    const bool primitive_restart_changed =
+        primitive_restart.enabled != cur_state.primitive_restart.enabled;
+    if (primitive_restart_changed) {
+        if (primitive_restart.enabled) {
+            glEnable(GL_PRIMITIVE_RESTART);
+        } else {
+            glDisable(GL_PRIMITIVE_RESTART);
+        }
+    }
+    if (primitive_restart_changed ||
+        (primitive_restart.enabled &&
+         primitive_restart.index != cur_state.primitive_restart.index)) {
+        glPrimitiveRestartIndex(primitive_restart.index);
     }
+}
 
-    // Stencil test
-    if (stencil.test_enabled != cur_state.stencil.test_enabled) {
+void OpenGLState::ApplyStencilTest() const {
+    const bool stencil_test_changed = stencil.test_enabled != cur_state.stencil.test_enabled;
+    if (stencil_test_changed) {
         if (stencil.test_enabled) {
             glEnable(GL_STENCIL_TEST);
         } else {
             glDisable(GL_STENCIL_TEST);
         }
     }
-    auto config_stencil = [](GLenum face, const auto& config, const auto& prev_config) {
-        if (config.test_func != prev_config.test_func || config.test_ref != prev_config.test_ref ||
-            config.test_mask != prev_config.test_mask) {
-            glStencilFuncSeparate(face, config.test_func, config.test_ref, config.test_mask);
+    if (stencil.test_enabled) {
+        auto config_stencil = [stencil_test_changed](GLenum face, const auto& config,
+                                                     const auto& prev_config) {
+            if (stencil_test_changed || config.test_func != prev_config.test_func ||
+                config.test_ref != prev_config.test_ref ||
+                config.test_mask != prev_config.test_mask) {
+                glStencilFuncSeparate(face, config.test_func, config.test_ref, config.test_mask);
+            }
+            if (stencil_test_changed || config.action_depth_fail != prev_config.action_depth_fail ||
+                config.action_depth_pass != prev_config.action_depth_pass ||
+                config.action_stencil_fail != prev_config.action_stencil_fail) {
+                glStencilOpSeparate(face, config.action_stencil_fail, config.action_depth_fail,
+                                    config.action_depth_pass);
+            }
+            if (config.write_mask != prev_config.write_mask) {
+                glStencilMaskSeparate(face, config.write_mask);
+            }
+        };
+        config_stencil(GL_FRONT, stencil.front, cur_state.stencil.front);
+        config_stencil(GL_BACK, stencil.back, cur_state.stencil.back);
+    }
+}
+
+void OpenGLState::ApplyScissor() const {
+    const bool scissor_changed = scissor.enabled != cur_state.scissor.enabled;
+    if (scissor_changed) {
+        if (scissor.enabled) {
+            glEnable(GL_SCISSOR_TEST);
+        } else {
+            glDisable(GL_SCISSOR_TEST);
         }
-        if (config.action_depth_fail != prev_config.action_depth_fail ||
-            config.action_depth_pass != prev_config.action_depth_pass ||
-            config.action_stencil_fail != prev_config.action_stencil_fail) {
-            glStencilOpSeparate(face, config.action_stencil_fail, config.action_depth_fail,
-                                config.action_depth_pass);
+    }
+    if (scissor.enabled &&
+        (scissor_changed || scissor.x != cur_state.scissor.x || scissor.y != cur_state.scissor.y ||
+         scissor.width != cur_state.scissor.width || scissor.height != cur_state.scissor.height)) {
+        glScissor(scissor.x, scissor.y, scissor.width, scissor.height);
+    }
+}
+
+void OpenGLState::ApplyViewport() const {
+    if (GLAD_GL_ARB_viewport_array) {
+        for (GLuint i = 0;
+             i < static_cast<GLuint>(Tegra::Engines::Maxwell3D::Regs::NumRenderTargets); i++) {
+            const auto& current = cur_state.viewports[i];
+            const auto& updated = viewports[i];
+            if (updated.x != current.x || updated.y != current.y ||
+                updated.width != current.width || updated.height != current.height) {
+                glViewportIndexedf(i, updated.x, updated.y, updated.width, updated.height);
+            }
+            if (updated.depth_range_near != current.depth_range_near ||
+                updated.depth_range_far != current.depth_range_far) {
+                glDepthRangeIndexed(i, updated.depth_range_near, updated.depth_range_far);
+            }
         }
-        if (config.write_mask != prev_config.write_mask) {
-            glStencilMaskSeparate(face, config.write_mask);
+    } else {
+        const auto& current = cur_state.viewports[0];
+        const auto& updated = viewports[0];
+        if (updated.x != current.x || updated.y != current.y || updated.width != current.width ||
+            updated.height != current.height) {
+            glViewport(static_cast<GLint>(updated.x), static_cast<GLint>(updated.y),
+                       static_cast<GLsizei>(updated.width), static_cast<GLsizei>(updated.height));
         }
-    };
-    config_stencil(GL_FRONT, stencil.front, cur_state.stencil.front);
-    config_stencil(GL_BACK, stencil.back, cur_state.stencil.back);
+        if (updated.depth_range_near != current.depth_range_near ||
+            updated.depth_range_far != current.depth_range_far) {
+            glDepthRange(updated.depth_range_near, updated.depth_range_far);
+        }
+    }
+}
 
-    // Blending
-    if (blend.enabled != cur_state.blend.enabled) {
-        if (blend.enabled) {
-            ASSERT(!logic_op.enabled);
+void OpenGLState::ApplyGlobalBlending() const {
+    const Blend& current = cur_state.blend[0];
+    const Blend& updated = blend[0];
+    const bool blend_changed = updated.enabled != current.enabled;
+    if (blend_changed) {
+        if (updated.enabled) {
             glEnable(GL_BLEND);
         } else {
             glDisable(GL_BLEND);
         }
     }
+    if (!updated.enabled) {
+        return;
+    }
+    if (updated.separate_alpha) {
+        if (blend_changed || updated.src_rgb_func != current.src_rgb_func ||
+            updated.dst_rgb_func != current.dst_rgb_func ||
+            updated.src_a_func != current.src_a_func || updated.dst_a_func != current.dst_a_func) {
+            glBlendFuncSeparate(updated.src_rgb_func, updated.dst_rgb_func, updated.src_a_func,
+                                updated.dst_a_func);
+        }
+
+        if (blend_changed || updated.rgb_equation != current.rgb_equation ||
+            updated.a_equation != current.a_equation) {
+            glBlendEquationSeparate(updated.rgb_equation, updated.a_equation);
+        }
+    } else {
+        if (blend_changed || updated.src_rgb_func != current.src_rgb_func ||
+            updated.dst_rgb_func != current.dst_rgb_func) {
+            glBlendFunc(updated.src_rgb_func, updated.dst_rgb_func);
+        }
+
+        if (blend_changed || updated.rgb_equation != current.rgb_equation) {
+            glBlendEquation(updated.rgb_equation);
+        }
+    }
+}
 
-    if (blend.color.red != cur_state.blend.color.red ||
-        blend.color.green != cur_state.blend.color.green ||
-        blend.color.blue != cur_state.blend.color.blue ||
-        blend.color.alpha != cur_state.blend.color.alpha) {
-        glBlendColor(blend.color.red, blend.color.green, blend.color.blue, blend.color.alpha);
+void OpenGLState::ApplyTargetBlending(std::size_t target, bool force) const {
+    const Blend& updated = blend[target];
+    const Blend& current = cur_state.blend[target];
+    const bool blend_changed = updated.enabled != current.enabled || force;
+    if (blend_changed) {
+        if (updated.enabled) {
+            glEnablei(GL_BLEND, static_cast<GLuint>(target));
+        } else {
+            glDisablei(GL_BLEND, static_cast<GLuint>(target));
+        }
+    }
+    if (!updated.enabled) {
+        return;
     }
+    if (updated.separate_alpha) {
+        if (blend_changed || updated.src_rgb_func != current.src_rgb_func ||
+            updated.dst_rgb_func != current.dst_rgb_func ||
+            updated.src_a_func != current.src_a_func || updated.dst_a_func != current.dst_a_func) {
+            glBlendFuncSeparateiARB(static_cast<GLuint>(target), updated.src_rgb_func,
+                                    updated.dst_rgb_func, updated.src_a_func, updated.dst_a_func);
+        }
+
+        if (blend_changed || updated.rgb_equation != current.rgb_equation ||
+            updated.a_equation != current.a_equation) {
+            glBlendEquationSeparateiARB(static_cast<GLuint>(target), updated.rgb_equation,
+                                        updated.a_equation);
+        }
+    } else {
+        if (blend_changed || updated.src_rgb_func != current.src_rgb_func ||
+            updated.dst_rgb_func != current.dst_rgb_func) {
+            glBlendFunciARB(static_cast<GLuint>(target), updated.src_rgb_func,
+                            updated.dst_rgb_func);
+        }
 
-    if (blend.src_rgb_func != cur_state.blend.src_rgb_func ||
-        blend.dst_rgb_func != cur_state.blend.dst_rgb_func ||
-        blend.src_a_func != cur_state.blend.src_a_func ||
-        blend.dst_a_func != cur_state.blend.dst_a_func) {
-        glBlendFuncSeparate(blend.src_rgb_func, blend.dst_rgb_func, blend.src_a_func,
-                            blend.dst_a_func);
+        if (blend_changed || updated.rgb_equation != current.rgb_equation) {
+            glBlendEquationiARB(static_cast<GLuint>(target), updated.rgb_equation);
+        }
     }
+}
 
-    if (blend.rgb_equation != cur_state.blend.rgb_equation ||
-        blend.a_equation != cur_state.blend.a_equation) {
-        glBlendEquationSeparate(blend.rgb_equation, blend.a_equation);
+void OpenGLState::ApplyBlending() const {
+    if (independant_blend.enabled) {
+        for (size_t i = 0; i < Tegra::Engines::Maxwell3D::Regs::NumRenderTargets; i++) {
+            ApplyTargetBlending(i,
+                                independant_blend.enabled != cur_state.independant_blend.enabled);
+        }
+    } else {
+        ApplyGlobalBlending();
+    }
+    if (blend_color.red != cur_state.blend_color.red ||
+        blend_color.green != cur_state.blend_color.green ||
+        blend_color.blue != cur_state.blend_color.blue ||
+        blend_color.alpha != cur_state.blend_color.alpha) {
+        glBlendColor(blend_color.red, blend_color.green, blend_color.blue, blend_color.alpha);
     }
+}
 
-    // Logic Operation
-    if (logic_op.enabled != cur_state.logic_op.enabled) {
+void OpenGLState::ApplyLogicOp() const {
+    const bool logic_op_changed = logic_op.enabled != cur_state.logic_op.enabled;
+    if (logic_op_changed) {
         if (logic_op.enabled) {
-            ASSERT(!blend.enabled);
             glEnable(GL_COLOR_LOGIC_OP);
         } else {
             glDisable(GL_COLOR_LOGIC_OP);
         }
     }
 
-    if (logic_op.operation != cur_state.logic_op.operation) {
+    if (logic_op.enabled &&
+        (logic_op_changed || logic_op.operation != cur_state.logic_op.operation)) {
         glLogicOp(logic_op.operation);
     }
+}
 
-    // Textures
+void OpenGLState::ApplyTextures() const {
     for (std::size_t i = 0; i < std::size(texture_units); ++i) {
         const auto& texture_unit = texture_units[i];
         const auto& cur_state_texture_unit = cur_state.texture_units[i];
@@ -217,28 +406,29 @@ void OpenGLState::Apply() const {
             glTexParameteriv(texture_unit.target, GL_TEXTURE_SWIZZLE_RGBA, mask.data());
         }
     }
+}
 
-    // Samplers
-    {
-        bool has_delta{};
-        std::size_t first{}, last{};
-        std::array<GLuint, Tegra::Engines::Maxwell3D::Regs::NumTextureSamplers> samplers;
-        for (std::size_t i = 0; i < std::size(samplers); ++i) {
-            samplers[i] = texture_units[i].sampler;
-            if (samplers[i] != cur_state.texture_units[i].sampler) {
-                if (!has_delta) {
-                    first = i;
-                    has_delta = true;
-                }
-                last = i;
+void OpenGLState::ApplySamplers() const {
+    bool has_delta{};
+    std::size_t first{}, last{};
+    std::array<GLuint, Tegra::Engines::Maxwell3D::Regs::NumTextureSamplers> samplers;
+    for (std::size_t i = 0; i < std::size(samplers); ++i) {
+        samplers[i] = texture_units[i].sampler;
+        if (samplers[i] != cur_state.texture_units[i].sampler) {
+            if (!has_delta) {
+                first = i;
+                has_delta = true;
             }
+            last = i;
         }
-        if (has_delta) {
-            glBindSamplers(static_cast<GLuint>(first), static_cast<GLsizei>(last - first + 1),
-                           samplers.data());
-        }
     }
+    if (has_delta) {
+        glBindSamplers(static_cast<GLuint>(first), static_cast<GLsizei>(last - first + 1),
+                       samplers.data());
+    }
+}
 
+void OpenGLState::ApplyFramebufferState() const {
     // Framebuffer
     if (draw.read_framebuffer != cur_state.draw.read_framebuffer) {
         glBindFramebuffer(GL_READ_FRAMEBUFFER, draw.read_framebuffer);
@@ -246,7 +436,9 @@ void OpenGLState::Apply() const {
     if (draw.draw_framebuffer != cur_state.draw.draw_framebuffer) {
         glBindFramebuffer(GL_DRAW_FRAMEBUFFER, draw.draw_framebuffer);
     }
+}
 
+void OpenGLState::ApplyVertexBufferState() const {
     // Vertex array
     if (draw.vertex_array != cur_state.draw.vertex_array) {
         glBindVertexArray(draw.vertex_array);
@@ -256,7 +448,11 @@ void OpenGLState::Apply() const {
     if (draw.vertex_buffer != cur_state.draw.vertex_buffer) {
         glBindBuffer(GL_ARRAY_BUFFER, draw.vertex_buffer);
     }
+}
 
+void OpenGLState::Apply() const {
+    ApplyFramebufferState();
+    ApplyVertexBufferState();
     // Uniform buffer
     if (draw.uniform_buffer != cur_state.draw.uniform_buffer) {
         glBindBuffer(GL_UNIFORM_BUFFER, draw.uniform_buffer);
@@ -271,27 +467,6 @@ void OpenGLState::Apply() const {
     if (draw.program_pipeline != cur_state.draw.program_pipeline) {
         glBindProgramPipeline(draw.program_pipeline);
     }
-
-    // Scissor test
-    if (scissor.enabled != cur_state.scissor.enabled) {
-        if (scissor.enabled) {
-            glEnable(GL_SCISSOR_TEST);
-        } else {
-            glDisable(GL_SCISSOR_TEST);
-        }
-    }
-
-    if (scissor.x != cur_state.scissor.x || scissor.y != cur_state.scissor.y ||
-        scissor.width != cur_state.scissor.width || scissor.height != cur_state.scissor.height) {
-        glScissor(scissor.x, scissor.y, scissor.width, scissor.height);
-    }
-
-    if (viewport.x != cur_state.viewport.x || viewport.y != cur_state.viewport.y ||
-        viewport.width != cur_state.viewport.width ||
-        viewport.height != cur_state.viewport.height) {
-        glViewport(viewport.x, viewport.y, viewport.width, viewport.height);
-    }
-
     // Clip distance
     for (std::size_t i = 0; i < clip_distance.size(); ++i) {
         if (clip_distance[i] != cur_state.clip_distance[i]) {
@@ -302,12 +477,22 @@ void OpenGLState::Apply() const {
             }
         }
     }
-
     // Point
     if (point.size != cur_state.point.size) {
         glPointSize(point.size);
     }
-
+    ApplyColorMask();
+    ApplyViewport();
+    ApplyScissor();
+    ApplyStencilTest();
+    ApplySRgb();
+    ApplyCulling();
+    ApplyDepth();
+    ApplyPrimitiveRestart();
+    ApplyBlending();
+    ApplyLogicOp();
+    ApplyTextures();
+    ApplySamplers();
     cur_state = *this;
 }
 
diff --git a/src/video_core/renderer_opengl/gl_state.h b/src/video_core/renderer_opengl/gl_state.h
index dc21a2ee3..e5d1baae6 100644
--- a/src/video_core/renderer_opengl/gl_state.h
+++ b/src/video_core/renderer_opengl/gl_state.h
@@ -36,6 +36,10 @@ constexpr TextureUnit ProcTexDiffLUT{9};
 class OpenGLState {
 public:
     struct {
+        bool enabled; // GL_FRAMEBUFFER_SRGB
+    } framebuffer_srgb;
+
+    struct {
         bool enabled;      // GL_CULL_FACE
         GLenum mode;       // GL_CULL_FACE_MODE
         GLenum front_face; // GL_FRONT_FACE
@@ -48,12 +52,18 @@ public:
     } depth;
 
     struct {
+        bool enabled;
+        GLuint index;
+    } primitive_restart; // GL_PRIMITIVE_RESTART
+
+    struct ColorMask {
         GLboolean red_enabled;
         GLboolean green_enabled;
         GLboolean blue_enabled;
         GLboolean alpha_enabled;
-    } color_mask; // GL_COLOR_WRITEMASK
-
+    };
+    std::array<ColorMask, Tegra::Engines::Maxwell3D::Regs::NumRenderTargets>
+        color_mask; // GL_COLOR_WRITEMASK
     struct {
         bool test_enabled; // GL_STENCIL_TEST
         struct {
@@ -67,22 +77,28 @@ public:
         } front, back;
     } stencil;
 
-    struct {
+    struct Blend {
         bool enabled;        // GL_BLEND
+        bool separate_alpha; // Independent blend enabled
         GLenum rgb_equation; // GL_BLEND_EQUATION_RGB
         GLenum a_equation;   // GL_BLEND_EQUATION_ALPHA
         GLenum src_rgb_func; // GL_BLEND_SRC_RGB
         GLenum dst_rgb_func; // GL_BLEND_DST_RGB
         GLenum src_a_func;   // GL_BLEND_SRC_ALPHA
         GLenum dst_a_func;   // GL_BLEND_DST_ALPHA
+    };
+    std::array<Blend, Tegra::Engines::Maxwell3D::Regs::NumRenderTargets> blend;
 
-        struct {
-            GLclampf red;
-            GLclampf green;
-            GLclampf blue;
-            GLclampf alpha;
-        } color; // GL_BLEND_COLOR
-    } blend;
+    struct {
+        bool enabled;
+    } independant_blend;
+
+    struct {
+        GLclampf red;
+        GLclampf green;
+        GLclampf blue;
+        GLclampf alpha;
+    } blend_color; // GL_BLEND_COLOR
 
     struct {
         bool enabled; // GL_LOGIC_OP_MODE
@@ -127,6 +143,16 @@ public:
         GLuint program_pipeline; // GL_PROGRAM_PIPELINE_BINDING
     } draw;
 
+    struct viewport {
+        GLfloat x;
+        GLfloat y;
+        GLfloat width;
+        GLfloat height;
+        GLfloat depth_range_near; // GL_DEPTH_RANGE
+        GLfloat depth_range_far;  // GL_DEPTH_RANGE
+    };
+    std::array<viewport, Tegra::Engines::Maxwell3D::Regs::NumRenderTargets> viewports;
+
     struct {
         bool enabled; // GL_SCISSOR_TEST
         GLint x;
@@ -136,13 +162,6 @@ public:
     } scissor;
 
     struct {
-        GLint x;
-        GLint y;
-        GLsizei width;
-        GLsizei height;
-    } viewport;
-
-    struct {
         float size; // GL_POINT_SIZE
     } point;
 
@@ -154,10 +173,20 @@ public:
     static OpenGLState GetCurState() {
         return cur_state;
     }
-
+    static bool GetsRGBUsed() {
+        return s_rgb_used;
+    }
+    static void ClearsRGBUsed() {
+        s_rgb_used = false;
+    }
     /// Apply this state as the current OpenGL state
     void Apply() const;
-
+    /// Apply only the state afecting the framebuffer
+    void ApplyFramebufferState() const;
+    /// Apply only the state afecting the vertex buffer
+    void ApplyVertexBufferState() const;
+    /// Set the initial OpenGL state
+    static void ApplyDefaultState();
     /// Resets any references to the given resource
     OpenGLState& UnbindTexture(GLuint handle);
     OpenGLState& ResetSampler(GLuint handle);
@@ -169,6 +198,23 @@ public:
 
 private:
     static OpenGLState cur_state;
+    // Workaround for sRGB problems caused by
+    // QT not supporting srgb output
+    static bool s_rgb_used;
+    void ApplySRgb() const;
+    void ApplyCulling() const;
+    void ApplyColorMask() const;
+    void ApplyDepth() const;
+    void ApplyPrimitiveRestart() const;
+    void ApplyStencilTest() const;
+    void ApplyViewport() const;
+    void ApplyTargetBlending(std::size_t target, bool force) const;
+    void ApplyGlobalBlending() const;
+    void ApplyBlending() const;
+    void ApplyLogicOp() const;
+    void ApplyTextures() const;
+    void ApplySamplers() const;
+    void ApplyScissor() const;
 };
 
 } // namespace OpenGL
diff --git a/src/video_core/renderer_opengl/gl_stream_buffer.cpp b/src/video_core/renderer_opengl/gl_stream_buffer.cpp
index e409228cc..b97b895a4 100644
--- a/src/video_core/renderer_opengl/gl_stream_buffer.cpp
+++ b/src/video_core/renderer_opengl/gl_stream_buffer.cpp
@@ -6,9 +6,13 @@
 #include <vector>
 #include "common/alignment.h"
 #include "common/assert.h"
+#include "common/microprofile.h"
 #include "video_core/renderer_opengl/gl_state.h"
 #include "video_core/renderer_opengl/gl_stream_buffer.h"
 
+MICROPROFILE_DEFINE(OpenGL_StreamBuffer, "OpenGL", "Stream Buffer Orphaning",
+                    MP_RGB(128, 128, 192));
+
 namespace OpenGL {
 
 OGLStreamBuffer::OGLStreamBuffer(GLenum target, GLsizeiptr size, bool prefer_coherent)
@@ -75,6 +79,7 @@ std::tuple<u8*, GLintptr, bool> OGLStreamBuffer::Map(GLsizeiptr size, GLintptr a
     }
 
     if (invalidate || !persistent) {
+        MICROPROFILE_SCOPE(OpenGL_StreamBuffer);
         GLbitfield flags = GL_MAP_WRITE_BIT | (persistent ? GL_MAP_PERSISTENT_BIT : 0) |
                            (coherent ? GL_MAP_COHERENT_BIT : GL_MAP_FLUSH_EXPLICIT_BIT) |
                            (invalidate ? GL_MAP_INVALIDATE_BUFFER_BIT : GL_MAP_UNSYNCHRONIZED_BIT);
diff --git a/src/video_core/renderer_opengl/maxwell_to_gl.h b/src/video_core/renderer_opengl/maxwell_to_gl.h
index 0f6dcab2b..3ce2cc6d2 100644
--- a/src/video_core/renderer_opengl/maxwell_to_gl.h
+++ b/src/video_core/renderer_opengl/maxwell_to_gl.h
@@ -135,17 +135,32 @@ inline GLenum PrimitiveTopology(Maxwell::PrimitiveTopology topology) {
     return {};
 }
 
-inline GLenum TextureFilterMode(Tegra::Texture::TextureFilter filter_mode) {
+inline GLenum TextureFilterMode(Tegra::Texture::TextureFilter filter_mode,
+                                Tegra::Texture::TextureMipmapFilter mip_filter_mode) {
     switch (filter_mode) {
-    case Tegra::Texture::TextureFilter::Linear:
-        return GL_LINEAR;
-    case Tegra::Texture::TextureFilter::Nearest:
-        return GL_NEAREST;
+    case Tegra::Texture::TextureFilter::Linear: {
+        switch (mip_filter_mode) {
+        case Tegra::Texture::TextureMipmapFilter::None:
+            return GL_LINEAR;
+        case Tegra::Texture::TextureMipmapFilter::Nearest:
+            return GL_NEAREST_MIPMAP_LINEAR;
+        case Tegra::Texture::TextureMipmapFilter::Linear:
+            return GL_LINEAR_MIPMAP_LINEAR;
+        }
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented texture filter mode={}",
-                 static_cast<u32>(filter_mode));
-    UNREACHABLE();
-    return {};
+    case Tegra::Texture::TextureFilter::Nearest: {
+        switch (mip_filter_mode) {
+        case Tegra::Texture::TextureMipmapFilter::None:
+            return GL_NEAREST;
+        case Tegra::Texture::TextureMipmapFilter::Nearest:
+            return GL_NEAREST_MIPMAP_NEAREST;
+        case Tegra::Texture::TextureMipmapFilter::Linear:
+            return GL_LINEAR_MIPMAP_NEAREST;
+        }
+    }
+    }
+    LOG_ERROR(Render_OpenGL, "Unimplemented texture filter mode={}", static_cast<u32>(filter_mode));
+    return GL_LINEAR;
 }
 
 inline GLenum WrapMode(Tegra::Texture::WrapMode wrap_mode) {
@@ -166,9 +181,8 @@ inline GLenum WrapMode(Tegra::Texture::WrapMode wrap_mode) {
     case Tegra::Texture::WrapMode::MirrorOnceClampToEdge:
         return GL_MIRROR_CLAMP_TO_EDGE;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented texture wrap mode={}", static_cast<u32>(wrap_mode));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented texture wrap mode={}", static_cast<u32>(wrap_mode));
+    return GL_REPEAT;
 }
 
 inline GLenum DepthCompareFunc(Tegra::Texture::DepthCompareFunc func) {
@@ -190,10 +204,9 @@ inline GLenum DepthCompareFunc(Tegra::Texture::DepthCompareFunc func) {
     case Tegra::Texture::DepthCompareFunc::Always:
         return GL_ALWAYS;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented texture depth compare function ={}",
-                 static_cast<u32>(func));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented texture depth compare function ={}",
+              static_cast<u32>(func));
+    return GL_GREATER;
 }
 
 inline GLenum BlendEquation(Maxwell::Blend::Equation equation) {
@@ -209,9 +222,8 @@ inline GLenum BlendEquation(Maxwell::Blend::Equation equation) {
     case Maxwell::Blend::Equation::Max:
         return GL_MAX;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented blend equation={}", static_cast<u32>(equation));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented blend equation={}", static_cast<u32>(equation));
+    return GL_FUNC_ADD;
 }
 
 inline GLenum BlendFunc(Maxwell::Blend::Factor factor) {
@@ -274,9 +286,8 @@ inline GLenum BlendFunc(Maxwell::Blend::Factor factor) {
     case Maxwell::Blend::Factor::OneMinusConstantAlphaGL:
         return GL_ONE_MINUS_CONSTANT_ALPHA;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented blend factor={}", static_cast<u32>(factor));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented blend factor={}", static_cast<u32>(factor));
+    return GL_ZERO;
 }
 
 inline GLenum SwizzleSource(Tegra::Texture::SwizzleSource source) {
@@ -295,9 +306,8 @@ inline GLenum SwizzleSource(Tegra::Texture::SwizzleSource source) {
     case Tegra::Texture::SwizzleSource::OneFloat:
         return GL_ONE;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented swizzle source={}", static_cast<u32>(source));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented swizzle source={}", static_cast<u32>(source));
+    return GL_ZERO;
 }
 
 inline GLenum ComparisonOp(Maxwell::ComparisonOp comparison) {
@@ -327,33 +337,39 @@ inline GLenum ComparisonOp(Maxwell::ComparisonOp comparison) {
     case Maxwell::ComparisonOp::AlwaysOld:
         return GL_ALWAYS;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented comparison op={}", static_cast<u32>(comparison));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented comparison op={}", static_cast<u32>(comparison));
+    return GL_ALWAYS;
 }
 
 inline GLenum StencilOp(Maxwell::StencilOp stencil) {
     switch (stencil) {
     case Maxwell::StencilOp::Keep:
+    case Maxwell::StencilOp::KeepOGL:
         return GL_KEEP;
     case Maxwell::StencilOp::Zero:
+    case Maxwell::StencilOp::ZeroOGL:
         return GL_ZERO;
     case Maxwell::StencilOp::Replace:
+    case Maxwell::StencilOp::ReplaceOGL:
         return GL_REPLACE;
     case Maxwell::StencilOp::Incr:
+    case Maxwell::StencilOp::IncrOGL:
         return GL_INCR;
     case Maxwell::StencilOp::Decr:
+    case Maxwell::StencilOp::DecrOGL:
         return GL_DECR;
     case Maxwell::StencilOp::Invert:
+    case Maxwell::StencilOp::InvertOGL:
         return GL_INVERT;
     case Maxwell::StencilOp::IncrWrap:
+    case Maxwell::StencilOp::IncrWrapOGL:
         return GL_INCR_WRAP;
     case Maxwell::StencilOp::DecrWrap:
+    case Maxwell::StencilOp::DecrWrapOGL:
         return GL_DECR_WRAP;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented stencil op={}", static_cast<u32>(stencil));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented stencil op={}", static_cast<u32>(stencil));
+    return GL_KEEP;
 }
 
 inline GLenum FrontFace(Maxwell::Cull::FrontFace front_face) {
@@ -363,9 +379,8 @@ inline GLenum FrontFace(Maxwell::Cull::FrontFace front_face) {
     case Maxwell::Cull::FrontFace::CounterClockWise:
         return GL_CCW;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented front face cull={}", static_cast<u32>(front_face));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented front face cull={}", static_cast<u32>(front_face));
+    return GL_CCW;
 }
 
 inline GLenum CullFace(Maxwell::Cull::CullFace cull_face) {
@@ -377,9 +392,8 @@ inline GLenum CullFace(Maxwell::Cull::CullFace cull_face) {
     case Maxwell::Cull::CullFace::FrontAndBack:
         return GL_FRONT_AND_BACK;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented cull face={}", static_cast<u32>(cull_face));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented cull face={}", static_cast<u32>(cull_face));
+    return GL_BACK;
 }
 
 inline GLenum LogicOp(Maxwell::LogicOperation operation) {
@@ -417,9 +431,8 @@ inline GLenum LogicOp(Maxwell::LogicOperation operation) {
     case Maxwell::LogicOperation::Set:
         return GL_SET;
     }
-    LOG_CRITICAL(Render_OpenGL, "Unimplemented logic operation={}", static_cast<u32>(operation));
-    UNREACHABLE();
-    return {};
+    LOG_ERROR(Render_OpenGL, "Unimplemented logic operation={}", static_cast<u32>(operation));
+    return GL_COPY;
 }
 
 } // namespace MaxwellToGL
diff --git a/src/video_core/renderer_opengl/renderer_opengl.cpp b/src/video_core/renderer_opengl/renderer_opengl.cpp
index 96d916b07..ea38da932 100644
--- a/src/video_core/renderer_opengl/renderer_opengl.cpp
+++ b/src/video_core/renderer_opengl/renderer_opengl.cpp
@@ -115,7 +115,8 @@ RendererOpenGL::RendererOpenGL(Core::Frontend::EmuWindow& window)
 RendererOpenGL::~RendererOpenGL() = default;
 
 /// Swap buffers (render frame)
-void RendererOpenGL::SwapBuffers(boost::optional<const Tegra::FramebufferConfig&> framebuffer) {
+void RendererOpenGL::SwapBuffers(
+    std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) {
     ScopeAcquireGLContext acquire_context{render_window};
 
     Core::System::GetInstance().GetPerfStats().EndSystemFrame();
@@ -124,11 +125,11 @@ void RendererOpenGL::SwapBuffers(boost::optional<const Tegra::FramebufferConfig&
     OpenGLState prev_state = OpenGLState::GetCurState();
     state.Apply();
 
-    if (framebuffer != boost::none) {
+    if (framebuffer) {
         // If framebuffer is provided, reload it from memory to a texture
-        if (screen_info.texture.width != (GLsizei)framebuffer->width ||
-            screen_info.texture.height != (GLsizei)framebuffer->height ||
-            screen_info.texture.pixel_format != framebuffer->pixel_format) {
+        if (screen_info.texture.width != (GLsizei)framebuffer->get().width ||
+            screen_info.texture.height != (GLsizei)framebuffer->get().height ||
+            screen_info.texture.pixel_format != framebuffer->get().pixel_format) {
             // Reallocate texture if the framebuffer size has changed.
             // This is expected to not happen very often and hence should not be a
             // performance problem.
@@ -283,7 +284,8 @@ void RendererOpenGL::CreateRasterizer() {
     if (rasterizer) {
         return;
     }
-
+    // Initialize sRGB Usage
+    OpenGLState::ClearsRGBUsed();
     rasterizer = std::make_unique<RasterizerOpenGL>(render_window, screen_info);
 }
 
@@ -356,13 +358,20 @@ void RendererOpenGL::DrawScreenTriangles(const ScreenInfo& screen_info, float x,
 
     state.texture_units[0].texture = screen_info.display_texture;
     state.texture_units[0].swizzle = {GL_RED, GL_GREEN, GL_BLUE, GL_ALPHA};
+    // Workaround brigthness problems in SMO by enabling sRGB in the final output
+    // if it has been used in the frame
+    // Needed because of this bug in QT
+    // QTBUG-50987
+    state.framebuffer_srgb.enabled = OpenGLState::GetsRGBUsed();
     state.Apply();
-
     glBufferSubData(GL_ARRAY_BUFFER, 0, sizeof(vertices), vertices.data());
     glDrawArrays(GL_TRIANGLE_STRIP, 0, 4);
-
+    // restore default state
+    state.framebuffer_srgb.enabled = false;
     state.texture_units[0].texture = 0;
     state.Apply();
+    // Clear sRGB state for the next frame
+    OpenGLState::ClearsRGBUsed();
 }
 
 /**
diff --git a/src/video_core/renderer_opengl/renderer_opengl.h b/src/video_core/renderer_opengl/renderer_opengl.h
index 961467a62..c0868c0e4 100644
--- a/src/video_core/renderer_opengl/renderer_opengl.h
+++ b/src/video_core/renderer_opengl/renderer_opengl.h
@@ -51,7 +51,8 @@ public:
     ~RendererOpenGL() override;
 
     /// Swap buffers (render frame)
-    void SwapBuffers(boost::optional<const Tegra::FramebufferConfig&> framebuffer) override;
+    void SwapBuffers(
+        std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) override;
 
     /// Initialize the renderer
     bool Init() override;
diff --git a/src/video_core/renderer_opengl/utils.cpp b/src/video_core/renderer_opengl/utils.cpp
new file mode 100644
index 000000000..d84634cb3
--- /dev/null
+++ b/src/video_core/renderer_opengl/utils.cpp
@@ -0,0 +1,38 @@
+// Copyright 2014 Citra Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <string>
+#include <fmt/format.h>
+#include <glad/glad.h>
+#include "common/common_types.h"
+#include "video_core/renderer_opengl/utils.h"
+
+namespace OpenGL {
+
+void LabelGLObject(GLenum identifier, GLuint handle, VAddr addr, std::string extra_info) {
+    if (!GLAD_GL_KHR_debug) {
+        return; // We don't need to throw an error as this is just for debugging
+    }
+    const std::string nice_addr = fmt::format("0x{:016x}", addr);
+    std::string object_label;
+
+    if (extra_info.empty()) {
+        switch (identifier) {
+        case GL_TEXTURE:
+            object_label = "Texture@" + nice_addr;
+            break;
+        case GL_PROGRAM:
+            object_label = "Shader@" + nice_addr;
+            break;
+        default:
+            object_label = fmt::format("Object(0x{:x})@{}", identifier, nice_addr);
+            break;
+        }
+    } else {
+        object_label = extra_info + '@' + nice_addr;
+    }
+    glObjectLabel(identifier, handle, -1, static_cast<const GLchar*>(object_label.c_str()));
+}
+
+} // namespace OpenGL
\ No newline at end of file
diff --git a/src/video_core/renderer_opengl/utils.h b/src/video_core/renderer_opengl/utils.h
new file mode 100644
index 000000000..1fcb6fc11
--- /dev/null
+++ b/src/video_core/renderer_opengl/utils.h
@@ -0,0 +1,15 @@
+// Copyright 2014 Citra Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <string>
+#include <glad/glad.h>
+#include "common/common_types.h"
+
+namespace OpenGL {
+
+void LabelGLObject(GLenum identifier, GLuint handle, VAddr addr, std::string extra_info = "");
+
+} // namespace OpenGL
\ No newline at end of file
diff --git a/src/video_core/surface.cpp b/src/video_core/surface.cpp
new file mode 100644
index 000000000..9582dd2ca
--- /dev/null
+++ b/src/video_core/surface.cpp
@@ -0,0 +1,490 @@
+// Copyright 2014 Citra Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "common/common_types.h"
+#include "common/math_util.h"
+#include "video_core/surface.h"
+
+namespace VideoCore::Surface {
+
+SurfaceTarget SurfaceTargetFromTextureType(Tegra::Texture::TextureType texture_type) {
+    switch (texture_type) {
+    case Tegra::Texture::TextureType::Texture1D:
+        return SurfaceTarget::Texture1D;
+    case Tegra::Texture::TextureType::Texture2D:
+    case Tegra::Texture::TextureType::Texture2DNoMipmap:
+        return SurfaceTarget::Texture2D;
+    case Tegra::Texture::TextureType::Texture3D:
+        return SurfaceTarget::Texture3D;
+    case Tegra::Texture::TextureType::TextureCubemap:
+        return SurfaceTarget::TextureCubemap;
+    case Tegra::Texture::TextureType::TextureCubeArray:
+        return SurfaceTarget::TextureCubeArray;
+    case Tegra::Texture::TextureType::Texture1DArray:
+        return SurfaceTarget::Texture1DArray;
+    case Tegra::Texture::TextureType::Texture2DArray:
+        return SurfaceTarget::Texture2DArray;
+    default:
+        LOG_CRITICAL(HW_GPU, "Unimplemented texture_type={}", static_cast<u32>(texture_type));
+        UNREACHABLE();
+        return SurfaceTarget::Texture2D;
+    }
+}
+
+bool SurfaceTargetIsLayered(SurfaceTarget target) {
+    switch (target) {
+    case SurfaceTarget::Texture1D:
+    case SurfaceTarget::Texture2D:
+    case SurfaceTarget::Texture3D:
+        return false;
+    case SurfaceTarget::Texture1DArray:
+    case SurfaceTarget::Texture2DArray:
+    case SurfaceTarget::TextureCubemap:
+    case SurfaceTarget::TextureCubeArray:
+        return true;
+    default:
+        LOG_CRITICAL(HW_GPU, "Unimplemented surface_target={}", static_cast<u32>(target));
+        UNREACHABLE();
+        return false;
+    }
+}
+
+PixelFormat PixelFormatFromDepthFormat(Tegra::DepthFormat format) {
+    switch (format) {
+    case Tegra::DepthFormat::S8_Z24_UNORM:
+        return PixelFormat::S8Z24;
+    case Tegra::DepthFormat::Z24_S8_UNORM:
+        return PixelFormat::Z24S8;
+    case Tegra::DepthFormat::Z32_FLOAT:
+        return PixelFormat::Z32F;
+    case Tegra::DepthFormat::Z16_UNORM:
+        return PixelFormat::Z16;
+    case Tegra::DepthFormat::Z32_S8_X24_FLOAT:
+        return PixelFormat::Z32FS8;
+    default:
+        LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));
+        UNREACHABLE();
+    }
+}
+
+PixelFormat PixelFormatFromRenderTargetFormat(Tegra::RenderTargetFormat format) {
+    switch (format) {
+        // TODO (Hexagon12): Converting SRGBA to RGBA is a hack and doesn't completely correct the
+        // gamma.
+    case Tegra::RenderTargetFormat::RGBA8_SRGB:
+        return PixelFormat::RGBA8_SRGB;
+    case Tegra::RenderTargetFormat::RGBA8_UNORM:
+        return PixelFormat::ABGR8U;
+    case Tegra::RenderTargetFormat::RGBA8_SNORM:
+        return PixelFormat::ABGR8S;
+    case Tegra::RenderTargetFormat::RGBA8_UINT:
+        return PixelFormat::ABGR8UI;
+    case Tegra::RenderTargetFormat::BGRA8_SRGB:
+        return PixelFormat::BGRA8_SRGB;
+    case Tegra::RenderTargetFormat::BGRA8_UNORM:
+        return PixelFormat::BGRA8;
+    case Tegra::RenderTargetFormat::RGB10_A2_UNORM:
+        return PixelFormat::A2B10G10R10U;
+    case Tegra::RenderTargetFormat::RGBA16_FLOAT:
+        return PixelFormat::RGBA16F;
+    case Tegra::RenderTargetFormat::RGBA16_UNORM:
+        return PixelFormat::RGBA16U;
+    case Tegra::RenderTargetFormat::RGBA16_UINT:
+        return PixelFormat::RGBA16UI;
+    case Tegra::RenderTargetFormat::RGBA32_FLOAT:
+        return PixelFormat::RGBA32F;
+    case Tegra::RenderTargetFormat::RG32_FLOAT:
+        return PixelFormat::RG32F;
+    case Tegra::RenderTargetFormat::R11G11B10_FLOAT:
+        return PixelFormat::R11FG11FB10F;
+    case Tegra::RenderTargetFormat::B5G6R5_UNORM:
+        return PixelFormat::B5G6R5U;
+    case Tegra::RenderTargetFormat::BGR5A1_UNORM:
+        return PixelFormat::A1B5G5R5U;
+    case Tegra::RenderTargetFormat::RGBA32_UINT:
+        return PixelFormat::RGBA32UI;
+    case Tegra::RenderTargetFormat::R8_UNORM:
+        return PixelFormat::R8U;
+    case Tegra::RenderTargetFormat::R8_UINT:
+        return PixelFormat::R8UI;
+    case Tegra::RenderTargetFormat::RG16_FLOAT:
+        return PixelFormat::RG16F;
+    case Tegra::RenderTargetFormat::RG16_UINT:
+        return PixelFormat::RG16UI;
+    case Tegra::RenderTargetFormat::RG16_SINT:
+        return PixelFormat::RG16I;
+    case Tegra::RenderTargetFormat::RG16_UNORM:
+        return PixelFormat::RG16;
+    case Tegra::RenderTargetFormat::RG16_SNORM:
+        return PixelFormat::RG16S;
+    case Tegra::RenderTargetFormat::RG8_UNORM:
+        return PixelFormat::RG8U;
+    case Tegra::RenderTargetFormat::RG8_SNORM:
+        return PixelFormat::RG8S;
+    case Tegra::RenderTargetFormat::R16_FLOAT:
+        return PixelFormat::R16F;
+    case Tegra::RenderTargetFormat::R16_UNORM:
+        return PixelFormat::R16U;
+    case Tegra::RenderTargetFormat::R16_SNORM:
+        return PixelFormat::R16S;
+    case Tegra::RenderTargetFormat::R16_UINT:
+        return PixelFormat::R16UI;
+    case Tegra::RenderTargetFormat::R16_SINT:
+        return PixelFormat::R16I;
+    case Tegra::RenderTargetFormat::R32_FLOAT:
+        return PixelFormat::R32F;
+    case Tegra::RenderTargetFormat::R32_UINT:
+        return PixelFormat::R32UI;
+    case Tegra::RenderTargetFormat::RG32_UINT:
+        return PixelFormat::RG32UI;
+    default:
+        LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));
+        UNREACHABLE();
+    }
+}
+
+PixelFormat PixelFormatFromTextureFormat(Tegra::Texture::TextureFormat format,
+                                         Tegra::Texture::ComponentType component_type,
+                                         bool is_srgb) {
+    // TODO(Subv): Properly implement this
+    switch (format) {
+    case Tegra::Texture::TextureFormat::A8R8G8B8:
+        if (is_srgb) {
+            return PixelFormat::RGBA8_SRGB;
+        }
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::UNORM:
+            return PixelFormat::ABGR8U;
+        case Tegra::Texture::ComponentType::SNORM:
+            return PixelFormat::ABGR8S;
+        case Tegra::Texture::ComponentType::UINT:
+            return PixelFormat::ABGR8UI;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::B5G6R5:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::UNORM:
+            return PixelFormat::B5G6R5U;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::A2B10G10R10:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::UNORM:
+            return PixelFormat::A2B10G10R10U;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::A1B5G5R5:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::UNORM:
+            return PixelFormat::A1B5G5R5U;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::R8:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::UNORM:
+            return PixelFormat::R8U;
+        case Tegra::Texture::ComponentType::UINT:
+            return PixelFormat::R8UI;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::G8R8:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::UNORM:
+            return PixelFormat::G8R8U;
+        case Tegra::Texture::ComponentType::SNORM:
+            return PixelFormat::G8R8S;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::R16_G16_B16_A16:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::UNORM:
+            return PixelFormat::RGBA16U;
+        case Tegra::Texture::ComponentType::FLOAT:
+            return PixelFormat::RGBA16F;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::BF10GF11RF11:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::FLOAT:
+            return PixelFormat::R11FG11FB10F;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::R32_G32_B32_A32:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::FLOAT:
+            return PixelFormat::RGBA32F;
+        case Tegra::Texture::ComponentType::UINT:
+            return PixelFormat::RGBA32UI;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::R32_G32:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::FLOAT:
+            return PixelFormat::RG32F;
+        case Tegra::Texture::ComponentType::UINT:
+            return PixelFormat::RG32UI;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::R32_G32_B32:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::FLOAT:
+            return PixelFormat::RGB32F;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::R16:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::FLOAT:
+            return PixelFormat::R16F;
+        case Tegra::Texture::ComponentType::UNORM:
+            return PixelFormat::R16U;
+        case Tegra::Texture::ComponentType::SNORM:
+            return PixelFormat::R16S;
+        case Tegra::Texture::ComponentType::UINT:
+            return PixelFormat::R16UI;
+        case Tegra::Texture::ComponentType::SINT:
+            return PixelFormat::R16I;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::R32:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::FLOAT:
+            return PixelFormat::R32F;
+        case Tegra::Texture::ComponentType::UINT:
+            return PixelFormat::R32UI;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::ZF32:
+        return PixelFormat::Z32F;
+    case Tegra::Texture::TextureFormat::Z16:
+        return PixelFormat::Z16;
+    case Tegra::Texture::TextureFormat::Z24S8:
+        return PixelFormat::Z24S8;
+    case Tegra::Texture::TextureFormat::DXT1:
+        return is_srgb ? PixelFormat::DXT1_SRGB : PixelFormat::DXT1;
+    case Tegra::Texture::TextureFormat::DXT23:
+        return is_srgb ? PixelFormat::DXT23_SRGB : PixelFormat::DXT23;
+    case Tegra::Texture::TextureFormat::DXT45:
+        return is_srgb ? PixelFormat::DXT45_SRGB : PixelFormat::DXT45;
+    case Tegra::Texture::TextureFormat::DXN1:
+        return PixelFormat::DXN1;
+    case Tegra::Texture::TextureFormat::DXN2:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::UNORM:
+            return PixelFormat::DXN2UNORM;
+        case Tegra::Texture::ComponentType::SNORM:
+            return PixelFormat::DXN2SNORM;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    case Tegra::Texture::TextureFormat::BC7U:
+        return is_srgb ? PixelFormat::BC7U_SRGB : PixelFormat::BC7U;
+    case Tegra::Texture::TextureFormat::BC6H_UF16:
+        return PixelFormat::BC6H_UF16;
+    case Tegra::Texture::TextureFormat::BC6H_SF16:
+        return PixelFormat::BC6H_SF16;
+    case Tegra::Texture::TextureFormat::ASTC_2D_4X4:
+        return is_srgb ? PixelFormat::ASTC_2D_4X4_SRGB : PixelFormat::ASTC_2D_4X4;
+    case Tegra::Texture::TextureFormat::ASTC_2D_5X4:
+        return is_srgb ? PixelFormat::ASTC_2D_5X4_SRGB : PixelFormat::ASTC_2D_5X4;
+    case Tegra::Texture::TextureFormat::ASTC_2D_5X5:
+        return is_srgb ? PixelFormat::ASTC_2D_5X5_SRGB : PixelFormat::ASTC_2D_5X5;
+    case Tegra::Texture::TextureFormat::ASTC_2D_8X8:
+        return is_srgb ? PixelFormat::ASTC_2D_8X8_SRGB : PixelFormat::ASTC_2D_8X8;
+    case Tegra::Texture::TextureFormat::ASTC_2D_8X5:
+        return is_srgb ? PixelFormat::ASTC_2D_8X5_SRGB : PixelFormat::ASTC_2D_8X5;
+    case Tegra::Texture::TextureFormat::ASTC_2D_10X8:
+        return is_srgb ? PixelFormat::ASTC_2D_10X8_SRGB : PixelFormat::ASTC_2D_10X8;
+    case Tegra::Texture::TextureFormat::R16_G16:
+        switch (component_type) {
+        case Tegra::Texture::ComponentType::FLOAT:
+            return PixelFormat::RG16F;
+        case Tegra::Texture::ComponentType::UNORM:
+            return PixelFormat::RG16;
+        case Tegra::Texture::ComponentType::SNORM:
+            return PixelFormat::RG16S;
+        case Tegra::Texture::ComponentType::UINT:
+            return PixelFormat::RG16UI;
+        case Tegra::Texture::ComponentType::SINT:
+            return PixelFormat::RG16I;
+        }
+        LOG_CRITICAL(HW_GPU, "Unimplemented component_type={}", static_cast<u32>(component_type));
+        UNREACHABLE();
+    default:
+        LOG_CRITICAL(HW_GPU, "Unimplemented format={}, component_type={}", static_cast<u32>(format),
+                     static_cast<u32>(component_type));
+        UNREACHABLE();
+    }
+}
+
+ComponentType ComponentTypeFromTexture(Tegra::Texture::ComponentType type) {
+    // TODO(Subv): Implement more component types
+    switch (type) {
+    case Tegra::Texture::ComponentType::UNORM:
+        return ComponentType::UNorm;
+    case Tegra::Texture::ComponentType::FLOAT:
+        return ComponentType::Float;
+    case Tegra::Texture::ComponentType::SNORM:
+        return ComponentType::SNorm;
+    case Tegra::Texture::ComponentType::UINT:
+        return ComponentType::UInt;
+    case Tegra::Texture::ComponentType::SINT:
+        return ComponentType::SInt;
+    default:
+        LOG_CRITICAL(HW_GPU, "Unimplemented component type={}", static_cast<u32>(type));
+        UNREACHABLE();
+    }
+}
+
+ComponentType ComponentTypeFromRenderTarget(Tegra::RenderTargetFormat format) {
+    // TODO(Subv): Implement more render targets
+    switch (format) {
+    case Tegra::RenderTargetFormat::RGBA8_UNORM:
+    case Tegra::RenderTargetFormat::RGBA8_SRGB:
+    case Tegra::RenderTargetFormat::BGRA8_UNORM:
+    case Tegra::RenderTargetFormat::BGRA8_SRGB:
+    case Tegra::RenderTargetFormat::RGB10_A2_UNORM:
+    case Tegra::RenderTargetFormat::R8_UNORM:
+    case Tegra::RenderTargetFormat::RG16_UNORM:
+    case Tegra::RenderTargetFormat::R16_UNORM:
+    case Tegra::RenderTargetFormat::B5G6R5_UNORM:
+    case Tegra::RenderTargetFormat::BGR5A1_UNORM:
+    case Tegra::RenderTargetFormat::RG8_UNORM:
+    case Tegra::RenderTargetFormat::RGBA16_UNORM:
+        return ComponentType::UNorm;
+    case Tegra::RenderTargetFormat::RGBA8_SNORM:
+    case Tegra::RenderTargetFormat::RG16_SNORM:
+    case Tegra::RenderTargetFormat::R16_SNORM:
+    case Tegra::RenderTargetFormat::RG8_SNORM:
+        return ComponentType::SNorm;
+    case Tegra::RenderTargetFormat::RGBA16_FLOAT:
+    case Tegra::RenderTargetFormat::R11G11B10_FLOAT:
+    case Tegra::RenderTargetFormat::RGBA32_FLOAT:
+    case Tegra::RenderTargetFormat::RG32_FLOAT:
+    case Tegra::RenderTargetFormat::RG16_FLOAT:
+    case Tegra::RenderTargetFormat::R16_FLOAT:
+    case Tegra::RenderTargetFormat::R32_FLOAT:
+        return ComponentType::Float;
+    case Tegra::RenderTargetFormat::RGBA32_UINT:
+    case Tegra::RenderTargetFormat::RGBA16_UINT:
+    case Tegra::RenderTargetFormat::RG16_UINT:
+    case Tegra::RenderTargetFormat::R8_UINT:
+    case Tegra::RenderTargetFormat::R16_UINT:
+    case Tegra::RenderTargetFormat::RG32_UINT:
+    case Tegra::RenderTargetFormat::R32_UINT:
+    case Tegra::RenderTargetFormat::RGBA8_UINT:
+        return ComponentType::UInt;
+    case Tegra::RenderTargetFormat::RG16_SINT:
+    case Tegra::RenderTargetFormat::R16_SINT:
+        return ComponentType::SInt;
+    default:
+        LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));
+        UNREACHABLE();
+    }
+}
+
+PixelFormat PixelFormatFromGPUPixelFormat(Tegra::FramebufferConfig::PixelFormat format) {
+    switch (format) {
+    case Tegra::FramebufferConfig::PixelFormat::ABGR8:
+        return PixelFormat::ABGR8U;
+    default:
+        LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));
+        UNREACHABLE();
+    }
+}
+
+ComponentType ComponentTypeFromDepthFormat(Tegra::DepthFormat format) {
+    switch (format) {
+    case Tegra::DepthFormat::Z16_UNORM:
+    case Tegra::DepthFormat::S8_Z24_UNORM:
+    case Tegra::DepthFormat::Z24_S8_UNORM:
+        return ComponentType::UNorm;
+    case Tegra::DepthFormat::Z32_FLOAT:
+    case Tegra::DepthFormat::Z32_S8_X24_FLOAT:
+        return ComponentType::Float;
+    default:
+        LOG_CRITICAL(HW_GPU, "Unimplemented format={}", static_cast<u32>(format));
+        UNREACHABLE();
+    }
+}
+
+SurfaceType GetFormatType(PixelFormat pixel_format) {
+    if (static_cast<std::size_t>(pixel_format) <
+        static_cast<std::size_t>(PixelFormat::MaxColorFormat)) {
+        return SurfaceType::ColorTexture;
+    }
+
+    if (static_cast<std::size_t>(pixel_format) <
+        static_cast<std::size_t>(PixelFormat::MaxDepthFormat)) {
+        return SurfaceType::Depth;
+    }
+
+    if (static_cast<std::size_t>(pixel_format) <
+        static_cast<std::size_t>(PixelFormat::MaxDepthStencilFormat)) {
+        return SurfaceType::DepthStencil;
+    }
+
+    // TODO(Subv): Implement the other formats
+    ASSERT(false);
+
+    return SurfaceType::Invalid;
+}
+
+bool IsPixelFormatASTC(PixelFormat format) {
+    switch (format) {
+    case PixelFormat::ASTC_2D_4X4:
+    case PixelFormat::ASTC_2D_5X4:
+    case PixelFormat::ASTC_2D_5X5:
+    case PixelFormat::ASTC_2D_8X8:
+    case PixelFormat::ASTC_2D_8X5:
+    case PixelFormat::ASTC_2D_4X4_SRGB:
+    case PixelFormat::ASTC_2D_5X4_SRGB:
+    case PixelFormat::ASTC_2D_5X5_SRGB:
+    case PixelFormat::ASTC_2D_8X8_SRGB:
+    case PixelFormat::ASTC_2D_8X5_SRGB:
+    case PixelFormat::ASTC_2D_10X8:
+    case PixelFormat::ASTC_2D_10X8_SRGB:
+        return true;
+    default:
+        return false;
+    }
+}
+
+std::pair<u32, u32> GetASTCBlockSize(PixelFormat format) {
+    return {GetDefaultBlockWidth(format), GetDefaultBlockHeight(format)};
+}
+
+bool IsFormatBCn(PixelFormat format) {
+    switch (format) {
+    case PixelFormat::DXT1:
+    case PixelFormat::DXT23:
+    case PixelFormat::DXT45:
+    case PixelFormat::DXN1:
+    case PixelFormat::DXN2SNORM:
+    case PixelFormat::DXN2UNORM:
+    case PixelFormat::BC7U:
+    case PixelFormat::BC6H_UF16:
+    case PixelFormat::BC6H_SF16:
+    case PixelFormat::DXT1_SRGB:
+    case PixelFormat::DXT23_SRGB:
+    case PixelFormat::DXT45_SRGB:
+    case PixelFormat::BC7U_SRGB:
+        return true;
+    }
+    return false;
+}
+
+} // namespace VideoCore::Surface
diff --git a/src/video_core/surface.h b/src/video_core/surface.h
new file mode 100644
index 000000000..0dd3eb2e4
--- /dev/null
+++ b/src/video_core/surface.h
@@ -0,0 +1,477 @@
+// Copyright 2014 Citra Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <climits>
+#include <utility>
+#include "common/assert.h"
+#include "common/common_types.h"
+#include "common/logging/log.h"
+#include "video_core/gpu.h"
+#include "video_core/textures/texture.h"
+
+namespace VideoCore::Surface {
+
+enum class PixelFormat {
+    ABGR8U = 0,
+    ABGR8S = 1,
+    ABGR8UI = 2,
+    B5G6R5U = 3,
+    A2B10G10R10U = 4,
+    A1B5G5R5U = 5,
+    R8U = 6,
+    R8UI = 7,
+    RGBA16F = 8,
+    RGBA16U = 9,
+    RGBA16UI = 10,
+    R11FG11FB10F = 11,
+    RGBA32UI = 12,
+    DXT1 = 13,
+    DXT23 = 14,
+    DXT45 = 15,
+    DXN1 = 16, // This is also known as BC4
+    DXN2UNORM = 17,
+    DXN2SNORM = 18,
+    BC7U = 19,
+    BC6H_UF16 = 20,
+    BC6H_SF16 = 21,
+    ASTC_2D_4X4 = 22,
+    G8R8U = 23,
+    G8R8S = 24,
+    BGRA8 = 25,
+    RGBA32F = 26,
+    RG32F = 27,
+    R32F = 28,
+    R16F = 29,
+    R16U = 30,
+    R16S = 31,
+    R16UI = 32,
+    R16I = 33,
+    RG16 = 34,
+    RG16F = 35,
+    RG16UI = 36,
+    RG16I = 37,
+    RG16S = 38,
+    RGB32F = 39,
+    RGBA8_SRGB = 40,
+    RG8U = 41,
+    RG8S = 42,
+    RG32UI = 43,
+    R32UI = 44,
+    ASTC_2D_8X8 = 45,
+    ASTC_2D_8X5 = 46,
+    ASTC_2D_5X4 = 47,
+    BGRA8_SRGB = 48,
+    DXT1_SRGB = 49,
+    DXT23_SRGB = 50,
+    DXT45_SRGB = 51,
+    BC7U_SRGB = 52,
+    ASTC_2D_4X4_SRGB = 53,
+    ASTC_2D_8X8_SRGB = 54,
+    ASTC_2D_8X5_SRGB = 55,
+    ASTC_2D_5X4_SRGB = 56,
+    ASTC_2D_5X5 = 57,
+    ASTC_2D_5X5_SRGB = 58,
+    ASTC_2D_10X8 = 59,
+    ASTC_2D_10X8_SRGB = 60,
+
+    MaxColorFormat,
+
+    // Depth formats
+    Z32F = 61,
+    Z16 = 62,
+
+    MaxDepthFormat,
+
+    // DepthStencil formats
+    Z24S8 = 63,
+    S8Z24 = 64,
+    Z32FS8 = 65,
+
+    MaxDepthStencilFormat,
+
+    Max = MaxDepthStencilFormat,
+    Invalid = 255,
+};
+
+static constexpr std::size_t MaxPixelFormat = static_cast<std::size_t>(PixelFormat::Max);
+
+enum class ComponentType {
+    Invalid = 0,
+    SNorm = 1,
+    UNorm = 2,
+    SInt = 3,
+    UInt = 4,
+    Float = 5,
+};
+
+enum class SurfaceType {
+    ColorTexture = 0,
+    Depth = 1,
+    DepthStencil = 2,
+    Fill = 3,
+    Invalid = 4,
+};
+
+enum class SurfaceTarget {
+    Texture1D,
+    Texture2D,
+    Texture3D,
+    Texture1DArray,
+    Texture2DArray,
+    TextureCubemap,
+    TextureCubeArray,
+};
+
+/**
+ * Gets the compression factor for the specified PixelFormat. This applies to just the
+ * "compressed width" and "compressed height", not the overall compression factor of a
+ * compressed image. This is used for maintaining proper surface sizes for compressed
+ * texture formats.
+ */
+static constexpr u32 GetCompressionFactor(PixelFormat format) {
+    if (format == PixelFormat::Invalid)
+        return 0;
+
+    constexpr std::array<u32, MaxPixelFormat> compression_factor_table = {{
+        1, // ABGR8U
+        1, // ABGR8S
+        1, // ABGR8UI
+        1, // B5G6R5U
+        1, // A2B10G10R10U
+        1, // A1B5G5R5U
+        1, // R8U
+        1, // R8UI
+        1, // RGBA16F
+        1, // RGBA16U
+        1, // RGBA16UI
+        1, // R11FG11FB10F
+        1, // RGBA32UI
+        4, // DXT1
+        4, // DXT23
+        4, // DXT45
+        4, // DXN1
+        4, // DXN2UNORM
+        4, // DXN2SNORM
+        4, // BC7U
+        4, // BC6H_UF16
+        4, // BC6H_SF16
+        4, // ASTC_2D_4X4
+        1, // G8R8U
+        1, // G8R8S
+        1, // BGRA8
+        1, // RGBA32F
+        1, // RG32F
+        1, // R32F
+        1, // R16F
+        1, // R16U
+        1, // R16S
+        1, // R16UI
+        1, // R16I
+        1, // RG16
+        1, // RG16F
+        1, // RG16UI
+        1, // RG16I
+        1, // RG16S
+        1, // RGB32F
+        1, // RGBA8_SRGB
+        1, // RG8U
+        1, // RG8S
+        1, // RG32UI
+        1, // R32UI
+        4, // ASTC_2D_8X8
+        4, // ASTC_2D_8X5
+        4, // ASTC_2D_5X4
+        1, // BGRA8_SRGB
+        4, // DXT1_SRGB
+        4, // DXT23_SRGB
+        4, // DXT45_SRGB
+        4, // BC7U_SRGB
+        4, // ASTC_2D_4X4_SRGB
+        4, // ASTC_2D_8X8_SRGB
+        4, // ASTC_2D_8X5_SRGB
+        4, // ASTC_2D_5X4_SRGB
+        4, // ASTC_2D_5X5
+        4, // ASTC_2D_5X5_SRGB
+        4, // ASTC_2D_10X8
+        4, // ASTC_2D_10X8_SRGB
+        1, // Z32F
+        1, // Z16
+        1, // Z24S8
+        1, // S8Z24
+        1, // Z32FS8
+    }};
+
+    ASSERT(static_cast<std::size_t>(format) < compression_factor_table.size());
+    return compression_factor_table[static_cast<std::size_t>(format)];
+}
+
+static constexpr u32 GetDefaultBlockWidth(PixelFormat format) {
+    if (format == PixelFormat::Invalid)
+        return 0;
+    constexpr std::array<u32, MaxPixelFormat> block_width_table = {{
+        1,  // ABGR8U
+        1,  // ABGR8S
+        1,  // ABGR8UI
+        1,  // B5G6R5U
+        1,  // A2B10G10R10U
+        1,  // A1B5G5R5U
+        1,  // R8U
+        1,  // R8UI
+        1,  // RGBA16F
+        1,  // RGBA16U
+        1,  // RGBA16UI
+        1,  // R11FG11FB10F
+        1,  // RGBA32UI
+        4,  // DXT1
+        4,  // DXT23
+        4,  // DXT45
+        4,  // DXN1
+        4,  // DXN2UNORM
+        4,  // DXN2SNORM
+        4,  // BC7U
+        4,  // BC6H_UF16
+        4,  // BC6H_SF16
+        4,  // ASTC_2D_4X4
+        1,  // G8R8U
+        1,  // G8R8S
+        1,  // BGRA8
+        1,  // RGBA32F
+        1,  // RG32F
+        1,  // R32F
+        1,  // R16F
+        1,  // R16U
+        1,  // R16S
+        1,  // R16UI
+        1,  // R16I
+        1,  // RG16
+        1,  // RG16F
+        1,  // RG16UI
+        1,  // RG16I
+        1,  // RG16S
+        1,  // RGB32F
+        1,  // RGBA8_SRGB
+        1,  // RG8U
+        1,  // RG8S
+        1,  // RG32UI
+        1,  // R32UI
+        8,  // ASTC_2D_8X8
+        8,  // ASTC_2D_8X5
+        5,  // ASTC_2D_5X4
+        1,  // BGRA8_SRGB
+        4,  // DXT1_SRGB
+        4,  // DXT23_SRGB
+        4,  // DXT45_SRGB
+        4,  // BC7U_SRGB
+        4,  // ASTC_2D_4X4_SRGB
+        8,  // ASTC_2D_8X8_SRGB
+        8,  // ASTC_2D_8X5_SRGB
+        5,  // ASTC_2D_5X4_SRGB
+        5,  // ASTC_2D_5X5
+        5,  // ASTC_2D_5X5_SRGB
+        10, // ASTC_2D_10X8
+        10, // ASTC_2D_10X8_SRGB
+        1,  // Z32F
+        1,  // Z16
+        1,  // Z24S8
+        1,  // S8Z24
+        1,  // Z32FS8
+    }};
+    ASSERT(static_cast<std::size_t>(format) < block_width_table.size());
+    return block_width_table[static_cast<std::size_t>(format)];
+}
+
+static constexpr u32 GetDefaultBlockHeight(PixelFormat format) {
+    if (format == PixelFormat::Invalid)
+        return 0;
+
+    constexpr std::array<u32, MaxPixelFormat> block_height_table = {{
+        1, // ABGR8U
+        1, // ABGR8S
+        1, // ABGR8UI
+        1, // B5G6R5U
+        1, // A2B10G10R10U
+        1, // A1B5G5R5U
+        1, // R8U
+        1, // R8UI
+        1, // RGBA16F
+        1, // RGBA16U
+        1, // RGBA16UI
+        1, // R11FG11FB10F
+        1, // RGBA32UI
+        4, // DXT1
+        4, // DXT23
+        4, // DXT45
+        4, // DXN1
+        4, // DXN2UNORM
+        4, // DXN2SNORM
+        4, // BC7U
+        4, // BC6H_UF16
+        4, // BC6H_SF16
+        4, // ASTC_2D_4X4
+        1, // G8R8U
+        1, // G8R8S
+        1, // BGRA8
+        1, // RGBA32F
+        1, // RG32F
+        1, // R32F
+        1, // R16F
+        1, // R16U
+        1, // R16S
+        1, // R16UI
+        1, // R16I
+        1, // RG16
+        1, // RG16F
+        1, // RG16UI
+        1, // RG16I
+        1, // RG16S
+        1, // RGB32F
+        1, // RGBA8_SRGB
+        1, // RG8U
+        1, // RG8S
+        1, // RG32UI
+        1, // R32UI
+        8, // ASTC_2D_8X8
+        5, // ASTC_2D_8X5
+        4, // ASTC_2D_5X4
+        1, // BGRA8_SRGB
+        4, // DXT1_SRGB
+        4, // DXT23_SRGB
+        4, // DXT45_SRGB
+        4, // BC7U_SRGB
+        4, // ASTC_2D_4X4_SRGB
+        8, // ASTC_2D_8X8_SRGB
+        5, // ASTC_2D_8X5_SRGB
+        4, // ASTC_2D_5X4_SRGB
+        5, // ASTC_2D_5X5
+        5, // ASTC_2D_5X5_SRGB
+        8, // ASTC_2D_10X8
+        8, // ASTC_2D_10X8_SRGB
+        1, // Z32F
+        1, // Z16
+        1, // Z24S8
+        1, // S8Z24
+        1, // Z32FS8
+    }};
+
+    ASSERT(static_cast<std::size_t>(format) < block_height_table.size());
+    return block_height_table[static_cast<std::size_t>(format)];
+}
+
+static constexpr u32 GetFormatBpp(PixelFormat format) {
+    if (format == PixelFormat::Invalid)
+        return 0;
+
+    constexpr std::array<u32, MaxPixelFormat> bpp_table = {{
+        32,  // ABGR8U
+        32,  // ABGR8S
+        32,  // ABGR8UI
+        16,  // B5G6R5U
+        32,  // A2B10G10R10U
+        16,  // A1B5G5R5U
+        8,   // R8U
+        8,   // R8UI
+        64,  // RGBA16F
+        64,  // RGBA16U
+        64,  // RGBA16UI
+        32,  // R11FG11FB10F
+        128, // RGBA32UI
+        64,  // DXT1
+        128, // DXT23
+        128, // DXT45
+        64,  // DXN1
+        128, // DXN2UNORM
+        128, // DXN2SNORM
+        128, // BC7U
+        128, // BC6H_UF16
+        128, // BC6H_SF16
+        128, // ASTC_2D_4X4
+        16,  // G8R8U
+        16,  // G8R8S
+        32,  // BGRA8
+        128, // RGBA32F
+        64,  // RG32F
+        32,  // R32F
+        16,  // R16F
+        16,  // R16U
+        16,  // R16S
+        16,  // R16UI
+        16,  // R16I
+        32,  // RG16
+        32,  // RG16F
+        32,  // RG16UI
+        32,  // RG16I
+        32,  // RG16S
+        96,  // RGB32F
+        32,  // RGBA8_SRGB
+        16,  // RG8U
+        16,  // RG8S
+        64,  // RG32UI
+        32,  // R32UI
+        128, // ASTC_2D_8X8
+        128, // ASTC_2D_8X5
+        128, // ASTC_2D_5X4
+        32,  // BGRA8_SRGB
+        64,  // DXT1_SRGB
+        128, // DXT23_SRGB
+        128, // DXT45_SRGB
+        128, // BC7U
+        128, // ASTC_2D_4X4_SRGB
+        128, // ASTC_2D_8X8_SRGB
+        128, // ASTC_2D_8X5_SRGB
+        128, // ASTC_2D_5X4_SRGB
+        128, // ASTC_2D_5X5
+        128, // ASTC_2D_5X5_SRGB
+        128, // ASTC_2D_10X8
+        128, // ASTC_2D_10X8_SRGB
+        32,  // Z32F
+        16,  // Z16
+        32,  // Z24S8
+        32,  // S8Z24
+        64,  // Z32FS8
+    }};
+
+    ASSERT(static_cast<std::size_t>(format) < bpp_table.size());
+    return bpp_table[static_cast<std::size_t>(format)];
+}
+
+/// Returns the sizer in bytes of the specified pixel format
+static constexpr u32 GetBytesPerPixel(PixelFormat pixel_format) {
+    if (pixel_format == PixelFormat::Invalid) {
+        return 0;
+    }
+    return GetFormatBpp(pixel_format) / CHAR_BIT;
+}
+
+SurfaceTarget SurfaceTargetFromTextureType(Tegra::Texture::TextureType texture_type);
+
+bool SurfaceTargetIsLayered(SurfaceTarget target);
+
+PixelFormat PixelFormatFromDepthFormat(Tegra::DepthFormat format);
+
+PixelFormat PixelFormatFromRenderTargetFormat(Tegra::RenderTargetFormat format);
+
+PixelFormat PixelFormatFromTextureFormat(Tegra::Texture::TextureFormat format,
+                                         Tegra::Texture::ComponentType component_type,
+                                         bool is_srgb);
+
+ComponentType ComponentTypeFromTexture(Tegra::Texture::ComponentType type);
+
+ComponentType ComponentTypeFromRenderTarget(Tegra::RenderTargetFormat format);
+
+PixelFormat PixelFormatFromGPUPixelFormat(Tegra::FramebufferConfig::PixelFormat format);
+
+ComponentType ComponentTypeFromDepthFormat(Tegra::DepthFormat format);
+
+SurfaceType GetFormatType(PixelFormat pixel_format);
+
+bool IsPixelFormatASTC(PixelFormat format);
+
+std::pair<u32, u32> GetASTCBlockSize(PixelFormat format);
+
+/// Returns true if the specified PixelFormat is a BCn format, e.g. DXT or DXN
+bool IsFormatBCn(PixelFormat format);
+
+} // namespace VideoCore::Surface
diff --git a/src/video_core/textures/astc.cpp b/src/video_core/textures/astc.cpp
index b1feacae9..bc50a4876 100644
--- a/src/video_core/textures/astc.cpp
+++ b/src/video_core/textures/astc.cpp
@@ -1598,27 +1598,29 @@ static void DecompressBlock(uint8_t inBuf[16], const uint32_t blockWidth,
 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 depth, uint32_t block_width, uint32_t block_height) {
     uint32_t blockIdx = 0;
-    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) {
+    std::vector<uint8_t> outData(height * width * depth * 4);
+    for (uint32_t k = 0; k < depth; k++) {
+        for (uint32_t j = 0; j < height; j += block_height) {
+            for (uint32_t i = 0; i < width; i += block_width) {
 
-            uint8_t* blockPtr = data.data() + blockIdx * 16;
+                uint8_t* blockPtr = data.data() + blockIdx * 16;
 
-            // Blocks can be at most 12x12
-            uint32_t uncompData[144];
-            ASTCC::DecompressBlock(blockPtr, block_width, block_height, uncompData);
+                // Blocks can be at most 12x12
+                uint32_t uncompData[144];
+                ASTCC::DecompressBlock(blockPtr, block_width, block_height, uncompData);
 
-            uint32_t decompWidth = std::min(block_width, width - i);
-            uint32_t decompHeight = std::min(block_height, height - j);
+                uint32_t decompWidth = std::min(block_width, width - i);
+                uint32_t decompHeight = std::min(block_height, height - j);
 
-            uint8_t* outRow = outData.data() + (j * width + i) * 4;
-            for (uint32_t jj = 0; jj < decompHeight; jj++) {
-                memcpy(outRow + jj * width * 4, uncompData + jj * block_width, decompWidth * 4);
-            }
+                uint8_t* outRow = outData.data() + (j * width + i) * 4;
+                for (uint32_t jj = 0; jj < decompHeight; jj++) {
+                    memcpy(outRow + jj * width * 4, uncompData + jj * block_width, decompWidth * 4);
+                }
 
-            blockIdx++;
+                blockIdx++;
+            }
         }
     }
 
diff --git a/src/video_core/textures/astc.h b/src/video_core/textures/astc.h
index f0d7c0e56..d419dd025 100644
--- a/src/video_core/textures/astc.h
+++ b/src/video_core/textures/astc.h
@@ -10,6 +10,6 @@
 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 depth, uint32_t block_width, uint32_t block_height);
 
 } // namespace Tegra::Texture::ASTC
diff --git a/src/video_core/textures/decoders.cpp b/src/video_core/textures/decoders.cpp
index f1b40e7f5..c9160b467 100644
--- a/src/video_core/textures/decoders.cpp
+++ b/src/video_core/textures/decoders.cpp
@@ -45,7 +45,7 @@ constexpr auto fast_swizzle_table = SwizzleTable<8, 4, 16>();
  * Instead of going gob by gob, we map the coordinates inside a block and manage from
  * those. Block_Width is assumed to be 1.
  */
-void PreciseProcessBlock(u8* swizzled_data, u8* unswizzled_data, const bool unswizzle,
+void PreciseProcessBlock(u8* const swizzled_data, u8* const unswizzled_data, const bool unswizzle,
                          const u32 x_start, const u32 y_start, const u32 z_start, const u32 x_end,
                          const u32 y_end, const u32 z_end, const u32 tile_offset,
                          const u32 xy_block_size, const u32 layer_z, const u32 stride_x,
@@ -81,7 +81,7 @@ void PreciseProcessBlock(u8* swizzled_data, u8* unswizzled_data, const bool unsw
  * Instead of going gob by gob, we map the coordinates inside a block and manage from
  * those. Block_Width is assumed to be 1.
  */
-void FastProcessBlock(u8* swizzled_data, u8* unswizzled_data, const bool unswizzle,
+void FastProcessBlock(u8* const swizzled_data, u8* const unswizzled_data, const bool unswizzle,
                       const u32 x_start, const u32 y_start, const u32 z_start, const u32 x_end,
                       const u32 y_end, const u32 z_end, const u32 tile_offset,
                       const u32 xy_block_size, const u32 layer_z, const u32 stride_x,
@@ -90,10 +90,10 @@ void FastProcessBlock(u8* swizzled_data, u8* unswizzled_data, const bool unswizz
     u32 z_address = tile_offset;
     const u32 x_startb = x_start * bytes_per_pixel;
     const u32 x_endb = x_end * bytes_per_pixel;
-    const u32 copy_size = 16;
-    const u32 gob_size_x = 64;
-    const u32 gob_size_y = 8;
-    const u32 gob_size_z = 1;
+    constexpr u32 copy_size = 16;
+    constexpr u32 gob_size_x = 64;
+    constexpr u32 gob_size_y = 8;
+    constexpr u32 gob_size_z = 1;
     const u32 gob_size = gob_size_x * gob_size_y * gob_size_z;
     for (u32 z = z_start; z < z_end; z++) {
         u32 y_address = z_address;
@@ -126,24 +126,23 @@ void FastProcessBlock(u8* swizzled_data, u8* unswizzled_data, const bool unswizz
  *  https://envytools.readthedocs.io/en/latest/hw/memory/g80-surface.html#blocklinear-surfaces
  */
 template <bool fast>
-void SwizzledData(u8* swizzled_data, u8* unswizzled_data, const bool unswizzle, const u32 width,
-                  const u32 height, const u32 depth, const u32 bytes_per_pixel,
+void SwizzledData(u8* const swizzled_data, u8* const unswizzled_data, const bool unswizzle,
+                  const u32 width, const u32 height, const u32 depth, const u32 bytes_per_pixel,
                   const u32 out_bytes_per_pixel, const u32 block_height, const u32 block_depth) {
     auto div_ceil = [](const u32 x, const u32 y) { return ((x + y - 1) / y); };
     const u32 stride_x = width * out_bytes_per_pixel;
     const u32 layer_z = height * stride_x;
-    const u32 gob_x_bytes = 64;
+    constexpr u32 gob_x_bytes = 64;
     const u32 gob_elements_x = gob_x_bytes / bytes_per_pixel;
-    const u32 gob_elements_y = 8;
-    const u32 gob_elements_z = 1;
+    constexpr u32 gob_elements_y = 8;
+    constexpr u32 gob_elements_z = 1;
     const u32 block_x_elements = gob_elements_x;
     const u32 block_y_elements = gob_elements_y * block_height;
     const u32 block_z_elements = gob_elements_z * block_depth;
     const u32 blocks_on_x = div_ceil(width, block_x_elements);
     const u32 blocks_on_y = div_ceil(height, block_y_elements);
     const u32 blocks_on_z = div_ceil(depth, block_z_elements);
-    const u32 blocks = blocks_on_x * blocks_on_y * blocks_on_z;
-    const u32 gob_size = gob_x_bytes * gob_elements_y * gob_elements_z;
+    constexpr u32 gob_size = gob_x_bytes * gob_elements_y * gob_elements_z;
     const u32 xy_block_size = gob_size * block_height;
     const u32 block_size = xy_block_size * block_depth;
     u32 tile_offset = 0;
@@ -172,7 +171,7 @@ void SwizzledData(u8* swizzled_data, u8* unswizzled_data, const bool unswizzle,
 }
 
 void CopySwizzledData(u32 width, u32 height, u32 depth, u32 bytes_per_pixel,
-                      u32 out_bytes_per_pixel, u8* swizzled_data, u8* unswizzled_data,
+                      u32 out_bytes_per_pixel, u8* const swizzled_data, u8* const unswizzled_data,
                       bool unswizzle, u32 block_height, u32 block_depth) {
     if (bytes_per_pixel % 3 != 0 && (width * bytes_per_pixel) % 16 == 0) {
         SwizzledData<true>(swizzled_data, unswizzled_data, unswizzle, width, height, depth,
@@ -203,6 +202,8 @@ u32 BytesPerPixel(TextureFormat format) {
     case TextureFormat::ASTC_2D_5X4:
     case TextureFormat::ASTC_2D_8X8:
     case TextureFormat::ASTC_2D_8X5:
+    case TextureFormat::ASTC_2D_10X8:
+    case TextureFormat::ASTC_2D_5X5:
     case TextureFormat::A8R8G8B8:
     case TextureFormat::A2B10G10R10:
     case TextureFormat::BF10GF11RF11:
@@ -228,12 +229,21 @@ u32 BytesPerPixel(TextureFormat format) {
     }
 }
 
-std::vector<u8> UnswizzleTexture(VAddr address, u32 tile_size, u32 bytes_per_pixel, u32 width,
-                                 u32 height, u32 depth, u32 block_height, u32 block_depth) {
+void UnswizzleTexture(u8* const unswizzled_data, VAddr address, u32 tile_size_x, u32 tile_size_y,
+                      u32 bytes_per_pixel, u32 width, u32 height, u32 depth, u32 block_height,
+                      u32 block_depth) {
+    CopySwizzledData((width + tile_size_x - 1) / tile_size_x,
+                     (height + tile_size_y - 1) / tile_size_y, depth, bytes_per_pixel,
+                     bytes_per_pixel, Memory::GetPointer(address), unswizzled_data, true,
+                     block_height, block_depth);
+}
+
+std::vector<u8> UnswizzleTexture(VAddr address, u32 tile_size_x, u32 tile_size_y,
+                                 u32 bytes_per_pixel, u32 width, u32 height, u32 depth,
+                                 u32 block_height, u32 block_depth) {
     std::vector<u8> unswizzled_data(width * height * depth * bytes_per_pixel);
-    CopySwizzledData(width / tile_size, height / tile_size, depth, bytes_per_pixel, bytes_per_pixel,
-                     Memory::GetPointer(address), unswizzled_data.data(), true, block_height,
-                     block_depth);
+    UnswizzleTexture(unswizzled_data.data(), address, tile_size_x, tile_size_y, bytes_per_pixel,
+                     width, height, depth, block_height, block_depth);
     return unswizzled_data;
 }
 
@@ -293,6 +303,8 @@ std::vector<u8> DecodeTexture(const std::vector<u8>& texture_data, TextureFormat
     case TextureFormat::BC6H_SF16:
     case TextureFormat::ASTC_2D_4X4:
     case TextureFormat::ASTC_2D_8X8:
+    case TextureFormat::ASTC_2D_5X5:
+    case TextureFormat::ASTC_2D_10X8:
     case TextureFormat::A8R8G8B8:
     case TextureFormat::A2B10G10R10:
     case TextureFormat::A1B5G5R5:
@@ -320,13 +332,13 @@ std::vector<u8> DecodeTexture(const std::vector<u8>& texture_data, TextureFormat
 std::size_t CalculateSize(bool tiled, u32 bytes_per_pixel, u32 width, u32 height, u32 depth,
                           u32 block_height, u32 block_depth) {
     if (tiled) {
-        const u32 gobs_in_x = 64 / bytes_per_pixel;
-        const u32 gobs_in_y = 8;
-        const u32 gobs_in_z = 1;
-        const u32 aligned_width = Common::AlignUp(width, gobs_in_x);
+        constexpr u32 gobs_in_x = 64;
+        constexpr u32 gobs_in_y = 8;
+        constexpr u32 gobs_in_z = 1;
+        const u32 aligned_width = Common::AlignUp(width * bytes_per_pixel, gobs_in_x);
         const u32 aligned_height = Common::AlignUp(height, gobs_in_y * block_height);
         const u32 aligned_depth = Common::AlignUp(depth, gobs_in_z * block_depth);
-        return aligned_width * aligned_height * aligned_depth * bytes_per_pixel;
+        return aligned_width * aligned_height * aligned_depth;
     } else {
         return width * height * depth * bytes_per_pixel;
     }
diff --git a/src/video_core/textures/decoders.h b/src/video_core/textures/decoders.h
index 4726f54a5..f4ef7c73e 100644
--- a/src/video_core/textures/decoders.h
+++ b/src/video_core/textures/decoders.h
@@ -10,11 +10,24 @@
 
 namespace Tegra::Texture {
 
+// GOBSize constant. Calculated by 64 bytes in x multiplied by 8 y coords, represents
+// an small rect of (64/bytes_per_pixel)X8.
+inline std::size_t GetGOBSize() {
+    return 512;
+}
+
+/**
+ * Unswizzles a swizzled texture without changing its format.
+ */
+void UnswizzleTexture(u8* unswizzled_data, VAddr address, u32 tile_size_x, u32 tile_size_y,
+                      u32 bytes_per_pixel, u32 width, u32 height, u32 depth,
+                      u32 block_height = TICEntry::DefaultBlockHeight,
+                      u32 block_depth = TICEntry::DefaultBlockHeight);
 /**
  * Unswizzles a swizzled texture without changing its format.
  */
-std::vector<u8> UnswizzleTexture(VAddr address, u32 tile_size, u32 bytes_per_pixel, u32 width,
-                                 u32 height, u32 depth,
+std::vector<u8> UnswizzleTexture(VAddr address, u32 tile_size_x, u32 tile_size_y,
+                                 u32 bytes_per_pixel, u32 width, u32 height, u32 depth,
                                  u32 block_height = TICEntry::DefaultBlockHeight,
                                  u32 block_depth = TICEntry::DefaultBlockHeight);
 
diff --git a/src/video_core/textures/texture.h b/src/video_core/textures/texture.h
index 5947bd2b9..e199d019a 100644
--- a/src/video_core/textures/texture.h
+++ b/src/video_core/textures/texture.h
@@ -168,19 +168,29 @@ struct TICEntry {
 
         // High 16 bits of the pitch value
         BitField<0, 16, u32> pitch_high;
-
+        BitField<26, 1, u32> use_header_opt_control;
+        BitField<27, 1, u32> depth_texture;
         BitField<28, 4, u32> max_mip_level;
     };
     union {
         BitField<0, 16, u32> width_minus_1;
+        BitField<22, 1, u32> srgb_conversion;
         BitField<23, 4, TextureType> texture_type;
+        BitField<29, 3, u32> border_size;
     };
     union {
         BitField<0, 16, u32> height_minus_1;
         BitField<16, 15, u32> depth_minus_1;
     };
+    union {
+        BitField<6, 13, u32> mip_lod_bias;
+        BitField<27, 3, u32> max_anisotropy;
+    };
 
-    INSERT_PADDING_BYTES(8);
+    union {
+        BitField<0, 4, u32> res_min_mip_level;
+        BitField<4, 4, u32> res_max_mip_level;
+    };
 
     GPUVAddr Address() const {
         return static_cast<GPUVAddr>((static_cast<GPUVAddr>(address_high) << 32) | address_low);
@@ -227,6 +237,10 @@ struct TICEntry {
         return header_version == TICHeaderVersion::BlockLinear ||
                header_version == TICHeaderVersion::BlockLinearColorKey;
     }
+
+    bool IsSrgbConversionEnabled() const {
+        return srgb_conversion != 0;
+    }
 };
 static_assert(sizeof(TICEntry) == 0x20, "TICEntry has wrong size");
 
diff --git a/src/video_core/utils.h b/src/video_core/utils.h
index 237cc1307..e0a14d48f 100644
--- a/src/video_core/utils.h
+++ b/src/video_core/utils.h
@@ -161,30 +161,4 @@ static inline void MortonCopyPixels128(u32 width, u32 height, u32 bytes_per_pixe
     }
 }
 
-static void LabelGLObject(GLenum identifier, GLuint handle, VAddr addr,
-                          std::string extra_info = "") {
-    if (!GLAD_GL_KHR_debug) {
-        return; // We don't need to throw an error as this is just for debugging
-    }
-    const std::string nice_addr = fmt::format("0x{:016x}", addr);
-    std::string object_label;
-
-    if (extra_info.empty()) {
-        switch (identifier) {
-        case GL_TEXTURE:
-            object_label = "Texture@" + nice_addr;
-            break;
-        case GL_PROGRAM:
-            object_label = "Shader@" + nice_addr;
-            break;
-        default:
-            object_label = fmt::format("Object(0x{:x})@{}", identifier, nice_addr);
-            break;
-        }
-    } else {
-        object_label = extra_info + '@' + nice_addr;
-    }
-    glObjectLabel(identifier, handle, -1, static_cast<const GLchar*>(object_label.c_str()));
-}
-
 } // namespace VideoCore
diff --git a/src/web_service/telemetry_json.cpp b/src/web_service/telemetry_json.cpp
index 0a8f2bd9e..9156ce802 100644
--- a/src/web_service/telemetry_json.cpp
+++ b/src/web_service/telemetry_json.cpp
@@ -102,16 +102,27 @@ void TelemetryJson::Complete() {
     impl->SerializeSection(Telemetry::FieldType::App, "App");
     impl->SerializeSection(Telemetry::FieldType::Session, "Session");
     impl->SerializeSection(Telemetry::FieldType::Performance, "Performance");
-    impl->SerializeSection(Telemetry::FieldType::UserFeedback, "UserFeedback");
     impl->SerializeSection(Telemetry::FieldType::UserConfig, "UserConfig");
     impl->SerializeSection(Telemetry::FieldType::UserSystem, "UserSystem");
 
     auto content = impl->TopSection().dump();
     // Send the telemetry async but don't handle the errors since they were written to the log
-    Common::DetachedTasks::AddTask(
-        [host{impl->host}, username{impl->username}, token{impl->token}, content]() {
-            Client{host, username, token}.PostJson("/telemetry", content, true);
-        });
+    Common::DetachedTasks::AddTask([host{impl->host}, content]() {
+        Client{host, "", ""}.PostJson("/telemetry", content, true);
+    });
+}
+
+bool TelemetryJson::SubmitTestcase() {
+    impl->SerializeSection(Telemetry::FieldType::App, "App");
+    impl->SerializeSection(Telemetry::FieldType::Session, "Session");
+    impl->SerializeSection(Telemetry::FieldType::UserFeedback, "UserFeedback");
+    impl->SerializeSection(Telemetry::FieldType::UserSystem, "UserSystem");
+
+    auto content = impl->TopSection().dump();
+    Client client(impl->host, impl->username, impl->token);
+    auto value = client.PostJson("/gamedb/testcase", content, false);
+
+    return value.result_code == Common::WebResult::Code::Success;
 }
 
 } // namespace WebService
diff --git a/src/web_service/telemetry_json.h b/src/web_service/telemetry_json.h
index 93371414a..dfd202829 100644
--- a/src/web_service/telemetry_json.h
+++ b/src/web_service/telemetry_json.h
@@ -35,6 +35,7 @@ public:
     void Visit(const Telemetry::Field<std::chrono::microseconds>& field) override;
 
     void Complete() override;
+    bool SubmitTestcase() override;
 
 private:
     struct Impl;
diff --git a/src/yuzu/CMakeLists.txt b/src/yuzu/CMakeLists.txt
index 9379d9110..f9ca2948e 100644
--- a/src/yuzu/CMakeLists.txt
+++ b/src/yuzu/CMakeLists.txt
@@ -56,6 +56,8 @@ add_executable(yuzu
     main.h
     ui_settings.cpp
     ui_settings.h
+    util/limitable_input_dialog.cpp
+    util/limitable_input_dialog.h
     util/spinbox.cpp
     util/spinbox.h
     util/util.cpp
diff --git a/src/yuzu/bootmanager.cpp b/src/yuzu/bootmanager.cpp
index e8ab23326..39eef8858 100644
--- a/src/yuzu/bootmanager.cpp
+++ b/src/yuzu/bootmanager.cpp
@@ -8,7 +8,6 @@
 
 #include "common/microprofile.h"
 #include "common/scm_rev.h"
-#include "common/string_util.h"
 #include "core/core.h"
 #include "core/frontend/framebuffer_layout.h"
 #include "core/settings.h"
@@ -107,9 +106,8 @@ private:
 GRenderWindow::GRenderWindow(QWidget* parent, EmuThread* emu_thread)
     : QWidget(parent), child(nullptr), emu_thread(emu_thread) {
 
-    std::string window_title = fmt::format("yuzu {} | {}-{}", Common::g_build_name,
-                                           Common::g_scm_branch, Common::g_scm_desc);
-    setWindowTitle(QString::fromStdString(window_title));
+    setWindowTitle(QStringLiteral("yuzu %1 | %2-%3")
+                       .arg(Common::g_build_name, Common::g_scm_branch, Common::g_scm_desc));
     setAttribute(Qt::WA_AcceptTouchEvents);
 
     InputCommon::Init();
diff --git a/src/yuzu/compatdb.cpp b/src/yuzu/compatdb.cpp
index 91e754274..5f0896f84 100644
--- a/src/yuzu/compatdb.cpp
+++ b/src/yuzu/compatdb.cpp
@@ -5,6 +5,7 @@
 #include <QButtonGroup>
 #include <QMessageBox>
 #include <QPushButton>
+#include <QtConcurrent/qtconcurrentrun.h>
 #include "common/logging/log.h"
 #include "common/telemetry.h"
 #include "core/core.h"
@@ -23,6 +24,8 @@ CompatDB::CompatDB(QWidget* parent)
     connect(ui->radioButton_IntroMenu, &QRadioButton::clicked, this, &CompatDB::EnableNext);
     connect(ui->radioButton_WontBoot, &QRadioButton::clicked, this, &CompatDB::EnableNext);
     connect(button(NextButton), &QPushButton::clicked, this, &CompatDB::Submit);
+    connect(&testcase_watcher, &QFutureWatcher<bool>::finished, this,
+            &CompatDB::OnTestcaseSubmitted);
 }
 
 CompatDB::~CompatDB() = default;
@@ -48,18 +51,38 @@ void CompatDB::Submit() {
         }
         break;
     case CompatDBPage::Final:
+        back();
         LOG_DEBUG(Frontend, "Compatibility Rating: {}", compatibility->checkedId());
         Core::Telemetry().AddField(Telemetry::FieldType::UserFeedback, "Compatibility",
                                    compatibility->checkedId());
-        // older versions of QT don't support the "NoCancelButtonOnLastPage" option, this is a
-        // workaround
+
+        button(NextButton)->setEnabled(false);
+        button(NextButton)->setText(tr("Submitting"));
         button(QWizard::CancelButton)->setVisible(false);
+
+        testcase_watcher.setFuture(QtConcurrent::run(
+            [this]() { return Core::System::GetInstance().TelemetrySession().SubmitTestcase(); }));
         break;
     default:
         LOG_ERROR(Frontend, "Unexpected page: {}", currentId());
     }
 }
 
+void CompatDB::OnTestcaseSubmitted() {
+    if (!testcase_watcher.result()) {
+        QMessageBox::critical(this, tr("Communication error"),
+                              tr("An error occured while sending the Testcase"));
+        button(NextButton)->setEnabled(true);
+        button(NextButton)->setText(tr("Next"));
+        button(QWizard::CancelButton)->setVisible(true);
+    } else {
+        next();
+        // older versions of QT don't support the "NoCancelButtonOnLastPage" option, this is a
+        // workaround
+        button(QWizard::CancelButton)->setVisible(false);
+    }
+}
+
 void CompatDB::EnableNext() {
     button(NextButton)->setEnabled(true);
 }
diff --git a/src/yuzu/compatdb.h b/src/yuzu/compatdb.h
index ca0dd11d6..5381f67f7 100644
--- a/src/yuzu/compatdb.h
+++ b/src/yuzu/compatdb.h
@@ -5,6 +5,7 @@
 #pragma once
 
 #include <memory>
+#include <QFutureWatcher>
 #include <QWizard>
 
 namespace Ui {
@@ -19,8 +20,11 @@ public:
     ~CompatDB();
 
 private:
+    QFutureWatcher<bool> testcase_watcher;
+
     std::unique_ptr<Ui::CompatDB> ui;
 
     void Submit();
+    void OnTestcaseSubmitted();
     void EnableNext();
 };
diff --git a/src/yuzu/configuration/config.cpp b/src/yuzu/configuration/config.cpp
index 71c6ebb41..be69fb831 100644
--- a/src/yuzu/configuration/config.cpp
+++ b/src/yuzu/configuration/config.cpp
@@ -4,6 +4,7 @@
 
 #include <QSettings>
 #include "common/file_util.h"
+#include "core/hle/service/acc/profile_manager.h"
 #include "input_common/main.h"
 #include "yuzu/configuration/config.h"
 #include "yuzu/ui_settings.h"
@@ -12,11 +13,16 @@ Config::Config() {
     // TODO: Don't hardcode the path; let the frontend decide where to put the config files.
     qt_config_loc = FileUtil::GetUserPath(FileUtil::UserPath::ConfigDir) + "qt-config.ini";
     FileUtil::CreateFullPath(qt_config_loc);
-    qt_config = new QSettings(QString::fromStdString(qt_config_loc), QSettings::IniFormat);
+    qt_config =
+        std::make_unique<QSettings>(QString::fromStdString(qt_config_loc), QSettings::IniFormat);
 
     Reload();
 }
 
+Config::~Config() {
+    Save();
+}
+
 const std::array<int, Settings::NativeButton::NumButtons> Config::default_buttons = {
     Qt::Key_A, Qt::Key_S, Qt::Key_Z,    Qt::Key_X,  Qt::Key_3,     Qt::Key_4,    Qt::Key_Q,
     Qt::Key_W, Qt::Key_1, Qt::Key_2,    Qt::Key_N,  Qt::Key_M,     Qt::Key_F,    Qt::Key_T,
@@ -122,8 +128,20 @@ void Config::ReadValues() {
 
     qt_config->beginGroup("System");
     Settings::values.use_docked_mode = qt_config->value("use_docked_mode", false).toBool();
-    Settings::values.username = qt_config->value("username", "yuzu").toString().toStdString();
+    Settings::values.enable_nfc = qt_config->value("enable_nfc", true).toBool();
+
+    Settings::values.current_user = std::clamp<int>(qt_config->value("current_user", 0).toInt(), 0,
+                                                    Service::Account::MAX_USERS - 1);
+
     Settings::values.language_index = qt_config->value("language_index", 1).toInt();
+
+    const auto enabled = qt_config->value("rng_seed_enabled", false).toBool();
+    if (enabled) {
+        Settings::values.rng_seed = qt_config->value("rng_seed", 0).toULongLong();
+    } else {
+        Settings::values.rng_seed = std::nullopt;
+    }
+
     qt_config->endGroup();
 
     qt_config->beginGroup("Miscellaneous");
@@ -135,6 +153,7 @@ void Config::ReadValues() {
     Settings::values.use_gdbstub = qt_config->value("use_gdbstub", false).toBool();
     Settings::values.gdbstub_port = qt_config->value("gdbstub_port", 24689).toInt();
     Settings::values.program_args = qt_config->value("program_args", "").toString().toStdString();
+    Settings::values.dump_nso = qt_config->value("dump_nso", false).toBool();
     qt_config->endGroup();
 
     qt_config->beginGroup("WebService");
@@ -152,6 +171,7 @@ void Config::ReadValues() {
 
     qt_config->beginGroup("UIGameList");
     UISettings::values.show_unknown = qt_config->value("show_unknown", true).toBool();
+    UISettings::values.show_add_ons = qt_config->value("show_add_ons", true).toBool();
     UISettings::values.icon_size = qt_config->value("icon_size", 64).toUInt();
     UISettings::values.row_1_text_id = qt_config->value("row_1_text_id", 3).toUInt();
     UISettings::values.row_2_text_id = qt_config->value("row_2_text_id", 2).toUInt();
@@ -258,8 +278,14 @@ void Config::SaveValues() {
 
     qt_config->beginGroup("System");
     qt_config->setValue("use_docked_mode", Settings::values.use_docked_mode);
-    qt_config->setValue("username", QString::fromStdString(Settings::values.username));
+    qt_config->setValue("enable_nfc", Settings::values.enable_nfc);
+    qt_config->setValue("current_user", Settings::values.current_user);
+
     qt_config->setValue("language_index", Settings::values.language_index);
+
+    qt_config->setValue("rng_seed_enabled", Settings::values.rng_seed.has_value());
+    qt_config->setValue("rng_seed", Settings::values.rng_seed.value_or(0));
+
     qt_config->endGroup();
 
     qt_config->beginGroup("Miscellaneous");
@@ -271,6 +297,7 @@ void Config::SaveValues() {
     qt_config->setValue("use_gdbstub", Settings::values.use_gdbstub);
     qt_config->setValue("gdbstub_port", Settings::values.gdbstub_port);
     qt_config->setValue("program_args", QString::fromStdString(Settings::values.program_args));
+    qt_config->setValue("dump_nso", Settings::values.dump_nso);
     qt_config->endGroup();
 
     qt_config->beginGroup("WebService");
@@ -286,6 +313,7 @@ void Config::SaveValues() {
 
     qt_config->beginGroup("UIGameList");
     qt_config->setValue("show_unknown", UISettings::values.show_unknown);
+    qt_config->setValue("show_add_ons", UISettings::values.show_add_ons);
     qt_config->setValue("icon_size", UISettings::values.icon_size);
     qt_config->setValue("row_1_text_id", UISettings::values.row_1_text_id);
     qt_config->setValue("row_2_text_id", UISettings::values.row_2_text_id);
@@ -335,9 +363,3 @@ void Config::Reload() {
 void Config::Save() {
     SaveValues();
 }
-
-Config::~Config() {
-    Save();
-
-    delete qt_config;
-}
diff --git a/src/yuzu/configuration/config.h b/src/yuzu/configuration/config.h
index cbf745ea2..9c99c1b75 100644
--- a/src/yuzu/configuration/config.h
+++ b/src/yuzu/configuration/config.h
@@ -5,6 +5,7 @@
 #pragma once
 
 #include <array>
+#include <memory>
 #include <string>
 #include <QVariant>
 #include "core/settings.h"
@@ -12,12 +13,6 @@
 class QSettings;
 
 class Config {
-    QSettings* qt_config;
-    std::string qt_config_loc;
-
-    void ReadValues();
-    void SaveValues();
-
 public:
     Config();
     ~Config();
@@ -27,4 +22,11 @@ public:
 
     static const std::array<int, Settings::NativeButton::NumButtons> default_buttons;
     static const std::array<std::array<int, 5>, Settings::NativeAnalog::NumAnalogs> default_analogs;
+
+private:
+    void ReadValues();
+    void SaveValues();
+
+    std::unique_ptr<QSettings> qt_config;
+    std::string qt_config_loc;
 };
diff --git a/src/yuzu/configuration/configure_debug.cpp b/src/yuzu/configuration/configure_debug.cpp
index 9e765fc93..fd5876b41 100644
--- a/src/yuzu/configuration/configure_debug.cpp
+++ b/src/yuzu/configuration/configure_debug.cpp
@@ -34,6 +34,7 @@ void ConfigureDebug::setConfiguration() {
     ui->toggle_console->setChecked(UISettings::values.show_console);
     ui->log_filter_edit->setText(QString::fromStdString(Settings::values.log_filter));
     ui->homebrew_args_edit->setText(QString::fromStdString(Settings::values.program_args));
+    ui->dump_decompressed_nso->setChecked(Settings::values.dump_nso);
 }
 
 void ConfigureDebug::applyConfiguration() {
@@ -42,6 +43,7 @@ void ConfigureDebug::applyConfiguration() {
     UISettings::values.show_console = ui->toggle_console->isChecked();
     Settings::values.log_filter = ui->log_filter_edit->text().toStdString();
     Settings::values.program_args = ui->homebrew_args_edit->text().toStdString();
+    Settings::values.dump_nso = ui->dump_decompressed_nso->isChecked();
     Debugger::ToggleConsole();
     Log::Filter filter;
     filter.ParseFilterString(Settings::values.log_filter);
diff --git a/src/yuzu/configuration/configure_debug.ui b/src/yuzu/configuration/configure_debug.ui
index ff4987604..9c5b702f8 100644
--- a/src/yuzu/configuration/configure_debug.ui
+++ b/src/yuzu/configuration/configure_debug.ui
@@ -7,7 +7,7 @@
     <x>0</x>
     <y>0</y>
     <width>400</width>
-    <height>300</height>
+    <height>357</height>
    </rect>
   </property>
   <property name="windowTitle">
@@ -130,6 +130,25 @@
     </widget>
    </item>
    <item>
+    <widget class="QGroupBox" name="groupBox_4">
+     <property name="title">
+      <string>Dump</string>
+     </property>
+     <layout class="QVBoxLayout" name="verticalLayout_4">
+      <item>
+       <widget class="QCheckBox" name="dump_decompressed_nso">
+        <property name="whatsThis">
+         <string>When checked, any NSO yuzu tries to load or patch will be copied decompressed to the yuzu/dump directory.</string>
+        </property>
+        <property name="text">
+         <string>Dump Decompressed NSOs</string>
+        </property>
+       </widget>
+      </item>
+     </layout>
+    </widget>
+   </item>
+   <item>
     <spacer name="verticalSpacer">
      <property name="orientation">
       <enum>Qt::Vertical</enum>
diff --git a/src/yuzu/configuration/configure_gamelist.cpp b/src/yuzu/configuration/configure_gamelist.cpp
index 8743ce982..ae8cac243 100644
--- a/src/yuzu/configuration/configure_gamelist.cpp
+++ b/src/yuzu/configuration/configure_gamelist.cpp
@@ -36,20 +36,36 @@ ConfigureGameList::ConfigureGameList(QWidget* parent)
     InitializeRowComboBoxes();
 
     this->setConfiguration();
+
+    // Force game list reload if any of the relevant settings are changed.
+    connect(ui->show_unknown, &QCheckBox::stateChanged, this,
+            &ConfigureGameList::RequestGameListUpdate);
+    connect(ui->icon_size_combobox, QOverload<int>::of(&QComboBox::currentIndexChanged), this,
+            &ConfigureGameList::RequestGameListUpdate);
+    connect(ui->row_1_text_combobox, QOverload<int>::of(&QComboBox::currentIndexChanged), this,
+            &ConfigureGameList::RequestGameListUpdate);
+    connect(ui->row_2_text_combobox, QOverload<int>::of(&QComboBox::currentIndexChanged), this,
+            &ConfigureGameList::RequestGameListUpdate);
 }
 
 ConfigureGameList::~ConfigureGameList() = default;
 
 void ConfigureGameList::applyConfiguration() {
     UISettings::values.show_unknown = ui->show_unknown->isChecked();
+    UISettings::values.show_add_ons = ui->show_add_ons->isChecked();
     UISettings::values.icon_size = ui->icon_size_combobox->currentData().toUInt();
     UISettings::values.row_1_text_id = ui->row_1_text_combobox->currentData().toUInt();
     UISettings::values.row_2_text_id = ui->row_2_text_combobox->currentData().toUInt();
     Settings::Apply();
 }
 
+void ConfigureGameList::RequestGameListUpdate() {
+    UISettings::values.is_game_list_reload_pending.exchange(true);
+}
+
 void ConfigureGameList::setConfiguration() {
     ui->show_unknown->setChecked(UISettings::values.show_unknown);
+    ui->show_add_ons->setChecked(UISettings::values.show_add_ons);
     ui->icon_size_combobox->setCurrentIndex(
         ui->icon_size_combobox->findData(UISettings::values.icon_size));
     ui->row_1_text_combobox->setCurrentIndex(
diff --git a/src/yuzu/configuration/configure_gamelist.h b/src/yuzu/configuration/configure_gamelist.h
index ff7406c60..bbf7e25f1 100644
--- a/src/yuzu/configuration/configure_gamelist.h
+++ b/src/yuzu/configuration/configure_gamelist.h
@@ -21,6 +21,8 @@ public:
     void applyConfiguration();
 
 private:
+    void RequestGameListUpdate();
+
     void setConfiguration();
 
     void changeEvent(QEvent*) override;
diff --git a/src/yuzu/configuration/configure_gamelist.ui b/src/yuzu/configuration/configure_gamelist.ui
index 7471fdb60..7a69377e7 100644
--- a/src/yuzu/configuration/configure_gamelist.ui
+++ b/src/yuzu/configuration/configure_gamelist.ui
@@ -1,126 +1,133 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <ui version="4.0">
  <class>ConfigureGameList</class>
-  <widget class="QWidget" name="ConfigureGeneral">
-    <property name="geometry">
-      <rect>
-        <x>0</x>
-        <y>0</y>
-        <width>300</width>
-        <height>377</height>
-      </rect>
-    </property>
-    <property name="windowTitle">
-      <string>Form</string>
-    </property>
-    <layout class="QHBoxLayout" name="HorizontalLayout">
-      <item>
-        <layout class="QVBoxLayout" name="VerticalLayout">
+ <widget class="QWidget" name="ConfigureGameList">
+  <property name="geometry">
+   <rect>
+    <x>0</x>
+    <y>0</y>
+    <width>300</width>
+    <height>377</height>
+   </rect>
+  </property>
+  <property name="windowTitle">
+   <string>Form</string>
+  </property>
+  <layout class="QHBoxLayout" name="HorizontalLayout">
+   <item>
+    <layout class="QVBoxLayout" name="VerticalLayout">
+     <item>
+      <widget class="QGroupBox" name="GeneralGroupBox">
+       <property name="title">
+        <string>General</string>
+       </property>
+       <layout class="QHBoxLayout" name="GeneralHorizontalLayout">
+        <item>
+         <layout class="QVBoxLayout" name="GeneralVerticalLayout">
           <item>
-            <widget class="QGroupBox" name="GeneralGroupBox">
-              <property name="title">
-                <string>General</string>
-              </property>
-              <layout class="QHBoxLayout" name="GeneralHorizontalLayout">
-                <item>
-                  <layout class="QVBoxLayout" name="GeneralVerticalLayout">
-                    <item>
-                      <widget class="QCheckBox" name="show_unknown">
-                        <property name="text">
-                          <string>Show files with type 'Unknown'</string>
-                        </property>
-                      </widget>
-                    </item>
-                  </layout>
-                </item>
-              </layout>
-            </widget>
+           <widget class="QCheckBox" name="show_unknown">
+            <property name="text">
+             <string>Show files with type 'Unknown'</string>
+            </property>
+           </widget>
           </item>
           <item>
-            <widget class="QGroupBox" name="IconSizeGroupBox">
-              <property name="title">
-                <string>Icon Size</string>
-              </property>
-              <layout class="QHBoxLayout" name="icon_size_qhbox_layout">
-                <item>
-                  <layout class="QVBoxLayout" name="icon_size_qvbox_layout">
-                    <item>
-                      <layout class="QHBoxLayout" name="icon_size_qhbox_layout_2">
-                        <item>
-                          <widget class="QLabel" name="icon_size_label">
-                            <property name="text">
-                              <string>Icon Size:</string>
-                            </property>
-                          </widget>
-                        </item>
-                        <item>
-                          <widget class="QComboBox" name="icon_size_combobox"/>
-                        </item>
-                      </layout>
-                    </item>
-                  </layout>
-                </item>
-              </layout>
-            </widget>
+           <widget class="QCheckBox" name="show_add_ons">
+            <property name="text">
+             <string>Show Add-Ons Column</string>
+            </property>
+           </widget>
           </item>
+         </layout>
+        </item>
+       </layout>
+      </widget>
+     </item>
+     <item>
+      <widget class="QGroupBox" name="IconSizeGroupBox">
+       <property name="title">
+        <string>Icon Size</string>
+       </property>
+       <layout class="QHBoxLayout" name="icon_size_qhbox_layout">
+        <item>
+         <layout class="QVBoxLayout" name="icon_size_qvbox_layout">
           <item>
-            <widget class="QGroupBox" name="RowGroupBox">
-              <property name="title">
-                <string>Row Text</string>
+           <layout class="QHBoxLayout" name="icon_size_qhbox_layout_2">
+            <item>
+             <widget class="QLabel" name="icon_size_label">
+              <property name="text">
+               <string>Icon Size:</string>
               </property>
-              <layout class="QHBoxLayout" name="RowHorizontalLayout">
-                <item>
-                  <layout class="QVBoxLayout" name="RowVerticalLayout">
-                    <item>
-                      <layout class="QHBoxLayout" name="row_1_qhbox_layout">
-                        <item>
-                          <widget class="QLabel" name="row_1_label">
-                            <property name="text">
-                              <string>Row 1 Text:</string>
-                            </property>
-                          </widget>
-                        </item>
-                        <item>
-                          <widget class="QComboBox" name="row_1_text_combobox"/>
-                        </item>
-                      </layout>
-                    </item>
-                    <item>
-                      <layout class="QHBoxLayout" name="row_2_qhbox_layout">
-                        <item>
-                          <widget class="QLabel" name="row_2_label">
-                            <property name="text">
-                              <string>Row 2 Text:</string>
-                            </property>
-                          </widget>
-                        </item>
-                        <item>
-                          <widget class="QComboBox" name="row_2_text_combobox"/>
-                        </item>
-                      </layout>
-                    </item>
-                  </layout>
-                </item>
-              </layout>
-            </widget>
+             </widget>
+            </item>
+            <item>
+             <widget class="QComboBox" name="icon_size_combobox"/>
+            </item>
+           </layout>
           </item>
+         </layout>
+        </item>
+       </layout>
+      </widget>
+     </item>
+     <item>
+      <widget class="QGroupBox" name="RowGroupBox">
+       <property name="title">
+        <string>Row Text</string>
+       </property>
+       <layout class="QHBoxLayout" name="RowHorizontalLayout">
+        <item>
+         <layout class="QVBoxLayout" name="RowVerticalLayout">
           <item>
-            <spacer name="verticalSpacer">
-              <property name="orientation">
-                <enum>Qt::Vertical</enum>
+           <layout class="QHBoxLayout" name="row_1_qhbox_layout">
+            <item>
+             <widget class="QLabel" name="row_1_label">
+              <property name="text">
+               <string>Row 1 Text:</string>
               </property>
-              <property name="sizeHint" stdset="0">
-                <size>
-                  <width>20</width>
-                  <height>40</height>
-                </size>
+             </widget>
+            </item>
+            <item>
+             <widget class="QComboBox" name="row_1_text_combobox"/>
+            </item>
+           </layout>
+          </item>
+          <item>
+           <layout class="QHBoxLayout" name="row_2_qhbox_layout">
+            <item>
+             <widget class="QLabel" name="row_2_label">
+              <property name="text">
+               <string>Row 2 Text:</string>
               </property>
-            </spacer>
+             </widget>
+            </item>
+            <item>
+             <widget class="QComboBox" name="row_2_text_combobox"/>
+            </item>
+           </layout>
           </item>
-        </layout>
-      </item>
+         </layout>
+        </item>
+       </layout>
+      </widget>
+     </item>
+     <item>
+      <spacer name="verticalSpacer">
+       <property name="orientation">
+        <enum>Qt::Vertical</enum>
+       </property>
+       <property name="sizeHint" stdset="0">
+        <size>
+         <width>20</width>
+         <height>40</height>
+        </size>
+       </property>
+      </spacer>
+     </item>
     </layout>
-  </widget>
+   </item>
+  </layout>
+ </widget>
  <resources/>
  <connections/>
 </ui>
diff --git a/src/yuzu/configuration/configure_general.cpp b/src/yuzu/configuration/configure_general.cpp
index f5db9e55b..c22742007 100644
--- a/src/yuzu/configuration/configure_general.cpp
+++ b/src/yuzu/configuration/configure_general.cpp
@@ -3,6 +3,10 @@
 // Refer to the license.txt file included.
 
 #include "core/core.h"
+#include "core/hle/service/am/am.h"
+#include "core/hle/service/am/applet_ae.h"
+#include "core/hle/service/am/applet_oe.h"
+#include "core/hle/service/sm/sm.h"
 #include "core/settings.h"
 #include "ui_configure_general.h"
 #include "yuzu/configuration/configure_general.h"
@@ -19,8 +23,10 @@ ConfigureGeneral::ConfigureGeneral(QWidget* parent)
 
     this->setConfiguration();
 
+    connect(ui->toggle_deepscan, &QCheckBox::stateChanged, this,
+            [] { UISettings::values.is_game_list_reload_pending.exchange(true); });
+
     ui->use_cpu_jit->setEnabled(!Core::System::GetInstance().IsPoweredOn());
-    ui->use_docked_mode->setEnabled(!Core::System::GetInstance().IsPoweredOn());
 }
 
 ConfigureGeneral::~ConfigureGeneral() = default;
@@ -31,12 +37,40 @@ void ConfigureGeneral::setConfiguration() {
     ui->theme_combobox->setCurrentIndex(ui->theme_combobox->findData(UISettings::values.theme));
     ui->use_cpu_jit->setChecked(Settings::values.use_cpu_jit);
     ui->use_docked_mode->setChecked(Settings::values.use_docked_mode);
+    ui->enable_nfc->setChecked(Settings::values.enable_nfc);
 }
 
 void ConfigureGeneral::PopulateHotkeyList(const HotkeyRegistry& registry) {
     ui->widget->Populate(registry);
 }
 
+void ConfigureGeneral::OnDockedModeChanged(bool last_state, bool new_state) {
+    if (last_state == new_state) {
+        return;
+    }
+
+    Core::System& system{Core::System::GetInstance()};
+    if (!system.IsPoweredOn()) {
+        return;
+    }
+    Service::SM::ServiceManager& sm = system.ServiceManager();
+
+    // Message queue is shared between these services, we just need to signal an operation
+    // change to one and it will handle both automatically
+    auto applet_oe = sm.GetService<Service::AM::AppletOE>("appletOE");
+    auto applet_ae = sm.GetService<Service::AM::AppletAE>("appletAE");
+    bool has_signalled = false;
+
+    if (applet_oe != nullptr) {
+        applet_oe->GetMessageQueue()->OperationModeChanged();
+        has_signalled = true;
+    }
+
+    if (applet_ae != nullptr && !has_signalled) {
+        applet_ae->GetMessageQueue()->OperationModeChanged();
+    }
+}
+
 void ConfigureGeneral::applyConfiguration() {
     UISettings::values.gamedir_deepscan = ui->toggle_deepscan->isChecked();
     UISettings::values.confirm_before_closing = ui->toggle_check_exit->isChecked();
@@ -44,5 +78,9 @@ void ConfigureGeneral::applyConfiguration() {
         ui->theme_combobox->itemData(ui->theme_combobox->currentIndex()).toString();
 
     Settings::values.use_cpu_jit = ui->use_cpu_jit->isChecked();
+    const bool pre_docked_mode = Settings::values.use_docked_mode;
     Settings::values.use_docked_mode = ui->use_docked_mode->isChecked();
+    OnDockedModeChanged(pre_docked_mode, Settings::values.use_docked_mode);
+
+    Settings::values.enable_nfc = ui->enable_nfc->isChecked();
 }
diff --git a/src/yuzu/configuration/configure_general.h b/src/yuzu/configuration/configure_general.h
index 4770034cc..2210d48da 100644
--- a/src/yuzu/configuration/configure_general.h
+++ b/src/yuzu/configuration/configure_general.h
@@ -25,6 +25,7 @@ public:
 
 private:
     void setConfiguration();
+    void OnDockedModeChanged(bool last_state, bool new_state);
 
     std::unique_ptr<Ui::ConfigureGeneral> ui;
 };
diff --git a/src/yuzu/configuration/configure_general.ui b/src/yuzu/configuration/configure_general.ui
index 1775c4d40..b82fffde8 100644
--- a/src/yuzu/configuration/configure_general.ui
+++ b/src/yuzu/configuration/configure_general.ui
@@ -68,19 +68,26 @@
        <property name="title">
         <string>Emulation</string>
        </property>
-        <layout class="QHBoxLayout" name="EmulationHorizontalLayout">
+       <layout class="QHBoxLayout" name="EmulationHorizontalLayout">
+        <item>
+         <layout class="QVBoxLayout" name="EmulationVerticalLayout">
+          <item>
+           <widget class="QCheckBox" name="use_docked_mode">
+            <property name="text">
+             <string>Enable docked mode</string>
+            </property>
+           </widget>
+          </item>
           <item>
-            <layout class="QVBoxLayout" name="EmulationVerticalLayout">
-              <item>
-                <widget class="QCheckBox" name="use_docked_mode">
-                  <property name="text">
-                    <string>Enable docked mode</string>
-                  </property>
-                </widget>
-              </item>
-            </layout>
+           <widget class="QCheckBox" name="enable_nfc">
+            <property name="text">
+             <string>Enable NFC</string>
+            </property>
+           </widget>
           </item>
-        </layout>
+         </layout>
+        </item>
+       </layout>
       </widget>
      </item>
      <item>
diff --git a/src/yuzu/configuration/configure_input.cpp b/src/yuzu/configuration/configure_input.cpp
index 94789c064..42a7beac6 100644
--- a/src/yuzu/configuration/configure_input.cpp
+++ b/src/yuzu/configuration/configure_input.cpp
@@ -322,7 +322,7 @@ void ConfigureInput::setPollingResult(const Common::ParamPackage& params, bool a
     }
 
     updateButtonLabels();
-    input_setter = boost::none;
+    input_setter = {};
 }
 
 void ConfigureInput::keyPressEvent(QKeyEvent* event) {
diff --git a/src/yuzu/configuration/configure_input.h b/src/yuzu/configuration/configure_input.h
index d1198db81..32c7183f9 100644
--- a/src/yuzu/configuration/configure_input.h
+++ b/src/yuzu/configuration/configure_input.h
@@ -7,11 +7,13 @@
 #include <array>
 #include <functional>
 #include <memory>
+#include <optional>
 #include <string>
 #include <unordered_map>
+
 #include <QKeyEvent>
 #include <QWidget>
-#include <boost/optional.hpp>
+
 #include "common/param_package.h"
 #include "core/settings.h"
 #include "input_common/main.h"
@@ -41,7 +43,7 @@ private:
     std::unique_ptr<QTimer> poll_timer;
 
     /// This will be the the setting function when an input is awaiting configuration.
-    boost::optional<std::function<void(const Common::ParamPackage&)>> input_setter;
+    std::optional<std::function<void(const Common::ParamPackage&)>> input_setter;
 
     std::array<Common::ParamPackage, Settings::NativeButton::NumButtons> buttons_param;
     std::array<Common::ParamPackage, Settings::NativeAnalog::NumAnalogs> analogs_param;
diff --git a/src/yuzu/configuration/configure_system.cpp b/src/yuzu/configuration/configure_system.cpp
index e9ed9c38f..ab5d46492 100644
--- a/src/yuzu/configuration/configure_system.cpp
+++ b/src/yuzu/configuration/configure_system.cpp
@@ -2,14 +2,27 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <algorithm>
+#include <QFileDialog>
+#include <QGraphicsItem>
+#include <QGraphicsScene>
+#include <QHeaderView>
 #include <QMessageBox>
+#include <QStandardItemModel>
+#include <QTreeView>
+#include <QVBoxLayout>
+#include "common/assert.h"
+#include "common/file_util.h"
+#include "common/string_util.h"
 #include "core/core.h"
+#include "core/hle/service/acc/profile_manager.h"
 #include "core/settings.h"
 #include "ui_configure_system.h"
 #include "yuzu/configuration/configure_system.h"
-#include "yuzu/main.h"
+#include "yuzu/util/limitable_input_dialog.h"
 
-static const std::array<int, 12> days_in_month = {{
+namespace {
+constexpr std::array<int, 12> days_in_month = {{
     31,
     29,
     31,
@@ -24,13 +37,114 @@ static const std::array<int, 12> days_in_month = {{
     31,
 }};
 
-ConfigureSystem::ConfigureSystem(QWidget* parent) : QWidget(parent), ui(new Ui::ConfigureSystem) {
+// Same backup JPEG used by acc IProfile::GetImage if no jpeg found
+constexpr std::array<u8, 107> backup_jpeg{
+    0xff, 0xd8, 0xff, 0xdb, 0x00, 0x43, 0x00, 0x03, 0x02, 0x02, 0x02, 0x02, 0x02, 0x03, 0x02, 0x02,
+    0x02, 0x03, 0x03, 0x03, 0x03, 0x04, 0x06, 0x04, 0x04, 0x04, 0x04, 0x04, 0x08, 0x06, 0x06, 0x05,
+    0x06, 0x09, 0x08, 0x0a, 0x0a, 0x09, 0x08, 0x09, 0x09, 0x0a, 0x0c, 0x0f, 0x0c, 0x0a, 0x0b, 0x0e,
+    0x0b, 0x09, 0x09, 0x0d, 0x11, 0x0d, 0x0e, 0x0f, 0x10, 0x10, 0x11, 0x10, 0x0a, 0x0c, 0x12, 0x13,
+    0x12, 0x10, 0x13, 0x0f, 0x10, 0x10, 0x10, 0xff, 0xc9, 0x00, 0x0b, 0x08, 0x00, 0x01, 0x00, 0x01,
+    0x01, 0x01, 0x11, 0x00, 0xff, 0xcc, 0x00, 0x06, 0x00, 0x10, 0x10, 0x05, 0xff, 0xda, 0x00, 0x08,
+    0x01, 0x01, 0x00, 0x00, 0x3f, 0x00, 0xd2, 0xcf, 0x20, 0xff, 0xd9,
+};
+
+QString GetImagePath(Service::Account::UUID uuid) {
+    const auto path = FileUtil::GetUserPath(FileUtil::UserPath::NANDDir) +
+                      "/system/save/8000000000000010/su/avators/" + uuid.FormatSwitch() + ".jpg";
+    return QString::fromStdString(path);
+}
+
+QString GetAccountUsername(const Service::Account::ProfileManager& manager,
+                           Service::Account::UUID uuid) {
+    Service::Account::ProfileBase profile;
+    if (!manager.GetProfileBase(uuid, profile)) {
+        return {};
+    }
+
+    const auto text = Common::StringFromFixedZeroTerminatedBuffer(
+        reinterpret_cast<const char*>(profile.username.data()), profile.username.size());
+    return QString::fromStdString(text);
+}
+
+QString FormatUserEntryText(const QString& username, Service::Account::UUID uuid) {
+    return ConfigureSystem::tr("%1\n%2",
+                               "%1 is the profile username, %2 is the formatted UUID (e.g. "
+                               "00112233-4455-6677-8899-AABBCCDDEEFF))")
+        .arg(username, QString::fromStdString(uuid.FormatSwitch()));
+}
+
+QPixmap GetIcon(Service::Account::UUID uuid) {
+    QPixmap icon{GetImagePath(uuid)};
+
+    if (!icon) {
+        icon.fill(Qt::black);
+        icon.loadFromData(backup_jpeg.data(), static_cast<u32>(backup_jpeg.size()));
+    }
+
+    return icon.scaled(64, 64, Qt::IgnoreAspectRatio, Qt::SmoothTransformation);
+}
+
+QString GetProfileUsernameFromUser(QWidget* parent, const QString& description_text) {
+    return LimitableInputDialog::GetText(parent, ConfigureSystem::tr("Enter Username"),
+                                         description_text, 1,
+                                         static_cast<int>(Service::Account::profile_username_size));
+}
+} // Anonymous namespace
+
+ConfigureSystem::ConfigureSystem(QWidget* parent)
+    : QWidget(parent), ui(new Ui::ConfigureSystem),
+      profile_manager(std::make_unique<Service::Account::ProfileManager>()) {
     ui->setupUi(this);
     connect(ui->combo_birthmonth,
             static_cast<void (QComboBox::*)(int)>(&QComboBox::currentIndexChanged), this,
-            &ConfigureSystem::updateBirthdayComboBox);
+            &ConfigureSystem::UpdateBirthdayComboBox);
     connect(ui->button_regenerate_console_id, &QPushButton::clicked, this,
-            &ConfigureSystem::refreshConsoleID);
+            &ConfigureSystem::RefreshConsoleID);
+
+    layout = new QVBoxLayout;
+    tree_view = new QTreeView;
+    item_model = new QStandardItemModel(tree_view);
+    tree_view->setModel(item_model);
+
+    tree_view->setAlternatingRowColors(true);
+    tree_view->setSelectionMode(QHeaderView::SingleSelection);
+    tree_view->setSelectionBehavior(QHeaderView::SelectRows);
+    tree_view->setVerticalScrollMode(QHeaderView::ScrollPerPixel);
+    tree_view->setHorizontalScrollMode(QHeaderView::ScrollPerPixel);
+    tree_view->setSortingEnabled(true);
+    tree_view->setEditTriggers(QHeaderView::NoEditTriggers);
+    tree_view->setUniformRowHeights(true);
+    tree_view->setIconSize({64, 64});
+    tree_view->setContextMenuPolicy(Qt::NoContextMenu);
+
+    item_model->insertColumns(0, 1);
+    item_model->setHeaderData(0, Qt::Horizontal, "Users");
+
+    // We must register all custom types with the Qt Automoc system so that we are able to use it
+    // with signals/slots. In this case, QList falls under the umbrells of custom types.
+    qRegisterMetaType<QList<QStandardItem*>>("QList<QStandardItem*>");
+
+    layout->setContentsMargins(0, 0, 0, 0);
+    layout->setSpacing(0);
+    layout->addWidget(tree_view);
+
+    ui->scrollArea->setLayout(layout);
+
+    connect(tree_view, &QTreeView::clicked, this, &ConfigureSystem::SelectUser);
+
+    connect(ui->pm_add, &QPushButton::pressed, this, &ConfigureSystem::AddUser);
+    connect(ui->pm_rename, &QPushButton::pressed, this, &ConfigureSystem::RenameUser);
+    connect(ui->pm_remove, &QPushButton::pressed, this, &ConfigureSystem::DeleteUser);
+    connect(ui->pm_set_image, &QPushButton::pressed, this, &ConfigureSystem::SetUserImage);
+
+    connect(ui->rng_seed_checkbox, &QCheckBox::stateChanged, this, [this](bool checked) {
+        ui->rng_seed_edit->setEnabled(checked);
+        if (!checked)
+            ui->rng_seed_edit->setText(QStringLiteral("00000000"));
+    });
+
+    scene = new QGraphicsScene;
+    ui->current_user_icon->setScene(scene);
 
     this->setConfiguration();
 }
@@ -39,8 +153,52 @@ ConfigureSystem::~ConfigureSystem() = default;
 
 void ConfigureSystem::setConfiguration() {
     enabled = !Core::System::GetInstance().IsPoweredOn();
-    ui->edit_username->setText(QString::fromStdString(Settings::values.username));
+
     ui->combo_language->setCurrentIndex(Settings::values.language_index);
+
+    item_model->removeRows(0, item_model->rowCount());
+    list_items.clear();
+
+    PopulateUserList();
+    UpdateCurrentUser();
+
+    ui->rng_seed_checkbox->setChecked(Settings::values.rng_seed.has_value());
+    ui->rng_seed_edit->setEnabled(Settings::values.rng_seed.has_value());
+
+    const auto rng_seed =
+        QString("%1").arg(Settings::values.rng_seed.value_or(0), 8, 16, QLatin1Char{'0'}).toUpper();
+    ui->rng_seed_edit->setText(rng_seed);
+}
+
+void ConfigureSystem::PopulateUserList() {
+    const auto& profiles = profile_manager->GetAllUsers();
+    for (const auto& user : profiles) {
+        Service::Account::ProfileBase profile;
+        if (!profile_manager->GetProfileBase(user, profile))
+            continue;
+
+        const auto username = Common::StringFromFixedZeroTerminatedBuffer(
+            reinterpret_cast<const char*>(profile.username.data()), profile.username.size());
+
+        list_items.push_back(QList<QStandardItem*>{new QStandardItem{
+            GetIcon(user), FormatUserEntryText(QString::fromStdString(username), user)}});
+    }
+
+    for (const auto& item : list_items)
+        item_model->appendRow(item);
+}
+
+void ConfigureSystem::UpdateCurrentUser() {
+    ui->pm_add->setEnabled(profile_manager->GetUserCount() < Service::Account::MAX_USERS);
+
+    const auto& current_user = profile_manager->GetUser(Settings::values.current_user);
+    ASSERT(current_user);
+    const auto username = GetAccountUsername(*profile_manager, *current_user);
+
+    scene->clear();
+    scene->addPixmap(
+        GetIcon(*current_user).scaled(48, 48, Qt::IgnoreAspectRatio, Qt::SmoothTransformation));
+    ui->current_user_username->setText(username);
 }
 
 void ConfigureSystem::ReadSystemSettings() {}
@@ -48,12 +206,18 @@ void ConfigureSystem::ReadSystemSettings() {}
 void ConfigureSystem::applyConfiguration() {
     if (!enabled)
         return;
-    Settings::values.username = ui->edit_username->text().toStdString();
+
     Settings::values.language_index = ui->combo_language->currentIndex();
+
+    if (ui->rng_seed_checkbox->isChecked())
+        Settings::values.rng_seed = ui->rng_seed_edit->text().toULongLong(nullptr, 16);
+    else
+        Settings::values.rng_seed = std::nullopt;
+
     Settings::Apply();
 }
 
-void ConfigureSystem::updateBirthdayComboBox(int birthmonth_index) {
+void ConfigureSystem::UpdateBirthdayComboBox(int birthmonth_index) {
     if (birthmonth_index < 0 || birthmonth_index >= 12)
         return;
 
@@ -78,7 +242,7 @@ void ConfigureSystem::updateBirthdayComboBox(int birthmonth_index) {
     ui->combo_birthday->setCurrentIndex(birthday_index);
 }
 
-void ConfigureSystem::refreshConsoleID() {
+void ConfigureSystem::RefreshConsoleID() {
     QMessageBox::StandardButton reply;
     QString warning_text = tr("This will replace your current virtual Switch with a new one. "
                               "Your current virtual Switch will not be recoverable. "
@@ -92,3 +256,130 @@ void ConfigureSystem::refreshConsoleID() {
     ui->label_console_id->setText(
         tr("Console ID: 0x%1").arg(QString::number(console_id, 16).toUpper()));
 }
+
+void ConfigureSystem::SelectUser(const QModelIndex& index) {
+    Settings::values.current_user =
+        std::clamp<s32>(index.row(), 0, static_cast<s32>(profile_manager->GetUserCount() - 1));
+
+    UpdateCurrentUser();
+
+    ui->pm_remove->setEnabled(profile_manager->GetUserCount() >= 2);
+    ui->pm_rename->setEnabled(true);
+    ui->pm_set_image->setEnabled(true);
+}
+
+void ConfigureSystem::AddUser() {
+    const auto username =
+        GetProfileUsernameFromUser(this, tr("Enter a username for the new user:"));
+    if (username.isEmpty()) {
+        return;
+    }
+
+    const auto uuid = Service::Account::UUID::Generate();
+    profile_manager->CreateNewUser(uuid, username.toStdString());
+
+    item_model->appendRow(new QStandardItem{GetIcon(uuid), FormatUserEntryText(username, uuid)});
+}
+
+void ConfigureSystem::RenameUser() {
+    const auto user = tree_view->currentIndex().row();
+    const auto uuid = profile_manager->GetUser(user);
+    ASSERT(uuid);
+
+    Service::Account::ProfileBase profile;
+    if (!profile_manager->GetProfileBase(*uuid, profile))
+        return;
+
+    const auto new_username = GetProfileUsernameFromUser(this, tr("Enter a new username:"));
+    if (new_username.isEmpty()) {
+        return;
+    }
+
+    const auto username_std = new_username.toStdString();
+    std::fill(profile.username.begin(), profile.username.end(), '\0');
+    std::copy(username_std.begin(), username_std.end(), profile.username.begin());
+
+    profile_manager->SetProfileBase(*uuid, profile);
+
+    item_model->setItem(
+        user, 0,
+        new QStandardItem{GetIcon(*uuid),
+                          FormatUserEntryText(QString::fromStdString(username_std), *uuid)});
+    UpdateCurrentUser();
+}
+
+void ConfigureSystem::DeleteUser() {
+    const auto index = tree_view->currentIndex().row();
+    const auto uuid = profile_manager->GetUser(index);
+    ASSERT(uuid);
+    const auto username = GetAccountUsername(*profile_manager, *uuid);
+
+    const auto confirm = QMessageBox::question(
+        this, tr("Confirm Delete"),
+        tr("You are about to delete user with name \"%1\". Are you sure?").arg(username));
+
+    if (confirm == QMessageBox::No)
+        return;
+
+    if (Settings::values.current_user == tree_view->currentIndex().row())
+        Settings::values.current_user = 0;
+    UpdateCurrentUser();
+
+    if (!profile_manager->RemoveUser(*uuid))
+        return;
+
+    item_model->removeRows(tree_view->currentIndex().row(), 1);
+    tree_view->clearSelection();
+
+    ui->pm_remove->setEnabled(false);
+    ui->pm_rename->setEnabled(false);
+}
+
+void ConfigureSystem::SetUserImage() {
+    const auto index = tree_view->currentIndex().row();
+    const auto uuid = profile_manager->GetUser(index);
+    ASSERT(uuid);
+
+    const auto file = QFileDialog::getOpenFileName(this, tr("Select User Image"), QString(),
+                                                   tr("JPEG Images (*.jpg *.jpeg)"));
+
+    if (file.isEmpty()) {
+        return;
+    }
+
+    const auto image_path = GetImagePath(*uuid);
+    if (QFile::exists(image_path) && !QFile::remove(image_path)) {
+        QMessageBox::warning(
+            this, tr("Error deleting image"),
+            tr("Error occurred attempting to overwrite previous image at: %1.").arg(image_path));
+        return;
+    }
+
+    const auto raw_path = QString::fromStdString(
+        FileUtil::GetUserPath(FileUtil::UserPath::NANDDir) + "/system/save/8000000000000010");
+    const QFileInfo raw_info{raw_path};
+    if (raw_info.exists() && !raw_info.isDir() && !QFile::remove(raw_path)) {
+        QMessageBox::warning(this, tr("Error deleting file"),
+                             tr("Unable to delete existing file: %1.").arg(raw_path));
+        return;
+    }
+
+    const QString absolute_dst_path = QFileInfo{image_path}.absolutePath();
+    if (!QDir{raw_path}.mkpath(absolute_dst_path)) {
+        QMessageBox::warning(
+            this, tr("Error creating user image directory"),
+            tr("Unable to create directory %1 for storing user images.").arg(absolute_dst_path));
+        return;
+    }
+
+    if (!QFile::copy(file, image_path)) {
+        QMessageBox::warning(this, tr("Error copying user image"),
+                             tr("Unable to copy image from %1 to %2").arg(file, image_path));
+        return;
+    }
+
+    const auto username = GetAccountUsername(*profile_manager, *uuid);
+    item_model->setItem(index, 0,
+                        new QStandardItem{GetIcon(*uuid), FormatUserEntryText(username, *uuid)});
+    UpdateCurrentUser();
+}
diff --git a/src/yuzu/configuration/configure_system.h b/src/yuzu/configuration/configure_system.h
index f13de17d4..07764e1f7 100644
--- a/src/yuzu/configuration/configure_system.h
+++ b/src/yuzu/configuration/configure_system.h
@@ -5,8 +5,20 @@
 #pragma once
 
 #include <memory>
+
+#include <QList>
 #include <QWidget>
 
+class QGraphicsScene;
+class QStandardItem;
+class QStandardItemModel;
+class QTreeView;
+class QVBoxLayout;
+
+namespace Service::Account {
+class ProfileManager;
+}
+
 namespace Ui {
 class ConfigureSystem;
 }
@@ -16,23 +28,39 @@ class ConfigureSystem : public QWidget {
 
 public:
     explicit ConfigureSystem(QWidget* parent = nullptr);
-    ~ConfigureSystem();
+    ~ConfigureSystem() override;
 
     void applyConfiguration();
     void setConfiguration();
 
-public slots:
-    void updateBirthdayComboBox(int birthmonth_index);
-    void refreshConsoleID();
-
 private:
     void ReadSystemSettings();
 
+    void UpdateBirthdayComboBox(int birthmonth_index);
+    void RefreshConsoleID();
+
+    void PopulateUserList();
+    void UpdateCurrentUser();
+    void SelectUser(const QModelIndex& index);
+    void AddUser();
+    void RenameUser();
+    void DeleteUser();
+    void SetUserImage();
+
+    QVBoxLayout* layout;
+    QTreeView* tree_view;
+    QStandardItemModel* item_model;
+    QGraphicsScene* scene;
+
+    std::vector<QList<QStandardItem*>> list_items;
+
     std::unique_ptr<Ui::ConfigureSystem> ui;
-    bool enabled;
+    bool enabled = false;
+
+    int birthmonth = 0;
+    int birthday = 0;
+    int language_index = 0;
+    int sound_index = 0;
 
-    std::u16string username;
-    int birthmonth, birthday;
-    int language_index;
-    int sound_index;
+    std::unique_ptr<Service::Account::ProfileManager> profile_manager;
 };
diff --git a/src/yuzu/configuration/configure_system.ui b/src/yuzu/configuration/configure_system.ui
index f3f8db038..a91580893 100644
--- a/src/yuzu/configuration/configure_system.ui
+++ b/src/yuzu/configuration/configure_system.ui
@@ -6,8 +6,8 @@
    <rect>
     <x>0</x>
     <y>0</y>
-    <width>360</width>
-    <height>377</height>
+    <width>366</width>
+    <height>483</height>
    </rect>
   </property>
   <property name="windowTitle">
@@ -22,34 +22,120 @@
         <string>System Settings</string>
        </property>
        <layout class="QGridLayout" name="gridLayout">
-        <item row="0" column="0">
-         <widget class="QLabel" name="label_username">
-          <property name="text">
-           <string>Username</string>
+        <item row="1" column="1">
+         <widget class="QComboBox" name="combo_language">
+          <property name="toolTip">
+           <string>Note: this can be overridden when region setting is auto-select</string>
           </property>
+          <item>
+           <property name="text">
+            <string>Japanese (日本語)</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>English</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>French (français)</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>German (Deutsch)</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Italian (italiano)</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Spanish (español)</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Chinese</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Korean (한국어)</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Dutch (Nederlands)</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Portuguese (português)</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Russian (Русский)</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Taiwanese</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>British English</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Canadian French</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Latin American Spanish</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Simplified Chinese</string>
+           </property>
+          </item>
+          <item>
+           <property name="text">
+            <string>Traditional Chinese (正體中文)</string>
+           </property>
+          </item>
          </widget>
         </item>
-        <item row="0" column="1">
-         <widget class="QLineEdit" name="edit_username">
-          <property name="sizePolicy">
-           <sizepolicy hsizetype="Preferred" vsizetype="Fixed">
-            <horstretch>0</horstretch>
-            <verstretch>0</verstretch>
-           </sizepolicy>
+        <item row="3" column="0">
+         <widget class="QLabel" name="label_console_id">
+          <property name="text">
+           <string>Console ID:</string>
           </property>
-          <property name="maxLength">
-           <number>32</number>
+         </widget>
+        </item>
+        <item row="2" column="0">
+         <widget class="QLabel" name="label_sound">
+          <property name="text">
+           <string>Sound output mode</string>
           </property>
          </widget>
         </item>
-        <item row="1" column="0">
+        <item row="0" column="0">
          <widget class="QLabel" name="label_birthday">
           <property name="text">
            <string>Birthday</string>
           </property>
          </widget>
         </item>
-        <item row="1" column="1">
+        <item row="0" column="1">
          <layout class="QHBoxLayout" name="horizontalLayout_birthday2">
           <item>
            <widget class="QComboBox" name="combo_birthmonth">
@@ -120,113 +206,23 @@
           </item>
          </layout>
         </item>
-        <item row="2" column="0">
-         <widget class="QLabel" name="label_language">
-          <property name="text">
-           <string>Language</string>
+        <item row="3" column="1">
+         <widget class="QPushButton" name="button_regenerate_console_id">
+          <property name="sizePolicy">
+           <sizepolicy hsizetype="Fixed" vsizetype="Fixed">
+            <horstretch>0</horstretch>
+            <verstretch>0</verstretch>
+           </sizepolicy>
           </property>
-         </widget>
-        </item>
-        <item row="2" column="1">
-         <widget class="QComboBox" name="combo_language">
-          <property name="toolTip">
-           <string>Note: this can be overridden when region setting is auto-select</string>
+          <property name="layoutDirection">
+           <enum>Qt::RightToLeft</enum>
           </property>
-          <item>
-           <property name="text">
-            <string>Japanese (日本語)</string>
-           </property>
-          </item>
-          <item>
-           <property name="text">
-            <string>English</string>
-           </property>
-          </item>
-          <item>
-           <property name="text">
-            <string>French (français)</string>
-           </property>
-          </item>
-          <item>
-           <property name="text">
-            <string>German (Deutsch)</string>
-           </property>
-          </item>
-          <item>
-           <property name="text">
-            <string>Italian (italiano)</string>
-           </property>
-          </item>
-          <item>
-           <property name="text">
-            <string>Spanish (español)</string>
-           </property>
-          </item>
-          <item>
-           <property name="text">
-            <string>Chinese</string>
-           </property>
-          </item>
-          <item>
-           <property name="text">
-            <string>Korean (한국어)</string>
-           </property>
-          </item>
-          <item>
-           <property name="text">
-            <string>Dutch (Nederlands)</string>
-           </property>
-          </item>
-          <item>
-           <property name="text">
-            <string>Portuguese (português)</string>
-           </property>
-          </item>
-          <item>
-           <property name="text">
-            <string>Russian (Русский)</string>
-           </property>
-          </item>
-           <item>
-             <property name="text">
-               <string>Taiwanese</string>
-             </property>
-           </item>
-           <item>
-             <property name="text">
-               <string>British English</string>
-             </property>
-           </item>
-           <item>
-             <property name="text">
-               <string>Canadian French</string>
-             </property>
-           </item>
-           <item>
-             <property name="text">
-               <string>Latin American Spanish</string>
-             </property>
-           </item>
-           <item>
-             <property name="text">
-               <string>Simplified Chinese</string>
-             </property>
-           </item>
-          <item>
-           <property name="text">
-            <string>Traditional Chinese (正體中文)</string>
-           </property>
-          </item>
-         </widget>
-        </item>
-        <item row="3" column="0">
-         <widget class="QLabel" name="label_sound">
           <property name="text">
-           <string>Sound output mode</string>
+           <string>Regenerate</string>
           </property>
          </widget>
         </item>
-        <item row="3" column="1">
+        <item row="2" column="1">
          <widget class="QComboBox" name="combo_sound">
           <item>
            <property name="text">
@@ -245,26 +241,38 @@
           </item>
          </widget>
         </item>
+        <item row="1" column="0">
+         <widget class="QLabel" name="label_language">
+          <property name="text">
+           <string>Language</string>
+          </property>
+         </widget>
+        </item>
         <item row="4" column="0">
-         <widget class="QLabel" name="label_console_id">
+         <widget class="QCheckBox" name="rng_seed_checkbox">
           <property name="text">
-           <string>Console ID:</string>
+           <string>RNG Seed</string>
           </property>
          </widget>
         </item>
         <item row="4" column="1">
-         <widget class="QPushButton" name="button_regenerate_console_id">
+         <widget class="QLineEdit" name="rng_seed_edit">
           <property name="sizePolicy">
-           <sizepolicy hsizetype="Fixed" vsizetype="Fixed">
+           <sizepolicy hsizetype="Minimum" vsizetype="Fixed">
             <horstretch>0</horstretch>
             <verstretch>0</verstretch>
            </sizepolicy>
           </property>
-          <property name="layoutDirection">
-           <enum>Qt::RightToLeft</enum>
+          <property name="font">
+           <font>
+            <family>Lucida Console</family>
+           </font>
           </property>
-          <property name="text">
-           <string>Regenerate</string>
+          <property name="inputMask">
+           <string notr="true">HHHHHHHH</string>
+          </property>
+          <property name="maxLength">
+           <number>8</number>
           </property>
          </widget>
         </item>
@@ -272,6 +280,143 @@
       </widget>
      </item>
      <item>
+      <widget class="QGroupBox" name="gridGroupBox">
+       <property name="title">
+        <string>Profile Manager</string>
+       </property>
+       <layout class="QGridLayout" name="gridLayout_2">
+        <property name="sizeConstraint">
+         <enum>QLayout::SetNoConstraint</enum>
+        </property>
+        <item row="0" column="0">
+         <layout class="QHBoxLayout" name="horizontalLayout_2">
+          <item>
+           <widget class="QLabel" name="label">
+            <property name="sizePolicy">
+             <sizepolicy hsizetype="Preferred" vsizetype="Minimum">
+              <horstretch>0</horstretch>
+              <verstretch>0</verstretch>
+             </sizepolicy>
+            </property>
+            <property name="text">
+             <string>Current User</string>
+            </property>
+           </widget>
+          </item>
+          <item>
+           <widget class="QGraphicsView" name="current_user_icon">
+            <property name="minimumSize">
+             <size>
+              <width>48</width>
+              <height>48</height>
+             </size>
+            </property>
+            <property name="maximumSize">
+             <size>
+              <width>48</width>
+              <height>48</height>
+             </size>
+            </property>
+            <property name="verticalScrollBarPolicy">
+             <enum>Qt::ScrollBarAlwaysOff</enum>
+            </property>
+            <property name="horizontalScrollBarPolicy">
+             <enum>Qt::ScrollBarAlwaysOff</enum>
+            </property>
+            <property name="interactive">
+             <bool>false</bool>
+            </property>
+           </widget>
+          </item>
+          <item>
+           <widget class="QLabel" name="current_user_username">
+            <property name="sizePolicy">
+             <sizepolicy hsizetype="Preferred" vsizetype="Minimum">
+              <horstretch>0</horstretch>
+              <verstretch>0</verstretch>
+             </sizepolicy>
+            </property>
+            <property name="text">
+             <string>Username</string>
+            </property>
+           </widget>
+          </item>
+         </layout>
+        </item>
+        <item row="1" column="0">
+         <widget class="QScrollArea" name="scrollArea">
+          <property name="sizePolicy">
+           <sizepolicy hsizetype="Expanding" vsizetype="Expanding">
+            <horstretch>0</horstretch>
+            <verstretch>0</verstretch>
+           </sizepolicy>
+          </property>
+          <property name="frameShape">
+           <enum>QFrame::StyledPanel</enum>
+          </property>
+          <property name="widgetResizable">
+           <bool>false</bool>
+          </property>
+         </widget>
+        </item>
+        <item row="2" column="0">
+         <layout class="QHBoxLayout" name="horizontalLayout_3">
+          <item>
+           <widget class="QPushButton" name="pm_set_image">
+            <property name="enabled">
+             <bool>false</bool>
+            </property>
+            <property name="text">
+             <string>Set Image</string>
+            </property>
+           </widget>
+          </item>
+          <item>
+           <spacer name="horizontalSpacer">
+            <property name="orientation">
+             <enum>Qt::Horizontal</enum>
+            </property>
+            <property name="sizeHint" stdset="0">
+             <size>
+              <width>40</width>
+              <height>20</height>
+             </size>
+            </property>
+           </spacer>
+          </item>
+          <item>
+           <widget class="QPushButton" name="pm_add">
+            <property name="text">
+             <string>Add</string>
+            </property>
+           </widget>
+          </item>
+          <item>
+           <widget class="QPushButton" name="pm_rename">
+            <property name="enabled">
+             <bool>false</bool>
+            </property>
+            <property name="text">
+             <string>Rename</string>
+            </property>
+           </widget>
+          </item>
+          <item>
+           <widget class="QPushButton" name="pm_remove">
+            <property name="enabled">
+             <bool>false</bool>
+            </property>
+            <property name="text">
+             <string>Remove</string>
+            </property>
+           </widget>
+          </item>
+         </layout>
+        </item>
+       </layout>
+      </widget>
+     </item>
+     <item>
       <widget class="QLabel" name="label_disable_info">
        <property name="text">
         <string>System settings are available only when game is not running.</string>
@@ -281,19 +426,6 @@
        </property>
       </widget>
      </item>
-     <item>
-      <spacer name="verticalSpacer">
-       <property name="orientation">
-        <enum>Qt::Vertical</enum>
-       </property>
-       <property name="sizeHint" stdset="0">
-        <size>
-         <width>20</width>
-         <height>40</height>
-        </size>
-       </property>
-      </spacer>
-     </item>
     </layout>
    </item>
   </layout>
diff --git a/src/yuzu/debugger/graphics/graphics_breakpoints.cpp b/src/yuzu/debugger/graphics/graphics_breakpoints.cpp
index b5c88f944..67ed0ba6d 100644
--- a/src/yuzu/debugger/graphics/graphics_breakpoints.cpp
+++ b/src/yuzu/debugger/graphics/graphics_breakpoints.cpp
@@ -2,7 +2,6 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
-#include <map>
 #include <QLabel>
 #include <QMetaType>
 #include <QPushButton>
@@ -32,21 +31,8 @@ QVariant BreakPointModel::data(const QModelIndex& index, int role) const {
     switch (role) {
     case Qt::DisplayRole: {
         if (index.column() == 0) {
-            static const std::map<Tegra::DebugContext::Event, QString> map = {
-                {Tegra::DebugContext::Event::MaxwellCommandLoaded, tr("Maxwell command loaded")},
-                {Tegra::DebugContext::Event::MaxwellCommandProcessed,
-                 tr("Maxwell command processed")},
-                {Tegra::DebugContext::Event::IncomingPrimitiveBatch,
-                 tr("Incoming primitive batch")},
-                {Tegra::DebugContext::Event::FinishedPrimitiveBatch,
-                 tr("Finished primitive batch")},
-            };
-
-            DEBUG_ASSERT(map.size() ==
-                         static_cast<std::size_t>(Tegra::DebugContext::Event::NumEvents));
-            return (map.find(event) != map.end()) ? map.at(event) : QString();
+            return DebugContextEventToString(event);
         }
-
         break;
     }
 
@@ -128,6 +114,23 @@ void BreakPointModel::OnResumed() {
     active_breakpoint = context->active_breakpoint;
 }
 
+QString BreakPointModel::DebugContextEventToString(Tegra::DebugContext::Event event) {
+    switch (event) {
+    case Tegra::DebugContext::Event::MaxwellCommandLoaded:
+        return tr("Maxwell command loaded");
+    case Tegra::DebugContext::Event::MaxwellCommandProcessed:
+        return tr("Maxwell command processed");
+    case Tegra::DebugContext::Event::IncomingPrimitiveBatch:
+        return tr("Incoming primitive batch");
+    case Tegra::DebugContext::Event::FinishedPrimitiveBatch:
+        return tr("Finished primitive batch");
+    case Tegra::DebugContext::Event::NumEvents:
+        break;
+    }
+
+    return tr("Unknown debug context event");
+}
+
 GraphicsBreakPointsWidget::GraphicsBreakPointsWidget(
     std::shared_ptr<Tegra::DebugContext> debug_context, QWidget* parent)
     : QDockWidget(tr("Maxwell Breakpoints"), parent), Tegra::DebugContext::BreakPointObserver(
diff --git a/src/yuzu/debugger/graphics/graphics_breakpoints_p.h b/src/yuzu/debugger/graphics/graphics_breakpoints_p.h
index 7112b87e6..fb488e38f 100644
--- a/src/yuzu/debugger/graphics/graphics_breakpoints_p.h
+++ b/src/yuzu/debugger/graphics/graphics_breakpoints_p.h
@@ -29,6 +29,8 @@ public:
     void OnResumed();
 
 private:
+    static QString DebugContextEventToString(Tegra::DebugContext::Event event);
+
     std::weak_ptr<Tegra::DebugContext> context_weak;
     bool at_breakpoint;
     Tegra::DebugContext::Event active_breakpoint;
diff --git a/src/yuzu/debugger/graphics/graphics_surface.cpp b/src/yuzu/debugger/graphics/graphics_surface.cpp
index 44d423da2..707747422 100644
--- a/src/yuzu/debugger/graphics/graphics_surface.cpp
+++ b/src/yuzu/debugger/graphics/graphics_surface.cpp
@@ -382,13 +382,13 @@ void GraphicsSurfaceWidget::OnUpdate() {
     // TODO: Implement a good way to visualize alpha components!
 
     QImage decoded_image(surface_width, surface_height, QImage::Format_ARGB32);
-    boost::optional<VAddr> address = gpu.MemoryManager().GpuToCpuAddress(surface_address);
+    std::optional<VAddr> address = gpu.MemoryManager().GpuToCpuAddress(surface_address);
 
     // TODO(bunnei): Will not work with BCn formats that swizzle 4x4 tiles.
     // Needs to be fixed if we plan to use this feature more, otherwise we may remove it.
-    auto unswizzled_data =
-        Tegra::Texture::UnswizzleTexture(*address, 1, Tegra::Texture::BytesPerPixel(surface_format),
-                                         surface_width, surface_height, 1U);
+    auto unswizzled_data = Tegra::Texture::UnswizzleTexture(
+        *address, 1, 1, Tegra::Texture::BytesPerPixel(surface_format), surface_width,
+        surface_height, 1U);
 
     auto texture_data = Tegra::Texture::DecodeTexture(unswizzled_data, surface_format,
                                                       surface_width, surface_height);
@@ -444,7 +444,7 @@ void GraphicsSurfaceWidget::SaveSurface() {
             pixmap->save(&file, "PNG");
     } else if (selectedFilter == bin_filter) {
         auto& gpu = Core::System::GetInstance().GPU();
-        boost::optional<VAddr> address = gpu.MemoryManager().GpuToCpuAddress(surface_address);
+        std::optional<VAddr> address = gpu.MemoryManager().GpuToCpuAddress(surface_address);
 
         const u8* buffer = Memory::GetPointer(*address);
         ASSERT_MSG(buffer != nullptr, "Memory not accessible");
diff --git a/src/yuzu/debugger/wait_tree.h b/src/yuzu/debugger/wait_tree.h
index defbf734f..331f89885 100644
--- a/src/yuzu/debugger/wait_tree.h
+++ b/src/yuzu/debugger/wait_tree.h
@@ -11,7 +11,6 @@
 #include <QAbstractItemModel>
 #include <QDockWidget>
 #include <QTreeView>
-#include <boost/container/flat_set.hpp>
 #include "common/common_types.h"
 #include "core/hle/kernel/object.h"
 
diff --git a/src/yuzu/game_list.cpp b/src/yuzu/game_list.cpp
index 67890455a..11a8c390b 100644
--- a/src/yuzu/game_list.cpp
+++ b/src/yuzu/game_list.cpp
@@ -16,7 +16,6 @@
 #include <fmt/format.h>
 #include "common/common_paths.h"
 #include "common/common_types.h"
-#include "common/file_util.h"
 #include "common/logging/log.h"
 #include "core/file_sys/patch_manager.h"
 #include "yuzu/compatibility_list.h"
@@ -216,12 +215,18 @@ GameList::GameList(FileSys::VirtualFilesystem vfs, GMainWindow* parent)
     tree_view->setUniformRowHeights(true);
     tree_view->setContextMenuPolicy(Qt::CustomContextMenu);
 
-    item_model->insertColumns(0, COLUMN_COUNT);
-    item_model->setHeaderData(COLUMN_NAME, Qt::Horizontal, "Name");
-    item_model->setHeaderData(COLUMN_COMPATIBILITY, Qt::Horizontal, "Compatibility");
-    item_model->setHeaderData(COLUMN_ADD_ONS, Qt::Horizontal, "Add-ons");
-    item_model->setHeaderData(COLUMN_FILE_TYPE, Qt::Horizontal, "File type");
-    item_model->setHeaderData(COLUMN_SIZE, Qt::Horizontal, "Size");
+    item_model->insertColumns(0, UISettings::values.show_add_ons ? COLUMN_COUNT : COLUMN_COUNT - 1);
+    item_model->setHeaderData(COLUMN_NAME, Qt::Horizontal, tr("Name"));
+    item_model->setHeaderData(COLUMN_COMPATIBILITY, Qt::Horizontal, tr("Compatibility"));
+
+    if (UISettings::values.show_add_ons) {
+        item_model->setHeaderData(COLUMN_ADD_ONS, Qt::Horizontal, tr("Add-ons"));
+        item_model->setHeaderData(COLUMN_FILE_TYPE, Qt::Horizontal, tr("File type"));
+        item_model->setHeaderData(COLUMN_SIZE, Qt::Horizontal, tr("Size"));
+    } else {
+        item_model->setHeaderData(COLUMN_FILE_TYPE - 1, Qt::Horizontal, tr("File type"));
+        item_model->setHeaderData(COLUMN_SIZE - 1, Qt::Horizontal, tr("Size"));
+    }
 
     connect(tree_view, &QTreeView::activated, this, &GameList::ValidateEntry);
     connect(tree_view, &QTreeView::customContextMenuRequested, this, &GameList::PopupContextMenu);
@@ -387,14 +392,33 @@ void GameList::LoadCompatibilityList() {
 }
 
 void GameList::PopulateAsync(const QString& dir_path, bool deep_scan) {
-    if (!FileUtil::Exists(dir_path.toStdString()) ||
-        !FileUtil::IsDirectory(dir_path.toStdString())) {
-        LOG_ERROR(Frontend, "Could not find game list folder at {}", dir_path.toLocal8Bit().data());
+    const QFileInfo dir_info{dir_path};
+    if (!dir_info.exists() || !dir_info.isDir()) {
+        LOG_ERROR(Frontend, "Could not find game list folder at {}", dir_path.toStdString());
         search_field->setFilterResult(0, 0);
         return;
     }
 
     tree_view->setEnabled(false);
+
+    // Update the columns in case UISettings has changed
+    item_model->removeColumns(0, item_model->columnCount());
+    item_model->insertColumns(0, UISettings::values.show_add_ons ? COLUMN_COUNT : COLUMN_COUNT - 1);
+    item_model->setHeaderData(COLUMN_NAME, Qt::Horizontal, tr("Name"));
+    item_model->setHeaderData(COLUMN_COMPATIBILITY, Qt::Horizontal, tr("Compatibility"));
+
+    if (UISettings::values.show_add_ons) {
+        item_model->setHeaderData(COLUMN_ADD_ONS, Qt::Horizontal, tr("Add-ons"));
+        item_model->setHeaderData(COLUMN_FILE_TYPE, Qt::Horizontal, tr("File type"));
+        item_model->setHeaderData(COLUMN_SIZE, Qt::Horizontal, tr("Size"));
+    } else {
+        item_model->setHeaderData(COLUMN_FILE_TYPE - 1, Qt::Horizontal, tr("File type"));
+        item_model->setHeaderData(COLUMN_SIZE - 1, Qt::Horizontal, tr("Size"));
+        item_model->removeColumns(COLUMN_COUNT - 1, 1);
+    }
+
+    LoadInterfaceLayout();
+
     // Delete any rows that might already exist if we're repopulating
     item_model->removeRows(0, item_model->rowCount());
 
diff --git a/src/yuzu/game_list_worker.cpp b/src/yuzu/game_list_worker.cpp
index 3881aba5f..362902e46 100644
--- a/src/yuzu/game_list_worker.cpp
+++ b/src/yuzu/game_list_worker.cpp
@@ -62,19 +62,24 @@ QString FormatPatchNameVersions(const FileSys::PatchManager& patch_manager,
     FileSys::VirtualFile update_raw;
     loader.ReadUpdateRaw(update_raw);
     for (const auto& kv : patch_manager.GetPatchVersionNames(update_raw)) {
-        if (!updatable && kv.first == "Update")
+        const bool is_update = kv.first == "Update";
+        if (!updatable && is_update) {
             continue;
+        }
+
+        const QString type = QString::fromStdString(kv.first);
 
         if (kv.second.empty()) {
-            out.append(fmt::format("{}\n", kv.first).c_str());
+            out.append(QStringLiteral("%1\n").arg(type));
         } else {
             auto ver = kv.second;
 
             // Display container name for packed updates
-            if (ver == "PACKED" && kv.first == "Update")
+            if (is_update && ver == "PACKED") {
                 ver = Loader::GetFileTypeString(loader.GetFileType());
+            }
 
-            out.append(fmt::format("{} ({})\n", kv.first, ver).c_str());
+            out.append(QStringLiteral("%1 (%2)\n").arg(type, QString::fromStdString(ver)));
         }
     }
 
@@ -118,17 +123,22 @@ void GameListWorker::AddInstalledTitlesToGameList() {
         if (it != compatibility_list.end())
             compatibility = it->second.first;
 
-        emit EntryReady({
+        QList<QStandardItem*> list{
             new GameListItemPath(
                 FormatGameName(file->GetFullPath()), icon, QString::fromStdString(name),
                 QString::fromStdString(Loader::GetFileTypeString(loader->GetFileType())),
                 program_id),
             new GameListItemCompat(compatibility),
-            new GameListItem(FormatPatchNameVersions(patch, *loader)),
             new GameListItem(
                 QString::fromStdString(Loader::GetFileTypeString(loader->GetFileType()))),
             new GameListItemSize(file->GetSize()),
-        });
+        };
+
+        if (UISettings::values.show_add_ons) {
+            list.insert(2, new GameListItem(FormatPatchNameVersions(patch, *loader)));
+        }
+
+        emit EntryReady(list);
     }
 
     const auto control_data = cache->ListEntriesFilter(FileSys::TitleType::Application,
@@ -211,18 +221,23 @@ void GameListWorker::AddFstEntriesToGameList(const std::string& dir_path, unsign
             if (it != compatibility_list.end())
                 compatibility = it->second.first;
 
-            emit EntryReady({
+            QList<QStandardItem*> list{
                 new GameListItemPath(
                     FormatGameName(physical_name), icon, QString::fromStdString(name),
                     QString::fromStdString(Loader::GetFileTypeString(loader->GetFileType())),
                     program_id),
                 new GameListItemCompat(compatibility),
                 new GameListItem(
-                    FormatPatchNameVersions(patch, *loader, loader->IsRomFSUpdatable())),
-                new GameListItem(
                     QString::fromStdString(Loader::GetFileTypeString(loader->GetFileType()))),
                 new GameListItemSize(FileUtil::GetSize(physical_name)),
-            });
+            };
+
+            if (UISettings::values.show_add_ons) {
+                list.insert(2, new GameListItem(FormatPatchNameVersions(
+                                   patch, *loader, loader->IsRomFSUpdatable())));
+            }
+
+            emit EntryReady(std::move(list));
         } else if (is_dir && recursion > 0) {
             watch_list.append(QString::fromStdString(physical_name));
             AddFstEntriesToGameList(physical_name, recursion - 1);
diff --git a/src/yuzu/main.cpp b/src/yuzu/main.cpp
index bef9df00d..4b969119c 100644
--- a/src/yuzu/main.cpp
+++ b/src/yuzu/main.cpp
@@ -10,8 +10,9 @@
 // VFS includes must be before glad as they will conflict with Windows file api, which uses defines.
 #include "core/file_sys/vfs.h"
 #include "core/file_sys/vfs_real.h"
+#include "core/hle/service/acc/profile_manager.h"
 
-// These are wrappers to avoid the calls to CreateDirectory and CreateFile becuase of the Windows
+// These are wrappers to avoid the calls to CreateDirectory and CreateFile because of the Windows
 // defines.
 static FileSys::VirtualDir VfsFilesystemCreateDirectoryWrapper(
     const FileSys::VirtualFilesystem& vfs, const std::string& path, FileSys::Mode mode) {
@@ -29,6 +30,7 @@ static FileSys::VirtualFile VfsDirectoryCreateFileWrapper(const FileSys::Virtual
 #define QT_NO_OPENGL
 #include <QDesktopWidget>
 #include <QDialogButtonBox>
+#include <QFile>
 #include <QFileDialog>
 #include <QMessageBox>
 #include <QtConcurrent/QtConcurrent>
@@ -60,6 +62,8 @@ static FileSys::VirtualFile VfsDirectoryCreateFileWrapper(const FileSys::Virtual
 #include "core/hle/kernel/process.h"
 #include "core/hle/service/filesystem/filesystem.h"
 #include "core/hle/service/filesystem/fsp_ldr.h"
+#include "core/hle/service/nfp/nfp.h"
+#include "core/hle/service/sm/sm.h"
 #include "core/loader/loader.h"
 #include "core/perf_stats.h"
 #include "core/settings.h"
@@ -100,6 +104,8 @@ __declspec(dllexport) int AmdPowerXpressRequestHighPerformance = 1;
 }
 #endif
 
+constexpr u64 DLC_BASE_TITLE_ID_MASK = 0xFFFFFFFFFFFFE000;
+
 /**
  * "Callouts" are one-time instructional messages shown to the user. In the config settings, there
  * is a bitfield "callout_flags" options, used to track if a message has already been shown to the
@@ -136,6 +142,9 @@ static void InitializeLogging() {
     const std::string& log_dir = FileUtil::GetUserPath(FileUtil::UserPath::LogDir);
     FileUtil::CreateFullPath(log_dir);
     Log::AddBackend(std::make_unique<Log::FileBackend>(log_dir + LOG_FILE));
+#ifdef _WIN32
+    Log::AddBackend(std::make_unique<Log::DebuggerBackend>());
+#endif
 }
 
 GMainWindow::GMainWindow()
@@ -299,6 +308,8 @@ void GMainWindow::InitializeHotkeys() {
                                    Qt::ApplicationShortcut);
     hotkey_registry.RegisterHotkey("Main Window", "Decrease Speed Limit", QKeySequence("-"),
                                    Qt::ApplicationShortcut);
+    hotkey_registry.RegisterHotkey("Main Window", "Load Amiibo", QKeySequence(Qt::Key_F2),
+                                   Qt::ApplicationShortcut);
     hotkey_registry.LoadHotkeys();
 
     connect(hotkey_registry.GetHotkey("Main Window", "Load File", this), &QShortcut::activated,
@@ -352,6 +363,12 @@ void GMainWindow::InitializeHotkeys() {
                     UpdateStatusBar();
                 }
             });
+    connect(hotkey_registry.GetHotkey("Main Window", "Load Amiibo", this), &QShortcut::activated,
+            this, [&] {
+                if (ui.action_Load_Amiibo->isEnabled()) {
+                    OnLoadAmiibo();
+                }
+            });
 }
 
 void GMainWindow::SetDefaultUIGeometry() {
@@ -422,6 +439,7 @@ void GMainWindow::ConnectMenuEvents() {
     connect(ui.action_Select_SDMC_Directory, &QAction::triggered, this,
             [this] { OnMenuSelectEmulatedDirectory(EmulatedDirectoryTarget::SDMC); });
     connect(ui.action_Exit, &QAction::triggered, this, &QMainWindow::close);
+    connect(ui.action_Load_Amiibo, &QAction::triggered, this, &GMainWindow::OnLoadAmiibo);
 
     // Emulation
     connect(ui.action_Start, &QAction::triggered, this, &GMainWindow::OnStartGame);
@@ -447,6 +465,7 @@ void GMainWindow::ConnectMenuEvents() {
     connect(ui.action_Fullscreen, &QAction::triggered, this, &GMainWindow::ToggleFullscreen);
 
     // Help
+    connect(ui.action_Open_yuzu_Folder, &QAction::triggered, this, &GMainWindow::OnOpenYuzuFolder);
     connect(ui.action_Rederive, &QAction::triggered, this,
             std::bind(&GMainWindow::OnReinitializeKeys, this, ReinitializeKeyBehavior::Warning));
     connect(ui.action_About, &QAction::triggered, this, &GMainWindow::OnAbout);
@@ -690,6 +709,7 @@ void GMainWindow::ShutdownGame() {
     ui.action_Stop->setEnabled(false);
     ui.action_Restart->setEnabled(false);
     ui.action_Report_Compatibility->setEnabled(false);
+    ui.action_Load_Amiibo->setEnabled(false);
     render_window->hide();
     game_list->show();
     game_list->setFilterFocus();
@@ -751,12 +771,43 @@ void GMainWindow::OnGameListOpenFolder(u64 program_id, GameListOpenTarget target
         open_target = "Save Data";
         const std::string nand_dir = FileUtil::GetUserPath(FileUtil::UserPath::NANDDir);
         ASSERT(program_id != 0);
-        // TODO(tech4me): Update this to work with arbitrary user profile
-        // Refer to core/hle/service/acc/profile_manager.cpp ProfileManager constructor
-        constexpr u128 user_id = {1, 0};
+
+        Service::Account::ProfileManager manager{};
+        const auto user_ids = manager.GetAllUsers();
+        QStringList list;
+        for (const auto& user_id : user_ids) {
+            if (user_id == Service::Account::UUID{})
+                continue;
+            Service::Account::ProfileBase base;
+            if (!manager.GetProfileBase(user_id, base))
+                continue;
+
+            list.push_back(QString::fromStdString(Common::StringFromFixedZeroTerminatedBuffer(
+                reinterpret_cast<const char*>(base.username.data()), base.username.size())));
+        }
+
+        bool ok = false;
+        const auto index_string =
+            QInputDialog::getItem(this, tr("Select User"),
+                                  tr("Please select the user's save data you would like to open."),
+                                  list, Settings::values.current_user, false, &ok);
+        if (!ok)
+            return;
+
+        const auto index = list.indexOf(index_string);
+        ASSERT(index != -1 && index < 8);
+
+        const auto user_id = manager.GetUser(index);
+        ASSERT(user_id);
         path = nand_dir + FileSys::SaveDataFactory::GetFullPath(FileSys::SaveDataSpaceId::NandUser,
                                                                 FileSys::SaveDataType::SaveData,
-                                                                program_id, user_id, 0);
+                                                                program_id, user_id->uuid, 0);
+
+        if (!FileUtil::Exists(path)) {
+            FileUtil::CreateFullPath(path);
+            FileUtil::CreateDir(path);
+        }
+
         break;
     }
     case GameListOpenTarget::ModData: {
@@ -823,14 +874,10 @@ static bool RomFSRawCopy(QProgressDialog& dialog, const FileSys::VirtualDir& src
 }
 
 void GMainWindow::OnGameListDumpRomFS(u64 program_id, const std::string& game_path) {
-    const auto path = fmt::format("{}{:016X}/romfs",
-                                  FileUtil::GetUserPath(FileUtil::UserPath::DumpDir), program_id);
-
-    const auto failed = [this, &path] {
+    const auto failed = [this] {
         QMessageBox::warning(this, tr("RomFS Extraction Failed!"),
                              tr("There was an error copying the RomFS files or the user "
                                 "cancelled the operation."));
-        vfs->DeleteDirectory(path);
     };
 
     const auto loader = Loader::GetLoader(vfs->OpenFile(game_path, FileSys::Mode::Read));
@@ -845,10 +892,24 @@ void GMainWindow::OnGameListDumpRomFS(u64 program_id, const std::string& game_pa
         return;
     }
 
-    const auto romfs =
-        loader->IsRomFSUpdatable()
-            ? FileSys::PatchManager(program_id).PatchRomFS(file, loader->ReadRomFSIVFCOffset())
-            : file;
+    const auto installed = Service::FileSystem::GetUnionContents();
+    auto romfs_title_id = SelectRomFSDumpTarget(*installed, program_id);
+
+    if (!romfs_title_id) {
+        failed();
+        return;
+    }
+
+    const auto path = fmt::format(
+        "{}{:016X}/romfs", FileUtil::GetUserPath(FileUtil::UserPath::DumpDir), *romfs_title_id);
+
+    FileSys::VirtualFile romfs;
+
+    if (*romfs_title_id == program_id) {
+        romfs = file;
+    } else {
+        romfs = installed->GetEntry(*romfs_title_id, FileSys::ContentRecordType::Data)->GetRomFS();
+    }
 
     const auto extracted = FileSys::ExtractRomFS(romfs, FileSys::RomFSExtractionType::Full);
     if (extracted == nullptr) {
@@ -860,6 +921,7 @@ void GMainWindow::OnGameListDumpRomFS(u64 program_id, const std::string& game_pa
 
     if (out == nullptr) {
         failed();
+        vfs->DeleteDirectory(path);
         return;
     }
 
@@ -870,13 +932,17 @@ void GMainWindow::OnGameListDumpRomFS(u64 program_id, const std::string& game_pa
            "files into the new directory while <br>skeleton will only create the directory "
            "structure."),
         {"Full", "Skeleton"}, 0, false, &ok);
-    if (!ok)
+    if (!ok) {
         failed();
+        vfs->DeleteDirectory(path);
+        return;
+    }
 
     const auto full = res == "Full";
     const auto entry_size = CalculateRomFSEntrySize(extracted, full);
 
-    QProgressDialog progress(tr("Extracting RomFS..."), tr("Cancel"), 0, entry_size, this);
+    QProgressDialog progress(tr("Extracting RomFS..."), tr("Cancel"), 0,
+                             static_cast<s32>(entry_size), this);
     progress.setWindowModality(Qt::WindowModal);
     progress.setMinimumDuration(100);
 
@@ -888,6 +954,7 @@ void GMainWindow::OnGameListDumpRomFS(u64 program_id, const std::string& game_pa
     } else {
         progress.close();
         failed();
+        vfs->DeleteDirectory(path);
     }
 }
 
@@ -1174,6 +1241,7 @@ void GMainWindow::OnStartGame() {
     ui.action_Report_Compatibility->setEnabled(true);
 
     discord_rpc->Update();
+    ui.action_Load_Amiibo->setEnabled(true);
 }
 
 void GMainWindow::OnPauseGame() {
@@ -1273,11 +1341,63 @@ void GMainWindow::OnConfigure() {
             UpdateUITheme();
         if (UISettings::values.enable_discord_presence != old_discord_presence)
             SetDiscordEnabled(UISettings::values.enable_discord_presence);
-        game_list->PopulateAsync(UISettings::values.gamedir, UISettings::values.gamedir_deepscan);
+
+        const auto reload = UISettings::values.is_game_list_reload_pending.exchange(false);
+        if (reload) {
+            game_list->PopulateAsync(UISettings::values.gamedir,
+                                     UISettings::values.gamedir_deepscan);
+        }
+
         config->Save();
     }
 }
 
+void GMainWindow::OnLoadAmiibo() {
+    const QString extensions{"*.bin"};
+    const QString file_filter = tr("Amiibo File (%1);; All Files (*.*)").arg(extensions);
+    const QString filename = QFileDialog::getOpenFileName(this, tr("Load Amiibo"), "", file_filter);
+
+    if (filename.isEmpty()) {
+        return;
+    }
+
+    Core::System& system{Core::System::GetInstance()};
+    Service::SM::ServiceManager& sm = system.ServiceManager();
+    auto nfc = sm.GetService<Service::NFP::Module::Interface>("nfp:user");
+    if (nfc == nullptr) {
+        return;
+    }
+
+    QFile nfc_file{filename};
+    if (!nfc_file.open(QIODevice::ReadOnly)) {
+        QMessageBox::warning(this, tr("Error opening Amiibo data file"),
+                             tr("Unable to open Amiibo file \"%1\" for reading.").arg(filename));
+        return;
+    }
+
+    const u64 nfc_file_size = nfc_file.size();
+    std::vector<u8> buffer(nfc_file_size);
+    const u64 read_size = nfc_file.read(reinterpret_cast<char*>(buffer.data()), nfc_file_size);
+    if (nfc_file_size != read_size) {
+        QMessageBox::warning(this, tr("Error reading Amiibo data file"),
+                             tr("Unable to fully read Amiibo data. Expected to read %1 bytes, but "
+                                "was only able to read %2 bytes.")
+                                 .arg(nfc_file_size)
+                                 .arg(read_size));
+        return;
+    }
+
+    if (!nfc->LoadAmiibo(buffer)) {
+        QMessageBox::warning(this, tr("Error loading Amiibo data"),
+                             tr("Unable to load Amiibo data."));
+    }
+}
+
+void GMainWindow::OnOpenYuzuFolder() {
+    QDesktopServices::openUrl(QUrl::fromLocalFile(
+        QString::fromStdString(FileUtil::GetUserPath(FileUtil::UserPath::UserDir))));
+}
+
 void GMainWindow::OnAbout() {
     AboutDialog aboutDialog(this);
     aboutDialog.exec();
@@ -1318,15 +1438,17 @@ void GMainWindow::UpdateStatusBar() {
 void GMainWindow::OnCoreError(Core::System::ResultStatus result, std::string details) {
     QMessageBox::StandardButton answer;
     QString status_message;
-    const QString common_message = tr(
-        "The game you are trying to load requires additional files from your Switch to be dumped "
-        "before playing.<br/><br/>For more information on dumping these files, please see the "
-        "following wiki page: <a "
-        "href='https://yuzu-emu.org/wiki/"
-        "dumping-system-archives-and-the-shared-fonts-from-a-switch-console/'>Dumping System "
-        "Archives and the Shared Fonts from a Switch Console</a>.<br/><br/>Would you like to quit "
-        "back to the game list? Continuing emulation may result in crashes, corrupted save "
-        "data, or other bugs.");
+    const QString common_message =
+        tr("The game you are trying to load requires additional files from your Switch to be "
+           "dumped "
+           "before playing.<br/><br/>For more information on dumping these files, please see the "
+           "following wiki page: <a "
+           "href='https://yuzu-emu.org/wiki/"
+           "dumping-system-archives-and-the-shared-fonts-from-a-switch-console/'>Dumping System "
+           "Archives and the Shared Fonts from a Switch Console</a>.<br/><br/>Would you like to "
+           "quit "
+           "back to the game list? Continuing emulation may result in crashes, corrupted save "
+           "data, or other bugs.");
     switch (result) {
     case Core::System::ResultStatus::ErrorSystemFiles: {
         QString message = "yuzu was unable to locate a Switch system archive";
@@ -1357,9 +1479,12 @@ void GMainWindow::OnCoreError(Core::System::ResultStatus result, std::string det
             this, tr("Fatal Error"),
             tr("yuzu has encountered a fatal error, please see the log for more details. "
                "For more information on accessing the log, please see the following page: "
-               "<a href='https://community.citra-emu.org/t/how-to-upload-the-log-file/296'>How to "
-               "Upload the Log File</a>.<br/><br/>Would you like to quit back to the game list? "
-               "Continuing emulation may result in crashes, corrupted save data, or other bugs."),
+               "<a href='https://community.citra-emu.org/t/how-to-upload-the-log-file/296'>How "
+               "to "
+               "Upload the Log File</a>.<br/><br/>Would you like to quit back to the game "
+               "list? "
+               "Continuing emulation may result in crashes, corrupted save data, or other "
+               "bugs."),
             QMessageBox::Yes | QMessageBox::No, QMessageBox::No);
         status_message = "Fatal Error encountered";
         break;
@@ -1431,7 +1556,7 @@ void GMainWindow::OnReinitializeKeys(ReinitializeKeyBehavior behavior) {
                    "derivation. It will be attempted but may not complete.<br><br>") +
                     errors +
                     tr("<br><br>You can get all of these and dump all of your games easily by "
-                       "following <a href='https://yuzu-emu.org/help/quickstart/quickstart/'>the "
+                       "following <a href='https://yuzu-emu.org/help/quickstart/'>the "
                        "quickstart guide</a>. Alternatively, you can use another method of dumping "
                        "to obtain all of your keys."));
         }
@@ -1459,6 +1584,42 @@ void GMainWindow::OnReinitializeKeys(ReinitializeKeyBehavior behavior) {
     }
 }
 
+std::optional<u64> GMainWindow::SelectRomFSDumpTarget(
+    const FileSys::RegisteredCacheUnion& installed, u64 program_id) {
+    const auto dlc_entries =
+        installed.ListEntriesFilter(FileSys::TitleType::AOC, FileSys::ContentRecordType::Data);
+    std::vector<FileSys::RegisteredCacheEntry> dlc_match;
+    dlc_match.reserve(dlc_entries.size());
+    std::copy_if(dlc_entries.begin(), dlc_entries.end(), std::back_inserter(dlc_match),
+                 [&program_id, &installed](const FileSys::RegisteredCacheEntry& entry) {
+                     return (entry.title_id & DLC_BASE_TITLE_ID_MASK) == program_id &&
+                            installed.GetEntry(entry)->GetStatus() == Loader::ResultStatus::Success;
+                 });
+
+    std::vector<u64> romfs_tids;
+    romfs_tids.push_back(program_id);
+    for (const auto& entry : dlc_match)
+        romfs_tids.push_back(entry.title_id);
+
+    if (romfs_tids.size() > 1) {
+        QStringList list{"Base"};
+        for (std::size_t i = 1; i < romfs_tids.size(); ++i)
+            list.push_back(QStringLiteral("DLC %1").arg(romfs_tids[i] & 0x7FF));
+
+        bool ok;
+        const auto res = QInputDialog::getItem(
+            this, tr("Select RomFS Dump Target"),
+            tr("Please select which RomFS you would like to dump."), list, 0, false, &ok);
+        if (!ok) {
+            return {};
+        }
+
+        return romfs_tids[list.indexOf(res)];
+    }
+
+    return program_id;
+}
+
 bool GMainWindow::ConfirmClose() {
     if (emu_thread == nullptr || !UISettings::values.confirm_before_closing)
         return true;
@@ -1475,7 +1636,7 @@ void GMainWindow::closeEvent(QCloseEvent* event) {
         return;
     }
 
-    if (ui.action_Fullscreen->isChecked()) {
+    if (!ui.action_Fullscreen->isChecked()) {
         UISettings::values.geometry = saveGeometry();
         UISettings::values.renderwindow_geometry = render_window->saveGeometry();
     }
diff --git a/src/yuzu/main.h b/src/yuzu/main.h
index 3663d6aed..929250e8c 100644
--- a/src/yuzu/main.h
+++ b/src/yuzu/main.h
@@ -5,6 +5,7 @@
 #pragma once
 
 #include <memory>
+#include <optional>
 #include <unordered_map>
 
 #include <QMainWindow>
@@ -29,8 +30,9 @@ class WaitTreeWidget;
 enum class GameListOpenTarget;
 
 namespace FileSys {
+class RegisteredCacheUnion;
 class VfsFilesystem;
-}
+} // namespace FileSys
 
 namespace Tegra {
 class DebugContext;
@@ -164,6 +166,8 @@ private slots:
     void OnMenuSelectEmulatedDirectory(EmulatedDirectoryTarget target);
     void OnMenuRecentFile();
     void OnConfigure();
+    void OnLoadAmiibo();
+    void OnOpenYuzuFolder();
     void OnAbout();
     void OnToggleFilterBar();
     void OnDisplayTitleBars(bool);
@@ -175,6 +179,7 @@ private slots:
     void OnReinitializeKeys(ReinitializeKeyBehavior behavior);
 
 private:
+    std::optional<u64> SelectRomFSDumpTarget(const FileSys::RegisteredCacheUnion&, u64 program_id);
     void UpdateStatusBar();
 
     Ui::MainWindow ui;
diff --git a/src/yuzu/main.ui b/src/yuzu/main.ui
index dffd9c788..75e96387f 100644
--- a/src/yuzu/main.ui
+++ b/src/yuzu/main.ui
@@ -57,8 +57,8 @@
       <string>Recent Files</string>
      </property>
     </widget>
-     <addaction name="action_Install_File_NAND" />
-     <addaction name="separator"/>
+    <addaction name="action_Install_File_NAND"/>
+    <addaction name="separator"/>
     <addaction name="action_Load_File"/>
     <addaction name="action_Load_Folder"/>
     <addaction name="separator"/>
@@ -68,6 +68,10 @@
     <addaction name="action_Select_NAND_Directory"/>
     <addaction name="action_Select_SDMC_Directory"/>
     <addaction name="separator"/>
+    <addaction name="action_Load_Amiibo"/>
+    <addaction name="separator"/>
+    <addaction name="action_Open_yuzu_Folder"/>
+    <addaction name="separator"/>
     <addaction name="action_Exit"/>
    </widget>
    <widget class="QMenu" name="menu_Emulation">
@@ -117,11 +121,14 @@
    <addaction name="menu_Tools" />
    <addaction name="menu_Help"/>
   </widget>
-   <action name="action_Install_File_NAND">
-     <property name="text">
-       <string>Install File to NAND...</string>
-     </property>
-   </action>
+  <action name="action_Install_File_NAND">
+   <property name="enabled">
+    <bool>true</bool>
+   </property>
+   <property name="text">
+    <string>Install File to NAND...</string>
+   </property>
+  </action>
   <action name="action_Load_File">
    <property name="text">
     <string>Load File...</string>
@@ -253,6 +260,14 @@
        <string>Restart</string>
      </property>
    </action>
+  <action name="action_Load_Amiibo">
+    <property name="enabled">
+      <bool>false</bool>
+    </property>
+    <property name="text">
+      <string>Load Amiibo...</string>
+    </property>
+  </action>
    <action name="action_Report_Compatibility">
      <property name="enabled">
        <bool>false</bool>
@@ -264,6 +279,11 @@
        <bool>false</bool>
      </property>
    </action>
+   <action name="action_Open_yuzu_Folder">
+     <property name="text">
+       <string>Open yuzu Folder</string>
+     </property>
+   </action>
   </widget>
  <resources/>
  <connections/>
diff --git a/src/yuzu/ui_settings.h b/src/yuzu/ui_settings.h
index 2e617d52a..e80aebc0a 100644
--- a/src/yuzu/ui_settings.h
+++ b/src/yuzu/ui_settings.h
@@ -5,6 +5,7 @@
 #pragma once
 
 #include <array>
+#include <atomic>
 #include <vector>
 #include <QByteArray>
 #include <QString>
@@ -59,9 +60,11 @@ struct Values {
 
     // Game List
     bool show_unknown;
+    bool show_add_ons;
     uint32_t icon_size;
     uint8_t row_1_text_id;
     uint8_t row_2_text_id;
+    std::atomic_bool is_game_list_reload_pending{false};
 };
 
 extern Values values;
diff --git a/src/yuzu/util/limitable_input_dialog.cpp b/src/yuzu/util/limitable_input_dialog.cpp
new file mode 100644
index 000000000..edd78e579
--- /dev/null
+++ b/src/yuzu/util/limitable_input_dialog.cpp
@@ -0,0 +1,59 @@
+// Copyright 2018 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <QDialogButtonBox>
+#include <QLabel>
+#include <QLineEdit>
+#include <QPushButton>
+#include <QVBoxLayout>
+#include "yuzu/util/limitable_input_dialog.h"
+
+LimitableInputDialog::LimitableInputDialog(QWidget* parent) : QDialog{parent} {
+    CreateUI();
+    ConnectEvents();
+}
+
+LimitableInputDialog::~LimitableInputDialog() = default;
+
+void LimitableInputDialog::CreateUI() {
+    setWindowFlags(windowFlags() & ~Qt::WindowContextHelpButtonHint);
+
+    text_label = new QLabel(this);
+    text_entry = new QLineEdit(this);
+    buttons = new QDialogButtonBox(QDialogButtonBox::Ok | QDialogButtonBox::Cancel, this);
+
+    auto* const layout = new QVBoxLayout;
+    layout->addWidget(text_label);
+    layout->addWidget(text_entry);
+    layout->addWidget(buttons);
+
+    setLayout(layout);
+}
+
+void LimitableInputDialog::ConnectEvents() {
+    connect(buttons, &QDialogButtonBox::accepted, this, &QDialog::accept);
+    connect(buttons, &QDialogButtonBox::rejected, this, &QDialog::reject);
+}
+
+QString LimitableInputDialog::GetText(QWidget* parent, const QString& title, const QString& text,
+                                      int min_character_limit, int max_character_limit) {
+    Q_ASSERT(min_character_limit <= max_character_limit);
+
+    LimitableInputDialog dialog{parent};
+    dialog.setWindowTitle(title);
+    dialog.text_label->setText(text);
+    dialog.text_entry->setMaxLength(max_character_limit);
+
+    auto* const ok_button = dialog.buttons->button(QDialogButtonBox::Ok);
+    ok_button->setEnabled(false);
+    connect(dialog.text_entry, &QLineEdit::textEdited, [&](const QString& new_text) {
+        ok_button->setEnabled(new_text.length() >= min_character_limit);
+    });
+
+    if (dialog.exec() != QDialog::Accepted) {
+        return {};
+    }
+
+    return dialog.text_entry->text();
+}
diff --git a/src/yuzu/util/limitable_input_dialog.h b/src/yuzu/util/limitable_input_dialog.h
new file mode 100644
index 000000000..164ad7301
--- /dev/null
+++ b/src/yuzu/util/limitable_input_dialog.h
@@ -0,0 +1,31 @@
+// Copyright 2018 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <QDialog>
+
+class QDialogButtonBox;
+class QLabel;
+class QLineEdit;
+
+/// A QDialog that functions similarly to QInputDialog, however, it allows
+/// restricting the minimum and total number of characters that can be entered.
+class LimitableInputDialog final : public QDialog {
+    Q_OBJECT
+public:
+    explicit LimitableInputDialog(QWidget* parent = nullptr);
+    ~LimitableInputDialog() override;
+
+    static QString GetText(QWidget* parent, const QString& title, const QString& text,
+                           int min_character_limit, int max_character_limit);
+
+private:
+    void CreateUI();
+    void ConnectEvents();
+
+    QLabel* text_label;
+    QLineEdit* text_entry;
+    QDialogButtonBox* buttons;
+};
diff --git a/src/yuzu_cmd/config.cpp b/src/yuzu_cmd/config.cpp
index 5e42e48b2..96f1ef636 100644
--- a/src/yuzu_cmd/config.cpp
+++ b/src/yuzu_cmd/config.cpp
@@ -8,6 +8,7 @@
 #include "common/file_util.h"
 #include "common/logging/log.h"
 #include "common/param_package.h"
+#include "core/hle/service/acc/profile_manager.h"
 #include "core/settings.h"
 #include "input_common/main.h"
 #include "yuzu_cmd/config.h"
@@ -125,11 +126,21 @@ void Config::ReadValues() {
 
     // System
     Settings::values.use_docked_mode = sdl2_config->GetBoolean("System", "use_docked_mode", false);
-    Settings::values.username = sdl2_config->Get("System", "username", "yuzu");
-    if (Settings::values.username.empty()) {
-        Settings::values.username = "yuzu";
+    Settings::values.enable_nfc = sdl2_config->GetBoolean("System", "enable_nfc", true);
+    const auto size = sdl2_config->GetInteger("System", "users_size", 0);
+
+    Settings::values.current_user = std::clamp<int>(
+        sdl2_config->GetInteger("System", "current_user", 0), 0, Service::Account::MAX_USERS - 1);
+
+    const auto enabled = sdl2_config->GetBoolean("System", "rng_seed_enabled", false);
+    if (enabled) {
+        Settings::values.rng_seed = sdl2_config->GetInteger("System", "rng_seed", 0);
+    } else {
+        Settings::values.rng_seed = std::nullopt;
     }
 
+    Settings::values.language_index = sdl2_config->GetInteger("System", "language_index", 1);
+
     // Miscellaneous
     Settings::values.log_filter = sdl2_config->Get("Miscellaneous", "log_filter", "*:Trace");
     Settings::values.use_dev_keys = sdl2_config->GetBoolean("Miscellaneous", "use_dev_keys", false);
@@ -139,6 +150,7 @@ void Config::ReadValues() {
     Settings::values.gdbstub_port =
         static_cast<u16>(sdl2_config->GetInteger("Debugging", "gdbstub_port", 24689));
     Settings::values.program_args = sdl2_config->Get("Debugging", "program_args", "");
+    Settings::values.dump_nso = sdl2_config->GetBoolean("Debugging", "dump_nso", false);
 
     // Web Service
     Settings::values.enable_telemetry =
diff --git a/src/yuzu_cmd/default_ini.h b/src/yuzu_cmd/default_ini.h
index a97b75f7b..ecf625e7b 100644
--- a/src/yuzu_cmd/default_ini.h
+++ b/src/yuzu_cmd/default_ini.h
@@ -174,6 +174,15 @@ use_virtual_sd =
 # 1: Yes, 0 (default): No
 use_docked_mode =
 
+# Allow the use of NFC in games
+# 1 (default): Yes, 0 : No
+enable_nfc =
+
+# Sets the seed for the RNG generator built into the switch
+# rng_seed will be ignored and randomly generated if rng_seed_enabled is false
+rng_seed_enabled =
+rng_seed =
+
 # Sets the account username, max length is 32 characters
 # yuzu (default)
 username = yuzu
@@ -197,6 +206,8 @@ log_filter = *:Trace
 # Port for listening to GDB connections.
 use_gdbstub=false
 gdbstub_port=24689
+# Determines whether or not yuzu will dump all NSOs it attempts to load while loading them
+dump_nso=false
 
 [WebService]
 # Whether or not to enable telemetry
diff --git a/src/yuzu_cmd/yuzu.cpp b/src/yuzu_cmd/yuzu.cpp
index c8b93b85b..806127b12 100644
--- a/src/yuzu_cmd/yuzu.cpp
+++ b/src/yuzu_cmd/yuzu.cpp
@@ -76,6 +76,9 @@ static void InitializeLogging() {
     const std::string& log_dir = FileUtil::GetUserPath(FileUtil::UserPath::LogDir);
     FileUtil::CreateFullPath(log_dir);
     Log::AddBackend(std::make_unique<Log::FileBackend>(log_dir + LOG_FILE));
+#ifdef _WIN32
+    Log::AddBackend(std::make_unique<Log::DebuggerBackend>());
+#endif
 }
 
 /// Application entry point